Adding upstream version 6.1.76.upstream/6.1.76upstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

58 files changed, 59896 insertions, 0 deletions

diff --git a/drivers/net/ethernet/chelsio/cxgb4/Makefile b/drivers/net/ethernet/chelsio/cxgb4/Makefile
new file mode 100644
index 000000000..a4b4d475a
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/Makefile
@@ -0,0 +1,16 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# Chelsio T4 driver
+#
+
+obj-$(CONFIG_CHELSIO_T4) += cxgb4.o
+
+cxgb4-objs := cxgb4_main.o l2t.o smt.o t4_hw.o sge.o clip_tbl.o cxgb4_ethtool.o \
+	      cxgb4_uld.o srq.o sched.o cxgb4_filter.o cxgb4_tc_u32.o \
+	      cxgb4_ptp.o cxgb4_tc_flower.o cxgb4_cudbg.o cxgb4_mps.o \
+	      cudbg_common.o cudbg_lib.o cudbg_zlib.o cxgb4_tc_mqprio.o \
+	      cxgb4_tc_matchall.o
+cxgb4-$(CONFIG_CHELSIO_T4_DCB) +=  cxgb4_dcb.o
+cxgb4-$(CONFIG_CHELSIO_T4_FCOE) +=  cxgb4_fcoe.o
+cxgb4-$(CONFIG_DEBUG_FS) += cxgb4_debugfs.o
+cxgb4-$(CONFIG_THERMAL) += cxgb4_thermal.o
diff --git a/drivers/net/ethernet/chelsio/cxgb4/clip_tbl.c b/drivers/net/ethernet/chelsio/cxgb4/clip_tbl.c
new file mode 100644
index 000000000..163efab27
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/clip_tbl.c
@@ -0,0 +1,328 @@
+/*
+ *  This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *  Copyright (C) 2003-2014 Chelsio Communications.  All rights reserved.
+ *
+ *  Written by Deepak (deepak.s@chelsio.com)
+ *
+ *  This program is distributed in the hope that it will be useful, but WITHOUT
+ *  ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ *  FITNESS FOR A PARTICULAR PURPOSE.  See the LICENSE file included in this
+ *  release for licensing terms and conditions.
+ */
+
+#include <linux/module.h>
+#include <linux/netdevice.h>
+#include <linux/jhash.h>
+#include <linux/if_vlan.h>
+#include <net/addrconf.h>
+#include "cxgb4.h"
+#include "clip_tbl.h"
+
+static inline unsigned int ipv4_clip_hash(struct clip_tbl *c, const u32 *key)
+{
+	unsigned int clipt_size_half = c->clipt_size / 2;
+
+	return jhash_1word(*key, 0) % clipt_size_half;
+}
+
+static inline unsigned int ipv6_clip_hash(struct clip_tbl *d, const u32 *key)
+{
+	unsigned int clipt_size_half = d->clipt_size / 2;
+	u32 xor = key[0] ^ key[1] ^ key[2] ^ key[3];
+
+	return clipt_size_half +
+		(jhash_1word(xor, 0) % clipt_size_half);
+}
+
+static unsigned int clip_addr_hash(struct clip_tbl *ctbl, const u32 *addr,
+				   u8 v6)
+{
+	return v6 ? ipv6_clip_hash(ctbl, addr) :
+			ipv4_clip_hash(ctbl, addr);
+}
+
+static int clip6_get_mbox(const struct net_device *dev,
+			  const struct in6_addr *lip)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct fw_clip_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_write = htonl(FW_CMD_OP_V(FW_CLIP_CMD) |
+			      FW_CMD_REQUEST_F | FW_CMD_WRITE_F);
+	c.alloc_to_len16 = htonl(FW_CLIP_CMD_ALLOC_F | FW_LEN16(c));
+	*(__be64 *)&c.ip_hi = *(__be64 *)(lip->s6_addr);
+	*(__be64 *)&c.ip_lo = *(__be64 *)(lip->s6_addr + 8);
+	return t4_wr_mbox_meat(adap, adap->mbox, &c, sizeof(c), &c, false);
+}
+
+static int clip6_release_mbox(const struct net_device *dev,
+			      const struct in6_addr *lip)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct fw_clip_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_write = htonl(FW_CMD_OP_V(FW_CLIP_CMD) |
+			      FW_CMD_REQUEST_F | FW_CMD_READ_F);
+	c.alloc_to_len16 = htonl(FW_CLIP_CMD_FREE_F | FW_LEN16(c));
+	*(__be64 *)&c.ip_hi = *(__be64 *)(lip->s6_addr);
+	*(__be64 *)&c.ip_lo = *(__be64 *)(lip->s6_addr + 8);
+	return t4_wr_mbox_meat(adap, adap->mbox, &c, sizeof(c), &c, false);
+}
+
+int cxgb4_clip_get(const struct net_device *dev, const u32 *lip, u8 v6)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct clip_tbl *ctbl = adap->clipt;
+	struct clip_entry *ce, *cte;
+	u32 *addr = (u32 *)lip;
+	int hash;
+	int ret = -1;
+
+	if (!ctbl)
+		return 0;
+
+	hash = clip_addr_hash(ctbl, addr, v6);
+
+	read_lock_bh(&ctbl->lock);
+	list_for_each_entry(cte, &ctbl->hash_list[hash], list) {
+		if (cte->addr6.sin6_family == AF_INET6 && v6)
+			ret = memcmp(lip, cte->addr6.sin6_addr.s6_addr,
+				     sizeof(struct in6_addr));
+		else if (cte->addr.sin_family == AF_INET && !v6)
+			ret = memcmp(lip, (char *)(&cte->addr.sin_addr),
+				     sizeof(struct in_addr));
+		if (!ret) {
+			ce = cte;
+			read_unlock_bh(&ctbl->lock);
+			refcount_inc(&ce->refcnt);
+			return 0;
+		}
+	}
+	read_unlock_bh(&ctbl->lock);
+
+	write_lock_bh(&ctbl->lock);
+	if (!list_empty(&ctbl->ce_free_head)) {
+		ce = list_first_entry(&ctbl->ce_free_head,
+				      struct clip_entry, list);
+		list_del_init(&ce->list);
+		spin_lock_init(&ce->lock);
+		refcount_set(&ce->refcnt, 0);
+		atomic_dec(&ctbl->nfree);
+		list_add_tail(&ce->list, &ctbl->hash_list[hash]);
+		if (v6) {
+			ce->addr6.sin6_family = AF_INET6;
+			memcpy(ce->addr6.sin6_addr.s6_addr,
+			       lip, sizeof(struct in6_addr));
+			ret = clip6_get_mbox(dev, (const struct in6_addr *)lip);
+			if (ret) {
+				write_unlock_bh(&ctbl->lock);
+				dev_err(adap->pdev_dev,
+					"CLIP FW cmd failed with error %d, "
+					"Connections using %pI6c wont be "
+					"offloaded",
+					ret, ce->addr6.sin6_addr.s6_addr);
+				return ret;
+			}
+		} else {
+			ce->addr.sin_family = AF_INET;
+			memcpy((char *)(&ce->addr.sin_addr), lip,
+			       sizeof(struct in_addr));
+		}
+	} else {
+		write_unlock_bh(&ctbl->lock);
+		dev_info(adap->pdev_dev, "CLIP table overflow, "
+			 "Connections using %pI6c wont be offloaded",
+			 (void *)lip);
+		return -ENOMEM;
+	}
+	write_unlock_bh(&ctbl->lock);
+	refcount_set(&ce->refcnt, 1);
+	return 0;
+}
+EXPORT_SYMBOL(cxgb4_clip_get);
+
+void cxgb4_clip_release(const struct net_device *dev, const u32 *lip, u8 v6)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct clip_tbl *ctbl = adap->clipt;
+	struct clip_entry *ce, *cte;
+	u32 *addr = (u32 *)lip;
+	int hash;
+	int ret = -1;
+
+	if (!ctbl)
+		return;
+
+	hash = clip_addr_hash(ctbl, addr, v6);
+
+	read_lock_bh(&ctbl->lock);
+	list_for_each_entry(cte, &ctbl->hash_list[hash], list) {
+		if (cte->addr6.sin6_family == AF_INET6 && v6)
+			ret = memcmp(lip, cte->addr6.sin6_addr.s6_addr,
+				     sizeof(struct in6_addr));
+		else if (cte->addr.sin_family == AF_INET && !v6)
+			ret = memcmp(lip, (char *)(&cte->addr.sin_addr),
+				     sizeof(struct in_addr));
+		if (!ret) {
+			ce = cte;
+			read_unlock_bh(&ctbl->lock);
+			goto found;
+		}
+	}
+	read_unlock_bh(&ctbl->lock);
+
+	return;
+found:
+	write_lock_bh(&ctbl->lock);
+	spin_lock_bh(&ce->lock);
+	if (refcount_dec_and_test(&ce->refcnt)) {
+		list_del_init(&ce->list);
+		list_add_tail(&ce->list, &ctbl->ce_free_head);
+		atomic_inc(&ctbl->nfree);
+		if (v6)
+			clip6_release_mbox(dev, (const struct in6_addr *)lip);
+	}
+	spin_unlock_bh(&ce->lock);
+	write_unlock_bh(&ctbl->lock);
+}
+EXPORT_SYMBOL(cxgb4_clip_release);
+
+/* Retrieves IPv6 addresses from a root device (bond, vlan) associated with
+ * a physical device.
+ * The physical device reference is needed to send the actul CLIP command.
+ */
+static int cxgb4_update_dev_clip(struct net_device *root_dev,
+				 struct net_device *dev)
+{
+	struct inet6_dev *idev = NULL;
+	struct inet6_ifaddr *ifa;
+	int ret = 0;
+
+	idev = __in6_dev_get(root_dev);
+	if (!idev)
+		return ret;
+
+	read_lock_bh(&idev->lock);
+	list_for_each_entry(ifa, &idev->addr_list, if_list) {
+		ret = cxgb4_clip_get(dev, (const u32 *)ifa->addr.s6_addr, 1);
+		if (ret < 0)
+			break;
+	}
+	read_unlock_bh(&idev->lock);
+
+	return ret;
+}
+
+int cxgb4_update_root_dev_clip(struct net_device *dev)
+{
+	struct net_device *root_dev = NULL;
+	int i, ret = 0;
+
+	/* First populate the real net device's IPv6 addresses */
+	ret = cxgb4_update_dev_clip(dev, dev);
+	if (ret)
+		return ret;
+
+	/* Parse all bond and vlan devices layered on top of the physical dev */
+	root_dev = netdev_master_upper_dev_get_rcu(dev);
+	if (root_dev) {
+		ret = cxgb4_update_dev_clip(root_dev, dev);
+		if (ret)
+			return ret;
+	}
+
+	for (i = 0; i < VLAN_N_VID; i++) {
+		root_dev = __vlan_find_dev_deep_rcu(dev, htons(ETH_P_8021Q), i);
+		if (!root_dev)
+			continue;
+
+		ret = cxgb4_update_dev_clip(root_dev, dev);
+		if (ret)
+			break;
+	}
+
+	return ret;
+}
+EXPORT_SYMBOL(cxgb4_update_root_dev_clip);
+
+int clip_tbl_show(struct seq_file *seq, void *v)
+{
+	struct adapter *adapter = seq->private;
+	struct clip_tbl *ctbl = adapter->clipt;
+	struct clip_entry *ce;
+	char ip[60];
+	int i;
+
+	read_lock_bh(&ctbl->lock);
+
+	seq_puts(seq, "IP Address                  Users\n");
+	for (i = 0 ; i < ctbl->clipt_size;  ++i) {
+		list_for_each_entry(ce, &ctbl->hash_list[i], list) {
+			ip[0] = '\0';
+			sprintf(ip, "%pISc", &ce->addr);
+			seq_printf(seq, "%-25s   %u\n", ip,
+				   refcount_read(&ce->refcnt));
+		}
+	}
+	seq_printf(seq, "Free clip entries : %d\n", atomic_read(&ctbl->nfree));
+
+	read_unlock_bh(&ctbl->lock);
+
+	return 0;
+}
+
+struct clip_tbl *t4_init_clip_tbl(unsigned int clipt_start,
+				  unsigned int clipt_end)
+{
+	struct clip_entry *cl_list;
+	struct clip_tbl *ctbl;
+	unsigned int clipt_size;
+	int i;
+
+	if (clipt_start >= clipt_end)
+		return NULL;
+	clipt_size = clipt_end - clipt_start + 1;
+	if (clipt_size < CLIPT_MIN_HASH_BUCKETS)
+		return NULL;
+
+	ctbl = kvzalloc(struct_size(ctbl, hash_list, clipt_size), GFP_KERNEL);
+	if (!ctbl)
+		return NULL;
+
+	ctbl->clipt_start = clipt_start;
+	ctbl->clipt_size = clipt_size;
+	INIT_LIST_HEAD(&ctbl->ce_free_head);
+
+	atomic_set(&ctbl->nfree, clipt_size);
+	rwlock_init(&ctbl->lock);
+
+	for (i = 0; i < ctbl->clipt_size; ++i)
+		INIT_LIST_HEAD(&ctbl->hash_list[i]);
+
+	cl_list = kvcalloc(clipt_size, sizeof(struct clip_entry), GFP_KERNEL);
+	if (!cl_list) {
+		kvfree(ctbl);
+		return NULL;
+	}
+	ctbl->cl_list = (void *)cl_list;
+
+	for (i = 0; i < clipt_size; i++) {
+		INIT_LIST_HEAD(&cl_list[i].list);
+		list_add_tail(&cl_list[i].list, &ctbl->ce_free_head);
+	}
+
+	return ctbl;
+}
+
+void t4_cleanup_clip_tbl(struct adapter *adap)
+{
+	struct clip_tbl *ctbl = adap->clipt;
+
+	if (ctbl) {
+		kvfree(ctbl->cl_list);
+		kvfree(ctbl);
+	}
+}
+EXPORT_SYMBOL(t4_cleanup_clip_tbl);
diff --git a/drivers/net/ethernet/chelsio/cxgb4/clip_tbl.h b/drivers/net/ethernet/chelsio/cxgb4/clip_tbl.h
new file mode 100644
index 000000000..290c10580
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/clip_tbl.h
@@ -0,0 +1,45 @@
+/*
+ *  This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *  Copyright (C) 2003-2014 Chelsio Communications.  All rights reserved.
+ *
+ *  Written by Deepak (deepak.s@chelsio.com)
+ *
+ *  This program is distributed in the hope that it will be useful, but WITHOUT
+ *  ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ *  FITNESS FOR A PARTICULAR PURPOSE.  See the LICENSE file included in this
+ *  release for licensing terms and conditions.
+ */
+
+#include <linux/refcount.h>
+
+struct clip_entry {
+	spinlock_t lock;	/* Hold while modifying clip reference */
+	refcount_t refcnt;
+	struct list_head list;
+	union {
+		struct sockaddr_in addr;
+		struct sockaddr_in6 addr6;
+	};
+};
+
+struct clip_tbl {
+	unsigned int clipt_start;
+	unsigned int clipt_size;
+	rwlock_t lock;
+	atomic_t nfree;
+	struct list_head ce_free_head;
+	void *cl_list;
+	struct list_head hash_list[];
+};
+
+enum {
+	CLIPT_MIN_HASH_BUCKETS = 2,
+};
+
+struct clip_tbl *t4_init_clip_tbl(unsigned int clipt_start,
+				  unsigned int clipt_end);
+int cxgb4_clip_get(const struct net_device *dev, const u32 *lip, u8 v6);
+void cxgb4_clip_release(const struct net_device *dev, const u32 *lip, u8 v6);
+int clip_tbl_show(struct seq_file *seq, void *v);
+int cxgb4_update_root_dev_clip(struct net_device *dev);
+void t4_cleanup_clip_tbl(struct adapter *adap);
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cudbg_common.c b/drivers/net/ethernet/chelsio/cxgb4/cudbg_common.c
new file mode 100644
index 000000000..175e1a675
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cudbg_common.c
@@ -0,0 +1,56 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  Copyright (C) 2017 Chelsio Communications.  All rights reserved.
+ */
+
+#include "cxgb4.h"
+#include "cudbg_if.h"
+#include "cudbg_lib_common.h"
+
+int cudbg_get_buff(struct cudbg_init *pdbg_init,
+		   struct cudbg_buffer *pdbg_buff, u32 size,
+		   struct cudbg_buffer *pin_buff)
+{
+	u32 offset;
+
+	offset = pdbg_buff->offset;
+	if (offset + size > pdbg_buff->size)
+		return CUDBG_STATUS_NO_MEM;
+
+	if (pdbg_init->compress_type != CUDBG_COMPRESSION_NONE) {
+		if (size > pdbg_init->compress_buff_size)
+			return CUDBG_STATUS_NO_MEM;
+
+		pin_buff->data = (char *)pdbg_init->compress_buff;
+		pin_buff->offset = 0;
+		pin_buff->size = size;
+		return 0;
+	}
+
+	pin_buff->data = (char *)pdbg_buff->data + offset;
+	pin_buff->offset = offset;
+	pin_buff->size = size;
+	return 0;
+}
+
+void cudbg_put_buff(struct cudbg_init *pdbg_init,
+		    struct cudbg_buffer *pin_buff)
+{
+	/* Clear compression buffer for re-use */
+	if (pdbg_init->compress_type != CUDBG_COMPRESSION_NONE)
+		memset(pdbg_init->compress_buff, 0,
+		       pdbg_init->compress_buff_size);
+
+	pin_buff->data = NULL;
+	pin_buff->offset = 0;
+	pin_buff->size = 0;
+}
+
+void cudbg_update_buff(struct cudbg_buffer *pin_buff,
+		       struct cudbg_buffer *pout_buff)
+{
+	/* We already write to buffer provided by ethool, so just
+	 * increment offset to next free space.
+	 */
+	pout_buff->offset += pin_buff->size;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cudbg_entity.h b/drivers/net/ethernet/chelsio/cxgb4/cudbg_entity.h
new file mode 100644
index 000000000..d5218e742
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cudbg_entity.h
@@ -0,0 +1,352 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ *  Copyright (C) 2017 Chelsio Communications.  All rights reserved.
+ */
+
+#ifndef __CUDBG_ENTITY_H__
+#define __CUDBG_ENTITY_H__
+
+#define EDC0_FLAG 0
+#define EDC1_FLAG 1
+#define MC_FLAG 2
+#define MC0_FLAG 3
+#define MC1_FLAG 4
+#define HMA_FLAG 5
+
+#define CUDBG_ENTITY_SIGNATURE 0xCCEDB001
+
+struct cudbg_mbox_log {
+	struct mbox_cmd entry;
+	u32 hi[MBOX_LEN / 8];
+	u32 lo[MBOX_LEN / 8];
+};
+
+struct cudbg_cim_qcfg {
+	u8 chip;
+	u16 base[CIM_NUM_IBQ + CIM_NUM_OBQ_T5];
+	u16 size[CIM_NUM_IBQ + CIM_NUM_OBQ_T5];
+	u16 thres[CIM_NUM_IBQ];
+	u32 obq_wr[2 * CIM_NUM_OBQ_T5];
+	u32 stat[4 * (CIM_NUM_IBQ + CIM_NUM_OBQ_T5)];
+};
+
+struct cudbg_rss_vf_conf {
+	u32 rss_vf_vfl;
+	u32 rss_vf_vfh;
+};
+
+struct cudbg_pm_stats {
+	u32 tx_cnt[T6_PM_NSTATS];
+	u32 rx_cnt[T6_PM_NSTATS];
+	u64 tx_cyc[T6_PM_NSTATS];
+	u64 rx_cyc[T6_PM_NSTATS];
+};
+
+struct cudbg_hw_sched {
+	u32 kbps[NTX_SCHED];
+	u32 ipg[NTX_SCHED];
+	u32 pace_tab[NTX_SCHED];
+	u32 mode;
+	u32 map;
+};
+
+#define SGE_QBASE_DATA_REG_NUM 4
+
+struct sge_qbase_reg_field {
+	u32 reg_addr;
+	u32 reg_data[SGE_QBASE_DATA_REG_NUM];
+	/* Max supported PFs */
+	u32 pf_data_value[PCIE_FW_MASTER_M + 1][SGE_QBASE_DATA_REG_NUM];
+	/* Max supported VFs */
+	u32 vf_data_value[T6_VF_M + 1][SGE_QBASE_DATA_REG_NUM];
+	u32 vfcount; /* Actual number of max vfs in current configuration */
+};
+
+struct ireg_field {
+	u32 ireg_addr;
+	u32 ireg_data;
+	u32 ireg_local_offset;
+	u32 ireg_offset_range;
+};
+
+struct ireg_buf {
+	struct ireg_field tp_pio;
+	u32 outbuf[32];
+};
+
+struct cudbg_ulprx_la {
+	u32 data[ULPRX_LA_SIZE * 8];
+	u32 size;
+};
+
+struct cudbg_tp_la {
+	u32 size;
+	u32 mode;
+	u8 data[];
+};
+
+static const char * const cudbg_region[] = {
+	"DBQ contexts:", "IMSG contexts:", "FLM cache:", "TCBs:",
+	"Pstructs:", "Timers:", "Rx FL:", "Tx FL:", "Pstruct FL:",
+	"Tx payload:", "Rx payload:", "LE hash:", "iSCSI region:",
+	"TDDP region:", "TPT region:", "STAG region:", "RQ region:",
+	"RQUDP region:", "PBL region:", "TXPBL region:",
+	"DBVFIFO region:", "ULPRX state:", "ULPTX state:",
+	"On-chip queues:"
+};
+
+/* Memory region info relative to current memory (i.e. wrt 0). */
+struct cudbg_region_info {
+	bool exist; /* Does region exists in current memory? */
+	u32 start;  /* Start wrt 0 */
+	u32 end;    /* End wrt 0 */
+};
+
+struct cudbg_mem_desc {
+	u32 base;
+	u32 limit;
+	u32 idx;
+};
+
+#define CUDBG_MEMINFO_REV 1
+
+struct cudbg_meminfo {
+	struct cudbg_mem_desc avail[4];
+	struct cudbg_mem_desc mem[ARRAY_SIZE(cudbg_region) + 3];
+	u32 avail_c;
+	u32 mem_c;
+	u32 up_ram_lo;
+	u32 up_ram_hi;
+	u32 up_extmem2_lo;
+	u32 up_extmem2_hi;
+	u32 rx_pages_data[3];
+	u32 tx_pages_data[4];
+	u32 p_structs;
+	u32 reserved[12];
+	u32 port_used[4];
+	u32 port_alloc[4];
+	u32 loopback_used[NCHAN];
+	u32 loopback_alloc[NCHAN];
+	u32 p_structs_free_cnt;
+	u32 free_rx_cnt;
+	u32 free_tx_cnt;
+};
+
+struct cudbg_cim_pif_la {
+	int size;
+	u8 data[];
+};
+
+struct cudbg_clk_info {
+	u64 retransmit_min;
+	u64 retransmit_max;
+	u64 persist_timer_min;
+	u64 persist_timer_max;
+	u64 keepalive_idle_timer;
+	u64 keepalive_interval;
+	u64 initial_srtt;
+	u64 finwait2_timer;
+	u32 dack_timer;
+	u32 res;
+	u32 cclk_ps;
+	u32 tre;
+	u32 dack_re;
+};
+
+struct cudbg_tid_info_region {
+	u32 ntids;
+	u32 nstids;
+	u32 stid_base;
+	u32 hash_base;
+
+	u32 natids;
+	u32 nftids;
+	u32 ftid_base;
+	u32 aftid_base;
+	u32 aftid_end;
+
+	u32 sftid_base;
+	u32 nsftids;
+
+	u32 uotid_base;
+	u32 nuotids;
+
+	u32 sb;
+	u32 flags;
+	u32 le_db_conf;
+	u32 ip_users;
+	u32 ipv6_users;
+
+	u32 hpftid_base;
+	u32 nhpftids;
+};
+
+#define CUDBG_TID_INFO_REV 1
+
+struct cudbg_tid_info_region_rev1 {
+	struct cudbg_ver_hdr ver_hdr;
+	struct cudbg_tid_info_region tid;
+	u32 tid_start;
+	u32 reserved[16];
+};
+
+#define CUDBG_LOWMEM_MAX_CTXT_QIDS 256
+#define CUDBG_MAX_FL_QIDS 1024
+
+struct cudbg_ch_cntxt {
+	u32 cntxt_type;
+	u32 cntxt_id;
+	u32 data[SGE_CTXT_SIZE / 4];
+};
+
+#define CUDBG_MAX_RPLC_SIZE 128
+
+struct cudbg_mps_tcam {
+	u64 mask;
+	u32 rplc[8];
+	u32 idx;
+	u32 cls_lo;
+	u32 cls_hi;
+	u32 rplc_size;
+	u32 vniy;
+	u32 vnix;
+	u32 dip_hit;
+	u32 vlan_vld;
+	u32 repli;
+	u16 ivlan;
+	u8 addr[ETH_ALEN];
+	u8 lookup_type;
+	u8 port_num;
+	u8 reserved[2];
+};
+
+#define CUDBG_VPD_VER_ADDR 0x18c7
+#define CUDBG_VPD_VER_LEN 2
+
+struct cudbg_vpd_data {
+	u8 sn[SERNUM_LEN + 1];
+	u8 bn[PN_LEN + 1];
+	u8 na[MACADDR_LEN + 1];
+	u8 mn[ID_LEN + 1];
+	u16 fw_major;
+	u16 fw_minor;
+	u16 fw_micro;
+	u16 fw_build;
+	u32 scfg_vers;
+	u32 vpd_vers;
+};
+
+#define CUDBG_MAX_TCAM_TID 0x800
+#define CUDBG_T6_CLIP 1536
+#define CUDBG_MAX_TID_COMP_EN 6144
+#define CUDBG_MAX_TID_COMP_DIS 3072
+
+enum cudbg_le_entry_types {
+	LE_ET_UNKNOWN = 0,
+	LE_ET_TCAM_CON = 1,
+	LE_ET_TCAM_SERVER = 2,
+	LE_ET_TCAM_FILTER = 3,
+	LE_ET_TCAM_CLIP = 4,
+	LE_ET_TCAM_ROUTING = 5,
+	LE_ET_HASH_CON = 6,
+	LE_ET_INVALID_TID = 8,
+};
+
+struct cudbg_tcam {
+	u32 filter_start;
+	u32 server_start;
+	u32 clip_start;
+	u32 routing_start;
+	u32 tid_hash_base;
+	u32 max_tid;
+};
+
+struct cudbg_tid_data {
+	u32 tid;
+	u32 dbig_cmd;
+	u32 dbig_conf;
+	u32 dbig_rsp_stat;
+	u32 data[NUM_LE_DB_DBGI_RSP_DATA_INSTANCES];
+};
+
+#define CUDBG_NUM_ULPTX 11
+#define CUDBG_NUM_ULPTX_READ 512
+#define CUDBG_NUM_ULPTX_ASIC 6
+#define CUDBG_NUM_ULPTX_ASIC_READ 128
+
+#define CUDBG_ULPTX_LA_REV 1
+
+struct cudbg_ulptx_la {
+	u32 rdptr[CUDBG_NUM_ULPTX];
+	u32 wrptr[CUDBG_NUM_ULPTX];
+	u32 rddata[CUDBG_NUM_ULPTX];
+	u32 rd_data[CUDBG_NUM_ULPTX][CUDBG_NUM_ULPTX_READ];
+	u32 rdptr_asic[CUDBG_NUM_ULPTX_ASIC_READ];
+	u32 rddata_asic[CUDBG_NUM_ULPTX_ASIC_READ][CUDBG_NUM_ULPTX_ASIC];
+};
+
+#define CUDBG_CHAC_PBT_ADDR 0x2800
+#define CUDBG_CHAC_PBT_LRF  0x3000
+#define CUDBG_CHAC_PBT_DATA 0x3800
+#define CUDBG_PBT_DYNAMIC_ENTRIES 8
+#define CUDBG_PBT_STATIC_ENTRIES 16
+#define CUDBG_LRF_ENTRIES 8
+#define CUDBG_PBT_DATA_ENTRIES 512
+
+struct cudbg_pbt_tables {
+	u32 pbt_dynamic[CUDBG_PBT_DYNAMIC_ENTRIES];
+	u32 pbt_static[CUDBG_PBT_STATIC_ENTRIES];
+	u32 lrf_table[CUDBG_LRF_ENTRIES];
+	u32 pbt_data[CUDBG_PBT_DATA_ENTRIES];
+};
+
+enum cudbg_qdesc_qtype {
+	CUDBG_QTYPE_UNKNOWN = 0,
+	CUDBG_QTYPE_NIC_TXQ,
+	CUDBG_QTYPE_NIC_RXQ,
+	CUDBG_QTYPE_NIC_FLQ,
+	CUDBG_QTYPE_CTRLQ,
+	CUDBG_QTYPE_FWEVTQ,
+	CUDBG_QTYPE_INTRQ,
+	CUDBG_QTYPE_PTP_TXQ,
+	CUDBG_QTYPE_OFLD_TXQ,
+	CUDBG_QTYPE_RDMA_RXQ,
+	CUDBG_QTYPE_RDMA_FLQ,
+	CUDBG_QTYPE_RDMA_CIQ,
+	CUDBG_QTYPE_ISCSI_RXQ,
+	CUDBG_QTYPE_ISCSI_FLQ,
+	CUDBG_QTYPE_ISCSIT_RXQ,
+	CUDBG_QTYPE_ISCSIT_FLQ,
+	CUDBG_QTYPE_CRYPTO_TXQ,
+	CUDBG_QTYPE_CRYPTO_RXQ,
+	CUDBG_QTYPE_CRYPTO_FLQ,
+	CUDBG_QTYPE_TLS_RXQ,
+	CUDBG_QTYPE_TLS_FLQ,
+	CUDBG_QTYPE_ETHOFLD_TXQ,
+	CUDBG_QTYPE_ETHOFLD_RXQ,
+	CUDBG_QTYPE_ETHOFLD_FLQ,
+	CUDBG_QTYPE_MAX,
+};
+
+#define CUDBG_QDESC_REV 1
+
+struct cudbg_qdesc_entry {
+	u32 data_size;
+	u32 qtype;
+	u32 qid;
+	u32 desc_size;
+	u32 num_desc;
+	u8 data[]; /* Must be last */
+};
+
+struct cudbg_qdesc_info {
+	u32 qdesc_entry_size;
+	u32 num_queues;
+	u8 data[]; /* Must be last */
+};
+
+#define IREG_NUM_ELEM 4
+
+#define CUDBG_NUM_PCIE_CONFIG_REGS 0x61
+
+#endif /* __CUDBG_ENTITY_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cudbg_if.h b/drivers/net/ethernet/chelsio/cxgb4/cudbg_if.h
new file mode 100644
index 000000000..c84719e3c
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cudbg_if.h
@@ -0,0 +1,91 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ *  Copyright (C) 2017 Chelsio Communications.  All rights reserved.
+ */
+
+#ifndef __CUDBG_IF_H__
+#define __CUDBG_IF_H__
+
+/* Error codes */
+#define CUDBG_STATUS_NO_MEM -19
+#define CUDBG_STATUS_ENTITY_NOT_FOUND -24
+#define CUDBG_STATUS_NOT_IMPLEMENTED -28
+#define CUDBG_SYSTEM_ERROR -29
+#define CUDBG_STATUS_CCLK_NOT_DEFINED -32
+#define CUDBG_STATUS_PARTIAL_DATA -41
+
+#define CUDBG_MAJOR_VERSION 1
+#define CUDBG_MINOR_VERSION 14
+
+enum cudbg_dbg_entity_type {
+	CUDBG_REG_DUMP = 1,
+	CUDBG_DEV_LOG = 2,
+	CUDBG_CIM_LA = 3,
+	CUDBG_CIM_MA_LA = 4,
+	CUDBG_CIM_QCFG = 5,
+	CUDBG_CIM_IBQ_TP0 = 6,
+	CUDBG_CIM_IBQ_TP1 = 7,
+	CUDBG_CIM_IBQ_ULP = 8,
+	CUDBG_CIM_IBQ_SGE0 = 9,
+	CUDBG_CIM_IBQ_SGE1 = 10,
+	CUDBG_CIM_IBQ_NCSI = 11,
+	CUDBG_CIM_OBQ_ULP0 = 12,
+	CUDBG_CIM_OBQ_ULP1 = 13,
+	CUDBG_CIM_OBQ_ULP2 = 14,
+	CUDBG_CIM_OBQ_ULP3 = 15,
+	CUDBG_CIM_OBQ_SGE = 16,
+	CUDBG_CIM_OBQ_NCSI = 17,
+	CUDBG_EDC0 = 18,
+	CUDBG_EDC1 = 19,
+	CUDBG_MC0 = 20,
+	CUDBG_MC1 = 21,
+	CUDBG_RSS = 22,
+	CUDBG_RSS_VF_CONF = 25,
+	CUDBG_PATH_MTU = 27,
+	CUDBG_PM_STATS = 30,
+	CUDBG_HW_SCHED = 31,
+	CUDBG_TP_INDIRECT = 36,
+	CUDBG_SGE_INDIRECT = 37,
+	CUDBG_ULPRX_LA = 41,
+	CUDBG_TP_LA = 43,
+	CUDBG_MEMINFO = 44,
+	CUDBG_CIM_PIF_LA = 45,
+	CUDBG_CLK = 46,
+	CUDBG_CIM_OBQ_RXQ0 = 47,
+	CUDBG_CIM_OBQ_RXQ1 = 48,
+	CUDBG_PCIE_INDIRECT = 50,
+	CUDBG_PM_INDIRECT = 51,
+	CUDBG_TID_INFO = 54,
+	CUDBG_PCIE_CONFIG = 55,
+	CUDBG_DUMP_CONTEXT = 56,
+	CUDBG_MPS_TCAM = 57,
+	CUDBG_VPD_DATA = 58,
+	CUDBG_LE_TCAM = 59,
+	CUDBG_CCTRL = 60,
+	CUDBG_MA_INDIRECT = 61,
+	CUDBG_ULPTX_LA = 62,
+	CUDBG_UP_CIM_INDIRECT = 64,
+	CUDBG_PBT_TABLE = 65,
+	CUDBG_MBOX_LOG = 66,
+	CUDBG_HMA_INDIRECT = 67,
+	CUDBG_HMA = 68,
+	CUDBG_QDESC = 70,
+	CUDBG_FLASH = 71,
+	CUDBG_MAX_ENTITY = 72,
+};
+
+struct cudbg_init {
+	struct adapter *adap; /* Pointer to adapter structure */
+	void *outbuf; /* Output buffer */
+	u32 outbuf_size;  /* Output buffer size */
+	u8 compress_type; /* Type of compression to use */
+	void *compress_buff; /* Compression buffer */
+	u32 compress_buff_size; /* Compression buffer size */
+	void *workspace; /* Workspace for zlib */
+};
+
+static inline unsigned int cudbg_mbytes_to_bytes(unsigned int size)
+{
+	return size * 1024 * 1024;
+}
+#endif /* __CUDBG_IF_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cudbg_lib.c b/drivers/net/ethernet/chelsio/cxgb4/cudbg_lib.c
new file mode 100644
index 000000000..557c591a6
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cudbg_lib.c
@@ -0,0 +1,3609 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  Copyright (C) 2017 Chelsio Communications.  All rights reserved.
+ */
+
+#include <linux/sort.h>
+#include <linux/string.h>
+
+#include "t4_regs.h"
+#include "cxgb4.h"
+#include "cxgb4_cudbg.h"
+#include "cudbg_if.h"
+#include "cudbg_lib_common.h"
+#include "cudbg_entity.h"
+#include "cudbg_lib.h"
+#include "cudbg_zlib.h"
+#include "cxgb4_tc_mqprio.h"
+
+static const u32 t6_tp_pio_array[][IREG_NUM_ELEM] = {
+	{0x7e40, 0x7e44, 0x020, 28}, /* t6_tp_pio_regs_20_to_3b */
+	{0x7e40, 0x7e44, 0x040, 10}, /* t6_tp_pio_regs_40_to_49 */
+	{0x7e40, 0x7e44, 0x050, 10}, /* t6_tp_pio_regs_50_to_59 */
+	{0x7e40, 0x7e44, 0x060, 14}, /* t6_tp_pio_regs_60_to_6d */
+	{0x7e40, 0x7e44, 0x06F, 1}, /* t6_tp_pio_regs_6f */
+	{0x7e40, 0x7e44, 0x070, 6}, /* t6_tp_pio_regs_70_to_75 */
+	{0x7e40, 0x7e44, 0x130, 18}, /* t6_tp_pio_regs_130_to_141 */
+	{0x7e40, 0x7e44, 0x145, 19}, /* t6_tp_pio_regs_145_to_157 */
+	{0x7e40, 0x7e44, 0x160, 1}, /* t6_tp_pio_regs_160 */
+	{0x7e40, 0x7e44, 0x230, 25}, /* t6_tp_pio_regs_230_to_248 */
+	{0x7e40, 0x7e44, 0x24a, 3}, /* t6_tp_pio_regs_24c */
+	{0x7e40, 0x7e44, 0x8C0, 1} /* t6_tp_pio_regs_8c0 */
+};
+
+static const u32 t5_tp_pio_array[][IREG_NUM_ELEM] = {
+	{0x7e40, 0x7e44, 0x020, 28}, /* t5_tp_pio_regs_20_to_3b */
+	{0x7e40, 0x7e44, 0x040, 19}, /* t5_tp_pio_regs_40_to_52 */
+	{0x7e40, 0x7e44, 0x054, 2}, /* t5_tp_pio_regs_54_to_55 */
+	{0x7e40, 0x7e44, 0x060, 13}, /* t5_tp_pio_regs_60_to_6c */
+	{0x7e40, 0x7e44, 0x06F, 1}, /* t5_tp_pio_regs_6f */
+	{0x7e40, 0x7e44, 0x120, 4}, /* t5_tp_pio_regs_120_to_123 */
+	{0x7e40, 0x7e44, 0x12b, 2}, /* t5_tp_pio_regs_12b_to_12c */
+	{0x7e40, 0x7e44, 0x12f, 21}, /* t5_tp_pio_regs_12f_to_143 */
+	{0x7e40, 0x7e44, 0x145, 19}, /* t5_tp_pio_regs_145_to_157 */
+	{0x7e40, 0x7e44, 0x230, 25}, /* t5_tp_pio_regs_230_to_248 */
+	{0x7e40, 0x7e44, 0x8C0, 1} /* t5_tp_pio_regs_8c0 */
+};
+
+static const u32 t6_tp_tm_pio_array[][IREG_NUM_ELEM] = {
+	{0x7e18, 0x7e1c, 0x0, 12}
+};
+
+static const u32 t5_tp_tm_pio_array[][IREG_NUM_ELEM] = {
+	{0x7e18, 0x7e1c, 0x0, 12}
+};
+
+static const u32 t6_tp_mib_index_array[6][IREG_NUM_ELEM] = {
+	{0x7e50, 0x7e54, 0x0, 13},
+	{0x7e50, 0x7e54, 0x10, 6},
+	{0x7e50, 0x7e54, 0x18, 21},
+	{0x7e50, 0x7e54, 0x30, 32},
+	{0x7e50, 0x7e54, 0x50, 22},
+	{0x7e50, 0x7e54, 0x68, 12}
+};
+
+static const u32 t5_tp_mib_index_array[9][IREG_NUM_ELEM] = {
+	{0x7e50, 0x7e54, 0x0, 13},
+	{0x7e50, 0x7e54, 0x10, 6},
+	{0x7e50, 0x7e54, 0x18, 8},
+	{0x7e50, 0x7e54, 0x20, 13},
+	{0x7e50, 0x7e54, 0x30, 16},
+	{0x7e50, 0x7e54, 0x40, 16},
+	{0x7e50, 0x7e54, 0x50, 16},
+	{0x7e50, 0x7e54, 0x60, 6},
+	{0x7e50, 0x7e54, 0x68, 4}
+};
+
+static const u32 t5_sge_dbg_index_array[2][IREG_NUM_ELEM] = {
+	{0x10cc, 0x10d0, 0x0, 16},
+	{0x10cc, 0x10d4, 0x0, 16},
+};
+
+static const u32 t6_sge_qbase_index_array[] = {
+	/* 1 addr reg SGE_QBASE_INDEX and 4 data reg SGE_QBASE_MAP[0-3] */
+	0x1250, 0x1240, 0x1244, 0x1248, 0x124c,
+};
+
+static const u32 t5_pcie_pdbg_array[][IREG_NUM_ELEM] = {
+	{0x5a04, 0x5a0c, 0x00, 0x20}, /* t5_pcie_pdbg_regs_00_to_20 */
+	{0x5a04, 0x5a0c, 0x21, 0x20}, /* t5_pcie_pdbg_regs_21_to_40 */
+	{0x5a04, 0x5a0c, 0x41, 0x10}, /* t5_pcie_pdbg_regs_41_to_50 */
+};
+
+static const u32 t5_pcie_cdbg_array[][IREG_NUM_ELEM] = {
+	{0x5a10, 0x5a18, 0x00, 0x20}, /* t5_pcie_cdbg_regs_00_to_20 */
+	{0x5a10, 0x5a18, 0x21, 0x18}, /* t5_pcie_cdbg_regs_21_to_37 */
+};
+
+static const u32 t5_pm_rx_array[][IREG_NUM_ELEM] = {
+	{0x8FD0, 0x8FD4, 0x10000, 0x20}, /* t5_pm_rx_regs_10000_to_10020 */
+	{0x8FD0, 0x8FD4, 0x10021, 0x0D}, /* t5_pm_rx_regs_10021_to_1002c */
+};
+
+static const u32 t5_pm_tx_array[][IREG_NUM_ELEM] = {
+	{0x8FF0, 0x8FF4, 0x10000, 0x20}, /* t5_pm_tx_regs_10000_to_10020 */
+	{0x8FF0, 0x8FF4, 0x10021, 0x1D}, /* t5_pm_tx_regs_10021_to_1003c */
+};
+
+static const u32 t5_pcie_config_array[][2] = {
+	{0x0, 0x34},
+	{0x3c, 0x40},
+	{0x50, 0x64},
+	{0x70, 0x80},
+	{0x94, 0xa0},
+	{0xb0, 0xb8},
+	{0xd0, 0xd4},
+	{0x100, 0x128},
+	{0x140, 0x148},
+	{0x150, 0x164},
+	{0x170, 0x178},
+	{0x180, 0x194},
+	{0x1a0, 0x1b8},
+	{0x1c0, 0x208},
+};
+
+static const u32 t6_ma_ireg_array[][IREG_NUM_ELEM] = {
+	{0x78f8, 0x78fc, 0xa000, 23}, /* t6_ma_regs_a000_to_a016 */
+	{0x78f8, 0x78fc, 0xa400, 30}, /* t6_ma_regs_a400_to_a41e */
+	{0x78f8, 0x78fc, 0xa800, 20} /* t6_ma_regs_a800_to_a813 */
+};
+
+static const u32 t6_ma_ireg_array2[][IREG_NUM_ELEM] = {
+	{0x78f8, 0x78fc, 0xe400, 17}, /* t6_ma_regs_e400_to_e600 */
+	{0x78f8, 0x78fc, 0xe640, 13} /* t6_ma_regs_e640_to_e7c0 */
+};
+
+static const u32 t6_up_cim_reg_array[][IREG_NUM_ELEM + 1] = {
+	{0x7b50, 0x7b54, 0x2000, 0x20, 0}, /* up_cim_2000_to_207c */
+	{0x7b50, 0x7b54, 0x2080, 0x1d, 0}, /* up_cim_2080_to_20fc */
+	{0x7b50, 0x7b54, 0x00, 0x20, 0}, /* up_cim_00_to_7c */
+	{0x7b50, 0x7b54, 0x80, 0x20, 0}, /* up_cim_80_to_fc */
+	{0x7b50, 0x7b54, 0x100, 0x11, 0}, /* up_cim_100_to_14c */
+	{0x7b50, 0x7b54, 0x200, 0x10, 0}, /* up_cim_200_to_23c */
+	{0x7b50, 0x7b54, 0x240, 0x2, 0}, /* up_cim_240_to_244 */
+	{0x7b50, 0x7b54, 0x250, 0x2, 0}, /* up_cim_250_to_254 */
+	{0x7b50, 0x7b54, 0x260, 0x2, 0}, /* up_cim_260_to_264 */
+	{0x7b50, 0x7b54, 0x270, 0x2, 0}, /* up_cim_270_to_274 */
+	{0x7b50, 0x7b54, 0x280, 0x20, 0}, /* up_cim_280_to_2fc */
+	{0x7b50, 0x7b54, 0x300, 0x20, 0}, /* up_cim_300_to_37c */
+	{0x7b50, 0x7b54, 0x380, 0x14, 0}, /* up_cim_380_to_3cc */
+	{0x7b50, 0x7b54, 0x4900, 0x4, 0x4}, /* up_cim_4900_to_4c60 */
+	{0x7b50, 0x7b54, 0x4904, 0x4, 0x4}, /* up_cim_4904_to_4c64 */
+	{0x7b50, 0x7b54, 0x4908, 0x4, 0x4}, /* up_cim_4908_to_4c68 */
+	{0x7b50, 0x7b54, 0x4910, 0x4, 0x4}, /* up_cim_4910_to_4c70 */
+	{0x7b50, 0x7b54, 0x4914, 0x4, 0x4}, /* up_cim_4914_to_4c74 */
+	{0x7b50, 0x7b54, 0x4920, 0x10, 0x10}, /* up_cim_4920_to_4a10 */
+	{0x7b50, 0x7b54, 0x4924, 0x10, 0x10}, /* up_cim_4924_to_4a14 */
+	{0x7b50, 0x7b54, 0x4928, 0x10, 0x10}, /* up_cim_4928_to_4a18 */
+	{0x7b50, 0x7b54, 0x492c, 0x10, 0x10}, /* up_cim_492c_to_4a1c */
+};
+
+static const u32 t5_up_cim_reg_array[][IREG_NUM_ELEM + 1] = {
+	{0x7b50, 0x7b54, 0x2000, 0x20, 0}, /* up_cim_2000_to_207c */
+	{0x7b50, 0x7b54, 0x2080, 0x19, 0}, /* up_cim_2080_to_20ec */
+	{0x7b50, 0x7b54, 0x00, 0x20, 0}, /* up_cim_00_to_7c */
+	{0x7b50, 0x7b54, 0x80, 0x20, 0}, /* up_cim_80_to_fc */
+	{0x7b50, 0x7b54, 0x100, 0x11, 0}, /* up_cim_100_to_14c */
+	{0x7b50, 0x7b54, 0x200, 0x10, 0}, /* up_cim_200_to_23c */
+	{0x7b50, 0x7b54, 0x240, 0x2, 0}, /* up_cim_240_to_244 */
+	{0x7b50, 0x7b54, 0x250, 0x2, 0}, /* up_cim_250_to_254 */
+	{0x7b50, 0x7b54, 0x260, 0x2, 0}, /* up_cim_260_to_264 */
+	{0x7b50, 0x7b54, 0x270, 0x2, 0}, /* up_cim_270_to_274 */
+	{0x7b50, 0x7b54, 0x280, 0x20, 0}, /* up_cim_280_to_2fc */
+	{0x7b50, 0x7b54, 0x300, 0x20, 0}, /* up_cim_300_to_37c */
+	{0x7b50, 0x7b54, 0x380, 0x14, 0}, /* up_cim_380_to_3cc */
+};
+
+static const u32 t6_hma_ireg_array[][IREG_NUM_ELEM] = {
+	{0x51320, 0x51324, 0xa000, 32} /* t6_hma_regs_a000_to_a01f */
+};
+
+u32 cudbg_get_entity_length(struct adapter *adap, u32 entity)
+{
+	struct cudbg_tcam tcam_region = { 0 };
+	u32 value, n = 0, len = 0;
+
+	switch (entity) {
+	case CUDBG_REG_DUMP:
+		switch (CHELSIO_CHIP_VERSION(adap->params.chip)) {
+		case CHELSIO_T4:
+			len = T4_REGMAP_SIZE;
+			break;
+		case CHELSIO_T5:
+		case CHELSIO_T6:
+			len = T5_REGMAP_SIZE;
+			break;
+		default:
+			break;
+		}
+		break;
+	case CUDBG_DEV_LOG:
+		len = adap->params.devlog.size;
+		break;
+	case CUDBG_CIM_LA:
+		if (is_t6(adap->params.chip)) {
+			len = adap->params.cim_la_size / 10 + 1;
+			len *= 10 * sizeof(u32);
+		} else {
+			len = adap->params.cim_la_size / 8;
+			len *= 8 * sizeof(u32);
+		}
+		len += sizeof(u32); /* for reading CIM LA configuration */
+		break;
+	case CUDBG_CIM_MA_LA:
+		len = 2 * CIM_MALA_SIZE * 5 * sizeof(u32);
+		break;
+	case CUDBG_CIM_QCFG:
+		len = sizeof(struct cudbg_cim_qcfg);
+		break;
+	case CUDBG_CIM_IBQ_TP0:
+	case CUDBG_CIM_IBQ_TP1:
+	case CUDBG_CIM_IBQ_ULP:
+	case CUDBG_CIM_IBQ_SGE0:
+	case CUDBG_CIM_IBQ_SGE1:
+	case CUDBG_CIM_IBQ_NCSI:
+		len = CIM_IBQ_SIZE * 4 * sizeof(u32);
+		break;
+	case CUDBG_CIM_OBQ_ULP0:
+		len = cudbg_cim_obq_size(adap, 0);
+		break;
+	case CUDBG_CIM_OBQ_ULP1:
+		len = cudbg_cim_obq_size(adap, 1);
+		break;
+	case CUDBG_CIM_OBQ_ULP2:
+		len = cudbg_cim_obq_size(adap, 2);
+		break;
+	case CUDBG_CIM_OBQ_ULP3:
+		len = cudbg_cim_obq_size(adap, 3);
+		break;
+	case CUDBG_CIM_OBQ_SGE:
+		len = cudbg_cim_obq_size(adap, 4);
+		break;
+	case CUDBG_CIM_OBQ_NCSI:
+		len = cudbg_cim_obq_size(adap, 5);
+		break;
+	case CUDBG_CIM_OBQ_RXQ0:
+		len = cudbg_cim_obq_size(adap, 6);
+		break;
+	case CUDBG_CIM_OBQ_RXQ1:
+		len = cudbg_cim_obq_size(adap, 7);
+		break;
+	case CUDBG_EDC0:
+		value = t4_read_reg(adap, MA_TARGET_MEM_ENABLE_A);
+		if (value & EDRAM0_ENABLE_F) {
+			value = t4_read_reg(adap, MA_EDRAM0_BAR_A);
+			len = EDRAM0_SIZE_G(value);
+		}
+		len = cudbg_mbytes_to_bytes(len);
+		break;
+	case CUDBG_EDC1:
+		value = t4_read_reg(adap, MA_TARGET_MEM_ENABLE_A);
+		if (value & EDRAM1_ENABLE_F) {
+			value = t4_read_reg(adap, MA_EDRAM1_BAR_A);
+			len = EDRAM1_SIZE_G(value);
+		}
+		len = cudbg_mbytes_to_bytes(len);
+		break;
+	case CUDBG_MC0:
+		value = t4_read_reg(adap, MA_TARGET_MEM_ENABLE_A);
+		if (value & EXT_MEM0_ENABLE_F) {
+			value = t4_read_reg(adap, MA_EXT_MEMORY0_BAR_A);
+			len = EXT_MEM0_SIZE_G(value);
+		}
+		len = cudbg_mbytes_to_bytes(len);
+		break;
+	case CUDBG_MC1:
+		value = t4_read_reg(adap, MA_TARGET_MEM_ENABLE_A);
+		if (value & EXT_MEM1_ENABLE_F) {
+			value = t4_read_reg(adap, MA_EXT_MEMORY1_BAR_A);
+			len = EXT_MEM1_SIZE_G(value);
+		}
+		len = cudbg_mbytes_to_bytes(len);
+		break;
+	case CUDBG_RSS:
+		len = t4_chip_rss_size(adap) * sizeof(u16);
+		break;
+	case CUDBG_RSS_VF_CONF:
+		len = adap->params.arch.vfcount *
+		      sizeof(struct cudbg_rss_vf_conf);
+		break;
+	case CUDBG_PATH_MTU:
+		len = NMTUS * sizeof(u16);
+		break;
+	case CUDBG_PM_STATS:
+		len = sizeof(struct cudbg_pm_stats);
+		break;
+	case CUDBG_HW_SCHED:
+		len = sizeof(struct cudbg_hw_sched);
+		break;
+	case CUDBG_TP_INDIRECT:
+		switch (CHELSIO_CHIP_VERSION(adap->params.chip)) {
+		case CHELSIO_T5:
+			n = sizeof(t5_tp_pio_array) +
+			    sizeof(t5_tp_tm_pio_array) +
+			    sizeof(t5_tp_mib_index_array);
+			break;
+		case CHELSIO_T6:
+			n = sizeof(t6_tp_pio_array) +
+			    sizeof(t6_tp_tm_pio_array) +
+			    sizeof(t6_tp_mib_index_array);
+			break;
+		default:
+			break;
+		}
+		n = n / (IREG_NUM_ELEM * sizeof(u32));
+		len = sizeof(struct ireg_buf) * n;
+		break;
+	case CUDBG_SGE_INDIRECT:
+		len = sizeof(struct ireg_buf) * 2 +
+		      sizeof(struct sge_qbase_reg_field);
+		break;
+	case CUDBG_ULPRX_LA:
+		len = sizeof(struct cudbg_ulprx_la);
+		break;
+	case CUDBG_TP_LA:
+		len = sizeof(struct cudbg_tp_la) + TPLA_SIZE * sizeof(u64);
+		break;
+	case CUDBG_MEMINFO:
+		len = sizeof(struct cudbg_ver_hdr) +
+		      sizeof(struct cudbg_meminfo);
+		break;
+	case CUDBG_CIM_PIF_LA:
+		len = sizeof(struct cudbg_cim_pif_la);
+		len += 2 * CIM_PIFLA_SIZE * 6 * sizeof(u32);
+		break;
+	case CUDBG_CLK:
+		len = sizeof(struct cudbg_clk_info);
+		break;
+	case CUDBG_PCIE_INDIRECT:
+		n = sizeof(t5_pcie_pdbg_array) / (IREG_NUM_ELEM * sizeof(u32));
+		len = sizeof(struct ireg_buf) * n * 2;
+		break;
+	case CUDBG_PM_INDIRECT:
+		n = sizeof(t5_pm_rx_array) / (IREG_NUM_ELEM * sizeof(u32));
+		len = sizeof(struct ireg_buf) * n * 2;
+		break;
+	case CUDBG_TID_INFO:
+		len = sizeof(struct cudbg_tid_info_region_rev1);
+		break;
+	case CUDBG_PCIE_CONFIG:
+		len = sizeof(u32) * CUDBG_NUM_PCIE_CONFIG_REGS;
+		break;
+	case CUDBG_DUMP_CONTEXT:
+		len = cudbg_dump_context_size(adap);
+		break;
+	case CUDBG_MPS_TCAM:
+		len = sizeof(struct cudbg_mps_tcam) *
+		      adap->params.arch.mps_tcam_size;
+		break;
+	case CUDBG_VPD_DATA:
+		len = sizeof(struct cudbg_vpd_data);
+		break;
+	case CUDBG_LE_TCAM:
+		cudbg_fill_le_tcam_info(adap, &tcam_region);
+		len = sizeof(struct cudbg_tcam) +
+		      sizeof(struct cudbg_tid_data) * tcam_region.max_tid;
+		break;
+	case CUDBG_CCTRL:
+		len = sizeof(u16) * NMTUS * NCCTRL_WIN;
+		break;
+	case CUDBG_MA_INDIRECT:
+		if (CHELSIO_CHIP_VERSION(adap->params.chip) > CHELSIO_T5) {
+			n = sizeof(t6_ma_ireg_array) /
+			    (IREG_NUM_ELEM * sizeof(u32));
+			len = sizeof(struct ireg_buf) * n * 2;
+		}
+		break;
+	case CUDBG_ULPTX_LA:
+		len = sizeof(struct cudbg_ver_hdr) +
+		      sizeof(struct cudbg_ulptx_la);
+		break;
+	case CUDBG_UP_CIM_INDIRECT:
+		n = 0;
+		if (is_t5(adap->params.chip))
+			n = sizeof(t5_up_cim_reg_array) /
+			    ((IREG_NUM_ELEM + 1) * sizeof(u32));
+		else if (is_t6(adap->params.chip))
+			n = sizeof(t6_up_cim_reg_array) /
+			    ((IREG_NUM_ELEM + 1) * sizeof(u32));
+		len = sizeof(struct ireg_buf) * n;
+		break;
+	case CUDBG_PBT_TABLE:
+		len = sizeof(struct cudbg_pbt_tables);
+		break;
+	case CUDBG_MBOX_LOG:
+		len = sizeof(struct cudbg_mbox_log) * adap->mbox_log->size;
+		break;
+	case CUDBG_HMA_INDIRECT:
+		if (CHELSIO_CHIP_VERSION(adap->params.chip) > CHELSIO_T5) {
+			n = sizeof(t6_hma_ireg_array) /
+			    (IREG_NUM_ELEM * sizeof(u32));
+			len = sizeof(struct ireg_buf) * n;
+		}
+		break;
+	case CUDBG_HMA:
+		value = t4_read_reg(adap, MA_TARGET_MEM_ENABLE_A);
+		if (value & HMA_MUX_F) {
+			/* In T6, there's no MC1.  So, HMA shares MC1
+			 * address space.
+			 */
+			value = t4_read_reg(adap, MA_EXT_MEMORY1_BAR_A);
+			len = EXT_MEM1_SIZE_G(value);
+		}
+		len = cudbg_mbytes_to_bytes(len);
+		break;
+	case CUDBG_QDESC:
+		cudbg_fill_qdesc_num_and_size(adap, NULL, &len);
+		break;
+	default:
+		break;
+	}
+
+	return len;
+}
+
+static int cudbg_do_compression(struct cudbg_init *pdbg_init,
+				struct cudbg_buffer *pin_buff,
+				struct cudbg_buffer *dbg_buff)
+{
+	struct cudbg_buffer temp_in_buff = { 0 };
+	int bytes_left, bytes_read, bytes;
+	u32 offset = dbg_buff->offset;
+	int rc;
+
+	temp_in_buff.offset = pin_buff->offset;
+	temp_in_buff.data = pin_buff->data;
+	temp_in_buff.size = pin_buff->size;
+
+	bytes_left = pin_buff->size;
+	bytes_read = 0;
+	while (bytes_left > 0) {
+		/* Do compression in smaller chunks */
+		bytes = min_t(unsigned long, bytes_left,
+			      (unsigned long)CUDBG_CHUNK_SIZE);
+		temp_in_buff.data = (char *)pin_buff->data + bytes_read;
+		temp_in_buff.size = bytes;
+		rc = cudbg_compress_buff(pdbg_init, &temp_in_buff, dbg_buff);
+		if (rc)
+			return rc;
+		bytes_left -= bytes;
+		bytes_read += bytes;
+	}
+
+	pin_buff->size = dbg_buff->offset - offset;
+	return 0;
+}
+
+static int cudbg_write_and_release_buff(struct cudbg_init *pdbg_init,
+					struct cudbg_buffer *pin_buff,
+					struct cudbg_buffer *dbg_buff)
+{
+	int rc = 0;
+
+	if (pdbg_init->compress_type == CUDBG_COMPRESSION_NONE) {
+		cudbg_update_buff(pin_buff, dbg_buff);
+	} else {
+		rc = cudbg_do_compression(pdbg_init, pin_buff, dbg_buff);
+		if (rc)
+			goto out;
+	}
+
+out:
+	cudbg_put_buff(pdbg_init, pin_buff);
+	return rc;
+}
+
+static int is_fw_attached(struct cudbg_init *pdbg_init)
+{
+	struct adapter *padap = pdbg_init->adap;
+
+	if (!(padap->flags & CXGB4_FW_OK) || padap->use_bd)
+		return 0;
+
+	return 1;
+}
+
+/* This function will add additional padding bytes into debug_buffer to make it
+ * 4 byte aligned.
+ */
+void cudbg_align_debug_buffer(struct cudbg_buffer *dbg_buff,
+			      struct cudbg_entity_hdr *entity_hdr)
+{
+	u8 zero_buf[4] = {0};
+	u8 padding, remain;
+
+	remain = (dbg_buff->offset - entity_hdr->start_offset) % 4;
+	padding = 4 - remain;
+	if (remain) {
+		memcpy(((u8 *)dbg_buff->data) + dbg_buff->offset, &zero_buf,
+		       padding);
+		dbg_buff->offset += padding;
+		entity_hdr->num_pad = padding;
+	}
+	entity_hdr->size = dbg_buff->offset - entity_hdr->start_offset;
+}
+
+struct cudbg_entity_hdr *cudbg_get_entity_hdr(void *outbuf, int i)
+{
+	struct cudbg_hdr *cudbg_hdr = (struct cudbg_hdr *)outbuf;
+
+	return (struct cudbg_entity_hdr *)
+	       ((char *)outbuf + cudbg_hdr->hdr_len +
+		(sizeof(struct cudbg_entity_hdr) * (i - 1)));
+}
+
+static int cudbg_read_vpd_reg(struct adapter *padap, u32 addr, u32 len,
+			      void *dest)
+{
+	int vaddr, rc;
+
+	vaddr = t4_eeprom_ptov(addr, padap->pf, EEPROMPFSIZE);
+	if (vaddr < 0)
+		return vaddr;
+
+	rc = pci_read_vpd(padap->pdev, vaddr, len, dest);
+	if (rc < 0)
+		return rc;
+
+	return 0;
+}
+
+static int cudbg_mem_desc_cmp(const void *a, const void *b)
+{
+	return ((const struct cudbg_mem_desc *)a)->base -
+	       ((const struct cudbg_mem_desc *)b)->base;
+}
+
+int cudbg_fill_meminfo(struct adapter *padap,
+		       struct cudbg_meminfo *meminfo_buff)
+{
+	struct cudbg_mem_desc *md;
+	u32 lo, hi, used, alloc;
+	int n, i;
+
+	memset(meminfo_buff->avail, 0,
+	       ARRAY_SIZE(meminfo_buff->avail) *
+	       sizeof(struct cudbg_mem_desc));
+	memset(meminfo_buff->mem, 0,
+	       (ARRAY_SIZE(cudbg_region) + 3) * sizeof(struct cudbg_mem_desc));
+	md  = meminfo_buff->mem;
+
+	for (i = 0; i < ARRAY_SIZE(meminfo_buff->mem); i++) {
+		meminfo_buff->mem[i].limit = 0;
+		meminfo_buff->mem[i].idx = i;
+	}
+
+	/* Find and sort the populated memory ranges */
+	i = 0;
+	lo = t4_read_reg(padap, MA_TARGET_MEM_ENABLE_A);
+	if (lo & EDRAM0_ENABLE_F) {
+		hi = t4_read_reg(padap, MA_EDRAM0_BAR_A);
+		meminfo_buff->avail[i].base =
+			cudbg_mbytes_to_bytes(EDRAM0_BASE_G(hi));
+		meminfo_buff->avail[i].limit =
+			meminfo_buff->avail[i].base +
+			cudbg_mbytes_to_bytes(EDRAM0_SIZE_G(hi));
+		meminfo_buff->avail[i].idx = 0;
+		i++;
+	}
+
+	if (lo & EDRAM1_ENABLE_F) {
+		hi =  t4_read_reg(padap, MA_EDRAM1_BAR_A);
+		meminfo_buff->avail[i].base =
+			cudbg_mbytes_to_bytes(EDRAM1_BASE_G(hi));
+		meminfo_buff->avail[i].limit =
+			meminfo_buff->avail[i].base +
+			cudbg_mbytes_to_bytes(EDRAM1_SIZE_G(hi));
+		meminfo_buff->avail[i].idx = 1;
+		i++;
+	}
+
+	if (is_t5(padap->params.chip)) {
+		if (lo & EXT_MEM0_ENABLE_F) {
+			hi = t4_read_reg(padap, MA_EXT_MEMORY0_BAR_A);
+			meminfo_buff->avail[i].base =
+				cudbg_mbytes_to_bytes(EXT_MEM_BASE_G(hi));
+			meminfo_buff->avail[i].limit =
+				meminfo_buff->avail[i].base +
+				cudbg_mbytes_to_bytes(EXT_MEM_SIZE_G(hi));
+			meminfo_buff->avail[i].idx = 3;
+			i++;
+		}
+
+		if (lo & EXT_MEM1_ENABLE_F) {
+			hi = t4_read_reg(padap, MA_EXT_MEMORY1_BAR_A);
+			meminfo_buff->avail[i].base =
+				cudbg_mbytes_to_bytes(EXT_MEM1_BASE_G(hi));
+			meminfo_buff->avail[i].limit =
+				meminfo_buff->avail[i].base +
+				cudbg_mbytes_to_bytes(EXT_MEM1_SIZE_G(hi));
+			meminfo_buff->avail[i].idx = 4;
+			i++;
+		}
+	} else {
+		if (lo & EXT_MEM_ENABLE_F) {
+			hi = t4_read_reg(padap, MA_EXT_MEMORY_BAR_A);
+			meminfo_buff->avail[i].base =
+				cudbg_mbytes_to_bytes(EXT_MEM_BASE_G(hi));
+			meminfo_buff->avail[i].limit =
+				meminfo_buff->avail[i].base +
+				cudbg_mbytes_to_bytes(EXT_MEM_SIZE_G(hi));
+			meminfo_buff->avail[i].idx = 2;
+			i++;
+		}
+
+		if (lo & HMA_MUX_F) {
+			hi = t4_read_reg(padap, MA_EXT_MEMORY1_BAR_A);
+			meminfo_buff->avail[i].base =
+				cudbg_mbytes_to_bytes(EXT_MEM1_BASE_G(hi));
+			meminfo_buff->avail[i].limit =
+				meminfo_buff->avail[i].base +
+				cudbg_mbytes_to_bytes(EXT_MEM1_SIZE_G(hi));
+			meminfo_buff->avail[i].idx = 5;
+			i++;
+		}
+	}
+
+	if (!i) /* no memory available */
+		return CUDBG_STATUS_ENTITY_NOT_FOUND;
+
+	meminfo_buff->avail_c = i;
+	sort(meminfo_buff->avail, i, sizeof(struct cudbg_mem_desc),
+	     cudbg_mem_desc_cmp, NULL);
+	(md++)->base = t4_read_reg(padap, SGE_DBQ_CTXT_BADDR_A);
+	(md++)->base = t4_read_reg(padap, SGE_IMSG_CTXT_BADDR_A);
+	(md++)->base = t4_read_reg(padap, SGE_FLM_CACHE_BADDR_A);
+	(md++)->base = t4_read_reg(padap, TP_CMM_TCB_BASE_A);
+	(md++)->base = t4_read_reg(padap, TP_CMM_MM_BASE_A);
+	(md++)->base = t4_read_reg(padap, TP_CMM_TIMER_BASE_A);
+	(md++)->base = t4_read_reg(padap, TP_CMM_MM_RX_FLST_BASE_A);
+	(md++)->base = t4_read_reg(padap, TP_CMM_MM_TX_FLST_BASE_A);
+	(md++)->base = t4_read_reg(padap, TP_CMM_MM_PS_FLST_BASE_A);
+
+	/* the next few have explicit upper bounds */
+	md->base = t4_read_reg(padap, TP_PMM_TX_BASE_A);
+	md->limit = md->base - 1 +
+		    t4_read_reg(padap, TP_PMM_TX_PAGE_SIZE_A) *
+		    PMTXMAXPAGE_G(t4_read_reg(padap, TP_PMM_TX_MAX_PAGE_A));
+	md++;
+
+	md->base = t4_read_reg(padap, TP_PMM_RX_BASE_A);
+	md->limit = md->base - 1 +
+		    t4_read_reg(padap, TP_PMM_RX_PAGE_SIZE_A) *
+		    PMRXMAXPAGE_G(t4_read_reg(padap, TP_PMM_RX_MAX_PAGE_A));
+	md++;
+
+	if (t4_read_reg(padap, LE_DB_CONFIG_A) & HASHEN_F) {
+		if (CHELSIO_CHIP_VERSION(padap->params.chip) <= CHELSIO_T5) {
+			hi = t4_read_reg(padap, LE_DB_TID_HASHBASE_A) / 4;
+			md->base = t4_read_reg(padap, LE_DB_HASH_TID_BASE_A);
+		} else {
+			hi = t4_read_reg(padap, LE_DB_HASH_TID_BASE_A);
+			md->base = t4_read_reg(padap,
+					       LE_DB_HASH_TBL_BASE_ADDR_A);
+		}
+		md->limit = 0;
+	} else {
+		md->base = 0;
+		md->idx = ARRAY_SIZE(cudbg_region);  /* hide it */
+	}
+	md++;
+
+#define ulp_region(reg) do { \
+	md->base = t4_read_reg(padap, ULP_ ## reg ## _LLIMIT_A);\
+	(md++)->limit = t4_read_reg(padap, ULP_ ## reg ## _ULIMIT_A);\
+} while (0)
+
+	ulp_region(RX_ISCSI);
+	ulp_region(RX_TDDP);
+	ulp_region(TX_TPT);
+	ulp_region(RX_STAG);
+	ulp_region(RX_RQ);
+	ulp_region(RX_RQUDP);
+	ulp_region(RX_PBL);
+	ulp_region(TX_PBL);
+#undef ulp_region
+	md->base = 0;
+	md->idx = ARRAY_SIZE(cudbg_region);
+	if (!is_t4(padap->params.chip)) {
+		u32 fifo_size = t4_read_reg(padap, SGE_DBVFIFO_SIZE_A);
+		u32 sge_ctrl = t4_read_reg(padap, SGE_CONTROL2_A);
+		u32 size = 0;
+
+		if (is_t5(padap->params.chip)) {
+			if (sge_ctrl & VFIFO_ENABLE_F)
+				size = DBVFIFO_SIZE_G(fifo_size);
+		} else {
+			size = T6_DBVFIFO_SIZE_G(fifo_size);
+		}
+
+		if (size) {
+			md->base = BASEADDR_G(t4_read_reg(padap,
+							  SGE_DBVFIFO_BADDR_A));
+			md->limit = md->base + (size << 2) - 1;
+		}
+	}
+
+	md++;
+
+	md->base = t4_read_reg(padap, ULP_RX_CTX_BASE_A);
+	md->limit = 0;
+	md++;
+	md->base = t4_read_reg(padap, ULP_TX_ERR_TABLE_BASE_A);
+	md->limit = 0;
+	md++;
+
+	md->base = padap->vres.ocq.start;
+	if (padap->vres.ocq.size)
+		md->limit = md->base + padap->vres.ocq.size - 1;
+	else
+		md->idx = ARRAY_SIZE(cudbg_region);  /* hide it */
+	md++;
+
+	/* add any address-space holes, there can be up to 3 */
+	for (n = 0; n < i - 1; n++)
+		if (meminfo_buff->avail[n].limit <
+		    meminfo_buff->avail[n + 1].base)
+			(md++)->base = meminfo_buff->avail[n].limit;
+
+	if (meminfo_buff->avail[n].limit)
+		(md++)->base = meminfo_buff->avail[n].limit;
+
+	n = md - meminfo_buff->mem;
+	meminfo_buff->mem_c = n;
+
+	sort(meminfo_buff->mem, n, sizeof(struct cudbg_mem_desc),
+	     cudbg_mem_desc_cmp, NULL);
+
+	lo = t4_read_reg(padap, CIM_SDRAM_BASE_ADDR_A);
+	hi = t4_read_reg(padap, CIM_SDRAM_ADDR_SIZE_A) + lo - 1;
+	meminfo_buff->up_ram_lo = lo;
+	meminfo_buff->up_ram_hi = hi;
+
+	lo = t4_read_reg(padap, CIM_EXTMEM2_BASE_ADDR_A);
+	hi = t4_read_reg(padap, CIM_EXTMEM2_ADDR_SIZE_A) + lo - 1;
+	meminfo_buff->up_extmem2_lo = lo;
+	meminfo_buff->up_extmem2_hi = hi;
+
+	lo = t4_read_reg(padap, TP_PMM_RX_MAX_PAGE_A);
+	for (i = 0, meminfo_buff->free_rx_cnt = 0; i < 2; i++)
+		meminfo_buff->free_rx_cnt +=
+			FREERXPAGECOUNT_G(t4_read_reg(padap,
+						      TP_FLM_FREE_RX_CNT_A));
+
+	meminfo_buff->rx_pages_data[0] =  PMRXMAXPAGE_G(lo);
+	meminfo_buff->rx_pages_data[1] =
+		t4_read_reg(padap, TP_PMM_RX_PAGE_SIZE_A) >> 10;
+	meminfo_buff->rx_pages_data[2] = (lo & PMRXNUMCHN_F) ? 2 : 1;
+
+	lo = t4_read_reg(padap, TP_PMM_TX_MAX_PAGE_A);
+	hi = t4_read_reg(padap, TP_PMM_TX_PAGE_SIZE_A);
+	for (i = 0, meminfo_buff->free_tx_cnt = 0; i < 4; i++)
+		meminfo_buff->free_tx_cnt +=
+			FREETXPAGECOUNT_G(t4_read_reg(padap,
+						      TP_FLM_FREE_TX_CNT_A));
+
+	meminfo_buff->tx_pages_data[0] = PMTXMAXPAGE_G(lo);
+	meminfo_buff->tx_pages_data[1] =
+		hi >= (1 << 20) ? (hi >> 20) : (hi >> 10);
+	meminfo_buff->tx_pages_data[2] =
+		hi >= (1 << 20) ? 'M' : 'K';
+	meminfo_buff->tx_pages_data[3] = 1 << PMTXNUMCHN_G(lo);
+
+	meminfo_buff->p_structs = t4_read_reg(padap, TP_CMM_MM_MAX_PSTRUCT_A);
+	meminfo_buff->p_structs_free_cnt =
+		FREEPSTRUCTCOUNT_G(t4_read_reg(padap, TP_FLM_FREE_PS_CNT_A));
+
+	for (i = 0; i < 4; i++) {
+		if (CHELSIO_CHIP_VERSION(padap->params.chip) > CHELSIO_T5)
+			lo = t4_read_reg(padap,
+					 MPS_RX_MAC_BG_PG_CNT0_A + i * 4);
+		else
+			lo = t4_read_reg(padap, MPS_RX_PG_RSV0_A + i * 4);
+		if (is_t5(padap->params.chip)) {
+			used = T5_USED_G(lo);
+			alloc = T5_ALLOC_G(lo);
+		} else {
+			used = USED_G(lo);
+			alloc = ALLOC_G(lo);
+		}
+		meminfo_buff->port_used[i] = used;
+		meminfo_buff->port_alloc[i] = alloc;
+	}
+
+	for (i = 0; i < padap->params.arch.nchan; i++) {
+		if (CHELSIO_CHIP_VERSION(padap->params.chip) > CHELSIO_T5)
+			lo = t4_read_reg(padap,
+					 MPS_RX_LPBK_BG_PG_CNT0_A + i * 4);
+		else
+			lo = t4_read_reg(padap, MPS_RX_PG_RSV4_A + i * 4);
+		if (is_t5(padap->params.chip)) {
+			used = T5_USED_G(lo);
+			alloc = T5_ALLOC_G(lo);
+		} else {
+			used = USED_G(lo);
+			alloc = ALLOC_G(lo);
+		}
+		meminfo_buff->loopback_used[i] = used;
+		meminfo_buff->loopback_alloc[i] = alloc;
+	}
+
+	return 0;
+}
+
+int cudbg_collect_reg_dump(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	u32 buf_size = 0;
+	int rc = 0;
+
+	if (is_t4(padap->params.chip))
+		buf_size = T4_REGMAP_SIZE;
+	else if (is_t5(padap->params.chip) || is_t6(padap->params.chip))
+		buf_size = T5_REGMAP_SIZE;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, buf_size, &temp_buff);
+	if (rc)
+		return rc;
+	t4_get_regs(padap, (void *)temp_buff.data, temp_buff.size);
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_fw_devlog(struct cudbg_init *pdbg_init,
+			    struct cudbg_buffer *dbg_buff,
+			    struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct devlog_params *dparams;
+	int rc = 0;
+
+	rc = t4_init_devlog_params(padap);
+	if (rc < 0) {
+		cudbg_err->sys_err = rc;
+		return rc;
+	}
+
+	dparams = &padap->params.devlog;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, dparams->size, &temp_buff);
+	if (rc)
+		return rc;
+
+	/* Collect FW devlog */
+	if (dparams->start != 0) {
+		spin_lock(&padap->win0_lock);
+		rc = t4_memory_rw(padap, padap->params.drv_memwin,
+				  dparams->memtype, dparams->start,
+				  dparams->size,
+				  (__be32 *)(char *)temp_buff.data,
+				  1);
+		spin_unlock(&padap->win0_lock);
+		if (rc) {
+			cudbg_err->sys_err = rc;
+			cudbg_put_buff(pdbg_init, &temp_buff);
+			return rc;
+		}
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_cim_la(struct cudbg_init *pdbg_init,
+			 struct cudbg_buffer *dbg_buff,
+			 struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	int size, rc;
+	u32 cfg = 0;
+
+	if (is_t6(padap->params.chip)) {
+		size = padap->params.cim_la_size / 10 + 1;
+		size *= 10 * sizeof(u32);
+	} else {
+		size = padap->params.cim_la_size / 8;
+		size *= 8 * sizeof(u32);
+	}
+
+	size += sizeof(cfg);
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	rc = t4_cim_read(padap, UP_UP_DBG_LA_CFG_A, 1, &cfg);
+	if (rc) {
+		cudbg_err->sys_err = rc;
+		cudbg_put_buff(pdbg_init, &temp_buff);
+		return rc;
+	}
+
+	memcpy((char *)temp_buff.data, &cfg, sizeof(cfg));
+	rc = t4_cim_read_la(padap,
+			    (u32 *)((char *)temp_buff.data + sizeof(cfg)),
+			    NULL);
+	if (rc < 0) {
+		cudbg_err->sys_err = rc;
+		cudbg_put_buff(pdbg_init, &temp_buff);
+		return rc;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_cim_ma_la(struct cudbg_init *pdbg_init,
+			    struct cudbg_buffer *dbg_buff,
+			    struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	int size, rc;
+
+	size = 2 * CIM_MALA_SIZE * 5 * sizeof(u32);
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	t4_cim_read_ma_la(padap,
+			  (u32 *)temp_buff.data,
+			  (u32 *)((char *)temp_buff.data +
+				  5 * CIM_MALA_SIZE));
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_cim_qcfg(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_cim_qcfg *cim_qcfg_data;
+	int rc;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, sizeof(struct cudbg_cim_qcfg),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	cim_qcfg_data = (struct cudbg_cim_qcfg *)temp_buff.data;
+	cim_qcfg_data->chip = padap->params.chip;
+	rc = t4_cim_read(padap, UP_IBQ_0_RDADDR_A,
+			 ARRAY_SIZE(cim_qcfg_data->stat), cim_qcfg_data->stat);
+	if (rc) {
+		cudbg_err->sys_err = rc;
+		cudbg_put_buff(pdbg_init, &temp_buff);
+		return rc;
+	}
+
+	rc = t4_cim_read(padap, UP_OBQ_0_REALADDR_A,
+			 ARRAY_SIZE(cim_qcfg_data->obq_wr),
+			 cim_qcfg_data->obq_wr);
+	if (rc) {
+		cudbg_err->sys_err = rc;
+		cudbg_put_buff(pdbg_init, &temp_buff);
+		return rc;
+	}
+
+	t4_read_cimq_cfg(padap, cim_qcfg_data->base, cim_qcfg_data->size,
+			 cim_qcfg_data->thres);
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+static int cudbg_read_cim_ibq(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err, int qid)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	int no_of_read_words, rc = 0;
+	u32 qsize;
+
+	/* collect CIM IBQ */
+	qsize = CIM_IBQ_SIZE * 4 * sizeof(u32);
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, qsize, &temp_buff);
+	if (rc)
+		return rc;
+
+	/* t4_read_cim_ibq will return no. of read words or error */
+	no_of_read_words = t4_read_cim_ibq(padap, qid,
+					   (u32 *)temp_buff.data, qsize);
+	/* no_of_read_words is less than or equal to 0 means error */
+	if (no_of_read_words <= 0) {
+		if (!no_of_read_words)
+			rc = CUDBG_SYSTEM_ERROR;
+		else
+			rc = no_of_read_words;
+		cudbg_err->sys_err = rc;
+		cudbg_put_buff(pdbg_init, &temp_buff);
+		return rc;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_cim_ibq_tp0(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_ibq(pdbg_init, dbg_buff, cudbg_err, 0);
+}
+
+int cudbg_collect_cim_ibq_tp1(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_ibq(pdbg_init, dbg_buff, cudbg_err, 1);
+}
+
+int cudbg_collect_cim_ibq_ulp(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_ibq(pdbg_init, dbg_buff, cudbg_err, 2);
+}
+
+int cudbg_collect_cim_ibq_sge0(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_ibq(pdbg_init, dbg_buff, cudbg_err, 3);
+}
+
+int cudbg_collect_cim_ibq_sge1(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_ibq(pdbg_init, dbg_buff, cudbg_err, 4);
+}
+
+int cudbg_collect_cim_ibq_ncsi(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_ibq(pdbg_init, dbg_buff, cudbg_err, 5);
+}
+
+u32 cudbg_cim_obq_size(struct adapter *padap, int qid)
+{
+	u32 value;
+
+	t4_write_reg(padap, CIM_QUEUE_CONFIG_REF_A, OBQSELECT_F |
+		     QUENUMSELECT_V(qid));
+	value = t4_read_reg(padap, CIM_QUEUE_CONFIG_CTRL_A);
+	value = CIMQSIZE_G(value) * 64; /* size in number of words */
+	return value * sizeof(u32);
+}
+
+static int cudbg_read_cim_obq(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err, int qid)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	int no_of_read_words, rc = 0;
+	u32 qsize;
+
+	/* collect CIM OBQ */
+	qsize =  cudbg_cim_obq_size(padap, qid);
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, qsize, &temp_buff);
+	if (rc)
+		return rc;
+
+	/* t4_read_cim_obq will return no. of read words or error */
+	no_of_read_words = t4_read_cim_obq(padap, qid,
+					   (u32 *)temp_buff.data, qsize);
+	/* no_of_read_words is less than or equal to 0 means error */
+	if (no_of_read_words <= 0) {
+		if (!no_of_read_words)
+			rc = CUDBG_SYSTEM_ERROR;
+		else
+			rc = no_of_read_words;
+		cudbg_err->sys_err = rc;
+		cudbg_put_buff(pdbg_init, &temp_buff);
+		return rc;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_cim_obq_ulp0(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_obq(pdbg_init, dbg_buff, cudbg_err, 0);
+}
+
+int cudbg_collect_cim_obq_ulp1(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_obq(pdbg_init, dbg_buff, cudbg_err, 1);
+}
+
+int cudbg_collect_cim_obq_ulp2(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_obq(pdbg_init, dbg_buff, cudbg_err, 2);
+}
+
+int cudbg_collect_cim_obq_ulp3(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_obq(pdbg_init, dbg_buff, cudbg_err, 3);
+}
+
+int cudbg_collect_cim_obq_sge(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_obq(pdbg_init, dbg_buff, cudbg_err, 4);
+}
+
+int cudbg_collect_cim_obq_ncsi(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_obq(pdbg_init, dbg_buff, cudbg_err, 5);
+}
+
+int cudbg_collect_obq_sge_rx_q0(struct cudbg_init *pdbg_init,
+				struct cudbg_buffer *dbg_buff,
+				struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_obq(pdbg_init, dbg_buff, cudbg_err, 6);
+}
+
+int cudbg_collect_obq_sge_rx_q1(struct cudbg_init *pdbg_init,
+				struct cudbg_buffer *dbg_buff,
+				struct cudbg_error *cudbg_err)
+{
+	return cudbg_read_cim_obq(pdbg_init, dbg_buff, cudbg_err, 7);
+}
+
+static int cudbg_meminfo_get_mem_index(struct adapter *padap,
+				       struct cudbg_meminfo *mem_info,
+				       u8 mem_type, u8 *idx)
+{
+	u8 i, flag;
+
+	switch (mem_type) {
+	case MEM_EDC0:
+		flag = EDC0_FLAG;
+		break;
+	case MEM_EDC1:
+		flag = EDC1_FLAG;
+		break;
+	case MEM_MC0:
+		/* Some T5 cards have both MC0 and MC1. */
+		flag = is_t5(padap->params.chip) ? MC0_FLAG : MC_FLAG;
+		break;
+	case MEM_MC1:
+		flag = MC1_FLAG;
+		break;
+	case MEM_HMA:
+		flag = HMA_FLAG;
+		break;
+	default:
+		return CUDBG_STATUS_ENTITY_NOT_FOUND;
+	}
+
+	for (i = 0; i < mem_info->avail_c; i++) {
+		if (mem_info->avail[i].idx == flag) {
+			*idx = i;
+			return 0;
+		}
+	}
+
+	return CUDBG_STATUS_ENTITY_NOT_FOUND;
+}
+
+/* Fetch the @region_name's start and end from @meminfo. */
+static int cudbg_get_mem_region(struct adapter *padap,
+				struct cudbg_meminfo *meminfo,
+				u8 mem_type, const char *region_name,
+				struct cudbg_mem_desc *mem_desc)
+{
+	u8 mc, found = 0;
+	u32 idx = 0;
+	int rc, i;
+
+	rc = cudbg_meminfo_get_mem_index(padap, meminfo, mem_type, &mc);
+	if (rc)
+		return rc;
+
+	i = match_string(cudbg_region, ARRAY_SIZE(cudbg_region), region_name);
+	if (i < 0)
+		return -EINVAL;
+
+	idx = i;
+	for (i = 0; i < meminfo->mem_c; i++) {
+		if (meminfo->mem[i].idx >= ARRAY_SIZE(cudbg_region))
+			continue; /* Skip holes */
+
+		if (!(meminfo->mem[i].limit))
+			meminfo->mem[i].limit =
+				i < meminfo->mem_c - 1 ?
+				meminfo->mem[i + 1].base - 1 : ~0;
+
+		if (meminfo->mem[i].idx == idx) {
+			/* Check if the region exists in @mem_type memory */
+			if (meminfo->mem[i].base < meminfo->avail[mc].base &&
+			    meminfo->mem[i].limit < meminfo->avail[mc].base)
+				return -EINVAL;
+
+			if (meminfo->mem[i].base > meminfo->avail[mc].limit)
+				return -EINVAL;
+
+			memcpy(mem_desc, &meminfo->mem[i],
+			       sizeof(struct cudbg_mem_desc));
+			found = 1;
+			break;
+		}
+	}
+	if (!found)
+		return -EINVAL;
+
+	return 0;
+}
+
+/* Fetch and update the start and end of the requested memory region w.r.t 0
+ * in the corresponding EDC/MC/HMA.
+ */
+static int cudbg_get_mem_relative(struct adapter *padap,
+				  struct cudbg_meminfo *meminfo,
+				  u8 mem_type, u32 *out_base, u32 *out_end)
+{
+	u8 mc_idx;
+	int rc;
+
+	rc = cudbg_meminfo_get_mem_index(padap, meminfo, mem_type, &mc_idx);
+	if (rc)
+		return rc;
+
+	if (*out_base < meminfo->avail[mc_idx].base)
+		*out_base = 0;
+	else
+		*out_base -= meminfo->avail[mc_idx].base;
+
+	if (*out_end > meminfo->avail[mc_idx].limit)
+		*out_end = meminfo->avail[mc_idx].limit;
+	else
+		*out_end -= meminfo->avail[mc_idx].base;
+
+	return 0;
+}
+
+/* Get TX and RX Payload region */
+static int cudbg_get_payload_range(struct adapter *padap, u8 mem_type,
+				   const char *region_name,
+				   struct cudbg_region_info *payload)
+{
+	struct cudbg_mem_desc mem_desc = { 0 };
+	struct cudbg_meminfo meminfo;
+	int rc;
+
+	rc = cudbg_fill_meminfo(padap, &meminfo);
+	if (rc)
+		return rc;
+
+	rc = cudbg_get_mem_region(padap, &meminfo, mem_type, region_name,
+				  &mem_desc);
+	if (rc) {
+		payload->exist = false;
+		return 0;
+	}
+
+	payload->exist = true;
+	payload->start = mem_desc.base;
+	payload->end = mem_desc.limit;
+
+	return cudbg_get_mem_relative(padap, &meminfo, mem_type,
+				      &payload->start, &payload->end);
+}
+
+static int cudbg_memory_read(struct cudbg_init *pdbg_init, int win,
+			     int mtype, u32 addr, u32 len, void *hbuf)
+{
+	u32 win_pf, memoffset, mem_aperture, mem_base;
+	struct adapter *adap = pdbg_init->adap;
+	u32 pos, offset, resid;
+	u32 *res_buf;
+	u64 *buf;
+	int ret;
+
+	/* Argument sanity checks ...
+	 */
+	if (addr & 0x3 || (uintptr_t)hbuf & 0x3)
+		return -EINVAL;
+
+	buf = (u64 *)hbuf;
+
+	/* Try to do 64-bit reads.  Residual will be handled later. */
+	resid = len & 0x7;
+	len -= resid;
+
+	ret = t4_memory_rw_init(adap, win, mtype, &memoffset, &mem_base,
+				&mem_aperture);
+	if (ret)
+		return ret;
+
+	addr = addr + memoffset;
+	win_pf = is_t4(adap->params.chip) ? 0 : PFNUM_V(adap->pf);
+
+	pos = addr & ~(mem_aperture - 1);
+	offset = addr - pos;
+
+	/* Set up initial PCI-E Memory Window to cover the start of our
+	 * transfer.
+	 */
+	t4_memory_update_win(adap, win, pos | win_pf);
+
+	/* Transfer data from the adapter */
+	while (len > 0) {
+		*buf++ = le64_to_cpu((__force __le64)
+				     t4_read_reg64(adap, mem_base + offset));
+		offset += sizeof(u64);
+		len -= sizeof(u64);
+
+		/* If we've reached the end of our current window aperture,
+		 * move the PCI-E Memory Window on to the next.
+		 */
+		if (offset == mem_aperture) {
+			pos += mem_aperture;
+			offset = 0;
+			t4_memory_update_win(adap, win, pos | win_pf);
+		}
+	}
+
+	res_buf = (u32 *)buf;
+	/* Read residual in 32-bit multiples */
+	while (resid > sizeof(u32)) {
+		*res_buf++ = le32_to_cpu((__force __le32)
+					 t4_read_reg(adap, mem_base + offset));
+		offset += sizeof(u32);
+		resid -= sizeof(u32);
+
+		/* If we've reached the end of our current window aperture,
+		 * move the PCI-E Memory Window on to the next.
+		 */
+		if (offset == mem_aperture) {
+			pos += mem_aperture;
+			offset = 0;
+			t4_memory_update_win(adap, win, pos | win_pf);
+		}
+	}
+
+	/* Transfer residual < 32-bits */
+	if (resid)
+		t4_memory_rw_residual(adap, resid, mem_base + offset,
+				      (u8 *)res_buf, T4_MEMORY_READ);
+
+	return 0;
+}
+
+#define CUDBG_YIELD_ITERATION 256
+
+static int cudbg_read_fw_mem(struct cudbg_init *pdbg_init,
+			     struct cudbg_buffer *dbg_buff, u8 mem_type,
+			     unsigned long tot_len,
+			     struct cudbg_error *cudbg_err)
+{
+	static const char * const region_name[] = { "Tx payload:",
+						    "Rx payload:" };
+	unsigned long bytes, bytes_left, bytes_read = 0;
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_region_info payload[2];
+	u32 yield_count = 0;
+	int rc = 0;
+	u8 i;
+
+	/* Get TX/RX Payload region range if they exist */
+	memset(payload, 0, sizeof(payload));
+	for (i = 0; i < ARRAY_SIZE(region_name); i++) {
+		rc = cudbg_get_payload_range(padap, mem_type, region_name[i],
+					     &payload[i]);
+		if (rc)
+			return rc;
+
+		if (payload[i].exist) {
+			/* Align start and end to avoid wrap around */
+			payload[i].start = roundup(payload[i].start,
+						   CUDBG_CHUNK_SIZE);
+			payload[i].end = rounddown(payload[i].end,
+						   CUDBG_CHUNK_SIZE);
+		}
+	}
+
+	bytes_left = tot_len;
+	while (bytes_left > 0) {
+		/* As MC size is huge and read through PIO access, this
+		 * loop will hold cpu for a longer time. OS may think that
+		 * the process is hanged and will generate CPU stall traces.
+		 * So yield the cpu regularly.
+		 */
+		yield_count++;
+		if (!(yield_count % CUDBG_YIELD_ITERATION))
+			schedule();
+
+		bytes = min_t(unsigned long, bytes_left,
+			      (unsigned long)CUDBG_CHUNK_SIZE);
+		rc = cudbg_get_buff(pdbg_init, dbg_buff, bytes, &temp_buff);
+		if (rc)
+			return rc;
+
+		for (i = 0; i < ARRAY_SIZE(payload); i++)
+			if (payload[i].exist &&
+			    bytes_read >= payload[i].start &&
+			    bytes_read + bytes <= payload[i].end)
+				/* TX and RX Payload regions can't overlap */
+				goto skip_read;
+
+		spin_lock(&padap->win0_lock);
+		rc = cudbg_memory_read(pdbg_init, MEMWIN_NIC, mem_type,
+				       bytes_read, bytes, temp_buff.data);
+		spin_unlock(&padap->win0_lock);
+		if (rc) {
+			cudbg_err->sys_err = rc;
+			cudbg_put_buff(pdbg_init, &temp_buff);
+			return rc;
+		}
+
+skip_read:
+		bytes_left -= bytes;
+		bytes_read += bytes;
+		rc = cudbg_write_and_release_buff(pdbg_init, &temp_buff,
+						  dbg_buff);
+		if (rc) {
+			cudbg_put_buff(pdbg_init, &temp_buff);
+			return rc;
+		}
+	}
+	return rc;
+}
+
+static void cudbg_t4_fwcache(struct cudbg_init *pdbg_init,
+			     struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	int rc;
+
+	if (is_fw_attached(pdbg_init)) {
+		/* Flush uP dcache before reading edcX/mcX  */
+		rc = t4_fwcache(padap, FW_PARAM_DEV_FWCACHE_FLUSH);
+		if (rc)
+			cudbg_err->sys_warn = rc;
+	}
+}
+
+static int cudbg_mem_region_size(struct cudbg_init *pdbg_init,
+				 struct cudbg_error *cudbg_err,
+				 u8 mem_type, unsigned long *region_size)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_meminfo mem_info;
+	u8 mc_idx;
+	int rc;
+
+	memset(&mem_info, 0, sizeof(struct cudbg_meminfo));
+	rc = cudbg_fill_meminfo(padap, &mem_info);
+	if (rc) {
+		cudbg_err->sys_err = rc;
+		return rc;
+	}
+
+	cudbg_t4_fwcache(pdbg_init, cudbg_err);
+	rc = cudbg_meminfo_get_mem_index(padap, &mem_info, mem_type, &mc_idx);
+	if (rc) {
+		cudbg_err->sys_err = rc;
+		return rc;
+	}
+
+	if (region_size)
+		*region_size = mem_info.avail[mc_idx].limit -
+			       mem_info.avail[mc_idx].base;
+
+	return 0;
+}
+
+static int cudbg_collect_mem_region(struct cudbg_init *pdbg_init,
+				    struct cudbg_buffer *dbg_buff,
+				    struct cudbg_error *cudbg_err,
+				    u8 mem_type)
+{
+	unsigned long size = 0;
+	int rc;
+
+	rc = cudbg_mem_region_size(pdbg_init, cudbg_err, mem_type, &size);
+	if (rc)
+		return rc;
+
+	return cudbg_read_fw_mem(pdbg_init, dbg_buff, mem_type, size,
+				 cudbg_err);
+}
+
+int cudbg_collect_edc0_meminfo(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	return cudbg_collect_mem_region(pdbg_init, dbg_buff, cudbg_err,
+					MEM_EDC0);
+}
+
+int cudbg_collect_edc1_meminfo(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	return cudbg_collect_mem_region(pdbg_init, dbg_buff, cudbg_err,
+					MEM_EDC1);
+}
+
+int cudbg_collect_mc0_meminfo(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err)
+{
+	return cudbg_collect_mem_region(pdbg_init, dbg_buff, cudbg_err,
+					MEM_MC0);
+}
+
+int cudbg_collect_mc1_meminfo(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err)
+{
+	return cudbg_collect_mem_region(pdbg_init, dbg_buff, cudbg_err,
+					MEM_MC1);
+}
+
+int cudbg_collect_hma_meminfo(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err)
+{
+	return cudbg_collect_mem_region(pdbg_init, dbg_buff, cudbg_err,
+					MEM_HMA);
+}
+
+int cudbg_collect_rss(struct cudbg_init *pdbg_init,
+		      struct cudbg_buffer *dbg_buff,
+		      struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	int rc, nentries;
+
+	nentries = t4_chip_rss_size(padap);
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, nentries * sizeof(u16),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	rc = t4_read_rss(padap, (u16 *)temp_buff.data);
+	if (rc) {
+		cudbg_err->sys_err = rc;
+		cudbg_put_buff(pdbg_init, &temp_buff);
+		return rc;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_rss_vf_config(struct cudbg_init *pdbg_init,
+				struct cudbg_buffer *dbg_buff,
+				struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_rss_vf_conf *vfconf;
+	int vf, rc, vf_count;
+
+	vf_count = padap->params.arch.vfcount;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff,
+			    vf_count * sizeof(struct cudbg_rss_vf_conf),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	vfconf = (struct cudbg_rss_vf_conf *)temp_buff.data;
+	for (vf = 0; vf < vf_count; vf++)
+		t4_read_rss_vf_config(padap, vf, &vfconf[vf].rss_vf_vfl,
+				      &vfconf[vf].rss_vf_vfh, true);
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_path_mtu(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	int rc;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, NMTUS * sizeof(u16),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	t4_read_mtu_tbl(padap, (u16 *)temp_buff.data, NULL);
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_pm_stats(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_pm_stats *pm_stats_buff;
+	int rc;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, sizeof(struct cudbg_pm_stats),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	pm_stats_buff = (struct cudbg_pm_stats *)temp_buff.data;
+	t4_pmtx_get_stats(padap, pm_stats_buff->tx_cnt, pm_stats_buff->tx_cyc);
+	t4_pmrx_get_stats(padap, pm_stats_buff->rx_cnt, pm_stats_buff->rx_cyc);
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_hw_sched(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_hw_sched *hw_sched_buff;
+	int i, rc = 0;
+
+	if (!padap->params.vpd.cclk)
+		return CUDBG_STATUS_CCLK_NOT_DEFINED;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, sizeof(struct cudbg_hw_sched),
+			    &temp_buff);
+
+	if (rc)
+		return rc;
+
+	hw_sched_buff = (struct cudbg_hw_sched *)temp_buff.data;
+	hw_sched_buff->map = t4_read_reg(padap, TP_TX_MOD_QUEUE_REQ_MAP_A);
+	hw_sched_buff->mode = TIMERMODE_G(t4_read_reg(padap, TP_MOD_CONFIG_A));
+	t4_read_pace_tbl(padap, hw_sched_buff->pace_tab);
+	for (i = 0; i < NTX_SCHED; ++i)
+		t4_get_tx_sched(padap, i, &hw_sched_buff->kbps[i],
+				&hw_sched_buff->ipg[i], true);
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_tp_indirect(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct ireg_buf *ch_tp_pio;
+	int i, rc, n = 0;
+	u32 size;
+
+	if (is_t5(padap->params.chip))
+		n = sizeof(t5_tp_pio_array) +
+		    sizeof(t5_tp_tm_pio_array) +
+		    sizeof(t5_tp_mib_index_array);
+	else
+		n = sizeof(t6_tp_pio_array) +
+		    sizeof(t6_tp_tm_pio_array) +
+		    sizeof(t6_tp_mib_index_array);
+
+	n = n / (IREG_NUM_ELEM * sizeof(u32));
+	size = sizeof(struct ireg_buf) * n;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	ch_tp_pio = (struct ireg_buf *)temp_buff.data;
+
+	/* TP_PIO */
+	if (is_t5(padap->params.chip))
+		n = sizeof(t5_tp_pio_array) / (IREG_NUM_ELEM * sizeof(u32));
+	else if (is_t6(padap->params.chip))
+		n = sizeof(t6_tp_pio_array) / (IREG_NUM_ELEM * sizeof(u32));
+
+	for (i = 0; i < n; i++) {
+		struct ireg_field *tp_pio = &ch_tp_pio->tp_pio;
+		u32 *buff = ch_tp_pio->outbuf;
+
+		if (is_t5(padap->params.chip)) {
+			tp_pio->ireg_addr = t5_tp_pio_array[i][0];
+			tp_pio->ireg_data = t5_tp_pio_array[i][1];
+			tp_pio->ireg_local_offset = t5_tp_pio_array[i][2];
+			tp_pio->ireg_offset_range = t5_tp_pio_array[i][3];
+		} else if (is_t6(padap->params.chip)) {
+			tp_pio->ireg_addr = t6_tp_pio_array[i][0];
+			tp_pio->ireg_data = t6_tp_pio_array[i][1];
+			tp_pio->ireg_local_offset = t6_tp_pio_array[i][2];
+			tp_pio->ireg_offset_range = t6_tp_pio_array[i][3];
+		}
+		t4_tp_pio_read(padap, buff, tp_pio->ireg_offset_range,
+			       tp_pio->ireg_local_offset, true);
+		ch_tp_pio++;
+	}
+
+	/* TP_TM_PIO */
+	if (is_t5(padap->params.chip))
+		n = sizeof(t5_tp_tm_pio_array) / (IREG_NUM_ELEM * sizeof(u32));
+	else if (is_t6(padap->params.chip))
+		n = sizeof(t6_tp_tm_pio_array) / (IREG_NUM_ELEM * sizeof(u32));
+
+	for (i = 0; i < n; i++) {
+		struct ireg_field *tp_pio = &ch_tp_pio->tp_pio;
+		u32 *buff = ch_tp_pio->outbuf;
+
+		if (is_t5(padap->params.chip)) {
+			tp_pio->ireg_addr = t5_tp_tm_pio_array[i][0];
+			tp_pio->ireg_data = t5_tp_tm_pio_array[i][1];
+			tp_pio->ireg_local_offset = t5_tp_tm_pio_array[i][2];
+			tp_pio->ireg_offset_range = t5_tp_tm_pio_array[i][3];
+		} else if (is_t6(padap->params.chip)) {
+			tp_pio->ireg_addr = t6_tp_tm_pio_array[i][0];
+			tp_pio->ireg_data = t6_tp_tm_pio_array[i][1];
+			tp_pio->ireg_local_offset = t6_tp_tm_pio_array[i][2];
+			tp_pio->ireg_offset_range = t6_tp_tm_pio_array[i][3];
+		}
+		t4_tp_tm_pio_read(padap, buff, tp_pio->ireg_offset_range,
+				  tp_pio->ireg_local_offset, true);
+		ch_tp_pio++;
+	}
+
+	/* TP_MIB_INDEX */
+	if (is_t5(padap->params.chip))
+		n = sizeof(t5_tp_mib_index_array) /
+		    (IREG_NUM_ELEM * sizeof(u32));
+	else if (is_t6(padap->params.chip))
+		n = sizeof(t6_tp_mib_index_array) /
+		    (IREG_NUM_ELEM * sizeof(u32));
+
+	for (i = 0; i < n ; i++) {
+		struct ireg_field *tp_pio = &ch_tp_pio->tp_pio;
+		u32 *buff = ch_tp_pio->outbuf;
+
+		if (is_t5(padap->params.chip)) {
+			tp_pio->ireg_addr = t5_tp_mib_index_array[i][0];
+			tp_pio->ireg_data = t5_tp_mib_index_array[i][1];
+			tp_pio->ireg_local_offset =
+				t5_tp_mib_index_array[i][2];
+			tp_pio->ireg_offset_range =
+				t5_tp_mib_index_array[i][3];
+		} else if (is_t6(padap->params.chip)) {
+			tp_pio->ireg_addr = t6_tp_mib_index_array[i][0];
+			tp_pio->ireg_data = t6_tp_mib_index_array[i][1];
+			tp_pio->ireg_local_offset =
+				t6_tp_mib_index_array[i][2];
+			tp_pio->ireg_offset_range =
+				t6_tp_mib_index_array[i][3];
+		}
+		t4_tp_mib_read(padap, buff, tp_pio->ireg_offset_range,
+			       tp_pio->ireg_local_offset, true);
+		ch_tp_pio++;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+static void cudbg_read_sge_qbase_indirect_reg(struct adapter *padap,
+					      struct sge_qbase_reg_field *qbase,
+					      u32 func, bool is_pf)
+{
+	u32 *buff, i;
+
+	if (is_pf) {
+		buff = qbase->pf_data_value[func];
+	} else {
+		buff = qbase->vf_data_value[func];
+		/* In SGE_QBASE_INDEX,
+		 * Entries 0->7 are PF0->7, Entries 8->263 are VFID0->256.
+		 */
+		func += 8;
+	}
+
+	t4_write_reg(padap, qbase->reg_addr, func);
+	for (i = 0; i < SGE_QBASE_DATA_REG_NUM; i++, buff++)
+		*buff = t4_read_reg(padap, qbase->reg_data[i]);
+}
+
+int cudbg_collect_sge_indirect(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct sge_qbase_reg_field *sge_qbase;
+	struct ireg_buf *ch_sge_dbg;
+	u8 padap_running = 0;
+	int i, rc;
+	u32 size;
+
+	/* Accessing SGE_QBASE_MAP[0-3] and SGE_QBASE_INDEX regs can
+	 * lead to SGE missing doorbells under heavy traffic. So, only
+	 * collect them when adapter is idle.
+	 */
+	for_each_port(padap, i) {
+		padap_running = netif_running(padap->port[i]);
+		if (padap_running)
+			break;
+	}
+
+	size = sizeof(*ch_sge_dbg) * 2;
+	if (!padap_running)
+		size += sizeof(*sge_qbase);
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	ch_sge_dbg = (struct ireg_buf *)temp_buff.data;
+	for (i = 0; i < 2; i++) {
+		struct ireg_field *sge_pio = &ch_sge_dbg->tp_pio;
+		u32 *buff = ch_sge_dbg->outbuf;
+
+		sge_pio->ireg_addr = t5_sge_dbg_index_array[i][0];
+		sge_pio->ireg_data = t5_sge_dbg_index_array[i][1];
+		sge_pio->ireg_local_offset = t5_sge_dbg_index_array[i][2];
+		sge_pio->ireg_offset_range = t5_sge_dbg_index_array[i][3];
+		t4_read_indirect(padap,
+				 sge_pio->ireg_addr,
+				 sge_pio->ireg_data,
+				 buff,
+				 sge_pio->ireg_offset_range,
+				 sge_pio->ireg_local_offset);
+		ch_sge_dbg++;
+	}
+
+	if (CHELSIO_CHIP_VERSION(padap->params.chip) > CHELSIO_T5 &&
+	    !padap_running) {
+		sge_qbase = (struct sge_qbase_reg_field *)ch_sge_dbg;
+		/* 1 addr reg SGE_QBASE_INDEX and 4 data reg
+		 * SGE_QBASE_MAP[0-3]
+		 */
+		sge_qbase->reg_addr = t6_sge_qbase_index_array[0];
+		for (i = 0; i < SGE_QBASE_DATA_REG_NUM; i++)
+			sge_qbase->reg_data[i] =
+				t6_sge_qbase_index_array[i + 1];
+
+		for (i = 0; i <= PCIE_FW_MASTER_M; i++)
+			cudbg_read_sge_qbase_indirect_reg(padap, sge_qbase,
+							  i, true);
+
+		for (i = 0; i < padap->params.arch.vfcount; i++)
+			cudbg_read_sge_qbase_indirect_reg(padap, sge_qbase,
+							  i, false);
+
+		sge_qbase->vfcount = padap->params.arch.vfcount;
+	}
+
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_ulprx_la(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_ulprx_la *ulprx_la_buff;
+	int rc;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, sizeof(struct cudbg_ulprx_la),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	ulprx_la_buff = (struct cudbg_ulprx_la *)temp_buff.data;
+	t4_ulprx_read_la(padap, (u32 *)ulprx_la_buff->data);
+	ulprx_la_buff->size = ULPRX_LA_SIZE;
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_tp_la(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *dbg_buff,
+			struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_tp_la *tp_la_buff;
+	int size, rc;
+
+	size = sizeof(struct cudbg_tp_la) + TPLA_SIZE *  sizeof(u64);
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	tp_la_buff = (struct cudbg_tp_la *)temp_buff.data;
+	tp_la_buff->mode = DBGLAMODE_G(t4_read_reg(padap, TP_DBG_LA_CONFIG_A));
+	t4_tp_read_la(padap, (u64 *)tp_la_buff->data, NULL);
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_meminfo(struct cudbg_init *pdbg_init,
+			  struct cudbg_buffer *dbg_buff,
+			  struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_meminfo *meminfo_buff;
+	struct cudbg_ver_hdr *ver_hdr;
+	int rc;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff,
+			    sizeof(struct cudbg_ver_hdr) +
+			    sizeof(struct cudbg_meminfo),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	ver_hdr = (struct cudbg_ver_hdr *)temp_buff.data;
+	ver_hdr->signature = CUDBG_ENTITY_SIGNATURE;
+	ver_hdr->revision = CUDBG_MEMINFO_REV;
+	ver_hdr->size = sizeof(struct cudbg_meminfo);
+
+	meminfo_buff = (struct cudbg_meminfo *)(temp_buff.data +
+						sizeof(*ver_hdr));
+	rc = cudbg_fill_meminfo(padap, meminfo_buff);
+	if (rc) {
+		cudbg_err->sys_err = rc;
+		cudbg_put_buff(pdbg_init, &temp_buff);
+		return rc;
+	}
+
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_cim_pif_la(struct cudbg_init *pdbg_init,
+			     struct cudbg_buffer *dbg_buff,
+			     struct cudbg_error *cudbg_err)
+{
+	struct cudbg_cim_pif_la *cim_pif_la_buff;
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	int size, rc;
+
+	size = sizeof(struct cudbg_cim_pif_la) +
+	       2 * CIM_PIFLA_SIZE * 6 * sizeof(u32);
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	cim_pif_la_buff = (struct cudbg_cim_pif_la *)temp_buff.data;
+	cim_pif_la_buff->size = CIM_PIFLA_SIZE;
+	t4_cim_read_pif_la(padap, (u32 *)cim_pif_la_buff->data,
+			   (u32 *)cim_pif_la_buff->data + 6 * CIM_PIFLA_SIZE,
+			   NULL, NULL);
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_clk_info(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_clk_info *clk_info_buff;
+	u64 tp_tick_us;
+	int rc;
+
+	if (!padap->params.vpd.cclk)
+		return CUDBG_STATUS_CCLK_NOT_DEFINED;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, sizeof(struct cudbg_clk_info),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	clk_info_buff = (struct cudbg_clk_info *)temp_buff.data;
+	clk_info_buff->cclk_ps = 1000000000 / padap->params.vpd.cclk; /* psec */
+	clk_info_buff->res = t4_read_reg(padap, TP_TIMER_RESOLUTION_A);
+	clk_info_buff->tre = TIMERRESOLUTION_G(clk_info_buff->res);
+	clk_info_buff->dack_re = DELAYEDACKRESOLUTION_G(clk_info_buff->res);
+	tp_tick_us = (clk_info_buff->cclk_ps << clk_info_buff->tre) / 1000000;
+
+	clk_info_buff->dack_timer =
+		(clk_info_buff->cclk_ps << clk_info_buff->dack_re) / 1000000 *
+		t4_read_reg(padap, TP_DACK_TIMER_A);
+	clk_info_buff->retransmit_min =
+		tp_tick_us * t4_read_reg(padap, TP_RXT_MIN_A);
+	clk_info_buff->retransmit_max =
+		tp_tick_us * t4_read_reg(padap, TP_RXT_MAX_A);
+	clk_info_buff->persist_timer_min =
+		tp_tick_us * t4_read_reg(padap, TP_PERS_MIN_A);
+	clk_info_buff->persist_timer_max =
+		tp_tick_us * t4_read_reg(padap, TP_PERS_MAX_A);
+	clk_info_buff->keepalive_idle_timer =
+		tp_tick_us * t4_read_reg(padap, TP_KEEP_IDLE_A);
+	clk_info_buff->keepalive_interval =
+		tp_tick_us * t4_read_reg(padap, TP_KEEP_INTVL_A);
+	clk_info_buff->initial_srtt =
+		tp_tick_us * INITSRTT_G(t4_read_reg(padap, TP_INIT_SRTT_A));
+	clk_info_buff->finwait2_timer =
+		tp_tick_us * t4_read_reg(padap, TP_FINWAIT2_TIMER_A);
+
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_pcie_indirect(struct cudbg_init *pdbg_init,
+				struct cudbg_buffer *dbg_buff,
+				struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct ireg_buf *ch_pcie;
+	int i, rc, n;
+	u32 size;
+
+	n = sizeof(t5_pcie_pdbg_array) / (IREG_NUM_ELEM * sizeof(u32));
+	size = sizeof(struct ireg_buf) * n * 2;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	ch_pcie = (struct ireg_buf *)temp_buff.data;
+	/* PCIE_PDBG */
+	for (i = 0; i < n; i++) {
+		struct ireg_field *pcie_pio = &ch_pcie->tp_pio;
+		u32 *buff = ch_pcie->outbuf;
+
+		pcie_pio->ireg_addr = t5_pcie_pdbg_array[i][0];
+		pcie_pio->ireg_data = t5_pcie_pdbg_array[i][1];
+		pcie_pio->ireg_local_offset = t5_pcie_pdbg_array[i][2];
+		pcie_pio->ireg_offset_range = t5_pcie_pdbg_array[i][3];
+		t4_read_indirect(padap,
+				 pcie_pio->ireg_addr,
+				 pcie_pio->ireg_data,
+				 buff,
+				 pcie_pio->ireg_offset_range,
+				 pcie_pio->ireg_local_offset);
+		ch_pcie++;
+	}
+
+	/* PCIE_CDBG */
+	n = sizeof(t5_pcie_cdbg_array) / (IREG_NUM_ELEM * sizeof(u32));
+	for (i = 0; i < n; i++) {
+		struct ireg_field *pcie_pio = &ch_pcie->tp_pio;
+		u32 *buff = ch_pcie->outbuf;
+
+		pcie_pio->ireg_addr = t5_pcie_cdbg_array[i][0];
+		pcie_pio->ireg_data = t5_pcie_cdbg_array[i][1];
+		pcie_pio->ireg_local_offset = t5_pcie_cdbg_array[i][2];
+		pcie_pio->ireg_offset_range = t5_pcie_cdbg_array[i][3];
+		t4_read_indirect(padap,
+				 pcie_pio->ireg_addr,
+				 pcie_pio->ireg_data,
+				 buff,
+				 pcie_pio->ireg_offset_range,
+				 pcie_pio->ireg_local_offset);
+		ch_pcie++;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_pm_indirect(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct ireg_buf *ch_pm;
+	int i, rc, n;
+	u32 size;
+
+	n = sizeof(t5_pm_rx_array) / (IREG_NUM_ELEM * sizeof(u32));
+	size = sizeof(struct ireg_buf) * n * 2;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	ch_pm = (struct ireg_buf *)temp_buff.data;
+	/* PM_RX */
+	for (i = 0; i < n; i++) {
+		struct ireg_field *pm_pio = &ch_pm->tp_pio;
+		u32 *buff = ch_pm->outbuf;
+
+		pm_pio->ireg_addr = t5_pm_rx_array[i][0];
+		pm_pio->ireg_data = t5_pm_rx_array[i][1];
+		pm_pio->ireg_local_offset = t5_pm_rx_array[i][2];
+		pm_pio->ireg_offset_range = t5_pm_rx_array[i][3];
+		t4_read_indirect(padap,
+				 pm_pio->ireg_addr,
+				 pm_pio->ireg_data,
+				 buff,
+				 pm_pio->ireg_offset_range,
+				 pm_pio->ireg_local_offset);
+		ch_pm++;
+	}
+
+	/* PM_TX */
+	n = sizeof(t5_pm_tx_array) / (IREG_NUM_ELEM * sizeof(u32));
+	for (i = 0; i < n; i++) {
+		struct ireg_field *pm_pio = &ch_pm->tp_pio;
+		u32 *buff = ch_pm->outbuf;
+
+		pm_pio->ireg_addr = t5_pm_tx_array[i][0];
+		pm_pio->ireg_data = t5_pm_tx_array[i][1];
+		pm_pio->ireg_local_offset = t5_pm_tx_array[i][2];
+		pm_pio->ireg_offset_range = t5_pm_tx_array[i][3];
+		t4_read_indirect(padap,
+				 pm_pio->ireg_addr,
+				 pm_pio->ireg_data,
+				 buff,
+				 pm_pio->ireg_offset_range,
+				 pm_pio->ireg_local_offset);
+		ch_pm++;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_tid(struct cudbg_init *pdbg_init,
+		      struct cudbg_buffer *dbg_buff,
+		      struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_tid_info_region_rev1 *tid1;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_tid_info_region *tid;
+	u32 para[2], val[2];
+	int rc;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff,
+			    sizeof(struct cudbg_tid_info_region_rev1),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	tid1 = (struct cudbg_tid_info_region_rev1 *)temp_buff.data;
+	tid = &tid1->tid;
+	tid1->ver_hdr.signature = CUDBG_ENTITY_SIGNATURE;
+	tid1->ver_hdr.revision = CUDBG_TID_INFO_REV;
+	tid1->ver_hdr.size = sizeof(struct cudbg_tid_info_region_rev1) -
+			     sizeof(struct cudbg_ver_hdr);
+
+	/* If firmware is not attached/alive, use backdoor register
+	 * access to collect dump.
+	 */
+	if (!is_fw_attached(pdbg_init))
+		goto fill_tid;
+
+#define FW_PARAM_PFVF_A(param) \
+	(FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_PFVF) | \
+	 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_PFVF_##param) | \
+	 FW_PARAMS_PARAM_Y_V(0) | \
+	 FW_PARAMS_PARAM_Z_V(0))
+
+	para[0] = FW_PARAM_PFVF_A(ETHOFLD_START);
+	para[1] = FW_PARAM_PFVF_A(ETHOFLD_END);
+	rc = t4_query_params(padap, padap->mbox, padap->pf, 0, 2, para, val);
+	if (rc <  0) {
+		cudbg_err->sys_err = rc;
+		cudbg_put_buff(pdbg_init, &temp_buff);
+		return rc;
+	}
+	tid->uotid_base = val[0];
+	tid->nuotids = val[1] - val[0] + 1;
+
+	if (is_t5(padap->params.chip)) {
+		tid->sb = t4_read_reg(padap, LE_DB_SERVER_INDEX_A) / 4;
+	} else if (is_t6(padap->params.chip)) {
+		tid1->tid_start =
+			t4_read_reg(padap, LE_DB_ACTIVE_TABLE_START_INDEX_A);
+		tid->sb = t4_read_reg(padap, LE_DB_SRVR_START_INDEX_A);
+
+		para[0] = FW_PARAM_PFVF_A(HPFILTER_START);
+		para[1] = FW_PARAM_PFVF_A(HPFILTER_END);
+		rc = t4_query_params(padap, padap->mbox, padap->pf, 0, 2,
+				     para, val);
+		if (rc < 0) {
+			cudbg_err->sys_err = rc;
+			cudbg_put_buff(pdbg_init, &temp_buff);
+			return rc;
+		}
+		tid->hpftid_base = val[0];
+		tid->nhpftids = val[1] - val[0] + 1;
+	}
+
+#undef FW_PARAM_PFVF_A
+
+fill_tid:
+	tid->ntids = padap->tids.ntids;
+	tid->nstids = padap->tids.nstids;
+	tid->stid_base = padap->tids.stid_base;
+	tid->hash_base = padap->tids.hash_base;
+
+	tid->natids = padap->tids.natids;
+	tid->nftids = padap->tids.nftids;
+	tid->ftid_base = padap->tids.ftid_base;
+	tid->aftid_base = padap->tids.aftid_base;
+	tid->aftid_end = padap->tids.aftid_end;
+
+	tid->sftid_base = padap->tids.sftid_base;
+	tid->nsftids = padap->tids.nsftids;
+
+	tid->flags = padap->flags;
+	tid->le_db_conf = t4_read_reg(padap, LE_DB_CONFIG_A);
+	tid->ip_users = t4_read_reg(padap, LE_DB_ACT_CNT_IPV4_A);
+	tid->ipv6_users = t4_read_reg(padap, LE_DB_ACT_CNT_IPV6_A);
+
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_pcie_config(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	u32 size, *value, j;
+	int i, rc, n;
+
+	size = sizeof(u32) * CUDBG_NUM_PCIE_CONFIG_REGS;
+	n = sizeof(t5_pcie_config_array) / (2 * sizeof(u32));
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	value = (u32 *)temp_buff.data;
+	for (i = 0; i < n; i++) {
+		for (j = t5_pcie_config_array[i][0];
+		     j <= t5_pcie_config_array[i][1]; j += 4) {
+			t4_hw_pci_read_cfg4(padap, j, value);
+			value++;
+		}
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+static int cudbg_sge_ctxt_check_valid(u32 *buf, int type)
+{
+	int index, bit, bit_pos = 0;
+
+	switch (type) {
+	case CTXT_EGRESS:
+		bit_pos = 176;
+		break;
+	case CTXT_INGRESS:
+		bit_pos = 141;
+		break;
+	case CTXT_FLM:
+		bit_pos = 89;
+		break;
+	}
+	index = bit_pos / 32;
+	bit =  bit_pos % 32;
+	return buf[index] & (1U << bit);
+}
+
+static int cudbg_get_ctxt_region_info(struct adapter *padap,
+				      struct cudbg_region_info *ctx_info,
+				      u8 *mem_type)
+{
+	struct cudbg_mem_desc mem_desc;
+	struct cudbg_meminfo meminfo;
+	u32 i, j, value, found;
+	u8 flq;
+	int rc;
+
+	rc = cudbg_fill_meminfo(padap, &meminfo);
+	if (rc)
+		return rc;
+
+	/* Get EGRESS and INGRESS context region size */
+	for (i = CTXT_EGRESS; i <= CTXT_INGRESS; i++) {
+		found = 0;
+		memset(&mem_desc, 0, sizeof(struct cudbg_mem_desc));
+		for (j = 0; j < ARRAY_SIZE(meminfo.avail); j++) {
+			rc = cudbg_get_mem_region(padap, &meminfo, j,
+						  cudbg_region[i],
+						  &mem_desc);
+			if (!rc) {
+				found = 1;
+				rc = cudbg_get_mem_relative(padap, &meminfo, j,
+							    &mem_desc.base,
+							    &mem_desc.limit);
+				if (rc) {
+					ctx_info[i].exist = false;
+					break;
+				}
+				ctx_info[i].exist = true;
+				ctx_info[i].start = mem_desc.base;
+				ctx_info[i].end = mem_desc.limit;
+				mem_type[i] = j;
+				break;
+			}
+		}
+		if (!found)
+			ctx_info[i].exist = false;
+	}
+
+	/* Get FLM and CNM max qid. */
+	value = t4_read_reg(padap, SGE_FLM_CFG_A);
+
+	/* Get number of data freelist queues */
+	flq = HDRSTARTFLQ_G(value);
+	ctx_info[CTXT_FLM].exist = true;
+	ctx_info[CTXT_FLM].end = (CUDBG_MAX_FL_QIDS >> flq) * SGE_CTXT_SIZE;
+
+	/* The number of CONM contexts are same as number of freelist
+	 * queues.
+	 */
+	ctx_info[CTXT_CNM].exist = true;
+	ctx_info[CTXT_CNM].end = ctx_info[CTXT_FLM].end;
+
+	return 0;
+}
+
+int cudbg_dump_context_size(struct adapter *padap)
+{
+	struct cudbg_region_info region_info[CTXT_CNM + 1] = { {0} };
+	u8 mem_type[CTXT_INGRESS + 1] = { 0 };
+	u32 i, size = 0;
+	int rc;
+
+	/* Get max valid qid for each type of queue */
+	rc = cudbg_get_ctxt_region_info(padap, region_info, mem_type);
+	if (rc)
+		return rc;
+
+	for (i = 0; i < CTXT_CNM; i++) {
+		if (!region_info[i].exist) {
+			if (i == CTXT_EGRESS || i == CTXT_INGRESS)
+				size += CUDBG_LOWMEM_MAX_CTXT_QIDS *
+					SGE_CTXT_SIZE;
+			continue;
+		}
+
+		size += (region_info[i].end - region_info[i].start + 1) /
+			SGE_CTXT_SIZE;
+	}
+	return size * sizeof(struct cudbg_ch_cntxt);
+}
+
+static void cudbg_read_sge_ctxt(struct cudbg_init *pdbg_init, u32 cid,
+				enum ctxt_type ctype, u32 *data)
+{
+	struct adapter *padap = pdbg_init->adap;
+	int rc = -1;
+
+	/* Under heavy traffic, the SGE Queue contexts registers will be
+	 * frequently accessed by firmware.
+	 *
+	 * To avoid conflicts with firmware, always ask firmware to fetch
+	 * the SGE Queue contexts via mailbox. On failure, fallback to
+	 * accessing hardware registers directly.
+	 */
+	if (is_fw_attached(pdbg_init))
+		rc = t4_sge_ctxt_rd(padap, padap->mbox, cid, ctype, data);
+	if (rc)
+		t4_sge_ctxt_rd_bd(padap, cid, ctype, data);
+}
+
+static void cudbg_get_sge_ctxt_fw(struct cudbg_init *pdbg_init, u32 max_qid,
+				  u8 ctxt_type,
+				  struct cudbg_ch_cntxt **out_buff)
+{
+	struct cudbg_ch_cntxt *buff = *out_buff;
+	int rc;
+	u32 j;
+
+	for (j = 0; j < max_qid; j++) {
+		cudbg_read_sge_ctxt(pdbg_init, j, ctxt_type, buff->data);
+		rc = cudbg_sge_ctxt_check_valid(buff->data, ctxt_type);
+		if (!rc)
+			continue;
+
+		buff->cntxt_type = ctxt_type;
+		buff->cntxt_id = j;
+		buff++;
+		if (ctxt_type == CTXT_FLM) {
+			cudbg_read_sge_ctxt(pdbg_init, j, CTXT_CNM, buff->data);
+			buff->cntxt_type = CTXT_CNM;
+			buff->cntxt_id = j;
+			buff++;
+		}
+	}
+
+	*out_buff = buff;
+}
+
+int cudbg_collect_dump_context(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	struct cudbg_region_info region_info[CTXT_CNM + 1] = { {0} };
+	struct adapter *padap = pdbg_init->adap;
+	u32 j, size, max_ctx_size, max_ctx_qid;
+	u8 mem_type[CTXT_INGRESS + 1] = { 0 };
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_ch_cntxt *buff;
+	u8 *ctx_buf;
+	u8 i, k;
+	int rc;
+
+	/* Get max valid qid for each type of queue */
+	rc = cudbg_get_ctxt_region_info(padap, region_info, mem_type);
+	if (rc)
+		return rc;
+
+	rc = cudbg_dump_context_size(padap);
+	if (rc <= 0)
+		return CUDBG_STATUS_ENTITY_NOT_FOUND;
+
+	size = rc;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	/* Get buffer with enough space to read the biggest context
+	 * region in memory.
+	 */
+	max_ctx_size = max(region_info[CTXT_EGRESS].end -
+			   region_info[CTXT_EGRESS].start + 1,
+			   region_info[CTXT_INGRESS].end -
+			   region_info[CTXT_INGRESS].start + 1);
+
+	ctx_buf = kvzalloc(max_ctx_size, GFP_KERNEL);
+	if (!ctx_buf) {
+		cudbg_put_buff(pdbg_init, &temp_buff);
+		return -ENOMEM;
+	}
+
+	buff = (struct cudbg_ch_cntxt *)temp_buff.data;
+
+	/* Collect EGRESS and INGRESS context data.
+	 * In case of failures, fallback to collecting via FW or
+	 * backdoor access.
+	 */
+	for (i = CTXT_EGRESS; i <= CTXT_INGRESS; i++) {
+		if (!region_info[i].exist) {
+			max_ctx_qid = CUDBG_LOWMEM_MAX_CTXT_QIDS;
+			cudbg_get_sge_ctxt_fw(pdbg_init, max_ctx_qid, i,
+					      &buff);
+			continue;
+		}
+
+		max_ctx_size = region_info[i].end - region_info[i].start + 1;
+		max_ctx_qid = max_ctx_size / SGE_CTXT_SIZE;
+
+		/* If firmware is not attached/alive, use backdoor register
+		 * access to collect dump.
+		 */
+		if (is_fw_attached(pdbg_init)) {
+			t4_sge_ctxt_flush(padap, padap->mbox, i);
+
+			rc = t4_memory_rw(padap, MEMWIN_NIC, mem_type[i],
+					  region_info[i].start, max_ctx_size,
+					  (__be32 *)ctx_buf, 1);
+		}
+
+		if (rc || !is_fw_attached(pdbg_init)) {
+			max_ctx_qid = CUDBG_LOWMEM_MAX_CTXT_QIDS;
+			cudbg_get_sge_ctxt_fw(pdbg_init, max_ctx_qid, i,
+					      &buff);
+			continue;
+		}
+
+		for (j = 0; j < max_ctx_qid; j++) {
+			__be64 *dst_off;
+			u64 *src_off;
+
+			src_off = (u64 *)(ctx_buf + j * SGE_CTXT_SIZE);
+			dst_off = (__be64 *)buff->data;
+
+			/* The data is stored in 64-bit cpu order.  Convert it
+			 * to big endian before parsing.
+			 */
+			for (k = 0; k < SGE_CTXT_SIZE / sizeof(u64); k++)
+				dst_off[k] = cpu_to_be64(src_off[k]);
+
+			rc = cudbg_sge_ctxt_check_valid(buff->data, i);
+			if (!rc)
+				continue;
+
+			buff->cntxt_type = i;
+			buff->cntxt_id = j;
+			buff++;
+		}
+	}
+
+	kvfree(ctx_buf);
+
+	/* Collect FREELIST and CONGESTION MANAGER contexts */
+	max_ctx_size = region_info[CTXT_FLM].end -
+		       region_info[CTXT_FLM].start + 1;
+	max_ctx_qid = max_ctx_size / SGE_CTXT_SIZE;
+	/* Since FLM and CONM are 1-to-1 mapped, the below function
+	 * will fetch both FLM and CONM contexts.
+	 */
+	cudbg_get_sge_ctxt_fw(pdbg_init, max_ctx_qid, CTXT_FLM, &buff);
+
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+static inline void cudbg_tcamxy2valmask(u64 x, u64 y, u8 *addr, u64 *mask)
+{
+	*mask = x | y;
+	y = (__force u64)cpu_to_be64(y);
+	memcpy(addr, (char *)&y + 2, ETH_ALEN);
+}
+
+static void cudbg_mps_rpl_backdoor(struct adapter *padap,
+				   struct fw_ldst_mps_rplc *mps_rplc)
+{
+	if (is_t5(padap->params.chip)) {
+		mps_rplc->rplc255_224 = htonl(t4_read_reg(padap,
+							  MPS_VF_RPLCT_MAP3_A));
+		mps_rplc->rplc223_192 = htonl(t4_read_reg(padap,
+							  MPS_VF_RPLCT_MAP2_A));
+		mps_rplc->rplc191_160 = htonl(t4_read_reg(padap,
+							  MPS_VF_RPLCT_MAP1_A));
+		mps_rplc->rplc159_128 = htonl(t4_read_reg(padap,
+							  MPS_VF_RPLCT_MAP0_A));
+	} else {
+		mps_rplc->rplc255_224 = htonl(t4_read_reg(padap,
+							  MPS_VF_RPLCT_MAP7_A));
+		mps_rplc->rplc223_192 = htonl(t4_read_reg(padap,
+							  MPS_VF_RPLCT_MAP6_A));
+		mps_rplc->rplc191_160 = htonl(t4_read_reg(padap,
+							  MPS_VF_RPLCT_MAP5_A));
+		mps_rplc->rplc159_128 = htonl(t4_read_reg(padap,
+							  MPS_VF_RPLCT_MAP4_A));
+	}
+	mps_rplc->rplc127_96 = htonl(t4_read_reg(padap, MPS_VF_RPLCT_MAP3_A));
+	mps_rplc->rplc95_64 = htonl(t4_read_reg(padap, MPS_VF_RPLCT_MAP2_A));
+	mps_rplc->rplc63_32 = htonl(t4_read_reg(padap, MPS_VF_RPLCT_MAP1_A));
+	mps_rplc->rplc31_0 = htonl(t4_read_reg(padap, MPS_VF_RPLCT_MAP0_A));
+}
+
+static int cudbg_collect_tcam_index(struct cudbg_init *pdbg_init,
+				    struct cudbg_mps_tcam *tcam, u32 idx)
+{
+	struct adapter *padap = pdbg_init->adap;
+	u64 tcamy, tcamx, val;
+	u32 ctl, data2;
+	int rc = 0;
+
+	if (CHELSIO_CHIP_VERSION(padap->params.chip) >= CHELSIO_T6) {
+		/* CtlReqID   - 1: use Host Driver Requester ID
+		 * CtlCmdType - 0: Read, 1: Write
+		 * CtlTcamSel - 0: TCAM0, 1: TCAM1
+		 * CtlXYBitSel- 0: Y bit, 1: X bit
+		 */
+
+		/* Read tcamy */
+		ctl = CTLREQID_V(1) | CTLCMDTYPE_V(0) | CTLXYBITSEL_V(0);
+		if (idx < 256)
+			ctl |= CTLTCAMINDEX_V(idx) | CTLTCAMSEL_V(0);
+		else
+			ctl |= CTLTCAMINDEX_V(idx - 256) | CTLTCAMSEL_V(1);
+
+		t4_write_reg(padap, MPS_CLS_TCAM_DATA2_CTL_A, ctl);
+		val = t4_read_reg(padap, MPS_CLS_TCAM_RDATA1_REQ_ID1_A);
+		tcamy = DMACH_G(val) << 32;
+		tcamy |= t4_read_reg(padap, MPS_CLS_TCAM_RDATA0_REQ_ID1_A);
+		data2 = t4_read_reg(padap, MPS_CLS_TCAM_RDATA2_REQ_ID1_A);
+		tcam->lookup_type = DATALKPTYPE_G(data2);
+
+		/* 0 - Outer header, 1 - Inner header
+		 * [71:48] bit locations are overloaded for
+		 * outer vs. inner lookup types.
+		 */
+		if (tcam->lookup_type && tcam->lookup_type != DATALKPTYPE_M) {
+			/* Inner header VNI */
+			tcam->vniy = (data2 & DATAVIDH2_F) | DATAVIDH1_G(data2);
+			tcam->vniy = (tcam->vniy << 16) | VIDL_G(val);
+			tcam->dip_hit = data2 & DATADIPHIT_F;
+		} else {
+			tcam->vlan_vld = data2 & DATAVIDH2_F;
+			tcam->ivlan = VIDL_G(val);
+		}
+
+		tcam->port_num = DATAPORTNUM_G(data2);
+
+		/* Read tcamx. Change the control param */
+		ctl |= CTLXYBITSEL_V(1);
+		t4_write_reg(padap, MPS_CLS_TCAM_DATA2_CTL_A, ctl);
+		val = t4_read_reg(padap, MPS_CLS_TCAM_RDATA1_REQ_ID1_A);
+		tcamx = DMACH_G(val) << 32;
+		tcamx |= t4_read_reg(padap, MPS_CLS_TCAM_RDATA0_REQ_ID1_A);
+		data2 = t4_read_reg(padap, MPS_CLS_TCAM_RDATA2_REQ_ID1_A);
+		if (tcam->lookup_type && tcam->lookup_type != DATALKPTYPE_M) {
+			/* Inner header VNI mask */
+			tcam->vnix = (data2 & DATAVIDH2_F) | DATAVIDH1_G(data2);
+			tcam->vnix = (tcam->vnix << 16) | VIDL_G(val);
+		}
+	} else {
+		tcamy = t4_read_reg64(padap, MPS_CLS_TCAM_Y_L(idx));
+		tcamx = t4_read_reg64(padap, MPS_CLS_TCAM_X_L(idx));
+	}
+
+	/* If no entry, return */
+	if (tcamx & tcamy)
+		return rc;
+
+	tcam->cls_lo = t4_read_reg(padap, MPS_CLS_SRAM_L(idx));
+	tcam->cls_hi = t4_read_reg(padap, MPS_CLS_SRAM_H(idx));
+
+	if (is_t5(padap->params.chip))
+		tcam->repli = (tcam->cls_lo & REPLICATE_F);
+	else if (is_t6(padap->params.chip))
+		tcam->repli = (tcam->cls_lo & T6_REPLICATE_F);
+
+	if (tcam->repli) {
+		struct fw_ldst_cmd ldst_cmd;
+		struct fw_ldst_mps_rplc mps_rplc;
+
+		memset(&ldst_cmd, 0, sizeof(ldst_cmd));
+		ldst_cmd.op_to_addrspace =
+			htonl(FW_CMD_OP_V(FW_LDST_CMD) |
+			      FW_CMD_REQUEST_F | FW_CMD_READ_F |
+			      FW_LDST_CMD_ADDRSPACE_V(FW_LDST_ADDRSPC_MPS));
+		ldst_cmd.cycles_to_len16 = htonl(FW_LEN16(ldst_cmd));
+		ldst_cmd.u.mps.rplc.fid_idx =
+			htons(FW_LDST_CMD_FID_V(FW_LDST_MPS_RPLC) |
+			      FW_LDST_CMD_IDX_V(idx));
+
+		/* If firmware is not attached/alive, use backdoor register
+		 * access to collect dump.
+		 */
+		if (is_fw_attached(pdbg_init))
+			rc = t4_wr_mbox(padap, padap->mbox, &ldst_cmd,
+					sizeof(ldst_cmd), &ldst_cmd);
+
+		if (rc || !is_fw_attached(pdbg_init)) {
+			cudbg_mps_rpl_backdoor(padap, &mps_rplc);
+			/* Ignore error since we collected directly from
+			 * reading registers.
+			 */
+			rc = 0;
+		} else {
+			mps_rplc = ldst_cmd.u.mps.rplc;
+		}
+
+		tcam->rplc[0] = ntohl(mps_rplc.rplc31_0);
+		tcam->rplc[1] = ntohl(mps_rplc.rplc63_32);
+		tcam->rplc[2] = ntohl(mps_rplc.rplc95_64);
+		tcam->rplc[3] = ntohl(mps_rplc.rplc127_96);
+		if (padap->params.arch.mps_rplc_size > CUDBG_MAX_RPLC_SIZE) {
+			tcam->rplc[4] = ntohl(mps_rplc.rplc159_128);
+			tcam->rplc[5] = ntohl(mps_rplc.rplc191_160);
+			tcam->rplc[6] = ntohl(mps_rplc.rplc223_192);
+			tcam->rplc[7] = ntohl(mps_rplc.rplc255_224);
+		}
+	}
+	cudbg_tcamxy2valmask(tcamx, tcamy, tcam->addr, &tcam->mask);
+	tcam->idx = idx;
+	tcam->rplc_size = padap->params.arch.mps_rplc_size;
+	return rc;
+}
+
+int cudbg_collect_mps_tcam(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	u32 size = 0, i, n, total_size = 0;
+	struct cudbg_mps_tcam *tcam;
+	int rc;
+
+	n = padap->params.arch.mps_tcam_size;
+	size = sizeof(struct cudbg_mps_tcam) * n;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	tcam = (struct cudbg_mps_tcam *)temp_buff.data;
+	for (i = 0; i < n; i++) {
+		rc = cudbg_collect_tcam_index(pdbg_init, tcam, i);
+		if (rc) {
+			cudbg_err->sys_err = rc;
+			cudbg_put_buff(pdbg_init, &temp_buff);
+			return rc;
+		}
+		total_size += sizeof(struct cudbg_mps_tcam);
+		tcam++;
+	}
+
+	if (!total_size) {
+		rc = CUDBG_SYSTEM_ERROR;
+		cudbg_err->sys_err = rc;
+		cudbg_put_buff(pdbg_init, &temp_buff);
+		return rc;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_vpd_data(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	char vpd_str[CUDBG_VPD_VER_LEN + 1];
+	struct cudbg_vpd_data *vpd_data;
+	struct vpd_params vpd = { 0 };
+	u32 vpd_vers, fw_vers;
+	int rc;
+
+	rc = t4_get_raw_vpd_params(padap, &vpd);
+	if (rc)
+		return rc;
+
+	rc = t4_get_fw_version(padap, &fw_vers);
+	if (rc)
+		return rc;
+
+	rc = cudbg_read_vpd_reg(padap, CUDBG_VPD_VER_ADDR, CUDBG_VPD_VER_LEN,
+				vpd_str);
+	if (rc)
+		return rc;
+
+	vpd_str[CUDBG_VPD_VER_LEN] = '\0';
+	rc = kstrtouint(vpd_str, 0, &vpd_vers);
+	if (rc)
+		return rc;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, sizeof(struct cudbg_vpd_data),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	vpd_data = (struct cudbg_vpd_data *)temp_buff.data;
+	memcpy(vpd_data->sn, vpd.sn, SERNUM_LEN + 1);
+	memcpy(vpd_data->bn, vpd.pn, PN_LEN + 1);
+	memcpy(vpd_data->na, vpd.na, MACADDR_LEN + 1);
+	memcpy(vpd_data->mn, vpd.id, ID_LEN + 1);
+	vpd_data->scfg_vers = t4_read_reg(padap, PCIE_STATIC_SPARE2_A);
+	vpd_data->vpd_vers = vpd_vers;
+	vpd_data->fw_major = FW_HDR_FW_VER_MAJOR_G(fw_vers);
+	vpd_data->fw_minor = FW_HDR_FW_VER_MINOR_G(fw_vers);
+	vpd_data->fw_micro = FW_HDR_FW_VER_MICRO_G(fw_vers);
+	vpd_data->fw_build = FW_HDR_FW_VER_BUILD_G(fw_vers);
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+static int cudbg_read_tid(struct cudbg_init *pdbg_init, u32 tid,
+			  struct cudbg_tid_data *tid_data)
+{
+	struct adapter *padap = pdbg_init->adap;
+	int i, cmd_retry = 8;
+	u32 val;
+
+	/* Fill REQ_DATA regs with 0's */
+	for (i = 0; i < NUM_LE_DB_DBGI_REQ_DATA_INSTANCES; i++)
+		t4_write_reg(padap, LE_DB_DBGI_REQ_DATA_A + (i << 2), 0);
+
+	/* Write DBIG command */
+	val = DBGICMD_V(4) | DBGITID_V(tid);
+	t4_write_reg(padap, LE_DB_DBGI_REQ_TCAM_CMD_A, val);
+	tid_data->dbig_cmd = val;
+
+	val = DBGICMDSTRT_F | DBGICMDMODE_V(1); /* LE mode */
+	t4_write_reg(padap, LE_DB_DBGI_CONFIG_A, val);
+	tid_data->dbig_conf = val;
+
+	/* Poll the DBGICMDBUSY bit */
+	val = 1;
+	while (val) {
+		val = t4_read_reg(padap, LE_DB_DBGI_CONFIG_A);
+		val = val & DBGICMDBUSY_F;
+		cmd_retry--;
+		if (!cmd_retry)
+			return CUDBG_SYSTEM_ERROR;
+	}
+
+	/* Check RESP status */
+	val = t4_read_reg(padap, LE_DB_DBGI_RSP_STATUS_A);
+	tid_data->dbig_rsp_stat = val;
+	if (!(val & 1))
+		return CUDBG_SYSTEM_ERROR;
+
+	/* Read RESP data */
+	for (i = 0; i < NUM_LE_DB_DBGI_RSP_DATA_INSTANCES; i++)
+		tid_data->data[i] = t4_read_reg(padap,
+						LE_DB_DBGI_RSP_DATA_A +
+						(i << 2));
+	tid_data->tid = tid;
+	return 0;
+}
+
+static int cudbg_get_le_type(u32 tid, struct cudbg_tcam tcam_region)
+{
+	int type = LE_ET_UNKNOWN;
+
+	if (tid < tcam_region.server_start)
+		type = LE_ET_TCAM_CON;
+	else if (tid < tcam_region.filter_start)
+		type = LE_ET_TCAM_SERVER;
+	else if (tid < tcam_region.clip_start)
+		type = LE_ET_TCAM_FILTER;
+	else if (tid < tcam_region.routing_start)
+		type = LE_ET_TCAM_CLIP;
+	else if (tid < tcam_region.tid_hash_base)
+		type = LE_ET_TCAM_ROUTING;
+	else if (tid < tcam_region.max_tid)
+		type = LE_ET_HASH_CON;
+	else
+		type = LE_ET_INVALID_TID;
+
+	return type;
+}
+
+static int cudbg_is_ipv6_entry(struct cudbg_tid_data *tid_data,
+			       struct cudbg_tcam tcam_region)
+{
+	int ipv6 = 0;
+	int le_type;
+
+	le_type = cudbg_get_le_type(tid_data->tid, tcam_region);
+	if (tid_data->tid & 1)
+		return 0;
+
+	if (le_type == LE_ET_HASH_CON) {
+		ipv6 = tid_data->data[16] & 0x8000;
+	} else if (le_type == LE_ET_TCAM_CON) {
+		ipv6 = tid_data->data[16] & 0x8000;
+		if (ipv6)
+			ipv6 = tid_data->data[9] == 0x00C00000;
+	} else {
+		ipv6 = 0;
+	}
+	return ipv6;
+}
+
+void cudbg_fill_le_tcam_info(struct adapter *padap,
+			     struct cudbg_tcam *tcam_region)
+{
+	u32 value;
+
+	/* Get the LE regions */
+	value = t4_read_reg(padap, LE_DB_TID_HASHBASE_A); /* hash base index */
+	tcam_region->tid_hash_base = value;
+
+	/* Get routing table index */
+	value = t4_read_reg(padap, LE_DB_ROUTING_TABLE_INDEX_A);
+	tcam_region->routing_start = value;
+
+	/* Get clip table index. For T6 there is separate CLIP TCAM */
+	if (is_t6(padap->params.chip))
+		value = t4_read_reg(padap, LE_DB_CLCAM_TID_BASE_A);
+	else
+		value = t4_read_reg(padap, LE_DB_CLIP_TABLE_INDEX_A);
+	tcam_region->clip_start = value;
+
+	/* Get filter table index */
+	value = t4_read_reg(padap, LE_DB_FILTER_TABLE_INDEX_A);
+	tcam_region->filter_start = value;
+
+	/* Get server table index */
+	value = t4_read_reg(padap, LE_DB_SERVER_INDEX_A);
+	tcam_region->server_start = value;
+
+	/* Check whether hash is enabled and calculate the max tids */
+	value = t4_read_reg(padap, LE_DB_CONFIG_A);
+	if ((value >> HASHEN_S) & 1) {
+		value = t4_read_reg(padap, LE_DB_HASH_CONFIG_A);
+		if (CHELSIO_CHIP_VERSION(padap->params.chip) > CHELSIO_T5) {
+			tcam_region->max_tid = (value & 0xFFFFF) +
+					       tcam_region->tid_hash_base;
+		} else {
+			value = HASHTIDSIZE_G(value);
+			value = 1 << value;
+			tcam_region->max_tid = value +
+					       tcam_region->tid_hash_base;
+		}
+	} else { /* hash not enabled */
+		if (is_t6(padap->params.chip))
+			tcam_region->max_tid = (value & ASLIPCOMPEN_F) ?
+					       CUDBG_MAX_TID_COMP_EN :
+					       CUDBG_MAX_TID_COMP_DIS;
+		else
+			tcam_region->max_tid = CUDBG_MAX_TCAM_TID;
+	}
+
+	if (is_t6(padap->params.chip))
+		tcam_region->max_tid += CUDBG_T6_CLIP;
+}
+
+int cudbg_collect_le_tcam(struct cudbg_init *pdbg_init,
+			  struct cudbg_buffer *dbg_buff,
+			  struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_tcam tcam_region = { 0 };
+	struct cudbg_tid_data *tid_data;
+	u32 bytes = 0;
+	int rc, size;
+	u32 i;
+
+	cudbg_fill_le_tcam_info(padap, &tcam_region);
+
+	size = sizeof(struct cudbg_tid_data) * tcam_region.max_tid;
+	size += sizeof(struct cudbg_tcam);
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	memcpy(temp_buff.data, &tcam_region, sizeof(struct cudbg_tcam));
+	bytes = sizeof(struct cudbg_tcam);
+	tid_data = (struct cudbg_tid_data *)(temp_buff.data + bytes);
+	/* read all tid */
+	for (i = 0; i < tcam_region.max_tid; ) {
+		rc = cudbg_read_tid(pdbg_init, i, tid_data);
+		if (rc) {
+			cudbg_err->sys_warn = CUDBG_STATUS_PARTIAL_DATA;
+			/* Update tcam header and exit */
+			tcam_region.max_tid = i;
+			memcpy(temp_buff.data, &tcam_region,
+			       sizeof(struct cudbg_tcam));
+			goto out;
+		}
+
+		if (cudbg_is_ipv6_entry(tid_data, tcam_region)) {
+			/* T6 CLIP TCAM: ipv6 takes 4 entries */
+			if (is_t6(padap->params.chip) &&
+			    i >= tcam_region.clip_start &&
+			    i < tcam_region.clip_start + CUDBG_T6_CLIP)
+				i += 4;
+			else /* Main TCAM: ipv6 takes two tids */
+				i += 2;
+		} else {
+			i++;
+		}
+
+		tid_data++;
+		bytes += sizeof(struct cudbg_tid_data);
+	}
+
+out:
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_cctrl(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *dbg_buff,
+			struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	u32 size;
+	int rc;
+
+	size = sizeof(u16) * NMTUS * NCCTRL_WIN;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	t4_read_cong_tbl(padap, (void *)temp_buff.data);
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_ma_indirect(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct ireg_buf *ma_indr;
+	int i, rc, n;
+	u32 size, j;
+
+	if (CHELSIO_CHIP_VERSION(padap->params.chip) < CHELSIO_T6)
+		return CUDBG_STATUS_ENTITY_NOT_FOUND;
+
+	n = sizeof(t6_ma_ireg_array) / (IREG_NUM_ELEM * sizeof(u32));
+	size = sizeof(struct ireg_buf) * n * 2;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	ma_indr = (struct ireg_buf *)temp_buff.data;
+	for (i = 0; i < n; i++) {
+		struct ireg_field *ma_fli = &ma_indr->tp_pio;
+		u32 *buff = ma_indr->outbuf;
+
+		ma_fli->ireg_addr = t6_ma_ireg_array[i][0];
+		ma_fli->ireg_data = t6_ma_ireg_array[i][1];
+		ma_fli->ireg_local_offset = t6_ma_ireg_array[i][2];
+		ma_fli->ireg_offset_range = t6_ma_ireg_array[i][3];
+		t4_read_indirect(padap, ma_fli->ireg_addr, ma_fli->ireg_data,
+				 buff, ma_fli->ireg_offset_range,
+				 ma_fli->ireg_local_offset);
+		ma_indr++;
+	}
+
+	n = sizeof(t6_ma_ireg_array2) / (IREG_NUM_ELEM * sizeof(u32));
+	for (i = 0; i < n; i++) {
+		struct ireg_field *ma_fli = &ma_indr->tp_pio;
+		u32 *buff = ma_indr->outbuf;
+
+		ma_fli->ireg_addr = t6_ma_ireg_array2[i][0];
+		ma_fli->ireg_data = t6_ma_ireg_array2[i][1];
+		ma_fli->ireg_local_offset = t6_ma_ireg_array2[i][2];
+		for (j = 0; j < t6_ma_ireg_array2[i][3]; j++) {
+			t4_read_indirect(padap, ma_fli->ireg_addr,
+					 ma_fli->ireg_data, buff, 1,
+					 ma_fli->ireg_local_offset);
+			buff++;
+			ma_fli->ireg_local_offset += 0x20;
+		}
+		ma_indr++;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_ulptx_la(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_ulptx_la *ulptx_la_buff;
+	struct cudbg_ver_hdr *ver_hdr;
+	u32 i, j;
+	int rc;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff,
+			    sizeof(struct cudbg_ver_hdr) +
+			    sizeof(struct cudbg_ulptx_la),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	ver_hdr = (struct cudbg_ver_hdr *)temp_buff.data;
+	ver_hdr->signature = CUDBG_ENTITY_SIGNATURE;
+	ver_hdr->revision = CUDBG_ULPTX_LA_REV;
+	ver_hdr->size = sizeof(struct cudbg_ulptx_la);
+
+	ulptx_la_buff = (struct cudbg_ulptx_la *)(temp_buff.data +
+						  sizeof(*ver_hdr));
+	for (i = 0; i < CUDBG_NUM_ULPTX; i++) {
+		ulptx_la_buff->rdptr[i] = t4_read_reg(padap,
+						      ULP_TX_LA_RDPTR_0_A +
+						      0x10 * i);
+		ulptx_la_buff->wrptr[i] = t4_read_reg(padap,
+						      ULP_TX_LA_WRPTR_0_A +
+						      0x10 * i);
+		ulptx_la_buff->rddata[i] = t4_read_reg(padap,
+						       ULP_TX_LA_RDDATA_0_A +
+						       0x10 * i);
+		for (j = 0; j < CUDBG_NUM_ULPTX_READ; j++)
+			ulptx_la_buff->rd_data[i][j] =
+				t4_read_reg(padap,
+					    ULP_TX_LA_RDDATA_0_A + 0x10 * i);
+	}
+
+	for (i = 0; i < CUDBG_NUM_ULPTX_ASIC_READ; i++) {
+		t4_write_reg(padap, ULP_TX_ASIC_DEBUG_CTRL_A, 0x1);
+		ulptx_la_buff->rdptr_asic[i] =
+				t4_read_reg(padap, ULP_TX_ASIC_DEBUG_CTRL_A);
+		ulptx_la_buff->rddata_asic[i][0] =
+				t4_read_reg(padap, ULP_TX_ASIC_DEBUG_0_A);
+		ulptx_la_buff->rddata_asic[i][1] =
+				t4_read_reg(padap, ULP_TX_ASIC_DEBUG_1_A);
+		ulptx_la_buff->rddata_asic[i][2] =
+				t4_read_reg(padap, ULP_TX_ASIC_DEBUG_2_A);
+		ulptx_la_buff->rddata_asic[i][3] =
+				t4_read_reg(padap, ULP_TX_ASIC_DEBUG_3_A);
+		ulptx_la_buff->rddata_asic[i][4] =
+				t4_read_reg(padap, ULP_TX_ASIC_DEBUG_4_A);
+		ulptx_la_buff->rddata_asic[i][5] =
+				t4_read_reg(padap, PM_RX_BASE_ADDR);
+	}
+
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_up_cim_indirect(struct cudbg_init *pdbg_init,
+				  struct cudbg_buffer *dbg_buff,
+				  struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	u32 local_offset, local_range;
+	struct ireg_buf *up_cim;
+	u32 size, j, iter;
+	u32 instance = 0;
+	int i, rc, n;
+
+	if (is_t5(padap->params.chip))
+		n = sizeof(t5_up_cim_reg_array) /
+		    ((IREG_NUM_ELEM + 1) * sizeof(u32));
+	else if (is_t6(padap->params.chip))
+		n = sizeof(t6_up_cim_reg_array) /
+		    ((IREG_NUM_ELEM + 1) * sizeof(u32));
+	else
+		return CUDBG_STATUS_NOT_IMPLEMENTED;
+
+	size = sizeof(struct ireg_buf) * n;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	up_cim = (struct ireg_buf *)temp_buff.data;
+	for (i = 0; i < n; i++) {
+		struct ireg_field *up_cim_reg = &up_cim->tp_pio;
+		u32 *buff = up_cim->outbuf;
+
+		if (is_t5(padap->params.chip)) {
+			up_cim_reg->ireg_addr = t5_up_cim_reg_array[i][0];
+			up_cim_reg->ireg_data = t5_up_cim_reg_array[i][1];
+			up_cim_reg->ireg_local_offset =
+						t5_up_cim_reg_array[i][2];
+			up_cim_reg->ireg_offset_range =
+						t5_up_cim_reg_array[i][3];
+			instance = t5_up_cim_reg_array[i][4];
+		} else if (is_t6(padap->params.chip)) {
+			up_cim_reg->ireg_addr = t6_up_cim_reg_array[i][0];
+			up_cim_reg->ireg_data = t6_up_cim_reg_array[i][1];
+			up_cim_reg->ireg_local_offset =
+						t6_up_cim_reg_array[i][2];
+			up_cim_reg->ireg_offset_range =
+						t6_up_cim_reg_array[i][3];
+			instance = t6_up_cim_reg_array[i][4];
+		}
+
+		switch (instance) {
+		case NUM_CIM_CTL_TSCH_CHANNEL_INSTANCES:
+			iter = up_cim_reg->ireg_offset_range;
+			local_offset = 0x120;
+			local_range = 1;
+			break;
+		case NUM_CIM_CTL_TSCH_CHANNEL_TSCH_CLASS_INSTANCES:
+			iter = up_cim_reg->ireg_offset_range;
+			local_offset = 0x10;
+			local_range = 1;
+			break;
+		default:
+			iter = 1;
+			local_offset = 0;
+			local_range = up_cim_reg->ireg_offset_range;
+			break;
+		}
+
+		for (j = 0; j < iter; j++, buff++) {
+			rc = t4_cim_read(padap,
+					 up_cim_reg->ireg_local_offset +
+					 (j * local_offset), local_range, buff);
+			if (rc) {
+				cudbg_put_buff(pdbg_init, &temp_buff);
+				return rc;
+			}
+		}
+		up_cim++;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_pbt_tables(struct cudbg_init *pdbg_init,
+			     struct cudbg_buffer *dbg_buff,
+			     struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_pbt_tables *pbt;
+	int i, rc;
+	u32 addr;
+
+	rc = cudbg_get_buff(pdbg_init, dbg_buff,
+			    sizeof(struct cudbg_pbt_tables),
+			    &temp_buff);
+	if (rc)
+		return rc;
+
+	pbt = (struct cudbg_pbt_tables *)temp_buff.data;
+	/* PBT dynamic entries */
+	addr = CUDBG_CHAC_PBT_ADDR;
+	for (i = 0; i < CUDBG_PBT_DYNAMIC_ENTRIES; i++) {
+		rc = t4_cim_read(padap, addr + (i * 4), 1,
+				 &pbt->pbt_dynamic[i]);
+		if (rc) {
+			cudbg_err->sys_err = rc;
+			cudbg_put_buff(pdbg_init, &temp_buff);
+			return rc;
+		}
+	}
+
+	/* PBT static entries */
+	/* static entries start when bit 6 is set */
+	addr = CUDBG_CHAC_PBT_ADDR + (1 << 6);
+	for (i = 0; i < CUDBG_PBT_STATIC_ENTRIES; i++) {
+		rc = t4_cim_read(padap, addr + (i * 4), 1,
+				 &pbt->pbt_static[i]);
+		if (rc) {
+			cudbg_err->sys_err = rc;
+			cudbg_put_buff(pdbg_init, &temp_buff);
+			return rc;
+		}
+	}
+
+	/* LRF entries */
+	addr = CUDBG_CHAC_PBT_LRF;
+	for (i = 0; i < CUDBG_LRF_ENTRIES; i++) {
+		rc = t4_cim_read(padap, addr + (i * 4), 1,
+				 &pbt->lrf_table[i]);
+		if (rc) {
+			cudbg_err->sys_err = rc;
+			cudbg_put_buff(pdbg_init, &temp_buff);
+			return rc;
+		}
+	}
+
+	/* PBT data entries */
+	addr = CUDBG_CHAC_PBT_DATA;
+	for (i = 0; i < CUDBG_PBT_DATA_ENTRIES; i++) {
+		rc = t4_cim_read(padap, addr + (i * 4), 1,
+				 &pbt->pbt_data[i]);
+		if (rc) {
+			cudbg_err->sys_err = rc;
+			cudbg_put_buff(pdbg_init, &temp_buff);
+			return rc;
+		}
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_mbox_log(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_mbox_log *mboxlog = NULL;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct mbox_cmd_log *log = NULL;
+	struct mbox_cmd *entry;
+	unsigned int entry_idx;
+	u16 mbox_cmds;
+	int i, k, rc;
+	u64 flit;
+	u32 size;
+
+	log = padap->mbox_log;
+	mbox_cmds = padap->mbox_log->size;
+	size = sizeof(struct cudbg_mbox_log) * mbox_cmds;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	mboxlog = (struct cudbg_mbox_log *)temp_buff.data;
+	for (k = 0; k < mbox_cmds; k++) {
+		entry_idx = log->cursor + k;
+		if (entry_idx >= log->size)
+			entry_idx -= log->size;
+
+		entry = mbox_cmd_log_entry(log, entry_idx);
+		/* skip over unused entries */
+		if (entry->timestamp == 0)
+			continue;
+
+		memcpy(&mboxlog->entry, entry, sizeof(struct mbox_cmd));
+		for (i = 0; i < MBOX_LEN / 8; i++) {
+			flit = entry->cmd[i];
+			mboxlog->hi[i] = (u32)(flit >> 32);
+			mboxlog->lo[i] = (u32)flit;
+		}
+		mboxlog++;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+int cudbg_collect_hma_indirect(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct ireg_buf *hma_indr;
+	int i, rc, n;
+	u32 size;
+
+	if (CHELSIO_CHIP_VERSION(padap->params.chip) < CHELSIO_T6)
+		return CUDBG_STATUS_ENTITY_NOT_FOUND;
+
+	n = sizeof(t6_hma_ireg_array) / (IREG_NUM_ELEM * sizeof(u32));
+	size = sizeof(struct ireg_buf) * n;
+	rc = cudbg_get_buff(pdbg_init, dbg_buff, size, &temp_buff);
+	if (rc)
+		return rc;
+
+	hma_indr = (struct ireg_buf *)temp_buff.data;
+	for (i = 0; i < n; i++) {
+		struct ireg_field *hma_fli = &hma_indr->tp_pio;
+		u32 *buff = hma_indr->outbuf;
+
+		hma_fli->ireg_addr = t6_hma_ireg_array[i][0];
+		hma_fli->ireg_data = t6_hma_ireg_array[i][1];
+		hma_fli->ireg_local_offset = t6_hma_ireg_array[i][2];
+		hma_fli->ireg_offset_range = t6_hma_ireg_array[i][3];
+		t4_read_indirect(padap, hma_fli->ireg_addr, hma_fli->ireg_data,
+				 buff, hma_fli->ireg_offset_range,
+				 hma_fli->ireg_local_offset);
+		hma_indr++;
+	}
+	return cudbg_write_and_release_buff(pdbg_init, &temp_buff, dbg_buff);
+}
+
+void cudbg_fill_qdesc_num_and_size(const struct adapter *padap,
+				   u32 *num, u32 *size)
+{
+	u32 tot_entries = 0, tot_size = 0;
+
+	/* NIC TXQ, RXQ, FLQ, and CTRLQ */
+	tot_entries += MAX_ETH_QSETS * 3;
+	tot_entries += MAX_CTRL_QUEUES;
+
+	tot_size += MAX_ETH_QSETS * MAX_TXQ_ENTRIES * MAX_TXQ_DESC_SIZE;
+	tot_size += MAX_ETH_QSETS * MAX_RSPQ_ENTRIES * MAX_RXQ_DESC_SIZE;
+	tot_size += MAX_ETH_QSETS * MAX_RX_BUFFERS * MAX_FL_DESC_SIZE;
+	tot_size += MAX_CTRL_QUEUES * MAX_CTRL_TXQ_ENTRIES *
+		    MAX_CTRL_TXQ_DESC_SIZE;
+
+	/* FW_EVTQ and INTRQ */
+	tot_entries += INGQ_EXTRAS;
+	tot_size += INGQ_EXTRAS * MAX_RSPQ_ENTRIES * MAX_RXQ_DESC_SIZE;
+
+	/* PTP_TXQ */
+	tot_entries += 1;
+	tot_size += MAX_TXQ_ENTRIES * MAX_TXQ_DESC_SIZE;
+
+	/* ULD TXQ, RXQ, and FLQ */
+	tot_entries += CXGB4_TX_MAX * MAX_OFLD_QSETS;
+	tot_entries += CXGB4_ULD_MAX * MAX_ULD_QSETS * 2;
+
+	tot_size += CXGB4_TX_MAX * MAX_OFLD_QSETS * MAX_TXQ_ENTRIES *
+		    MAX_TXQ_DESC_SIZE;
+	tot_size += CXGB4_ULD_MAX * MAX_ULD_QSETS * MAX_RSPQ_ENTRIES *
+		    MAX_RXQ_DESC_SIZE;
+	tot_size += CXGB4_ULD_MAX * MAX_ULD_QSETS * MAX_RX_BUFFERS *
+		    MAX_FL_DESC_SIZE;
+
+	/* ULD CIQ */
+	tot_entries += CXGB4_ULD_MAX * MAX_ULD_QSETS;
+	tot_size += CXGB4_ULD_MAX * MAX_ULD_QSETS * SGE_MAX_IQ_SIZE *
+		    MAX_RXQ_DESC_SIZE;
+
+	/* ETHOFLD TXQ, RXQ, and FLQ */
+	tot_entries += MAX_OFLD_QSETS * 3;
+	tot_size += MAX_OFLD_QSETS * MAX_TXQ_ENTRIES * MAX_TXQ_DESC_SIZE;
+
+	tot_size += sizeof(struct cudbg_ver_hdr) +
+		    sizeof(struct cudbg_qdesc_info) +
+		    sizeof(struct cudbg_qdesc_entry) * tot_entries;
+
+	if (num)
+		*num = tot_entries;
+
+	if (size)
+		*size = tot_size;
+}
+
+int cudbg_collect_qdesc(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *dbg_buff,
+			struct cudbg_error *cudbg_err)
+{
+	u32 num_queues = 0, tot_entries = 0, size = 0;
+	struct adapter *padap = pdbg_init->adap;
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_qdesc_entry *qdesc_entry;
+	struct cudbg_qdesc_info *qdesc_info;
+	struct cudbg_ver_hdr *ver_hdr;
+	struct sge *s = &padap->sge;
+	u32 i, j, cur_off, tot_len;
+	u8 *data;
+	int rc;
+
+	cudbg_fill_qdesc_num_and_size(padap, &tot_entries, &size);
+	size = min_t(u32, size, CUDBG_DUMP_BUFF_SIZE);
+	tot_len = size;
+	data = kvzalloc(size, GFP_KERNEL);
+	if (!data)
+		return -ENOMEM;
+
+	ver_hdr = (struct cudbg_ver_hdr *)data;
+	ver_hdr->signature = CUDBG_ENTITY_SIGNATURE;
+	ver_hdr->revision = CUDBG_QDESC_REV;
+	ver_hdr->size = sizeof(struct cudbg_qdesc_info);
+	size -= sizeof(*ver_hdr);
+
+	qdesc_info = (struct cudbg_qdesc_info *)(data +
+						 sizeof(*ver_hdr));
+	size -= sizeof(*qdesc_info);
+	qdesc_entry = (struct cudbg_qdesc_entry *)qdesc_info->data;
+
+#define QDESC_GET(q, desc, type, label) do { \
+	if (size <= 0) { \
+		goto label; \
+	} \
+	if (desc) { \
+		cudbg_fill_qdesc_##q(q, type, qdesc_entry); \
+		size -= sizeof(*qdesc_entry) + qdesc_entry->data_size; \
+		num_queues++; \
+		qdesc_entry = cudbg_next_qdesc(qdesc_entry); \
+	} \
+} while (0)
+
+#define QDESC_GET_TXQ(q, type, label) do { \
+	struct sge_txq *txq = (struct sge_txq *)q; \
+	QDESC_GET(txq, txq->desc, type, label); \
+} while (0)
+
+#define QDESC_GET_RXQ(q, type, label) do { \
+	struct sge_rspq *rxq = (struct sge_rspq *)q; \
+	QDESC_GET(rxq, rxq->desc, type, label); \
+} while (0)
+
+#define QDESC_GET_FLQ(q, type, label) do { \
+	struct sge_fl *flq = (struct sge_fl *)q; \
+	QDESC_GET(flq, flq->desc, type, label); \
+} while (0)
+
+	/* NIC TXQ */
+	for (i = 0; i < s->ethqsets; i++)
+		QDESC_GET_TXQ(&s->ethtxq[i].q, CUDBG_QTYPE_NIC_TXQ, out);
+
+	/* NIC RXQ */
+	for (i = 0; i < s->ethqsets; i++)
+		QDESC_GET_RXQ(&s->ethrxq[i].rspq, CUDBG_QTYPE_NIC_RXQ, out);
+
+	/* NIC FLQ */
+	for (i = 0; i < s->ethqsets; i++)
+		QDESC_GET_FLQ(&s->ethrxq[i].fl, CUDBG_QTYPE_NIC_FLQ, out);
+
+	/* NIC CTRLQ */
+	for (i = 0; i < padap->params.nports; i++)
+		QDESC_GET_TXQ(&s->ctrlq[i].q, CUDBG_QTYPE_CTRLQ, out);
+
+	/* FW_EVTQ */
+	QDESC_GET_RXQ(&s->fw_evtq, CUDBG_QTYPE_FWEVTQ, out);
+
+	/* INTRQ */
+	QDESC_GET_RXQ(&s->intrq, CUDBG_QTYPE_INTRQ, out);
+
+	/* PTP_TXQ */
+	QDESC_GET_TXQ(&s->ptptxq.q, CUDBG_QTYPE_PTP_TXQ, out);
+
+	/* ULD Queues */
+	mutex_lock(&uld_mutex);
+
+	if (s->uld_txq_info) {
+		struct sge_uld_txq_info *utxq;
+
+		/* ULD TXQ */
+		for (j = 0; j < CXGB4_TX_MAX; j++) {
+			if (!s->uld_txq_info[j])
+				continue;
+
+			utxq = s->uld_txq_info[j];
+			for (i = 0; i < utxq->ntxq; i++)
+				QDESC_GET_TXQ(&utxq->uldtxq[i].q,
+					      cudbg_uld_txq_to_qtype(j),
+					      out_unlock_uld);
+		}
+	}
+
+	if (s->uld_rxq_info) {
+		struct sge_uld_rxq_info *urxq;
+		u32 base;
+
+		/* ULD RXQ */
+		for (j = 0; j < CXGB4_ULD_MAX; j++) {
+			if (!s->uld_rxq_info[j])
+				continue;
+
+			urxq = s->uld_rxq_info[j];
+			for (i = 0; i < urxq->nrxq; i++)
+				QDESC_GET_RXQ(&urxq->uldrxq[i].rspq,
+					      cudbg_uld_rxq_to_qtype(j),
+					      out_unlock_uld);
+		}
+
+		/* ULD FLQ */
+		for (j = 0; j < CXGB4_ULD_MAX; j++) {
+			if (!s->uld_rxq_info[j])
+				continue;
+
+			urxq = s->uld_rxq_info[j];
+			for (i = 0; i < urxq->nrxq; i++)
+				QDESC_GET_FLQ(&urxq->uldrxq[i].fl,
+					      cudbg_uld_flq_to_qtype(j),
+					      out_unlock_uld);
+		}
+
+		/* ULD CIQ */
+		for (j = 0; j < CXGB4_ULD_MAX; j++) {
+			if (!s->uld_rxq_info[j])
+				continue;
+
+			urxq = s->uld_rxq_info[j];
+			base = urxq->nrxq;
+			for (i = 0; i < urxq->nciq; i++)
+				QDESC_GET_RXQ(&urxq->uldrxq[base + i].rspq,
+					      cudbg_uld_ciq_to_qtype(j),
+					      out_unlock_uld);
+		}
+	}
+	mutex_unlock(&uld_mutex);
+
+	if (!padap->tc_mqprio)
+		goto out;
+
+	mutex_lock(&padap->tc_mqprio->mqprio_mutex);
+	/* ETHOFLD TXQ */
+	if (s->eohw_txq)
+		for (i = 0; i < s->eoqsets; i++)
+			QDESC_GET_TXQ(&s->eohw_txq[i].q,
+				      CUDBG_QTYPE_ETHOFLD_TXQ, out_unlock_mqprio);
+
+	/* ETHOFLD RXQ and FLQ */
+	if (s->eohw_rxq) {
+		for (i = 0; i < s->eoqsets; i++)
+			QDESC_GET_RXQ(&s->eohw_rxq[i].rspq,
+				      CUDBG_QTYPE_ETHOFLD_RXQ, out_unlock_mqprio);
+
+		for (i = 0; i < s->eoqsets; i++)
+			QDESC_GET_FLQ(&s->eohw_rxq[i].fl,
+				      CUDBG_QTYPE_ETHOFLD_FLQ, out_unlock_mqprio);
+	}
+
+out_unlock_mqprio:
+	mutex_unlock(&padap->tc_mqprio->mqprio_mutex);
+
+out:
+	qdesc_info->qdesc_entry_size = sizeof(*qdesc_entry);
+	qdesc_info->num_queues = num_queues;
+	cur_off = 0;
+	while (tot_len) {
+		u32 chunk_size = min_t(u32, tot_len, CUDBG_CHUNK_SIZE);
+
+		rc = cudbg_get_buff(pdbg_init, dbg_buff, chunk_size,
+				    &temp_buff);
+		if (rc) {
+			cudbg_err->sys_warn = CUDBG_STATUS_PARTIAL_DATA;
+			goto out_free;
+		}
+
+		memcpy(temp_buff.data, data + cur_off, chunk_size);
+		tot_len -= chunk_size;
+		cur_off += chunk_size;
+		rc = cudbg_write_and_release_buff(pdbg_init, &temp_buff,
+						  dbg_buff);
+		if (rc) {
+			cudbg_put_buff(pdbg_init, &temp_buff);
+			cudbg_err->sys_warn = CUDBG_STATUS_PARTIAL_DATA;
+			goto out_free;
+		}
+	}
+
+out_free:
+	kvfree(data);
+
+#undef QDESC_GET_FLQ
+#undef QDESC_GET_RXQ
+#undef QDESC_GET_TXQ
+#undef QDESC_GET
+
+	return rc;
+
+out_unlock_uld:
+	mutex_unlock(&uld_mutex);
+	goto out;
+}
+
+int cudbg_collect_flash(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *dbg_buff,
+			struct cudbg_error *cudbg_err)
+{
+	struct adapter *padap = pdbg_init->adap;
+	u32 count = padap->params.sf_size, n;
+	struct cudbg_buffer temp_buff = {0};
+	u32 addr, i;
+	int rc;
+
+	addr = FLASH_EXP_ROM_START;
+
+	for (i = 0; i < count; i += SF_PAGE_SIZE) {
+		n = min_t(u32, count - i, SF_PAGE_SIZE);
+
+		rc = cudbg_get_buff(pdbg_init, dbg_buff, n, &temp_buff);
+		if (rc) {
+			cudbg_err->sys_warn = CUDBG_STATUS_PARTIAL_DATA;
+			goto out;
+		}
+		rc = t4_read_flash(padap, addr, n, (u32 *)temp_buff.data, 0);
+		if (rc)
+			goto out;
+
+		addr += (n * 4);
+		rc = cudbg_write_and_release_buff(pdbg_init, &temp_buff,
+						  dbg_buff);
+		if (rc) {
+			cudbg_err->sys_warn = CUDBG_STATUS_PARTIAL_DATA;
+			goto out;
+		}
+	}
+
+out:
+	return rc;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cudbg_lib.h b/drivers/net/ethernet/chelsio/cxgb4/cudbg_lib.h
new file mode 100644
index 000000000..d6d6cd298
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cudbg_lib.h
@@ -0,0 +1,283 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ *  Copyright (C) 2017 Chelsio Communications.  All rights reserved.
+ */
+
+#ifndef __CUDBG_LIB_H__
+#define __CUDBG_LIB_H__
+
+int cudbg_collect_reg_dump(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err);
+int cudbg_collect_fw_devlog(struct cudbg_init *pdbg_init,
+			    struct cudbg_buffer *dbg_buff,
+			    struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_la(struct cudbg_init *pdbg_init,
+			 struct cudbg_buffer *dbg_buff,
+			 struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_ma_la(struct cudbg_init *pdbg_init,
+			    struct cudbg_buffer *dbg_buff,
+			    struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_qcfg(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_ibq_tp0(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_ibq_tp1(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_ibq_ulp(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_ibq_sge0(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_ibq_sge1(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_ibq_ncsi(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_obq_ulp0(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_obq_ulp1(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_obq_ulp2(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_obq_ulp3(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_obq_sge(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_obq_ncsi(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_edc0_meminfo(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_edc1_meminfo(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_mc0_meminfo(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err);
+int cudbg_collect_mc1_meminfo(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err);
+int cudbg_collect_rss(struct cudbg_init *pdbg_init,
+		      struct cudbg_buffer *dbg_buff,
+		      struct cudbg_error *cudbg_err);
+int cudbg_collect_rss_vf_config(struct cudbg_init *pdbg_init,
+				struct cudbg_buffer *dbg_buff,
+				struct cudbg_error *cudbg_err);
+int cudbg_collect_tp_indirect(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err);
+int cudbg_collect_path_mtu(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err);
+int cudbg_collect_pm_stats(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err);
+int cudbg_collect_hw_sched(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err);
+int cudbg_collect_sge_indirect(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_ulprx_la(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err);
+int cudbg_collect_tp_la(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *dbg_buff,
+			struct cudbg_error *cudbg_err);
+int cudbg_collect_meminfo(struct cudbg_init *pdbg_init,
+			  struct cudbg_buffer *dbg_buff,
+			  struct cudbg_error *cudbg_err);
+int cudbg_collect_cim_pif_la(struct cudbg_init *pdbg_init,
+			     struct cudbg_buffer *dbg_buff,
+			     struct cudbg_error *cudbg_err);
+int cudbg_collect_clk_info(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err);
+int cudbg_collect_obq_sge_rx_q0(struct cudbg_init *pdbg_init,
+				struct cudbg_buffer *dbg_buff,
+				struct cudbg_error *cudbg_err);
+int cudbg_collect_obq_sge_rx_q1(struct cudbg_init *pdbg_init,
+				struct cudbg_buffer *dbg_buff,
+				struct cudbg_error *cudbg_err);
+int cudbg_collect_pcie_indirect(struct cudbg_init *pdbg_init,
+				struct cudbg_buffer *dbg_buff,
+				struct cudbg_error *cudbg_err);
+int cudbg_collect_pm_indirect(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err);
+int cudbg_collect_tid(struct cudbg_init *pdbg_init,
+		      struct cudbg_buffer *dbg_buff,
+		      struct cudbg_error *cudbg_err);
+int cudbg_collect_pcie_config(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err);
+int cudbg_collect_dump_context(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_mps_tcam(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err);
+int cudbg_collect_vpd_data(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err);
+int cudbg_collect_le_tcam(struct cudbg_init *pdbg_init,
+			  struct cudbg_buffer *dbg_buff,
+			  struct cudbg_error *cudbg_err);
+int cudbg_collect_cctrl(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *dbg_buff,
+			struct cudbg_error *cudbg_err);
+int cudbg_collect_ma_indirect(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err);
+int cudbg_collect_ulptx_la(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err);
+int cudbg_collect_up_cim_indirect(struct cudbg_init *pdbg_init,
+				  struct cudbg_buffer *dbg_buff,
+				  struct cudbg_error *cudbg_err);
+int cudbg_collect_pbt_tables(struct cudbg_init *pdbg_init,
+			     struct cudbg_buffer *dbg_buff,
+			     struct cudbg_error *cudbg_err);
+int cudbg_collect_mbox_log(struct cudbg_init *pdbg_init,
+			   struct cudbg_buffer *dbg_buff,
+			   struct cudbg_error *cudbg_err);
+int cudbg_collect_hma_indirect(struct cudbg_init *pdbg_init,
+			       struct cudbg_buffer *dbg_buff,
+			       struct cudbg_error *cudbg_err);
+int cudbg_collect_hma_meminfo(struct cudbg_init *pdbg_init,
+			      struct cudbg_buffer *dbg_buff,
+			      struct cudbg_error *cudbg_err);
+int cudbg_collect_qdesc(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *dbg_buff,
+			struct cudbg_error *cudbg_err);
+int cudbg_collect_flash(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *dbg_buff,
+			struct cudbg_error *cudbg_err);
+
+u32 cudbg_get_entity_length(struct adapter *adap, u32 entity);
+struct cudbg_entity_hdr *cudbg_get_entity_hdr(void *outbuf, int i);
+void cudbg_align_debug_buffer(struct cudbg_buffer *dbg_buff,
+			      struct cudbg_entity_hdr *entity_hdr);
+u32 cudbg_cim_obq_size(struct adapter *padap, int qid);
+int cudbg_dump_context_size(struct adapter *padap);
+
+int cudbg_fill_meminfo(struct adapter *padap,
+		       struct cudbg_meminfo *meminfo_buff);
+void cudbg_fill_le_tcam_info(struct adapter *padap,
+			     struct cudbg_tcam *tcam_region);
+void cudbg_fill_qdesc_num_and_size(const struct adapter *padap,
+				   u32 *num, u32 *size);
+
+static inline u32 cudbg_uld_txq_to_qtype(u32 uld)
+{
+	switch (uld) {
+	case CXGB4_TX_OFLD:
+		return CUDBG_QTYPE_OFLD_TXQ;
+	case CXGB4_TX_CRYPTO:
+		return CUDBG_QTYPE_CRYPTO_TXQ;
+	}
+
+	return CUDBG_QTYPE_UNKNOWN;
+}
+
+static inline u32 cudbg_uld_rxq_to_qtype(u32 uld)
+{
+	switch (uld) {
+	case CXGB4_ULD_RDMA:
+		return CUDBG_QTYPE_RDMA_RXQ;
+	case CXGB4_ULD_ISCSI:
+		return CUDBG_QTYPE_ISCSI_RXQ;
+	case CXGB4_ULD_ISCSIT:
+		return CUDBG_QTYPE_ISCSIT_RXQ;
+	case CXGB4_ULD_CRYPTO:
+		return CUDBG_QTYPE_CRYPTO_RXQ;
+	case CXGB4_ULD_TLS:
+		return CUDBG_QTYPE_TLS_RXQ;
+	}
+
+	return CUDBG_QTYPE_UNKNOWN;
+}
+
+static inline u32 cudbg_uld_flq_to_qtype(u32 uld)
+{
+	switch (uld) {
+	case CXGB4_ULD_RDMA:
+		return CUDBG_QTYPE_RDMA_FLQ;
+	case CXGB4_ULD_ISCSI:
+		return CUDBG_QTYPE_ISCSI_FLQ;
+	case CXGB4_ULD_ISCSIT:
+		return CUDBG_QTYPE_ISCSIT_FLQ;
+	case CXGB4_ULD_CRYPTO:
+		return CUDBG_QTYPE_CRYPTO_FLQ;
+	case CXGB4_ULD_TLS:
+		return CUDBG_QTYPE_TLS_FLQ;
+	}
+
+	return CUDBG_QTYPE_UNKNOWN;
+}
+
+static inline u32 cudbg_uld_ciq_to_qtype(u32 uld)
+{
+	switch (uld) {
+	case CXGB4_ULD_RDMA:
+		return CUDBG_QTYPE_RDMA_CIQ;
+	}
+
+	return CUDBG_QTYPE_UNKNOWN;
+}
+
+static inline void cudbg_fill_qdesc_txq(const struct sge_txq *txq,
+					enum cudbg_qdesc_qtype type,
+					struct cudbg_qdesc_entry *entry)
+{
+	entry->qtype = type;
+	entry->qid = txq->cntxt_id;
+	entry->desc_size = sizeof(struct tx_desc);
+	entry->num_desc = txq->size;
+	entry->data_size = txq->size * sizeof(struct tx_desc);
+	memcpy(entry->data, txq->desc, entry->data_size);
+}
+
+static inline void cudbg_fill_qdesc_rxq(const struct sge_rspq *rxq,
+					enum cudbg_qdesc_qtype type,
+					struct cudbg_qdesc_entry *entry)
+{
+	entry->qtype = type;
+	entry->qid = rxq->cntxt_id;
+	entry->desc_size = rxq->iqe_len;
+	entry->num_desc = rxq->size;
+	entry->data_size = rxq->size * rxq->iqe_len;
+	memcpy(entry->data, rxq->desc, entry->data_size);
+}
+
+static inline void cudbg_fill_qdesc_flq(const struct sge_fl *flq,
+					enum cudbg_qdesc_qtype type,
+					struct cudbg_qdesc_entry *entry)
+{
+	entry->qtype = type;
+	entry->qid = flq->cntxt_id;
+	entry->desc_size = sizeof(__be64);
+	entry->num_desc = flq->size;
+	entry->data_size = flq->size * sizeof(__be64);
+	memcpy(entry->data, flq->desc, entry->data_size);
+}
+
+static inline
+struct cudbg_qdesc_entry *cudbg_next_qdesc(struct cudbg_qdesc_entry *e)
+{
+	return (struct cudbg_qdesc_entry *)
+	       ((u8 *)e + sizeof(*e) + e->data_size);
+}
+#endif /* __CUDBG_LIB_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cudbg_lib_common.h b/drivers/net/ethernet/chelsio/cxgb4/cudbg_lib_common.h
new file mode 100644
index 000000000..9fac777b0
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cudbg_lib_common.h
@@ -0,0 +1,77 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ *  Copyright (C) 2017 Chelsio Communications.  All rights reserved.
+ */
+
+#ifndef __CUDBG_LIB_COMMON_H__
+#define __CUDBG_LIB_COMMON_H__
+
+#define CUDBG_SIGNATURE 67856866 /* CUDB in ascii */
+
+enum cudbg_dump_type {
+	CUDBG_DUMP_TYPE_MINI = 1,
+};
+
+enum cudbg_compression_type {
+	CUDBG_COMPRESSION_NONE = 1,
+	CUDBG_COMPRESSION_ZLIB,
+};
+
+struct cudbg_hdr {
+	u32 signature;
+	u32 hdr_len;
+	u16 major_ver;
+	u16 minor_ver;
+	u32 data_len;
+	u32 hdr_flags;
+	u16 max_entities;
+	u8 chip_ver;
+	u8 dump_type:3;
+	u8 reserved1:1;
+	u8 compress_type:4;
+	u32 reserved[8];
+};
+
+struct cudbg_entity_hdr {
+	u32 entity_type;
+	u32 start_offset;
+	u32 size;
+	int hdr_flags;
+	u32 sys_warn;
+	u32 sys_err;
+	u8 num_pad;
+	u8 flag;             /* bit 0 is used to indicate ext data */
+	u8 reserved1[2];
+	u32 next_ext_offset; /* pointer to next extended entity meta data */
+	u32 reserved[5];
+};
+
+struct cudbg_ver_hdr {
+	u32 signature;
+	u16 revision;
+	u16 size;
+};
+
+struct cudbg_buffer {
+	u32 size;
+	u32 offset;
+	char *data;
+};
+
+struct cudbg_error {
+	int sys_err;
+	int sys_warn;
+	int app_err;
+};
+
+#define CDUMP_MAX_COMP_BUF_SIZE ((64 * 1024) - 1)
+#define CUDBG_CHUNK_SIZE ((CDUMP_MAX_COMP_BUF_SIZE / 1024) * 1024)
+
+int cudbg_get_buff(struct cudbg_init *pdbg_init,
+		   struct cudbg_buffer *pdbg_buff, u32 size,
+		   struct cudbg_buffer *pin_buff);
+void cudbg_put_buff(struct cudbg_init *pdbg_init,
+		    struct cudbg_buffer *pin_buff);
+void cudbg_update_buff(struct cudbg_buffer *pin_buff,
+		       struct cudbg_buffer *pout_buff);
+#endif /* __CUDBG_LIB_COMMON_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cudbg_zlib.c b/drivers/net/ethernet/chelsio/cxgb4/cudbg_zlib.c
new file mode 100644
index 000000000..aad55fb35
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cudbg_zlib.c
@@ -0,0 +1,70 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  Copyright (C) 2018 Chelsio Communications.  All rights reserved.
+ */
+
+#include <linux/zlib.h>
+
+#include "cxgb4.h"
+#include "cudbg_if.h"
+#include "cudbg_lib_common.h"
+#include "cudbg_zlib.h"
+
+static int cudbg_get_compress_hdr(struct cudbg_buffer *pdbg_buff,
+				  struct cudbg_buffer *pin_buff)
+{
+	if (pdbg_buff->offset + sizeof(struct cudbg_compress_hdr) >
+	    pdbg_buff->size)
+		return CUDBG_STATUS_NO_MEM;
+
+	pin_buff->data = (char *)pdbg_buff->data + pdbg_buff->offset;
+	pin_buff->offset = 0;
+	pin_buff->size = sizeof(struct cudbg_compress_hdr);
+	pdbg_buff->offset += sizeof(struct cudbg_compress_hdr);
+	return 0;
+}
+
+int cudbg_compress_buff(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *pin_buff,
+			struct cudbg_buffer *pout_buff)
+{
+	struct cudbg_buffer temp_buff = { 0 };
+	struct z_stream_s compress_stream;
+	struct cudbg_compress_hdr *c_hdr;
+	int rc;
+
+	/* Write compression header to output buffer before compression */
+	rc = cudbg_get_compress_hdr(pout_buff, &temp_buff);
+	if (rc)
+		return rc;
+
+	c_hdr = (struct cudbg_compress_hdr *)temp_buff.data;
+	c_hdr->compress_id = CUDBG_ZLIB_COMPRESS_ID;
+
+	memset(&compress_stream, 0, sizeof(struct z_stream_s));
+	compress_stream.workspace = pdbg_init->workspace;
+	rc = zlib_deflateInit2(&compress_stream, Z_DEFAULT_COMPRESSION,
+			       Z_DEFLATED, CUDBG_ZLIB_WIN_BITS,
+			       CUDBG_ZLIB_MEM_LVL, Z_DEFAULT_STRATEGY);
+	if (rc != Z_OK)
+		return CUDBG_SYSTEM_ERROR;
+
+	compress_stream.next_in = pin_buff->data;
+	compress_stream.avail_in = pin_buff->size;
+	compress_stream.next_out = pout_buff->data + pout_buff->offset;
+	compress_stream.avail_out = pout_buff->size - pout_buff->offset;
+
+	rc = zlib_deflate(&compress_stream, Z_FINISH);
+	if (rc != Z_STREAM_END)
+		return CUDBG_SYSTEM_ERROR;
+
+	rc = zlib_deflateEnd(&compress_stream);
+	if (rc != Z_OK)
+		return CUDBG_SYSTEM_ERROR;
+
+	c_hdr->compress_size = compress_stream.total_out;
+	c_hdr->decompress_size = pin_buff->size;
+	pout_buff->offset += compress_stream.total_out;
+
+	return 0;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cudbg_zlib.h b/drivers/net/ethernet/chelsio/cxgb4/cudbg_zlib.h
new file mode 100644
index 000000000..f6d83289f
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cudbg_zlib.h
@@ -0,0 +1,31 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ *  Copyright (C) 2018 Chelsio Communications.  All rights reserved.
+ */
+
+#ifndef __CUDBG_ZLIB_H__
+#define __CUDBG_ZLIB_H__
+
+#include <linux/zlib.h>
+
+#define CUDBG_ZLIB_COMPRESS_ID 17
+#define CUDBG_ZLIB_WIN_BITS 12
+#define CUDBG_ZLIB_MEM_LVL 4
+
+struct cudbg_compress_hdr {
+	u32 compress_id;
+	u64 decompress_size;
+	u64 compress_size;
+	u64 rsvd[32];
+};
+
+static inline int cudbg_get_workspace_size(void)
+{
+	return zlib_deflate_workspacesize(CUDBG_ZLIB_WIN_BITS,
+					  CUDBG_ZLIB_MEM_LVL);
+}
+
+int cudbg_compress_buff(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *pin_buff,
+			struct cudbg_buffer *pout_buff);
+#endif /* __CUDBG_ZLIB_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
new file mode 100644
index 000000000..5657ac8cf
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
@@ -0,0 +1,2185 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_H__
+#define __CXGB4_H__
+
+#include "t4_hw.h"
+
+#include <linux/bitops.h>
+#include <linux/cache.h>
+#include <linux/ethtool.h>
+#include <linux/interrupt.h>
+#include <linux/list.h>
+#include <linux/netdevice.h>
+#include <linux/pci.h>
+#include <linux/spinlock.h>
+#include <linux/timer.h>
+#include <linux/vmalloc.h>
+#include <linux/rhashtable.h>
+#include <linux/etherdevice.h>
+#include <linux/net_tstamp.h>
+#include <linux/ptp_clock_kernel.h>
+#include <linux/ptp_classify.h>
+#include <linux/crash_dump.h>
+#include <linux/thermal.h>
+#include <asm/io.h>
+#include "t4_chip_type.h"
+#include "cxgb4_uld.h"
+#include "t4fw_api.h"
+
+#define CH_WARN(adap, fmt, ...) dev_warn(adap->pdev_dev, fmt, ## __VA_ARGS__)
+extern struct list_head adapter_list;
+extern struct list_head uld_list;
+extern struct mutex uld_mutex;
+
+/* Suspend an Ethernet Tx queue with fewer available descriptors than this.
+ * This is the same as calc_tx_descs() for a TSO packet with
+ * nr_frags == MAX_SKB_FRAGS.
+ */
+#define ETHTXQ_STOP_THRES \
+	(1 + DIV_ROUND_UP((3 * MAX_SKB_FRAGS) / 2 + (MAX_SKB_FRAGS & 1), 8))
+
+#define FW_PARAM_DEV(param) \
+	(FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) | \
+	 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_##param))
+
+#define FW_PARAM_PFVF(param) \
+	(FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_PFVF) | \
+	 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_PFVF_##param) |  \
+	 FW_PARAMS_PARAM_Y_V(0) | \
+	 FW_PARAMS_PARAM_Z_V(0))
+
+enum {
+	MAX_NPORTS	= 4,     /* max # of ports */
+	SERNUM_LEN	= 24,    /* Serial # length */
+	ID_LEN		= 16,    /* ID length */
+	PN_LEN		= 16,    /* Part Number length */
+	MACADDR_LEN	= 12,    /* MAC Address length */
+};
+
+enum {
+	T4_REGMAP_SIZE = (160 * 1024),
+	T5_REGMAP_SIZE = (332 * 1024),
+};
+
+enum {
+	MEM_EDC0,
+	MEM_EDC1,
+	MEM_MC,
+	MEM_MC0 = MEM_MC,
+	MEM_MC1,
+	MEM_HMA,
+};
+
+enum {
+	MEMWIN0_APERTURE = 2048,
+	MEMWIN0_BASE     = 0x1b800,
+	MEMWIN1_APERTURE = 32768,
+	MEMWIN1_BASE     = 0x28000,
+	MEMWIN1_BASE_T5  = 0x52000,
+	MEMWIN2_APERTURE = 65536,
+	MEMWIN2_BASE     = 0x30000,
+	MEMWIN2_APERTURE_T5 = 131072,
+	MEMWIN2_BASE_T5  = 0x60000,
+};
+
+enum dev_master {
+	MASTER_CANT,
+	MASTER_MAY,
+	MASTER_MUST
+};
+
+enum dev_state {
+	DEV_STATE_UNINIT,
+	DEV_STATE_INIT,
+	DEV_STATE_ERR
+};
+
+enum cc_pause {
+	PAUSE_RX      = 1 << 0,
+	PAUSE_TX      = 1 << 1,
+	PAUSE_AUTONEG = 1 << 2
+};
+
+enum cc_fec {
+	FEC_AUTO      = 1 << 0,	 /* IEEE 802.3 "automatic" */
+	FEC_RS        = 1 << 1,  /* Reed-Solomon */
+	FEC_BASER_RS  = 1 << 2   /* BaseR/Reed-Solomon */
+};
+
+enum {
+	CXGB4_ETHTOOL_FLASH_FW = 1,
+	CXGB4_ETHTOOL_FLASH_PHY = 2,
+	CXGB4_ETHTOOL_FLASH_BOOT = 3,
+	CXGB4_ETHTOOL_FLASH_BOOTCFG = 4
+};
+
+enum cxgb4_netdev_tls_ops {
+	CXGB4_TLSDEV_OPS  = 1,
+	CXGB4_XFRMDEV_OPS
+};
+
+struct cxgb4_bootcfg_data {
+	__le16 signature;
+	__u8 reserved[2];
+};
+
+struct cxgb4_pcir_data {
+	__le32 signature;	/* Signature. The string "PCIR" */
+	__le16 vendor_id;	/* Vendor Identification */
+	__le16 device_id;	/* Device Identification */
+	__u8 vital_product[2];	/* Pointer to Vital Product Data */
+	__u8 length[2];		/* PCIR Data Structure Length */
+	__u8 revision;		/* PCIR Data Structure Revision */
+	__u8 class_code[3];	/* Class Code */
+	__u8 image_length[2];	/* Image Length. Multiple of 512B */
+	__u8 code_revision[2];	/* Revision Level of Code/Data */
+	__u8 code_type;
+	__u8 indicator;
+	__u8 reserved[2];
+};
+
+/* BIOS boot headers */
+struct cxgb4_pci_exp_rom_header {
+	__le16 signature;	/* ROM Signature. Should be 0xaa55 */
+	__u8 reserved[22];	/* Reserved per processor Architecture data */
+	__le16 pcir_offset;	/* Offset to PCI Data Structure */
+};
+
+/* Legacy PCI Expansion ROM Header */
+struct legacy_pci_rom_hdr {
+	__u8 signature[2];	/* ROM Signature. Should be 0xaa55 */
+	__u8 size512;		/* Current Image Size in units of 512 bytes */
+	__u8 initentry_point[4];
+	__u8 cksum;		/* Checksum computed on the entire Image */
+	__u8 reserved[16];	/* Reserved */
+	__le16 pcir_offset;	/* Offset to PCI Data Struture */
+};
+
+#define CXGB4_HDR_CODE1 0x00
+#define CXGB4_HDR_CODE2 0x03
+#define CXGB4_HDR_INDI 0x80
+
+/* BOOT constants */
+enum {
+	BOOT_CFG_SIG = 0x4243,
+	BOOT_SIZE_INC = 512,
+	BOOT_SIGNATURE = 0xaa55,
+	BOOT_MIN_SIZE = sizeof(struct cxgb4_pci_exp_rom_header),
+	BOOT_MAX_SIZE = 1024 * BOOT_SIZE_INC,
+	PCIR_SIGNATURE = 0x52494350
+};
+
+struct port_stats {
+	u64 tx_octets;            /* total # of octets in good frames */
+	u64 tx_frames;            /* all good frames */
+	u64 tx_bcast_frames;      /* all broadcast frames */
+	u64 tx_mcast_frames;      /* all multicast frames */
+	u64 tx_ucast_frames;      /* all unicast frames */
+	u64 tx_error_frames;      /* all error frames */
+
+	u64 tx_frames_64;         /* # of Tx frames in a particular range */
+	u64 tx_frames_65_127;
+	u64 tx_frames_128_255;
+	u64 tx_frames_256_511;
+	u64 tx_frames_512_1023;
+	u64 tx_frames_1024_1518;
+	u64 tx_frames_1519_max;
+
+	u64 tx_drop;              /* # of dropped Tx frames */
+	u64 tx_pause;             /* # of transmitted pause frames */
+	u64 tx_ppp0;              /* # of transmitted PPP prio 0 frames */
+	u64 tx_ppp1;              /* # of transmitted PPP prio 1 frames */
+	u64 tx_ppp2;              /* # of transmitted PPP prio 2 frames */
+	u64 tx_ppp3;              /* # of transmitted PPP prio 3 frames */
+	u64 tx_ppp4;              /* # of transmitted PPP prio 4 frames */
+	u64 tx_ppp5;              /* # of transmitted PPP prio 5 frames */
+	u64 tx_ppp6;              /* # of transmitted PPP prio 6 frames */
+	u64 tx_ppp7;              /* # of transmitted PPP prio 7 frames */
+
+	u64 rx_octets;            /* total # of octets in good frames */
+	u64 rx_frames;            /* all good frames */
+	u64 rx_bcast_frames;      /* all broadcast frames */
+	u64 rx_mcast_frames;      /* all multicast frames */
+	u64 rx_ucast_frames;      /* all unicast frames */
+	u64 rx_too_long;          /* # of frames exceeding MTU */
+	u64 rx_jabber;            /* # of jabber frames */
+	u64 rx_fcs_err;           /* # of received frames with bad FCS */
+	u64 rx_len_err;           /* # of received frames with length error */
+	u64 rx_symbol_err;        /* symbol errors */
+	u64 rx_runt;              /* # of short frames */
+
+	u64 rx_frames_64;         /* # of Rx frames in a particular range */
+	u64 rx_frames_65_127;
+	u64 rx_frames_128_255;
+	u64 rx_frames_256_511;
+	u64 rx_frames_512_1023;
+	u64 rx_frames_1024_1518;
+	u64 rx_frames_1519_max;
+
+	u64 rx_pause;             /* # of received pause frames */
+	u64 rx_ppp0;              /* # of received PPP prio 0 frames */
+	u64 rx_ppp1;              /* # of received PPP prio 1 frames */
+	u64 rx_ppp2;              /* # of received PPP prio 2 frames */
+	u64 rx_ppp3;              /* # of received PPP prio 3 frames */
+	u64 rx_ppp4;              /* # of received PPP prio 4 frames */
+	u64 rx_ppp5;              /* # of received PPP prio 5 frames */
+	u64 rx_ppp6;              /* # of received PPP prio 6 frames */
+	u64 rx_ppp7;              /* # of received PPP prio 7 frames */
+
+	u64 rx_ovflow0;           /* drops due to buffer-group 0 overflows */
+	u64 rx_ovflow1;           /* drops due to buffer-group 1 overflows */
+	u64 rx_ovflow2;           /* drops due to buffer-group 2 overflows */
+	u64 rx_ovflow3;           /* drops due to buffer-group 3 overflows */
+	u64 rx_trunc0;            /* buffer-group 0 truncated packets */
+	u64 rx_trunc1;            /* buffer-group 1 truncated packets */
+	u64 rx_trunc2;            /* buffer-group 2 truncated packets */
+	u64 rx_trunc3;            /* buffer-group 3 truncated packets */
+};
+
+struct lb_port_stats {
+	u64 octets;
+	u64 frames;
+	u64 bcast_frames;
+	u64 mcast_frames;
+	u64 ucast_frames;
+	u64 error_frames;
+
+	u64 frames_64;
+	u64 frames_65_127;
+	u64 frames_128_255;
+	u64 frames_256_511;
+	u64 frames_512_1023;
+	u64 frames_1024_1518;
+	u64 frames_1519_max;
+
+	u64 drop;
+
+	u64 ovflow0;
+	u64 ovflow1;
+	u64 ovflow2;
+	u64 ovflow3;
+	u64 trunc0;
+	u64 trunc1;
+	u64 trunc2;
+	u64 trunc3;
+};
+
+struct tp_tcp_stats {
+	u32 tcp_out_rsts;
+	u64 tcp_in_segs;
+	u64 tcp_out_segs;
+	u64 tcp_retrans_segs;
+};
+
+struct tp_usm_stats {
+	u32 frames;
+	u32 drops;
+	u64 octets;
+};
+
+struct tp_fcoe_stats {
+	u32 frames_ddp;
+	u32 frames_drop;
+	u64 octets_ddp;
+};
+
+struct tp_err_stats {
+	u32 mac_in_errs[4];
+	u32 hdr_in_errs[4];
+	u32 tcp_in_errs[4];
+	u32 tnl_cong_drops[4];
+	u32 ofld_chan_drops[4];
+	u32 tnl_tx_drops[4];
+	u32 ofld_vlan_drops[4];
+	u32 tcp6_in_errs[4];
+	u32 ofld_no_neigh;
+	u32 ofld_cong_defer;
+};
+
+struct tp_cpl_stats {
+	u32 req[4];
+	u32 rsp[4];
+};
+
+struct tp_rdma_stats {
+	u32 rqe_dfr_pkt;
+	u32 rqe_dfr_mod;
+};
+
+struct sge_params {
+	u32 hps;			/* host page size for our PF/VF */
+	u32 eq_qpp;			/* egress queues/page for our PF/VF */
+	u32 iq_qpp;			/* egress queues/page for our PF/VF */
+};
+
+struct tp_params {
+	unsigned int tre;            /* log2 of core clocks per TP tick */
+	unsigned int la_mask;        /* what events are recorded by TP LA */
+	unsigned short tx_modq_map;  /* TX modulation scheduler queue to */
+				     /* channel map */
+
+	uint32_t dack_re;            /* DACK timer resolution */
+	unsigned short tx_modq[NCHAN];	/* channel to modulation queue map */
+
+	u32 vlan_pri_map;               /* cached TP_VLAN_PRI_MAP */
+	u32 filter_mask;
+	u32 ingress_config;             /* cached TP_INGRESS_CONFIG */
+
+	/* cached TP_OUT_CONFIG compressed error vector
+	 * and passing outer header info for encapsulated packets.
+	 */
+	int rx_pkt_encap;
+
+	/* TP_VLAN_PRI_MAP Compressed Filter Tuple field offsets.  This is a
+	 * subset of the set of fields which may be present in the Compressed
+	 * Filter Tuple portion of filters and TCP TCB connections.  The
+	 * fields which are present are controlled by the TP_VLAN_PRI_MAP.
+	 * Since a variable number of fields may or may not be present, their
+	 * shifted field positions within the Compressed Filter Tuple may
+	 * vary, or not even be present if the field isn't selected in
+	 * TP_VLAN_PRI_MAP.  Since some of these fields are needed in various
+	 * places we store their offsets here, or a -1 if the field isn't
+	 * present.
+	 */
+	int fcoe_shift;
+	int port_shift;
+	int vnic_shift;
+	int vlan_shift;
+	int tos_shift;
+	int protocol_shift;
+	int ethertype_shift;
+	int macmatch_shift;
+	int matchtype_shift;
+	int frag_shift;
+
+	u64 hash_filter_mask;
+};
+
+struct vpd_params {
+	unsigned int cclk;
+	u8 sn[SERNUM_LEN + 1];
+	u8 id[ID_LEN + 1];
+	u8 pn[PN_LEN + 1];
+	u8 na[MACADDR_LEN + 1];
+};
+
+/* Maximum resources provisioned for a PCI PF.
+ */
+struct pf_resources {
+	unsigned int nvi;		/* N virtual interfaces */
+	unsigned int neq;		/* N egress Qs */
+	unsigned int nethctrl;		/* N egress ETH or CTRL Qs */
+	unsigned int niqflint;		/* N ingress Qs/w free list(s) & intr */
+	unsigned int niq;		/* N ingress Qs */
+	unsigned int tc;		/* PCI-E traffic class */
+	unsigned int pmask;		/* port access rights mask */
+	unsigned int nexactf;		/* N exact MPS filters */
+	unsigned int r_caps;		/* read capabilities */
+	unsigned int wx_caps;		/* write/execute capabilities */
+};
+
+struct pci_params {
+	unsigned char speed;
+	unsigned char width;
+};
+
+struct devlog_params {
+	u32 memtype;                    /* which memory (EDC0, EDC1, MC) */
+	u32 start;                      /* start of log in firmware memory */
+	u32 size;                       /* size of log */
+};
+
+/* Stores chip specific parameters */
+struct arch_specific_params {
+	u8 nchan;
+	u8 pm_stats_cnt;
+	u8 cng_ch_bits_log;		/* congestion channel map bits width */
+	u16 mps_rplc_size;
+	u16 vfcount;
+	u32 sge_fl_db;
+	u16 mps_tcam_size;
+};
+
+struct adapter_params {
+	struct sge_params sge;
+	struct tp_params  tp;
+	struct vpd_params vpd;
+	struct pf_resources pfres;
+	struct pci_params pci;
+	struct devlog_params devlog;
+	enum pcie_memwin drv_memwin;
+
+	unsigned int cim_la_size;
+
+	unsigned int sf_size;             /* serial flash size in bytes */
+	unsigned int sf_nsec;             /* # of flash sectors */
+
+	unsigned int fw_vers;		  /* firmware version */
+	unsigned int bs_vers;		  /* bootstrap version */
+	unsigned int tp_vers;		  /* TP microcode version */
+	unsigned int er_vers;		  /* expansion ROM version */
+	unsigned int scfg_vers;		  /* Serial Configuration version */
+	unsigned int vpd_vers;		  /* VPD Version */
+	u8 api_vers[7];
+
+	unsigned short mtus[NMTUS];
+	unsigned short a_wnd[NCCTRL_WIN];
+	unsigned short b_wnd[NCCTRL_WIN];
+
+	unsigned char nports;             /* # of ethernet ports */
+	unsigned char portvec;
+	enum chip_type chip;               /* chip code */
+	struct arch_specific_params arch;  /* chip specific params */
+	unsigned char offload;
+	unsigned char crypto;		/* HW capability for crypto */
+	unsigned char ethofld;		/* QoS support */
+
+	unsigned char bypass;
+	unsigned char hash_filter;
+
+	unsigned int ofldq_wr_cred;
+	bool ulptx_memwrite_dsgl;          /* use of T5 DSGL allowed */
+
+	unsigned int nsched_cls;          /* number of traffic classes */
+	unsigned int max_ordird_qp;       /* Max read depth per RDMA QP */
+	unsigned int max_ird_adapter;     /* Max read depth per adapter */
+	bool fr_nsmr_tpte_wr_support;	  /* FW support for FR_NSMR_TPTE_WR */
+	u8 fw_caps_support;		/* 32-bit Port Capabilities */
+	bool filter2_wr_support;	/* FW support for FILTER2_WR */
+	unsigned int viid_smt_extn_support:1; /* FW returns vin and smt index */
+
+	/* MPS Buffer Group Map[per Port].  Bit i is set if buffer group i is
+	 * used by the Port
+	 */
+	u8 mps_bg_map[MAX_NPORTS];	/* MPS Buffer Group Map */
+	bool write_w_imm_support;       /* FW supports WRITE_WITH_IMMEDIATE */
+	bool write_cmpl_support;        /* FW supports WRITE_CMPL */
+};
+
+/* State needed to monitor the forward progress of SGE Ingress DMA activities
+ * and possible hangs.
+ */
+struct sge_idma_monitor_state {
+	unsigned int idma_1s_thresh;	/* 1s threshold in Core Clock ticks */
+	unsigned int idma_stalled[2];	/* synthesized stalled timers in HZ */
+	unsigned int idma_state[2];	/* IDMA Hang detect state */
+	unsigned int idma_qid[2];	/* IDMA Hung Ingress Queue ID */
+	unsigned int idma_warn[2];	/* time to warning in HZ */
+};
+
+/* Firmware Mailbox Command/Reply log.  All values are in Host-Endian format.
+ * The access and execute times are signed in order to accommodate negative
+ * error returns.
+ */
+struct mbox_cmd {
+	u64 cmd[MBOX_LEN / 8];		/* a Firmware Mailbox Command/Reply */
+	u64 timestamp;			/* OS-dependent timestamp */
+	u32 seqno;			/* sequence number */
+	s16 access;			/* time (ms) to access mailbox */
+	s16 execute;			/* time (ms) to execute */
+};
+
+struct mbox_cmd_log {
+	unsigned int size;		/* number of entries in the log */
+	unsigned int cursor;		/* next position in the log to write */
+	u32 seqno;			/* next sequence number */
+	/* variable length mailbox command log starts here */
+};
+
+/* Given a pointer to a Firmware Mailbox Command Log and a log entry index,
+ * return a pointer to the specified entry.
+ */
+static inline struct mbox_cmd *mbox_cmd_log_entry(struct mbox_cmd_log *log,
+						  unsigned int entry_idx)
+{
+	return &((struct mbox_cmd *)&(log)[1])[entry_idx];
+}
+
+#define FW_VERSION(chip) ( \
+		FW_HDR_FW_VER_MAJOR_G(chip##FW_VERSION_MAJOR) | \
+		FW_HDR_FW_VER_MINOR_G(chip##FW_VERSION_MINOR) | \
+		FW_HDR_FW_VER_MICRO_G(chip##FW_VERSION_MICRO) | \
+		FW_HDR_FW_VER_BUILD_G(chip##FW_VERSION_BUILD))
+#define FW_INTFVER(chip, intf) (FW_HDR_INTFVER_##intf)
+
+struct cxgb4_ethtool_lb_test {
+	struct completion completion;
+	int result;
+	int loopback;
+};
+
+struct fw_info {
+	u8 chip;
+	char *fs_name;
+	char *fw_mod_name;
+	struct fw_hdr fw_hdr;
+};
+
+struct trace_params {
+	u32 data[TRACE_LEN / 4];
+	u32 mask[TRACE_LEN / 4];
+	unsigned short snap_len;
+	unsigned short min_len;
+	unsigned char skip_ofst;
+	unsigned char skip_len;
+	unsigned char invert;
+	unsigned char port;
+};
+
+struct cxgb4_fw_data {
+	__be32 signature;
+	__u8 reserved[4];
+};
+
+/* Firmware Port Capabilities types. */
+
+typedef u16 fw_port_cap16_t;	/* 16-bit Port Capabilities integral value */
+typedef u32 fw_port_cap32_t;	/* 32-bit Port Capabilities integral value */
+
+enum fw_caps {
+	FW_CAPS_UNKNOWN	= 0,	/* 0'ed out initial state */
+	FW_CAPS16	= 1,	/* old Firmware: 16-bit Port Capabilities */
+	FW_CAPS32	= 2,	/* new Firmware: 32-bit Port Capabilities */
+};
+
+struct link_config {
+	fw_port_cap32_t pcaps;           /* link capabilities */
+	fw_port_cap32_t def_acaps;       /* default advertised capabilities */
+	fw_port_cap32_t acaps;           /* advertised capabilities */
+	fw_port_cap32_t lpacaps;         /* peer advertised capabilities */
+
+	fw_port_cap32_t speed_caps;      /* speed(s) user has requested */
+	unsigned int   speed;            /* actual link speed (Mb/s) */
+
+	enum cc_pause  requested_fc;     /* flow control user has requested */
+	enum cc_pause  fc;               /* actual link flow control */
+	enum cc_pause  advertised_fc;    /* actual advertised flow control */
+
+	enum cc_fec    requested_fec;	 /* Forward Error Correction: */
+	enum cc_fec    fec;		 /* requested and actual in use */
+
+	unsigned char  autoneg;          /* autonegotiating? */
+
+	unsigned char  link_ok;          /* link up? */
+	unsigned char  link_down_rc;     /* link down reason */
+
+	bool new_module;		 /* ->OS Transceiver Module inserted */
+	bool redo_l1cfg;		 /* ->CC redo current "sticky" L1 CFG */
+};
+
+#define FW_LEN16(fw_struct) FW_CMD_LEN16_V(sizeof(fw_struct) / 16)
+
+enum {
+	MAX_ETH_QSETS = 32,           /* # of Ethernet Tx/Rx queue sets */
+	MAX_OFLD_QSETS = 16,          /* # of offload Tx, iscsi Rx queue sets */
+	MAX_CTRL_QUEUES = NCHAN,      /* # of control Tx queues */
+};
+
+enum {
+	MAX_TXQ_ENTRIES      = 16384,
+	MAX_CTRL_TXQ_ENTRIES = 1024,
+	MAX_RSPQ_ENTRIES     = 16384,
+	MAX_RX_BUFFERS       = 16384,
+	MIN_TXQ_ENTRIES      = 32,
+	MIN_CTRL_TXQ_ENTRIES = 32,
+	MIN_RSPQ_ENTRIES     = 128,
+	MIN_FL_ENTRIES       = 16
+};
+
+enum {
+	MAX_TXQ_DESC_SIZE      = 64,
+	MAX_RXQ_DESC_SIZE      = 128,
+	MAX_FL_DESC_SIZE       = 8,
+	MAX_CTRL_TXQ_DESC_SIZE = 64,
+};
+
+enum {
+	INGQ_EXTRAS = 2,        /* firmware event queue and */
+				/*   forwarded interrupts */
+	MAX_INGQ = MAX_ETH_QSETS + INGQ_EXTRAS,
+};
+
+enum {
+	PRIV_FLAG_PORT_TX_VM_BIT,
+};
+
+#define PRIV_FLAG_PORT_TX_VM		BIT(PRIV_FLAG_PORT_TX_VM_BIT)
+
+#define PRIV_FLAGS_ADAP			0
+#define PRIV_FLAGS_PORT			PRIV_FLAG_PORT_TX_VM
+
+struct adapter;
+struct sge_rspq;
+
+#include "cxgb4_dcb.h"
+
+#ifdef CONFIG_CHELSIO_T4_FCOE
+#include "cxgb4_fcoe.h"
+#endif /* CONFIG_CHELSIO_T4_FCOE */
+
+struct port_info {
+	struct adapter *adapter;
+	u16    viid;
+	int    xact_addr_filt;        /* index of exact MAC address filter */
+	u16    rss_size;              /* size of VI's RSS table slice */
+	s8     mdio_addr;
+	enum fw_port_type port_type;
+	u8     mod_type;
+	u8     port_id;
+	u8     tx_chan;
+	u8     lport;                 /* associated offload logical port */
+	u8     nqsets;                /* # of qsets */
+	u8     first_qset;            /* index of first qset */
+	u8     rss_mode;
+	struct link_config link_cfg;
+	u16   *rss;
+	struct port_stats stats_base;
+#ifdef CONFIG_CHELSIO_T4_DCB
+	struct port_dcb_info dcb;     /* Data Center Bridging support */
+#endif
+#ifdef CONFIG_CHELSIO_T4_FCOE
+	struct cxgb_fcoe fcoe;
+#endif /* CONFIG_CHELSIO_T4_FCOE */
+	bool rxtstamp;  /* Enable TS */
+	struct hwtstamp_config tstamp_config;
+	bool ptp_enable;
+	struct sched_table *sched_tbl;
+	u32 eth_flags;
+
+	/* viid and smt fields either returned by fw
+	 * or decoded by parsing viid by driver.
+	 */
+	u8 vin;
+	u8 vivld;
+	u8 smt_idx;
+	u8 rx_cchan;
+
+	bool tc_block_shared;
+
+	/* Mirror VI information */
+	u16 viid_mirror;
+	u16 nmirrorqsets;
+	u32 vi_mirror_count;
+	struct mutex vi_mirror_mutex; /* Sync access to Mirror VI info */
+	struct cxgb4_ethtool_lb_test ethtool_lb;
+};
+
+struct dentry;
+struct work_struct;
+
+enum {                                 /* adapter flags */
+	CXGB4_FULL_INIT_DONE		= (1 << 0),
+	CXGB4_DEV_ENABLED		= (1 << 1),
+	CXGB4_USING_MSI			= (1 << 2),
+	CXGB4_USING_MSIX		= (1 << 3),
+	CXGB4_FW_OK			= (1 << 4),
+	CXGB4_RSS_TNLALLLOOKUP		= (1 << 5),
+	CXGB4_USING_SOFT_PARAMS		= (1 << 6),
+	CXGB4_MASTER_PF			= (1 << 7),
+	CXGB4_FW_OFLD_CONN		= (1 << 9),
+	CXGB4_ROOT_NO_RELAXED_ORDERING	= (1 << 10),
+	CXGB4_SHUTTING_DOWN		= (1 << 11),
+	CXGB4_SGE_DBQ_TIMER		= (1 << 12),
+};
+
+enum {
+	ULP_CRYPTO_LOOKASIDE = 1 << 0,
+	ULP_CRYPTO_IPSEC_INLINE = 1 << 1,
+	ULP_CRYPTO_KTLS_INLINE  = 1 << 3,
+};
+
+#define CXGB4_MIRROR_RXQ_DEFAULT_DESC_NUM 1024
+#define CXGB4_MIRROR_RXQ_DEFAULT_DESC_SIZE 64
+#define CXGB4_MIRROR_RXQ_DEFAULT_INTR_USEC 5
+#define CXGB4_MIRROR_RXQ_DEFAULT_PKT_CNT 8
+
+#define CXGB4_MIRROR_FLQ_DEFAULT_DESC_NUM 72
+
+struct rx_sw_desc;
+
+struct sge_fl {                     /* SGE free-buffer queue state */
+	unsigned int avail;         /* # of available Rx buffers */
+	unsigned int pend_cred;     /* new buffers since last FL DB ring */
+	unsigned int cidx;          /* consumer index */
+	unsigned int pidx;          /* producer index */
+	unsigned long alloc_failed; /* # of times buffer allocation failed */
+	unsigned long large_alloc_failed;
+	unsigned long mapping_err;  /* # of RX Buffer DMA Mapping failures */
+	unsigned long low;          /* # of times momentarily starving */
+	unsigned long starving;
+	/* RO fields */
+	unsigned int cntxt_id;      /* SGE context id for the free list */
+	unsigned int size;          /* capacity of free list */
+	struct rx_sw_desc *sdesc;   /* address of SW Rx descriptor ring */
+	__be64 *desc;               /* address of HW Rx descriptor ring */
+	dma_addr_t addr;            /* bus address of HW ring start */
+	void __iomem *bar2_addr;    /* address of BAR2 Queue registers */
+	unsigned int bar2_qid;      /* Queue ID for BAR2 Queue registers */
+};
+
+/* A packet gather list */
+struct pkt_gl {
+	u64 sgetstamp;		    /* SGE Time Stamp for Ingress Packet */
+	struct page_frag frags[MAX_SKB_FRAGS];
+	void *va;                         /* virtual address of first byte */
+	unsigned int nfrags;              /* # of fragments */
+	unsigned int tot_len;             /* total length of fragments */
+};
+
+typedef int (*rspq_handler_t)(struct sge_rspq *q, const __be64 *rsp,
+			      const struct pkt_gl *gl);
+typedef void (*rspq_flush_handler_t)(struct sge_rspq *q);
+/* LRO related declarations for ULD */
+struct t4_lro_mgr {
+#define MAX_LRO_SESSIONS		64
+	u8 lro_session_cnt;         /* # of sessions to aggregate */
+	unsigned long lro_pkts;     /* # of LRO super packets */
+	unsigned long lro_merged;   /* # of wire packets merged by LRO */
+	struct sk_buff_head lroq;   /* list of aggregated sessions */
+};
+
+struct sge_rspq {                   /* state for an SGE response queue */
+	struct napi_struct napi;
+	const __be64 *cur_desc;     /* current descriptor in queue */
+	unsigned int cidx;          /* consumer index */
+	u8 gen;                     /* current generation bit */
+	u8 intr_params;             /* interrupt holdoff parameters */
+	u8 next_intr_params;        /* holdoff params for next interrupt */
+	u8 adaptive_rx;
+	u8 pktcnt_idx;              /* interrupt packet threshold */
+	u8 uld;                     /* ULD handling this queue */
+	u8 idx;                     /* queue index within its group */
+	int offset;                 /* offset into current Rx buffer */
+	u16 cntxt_id;               /* SGE context id for the response q */
+	u16 abs_id;                 /* absolute SGE id for the response q */
+	__be64 *desc;               /* address of HW response ring */
+	dma_addr_t phys_addr;       /* physical address of the ring */
+	void __iomem *bar2_addr;    /* address of BAR2 Queue registers */
+	unsigned int bar2_qid;      /* Queue ID for BAR2 Queue registers */
+	unsigned int iqe_len;       /* entry size */
+	unsigned int size;          /* capacity of response queue */
+	struct adapter *adap;
+	struct net_device *netdev;  /* associated net device */
+	rspq_handler_t handler;
+	rspq_flush_handler_t flush_handler;
+	struct t4_lro_mgr lro_mgr;
+};
+
+struct sge_eth_stats {              /* Ethernet queue statistics */
+	unsigned long pkts;         /* # of ethernet packets */
+	unsigned long lro_pkts;     /* # of LRO super packets */
+	unsigned long lro_merged;   /* # of wire packets merged by LRO */
+	unsigned long rx_cso;       /* # of Rx checksum offloads */
+	unsigned long vlan_ex;      /* # of Rx VLAN extractions */
+	unsigned long rx_drops;     /* # of packets dropped due to no mem */
+	unsigned long bad_rx_pkts;  /* # of packets with err_vec!=0 */
+};
+
+struct sge_eth_rxq {                /* SW Ethernet Rx queue */
+	struct sge_rspq rspq;
+	struct sge_fl fl;
+	struct sge_eth_stats stats;
+	struct msix_info *msix;
+} ____cacheline_aligned_in_smp;
+
+struct sge_ofld_stats {             /* offload queue statistics */
+	unsigned long pkts;         /* # of packets */
+	unsigned long imm;          /* # of immediate-data packets */
+	unsigned long an;           /* # of asynchronous notifications */
+	unsigned long nomem;        /* # of responses deferred due to no mem */
+};
+
+struct sge_ofld_rxq {               /* SW offload Rx queue */
+	struct sge_rspq rspq;
+	struct sge_fl fl;
+	struct sge_ofld_stats stats;
+	struct msix_info *msix;
+} ____cacheline_aligned_in_smp;
+
+struct tx_desc {
+	__be64 flit[8];
+};
+
+struct ulptx_sgl;
+
+struct tx_sw_desc {
+	struct sk_buff *skb; /* SKB to free after getting completion */
+	dma_addr_t addr[MAX_SKB_FRAGS + 1]; /* DMA mapped addresses */
+};
+
+struct sge_txq {
+	unsigned int  in_use;       /* # of in-use Tx descriptors */
+	unsigned int  q_type;	    /* Q type Eth/Ctrl/Ofld */
+	unsigned int  size;         /* # of descriptors */
+	unsigned int  cidx;         /* SW consumer index */
+	unsigned int  pidx;         /* producer index */
+	unsigned long stops;        /* # of times q has been stopped */
+	unsigned long restarts;     /* # of queue restarts */
+	unsigned int  cntxt_id;     /* SGE context id for the Tx q */
+	struct tx_desc *desc;       /* address of HW Tx descriptor ring */
+	struct tx_sw_desc *sdesc;   /* address of SW Tx descriptor ring */
+	struct sge_qstat *stat;     /* queue status entry */
+	dma_addr_t    phys_addr;    /* physical address of the ring */
+	spinlock_t db_lock;
+	int db_disabled;
+	unsigned short db_pidx;
+	unsigned short db_pidx_inc;
+	void __iomem *bar2_addr;    /* address of BAR2 Queue registers */
+	unsigned int bar2_qid;      /* Queue ID for BAR2 Queue registers */
+};
+
+struct sge_eth_txq {                /* state for an SGE Ethernet Tx queue */
+	struct sge_txq q;
+	struct netdev_queue *txq;   /* associated netdev TX queue */
+#ifdef CONFIG_CHELSIO_T4_DCB
+	u8 dcb_prio;		    /* DCB Priority bound to queue */
+#endif
+	u8 dbqt;                    /* SGE Doorbell Queue Timer in use */
+	unsigned int dbqtimerix;    /* SGE Doorbell Queue Timer Index */
+	unsigned long tso;          /* # of TSO requests */
+	unsigned long uso;          /* # of USO requests */
+	unsigned long tx_cso;       /* # of Tx checksum offloads */
+	unsigned long vlan_ins;     /* # of Tx VLAN insertions */
+	unsigned long mapping_err;  /* # of I/O MMU packet mapping errors */
+} ____cacheline_aligned_in_smp;
+
+struct sge_uld_txq {               /* state for an SGE offload Tx queue */
+	struct sge_txq q;
+	struct adapter *adap;
+	struct sk_buff_head sendq;  /* list of backpressured packets */
+	struct tasklet_struct qresume_tsk; /* restarts the queue */
+	bool service_ofldq_running; /* service_ofldq() is processing sendq */
+	u8 full;                    /* the Tx ring is full */
+	unsigned long mapping_err;  /* # of I/O MMU packet mapping errors */
+} ____cacheline_aligned_in_smp;
+
+struct sge_ctrl_txq {               /* state for an SGE control Tx queue */
+	struct sge_txq q;
+	struct adapter *adap;
+	struct sk_buff_head sendq;  /* list of backpressured packets */
+	struct tasklet_struct qresume_tsk; /* restarts the queue */
+	u8 full;                    /* the Tx ring is full */
+} ____cacheline_aligned_in_smp;
+
+struct sge_uld_rxq_info {
+	char name[IFNAMSIZ];	/* name of ULD driver */
+	struct sge_ofld_rxq *uldrxq; /* Rxq's for ULD */
+	u16 *rspq_id;		/* response queue id's of rxq */
+	u16 nrxq;		/* # of ingress uld queues */
+	u16 nciq;		/* # of completion queues */
+	u8 uld;			/* uld type */
+};
+
+struct sge_uld_txq_info {
+	struct sge_uld_txq *uldtxq; /* Txq's for ULD */
+	atomic_t users;		/* num users */
+	u16 ntxq;		/* # of egress uld queues */
+};
+
+/* struct to maintain ULD list to reallocate ULD resources on hotplug */
+struct cxgb4_uld_list {
+	struct cxgb4_uld_info uld_info;
+	struct list_head list_node;
+	enum cxgb4_uld uld_type;
+};
+
+enum sge_eosw_state {
+	CXGB4_EO_STATE_CLOSED = 0, /* Not ready to accept traffic */
+	CXGB4_EO_STATE_FLOWC_OPEN_SEND, /* Send FLOWC open request */
+	CXGB4_EO_STATE_FLOWC_OPEN_REPLY, /* Waiting for FLOWC open reply */
+	CXGB4_EO_STATE_ACTIVE, /* Ready to accept traffic */
+	CXGB4_EO_STATE_FLOWC_CLOSE_SEND, /* Send FLOWC close request */
+	CXGB4_EO_STATE_FLOWC_CLOSE_REPLY, /* Waiting for FLOWC close reply */
+};
+
+struct sge_eosw_txq {
+	spinlock_t lock; /* Per queue lock to synchronize completions */
+	enum sge_eosw_state state; /* Current ETHOFLD State */
+	struct tx_sw_desc *desc; /* Descriptor ring to hold packets */
+	u32 ndesc; /* Number of descriptors */
+	u32 pidx; /* Current Producer Index */
+	u32 last_pidx; /* Last successfully transmitted Producer Index */
+	u32 cidx; /* Current Consumer Index */
+	u32 last_cidx; /* Last successfully reclaimed Consumer Index */
+	u32 flowc_idx; /* Descriptor containing a FLOWC request */
+	u32 inuse; /* Number of packets held in ring */
+
+	u32 cred; /* Current available credits */
+	u32 ncompl; /* # of completions posted */
+	u32 last_compl; /* # of credits consumed since last completion req */
+
+	u32 eotid; /* Index into EOTID table in software */
+	u32 hwtid; /* Hardware EOTID index */
+
+	u32 hwqid; /* Underlying hardware queue index */
+	struct net_device *netdev; /* Pointer to netdevice */
+	struct tasklet_struct qresume_tsk; /* Restarts the queue */
+	struct completion completion; /* completion for FLOWC rendezvous */
+};
+
+struct sge_eohw_txq {
+	spinlock_t lock; /* Per queue lock */
+	struct sge_txq q; /* HW Txq */
+	struct adapter *adap; /* Backpointer to adapter */
+	unsigned long tso; /* # of TSO requests */
+	unsigned long uso; /* # of USO requests */
+	unsigned long tx_cso; /* # of Tx checksum offloads */
+	unsigned long vlan_ins; /* # of Tx VLAN insertions */
+	unsigned long mapping_err; /* # of I/O MMU packet mapping errors */
+};
+
+struct sge {
+	struct sge_eth_txq ethtxq[MAX_ETH_QSETS];
+	struct sge_eth_txq ptptxq;
+	struct sge_ctrl_txq ctrlq[MAX_CTRL_QUEUES];
+
+	struct sge_eth_rxq ethrxq[MAX_ETH_QSETS];
+	struct sge_rspq fw_evtq ____cacheline_aligned_in_smp;
+	struct sge_uld_rxq_info **uld_rxq_info;
+	struct sge_uld_txq_info **uld_txq_info;
+
+	struct sge_rspq intrq ____cacheline_aligned_in_smp;
+	spinlock_t intrq_lock;
+
+	struct sge_eohw_txq *eohw_txq;
+	struct sge_ofld_rxq *eohw_rxq;
+
+	struct sge_eth_rxq *mirror_rxq[NCHAN];
+
+	u16 max_ethqsets;           /* # of available Ethernet queue sets */
+	u16 ethqsets;               /* # of active Ethernet queue sets */
+	u16 ethtxq_rover;           /* Tx queue to clean up next */
+	u16 ofldqsets;              /* # of active ofld queue sets */
+	u16 nqs_per_uld;	    /* # of Rx queues per ULD */
+	u16 eoqsets;                /* # of ETHOFLD queues */
+	u16 mirrorqsets;            /* # of Mirror queues */
+
+	u16 timer_val[SGE_NTIMERS];
+	u8 counter_val[SGE_NCOUNTERS];
+	u16 dbqtimer_tick;
+	u16 dbqtimer_val[SGE_NDBQTIMERS];
+	u32 fl_pg_order;            /* large page allocation size */
+	u32 stat_len;               /* length of status page at ring end */
+	u32 pktshift;               /* padding between CPL & packet data */
+	u32 fl_align;               /* response queue message alignment */
+	u32 fl_starve_thres;        /* Free List starvation threshold */
+
+	struct sge_idma_monitor_state idma_monitor;
+	unsigned int egr_start;
+	unsigned int egr_sz;
+	unsigned int ingr_start;
+	unsigned int ingr_sz;
+	void **egr_map;    /* qid->queue egress queue map */
+	struct sge_rspq **ingr_map; /* qid->queue ingress queue map */
+	unsigned long *starving_fl;
+	unsigned long *txq_maperr;
+	unsigned long *blocked_fl;
+	struct timer_list rx_timer; /* refills starving FLs */
+	struct timer_list tx_timer; /* checks Tx queues */
+
+	int fwevtq_msix_idx; /* Index to firmware event queue MSI-X info */
+	int nd_msix_idx; /* Index to non-data interrupts MSI-X info */
+};
+
+#define for_each_ethrxq(sge, i) for (i = 0; i < (sge)->ethqsets; i++)
+#define for_each_ofldtxq(sge, i) for (i = 0; i < (sge)->ofldqsets; i++)
+
+struct l2t_data;
+
+#ifdef CONFIG_PCI_IOV
+
+/* T4 supports SRIOV on PF0-3 and T5 on PF0-7.  However, the Serial
+ * Configuration initialization for T5 only has SR-IOV functionality enabled
+ * on PF0-3 in order to simplify everything.
+ */
+#define NUM_OF_PF_WITH_SRIOV 4
+
+#endif
+
+struct doorbell_stats {
+	u32 db_drop;
+	u32 db_empty;
+	u32 db_full;
+};
+
+struct hash_mac_addr {
+	struct list_head list;
+	u8 addr[ETH_ALEN];
+	unsigned int iface_mac;
+};
+
+struct msix_bmap {
+	unsigned long *msix_bmap;
+	unsigned int mapsize;
+	spinlock_t lock; /* lock for acquiring bitmap */
+};
+
+struct msix_info {
+	unsigned short vec;
+	char desc[IFNAMSIZ + 10];
+	unsigned int idx;
+	cpumask_var_t aff_mask;
+};
+
+struct vf_info {
+	unsigned char vf_mac_addr[ETH_ALEN];
+	unsigned int tx_rate;
+	bool pf_set_mac;
+	u16 vlan;
+	int link_state;
+};
+
+enum {
+	HMA_DMA_MAPPED_FLAG = 1
+};
+
+struct hma_data {
+	unsigned char flags;
+	struct sg_table *sgt;
+	dma_addr_t *phy_addr;	/* physical address of the page */
+};
+
+struct mbox_list {
+	struct list_head list;
+};
+
+#if IS_ENABLED(CONFIG_THERMAL)
+struct ch_thermal {
+	struct thermal_zone_device *tzdev;
+	int trip_temp;
+	int trip_type;
+};
+#endif
+
+struct mps_entries_ref {
+	struct list_head list;
+	u8 addr[ETH_ALEN];
+	u8 mask[ETH_ALEN];
+	u16 idx;
+	refcount_t refcnt;
+};
+
+struct cxgb4_ethtool_filter_info {
+	u32 *loc_array; /* Array holding the actual TIDs set to filters */
+	unsigned long *bmap; /* Bitmap for managing filters in use */
+	u32 in_use; /* # of filters in use */
+};
+
+struct cxgb4_ethtool_filter {
+	u32 nentries; /* Adapter wide number of supported filters */
+	struct cxgb4_ethtool_filter_info *port; /* Per port entry */
+};
+
+struct adapter {
+	void __iomem *regs;
+	void __iomem *bar2;
+	u32 t4_bar0;
+	struct pci_dev *pdev;
+	struct device *pdev_dev;
+	const char *name;
+	unsigned int mbox;
+	unsigned int pf;
+	unsigned int flags;
+	unsigned int adap_idx;
+	enum chip_type chip;
+	u32 eth_flags;
+
+	int msg_enable;
+	__be16 vxlan_port;
+	__be16 geneve_port;
+
+	struct adapter_params params;
+	struct cxgb4_virt_res vres;
+	unsigned int swintr;
+
+	/* MSI-X Info for NIC and OFLD queues */
+	struct msix_info *msix_info;
+	struct msix_bmap msix_bmap;
+
+	struct doorbell_stats db_stats;
+	struct sge sge;
+
+	struct net_device *port[MAX_NPORTS];
+	u8 chan_map[NCHAN];                   /* channel -> port map */
+
+	struct vf_info *vfinfo;
+	u8 num_vfs;
+
+	u32 filter_mode;
+	unsigned int l2t_start;
+	unsigned int l2t_end;
+	struct l2t_data *l2t;
+	unsigned int clipt_start;
+	unsigned int clipt_end;
+	struct clip_tbl *clipt;
+	unsigned int rawf_start;
+	unsigned int rawf_cnt;
+	struct smt_data *smt;
+	struct cxgb4_uld_info *uld;
+	void *uld_handle[CXGB4_ULD_MAX];
+	unsigned int num_uld;
+	unsigned int num_ofld_uld;
+	struct list_head list_node;
+	struct list_head rcu_node;
+	struct list_head mac_hlist; /* list of MAC addresses in MPS Hash */
+	struct list_head mps_ref;
+	spinlock_t mps_ref_lock; /* lock for syncing mps ref/def activities */
+
+	void *iscsi_ppm;
+
+	struct tid_info tids;
+	void **tid_release_head;
+	spinlock_t tid_release_lock;
+	struct workqueue_struct *workq;
+	struct work_struct tid_release_task;
+	struct work_struct db_full_task;
+	struct work_struct db_drop_task;
+	struct work_struct fatal_err_notify_task;
+	bool tid_release_task_busy;
+
+	/* lock for mailbox cmd list */
+	spinlock_t mbox_lock;
+	struct mbox_list mlist;
+
+	/* support for mailbox command/reply logging */
+#define T4_OS_LOG_MBOX_CMDS 256
+	struct mbox_cmd_log *mbox_log;
+
+	struct mutex uld_mutex;
+
+	struct dentry *debugfs_root;
+	bool use_bd;     /* Use SGE Back Door intfc for reading SGE Contexts */
+	bool trace_rss;	/* 1 implies that different RSS flit per filter is
+			 * used per filter else if 0 default RSS flit is
+			 * used for all 4 filters.
+			 */
+
+	struct ptp_clock *ptp_clock;
+	struct ptp_clock_info ptp_clock_info;
+	struct sk_buff *ptp_tx_skb;
+	/* ptp lock */
+	spinlock_t ptp_lock;
+	spinlock_t stats_lock;
+	spinlock_t win0_lock ____cacheline_aligned_in_smp;
+
+	/* TC u32 offload */
+	struct cxgb4_tc_u32_table *tc_u32;
+	struct chcr_ktls chcr_ktls;
+	struct chcr_stats_debug chcr_stats;
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+	struct ch_ktls_stats_debug ch_ktls_stats;
+#endif
+#if IS_ENABLED(CONFIG_CHELSIO_IPSEC_INLINE)
+	struct ch_ipsec_stats_debug ch_ipsec_stats;
+#endif
+
+	/* TC flower offload */
+	bool tc_flower_initialized;
+	struct rhashtable flower_tbl;
+	struct rhashtable_params flower_ht_params;
+	struct timer_list flower_stats_timer;
+	struct work_struct flower_stats_work;
+
+	/* Ethtool Dump */
+	struct ethtool_dump eth_dump;
+
+	/* HMA */
+	struct hma_data hma;
+
+	struct srq_data *srq;
+
+	/* Dump buffer for collecting logs in kdump kernel */
+	struct vmcoredd_data vmcoredd;
+#if IS_ENABLED(CONFIG_THERMAL)
+	struct ch_thermal ch_thermal;
+#endif
+
+	/* TC MQPRIO offload */
+	struct cxgb4_tc_mqprio *tc_mqprio;
+
+	/* TC MATCHALL classifier offload */
+	struct cxgb4_tc_matchall *tc_matchall;
+
+	/* Ethtool n-tuple */
+	struct cxgb4_ethtool_filter *ethtool_filters;
+};
+
+/* Support for "sched-class" command to allow a TX Scheduling Class to be
+ * programmed with various parameters.
+ */
+struct ch_sched_params {
+	u8   type;                     /* packet or flow */
+	union {
+		struct {
+			u8   level;    /* scheduler hierarchy level */
+			u8   mode;     /* per-class or per-flow */
+			u8   rateunit; /* bit or packet rate */
+			u8   ratemode; /* %port relative or kbps absolute */
+			u8   channel;  /* scheduler channel [0..N] */
+			u8   class;    /* scheduler class [0..N] */
+			u32  minrate;  /* minimum rate */
+			u32  maxrate;  /* maximum rate */
+			u16  weight;   /* percent weight */
+			u16  pktsize;  /* average packet size */
+			u16  burstsize;  /* burst buffer size */
+		} params;
+	} u;
+};
+
+enum {
+	SCHED_CLASS_TYPE_PACKET = 0,    /* class type */
+};
+
+enum {
+	SCHED_CLASS_LEVEL_CL_RL = 0,    /* class rate limiter */
+	SCHED_CLASS_LEVEL_CH_RL = 2,    /* channel rate limiter */
+};
+
+enum {
+	SCHED_CLASS_MODE_CLASS = 0,     /* per-class scheduling */
+	SCHED_CLASS_MODE_FLOW,          /* per-flow scheduling */
+};
+
+enum {
+	SCHED_CLASS_RATEUNIT_BITS = 0,  /* bit rate scheduling */
+};
+
+enum {
+	SCHED_CLASS_RATEMODE_ABS = 1,   /* Kb/s */
+};
+
+/* Support for "sched_queue" command to allow one or more NIC TX Queues
+ * to be bound to a TX Scheduling Class.
+ */
+struct ch_sched_queue {
+	s8   queue;    /* queue index */
+	s8   class;    /* class index */
+};
+
+/* Support for "sched_flowc" command to allow one or more FLOWC
+ * to be bound to a TX Scheduling Class.
+ */
+struct ch_sched_flowc {
+	s32 tid;   /* TID to bind */
+	s8  class; /* class index */
+};
+
+/* Defined bit width of user definable filter tuples
+ */
+#define ETHTYPE_BITWIDTH 16
+#define FRAG_BITWIDTH 1
+#define MACIDX_BITWIDTH 9
+#define FCOE_BITWIDTH 1
+#define IPORT_BITWIDTH 3
+#define MATCHTYPE_BITWIDTH 3
+#define PROTO_BITWIDTH 8
+#define TOS_BITWIDTH 8
+#define PF_BITWIDTH 8
+#define VF_BITWIDTH 8
+#define IVLAN_BITWIDTH 16
+#define OVLAN_BITWIDTH 16
+#define ENCAP_VNI_BITWIDTH 24
+
+/* Filter matching rules.  These consist of a set of ingress packet field
+ * (value, mask) tuples.  The associated ingress packet field matches the
+ * tuple when ((field & mask) == value).  (Thus a wildcard "don't care" field
+ * rule can be constructed by specifying a tuple of (0, 0).)  A filter rule
+ * matches an ingress packet when all of the individual individual field
+ * matching rules are true.
+ *
+ * Partial field masks are always valid, however, while it may be easy to
+ * understand their meanings for some fields (e.g. IP address to match a
+ * subnet), for others making sensible partial masks is less intuitive (e.g.
+ * MPS match type) ...
+ *
+ * Most of the following data structures are modeled on T4 capabilities.
+ * Drivers for earlier chips use the subsets which make sense for those chips.
+ * We really need to come up with a hardware-independent mechanism to
+ * represent hardware filter capabilities ...
+ */
+struct ch_filter_tuple {
+	/* Compressed header matching field rules.  The TP_VLAN_PRI_MAP
+	 * register selects which of these fields will participate in the
+	 * filter match rules -- up to a maximum of 36 bits.  Because
+	 * TP_VLAN_PRI_MAP is a global register, all filters must use the same
+	 * set of fields.
+	 */
+	uint32_t ethtype:ETHTYPE_BITWIDTH;      /* Ethernet type */
+	uint32_t frag:FRAG_BITWIDTH;            /* IP fragmentation header */
+	uint32_t ivlan_vld:1;                   /* inner VLAN valid */
+	uint32_t ovlan_vld:1;                   /* outer VLAN valid */
+	uint32_t pfvf_vld:1;                    /* PF/VF valid */
+	uint32_t encap_vld:1;			/* Encapsulation valid */
+	uint32_t macidx:MACIDX_BITWIDTH;        /* exact match MAC index */
+	uint32_t fcoe:FCOE_BITWIDTH;            /* FCoE packet */
+	uint32_t iport:IPORT_BITWIDTH;          /* ingress port */
+	uint32_t matchtype:MATCHTYPE_BITWIDTH;  /* MPS match type */
+	uint32_t proto:PROTO_BITWIDTH;          /* protocol type */
+	uint32_t tos:TOS_BITWIDTH;              /* TOS/Traffic Type */
+	uint32_t pf:PF_BITWIDTH;                /* PCI-E PF ID */
+	uint32_t vf:VF_BITWIDTH;                /* PCI-E VF ID */
+	uint32_t ivlan:IVLAN_BITWIDTH;          /* inner VLAN */
+	uint32_t ovlan:OVLAN_BITWIDTH;          /* outer VLAN */
+	uint32_t vni:ENCAP_VNI_BITWIDTH;	/* VNI of tunnel */
+
+	/* Uncompressed header matching field rules.  These are always
+	 * available for field rules.
+	 */
+	uint8_t lip[16];        /* local IP address (IPv4 in [3:0]) */
+	uint8_t fip[16];        /* foreign IP address (IPv4 in [3:0]) */
+	uint16_t lport;         /* local port */
+	uint16_t fport;         /* foreign port */
+};
+
+/* A filter ioctl command.
+ */
+struct ch_filter_specification {
+	/* Administrative fields for filter.
+	 */
+	uint32_t hitcnts:1;     /* count filter hits in TCB */
+	uint32_t prio:1;        /* filter has priority over active/server */
+
+	/* Fundamental filter typing.  This is the one element of filter
+	 * matching that doesn't exist as a (value, mask) tuple.
+	 */
+	uint32_t type:1;        /* 0 => IPv4, 1 => IPv6 */
+	u32 hash:1;		/* 0 => wild-card, 1 => exact-match */
+
+	/* Packet dispatch information.  Ingress packets which match the
+	 * filter rules will be dropped, passed to the host or switched back
+	 * out as egress packets.
+	 */
+	uint32_t action:2;      /* drop, pass, switch */
+
+	uint32_t rpttid:1;      /* report TID in RSS hash field */
+
+	uint32_t dirsteer:1;    /* 0 => RSS, 1 => steer to iq */
+	uint32_t iq:10;         /* ingress queue */
+
+	uint32_t maskhash:1;    /* dirsteer=0: store RSS hash in TCB */
+	uint32_t dirsteerhash:1;/* dirsteer=1: 0 => TCB contains RSS hash */
+				/*             1 => TCB contains IQ ID */
+
+	/* Switch proxy/rewrite fields.  An ingress packet which matches a
+	 * filter with "switch" set will be looped back out as an egress
+	 * packet -- potentially with some Ethernet header rewriting.
+	 */
+	uint32_t eport:2;       /* egress port to switch packet out */
+	uint32_t newdmac:1;     /* rewrite destination MAC address */
+	uint32_t newsmac:1;     /* rewrite source MAC address */
+	uint32_t newvlan:2;     /* rewrite VLAN Tag */
+	uint32_t nat_mode:3;    /* specify NAT operation mode */
+	uint8_t dmac[ETH_ALEN]; /* new destination MAC address */
+	uint8_t smac[ETH_ALEN]; /* new source MAC address */
+	uint16_t vlan;          /* VLAN Tag to insert */
+
+	u8 nat_lip[16];		/* local IP to use after NAT'ing */
+	u8 nat_fip[16];		/* foreign IP to use after NAT'ing */
+	u16 nat_lport;		/* local port to use after NAT'ing */
+	u16 nat_fport;		/* foreign port to use after NAT'ing */
+
+	u32 tc_prio;		/* TC's filter priority index */
+	u64 tc_cookie;		/* Unique cookie identifying TC rules */
+
+	/* reservation for future additions */
+	u8 rsvd[12];
+
+	/* Filter rule value/mask pairs.
+	 */
+	struct ch_filter_tuple val;
+	struct ch_filter_tuple mask;
+};
+
+enum {
+	FILTER_PASS = 0,        /* default */
+	FILTER_DROP,
+	FILTER_SWITCH
+};
+
+enum {
+	VLAN_NOCHANGE = 0,      /* default */
+	VLAN_REMOVE,
+	VLAN_INSERT,
+	VLAN_REWRITE
+};
+
+enum {
+	NAT_MODE_NONE = 0,	/* No NAT performed */
+	NAT_MODE_DIP,		/* NAT on Dst IP */
+	NAT_MODE_DIP_DP,	/* NAT on Dst IP, Dst Port */
+	NAT_MODE_DIP_DP_SIP,	/* NAT on Dst IP, Dst Port and Src IP */
+	NAT_MODE_DIP_DP_SP,	/* NAT on Dst IP, Dst Port and Src Port */
+	NAT_MODE_SIP_SP,	/* NAT on Src IP and Src Port */
+	NAT_MODE_DIP_SIP_SP,	/* NAT on Dst IP, Src IP and Src Port */
+	NAT_MODE_ALL		/* NAT on entire 4-tuple */
+};
+
+#define CXGB4_FILTER_TYPE_MAX 2
+
+/* Host shadow copy of ingress filter entry.  This is in host native format
+ * and doesn't match the ordering or bit order, etc. of the hardware of the
+ * firmware command.  The use of bit-field structure elements is purely to
+ * remind ourselves of the field size limitations and save memory in the case
+ * where the filter table is large.
+ */
+struct filter_entry {
+	/* Administrative fields for filter. */
+	u32 valid:1;            /* filter allocated and valid */
+	u32 locked:1;           /* filter is administratively locked */
+
+	u32 pending:1;          /* filter action is pending firmware reply */
+	struct filter_ctx *ctx; /* Caller's completion hook */
+	struct l2t_entry *l2t;  /* Layer Two Table entry for dmac */
+	struct smt_entry *smt;  /* Source Mac Table entry for smac */
+	struct net_device *dev; /* Associated net device */
+	u32 tid;                /* This will store the actual tid */
+
+	/* The filter itself.  Most of this is a straight copy of information
+	 * provided by the extended ioctl().  Some fields are translated to
+	 * internal forms -- for instance the Ingress Queue ID passed in from
+	 * the ioctl() is translated into the Absolute Ingress Queue ID.
+	 */
+	struct ch_filter_specification fs;
+};
+
+static inline int is_offload(const struct adapter *adap)
+{
+	return adap->params.offload;
+}
+
+static inline int is_hashfilter(const struct adapter *adap)
+{
+	return adap->params.hash_filter;
+}
+
+static inline int is_pci_uld(const struct adapter *adap)
+{
+	return adap->params.crypto;
+}
+
+static inline int is_uld(const struct adapter *adap)
+{
+	return (adap->params.offload || adap->params.crypto);
+}
+
+static inline int is_ethofld(const struct adapter *adap)
+{
+	return adap->params.ethofld;
+}
+
+static inline u32 t4_read_reg(struct adapter *adap, u32 reg_addr)
+{
+	return readl(adap->regs + reg_addr);
+}
+
+static inline void t4_write_reg(struct adapter *adap, u32 reg_addr, u32 val)
+{
+	writel(val, adap->regs + reg_addr);
+}
+
+#ifndef readq
+static inline u64 readq(const volatile void __iomem *addr)
+{
+	return readl(addr) + ((u64)readl(addr + 4) << 32);
+}
+
+static inline void writeq(u64 val, volatile void __iomem *addr)
+{
+	writel(val, addr);
+	writel(val >> 32, addr + 4);
+}
+#endif
+
+static inline u64 t4_read_reg64(struct adapter *adap, u32 reg_addr)
+{
+	return readq(adap->regs + reg_addr);
+}
+
+static inline void t4_write_reg64(struct adapter *adap, u32 reg_addr, u64 val)
+{
+	writeq(val, adap->regs + reg_addr);
+}
+
+/**
+ * t4_set_hw_addr - store a port's MAC address in SW
+ * @adapter: the adapter
+ * @port_idx: the port index
+ * @hw_addr: the Ethernet address
+ *
+ * Store the Ethernet address of the given port in SW.  Called by the common
+ * code when it retrieves a port's Ethernet address from EEPROM.
+ */
+static inline void t4_set_hw_addr(struct adapter *adapter, int port_idx,
+				  u8 hw_addr[])
+{
+	eth_hw_addr_set(adapter->port[port_idx], hw_addr);
+	ether_addr_copy(adapter->port[port_idx]->perm_addr, hw_addr);
+}
+
+/**
+ * netdev2pinfo - return the port_info structure associated with a net_device
+ * @dev: the netdev
+ *
+ * Return the struct port_info associated with a net_device
+ */
+static inline struct port_info *netdev2pinfo(const struct net_device *dev)
+{
+	return netdev_priv(dev);
+}
+
+/**
+ * adap2pinfo - return the port_info of a port
+ * @adap: the adapter
+ * @idx: the port index
+ *
+ * Return the port_info structure for the port of the given index.
+ */
+static inline struct port_info *adap2pinfo(struct adapter *adap, int idx)
+{
+	return netdev_priv(adap->port[idx]);
+}
+
+/**
+ * netdev2adap - return the adapter structure associated with a net_device
+ * @dev: the netdev
+ *
+ * Return the struct adapter associated with a net_device
+ */
+static inline struct adapter *netdev2adap(const struct net_device *dev)
+{
+	return netdev2pinfo(dev)->adapter;
+}
+
+/* Return a version number to identify the type of adapter.  The scheme is:
+ * - bits 0..9: chip version
+ * - bits 10..15: chip revision
+ * - bits 16..23: register dump version
+ */
+static inline unsigned int mk_adap_vers(struct adapter *ap)
+{
+	return CHELSIO_CHIP_VERSION(ap->params.chip) |
+		(CHELSIO_CHIP_RELEASE(ap->params.chip) << 10) | (1 << 16);
+}
+
+/* Return a queue's interrupt hold-off time in us.  0 means no timer. */
+static inline unsigned int qtimer_val(const struct adapter *adap,
+				      const struct sge_rspq *q)
+{
+	unsigned int idx = q->intr_params >> 1;
+
+	return idx < SGE_NTIMERS ? adap->sge.timer_val[idx] : 0;
+}
+
+/* driver name used for ethtool_drvinfo */
+extern char cxgb4_driver_name[];
+
+void t4_os_portmod_changed(struct adapter *adap, int port_id);
+void t4_os_link_changed(struct adapter *adap, int port_id, int link_stat);
+
+void t4_free_sge_resources(struct adapter *adap);
+void t4_free_ofld_rxqs(struct adapter *adap, int n, struct sge_ofld_rxq *q);
+irq_handler_t t4_intr_handler(struct adapter *adap);
+netdev_tx_t t4_start_xmit(struct sk_buff *skb, struct net_device *dev);
+int cxgb4_selftest_lb_pkt(struct net_device *netdev);
+int t4_ethrx_handler(struct sge_rspq *q, const __be64 *rsp,
+		     const struct pkt_gl *gl);
+int t4_mgmt_tx(struct adapter *adap, struct sk_buff *skb);
+int t4_ofld_send(struct adapter *adap, struct sk_buff *skb);
+int t4_sge_alloc_rxq(struct adapter *adap, struct sge_rspq *iq, bool fwevtq,
+		     struct net_device *dev, int intr_idx,
+		     struct sge_fl *fl, rspq_handler_t hnd,
+		     rspq_flush_handler_t flush_handler, int cong);
+int t4_sge_alloc_eth_txq(struct adapter *adap, struct sge_eth_txq *txq,
+			 struct net_device *dev, struct netdev_queue *netdevq,
+			 unsigned int iqid, u8 dbqt);
+int t4_sge_alloc_ctrl_txq(struct adapter *adap, struct sge_ctrl_txq *txq,
+			  struct net_device *dev, unsigned int iqid,
+			  unsigned int cmplqid);
+int t4_sge_mod_ctrl_txq(struct adapter *adap, unsigned int eqid,
+			unsigned int cmplqid);
+int t4_sge_alloc_uld_txq(struct adapter *adap, struct sge_uld_txq *txq,
+			 struct net_device *dev, unsigned int iqid,
+			 unsigned int uld_type);
+int t4_sge_alloc_ethofld_txq(struct adapter *adap, struct sge_eohw_txq *txq,
+			     struct net_device *dev, u32 iqid);
+void t4_sge_free_ethofld_txq(struct adapter *adap, struct sge_eohw_txq *txq);
+irqreturn_t t4_sge_intr_msix(int irq, void *cookie);
+int t4_sge_init(struct adapter *adap);
+void t4_sge_start(struct adapter *adap);
+void t4_sge_stop(struct adapter *adap);
+int t4_sge_eth_txq_egress_update(struct adapter *adap, struct sge_eth_txq *q,
+				 int maxreclaim);
+void cxgb4_set_ethtool_ops(struct net_device *netdev);
+int cxgb4_write_rss(const struct port_info *pi, const u16 *queues);
+enum cpl_tx_tnl_lso_type cxgb_encap_offload_supported(struct sk_buff *skb);
+extern int dbfifo_int_thresh;
+
+#define for_each_port(adapter, iter) \
+	for (iter = 0; iter < (adapter)->params.nports; ++iter)
+
+static inline int is_bypass(struct adapter *adap)
+{
+	return adap->params.bypass;
+}
+
+static inline int is_bypass_device(int device)
+{
+	/* this should be set based upon device capabilities */
+	switch (device) {
+	case 0x440b:
+	case 0x440c:
+		return 1;
+	default:
+		return 0;
+	}
+}
+
+static inline int is_10gbt_device(int device)
+{
+	/* this should be set based upon device capabilities */
+	switch (device) {
+	case 0x4409:
+	case 0x4486:
+		return 1;
+
+	default:
+		return 0;
+	}
+}
+
+static inline unsigned int core_ticks_per_usec(const struct adapter *adap)
+{
+	return adap->params.vpd.cclk / 1000;
+}
+
+static inline unsigned int us_to_core_ticks(const struct adapter *adap,
+					    unsigned int us)
+{
+	return (us * adap->params.vpd.cclk) / 1000;
+}
+
+static inline unsigned int core_ticks_to_us(const struct adapter *adapter,
+					    unsigned int ticks)
+{
+	/* add Core Clock / 2 to round ticks to nearest uS */
+	return ((ticks * 1000 + adapter->params.vpd.cclk/2) /
+		adapter->params.vpd.cclk);
+}
+
+static inline unsigned int dack_ticks_to_usec(const struct adapter *adap,
+					      unsigned int ticks)
+{
+	return (ticks << adap->params.tp.dack_re) / core_ticks_per_usec(adap);
+}
+
+void t4_set_reg_field(struct adapter *adap, unsigned int addr, u32 mask,
+		      u32 val);
+
+int t4_wr_mbox_meat_timeout(struct adapter *adap, int mbox, const void *cmd,
+			    int size, void *rpl, bool sleep_ok, int timeout);
+int t4_wr_mbox_meat(struct adapter *adap, int mbox, const void *cmd, int size,
+		    void *rpl, bool sleep_ok);
+
+static inline int t4_wr_mbox_timeout(struct adapter *adap, int mbox,
+				     const void *cmd, int size, void *rpl,
+				     int timeout)
+{
+	return t4_wr_mbox_meat_timeout(adap, mbox, cmd, size, rpl, true,
+				       timeout);
+}
+
+static inline int t4_wr_mbox(struct adapter *adap, int mbox, const void *cmd,
+			     int size, void *rpl)
+{
+	return t4_wr_mbox_meat(adap, mbox, cmd, size, rpl, true);
+}
+
+static inline int t4_wr_mbox_ns(struct adapter *adap, int mbox, const void *cmd,
+				int size, void *rpl)
+{
+	return t4_wr_mbox_meat(adap, mbox, cmd, size, rpl, false);
+}
+
+/**
+ *	hash_mac_addr - return the hash value of a MAC address
+ *	@addr: the 48-bit Ethernet MAC address
+ *
+ *	Hashes a MAC address according to the hash function used by HW inexact
+ *	(hash) address matching.
+ */
+static inline int hash_mac_addr(const u8 *addr)
+{
+	u32 a = ((u32)addr[0] << 16) | ((u32)addr[1] << 8) | addr[2];
+	u32 b = ((u32)addr[3] << 16) | ((u32)addr[4] << 8) | addr[5];
+
+	a ^= b;
+	a ^= (a >> 12);
+	a ^= (a >> 6);
+	return a & 0x3f;
+}
+
+int cxgb4_set_rspq_intr_params(struct sge_rspq *q, unsigned int us,
+			       unsigned int cnt);
+static inline void init_rspq(struct adapter *adap, struct sge_rspq *q,
+			     unsigned int us, unsigned int cnt,
+			     unsigned int size, unsigned int iqe_size)
+{
+	q->adap = adap;
+	cxgb4_set_rspq_intr_params(q, us, cnt);
+	q->iqe_len = iqe_size;
+	q->size = size;
+}
+
+/**
+ *     t4_is_inserted_mod_type - is a plugged in Firmware Module Type
+ *     @fw_mod_type: the Firmware Mofule Type
+ *
+ *     Return whether the Firmware Module Type represents a real Transceiver
+ *     Module/Cable Module Type which has been inserted.
+ */
+static inline bool t4_is_inserted_mod_type(unsigned int fw_mod_type)
+{
+	return (fw_mod_type != FW_PORT_MOD_TYPE_NONE &&
+		fw_mod_type != FW_PORT_MOD_TYPE_NOTSUPPORTED &&
+		fw_mod_type != FW_PORT_MOD_TYPE_UNKNOWN &&
+		fw_mod_type != FW_PORT_MOD_TYPE_ERROR);
+}
+
+void t4_write_indirect(struct adapter *adap, unsigned int addr_reg,
+		       unsigned int data_reg, const u32 *vals,
+		       unsigned int nregs, unsigned int start_idx);
+void t4_read_indirect(struct adapter *adap, unsigned int addr_reg,
+		      unsigned int data_reg, u32 *vals, unsigned int nregs,
+		      unsigned int start_idx);
+void t4_hw_pci_read_cfg4(struct adapter *adapter, int reg, u32 *val);
+
+struct fw_filter_wr;
+
+void t4_intr_enable(struct adapter *adapter);
+void t4_intr_disable(struct adapter *adapter);
+int t4_slow_intr_handler(struct adapter *adapter);
+
+int t4_wait_dev_ready(void __iomem *regs);
+
+fw_port_cap32_t t4_link_acaps(struct adapter *adapter, unsigned int port,
+			      struct link_config *lc);
+int t4_link_l1cfg_core(struct adapter *adap, unsigned int mbox,
+		       unsigned int port, struct link_config *lc,
+		       u8 sleep_ok, int timeout);
+
+static inline int t4_link_l1cfg(struct adapter *adapter, unsigned int mbox,
+				unsigned int port, struct link_config *lc)
+{
+	return t4_link_l1cfg_core(adapter, mbox, port, lc,
+				  true, FW_CMD_MAX_TIMEOUT);
+}
+
+static inline int t4_link_l1cfg_ns(struct adapter *adapter, unsigned int mbox,
+				   unsigned int port, struct link_config *lc)
+{
+	return t4_link_l1cfg_core(adapter, mbox, port, lc,
+				  false, FW_CMD_MAX_TIMEOUT);
+}
+
+int t4_restart_aneg(struct adapter *adap, unsigned int mbox, unsigned int port);
+
+u32 t4_read_pcie_cfg4(struct adapter *adap, int reg);
+u32 t4_get_util_window(struct adapter *adap);
+void t4_setup_memwin(struct adapter *adap, u32 memwin_base, u32 window);
+
+int t4_memory_rw_init(struct adapter *adap, int win, int mtype, u32 *mem_off,
+		      u32 *mem_base, u32 *mem_aperture);
+void t4_memory_update_win(struct adapter *adap, int win, u32 addr);
+void t4_memory_rw_residual(struct adapter *adap, u32 off, u32 addr, u8 *buf,
+			   int dir);
+#define T4_MEMORY_WRITE	0
+#define T4_MEMORY_READ	1
+int t4_memory_rw(struct adapter *adap, int win, int mtype, u32 addr, u32 len,
+		 void *buf, int dir);
+static inline int t4_memory_write(struct adapter *adap, int mtype, u32 addr,
+				  u32 len, __be32 *buf)
+{
+	return t4_memory_rw(adap, 0, mtype, addr, len, buf, 0);
+}
+
+unsigned int t4_get_regs_len(struct adapter *adapter);
+void t4_get_regs(struct adapter *adap, void *buf, size_t buf_size);
+
+int t4_eeprom_ptov(unsigned int phys_addr, unsigned int fn, unsigned int sz);
+int t4_seeprom_wp(struct adapter *adapter, bool enable);
+int t4_get_raw_vpd_params(struct adapter *adapter, struct vpd_params *p);
+int t4_get_vpd_params(struct adapter *adapter, struct vpd_params *p);
+int t4_get_pfres(struct adapter *adapter);
+int t4_read_flash(struct adapter *adapter, unsigned int addr,
+		  unsigned int nwords, u32 *data, int byte_oriented);
+int t4_load_fw(struct adapter *adapter, const u8 *fw_data, unsigned int size);
+int t4_load_phy_fw(struct adapter *adap, int win,
+		   int (*phy_fw_version)(const u8 *, size_t),
+		   const u8 *phy_fw_data, size_t phy_fw_size);
+int t4_phy_fw_ver(struct adapter *adap, int *phy_fw_ver);
+int t4_fwcache(struct adapter *adap, enum fw_params_param_dev_fwcache op);
+int t4_fw_upgrade(struct adapter *adap, unsigned int mbox,
+		  const u8 *fw_data, unsigned int size, int force);
+int t4_fl_pkt_align(struct adapter *adap);
+unsigned int t4_flash_cfg_addr(struct adapter *adapter);
+int t4_check_fw_version(struct adapter *adap);
+int t4_load_cfg(struct adapter *adapter, const u8 *cfg_data, unsigned int size);
+int t4_get_fw_version(struct adapter *adapter, u32 *vers);
+int t4_get_bs_version(struct adapter *adapter, u32 *vers);
+int t4_get_tp_version(struct adapter *adapter, u32 *vers);
+int t4_get_exprom_version(struct adapter *adapter, u32 *vers);
+int t4_get_scfg_version(struct adapter *adapter, u32 *vers);
+int t4_get_vpd_version(struct adapter *adapter, u32 *vers);
+int t4_get_version_info(struct adapter *adapter);
+void t4_dump_version_info(struct adapter *adapter);
+int t4_prep_fw(struct adapter *adap, struct fw_info *fw_info,
+	       const u8 *fw_data, unsigned int fw_size,
+	       struct fw_hdr *card_fw, enum dev_state state, int *reset);
+int t4_prep_adapter(struct adapter *adapter);
+int t4_shutdown_adapter(struct adapter *adapter);
+
+enum t4_bar2_qtype { T4_BAR2_QTYPE_EGRESS, T4_BAR2_QTYPE_INGRESS };
+int t4_bar2_sge_qregs(struct adapter *adapter,
+		      unsigned int qid,
+		      enum t4_bar2_qtype qtype,
+		      int user,
+		      u64 *pbar2_qoffset,
+		      unsigned int *pbar2_qid);
+
+unsigned int qtimer_val(const struct adapter *adap,
+			const struct sge_rspq *q);
+
+int t4_init_devlog_params(struct adapter *adapter);
+int t4_init_sge_params(struct adapter *adapter);
+int t4_init_tp_params(struct adapter *adap, bool sleep_ok);
+int t4_filter_field_shift(const struct adapter *adap, int filter_sel);
+int t4_init_rss_mode(struct adapter *adap, int mbox);
+int t4_init_portinfo(struct port_info *pi, int mbox,
+		     int port, int pf, int vf, u8 mac[]);
+int t4_port_init(struct adapter *adap, int mbox, int pf, int vf);
+int t4_init_port_mirror(struct port_info *pi, u8 mbox, u8 port, u8 pf, u8 vf,
+			u16 *mirror_viid);
+void t4_fatal_err(struct adapter *adapter);
+unsigned int t4_chip_rss_size(struct adapter *adapter);
+int t4_config_rss_range(struct adapter *adapter, int mbox, unsigned int viid,
+			int start, int n, const u16 *rspq, unsigned int nrspq);
+int t4_config_glbl_rss(struct adapter *adapter, int mbox, unsigned int mode,
+		       unsigned int flags);
+int t4_config_vi_rss(struct adapter *adapter, int mbox, unsigned int viid,
+		     unsigned int flags, unsigned int defq);
+int t4_read_rss(struct adapter *adapter, u16 *entries);
+void t4_read_rss_key(struct adapter *adapter, u32 *key, bool sleep_ok);
+void t4_write_rss_key(struct adapter *adap, const u32 *key, int idx,
+		      bool sleep_ok);
+void t4_read_rss_pf_config(struct adapter *adapter, unsigned int index,
+			   u32 *valp, bool sleep_ok);
+void t4_read_rss_vf_config(struct adapter *adapter, unsigned int index,
+			   u32 *vfl, u32 *vfh, bool sleep_ok);
+u32 t4_read_rss_pf_map(struct adapter *adapter, bool sleep_ok);
+u32 t4_read_rss_pf_mask(struct adapter *adapter, bool sleep_ok);
+
+unsigned int t4_get_mps_bg_map(struct adapter *adapter, int pidx);
+unsigned int t4_get_tp_ch_map(struct adapter *adapter, int pidx);
+void t4_pmtx_get_stats(struct adapter *adap, u32 cnt[], u64 cycles[]);
+void t4_pmrx_get_stats(struct adapter *adap, u32 cnt[], u64 cycles[]);
+int t4_read_cim_ibq(struct adapter *adap, unsigned int qid, u32 *data,
+		    size_t n);
+int t4_read_cim_obq(struct adapter *adap, unsigned int qid, u32 *data,
+		    size_t n);
+int t4_cim_read(struct adapter *adap, unsigned int addr, unsigned int n,
+		unsigned int *valp);
+int t4_cim_write(struct adapter *adap, unsigned int addr, unsigned int n,
+		 const unsigned int *valp);
+int t4_cim_read_la(struct adapter *adap, u32 *la_buf, unsigned int *wrptr);
+void t4_cim_read_pif_la(struct adapter *adap, u32 *pif_req, u32 *pif_rsp,
+			unsigned int *pif_req_wrptr,
+			unsigned int *pif_rsp_wrptr);
+void t4_cim_read_ma_la(struct adapter *adap, u32 *ma_req, u32 *ma_rsp);
+void t4_read_cimq_cfg(struct adapter *adap, u16 *base, u16 *size, u16 *thres);
+const char *t4_get_port_type_description(enum fw_port_type port_type);
+void t4_get_port_stats(struct adapter *adap, int idx, struct port_stats *p);
+void t4_get_port_stats_offset(struct adapter *adap, int idx,
+			      struct port_stats *stats,
+			      struct port_stats *offset);
+void t4_get_lb_stats(struct adapter *adap, int idx, struct lb_port_stats *p);
+void t4_read_mtu_tbl(struct adapter *adap, u16 *mtus, u8 *mtu_log);
+void t4_read_cong_tbl(struct adapter *adap, u16 incr[NMTUS][NCCTRL_WIN]);
+void t4_tp_wr_bits_indirect(struct adapter *adap, unsigned int addr,
+			    unsigned int mask, unsigned int val);
+void t4_tp_read_la(struct adapter *adap, u64 *la_buf, unsigned int *wrptr);
+void t4_tp_get_err_stats(struct adapter *adap, struct tp_err_stats *st,
+			 bool sleep_ok);
+void t4_tp_get_cpl_stats(struct adapter *adap, struct tp_cpl_stats *st,
+			 bool sleep_ok);
+void t4_tp_get_rdma_stats(struct adapter *adap, struct tp_rdma_stats *st,
+			  bool sleep_ok);
+void t4_get_usm_stats(struct adapter *adap, struct tp_usm_stats *st,
+		      bool sleep_ok);
+void t4_tp_get_tcp_stats(struct adapter *adap, struct tp_tcp_stats *v4,
+			 struct tp_tcp_stats *v6, bool sleep_ok);
+void t4_get_fcoe_stats(struct adapter *adap, unsigned int idx,
+		       struct tp_fcoe_stats *st, bool sleep_ok);
+void t4_load_mtus(struct adapter *adap, const unsigned short *mtus,
+		  const unsigned short *alpha, const unsigned short *beta);
+
+void t4_ulprx_read_la(struct adapter *adap, u32 *la_buf);
+
+void t4_get_chan_txrate(struct adapter *adap, u64 *nic_rate, u64 *ofld_rate);
+void t4_mk_filtdelwr(unsigned int ftid, struct fw_filter_wr *wr, int qid);
+
+void t4_wol_magic_enable(struct adapter *adap, unsigned int port,
+			 const u8 *addr);
+int t4_wol_pat_enable(struct adapter *adap, unsigned int port, unsigned int map,
+		      u64 mask0, u64 mask1, unsigned int crc, bool enable);
+
+int t4_fw_hello(struct adapter *adap, unsigned int mbox, unsigned int evt_mbox,
+		enum dev_master master, enum dev_state *state);
+int t4_fw_bye(struct adapter *adap, unsigned int mbox);
+int t4_early_init(struct adapter *adap, unsigned int mbox);
+int t4_fw_reset(struct adapter *adap, unsigned int mbox, int reset);
+int t4_fixup_host_params(struct adapter *adap, unsigned int page_size,
+			  unsigned int cache_line_size);
+int t4_fw_initialize(struct adapter *adap, unsigned int mbox);
+int t4_query_params(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		    unsigned int vf, unsigned int nparams, const u32 *params,
+		    u32 *val);
+int t4_query_params_ns(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		       unsigned int vf, unsigned int nparams, const u32 *params,
+		       u32 *val);
+int t4_query_params_rw(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		       unsigned int vf, unsigned int nparams, const u32 *params,
+		       u32 *val, int rw, bool sleep_ok);
+int t4_set_params_timeout(struct adapter *adap, unsigned int mbox,
+			  unsigned int pf, unsigned int vf,
+			  unsigned int nparams, const u32 *params,
+			  const u32 *val, int timeout);
+int t4_set_params(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		  unsigned int vf, unsigned int nparams, const u32 *params,
+		  const u32 *val);
+int t4_cfg_pfvf(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		unsigned int vf, unsigned int txq, unsigned int txq_eth_ctrl,
+		unsigned int rxqi, unsigned int rxq, unsigned int tc,
+		unsigned int vi, unsigned int cmask, unsigned int pmask,
+		unsigned int nexact, unsigned int rcaps, unsigned int wxcaps);
+int t4_alloc_vi(struct adapter *adap, unsigned int mbox, unsigned int port,
+		unsigned int pf, unsigned int vf, unsigned int nmac, u8 *mac,
+		unsigned int *rss_size, u8 *vivld, u8 *vin);
+int t4_free_vi(struct adapter *adap, unsigned int mbox,
+	       unsigned int pf, unsigned int vf,
+	       unsigned int viid);
+int t4_set_rxmode(struct adapter *adap, unsigned int mbox, unsigned int viid,
+		  unsigned int viid_mirror, int mtu, int promisc, int all_multi,
+		  int bcast, int vlanex, bool sleep_ok);
+int t4_free_raw_mac_filt(struct adapter *adap, unsigned int viid,
+			 const u8 *addr, const u8 *mask, unsigned int idx,
+			 u8 lookup_type, u8 port_id, bool sleep_ok);
+int t4_free_encap_mac_filt(struct adapter *adap, unsigned int viid, int idx,
+			   bool sleep_ok);
+int t4_alloc_encap_mac_filt(struct adapter *adap, unsigned int viid,
+			    const u8 *addr, const u8 *mask, unsigned int vni,
+			    unsigned int vni_mask, u8 dip_hit, u8 lookup_type,
+			    bool sleep_ok);
+int t4_alloc_raw_mac_filt(struct adapter *adap, unsigned int viid,
+			  const u8 *addr, const u8 *mask, unsigned int idx,
+			  u8 lookup_type, u8 port_id, bool sleep_ok);
+int t4_alloc_mac_filt(struct adapter *adap, unsigned int mbox,
+		      unsigned int viid, bool free, unsigned int naddr,
+		      const u8 **addr, u16 *idx, u64 *hash, bool sleep_ok);
+int t4_free_mac_filt(struct adapter *adap, unsigned int mbox,
+		     unsigned int viid, unsigned int naddr,
+		     const u8 **addr, bool sleep_ok);
+int t4_change_mac(struct adapter *adap, unsigned int mbox, unsigned int viid,
+		  int idx, const u8 *addr, bool persist, u8 *smt_idx);
+int t4_set_addr_hash(struct adapter *adap, unsigned int mbox, unsigned int viid,
+		     bool ucast, u64 vec, bool sleep_ok);
+int t4_enable_vi_params(struct adapter *adap, unsigned int mbox,
+			unsigned int viid, bool rx_en, bool tx_en, bool dcb_en);
+int t4_enable_pi_params(struct adapter *adap, unsigned int mbox,
+			struct port_info *pi,
+			bool rx_en, bool tx_en, bool dcb_en);
+int t4_enable_vi(struct adapter *adap, unsigned int mbox, unsigned int viid,
+		 bool rx_en, bool tx_en);
+int t4_identify_port(struct adapter *adap, unsigned int mbox, unsigned int viid,
+		     unsigned int nblinks);
+int t4_mdio_rd(struct adapter *adap, unsigned int mbox, unsigned int phy_addr,
+	       unsigned int mmd, unsigned int reg, u16 *valp);
+int t4_mdio_wr(struct adapter *adap, unsigned int mbox, unsigned int phy_addr,
+	       unsigned int mmd, unsigned int reg, u16 val);
+int t4_iq_stop(struct adapter *adap, unsigned int mbox, unsigned int pf,
+	       unsigned int vf, unsigned int iqtype, unsigned int iqid,
+	       unsigned int fl0id, unsigned int fl1id);
+int t4_iq_free(struct adapter *adap, unsigned int mbox, unsigned int pf,
+	       unsigned int vf, unsigned int iqtype, unsigned int iqid,
+	       unsigned int fl0id, unsigned int fl1id);
+int t4_eth_eq_free(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		   unsigned int vf, unsigned int eqid);
+int t4_ctrl_eq_free(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		    unsigned int vf, unsigned int eqid);
+int t4_ofld_eq_free(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		    unsigned int vf, unsigned int eqid);
+int t4_sge_ctxt_flush(struct adapter *adap, unsigned int mbox, int ctxt_type);
+int t4_read_sge_dbqtimers(struct adapter *adap, unsigned int ndbqtimers,
+			  u16 *dbqtimers);
+void t4_handle_get_port_info(struct port_info *pi, const __be64 *rpl);
+int t4_update_port_info(struct port_info *pi);
+int t4_get_link_params(struct port_info *pi, unsigned int *link_okp,
+		       unsigned int *speedp, unsigned int *mtup);
+int t4_handle_fw_rpl(struct adapter *adap, const __be64 *rpl);
+void t4_db_full(struct adapter *adapter);
+void t4_db_dropped(struct adapter *adapter);
+int t4_set_trace_filter(struct adapter *adapter, const struct trace_params *tp,
+			int filter_index, int enable);
+void t4_get_trace_filter(struct adapter *adapter, struct trace_params *tp,
+			 int filter_index, int *enabled);
+int t4_fwaddrspace_write(struct adapter *adap, unsigned int mbox,
+			 u32 addr, u32 val);
+void t4_read_pace_tbl(struct adapter *adap, unsigned int pace_vals[NTX_SCHED]);
+void t4_get_tx_sched(struct adapter *adap, unsigned int sched,
+		     unsigned int *kbps, unsigned int *ipg, bool sleep_ok);
+int t4_sge_ctxt_rd(struct adapter *adap, unsigned int mbox, unsigned int cid,
+		   enum ctxt_type ctype, u32 *data);
+int t4_sge_ctxt_rd_bd(struct adapter *adap, unsigned int cid,
+		      enum ctxt_type ctype, u32 *data);
+int t4_sched_params(struct adapter *adapter, u8 type, u8 level, u8 mode,
+		    u8 rateunit, u8 ratemode, u8 channel, u8 class,
+		    u32 minrate, u32 maxrate, u16 weight, u16 pktsize,
+		    u16 burstsize);
+void t4_sge_decode_idma_state(struct adapter *adapter, int state);
+void t4_idma_monitor_init(struct adapter *adapter,
+			  struct sge_idma_monitor_state *idma);
+void t4_idma_monitor(struct adapter *adapter,
+		     struct sge_idma_monitor_state *idma,
+		     int hz, int ticks);
+int t4_set_vf_mac_acl(struct adapter *adapter, unsigned int vf,
+		      unsigned int naddr, u8 *addr);
+void t4_tp_pio_read(struct adapter *adap, u32 *buff, u32 nregs,
+		    u32 start_index, bool sleep_ok);
+void t4_tp_tm_pio_read(struct adapter *adap, u32 *buff, u32 nregs,
+		       u32 start_index, bool sleep_ok);
+void t4_tp_mib_read(struct adapter *adap, u32 *buff, u32 nregs,
+		    u32 start_index, bool sleep_ok);
+
+void t4_uld_mem_free(struct adapter *adap);
+int t4_uld_mem_alloc(struct adapter *adap);
+void t4_uld_clean_up(struct adapter *adap);
+void t4_register_netevent_notifier(void);
+int t4_i2c_rd(struct adapter *adap, unsigned int mbox, int port,
+	      unsigned int devid, unsigned int offset,
+	      unsigned int len, u8 *buf);
+int t4_load_boot(struct adapter *adap, u8 *boot_data,
+		 unsigned int boot_addr, unsigned int size);
+int t4_load_bootcfg(struct adapter *adap,
+		    const u8 *cfg_data, unsigned int size);
+void free_rspq_fl(struct adapter *adap, struct sge_rspq *rq, struct sge_fl *fl);
+void free_tx_desc(struct adapter *adap, struct sge_txq *q,
+		  unsigned int n, bool unmap);
+void cxgb4_eosw_txq_free_desc(struct adapter *adap, struct sge_eosw_txq *txq,
+			      u32 ndesc);
+int cxgb4_ethofld_send_flowc(struct net_device *dev, u32 eotid, u32 tc);
+void cxgb4_ethofld_restart(struct tasklet_struct *t);
+int cxgb4_ethofld_rx_handler(struct sge_rspq *q, const __be64 *rsp,
+			     const struct pkt_gl *si);
+void free_txq(struct adapter *adap, struct sge_txq *q);
+void cxgb4_reclaim_completed_tx(struct adapter *adap,
+				struct sge_txq *q, bool unmap);
+int cxgb4_map_skb(struct device *dev, const struct sk_buff *skb,
+		  dma_addr_t *addr);
+void cxgb4_inline_tx_skb(const struct sk_buff *skb, const struct sge_txq *q,
+			 void *pos);
+void cxgb4_write_sgl(const struct sk_buff *skb, struct sge_txq *q,
+		     struct ulptx_sgl *sgl, u64 *end, unsigned int start,
+		     const dma_addr_t *addr);
+void cxgb4_write_partial_sgl(const struct sk_buff *skb, struct sge_txq *q,
+			     struct ulptx_sgl *sgl, u64 *end,
+			     const dma_addr_t *addr, u32 start, u32 send_len);
+void cxgb4_ring_tx_db(struct adapter *adap, struct sge_txq *q, int n);
+int t4_set_vlan_acl(struct adapter *adap, unsigned int mbox, unsigned int vf,
+		    u16 vlan);
+int cxgb4_dcb_enabled(const struct net_device *dev);
+
+int cxgb4_thermal_init(struct adapter *adap);
+int cxgb4_thermal_remove(struct adapter *adap);
+int cxgb4_set_msix_aff(struct adapter *adap, unsigned short vec,
+		       cpumask_var_t *aff_mask, int idx);
+void cxgb4_clear_msix_aff(unsigned short vec, cpumask_var_t aff_mask);
+
+int cxgb4_change_mac(struct port_info *pi, unsigned int viid,
+		     int *tcam_idx, const u8 *addr,
+		     bool persistent, u8 *smt_idx);
+
+int cxgb4_alloc_mac_filt(struct adapter *adap, unsigned int viid,
+			 bool free, unsigned int naddr,
+			 const u8 **addr, u16 *idx,
+			 u64 *hash, bool sleep_ok);
+int cxgb4_free_mac_filt(struct adapter *adap, unsigned int viid,
+			unsigned int naddr, const u8 **addr, bool sleep_ok);
+int cxgb4_init_mps_ref_entries(struct adapter *adap);
+void cxgb4_free_mps_ref_entries(struct adapter *adap);
+int cxgb4_alloc_encap_mac_filt(struct adapter *adap, unsigned int viid,
+			       const u8 *addr, const u8 *mask,
+			       unsigned int vni, unsigned int vni_mask,
+			       u8 dip_hit, u8 lookup_type, bool sleep_ok);
+int cxgb4_free_encap_mac_filt(struct adapter *adap, unsigned int viid,
+			      int idx, bool sleep_ok);
+int cxgb4_free_raw_mac_filt(struct adapter *adap,
+			    unsigned int viid,
+			    const u8 *addr,
+			    const u8 *mask,
+			    unsigned int idx,
+			    u8 lookup_type,
+			    u8 port_id,
+			    bool sleep_ok);
+int cxgb4_alloc_raw_mac_filt(struct adapter *adap,
+			     unsigned int viid,
+			     const u8 *addr,
+			     const u8 *mask,
+			     unsigned int idx,
+			     u8 lookup_type,
+			     u8 port_id,
+			     bool sleep_ok);
+int cxgb4_update_mac_filt(struct port_info *pi, unsigned int viid,
+			  int *tcam_idx, const u8 *addr,
+			  bool persistent, u8 *smt_idx);
+int cxgb4_get_msix_idx_from_bmap(struct adapter *adap);
+void cxgb4_free_msix_idx_in_bmap(struct adapter *adap, u32 msix_idx);
+void cxgb4_enable_rx(struct adapter *adap, struct sge_rspq *q);
+void cxgb4_quiesce_rx(struct sge_rspq *q);
+int cxgb4_port_mirror_alloc(struct net_device *dev);
+void cxgb4_port_mirror_free(struct net_device *dev);
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+int cxgb4_set_ktls_feature(struct adapter *adap, bool enable);
+#endif
+#endif /* __CXGB4_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_cudbg.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_cudbg.c
new file mode 100644
index 000000000..dd66b2444
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_cudbg.c
@@ -0,0 +1,279 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  Copyright (C) 2017 Chelsio Communications.  All rights reserved.
+ */
+
+#include "t4_regs.h"
+#include "cxgb4.h"
+#include "cxgb4_cudbg.h"
+#include "cudbg_zlib.h"
+
+static const struct cxgb4_collect_entity cxgb4_collect_mem_dump[] = {
+	{ CUDBG_EDC0, cudbg_collect_edc0_meminfo },
+	{ CUDBG_EDC1, cudbg_collect_edc1_meminfo },
+	{ CUDBG_MC0, cudbg_collect_mc0_meminfo },
+	{ CUDBG_MC1, cudbg_collect_mc1_meminfo },
+	{ CUDBG_HMA, cudbg_collect_hma_meminfo },
+};
+
+static const struct cxgb4_collect_entity cxgb4_collect_hw_dump[] = {
+	{ CUDBG_MBOX_LOG, cudbg_collect_mbox_log },
+	{ CUDBG_QDESC, cudbg_collect_qdesc },
+	{ CUDBG_DEV_LOG, cudbg_collect_fw_devlog },
+	{ CUDBG_REG_DUMP, cudbg_collect_reg_dump },
+	{ CUDBG_CIM_LA, cudbg_collect_cim_la },
+	{ CUDBG_CIM_MA_LA, cudbg_collect_cim_ma_la },
+	{ CUDBG_CIM_QCFG, cudbg_collect_cim_qcfg },
+	{ CUDBG_CIM_IBQ_TP0, cudbg_collect_cim_ibq_tp0 },
+	{ CUDBG_CIM_IBQ_TP1, cudbg_collect_cim_ibq_tp1 },
+	{ CUDBG_CIM_IBQ_ULP, cudbg_collect_cim_ibq_ulp },
+	{ CUDBG_CIM_IBQ_SGE0, cudbg_collect_cim_ibq_sge0 },
+	{ CUDBG_CIM_IBQ_SGE1, cudbg_collect_cim_ibq_sge1 },
+	{ CUDBG_CIM_IBQ_NCSI, cudbg_collect_cim_ibq_ncsi },
+	{ CUDBG_CIM_OBQ_ULP0, cudbg_collect_cim_obq_ulp0 },
+	{ CUDBG_CIM_OBQ_ULP1, cudbg_collect_cim_obq_ulp1 },
+	{ CUDBG_CIM_OBQ_ULP2, cudbg_collect_cim_obq_ulp2 },
+	{ CUDBG_CIM_OBQ_ULP3, cudbg_collect_cim_obq_ulp3 },
+	{ CUDBG_CIM_OBQ_SGE, cudbg_collect_cim_obq_sge },
+	{ CUDBG_CIM_OBQ_NCSI, cudbg_collect_cim_obq_ncsi },
+	{ CUDBG_RSS, cudbg_collect_rss },
+	{ CUDBG_RSS_VF_CONF, cudbg_collect_rss_vf_config },
+	{ CUDBG_PATH_MTU, cudbg_collect_path_mtu },
+	{ CUDBG_PM_STATS, cudbg_collect_pm_stats },
+	{ CUDBG_HW_SCHED, cudbg_collect_hw_sched },
+	{ CUDBG_TP_INDIRECT, cudbg_collect_tp_indirect },
+	{ CUDBG_SGE_INDIRECT, cudbg_collect_sge_indirect },
+	{ CUDBG_ULPRX_LA, cudbg_collect_ulprx_la },
+	{ CUDBG_TP_LA, cudbg_collect_tp_la },
+	{ CUDBG_MEMINFO, cudbg_collect_meminfo },
+	{ CUDBG_CIM_PIF_LA, cudbg_collect_cim_pif_la },
+	{ CUDBG_CLK, cudbg_collect_clk_info },
+	{ CUDBG_CIM_OBQ_RXQ0, cudbg_collect_obq_sge_rx_q0 },
+	{ CUDBG_CIM_OBQ_RXQ1, cudbg_collect_obq_sge_rx_q1 },
+	{ CUDBG_PCIE_INDIRECT, cudbg_collect_pcie_indirect },
+	{ CUDBG_PM_INDIRECT, cudbg_collect_pm_indirect },
+	{ CUDBG_TID_INFO, cudbg_collect_tid },
+	{ CUDBG_PCIE_CONFIG, cudbg_collect_pcie_config },
+	{ CUDBG_DUMP_CONTEXT, cudbg_collect_dump_context },
+	{ CUDBG_MPS_TCAM, cudbg_collect_mps_tcam },
+	{ CUDBG_VPD_DATA, cudbg_collect_vpd_data },
+	{ CUDBG_LE_TCAM, cudbg_collect_le_tcam },
+	{ CUDBG_CCTRL, cudbg_collect_cctrl },
+	{ CUDBG_MA_INDIRECT, cudbg_collect_ma_indirect },
+	{ CUDBG_ULPTX_LA, cudbg_collect_ulptx_la },
+	{ CUDBG_UP_CIM_INDIRECT, cudbg_collect_up_cim_indirect },
+	{ CUDBG_PBT_TABLE, cudbg_collect_pbt_tables },
+	{ CUDBG_HMA_INDIRECT, cudbg_collect_hma_indirect },
+};
+
+static const struct cxgb4_collect_entity cxgb4_collect_flash_dump[] = {
+	{ CUDBG_FLASH, cudbg_collect_flash },
+};
+
+u32 cxgb4_get_dump_length(struct adapter *adap, u32 flag)
+{
+	u32 i, entity;
+	u32 len = 0;
+	u32 wsize;
+
+	if (flag & CXGB4_ETH_DUMP_HW) {
+		for (i = 0; i < ARRAY_SIZE(cxgb4_collect_hw_dump); i++) {
+			entity = cxgb4_collect_hw_dump[i].entity;
+			len += cudbg_get_entity_length(adap, entity);
+		}
+	}
+
+	if (flag & CXGB4_ETH_DUMP_MEM) {
+		for (i = 0; i < ARRAY_SIZE(cxgb4_collect_mem_dump); i++) {
+			entity = cxgb4_collect_mem_dump[i].entity;
+			len += cudbg_get_entity_length(adap, entity);
+		}
+	}
+
+	if (flag & CXGB4_ETH_DUMP_FLASH)
+		len += adap->params.sf_size;
+
+	/* If compression is enabled, a smaller destination buffer is enough */
+	wsize = cudbg_get_workspace_size();
+	if (wsize && len > CUDBG_DUMP_BUFF_SIZE)
+		len = CUDBG_DUMP_BUFF_SIZE;
+
+	return len;
+}
+
+static void cxgb4_cudbg_collect_entity(struct cudbg_init *pdbg_init,
+				       struct cudbg_buffer *dbg_buff,
+				       const struct cxgb4_collect_entity *e_arr,
+				       u32 arr_size, void *buf, u32 *tot_size)
+{
+	struct cudbg_error cudbg_err = { 0 };
+	struct cudbg_entity_hdr *entity_hdr;
+	u32 i, total_size = 0;
+	int ret;
+
+	for (i = 0; i < arr_size; i++) {
+		const struct cxgb4_collect_entity *e = &e_arr[i];
+
+		entity_hdr = cudbg_get_entity_hdr(buf, e->entity);
+		entity_hdr->entity_type = e->entity;
+		entity_hdr->start_offset = dbg_buff->offset;
+		memset(&cudbg_err, 0, sizeof(struct cudbg_error));
+		ret = e->collect_cb(pdbg_init, dbg_buff, &cudbg_err);
+		if (ret) {
+			entity_hdr->size = 0;
+			dbg_buff->offset = entity_hdr->start_offset;
+		} else {
+			cudbg_align_debug_buffer(dbg_buff, entity_hdr);
+		}
+
+		/* Log error and continue with next entity */
+		if (cudbg_err.sys_err)
+			ret = CUDBG_SYSTEM_ERROR;
+
+		entity_hdr->hdr_flags = ret;
+		entity_hdr->sys_err = cudbg_err.sys_err;
+		entity_hdr->sys_warn = cudbg_err.sys_warn;
+		total_size += entity_hdr->size;
+	}
+
+	*tot_size += total_size;
+}
+
+static int cudbg_alloc_compress_buff(struct cudbg_init *pdbg_init)
+{
+	u32 workspace_size;
+
+	workspace_size = cudbg_get_workspace_size();
+	pdbg_init->compress_buff = vzalloc(CUDBG_COMPRESS_BUFF_SIZE +
+					   workspace_size);
+	if (!pdbg_init->compress_buff)
+		return -ENOMEM;
+
+	pdbg_init->compress_buff_size = CUDBG_COMPRESS_BUFF_SIZE;
+	pdbg_init->workspace = (u8 *)pdbg_init->compress_buff +
+			       CUDBG_COMPRESS_BUFF_SIZE - workspace_size;
+	return 0;
+}
+
+static void cudbg_free_compress_buff(struct cudbg_init *pdbg_init)
+{
+	vfree(pdbg_init->compress_buff);
+}
+
+int cxgb4_cudbg_collect(struct adapter *adap, void *buf, u32 *buf_size,
+			u32 flag)
+{
+	struct cudbg_buffer dbg_buff = { 0 };
+	u32 size, min_size, total_size = 0;
+	struct cudbg_init cudbg_init;
+	struct cudbg_hdr *cudbg_hdr;
+	int rc;
+
+	size = *buf_size;
+
+	memset(&cudbg_init, 0, sizeof(struct cudbg_init));
+	cudbg_init.adap = adap;
+	cudbg_init.outbuf = buf;
+	cudbg_init.outbuf_size = size;
+
+	dbg_buff.data = buf;
+	dbg_buff.size = size;
+	dbg_buff.offset = 0;
+
+	cudbg_hdr = (struct cudbg_hdr *)buf;
+	cudbg_hdr->signature = CUDBG_SIGNATURE;
+	cudbg_hdr->hdr_len = sizeof(struct cudbg_hdr);
+	cudbg_hdr->major_ver = CUDBG_MAJOR_VERSION;
+	cudbg_hdr->minor_ver = CUDBG_MINOR_VERSION;
+	cudbg_hdr->max_entities = CUDBG_MAX_ENTITY;
+	cudbg_hdr->chip_ver = adap->params.chip;
+	cudbg_hdr->dump_type = CUDBG_DUMP_TYPE_MINI;
+
+	min_size = sizeof(struct cudbg_hdr) +
+		   sizeof(struct cudbg_entity_hdr) *
+		   cudbg_hdr->max_entities;
+	if (size < min_size)
+		return -ENOMEM;
+
+	rc = cudbg_get_workspace_size();
+	if (rc) {
+		/* Zlib available.  So, use zlib deflate */
+		cudbg_init.compress_type = CUDBG_COMPRESSION_ZLIB;
+		rc = cudbg_alloc_compress_buff(&cudbg_init);
+		if (rc) {
+			/* Ignore error and continue without compression. */
+			dev_warn(adap->pdev_dev,
+				 "Fail allocating compression buffer ret: %d.  Continuing without compression.\n",
+				 rc);
+			cudbg_init.compress_type = CUDBG_COMPRESSION_NONE;
+			rc = 0;
+		}
+	} else {
+		cudbg_init.compress_type = CUDBG_COMPRESSION_NONE;
+	}
+
+	cudbg_hdr->compress_type = cudbg_init.compress_type;
+	dbg_buff.offset += min_size;
+	total_size = dbg_buff.offset;
+
+	if (flag & CXGB4_ETH_DUMP_HW)
+		cxgb4_cudbg_collect_entity(&cudbg_init, &dbg_buff,
+					   cxgb4_collect_hw_dump,
+					   ARRAY_SIZE(cxgb4_collect_hw_dump),
+					   buf,
+					   &total_size);
+
+	if (flag & CXGB4_ETH_DUMP_MEM)
+		cxgb4_cudbg_collect_entity(&cudbg_init, &dbg_buff,
+					   cxgb4_collect_mem_dump,
+					   ARRAY_SIZE(cxgb4_collect_mem_dump),
+					   buf,
+					   &total_size);
+
+	if (flag & CXGB4_ETH_DUMP_FLASH)
+		cxgb4_cudbg_collect_entity(&cudbg_init, &dbg_buff,
+					   cxgb4_collect_flash_dump,
+					   ARRAY_SIZE(cxgb4_collect_flash_dump),
+					   buf,
+					   &total_size);
+
+	cudbg_free_compress_buff(&cudbg_init);
+	cudbg_hdr->data_len = total_size;
+	if (cudbg_init.compress_type != CUDBG_COMPRESSION_NONE)
+		*buf_size = size;
+	else
+		*buf_size = total_size;
+	return 0;
+}
+
+void cxgb4_init_ethtool_dump(struct adapter *adapter)
+{
+	adapter->eth_dump.flag = CXGB4_ETH_DUMP_NONE;
+	adapter->eth_dump.version = adapter->params.fw_vers;
+	adapter->eth_dump.len = 0;
+}
+
+static int cxgb4_cudbg_vmcoredd_collect(struct vmcoredd_data *data, void *buf)
+{
+	struct adapter *adap = container_of(data, struct adapter, vmcoredd);
+	u32 len = data->size;
+
+	return cxgb4_cudbg_collect(adap, buf, &len, CXGB4_ETH_DUMP_ALL);
+}
+
+int cxgb4_cudbg_vmcore_add_dump(struct adapter *adap)
+{
+	struct vmcoredd_data *data = &adap->vmcoredd;
+	u32 len;
+
+	len = sizeof(struct cudbg_hdr) +
+	      sizeof(struct cudbg_entity_hdr) * CUDBG_MAX_ENTITY;
+	len += CUDBG_DUMP_BUFF_SIZE;
+
+	data->size = len;
+	snprintf(data->dump_name, sizeof(data->dump_name), "%s_%s",
+		 cxgb4_driver_name, adap->name);
+	data->vmcoredd_callback = cxgb4_cudbg_vmcoredd_collect;
+
+	return vmcore_add_device_dump(data);
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_cudbg.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_cudbg.h
new file mode 100644
index 000000000..c04a49b63
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_cudbg.h
@@ -0,0 +1,40 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ *  Copyright (C) 2017 Chelsio Communications.  All rights reserved.
+ */
+
+#ifndef __CXGB4_CUDBG_H__
+#define __CXGB4_CUDBG_H__
+
+#include "cudbg_if.h"
+#include "cudbg_lib_common.h"
+#include "cudbg_entity.h"
+#include "cudbg_lib.h"
+
+#define CUDBG_DUMP_BUFF_SIZE (32 * 1024 * 1024) /* 32 MB */
+#define CUDBG_COMPRESS_BUFF_SIZE (4 * 1024 * 1024) /* 4 MB */
+
+typedef int (*cudbg_collect_callback_t)(struct cudbg_init *pdbg_init,
+					struct cudbg_buffer *dbg_buff,
+					struct cudbg_error *cudbg_err);
+
+struct cxgb4_collect_entity {
+	enum cudbg_dbg_entity_type entity;
+	cudbg_collect_callback_t collect_cb;
+};
+
+enum CXGB4_ETHTOOL_DUMP_FLAGS {
+	CXGB4_ETH_DUMP_NONE = ETH_FW_DUMP_DISABLE,
+	CXGB4_ETH_DUMP_MEM = (1 << 0), /* On-Chip Memory Dumps */
+	CXGB4_ETH_DUMP_HW = (1 << 1), /* various FW and HW dumps */
+	CXGB4_ETH_DUMP_FLASH = (1 << 2), /* Dump flash memory */
+};
+
+#define CXGB4_ETH_DUMP_ALL (CXGB4_ETH_DUMP_MEM | CXGB4_ETH_DUMP_HW)
+
+u32 cxgb4_get_dump_length(struct adapter *adap, u32 flag);
+int cxgb4_cudbg_collect(struct adapter *adap, void *buf, u32 *buf_size,
+			u32 flag);
+void cxgb4_init_ethtool_dump(struct adapter *adapter);
+int cxgb4_cudbg_vmcore_add_dump(struct adapter *adap);
+#endif /* __CXGB4_CUDBG_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_dcb.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_dcb.c
new file mode 100644
index 000000000..7d5204834
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_dcb.c
@@ -0,0 +1,1281 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  Copyright (C) 2013-2014 Chelsio Communications.  All rights reserved.
+ *
+ *  Written by Anish Bhatt (anish@chelsio.com)
+ *	       Casey Leedom (leedom@chelsio.com)
+ */
+
+#include "cxgb4.h"
+
+/* DCBx version control
+ */
+const char * const dcb_ver_array[] = {
+	"Unknown",
+	"DCBx-CIN",
+	"DCBx-CEE 1.01",
+	"DCBx-IEEE",
+	"", "", "",
+	"Auto Negotiated"
+};
+
+static inline bool cxgb4_dcb_state_synced(enum cxgb4_dcb_state state)
+{
+	if (state == CXGB4_DCB_STATE_FW_ALLSYNCED ||
+	    state == CXGB4_DCB_STATE_HOST)
+		return true;
+	else
+		return false;
+}
+
+/* Initialize a port's Data Center Bridging state.
+ */
+void cxgb4_dcb_state_init(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct port_dcb_info *dcb = &pi->dcb;
+	int version_temp = dcb->dcb_version;
+
+	memset(dcb, 0, sizeof(struct port_dcb_info));
+	dcb->state = CXGB4_DCB_STATE_START;
+	if (version_temp)
+		dcb->dcb_version = version_temp;
+
+	netdev_dbg(dev, "%s: Initializing DCB state for port[%d]\n",
+		    __func__, pi->port_id);
+}
+
+void cxgb4_dcb_version_init(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct port_dcb_info *dcb = &pi->dcb;
+
+	/* Any writes here are only done on kernels that exlicitly need
+	 * a specific version, say < 2.6.38 which only support CEE
+	 */
+	dcb->dcb_version = FW_PORT_DCB_VER_AUTO;
+}
+
+static void cxgb4_dcb_cleanup_apps(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = pi->adapter;
+	struct port_dcb_info *dcb = &pi->dcb;
+	struct dcb_app app;
+	int i, err;
+
+	/* zero priority implies remove */
+	app.priority = 0;
+
+	for (i = 0; i < CXGB4_MAX_DCBX_APP_SUPPORTED; i++) {
+		/* Check if app list is exhausted */
+		if (!dcb->app_priority[i].protocolid)
+			break;
+
+		app.protocol = dcb->app_priority[i].protocolid;
+
+		if (dcb->dcb_version == FW_PORT_DCB_VER_IEEE) {
+			app.priority = dcb->app_priority[i].user_prio_map;
+			app.selector = dcb->app_priority[i].sel_field + 1;
+			err = dcb_ieee_delapp(dev, &app);
+		} else {
+			app.selector = !!(dcb->app_priority[i].sel_field);
+			err = dcb_setapp(dev, &app);
+		}
+
+		if (err) {
+			dev_err(adap->pdev_dev,
+				"Failed DCB Clear %s Application Priority: sel=%d, prot=%d, err=%d\n",
+				dcb_ver_array[dcb->dcb_version], app.selector,
+				app.protocol, -err);
+			break;
+		}
+	}
+}
+
+/* Reset a port's Data Center Bridging state.  Typically used after a
+ * Link Down event.
+ */
+void cxgb4_dcb_reset(struct net_device *dev)
+{
+	cxgb4_dcb_cleanup_apps(dev);
+	cxgb4_dcb_state_init(dev);
+}
+
+/* update the dcb port support, if version is IEEE then set it to
+ * FW_PORT_DCB_VER_IEEE and if DCB_CAP_DCBX_VER_CEE is already set then
+ * clear that. and if it is set to CEE then set dcb supported to
+ * DCB_CAP_DCBX_VER_CEE & if DCB_CAP_DCBX_VER_IEEE is set, clear it
+ */
+static inline void cxgb4_dcb_update_support(struct port_dcb_info *dcb)
+{
+	if (dcb->dcb_version == FW_PORT_DCB_VER_IEEE) {
+		if (dcb->supported & DCB_CAP_DCBX_VER_CEE)
+			dcb->supported &= ~DCB_CAP_DCBX_VER_CEE;
+		dcb->supported |= DCB_CAP_DCBX_VER_IEEE;
+	} else if (dcb->dcb_version == FW_PORT_DCB_VER_CEE1D01) {
+		if (dcb->supported & DCB_CAP_DCBX_VER_IEEE)
+			dcb->supported &= ~DCB_CAP_DCBX_VER_IEEE;
+		dcb->supported |= DCB_CAP_DCBX_VER_CEE;
+	}
+}
+
+/* Finite State machine for Data Center Bridging.
+ */
+void cxgb4_dcb_state_fsm(struct net_device *dev,
+			 enum cxgb4_dcb_state_input transition_to)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct port_dcb_info *dcb = &pi->dcb;
+	struct adapter *adap = pi->adapter;
+	enum cxgb4_dcb_state current_state = dcb->state;
+
+	netdev_dbg(dev, "%s: State change from %d to %d for %s\n",
+		    __func__, dcb->state, transition_to, dev->name);
+
+	switch (current_state) {
+	case CXGB4_DCB_STATE_START: {
+		switch (transition_to) {
+		case CXGB4_DCB_INPUT_FW_DISABLED: {
+			/* we're going to use Host DCB */
+			dcb->state = CXGB4_DCB_STATE_HOST;
+			dcb->supported = CXGB4_DCBX_HOST_SUPPORT;
+			break;
+		}
+
+		case CXGB4_DCB_INPUT_FW_ENABLED: {
+			/* we're going to use Firmware DCB */
+			dcb->state = CXGB4_DCB_STATE_FW_INCOMPLETE;
+			dcb->supported = DCB_CAP_DCBX_LLD_MANAGED;
+			if (dcb->dcb_version == FW_PORT_DCB_VER_IEEE)
+				dcb->supported |= DCB_CAP_DCBX_VER_IEEE;
+			else
+				dcb->supported |= DCB_CAP_DCBX_VER_CEE;
+			break;
+		}
+
+		case CXGB4_DCB_INPUT_FW_INCOMPLETE: {
+			/* expected transition */
+			break;
+		}
+
+		case CXGB4_DCB_INPUT_FW_ALLSYNCED: {
+			dcb->state = CXGB4_DCB_STATE_FW_ALLSYNCED;
+			break;
+		}
+
+		default:
+			goto bad_state_input;
+		}
+		break;
+	}
+
+	case CXGB4_DCB_STATE_FW_INCOMPLETE: {
+		if (transition_to != CXGB4_DCB_INPUT_FW_DISABLED) {
+			/* during this CXGB4_DCB_STATE_FW_INCOMPLETE state,
+			 * check if the dcb version is changed (there can be
+			 * mismatch in default config & the negotiated switch
+			 * configuration at FW, so update the dcb support
+			 * accordingly.
+			 */
+			cxgb4_dcb_update_support(dcb);
+		}
+		switch (transition_to) {
+		case CXGB4_DCB_INPUT_FW_ENABLED: {
+			/* we're alreaady in firmware DCB mode */
+			break;
+		}
+
+		case CXGB4_DCB_INPUT_FW_INCOMPLETE: {
+			/* we're already incomplete */
+			break;
+		}
+
+		case CXGB4_DCB_INPUT_FW_ALLSYNCED: {
+			dcb->state = CXGB4_DCB_STATE_FW_ALLSYNCED;
+			dcb->enabled = 1;
+			linkwatch_fire_event(dev);
+			break;
+		}
+
+		default:
+			goto bad_state_input;
+		}
+		break;
+	}
+
+	case CXGB4_DCB_STATE_FW_ALLSYNCED: {
+		switch (transition_to) {
+		case CXGB4_DCB_INPUT_FW_ENABLED: {
+			/* we're alreaady in firmware DCB mode */
+			break;
+		}
+
+		case CXGB4_DCB_INPUT_FW_INCOMPLETE: {
+			/* We were successfully running with firmware DCB but
+			 * now it's telling us that it's in an "incomplete
+			 * state.  We need to reset back to a ground state
+			 * of incomplete.
+			 */
+			cxgb4_dcb_reset(dev);
+			dcb->state = CXGB4_DCB_STATE_FW_INCOMPLETE;
+			dcb->supported = CXGB4_DCBX_FW_SUPPORT;
+			linkwatch_fire_event(dev);
+			break;
+		}
+
+		case CXGB4_DCB_INPUT_FW_ALLSYNCED: {
+			/* we're already all sync'ed
+			 * this is only applicable for IEEE or
+			 * when another VI already completed negotiaton
+			 */
+			dcb->enabled = 1;
+			linkwatch_fire_event(dev);
+			break;
+		}
+
+		default:
+			goto bad_state_input;
+		}
+		break;
+	}
+
+	case CXGB4_DCB_STATE_HOST: {
+		switch (transition_to) {
+		case CXGB4_DCB_INPUT_FW_DISABLED: {
+			/* we're alreaady in Host DCB mode */
+			break;
+		}
+
+		default:
+			goto bad_state_input;
+		}
+		break;
+	}
+
+	default:
+		goto bad_state_transition;
+	}
+	return;
+
+bad_state_input:
+	dev_err(adap->pdev_dev, "cxgb4_dcb_state_fsm: illegal input symbol %d\n",
+		transition_to);
+	return;
+
+bad_state_transition:
+	dev_err(adap->pdev_dev, "cxgb4_dcb_state_fsm: bad state transition, state = %d, input = %d\n",
+		current_state, transition_to);
+}
+
+/* Handle a DCB/DCBX update message from the firmware.
+ */
+void cxgb4_dcb_handle_fw_update(struct adapter *adap,
+				const struct fw_port_cmd *pcmd)
+{
+	const union fw_port_dcb *fwdcb = &pcmd->u.dcb;
+	int port = FW_PORT_CMD_PORTID_G(be32_to_cpu(pcmd->op_to_portid));
+	struct net_device *dev = adap->port[adap->chan_map[port]];
+	struct port_info *pi = netdev_priv(dev);
+	struct port_dcb_info *dcb = &pi->dcb;
+	int dcb_type = pcmd->u.dcb.pgid.type;
+	int dcb_running_version;
+
+	/* Handle Firmware DCB Control messages separately since they drive
+	 * our state machine.
+	 */
+	if (dcb_type == FW_PORT_DCB_TYPE_CONTROL) {
+		enum cxgb4_dcb_state_input input =
+			((pcmd->u.dcb.control.all_syncd_pkd &
+			  FW_PORT_CMD_ALL_SYNCD_F)
+			 ? CXGB4_DCB_INPUT_FW_ALLSYNCED
+			 : CXGB4_DCB_INPUT_FW_INCOMPLETE);
+
+		if (dcb->dcb_version != FW_PORT_DCB_VER_UNKNOWN) {
+			dcb_running_version = FW_PORT_CMD_DCB_VERSION_G(
+				be16_to_cpu(
+				pcmd->u.dcb.control.dcb_version_to_app_state));
+			if (dcb_running_version == FW_PORT_DCB_VER_CEE1D01 ||
+			    dcb_running_version == FW_PORT_DCB_VER_IEEE) {
+				dcb->dcb_version = dcb_running_version;
+				dev_warn(adap->pdev_dev, "Interface %s is running %s\n",
+					 dev->name,
+					 dcb_ver_array[dcb->dcb_version]);
+			} else {
+				dev_warn(adap->pdev_dev,
+					 "Something screwed up, requested firmware for %s, but firmware returned %s instead\n",
+					 dcb_ver_array[dcb->dcb_version],
+					 dcb_ver_array[dcb_running_version]);
+				dcb->dcb_version = FW_PORT_DCB_VER_UNKNOWN;
+			}
+		}
+
+		cxgb4_dcb_state_fsm(dev, input);
+		return;
+	}
+
+	/* It's weird, and almost certainly an error, to get Firmware DCB
+	 * messages when we either haven't been told whether we're going to be
+	 * doing Host or Firmware DCB; and even worse when we've been told
+	 * that we're doing Host DCB!
+	 */
+	if (dcb->state == CXGB4_DCB_STATE_START ||
+	    dcb->state == CXGB4_DCB_STATE_HOST) {
+		dev_err(adap->pdev_dev, "Receiving Firmware DCB messages in State %d\n",
+			dcb->state);
+		return;
+	}
+
+	/* Now handle the general Firmware DCB update messages ...
+	 */
+	switch (dcb_type) {
+	case FW_PORT_DCB_TYPE_PGID:
+		dcb->pgid = be32_to_cpu(fwdcb->pgid.pgid);
+		dcb->msgs |= CXGB4_DCB_FW_PGID;
+		break;
+
+	case FW_PORT_DCB_TYPE_PGRATE:
+		dcb->pg_num_tcs_supported = fwdcb->pgrate.num_tcs_supported;
+		memcpy(dcb->pgrate, &fwdcb->pgrate.pgrate,
+		       sizeof(dcb->pgrate));
+		memcpy(dcb->tsa, &fwdcb->pgrate.tsa,
+		       sizeof(dcb->tsa));
+		dcb->msgs |= CXGB4_DCB_FW_PGRATE;
+		if (dcb->msgs & CXGB4_DCB_FW_PGID)
+			IEEE_FAUX_SYNC(dev, dcb);
+		break;
+
+	case FW_PORT_DCB_TYPE_PRIORATE:
+		memcpy(dcb->priorate, &fwdcb->priorate.strict_priorate,
+		       sizeof(dcb->priorate));
+		dcb->msgs |= CXGB4_DCB_FW_PRIORATE;
+		break;
+
+	case FW_PORT_DCB_TYPE_PFC:
+		dcb->pfcen = fwdcb->pfc.pfcen;
+		dcb->pfc_num_tcs_supported = fwdcb->pfc.max_pfc_tcs;
+		dcb->msgs |= CXGB4_DCB_FW_PFC;
+		IEEE_FAUX_SYNC(dev, dcb);
+		break;
+
+	case FW_PORT_DCB_TYPE_APP_ID: {
+		const struct fw_port_app_priority *fwap = &fwdcb->app_priority;
+		int idx = fwap->idx;
+		struct app_priority *ap = &dcb->app_priority[idx];
+
+		struct dcb_app app = {
+			.protocol = be16_to_cpu(fwap->protocolid),
+		};
+		int err;
+
+		/* Convert from firmware format to relevant format
+		 * when using app selector
+		 */
+		if (dcb->dcb_version == FW_PORT_DCB_VER_IEEE) {
+			app.selector = (fwap->sel_field + 1);
+			app.priority = ffs(fwap->user_prio_map) - 1;
+			err = dcb_ieee_setapp(dev, &app);
+			IEEE_FAUX_SYNC(dev, dcb);
+		} else {
+			/* Default is CEE */
+			app.selector = !!(fwap->sel_field);
+			app.priority = fwap->user_prio_map;
+			err = dcb_setapp(dev, &app);
+		}
+
+		if (err)
+			dev_err(adap->pdev_dev,
+				"Failed DCB Set Application Priority: sel=%d, prot=%d, prio=%d, err=%d\n",
+				app.selector, app.protocol, app.priority, -err);
+
+		ap->user_prio_map = fwap->user_prio_map;
+		ap->sel_field = fwap->sel_field;
+		ap->protocolid = be16_to_cpu(fwap->protocolid);
+		dcb->msgs |= CXGB4_DCB_FW_APP_ID;
+		break;
+	}
+
+	default:
+		dev_err(adap->pdev_dev, "Unknown DCB update type received %x\n",
+			dcb_type);
+		break;
+	}
+}
+
+/* Data Center Bridging netlink operations.
+ */
+
+
+/* Get current DCB enabled/disabled state.
+ */
+static u8 cxgb4_getstate(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+
+	return pi->dcb.enabled;
+}
+
+/* Set DCB enabled/disabled.
+ */
+static u8 cxgb4_setstate(struct net_device *dev, u8 enabled)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+
+	/* If DCBx is host-managed, dcb is enabled by outside lldp agents */
+	if (pi->dcb.state == CXGB4_DCB_STATE_HOST) {
+		pi->dcb.enabled = enabled;
+		return 0;
+	}
+
+	/* Firmware doesn't provide any mechanism to control the DCB state.
+	 */
+	if (enabled != (pi->dcb.state == CXGB4_DCB_STATE_FW_ALLSYNCED))
+		return 1;
+
+	return 0;
+}
+
+static void cxgb4_getpgtccfg(struct net_device *dev, int tc,
+			     u8 *prio_type, u8 *pgid, u8 *bw_per,
+			     u8 *up_tc_map, int local)
+{
+	struct fw_port_cmd pcmd;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = pi->adapter;
+	int err;
+
+	*prio_type = *pgid = *bw_per = *up_tc_map = 0;
+
+	if (local)
+		INIT_PORT_DCB_READ_LOCAL_CMD(pcmd, pi->port_id);
+	else
+		INIT_PORT_DCB_READ_PEER_CMD(pcmd, pi->port_id);
+
+	pcmd.u.dcb.pgid.type = FW_PORT_DCB_TYPE_PGID;
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB read PGID failed with %d\n", -err);
+		return;
+	}
+	*pgid = (be32_to_cpu(pcmd.u.dcb.pgid.pgid) >> (tc * 4)) & 0xf;
+
+	if (local)
+		INIT_PORT_DCB_READ_LOCAL_CMD(pcmd, pi->port_id);
+	else
+		INIT_PORT_DCB_READ_PEER_CMD(pcmd, pi->port_id);
+	pcmd.u.dcb.pgrate.type = FW_PORT_DCB_TYPE_PGRATE;
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB read PGRATE failed with %d\n",
+			-err);
+		return;
+	}
+
+	*bw_per = pcmd.u.dcb.pgrate.pgrate[*pgid];
+	*up_tc_map = (1 << tc);
+
+	/* prio_type is link strict */
+	if (*pgid != 0xF)
+		*prio_type = 0x2;
+}
+
+static void cxgb4_getpgtccfg_tx(struct net_device *dev, int tc,
+				u8 *prio_type, u8 *pgid, u8 *bw_per,
+				u8 *up_tc_map)
+{
+	/* tc 0 is written at MSB position */
+	return cxgb4_getpgtccfg(dev, (7 - tc), prio_type, pgid, bw_per,
+				up_tc_map, 1);
+}
+
+
+static void cxgb4_getpgtccfg_rx(struct net_device *dev, int tc,
+				u8 *prio_type, u8 *pgid, u8 *bw_per,
+				u8 *up_tc_map)
+{
+	/* tc 0 is written at MSB position */
+	return cxgb4_getpgtccfg(dev, (7 - tc), prio_type, pgid, bw_per,
+				up_tc_map, 0);
+}
+
+static void cxgb4_setpgtccfg_tx(struct net_device *dev, int tc,
+				u8 prio_type, u8 pgid, u8 bw_per,
+				u8 up_tc_map)
+{
+	struct fw_port_cmd pcmd;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = pi->adapter;
+	int fw_tc = 7 - tc;
+	u32 _pgid;
+	int err;
+
+	if (pgid == DCB_ATTR_VALUE_UNDEFINED)
+		return;
+	if (bw_per == DCB_ATTR_VALUE_UNDEFINED)
+		return;
+
+	INIT_PORT_DCB_READ_LOCAL_CMD(pcmd, pi->port_id);
+	pcmd.u.dcb.pgid.type = FW_PORT_DCB_TYPE_PGID;
+
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB read PGID failed with %d\n", -err);
+		return;
+	}
+
+	_pgid = be32_to_cpu(pcmd.u.dcb.pgid.pgid);
+	_pgid &= ~(0xF << (fw_tc * 4));
+	_pgid |= pgid << (fw_tc * 4);
+	pcmd.u.dcb.pgid.pgid = cpu_to_be32(_pgid);
+
+	INIT_PORT_DCB_WRITE_CMD(pcmd, pi->port_id);
+
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB write PGID failed with %d\n",
+			-err);
+		return;
+	}
+
+	memset(&pcmd, 0, sizeof(struct fw_port_cmd));
+
+	INIT_PORT_DCB_READ_LOCAL_CMD(pcmd, pi->port_id);
+	pcmd.u.dcb.pgrate.type = FW_PORT_DCB_TYPE_PGRATE;
+
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB read PGRATE failed with %d\n",
+			-err);
+		return;
+	}
+
+	pcmd.u.dcb.pgrate.pgrate[pgid] = bw_per;
+
+	INIT_PORT_DCB_WRITE_CMD(pcmd, pi->port_id);
+	if (pi->dcb.state == CXGB4_DCB_STATE_HOST)
+		pcmd.op_to_portid |= cpu_to_be32(FW_PORT_CMD_APPLY_F);
+
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS)
+		dev_err(adap->pdev_dev, "DCB write PGRATE failed with %d\n",
+			-err);
+}
+
+static void cxgb4_getpgbwgcfg(struct net_device *dev, int pgid, u8 *bw_per,
+			      int local)
+{
+	struct fw_port_cmd pcmd;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = pi->adapter;
+	int err;
+
+	if (local)
+		INIT_PORT_DCB_READ_LOCAL_CMD(pcmd, pi->port_id);
+	else
+		INIT_PORT_DCB_READ_PEER_CMD(pcmd, pi->port_id);
+
+	pcmd.u.dcb.pgrate.type = FW_PORT_DCB_TYPE_PGRATE;
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB read PGRATE failed with %d\n",
+			-err);
+		return;
+	}
+
+	*bw_per = pcmd.u.dcb.pgrate.pgrate[pgid];
+}
+
+static void cxgb4_getpgbwgcfg_tx(struct net_device *dev, int pgid, u8 *bw_per)
+{
+	return cxgb4_getpgbwgcfg(dev, pgid, bw_per, 1);
+}
+
+static void cxgb4_getpgbwgcfg_rx(struct net_device *dev, int pgid, u8 *bw_per)
+{
+	return cxgb4_getpgbwgcfg(dev, pgid, bw_per, 0);
+}
+
+static void cxgb4_setpgbwgcfg_tx(struct net_device *dev, int pgid,
+				 u8 bw_per)
+{
+	struct fw_port_cmd pcmd;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = pi->adapter;
+	int err;
+
+	INIT_PORT_DCB_READ_LOCAL_CMD(pcmd, pi->port_id);
+	pcmd.u.dcb.pgrate.type = FW_PORT_DCB_TYPE_PGRATE;
+
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB read PGRATE failed with %d\n",
+			-err);
+		return;
+	}
+
+	pcmd.u.dcb.pgrate.pgrate[pgid] = bw_per;
+
+	INIT_PORT_DCB_WRITE_CMD(pcmd, pi->port_id);
+	if (pi->dcb.state == CXGB4_DCB_STATE_HOST)
+		pcmd.op_to_portid |= cpu_to_be32(FW_PORT_CMD_APPLY_F);
+
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+
+	if (err != FW_PORT_DCB_CFG_SUCCESS)
+		dev_err(adap->pdev_dev, "DCB write PGRATE failed with %d\n",
+			-err);
+}
+
+/* Return whether the specified Traffic Class Priority has Priority Pause
+ * Frames enabled.
+ */
+static void cxgb4_getpfccfg(struct net_device *dev, int priority, u8 *pfccfg)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct port_dcb_info *dcb = &pi->dcb;
+
+	if (!cxgb4_dcb_state_synced(dcb->state) ||
+	    priority >= CXGB4_MAX_PRIORITY)
+		*pfccfg = 0;
+	else
+		*pfccfg = (pi->dcb.pfcen >> (7 - priority)) & 1;
+}
+
+/* Enable/disable Priority Pause Frames for the specified Traffic Class
+ * Priority.
+ */
+static void cxgb4_setpfccfg(struct net_device *dev, int priority, u8 pfccfg)
+{
+	struct fw_port_cmd pcmd;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = pi->adapter;
+	int err;
+
+	if (!cxgb4_dcb_state_synced(pi->dcb.state) ||
+	    priority >= CXGB4_MAX_PRIORITY)
+		return;
+
+	INIT_PORT_DCB_WRITE_CMD(pcmd, pi->port_id);
+	if (pi->dcb.state == CXGB4_DCB_STATE_HOST)
+		pcmd.op_to_portid |= cpu_to_be32(FW_PORT_CMD_APPLY_F);
+
+	pcmd.u.dcb.pfc.type = FW_PORT_DCB_TYPE_PFC;
+	pcmd.u.dcb.pfc.pfcen = pi->dcb.pfcen;
+
+	if (pfccfg)
+		pcmd.u.dcb.pfc.pfcen |= (1 << (7 - priority));
+	else
+		pcmd.u.dcb.pfc.pfcen &= (~(1 << (7 - priority)));
+
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB PFC write failed with %d\n", -err);
+		return;
+	}
+
+	pi->dcb.pfcen = pcmd.u.dcb.pfc.pfcen;
+}
+
+static u8 cxgb4_setall(struct net_device *dev)
+{
+	return 0;
+}
+
+/* Return DCB capabilities.
+ */
+static u8 cxgb4_getcap(struct net_device *dev, int cap_id, u8 *caps)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+
+	switch (cap_id) {
+	case DCB_CAP_ATTR_PG:
+	case DCB_CAP_ATTR_PFC:
+		*caps = true;
+		break;
+
+	case DCB_CAP_ATTR_PG_TCS:
+		/* 8 priorities for PG represented by bitmap */
+		*caps = 0x80;
+		break;
+
+	case DCB_CAP_ATTR_PFC_TCS:
+		/* 8 priorities for PFC represented by bitmap */
+		*caps = 0x80;
+		break;
+
+	case DCB_CAP_ATTR_GSP:
+		*caps = true;
+		break;
+
+	case DCB_CAP_ATTR_UP2TC:
+	case DCB_CAP_ATTR_BCN:
+		*caps = false;
+		break;
+
+	case DCB_CAP_ATTR_DCBX:
+		*caps = pi->dcb.supported;
+		break;
+
+	default:
+		*caps = false;
+	}
+
+	return 0;
+}
+
+/* Return the number of Traffic Classes for the indicated Traffic Class ID.
+ */
+static int cxgb4_getnumtcs(struct net_device *dev, int tcs_id, u8 *num)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+
+	switch (tcs_id) {
+	case DCB_NUMTCS_ATTR_PG:
+		if (pi->dcb.msgs & CXGB4_DCB_FW_PGRATE)
+			*num = pi->dcb.pg_num_tcs_supported;
+		else
+			*num = 0x8;
+		break;
+
+	case DCB_NUMTCS_ATTR_PFC:
+		*num = 0x8;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+/* Set the number of Traffic Classes supported for the indicated Traffic Class
+ * ID.
+ */
+static int cxgb4_setnumtcs(struct net_device *dev, int tcs_id, u8 num)
+{
+	/* Setting the number of Traffic Classes isn't supported.
+	 */
+	return -ENOSYS;
+}
+
+/* Return whether Priority Flow Control is enabled.  */
+static u8 cxgb4_getpfcstate(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+
+	if (!cxgb4_dcb_state_synced(pi->dcb.state))
+		return false;
+
+	return pi->dcb.pfcen != 0;
+}
+
+/* Enable/disable Priority Flow Control. */
+static void cxgb4_setpfcstate(struct net_device *dev, u8 state)
+{
+	/* We can't enable/disable Priority Flow Control but we also can't
+	 * return an error ...
+	 */
+}
+
+/* Return the Application User Priority Map associated with the specified
+ * Application ID.
+ */
+static int __cxgb4_getapp(struct net_device *dev, u8 app_idtype, u16 app_id,
+			  int peer)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = pi->adapter;
+	int i;
+
+	if (!cxgb4_dcb_state_synced(pi->dcb.state))
+		return 0;
+
+	for (i = 0; i < CXGB4_MAX_DCBX_APP_SUPPORTED; i++) {
+		struct fw_port_cmd pcmd;
+		int err;
+
+		if (peer)
+			INIT_PORT_DCB_READ_PEER_CMD(pcmd, pi->port_id);
+		else
+			INIT_PORT_DCB_READ_LOCAL_CMD(pcmd, pi->port_id);
+
+		pcmd.u.dcb.app_priority.type = FW_PORT_DCB_TYPE_APP_ID;
+		pcmd.u.dcb.app_priority.idx = i;
+
+		err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+		if (err != FW_PORT_DCB_CFG_SUCCESS) {
+			dev_err(adap->pdev_dev, "DCB APP read failed with %d\n",
+				-err);
+			return err;
+		}
+		if (be16_to_cpu(pcmd.u.dcb.app_priority.protocolid) == app_id)
+			if (pcmd.u.dcb.app_priority.sel_field == app_idtype)
+				return pcmd.u.dcb.app_priority.user_prio_map;
+
+		/* exhausted app list */
+		if (!pcmd.u.dcb.app_priority.protocolid)
+			break;
+	}
+
+	return -EEXIST;
+}
+
+/* Return the Application User Priority Map associated with the specified
+ * Application ID.
+ */
+static int cxgb4_getapp(struct net_device *dev, u8 app_idtype, u16 app_id)
+{
+	/* Convert app_idtype to firmware format before querying */
+	return __cxgb4_getapp(dev, app_idtype == DCB_APP_IDTYPE_ETHTYPE ?
+			      app_idtype : 3, app_id, 0);
+}
+
+/* Write a new Application User Priority Map for the specified Application ID
+ */
+static int __cxgb4_setapp(struct net_device *dev, u8 app_idtype, u16 app_id,
+			  u8 app_prio)
+{
+	struct fw_port_cmd pcmd;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = pi->adapter;
+	int i, err;
+
+
+	if (!cxgb4_dcb_state_synced(pi->dcb.state))
+		return -EINVAL;
+
+	/* DCB info gets thrown away on link up */
+	if (!netif_carrier_ok(dev))
+		return -ENOLINK;
+
+	for (i = 0; i < CXGB4_MAX_DCBX_APP_SUPPORTED; i++) {
+		INIT_PORT_DCB_READ_LOCAL_CMD(pcmd, pi->port_id);
+		pcmd.u.dcb.app_priority.type = FW_PORT_DCB_TYPE_APP_ID;
+		pcmd.u.dcb.app_priority.idx = i;
+		err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+
+		if (err != FW_PORT_DCB_CFG_SUCCESS) {
+			dev_err(adap->pdev_dev, "DCB app table read failed with %d\n",
+				-err);
+			return err;
+		}
+		if (be16_to_cpu(pcmd.u.dcb.app_priority.protocolid) == app_id) {
+			/* overwrite existing app table */
+			pcmd.u.dcb.app_priority.protocolid = 0;
+			break;
+		}
+		/* find first empty slot */
+		if (!pcmd.u.dcb.app_priority.protocolid)
+			break;
+	}
+
+	if (i == CXGB4_MAX_DCBX_APP_SUPPORTED) {
+		/* no empty slots available */
+		dev_err(adap->pdev_dev, "DCB app table full\n");
+		return -EBUSY;
+	}
+
+	/* write out new app table entry */
+	INIT_PORT_DCB_WRITE_CMD(pcmd, pi->port_id);
+	if (pi->dcb.state == CXGB4_DCB_STATE_HOST)
+		pcmd.op_to_portid |= cpu_to_be32(FW_PORT_CMD_APPLY_F);
+
+	pcmd.u.dcb.app_priority.type = FW_PORT_DCB_TYPE_APP_ID;
+	pcmd.u.dcb.app_priority.protocolid = cpu_to_be16(app_id);
+	pcmd.u.dcb.app_priority.sel_field = app_idtype;
+	pcmd.u.dcb.app_priority.user_prio_map = app_prio;
+	pcmd.u.dcb.app_priority.idx = i;
+
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB app table write failed with %d\n",
+			-err);
+		return err;
+	}
+
+	return 0;
+}
+
+/* Priority for CEE inside dcb_app is bitmask, with 0 being an invalid value */
+static int cxgb4_setapp(struct net_device *dev, u8 app_idtype, u16 app_id,
+			u8 app_prio)
+{
+	int ret;
+	struct dcb_app app = {
+		.selector = app_idtype,
+		.protocol = app_id,
+		.priority = app_prio,
+	};
+
+	if (app_idtype != DCB_APP_IDTYPE_ETHTYPE &&
+	    app_idtype != DCB_APP_IDTYPE_PORTNUM)
+		return -EINVAL;
+
+	/* Convert app_idtype to a format that firmware understands */
+	ret = __cxgb4_setapp(dev, app_idtype == DCB_APP_IDTYPE_ETHTYPE ?
+			      app_idtype : 3, app_id, app_prio);
+	if (ret)
+		return ret;
+
+	return dcb_setapp(dev, &app);
+}
+
+/* Return whether IEEE Data Center Bridging has been negotiated.
+ */
+static inline int
+cxgb4_ieee_negotiation_complete(struct net_device *dev,
+				enum cxgb4_dcb_fw_msgs dcb_subtype)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct port_dcb_info *dcb = &pi->dcb;
+
+	if (dcb->state == CXGB4_DCB_STATE_FW_ALLSYNCED)
+		if (dcb_subtype && !(dcb->msgs & dcb_subtype))
+			return 0;
+
+	return (cxgb4_dcb_state_synced(dcb->state) &&
+		(dcb->supported & DCB_CAP_DCBX_VER_IEEE));
+}
+
+static int cxgb4_ieee_read_ets(struct net_device *dev, struct ieee_ets *ets,
+			       int local)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct port_dcb_info *dcb = &pi->dcb;
+	struct adapter *adap = pi->adapter;
+	uint32_t tc_info;
+	struct fw_port_cmd pcmd;
+	int i, bwg, err;
+
+	if (!(dcb->msgs & (CXGB4_DCB_FW_PGID | CXGB4_DCB_FW_PGRATE)))
+		return 0;
+
+	ets->ets_cap =  dcb->pg_num_tcs_supported;
+
+	if (local) {
+		ets->willing = 1;
+		INIT_PORT_DCB_READ_LOCAL_CMD(pcmd, pi->port_id);
+	} else {
+		INIT_PORT_DCB_READ_PEER_CMD(pcmd, pi->port_id);
+	}
+
+	pcmd.u.dcb.pgid.type = FW_PORT_DCB_TYPE_PGID;
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB read PGID failed with %d\n", -err);
+		return err;
+	}
+
+	tc_info = be32_to_cpu(pcmd.u.dcb.pgid.pgid);
+
+	if (local)
+		INIT_PORT_DCB_READ_LOCAL_CMD(pcmd, pi->port_id);
+	else
+		INIT_PORT_DCB_READ_PEER_CMD(pcmd, pi->port_id);
+
+	pcmd.u.dcb.pgrate.type = FW_PORT_DCB_TYPE_PGRATE;
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB read PGRATE failed with %d\n",
+			-err);
+		return err;
+	}
+
+	for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) {
+		bwg = (tc_info >> ((7 - i) * 4)) & 0xF;
+		ets->prio_tc[i] = bwg;
+		ets->tc_tx_bw[i] = pcmd.u.dcb.pgrate.pgrate[i];
+		ets->tc_rx_bw[i] = ets->tc_tx_bw[i];
+		ets->tc_tsa[i] = pcmd.u.dcb.pgrate.tsa[i];
+	}
+
+	return 0;
+}
+
+static int cxgb4_ieee_get_ets(struct net_device *dev, struct ieee_ets *ets)
+{
+	return cxgb4_ieee_read_ets(dev, ets, 1);
+}
+
+/* We reuse this for peer PFC as well, as we can't have it enabled one way */
+static int cxgb4_ieee_get_pfc(struct net_device *dev, struct ieee_pfc *pfc)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct port_dcb_info *dcb = &pi->dcb;
+
+	memset(pfc, 0, sizeof(struct ieee_pfc));
+
+	if (!(dcb->msgs & CXGB4_DCB_FW_PFC))
+		return 0;
+
+	pfc->pfc_cap = dcb->pfc_num_tcs_supported;
+	pfc->pfc_en = bitswap_1(dcb->pfcen);
+
+	return 0;
+}
+
+static int cxgb4_ieee_peer_ets(struct net_device *dev, struct ieee_ets *ets)
+{
+	return cxgb4_ieee_read_ets(dev, ets, 0);
+}
+
+/* Fill in the Application User Priority Map associated with the
+ * specified Application.
+ * Priority for IEEE dcb_app is an integer, with 0 being a valid value
+ */
+static int cxgb4_ieee_getapp(struct net_device *dev, struct dcb_app *app)
+{
+	int prio;
+
+	if (!cxgb4_ieee_negotiation_complete(dev, CXGB4_DCB_FW_APP_ID))
+		return -EINVAL;
+	if (!(app->selector && app->protocol))
+		return -EINVAL;
+
+	/* Try querying firmware first, use firmware format */
+	prio = __cxgb4_getapp(dev, app->selector - 1, app->protocol, 0);
+
+	if (prio < 0)
+		prio = dcb_ieee_getapp_mask(dev, app);
+
+	app->priority = ffs(prio) - 1;
+	return 0;
+}
+
+/* Write a new Application User Priority Map for the specified Application ID.
+ * Priority for IEEE dcb_app is an integer, with 0 being a valid value
+ */
+static int cxgb4_ieee_setapp(struct net_device *dev, struct dcb_app *app)
+{
+	int ret;
+
+	if (!cxgb4_ieee_negotiation_complete(dev, CXGB4_DCB_FW_APP_ID))
+		return -EINVAL;
+	if (!(app->selector && app->protocol))
+		return -EINVAL;
+
+	if (!(app->selector > IEEE_8021QAZ_APP_SEL_ETHERTYPE  &&
+	      app->selector < IEEE_8021QAZ_APP_SEL_ANY))
+		return -EINVAL;
+
+	/* change selector to a format that firmware understands */
+	ret = __cxgb4_setapp(dev, app->selector - 1, app->protocol,
+			     (1 << app->priority));
+	if (ret)
+		return ret;
+
+	return dcb_ieee_setapp(dev, app);
+}
+
+/* Return our DCBX parameters.
+ */
+static u8 cxgb4_getdcbx(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+
+	/* This is already set by cxgb4_set_dcb_caps, so just return it */
+	return pi->dcb.supported;
+}
+
+/* Set our DCBX parameters.
+ */
+static u8 cxgb4_setdcbx(struct net_device *dev, u8 dcb_request)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+
+	/* Filter out requests which exceed our capabilities.
+	 */
+	if ((dcb_request & (CXGB4_DCBX_FW_SUPPORT | CXGB4_DCBX_HOST_SUPPORT))
+	    != dcb_request)
+		return 1;
+
+	/* Can't enable DCB if we haven't successfully negotiated it.
+	 */
+	if (!cxgb4_dcb_state_synced(pi->dcb.state))
+		return 1;
+
+	/* There's currently no mechanism to allow for the firmware DCBX
+	 * negotiation to be changed from the Host Driver.  If the caller
+	 * requests exactly the same parameters that we already have then
+	 * we'll allow them to be successfully "set" ...
+	 */
+	if (dcb_request != pi->dcb.supported)
+		return 1;
+
+	pi->dcb.supported = dcb_request;
+	return 0;
+}
+
+static int cxgb4_getpeer_app(struct net_device *dev,
+			     struct dcb_peer_app_info *info, u16 *app_count)
+{
+	struct fw_port_cmd pcmd;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = pi->adapter;
+	int i, err = 0;
+
+	if (!cxgb4_dcb_state_synced(pi->dcb.state))
+		return 1;
+
+	info->willing = 0;
+	info->error = 0;
+
+	*app_count = 0;
+	for (i = 0; i < CXGB4_MAX_DCBX_APP_SUPPORTED; i++) {
+		INIT_PORT_DCB_READ_PEER_CMD(pcmd, pi->port_id);
+		pcmd.u.dcb.app_priority.type = FW_PORT_DCB_TYPE_APP_ID;
+		pcmd.u.dcb.app_priority.idx = *app_count;
+		err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+
+		if (err != FW_PORT_DCB_CFG_SUCCESS) {
+			dev_err(adap->pdev_dev, "DCB app table read failed with %d\n",
+				-err);
+			return err;
+		}
+
+		/* find first empty slot */
+		if (!pcmd.u.dcb.app_priority.protocolid)
+			break;
+	}
+	*app_count = i;
+	return err;
+}
+
+static int cxgb4_getpeerapp_tbl(struct net_device *dev, struct dcb_app *table)
+{
+	struct fw_port_cmd pcmd;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = pi->adapter;
+	int i, err = 0;
+
+	if (!cxgb4_dcb_state_synced(pi->dcb.state))
+		return 1;
+
+	for (i = 0; i < CXGB4_MAX_DCBX_APP_SUPPORTED; i++) {
+		INIT_PORT_DCB_READ_PEER_CMD(pcmd, pi->port_id);
+		pcmd.u.dcb.app_priority.type = FW_PORT_DCB_TYPE_APP_ID;
+		pcmd.u.dcb.app_priority.idx = i;
+		err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+
+		if (err != FW_PORT_DCB_CFG_SUCCESS) {
+			dev_err(adap->pdev_dev, "DCB app table read failed with %d\n",
+				-err);
+			return err;
+		}
+
+		/* find first empty slot */
+		if (!pcmd.u.dcb.app_priority.protocolid)
+			break;
+
+		table[i].selector = (pcmd.u.dcb.app_priority.sel_field + 1);
+		table[i].protocol =
+			be16_to_cpu(pcmd.u.dcb.app_priority.protocolid);
+		table[i].priority =
+			ffs(pcmd.u.dcb.app_priority.user_prio_map) - 1;
+	}
+	return err;
+}
+
+/* Return Priority Group information.
+ */
+static int cxgb4_cee_peer_getpg(struct net_device *dev, struct cee_pg *pg)
+{
+	struct fw_port_cmd pcmd;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = pi->adapter;
+	u32 pgid;
+	int i, err;
+
+	/* We're always "willing" -- the Switch Fabric always dictates the
+	 * DCBX parameters to us.
+	 */
+	pg->willing = true;
+
+	INIT_PORT_DCB_READ_PEER_CMD(pcmd, pi->port_id);
+	pcmd.u.dcb.pgid.type = FW_PORT_DCB_TYPE_PGID;
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB read PGID failed with %d\n", -err);
+		return err;
+	}
+	pgid = be32_to_cpu(pcmd.u.dcb.pgid.pgid);
+
+	for (i = 0; i < CXGB4_MAX_PRIORITY; i++)
+		pg->prio_pg[7 - i] = (pgid >> (i * 4)) & 0xF;
+
+	INIT_PORT_DCB_READ_PEER_CMD(pcmd, pi->port_id);
+	pcmd.u.dcb.pgrate.type = FW_PORT_DCB_TYPE_PGRATE;
+	err = t4_wr_mbox(adap, adap->mbox, &pcmd, sizeof(pcmd), &pcmd);
+	if (err != FW_PORT_DCB_CFG_SUCCESS) {
+		dev_err(adap->pdev_dev, "DCB read PGRATE failed with %d\n",
+			-err);
+		return err;
+	}
+
+	for (i = 0; i < CXGB4_MAX_PRIORITY; i++)
+		pg->pg_bw[i] = pcmd.u.dcb.pgrate.pgrate[i];
+
+	pg->tcs_supported = pcmd.u.dcb.pgrate.num_tcs_supported;
+
+	return 0;
+}
+
+/* Return Priority Flow Control information.
+ */
+static int cxgb4_cee_peer_getpfc(struct net_device *dev, struct cee_pfc *pfc)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+
+	cxgb4_getnumtcs(dev, DCB_NUMTCS_ATTR_PFC, &(pfc->tcs_supported));
+
+	/* Firmware sends this to us in a formwat that is a bit flipped version
+	 * of spec, correct it before we send it to host. This is taken care of
+	 * by bit shifting in other uses of pfcen
+	 */
+	pfc->pfc_en = bitswap_1(pi->dcb.pfcen);
+
+	pfc->tcs_supported = pi->dcb.pfc_num_tcs_supported;
+
+	return 0;
+}
+
+const struct dcbnl_rtnl_ops cxgb4_dcb_ops = {
+	.ieee_getets		= cxgb4_ieee_get_ets,
+	.ieee_getpfc		= cxgb4_ieee_get_pfc,
+	.ieee_getapp		= cxgb4_ieee_getapp,
+	.ieee_setapp		= cxgb4_ieee_setapp,
+	.ieee_peer_getets	= cxgb4_ieee_peer_ets,
+	.ieee_peer_getpfc	= cxgb4_ieee_get_pfc,
+
+	/* CEE std */
+	.getstate		= cxgb4_getstate,
+	.setstate		= cxgb4_setstate,
+	.getpgtccfgtx		= cxgb4_getpgtccfg_tx,
+	.getpgbwgcfgtx		= cxgb4_getpgbwgcfg_tx,
+	.getpgtccfgrx		= cxgb4_getpgtccfg_rx,
+	.getpgbwgcfgrx		= cxgb4_getpgbwgcfg_rx,
+	.setpgtccfgtx		= cxgb4_setpgtccfg_tx,
+	.setpgbwgcfgtx		= cxgb4_setpgbwgcfg_tx,
+	.setpfccfg		= cxgb4_setpfccfg,
+	.getpfccfg		= cxgb4_getpfccfg,
+	.setall			= cxgb4_setall,
+	.getcap			= cxgb4_getcap,
+	.getnumtcs		= cxgb4_getnumtcs,
+	.setnumtcs		= cxgb4_setnumtcs,
+	.getpfcstate		= cxgb4_getpfcstate,
+	.setpfcstate		= cxgb4_setpfcstate,
+	.getapp			= cxgb4_getapp,
+	.setapp			= cxgb4_setapp,
+
+	/* DCBX configuration */
+	.getdcbx		= cxgb4_getdcbx,
+	.setdcbx		= cxgb4_setdcbx,
+
+	/* peer apps */
+	.peer_getappinfo	= cxgb4_getpeer_app,
+	.peer_getapptable	= cxgb4_getpeerapp_tbl,
+
+	/* CEE peer */
+	.cee_peer_getpg		= cxgb4_cee_peer_getpg,
+	.cee_peer_getpfc	= cxgb4_cee_peer_getpfc,
+};
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_dcb.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_dcb.h
new file mode 100644
index 000000000..80c6627fe
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_dcb.h
@@ -0,0 +1,154 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ *  Copyright (C) 2013-2014 Chelsio Communications.  All rights reserved.
+ *
+ *  Written by Anish Bhatt (anish@chelsio.com)
+ */
+
+#ifndef __CXGB4_DCB_H
+#define __CXGB4_DCB_H
+
+#include <linux/netdevice.h>
+#include <linux/dcbnl.h>
+#include <net/dcbnl.h>
+
+#ifdef CONFIG_CHELSIO_T4_DCB
+
+#define CXGB4_DCBX_FW_SUPPORT \
+	(DCB_CAP_DCBX_VER_CEE | \
+	 DCB_CAP_DCBX_VER_IEEE | \
+	 DCB_CAP_DCBX_LLD_MANAGED)
+#define CXGB4_DCBX_HOST_SUPPORT \
+	(DCB_CAP_DCBX_VER_CEE | \
+	 DCB_CAP_DCBX_VER_IEEE | \
+	 DCB_CAP_DCBX_HOST)
+
+#define CXGB4_MAX_PRIORITY      CXGB4_MAX_DCBX_APP_SUPPORTED
+#define CXGB4_MAX_TCS           CXGB4_MAX_DCBX_APP_SUPPORTED
+
+#define INIT_PORT_DCB_CMD(__pcmd, __port, __op, __action) \
+	do { \
+		memset(&(__pcmd), 0, sizeof(__pcmd)); \
+		(__pcmd).op_to_portid = \
+			cpu_to_be32(FW_CMD_OP_V(FW_PORT_CMD) | \
+				    FW_CMD_REQUEST_F | \
+				    FW_CMD_##__op##_F | \
+				    FW_PORT_CMD_PORTID_V(__port)); \
+		(__pcmd).action_to_len16 = \
+			cpu_to_be32(FW_PORT_CMD_ACTION_V(__action) | \
+				    FW_LEN16(pcmd)); \
+	} while (0)
+
+#define INIT_PORT_DCB_READ_PEER_CMD(__pcmd, __port) \
+	INIT_PORT_DCB_CMD(__pcmd, __port, READ, FW_PORT_ACTION_DCB_READ_RECV)
+
+#define INIT_PORT_DCB_READ_LOCAL_CMD(__pcmd, __port) \
+	INIT_PORT_DCB_CMD(__pcmd, __port, READ, FW_PORT_ACTION_DCB_READ_TRANS)
+
+#define INIT_PORT_DCB_READ_SYNC_CMD(__pcmd, __port) \
+	INIT_PORT_DCB_CMD(__pcmd, __port, READ, FW_PORT_ACTION_DCB_READ_DET)
+
+#define INIT_PORT_DCB_WRITE_CMD(__pcmd, __port) \
+	INIT_PORT_DCB_CMD(__pcmd, __port, EXEC, FW_PORT_ACTION_L2_DCB_CFG)
+
+#define IEEE_FAUX_SYNC(__dev, __dcb) \
+	do { \
+		if ((__dcb)->dcb_version == FW_PORT_DCB_VER_IEEE) \
+			cxgb4_dcb_state_fsm((__dev), \
+					    CXGB4_DCB_INPUT_FW_ALLSYNCED); \
+	} while (0)
+
+/* States we can be in for a port's Data Center Bridging.
+ */
+enum cxgb4_dcb_state {
+	CXGB4_DCB_STATE_START,		/* initial unknown state */
+	CXGB4_DCB_STATE_HOST,		/* we're using Host DCB (if at all) */
+	CXGB4_DCB_STATE_FW_INCOMPLETE,	/* using firmware DCB, incomplete */
+	CXGB4_DCB_STATE_FW_ALLSYNCED,	/* using firmware DCB, all sync'ed */
+};
+
+/* Data Center Bridging state input for the Finite State Machine.
+ */
+enum cxgb4_dcb_state_input {
+	/* Input from the firmware.
+	 */
+	CXGB4_DCB_INPUT_FW_DISABLED,	/* firmware DCB disabled */
+	CXGB4_DCB_INPUT_FW_ENABLED,	/* firmware DCB enabled */
+	CXGB4_DCB_INPUT_FW_INCOMPLETE,	/* firmware reports incomplete DCB */
+	CXGB4_DCB_INPUT_FW_ALLSYNCED,	/* firmware reports all sync'ed */
+
+};
+
+/* Firmware DCB messages that we've received so far ...
+ */
+enum cxgb4_dcb_fw_msgs {
+	CXGB4_DCB_FW_PGID	= 0x01,
+	CXGB4_DCB_FW_PGRATE	= 0x02,
+	CXGB4_DCB_FW_PRIORATE	= 0x04,
+	CXGB4_DCB_FW_PFC	= 0x08,
+	CXGB4_DCB_FW_APP_ID	= 0x10,
+};
+
+#define CXGB4_MAX_DCBX_APP_SUPPORTED 8
+
+/* Data Center Bridging support;
+ */
+struct port_dcb_info {
+	enum cxgb4_dcb_state state;	/* DCB State Machine */
+	enum cxgb4_dcb_fw_msgs msgs;	/* DCB Firmware messages received */
+	unsigned int supported;		/* OS DCB capabilities supported */
+	bool enabled;			/* OS Enabled state */
+
+	/* Cached copies of DCB information sent by the firmware (in Host
+	 * Native Endian format).
+	 */
+	u32	pgid;			/* Priority Group[0..7] */
+	u8	dcb_version;		/* Running DCBx version */
+	u8	pfcen;			/* Priority Flow Control[0..7] */
+	u8	pg_num_tcs_supported;	/* max PG Traffic Classes */
+	u8	pfc_num_tcs_supported;	/* max PFC Traffic Classes */
+	u8	pgrate[8];		/* Priority Group Rate[0..7] */
+	u8	priorate[8];		/* Priority Rate[0..7] */
+	u8	tsa[8];			/* TSA Algorithm[0..7] */
+	struct app_priority { /* Application Information */
+		u8	user_prio_map;	/* Priority Map bitfield */
+		u8	sel_field;	/* Protocol ID interpretation */
+		u16	protocolid;	/* Protocol ID */
+	} app_priority[CXGB4_MAX_DCBX_APP_SUPPORTED];
+};
+
+void cxgb4_dcb_state_init(struct net_device *);
+void cxgb4_dcb_version_init(struct net_device *);
+void cxgb4_dcb_reset(struct net_device *dev);
+void cxgb4_dcb_state_fsm(struct net_device *, enum cxgb4_dcb_state_input);
+void cxgb4_dcb_handle_fw_update(struct adapter *, const struct fw_port_cmd *);
+void cxgb4_dcb_set_caps(struct adapter *, const struct fw_port_cmd *);
+extern const struct dcbnl_rtnl_ops cxgb4_dcb_ops;
+
+static inline __u8 bitswap_1(unsigned char val)
+{
+	return ((val & 0x80) >> 7) |
+	       ((val & 0x40) >> 5) |
+	       ((val & 0x20) >> 3) |
+	       ((val & 0x10) >> 1) |
+	       ((val & 0x08) << 1) |
+	       ((val & 0x04) << 3) |
+	       ((val & 0x02) << 5) |
+	       ((val & 0x01) << 7);
+}
+
+extern const char * const dcb_ver_array[];
+
+#define CXGB4_DCB_ENABLED true
+
+#else /* !CONFIG_CHELSIO_T4_DCB */
+
+static inline void cxgb4_dcb_state_init(struct net_device *dev)
+{
+}
+
+#define CXGB4_DCB_ENABLED false
+
+#endif /* !CONFIG_CHELSIO_T4_DCB */
+
+#endif /* __CXGB4_DCB_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
new file mode 100644
index 000000000..14e0d989c
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
@@ -0,0 +1,3893 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/seq_file.h>
+#include <linux/debugfs.h>
+#include <linux/string_helpers.h>
+#include <linux/sort.h>
+#include <linux/ctype.h>
+
+#include "cxgb4.h"
+#include "t4_regs.h"
+#include "t4_values.h"
+#include "t4fw_api.h"
+#include "cxgb4_debugfs.h"
+#include "clip_tbl.h"
+#include "l2t.h"
+#include "cudbg_if.h"
+#include "cudbg_lib_common.h"
+#include "cudbg_entity.h"
+#include "cudbg_lib.h"
+#include "cxgb4_tc_mqprio.h"
+
+/* generic seq_file support for showing a table of size rows x width. */
+static void *seq_tab_get_idx(struct seq_tab *tb, loff_t pos)
+{
+	pos -= tb->skip_first;
+	return pos >= tb->rows ? NULL : &tb->data[pos * tb->width];
+}
+
+static void *seq_tab_start(struct seq_file *seq, loff_t *pos)
+{
+	struct seq_tab *tb = seq->private;
+
+	if (tb->skip_first && *pos == 0)
+		return SEQ_START_TOKEN;
+
+	return seq_tab_get_idx(tb, *pos);
+}
+
+static void *seq_tab_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+	v = seq_tab_get_idx(seq->private, *pos + 1);
+	++(*pos);
+	return v;
+}
+
+static void seq_tab_stop(struct seq_file *seq, void *v)
+{
+}
+
+static int seq_tab_show(struct seq_file *seq, void *v)
+{
+	const struct seq_tab *tb = seq->private;
+
+	return tb->show(seq, v, ((char *)v - tb->data) / tb->width);
+}
+
+static const struct seq_operations seq_tab_ops = {
+	.start = seq_tab_start,
+	.next  = seq_tab_next,
+	.stop  = seq_tab_stop,
+	.show  = seq_tab_show
+};
+
+struct seq_tab *seq_open_tab(struct file *f, unsigned int rows,
+			     unsigned int width, unsigned int have_header,
+			     int (*show)(struct seq_file *seq, void *v, int i))
+{
+	struct seq_tab *p;
+
+	p = __seq_open_private(f, &seq_tab_ops, sizeof(*p) + rows * width);
+	if (p) {
+		p->show = show;
+		p->rows = rows;
+		p->width = width;
+		p->skip_first = have_header != 0;
+	}
+	return p;
+}
+
+/* Trim the size of a seq_tab to the supplied number of rows.  The operation is
+ * irreversible.
+ */
+static int seq_tab_trim(struct seq_tab *p, unsigned int new_rows)
+{
+	if (new_rows > p->rows)
+		return -EINVAL;
+	p->rows = new_rows;
+	return 0;
+}
+
+static int cim_la_show(struct seq_file *seq, void *v, int idx)
+{
+	if (v == SEQ_START_TOKEN)
+		seq_puts(seq, "Status   Data      PC     LS0Stat  LS0Addr "
+			 "            LS0Data\n");
+	else {
+		const u32 *p = v;
+
+		seq_printf(seq,
+			   "  %02x  %x%07x %x%07x %08x %08x %08x%08x%08x%08x\n",
+			   (p[0] >> 4) & 0xff, p[0] & 0xf, p[1] >> 4,
+			   p[1] & 0xf, p[2] >> 4, p[2] & 0xf, p[3], p[4], p[5],
+			   p[6], p[7]);
+	}
+	return 0;
+}
+
+static int cim_la_show_3in1(struct seq_file *seq, void *v, int idx)
+{
+	if (v == SEQ_START_TOKEN) {
+		seq_puts(seq, "Status   Data      PC\n");
+	} else {
+		const u32 *p = v;
+
+		seq_printf(seq, "  %02x   %08x %08x\n", p[5] & 0xff, p[6],
+			   p[7]);
+		seq_printf(seq, "  %02x   %02x%06x %02x%06x\n",
+			   (p[3] >> 8) & 0xff, p[3] & 0xff, p[4] >> 8,
+			   p[4] & 0xff, p[5] >> 8);
+		seq_printf(seq, "  %02x   %x%07x %x%07x\n", (p[0] >> 4) & 0xff,
+			   p[0] & 0xf, p[1] >> 4, p[1] & 0xf, p[2] >> 4);
+	}
+	return 0;
+}
+
+static int cim_la_show_t6(struct seq_file *seq, void *v, int idx)
+{
+	if (v == SEQ_START_TOKEN) {
+		seq_puts(seq, "Status   Inst    Data      PC     LS0Stat  "
+			 "LS0Addr  LS0Data  LS1Stat  LS1Addr  LS1Data\n");
+	} else {
+		const u32 *p = v;
+
+		seq_printf(seq, "  %02x   %04x%04x %04x%04x %04x%04x %08x %08x %08x %08x %08x %08x\n",
+			   (p[9] >> 16) & 0xff,       /* Status */
+			   p[9] & 0xffff, p[8] >> 16, /* Inst */
+			   p[8] & 0xffff, p[7] >> 16, /* Data */
+			   p[7] & 0xffff, p[6] >> 16, /* PC */
+			   p[2], p[1], p[0],      /* LS0 Stat, Addr and Data */
+			   p[5], p[4], p[3]);     /* LS1 Stat, Addr and Data */
+	}
+	return 0;
+}
+
+static int cim_la_show_pc_t6(struct seq_file *seq, void *v, int idx)
+{
+	if (v == SEQ_START_TOKEN) {
+		seq_puts(seq, "Status   Inst    Data      PC\n");
+	} else {
+		const u32 *p = v;
+
+		seq_printf(seq, "  %02x   %08x %08x %08x\n",
+			   p[3] & 0xff, p[2], p[1], p[0]);
+		seq_printf(seq, "  %02x   %02x%06x %02x%06x %02x%06x\n",
+			   (p[6] >> 8) & 0xff, p[6] & 0xff, p[5] >> 8,
+			   p[5] & 0xff, p[4] >> 8, p[4] & 0xff, p[3] >> 8);
+		seq_printf(seq, "  %02x   %04x%04x %04x%04x %04x%04x\n",
+			   (p[9] >> 16) & 0xff, p[9] & 0xffff, p[8] >> 16,
+			   p[8] & 0xffff, p[7] >> 16, p[7] & 0xffff,
+			   p[6] >> 16);
+	}
+	return 0;
+}
+
+static int cim_la_open(struct inode *inode, struct file *file)
+{
+	int ret;
+	unsigned int cfg;
+	struct seq_tab *p;
+	struct adapter *adap = inode->i_private;
+
+	ret = t4_cim_read(adap, UP_UP_DBG_LA_CFG_A, 1, &cfg);
+	if (ret)
+		return ret;
+
+	if (is_t6(adap->params.chip)) {
+		/* +1 to account for integer division of CIMLA_SIZE/10 */
+		p = seq_open_tab(file, (adap->params.cim_la_size / 10) + 1,
+				 10 * sizeof(u32), 1,
+				 cfg & UPDBGLACAPTPCONLY_F ?
+					cim_la_show_pc_t6 : cim_la_show_t6);
+	} else {
+		p = seq_open_tab(file, adap->params.cim_la_size / 8,
+				 8 * sizeof(u32), 1,
+				 cfg & UPDBGLACAPTPCONLY_F ? cim_la_show_3in1 :
+							     cim_la_show);
+	}
+	if (!p)
+		return -ENOMEM;
+
+	ret = t4_cim_read_la(adap, (u32 *)p->data, NULL);
+	if (ret)
+		seq_release_private(inode, file);
+	return ret;
+}
+
+static const struct file_operations cim_la_fops = {
+	.owner   = THIS_MODULE,
+	.open    = cim_la_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private
+};
+
+static int cim_pif_la_show(struct seq_file *seq, void *v, int idx)
+{
+	const u32 *p = v;
+
+	if (v == SEQ_START_TOKEN) {
+		seq_puts(seq, "Cntl ID DataBE   Addr                 Data\n");
+	} else if (idx < CIM_PIFLA_SIZE) {
+		seq_printf(seq, " %02x  %02x  %04x  %08x %08x%08x%08x%08x\n",
+			   (p[5] >> 22) & 0xff, (p[5] >> 16) & 0x3f,
+			   p[5] & 0xffff, p[4], p[3], p[2], p[1], p[0]);
+	} else {
+		if (idx == CIM_PIFLA_SIZE)
+			seq_puts(seq, "\nCntl ID               Data\n");
+		seq_printf(seq, " %02x  %02x %08x%08x%08x%08x\n",
+			   (p[4] >> 6) & 0xff, p[4] & 0x3f,
+			   p[3], p[2], p[1], p[0]);
+	}
+	return 0;
+}
+
+static int cim_pif_la_open(struct inode *inode, struct file *file)
+{
+	struct seq_tab *p;
+	struct adapter *adap = inode->i_private;
+
+	p = seq_open_tab(file, 2 * CIM_PIFLA_SIZE, 6 * sizeof(u32), 1,
+			 cim_pif_la_show);
+	if (!p)
+		return -ENOMEM;
+
+	t4_cim_read_pif_la(adap, (u32 *)p->data,
+			   (u32 *)p->data + 6 * CIM_PIFLA_SIZE, NULL, NULL);
+	return 0;
+}
+
+static const struct file_operations cim_pif_la_fops = {
+	.owner   = THIS_MODULE,
+	.open    = cim_pif_la_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private
+};
+
+static int cim_ma_la_show(struct seq_file *seq, void *v, int idx)
+{
+	const u32 *p = v;
+
+	if (v == SEQ_START_TOKEN) {
+		seq_puts(seq, "\n");
+	} else if (idx < CIM_MALA_SIZE) {
+		seq_printf(seq, "%02x%08x%08x%08x%08x\n",
+			   p[4], p[3], p[2], p[1], p[0]);
+	} else {
+		if (idx == CIM_MALA_SIZE)
+			seq_puts(seq,
+				 "\nCnt ID Tag UE       Data       RDY VLD\n");
+		seq_printf(seq, "%3u %2u  %x   %u %08x%08x  %u   %u\n",
+			   (p[2] >> 10) & 0xff, (p[2] >> 7) & 7,
+			   (p[2] >> 3) & 0xf, (p[2] >> 2) & 1,
+			   (p[1] >> 2) | ((p[2] & 3) << 30),
+			   (p[0] >> 2) | ((p[1] & 3) << 30), (p[0] >> 1) & 1,
+			   p[0] & 1);
+	}
+	return 0;
+}
+
+static int cim_ma_la_open(struct inode *inode, struct file *file)
+{
+	struct seq_tab *p;
+	struct adapter *adap = inode->i_private;
+
+	p = seq_open_tab(file, 2 * CIM_MALA_SIZE, 5 * sizeof(u32), 1,
+			 cim_ma_la_show);
+	if (!p)
+		return -ENOMEM;
+
+	t4_cim_read_ma_la(adap, (u32 *)p->data,
+			  (u32 *)p->data + 5 * CIM_MALA_SIZE);
+	return 0;
+}
+
+static const struct file_operations cim_ma_la_fops = {
+	.owner   = THIS_MODULE,
+	.open    = cim_ma_la_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private
+};
+
+static int cim_qcfg_show(struct seq_file *seq, void *v)
+{
+	static const char * const qname[] = {
+		"TP0", "TP1", "ULP", "SGE0", "SGE1", "NC-SI",
+		"ULP0", "ULP1", "ULP2", "ULP3", "SGE", "NC-SI",
+		"SGE0-RX", "SGE1-RX"
+	};
+
+	int i;
+	struct adapter *adap = seq->private;
+	u16 base[CIM_NUM_IBQ + CIM_NUM_OBQ_T5];
+	u16 size[CIM_NUM_IBQ + CIM_NUM_OBQ_T5];
+	u32 stat[(4 * (CIM_NUM_IBQ + CIM_NUM_OBQ_T5))];
+	u16 thres[CIM_NUM_IBQ];
+	u32 obq_wr_t4[2 * CIM_NUM_OBQ], *wr;
+	u32 obq_wr_t5[2 * CIM_NUM_OBQ_T5];
+	u32 *p = stat;
+	int cim_num_obq = is_t4(adap->params.chip) ?
+				CIM_NUM_OBQ : CIM_NUM_OBQ_T5;
+
+	i = t4_cim_read(adap, is_t4(adap->params.chip) ? UP_IBQ_0_RDADDR_A :
+			UP_IBQ_0_SHADOW_RDADDR_A,
+			ARRAY_SIZE(stat), stat);
+	if (!i) {
+		if (is_t4(adap->params.chip)) {
+			i = t4_cim_read(adap, UP_OBQ_0_REALADDR_A,
+					ARRAY_SIZE(obq_wr_t4), obq_wr_t4);
+			wr = obq_wr_t4;
+		} else {
+			i = t4_cim_read(adap, UP_OBQ_0_SHADOW_REALADDR_A,
+					ARRAY_SIZE(obq_wr_t5), obq_wr_t5);
+			wr = obq_wr_t5;
+		}
+	}
+	if (i)
+		return i;
+
+	t4_read_cimq_cfg(adap, base, size, thres);
+
+	seq_printf(seq,
+		   "  Queue  Base  Size Thres  RdPtr WrPtr  SOP  EOP Avail\n");
+	for (i = 0; i < CIM_NUM_IBQ; i++, p += 4)
+		seq_printf(seq, "%7s %5x %5u %5u %6x  %4x %4u %4u %5u\n",
+			   qname[i], base[i], size[i], thres[i],
+			   IBQRDADDR_G(p[0]), IBQWRADDR_G(p[1]),
+			   QUESOPCNT_G(p[3]), QUEEOPCNT_G(p[3]),
+			   QUEREMFLITS_G(p[2]) * 16);
+	for ( ; i < CIM_NUM_IBQ + cim_num_obq; i++, p += 4, wr += 2)
+		seq_printf(seq, "%7s %5x %5u %12x  %4x %4u %4u %5u\n",
+			   qname[i], base[i], size[i],
+			   QUERDADDR_G(p[0]) & 0x3fff, wr[0] - base[i],
+			   QUESOPCNT_G(p[3]), QUEEOPCNT_G(p[3]),
+			   QUEREMFLITS_G(p[2]) * 16);
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(cim_qcfg);
+
+static int cimq_show(struct seq_file *seq, void *v, int idx)
+{
+	const u32 *p = v;
+
+	seq_printf(seq, "%#06x: %08x %08x %08x %08x\n", idx * 16, p[0], p[1],
+		   p[2], p[3]);
+	return 0;
+}
+
+static int cim_ibq_open(struct inode *inode, struct file *file)
+{
+	int ret;
+	struct seq_tab *p;
+	unsigned int qid = (uintptr_t)inode->i_private & 7;
+	struct adapter *adap = inode->i_private - qid;
+
+	p = seq_open_tab(file, CIM_IBQ_SIZE, 4 * sizeof(u32), 0, cimq_show);
+	if (!p)
+		return -ENOMEM;
+
+	ret = t4_read_cim_ibq(adap, qid, (u32 *)p->data, CIM_IBQ_SIZE * 4);
+	if (ret < 0)
+		seq_release_private(inode, file);
+	else
+		ret = 0;
+	return ret;
+}
+
+static const struct file_operations cim_ibq_fops = {
+	.owner   = THIS_MODULE,
+	.open    = cim_ibq_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private
+};
+
+static int cim_obq_open(struct inode *inode, struct file *file)
+{
+	int ret;
+	struct seq_tab *p;
+	unsigned int qid = (uintptr_t)inode->i_private & 7;
+	struct adapter *adap = inode->i_private - qid;
+
+	p = seq_open_tab(file, 6 * CIM_OBQ_SIZE, 4 * sizeof(u32), 0, cimq_show);
+	if (!p)
+		return -ENOMEM;
+
+	ret = t4_read_cim_obq(adap, qid, (u32 *)p->data, 6 * CIM_OBQ_SIZE * 4);
+	if (ret < 0) {
+		seq_release_private(inode, file);
+	} else {
+		seq_tab_trim(p, ret / 4);
+		ret = 0;
+	}
+	return ret;
+}
+
+static const struct file_operations cim_obq_fops = {
+	.owner   = THIS_MODULE,
+	.open    = cim_obq_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private
+};
+
+struct field_desc {
+	const char *name;
+	unsigned int start;
+	unsigned int width;
+};
+
+static void field_desc_show(struct seq_file *seq, u64 v,
+			    const struct field_desc *p)
+{
+	char buf[32];
+	int line_size = 0;
+
+	while (p->name) {
+		u64 mask = (1ULL << p->width) - 1;
+		int len = scnprintf(buf, sizeof(buf), "%s: %llu", p->name,
+				    ((unsigned long long)v >> p->start) & mask);
+
+		if (line_size + len >= 79) {
+			line_size = 8;
+			seq_puts(seq, "\n        ");
+		}
+		seq_printf(seq, "%s ", buf);
+		line_size += len + 1;
+		p++;
+	}
+	seq_putc(seq, '\n');
+}
+
+static struct field_desc tp_la0[] = {
+	{ "RcfOpCodeOut", 60, 4 },
+	{ "State", 56, 4 },
+	{ "WcfState", 52, 4 },
+	{ "RcfOpcSrcOut", 50, 2 },
+	{ "CRxError", 49, 1 },
+	{ "ERxError", 48, 1 },
+	{ "SanityFailed", 47, 1 },
+	{ "SpuriousMsg", 46, 1 },
+	{ "FlushInputMsg", 45, 1 },
+	{ "FlushInputCpl", 44, 1 },
+	{ "RssUpBit", 43, 1 },
+	{ "RssFilterHit", 42, 1 },
+	{ "Tid", 32, 10 },
+	{ "InitTcb", 31, 1 },
+	{ "LineNumber", 24, 7 },
+	{ "Emsg", 23, 1 },
+	{ "EdataOut", 22, 1 },
+	{ "Cmsg", 21, 1 },
+	{ "CdataOut", 20, 1 },
+	{ "EreadPdu", 19, 1 },
+	{ "CreadPdu", 18, 1 },
+	{ "TunnelPkt", 17, 1 },
+	{ "RcfPeerFin", 16, 1 },
+	{ "RcfReasonOut", 12, 4 },
+	{ "TxCchannel", 10, 2 },
+	{ "RcfTxChannel", 8, 2 },
+	{ "RxEchannel", 6, 2 },
+	{ "RcfRxChannel", 5, 1 },
+	{ "RcfDataOutSrdy", 4, 1 },
+	{ "RxDvld", 3, 1 },
+	{ "RxOoDvld", 2, 1 },
+	{ "RxCongestion", 1, 1 },
+	{ "TxCongestion", 0, 1 },
+	{ NULL }
+};
+
+static int tp_la_show(struct seq_file *seq, void *v, int idx)
+{
+	const u64 *p = v;
+
+	field_desc_show(seq, *p, tp_la0);
+	return 0;
+}
+
+static int tp_la_show2(struct seq_file *seq, void *v, int idx)
+{
+	const u64 *p = v;
+
+	if (idx)
+		seq_putc(seq, '\n');
+	field_desc_show(seq, p[0], tp_la0);
+	if (idx < (TPLA_SIZE / 2 - 1) || p[1] != ~0ULL)
+		field_desc_show(seq, p[1], tp_la0);
+	return 0;
+}
+
+static int tp_la_show3(struct seq_file *seq, void *v, int idx)
+{
+	static struct field_desc tp_la1[] = {
+		{ "CplCmdIn", 56, 8 },
+		{ "CplCmdOut", 48, 8 },
+		{ "ESynOut", 47, 1 },
+		{ "EAckOut", 46, 1 },
+		{ "EFinOut", 45, 1 },
+		{ "ERstOut", 44, 1 },
+		{ "SynIn", 43, 1 },
+		{ "AckIn", 42, 1 },
+		{ "FinIn", 41, 1 },
+		{ "RstIn", 40, 1 },
+		{ "DataIn", 39, 1 },
+		{ "DataInVld", 38, 1 },
+		{ "PadIn", 37, 1 },
+		{ "RxBufEmpty", 36, 1 },
+		{ "RxDdp", 35, 1 },
+		{ "RxFbCongestion", 34, 1 },
+		{ "TxFbCongestion", 33, 1 },
+		{ "TxPktSumSrdy", 32, 1 },
+		{ "RcfUlpType", 28, 4 },
+		{ "Eread", 27, 1 },
+		{ "Ebypass", 26, 1 },
+		{ "Esave", 25, 1 },
+		{ "Static0", 24, 1 },
+		{ "Cread", 23, 1 },
+		{ "Cbypass", 22, 1 },
+		{ "Csave", 21, 1 },
+		{ "CPktOut", 20, 1 },
+		{ "RxPagePoolFull", 18, 2 },
+		{ "RxLpbkPkt", 17, 1 },
+		{ "TxLpbkPkt", 16, 1 },
+		{ "RxVfValid", 15, 1 },
+		{ "SynLearned", 14, 1 },
+		{ "SetDelEntry", 13, 1 },
+		{ "SetInvEntry", 12, 1 },
+		{ "CpcmdDvld", 11, 1 },
+		{ "CpcmdSave", 10, 1 },
+		{ "RxPstructsFull", 8, 2 },
+		{ "EpcmdDvld", 7, 1 },
+		{ "EpcmdFlush", 6, 1 },
+		{ "EpcmdTrimPrefix", 5, 1 },
+		{ "EpcmdTrimPostfix", 4, 1 },
+		{ "ERssIp4Pkt", 3, 1 },
+		{ "ERssIp6Pkt", 2, 1 },
+		{ "ERssTcpUdpPkt", 1, 1 },
+		{ "ERssFceFipPkt", 0, 1 },
+		{ NULL }
+	};
+	static struct field_desc tp_la2[] = {
+		{ "CplCmdIn", 56, 8 },
+		{ "MpsVfVld", 55, 1 },
+		{ "MpsPf", 52, 3 },
+		{ "MpsVf", 44, 8 },
+		{ "SynIn", 43, 1 },
+		{ "AckIn", 42, 1 },
+		{ "FinIn", 41, 1 },
+		{ "RstIn", 40, 1 },
+		{ "DataIn", 39, 1 },
+		{ "DataInVld", 38, 1 },
+		{ "PadIn", 37, 1 },
+		{ "RxBufEmpty", 36, 1 },
+		{ "RxDdp", 35, 1 },
+		{ "RxFbCongestion", 34, 1 },
+		{ "TxFbCongestion", 33, 1 },
+		{ "TxPktSumSrdy", 32, 1 },
+		{ "RcfUlpType", 28, 4 },
+		{ "Eread", 27, 1 },
+		{ "Ebypass", 26, 1 },
+		{ "Esave", 25, 1 },
+		{ "Static0", 24, 1 },
+		{ "Cread", 23, 1 },
+		{ "Cbypass", 22, 1 },
+		{ "Csave", 21, 1 },
+		{ "CPktOut", 20, 1 },
+		{ "RxPagePoolFull", 18, 2 },
+		{ "RxLpbkPkt", 17, 1 },
+		{ "TxLpbkPkt", 16, 1 },
+		{ "RxVfValid", 15, 1 },
+		{ "SynLearned", 14, 1 },
+		{ "SetDelEntry", 13, 1 },
+		{ "SetInvEntry", 12, 1 },
+		{ "CpcmdDvld", 11, 1 },
+		{ "CpcmdSave", 10, 1 },
+		{ "RxPstructsFull", 8, 2 },
+		{ "EpcmdDvld", 7, 1 },
+		{ "EpcmdFlush", 6, 1 },
+		{ "EpcmdTrimPrefix", 5, 1 },
+		{ "EpcmdTrimPostfix", 4, 1 },
+		{ "ERssIp4Pkt", 3, 1 },
+		{ "ERssIp6Pkt", 2, 1 },
+		{ "ERssTcpUdpPkt", 1, 1 },
+		{ "ERssFceFipPkt", 0, 1 },
+		{ NULL }
+	};
+	const u64 *p = v;
+
+	if (idx)
+		seq_putc(seq, '\n');
+	field_desc_show(seq, p[0], tp_la0);
+	if (idx < (TPLA_SIZE / 2 - 1) || p[1] != ~0ULL)
+		field_desc_show(seq, p[1], (p[0] & BIT(17)) ? tp_la2 : tp_la1);
+	return 0;
+}
+
+static int tp_la_open(struct inode *inode, struct file *file)
+{
+	struct seq_tab *p;
+	struct adapter *adap = inode->i_private;
+
+	switch (DBGLAMODE_G(t4_read_reg(adap, TP_DBG_LA_CONFIG_A))) {
+	case 2:
+		p = seq_open_tab(file, TPLA_SIZE / 2, 2 * sizeof(u64), 0,
+				 tp_la_show2);
+		break;
+	case 3:
+		p = seq_open_tab(file, TPLA_SIZE / 2, 2 * sizeof(u64), 0,
+				 tp_la_show3);
+		break;
+	default:
+		p = seq_open_tab(file, TPLA_SIZE, sizeof(u64), 0, tp_la_show);
+	}
+	if (!p)
+		return -ENOMEM;
+
+	t4_tp_read_la(adap, (u64 *)p->data, NULL);
+	return 0;
+}
+
+static ssize_t tp_la_write(struct file *file, const char __user *buf,
+			   size_t count, loff_t *pos)
+{
+	int err;
+	char s[32];
+	unsigned long val;
+	size_t size = min(sizeof(s) - 1, count);
+	struct adapter *adap = file_inode(file)->i_private;
+
+	if (copy_from_user(s, buf, size))
+		return -EFAULT;
+	s[size] = '\0';
+	err = kstrtoul(s, 0, &val);
+	if (err)
+		return err;
+	if (val > 0xffff)
+		return -EINVAL;
+	adap->params.tp.la_mask = val << 16;
+	t4_set_reg_field(adap, TP_DBG_LA_CONFIG_A, 0xffff0000U,
+			 adap->params.tp.la_mask);
+	return count;
+}
+
+static const struct file_operations tp_la_fops = {
+	.owner   = THIS_MODULE,
+	.open    = tp_la_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private,
+	.write   = tp_la_write
+};
+
+static int ulprx_la_show(struct seq_file *seq, void *v, int idx)
+{
+	const u32 *p = v;
+
+	if (v == SEQ_START_TOKEN)
+		seq_puts(seq, "      Pcmd        Type   Message"
+			 "                Data\n");
+	else
+		seq_printf(seq, "%08x%08x  %4x  %08x  %08x%08x%08x%08x\n",
+			   p[1], p[0], p[2], p[3], p[7], p[6], p[5], p[4]);
+	return 0;
+}
+
+static int ulprx_la_open(struct inode *inode, struct file *file)
+{
+	struct seq_tab *p;
+	struct adapter *adap = inode->i_private;
+
+	p = seq_open_tab(file, ULPRX_LA_SIZE, 8 * sizeof(u32), 1,
+			 ulprx_la_show);
+	if (!p)
+		return -ENOMEM;
+
+	t4_ulprx_read_la(adap, (u32 *)p->data);
+	return 0;
+}
+
+static const struct file_operations ulprx_la_fops = {
+	.owner   = THIS_MODULE,
+	.open    = ulprx_la_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private
+};
+
+/* Show the PM memory stats.  These stats include:
+ *
+ * TX:
+ *   Read: memory read operation
+ *   Write Bypass: cut-through
+ *   Bypass + mem: cut-through and save copy
+ *
+ * RX:
+ *   Read: memory read
+ *   Write Bypass: cut-through
+ *   Flush: payload trim or drop
+ */
+static int pm_stats_show(struct seq_file *seq, void *v)
+{
+	static const char * const tx_pm_stats[] = {
+		"Read:", "Write bypass:", "Write mem:", "Bypass + mem:"
+	};
+	static const char * const rx_pm_stats[] = {
+		"Read:", "Write bypass:", "Write mem:", "Flush:"
+	};
+
+	int i;
+	u32 tx_cnt[T6_PM_NSTATS], rx_cnt[T6_PM_NSTATS];
+	u64 tx_cyc[T6_PM_NSTATS], rx_cyc[T6_PM_NSTATS];
+	struct adapter *adap = seq->private;
+
+	t4_pmtx_get_stats(adap, tx_cnt, tx_cyc);
+	t4_pmrx_get_stats(adap, rx_cnt, rx_cyc);
+
+	seq_printf(seq, "%13s %10s  %20s\n", " ", "Tx pcmds", "Tx bytes");
+	for (i = 0; i < PM_NSTATS - 1; i++)
+		seq_printf(seq, "%-13s %10u  %20llu\n",
+			   tx_pm_stats[i], tx_cnt[i], tx_cyc[i]);
+
+	seq_printf(seq, "%13s %10s  %20s\n", " ", "Rx pcmds", "Rx bytes");
+	for (i = 0; i < PM_NSTATS - 1; i++)
+		seq_printf(seq, "%-13s %10u  %20llu\n",
+			   rx_pm_stats[i], rx_cnt[i], rx_cyc[i]);
+
+	if (CHELSIO_CHIP_VERSION(adap->params.chip) > CHELSIO_T5) {
+		/* In T5 the granularity of the total wait is too fine.
+		 * It is not useful as it reaches the max value too fast.
+		 * Hence display this Input FIFO wait for T6 onwards.
+		 */
+		seq_printf(seq, "%13s %10s  %20s\n",
+			   " ", "Total wait", "Total Occupancy");
+		seq_printf(seq, "Tx FIFO wait  %10u  %20llu\n",
+			   tx_cnt[i], tx_cyc[i]);
+		seq_printf(seq, "Rx FIFO wait  %10u  %20llu\n",
+			   rx_cnt[i], rx_cyc[i]);
+
+		/* Skip index 6 as there is nothing useful ihere */
+		i += 2;
+
+		/* At index 7, a new stat for read latency (count, total wait)
+		 * is added.
+		 */
+		seq_printf(seq, "%13s %10s  %20s\n",
+			   " ", "Reads", "Total wait");
+		seq_printf(seq, "Tx latency    %10u  %20llu\n",
+			   tx_cnt[i], tx_cyc[i]);
+		seq_printf(seq, "Rx latency    %10u  %20llu\n",
+			   rx_cnt[i], rx_cyc[i]);
+	}
+	return 0;
+}
+
+static int pm_stats_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, pm_stats_show, inode->i_private);
+}
+
+static ssize_t pm_stats_clear(struct file *file, const char __user *buf,
+			      size_t count, loff_t *pos)
+{
+	struct adapter *adap = file_inode(file)->i_private;
+
+	t4_write_reg(adap, PM_RX_STAT_CONFIG_A, 0);
+	t4_write_reg(adap, PM_TX_STAT_CONFIG_A, 0);
+	return count;
+}
+
+static const struct file_operations pm_stats_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = pm_stats_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = single_release,
+	.write   = pm_stats_clear
+};
+
+static int tx_rate_show(struct seq_file *seq, void *v)
+{
+	u64 nrate[NCHAN], orate[NCHAN];
+	struct adapter *adap = seq->private;
+
+	t4_get_chan_txrate(adap, nrate, orate);
+	if (adap->params.arch.nchan == NCHAN) {
+		seq_puts(seq, "              channel 0   channel 1   "
+			 "channel 2   channel 3\n");
+		seq_printf(seq, "NIC B/s:     %10llu  %10llu  %10llu  %10llu\n",
+			   (unsigned long long)nrate[0],
+			   (unsigned long long)nrate[1],
+			   (unsigned long long)nrate[2],
+			   (unsigned long long)nrate[3]);
+		seq_printf(seq, "Offload B/s: %10llu  %10llu  %10llu  %10llu\n",
+			   (unsigned long long)orate[0],
+			   (unsigned long long)orate[1],
+			   (unsigned long long)orate[2],
+			   (unsigned long long)orate[3]);
+	} else {
+		seq_puts(seq, "              channel 0   channel 1\n");
+		seq_printf(seq, "NIC B/s:     %10llu  %10llu\n",
+			   (unsigned long long)nrate[0],
+			   (unsigned long long)nrate[1]);
+		seq_printf(seq, "Offload B/s: %10llu  %10llu\n",
+			   (unsigned long long)orate[0],
+			   (unsigned long long)orate[1]);
+	}
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(tx_rate);
+
+static int cctrl_tbl_show(struct seq_file *seq, void *v)
+{
+	static const char * const dec_fac[] = {
+		"0.5", "0.5625", "0.625", "0.6875", "0.75", "0.8125", "0.875",
+		"0.9375" };
+
+	int i;
+	u16 (*incr)[NCCTRL_WIN];
+	struct adapter *adap = seq->private;
+
+	incr = kmalloc_array(NMTUS, sizeof(*incr), GFP_KERNEL);
+	if (!incr)
+		return -ENOMEM;
+
+	t4_read_cong_tbl(adap, incr);
+
+	for (i = 0; i < NCCTRL_WIN; ++i) {
+		seq_printf(seq, "%2d: %4u %4u %4u %4u %4u %4u %4u %4u\n", i,
+			   incr[0][i], incr[1][i], incr[2][i], incr[3][i],
+			   incr[4][i], incr[5][i], incr[6][i], incr[7][i]);
+		seq_printf(seq, "%8u %4u %4u %4u %4u %4u %4u %4u %5u %s\n",
+			   incr[8][i], incr[9][i], incr[10][i], incr[11][i],
+			   incr[12][i], incr[13][i], incr[14][i], incr[15][i],
+			   adap->params.a_wnd[i],
+			   dec_fac[adap->params.b_wnd[i]]);
+	}
+
+	kfree(incr);
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(cctrl_tbl);
+
+/* Format a value in a unit that differs from the value's native unit by the
+ * given factor.
+ */
+static char *unit_conv(char *buf, size_t len, unsigned int val,
+		       unsigned int factor)
+{
+	unsigned int rem = val % factor;
+
+	if (rem == 0) {
+		snprintf(buf, len, "%u", val / factor);
+	} else {
+		while (rem % 10 == 0)
+			rem /= 10;
+		snprintf(buf, len, "%u.%u", val / factor, rem);
+	}
+	return buf;
+}
+
+static int clk_show(struct seq_file *seq, void *v)
+{
+	char buf[32];
+	struct adapter *adap = seq->private;
+	unsigned int cclk_ps = 1000000000 / adap->params.vpd.cclk;  /* in ps */
+	u32 res = t4_read_reg(adap, TP_TIMER_RESOLUTION_A);
+	unsigned int tre = TIMERRESOLUTION_G(res);
+	unsigned int dack_re = DELAYEDACKRESOLUTION_G(res);
+	unsigned long long tp_tick_us = (cclk_ps << tre) / 1000000; /* in us */
+
+	seq_printf(seq, "Core clock period: %s ns\n",
+		   unit_conv(buf, sizeof(buf), cclk_ps, 1000));
+	seq_printf(seq, "TP timer tick: %s us\n",
+		   unit_conv(buf, sizeof(buf), (cclk_ps << tre), 1000000));
+	seq_printf(seq, "TCP timestamp tick: %s us\n",
+		   unit_conv(buf, sizeof(buf),
+			     (cclk_ps << TIMESTAMPRESOLUTION_G(res)), 1000000));
+	seq_printf(seq, "DACK tick: %s us\n",
+		   unit_conv(buf, sizeof(buf), (cclk_ps << dack_re), 1000000));
+	seq_printf(seq, "DACK timer: %u us\n",
+		   ((cclk_ps << dack_re) / 1000000) *
+		   t4_read_reg(adap, TP_DACK_TIMER_A));
+	seq_printf(seq, "Retransmit min: %llu us\n",
+		   tp_tick_us * t4_read_reg(adap, TP_RXT_MIN_A));
+	seq_printf(seq, "Retransmit max: %llu us\n",
+		   tp_tick_us * t4_read_reg(adap, TP_RXT_MAX_A));
+	seq_printf(seq, "Persist timer min: %llu us\n",
+		   tp_tick_us * t4_read_reg(adap, TP_PERS_MIN_A));
+	seq_printf(seq, "Persist timer max: %llu us\n",
+		   tp_tick_us * t4_read_reg(adap, TP_PERS_MAX_A));
+	seq_printf(seq, "Keepalive idle timer: %llu us\n",
+		   tp_tick_us * t4_read_reg(adap, TP_KEEP_IDLE_A));
+	seq_printf(seq, "Keepalive interval: %llu us\n",
+		   tp_tick_us * t4_read_reg(adap, TP_KEEP_INTVL_A));
+	seq_printf(seq, "Initial SRTT: %llu us\n",
+		   tp_tick_us * INITSRTT_G(t4_read_reg(adap, TP_INIT_SRTT_A)));
+	seq_printf(seq, "FINWAIT2 timer: %llu us\n",
+		   tp_tick_us * t4_read_reg(adap, TP_FINWAIT2_TIMER_A));
+
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(clk);
+
+/* Firmware Device Log dump. */
+static const char * const devlog_level_strings[] = {
+	[FW_DEVLOG_LEVEL_EMERG]		= "EMERG",
+	[FW_DEVLOG_LEVEL_CRIT]		= "CRIT",
+	[FW_DEVLOG_LEVEL_ERR]		= "ERR",
+	[FW_DEVLOG_LEVEL_NOTICE]	= "NOTICE",
+	[FW_DEVLOG_LEVEL_INFO]		= "INFO",
+	[FW_DEVLOG_LEVEL_DEBUG]		= "DEBUG"
+};
+
+static const char * const devlog_facility_strings[] = {
+	[FW_DEVLOG_FACILITY_CORE]	= "CORE",
+	[FW_DEVLOG_FACILITY_CF]         = "CF",
+	[FW_DEVLOG_FACILITY_SCHED]	= "SCHED",
+	[FW_DEVLOG_FACILITY_TIMER]	= "TIMER",
+	[FW_DEVLOG_FACILITY_RES]	= "RES",
+	[FW_DEVLOG_FACILITY_HW]		= "HW",
+	[FW_DEVLOG_FACILITY_FLR]	= "FLR",
+	[FW_DEVLOG_FACILITY_DMAQ]	= "DMAQ",
+	[FW_DEVLOG_FACILITY_PHY]	= "PHY",
+	[FW_DEVLOG_FACILITY_MAC]	= "MAC",
+	[FW_DEVLOG_FACILITY_PORT]	= "PORT",
+	[FW_DEVLOG_FACILITY_VI]		= "VI",
+	[FW_DEVLOG_FACILITY_FILTER]	= "FILTER",
+	[FW_DEVLOG_FACILITY_ACL]	= "ACL",
+	[FW_DEVLOG_FACILITY_TM]		= "TM",
+	[FW_DEVLOG_FACILITY_QFC]	= "QFC",
+	[FW_DEVLOG_FACILITY_DCB]	= "DCB",
+	[FW_DEVLOG_FACILITY_ETH]	= "ETH",
+	[FW_DEVLOG_FACILITY_OFLD]	= "OFLD",
+	[FW_DEVLOG_FACILITY_RI]		= "RI",
+	[FW_DEVLOG_FACILITY_ISCSI]	= "ISCSI",
+	[FW_DEVLOG_FACILITY_FCOE]	= "FCOE",
+	[FW_DEVLOG_FACILITY_FOISCSI]	= "FOISCSI",
+	[FW_DEVLOG_FACILITY_FOFCOE]	= "FOFCOE"
+};
+
+/* Information gathered by Device Log Open routine for the display routine.
+ */
+struct devlog_info {
+	unsigned int nentries;		/* number of entries in log[] */
+	unsigned int first;		/* first [temporal] entry in log[] */
+	struct fw_devlog_e log[];	/* Firmware Device Log */
+};
+
+/* Dump a Firmaware Device Log entry.
+ */
+static int devlog_show(struct seq_file *seq, void *v)
+{
+	if (v == SEQ_START_TOKEN)
+		seq_printf(seq, "%10s  %15s  %8s  %8s  %s\n",
+			   "Seq#", "Tstamp", "Level", "Facility", "Message");
+	else {
+		struct devlog_info *dinfo = seq->private;
+		int fidx = (uintptr_t)v - 2;
+		unsigned long index;
+		struct fw_devlog_e *e;
+
+		/* Get a pointer to the log entry to display.  Skip unused log
+		 * entries.
+		 */
+		index = dinfo->first + fidx;
+		if (index >= dinfo->nentries)
+			index -= dinfo->nentries;
+		e = &dinfo->log[index];
+		if (e->timestamp == 0)
+			return 0;
+
+		/* Print the message.  This depends on the firmware using
+		 * exactly the same formating strings as the kernel so we may
+		 * eventually have to put a format interpreter in here ...
+		 */
+		seq_printf(seq, "%10d  %15llu  %8s  %8s  ",
+			   be32_to_cpu(e->seqno),
+			   be64_to_cpu(e->timestamp),
+			   (e->level < ARRAY_SIZE(devlog_level_strings)
+			    ? devlog_level_strings[e->level]
+			    : "UNKNOWN"),
+			   (e->facility < ARRAY_SIZE(devlog_facility_strings)
+			    ? devlog_facility_strings[e->facility]
+			    : "UNKNOWN"));
+		seq_printf(seq, e->fmt,
+			   be32_to_cpu(e->params[0]),
+			   be32_to_cpu(e->params[1]),
+			   be32_to_cpu(e->params[2]),
+			   be32_to_cpu(e->params[3]),
+			   be32_to_cpu(e->params[4]),
+			   be32_to_cpu(e->params[5]),
+			   be32_to_cpu(e->params[6]),
+			   be32_to_cpu(e->params[7]));
+	}
+	return 0;
+}
+
+/* Sequential File Operations for Device Log.
+ */
+static inline void *devlog_get_idx(struct devlog_info *dinfo, loff_t pos)
+{
+	if (pos > dinfo->nentries)
+		return NULL;
+
+	return (void *)(uintptr_t)(pos + 1);
+}
+
+static void *devlog_start(struct seq_file *seq, loff_t *pos)
+{
+	struct devlog_info *dinfo = seq->private;
+
+	return (*pos
+		? devlog_get_idx(dinfo, *pos)
+		: SEQ_START_TOKEN);
+}
+
+static void *devlog_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+	struct devlog_info *dinfo = seq->private;
+
+	(*pos)++;
+	return devlog_get_idx(dinfo, *pos);
+}
+
+static void devlog_stop(struct seq_file *seq, void *v)
+{
+}
+
+static const struct seq_operations devlog_seq_ops = {
+	.start = devlog_start,
+	.next  = devlog_next,
+	.stop  = devlog_stop,
+	.show  = devlog_show
+};
+
+/* Set up for reading the firmware's device log.  We read the entire log here
+ * and then display it incrementally in devlog_show().
+ */
+static int devlog_open(struct inode *inode, struct file *file)
+{
+	struct adapter *adap = inode->i_private;
+	struct devlog_params *dparams = &adap->params.devlog;
+	struct devlog_info *dinfo;
+	unsigned int index;
+	u32 fseqno;
+	int ret;
+
+	/* If we don't know where the log is we can't do anything.
+	 */
+	if (dparams->start == 0)
+		return -ENXIO;
+
+	/* Allocate the space to read in the firmware's device log and set up
+	 * for the iterated call to our display function.
+	 */
+	dinfo = __seq_open_private(file, &devlog_seq_ops,
+				   sizeof(*dinfo) + dparams->size);
+	if (!dinfo)
+		return -ENOMEM;
+
+	/* Record the basic log buffer information and read in the raw log.
+	 */
+	dinfo->nentries = (dparams->size / sizeof(struct fw_devlog_e));
+	dinfo->first = 0;
+	spin_lock(&adap->win0_lock);
+	ret = t4_memory_rw(adap, adap->params.drv_memwin, dparams->memtype,
+			   dparams->start, dparams->size, (__be32 *)dinfo->log,
+			   T4_MEMORY_READ);
+	spin_unlock(&adap->win0_lock);
+	if (ret) {
+		seq_release_private(inode, file);
+		return ret;
+	}
+
+	/* Find the earliest (lowest Sequence Number) log entry in the
+	 * circular Device Log.
+	 */
+	for (fseqno = ~((u32)0), index = 0; index < dinfo->nentries; index++) {
+		struct fw_devlog_e *e = &dinfo->log[index];
+		__u32 seqno;
+
+		if (e->timestamp == 0)
+			continue;
+
+		seqno = be32_to_cpu(e->seqno);
+		if (seqno < fseqno) {
+			fseqno = seqno;
+			dinfo->first = index;
+		}
+	}
+	return 0;
+}
+
+static const struct file_operations devlog_fops = {
+	.owner   = THIS_MODULE,
+	.open    = devlog_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private
+};
+
+/* Show Firmware Mailbox Command/Reply Log
+ *
+ * Note that we don't do any locking when dumping the Firmware Mailbox Log so
+ * it's possible that we can catch things during a log update and therefore
+ * see partially corrupted log entries.  But it's probably Good Enough(tm).
+ * If we ever decide that we want to make sure that we're dumping a coherent
+ * log, we'd need to perform locking in the mailbox logging and in
+ * mboxlog_open() where we'd need to grab the entire mailbox log in one go
+ * like we do for the Firmware Device Log.
+ */
+static int mboxlog_show(struct seq_file *seq, void *v)
+{
+	struct adapter *adapter = seq->private;
+	struct mbox_cmd_log *log = adapter->mbox_log;
+	struct mbox_cmd *entry;
+	int entry_idx, i;
+
+	if (v == SEQ_START_TOKEN) {
+		seq_printf(seq,
+			   "%10s  %15s  %5s  %5s  %s\n",
+			   "Seq#", "Tstamp", "Atime", "Etime",
+			   "Command/Reply");
+		return 0;
+	}
+
+	entry_idx = log->cursor + ((uintptr_t)v - 2);
+	if (entry_idx >= log->size)
+		entry_idx -= log->size;
+	entry = mbox_cmd_log_entry(log, entry_idx);
+
+	/* skip over unused entries */
+	if (entry->timestamp == 0)
+		return 0;
+
+	seq_printf(seq, "%10u  %15llu  %5d  %5d",
+		   entry->seqno, entry->timestamp,
+		   entry->access, entry->execute);
+	for (i = 0; i < MBOX_LEN / 8; i++) {
+		u64 flit = entry->cmd[i];
+		u32 hi = (u32)(flit >> 32);
+		u32 lo = (u32)flit;
+
+		seq_printf(seq, "  %08x %08x", hi, lo);
+	}
+	seq_puts(seq, "\n");
+	return 0;
+}
+
+static inline void *mboxlog_get_idx(struct seq_file *seq, loff_t pos)
+{
+	struct adapter *adapter = seq->private;
+	struct mbox_cmd_log *log = adapter->mbox_log;
+
+	return ((pos <= log->size) ? (void *)(uintptr_t)(pos + 1) : NULL);
+}
+
+static void *mboxlog_start(struct seq_file *seq, loff_t *pos)
+{
+	return *pos ? mboxlog_get_idx(seq, *pos) : SEQ_START_TOKEN;
+}
+
+static void *mboxlog_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+	++*pos;
+	return mboxlog_get_idx(seq, *pos);
+}
+
+static void mboxlog_stop(struct seq_file *seq, void *v)
+{
+}
+
+static const struct seq_operations mboxlog_seq_ops = {
+	.start = mboxlog_start,
+	.next  = mboxlog_next,
+	.stop  = mboxlog_stop,
+	.show  = mboxlog_show
+};
+
+static int mboxlog_open(struct inode *inode, struct file *file)
+{
+	int res = seq_open(file, &mboxlog_seq_ops);
+
+	if (!res) {
+		struct seq_file *seq = file->private_data;
+
+		seq->private = inode->i_private;
+	}
+	return res;
+}
+
+static const struct file_operations mboxlog_fops = {
+	.owner   = THIS_MODULE,
+	.open    = mboxlog_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release,
+};
+
+static int mbox_show(struct seq_file *seq, void *v)
+{
+	static const char * const owner[] = { "none", "FW", "driver",
+					      "unknown", "<unread>" };
+
+	int i;
+	unsigned int mbox = (uintptr_t)seq->private & 7;
+	struct adapter *adap = seq->private - mbox;
+	void __iomem *addr = adap->regs + PF_REG(mbox, CIM_PF_MAILBOX_DATA_A);
+
+	/* For T4 we don't have a shadow copy of the Mailbox Control register.
+	 * And since reading that real register causes a side effect of
+	 * granting ownership, we're best of simply not reading it at all.
+	 */
+	if (is_t4(adap->params.chip)) {
+		i = 4; /* index of "<unread>" */
+	} else {
+		unsigned int ctrl_reg = CIM_PF_MAILBOX_CTRL_SHADOW_COPY_A;
+		void __iomem *ctrl = adap->regs + PF_REG(mbox, ctrl_reg);
+
+		i = MBOWNER_G(readl(ctrl));
+	}
+
+	seq_printf(seq, "mailbox owned by %s\n\n", owner[i]);
+
+	for (i = 0; i < MBOX_LEN; i += 8)
+		seq_printf(seq, "%016llx\n",
+			   (unsigned long long)readq(addr + i));
+	return 0;
+}
+
+static int mbox_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, mbox_show, inode->i_private);
+}
+
+static ssize_t mbox_write(struct file *file, const char __user *buf,
+			  size_t count, loff_t *pos)
+{
+	int i;
+	char c = '\n', s[256];
+	unsigned long long data[8];
+	const struct inode *ino;
+	unsigned int mbox;
+	struct adapter *adap;
+	void __iomem *addr;
+	void __iomem *ctrl;
+
+	if (count > sizeof(s) - 1 || !count)
+		return -EINVAL;
+	if (copy_from_user(s, buf, count))
+		return -EFAULT;
+	s[count] = '\0';
+
+	if (sscanf(s, "%llx %llx %llx %llx %llx %llx %llx %llx%c", &data[0],
+		   &data[1], &data[2], &data[3], &data[4], &data[5], &data[6],
+		   &data[7], &c) < 8 || c != '\n')
+		return -EINVAL;
+
+	ino = file_inode(file);
+	mbox = (uintptr_t)ino->i_private & 7;
+	adap = ino->i_private - mbox;
+	addr = adap->regs + PF_REG(mbox, CIM_PF_MAILBOX_DATA_A);
+	ctrl = addr + MBOX_LEN;
+
+	if (MBOWNER_G(readl(ctrl)) != X_MBOWNER_PL)
+		return -EBUSY;
+
+	for (i = 0; i < 8; i++)
+		writeq(data[i], addr + 8 * i);
+
+	writel(MBMSGVALID_F | MBOWNER_V(X_MBOWNER_FW), ctrl);
+	return count;
+}
+
+static const struct file_operations mbox_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = mbox_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = single_release,
+	.write   = mbox_write
+};
+
+static int mps_trc_show(struct seq_file *seq, void *v)
+{
+	int enabled, i;
+	struct trace_params tp;
+	unsigned int trcidx = (uintptr_t)seq->private & 3;
+	struct adapter *adap = seq->private - trcidx;
+
+	t4_get_trace_filter(adap, &tp, trcidx, &enabled);
+	if (!enabled) {
+		seq_puts(seq, "tracer is disabled\n");
+		return 0;
+	}
+
+	if (tp.skip_ofst * 8 >= TRACE_LEN) {
+		dev_err(adap->pdev_dev, "illegal trace pattern skip offset\n");
+		return -EINVAL;
+	}
+	if (tp.port < 8) {
+		i = adap->chan_map[tp.port & 3];
+		if (i >= MAX_NPORTS) {
+			dev_err(adap->pdev_dev, "tracer %u is assigned "
+				"to non-existing port\n", trcidx);
+			return -EINVAL;
+		}
+		seq_printf(seq, "tracer is capturing %s %s, ",
+			   adap->port[i]->name, tp.port < 4 ? "Rx" : "Tx");
+	} else
+		seq_printf(seq, "tracer is capturing loopback %d, ",
+			   tp.port - 8);
+	seq_printf(seq, "snap length: %u, min length: %u\n", tp.snap_len,
+		   tp.min_len);
+	seq_printf(seq, "packets captured %smatch filter\n",
+		   tp.invert ? "do not " : "");
+
+	if (tp.skip_ofst) {
+		seq_puts(seq, "filter pattern: ");
+		for (i = 0; i < tp.skip_ofst * 2; i += 2)
+			seq_printf(seq, "%08x%08x", tp.data[i], tp.data[i + 1]);
+		seq_putc(seq, '/');
+		for (i = 0; i < tp.skip_ofst * 2; i += 2)
+			seq_printf(seq, "%08x%08x", tp.mask[i], tp.mask[i + 1]);
+		seq_puts(seq, "@0\n");
+	}
+
+	seq_puts(seq, "filter pattern: ");
+	for (i = tp.skip_ofst * 2; i < TRACE_LEN / 4; i += 2)
+		seq_printf(seq, "%08x%08x", tp.data[i], tp.data[i + 1]);
+	seq_putc(seq, '/');
+	for (i = tp.skip_ofst * 2; i < TRACE_LEN / 4; i += 2)
+		seq_printf(seq, "%08x%08x", tp.mask[i], tp.mask[i + 1]);
+	seq_printf(seq, "@%u\n", (tp.skip_ofst + tp.skip_len) * 8);
+	return 0;
+}
+
+static int mps_trc_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, mps_trc_show, inode->i_private);
+}
+
+static unsigned int xdigit2int(unsigned char c)
+{
+	return isdigit(c) ? c - '0' : tolower(c) - 'a' + 10;
+}
+
+#define TRC_PORT_NONE 0xff
+#define TRC_RSS_ENABLE 0x33
+#define TRC_RSS_DISABLE 0x13
+
+/* Set an MPS trace filter.  Syntax is:
+ *
+ * disable
+ *
+ * to disable tracing, or
+ *
+ * interface qid=<qid no> [snaplen=<val>] [minlen=<val>] [not] [<pattern>]...
+ *
+ * where interface is one of rxN, txN, or loopbackN, N = 0..3, qid can be one
+ * of the NIC's response qid obtained from sge_qinfo and pattern has the form
+ *
+ * <pattern data>[/<pattern mask>][@<anchor>]
+ *
+ * Up to 2 filter patterns can be specified.  If 2 are supplied the first one
+ * must be anchored at 0.  An omitted mask is taken as a mask of 1s, an omitted
+ * anchor is taken as 0.
+ */
+static ssize_t mps_trc_write(struct file *file, const char __user *buf,
+			     size_t count, loff_t *pos)
+{
+	int i, enable, ret;
+	u32 *data, *mask;
+	struct trace_params tp;
+	const struct inode *ino;
+	unsigned int trcidx;
+	char *s, *p, *word, *end;
+	struct adapter *adap;
+	u32 j;
+
+	ino = file_inode(file);
+	trcidx = (uintptr_t)ino->i_private & 3;
+	adap = ino->i_private - trcidx;
+
+	/* Don't accept input more than 1K, can't be anything valid except lots
+	 * of whitespace.  Well, use less.
+	 */
+	if (count > 1024)
+		return -EFBIG;
+	p = s = kzalloc(count + 1, GFP_USER);
+	if (!s)
+		return -ENOMEM;
+	if (copy_from_user(s, buf, count)) {
+		count = -EFAULT;
+		goto out;
+	}
+
+	if (s[count - 1] == '\n')
+		s[count - 1] = '\0';
+
+	enable = strcmp("disable", s) != 0;
+	if (!enable)
+		goto apply;
+
+	/* enable or disable trace multi rss filter */
+	if (adap->trace_rss)
+		t4_write_reg(adap, MPS_TRC_CFG_A, TRC_RSS_ENABLE);
+	else
+		t4_write_reg(adap, MPS_TRC_CFG_A, TRC_RSS_DISABLE);
+
+	memset(&tp, 0, sizeof(tp));
+	tp.port = TRC_PORT_NONE;
+	i = 0;	/* counts pattern nibbles */
+
+	while (p) {
+		while (isspace(*p))
+			p++;
+		word = strsep(&p, " ");
+		if (!*word)
+			break;
+
+		if (!strncmp(word, "qid=", 4)) {
+			end = (char *)word + 4;
+			ret = kstrtouint(end, 10, &j);
+			if (ret)
+				goto out;
+			if (!adap->trace_rss) {
+				t4_write_reg(adap, MPS_T5_TRC_RSS_CONTROL_A, j);
+				continue;
+			}
+
+			switch (trcidx) {
+			case 0:
+				t4_write_reg(adap, MPS_TRC_RSS_CONTROL_A, j);
+				break;
+			case 1:
+				t4_write_reg(adap,
+					     MPS_TRC_FILTER1_RSS_CONTROL_A, j);
+				break;
+			case 2:
+				t4_write_reg(adap,
+					     MPS_TRC_FILTER2_RSS_CONTROL_A, j);
+				break;
+			case 3:
+				t4_write_reg(adap,
+					     MPS_TRC_FILTER3_RSS_CONTROL_A, j);
+				break;
+			}
+			continue;
+		}
+		if (!strncmp(word, "snaplen=", 8)) {
+			end = (char *)word + 8;
+			ret = kstrtouint(end, 10, &j);
+			if (ret || j > 9600) {
+inval:				count = -EINVAL;
+				goto out;
+			}
+			tp.snap_len = j;
+			continue;
+		}
+		if (!strncmp(word, "minlen=", 7)) {
+			end = (char *)word + 7;
+			ret = kstrtouint(end, 10, &j);
+			if (ret || j > TFMINPKTSIZE_M)
+				goto inval;
+			tp.min_len = j;
+			continue;
+		}
+		if (!strcmp(word, "not")) {
+			tp.invert = !tp.invert;
+			continue;
+		}
+		if (!strncmp(word, "loopback", 8) && tp.port == TRC_PORT_NONE) {
+			if (word[8] < '0' || word[8] > '3' || word[9])
+				goto inval;
+			tp.port = word[8] - '0' + 8;
+			continue;
+		}
+		if (!strncmp(word, "tx", 2) && tp.port == TRC_PORT_NONE) {
+			if (word[2] < '0' || word[2] > '3' || word[3])
+				goto inval;
+			tp.port = word[2] - '0' + 4;
+			if (adap->chan_map[tp.port & 3] >= MAX_NPORTS)
+				goto inval;
+			continue;
+		}
+		if (!strncmp(word, "rx", 2) && tp.port == TRC_PORT_NONE) {
+			if (word[2] < '0' || word[2] > '3' || word[3])
+				goto inval;
+			tp.port = word[2] - '0';
+			if (adap->chan_map[tp.port] >= MAX_NPORTS)
+				goto inval;
+			continue;
+		}
+		if (!isxdigit(*word))
+			goto inval;
+
+		/* we have found a trace pattern */
+		if (i) {                            /* split pattern */
+			if (tp.skip_len)            /* too many splits */
+				goto inval;
+			tp.skip_ofst = i / 16;
+		}
+
+		data = &tp.data[i / 8];
+		mask = &tp.mask[i / 8];
+		j = i;
+
+		while (isxdigit(*word)) {
+			if (i >= TRACE_LEN * 2) {
+				count = -EFBIG;
+				goto out;
+			}
+			*data = (*data << 4) + xdigit2int(*word++);
+			if (++i % 8 == 0)
+				data++;
+		}
+		if (*word == '/') {
+			word++;
+			while (isxdigit(*word)) {
+				if (j >= i)         /* mask longer than data */
+					goto inval;
+				*mask = (*mask << 4) + xdigit2int(*word++);
+				if (++j % 8 == 0)
+					mask++;
+			}
+			if (i != j)                 /* mask shorter than data */
+				goto inval;
+		} else {                            /* no mask, use all 1s */
+			for ( ; i - j >= 8; j += 8)
+				*mask++ = 0xffffffff;
+			if (i % 8)
+				*mask = (1 << (i % 8) * 4) - 1;
+		}
+		if (*word == '@') {
+			end = (char *)word + 1;
+			ret = kstrtouint(end, 10, &j);
+			if (*end && *end != '\n')
+				goto inval;
+			if (j & 7)          /* doesn't start at multiple of 8 */
+				goto inval;
+			j /= 8;
+			if (j < tp.skip_ofst)     /* overlaps earlier pattern */
+				goto inval;
+			if (j - tp.skip_ofst > 31)            /* skip too big */
+				goto inval;
+			tp.skip_len = j - tp.skip_ofst;
+		}
+		if (i % 8) {
+			*data <<= (8 - i % 8) * 4;
+			*mask <<= (8 - i % 8) * 4;
+			i = (i + 15) & ~15;         /* 8-byte align */
+		}
+	}
+
+	if (tp.port == TRC_PORT_NONE)
+		goto inval;
+
+apply:
+	i = t4_set_trace_filter(adap, &tp, trcidx, enable);
+	if (i)
+		count = i;
+out:
+	kfree(s);
+	return count;
+}
+
+static const struct file_operations mps_trc_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = mps_trc_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = single_release,
+	.write   = mps_trc_write
+};
+
+static ssize_t flash_read(struct file *file, char __user *buf, size_t count,
+			  loff_t *ppos)
+{
+	loff_t pos = *ppos;
+	loff_t avail = file_inode(file)->i_size;
+	struct adapter *adap = file->private_data;
+
+	if (pos < 0)
+		return -EINVAL;
+	if (pos >= avail)
+		return 0;
+	if (count > avail - pos)
+		count = avail - pos;
+
+	while (count) {
+		size_t len;
+		int ret, ofst;
+		u8 data[256];
+
+		ofst = pos & 3;
+		len = min(count + ofst, sizeof(data));
+		ret = t4_read_flash(adap, pos - ofst, (len + 3) / 4,
+				    (u32 *)data, 1);
+		if (ret)
+			return ret;
+
+		len -= ofst;
+		if (copy_to_user(buf, data + ofst, len))
+			return -EFAULT;
+
+		buf += len;
+		pos += len;
+		count -= len;
+	}
+	count = pos - *ppos;
+	*ppos = pos;
+	return count;
+}
+
+static const struct file_operations flash_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = mem_open,
+	.read    = flash_read,
+	.llseek  = default_llseek,
+};
+
+static inline void tcamxy2valmask(u64 x, u64 y, u8 *addr, u64 *mask)
+{
+	*mask = x | y;
+	y = (__force u64)cpu_to_be64(y);
+	memcpy(addr, (char *)&y + 2, ETH_ALEN);
+}
+
+static int mps_tcam_show(struct seq_file *seq, void *v)
+{
+	struct adapter *adap = seq->private;
+	unsigned int chip_ver = CHELSIO_CHIP_VERSION(adap->params.chip);
+	if (v == SEQ_START_TOKEN) {
+		if (chip_ver > CHELSIO_T5) {
+			seq_puts(seq, "Idx  Ethernet address     Mask     "
+				 "  VNI   Mask   IVLAN Vld "
+				 "DIP_Hit   Lookup  Port "
+				 "Vld Ports PF  VF                           "
+				 "Replication                                "
+				 "    P0 P1 P2 P3  ML\n");
+		} else {
+			if (adap->params.arch.mps_rplc_size > 128)
+				seq_puts(seq, "Idx  Ethernet address     Mask     "
+					 "Vld Ports PF  VF                           "
+					 "Replication                                "
+					 "    P0 P1 P2 P3  ML\n");
+			else
+				seq_puts(seq, "Idx  Ethernet address     Mask     "
+					 "Vld Ports PF  VF              Replication"
+					 "	         P0 P1 P2 P3  ML\n");
+		}
+	} else {
+		u64 mask;
+		u8 addr[ETH_ALEN];
+		bool replicate, dip_hit = false, vlan_vld = false;
+		unsigned int idx = (uintptr_t)v - 2;
+		u64 tcamy, tcamx, val;
+		u32 cls_lo, cls_hi, ctl, data2, vnix = 0, vniy = 0;
+		u32 rplc[8] = {0};
+		u8 lookup_type = 0, port_num = 0;
+		u16 ivlan = 0;
+
+		if (chip_ver > CHELSIO_T5) {
+			/* CtlCmdType - 0: Read, 1: Write
+			 * CtlTcamSel - 0: TCAM0, 1: TCAM1
+			 * CtlXYBitSel- 0: Y bit, 1: X bit
+			 */
+
+			/* Read tcamy */
+			ctl = CTLCMDTYPE_V(0) | CTLXYBITSEL_V(0);
+			if (idx < 256)
+				ctl |= CTLTCAMINDEX_V(idx) | CTLTCAMSEL_V(0);
+			else
+				ctl |= CTLTCAMINDEX_V(idx - 256) |
+				       CTLTCAMSEL_V(1);
+			t4_write_reg(adap, MPS_CLS_TCAM_DATA2_CTL_A, ctl);
+			val = t4_read_reg(adap, MPS_CLS_TCAM_DATA1_A);
+			tcamy = DMACH_G(val) << 32;
+			tcamy |= t4_read_reg(adap, MPS_CLS_TCAM_DATA0_A);
+			data2 = t4_read_reg(adap, MPS_CLS_TCAM_DATA2_CTL_A);
+			lookup_type = DATALKPTYPE_G(data2);
+			/* 0 - Outer header, 1 - Inner header
+			 * [71:48] bit locations are overloaded for
+			 * outer vs. inner lookup types.
+			 */
+			if (lookup_type && (lookup_type != DATALKPTYPE_M)) {
+				/* Inner header VNI */
+				vniy = (data2 & DATAVIDH2_F) |
+				       (DATAVIDH1_G(data2) << 16) | VIDL_G(val);
+				dip_hit = data2 & DATADIPHIT_F;
+			} else {
+				vlan_vld = data2 & DATAVIDH2_F;
+				ivlan = VIDL_G(val);
+			}
+			port_num = DATAPORTNUM_G(data2);
+
+			/* Read tcamx. Change the control param */
+			vnix = 0;
+			ctl |= CTLXYBITSEL_V(1);
+			t4_write_reg(adap, MPS_CLS_TCAM_DATA2_CTL_A, ctl);
+			val = t4_read_reg(adap, MPS_CLS_TCAM_DATA1_A);
+			tcamx = DMACH_G(val) << 32;
+			tcamx |= t4_read_reg(adap, MPS_CLS_TCAM_DATA0_A);
+			data2 = t4_read_reg(adap, MPS_CLS_TCAM_DATA2_CTL_A);
+			if (lookup_type && (lookup_type != DATALKPTYPE_M)) {
+				/* Inner header VNI mask */
+				vnix = (data2 & DATAVIDH2_F) |
+				       (DATAVIDH1_G(data2) << 16) | VIDL_G(val);
+			}
+		} else {
+			tcamy = t4_read_reg64(adap, MPS_CLS_TCAM_Y_L(idx));
+			tcamx = t4_read_reg64(adap, MPS_CLS_TCAM_X_L(idx));
+		}
+
+		cls_lo = t4_read_reg(adap, MPS_CLS_SRAM_L(idx));
+		cls_hi = t4_read_reg(adap, MPS_CLS_SRAM_H(idx));
+
+		if (tcamx & tcamy) {
+			seq_printf(seq, "%3u         -\n", idx);
+			goto out;
+		}
+
+		rplc[0] = rplc[1] = rplc[2] = rplc[3] = 0;
+		if (chip_ver > CHELSIO_T5)
+			replicate = (cls_lo & T6_REPLICATE_F);
+		else
+			replicate = (cls_lo & REPLICATE_F);
+
+		if (replicate) {
+			struct fw_ldst_cmd ldst_cmd;
+			int ret;
+			struct fw_ldst_mps_rplc mps_rplc;
+			u32 ldst_addrspc;
+
+			memset(&ldst_cmd, 0, sizeof(ldst_cmd));
+			ldst_addrspc =
+				FW_LDST_CMD_ADDRSPACE_V(FW_LDST_ADDRSPC_MPS);
+			ldst_cmd.op_to_addrspace =
+				htonl(FW_CMD_OP_V(FW_LDST_CMD) |
+				      FW_CMD_REQUEST_F |
+				      FW_CMD_READ_F |
+				      ldst_addrspc);
+			ldst_cmd.cycles_to_len16 = htonl(FW_LEN16(ldst_cmd));
+			ldst_cmd.u.mps.rplc.fid_idx =
+				htons(FW_LDST_CMD_FID_V(FW_LDST_MPS_RPLC) |
+				      FW_LDST_CMD_IDX_V(idx));
+			ret = t4_wr_mbox(adap, adap->mbox, &ldst_cmd,
+					 sizeof(ldst_cmd), &ldst_cmd);
+			if (ret)
+				dev_warn(adap->pdev_dev, "Can't read MPS "
+					 "replication map for idx %d: %d\n",
+					 idx, -ret);
+			else {
+				mps_rplc = ldst_cmd.u.mps.rplc;
+				rplc[0] = ntohl(mps_rplc.rplc31_0);
+				rplc[1] = ntohl(mps_rplc.rplc63_32);
+				rplc[2] = ntohl(mps_rplc.rplc95_64);
+				rplc[3] = ntohl(mps_rplc.rplc127_96);
+				if (adap->params.arch.mps_rplc_size > 128) {
+					rplc[4] = ntohl(mps_rplc.rplc159_128);
+					rplc[5] = ntohl(mps_rplc.rplc191_160);
+					rplc[6] = ntohl(mps_rplc.rplc223_192);
+					rplc[7] = ntohl(mps_rplc.rplc255_224);
+				}
+			}
+		}
+
+		tcamxy2valmask(tcamx, tcamy, addr, &mask);
+		if (chip_ver > CHELSIO_T5) {
+			/* Inner header lookup */
+			if (lookup_type && (lookup_type != DATALKPTYPE_M)) {
+				seq_printf(seq,
+					   "%3u %pM %012llx %06x %06x    -    -   %3c      'I'  %4x   %3c   %#x%4u%4d",
+					   idx, addr,
+					   (unsigned long long)mask,
+					   vniy, (vnix | vniy),
+					   dip_hit ? 'Y' : 'N',
+					   port_num,
+					   (cls_lo & T6_SRAM_VLD_F) ? 'Y' : 'N',
+					   PORTMAP_G(cls_hi),
+					   T6_PF_G(cls_lo),
+					   (cls_lo & T6_VF_VALID_F) ?
+					   T6_VF_G(cls_lo) : -1);
+			} else {
+				seq_printf(seq,
+					   "%3u %pM %012llx    -       -   ",
+					   idx, addr,
+					   (unsigned long long)mask);
+
+				if (vlan_vld)
+					seq_printf(seq, "%4u   Y     ", ivlan);
+				else
+					seq_puts(seq, "  -    N     ");
+
+				seq_printf(seq,
+					   "-      %3c  %4x   %3c   %#x%4u%4d",
+					   lookup_type ? 'I' : 'O', port_num,
+					   (cls_lo & T6_SRAM_VLD_F) ? 'Y' : 'N',
+					   PORTMAP_G(cls_hi),
+					   T6_PF_G(cls_lo),
+					   (cls_lo & T6_VF_VALID_F) ?
+					   T6_VF_G(cls_lo) : -1);
+			}
+		} else
+			seq_printf(seq, "%3u %pM %012llx%3c   %#x%4u%4d",
+				   idx, addr, (unsigned long long)mask,
+				   (cls_lo & SRAM_VLD_F) ? 'Y' : 'N',
+				   PORTMAP_G(cls_hi),
+				   PF_G(cls_lo),
+				   (cls_lo & VF_VALID_F) ? VF_G(cls_lo) : -1);
+
+		if (replicate) {
+			if (adap->params.arch.mps_rplc_size > 128)
+				seq_printf(seq, " %08x %08x %08x %08x "
+					   "%08x %08x %08x %08x",
+					   rplc[7], rplc[6], rplc[5], rplc[4],
+					   rplc[3], rplc[2], rplc[1], rplc[0]);
+			else
+				seq_printf(seq, " %08x %08x %08x %08x",
+					   rplc[3], rplc[2], rplc[1], rplc[0]);
+		} else {
+			if (adap->params.arch.mps_rplc_size > 128)
+				seq_printf(seq, "%72c", ' ');
+			else
+				seq_printf(seq, "%36c", ' ');
+		}
+
+		if (chip_ver > CHELSIO_T5)
+			seq_printf(seq, "%4u%3u%3u%3u %#x\n",
+				   T6_SRAM_PRIO0_G(cls_lo),
+				   T6_SRAM_PRIO1_G(cls_lo),
+				   T6_SRAM_PRIO2_G(cls_lo),
+				   T6_SRAM_PRIO3_G(cls_lo),
+				   (cls_lo >> T6_MULTILISTEN0_S) & 0xf);
+		else
+			seq_printf(seq, "%4u%3u%3u%3u %#x\n",
+				   SRAM_PRIO0_G(cls_lo), SRAM_PRIO1_G(cls_lo),
+				   SRAM_PRIO2_G(cls_lo), SRAM_PRIO3_G(cls_lo),
+				   (cls_lo >> MULTILISTEN0_S) & 0xf);
+	}
+out:	return 0;
+}
+
+static inline void *mps_tcam_get_idx(struct seq_file *seq, loff_t pos)
+{
+	struct adapter *adap = seq->private;
+	int max_mac_addr = is_t4(adap->params.chip) ?
+				NUM_MPS_CLS_SRAM_L_INSTANCES :
+				NUM_MPS_T5_CLS_SRAM_L_INSTANCES;
+	return ((pos <= max_mac_addr) ? (void *)(uintptr_t)(pos + 1) : NULL);
+}
+
+static void *mps_tcam_start(struct seq_file *seq, loff_t *pos)
+{
+	return *pos ? mps_tcam_get_idx(seq, *pos) : SEQ_START_TOKEN;
+}
+
+static void *mps_tcam_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+	++*pos;
+	return mps_tcam_get_idx(seq, *pos);
+}
+
+static void mps_tcam_stop(struct seq_file *seq, void *v)
+{
+}
+
+static const struct seq_operations mps_tcam_seq_ops = {
+	.start = mps_tcam_start,
+	.next  = mps_tcam_next,
+	.stop  = mps_tcam_stop,
+	.show  = mps_tcam_show
+};
+
+static int mps_tcam_open(struct inode *inode, struct file *file)
+{
+	int res = seq_open(file, &mps_tcam_seq_ops);
+
+	if (!res) {
+		struct seq_file *seq = file->private_data;
+
+		seq->private = inode->i_private;
+	}
+	return res;
+}
+
+static const struct file_operations mps_tcam_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = mps_tcam_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release,
+};
+
+/* Display various sensor information.
+ */
+static int sensors_show(struct seq_file *seq, void *v)
+{
+	struct adapter *adap = seq->private;
+	u32 param[7], val[7];
+	int ret;
+
+	/* Note that if the sensors haven't been initialized and turned on
+	 * we'll get values of 0, so treat those as "<unknown>" ...
+	 */
+	param[0] = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		    FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_DIAG) |
+		    FW_PARAMS_PARAM_Y_V(FW_PARAM_DEV_DIAG_TMP));
+	param[1] = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		    FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_DIAG) |
+		    FW_PARAMS_PARAM_Y_V(FW_PARAM_DEV_DIAG_VDD));
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2,
+			      param, val);
+
+	if (ret < 0 || val[0] == 0)
+		seq_puts(seq, "Temperature: <unknown>\n");
+	else
+		seq_printf(seq, "Temperature: %dC\n", val[0]);
+
+	if (ret < 0 || val[1] == 0)
+		seq_puts(seq, "Core VDD:    <unknown>\n");
+	else
+		seq_printf(seq, "Core VDD:    %dmV\n", val[1]);
+
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(sensors);
+
+#if IS_ENABLED(CONFIG_IPV6)
+DEFINE_SHOW_ATTRIBUTE(clip_tbl);
+#endif
+
+/*RSS Table.
+ */
+
+static int rss_show(struct seq_file *seq, void *v, int idx)
+{
+	u16 *entry = v;
+
+	seq_printf(seq, "%4d:  %4u  %4u  %4u  %4u  %4u  %4u  %4u  %4u\n",
+		   idx * 8, entry[0], entry[1], entry[2], entry[3], entry[4],
+		   entry[5], entry[6], entry[7]);
+	return 0;
+}
+
+static int rss_open(struct inode *inode, struct file *file)
+{
+	struct adapter *adap = inode->i_private;
+	int ret, nentries;
+	struct seq_tab *p;
+
+	nentries = t4_chip_rss_size(adap);
+	p = seq_open_tab(file, nentries / 8, 8 * sizeof(u16), 0, rss_show);
+	if (!p)
+		return -ENOMEM;
+
+	ret = t4_read_rss(adap, (u16 *)p->data);
+	if (ret)
+		seq_release_private(inode, file);
+
+	return ret;
+}
+
+static const struct file_operations rss_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = rss_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private
+};
+
+/* RSS Configuration.
+ */
+
+/* Small utility function to return the strings "yes" or "no" if the supplied
+ * argument is non-zero.
+ */
+static const char *yesno(int x)
+{
+	static const char *yes = "yes";
+	static const char *no = "no";
+
+	return x ? yes : no;
+}
+
+static int rss_config_show(struct seq_file *seq, void *v)
+{
+	struct adapter *adapter = seq->private;
+	static const char * const keymode[] = {
+		"global",
+		"global and per-VF scramble",
+		"per-PF and per-VF scramble",
+		"per-VF and per-VF scramble",
+	};
+	u32 rssconf;
+
+	rssconf = t4_read_reg(adapter, TP_RSS_CONFIG_A);
+	seq_printf(seq, "TP_RSS_CONFIG: %#x\n", rssconf);
+	seq_printf(seq, "  Tnl4TupEnIpv6: %3s\n", yesno(rssconf &
+							TNL4TUPENIPV6_F));
+	seq_printf(seq, "  Tnl2TupEnIpv6: %3s\n", yesno(rssconf &
+							TNL2TUPENIPV6_F));
+	seq_printf(seq, "  Tnl4TupEnIpv4: %3s\n", yesno(rssconf &
+							TNL4TUPENIPV4_F));
+	seq_printf(seq, "  Tnl2TupEnIpv4: %3s\n", yesno(rssconf &
+							TNL2TUPENIPV4_F));
+	seq_printf(seq, "  TnlTcpSel:     %3s\n", yesno(rssconf & TNLTCPSEL_F));
+	seq_printf(seq, "  TnlIp6Sel:     %3s\n", yesno(rssconf & TNLIP6SEL_F));
+	seq_printf(seq, "  TnlVrtSel:     %3s\n", yesno(rssconf & TNLVRTSEL_F));
+	seq_printf(seq, "  TnlMapEn:      %3s\n", yesno(rssconf & TNLMAPEN_F));
+	seq_printf(seq, "  OfdHashSave:   %3s\n", yesno(rssconf &
+							OFDHASHSAVE_F));
+	seq_printf(seq, "  OfdVrtSel:     %3s\n", yesno(rssconf & OFDVRTSEL_F));
+	seq_printf(seq, "  OfdMapEn:      %3s\n", yesno(rssconf & OFDMAPEN_F));
+	seq_printf(seq, "  OfdLkpEn:      %3s\n", yesno(rssconf & OFDLKPEN_F));
+	seq_printf(seq, "  Syn4TupEnIpv6: %3s\n", yesno(rssconf &
+							SYN4TUPENIPV6_F));
+	seq_printf(seq, "  Syn2TupEnIpv6: %3s\n", yesno(rssconf &
+							SYN2TUPENIPV6_F));
+	seq_printf(seq, "  Syn4TupEnIpv4: %3s\n", yesno(rssconf &
+							SYN4TUPENIPV4_F));
+	seq_printf(seq, "  Syn2TupEnIpv4: %3s\n", yesno(rssconf &
+							SYN2TUPENIPV4_F));
+	seq_printf(seq, "  Syn4TupEnIpv6: %3s\n", yesno(rssconf &
+							SYN4TUPENIPV6_F));
+	seq_printf(seq, "  SynIp6Sel:     %3s\n", yesno(rssconf & SYNIP6SEL_F));
+	seq_printf(seq, "  SynVrt6Sel:    %3s\n", yesno(rssconf & SYNVRTSEL_F));
+	seq_printf(seq, "  SynMapEn:      %3s\n", yesno(rssconf & SYNMAPEN_F));
+	seq_printf(seq, "  SynLkpEn:      %3s\n", yesno(rssconf & SYNLKPEN_F));
+	seq_printf(seq, "  ChnEn:         %3s\n", yesno(rssconf &
+							CHANNELENABLE_F));
+	seq_printf(seq, "  PrtEn:         %3s\n", yesno(rssconf &
+							PORTENABLE_F));
+	seq_printf(seq, "  TnlAllLkp:     %3s\n", yesno(rssconf &
+							TNLALLLOOKUP_F));
+	seq_printf(seq, "  VrtEn:         %3s\n", yesno(rssconf &
+							VIRTENABLE_F));
+	seq_printf(seq, "  CngEn:         %3s\n", yesno(rssconf &
+							CONGESTIONENABLE_F));
+	seq_printf(seq, "  HashToeplitz:  %3s\n", yesno(rssconf &
+							HASHTOEPLITZ_F));
+	seq_printf(seq, "  Udp4En:        %3s\n", yesno(rssconf & UDPENABLE_F));
+	seq_printf(seq, "  Disable:       %3s\n", yesno(rssconf & DISABLE_F));
+
+	seq_puts(seq, "\n");
+
+	rssconf = t4_read_reg(adapter, TP_RSS_CONFIG_TNL_A);
+	seq_printf(seq, "TP_RSS_CONFIG_TNL: %#x\n", rssconf);
+	seq_printf(seq, "  MaskSize:      %3d\n", MASKSIZE_G(rssconf));
+	seq_printf(seq, "  MaskFilter:    %3d\n", MASKFILTER_G(rssconf));
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) > CHELSIO_T5) {
+		seq_printf(seq, "  HashAll:     %3s\n",
+			   yesno(rssconf & HASHALL_F));
+		seq_printf(seq, "  HashEth:     %3s\n",
+			   yesno(rssconf & HASHETH_F));
+	}
+	seq_printf(seq, "  UseWireCh:     %3s\n", yesno(rssconf & USEWIRECH_F));
+
+	seq_puts(seq, "\n");
+
+	rssconf = t4_read_reg(adapter, TP_RSS_CONFIG_OFD_A);
+	seq_printf(seq, "TP_RSS_CONFIG_OFD: %#x\n", rssconf);
+	seq_printf(seq, "  MaskSize:      %3d\n", MASKSIZE_G(rssconf));
+	seq_printf(seq, "  RRCplMapEn:    %3s\n", yesno(rssconf &
+							RRCPLMAPEN_F));
+	seq_printf(seq, "  RRCplQueWidth: %3d\n", RRCPLQUEWIDTH_G(rssconf));
+
+	seq_puts(seq, "\n");
+
+	rssconf = t4_read_reg(adapter, TP_RSS_CONFIG_SYN_A);
+	seq_printf(seq, "TP_RSS_CONFIG_SYN: %#x\n", rssconf);
+	seq_printf(seq, "  MaskSize:      %3d\n", MASKSIZE_G(rssconf));
+	seq_printf(seq, "  UseWireCh:     %3s\n", yesno(rssconf & USEWIRECH_F));
+
+	seq_puts(seq, "\n");
+
+	rssconf = t4_read_reg(adapter, TP_RSS_CONFIG_VRT_A);
+	seq_printf(seq, "TP_RSS_CONFIG_VRT: %#x\n", rssconf);
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) > CHELSIO_T5) {
+		seq_printf(seq, "  KeyWrAddrX:     %3d\n",
+			   KEYWRADDRX_G(rssconf));
+		seq_printf(seq, "  KeyExtend:      %3s\n",
+			   yesno(rssconf & KEYEXTEND_F));
+	}
+	seq_printf(seq, "  VfRdRg:        %3s\n", yesno(rssconf & VFRDRG_F));
+	seq_printf(seq, "  VfRdEn:        %3s\n", yesno(rssconf & VFRDEN_F));
+	seq_printf(seq, "  VfPerrEn:      %3s\n", yesno(rssconf & VFPERREN_F));
+	seq_printf(seq, "  KeyPerrEn:     %3s\n", yesno(rssconf & KEYPERREN_F));
+	seq_printf(seq, "  DisVfVlan:     %3s\n", yesno(rssconf &
+							DISABLEVLAN_F));
+	seq_printf(seq, "  EnUpSwt:       %3s\n", yesno(rssconf & ENABLEUP0_F));
+	seq_printf(seq, "  HashDelay:     %3d\n", HASHDELAY_G(rssconf));
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) <= CHELSIO_T5)
+		seq_printf(seq, "  VfWrAddr:      %3d\n", VFWRADDR_G(rssconf));
+	else
+		seq_printf(seq, "  VfWrAddr:      %3d\n",
+			   T6_VFWRADDR_G(rssconf));
+	seq_printf(seq, "  KeyMode:       %s\n", keymode[KEYMODE_G(rssconf)]);
+	seq_printf(seq, "  VfWrEn:        %3s\n", yesno(rssconf & VFWREN_F));
+	seq_printf(seq, "  KeyWrEn:       %3s\n", yesno(rssconf & KEYWREN_F));
+	seq_printf(seq, "  KeyWrAddr:     %3d\n", KEYWRADDR_G(rssconf));
+
+	seq_puts(seq, "\n");
+
+	rssconf = t4_read_reg(adapter, TP_RSS_CONFIG_CNG_A);
+	seq_printf(seq, "TP_RSS_CONFIG_CNG: %#x\n", rssconf);
+	seq_printf(seq, "  ChnCount3:     %3s\n", yesno(rssconf & CHNCOUNT3_F));
+	seq_printf(seq, "  ChnCount2:     %3s\n", yesno(rssconf & CHNCOUNT2_F));
+	seq_printf(seq, "  ChnCount1:     %3s\n", yesno(rssconf & CHNCOUNT1_F));
+	seq_printf(seq, "  ChnCount0:     %3s\n", yesno(rssconf & CHNCOUNT0_F));
+	seq_printf(seq, "  ChnUndFlow3:   %3s\n", yesno(rssconf &
+							CHNUNDFLOW3_F));
+	seq_printf(seq, "  ChnUndFlow2:   %3s\n", yesno(rssconf &
+							CHNUNDFLOW2_F));
+	seq_printf(seq, "  ChnUndFlow1:   %3s\n", yesno(rssconf &
+							CHNUNDFLOW1_F));
+	seq_printf(seq, "  ChnUndFlow0:   %3s\n", yesno(rssconf &
+							CHNUNDFLOW0_F));
+	seq_printf(seq, "  RstChn3:       %3s\n", yesno(rssconf & RSTCHN3_F));
+	seq_printf(seq, "  RstChn2:       %3s\n", yesno(rssconf & RSTCHN2_F));
+	seq_printf(seq, "  RstChn1:       %3s\n", yesno(rssconf & RSTCHN1_F));
+	seq_printf(seq, "  RstChn0:       %3s\n", yesno(rssconf & RSTCHN0_F));
+	seq_printf(seq, "  UpdVld:        %3s\n", yesno(rssconf & UPDVLD_F));
+	seq_printf(seq, "  Xoff:          %3s\n", yesno(rssconf & XOFF_F));
+	seq_printf(seq, "  UpdChn3:       %3s\n", yesno(rssconf & UPDCHN3_F));
+	seq_printf(seq, "  UpdChn2:       %3s\n", yesno(rssconf & UPDCHN2_F));
+	seq_printf(seq, "  UpdChn1:       %3s\n", yesno(rssconf & UPDCHN1_F));
+	seq_printf(seq, "  UpdChn0:       %3s\n", yesno(rssconf & UPDCHN0_F));
+	seq_printf(seq, "  Queue:         %3d\n", QUEUE_G(rssconf));
+
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(rss_config);
+
+/* RSS Secret Key.
+ */
+
+static int rss_key_show(struct seq_file *seq, void *v)
+{
+	u32 key[10];
+
+	t4_read_rss_key(seq->private, key, true);
+	seq_printf(seq, "%08x%08x%08x%08x%08x%08x%08x%08x%08x%08x\n",
+		   key[9], key[8], key[7], key[6], key[5], key[4], key[3],
+		   key[2], key[1], key[0]);
+	return 0;
+}
+
+static int rss_key_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, rss_key_show, inode->i_private);
+}
+
+static ssize_t rss_key_write(struct file *file, const char __user *buf,
+			     size_t count, loff_t *pos)
+{
+	int i, j;
+	u32 key[10];
+	char s[100], *p;
+	struct adapter *adap = file_inode(file)->i_private;
+
+	if (count > sizeof(s) - 1)
+		return -EINVAL;
+	if (copy_from_user(s, buf, count))
+		return -EFAULT;
+	for (i = count; i > 0 && isspace(s[i - 1]); i--)
+		;
+	s[i] = '\0';
+
+	for (p = s, i = 9; i >= 0; i--) {
+		key[i] = 0;
+		for (j = 0; j < 8; j++, p++) {
+			if (!isxdigit(*p))
+				return -EINVAL;
+			key[i] = (key[i] << 4) | hex2val(*p);
+		}
+	}
+
+	t4_write_rss_key(adap, key, -1, true);
+	return count;
+}
+
+static const struct file_operations rss_key_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = rss_key_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = single_release,
+	.write   = rss_key_write
+};
+
+/* PF RSS Configuration.
+ */
+
+struct rss_pf_conf {
+	u32 rss_pf_map;
+	u32 rss_pf_mask;
+	u32 rss_pf_config;
+};
+
+static int rss_pf_config_show(struct seq_file *seq, void *v, int idx)
+{
+	struct rss_pf_conf *pfconf;
+
+	if (v == SEQ_START_TOKEN) {
+		/* use the 0th entry to dump the PF Map Index Size */
+		pfconf = seq->private + offsetof(struct seq_tab, data);
+		seq_printf(seq, "PF Map Index Size = %d\n\n",
+			   LKPIDXSIZE_G(pfconf->rss_pf_map));
+
+		seq_puts(seq, "     RSS              PF   VF    Hash Tuple Enable         Default\n");
+		seq_puts(seq, "     Enable       IPF Mask Mask  IPv6      IPv4      UDP   Queue\n");
+		seq_puts(seq, " PF  Map Chn Prt  Map Size Size  Four Two  Four Two  Four  Ch1  Ch0\n");
+	} else {
+		#define G_PFnLKPIDX(map, n) \
+			(((map) >> PF1LKPIDX_S*(n)) & PF0LKPIDX_M)
+		#define G_PFnMSKSIZE(mask, n) \
+			(((mask) >> PF1MSKSIZE_S*(n)) & PF1MSKSIZE_M)
+
+		pfconf = v;
+		seq_printf(seq, "%3d  %3s %3s %3s  %3d  %3d  %3d   %3s %3s   %3s %3s   %3s  %3d  %3d\n",
+			   idx,
+			   yesno(pfconf->rss_pf_config & MAPENABLE_F),
+			   yesno(pfconf->rss_pf_config & CHNENABLE_F),
+			   yesno(pfconf->rss_pf_config & PRTENABLE_F),
+			   G_PFnLKPIDX(pfconf->rss_pf_map, idx),
+			   G_PFnMSKSIZE(pfconf->rss_pf_mask, idx),
+			   IVFWIDTH_G(pfconf->rss_pf_config),
+			   yesno(pfconf->rss_pf_config & IP6FOURTUPEN_F),
+			   yesno(pfconf->rss_pf_config & IP6TWOTUPEN_F),
+			   yesno(pfconf->rss_pf_config & IP4FOURTUPEN_F),
+			   yesno(pfconf->rss_pf_config & IP4TWOTUPEN_F),
+			   yesno(pfconf->rss_pf_config & UDPFOURTUPEN_F),
+			   CH1DEFAULTQUEUE_G(pfconf->rss_pf_config),
+			   CH0DEFAULTQUEUE_G(pfconf->rss_pf_config));
+
+		#undef G_PFnLKPIDX
+		#undef G_PFnMSKSIZE
+	}
+	return 0;
+}
+
+static int rss_pf_config_open(struct inode *inode, struct file *file)
+{
+	struct adapter *adapter = inode->i_private;
+	struct seq_tab *p;
+	u32 rss_pf_map, rss_pf_mask;
+	struct rss_pf_conf *pfconf;
+	int pf;
+
+	p = seq_open_tab(file, 8, sizeof(*pfconf), 1, rss_pf_config_show);
+	if (!p)
+		return -ENOMEM;
+
+	pfconf = (struct rss_pf_conf *)p->data;
+	rss_pf_map = t4_read_rss_pf_map(adapter, true);
+	rss_pf_mask = t4_read_rss_pf_mask(adapter, true);
+	for (pf = 0; pf < 8; pf++) {
+		pfconf[pf].rss_pf_map = rss_pf_map;
+		pfconf[pf].rss_pf_mask = rss_pf_mask;
+		t4_read_rss_pf_config(adapter, pf, &pfconf[pf].rss_pf_config,
+				      true);
+	}
+	return 0;
+}
+
+static const struct file_operations rss_pf_config_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = rss_pf_config_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private
+};
+
+/* VF RSS Configuration.
+ */
+
+struct rss_vf_conf {
+	u32 rss_vf_vfl;
+	u32 rss_vf_vfh;
+};
+
+static int rss_vf_config_show(struct seq_file *seq, void *v, int idx)
+{
+	if (v == SEQ_START_TOKEN) {
+		seq_puts(seq, "     RSS                     Hash Tuple Enable\n");
+		seq_puts(seq, "     Enable   IVF  Dis  Enb  IPv6      IPv4      UDP    Def  Secret Key\n");
+		seq_puts(seq, " VF  Chn Prt  Map  VLAN  uP  Four Two  Four Two  Four   Que  Idx       Hash\n");
+	} else {
+		struct rss_vf_conf *vfconf = v;
+
+		seq_printf(seq, "%3d  %3s %3s  %3d   %3s %3s   %3s %3s   %3s  %3s   %3s  %4d  %3d %#10x\n",
+			   idx,
+			   yesno(vfconf->rss_vf_vfh & VFCHNEN_F),
+			   yesno(vfconf->rss_vf_vfh & VFPRTEN_F),
+			   VFLKPIDX_G(vfconf->rss_vf_vfh),
+			   yesno(vfconf->rss_vf_vfh & VFVLNEX_F),
+			   yesno(vfconf->rss_vf_vfh & VFUPEN_F),
+			   yesno(vfconf->rss_vf_vfh & VFIP4FOURTUPEN_F),
+			   yesno(vfconf->rss_vf_vfh & VFIP6TWOTUPEN_F),
+			   yesno(vfconf->rss_vf_vfh & VFIP4FOURTUPEN_F),
+			   yesno(vfconf->rss_vf_vfh & VFIP4TWOTUPEN_F),
+			   yesno(vfconf->rss_vf_vfh & ENABLEUDPHASH_F),
+			   DEFAULTQUEUE_G(vfconf->rss_vf_vfh),
+			   KEYINDEX_G(vfconf->rss_vf_vfh),
+			   vfconf->rss_vf_vfl);
+	}
+	return 0;
+}
+
+static int rss_vf_config_open(struct inode *inode, struct file *file)
+{
+	struct adapter *adapter = inode->i_private;
+	struct seq_tab *p;
+	struct rss_vf_conf *vfconf;
+	int vf, vfcount = adapter->params.arch.vfcount;
+
+	p = seq_open_tab(file, vfcount, sizeof(*vfconf), 1, rss_vf_config_show);
+	if (!p)
+		return -ENOMEM;
+
+	vfconf = (struct rss_vf_conf *)p->data;
+	for (vf = 0; vf < vfcount; vf++) {
+		t4_read_rss_vf_config(adapter, vf, &vfconf[vf].rss_vf_vfl,
+				      &vfconf[vf].rss_vf_vfh, true);
+	}
+	return 0;
+}
+
+static const struct file_operations rss_vf_config_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = rss_vf_config_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release_private
+};
+
+#ifdef CONFIG_CHELSIO_T4_DCB
+
+/* Data Center Briging information for each port.
+ */
+static int dcb_info_show(struct seq_file *seq, void *v)
+{
+	struct adapter *adap = seq->private;
+
+	if (v == SEQ_START_TOKEN) {
+		seq_puts(seq, "Data Center Bridging Information\n");
+	} else {
+		int port = (uintptr_t)v - 2;
+		struct net_device *dev = adap->port[port];
+		struct port_info *pi = netdev2pinfo(dev);
+		struct port_dcb_info *dcb = &pi->dcb;
+
+		seq_puts(seq, "\n");
+		seq_printf(seq, "Port: %d (DCB negotiated: %s)\n",
+			   port,
+			   cxgb4_dcb_enabled(dev) ? "yes" : "no");
+
+		if (cxgb4_dcb_enabled(dev))
+			seq_printf(seq, "[ DCBx Version %s ]\n",
+				   dcb_ver_array[dcb->dcb_version]);
+
+		if (dcb->msgs) {
+			int i;
+
+			seq_puts(seq, "\n  Index\t\t\t  :\t");
+			for (i = 0; i < 8; i++)
+				seq_printf(seq, " %3d", i);
+			seq_puts(seq, "\n\n");
+		}
+
+		if (dcb->msgs & CXGB4_DCB_FW_PGID) {
+			int prio, pgid;
+
+			seq_puts(seq, "  Priority Group IDs\t  :\t");
+			for (prio = 0; prio < 8; prio++) {
+				pgid = (dcb->pgid >> 4 * (7 - prio)) & 0xf;
+				seq_printf(seq, " %3d", pgid);
+			}
+			seq_puts(seq, "\n");
+		}
+
+		if (dcb->msgs & CXGB4_DCB_FW_PGRATE) {
+			int pg;
+
+			seq_puts(seq, "  Priority Group BW(%)\t  :\t");
+			for (pg = 0; pg < 8; pg++)
+				seq_printf(seq, " %3d", dcb->pgrate[pg]);
+			seq_puts(seq, "\n");
+
+			if (dcb->dcb_version == FW_PORT_DCB_VER_IEEE) {
+				seq_puts(seq, "  TSA Algorithm\t\t  :\t");
+				for (pg = 0; pg < 8; pg++)
+					seq_printf(seq, " %3d", dcb->tsa[pg]);
+				seq_puts(seq, "\n");
+			}
+
+			seq_printf(seq, "  Max PG Traffic Classes  [%3d  ]\n",
+				   dcb->pg_num_tcs_supported);
+
+			seq_puts(seq, "\n");
+		}
+
+		if (dcb->msgs & CXGB4_DCB_FW_PRIORATE) {
+			int prio;
+
+			seq_puts(seq, "  Priority Rate\t:\t");
+			for (prio = 0; prio < 8; prio++)
+				seq_printf(seq, " %3d", dcb->priorate[prio]);
+			seq_puts(seq, "\n");
+		}
+
+		if (dcb->msgs & CXGB4_DCB_FW_PFC) {
+			int prio;
+
+			seq_puts(seq, "  Priority Flow Control   :\t");
+			for (prio = 0; prio < 8; prio++) {
+				int pfcen = (dcb->pfcen >> 1 * (7 - prio))
+					    & 0x1;
+				seq_printf(seq, " %3d", pfcen);
+			}
+			seq_puts(seq, "\n");
+
+			seq_printf(seq, "  Max PFC Traffic Classes [%3d  ]\n",
+				   dcb->pfc_num_tcs_supported);
+
+			seq_puts(seq, "\n");
+		}
+
+		if (dcb->msgs & CXGB4_DCB_FW_APP_ID) {
+			int app, napps;
+
+			seq_puts(seq, "  Application Information:\n");
+			seq_puts(seq, "  App    Priority    Selection         Protocol\n");
+			seq_puts(seq, "  Index  Map         Field             ID\n");
+			for (app = 0, napps = 0;
+			     app < CXGB4_MAX_DCBX_APP_SUPPORTED; app++) {
+				struct app_priority *ap;
+				static const char * const sel_names[] = {
+					"Ethertype",
+					"Socket TCP",
+					"Socket UDP",
+					"Socket All",
+				};
+				const char *sel_name;
+
+				ap = &dcb->app_priority[app];
+				/* skip empty slots */
+				if (ap->protocolid == 0)
+					continue;
+				napps++;
+
+				if (ap->sel_field < ARRAY_SIZE(sel_names))
+					sel_name = sel_names[ap->sel_field];
+				else
+					sel_name = "UNKNOWN";
+
+				seq_printf(seq, "  %3d    %#04x        %-10s (%d)    %#06x (%d)\n",
+					   app,
+					   ap->user_prio_map,
+					   sel_name, ap->sel_field,
+					   ap->protocolid, ap->protocolid);
+			}
+			if (napps == 0)
+				seq_puts(seq, "    --- None ---\n");
+		}
+	}
+	return 0;
+}
+
+static inline void *dcb_info_get_idx(struct adapter *adap, loff_t pos)
+{
+	return (pos <= adap->params.nports
+		? (void *)((uintptr_t)pos + 1)
+		: NULL);
+}
+
+static void *dcb_info_start(struct seq_file *seq, loff_t *pos)
+{
+	struct adapter *adap = seq->private;
+
+	return (*pos
+		? dcb_info_get_idx(adap, *pos)
+		: SEQ_START_TOKEN);
+}
+
+static void dcb_info_stop(struct seq_file *seq, void *v)
+{
+}
+
+static void *dcb_info_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+	struct adapter *adap = seq->private;
+
+	(*pos)++;
+	return dcb_info_get_idx(adap, *pos);
+}
+
+static const struct seq_operations dcb_info_seq_ops = {
+	.start = dcb_info_start,
+	.next  = dcb_info_next,
+	.stop  = dcb_info_stop,
+	.show  = dcb_info_show
+};
+
+static int dcb_info_open(struct inode *inode, struct file *file)
+{
+	int res = seq_open(file, &dcb_info_seq_ops);
+
+	if (!res) {
+		struct seq_file *seq = file->private_data;
+
+		seq->private = inode->i_private;
+	}
+	return res;
+}
+
+static const struct file_operations dcb_info_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = dcb_info_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release,
+};
+#endif /* CONFIG_CHELSIO_T4_DCB */
+
+static int resources_show(struct seq_file *seq, void *v)
+{
+	struct adapter *adapter = seq->private;
+	struct pf_resources *pfres = &adapter->params.pfres;
+
+	#define S(desc, fmt, var) \
+		seq_printf(seq, "%-60s " fmt "\n", \
+			   desc " (" #var "):", pfres->var)
+
+	S("Virtual Interfaces", "%d", nvi);
+	S("Egress Queues", "%d", neq);
+	S("Ethernet Control", "%d", nethctrl);
+	S("Ingress Queues/w Free Lists/Interrupts", "%d", niqflint);
+	S("Ingress Queues", "%d", niq);
+	S("Traffic Class", "%d", tc);
+	S("Port Access Rights Mask", "%#x", pmask);
+	S("MAC Address Filters", "%d", nexactf);
+	S("Firmware Command Read Capabilities", "%#x", r_caps);
+	S("Firmware Command Write/Execute Capabilities", "%#x", wx_caps);
+
+	#undef S
+
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(resources);
+
+/**
+ * ethqset2pinfo - return port_info of an Ethernet Queue Set
+ * @adap: the adapter
+ * @qset: Ethernet Queue Set
+ */
+static inline struct port_info *ethqset2pinfo(struct adapter *adap, int qset)
+{
+	int pidx;
+
+	for_each_port(adap, pidx) {
+		struct port_info *pi = adap2pinfo(adap, pidx);
+
+		if (qset >= pi->first_qset &&
+		    qset < pi->first_qset + pi->nqsets)
+			return pi;
+	}
+
+	/* should never happen! */
+	BUG();
+	return NULL;
+}
+
+static int sge_qinfo_uld_txq_entries(const struct adapter *adap, int uld)
+{
+	const struct sge_uld_txq_info *utxq_info = adap->sge.uld_txq_info[uld];
+
+	if (!utxq_info)
+		return 0;
+
+	return DIV_ROUND_UP(utxq_info->ntxq, 4);
+}
+
+static int sge_qinfo_uld_rspq_entries(const struct adapter *adap, int uld,
+				      bool ciq)
+{
+	const struct sge_uld_rxq_info *urxq_info = adap->sge.uld_rxq_info[uld];
+
+	if (!urxq_info)
+		return 0;
+
+	return ciq ? DIV_ROUND_UP(urxq_info->nciq, 4) :
+		     DIV_ROUND_UP(urxq_info->nrxq, 4);
+}
+
+static int sge_qinfo_uld_rxq_entries(const struct adapter *adap, int uld)
+{
+	return sge_qinfo_uld_rspq_entries(adap, uld, false);
+}
+
+static int sge_qinfo_uld_ciq_entries(const struct adapter *adap, int uld)
+{
+	return sge_qinfo_uld_rspq_entries(adap, uld, true);
+}
+
+static int sge_qinfo_show(struct seq_file *seq, void *v)
+{
+	int eth_entries, ctrl_entries, eohw_entries = 0, eosw_entries = 0;
+	int uld_rxq_entries[CXGB4_ULD_MAX] = { 0 };
+	int uld_ciq_entries[CXGB4_ULD_MAX] = { 0 };
+	int uld_txq_entries[CXGB4_TX_MAX] = { 0 };
+	const struct sge_uld_txq_info *utxq_info;
+	const struct sge_uld_rxq_info *urxq_info;
+	struct cxgb4_tc_port_mqprio *port_mqprio;
+	struct adapter *adap = seq->private;
+	int i, j, n, r = (uintptr_t)v - 1;
+	struct sge *s = &adap->sge;
+
+	eth_entries = DIV_ROUND_UP(adap->sge.ethqsets, 4);
+	ctrl_entries = DIV_ROUND_UP(MAX_CTRL_QUEUES, 4);
+
+	if (r)
+		seq_putc(seq, '\n');
+
+#define S3(fmt_spec, s, v) \
+do { \
+	seq_printf(seq, "%-12s", s); \
+	for (i = 0; i < n; ++i) \
+		seq_printf(seq, " %16" fmt_spec, v); \
+	seq_putc(seq, '\n'); \
+} while (0)
+#define S(s, v) S3("s", s, v)
+#define T3(fmt_spec, s, v) S3(fmt_spec, s, tx[i].v)
+#define T(s, v) S3("u", s, tx[i].v)
+#define TL(s, v) T3("lu", s, v)
+#define R3(fmt_spec, s, v) S3(fmt_spec, s, rx[i].v)
+#define R(s, v) S3("u", s, rx[i].v)
+#define RL(s, v) R3("lu", s, v)
+
+	if (r < eth_entries) {
+		int base_qset = r * 4;
+		const struct sge_eth_rxq *rx = &s->ethrxq[base_qset];
+		const struct sge_eth_txq *tx = &s->ethtxq[base_qset];
+
+		n = min(4, s->ethqsets - 4 * r);
+
+		S("QType:", "Ethernet");
+		S("Interface:",
+		  rx[i].rspq.netdev ? rx[i].rspq.netdev->name : "N/A");
+		T("TxQ ID:", q.cntxt_id);
+		T("TxQ size:", q.size);
+		T("TxQ inuse:", q.in_use);
+		T("TxQ CIDX:", q.cidx);
+		T("TxQ PIDX:", q.pidx);
+#ifdef CONFIG_CHELSIO_T4_DCB
+		T("DCB Prio:", dcb_prio);
+		S3("u", "DCB PGID:",
+		   (ethqset2pinfo(adap, base_qset + i)->dcb.pgid >>
+		    4*(7-tx[i].dcb_prio)) & 0xf);
+		S3("u", "DCB PFC:",
+		   (ethqset2pinfo(adap, base_qset + i)->dcb.pfcen >>
+		    1*(7-tx[i].dcb_prio)) & 0x1);
+#endif
+		R("RspQ ID:", rspq.abs_id);
+		R("RspQ size:", rspq.size);
+		R("RspQE size:", rspq.iqe_len);
+		R("RspQ CIDX:", rspq.cidx);
+		R("RspQ Gen:", rspq.gen);
+		S3("u", "Intr delay:", qtimer_val(adap, &rx[i].rspq));
+		S3("u", "Intr pktcnt:", s->counter_val[rx[i].rspq.pktcnt_idx]);
+		R("FL ID:", fl.cntxt_id);
+		R("FL size:", fl.size - 8);
+		R("FL pend:", fl.pend_cred);
+		R("FL avail:", fl.avail);
+		R("FL PIDX:", fl.pidx);
+		R("FL CIDX:", fl.cidx);
+		RL("RxPackets:", stats.pkts);
+		RL("RxCSO:", stats.rx_cso);
+		RL("VLANxtract:", stats.vlan_ex);
+		RL("LROmerged:", stats.lro_merged);
+		RL("LROpackets:", stats.lro_pkts);
+		RL("RxDrops:", stats.rx_drops);
+		RL("RxBadPkts:", stats.bad_rx_pkts);
+		TL("TSO:", tso);
+		TL("USO:", uso);
+		TL("TxCSO:", tx_cso);
+		TL("VLANins:", vlan_ins);
+		TL("TxQFull:", q.stops);
+		TL("TxQRestarts:", q.restarts);
+		TL("TxMapErr:", mapping_err);
+		RL("FLAllocErr:", fl.alloc_failed);
+		RL("FLLrgAlcErr:", fl.large_alloc_failed);
+		RL("FLMapErr:", fl.mapping_err);
+		RL("FLLow:", fl.low);
+		RL("FLStarving:", fl.starving);
+
+		goto out;
+	}
+
+	r -= eth_entries;
+	for_each_port(adap, j) {
+		struct port_info *pi = adap2pinfo(adap, j);
+		const struct sge_eth_rxq *rx;
+
+		mutex_lock(&pi->vi_mirror_mutex);
+		if (!pi->vi_mirror_count) {
+			mutex_unlock(&pi->vi_mirror_mutex);
+			continue;
+		}
+
+		if (r >= DIV_ROUND_UP(pi->nmirrorqsets, 4)) {
+			r -= DIV_ROUND_UP(pi->nmirrorqsets, 4);
+			mutex_unlock(&pi->vi_mirror_mutex);
+			continue;
+		}
+
+		rx = &s->mirror_rxq[j][r * 4];
+		n = min(4, pi->nmirrorqsets - 4 * r);
+
+		S("QType:", "Mirror-Rxq");
+		S("Interface:",
+		  rx[i].rspq.netdev ? rx[i].rspq.netdev->name : "N/A");
+		R("RspQ ID:", rspq.abs_id);
+		R("RspQ size:", rspq.size);
+		R("RspQE size:", rspq.iqe_len);
+		R("RspQ CIDX:", rspq.cidx);
+		R("RspQ Gen:", rspq.gen);
+		S3("u", "Intr delay:", qtimer_val(adap, &rx[i].rspq));
+		S3("u", "Intr pktcnt:", s->counter_val[rx[i].rspq.pktcnt_idx]);
+		R("FL ID:", fl.cntxt_id);
+		R("FL size:", fl.size - 8);
+		R("FL pend:", fl.pend_cred);
+		R("FL avail:", fl.avail);
+		R("FL PIDX:", fl.pidx);
+		R("FL CIDX:", fl.cidx);
+		RL("RxPackets:", stats.pkts);
+		RL("RxCSO:", stats.rx_cso);
+		RL("VLANxtract:", stats.vlan_ex);
+		RL("LROmerged:", stats.lro_merged);
+		RL("LROpackets:", stats.lro_pkts);
+		RL("RxDrops:", stats.rx_drops);
+		RL("RxBadPkts:", stats.bad_rx_pkts);
+		RL("FLAllocErr:", fl.alloc_failed);
+		RL("FLLrgAlcErr:", fl.large_alloc_failed);
+		RL("FLMapErr:", fl.mapping_err);
+		RL("FLLow:", fl.low);
+		RL("FLStarving:", fl.starving);
+
+		mutex_unlock(&pi->vi_mirror_mutex);
+		goto out;
+	}
+
+	if (!adap->tc_mqprio)
+		goto skip_mqprio;
+
+	mutex_lock(&adap->tc_mqprio->mqprio_mutex);
+	if (!refcount_read(&adap->tc_mqprio->refcnt)) {
+		mutex_unlock(&adap->tc_mqprio->mqprio_mutex);
+		goto skip_mqprio;
+	}
+
+	eohw_entries = DIV_ROUND_UP(adap->sge.eoqsets, 4);
+	if (r < eohw_entries) {
+		int base_qset = r * 4;
+		const struct sge_ofld_rxq *rx = &s->eohw_rxq[base_qset];
+		const struct sge_eohw_txq *tx = &s->eohw_txq[base_qset];
+
+		n = min(4, s->eoqsets - 4 * r);
+
+		S("QType:", "ETHOFLD");
+		S("Interface:",
+		  rx[i].rspq.netdev ? rx[i].rspq.netdev->name : "N/A");
+		T("TxQ ID:", q.cntxt_id);
+		T("TxQ size:", q.size);
+		T("TxQ inuse:", q.in_use);
+		T("TxQ CIDX:", q.cidx);
+		T("TxQ PIDX:", q.pidx);
+		R("RspQ ID:", rspq.abs_id);
+		R("RspQ size:", rspq.size);
+		R("RspQE size:", rspq.iqe_len);
+		R("RspQ CIDX:", rspq.cidx);
+		R("RspQ Gen:", rspq.gen);
+		S3("u", "Intr delay:", qtimer_val(adap, &rx[i].rspq));
+		S3("u", "Intr pktcnt:", s->counter_val[rx[i].rspq.pktcnt_idx]);
+		R("FL ID:", fl.cntxt_id);
+		S3("u", "FL size:", rx->fl.size ? rx->fl.size - 8 : 0);
+		R("FL pend:", fl.pend_cred);
+		R("FL avail:", fl.avail);
+		R("FL PIDX:", fl.pidx);
+		R("FL CIDX:", fl.cidx);
+		RL("RxPackets:", stats.pkts);
+		RL("RxImm:", stats.imm);
+		RL("RxAN", stats.an);
+		RL("RxNoMem", stats.nomem);
+		TL("TSO:", tso);
+		TL("USO:", uso);
+		TL("TxCSO:", tx_cso);
+		TL("VLANins:", vlan_ins);
+		TL("TxQFull:", q.stops);
+		TL("TxQRestarts:", q.restarts);
+		TL("TxMapErr:", mapping_err);
+		RL("FLAllocErr:", fl.alloc_failed);
+		RL("FLLrgAlcErr:", fl.large_alloc_failed);
+		RL("FLMapErr:", fl.mapping_err);
+		RL("FLLow:", fl.low);
+		RL("FLStarving:", fl.starving);
+
+		mutex_unlock(&adap->tc_mqprio->mqprio_mutex);
+		goto out;
+	}
+
+	r -= eohw_entries;
+	for (j = 0; j < adap->params.nports; j++) {
+		int entries;
+		u8 tc;
+
+		port_mqprio = &adap->tc_mqprio->port_mqprio[j];
+		entries = 0;
+		for (tc = 0; tc < port_mqprio->mqprio.qopt.num_tc; tc++)
+			entries += port_mqprio->mqprio.qopt.count[tc];
+
+		if (!entries)
+			continue;
+
+		eosw_entries = DIV_ROUND_UP(entries, 4);
+		if (r < eosw_entries) {
+			const struct sge_eosw_txq *tx;
+
+			n = min(4, entries - 4 * r);
+			tx = &port_mqprio->eosw_txq[4 * r];
+
+			S("QType:", "EOSW-TXQ");
+			S("Interface:",
+			  adap->port[j] ? adap->port[j]->name : "N/A");
+			T("EOTID:", hwtid);
+			T("HWQID:", hwqid);
+			T("State:", state);
+			T("Size:", ndesc);
+			T("In-Use:", inuse);
+			T("Credits:", cred);
+			T("Compl:", ncompl);
+			T("Last-Compl:", last_compl);
+			T("PIDX:", pidx);
+			T("Last-PIDX:", last_pidx);
+			T("CIDX:", cidx);
+			T("Last-CIDX:", last_cidx);
+			T("FLOWC-IDX:", flowc_idx);
+
+			mutex_unlock(&adap->tc_mqprio->mqprio_mutex);
+			goto out;
+		}
+
+		r -= eosw_entries;
+	}
+	mutex_unlock(&adap->tc_mqprio->mqprio_mutex);
+
+skip_mqprio:
+	if (!is_uld(adap))
+		goto skip_uld;
+
+	mutex_lock(&uld_mutex);
+	if (s->uld_txq_info)
+		for (i = 0; i < ARRAY_SIZE(uld_txq_entries); i++)
+			uld_txq_entries[i] = sge_qinfo_uld_txq_entries(adap, i);
+
+	if (s->uld_rxq_info) {
+		for (i = 0; i < ARRAY_SIZE(uld_rxq_entries); i++) {
+			uld_rxq_entries[i] = sge_qinfo_uld_rxq_entries(adap, i);
+			uld_ciq_entries[i] = sge_qinfo_uld_ciq_entries(adap, i);
+		}
+	}
+
+	if (r < uld_txq_entries[CXGB4_TX_OFLD]) {
+		const struct sge_uld_txq *tx;
+
+		utxq_info = s->uld_txq_info[CXGB4_TX_OFLD];
+		tx = &utxq_info->uldtxq[r * 4];
+		n = min(4, utxq_info->ntxq - 4 * r);
+
+		S("QType:", "OFLD-TXQ");
+		T("TxQ ID:", q.cntxt_id);
+		T("TxQ size:", q.size);
+		T("TxQ inuse:", q.in_use);
+		T("TxQ CIDX:", q.cidx);
+		T("TxQ PIDX:", q.pidx);
+
+		goto unlock;
+	}
+
+	r -= uld_txq_entries[CXGB4_TX_OFLD];
+	if (r < uld_rxq_entries[CXGB4_ULD_RDMA]) {
+		const struct sge_ofld_rxq *rx;
+
+		urxq_info = s->uld_rxq_info[CXGB4_ULD_RDMA];
+		rx = &urxq_info->uldrxq[r * 4];
+		n = min(4, urxq_info->nrxq - 4 * r);
+
+		S("QType:", "RDMA-CPL");
+		S("Interface:",
+		  rx[i].rspq.netdev ? rx[i].rspq.netdev->name : "N/A");
+		R("RspQ ID:", rspq.abs_id);
+		R("RspQ size:", rspq.size);
+		R("RspQE size:", rspq.iqe_len);
+		R("RspQ CIDX:", rspq.cidx);
+		R("RspQ Gen:", rspq.gen);
+		S3("u", "Intr delay:", qtimer_val(adap, &rx[i].rspq));
+		S3("u", "Intr pktcnt:",	s->counter_val[rx[i].rspq.pktcnt_idx]);
+		R("FL ID:", fl.cntxt_id);
+		R("FL size:", fl.size - 8);
+		R("FL pend:", fl.pend_cred);
+		R("FL avail:", fl.avail);
+		R("FL PIDX:", fl.pidx);
+		R("FL CIDX:", fl.cidx);
+
+		goto unlock;
+	}
+
+	r -= uld_rxq_entries[CXGB4_ULD_RDMA];
+	if (r < uld_ciq_entries[CXGB4_ULD_RDMA]) {
+		const struct sge_ofld_rxq *rx;
+		int ciq_idx = 0;
+
+		urxq_info = s->uld_rxq_info[CXGB4_ULD_RDMA];
+		ciq_idx = urxq_info->nrxq + (r * 4);
+		rx = &urxq_info->uldrxq[ciq_idx];
+		n = min(4, urxq_info->nciq - 4 * r);
+
+		S("QType:", "RDMA-CIQ");
+		S("Interface:",
+		  rx[i].rspq.netdev ? rx[i].rspq.netdev->name : "N/A");
+		R("RspQ ID:", rspq.abs_id);
+		R("RspQ size:", rspq.size);
+		R("RspQE size:", rspq.iqe_len);
+		R("RspQ CIDX:", rspq.cidx);
+		R("RspQ Gen:", rspq.gen);
+		S3("u", "Intr delay:", qtimer_val(adap, &rx[i].rspq));
+		S3("u", "Intr pktcnt:",	s->counter_val[rx[i].rspq.pktcnt_idx]);
+
+		goto unlock;
+	}
+
+	r -= uld_ciq_entries[CXGB4_ULD_RDMA];
+	if (r < uld_rxq_entries[CXGB4_ULD_ISCSI]) {
+		const struct sge_ofld_rxq *rx;
+
+		urxq_info = s->uld_rxq_info[CXGB4_ULD_ISCSI];
+		rx = &urxq_info->uldrxq[r * 4];
+		n = min(4, urxq_info->nrxq - 4 * r);
+
+		S("QType:", "iSCSI");
+		R("RspQ ID:", rspq.abs_id);
+		R("RspQ size:", rspq.size);
+		R("RspQE size:", rspq.iqe_len);
+		R("RspQ CIDX:", rspq.cidx);
+		R("RspQ Gen:", rspq.gen);
+		S3("u", "Intr delay:", qtimer_val(adap, &rx[i].rspq));
+		S3("u", "Intr pktcnt:",	s->counter_val[rx[i].rspq.pktcnt_idx]);
+		R("FL ID:", fl.cntxt_id);
+		R("FL size:", fl.size - 8);
+		R("FL pend:", fl.pend_cred);
+		R("FL avail:", fl.avail);
+		R("FL PIDX:", fl.pidx);
+		R("FL CIDX:", fl.cidx);
+
+		goto unlock;
+	}
+
+	r -= uld_rxq_entries[CXGB4_ULD_ISCSI];
+	if (r < uld_rxq_entries[CXGB4_ULD_ISCSIT]) {
+		const struct sge_ofld_rxq *rx;
+
+		urxq_info = s->uld_rxq_info[CXGB4_ULD_ISCSIT];
+		rx = &urxq_info->uldrxq[r * 4];
+		n = min(4, urxq_info->nrxq - 4 * r);
+
+		S("QType:", "iSCSIT");
+		R("RspQ ID:", rspq.abs_id);
+		R("RspQ size:", rspq.size);
+		R("RspQE size:", rspq.iqe_len);
+		R("RspQ CIDX:", rspq.cidx);
+		R("RspQ Gen:", rspq.gen);
+		S3("u", "Intr delay:", qtimer_val(adap, &rx[i].rspq));
+		S3("u", "Intr pktcnt:",	s->counter_val[rx[i].rspq.pktcnt_idx]);
+		R("FL ID:", fl.cntxt_id);
+		R("FL size:", fl.size - 8);
+		R("FL pend:", fl.pend_cred);
+		R("FL avail:", fl.avail);
+		R("FL PIDX:", fl.pidx);
+		R("FL CIDX:", fl.cidx);
+
+		goto unlock;
+	}
+
+	r -= uld_rxq_entries[CXGB4_ULD_ISCSIT];
+	if (r < uld_rxq_entries[CXGB4_ULD_TLS]) {
+		const struct sge_ofld_rxq *rx;
+
+		urxq_info = s->uld_rxq_info[CXGB4_ULD_TLS];
+		rx = &urxq_info->uldrxq[r * 4];
+		n = min(4, urxq_info->nrxq - 4 * r);
+
+		S("QType:", "TLS");
+		R("RspQ ID:", rspq.abs_id);
+		R("RspQ size:", rspq.size);
+		R("RspQE size:", rspq.iqe_len);
+		R("RspQ CIDX:", rspq.cidx);
+		R("RspQ Gen:", rspq.gen);
+		S3("u", "Intr delay:", qtimer_val(adap, &rx[i].rspq));
+		S3("u", "Intr pktcnt:",	s->counter_val[rx[i].rspq.pktcnt_idx]);
+		R("FL ID:", fl.cntxt_id);
+		R("FL size:", fl.size - 8);
+		R("FL pend:", fl.pend_cred);
+		R("FL avail:", fl.avail);
+		R("FL PIDX:", fl.pidx);
+		R("FL CIDX:", fl.cidx);
+
+		goto unlock;
+	}
+
+	r -= uld_rxq_entries[CXGB4_ULD_TLS];
+	if (r < uld_txq_entries[CXGB4_TX_CRYPTO]) {
+		const struct sge_ofld_rxq *rx;
+		const struct sge_uld_txq *tx;
+
+		utxq_info = s->uld_txq_info[CXGB4_TX_CRYPTO];
+		urxq_info = s->uld_rxq_info[CXGB4_ULD_CRYPTO];
+		tx = &utxq_info->uldtxq[r * 4];
+		rx = &urxq_info->uldrxq[r * 4];
+		n = min(4, utxq_info->ntxq - 4 * r);
+
+		S("QType:", "Crypto");
+		T("TxQ ID:", q.cntxt_id);
+		T("TxQ size:", q.size);
+		T("TxQ inuse:", q.in_use);
+		T("TxQ CIDX:", q.cidx);
+		T("TxQ PIDX:", q.pidx);
+		R("RspQ ID:", rspq.abs_id);
+		R("RspQ size:", rspq.size);
+		R("RspQE size:", rspq.iqe_len);
+		R("RspQ CIDX:", rspq.cidx);
+		R("RspQ Gen:", rspq.gen);
+		S3("u", "Intr delay:", qtimer_val(adap, &rx[i].rspq));
+		S3("u", "Intr pktcnt:",	s->counter_val[rx[i].rspq.pktcnt_idx]);
+		R("FL ID:", fl.cntxt_id);
+		R("FL size:", fl.size - 8);
+		R("FL pend:", fl.pend_cred);
+		R("FL avail:", fl.avail);
+		R("FL PIDX:", fl.pidx);
+		R("FL CIDX:", fl.cidx);
+
+		goto unlock;
+	}
+
+	r -= uld_txq_entries[CXGB4_TX_CRYPTO];
+	mutex_unlock(&uld_mutex);
+
+skip_uld:
+	if (r < ctrl_entries) {
+		const struct sge_ctrl_txq *tx = &s->ctrlq[r * 4];
+
+		n = min(4, adap->params.nports - 4 * r);
+
+		S("QType:", "Control");
+		T("TxQ ID:", q.cntxt_id);
+		T("TxQ size:", q.size);
+		T("TxQ inuse:", q.in_use);
+		T("TxQ CIDX:", q.cidx);
+		T("TxQ PIDX:", q.pidx);
+		TL("TxQFull:", q.stops);
+		TL("TxQRestarts:", q.restarts);
+
+		goto out;
+	}
+
+	r -= ctrl_entries;
+	if (r < 1) {
+		const struct sge_rspq *evtq = &s->fw_evtq;
+
+		seq_printf(seq, "%-12s %16s\n", "QType:", "FW event queue");
+		seq_printf(seq, "%-12s %16u\n", "RspQ ID:", evtq->abs_id);
+		seq_printf(seq, "%-12s %16u\n", "RspQ size:", evtq->size);
+		seq_printf(seq, "%-12s %16u\n", "RspQE size:", evtq->iqe_len);
+		seq_printf(seq, "%-12s %16u\n", "RspQ CIDX:", evtq->cidx);
+		seq_printf(seq, "%-12s %16u\n", "RspQ Gen:", evtq->gen);
+		seq_printf(seq, "%-12s %16u\n", "Intr delay:",
+			   qtimer_val(adap, evtq));
+		seq_printf(seq, "%-12s %16u\n", "Intr pktcnt:",
+			   s->counter_val[evtq->pktcnt_idx]);
+
+		goto out;
+	}
+
+#undef R
+#undef RL
+#undef T
+#undef TL
+#undef S
+#undef R3
+#undef T3
+#undef S3
+out:
+	return 0;
+
+unlock:
+	mutex_unlock(&uld_mutex);
+	return 0;
+}
+
+static int sge_queue_entries(struct adapter *adap)
+{
+	int i, tot_uld_entries = 0, eohw_entries = 0, eosw_entries = 0;
+	int mirror_rxq_entries = 0;
+
+	if (adap->tc_mqprio) {
+		struct cxgb4_tc_port_mqprio *port_mqprio;
+		u8 tc;
+
+		mutex_lock(&adap->tc_mqprio->mqprio_mutex);
+		if (adap->sge.eohw_txq)
+			eohw_entries = DIV_ROUND_UP(adap->sge.eoqsets, 4);
+
+		for (i = 0; i < adap->params.nports; i++) {
+			u32 entries = 0;
+
+			port_mqprio = &adap->tc_mqprio->port_mqprio[i];
+			for (tc = 0; tc < port_mqprio->mqprio.qopt.num_tc; tc++)
+				entries += port_mqprio->mqprio.qopt.count[tc];
+
+			if (entries)
+				eosw_entries += DIV_ROUND_UP(entries, 4);
+		}
+		mutex_unlock(&adap->tc_mqprio->mqprio_mutex);
+	}
+
+	for_each_port(adap, i) {
+		struct port_info *pi = adap2pinfo(adap, i);
+
+		mutex_lock(&pi->vi_mirror_mutex);
+		if (pi->vi_mirror_count)
+			mirror_rxq_entries += DIV_ROUND_UP(pi->nmirrorqsets, 4);
+		mutex_unlock(&pi->vi_mirror_mutex);
+	}
+
+	if (!is_uld(adap))
+		goto lld_only;
+
+	mutex_lock(&uld_mutex);
+	for (i = 0; i < CXGB4_TX_MAX; i++)
+		tot_uld_entries += sge_qinfo_uld_txq_entries(adap, i);
+
+	for (i = 0; i < CXGB4_ULD_MAX; i++) {
+		tot_uld_entries += sge_qinfo_uld_rxq_entries(adap, i);
+		tot_uld_entries += sge_qinfo_uld_ciq_entries(adap, i);
+	}
+	mutex_unlock(&uld_mutex);
+
+lld_only:
+	return DIV_ROUND_UP(adap->sge.ethqsets, 4) + mirror_rxq_entries +
+	       eohw_entries + eosw_entries + tot_uld_entries +
+	       DIV_ROUND_UP(MAX_CTRL_QUEUES, 4) + 1;
+}
+
+static void *sge_queue_start(struct seq_file *seq, loff_t *pos)
+{
+	int entries = sge_queue_entries(seq->private);
+
+	return *pos < entries ? (void *)((uintptr_t)*pos + 1) : NULL;
+}
+
+static void sge_queue_stop(struct seq_file *seq, void *v)
+{
+}
+
+static void *sge_queue_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+	int entries = sge_queue_entries(seq->private);
+
+	++*pos;
+	return *pos < entries ? (void *)((uintptr_t)*pos + 1) : NULL;
+}
+
+static const struct seq_operations sge_qinfo_seq_ops = {
+	.start = sge_queue_start,
+	.next  = sge_queue_next,
+	.stop  = sge_queue_stop,
+	.show  = sge_qinfo_show
+};
+
+static int sge_qinfo_open(struct inode *inode, struct file *file)
+{
+	int res = seq_open(file, &sge_qinfo_seq_ops);
+
+	if (!res) {
+		struct seq_file *seq = file->private_data;
+
+		seq->private = inode->i_private;
+	}
+	return res;
+}
+
+static const struct file_operations sge_qinfo_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = sge_qinfo_open,
+	.read    = seq_read,
+	.llseek  = seq_lseek,
+	.release = seq_release,
+};
+
+int mem_open(struct inode *inode, struct file *file)
+{
+	unsigned int mem;
+	struct adapter *adap;
+
+	file->private_data = inode->i_private;
+
+	mem = (uintptr_t)file->private_data & 0x7;
+	adap = file->private_data - mem;
+
+	(void)t4_fwcache(adap, FW_PARAM_DEV_FWCACHE_FLUSH);
+
+	return 0;
+}
+
+static ssize_t mem_read(struct file *file, char __user *buf, size_t count,
+			loff_t *ppos)
+{
+	loff_t pos = *ppos;
+	loff_t avail = file_inode(file)->i_size;
+	unsigned int mem = (uintptr_t)file->private_data & 0x7;
+	struct adapter *adap = file->private_data - mem;
+	__be32 *data;
+	int ret;
+
+	if (pos < 0)
+		return -EINVAL;
+	if (pos >= avail)
+		return 0;
+	if (count > avail - pos)
+		count = avail - pos;
+
+	data = kvzalloc(count, GFP_KERNEL);
+	if (!data)
+		return -ENOMEM;
+
+	spin_lock(&adap->win0_lock);
+	ret = t4_memory_rw(adap, 0, mem, pos, count, data, T4_MEMORY_READ);
+	spin_unlock(&adap->win0_lock);
+	if (ret) {
+		kvfree(data);
+		return ret;
+	}
+	ret = copy_to_user(buf, data, count);
+
+	kvfree(data);
+	if (ret)
+		return -EFAULT;
+
+	*ppos = pos + count;
+	return count;
+}
+static const struct file_operations mem_debugfs_fops = {
+	.owner   = THIS_MODULE,
+	.open    = simple_open,
+	.read    = mem_read,
+	.llseek  = default_llseek,
+};
+
+static int tid_info_show(struct seq_file *seq, void *v)
+{
+	struct adapter *adap = seq->private;
+	const struct tid_info *t;
+	enum chip_type chip;
+
+	t = &adap->tids;
+	chip = CHELSIO_CHIP_VERSION(adap->params.chip);
+	if (t4_read_reg(adap, LE_DB_CONFIG_A) & HASHEN_F) {
+		unsigned int sb;
+		seq_printf(seq, "Connections in use: %u\n",
+			   atomic_read(&t->conns_in_use));
+
+		if (chip <= CHELSIO_T5)
+			sb = t4_read_reg(adap, LE_DB_SERVER_INDEX_A) / 4;
+		else
+			sb = t4_read_reg(adap, LE_DB_SRVR_START_INDEX_A);
+
+		if (sb) {
+			seq_printf(seq, "TID range: %u..%u/%u..%u", t->tid_base,
+				   sb - 1, adap->tids.hash_base,
+				   t->tid_base + t->ntids - 1);
+			seq_printf(seq, ", in use: %u/%u\n",
+				   atomic_read(&t->tids_in_use),
+				   atomic_read(&t->hash_tids_in_use));
+		} else if (adap->flags & CXGB4_FW_OFLD_CONN) {
+			seq_printf(seq, "TID range: %u..%u/%u..%u",
+				   t->aftid_base,
+				   t->aftid_end,
+				   adap->tids.hash_base,
+				   t->tid_base + t->ntids - 1);
+			seq_printf(seq, ", in use: %u/%u\n",
+				   atomic_read(&t->tids_in_use),
+				   atomic_read(&t->hash_tids_in_use));
+		} else {
+			seq_printf(seq, "TID range: %u..%u",
+				   adap->tids.hash_base,
+				   t->tid_base + t->ntids - 1);
+			seq_printf(seq, ", in use: %u\n",
+				   atomic_read(&t->hash_tids_in_use));
+		}
+	} else if (t->ntids) {
+		seq_printf(seq, "Connections in use: %u\n",
+			   atomic_read(&t->conns_in_use));
+
+		seq_printf(seq, "TID range: %u..%u", t->tid_base,
+			   t->tid_base + t->ntids - 1);
+		seq_printf(seq, ", in use: %u\n",
+			   atomic_read(&t->tids_in_use));
+	}
+
+	if (t->nstids)
+		seq_printf(seq, "STID range: %u..%u, in use-IPv4/IPv6: %u/%u\n",
+			   (!t->stid_base &&
+			   (chip <= CHELSIO_T5)) ?
+			   t->stid_base + 1 : t->stid_base,
+			   t->stid_base + t->nstids - 1,
+			   t->stids_in_use - t->v6_stids_in_use,
+			   t->v6_stids_in_use);
+
+	if (t->natids)
+		seq_printf(seq, "ATID range: 0..%u, in use: %u\n",
+			   t->natids - 1, t->atids_in_use);
+	seq_printf(seq, "FTID range: %u..%u\n", t->ftid_base,
+		   t->ftid_base + t->nftids - 1);
+	if (t->nsftids)
+		seq_printf(seq, "SFTID range: %u..%u in use: %u\n",
+			   t->sftid_base, t->sftid_base + t->nsftids - 2,
+			   t->sftids_in_use);
+	if (t->nhpftids)
+		seq_printf(seq, "HPFTID range: %u..%u\n", t->hpftid_base,
+			   t->hpftid_base + t->nhpftids - 1);
+	if (t->neotids)
+		seq_printf(seq, "EOTID range: %u..%u, in use: %u\n",
+			   t->eotid_base, t->eotid_base + t->neotids - 1,
+			   atomic_read(&t->eotids_in_use));
+	if (t->ntids)
+		seq_printf(seq, "HW TID usage: %u IP users, %u IPv6 users\n",
+			   t4_read_reg(adap, LE_DB_ACT_CNT_IPV4_A),
+			   t4_read_reg(adap, LE_DB_ACT_CNT_IPV6_A));
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(tid_info);
+
+static void add_debugfs_mem(struct adapter *adap, const char *name,
+			    unsigned int idx, unsigned int size_mb)
+{
+	debugfs_create_file_size(name, 0400, adap->debugfs_root,
+				 (void *)adap + idx, &mem_debugfs_fops,
+				 size_mb << 20);
+}
+
+static ssize_t blocked_fl_read(struct file *filp, char __user *ubuf,
+			       size_t count, loff_t *ppos)
+{
+	int len;
+	const struct adapter *adap = filp->private_data;
+	char *buf;
+	ssize_t size = (adap->sge.egr_sz + 3) / 4 +
+			adap->sge.egr_sz / 32 + 2; /* includes ,/\n/\0 */
+
+	buf = kzalloc(size, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	len = snprintf(buf, size - 1, "%*pb\n",
+		       adap->sge.egr_sz, adap->sge.blocked_fl);
+	len += sprintf(buf + len, "\n");
+	size = simple_read_from_buffer(ubuf, count, ppos, buf, len);
+	kfree(buf);
+	return size;
+}
+
+static ssize_t blocked_fl_write(struct file *filp, const char __user *ubuf,
+				size_t count, loff_t *ppos)
+{
+	int err;
+	unsigned long *t;
+	struct adapter *adap = filp->private_data;
+
+	t = bitmap_zalloc(adap->sge.egr_sz, GFP_KERNEL);
+	if (!t)
+		return -ENOMEM;
+
+	err = bitmap_parse_user(ubuf, count, t, adap->sge.egr_sz);
+	if (err) {
+		bitmap_free(t);
+		return err;
+	}
+
+	bitmap_copy(adap->sge.blocked_fl, t, adap->sge.egr_sz);
+	bitmap_free(t);
+	return count;
+}
+
+static const struct file_operations blocked_fl_fops = {
+	.owner   = THIS_MODULE,
+	.open    = simple_open,
+	.read    = blocked_fl_read,
+	.write   = blocked_fl_write,
+	.llseek  = generic_file_llseek,
+};
+
+static void mem_region_show(struct seq_file *seq, const char *name,
+			    unsigned int from, unsigned int to)
+{
+	char buf[40];
+
+	string_get_size((u64)to - from + 1, 1, STRING_UNITS_2, buf,
+			sizeof(buf));
+	seq_printf(seq, "%-15s %#x-%#x [%s]\n", name, from, to, buf);
+}
+
+static int meminfo_show(struct seq_file *seq, void *v)
+{
+	static const char * const memory[] = { "EDC0:", "EDC1:", "MC:",
+					       "MC0:", "MC1:", "HMA:"};
+	struct adapter *adap = seq->private;
+	struct cudbg_meminfo meminfo;
+	int i, rc;
+
+	memset(&meminfo, 0, sizeof(struct cudbg_meminfo));
+	rc = cudbg_fill_meminfo(adap, &meminfo);
+	if (rc)
+		return -ENXIO;
+
+	for (i = 0; i < meminfo.avail_c; i++)
+		mem_region_show(seq, memory[meminfo.avail[i].idx],
+				meminfo.avail[i].base,
+				meminfo.avail[i].limit - 1);
+
+	seq_putc(seq, '\n');
+	for (i = 0; i < meminfo.mem_c; i++) {
+		if (meminfo.mem[i].idx >= ARRAY_SIZE(cudbg_region))
+			continue;                        /* skip holes */
+		if (!meminfo.mem[i].limit)
+			meminfo.mem[i].limit =
+				i < meminfo.mem_c - 1 ?
+				meminfo.mem[i + 1].base - 1 : ~0;
+		mem_region_show(seq, cudbg_region[meminfo.mem[i].idx],
+				meminfo.mem[i].base, meminfo.mem[i].limit);
+	}
+
+	seq_putc(seq, '\n');
+	mem_region_show(seq, "uP RAM:", meminfo.up_ram_lo, meminfo.up_ram_hi);
+	mem_region_show(seq, "uP Extmem2:", meminfo.up_extmem2_lo,
+			meminfo.up_extmem2_hi);
+
+	seq_printf(seq, "\n%u Rx pages (%u free) of size %uKiB for %u channels\n",
+		   meminfo.rx_pages_data[0], meminfo.free_rx_cnt,
+		   meminfo.rx_pages_data[1], meminfo.rx_pages_data[2]);
+
+	seq_printf(seq, "%u Tx pages (%u free) of size %u%ciB for %u channels\n",
+		   meminfo.tx_pages_data[0], meminfo.free_tx_cnt,
+		   meminfo.tx_pages_data[1], meminfo.tx_pages_data[2],
+		   meminfo.tx_pages_data[3]);
+
+	seq_printf(seq, "%u p-structs (%u free)\n\n",
+		   meminfo.p_structs, meminfo.p_structs_free_cnt);
+
+	for (i = 0; i < 4; i++)
+		/* For T6 these are MAC buffer groups */
+		seq_printf(seq, "Port %d using %u pages out of %u allocated\n",
+			   i, meminfo.port_used[i], meminfo.port_alloc[i]);
+
+	for (i = 0; i < adap->params.arch.nchan; i++)
+		/* For T6 these are MAC buffer groups */
+		seq_printf(seq,
+			   "Loopback %d using %u pages out of %u allocated\n",
+			   i, meminfo.loopback_used[i],
+			   meminfo.loopback_alloc[i]);
+
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(meminfo);
+
+static int chcr_stats_show(struct seq_file *seq, void *v)
+{
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+	struct ch_ktls_port_stats_debug *ktls_port;
+	int i = 0;
+#endif
+	struct adapter *adap = seq->private;
+
+	seq_puts(seq, "Chelsio Crypto Accelerator Stats \n");
+	seq_printf(seq, "Cipher Ops: %10u \n",
+		   atomic_read(&adap->chcr_stats.cipher_rqst));
+	seq_printf(seq, "Digest Ops: %10u \n",
+		   atomic_read(&adap->chcr_stats.digest_rqst));
+	seq_printf(seq, "Aead Ops: %10u \n",
+		   atomic_read(&adap->chcr_stats.aead_rqst));
+	seq_printf(seq, "Completion: %10u \n",
+		   atomic_read(&adap->chcr_stats.complete));
+	seq_printf(seq, "Error: %10u \n",
+		   atomic_read(&adap->chcr_stats.error));
+	seq_printf(seq, "Fallback: %10u \n",
+		   atomic_read(&adap->chcr_stats.fallback));
+	seq_printf(seq, "TLS PDU Tx: %10u\n",
+		   atomic_read(&adap->chcr_stats.tls_pdu_tx));
+	seq_printf(seq, "TLS PDU Rx: %10u\n",
+		   atomic_read(&adap->chcr_stats.tls_pdu_rx));
+	seq_printf(seq, "TLS Keys (DDR) Count: %10u\n",
+		   atomic_read(&adap->chcr_stats.tls_key));
+#if IS_ENABLED(CONFIG_CHELSIO_IPSEC_INLINE)
+	seq_puts(seq, "\nChelsio Inline IPsec Crypto Accelerator Stats\n");
+	seq_printf(seq, "IPSec PDU: %10u\n",
+		   atomic_read(&adap->ch_ipsec_stats.ipsec_cnt));
+#endif
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+	seq_puts(seq, "\nChelsio KTLS Crypto Accelerator Stats\n");
+	seq_printf(seq, "Tx TLS offload refcount:          %20u\n",
+		   refcount_read(&adap->chcr_ktls.ktls_refcount));
+	seq_printf(seq, "Tx records send:                  %20llu\n",
+		   atomic64_read(&adap->ch_ktls_stats.ktls_tx_send_records));
+	seq_printf(seq, "Tx partial start of records:      %20llu\n",
+		   atomic64_read(&adap->ch_ktls_stats.ktls_tx_start_pkts));
+	seq_printf(seq, "Tx partial middle of records:     %20llu\n",
+		   atomic64_read(&adap->ch_ktls_stats.ktls_tx_middle_pkts));
+	seq_printf(seq, "Tx partial end of record:         %20llu\n",
+		   atomic64_read(&adap->ch_ktls_stats.ktls_tx_end_pkts));
+	seq_printf(seq, "Tx complete records:              %20llu\n",
+		   atomic64_read(&adap->ch_ktls_stats.ktls_tx_complete_pkts));
+	seq_printf(seq, "TX trim pkts :                    %20llu\n",
+		   atomic64_read(&adap->ch_ktls_stats.ktls_tx_trimmed_pkts));
+	seq_printf(seq, "TX sw fallback :                  %20llu\n",
+		   atomic64_read(&adap->ch_ktls_stats.ktls_tx_fallback));
+	while (i < MAX_NPORTS) {
+		ktls_port = &adap->ch_ktls_stats.ktls_port[i];
+		seq_printf(seq, "Port %d\n", i);
+		seq_printf(seq, "Tx connection created:            %20llu\n",
+			   atomic64_read(&ktls_port->ktls_tx_connection_open));
+		seq_printf(seq, "Tx connection failed:             %20llu\n",
+			   atomic64_read(&ktls_port->ktls_tx_connection_fail));
+		seq_printf(seq, "Tx connection closed:             %20llu\n",
+			   atomic64_read(&ktls_port->ktls_tx_connection_close));
+		i++;
+	}
+#endif
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(chcr_stats);
+
+#define PRINT_ADAP_STATS(string, value) \
+	seq_printf(seq, "%-25s %-20llu\n", (string), \
+		   (unsigned long long)(value))
+
+#define PRINT_CH_STATS(string, value) \
+do { \
+	seq_printf(seq, "%-25s ", (string)); \
+	for (i = 0; i < adap->params.arch.nchan; i++) \
+		seq_printf(seq, "%-20llu ", \
+			   (unsigned long long)stats.value[i]); \
+	seq_printf(seq, "\n"); \
+} while (0)
+
+#define PRINT_CH_STATS2(string, value) \
+do { \
+	seq_printf(seq, "%-25s ", (string)); \
+	for (i = 0; i < adap->params.arch.nchan; i++) \
+		seq_printf(seq, "%-20llu ", \
+			   (unsigned long long)stats[i].value); \
+	seq_printf(seq, "\n"); \
+} while (0)
+
+static void show_tcp_stats(struct seq_file *seq)
+{
+	struct adapter *adap = seq->private;
+	struct tp_tcp_stats v4, v6;
+
+	spin_lock(&adap->stats_lock);
+	t4_tp_get_tcp_stats(adap, &v4, &v6, false);
+	spin_unlock(&adap->stats_lock);
+
+	PRINT_ADAP_STATS("tcp_ipv4_out_rsts:", v4.tcp_out_rsts);
+	PRINT_ADAP_STATS("tcp_ipv4_in_segs:", v4.tcp_in_segs);
+	PRINT_ADAP_STATS("tcp_ipv4_out_segs:", v4.tcp_out_segs);
+	PRINT_ADAP_STATS("tcp_ipv4_retrans_segs:", v4.tcp_retrans_segs);
+	PRINT_ADAP_STATS("tcp_ipv6_out_rsts:", v6.tcp_out_rsts);
+	PRINT_ADAP_STATS("tcp_ipv6_in_segs:", v6.tcp_in_segs);
+	PRINT_ADAP_STATS("tcp_ipv6_out_segs:", v6.tcp_out_segs);
+	PRINT_ADAP_STATS("tcp_ipv6_retrans_segs:", v6.tcp_retrans_segs);
+}
+
+static void show_ddp_stats(struct seq_file *seq)
+{
+	struct adapter *adap = seq->private;
+	struct tp_usm_stats stats;
+
+	spin_lock(&adap->stats_lock);
+	t4_get_usm_stats(adap, &stats, false);
+	spin_unlock(&adap->stats_lock);
+
+	PRINT_ADAP_STATS("usm_ddp_frames:", stats.frames);
+	PRINT_ADAP_STATS("usm_ddp_octets:", stats.octets);
+	PRINT_ADAP_STATS("usm_ddp_drops:", stats.drops);
+}
+
+static void show_rdma_stats(struct seq_file *seq)
+{
+	struct adapter *adap = seq->private;
+	struct tp_rdma_stats stats;
+
+	spin_lock(&adap->stats_lock);
+	t4_tp_get_rdma_stats(adap, &stats, false);
+	spin_unlock(&adap->stats_lock);
+
+	PRINT_ADAP_STATS("rdma_no_rqe_mod_defer:", stats.rqe_dfr_mod);
+	PRINT_ADAP_STATS("rdma_no_rqe_pkt_defer:", stats.rqe_dfr_pkt);
+}
+
+static void show_tp_err_adapter_stats(struct seq_file *seq)
+{
+	struct adapter *adap = seq->private;
+	struct tp_err_stats stats;
+
+	spin_lock(&adap->stats_lock);
+	t4_tp_get_err_stats(adap, &stats, false);
+	spin_unlock(&adap->stats_lock);
+
+	PRINT_ADAP_STATS("tp_err_ofld_no_neigh:", stats.ofld_no_neigh);
+	PRINT_ADAP_STATS("tp_err_ofld_cong_defer:", stats.ofld_cong_defer);
+}
+
+static void show_cpl_stats(struct seq_file *seq)
+{
+	struct adapter *adap = seq->private;
+	struct tp_cpl_stats stats;
+	u8 i;
+
+	spin_lock(&adap->stats_lock);
+	t4_tp_get_cpl_stats(adap, &stats, false);
+	spin_unlock(&adap->stats_lock);
+
+	PRINT_CH_STATS("tp_cpl_requests:", req);
+	PRINT_CH_STATS("tp_cpl_responses:", rsp);
+}
+
+static void show_tp_err_channel_stats(struct seq_file *seq)
+{
+	struct adapter *adap = seq->private;
+	struct tp_err_stats stats;
+	u8 i;
+
+	spin_lock(&adap->stats_lock);
+	t4_tp_get_err_stats(adap, &stats, false);
+	spin_unlock(&adap->stats_lock);
+
+	PRINT_CH_STATS("tp_mac_in_errs:", mac_in_errs);
+	PRINT_CH_STATS("tp_hdr_in_errs:", hdr_in_errs);
+	PRINT_CH_STATS("tp_tcp_in_errs:", tcp_in_errs);
+	PRINT_CH_STATS("tp_tcp6_in_errs:", tcp6_in_errs);
+	PRINT_CH_STATS("tp_tnl_cong_drops:", tnl_cong_drops);
+	PRINT_CH_STATS("tp_tnl_tx_drops:", tnl_tx_drops);
+	PRINT_CH_STATS("tp_ofld_vlan_drops:", ofld_vlan_drops);
+	PRINT_CH_STATS("tp_ofld_chan_drops:", ofld_chan_drops);
+}
+
+static void show_fcoe_stats(struct seq_file *seq)
+{
+	struct adapter *adap = seq->private;
+	struct tp_fcoe_stats stats[NCHAN];
+	u8 i;
+
+	spin_lock(&adap->stats_lock);
+	for (i = 0; i < adap->params.arch.nchan; i++)
+		t4_get_fcoe_stats(adap, i, &stats[i], false);
+	spin_unlock(&adap->stats_lock);
+
+	PRINT_CH_STATS2("fcoe_octets_ddp", octets_ddp);
+	PRINT_CH_STATS2("fcoe_frames_ddp", frames_ddp);
+	PRINT_CH_STATS2("fcoe_frames_drop", frames_drop);
+}
+
+#undef PRINT_CH_STATS2
+#undef PRINT_CH_STATS
+#undef PRINT_ADAP_STATS
+
+static int tp_stats_show(struct seq_file *seq, void *v)
+{
+	struct adapter *adap = seq->private;
+
+	seq_puts(seq, "\n--------Adapter Stats--------\n");
+	show_tcp_stats(seq);
+	show_ddp_stats(seq);
+	show_rdma_stats(seq);
+	show_tp_err_adapter_stats(seq);
+
+	seq_puts(seq, "\n-------- Channel Stats --------\n");
+	if (adap->params.arch.nchan == NCHAN)
+		seq_printf(seq, "%-25s %-20s %-20s %-20s %-20s\n",
+			   " ", "channel 0", "channel 1",
+			   "channel 2", "channel 3");
+	else
+		seq_printf(seq, "%-25s %-20s %-20s\n",
+			   " ", "channel 0", "channel 1");
+	show_cpl_stats(seq);
+	show_tp_err_channel_stats(seq);
+	show_fcoe_stats(seq);
+
+	return 0;
+}
+DEFINE_SHOW_ATTRIBUTE(tp_stats);
+
+/* Add an array of Debug FS files.
+ */
+void add_debugfs_files(struct adapter *adap,
+		       struct t4_debugfs_entry *files,
+		       unsigned int nfiles)
+{
+	int i;
+
+	/* debugfs support is best effort */
+	for (i = 0; i < nfiles; i++)
+		debugfs_create_file(files[i].name, files[i].mode,
+				    adap->debugfs_root,
+				    (void *)adap + files[i].data,
+				    files[i].ops);
+}
+
+int t4_setup_debugfs(struct adapter *adap)
+{
+	int i;
+	u32 size = 0;
+
+	static struct t4_debugfs_entry t4_debugfs_files[] = {
+		{ "cim_la", &cim_la_fops, 0400, 0 },
+		{ "cim_pif_la", &cim_pif_la_fops, 0400, 0 },
+		{ "cim_ma_la", &cim_ma_la_fops, 0400, 0 },
+		{ "cim_qcfg", &cim_qcfg_fops, 0400, 0 },
+		{ "clk", &clk_fops, 0400, 0 },
+		{ "devlog", &devlog_fops, 0400, 0 },
+		{ "mboxlog", &mboxlog_fops, 0400, 0 },
+		{ "mbox0", &mbox_debugfs_fops, 0600, 0 },
+		{ "mbox1", &mbox_debugfs_fops, 0600, 1 },
+		{ "mbox2", &mbox_debugfs_fops, 0600, 2 },
+		{ "mbox3", &mbox_debugfs_fops, 0600, 3 },
+		{ "mbox4", &mbox_debugfs_fops, 0600, 4 },
+		{ "mbox5", &mbox_debugfs_fops, 0600, 5 },
+		{ "mbox6", &mbox_debugfs_fops, 0600, 6 },
+		{ "mbox7", &mbox_debugfs_fops, 0600, 7 },
+		{ "trace0", &mps_trc_debugfs_fops, 0600, 0 },
+		{ "trace1", &mps_trc_debugfs_fops, 0600, 1 },
+		{ "trace2", &mps_trc_debugfs_fops, 0600, 2 },
+		{ "trace3", &mps_trc_debugfs_fops, 0600, 3 },
+		{ "l2t", &t4_l2t_fops, 0400, 0},
+		{ "mps_tcam", &mps_tcam_debugfs_fops, 0400, 0 },
+		{ "rss", &rss_debugfs_fops, 0400, 0 },
+		{ "rss_config", &rss_config_fops, 0400, 0 },
+		{ "rss_key", &rss_key_debugfs_fops, 0400, 0 },
+		{ "rss_pf_config", &rss_pf_config_debugfs_fops, 0400, 0 },
+		{ "rss_vf_config", &rss_vf_config_debugfs_fops, 0400, 0 },
+		{ "resources", &resources_fops, 0400, 0 },
+#ifdef CONFIG_CHELSIO_T4_DCB
+		{ "dcb_info", &dcb_info_debugfs_fops, 0400, 0 },
+#endif
+		{ "sge_qinfo", &sge_qinfo_debugfs_fops, 0400, 0 },
+		{ "ibq_tp0",  &cim_ibq_fops, 0400, 0 },
+		{ "ibq_tp1",  &cim_ibq_fops, 0400, 1 },
+		{ "ibq_ulp",  &cim_ibq_fops, 0400, 2 },
+		{ "ibq_sge0", &cim_ibq_fops, 0400, 3 },
+		{ "ibq_sge1", &cim_ibq_fops, 0400, 4 },
+		{ "ibq_ncsi", &cim_ibq_fops, 0400, 5 },
+		{ "obq_ulp0", &cim_obq_fops, 0400, 0 },
+		{ "obq_ulp1", &cim_obq_fops, 0400, 1 },
+		{ "obq_ulp2", &cim_obq_fops, 0400, 2 },
+		{ "obq_ulp3", &cim_obq_fops, 0400, 3 },
+		{ "obq_sge",  &cim_obq_fops, 0400, 4 },
+		{ "obq_ncsi", &cim_obq_fops, 0400, 5 },
+		{ "tp_la", &tp_la_fops, 0400, 0 },
+		{ "ulprx_la", &ulprx_la_fops, 0400, 0 },
+		{ "sensors", &sensors_fops, 0400, 0 },
+		{ "pm_stats", &pm_stats_debugfs_fops, 0400, 0 },
+		{ "tx_rate", &tx_rate_fops, 0400, 0 },
+		{ "cctrl", &cctrl_tbl_fops, 0400, 0 },
+#if IS_ENABLED(CONFIG_IPV6)
+		{ "clip_tbl", &clip_tbl_fops, 0400, 0 },
+#endif
+		{ "tids", &tid_info_fops, 0400, 0},
+		{ "blocked_fl", &blocked_fl_fops, 0600, 0 },
+		{ "meminfo", &meminfo_fops, 0400, 0 },
+		{ "crypto", &chcr_stats_fops, 0400, 0 },
+		{ "tp_stats", &tp_stats_fops, 0400, 0 },
+	};
+
+	/* Debug FS nodes common to all T5 and later adapters.
+	 */
+	static struct t4_debugfs_entry t5_debugfs_files[] = {
+		{ "obq_sge_rx_q0", &cim_obq_fops, 0400, 6 },
+		{ "obq_sge_rx_q1", &cim_obq_fops, 0400, 7 },
+	};
+
+	add_debugfs_files(adap,
+			  t4_debugfs_files,
+			  ARRAY_SIZE(t4_debugfs_files));
+	if (!is_t4(adap->params.chip))
+		add_debugfs_files(adap,
+				  t5_debugfs_files,
+				  ARRAY_SIZE(t5_debugfs_files));
+
+	i = t4_read_reg(adap, MA_TARGET_MEM_ENABLE_A);
+	if (i & EDRAM0_ENABLE_F) {
+		size = t4_read_reg(adap, MA_EDRAM0_BAR_A);
+		add_debugfs_mem(adap, "edc0", MEM_EDC0, EDRAM0_SIZE_G(size));
+	}
+	if (i & EDRAM1_ENABLE_F) {
+		size = t4_read_reg(adap, MA_EDRAM1_BAR_A);
+		add_debugfs_mem(adap, "edc1", MEM_EDC1, EDRAM1_SIZE_G(size));
+	}
+	if (is_t5(adap->params.chip)) {
+		if (i & EXT_MEM0_ENABLE_F) {
+			size = t4_read_reg(adap, MA_EXT_MEMORY0_BAR_A);
+			add_debugfs_mem(adap, "mc0", MEM_MC0,
+					EXT_MEM0_SIZE_G(size));
+		}
+		if (i & EXT_MEM1_ENABLE_F) {
+			size = t4_read_reg(adap, MA_EXT_MEMORY1_BAR_A);
+			add_debugfs_mem(adap, "mc1", MEM_MC1,
+					EXT_MEM1_SIZE_G(size));
+		}
+	} else {
+		if (i & EXT_MEM_ENABLE_F) {
+			size = t4_read_reg(adap, MA_EXT_MEMORY_BAR_A);
+			add_debugfs_mem(adap, "mc", MEM_MC,
+					EXT_MEM_SIZE_G(size));
+		}
+
+		if (i & HMA_MUX_F) {
+			size = t4_read_reg(adap, MA_EXT_MEMORY1_BAR_A);
+			add_debugfs_mem(adap, "hma", MEM_HMA,
+					EXT_MEM1_SIZE_G(size));
+		}
+	}
+
+	debugfs_create_file_size("flash", 0400, adap->debugfs_root, adap,
+				 &flash_debugfs_fops, adap->params.sf_size);
+	debugfs_create_bool("use_backdoor", 0600,
+			    adap->debugfs_root, &adap->use_bd);
+	debugfs_create_bool("trace_rss", 0600,
+			    adap->debugfs_root, &adap->trace_rss);
+
+	return 0;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h
new file mode 100644
index 000000000..1471cf0de
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h
@@ -0,0 +1,70 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_DEBUGFS_H
+#define __CXGB4_DEBUGFS_H
+
+#include <linux/export.h>
+
+struct t4_debugfs_entry {
+	const char *name;
+	const struct file_operations *ops;
+	umode_t mode;
+	unsigned char data;
+};
+
+struct seq_tab {
+	int (*show)(struct seq_file *seq, void *v, int idx);
+	unsigned int rows;        /* # of entries */
+	unsigned char width;      /* size in bytes of each entry */
+	unsigned char skip_first; /* whether the first line is a header */
+	char data[];             /* the table data */
+};
+
+static inline unsigned int hex2val(char c)
+{
+	return isdigit(c) ? c - '0' : tolower(c) - 'a' + 10;
+}
+
+struct seq_tab *seq_open_tab(struct file *f, unsigned int rows,
+			     unsigned int width, unsigned int have_header,
+			     int (*show)(struct seq_file *seq, void *v, int i));
+
+int t4_setup_debugfs(struct adapter *adap);
+void add_debugfs_files(struct adapter *adap,
+		       struct t4_debugfs_entry *files,
+		       unsigned int nfiles);
+int mem_open(struct inode *inode, struct file *file);
+
+#endif
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c
new file mode 100644
index 000000000..8477a93ce
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c
@@ -0,0 +1,2299 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  Copyright (C) 2013-2015 Chelsio Communications.  All rights reserved.
+ */
+
+#include <linux/firmware.h>
+#include <linux/mdio.h>
+
+#include "cxgb4.h"
+#include "t4_regs.h"
+#include "t4fw_api.h"
+#include "cxgb4_cudbg.h"
+#include "cxgb4_filter.h"
+#include "cxgb4_tc_flower.h"
+
+#define EEPROM_MAGIC 0x38E2F10C
+
+static u32 get_msglevel(struct net_device *dev)
+{
+	return netdev2adap(dev)->msg_enable;
+}
+
+static void set_msglevel(struct net_device *dev, u32 val)
+{
+	netdev2adap(dev)->msg_enable = val;
+}
+
+enum cxgb4_ethtool_tests {
+	CXGB4_ETHTOOL_LB_TEST,
+	CXGB4_ETHTOOL_MAX_TEST,
+};
+
+static const char cxgb4_selftest_strings[CXGB4_ETHTOOL_MAX_TEST][ETH_GSTRING_LEN] = {
+	"Loop back test (offline)",
+};
+
+static const char * const flash_region_strings[] = {
+	"All",
+	"Firmware",
+	"PHY Firmware",
+	"Boot",
+	"Boot CFG",
+};
+
+static const char stats_strings[][ETH_GSTRING_LEN] = {
+	"tx_octets_ok           ",
+	"tx_frames_ok           ",
+	"tx_broadcast_frames    ",
+	"tx_multicast_frames    ",
+	"tx_unicast_frames      ",
+	"tx_error_frames        ",
+
+	"tx_frames_64           ",
+	"tx_frames_65_to_127    ",
+	"tx_frames_128_to_255   ",
+	"tx_frames_256_to_511   ",
+	"tx_frames_512_to_1023  ",
+	"tx_frames_1024_to_1518 ",
+	"tx_frames_1519_to_max  ",
+
+	"tx_frames_dropped      ",
+	"tx_pause_frames        ",
+	"tx_ppp0_frames         ",
+	"tx_ppp1_frames         ",
+	"tx_ppp2_frames         ",
+	"tx_ppp3_frames         ",
+	"tx_ppp4_frames         ",
+	"tx_ppp5_frames         ",
+	"tx_ppp6_frames         ",
+	"tx_ppp7_frames         ",
+
+	"rx_octets_ok           ",
+	"rx_frames_ok           ",
+	"rx_broadcast_frames    ",
+	"rx_multicast_frames    ",
+	"rx_unicast_frames      ",
+
+	"rx_frames_too_long     ",
+	"rx_jabber_errors       ",
+	"rx_fcs_errors          ",
+	"rx_length_errors       ",
+	"rx_symbol_errors       ",
+	"rx_runt_frames         ",
+
+	"rx_frames_64           ",
+	"rx_frames_65_to_127    ",
+	"rx_frames_128_to_255   ",
+	"rx_frames_256_to_511   ",
+	"rx_frames_512_to_1023  ",
+	"rx_frames_1024_to_1518 ",
+	"rx_frames_1519_to_max  ",
+
+	"rx_pause_frames        ",
+	"rx_ppp0_frames         ",
+	"rx_ppp1_frames         ",
+	"rx_ppp2_frames         ",
+	"rx_ppp3_frames         ",
+	"rx_ppp4_frames         ",
+	"rx_ppp5_frames         ",
+	"rx_ppp6_frames         ",
+	"rx_ppp7_frames         ",
+
+	"rx_bg0_frames_dropped  ",
+	"rx_bg1_frames_dropped  ",
+	"rx_bg2_frames_dropped  ",
+	"rx_bg3_frames_dropped  ",
+	"rx_bg0_frames_trunc    ",
+	"rx_bg1_frames_trunc    ",
+	"rx_bg2_frames_trunc    ",
+	"rx_bg3_frames_trunc    ",
+
+	"tso                    ",
+	"uso                    ",
+	"tx_csum_offload        ",
+	"rx_csum_good           ",
+	"vlan_extractions       ",
+	"vlan_insertions        ",
+	"gro_packets            ",
+	"gro_merged             ",
+#if  IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+	"tx_tls_encrypted_packets",
+	"tx_tls_encrypted_bytes  ",
+	"tx_tls_ctx              ",
+	"tx_tls_ooo              ",
+	"tx_tls_skip_no_sync_data",
+	"tx_tls_drop_no_sync_data",
+	"tx_tls_drop_bypass_req  ",
+#endif
+};
+
+static char adapter_stats_strings[][ETH_GSTRING_LEN] = {
+	"db_drop                ",
+	"db_full                ",
+	"db_empty               ",
+	"write_coal_success     ",
+	"write_coal_fail        ",
+};
+
+static char loopback_stats_strings[][ETH_GSTRING_LEN] = {
+	"-------Loopback----------- ",
+	"octets_ok              ",
+	"frames_ok              ",
+	"bcast_frames           ",
+	"mcast_frames           ",
+	"ucast_frames           ",
+	"error_frames           ",
+	"frames_64              ",
+	"frames_65_to_127       ",
+	"frames_128_to_255      ",
+	"frames_256_to_511      ",
+	"frames_512_to_1023     ",
+	"frames_1024_to_1518    ",
+	"frames_1519_to_max     ",
+	"frames_dropped         ",
+	"bg0_frames_dropped     ",
+	"bg1_frames_dropped     ",
+	"bg2_frames_dropped     ",
+	"bg3_frames_dropped     ",
+	"bg0_frames_trunc       ",
+	"bg1_frames_trunc       ",
+	"bg2_frames_trunc       ",
+	"bg3_frames_trunc       ",
+};
+
+static const char cxgb4_priv_flags_strings[][ETH_GSTRING_LEN] = {
+	[PRIV_FLAG_PORT_TX_VM_BIT] = "port_tx_vm_wr",
+};
+
+static int get_sset_count(struct net_device *dev, int sset)
+{
+	switch (sset) {
+	case ETH_SS_STATS:
+		return ARRAY_SIZE(stats_strings) +
+		       ARRAY_SIZE(adapter_stats_strings) +
+		       ARRAY_SIZE(loopback_stats_strings);
+	case ETH_SS_PRIV_FLAGS:
+		return ARRAY_SIZE(cxgb4_priv_flags_strings);
+	case ETH_SS_TEST:
+		return ARRAY_SIZE(cxgb4_selftest_strings);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static int get_regs_len(struct net_device *dev)
+{
+	struct adapter *adap = netdev2adap(dev);
+
+	return t4_get_regs_len(adap);
+}
+
+static int get_eeprom_len(struct net_device *dev)
+{
+	return EEPROMSIZE;
+}
+
+static void get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info)
+{
+	struct adapter *adapter = netdev2adap(dev);
+	u32 exprom_vers;
+
+	strscpy(info->driver, cxgb4_driver_name, sizeof(info->driver));
+	strscpy(info->bus_info, pci_name(adapter->pdev),
+		sizeof(info->bus_info));
+	info->regdump_len = get_regs_len(dev);
+
+	if (adapter->params.fw_vers)
+		snprintf(info->fw_version, sizeof(info->fw_version),
+			 "%u.%u.%u.%u, TP %u.%u.%u.%u",
+			 FW_HDR_FW_VER_MAJOR_G(adapter->params.fw_vers),
+			 FW_HDR_FW_VER_MINOR_G(adapter->params.fw_vers),
+			 FW_HDR_FW_VER_MICRO_G(adapter->params.fw_vers),
+			 FW_HDR_FW_VER_BUILD_G(adapter->params.fw_vers),
+			 FW_HDR_FW_VER_MAJOR_G(adapter->params.tp_vers),
+			 FW_HDR_FW_VER_MINOR_G(adapter->params.tp_vers),
+			 FW_HDR_FW_VER_MICRO_G(adapter->params.tp_vers),
+			 FW_HDR_FW_VER_BUILD_G(adapter->params.tp_vers));
+
+	if (!t4_get_exprom_version(adapter, &exprom_vers))
+		snprintf(info->erom_version, sizeof(info->erom_version),
+			 "%u.%u.%u.%u",
+			 FW_HDR_FW_VER_MAJOR_G(exprom_vers),
+			 FW_HDR_FW_VER_MINOR_G(exprom_vers),
+			 FW_HDR_FW_VER_MICRO_G(exprom_vers),
+			 FW_HDR_FW_VER_BUILD_G(exprom_vers));
+	info->n_priv_flags = ARRAY_SIZE(cxgb4_priv_flags_strings);
+}
+
+static void get_strings(struct net_device *dev, u32 stringset, u8 *data)
+{
+	if (stringset == ETH_SS_STATS) {
+		memcpy(data, stats_strings, sizeof(stats_strings));
+		data += sizeof(stats_strings);
+		memcpy(data, adapter_stats_strings,
+		       sizeof(adapter_stats_strings));
+		data += sizeof(adapter_stats_strings);
+		memcpy(data, loopback_stats_strings,
+		       sizeof(loopback_stats_strings));
+	} else if (stringset == ETH_SS_PRIV_FLAGS) {
+		memcpy(data, cxgb4_priv_flags_strings,
+		       sizeof(cxgb4_priv_flags_strings));
+	} else if (stringset == ETH_SS_TEST) {
+		memcpy(data, cxgb4_selftest_strings,
+		       sizeof(cxgb4_selftest_strings));
+	}
+}
+
+/* port stats maintained per queue of the port. They should be in the same
+ * order as in stats_strings above.
+ */
+struct queue_port_stats {
+	u64 tso;
+	u64 uso;
+	u64 tx_csum;
+	u64 rx_csum;
+	u64 vlan_ex;
+	u64 vlan_ins;
+	u64 gro_pkts;
+	u64 gro_merged;
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+	u64 tx_tls_encrypted_packets;
+	u64 tx_tls_encrypted_bytes;
+	u64 tx_tls_ctx;
+	u64 tx_tls_ooo;
+	u64 tx_tls_skip_no_sync_data;
+	u64 tx_tls_drop_no_sync_data;
+	u64 tx_tls_drop_bypass_req;
+#endif
+};
+
+struct adapter_stats {
+	u64 db_drop;
+	u64 db_full;
+	u64 db_empty;
+	u64 wc_success;
+	u64 wc_fail;
+};
+
+static void collect_sge_port_stats(const struct adapter *adap,
+				   const struct port_info *p,
+				   struct queue_port_stats *s)
+{
+	const struct sge_eth_txq *tx = &adap->sge.ethtxq[p->first_qset];
+	const struct sge_eth_rxq *rx = &adap->sge.ethrxq[p->first_qset];
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+	const struct ch_ktls_port_stats_debug *ktls_stats;
+#endif
+	struct sge_eohw_txq *eohw_tx;
+	unsigned int i;
+
+	memset(s, 0, sizeof(*s));
+	for (i = 0; i < p->nqsets; i++, rx++, tx++) {
+		s->tso += tx->tso;
+		s->uso += tx->uso;
+		s->tx_csum += tx->tx_cso;
+		s->rx_csum += rx->stats.rx_cso;
+		s->vlan_ex += rx->stats.vlan_ex;
+		s->vlan_ins += tx->vlan_ins;
+		s->gro_pkts += rx->stats.lro_pkts;
+		s->gro_merged += rx->stats.lro_merged;
+	}
+
+	if (adap->sge.eohw_txq) {
+		eohw_tx = &adap->sge.eohw_txq[p->first_qset];
+		for (i = 0; i < p->nqsets; i++, eohw_tx++) {
+			s->tso += eohw_tx->tso;
+			s->uso += eohw_tx->uso;
+			s->tx_csum += eohw_tx->tx_cso;
+			s->vlan_ins += eohw_tx->vlan_ins;
+		}
+	}
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+	ktls_stats = &adap->ch_ktls_stats.ktls_port[p->port_id];
+	s->tx_tls_encrypted_packets =
+		atomic64_read(&ktls_stats->ktls_tx_encrypted_packets);
+	s->tx_tls_encrypted_bytes =
+		atomic64_read(&ktls_stats->ktls_tx_encrypted_bytes);
+	s->tx_tls_ctx = atomic64_read(&ktls_stats->ktls_tx_ctx);
+	s->tx_tls_ooo = atomic64_read(&ktls_stats->ktls_tx_ooo);
+	s->tx_tls_skip_no_sync_data =
+		atomic64_read(&ktls_stats->ktls_tx_skip_no_sync_data);
+	s->tx_tls_drop_no_sync_data =
+		atomic64_read(&ktls_stats->ktls_tx_drop_no_sync_data);
+	s->tx_tls_drop_bypass_req =
+		atomic64_read(&ktls_stats->ktls_tx_drop_bypass_req);
+#endif
+}
+
+static void collect_adapter_stats(struct adapter *adap, struct adapter_stats *s)
+{
+	u64 val1, val2;
+
+	memset(s, 0, sizeof(*s));
+
+	s->db_drop = adap->db_stats.db_drop;
+	s->db_full = adap->db_stats.db_full;
+	s->db_empty = adap->db_stats.db_empty;
+
+	if (!is_t4(adap->params.chip)) {
+		int v;
+
+		v = t4_read_reg(adap, SGE_STAT_CFG_A);
+		if (STATSOURCE_T5_G(v) == 7) {
+			val2 = t4_read_reg(adap, SGE_STAT_MATCH_A);
+			val1 = t4_read_reg(adap, SGE_STAT_TOTAL_A);
+			s->wc_success = val1 - val2;
+			s->wc_fail = val2;
+		}
+	}
+}
+
+static void get_stats(struct net_device *dev, struct ethtool_stats *stats,
+		      u64 *data)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adapter = pi->adapter;
+	struct lb_port_stats s;
+	int i;
+	u64 *p0;
+
+	t4_get_port_stats_offset(adapter, pi->tx_chan,
+				 (struct port_stats *)data,
+				 &pi->stats_base);
+
+	data += sizeof(struct port_stats) / sizeof(u64);
+	collect_sge_port_stats(adapter, pi, (struct queue_port_stats *)data);
+	data += sizeof(struct queue_port_stats) / sizeof(u64);
+	collect_adapter_stats(adapter, (struct adapter_stats *)data);
+	data += sizeof(struct adapter_stats) / sizeof(u64);
+
+	*data++ = (u64)pi->port_id;
+	memset(&s, 0, sizeof(s));
+	t4_get_lb_stats(adapter, pi->port_id, &s);
+
+	p0 = &s.octets;
+	for (i = 0; i < ARRAY_SIZE(loopback_stats_strings) - 1; i++)
+		*data++ = (unsigned long long)*p0++;
+}
+
+static void get_regs(struct net_device *dev, struct ethtool_regs *regs,
+		     void *buf)
+{
+	struct adapter *adap = netdev2adap(dev);
+	size_t buf_size;
+
+	buf_size = t4_get_regs_len(adap);
+	regs->version = mk_adap_vers(adap);
+	t4_get_regs(adap, buf, buf_size);
+}
+
+static int restart_autoneg(struct net_device *dev)
+{
+	struct port_info *p = netdev_priv(dev);
+
+	if (!netif_running(dev))
+		return -EAGAIN;
+	if (p->link_cfg.autoneg != AUTONEG_ENABLE)
+		return -EINVAL;
+	t4_restart_aneg(p->adapter, p->adapter->pf, p->tx_chan);
+	return 0;
+}
+
+static int identify_port(struct net_device *dev,
+			 enum ethtool_phys_id_state state)
+{
+	unsigned int val;
+	struct adapter *adap = netdev2adap(dev);
+
+	if (state == ETHTOOL_ID_ACTIVE)
+		val = 0xffff;
+	else if (state == ETHTOOL_ID_INACTIVE)
+		val = 0;
+	else
+		return -EINVAL;
+
+	return t4_identify_port(adap, adap->pf, netdev2pinfo(dev)->viid, val);
+}
+
+/**
+ *	from_fw_port_mod_type - translate Firmware Port/Module type to Ethtool
+ *	@port_type: Firmware Port Type
+ *	@mod_type: Firmware Module Type
+ *
+ *	Translate Firmware Port/Module type to Ethtool Port Type.
+ */
+static int from_fw_port_mod_type(enum fw_port_type port_type,
+				 enum fw_port_module_type mod_type)
+{
+	if (port_type == FW_PORT_TYPE_BT_SGMII ||
+	    port_type == FW_PORT_TYPE_BT_XFI ||
+	    port_type == FW_PORT_TYPE_BT_XAUI) {
+		return PORT_TP;
+	} else if (port_type == FW_PORT_TYPE_FIBER_XFI ||
+		   port_type == FW_PORT_TYPE_FIBER_XAUI) {
+		return PORT_FIBRE;
+	} else if (port_type == FW_PORT_TYPE_SFP ||
+		   port_type == FW_PORT_TYPE_QSFP_10G ||
+		   port_type == FW_PORT_TYPE_QSA ||
+		   port_type == FW_PORT_TYPE_QSFP ||
+		   port_type == FW_PORT_TYPE_CR4_QSFP ||
+		   port_type == FW_PORT_TYPE_CR_QSFP ||
+		   port_type == FW_PORT_TYPE_CR2_QSFP ||
+		   port_type == FW_PORT_TYPE_SFP28) {
+		if (mod_type == FW_PORT_MOD_TYPE_LR ||
+		    mod_type == FW_PORT_MOD_TYPE_SR ||
+		    mod_type == FW_PORT_MOD_TYPE_ER ||
+		    mod_type == FW_PORT_MOD_TYPE_LRM)
+			return PORT_FIBRE;
+		else if (mod_type == FW_PORT_MOD_TYPE_TWINAX_PASSIVE ||
+			 mod_type == FW_PORT_MOD_TYPE_TWINAX_ACTIVE)
+			return PORT_DA;
+		else
+			return PORT_OTHER;
+	} else if (port_type == FW_PORT_TYPE_KR4_100G ||
+		   port_type == FW_PORT_TYPE_KR_SFP28 ||
+		   port_type == FW_PORT_TYPE_KR_XLAUI) {
+		return PORT_NONE;
+	}
+
+	return PORT_OTHER;
+}
+
+/**
+ *	speed_to_fw_caps - translate Port Speed to Firmware Port Capabilities
+ *	@speed: speed in Kb/s
+ *
+ *	Translates a specific Port Speed into a Firmware Port Capabilities
+ *	value.
+ */
+static unsigned int speed_to_fw_caps(int speed)
+{
+	if (speed == 100)
+		return FW_PORT_CAP32_SPEED_100M;
+	if (speed == 1000)
+		return FW_PORT_CAP32_SPEED_1G;
+	if (speed == 10000)
+		return FW_PORT_CAP32_SPEED_10G;
+	if (speed == 25000)
+		return FW_PORT_CAP32_SPEED_25G;
+	if (speed == 40000)
+		return FW_PORT_CAP32_SPEED_40G;
+	if (speed == 50000)
+		return FW_PORT_CAP32_SPEED_50G;
+	if (speed == 100000)
+		return FW_PORT_CAP32_SPEED_100G;
+	if (speed == 200000)
+		return FW_PORT_CAP32_SPEED_200G;
+	if (speed == 400000)
+		return FW_PORT_CAP32_SPEED_400G;
+	return 0;
+}
+
+/**
+ *	fw_caps_to_lmm - translate Firmware to ethtool Link Mode Mask
+ *	@port_type: Firmware Port Type
+ *	@fw_caps: Firmware Port Capabilities
+ *	@link_mode_mask: ethtool Link Mode Mask
+ *
+ *	Translate a Firmware Port Capabilities specification to an ethtool
+ *	Link Mode Mask.
+ */
+static void fw_caps_to_lmm(enum fw_port_type port_type,
+			   fw_port_cap32_t fw_caps,
+			   unsigned long *link_mode_mask)
+{
+	#define SET_LMM(__lmm_name) \
+		do { \
+			__set_bit(ETHTOOL_LINK_MODE_ ## __lmm_name ## _BIT, \
+				  link_mode_mask); \
+		} while (0)
+
+	#define FW_CAPS_TO_LMM(__fw_name, __lmm_name) \
+		do { \
+			if (fw_caps & FW_PORT_CAP32_ ## __fw_name) \
+				SET_LMM(__lmm_name); \
+		} while (0)
+
+	switch (port_type) {
+	case FW_PORT_TYPE_BT_SGMII:
+	case FW_PORT_TYPE_BT_XFI:
+	case FW_PORT_TYPE_BT_XAUI:
+		SET_LMM(TP);
+		FW_CAPS_TO_LMM(SPEED_100M, 100baseT_Full);
+		FW_CAPS_TO_LMM(SPEED_1G, 1000baseT_Full);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseT_Full);
+		break;
+
+	case FW_PORT_TYPE_KX4:
+	case FW_PORT_TYPE_KX:
+		SET_LMM(Backplane);
+		FW_CAPS_TO_LMM(SPEED_1G, 1000baseKX_Full);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseKX4_Full);
+		break;
+
+	case FW_PORT_TYPE_KR:
+		SET_LMM(Backplane);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseKR_Full);
+		break;
+
+	case FW_PORT_TYPE_BP_AP:
+		SET_LMM(Backplane);
+		FW_CAPS_TO_LMM(SPEED_1G, 1000baseKX_Full);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseR_FEC);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseKR_Full);
+		break;
+
+	case FW_PORT_TYPE_BP4_AP:
+		SET_LMM(Backplane);
+		FW_CAPS_TO_LMM(SPEED_1G, 1000baseKX_Full);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseR_FEC);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseKR_Full);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseKX4_Full);
+		break;
+
+	case FW_PORT_TYPE_FIBER_XFI:
+	case FW_PORT_TYPE_FIBER_XAUI:
+	case FW_PORT_TYPE_SFP:
+	case FW_PORT_TYPE_QSFP_10G:
+	case FW_PORT_TYPE_QSA:
+		SET_LMM(FIBRE);
+		FW_CAPS_TO_LMM(SPEED_1G, 1000baseT_Full);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseT_Full);
+		break;
+
+	case FW_PORT_TYPE_BP40_BA:
+	case FW_PORT_TYPE_QSFP:
+		SET_LMM(FIBRE);
+		FW_CAPS_TO_LMM(SPEED_1G, 1000baseT_Full);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseT_Full);
+		FW_CAPS_TO_LMM(SPEED_40G, 40000baseSR4_Full);
+		break;
+
+	case FW_PORT_TYPE_CR_QSFP:
+	case FW_PORT_TYPE_SFP28:
+		SET_LMM(FIBRE);
+		FW_CAPS_TO_LMM(SPEED_1G, 1000baseT_Full);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseT_Full);
+		FW_CAPS_TO_LMM(SPEED_25G, 25000baseCR_Full);
+		break;
+
+	case FW_PORT_TYPE_KR_SFP28:
+		SET_LMM(Backplane);
+		FW_CAPS_TO_LMM(SPEED_1G, 1000baseT_Full);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseKR_Full);
+		FW_CAPS_TO_LMM(SPEED_25G, 25000baseKR_Full);
+		break;
+
+	case FW_PORT_TYPE_KR_XLAUI:
+		SET_LMM(Backplane);
+		FW_CAPS_TO_LMM(SPEED_1G, 1000baseKX_Full);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseKR_Full);
+		FW_CAPS_TO_LMM(SPEED_40G, 40000baseKR4_Full);
+		break;
+
+	case FW_PORT_TYPE_CR2_QSFP:
+		SET_LMM(FIBRE);
+		FW_CAPS_TO_LMM(SPEED_50G, 50000baseSR2_Full);
+		break;
+
+	case FW_PORT_TYPE_KR4_100G:
+	case FW_PORT_TYPE_CR4_QSFP:
+		SET_LMM(FIBRE);
+		FW_CAPS_TO_LMM(SPEED_1G,  1000baseT_Full);
+		FW_CAPS_TO_LMM(SPEED_10G, 10000baseKR_Full);
+		FW_CAPS_TO_LMM(SPEED_40G, 40000baseSR4_Full);
+		FW_CAPS_TO_LMM(SPEED_25G, 25000baseCR_Full);
+		FW_CAPS_TO_LMM(SPEED_50G, 50000baseCR2_Full);
+		FW_CAPS_TO_LMM(SPEED_100G, 100000baseCR4_Full);
+		break;
+
+	default:
+		break;
+	}
+
+	if (fw_caps & FW_PORT_CAP32_FEC_V(FW_PORT_CAP32_FEC_M)) {
+		FW_CAPS_TO_LMM(FEC_RS, FEC_RS);
+		FW_CAPS_TO_LMM(FEC_BASER_RS, FEC_BASER);
+	} else {
+		SET_LMM(FEC_NONE);
+	}
+
+	FW_CAPS_TO_LMM(ANEG, Autoneg);
+	FW_CAPS_TO_LMM(802_3_PAUSE, Pause);
+	FW_CAPS_TO_LMM(802_3_ASM_DIR, Asym_Pause);
+
+	#undef FW_CAPS_TO_LMM
+	#undef SET_LMM
+}
+
+/**
+ *	lmm_to_fw_caps - translate ethtool Link Mode Mask to Firmware
+ *	capabilities
+ *	@link_mode_mask: ethtool Link Mode Mask
+ *
+ *	Translate ethtool Link Mode Mask into a Firmware Port capabilities
+ *	value.
+ */
+static unsigned int lmm_to_fw_caps(const unsigned long *link_mode_mask)
+{
+	unsigned int fw_caps = 0;
+
+	#define LMM_TO_FW_CAPS(__lmm_name, __fw_name) \
+		do { \
+			if (test_bit(ETHTOOL_LINK_MODE_ ## __lmm_name ## _BIT, \
+				     link_mode_mask)) \
+				fw_caps |= FW_PORT_CAP32_ ## __fw_name; \
+		} while (0)
+
+	LMM_TO_FW_CAPS(100baseT_Full, SPEED_100M);
+	LMM_TO_FW_CAPS(1000baseT_Full, SPEED_1G);
+	LMM_TO_FW_CAPS(10000baseT_Full, SPEED_10G);
+	LMM_TO_FW_CAPS(40000baseSR4_Full, SPEED_40G);
+	LMM_TO_FW_CAPS(25000baseCR_Full, SPEED_25G);
+	LMM_TO_FW_CAPS(50000baseCR2_Full, SPEED_50G);
+	LMM_TO_FW_CAPS(100000baseCR4_Full, SPEED_100G);
+
+	#undef LMM_TO_FW_CAPS
+
+	return fw_caps;
+}
+
+static int get_link_ksettings(struct net_device *dev,
+			      struct ethtool_link_ksettings *link_ksettings)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct ethtool_link_settings *base = &link_ksettings->base;
+
+	/* For the nonce, the Firmware doesn't send up Port State changes
+	 * when the Virtual Interface attached to the Port is down.  So
+	 * if it's down, let's grab any changes.
+	 */
+	if (!netif_running(dev))
+		(void)t4_update_port_info(pi);
+
+	ethtool_link_ksettings_zero_link_mode(link_ksettings, supported);
+	ethtool_link_ksettings_zero_link_mode(link_ksettings, advertising);
+	ethtool_link_ksettings_zero_link_mode(link_ksettings, lp_advertising);
+
+	base->port = from_fw_port_mod_type(pi->port_type, pi->mod_type);
+
+	if (pi->mdio_addr >= 0) {
+		base->phy_address = pi->mdio_addr;
+		base->mdio_support = (pi->port_type == FW_PORT_TYPE_BT_SGMII
+				      ? ETH_MDIO_SUPPORTS_C22
+				      : ETH_MDIO_SUPPORTS_C45);
+	} else {
+		base->phy_address = 255;
+		base->mdio_support = 0;
+	}
+
+	fw_caps_to_lmm(pi->port_type, pi->link_cfg.pcaps,
+		       link_ksettings->link_modes.supported);
+	fw_caps_to_lmm(pi->port_type,
+		       t4_link_acaps(pi->adapter,
+				     pi->lport,
+				     &pi->link_cfg),
+		       link_ksettings->link_modes.advertising);
+	fw_caps_to_lmm(pi->port_type, pi->link_cfg.lpacaps,
+		       link_ksettings->link_modes.lp_advertising);
+
+	base->speed = (netif_carrier_ok(dev)
+		       ? pi->link_cfg.speed
+		       : SPEED_UNKNOWN);
+	base->duplex = DUPLEX_FULL;
+
+	base->autoneg = pi->link_cfg.autoneg;
+	if (pi->link_cfg.pcaps & FW_PORT_CAP32_ANEG)
+		ethtool_link_ksettings_add_link_mode(link_ksettings,
+						     supported, Autoneg);
+	if (pi->link_cfg.autoneg)
+		ethtool_link_ksettings_add_link_mode(link_ksettings,
+						     advertising, Autoneg);
+
+	return 0;
+}
+
+static int set_link_ksettings(struct net_device *dev,
+			    const struct ethtool_link_ksettings *link_ksettings)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct link_config *lc = &pi->link_cfg;
+	const struct ethtool_link_settings *base = &link_ksettings->base;
+	struct link_config old_lc;
+	unsigned int fw_caps;
+	int ret = 0;
+
+	/* only full-duplex supported */
+	if (base->duplex != DUPLEX_FULL)
+		return -EINVAL;
+
+	old_lc = *lc;
+	if (!(lc->pcaps & FW_PORT_CAP32_ANEG) ||
+	    base->autoneg == AUTONEG_DISABLE) {
+		fw_caps = speed_to_fw_caps(base->speed);
+
+		/* Speed must be supported by Physical Port Capabilities. */
+		if (!(lc->pcaps & fw_caps))
+			return -EINVAL;
+
+		lc->speed_caps = fw_caps;
+		lc->acaps = fw_caps;
+	} else {
+		fw_caps =
+			lmm_to_fw_caps(link_ksettings->link_modes.advertising);
+		if (!(lc->pcaps & fw_caps))
+			return -EINVAL;
+		lc->speed_caps = 0;
+		lc->acaps = fw_caps | FW_PORT_CAP32_ANEG;
+	}
+	lc->autoneg = base->autoneg;
+
+	/* If the firmware rejects the Link Configuration request, back out
+	 * the changes and report the error.
+	 */
+	ret = t4_link_l1cfg(pi->adapter, pi->adapter->mbox, pi->tx_chan, lc);
+	if (ret)
+		*lc = old_lc;
+
+	return ret;
+}
+
+/* Translate the Firmware FEC value into the ethtool value. */
+static inline unsigned int fwcap_to_eth_fec(unsigned int fw_fec)
+{
+	unsigned int eth_fec = 0;
+
+	if (fw_fec & FW_PORT_CAP32_FEC_RS)
+		eth_fec |= ETHTOOL_FEC_RS;
+	if (fw_fec & FW_PORT_CAP32_FEC_BASER_RS)
+		eth_fec |= ETHTOOL_FEC_BASER;
+
+	/* if nothing is set, then FEC is off */
+	if (!eth_fec)
+		eth_fec = ETHTOOL_FEC_OFF;
+
+	return eth_fec;
+}
+
+/* Translate Common Code FEC value into ethtool value. */
+static inline unsigned int cc_to_eth_fec(unsigned int cc_fec)
+{
+	unsigned int eth_fec = 0;
+
+	if (cc_fec & FEC_AUTO)
+		eth_fec |= ETHTOOL_FEC_AUTO;
+	if (cc_fec & FEC_RS)
+		eth_fec |= ETHTOOL_FEC_RS;
+	if (cc_fec & FEC_BASER_RS)
+		eth_fec |= ETHTOOL_FEC_BASER;
+
+	/* if nothing is set, then FEC is off */
+	if (!eth_fec)
+		eth_fec = ETHTOOL_FEC_OFF;
+
+	return eth_fec;
+}
+
+/* Translate ethtool FEC value into Common Code value. */
+static inline unsigned int eth_to_cc_fec(unsigned int eth_fec)
+{
+	unsigned int cc_fec = 0;
+
+	if (eth_fec & ETHTOOL_FEC_OFF)
+		return cc_fec;
+
+	if (eth_fec & ETHTOOL_FEC_AUTO)
+		cc_fec |= FEC_AUTO;
+	if (eth_fec & ETHTOOL_FEC_RS)
+		cc_fec |= FEC_RS;
+	if (eth_fec & ETHTOOL_FEC_BASER)
+		cc_fec |= FEC_BASER_RS;
+
+	return cc_fec;
+}
+
+static int get_fecparam(struct net_device *dev, struct ethtool_fecparam *fec)
+{
+	const struct port_info *pi = netdev_priv(dev);
+	const struct link_config *lc = &pi->link_cfg;
+
+	/* Translate the Firmware FEC Support into the ethtool value.  We
+	 * always support IEEE 802.3 "automatic" selection of Link FEC type if
+	 * any FEC is supported.
+	 */
+	fec->fec = fwcap_to_eth_fec(lc->pcaps);
+	if (fec->fec != ETHTOOL_FEC_OFF)
+		fec->fec |= ETHTOOL_FEC_AUTO;
+
+	/* Translate the current internal FEC parameters into the
+	 * ethtool values.
+	 */
+	fec->active_fec = cc_to_eth_fec(lc->fec);
+
+	return 0;
+}
+
+static int set_fecparam(struct net_device *dev, struct ethtool_fecparam *fec)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct link_config *lc = &pi->link_cfg;
+	struct link_config old_lc;
+	int ret;
+
+	/* Save old Link Configuration in case the L1 Configure below
+	 * fails.
+	 */
+	old_lc = *lc;
+
+	/* Try to perform the L1 Configure and return the result of that
+	 * effort.  If it fails, revert the attempted change.
+	 */
+	lc->requested_fec = eth_to_cc_fec(fec->fec);
+	ret = t4_link_l1cfg(pi->adapter, pi->adapter->mbox,
+			    pi->tx_chan, lc);
+	if (ret)
+		*lc = old_lc;
+	return ret;
+}
+
+static void get_pauseparam(struct net_device *dev,
+			   struct ethtool_pauseparam *epause)
+{
+	struct port_info *p = netdev_priv(dev);
+
+	epause->autoneg = (p->link_cfg.requested_fc & PAUSE_AUTONEG) != 0;
+	epause->rx_pause = (p->link_cfg.advertised_fc & PAUSE_RX) != 0;
+	epause->tx_pause = (p->link_cfg.advertised_fc & PAUSE_TX) != 0;
+}
+
+static int set_pauseparam(struct net_device *dev,
+			  struct ethtool_pauseparam *epause)
+{
+	struct port_info *p = netdev_priv(dev);
+	struct link_config *lc = &p->link_cfg;
+
+	if (epause->autoneg == AUTONEG_DISABLE)
+		lc->requested_fc = 0;
+	else if (lc->pcaps & FW_PORT_CAP32_ANEG)
+		lc->requested_fc = PAUSE_AUTONEG;
+	else
+		return -EINVAL;
+
+	if (epause->rx_pause)
+		lc->requested_fc |= PAUSE_RX;
+	if (epause->tx_pause)
+		lc->requested_fc |= PAUSE_TX;
+	if (netif_running(dev))
+		return t4_link_l1cfg(p->adapter, p->adapter->mbox, p->tx_chan,
+				     lc);
+	return 0;
+}
+
+static void get_sge_param(struct net_device *dev, struct ethtool_ringparam *e,
+			  struct kernel_ethtool_ringparam *kernel_e,
+			  struct netlink_ext_ack *extack)
+{
+	const struct port_info *pi = netdev_priv(dev);
+	const struct sge *s = &pi->adapter->sge;
+
+	e->rx_max_pending = MAX_RX_BUFFERS;
+	e->rx_mini_max_pending = MAX_RSPQ_ENTRIES;
+	e->rx_jumbo_max_pending = 0;
+	e->tx_max_pending = MAX_TXQ_ENTRIES;
+
+	e->rx_pending = s->ethrxq[pi->first_qset].fl.size - 8;
+	e->rx_mini_pending = s->ethrxq[pi->first_qset].rspq.size;
+	e->rx_jumbo_pending = 0;
+	e->tx_pending = s->ethtxq[pi->first_qset].q.size;
+}
+
+static int set_sge_param(struct net_device *dev, struct ethtool_ringparam *e,
+			 struct kernel_ethtool_ringparam *kernel_e,
+			 struct netlink_ext_ack *extack)
+{
+	int i;
+	const struct port_info *pi = netdev_priv(dev);
+	struct adapter *adapter = pi->adapter;
+	struct sge *s = &adapter->sge;
+
+	if (e->rx_pending > MAX_RX_BUFFERS || e->rx_jumbo_pending ||
+	    e->tx_pending > MAX_TXQ_ENTRIES ||
+	    e->rx_mini_pending > MAX_RSPQ_ENTRIES ||
+	    e->rx_mini_pending < MIN_RSPQ_ENTRIES ||
+	    e->rx_pending < MIN_FL_ENTRIES || e->tx_pending < MIN_TXQ_ENTRIES)
+		return -EINVAL;
+
+	if (adapter->flags & CXGB4_FULL_INIT_DONE)
+		return -EBUSY;
+
+	for (i = 0; i < pi->nqsets; ++i) {
+		s->ethtxq[pi->first_qset + i].q.size = e->tx_pending;
+		s->ethrxq[pi->first_qset + i].fl.size = e->rx_pending + 8;
+		s->ethrxq[pi->first_qset + i].rspq.size = e->rx_mini_pending;
+	}
+	return 0;
+}
+
+/**
+ * set_rx_intr_params - set a net devices's RX interrupt holdoff paramete!
+ * @dev: the network device
+ * @us: the hold-off time in us, or 0 to disable timer
+ * @cnt: the hold-off packet count, or 0 to disable counter
+ *
+ * Set the RX interrupt hold-off parameters for a network device.
+ */
+static int set_rx_intr_params(struct net_device *dev,
+			      unsigned int us, unsigned int cnt)
+{
+	int i, err;
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	struct sge_eth_rxq *q = &adap->sge.ethrxq[pi->first_qset];
+
+	for (i = 0; i < pi->nqsets; i++, q++) {
+		err = cxgb4_set_rspq_intr_params(&q->rspq, us, cnt);
+		if (err)
+			return err;
+	}
+	return 0;
+}
+
+static int set_adaptive_rx_setting(struct net_device *dev, int adaptive_rx)
+{
+	int i;
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	struct sge_eth_rxq *q = &adap->sge.ethrxq[pi->first_qset];
+
+	for (i = 0; i < pi->nqsets; i++, q++)
+		q->rspq.adaptive_rx = adaptive_rx;
+
+	return 0;
+}
+
+static int get_adaptive_rx_setting(struct net_device *dev)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	struct sge_eth_rxq *q = &adap->sge.ethrxq[pi->first_qset];
+
+	return q->rspq.adaptive_rx;
+}
+
+/* Return the current global Adapter SGE Doorbell Queue Timer Tick for all
+ * Ethernet TX Queues.
+ */
+static int get_dbqtimer_tick(struct net_device *dev)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+
+	if (!(adap->flags & CXGB4_SGE_DBQ_TIMER))
+		return 0;
+
+	return adap->sge.dbqtimer_tick;
+}
+
+/* Return the SGE Doorbell Queue Timer Value for the Ethernet TX Queues
+ * associated with a Network Device.
+ */
+static int get_dbqtimer(struct net_device *dev)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	struct sge_eth_txq *txq;
+
+	txq = &adap->sge.ethtxq[pi->first_qset];
+
+	if (!(adap->flags & CXGB4_SGE_DBQ_TIMER))
+		return 0;
+
+	/* all of the TX Queues use the same Timer Index */
+	return adap->sge.dbqtimer_val[txq->dbqtimerix];
+}
+
+/* Set the global Adapter SGE Doorbell Queue Timer Tick for all Ethernet TX
+ * Queues.  This is the fundamental "Tick" that sets the scale of values which
+ * can be used.  Individual Ethernet TX Queues index into a relatively small
+ * array of Tick Multipliers.  Changing the base Tick will thus change all of
+ * the resulting Timer Values associated with those multipliers for all
+ * Ethernet TX Queues.
+ */
+static int set_dbqtimer_tick(struct net_device *dev, int usecs)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	struct sge *s = &adap->sge;
+	u32 param, val;
+	int ret;
+
+	if (!(adap->flags & CXGB4_SGE_DBQ_TIMER))
+		return 0;
+
+	/* return early if it's the same Timer Tick we're already using */
+	if (s->dbqtimer_tick == usecs)
+		return 0;
+
+	/* attempt to set the new Timer Tick value */
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_DBQ_TIMERTICK));
+	val = usecs;
+	ret = t4_set_params(adap, adap->mbox, adap->pf, 0, 1, &param, &val);
+	if (ret)
+		return ret;
+	s->dbqtimer_tick = usecs;
+
+	/* if successful, reread resulting dependent Timer values */
+	ret = t4_read_sge_dbqtimers(adap, ARRAY_SIZE(s->dbqtimer_val),
+				    s->dbqtimer_val);
+	return ret;
+}
+
+/* Set the SGE Doorbell Queue Timer Value for the Ethernet TX Queues
+ * associated with a Network Device.  There is a relatively small array of
+ * possible Timer Values so we need to pick the closest value available.
+ */
+static int set_dbqtimer(struct net_device *dev, int usecs)
+{
+	int qix, timerix, min_timerix, delta, min_delta;
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	struct sge *s = &adap->sge;
+	struct sge_eth_txq *txq;
+	u32 param, val;
+	int ret;
+
+	if (!(adap->flags & CXGB4_SGE_DBQ_TIMER))
+		return 0;
+
+	/* Find the SGE Doorbell Timer Value that's closest to the requested
+	 * value.
+	 */
+	min_delta = INT_MAX;
+	min_timerix = 0;
+	for (timerix = 0; timerix < ARRAY_SIZE(s->dbqtimer_val); timerix++) {
+		delta = s->dbqtimer_val[timerix] - usecs;
+		if (delta < 0)
+			delta = -delta;
+		if (delta < min_delta) {
+			min_delta = delta;
+			min_timerix = timerix;
+		}
+	}
+
+	/* Return early if it's the same Timer Index we're already using.
+	 * We use the same Timer Index for all of the TX Queues for an
+	 * interface so it's only necessary to check the first one.
+	 */
+	txq = &s->ethtxq[pi->first_qset];
+	if (txq->dbqtimerix == min_timerix)
+		return 0;
+
+	for (qix = 0; qix < pi->nqsets; qix++, txq++) {
+		if (adap->flags & CXGB4_FULL_INIT_DONE) {
+			param =
+			 (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DMAQ) |
+			  FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DMAQ_EQ_TIMERIX) |
+			  FW_PARAMS_PARAM_YZ_V(txq->q.cntxt_id));
+			val = min_timerix;
+			ret = t4_set_params(adap, adap->mbox, adap->pf, 0,
+					    1, &param, &val);
+			if (ret)
+				return ret;
+		}
+		txq->dbqtimerix = min_timerix;
+	}
+	return 0;
+}
+
+/* Set the global Adapter SGE Doorbell Queue Timer Tick for all Ethernet TX
+ * Queues and the Timer Value for the Ethernet TX Queues associated with a
+ * Network Device.  Since changing the global Tick changes all of the
+ * available Timer Values, we need to do this first before selecting the
+ * resulting closest Timer Value.  Moreover, since the Tick is global,
+ * changing it affects the Timer Values for all Network Devices on the
+ * adapter.  So, before changing the Tick, we grab all of the current Timer
+ * Values for other Network Devices on this Adapter and then attempt to select
+ * new Timer Values which are close to the old values ...
+ */
+static int set_dbqtimer_tickval(struct net_device *dev,
+				int tick_usecs, int timer_usecs)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	int timer[MAX_NPORTS];
+	unsigned int port;
+	int ret;
+
+	/* Grab the other adapter Network Interface current timers and fill in
+	 * the new one for this Network Interface.
+	 */
+	for_each_port(adap, port)
+		if (port == pi->port_id)
+			timer[port] = timer_usecs;
+		else
+			timer[port] = get_dbqtimer(adap->port[port]);
+
+	/* Change the global Tick first ... */
+	ret = set_dbqtimer_tick(dev, tick_usecs);
+	if (ret)
+		return ret;
+
+	/* ... and then set all of the Network Interface Timer Values ... */
+	for_each_port(adap, port) {
+		ret = set_dbqtimer(adap->port[port], timer[port]);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static int set_coalesce(struct net_device *dev,
+			struct ethtool_coalesce *coalesce,
+			struct kernel_ethtool_coalesce *kernel_coal,
+			struct netlink_ext_ack *extack)
+{
+	int ret;
+
+	set_adaptive_rx_setting(dev, coalesce->use_adaptive_rx_coalesce);
+
+	ret = set_rx_intr_params(dev, coalesce->rx_coalesce_usecs,
+				 coalesce->rx_max_coalesced_frames);
+	if (ret)
+		return ret;
+
+	return set_dbqtimer_tickval(dev,
+				    coalesce->tx_coalesce_usecs_irq,
+				    coalesce->tx_coalesce_usecs);
+}
+
+static int get_coalesce(struct net_device *dev, struct ethtool_coalesce *c,
+			struct kernel_ethtool_coalesce *kernel_coal,
+			struct netlink_ext_ack *extack)
+{
+	const struct port_info *pi = netdev_priv(dev);
+	const struct adapter *adap = pi->adapter;
+	const struct sge_rspq *rq = &adap->sge.ethrxq[pi->first_qset].rspq;
+
+	c->rx_coalesce_usecs = qtimer_val(adap, rq);
+	c->rx_max_coalesced_frames = (rq->intr_params & QINTR_CNT_EN_F) ?
+		adap->sge.counter_val[rq->pktcnt_idx] : 0;
+	c->use_adaptive_rx_coalesce = get_adaptive_rx_setting(dev);
+	c->tx_coalesce_usecs_irq = get_dbqtimer_tick(dev);
+	c->tx_coalesce_usecs = get_dbqtimer(dev);
+	return 0;
+}
+
+/* The next two routines implement eeprom read/write from physical addresses.
+ */
+static int eeprom_rd_phys(struct adapter *adap, unsigned int phys_addr, u32 *v)
+{
+	int vaddr = t4_eeprom_ptov(phys_addr, adap->pf, EEPROMPFSIZE);
+
+	if (vaddr >= 0)
+		vaddr = pci_read_vpd(adap->pdev, vaddr, sizeof(u32), v);
+	return vaddr < 0 ? vaddr : 0;
+}
+
+static int eeprom_wr_phys(struct adapter *adap, unsigned int phys_addr, u32 v)
+{
+	int vaddr = t4_eeprom_ptov(phys_addr, adap->pf, EEPROMPFSIZE);
+
+	if (vaddr >= 0)
+		vaddr = pci_write_vpd(adap->pdev, vaddr, sizeof(u32), &v);
+	return vaddr < 0 ? vaddr : 0;
+}
+
+#define EEPROM_MAGIC 0x38E2F10C
+
+static int get_eeprom(struct net_device *dev, struct ethtool_eeprom *e,
+		      u8 *data)
+{
+	int i, err = 0;
+	struct adapter *adapter = netdev2adap(dev);
+	u8 *buf = kvzalloc(EEPROMSIZE, GFP_KERNEL);
+
+	if (!buf)
+		return -ENOMEM;
+
+	e->magic = EEPROM_MAGIC;
+	for (i = e->offset & ~3; !err && i < e->offset + e->len; i += 4)
+		err = eeprom_rd_phys(adapter, i, (u32 *)&buf[i]);
+
+	if (!err)
+		memcpy(data, buf + e->offset, e->len);
+	kvfree(buf);
+	return err;
+}
+
+static int set_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom,
+		      u8 *data)
+{
+	u8 *buf;
+	int err = 0;
+	u32 aligned_offset, aligned_len, *p;
+	struct adapter *adapter = netdev2adap(dev);
+
+	if (eeprom->magic != EEPROM_MAGIC)
+		return -EINVAL;
+
+	aligned_offset = eeprom->offset & ~3;
+	aligned_len = (eeprom->len + (eeprom->offset & 3) + 3) & ~3;
+
+	if (adapter->pf > 0) {
+		u32 start = 1024 + adapter->pf * EEPROMPFSIZE;
+
+		if (aligned_offset < start ||
+		    aligned_offset + aligned_len > start + EEPROMPFSIZE)
+			return -EPERM;
+	}
+
+	if (aligned_offset != eeprom->offset || aligned_len != eeprom->len) {
+		/* RMW possibly needed for first or last words.
+		 */
+		buf = kvzalloc(aligned_len, GFP_KERNEL);
+		if (!buf)
+			return -ENOMEM;
+		err = eeprom_rd_phys(adapter, aligned_offset, (u32 *)buf);
+		if (!err && aligned_len > 4)
+			err = eeprom_rd_phys(adapter,
+					     aligned_offset + aligned_len - 4,
+					     (u32 *)&buf[aligned_len - 4]);
+		if (err)
+			goto out;
+		memcpy(buf + (eeprom->offset & 3), data, eeprom->len);
+	} else {
+		buf = data;
+	}
+
+	err = t4_seeprom_wp(adapter, false);
+	if (err)
+		goto out;
+
+	for (p = (u32 *)buf; !err && aligned_len; aligned_len -= 4, p++) {
+		err = eeprom_wr_phys(adapter, aligned_offset, *p);
+		aligned_offset += 4;
+	}
+
+	if (!err)
+		err = t4_seeprom_wp(adapter, true);
+out:
+	if (buf != data)
+		kvfree(buf);
+	return err;
+}
+
+static int cxgb4_ethtool_flash_bootcfg(struct net_device *netdev,
+				       const u8 *data, u32 size)
+{
+	struct adapter *adap = netdev2adap(netdev);
+	int ret;
+
+	ret = t4_load_bootcfg(adap, data, size);
+	if (ret)
+		dev_err(adap->pdev_dev, "Failed to load boot cfg image\n");
+
+	return ret;
+}
+
+static int cxgb4_ethtool_flash_boot(struct net_device *netdev,
+				    const u8 *bdata, u32 size)
+{
+	struct adapter *adap = netdev2adap(netdev);
+	unsigned int offset;
+	u8 *data;
+	int ret;
+
+	data = kmemdup(bdata, size, GFP_KERNEL);
+	if (!data)
+		return -ENOMEM;
+
+	offset = OFFSET_G(t4_read_reg(adap, PF_REG(0, PCIE_PF_EXPROM_OFST_A)));
+
+	ret = t4_load_boot(adap, data, offset, size);
+	if (ret)
+		dev_err(adap->pdev_dev, "Failed to load boot image\n");
+
+	kfree(data);
+	return ret;
+}
+
+#define CXGB4_PHY_SIG 0x130000ea
+
+static int cxgb4_validate_phy_image(const u8 *data, u32 *size)
+{
+	struct cxgb4_fw_data *header;
+
+	header = (struct cxgb4_fw_data *)data;
+	if (be32_to_cpu(header->signature) != CXGB4_PHY_SIG)
+		return -EINVAL;
+
+	return 0;
+}
+
+static int cxgb4_ethtool_flash_phy(struct net_device *netdev,
+				   const u8 *data, u32 size)
+{
+	struct adapter *adap = netdev2adap(netdev);
+	int ret;
+
+	ret = cxgb4_validate_phy_image(data, NULL);
+	if (ret) {
+		dev_err(adap->pdev_dev, "PHY signature mismatch\n");
+		return ret;
+	}
+
+	/* We have to RESET the chip/firmware because we need the
+	 * chip in uninitialized state for loading new PHY image.
+	 * Otherwise, the running firmware will only store the PHY
+	 * image in local RAM which will be lost after next reset.
+	 */
+	ret = t4_fw_reset(adap, adap->mbox, PIORSTMODE_F | PIORST_F);
+	if (ret < 0) {
+		dev_err(adap->pdev_dev,
+			"Set FW to RESET for flashing PHY FW failed. ret: %d\n",
+			ret);
+		return ret;
+	}
+
+	ret = t4_load_phy_fw(adap, MEMWIN_NIC, NULL, data, size);
+	if (ret < 0) {
+		dev_err(adap->pdev_dev, "Failed to load PHY FW. ret: %d\n",
+			ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static int cxgb4_ethtool_flash_fw(struct net_device *netdev,
+				  const u8 *data, u32 size)
+{
+	struct adapter *adap = netdev2adap(netdev);
+	unsigned int mbox = PCIE_FW_MASTER_M + 1;
+	int ret;
+
+	/* If the adapter has been fully initialized then we'll go ahead and
+	 * try to get the firmware's cooperation in upgrading to the new
+	 * firmware image otherwise we'll try to do the entire job from the
+	 * host ... and we always "force" the operation in this path.
+	 */
+	if (adap->flags & CXGB4_FULL_INIT_DONE)
+		mbox = adap->mbox;
+
+	ret = t4_fw_upgrade(adap, mbox, data, size, 1);
+	if (ret)
+		dev_err(adap->pdev_dev,
+			"Failed to flash firmware\n");
+
+	return ret;
+}
+
+static int cxgb4_ethtool_flash_region(struct net_device *netdev,
+				      const u8 *data, u32 size, u32 region)
+{
+	struct adapter *adap = netdev2adap(netdev);
+	int ret;
+
+	switch (region) {
+	case CXGB4_ETHTOOL_FLASH_FW:
+		ret = cxgb4_ethtool_flash_fw(netdev, data, size);
+		break;
+	case CXGB4_ETHTOOL_FLASH_PHY:
+		ret = cxgb4_ethtool_flash_phy(netdev, data, size);
+		break;
+	case CXGB4_ETHTOOL_FLASH_BOOT:
+		ret = cxgb4_ethtool_flash_boot(netdev, data, size);
+		break;
+	case CXGB4_ETHTOOL_FLASH_BOOTCFG:
+		ret = cxgb4_ethtool_flash_bootcfg(netdev, data, size);
+		break;
+	default:
+		ret = -EOPNOTSUPP;
+		break;
+	}
+
+	if (!ret)
+		dev_info(adap->pdev_dev,
+			 "loading %s successful, reload cxgb4 driver\n",
+			 flash_region_strings[region]);
+	return ret;
+}
+
+#define CXGB4_FW_SIG 0x4368656c
+#define CXGB4_FW_SIG_OFFSET 0x160
+
+static int cxgb4_validate_fw_image(const u8 *data, u32 *size)
+{
+	struct cxgb4_fw_data *header;
+
+	header = (struct cxgb4_fw_data *)&data[CXGB4_FW_SIG_OFFSET];
+	if (be32_to_cpu(header->signature) != CXGB4_FW_SIG)
+		return -EINVAL;
+
+	if (size)
+		*size = be16_to_cpu(((struct fw_hdr *)data)->len512) * 512;
+
+	return 0;
+}
+
+static int cxgb4_validate_bootcfg_image(const u8 *data, u32 *size)
+{
+	struct cxgb4_bootcfg_data *header;
+
+	header = (struct cxgb4_bootcfg_data *)data;
+	if (le16_to_cpu(header->signature) != BOOT_CFG_SIG)
+		return -EINVAL;
+
+	return 0;
+}
+
+static int cxgb4_validate_boot_image(const u8 *data, u32 *size)
+{
+	struct cxgb4_pci_exp_rom_header *exp_header;
+	struct cxgb4_pcir_data *pcir_header;
+	struct legacy_pci_rom_hdr *header;
+	const u8 *cur_header = data;
+	u16 pcir_offset;
+
+	exp_header = (struct cxgb4_pci_exp_rom_header *)data;
+
+	if (le16_to_cpu(exp_header->signature) != BOOT_SIGNATURE)
+		return -EINVAL;
+
+	if (size) {
+		do {
+			header = (struct legacy_pci_rom_hdr *)cur_header;
+			pcir_offset = le16_to_cpu(header->pcir_offset);
+			pcir_header = (struct cxgb4_pcir_data *)(cur_header +
+				      pcir_offset);
+
+			*size += header->size512 * 512;
+			cur_header += header->size512 * 512;
+		} while (!(pcir_header->indicator & CXGB4_HDR_INDI));
+	}
+
+	return 0;
+}
+
+static int cxgb4_ethtool_get_flash_region(const u8 *data, u32 *size)
+{
+	if (!cxgb4_validate_fw_image(data, size))
+		return CXGB4_ETHTOOL_FLASH_FW;
+	if (!cxgb4_validate_boot_image(data, size))
+		return CXGB4_ETHTOOL_FLASH_BOOT;
+	if (!cxgb4_validate_phy_image(data, size))
+		return CXGB4_ETHTOOL_FLASH_PHY;
+	if (!cxgb4_validate_bootcfg_image(data, size))
+		return CXGB4_ETHTOOL_FLASH_BOOTCFG;
+
+	return -EOPNOTSUPP;
+}
+
+static int set_flash(struct net_device *netdev, struct ethtool_flash *ef)
+{
+	struct adapter *adap = netdev2adap(netdev);
+	const struct firmware *fw;
+	unsigned int master;
+	u8 master_vld = 0;
+	const u8 *fw_data;
+	size_t fw_size;
+	u32 size = 0;
+	u32 pcie_fw;
+	int region;
+	int ret;
+
+	pcie_fw = t4_read_reg(adap, PCIE_FW_A);
+	master = PCIE_FW_MASTER_G(pcie_fw);
+	if (pcie_fw & PCIE_FW_MASTER_VLD_F)
+		master_vld = 1;
+	/* if csiostor is the master return */
+	if (master_vld && (master != adap->pf)) {
+		dev_warn(adap->pdev_dev,
+			 "cxgb4 driver needs to be loaded as MASTER to support FW flash\n");
+		return -EOPNOTSUPP;
+	}
+
+	ef->data[sizeof(ef->data) - 1] = '\0';
+	ret = request_firmware(&fw, ef->data, adap->pdev_dev);
+	if (ret < 0)
+		return ret;
+
+	fw_data = fw->data;
+	fw_size = fw->size;
+	if (ef->region == ETHTOOL_FLASH_ALL_REGIONS) {
+		while (fw_size > 0) {
+			size = 0;
+			region = cxgb4_ethtool_get_flash_region(fw_data, &size);
+			if (region < 0 || !size) {
+				ret = region;
+				goto out_free_fw;
+			}
+
+			ret = cxgb4_ethtool_flash_region(netdev, fw_data, size,
+							 region);
+			if (ret)
+				goto out_free_fw;
+
+			fw_data += size;
+			fw_size -= size;
+		}
+	} else {
+		ret = cxgb4_ethtool_flash_region(netdev, fw_data, fw_size,
+						 ef->region);
+	}
+
+out_free_fw:
+	release_firmware(fw);
+	return ret;
+}
+
+static int get_ts_info(struct net_device *dev, struct ethtool_ts_info *ts_info)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct  adapter *adapter = pi->adapter;
+
+	ts_info->so_timestamping = SOF_TIMESTAMPING_TX_SOFTWARE |
+				   SOF_TIMESTAMPING_RX_SOFTWARE |
+				   SOF_TIMESTAMPING_SOFTWARE;
+
+	ts_info->so_timestamping |= SOF_TIMESTAMPING_RX_HARDWARE |
+				    SOF_TIMESTAMPING_TX_HARDWARE |
+				    SOF_TIMESTAMPING_RAW_HARDWARE;
+
+	ts_info->tx_types = (1 << HWTSTAMP_TX_OFF) |
+			    (1 << HWTSTAMP_TX_ON);
+
+	ts_info->rx_filters = (1 << HWTSTAMP_FILTER_NONE) |
+			      (1 << HWTSTAMP_FILTER_PTP_V2_L4_EVENT) |
+			      (1 << HWTSTAMP_FILTER_PTP_V1_L4_SYNC) |
+			      (1 << HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ) |
+			      (1 << HWTSTAMP_FILTER_PTP_V2_L4_SYNC) |
+			      (1 << HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ);
+
+	if (adapter->ptp_clock)
+		ts_info->phc_index = ptp_clock_index(adapter->ptp_clock);
+	else
+		ts_info->phc_index = -1;
+
+	return 0;
+}
+
+static u32 get_rss_table_size(struct net_device *dev)
+{
+	const struct port_info *pi = netdev_priv(dev);
+
+	return pi->rss_size;
+}
+
+static int get_rss_table(struct net_device *dev, u32 *p, u8 *key, u8 *hfunc)
+{
+	const struct port_info *pi = netdev_priv(dev);
+	unsigned int n = pi->rss_size;
+
+	if (hfunc)
+		*hfunc = ETH_RSS_HASH_TOP;
+	if (!p)
+		return 0;
+	while (n--)
+		p[n] = pi->rss[n];
+	return 0;
+}
+
+static int set_rss_table(struct net_device *dev, const u32 *p, const u8 *key,
+			 const u8 hfunc)
+{
+	unsigned int i;
+	struct port_info *pi = netdev_priv(dev);
+
+	/* We require at least one supported parameter to be changed and no
+	 * change in any of the unsupported parameters
+	 */
+	if (key ||
+	    (hfunc != ETH_RSS_HASH_NO_CHANGE && hfunc != ETH_RSS_HASH_TOP))
+		return -EOPNOTSUPP;
+	if (!p)
+		return 0;
+
+	/* Interface must be brought up atleast once */
+	if (pi->adapter->flags & CXGB4_FULL_INIT_DONE) {
+		for (i = 0; i < pi->rss_size; i++)
+			pi->rss[i] = p[i];
+
+		return cxgb4_write_rss(pi, pi->rss);
+	}
+
+	return -EPERM;
+}
+
+static struct filter_entry *cxgb4_get_filter_entry(struct adapter *adap,
+						   u32 ftid)
+{
+	struct tid_info *t = &adap->tids;
+
+	if (ftid >= t->hpftid_base && ftid < t->hpftid_base + t->nhpftids)
+		return &t->hpftid_tab[ftid - t->hpftid_base];
+
+	if (ftid >= t->ftid_base && ftid < t->ftid_base + t->nftids)
+		return &t->ftid_tab[ftid - t->ftid_base];
+
+	return lookup_tid(t, ftid);
+}
+
+static void cxgb4_fill_filter_rule(struct ethtool_rx_flow_spec *fs,
+				   struct ch_filter_specification *dfs)
+{
+	switch (dfs->val.proto) {
+	case IPPROTO_TCP:
+		if (dfs->type)
+			fs->flow_type = TCP_V6_FLOW;
+		else
+			fs->flow_type = TCP_V4_FLOW;
+		break;
+	case IPPROTO_UDP:
+		if (dfs->type)
+			fs->flow_type = UDP_V6_FLOW;
+		else
+			fs->flow_type = UDP_V4_FLOW;
+		break;
+	}
+
+	if (dfs->type) {
+		fs->h_u.tcp_ip6_spec.psrc = cpu_to_be16(dfs->val.fport);
+		fs->m_u.tcp_ip6_spec.psrc = cpu_to_be16(dfs->mask.fport);
+		fs->h_u.tcp_ip6_spec.pdst = cpu_to_be16(dfs->val.lport);
+		fs->m_u.tcp_ip6_spec.pdst = cpu_to_be16(dfs->mask.lport);
+		memcpy(&fs->h_u.tcp_ip6_spec.ip6src, &dfs->val.fip[0],
+		       sizeof(fs->h_u.tcp_ip6_spec.ip6src));
+		memcpy(&fs->m_u.tcp_ip6_spec.ip6src, &dfs->mask.fip[0],
+		       sizeof(fs->m_u.tcp_ip6_spec.ip6src));
+		memcpy(&fs->h_u.tcp_ip6_spec.ip6dst, &dfs->val.lip[0],
+		       sizeof(fs->h_u.tcp_ip6_spec.ip6dst));
+		memcpy(&fs->m_u.tcp_ip6_spec.ip6dst, &dfs->mask.lip[0],
+		       sizeof(fs->m_u.tcp_ip6_spec.ip6dst));
+		fs->h_u.tcp_ip6_spec.tclass = dfs->val.tos;
+		fs->m_u.tcp_ip6_spec.tclass = dfs->mask.tos;
+	} else {
+		fs->h_u.tcp_ip4_spec.psrc = cpu_to_be16(dfs->val.fport);
+		fs->m_u.tcp_ip4_spec.psrc = cpu_to_be16(dfs->mask.fport);
+		fs->h_u.tcp_ip4_spec.pdst = cpu_to_be16(dfs->val.lport);
+		fs->m_u.tcp_ip4_spec.pdst = cpu_to_be16(dfs->mask.lport);
+		memcpy(&fs->h_u.tcp_ip4_spec.ip4src, &dfs->val.fip[0],
+		       sizeof(fs->h_u.tcp_ip4_spec.ip4src));
+		memcpy(&fs->m_u.tcp_ip4_spec.ip4src, &dfs->mask.fip[0],
+		       sizeof(fs->m_u.tcp_ip4_spec.ip4src));
+		memcpy(&fs->h_u.tcp_ip4_spec.ip4dst, &dfs->val.lip[0],
+		       sizeof(fs->h_u.tcp_ip4_spec.ip4dst));
+		memcpy(&fs->m_u.tcp_ip4_spec.ip4dst, &dfs->mask.lip[0],
+		       sizeof(fs->m_u.tcp_ip4_spec.ip4dst));
+		fs->h_u.tcp_ip4_spec.tos = dfs->val.tos;
+		fs->m_u.tcp_ip4_spec.tos = dfs->mask.tos;
+	}
+	fs->h_ext.vlan_tci = cpu_to_be16(dfs->val.ivlan);
+	fs->m_ext.vlan_tci = cpu_to_be16(dfs->mask.ivlan);
+	fs->flow_type |= FLOW_EXT;
+
+	if (dfs->action == FILTER_DROP)
+		fs->ring_cookie = RX_CLS_FLOW_DISC;
+	else
+		fs->ring_cookie = dfs->iq;
+}
+
+static int cxgb4_ntuple_get_filter(struct net_device *dev,
+				   struct ethtool_rxnfc *cmd,
+				   unsigned int loc)
+{
+	const struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct filter_entry *f;
+	int ftid;
+
+	if (!(adap->flags & CXGB4_FULL_INIT_DONE))
+		return -EAGAIN;
+
+	/* Check for maximum filter range */
+	if (!adap->ethtool_filters)
+		return -EOPNOTSUPP;
+
+	if (loc >= adap->ethtool_filters->nentries)
+		return -ERANGE;
+
+	if (!test_bit(loc, adap->ethtool_filters->port[pi->port_id].bmap))
+		return -ENOENT;
+
+	ftid = adap->ethtool_filters->port[pi->port_id].loc_array[loc];
+
+	/* Fetch filter_entry */
+	f = cxgb4_get_filter_entry(adap, ftid);
+
+	cxgb4_fill_filter_rule(&cmd->fs, &f->fs);
+
+	return 0;
+}
+
+static int get_rxnfc(struct net_device *dev, struct ethtool_rxnfc *info,
+		     u32 *rules)
+{
+	const struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = netdev2adap(dev);
+	unsigned int count = 0, index = 0;
+	int ret = 0;
+
+	switch (info->cmd) {
+	case ETHTOOL_GRXFH: {
+		unsigned int v = pi->rss_mode;
+
+		info->data = 0;
+		switch (info->flow_type) {
+		case TCP_V4_FLOW:
+			if (v & FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN_F)
+				info->data = RXH_IP_SRC | RXH_IP_DST |
+					     RXH_L4_B_0_1 | RXH_L4_B_2_3;
+			else if (v & FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN_F)
+				info->data = RXH_IP_SRC | RXH_IP_DST;
+			break;
+		case UDP_V4_FLOW:
+			if ((v & FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN_F) &&
+			    (v & FW_RSS_VI_CONFIG_CMD_UDPEN_F))
+				info->data = RXH_IP_SRC | RXH_IP_DST |
+					     RXH_L4_B_0_1 | RXH_L4_B_2_3;
+			else if (v & FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN_F)
+				info->data = RXH_IP_SRC | RXH_IP_DST;
+			break;
+		case SCTP_V4_FLOW:
+		case AH_ESP_V4_FLOW:
+		case IPV4_FLOW:
+			if (v & FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN_F)
+				info->data = RXH_IP_SRC | RXH_IP_DST;
+			break;
+		case TCP_V6_FLOW:
+			if (v & FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN_F)
+				info->data = RXH_IP_SRC | RXH_IP_DST |
+					     RXH_L4_B_0_1 | RXH_L4_B_2_3;
+			else if (v & FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN_F)
+				info->data = RXH_IP_SRC | RXH_IP_DST;
+			break;
+		case UDP_V6_FLOW:
+			if ((v & FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN_F) &&
+			    (v & FW_RSS_VI_CONFIG_CMD_UDPEN_F))
+				info->data = RXH_IP_SRC | RXH_IP_DST |
+					     RXH_L4_B_0_1 | RXH_L4_B_2_3;
+			else if (v & FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN_F)
+				info->data = RXH_IP_SRC | RXH_IP_DST;
+			break;
+		case SCTP_V6_FLOW:
+		case AH_ESP_V6_FLOW:
+		case IPV6_FLOW:
+			if (v & FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN_F)
+				info->data = RXH_IP_SRC | RXH_IP_DST;
+			break;
+		}
+		return 0;
+	}
+	case ETHTOOL_GRXRINGS:
+		info->data = pi->nqsets;
+		return 0;
+	case ETHTOOL_GRXCLSRLCNT:
+		info->rule_cnt =
+		       adap->ethtool_filters->port[pi->port_id].in_use;
+		return 0;
+	case ETHTOOL_GRXCLSRULE:
+		return cxgb4_ntuple_get_filter(dev, info, info->fs.location);
+	case ETHTOOL_GRXCLSRLALL:
+		info->data = adap->ethtool_filters->nentries;
+		while (count < info->rule_cnt) {
+			ret = cxgb4_ntuple_get_filter(dev, info, index);
+			if (!ret)
+				rules[count++] = index;
+			index++;
+		}
+		return 0;
+	}
+
+	return -EOPNOTSUPP;
+}
+
+static int cxgb4_ntuple_del_filter(struct net_device *dev,
+				   struct ethtool_rxnfc *cmd)
+{
+	struct cxgb4_ethtool_filter_info *filter_info;
+	struct adapter *adapter = netdev2adap(dev);
+	struct port_info *pi = netdev_priv(dev);
+	struct filter_entry *f;
+	u32 filter_id;
+	int ret;
+
+	if (!(adapter->flags & CXGB4_FULL_INIT_DONE))
+		return -EAGAIN;  /* can still change nfilters */
+
+	if (!adapter->ethtool_filters)
+		return -EOPNOTSUPP;
+
+	if (cmd->fs.location >= adapter->ethtool_filters->nentries) {
+		dev_err(adapter->pdev_dev,
+			"Location must be < %u",
+			adapter->ethtool_filters->nentries);
+		return -ERANGE;
+	}
+
+	filter_info = &adapter->ethtool_filters->port[pi->port_id];
+
+	if (!test_bit(cmd->fs.location, filter_info->bmap))
+		return -ENOENT;
+
+	filter_id = filter_info->loc_array[cmd->fs.location];
+	f = cxgb4_get_filter_entry(adapter, filter_id);
+
+	if (f->fs.prio)
+		filter_id -= adapter->tids.hpftid_base;
+	else if (!f->fs.hash)
+		filter_id -= (adapter->tids.ftid_base - adapter->tids.nhpftids);
+
+	ret = cxgb4_flow_rule_destroy(dev, f->fs.tc_prio, &f->fs, filter_id);
+	if (ret)
+		goto err;
+
+	clear_bit(cmd->fs.location, filter_info->bmap);
+	filter_info->in_use--;
+
+err:
+	return ret;
+}
+
+/* Add Ethtool n-tuple filters. */
+static int cxgb4_ntuple_set_filter(struct net_device *netdev,
+				   struct ethtool_rxnfc *cmd)
+{
+	struct ethtool_rx_flow_spec_input input = {};
+	struct cxgb4_ethtool_filter_info *filter_info;
+	struct adapter *adapter = netdev2adap(netdev);
+	struct port_info *pi = netdev_priv(netdev);
+	struct ch_filter_specification fs;
+	struct ethtool_rx_flow_rule *flow;
+	u32 tid;
+	int ret;
+
+	if (!(adapter->flags & CXGB4_FULL_INIT_DONE))
+		return -EAGAIN;  /* can still change nfilters */
+
+	if (!adapter->ethtool_filters)
+		return -EOPNOTSUPP;
+
+	if (cmd->fs.location >= adapter->ethtool_filters->nentries) {
+		dev_err(adapter->pdev_dev,
+			"Location must be < %u",
+			adapter->ethtool_filters->nentries);
+		return -ERANGE;
+	}
+
+	if (test_bit(cmd->fs.location,
+		     adapter->ethtool_filters->port[pi->port_id].bmap))
+		return -EEXIST;
+
+	memset(&fs, 0, sizeof(fs));
+
+	input.fs = &cmd->fs;
+	flow = ethtool_rx_flow_rule_create(&input);
+	if (IS_ERR(flow)) {
+		ret = PTR_ERR(flow);
+		goto exit;
+	}
+
+	fs.hitcnts = 1;
+
+	ret = cxgb4_flow_rule_replace(netdev, flow->rule, cmd->fs.location,
+				      NULL, &fs, &tid);
+	if (ret)
+		goto free;
+
+	filter_info = &adapter->ethtool_filters->port[pi->port_id];
+
+	if (fs.prio)
+		tid += adapter->tids.hpftid_base;
+	else if (!fs.hash)
+		tid += (adapter->tids.ftid_base - adapter->tids.nhpftids);
+
+	filter_info->loc_array[cmd->fs.location] = tid;
+	set_bit(cmd->fs.location, filter_info->bmap);
+	filter_info->in_use++;
+
+free:
+	ethtool_rx_flow_rule_destroy(flow);
+exit:
+	return ret;
+}
+
+static int set_rxnfc(struct net_device *dev, struct ethtool_rxnfc *cmd)
+{
+	int ret = -EOPNOTSUPP;
+
+	switch (cmd->cmd) {
+	case ETHTOOL_SRXCLSRLINS:
+		ret = cxgb4_ntuple_set_filter(dev, cmd);
+		break;
+	case ETHTOOL_SRXCLSRLDEL:
+		ret = cxgb4_ntuple_del_filter(dev, cmd);
+		break;
+	default:
+		break;
+	}
+
+	return ret;
+}
+
+static int set_dump(struct net_device *dev, struct ethtool_dump *eth_dump)
+{
+	struct adapter *adapter = netdev2adap(dev);
+	u32 len = 0;
+
+	len = sizeof(struct cudbg_hdr) +
+	      sizeof(struct cudbg_entity_hdr) * CUDBG_MAX_ENTITY;
+	len += cxgb4_get_dump_length(adapter, eth_dump->flag);
+
+	adapter->eth_dump.flag = eth_dump->flag;
+	adapter->eth_dump.len = len;
+	return 0;
+}
+
+static int get_dump_flag(struct net_device *dev, struct ethtool_dump *eth_dump)
+{
+	struct adapter *adapter = netdev2adap(dev);
+
+	eth_dump->flag = adapter->eth_dump.flag;
+	eth_dump->len = adapter->eth_dump.len;
+	eth_dump->version = adapter->eth_dump.version;
+	return 0;
+}
+
+static int get_dump_data(struct net_device *dev, struct ethtool_dump *eth_dump,
+			 void *buf)
+{
+	struct adapter *adapter = netdev2adap(dev);
+	u32 len = 0;
+	int ret = 0;
+
+	if (adapter->eth_dump.flag == CXGB4_ETH_DUMP_NONE)
+		return -ENOENT;
+
+	len = sizeof(struct cudbg_hdr) +
+	      sizeof(struct cudbg_entity_hdr) * CUDBG_MAX_ENTITY;
+	len += cxgb4_get_dump_length(adapter, adapter->eth_dump.flag);
+	if (eth_dump->len < len)
+		return -ENOMEM;
+
+	ret = cxgb4_cudbg_collect(adapter, buf, &len, adapter->eth_dump.flag);
+	if (ret)
+		return ret;
+
+	eth_dump->flag = adapter->eth_dump.flag;
+	eth_dump->len = len;
+	eth_dump->version = adapter->eth_dump.version;
+	return 0;
+}
+
+static bool cxgb4_fw_mod_type_info_available(unsigned int fw_mod_type)
+{
+	/* Read port module EEPROM as long as it is plugged-in and
+	 * safe to read.
+	 */
+	return (fw_mod_type != FW_PORT_MOD_TYPE_NONE &&
+		fw_mod_type != FW_PORT_MOD_TYPE_ERROR);
+}
+
+static int cxgb4_get_module_info(struct net_device *dev,
+				 struct ethtool_modinfo *modinfo)
+{
+	struct port_info *pi = netdev_priv(dev);
+	u8 sff8472_comp, sff_diag_type, sff_rev;
+	struct adapter *adapter = pi->adapter;
+	int ret;
+
+	if (!cxgb4_fw_mod_type_info_available(pi->mod_type))
+		return -EINVAL;
+
+	switch (pi->port_type) {
+	case FW_PORT_TYPE_SFP:
+	case FW_PORT_TYPE_QSA:
+	case FW_PORT_TYPE_SFP28:
+		ret = t4_i2c_rd(adapter, adapter->mbox, pi->tx_chan,
+				I2C_DEV_ADDR_A0, SFF_8472_COMP_ADDR,
+				SFF_8472_COMP_LEN, &sff8472_comp);
+		if (ret)
+			return ret;
+		ret = t4_i2c_rd(adapter, adapter->mbox, pi->tx_chan,
+				I2C_DEV_ADDR_A0, SFP_DIAG_TYPE_ADDR,
+				SFP_DIAG_TYPE_LEN, &sff_diag_type);
+		if (ret)
+			return ret;
+
+		if (!sff8472_comp || (sff_diag_type & SFP_DIAG_ADDRMODE)) {
+			modinfo->type = ETH_MODULE_SFF_8079;
+			modinfo->eeprom_len = ETH_MODULE_SFF_8079_LEN;
+		} else {
+			modinfo->type = ETH_MODULE_SFF_8472;
+			if (sff_diag_type & SFP_DIAG_IMPLEMENTED)
+				modinfo->eeprom_len = ETH_MODULE_SFF_8472_LEN;
+			else
+				modinfo->eeprom_len = ETH_MODULE_SFF_8472_LEN / 2;
+		}
+		break;
+
+	case FW_PORT_TYPE_QSFP:
+	case FW_PORT_TYPE_QSFP_10G:
+	case FW_PORT_TYPE_CR_QSFP:
+	case FW_PORT_TYPE_CR2_QSFP:
+	case FW_PORT_TYPE_CR4_QSFP:
+		ret = t4_i2c_rd(adapter, adapter->mbox, pi->tx_chan,
+				I2C_DEV_ADDR_A0, SFF_REV_ADDR,
+				SFF_REV_LEN, &sff_rev);
+		/* For QSFP type ports, revision value >= 3
+		 * means the SFP is 8636 compliant.
+		 */
+		if (ret)
+			return ret;
+		if (sff_rev >= 0x3) {
+			modinfo->type = ETH_MODULE_SFF_8636;
+			modinfo->eeprom_len = ETH_MODULE_SFF_8636_LEN;
+		} else {
+			modinfo->type = ETH_MODULE_SFF_8436;
+			modinfo->eeprom_len = ETH_MODULE_SFF_8436_LEN;
+		}
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int cxgb4_get_module_eeprom(struct net_device *dev,
+				   struct ethtool_eeprom *eprom, u8 *data)
+{
+	int ret = 0, offset = eprom->offset, len = eprom->len;
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adapter = pi->adapter;
+
+	memset(data, 0, eprom->len);
+	if (offset + len <= I2C_PAGE_SIZE)
+		return t4_i2c_rd(adapter, adapter->mbox, pi->tx_chan,
+				 I2C_DEV_ADDR_A0, offset, len, data);
+
+	/* offset + len spans 0xa0 and 0xa1 pages */
+	if (offset <= I2C_PAGE_SIZE) {
+		/* read 0xa0 page */
+		len = I2C_PAGE_SIZE - offset;
+		ret =  t4_i2c_rd(adapter, adapter->mbox, pi->tx_chan,
+				 I2C_DEV_ADDR_A0, offset, len, data);
+		if (ret)
+			return ret;
+		offset = I2C_PAGE_SIZE;
+		/* Remaining bytes to be read from second page =
+		 * Total length - bytes read from first page
+		 */
+		len = eprom->len - len;
+	}
+	/* Read additional optical diagnostics from page 0xa2 if supported */
+	return t4_i2c_rd(adapter, adapter->mbox, pi->tx_chan, I2C_DEV_ADDR_A2,
+			 offset, len, &data[eprom->len - len]);
+}
+
+static u32 cxgb4_get_priv_flags(struct net_device *netdev)
+{
+	struct port_info *pi = netdev_priv(netdev);
+	struct adapter *adapter = pi->adapter;
+
+	return (adapter->eth_flags | pi->eth_flags);
+}
+
+/**
+ *	set_flags - set/unset specified flags if passed in new_flags
+ *	@cur_flags: pointer to current flags
+ *	@new_flags: new incoming flags
+ *	@flags: set of flags to set/unset
+ */
+static inline void set_flags(u32 *cur_flags, u32 new_flags, u32 flags)
+{
+	*cur_flags = (*cur_flags & ~flags) | (new_flags & flags);
+}
+
+static int cxgb4_set_priv_flags(struct net_device *netdev, u32 flags)
+{
+	struct port_info *pi = netdev_priv(netdev);
+	struct adapter *adapter = pi->adapter;
+
+	set_flags(&adapter->eth_flags, flags, PRIV_FLAGS_ADAP);
+	set_flags(&pi->eth_flags, flags, PRIV_FLAGS_PORT);
+
+	return 0;
+}
+
+static void cxgb4_lb_test(struct net_device *netdev, u64 *lb_status)
+{
+	int dev_state = netif_running(netdev);
+
+	if (dev_state) {
+		netif_tx_stop_all_queues(netdev);
+		netif_carrier_off(netdev);
+	}
+
+	*lb_status = cxgb4_selftest_lb_pkt(netdev);
+
+	if (dev_state) {
+		netif_tx_start_all_queues(netdev);
+		netif_carrier_on(netdev);
+	}
+}
+
+static void cxgb4_self_test(struct net_device *netdev,
+			    struct ethtool_test *eth_test, u64 *data)
+{
+	struct port_info *pi = netdev_priv(netdev);
+	struct adapter *adap = pi->adapter;
+
+	memset(data, 0, sizeof(u64) * CXGB4_ETHTOOL_MAX_TEST);
+
+	if (!(adap->flags & CXGB4_FULL_INIT_DONE) ||
+	    !(adap->flags & CXGB4_FW_OK)) {
+		eth_test->flags |= ETH_TEST_FL_FAILED;
+		return;
+	}
+
+	if (eth_test->flags & ETH_TEST_FL_OFFLINE)
+		cxgb4_lb_test(netdev, &data[CXGB4_ETHTOOL_LB_TEST]);
+
+	if (data[CXGB4_ETHTOOL_LB_TEST])
+		eth_test->flags |= ETH_TEST_FL_FAILED;
+}
+
+static const struct ethtool_ops cxgb_ethtool_ops = {
+	.supported_coalesce_params = ETHTOOL_COALESCE_USECS |
+				     ETHTOOL_COALESCE_RX_MAX_FRAMES |
+				     ETHTOOL_COALESCE_TX_USECS_IRQ |
+				     ETHTOOL_COALESCE_USE_ADAPTIVE_RX,
+	.get_link_ksettings = get_link_ksettings,
+	.set_link_ksettings = set_link_ksettings,
+	.get_fecparam      = get_fecparam,
+	.set_fecparam      = set_fecparam,
+	.get_drvinfo       = get_drvinfo,
+	.get_msglevel      = get_msglevel,
+	.set_msglevel      = set_msglevel,
+	.get_ringparam     = get_sge_param,
+	.set_ringparam     = set_sge_param,
+	.get_coalesce      = get_coalesce,
+	.set_coalesce      = set_coalesce,
+	.get_eeprom_len    = get_eeprom_len,
+	.get_eeprom        = get_eeprom,
+	.set_eeprom        = set_eeprom,
+	.get_pauseparam    = get_pauseparam,
+	.set_pauseparam    = set_pauseparam,
+	.get_link          = ethtool_op_get_link,
+	.get_strings       = get_strings,
+	.set_phys_id       = identify_port,
+	.nway_reset        = restart_autoneg,
+	.get_sset_count    = get_sset_count,
+	.get_ethtool_stats = get_stats,
+	.get_regs_len      = get_regs_len,
+	.get_regs          = get_regs,
+	.get_rxnfc         = get_rxnfc,
+	.set_rxnfc         = set_rxnfc,
+	.get_rxfh_indir_size = get_rss_table_size,
+	.get_rxfh	   = get_rss_table,
+	.set_rxfh	   = set_rss_table,
+	.self_test	   = cxgb4_self_test,
+	.flash_device      = set_flash,
+	.get_ts_info       = get_ts_info,
+	.set_dump          = set_dump,
+	.get_dump_flag     = get_dump_flag,
+	.get_dump_data     = get_dump_data,
+	.get_module_info   = cxgb4_get_module_info,
+	.get_module_eeprom = cxgb4_get_module_eeprom,
+	.get_priv_flags    = cxgb4_get_priv_flags,
+	.set_priv_flags    = cxgb4_set_priv_flags,
+};
+
+void cxgb4_cleanup_ethtool_filters(struct adapter *adap)
+{
+	struct cxgb4_ethtool_filter_info *eth_filter_info;
+	u8 i;
+
+	if (!adap->ethtool_filters)
+		return;
+
+	eth_filter_info = adap->ethtool_filters->port;
+
+	if (eth_filter_info) {
+		for (i = 0; i < adap->params.nports; i++) {
+			kvfree(eth_filter_info[i].loc_array);
+			bitmap_free(eth_filter_info[i].bmap);
+		}
+		kfree(eth_filter_info);
+	}
+
+	kfree(adap->ethtool_filters);
+}
+
+int cxgb4_init_ethtool_filters(struct adapter *adap)
+{
+	struct cxgb4_ethtool_filter_info *eth_filter_info;
+	struct cxgb4_ethtool_filter *eth_filter;
+	struct tid_info *tids = &adap->tids;
+	u32 nentries, i;
+	int ret;
+
+	eth_filter = kzalloc(sizeof(*eth_filter), GFP_KERNEL);
+	if (!eth_filter)
+		return -ENOMEM;
+
+	eth_filter_info = kcalloc(adap->params.nports,
+				  sizeof(*eth_filter_info),
+				  GFP_KERNEL);
+	if (!eth_filter_info) {
+		ret = -ENOMEM;
+		goto free_eth_filter;
+	}
+
+	eth_filter->port = eth_filter_info;
+
+	nentries = tids->nhpftids + tids->nftids;
+	if (is_hashfilter(adap))
+		nentries += tids->nhash +
+			    (adap->tids.stid_base - adap->tids.tid_base);
+	eth_filter->nentries = nentries;
+
+	for (i = 0; i < adap->params.nports; i++) {
+		eth_filter->port[i].loc_array = kvzalloc(nentries, GFP_KERNEL);
+		if (!eth_filter->port[i].loc_array) {
+			ret = -ENOMEM;
+			goto free_eth_finfo;
+		}
+
+		eth_filter->port[i].bmap = bitmap_zalloc(nentries, GFP_KERNEL);
+		if (!eth_filter->port[i].bmap) {
+			ret = -ENOMEM;
+			goto free_eth_finfo;
+		}
+	}
+
+	adap->ethtool_filters = eth_filter;
+	return 0;
+
+free_eth_finfo:
+	while (i-- > 0) {
+		bitmap_free(eth_filter->port[i].bmap);
+		kvfree(eth_filter->port[i].loc_array);
+	}
+	kfree(eth_filter_info);
+
+free_eth_filter:
+	kfree(eth_filter);
+
+	return ret;
+}
+
+void cxgb4_set_ethtool_ops(struct net_device *netdev)
+{
+	netdev->ethtool_ops = &cxgb_ethtool_ops;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_fcoe.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_fcoe.c
new file mode 100644
index 000000000..33b2c0c45
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_fcoe.c
@@ -0,0 +1,122 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2015 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifdef CONFIG_CHELSIO_T4_FCOE
+
+#include <scsi/fc/fc_fs.h>
+#include <scsi/libfcoe.h>
+#include "cxgb4.h"
+
+bool cxgb_fcoe_sof_eof_supported(struct adapter *adap, struct sk_buff *skb)
+{
+	struct fcoe_hdr *fcoeh = (struct fcoe_hdr *)skb_network_header(skb);
+	u8 sof = fcoeh->fcoe_sof;
+	u8 eof = 0;
+
+	if ((sof != FC_SOF_I3) && (sof != FC_SOF_N3)) {
+		dev_err(adap->pdev_dev, "Unsupported SOF 0x%x\n", sof);
+		return false;
+	}
+
+	skb_copy_bits(skb, skb->len - 4, &eof, 1);
+
+	if ((eof != FC_EOF_N) && (eof != FC_EOF_T)) {
+		dev_err(adap->pdev_dev, "Unsupported EOF 0x%x\n", eof);
+		return false;
+	}
+
+	return true;
+}
+
+/**
+ * cxgb_fcoe_enable - enable FCoE offload features
+ * @netdev: net device
+ *
+ * Returns 0 on success or -EINVAL on failure.
+ */
+int cxgb_fcoe_enable(struct net_device *netdev)
+{
+	struct port_info *pi = netdev_priv(netdev);
+	struct adapter *adap = pi->adapter;
+	struct cxgb_fcoe *fcoe = &pi->fcoe;
+
+	if (is_t4(adap->params.chip))
+		return -EINVAL;
+
+	if (!(adap->flags & CXGB4_FULL_INIT_DONE))
+		return -EINVAL;
+
+	dev_info(adap->pdev_dev, "Enabling FCoE offload features\n");
+
+	netdev->features |= NETIF_F_FCOE_CRC;
+	netdev->vlan_features |= NETIF_F_FCOE_CRC;
+	netdev->features |= NETIF_F_FCOE_MTU;
+	netdev->vlan_features |= NETIF_F_FCOE_MTU;
+
+	netdev_features_change(netdev);
+
+	fcoe->flags |= CXGB_FCOE_ENABLED;
+
+	return 0;
+}
+
+/**
+ * cxgb_fcoe_disable - disable FCoE offload
+ * @netdev: net device
+ *
+ * Returns 0 on success or -EINVAL on failure.
+ */
+int cxgb_fcoe_disable(struct net_device *netdev)
+{
+	struct port_info *pi = netdev_priv(netdev);
+	struct adapter *adap = pi->adapter;
+	struct cxgb_fcoe *fcoe = &pi->fcoe;
+
+	if (!(fcoe->flags & CXGB_FCOE_ENABLED))
+		return -EINVAL;
+
+	dev_info(adap->pdev_dev, "Disabling FCoE offload features\n");
+
+	fcoe->flags &= ~CXGB_FCOE_ENABLED;
+
+	netdev->features &= ~NETIF_F_FCOE_CRC;
+	netdev->vlan_features &= ~NETIF_F_FCOE_CRC;
+	netdev->features &= ~NETIF_F_FCOE_MTU;
+	netdev->vlan_features &= ~NETIF_F_FCOE_MTU;
+
+	netdev_features_change(netdev);
+
+	return 0;
+}
+#endif /* CONFIG_CHELSIO_T4_FCOE */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_fcoe.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_fcoe.h
new file mode 100644
index 000000000..bf9258a56
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_fcoe.h
@@ -0,0 +1,57 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2015 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_FCOE_H__
+#define __CXGB4_FCOE_H__
+
+#ifdef CONFIG_CHELSIO_T4_FCOE
+
+#define CXGB_FCOE_TXPKT_CSUM_START	28
+#define CXGB_FCOE_TXPKT_CSUM_END	8
+
+/* fcoe flags */
+enum {
+	CXGB_FCOE_ENABLED     = (1 << 0),
+};
+
+struct cxgb_fcoe {
+	u8	flags;
+};
+
+int cxgb_fcoe_enable(struct net_device *);
+int cxgb_fcoe_disable(struct net_device *);
+bool cxgb_fcoe_sof_eof_supported(struct adapter *, struct sk_buff *);
+
+#endif /* CONFIG_CHELSIO_T4_FCOE */
+#endif /* __CXGB4_FCOE_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c
new file mode 100644
index 000000000..786ceae34
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c
@@ -0,0 +1,2174 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include <net/ipv6.h>
+
+#include "cxgb4.h"
+#include "t4_regs.h"
+#include "t4_tcb.h"
+#include "t4_values.h"
+#include "clip_tbl.h"
+#include "l2t.h"
+#include "smt.h"
+#include "t4fw_api.h"
+#include "cxgb4_filter.h"
+
+static inline bool is_field_set(u32 val, u32 mask)
+{
+	return val || mask;
+}
+
+static inline bool unsupported(u32 conf, u32 conf_mask, u32 val, u32 mask)
+{
+	return !(conf & conf_mask) && is_field_set(val, mask);
+}
+
+static int set_tcb_field(struct adapter *adap, struct filter_entry *f,
+			 unsigned int ftid,  u16 word, u64 mask, u64 val,
+			 int no_reply)
+{
+	struct cpl_set_tcb_field *req;
+	struct sk_buff *skb;
+
+	skb = alloc_skb(sizeof(struct cpl_set_tcb_field), GFP_ATOMIC);
+	if (!skb)
+		return -ENOMEM;
+
+	req = (struct cpl_set_tcb_field *)__skb_put_zero(skb, sizeof(*req));
+	INIT_TP_WR_CPL(req, CPL_SET_TCB_FIELD, ftid);
+	req->reply_ctrl = htons(REPLY_CHAN_V(0) |
+				QUEUENO_V(adap->sge.fw_evtq.abs_id) |
+				NO_REPLY_V(no_reply));
+	req->word_cookie = htons(TCB_WORD_V(word) | TCB_COOKIE_V(ftid));
+	req->mask = cpu_to_be64(mask);
+	req->val = cpu_to_be64(val);
+	set_wr_txq(skb, CPL_PRIORITY_CONTROL, f->fs.val.iport & 0x3);
+	t4_ofld_send(adap, skb);
+	return 0;
+}
+
+/* Set one of the t_flags bits in the TCB.
+ */
+static int set_tcb_tflag(struct adapter *adap, struct filter_entry *f,
+			 unsigned int ftid, unsigned int bit_pos,
+			 unsigned int val, int no_reply)
+{
+	return set_tcb_field(adap, f, ftid,  TCB_T_FLAGS_W, 1ULL << bit_pos,
+			     (unsigned long long)val << bit_pos, no_reply);
+}
+
+static void mk_abort_req_ulp(struct cpl_abort_req *abort_req, unsigned int tid)
+{
+	struct ulp_txpkt *txpkt = (struct ulp_txpkt *)abort_req;
+	struct ulptx_idata *sc = (struct ulptx_idata *)(txpkt + 1);
+
+	txpkt->cmd_dest = htonl(ULPTX_CMD_V(ULP_TX_PKT) | ULP_TXPKT_DEST_V(0));
+	txpkt->len = htonl(DIV_ROUND_UP(sizeof(*abort_req), 16));
+	sc->cmd_more = htonl(ULPTX_CMD_V(ULP_TX_SC_IMM));
+	sc->len = htonl(sizeof(*abort_req) - sizeof(struct work_request_hdr));
+	OPCODE_TID(abort_req) = htonl(MK_OPCODE_TID(CPL_ABORT_REQ, tid));
+	abort_req->rsvd0 = htonl(0);
+	abort_req->rsvd1 = 0;
+	abort_req->cmd = CPL_ABORT_NO_RST;
+}
+
+static void mk_abort_rpl_ulp(struct cpl_abort_rpl *abort_rpl, unsigned int tid)
+{
+	struct ulp_txpkt *txpkt = (struct ulp_txpkt *)abort_rpl;
+	struct ulptx_idata *sc = (struct ulptx_idata *)(txpkt + 1);
+
+	txpkt->cmd_dest = htonl(ULPTX_CMD_V(ULP_TX_PKT) | ULP_TXPKT_DEST_V(0));
+	txpkt->len = htonl(DIV_ROUND_UP(sizeof(*abort_rpl), 16));
+	sc->cmd_more = htonl(ULPTX_CMD_V(ULP_TX_SC_IMM));
+	sc->len = htonl(sizeof(*abort_rpl) - sizeof(struct work_request_hdr));
+	OPCODE_TID(abort_rpl) = htonl(MK_OPCODE_TID(CPL_ABORT_RPL, tid));
+	abort_rpl->rsvd0 = htonl(0);
+	abort_rpl->rsvd1 = 0;
+	abort_rpl->cmd = CPL_ABORT_NO_RST;
+}
+
+static void mk_set_tcb_ulp(struct filter_entry *f,
+			   struct cpl_set_tcb_field *req,
+			   unsigned int word, u64 mask, u64 val,
+			   u8 cookie, int no_reply)
+{
+	struct ulp_txpkt *txpkt = (struct ulp_txpkt *)req;
+	struct ulptx_idata *sc = (struct ulptx_idata *)(txpkt + 1);
+
+	txpkt->cmd_dest = htonl(ULPTX_CMD_V(ULP_TX_PKT) | ULP_TXPKT_DEST_V(0));
+	txpkt->len = htonl(DIV_ROUND_UP(sizeof(*req), 16));
+	sc->cmd_more = htonl(ULPTX_CMD_V(ULP_TX_SC_IMM));
+	sc->len = htonl(sizeof(*req) - sizeof(struct work_request_hdr));
+	OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_SET_TCB_FIELD, f->tid));
+	req->reply_ctrl = htons(NO_REPLY_V(no_reply) | REPLY_CHAN_V(0) |
+				QUEUENO_V(0));
+	req->word_cookie = htons(TCB_WORD_V(word) | TCB_COOKIE_V(cookie));
+	req->mask = cpu_to_be64(mask);
+	req->val = cpu_to_be64(val);
+	sc = (struct ulptx_idata *)(req + 1);
+	sc->cmd_more = htonl(ULPTX_CMD_V(ULP_TX_SC_NOOP));
+	sc->len = htonl(0);
+}
+
+static int configure_filter_smac(struct adapter *adap, struct filter_entry *f)
+{
+	int err;
+
+	/* do a set-tcb for smac-sel and CWR bit.. */
+	err = set_tcb_field(adap, f, f->tid, TCB_SMAC_SEL_W,
+			    TCB_SMAC_SEL_V(TCB_SMAC_SEL_M),
+			    TCB_SMAC_SEL_V(f->smt->idx), 1);
+	if (err)
+		goto smac_err;
+
+	err = set_tcb_tflag(adap, f, f->tid, TF_CCTRL_CWR_S, 1, 1);
+	if (!err)
+		return 0;
+
+smac_err:
+	dev_err(adap->pdev_dev, "filter %u smac config failed with error %u\n",
+		f->tid, err);
+	return err;
+}
+
+static void set_nat_params(struct adapter *adap, struct filter_entry *f,
+			   unsigned int tid, bool dip, bool sip, bool dp,
+			   bool sp)
+{
+	u8 *nat_lp = (u8 *)&f->fs.nat_lport;
+	u8 *nat_fp = (u8 *)&f->fs.nat_fport;
+
+	if (dip) {
+		if (f->fs.type) {
+			set_tcb_field(adap, f, tid, TCB_SND_UNA_RAW_W,
+				      WORD_MASK, f->fs.nat_lip[15] |
+				      f->fs.nat_lip[14] << 8 |
+				      f->fs.nat_lip[13] << 16 |
+				      (u64)f->fs.nat_lip[12] << 24, 1);
+
+			set_tcb_field(adap, f, tid, TCB_SND_UNA_RAW_W + 1,
+				      WORD_MASK, f->fs.nat_lip[11] |
+				      f->fs.nat_lip[10] << 8 |
+				      f->fs.nat_lip[9] << 16 |
+				      (u64)f->fs.nat_lip[8] << 24, 1);
+
+			set_tcb_field(adap, f, tid, TCB_SND_UNA_RAW_W + 2,
+				      WORD_MASK, f->fs.nat_lip[7] |
+				      f->fs.nat_lip[6] << 8 |
+				      f->fs.nat_lip[5] << 16 |
+				      (u64)f->fs.nat_lip[4] << 24, 1);
+
+			set_tcb_field(adap, f, tid, TCB_SND_UNA_RAW_W + 3,
+				      WORD_MASK, f->fs.nat_lip[3] |
+				      f->fs.nat_lip[2] << 8 |
+				      f->fs.nat_lip[1] << 16 |
+				      (u64)f->fs.nat_lip[0] << 24, 1);
+		} else {
+			set_tcb_field(adap, f, tid, TCB_RX_FRAG3_LEN_RAW_W,
+				      WORD_MASK, f->fs.nat_lip[3] |
+				      f->fs.nat_lip[2] << 8 |
+				      f->fs.nat_lip[1] << 16 |
+				      (u64)f->fs.nat_lip[0] << 24, 1);
+		}
+	}
+
+	if (sip) {
+		if (f->fs.type) {
+			set_tcb_field(adap, f, tid, TCB_RX_FRAG2_PTR_RAW_W,
+				      WORD_MASK, f->fs.nat_fip[15] |
+				      f->fs.nat_fip[14] << 8 |
+				      f->fs.nat_fip[13] << 16 |
+				      (u64)f->fs.nat_fip[12] << 24, 1);
+
+			set_tcb_field(adap, f, tid, TCB_RX_FRAG2_PTR_RAW_W + 1,
+				      WORD_MASK, f->fs.nat_fip[11] |
+				      f->fs.nat_fip[10] << 8 |
+				      f->fs.nat_fip[9] << 16 |
+				      (u64)f->fs.nat_fip[8] << 24, 1);
+
+			set_tcb_field(adap, f, tid, TCB_RX_FRAG2_PTR_RAW_W + 2,
+				      WORD_MASK, f->fs.nat_fip[7] |
+				      f->fs.nat_fip[6] << 8 |
+				      f->fs.nat_fip[5] << 16 |
+				      (u64)f->fs.nat_fip[4] << 24, 1);
+
+			set_tcb_field(adap, f, tid, TCB_RX_FRAG2_PTR_RAW_W + 3,
+				      WORD_MASK, f->fs.nat_fip[3] |
+				      f->fs.nat_fip[2] << 8 |
+				      f->fs.nat_fip[1] << 16 |
+				      (u64)f->fs.nat_fip[0] << 24, 1);
+
+		} else {
+			set_tcb_field(adap, f, tid,
+				      TCB_RX_FRAG3_START_IDX_OFFSET_RAW_W,
+				      WORD_MASK, f->fs.nat_fip[3] |
+				      f->fs.nat_fip[2] << 8 |
+				      f->fs.nat_fip[1] << 16 |
+				      (u64)f->fs.nat_fip[0] << 24, 1);
+		}
+	}
+
+	set_tcb_field(adap, f, tid, TCB_PDU_HDR_LEN_W, WORD_MASK,
+		      (dp ? (nat_lp[1] | nat_lp[0] << 8) : 0) |
+		      (sp ? (nat_fp[1] << 16 | (u64)nat_fp[0] << 24) : 0),
+		      1);
+}
+
+/* Validate filter spec against configuration done on the card. */
+static int validate_filter(struct net_device *dev,
+			   struct ch_filter_specification *fs)
+{
+	struct adapter *adapter = netdev2adap(dev);
+	u32 fconf, iconf;
+
+	/* Check for unconfigured fields being used. */
+	iconf = adapter->params.tp.ingress_config;
+	fconf = fs->hash ? adapter->params.tp.filter_mask :
+			   adapter->params.tp.vlan_pri_map;
+
+	if (unsupported(fconf, FCOE_F, fs->val.fcoe, fs->mask.fcoe) ||
+	    unsupported(fconf, PORT_F, fs->val.iport, fs->mask.iport) ||
+	    unsupported(fconf, TOS_F, fs->val.tos, fs->mask.tos) ||
+	    unsupported(fconf, ETHERTYPE_F, fs->val.ethtype,
+			fs->mask.ethtype) ||
+	    unsupported(fconf, MACMATCH_F, fs->val.macidx, fs->mask.macidx) ||
+	    unsupported(fconf, MPSHITTYPE_F, fs->val.matchtype,
+			fs->mask.matchtype) ||
+	    unsupported(fconf, FRAGMENTATION_F, fs->val.frag, fs->mask.frag) ||
+	    unsupported(fconf, PROTOCOL_F, fs->val.proto, fs->mask.proto) ||
+	    unsupported(fconf, VNIC_ID_F, fs->val.pfvf_vld,
+			fs->mask.pfvf_vld) ||
+	    unsupported(fconf, VNIC_ID_F, fs->val.ovlan_vld,
+			fs->mask.ovlan_vld) ||
+	    unsupported(fconf, VNIC_ID_F, fs->val.encap_vld,
+			fs->mask.encap_vld) ||
+	    unsupported(fconf, VLAN_F, fs->val.ivlan_vld, fs->mask.ivlan_vld))
+		return -EOPNOTSUPP;
+
+	/* T4 inconveniently uses the same FT_VNIC_ID_W bits for both the Outer
+	 * VLAN Tag and PF/VF/VFvld fields based on VNIC_F being set
+	 * in TP_INGRESS_CONFIG.  Hense the somewhat crazy checks
+	 * below.  Additionally, since the T4 firmware interface also
+	 * carries that overlap, we need to translate any PF/VF
+	 * specification into that internal format below.
+	 */
+	if ((is_field_set(fs->val.pfvf_vld, fs->mask.pfvf_vld) &&
+	     is_field_set(fs->val.ovlan_vld, fs->mask.ovlan_vld)) ||
+	    (is_field_set(fs->val.pfvf_vld, fs->mask.pfvf_vld) &&
+	     is_field_set(fs->val.encap_vld, fs->mask.encap_vld)) ||
+	    (is_field_set(fs->val.ovlan_vld, fs->mask.ovlan_vld) &&
+	     is_field_set(fs->val.encap_vld, fs->mask.encap_vld)))
+		return -EOPNOTSUPP;
+	if (unsupported(iconf, VNIC_F, fs->val.pfvf_vld, fs->mask.pfvf_vld) ||
+	    (is_field_set(fs->val.ovlan_vld, fs->mask.ovlan_vld) &&
+	     (iconf & VNIC_F)))
+		return -EOPNOTSUPP;
+	if (fs->val.pf > 0x7 || fs->val.vf > 0x7f)
+		return -ERANGE;
+	fs->mask.pf &= 0x7;
+	fs->mask.vf &= 0x7f;
+
+	/* If the user is requesting that the filter action loop
+	 * matching packets back out one of our ports, make sure that
+	 * the egress port is in range.
+	 */
+	if (fs->action == FILTER_SWITCH &&
+	    fs->eport >= adapter->params.nports)
+		return -ERANGE;
+
+	/* Don't allow various trivially obvious bogus out-of-range values... */
+	if (fs->val.iport >= adapter->params.nports)
+		return -ERANGE;
+
+	/* T4 doesn't support removing VLAN Tags for loop back filters. */
+	if (is_t4(adapter->params.chip) &&
+	    fs->action == FILTER_SWITCH &&
+	    (fs->newvlan == VLAN_REMOVE ||
+	     fs->newvlan == VLAN_REWRITE))
+		return -EOPNOTSUPP;
+
+	if (fs->val.encap_vld &&
+	    CHELSIO_CHIP_VERSION(adapter->params.chip) < CHELSIO_T6)
+		return -EOPNOTSUPP;
+	return 0;
+}
+
+static int get_filter_steerq(struct net_device *dev,
+			     struct ch_filter_specification *fs)
+{
+	struct adapter *adapter = netdev2adap(dev);
+	int iq;
+
+	/* If the user has requested steering matching Ingress Packets
+	 * to a specific Queue Set, we need to make sure it's in range
+	 * for the port and map that into the Absolute Queue ID of the
+	 * Queue Set's Response Queue.
+	 */
+	if (!fs->dirsteer) {
+		if (fs->iq)
+			return -EINVAL;
+		iq = 0;
+	} else {
+		struct port_info *pi = netdev_priv(dev);
+
+		/* If the iq id is greater than the number of qsets,
+		 * then assume it is an absolute qid.
+		 */
+		if (fs->iq < pi->nqsets)
+			iq = adapter->sge.ethrxq[pi->first_qset +
+						 fs->iq].rspq.abs_id;
+		else
+			iq = fs->iq;
+	}
+
+	return iq;
+}
+
+static int get_filter_count(struct adapter *adapter, unsigned int fidx,
+			    u64 *pkts, u64 *bytes, bool hash)
+{
+	unsigned int tcb_base, tcbaddr;
+	unsigned int word_offset;
+	struct filter_entry *f;
+	__be64 be64_byte_count;
+	int ret;
+
+	tcb_base = t4_read_reg(adapter, TP_CMM_TCB_BASE_A);
+	if (is_hashfilter(adapter) && hash) {
+		if (tid_out_of_range(&adapter->tids, fidx))
+			return -E2BIG;
+		f = adapter->tids.tid_tab[fidx - adapter->tids.tid_base];
+		if (!f)
+			return -EINVAL;
+	} else {
+		if ((fidx != (adapter->tids.nftids + adapter->tids.nsftids +
+			      adapter->tids.nhpftids - 1)) &&
+		    fidx >= (adapter->tids.nftids + adapter->tids.nhpftids))
+			return -E2BIG;
+
+		if (fidx < adapter->tids.nhpftids)
+			f = &adapter->tids.hpftid_tab[fidx];
+		else
+			f = &adapter->tids.ftid_tab[fidx -
+						    adapter->tids.nhpftids];
+		if (!f->valid)
+			return -EINVAL;
+	}
+	tcbaddr = tcb_base + f->tid * TCB_SIZE;
+
+	spin_lock(&adapter->win0_lock);
+	if (is_t4(adapter->params.chip)) {
+		__be64 be64_count;
+
+		/* T4 doesn't maintain byte counts in hw */
+		*bytes = 0;
+
+		/* Get pkts */
+		word_offset = 4;
+		ret = t4_memory_rw(adapter, MEMWIN_NIC, MEM_EDC0,
+				   tcbaddr + (word_offset * sizeof(__be32)),
+				   sizeof(be64_count),
+				   (__be32 *)&be64_count,
+				   T4_MEMORY_READ);
+		if (ret < 0)
+			goto out;
+		*pkts = be64_to_cpu(be64_count);
+	} else {
+		__be32 be32_count;
+
+		/* Get bytes */
+		word_offset = 4;
+		ret = t4_memory_rw(adapter, MEMWIN_NIC, MEM_EDC0,
+				   tcbaddr + (word_offset * sizeof(__be32)),
+				   sizeof(be64_byte_count),
+				   &be64_byte_count,
+				   T4_MEMORY_READ);
+		if (ret < 0)
+			goto out;
+		*bytes = be64_to_cpu(be64_byte_count);
+
+		/* Get pkts */
+		word_offset = 6;
+		ret = t4_memory_rw(adapter, MEMWIN_NIC, MEM_EDC0,
+				   tcbaddr + (word_offset * sizeof(__be32)),
+				   sizeof(be32_count),
+				   &be32_count,
+				   T4_MEMORY_READ);
+		if (ret < 0)
+			goto out;
+		*pkts = (u64)be32_to_cpu(be32_count);
+	}
+
+out:
+	spin_unlock(&adapter->win0_lock);
+	return ret;
+}
+
+int cxgb4_get_filter_counters(struct net_device *dev, unsigned int fidx,
+			      u64 *hitcnt, u64 *bytecnt, bool hash)
+{
+	struct adapter *adapter = netdev2adap(dev);
+
+	return get_filter_count(adapter, fidx, hitcnt, bytecnt, hash);
+}
+
+static bool cxgb4_filter_prio_in_range(struct tid_info *t, u32 idx, u8 nslots,
+				       u32 prio)
+{
+	struct filter_entry *prev_tab, *next_tab, *prev_fe, *next_fe;
+	u32 prev_ftid, next_ftid;
+
+	/* Only insert the rule if both of the following conditions
+	 * are met:
+	 * 1. The immediate previous rule has priority <= @prio.
+	 * 2. The immediate next rule has priority >= @prio.
+	 */
+
+	/* High Priority (HPFILTER) region always has higher priority
+	 * than normal FILTER region. So, all rules in HPFILTER region
+	 * must have prio value <= rules in normal FILTER region.
+	 */
+	if (idx < t->nhpftids) {
+		/* Don't insert if there's a rule already present at @idx
+		 * in HPFILTER region.
+		 */
+		if (test_bit(idx, t->hpftid_bmap))
+			return false;
+
+		next_tab = t->hpftid_tab;
+		next_ftid = find_next_bit(t->hpftid_bmap, t->nhpftids, idx);
+		if (next_ftid >= t->nhpftids) {
+			/* No next entry found in HPFILTER region.
+			 * See if there's any next entry in normal
+			 * FILTER region.
+			 */
+			next_ftid = find_first_bit(t->ftid_bmap, t->nftids);
+			if (next_ftid >= t->nftids)
+				next_ftid = idx;
+			else
+				next_tab = t->ftid_tab;
+		}
+
+		/* Search for the closest previous filter entry in HPFILTER
+		 * region. No need to search in normal FILTER region because
+		 * there can never be any entry in normal FILTER region whose
+		 * prio value is < last entry in HPFILTER region.
+		 */
+		prev_ftid = find_last_bit(t->hpftid_bmap, idx);
+		if (prev_ftid >= idx)
+			prev_ftid = idx;
+
+		prev_tab = t->hpftid_tab;
+	} else {
+		idx -= t->nhpftids;
+
+		/* Don't insert if there's a rule already present at @idx
+		 * in normal FILTER region.
+		 */
+		if (test_bit(idx, t->ftid_bmap))
+			return false;
+
+		prev_tab = t->ftid_tab;
+		prev_ftid = find_last_bit(t->ftid_bmap, idx);
+		if (prev_ftid >= idx) {
+			/* No previous entry found in normal FILTER
+			 * region. See if there's any previous entry
+			 * in HPFILTER region.
+			 */
+			prev_ftid = find_last_bit(t->hpftid_bmap, t->nhpftids);
+			if (prev_ftid >= t->nhpftids)
+				prev_ftid = idx;
+			else
+				prev_tab = t->hpftid_tab;
+		}
+
+		/* Search for the closest next filter entry in normal
+		 * FILTER region. No need to search in HPFILTER region
+		 * because there can never be any entry in HPFILTER
+		 * region whose prio value is > first entry in normal
+		 * FILTER region.
+		 */
+		next_ftid = find_next_bit(t->ftid_bmap, t->nftids, idx);
+		if (next_ftid >= t->nftids)
+			next_ftid = idx;
+
+		next_tab = t->ftid_tab;
+	}
+
+	next_fe = &next_tab[next_ftid];
+
+	/* See if the filter entry belongs to an IPv6 rule, which
+	 * occupy 4 slots on T5 and 2 slots on T6. Adjust the
+	 * reference to the previously inserted filter entry
+	 * accordingly.
+	 */
+	prev_fe = &prev_tab[prev_ftid & ~(nslots - 1)];
+	if (!prev_fe->fs.type)
+		prev_fe = &prev_tab[prev_ftid];
+
+	if ((prev_fe->valid && prev_fe->fs.tc_prio > prio) ||
+	    (next_fe->valid && next_fe->fs.tc_prio < prio))
+		return false;
+
+	return true;
+}
+
+int cxgb4_get_free_ftid(struct net_device *dev, u8 family, bool hash_en,
+			u32 tc_prio)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct tid_info *t = &adap->tids;
+	u32 bmap_ftid, max_ftid;
+	struct filter_entry *f;
+	unsigned long *bmap;
+	bool found = false;
+	u8 i, cnt, n;
+	int ftid = 0;
+
+	/* IPv4 occupy 1 slot. IPv6 occupy 2 slots on T6 and 4 slots
+	 * on T5.
+	 */
+	n = 1;
+	if (family == PF_INET6) {
+		n++;
+		if (CHELSIO_CHIP_VERSION(adap->params.chip) < CHELSIO_T6)
+			n += 2;
+	}
+
+	/* There are 3 filter regions available in hardware in
+	 * following order of priority:
+	 *
+	 * 1. High Priority (HPFILTER) region (Highest Priority).
+	 * 2. HASH region.
+	 * 3. Normal FILTER region (Lowest Priority).
+	 *
+	 * Entries in HPFILTER and normal FILTER region have index
+	 * 0 as the highest priority and the rules will be scanned
+	 * in ascending order until either a rule hits or end of
+	 * the region is reached.
+	 *
+	 * All HASH region entries have same priority. The set of
+	 * fields to match in headers are pre-determined. The same
+	 * set of header match fields must be compulsorily specified
+	 * in all the rules wanting to get inserted in HASH region.
+	 * Hence, HASH region is an exact-match region. A HASH is
+	 * generated for a rule based on the values in the
+	 * pre-determined set of header match fields. The generated
+	 * HASH serves as an index into the HASH region. There can
+	 * never be 2 rules having the same HASH. Hardware will
+	 * compute a HASH for every incoming packet based on the
+	 * values in the pre-determined set of header match fields
+	 * and uses it as an index to check if there's a rule
+	 * inserted in the HASH region at the specified index. If
+	 * there's a rule inserted, then it's considered as a filter
+	 * hit. Otherwise, it's a filter miss and normal FILTER region
+	 * is scanned afterwards.
+	 */
+
+	spin_lock_bh(&t->ftid_lock);
+
+	ftid = (tc_prio <= t->nhpftids) ? 0 : t->nhpftids;
+	max_ftid = t->nftids + t->nhpftids;
+	while (ftid < max_ftid) {
+		if (ftid < t->nhpftids) {
+			/* If the new rule wants to get inserted into
+			 * HPFILTER region, but its prio is greater
+			 * than the rule with the highest prio in HASH
+			 * region, or if there's not enough slots
+			 * available in HPFILTER region, then skip
+			 * trying to insert this rule into HPFILTER
+			 * region and directly go to the next region.
+			 */
+			if ((t->tc_hash_tids_max_prio &&
+			     tc_prio > t->tc_hash_tids_max_prio) ||
+			     (ftid + n) > t->nhpftids) {
+				ftid = t->nhpftids;
+				continue;
+			}
+
+			bmap = t->hpftid_bmap;
+			bmap_ftid = ftid;
+		} else if (hash_en) {
+			/* Ensure priority is >= last rule in HPFILTER
+			 * region.
+			 */
+			ftid = find_last_bit(t->hpftid_bmap, t->nhpftids);
+			if (ftid < t->nhpftids) {
+				f = &t->hpftid_tab[ftid];
+				if (f->valid && tc_prio < f->fs.tc_prio)
+					break;
+			}
+
+			/* Ensure priority is <= first rule in normal
+			 * FILTER region.
+			 */
+			ftid = find_first_bit(t->ftid_bmap, t->nftids);
+			if (ftid < t->nftids) {
+				f = &t->ftid_tab[ftid];
+				if (f->valid && tc_prio > f->fs.tc_prio)
+					break;
+			}
+
+			found = true;
+			ftid = t->nhpftids;
+			goto out_unlock;
+		} else {
+			/* If the new rule wants to get inserted into
+			 * normal FILTER region, but its prio is less
+			 * than the rule with the highest prio in HASH
+			 * region, then reject the rule.
+			 */
+			if (t->tc_hash_tids_max_prio &&
+			    tc_prio < t->tc_hash_tids_max_prio)
+				break;
+
+			if (ftid + n > max_ftid)
+				break;
+
+			bmap = t->ftid_bmap;
+			bmap_ftid = ftid - t->nhpftids;
+		}
+
+		cnt = 0;
+		for (i = 0; i < n; i++) {
+			if (test_bit(bmap_ftid + i, bmap))
+				break;
+			cnt++;
+		}
+
+		if (cnt == n) {
+			/* Ensure the new rule's prio doesn't conflict
+			 * with existing rules.
+			 */
+			if (cxgb4_filter_prio_in_range(t, ftid, n,
+						       tc_prio)) {
+				ftid &= ~(n - 1);
+				found = true;
+				break;
+			}
+		}
+
+		ftid += n;
+	}
+
+out_unlock:
+	spin_unlock_bh(&t->ftid_lock);
+	return found ? ftid : -ENOMEM;
+}
+
+static int cxgb4_set_ftid(struct tid_info *t, int fidx, int family,
+			  unsigned int chip_ver)
+{
+	spin_lock_bh(&t->ftid_lock);
+
+	if (test_bit(fidx, t->ftid_bmap)) {
+		spin_unlock_bh(&t->ftid_lock);
+		return -EBUSY;
+	}
+
+	if (family == PF_INET) {
+		__set_bit(fidx, t->ftid_bmap);
+	} else {
+		if (chip_ver < CHELSIO_T6)
+			bitmap_allocate_region(t->ftid_bmap, fidx, 2);
+		else
+			bitmap_allocate_region(t->ftid_bmap, fidx, 1);
+	}
+
+	spin_unlock_bh(&t->ftid_lock);
+	return 0;
+}
+
+static int cxgb4_set_hpftid(struct tid_info *t, int fidx, int family)
+{
+	spin_lock_bh(&t->ftid_lock);
+
+	if (test_bit(fidx, t->hpftid_bmap)) {
+		spin_unlock_bh(&t->ftid_lock);
+		return -EBUSY;
+	}
+
+	if (family == PF_INET)
+		__set_bit(fidx, t->hpftid_bmap);
+	else
+		bitmap_allocate_region(t->hpftid_bmap, fidx, 1);
+
+	spin_unlock_bh(&t->ftid_lock);
+	return 0;
+}
+
+static void cxgb4_clear_ftid(struct tid_info *t, int fidx, int family,
+			     unsigned int chip_ver)
+{
+	spin_lock_bh(&t->ftid_lock);
+	if (family == PF_INET) {
+		__clear_bit(fidx, t->ftid_bmap);
+	} else {
+		if (chip_ver < CHELSIO_T6)
+			bitmap_release_region(t->ftid_bmap, fidx, 2);
+		else
+			bitmap_release_region(t->ftid_bmap, fidx, 1);
+	}
+	spin_unlock_bh(&t->ftid_lock);
+}
+
+static void cxgb4_clear_hpftid(struct tid_info *t, int fidx, int family)
+{
+	spin_lock_bh(&t->ftid_lock);
+
+	if (family == PF_INET)
+		__clear_bit(fidx, t->hpftid_bmap);
+	else
+		bitmap_release_region(t->hpftid_bmap, fidx, 1);
+
+	spin_unlock_bh(&t->ftid_lock);
+}
+
+/* Delete the filter at a specified index. */
+static int del_filter_wr(struct adapter *adapter, int fidx)
+{
+	struct fw_filter_wr *fwr;
+	struct filter_entry *f;
+	struct sk_buff *skb;
+	unsigned int len;
+
+	if (fidx < adapter->tids.nhpftids)
+		f = &adapter->tids.hpftid_tab[fidx];
+	else
+		f = &adapter->tids.ftid_tab[fidx - adapter->tids.nhpftids];
+
+	len = sizeof(*fwr);
+
+	skb = alloc_skb(len, GFP_KERNEL);
+	if (!skb)
+		return -ENOMEM;
+
+	fwr = __skb_put(skb, len);
+	t4_mk_filtdelwr(f->tid, fwr, adapter->sge.fw_evtq.abs_id);
+
+	/* Mark the filter as "pending" and ship off the Filter Work Request.
+	 * When we get the Work Request Reply we'll clear the pending status.
+	 */
+	f->pending = 1;
+	t4_mgmt_tx(adapter, skb);
+	return 0;
+}
+
+/* Send a Work Request to write the filter at a specified index.  We construct
+ * a Firmware Filter Work Request to have the work done and put the indicated
+ * filter into "pending" mode which will prevent any further actions against
+ * it till we get a reply from the firmware on the completion status of the
+ * request.
+ */
+int set_filter_wr(struct adapter *adapter, int fidx)
+{
+	struct fw_filter2_wr *fwr;
+	struct filter_entry *f;
+	struct sk_buff *skb;
+
+	if (fidx < adapter->tids.nhpftids)
+		f = &adapter->tids.hpftid_tab[fidx];
+	else
+		f = &adapter->tids.ftid_tab[fidx - adapter->tids.nhpftids];
+
+	skb = alloc_skb(sizeof(*fwr), GFP_KERNEL);
+	if (!skb)
+		return -ENOMEM;
+
+	/* If the new filter requires loopback Destination MAC and/or VLAN
+	 * rewriting then we need to allocate a Layer 2 Table (L2T) entry for
+	 * the filter.
+	 */
+	if (f->fs.newdmac || f->fs.newvlan) {
+		/* allocate L2T entry for new filter */
+		f->l2t = t4_l2t_alloc_switching(adapter, f->fs.vlan,
+						f->fs.eport, f->fs.dmac);
+		if (!f->l2t) {
+			kfree_skb(skb);
+			return -ENOMEM;
+		}
+	}
+
+	/* If the new filter requires loopback Source MAC rewriting then
+	 * we need to allocate a SMT entry for the filter.
+	 */
+	if (f->fs.newsmac) {
+		f->smt = cxgb4_smt_alloc_switching(f->dev, f->fs.smac);
+		if (!f->smt) {
+			if (f->l2t) {
+				cxgb4_l2t_release(f->l2t);
+				f->l2t = NULL;
+			}
+			kfree_skb(skb);
+			return -ENOMEM;
+		}
+	}
+
+	fwr = __skb_put_zero(skb, sizeof(*fwr));
+
+	/* It would be nice to put most of the following in t4_hw.c but most
+	 * of the work is translating the cxgbtool ch_filter_specification
+	 * into the Work Request and the definition of that structure is
+	 * currently in cxgbtool.h which isn't appropriate to pull into the
+	 * common code.  We may eventually try to come up with a more neutral
+	 * filter specification structure but for now it's easiest to simply
+	 * put this fairly direct code in line ...
+	 */
+	if (adapter->params.filter2_wr_support)
+		fwr->op_pkd = htonl(FW_WR_OP_V(FW_FILTER2_WR));
+	else
+		fwr->op_pkd = htonl(FW_WR_OP_V(FW_FILTER_WR));
+	fwr->len16_pkd = htonl(FW_WR_LEN16_V(sizeof(*fwr) / 16));
+	fwr->tid_to_iq =
+		htonl(FW_FILTER_WR_TID_V(f->tid) |
+		      FW_FILTER_WR_RQTYPE_V(f->fs.type) |
+		      FW_FILTER_WR_NOREPLY_V(0) |
+		      FW_FILTER_WR_IQ_V(f->fs.iq));
+	fwr->del_filter_to_l2tix =
+		htonl(FW_FILTER_WR_RPTTID_V(f->fs.rpttid) |
+		      FW_FILTER_WR_DROP_V(f->fs.action == FILTER_DROP) |
+		      FW_FILTER_WR_DIRSTEER_V(f->fs.dirsteer) |
+		      FW_FILTER_WR_MASKHASH_V(f->fs.maskhash) |
+		      FW_FILTER_WR_DIRSTEERHASH_V(f->fs.dirsteerhash) |
+		      FW_FILTER_WR_LPBK_V(f->fs.action == FILTER_SWITCH) |
+		      FW_FILTER_WR_DMAC_V(f->fs.newdmac) |
+		      FW_FILTER_WR_SMAC_V(f->fs.newsmac) |
+		      FW_FILTER_WR_INSVLAN_V(f->fs.newvlan == VLAN_INSERT ||
+					     f->fs.newvlan == VLAN_REWRITE) |
+		      FW_FILTER_WR_RMVLAN_V(f->fs.newvlan == VLAN_REMOVE ||
+					    f->fs.newvlan == VLAN_REWRITE) |
+		      FW_FILTER_WR_HITCNTS_V(f->fs.hitcnts) |
+		      FW_FILTER_WR_TXCHAN_V(f->fs.eport) |
+		      FW_FILTER_WR_PRIO_V(f->fs.prio) |
+		      FW_FILTER_WR_L2TIX_V(f->l2t ? f->l2t->idx : 0));
+	fwr->ethtype = htons(f->fs.val.ethtype);
+	fwr->ethtypem = htons(f->fs.mask.ethtype);
+	fwr->frag_to_ovlan_vldm =
+		(FW_FILTER_WR_FRAG_V(f->fs.val.frag) |
+		 FW_FILTER_WR_FRAGM_V(f->fs.mask.frag) |
+		 FW_FILTER_WR_IVLAN_VLD_V(f->fs.val.ivlan_vld) |
+		 FW_FILTER_WR_OVLAN_VLD_V(f->fs.val.ovlan_vld) |
+		 FW_FILTER_WR_IVLAN_VLDM_V(f->fs.mask.ivlan_vld) |
+		 FW_FILTER_WR_OVLAN_VLDM_V(f->fs.mask.ovlan_vld));
+	if (f->fs.newsmac)
+		fwr->smac_sel = f->smt->idx;
+	fwr->rx_chan_rx_rpl_iq =
+		htons(FW_FILTER_WR_RX_CHAN_V(0) |
+		      FW_FILTER_WR_RX_RPL_IQ_V(adapter->sge.fw_evtq.abs_id));
+	fwr->maci_to_matchtypem =
+		htonl(FW_FILTER_WR_MACI_V(f->fs.val.macidx) |
+		      FW_FILTER_WR_MACIM_V(f->fs.mask.macidx) |
+		      FW_FILTER_WR_FCOE_V(f->fs.val.fcoe) |
+		      FW_FILTER_WR_FCOEM_V(f->fs.mask.fcoe) |
+		      FW_FILTER_WR_PORT_V(f->fs.val.iport) |
+		      FW_FILTER_WR_PORTM_V(f->fs.mask.iport) |
+		      FW_FILTER_WR_MATCHTYPE_V(f->fs.val.matchtype) |
+		      FW_FILTER_WR_MATCHTYPEM_V(f->fs.mask.matchtype));
+	fwr->ptcl = f->fs.val.proto;
+	fwr->ptclm = f->fs.mask.proto;
+	fwr->ttyp = f->fs.val.tos;
+	fwr->ttypm = f->fs.mask.tos;
+	fwr->ivlan = htons(f->fs.val.ivlan);
+	fwr->ivlanm = htons(f->fs.mask.ivlan);
+	fwr->ovlan = htons(f->fs.val.ovlan);
+	fwr->ovlanm = htons(f->fs.mask.ovlan);
+	memcpy(fwr->lip, f->fs.val.lip, sizeof(fwr->lip));
+	memcpy(fwr->lipm, f->fs.mask.lip, sizeof(fwr->lipm));
+	memcpy(fwr->fip, f->fs.val.fip, sizeof(fwr->fip));
+	memcpy(fwr->fipm, f->fs.mask.fip, sizeof(fwr->fipm));
+	fwr->lp = htons(f->fs.val.lport);
+	fwr->lpm = htons(f->fs.mask.lport);
+	fwr->fp = htons(f->fs.val.fport);
+	fwr->fpm = htons(f->fs.mask.fport);
+
+	if (adapter->params.filter2_wr_support) {
+		u8 *nat_lp = (u8 *)&f->fs.nat_lport;
+		u8 *nat_fp = (u8 *)&f->fs.nat_fport;
+
+		fwr->natmode_to_ulp_type =
+			FW_FILTER2_WR_ULP_TYPE_V(f->fs.nat_mode ?
+						 ULP_MODE_TCPDDP :
+						 ULP_MODE_NONE) |
+			FW_FILTER2_WR_NATMODE_V(f->fs.nat_mode);
+		memcpy(fwr->newlip, f->fs.nat_lip, sizeof(fwr->newlip));
+		memcpy(fwr->newfip, f->fs.nat_fip, sizeof(fwr->newfip));
+		fwr->newlport = htons(nat_lp[1] | nat_lp[0] << 8);
+		fwr->newfport = htons(nat_fp[1] | nat_fp[0] << 8);
+	}
+
+	/* Mark the filter as "pending" and ship off the Filter Work Request.
+	 * When we get the Work Request Reply we'll clear the pending status.
+	 */
+	f->pending = 1;
+	set_wr_txq(skb, CPL_PRIORITY_CONTROL, f->fs.val.iport & 0x3);
+	t4_ofld_send(adapter, skb);
+	return 0;
+}
+
+/* Return an error number if the indicated filter isn't writable ... */
+int writable_filter(struct filter_entry *f)
+{
+	if (f->locked)
+		return -EPERM;
+	if (f->pending)
+		return -EBUSY;
+
+	return 0;
+}
+
+/* Delete the filter at the specified index (if valid).  The checks for all
+ * the common problems with doing this like the filter being locked, currently
+ * pending in another operation, etc.
+ */
+int delete_filter(struct adapter *adapter, unsigned int fidx)
+{
+	struct filter_entry *f;
+	int ret;
+
+	if (fidx >= adapter->tids.nftids + adapter->tids.nsftids +
+		    adapter->tids.nhpftids)
+		return -EINVAL;
+
+	if (fidx < adapter->tids.nhpftids)
+		f = &adapter->tids.hpftid_tab[fidx];
+	else
+		f = &adapter->tids.ftid_tab[fidx - adapter->tids.nhpftids];
+	ret = writable_filter(f);
+	if (ret)
+		return ret;
+	if (f->valid)
+		return del_filter_wr(adapter, fidx);
+
+	return 0;
+}
+
+/* Clear a filter and release any of its resources that we own.  This also
+ * clears the filter's "pending" status.
+ */
+void clear_filter(struct adapter *adap, struct filter_entry *f)
+{
+	struct port_info *pi = netdev_priv(f->dev);
+
+	/* If the new or old filter have loopback rewriting rules then we'll
+	 * need to free any existing L2T, SMT, CLIP entries of filter
+	 * rule.
+	 */
+	if (f->l2t)
+		cxgb4_l2t_release(f->l2t);
+
+	if (f->smt)
+		cxgb4_smt_release(f->smt);
+
+	if (f->fs.val.encap_vld && f->fs.val.ovlan_vld)
+		t4_free_encap_mac_filt(adap, pi->viid,
+				       f->fs.val.ovlan & 0x1ff, 0);
+
+	if ((f->fs.hash || is_t6(adap->params.chip)) && f->fs.type)
+		cxgb4_clip_release(f->dev, (const u32 *)&f->fs.val.lip, 1);
+
+	/* The zeroing of the filter rule below clears the filter valid,
+	 * pending, locked flags, l2t pointer, etc. so it's all we need for
+	 * this operation.
+	 */
+	memset(f, 0, sizeof(*f));
+}
+
+void clear_all_filters(struct adapter *adapter)
+{
+	struct net_device *dev = adapter->port[0];
+	unsigned int i;
+
+	if (adapter->tids.hpftid_tab) {
+		struct filter_entry *f = &adapter->tids.hpftid_tab[0];
+
+		for (i = 0; i < adapter->tids.nhpftids; i++, f++)
+			if (f->valid || f->pending)
+				cxgb4_del_filter(dev, i, &f->fs);
+	}
+
+	if (adapter->tids.ftid_tab) {
+		struct filter_entry *f = &adapter->tids.ftid_tab[0];
+		unsigned int max_ftid = adapter->tids.nftids +
+					adapter->tids.nsftids +
+					adapter->tids.nhpftids;
+
+		/* Clear all TCAM filters */
+		for (i = adapter->tids.nhpftids; i < max_ftid; i++, f++)
+			if (f->valid || f->pending)
+				cxgb4_del_filter(dev, i, &f->fs);
+	}
+
+	/* Clear all hash filters */
+	if (is_hashfilter(adapter) && adapter->tids.tid_tab) {
+		struct filter_entry *f;
+		unsigned int sb;
+
+		for (i = adapter->tids.hash_base;
+		     i <= adapter->tids.ntids; i++) {
+			f = (struct filter_entry *)
+				adapter->tids.tid_tab[i];
+
+			if (f && (f->valid || f->pending))
+				cxgb4_del_filter(dev, f->tid, &f->fs);
+		}
+
+		sb = adapter->tids.stid_base;
+		for (i = 0; i < sb; i++) {
+			f = (struct filter_entry *)adapter->tids.tid_tab[i];
+
+			if (f && (f->valid || f->pending))
+				cxgb4_del_filter(dev, f->tid, &f->fs);
+		}
+	}
+}
+
+/* Fill up default masks for set match fields. */
+static void fill_default_mask(struct ch_filter_specification *fs)
+{
+	unsigned int lip = 0, lip_mask = 0;
+	unsigned int fip = 0, fip_mask = 0;
+	unsigned int i;
+
+	if (fs->val.iport && !fs->mask.iport)
+		fs->mask.iport |= ~0;
+	if (fs->val.fcoe && !fs->mask.fcoe)
+		fs->mask.fcoe |= ~0;
+	if (fs->val.matchtype && !fs->mask.matchtype)
+		fs->mask.matchtype |= ~0;
+	if (fs->val.macidx && !fs->mask.macidx)
+		fs->mask.macidx |= ~0;
+	if (fs->val.ethtype && !fs->mask.ethtype)
+		fs->mask.ethtype |= ~0;
+	if (fs->val.ivlan && !fs->mask.ivlan)
+		fs->mask.ivlan |= ~0;
+	if (fs->val.ovlan && !fs->mask.ovlan)
+		fs->mask.ovlan |= ~0;
+	if (fs->val.frag && !fs->mask.frag)
+		fs->mask.frag |= ~0;
+	if (fs->val.tos && !fs->mask.tos)
+		fs->mask.tos |= ~0;
+	if (fs->val.proto && !fs->mask.proto)
+		fs->mask.proto |= ~0;
+	if (fs->val.pfvf_vld && !fs->mask.pfvf_vld)
+		fs->mask.pfvf_vld |= ~0;
+	if (fs->val.pf && !fs->mask.pf)
+		fs->mask.pf |= ~0;
+	if (fs->val.vf && !fs->mask.vf)
+		fs->mask.vf |= ~0;
+
+	for (i = 0; i < ARRAY_SIZE(fs->val.lip); i++) {
+		lip |= fs->val.lip[i];
+		lip_mask |= fs->mask.lip[i];
+		fip |= fs->val.fip[i];
+		fip_mask |= fs->mask.fip[i];
+	}
+
+	if (lip && !lip_mask)
+		memset(fs->mask.lip, ~0, sizeof(fs->mask.lip));
+
+	if (fip && !fip_mask)
+		memset(fs->mask.fip, ~0, sizeof(fs->mask.lip));
+
+	if (fs->val.lport && !fs->mask.lport)
+		fs->mask.lport = ~0;
+	if (fs->val.fport && !fs->mask.fport)
+		fs->mask.fport = ~0;
+}
+
+static bool is_addr_all_mask(u8 *ipmask, int family)
+{
+	if (family == AF_INET) {
+		struct in_addr *addr;
+
+		addr = (struct in_addr *)ipmask;
+		if (addr->s_addr == htonl(0xffffffff))
+			return true;
+	} else if (family == AF_INET6) {
+		struct in6_addr *addr6;
+
+		addr6 = (struct in6_addr *)ipmask;
+		if (addr6->s6_addr32[0] == htonl(0xffffffff) &&
+		    addr6->s6_addr32[1] == htonl(0xffffffff) &&
+		    addr6->s6_addr32[2] == htonl(0xffffffff) &&
+		    addr6->s6_addr32[3] == htonl(0xffffffff))
+			return true;
+	}
+	return false;
+}
+
+static bool is_inaddr_any(u8 *ip, int family)
+{
+	int addr_type;
+
+	if (family == AF_INET) {
+		struct in_addr *addr;
+
+		addr = (struct in_addr *)ip;
+		if (addr->s_addr == htonl(INADDR_ANY))
+			return true;
+	} else if (family == AF_INET6) {
+		struct in6_addr *addr6;
+
+		addr6 = (struct in6_addr *)ip;
+		addr_type = ipv6_addr_type((const struct in6_addr *)
+					   &addr6);
+		if (addr_type == IPV6_ADDR_ANY)
+			return true;
+	}
+	return false;
+}
+
+bool is_filter_exact_match(struct adapter *adap,
+			   struct ch_filter_specification *fs)
+{
+	struct tp_params *tp = &adap->params.tp;
+	u64 hash_filter_mask = tp->hash_filter_mask;
+	u64 ntuple_mask = 0;
+
+	if (!is_hashfilter(adap))
+		return false;
+
+	if ((atomic_read(&adap->tids.hash_tids_in_use) +
+	     atomic_read(&adap->tids.tids_in_use)) >=
+	    (adap->tids.nhash + (adap->tids.stid_base - adap->tids.tid_base)))
+		return false;
+
+	 /* Keep tunnel VNI match disabled for hash-filters for now */
+	if (fs->mask.encap_vld)
+		return false;
+
+	if (fs->type) {
+		if (is_inaddr_any(fs->val.fip, AF_INET6) ||
+		    !is_addr_all_mask(fs->mask.fip, AF_INET6))
+			return false;
+
+		if (is_inaddr_any(fs->val.lip, AF_INET6) ||
+		    !is_addr_all_mask(fs->mask.lip, AF_INET6))
+			return false;
+	} else {
+		if (is_inaddr_any(fs->val.fip, AF_INET) ||
+		    !is_addr_all_mask(fs->mask.fip, AF_INET))
+			return false;
+
+		if (is_inaddr_any(fs->val.lip, AF_INET) ||
+		    !is_addr_all_mask(fs->mask.lip, AF_INET))
+			return false;
+	}
+
+	if (!fs->val.lport || fs->mask.lport != 0xffff)
+		return false;
+
+	if (!fs->val.fport || fs->mask.fport != 0xffff)
+		return false;
+
+	/* calculate tuple mask and compare with mask configured in hw */
+	if (tp->fcoe_shift >= 0)
+		ntuple_mask |= (u64)fs->mask.fcoe << tp->fcoe_shift;
+
+	if (tp->port_shift >= 0)
+		ntuple_mask |= (u64)fs->mask.iport << tp->port_shift;
+
+	if (tp->vnic_shift >= 0) {
+		if ((adap->params.tp.ingress_config & VNIC_F))
+			ntuple_mask |= (u64)fs->mask.pfvf_vld << tp->vnic_shift;
+		else
+			ntuple_mask |= (u64)fs->mask.ovlan_vld <<
+				tp->vnic_shift;
+	}
+
+	if (tp->vlan_shift >= 0)
+		ntuple_mask |= (u64)fs->mask.ivlan << tp->vlan_shift;
+
+	if (tp->tos_shift >= 0)
+		ntuple_mask |= (u64)fs->mask.tos << tp->tos_shift;
+
+	if (tp->protocol_shift >= 0)
+		ntuple_mask |= (u64)fs->mask.proto << tp->protocol_shift;
+
+	if (tp->ethertype_shift >= 0)
+		ntuple_mask |= (u64)fs->mask.ethtype << tp->ethertype_shift;
+
+	if (tp->macmatch_shift >= 0)
+		ntuple_mask |= (u64)fs->mask.macidx << tp->macmatch_shift;
+
+	if (tp->matchtype_shift >= 0)
+		ntuple_mask |= (u64)fs->mask.matchtype << tp->matchtype_shift;
+
+	if (tp->frag_shift >= 0)
+		ntuple_mask |= (u64)fs->mask.frag << tp->frag_shift;
+
+	if (ntuple_mask != hash_filter_mask)
+		return false;
+
+	return true;
+}
+
+static u64 hash_filter_ntuple(struct ch_filter_specification *fs,
+			      struct net_device *dev)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct tp_params *tp = &adap->params.tp;
+	u64 ntuple = 0;
+
+	/* Initialize each of the fields which we care about which are present
+	 * in the Compressed Filter Tuple.
+	 */
+	if (tp->vlan_shift >= 0 && fs->mask.ivlan)
+		ntuple |= (FT_VLAN_VLD_F | fs->val.ivlan) << tp->vlan_shift;
+
+	if (tp->port_shift >= 0 && fs->mask.iport)
+		ntuple |= (u64)fs->val.iport << tp->port_shift;
+
+	if (tp->protocol_shift >= 0) {
+		if (!fs->val.proto)
+			ntuple |= (u64)IPPROTO_TCP << tp->protocol_shift;
+		else
+			ntuple |= (u64)fs->val.proto << tp->protocol_shift;
+	}
+
+	if (tp->tos_shift >= 0 && fs->mask.tos)
+		ntuple |= (u64)(fs->val.tos) << tp->tos_shift;
+
+	if (tp->vnic_shift >= 0) {
+		if ((adap->params.tp.ingress_config & USE_ENC_IDX_F) &&
+		    fs->mask.encap_vld)
+			ntuple |= (u64)((fs->val.encap_vld << 16) |
+					(fs->val.ovlan)) << tp->vnic_shift;
+		else if ((adap->params.tp.ingress_config & VNIC_F) &&
+			 fs->mask.pfvf_vld)
+			ntuple |= (u64)((fs->val.pfvf_vld << 16) |
+					(fs->val.pf << 13) |
+					(fs->val.vf)) << tp->vnic_shift;
+		else
+			ntuple |= (u64)((fs->val.ovlan_vld << 16) |
+					(fs->val.ovlan)) << tp->vnic_shift;
+	}
+
+	if (tp->macmatch_shift >= 0 && fs->mask.macidx)
+		ntuple |= (u64)(fs->val.macidx) << tp->macmatch_shift;
+
+	if (tp->ethertype_shift >= 0 && fs->mask.ethtype)
+		ntuple |= (u64)(fs->val.ethtype) << tp->ethertype_shift;
+
+	if (tp->matchtype_shift >= 0 && fs->mask.matchtype)
+		ntuple |= (u64)(fs->val.matchtype) << tp->matchtype_shift;
+
+	if (tp->frag_shift >= 0 && fs->mask.frag)
+		ntuple |= (u64)(fs->val.frag) << tp->frag_shift;
+
+	if (tp->fcoe_shift >= 0 && fs->mask.fcoe)
+		ntuple |= (u64)(fs->val.fcoe) << tp->fcoe_shift;
+	return ntuple;
+}
+
+static void mk_act_open_req6(struct filter_entry *f, struct sk_buff *skb,
+			     unsigned int qid_filterid, struct adapter *adap)
+{
+	struct cpl_t6_act_open_req6 *t6req = NULL;
+	struct cpl_act_open_req6 *req = NULL;
+
+	t6req = (struct cpl_t6_act_open_req6 *)__skb_put(skb, sizeof(*t6req));
+	INIT_TP_WR(t6req, 0);
+	req = (struct cpl_act_open_req6 *)t6req;
+	OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_ACT_OPEN_REQ6, qid_filterid));
+	req->local_port = cpu_to_be16(f->fs.val.lport);
+	req->peer_port = cpu_to_be16(f->fs.val.fport);
+	req->local_ip_hi = *(__be64 *)(&f->fs.val.lip);
+	req->local_ip_lo = *(((__be64 *)&f->fs.val.lip) + 1);
+	req->peer_ip_hi = *(__be64 *)(&f->fs.val.fip);
+	req->peer_ip_lo = *(((__be64 *)&f->fs.val.fip) + 1);
+	req->opt0 = cpu_to_be64(NAGLE_V(f->fs.newvlan == VLAN_REMOVE ||
+					f->fs.newvlan == VLAN_REWRITE) |
+				DELACK_V(f->fs.hitcnts) |
+				L2T_IDX_V(f->l2t ? f->l2t->idx : 0) |
+				SMAC_SEL_V((cxgb4_port_viid(f->dev) &
+					    0x7F) << 1) |
+				TX_CHAN_V(f->fs.eport) |
+				NO_CONG_V(f->fs.rpttid) |
+				ULP_MODE_V(f->fs.nat_mode ?
+					   ULP_MODE_TCPDDP : ULP_MODE_NONE) |
+				TCAM_BYPASS_F | NON_OFFLOAD_F);
+	t6req->params = cpu_to_be64(FILTER_TUPLE_V(hash_filter_ntuple(&f->fs,
+								      f->dev)));
+	t6req->opt2 = htonl(RSS_QUEUE_VALID_F |
+			    RSS_QUEUE_V(f->fs.iq) |
+			    TX_QUEUE_V(f->fs.nat_mode) |
+			    T5_OPT_2_VALID_F |
+			    RX_CHANNEL_V(cxgb4_port_e2cchan(f->dev)) |
+			    PACE_V((f->fs.maskhash) |
+				   ((f->fs.dirsteerhash) << 1)));
+}
+
+static void mk_act_open_req(struct filter_entry *f, struct sk_buff *skb,
+			    unsigned int qid_filterid, struct adapter *adap)
+{
+	struct cpl_t6_act_open_req *t6req = NULL;
+	struct cpl_act_open_req *req = NULL;
+
+	t6req = (struct cpl_t6_act_open_req *)__skb_put(skb, sizeof(*t6req));
+	INIT_TP_WR(t6req, 0);
+	req = (struct cpl_act_open_req *)t6req;
+	OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_ACT_OPEN_REQ, qid_filterid));
+	req->local_port = cpu_to_be16(f->fs.val.lport);
+	req->peer_port = cpu_to_be16(f->fs.val.fport);
+	memcpy(&req->local_ip, f->fs.val.lip, 4);
+	memcpy(&req->peer_ip, f->fs.val.fip, 4);
+	req->opt0 = cpu_to_be64(NAGLE_V(f->fs.newvlan == VLAN_REMOVE ||
+					f->fs.newvlan == VLAN_REWRITE) |
+				DELACK_V(f->fs.hitcnts) |
+				L2T_IDX_V(f->l2t ? f->l2t->idx : 0) |
+				SMAC_SEL_V((cxgb4_port_viid(f->dev) &
+					    0x7F) << 1) |
+				TX_CHAN_V(f->fs.eport) |
+				NO_CONG_V(f->fs.rpttid) |
+				ULP_MODE_V(f->fs.nat_mode ?
+					   ULP_MODE_TCPDDP : ULP_MODE_NONE) |
+				TCAM_BYPASS_F | NON_OFFLOAD_F);
+
+	t6req->params = cpu_to_be64(FILTER_TUPLE_V(hash_filter_ntuple(&f->fs,
+								      f->dev)));
+	t6req->opt2 = htonl(RSS_QUEUE_VALID_F |
+			    RSS_QUEUE_V(f->fs.iq) |
+			    TX_QUEUE_V(f->fs.nat_mode) |
+			    T5_OPT_2_VALID_F |
+			    RX_CHANNEL_V(cxgb4_port_e2cchan(f->dev)) |
+			    PACE_V((f->fs.maskhash) |
+				   ((f->fs.dirsteerhash) << 1)));
+}
+
+static int cxgb4_set_hash_filter(struct net_device *dev,
+				 struct ch_filter_specification *fs,
+				 struct filter_ctx *ctx)
+{
+	struct adapter *adapter = netdev2adap(dev);
+	struct port_info *pi = netdev_priv(dev);
+	struct tid_info *t = &adapter->tids;
+	struct filter_entry *f;
+	struct sk_buff *skb;
+	int iq, atid, size;
+	int ret = 0;
+	u32 iconf;
+
+	fill_default_mask(fs);
+	ret = validate_filter(dev, fs);
+	if (ret)
+		return ret;
+
+	iq = get_filter_steerq(dev, fs);
+	if (iq < 0)
+		return iq;
+
+	f = kzalloc(sizeof(*f), GFP_KERNEL);
+	if (!f)
+		return -ENOMEM;
+
+	f->fs = *fs;
+	f->ctx = ctx;
+	f->dev = dev;
+	f->fs.iq = iq;
+
+	/* If the new filter requires loopback Destination MAC and/or VLAN
+	 * rewriting then we need to allocate a Layer 2 Table (L2T) entry for
+	 * the filter.
+	 */
+	if (f->fs.newdmac || f->fs.newvlan) {
+		/* allocate L2T entry for new filter */
+		f->l2t = t4_l2t_alloc_switching(adapter, f->fs.vlan,
+						f->fs.eport, f->fs.dmac);
+		if (!f->l2t) {
+			ret = -ENOMEM;
+			goto out_err;
+		}
+	}
+
+	/* If the new filter requires loopback Source MAC rewriting then
+	 * we need to allocate a SMT entry for the filter.
+	 */
+	if (f->fs.newsmac) {
+		f->smt = cxgb4_smt_alloc_switching(f->dev, f->fs.smac);
+		if (!f->smt) {
+			if (f->l2t) {
+				cxgb4_l2t_release(f->l2t);
+				f->l2t = NULL;
+			}
+			ret = -ENOMEM;
+			goto free_l2t;
+		}
+	}
+
+	atid = cxgb4_alloc_atid(t, f);
+	if (atid < 0) {
+		ret = atid;
+		goto free_smt;
+	}
+
+	iconf = adapter->params.tp.ingress_config;
+	if (iconf & VNIC_F) {
+		f->fs.val.ovlan = (fs->val.pf << 13) | fs->val.vf;
+		f->fs.mask.ovlan = (fs->mask.pf << 13) | fs->mask.vf;
+		f->fs.val.ovlan_vld = fs->val.pfvf_vld;
+		f->fs.mask.ovlan_vld = fs->mask.pfvf_vld;
+	} else if (iconf & USE_ENC_IDX_F) {
+		if (f->fs.val.encap_vld) {
+			struct port_info *pi = netdev_priv(f->dev);
+			static const u8 match_all_mac[] = { 0, 0, 0, 0, 0, 0 };
+
+			/* allocate MPS TCAM entry */
+			ret = t4_alloc_encap_mac_filt(adapter, pi->viid,
+						      match_all_mac,
+						      match_all_mac,
+						      f->fs.val.vni,
+						      f->fs.mask.vni,
+						      0, 1, 1);
+			if (ret < 0)
+				goto free_atid;
+
+			f->fs.val.ovlan = ret;
+			f->fs.mask.ovlan = 0xffff;
+			f->fs.val.ovlan_vld = 1;
+			f->fs.mask.ovlan_vld = 1;
+		}
+	}
+
+	size = sizeof(struct cpl_t6_act_open_req);
+	if (f->fs.type) {
+		ret = cxgb4_clip_get(f->dev, (const u32 *)&f->fs.val.lip, 1);
+		if (ret)
+			goto free_mps;
+
+		skb = alloc_skb(size, GFP_KERNEL);
+		if (!skb) {
+			ret = -ENOMEM;
+			goto free_clip;
+		}
+
+		mk_act_open_req6(f, skb,
+				 ((adapter->sge.fw_evtq.abs_id << 14) | atid),
+				 adapter);
+	} else {
+		skb = alloc_skb(size, GFP_KERNEL);
+		if (!skb) {
+			ret = -ENOMEM;
+			goto free_mps;
+		}
+
+		mk_act_open_req(f, skb,
+				((adapter->sge.fw_evtq.abs_id << 14) | atid),
+				adapter);
+	}
+
+	f->pending = 1;
+	set_wr_txq(skb, CPL_PRIORITY_SETUP, f->fs.val.iport & 0x3);
+	t4_ofld_send(adapter, skb);
+	return 0;
+
+free_clip:
+	cxgb4_clip_release(f->dev, (const u32 *)&f->fs.val.lip, 1);
+
+free_mps:
+	if (f->fs.val.encap_vld && f->fs.val.ovlan_vld)
+		t4_free_encap_mac_filt(adapter, pi->viid, f->fs.val.ovlan, 1);
+
+free_atid:
+	cxgb4_free_atid(t, atid);
+
+free_smt:
+	if (f->smt) {
+		cxgb4_smt_release(f->smt);
+		f->smt = NULL;
+	}
+
+free_l2t:
+	if (f->l2t) {
+		cxgb4_l2t_release(f->l2t);
+		f->l2t = NULL;
+	}
+
+out_err:
+	kfree(f);
+	return ret;
+}
+
+/* Check a Chelsio Filter Request for validity, convert it into our internal
+ * format and send it to the hardware.  Return 0 on success, an error number
+ * otherwise.  We attach any provided filter operation context to the internal
+ * filter specification in order to facilitate signaling completion of the
+ * operation.
+ */
+int __cxgb4_set_filter(struct net_device *dev, int ftid,
+		       struct ch_filter_specification *fs,
+		       struct filter_ctx *ctx)
+{
+	struct adapter *adapter = netdev2adap(dev);
+	unsigned int max_fidx, fidx, chip_ver;
+	int iq, ret, filter_id = ftid;
+	struct filter_entry *f, *tab;
+	u32 iconf;
+
+	chip_ver = CHELSIO_CHIP_VERSION(adapter->params.chip);
+	if (fs->hash) {
+		if (is_hashfilter(adapter))
+			return cxgb4_set_hash_filter(dev, fs, ctx);
+		netdev_err(dev, "%s: Exact-match filters only supported with Hash Filter configuration\n",
+			   __func__);
+		return -EINVAL;
+	}
+
+	max_fidx = adapter->tids.nftids + adapter->tids.nhpftids;
+	if (filter_id != (max_fidx + adapter->tids.nsftids - 1) &&
+	    filter_id >= max_fidx)
+		return -E2BIG;
+
+	fill_default_mask(fs);
+
+	ret = validate_filter(dev, fs);
+	if (ret)
+		return ret;
+
+	iq = get_filter_steerq(dev, fs);
+	if (iq < 0)
+		return iq;
+
+	if (fs->prio) {
+		tab = &adapter->tids.hpftid_tab[0];
+	} else {
+		tab = &adapter->tids.ftid_tab[0];
+		filter_id = ftid - adapter->tids.nhpftids;
+	}
+
+	/* IPv6 filters occupy four slots and must be aligned on
+	 * four-slot boundaries.  IPv4 filters only occupy a single
+	 * slot and have no alignment requirements but writing a new
+	 * IPv4 filter into the middle of an existing IPv6 filter
+	 * requires clearing the old IPv6 filter and hence we prevent
+	 * insertion.
+	 */
+	if (fs->type == 0) { /* IPv4 */
+		/* For T6, If our IPv4 filter isn't being written to a
+		 * multiple of two filter index and there's an IPv6
+		 * filter at the multiple of 2 base slot, then we need
+		 * to delete that IPv6 filter ...
+		 * For adapters below T6, IPv6 filter occupies 4 entries.
+		 * Hence we need to delete the filter in multiple of 4 slot.
+		 */
+		if (chip_ver < CHELSIO_T6)
+			fidx = filter_id & ~0x3;
+		else
+			fidx = filter_id & ~0x1;
+
+		if (fidx != filter_id && tab[fidx].fs.type) {
+			f = &tab[fidx];
+			if (f->valid) {
+				dev_err(adapter->pdev_dev,
+					"Invalid location. IPv6 requires 4 slots and is occupying slots %u to %u\n",
+					fidx, fidx + 3);
+				return -EINVAL;
+			}
+		}
+	} else { /* IPv6 */
+		if (chip_ver < CHELSIO_T6) {
+			/* Ensure that the IPv6 filter is aligned on a
+			 * multiple of 4 boundary.
+			 */
+			if (filter_id & 0x3) {
+				dev_err(adapter->pdev_dev,
+					"Invalid location. IPv6 must be aligned on a 4-slot boundary\n");
+				return -EINVAL;
+			}
+
+			/* Check all except the base overlapping IPv4 filter
+			 * slots.
+			 */
+			for (fidx = filter_id + 1; fidx < filter_id + 4;
+			     fidx++) {
+				f = &tab[fidx];
+				if (f->valid) {
+					dev_err(adapter->pdev_dev,
+						"Invalid location.  IPv6 requires 4 slots and an IPv4 filter exists at %u\n",
+						fidx);
+					return -EBUSY;
+				}
+			}
+		} else {
+			/* For T6, CLIP being enabled, IPv6 filter would occupy
+			 * 2 entries.
+			 */
+			if (filter_id & 0x1)
+				return -EINVAL;
+			/* Check overlapping IPv4 filter slot */
+			fidx = filter_id + 1;
+			f = &tab[fidx];
+			if (f->valid) {
+				pr_err("%s: IPv6 filter requires 2 indices. IPv4 filter already present at %d. Please remove IPv4 filter first.\n",
+				       __func__, fidx);
+				return -EBUSY;
+			}
+		}
+	}
+
+	/* Check to make sure that provided filter index is not
+	 * already in use by someone else
+	 */
+	f = &tab[filter_id];
+	if (f->valid)
+		return -EBUSY;
+
+	if (fs->prio) {
+		fidx = filter_id + adapter->tids.hpftid_base;
+		ret = cxgb4_set_hpftid(&adapter->tids, filter_id,
+				       fs->type ? PF_INET6 : PF_INET);
+	} else {
+		fidx = filter_id + adapter->tids.ftid_base;
+		ret = cxgb4_set_ftid(&adapter->tids, filter_id,
+				     fs->type ? PF_INET6 : PF_INET,
+				     chip_ver);
+	}
+
+	if (ret)
+		return ret;
+
+	/* Check t  make sure the filter requested is writable ... */
+	ret = writable_filter(f);
+	if (ret)
+		goto free_tid;
+
+	if (is_t6(adapter->params.chip) && fs->type &&
+	    ipv6_addr_type((const struct in6_addr *)fs->val.lip) !=
+	    IPV6_ADDR_ANY) {
+		ret = cxgb4_clip_get(dev, (const u32 *)&fs->val.lip, 1);
+		if (ret)
+			goto free_tid;
+	}
+
+	/* Convert the filter specification into our internal format.
+	 * We copy the PF/VF specification into the Outer VLAN field
+	 * here so the rest of the code -- including the interface to
+	 * the firmware -- doesn't have to constantly do these checks.
+	 */
+	f->fs = *fs;
+	f->fs.iq = iq;
+	f->dev = dev;
+
+	iconf = adapter->params.tp.ingress_config;
+	if (iconf & VNIC_F) {
+		f->fs.val.ovlan = (fs->val.pf << 13) | fs->val.vf;
+		f->fs.mask.ovlan = (fs->mask.pf << 13) | fs->mask.vf;
+		f->fs.val.ovlan_vld = fs->val.pfvf_vld;
+		f->fs.mask.ovlan_vld = fs->mask.pfvf_vld;
+	} else if (iconf & USE_ENC_IDX_F) {
+		if (f->fs.val.encap_vld) {
+			struct port_info *pi = netdev_priv(f->dev);
+			static const u8 match_all_mac[] = { 0, 0, 0, 0, 0, 0 };
+
+			/* allocate MPS TCAM entry */
+			ret = t4_alloc_encap_mac_filt(adapter, pi->viid,
+						      match_all_mac,
+						      match_all_mac,
+						      f->fs.val.vni,
+						      f->fs.mask.vni,
+						      0, 1, 1);
+			if (ret < 0)
+				goto free_tid;
+
+			f->fs.val.ovlan = ret;
+			f->fs.mask.ovlan = 0x1ff;
+			f->fs.val.ovlan_vld = 1;
+			f->fs.mask.ovlan_vld = 1;
+		}
+	}
+
+	/* Attempt to set the filter.  If we don't succeed, we clear
+	 * it and return the failure.
+	 */
+	f->ctx = ctx;
+	f->tid = fidx; /* Save the actual tid */
+	ret = set_filter_wr(adapter, ftid);
+	if (ret)
+		goto free_tid;
+
+	return ret;
+
+free_tid:
+	if (f->fs.prio)
+		cxgb4_clear_hpftid(&adapter->tids, filter_id,
+				   fs->type ? PF_INET6 : PF_INET);
+	else
+		cxgb4_clear_ftid(&adapter->tids, filter_id,
+				 fs->type ? PF_INET6 : PF_INET,
+				 chip_ver);
+
+	clear_filter(adapter, f);
+	return ret;
+}
+
+static int cxgb4_del_hash_filter(struct net_device *dev, int filter_id,
+				 struct filter_ctx *ctx)
+{
+	struct adapter *adapter = netdev2adap(dev);
+	struct tid_info *t = &adapter->tids;
+	struct cpl_abort_req *abort_req;
+	struct cpl_abort_rpl *abort_rpl;
+	struct cpl_set_tcb_field *req;
+	struct ulptx_idata *aligner;
+	struct work_request_hdr *wr;
+	struct filter_entry *f;
+	struct sk_buff *skb;
+	unsigned int wrlen;
+	int ret;
+
+	netdev_dbg(dev, "%s: filter_id = %d ; nftids = %d\n",
+		   __func__, filter_id, adapter->tids.nftids);
+
+	if (tid_out_of_range(t, filter_id))
+		return -E2BIG;
+
+	f = lookup_tid(t, filter_id);
+	if (!f) {
+		netdev_err(dev, "%s: no filter entry for filter_id = %d",
+			   __func__, filter_id);
+		return -EINVAL;
+	}
+
+	ret = writable_filter(f);
+	if (ret)
+		return ret;
+
+	if (!f->valid)
+		return -EINVAL;
+
+	f->ctx = ctx;
+	f->pending = 1;
+	wrlen = roundup(sizeof(*wr) + (sizeof(*req) + sizeof(*aligner))
+			+ sizeof(*abort_req) + sizeof(*abort_rpl), 16);
+	skb = alloc_skb(wrlen, GFP_KERNEL);
+	if (!skb) {
+		netdev_err(dev, "%s: could not allocate skb ..\n", __func__);
+		return -ENOMEM;
+	}
+	set_wr_txq(skb, CPL_PRIORITY_CONTROL, f->fs.val.iport & 0x3);
+	req = (struct cpl_set_tcb_field *)__skb_put(skb, wrlen);
+	INIT_ULPTX_WR(req, wrlen, 0, 0);
+	wr = (struct work_request_hdr *)req;
+	wr++;
+	req = (struct cpl_set_tcb_field *)wr;
+	mk_set_tcb_ulp(f, req, TCB_RSS_INFO_W, TCB_RSS_INFO_V(TCB_RSS_INFO_M),
+		       TCB_RSS_INFO_V(adapter->sge.fw_evtq.abs_id), 0, 1);
+	aligner = (struct ulptx_idata *)(req + 1);
+	abort_req = (struct cpl_abort_req *)(aligner + 1);
+	mk_abort_req_ulp(abort_req, f->tid);
+	abort_rpl = (struct cpl_abort_rpl *)(abort_req + 1);
+	mk_abort_rpl_ulp(abort_rpl, f->tid);
+	t4_ofld_send(adapter, skb);
+	return 0;
+}
+
+/* Check a delete filter request for validity and send it to the hardware.
+ * Return 0 on success, an error number otherwise.  We attach any provided
+ * filter operation context to the internal filter specification in order to
+ * facilitate signaling completion of the operation.
+ */
+int __cxgb4_del_filter(struct net_device *dev, int filter_id,
+		       struct ch_filter_specification *fs,
+		       struct filter_ctx *ctx)
+{
+	struct adapter *adapter = netdev2adap(dev);
+	unsigned int max_fidx, chip_ver;
+	struct filter_entry *f;
+	int ret;
+
+	chip_ver = CHELSIO_CHIP_VERSION(adapter->params.chip);
+	if (fs && fs->hash) {
+		if (is_hashfilter(adapter))
+			return cxgb4_del_hash_filter(dev, filter_id, ctx);
+		netdev_err(dev, "%s: Exact-match filters only supported with Hash Filter configuration\n",
+			   __func__);
+		return -EINVAL;
+	}
+
+	max_fidx = adapter->tids.nftids + adapter->tids.nhpftids;
+	if (filter_id != (max_fidx + adapter->tids.nsftids - 1) &&
+	    filter_id >= max_fidx)
+		return -E2BIG;
+
+	if (filter_id < adapter->tids.nhpftids)
+		f = &adapter->tids.hpftid_tab[filter_id];
+	else
+		f = &adapter->tids.ftid_tab[filter_id - adapter->tids.nhpftids];
+
+	ret = writable_filter(f);
+	if (ret)
+		return ret;
+
+	if (f->valid) {
+		f->ctx = ctx;
+		if (f->fs.prio)
+			cxgb4_clear_hpftid(&adapter->tids,
+					   f->tid - adapter->tids.hpftid_base,
+					   f->fs.type ? PF_INET6 : PF_INET);
+		else
+			cxgb4_clear_ftid(&adapter->tids,
+					 f->tid - adapter->tids.ftid_base,
+					 f->fs.type ? PF_INET6 : PF_INET,
+					 chip_ver);
+		return del_filter_wr(adapter, filter_id);
+	}
+
+	/* If the caller has passed in a Completion Context then we need to
+	 * mark it as a successful completion so they don't stall waiting
+	 * for it.
+	 */
+	if (ctx) {
+		ctx->result = 0;
+		complete(&ctx->completion);
+	}
+	return ret;
+}
+
+int cxgb4_set_filter(struct net_device *dev, int filter_id,
+		     struct ch_filter_specification *fs)
+{
+	struct filter_ctx ctx;
+	int ret;
+
+	init_completion(&ctx.completion);
+
+	ret = __cxgb4_set_filter(dev, filter_id, fs, &ctx);
+	if (ret)
+		goto out;
+
+	/* Wait for reply */
+	ret = wait_for_completion_timeout(&ctx.completion, 10 * HZ);
+	if (!ret)
+		return -ETIMEDOUT;
+
+	ret = ctx.result;
+out:
+	return ret;
+}
+
+int cxgb4_del_filter(struct net_device *dev, int filter_id,
+		     struct ch_filter_specification *fs)
+{
+	struct filter_ctx ctx;
+	int ret;
+
+	if (netdev2adap(dev)->flags & CXGB4_SHUTTING_DOWN)
+		return 0;
+
+	init_completion(&ctx.completion);
+
+	ret = __cxgb4_del_filter(dev, filter_id, fs, &ctx);
+	if (ret)
+		goto out;
+
+	/* Wait for reply */
+	ret = wait_for_completion_timeout(&ctx.completion, 10 * HZ);
+	if (!ret)
+		return -ETIMEDOUT;
+
+	ret = ctx.result;
+out:
+	return ret;
+}
+
+static int configure_filter_tcb(struct adapter *adap, unsigned int tid,
+				struct filter_entry *f)
+{
+	if (f->fs.hitcnts) {
+		set_tcb_field(adap, f, tid, TCB_TIMESTAMP_W,
+			      TCB_TIMESTAMP_V(TCB_TIMESTAMP_M),
+			      TCB_TIMESTAMP_V(0ULL),
+			      1);
+		set_tcb_field(adap, f, tid, TCB_RTT_TS_RECENT_AGE_W,
+			      TCB_RTT_TS_RECENT_AGE_V(TCB_RTT_TS_RECENT_AGE_M),
+			      TCB_RTT_TS_RECENT_AGE_V(0ULL),
+			      1);
+	}
+
+	if (f->fs.newdmac)
+		set_tcb_tflag(adap, f, tid, TF_CCTRL_ECE_S, 1,
+			      1);
+
+	if (f->fs.newvlan == VLAN_INSERT ||
+	    f->fs.newvlan == VLAN_REWRITE)
+		set_tcb_tflag(adap, f, tid, TF_CCTRL_RFR_S, 1,
+			      1);
+	if (f->fs.newsmac)
+		configure_filter_smac(adap, f);
+
+	if (f->fs.nat_mode) {
+		switch (f->fs.nat_mode) {
+		case NAT_MODE_DIP:
+			set_nat_params(adap, f, tid, true, false, false, false);
+			break;
+
+		case NAT_MODE_DIP_DP:
+			set_nat_params(adap, f, tid, true, false, true, false);
+			break;
+
+		case NAT_MODE_DIP_DP_SIP:
+			set_nat_params(adap, f, tid, true, true, true, false);
+			break;
+		case NAT_MODE_DIP_DP_SP:
+			set_nat_params(adap, f, tid, true, false, true, true);
+			break;
+
+		case NAT_MODE_SIP_SP:
+			set_nat_params(adap, f, tid, false, true, false, true);
+			break;
+
+		case NAT_MODE_DIP_SIP_SP:
+			set_nat_params(adap, f, tid, true, true, false, true);
+			break;
+
+		case NAT_MODE_ALL:
+			set_nat_params(adap, f, tid, true, true, true, true);
+			break;
+
+		default:
+			pr_err("%s: Invalid NAT mode: %d\n",
+			       __func__, f->fs.nat_mode);
+			return -EINVAL;
+		}
+	}
+	return 0;
+}
+
+void hash_del_filter_rpl(struct adapter *adap,
+			 const struct cpl_abort_rpl_rss *rpl)
+{
+	unsigned int status = rpl->status;
+	struct tid_info *t = &adap->tids;
+	unsigned int tid = GET_TID(rpl);
+	struct filter_ctx *ctx = NULL;
+	struct filter_entry *f;
+
+	dev_dbg(adap->pdev_dev, "%s: status = %u; tid = %u\n",
+		__func__, status, tid);
+
+	f = lookup_tid(t, tid);
+	if (!f) {
+		dev_err(adap->pdev_dev, "%s:could not find filter entry",
+			__func__);
+		return;
+	}
+	ctx = f->ctx;
+	f->ctx = NULL;
+	clear_filter(adap, f);
+	cxgb4_remove_tid(t, 0, tid, 0);
+	kfree(f);
+	if (ctx) {
+		ctx->result = 0;
+		complete(&ctx->completion);
+	}
+}
+
+void hash_filter_rpl(struct adapter *adap, const struct cpl_act_open_rpl *rpl)
+{
+	unsigned int ftid = TID_TID_G(AOPEN_ATID_G(ntohl(rpl->atid_status)));
+	unsigned int status  = AOPEN_STATUS_G(ntohl(rpl->atid_status));
+	struct tid_info *t = &adap->tids;
+	unsigned int tid = GET_TID(rpl);
+	struct filter_ctx *ctx = NULL;
+	struct filter_entry *f;
+
+	dev_dbg(adap->pdev_dev, "%s: tid = %u; atid = %u; status = %u\n",
+		__func__, tid, ftid, status);
+
+	f = lookup_atid(t, ftid);
+	if (!f) {
+		dev_err(adap->pdev_dev, "%s:could not find filter entry",
+			__func__);
+		return;
+	}
+	ctx = f->ctx;
+	f->ctx = NULL;
+
+	switch (status) {
+	case CPL_ERR_NONE:
+		f->tid = tid;
+		f->pending = 0;
+		f->valid = 1;
+		cxgb4_insert_tid(t, f, f->tid, 0);
+		cxgb4_free_atid(t, ftid);
+		if (ctx) {
+			ctx->tid = f->tid;
+			ctx->result = 0;
+		}
+		if (configure_filter_tcb(adap, tid, f)) {
+			clear_filter(adap, f);
+			cxgb4_remove_tid(t, 0, tid, 0);
+			kfree(f);
+			if (ctx) {
+				ctx->result = -EINVAL;
+				complete(&ctx->completion);
+			}
+			return;
+		}
+		switch (f->fs.action) {
+		case FILTER_PASS:
+			if (f->fs.dirsteer)
+				set_tcb_tflag(adap, f, tid,
+					      TF_DIRECT_STEER_S, 1, 1);
+			break;
+		case FILTER_DROP:
+			set_tcb_tflag(adap, f, tid, TF_DROP_S, 1, 1);
+			break;
+		case FILTER_SWITCH:
+			set_tcb_tflag(adap, f, tid, TF_LPBK_S, 1, 1);
+			break;
+		}
+
+		break;
+
+	default:
+		if (status != CPL_ERR_TCAM_FULL)
+			dev_err(adap->pdev_dev, "%s: filter creation PROBLEM; status = %u\n",
+				__func__, status);
+
+		if (ctx) {
+			if (status == CPL_ERR_TCAM_FULL)
+				ctx->result = -ENOSPC;
+			else
+				ctx->result = -EINVAL;
+		}
+		clear_filter(adap, f);
+		cxgb4_free_atid(t, ftid);
+		kfree(f);
+	}
+	if (ctx)
+		complete(&ctx->completion);
+}
+
+/* Handle a filter write/deletion reply. */
+void filter_rpl(struct adapter *adap, const struct cpl_set_tcb_rpl *rpl)
+{
+	unsigned int tid = GET_TID(rpl);
+	struct filter_entry *f = NULL;
+	unsigned int max_fidx;
+	int idx;
+
+	max_fidx = adap->tids.nftids + adap->tids.nsftids;
+	/* Get the corresponding filter entry for this tid */
+	if (adap->tids.ftid_tab) {
+		idx = tid - adap->tids.hpftid_base;
+		if (idx < adap->tids.nhpftids) {
+			f = &adap->tids.hpftid_tab[idx];
+		} else {
+			/* Check this in normal filter region */
+			idx = tid - adap->tids.ftid_base;
+			if (idx >= max_fidx)
+				return;
+			f = &adap->tids.ftid_tab[idx];
+			idx += adap->tids.nhpftids;
+		}
+
+		if (f->tid != tid)
+			return;
+	}
+
+	/* We found the filter entry for this tid */
+	if (f) {
+		unsigned int ret = TCB_COOKIE_G(rpl->cookie);
+		struct filter_ctx *ctx;
+
+		/* Pull off any filter operation context attached to the
+		 * filter.
+		 */
+		ctx = f->ctx;
+		f->ctx = NULL;
+
+		if (ret == FW_FILTER_WR_FLT_DELETED) {
+			/* Clear the filter when we get confirmation from the
+			 * hardware that the filter has been deleted.
+			 */
+			clear_filter(adap, f);
+			if (ctx)
+				ctx->result = 0;
+		} else if (ret == FW_FILTER_WR_FLT_ADDED) {
+			f->pending = 0;  /* async setup completed */
+			f->valid = 1;
+			if (ctx) {
+				ctx->result = 0;
+				ctx->tid = idx;
+			}
+		} else {
+			/* Something went wrong.  Issue a warning about the
+			 * problem and clear everything out.
+			 */
+			dev_err(adap->pdev_dev, "filter %u setup failed with error %u\n",
+				idx, ret);
+			clear_filter(adap, f);
+			if (ctx)
+				ctx->result = -EINVAL;
+		}
+		if (ctx)
+			complete(&ctx->completion);
+	}
+}
+
+void init_hash_filter(struct adapter *adap)
+{
+	u32 reg;
+
+	/* On T6, verify the necessary register configs and warn the user in
+	 * case of improper config
+	 */
+	if (is_t6(adap->params.chip)) {
+		if (is_offload(adap)) {
+			if (!(t4_read_reg(adap, TP_GLOBAL_CONFIG_A)
+			   & ACTIVEFILTERCOUNTS_F)) {
+				dev_err(adap->pdev_dev, "Invalid hash filter + ofld config\n");
+				return;
+			}
+		} else {
+			reg = t4_read_reg(adap, LE_DB_RSP_CODE_0_A);
+			if (TCAM_ACTV_HIT_G(reg) != 4) {
+				dev_err(adap->pdev_dev, "Invalid hash filter config\n");
+				return;
+			}
+
+			reg = t4_read_reg(adap, LE_DB_RSP_CODE_1_A);
+			if (HASH_ACTV_HIT_G(reg) != 4) {
+				dev_err(adap->pdev_dev, "Invalid hash filter config\n");
+				return;
+			}
+		}
+
+	} else {
+		dev_err(adap->pdev_dev, "Hash filter supported only on T6\n");
+		return;
+	}
+
+	adap->params.hash_filter = 1;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.h
new file mode 100644
index 000000000..807a8dafe
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.h
@@ -0,0 +1,58 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_FILTER_H
+#define __CXGB4_FILTER_H
+
+#include "t4_msg.h"
+
+#define WORD_MASK	0xffffffff
+
+void filter_rpl(struct adapter *adap, const struct cpl_set_tcb_rpl *rpl);
+void hash_filter_rpl(struct adapter *adap, const struct cpl_act_open_rpl *rpl);
+void hash_del_filter_rpl(struct adapter *adap,
+			 const struct cpl_abort_rpl_rss *rpl);
+void clear_filter(struct adapter *adap, struct filter_entry *f);
+
+int set_filter_wr(struct adapter *adapter, int fidx);
+int delete_filter(struct adapter *adapter, unsigned int fidx);
+
+int writable_filter(struct filter_entry *f);
+void clear_all_filters(struct adapter *adapter);
+void init_hash_filter(struct adapter *adap);
+bool is_filter_exact_match(struct adapter *adap,
+			   struct ch_filter_specification *fs);
+void cxgb4_cleanup_ethtool_filters(struct adapter *adap);
+int cxgb4_init_ethtool_filters(struct adapter *adap);
+#endif /* __CXGB4_FILTER_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
new file mode 100644
index 000000000..9cbce1faa
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -0,0 +1,7284 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/bitmap.h>
+#include <linux/crc32.h>
+#include <linux/ctype.h>
+#include <linux/debugfs.h>
+#include <linux/err.h>
+#include <linux/etherdevice.h>
+#include <linux/firmware.h>
+#include <linux/if.h>
+#include <linux/if_vlan.h>
+#include <linux/init.h>
+#include <linux/log2.h>
+#include <linux/mdio.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/mutex.h>
+#include <linux/netdevice.h>
+#include <linux/pci.h>
+#include <linux/aer.h>
+#include <linux/rtnetlink.h>
+#include <linux/sched.h>
+#include <linux/seq_file.h>
+#include <linux/sockios.h>
+#include <linux/vmalloc.h>
+#include <linux/workqueue.h>
+#include <net/neighbour.h>
+#include <net/netevent.h>
+#include <net/addrconf.h>
+#include <net/bonding.h>
+#include <linux/uaccess.h>
+#include <linux/crash_dump.h>
+#include <net/udp_tunnel.h>
+#include <net/xfrm.h>
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+#include <net/tls.h>
+#endif
+
+#include "cxgb4.h"
+#include "cxgb4_filter.h"
+#include "t4_regs.h"
+#include "t4_values.h"
+#include "t4_msg.h"
+#include "t4fw_api.h"
+#include "t4fw_version.h"
+#include "cxgb4_dcb.h"
+#include "srq.h"
+#include "cxgb4_debugfs.h"
+#include "clip_tbl.h"
+#include "l2t.h"
+#include "smt.h"
+#include "sched.h"
+#include "cxgb4_tc_u32.h"
+#include "cxgb4_tc_flower.h"
+#include "cxgb4_tc_mqprio.h"
+#include "cxgb4_tc_matchall.h"
+#include "cxgb4_ptp.h"
+#include "cxgb4_cudbg.h"
+
+char cxgb4_driver_name[] = KBUILD_MODNAME;
+
+#define DRV_DESC "Chelsio T4/T5/T6 Network Driver"
+
+#define DFLT_MSG_ENABLE (NETIF_MSG_DRV | NETIF_MSG_PROBE | NETIF_MSG_LINK | \
+			 NETIF_MSG_TIMER | NETIF_MSG_IFDOWN | NETIF_MSG_IFUP |\
+			 NETIF_MSG_RX_ERR | NETIF_MSG_TX_ERR)
+
+/* Macros needed to support the PCI Device ID Table ...
+ */
+#define CH_PCI_DEVICE_ID_TABLE_DEFINE_BEGIN \
+	static const struct pci_device_id cxgb4_pci_tbl[] = {
+#define CXGB4_UNIFIED_PF 0x4
+
+#define CH_PCI_DEVICE_ID_FUNCTION CXGB4_UNIFIED_PF
+
+/* Include PCI Device IDs for both PF4 and PF0-3 so our PCI probe() routine is
+ * called for both.
+ */
+#define CH_PCI_DEVICE_ID_FUNCTION2 0x0
+
+#define CH_PCI_ID_TABLE_ENTRY(devid) \
+		{PCI_VDEVICE(CHELSIO, (devid)), CXGB4_UNIFIED_PF}
+
+#define CH_PCI_DEVICE_ID_TABLE_DEFINE_END \
+		{ 0, } \
+	}
+
+#include "t4_pci_id_tbl.h"
+
+#define FW4_FNAME "cxgb4/t4fw.bin"
+#define FW5_FNAME "cxgb4/t5fw.bin"
+#define FW6_FNAME "cxgb4/t6fw.bin"
+#define FW4_CFNAME "cxgb4/t4-config.txt"
+#define FW5_CFNAME "cxgb4/t5-config.txt"
+#define FW6_CFNAME "cxgb4/t6-config.txt"
+#define PHY_AQ1202_FIRMWARE "cxgb4/aq1202_fw.cld"
+#define PHY_BCM84834_FIRMWARE "cxgb4/bcm8483.bin"
+#define PHY_AQ1202_DEVICEID 0x4409
+#define PHY_BCM84834_DEVICEID 0x4486
+
+MODULE_DESCRIPTION(DRV_DESC);
+MODULE_AUTHOR("Chelsio Communications");
+MODULE_LICENSE("Dual BSD/GPL");
+MODULE_DEVICE_TABLE(pci, cxgb4_pci_tbl);
+MODULE_FIRMWARE(FW4_FNAME);
+MODULE_FIRMWARE(FW5_FNAME);
+MODULE_FIRMWARE(FW6_FNAME);
+
+/*
+ * The driver uses the best interrupt scheme available on a platform in the
+ * order MSI-X, MSI, legacy INTx interrupts.  This parameter determines which
+ * of these schemes the driver may consider as follows:
+ *
+ * msi = 2: choose from among all three options
+ * msi = 1: only consider MSI and INTx interrupts
+ * msi = 0: force INTx interrupts
+ */
+static int msi = 2;
+
+module_param(msi, int, 0644);
+MODULE_PARM_DESC(msi, "whether to use INTx (0), MSI (1) or MSI-X (2)");
+
+/*
+ * Normally we tell the chip to deliver Ingress Packets into our DMA buffers
+ * offset by 2 bytes in order to have the IP headers line up on 4-byte
+ * boundaries.  This is a requirement for many architectures which will throw
+ * a machine check fault if an attempt is made to access one of the 4-byte IP
+ * header fields on a non-4-byte boundary.  And it's a major performance issue
+ * even on some architectures which allow it like some implementations of the
+ * x86 ISA.  However, some architectures don't mind this and for some very
+ * edge-case performance sensitive applications (like forwarding large volumes
+ * of small packets), setting this DMA offset to 0 will decrease the number of
+ * PCI-E Bus transfers enough to measurably affect performance.
+ */
+static int rx_dma_offset = 2;
+
+/* TX Queue select used to determine what algorithm to use for selecting TX
+ * queue. Select between the kernel provided function (select_queue=0) or user
+ * cxgb_select_queue function (select_queue=1)
+ *
+ * Default: select_queue=0
+ */
+static int select_queue;
+module_param(select_queue, int, 0644);
+MODULE_PARM_DESC(select_queue,
+		 "Select between kernel provided method of selecting or driver method of selecting TX queue. Default is kernel method.");
+
+static struct dentry *cxgb4_debugfs_root;
+
+LIST_HEAD(adapter_list);
+DEFINE_MUTEX(uld_mutex);
+LIST_HEAD(uld_list);
+
+static int cfg_queues(struct adapter *adap);
+
+static void link_report(struct net_device *dev)
+{
+	if (!netif_carrier_ok(dev))
+		netdev_info(dev, "link down\n");
+	else {
+		static const char *fc[] = { "no", "Rx", "Tx", "Tx/Rx" };
+
+		const char *s;
+		const struct port_info *p = netdev_priv(dev);
+
+		switch (p->link_cfg.speed) {
+		case 100:
+			s = "100Mbps";
+			break;
+		case 1000:
+			s = "1Gbps";
+			break;
+		case 10000:
+			s = "10Gbps";
+			break;
+		case 25000:
+			s = "25Gbps";
+			break;
+		case 40000:
+			s = "40Gbps";
+			break;
+		case 50000:
+			s = "50Gbps";
+			break;
+		case 100000:
+			s = "100Gbps";
+			break;
+		default:
+			pr_info("%s: unsupported speed: %d\n",
+				dev->name, p->link_cfg.speed);
+			return;
+		}
+
+		netdev_info(dev, "link up, %s, full-duplex, %s PAUSE\n", s,
+			    fc[p->link_cfg.fc]);
+	}
+}
+
+#ifdef CONFIG_CHELSIO_T4_DCB
+/* Set up/tear down Data Center Bridging Priority mapping for a net device. */
+static void dcb_tx_queue_prio_enable(struct net_device *dev, int enable)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	struct sge_eth_txq *txq = &adap->sge.ethtxq[pi->first_qset];
+	int i;
+
+	/* We use a simple mapping of Port TX Queue Index to DCB
+	 * Priority when we're enabling DCB.
+	 */
+	for (i = 0; i < pi->nqsets; i++, txq++) {
+		u32 name, value;
+		int err;
+
+		name = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DMAQ) |
+			FW_PARAMS_PARAM_X_V(
+				FW_PARAMS_PARAM_DMAQ_EQ_DCBPRIO_ETH) |
+			FW_PARAMS_PARAM_YZ_V(txq->q.cntxt_id));
+		value = enable ? i : 0xffffffff;
+
+		/* Since we can be called while atomic (from "interrupt
+		 * level") we need to issue the Set Parameters Commannd
+		 * without sleeping (timeout < 0).
+		 */
+		err = t4_set_params_timeout(adap, adap->mbox, adap->pf, 0, 1,
+					    &name, &value,
+					    -FW_CMD_MAX_TIMEOUT);
+
+		if (err)
+			dev_err(adap->pdev_dev,
+				"Can't %s DCB Priority on port %d, TX Queue %d: err=%d\n",
+				enable ? "set" : "unset", pi->port_id, i, -err);
+		else
+			txq->dcb_prio = enable ? value : 0;
+	}
+}
+
+int cxgb4_dcb_enabled(const struct net_device *dev)
+{
+	struct port_info *pi = netdev_priv(dev);
+
+	if (!pi->dcb.enabled)
+		return 0;
+
+	return ((pi->dcb.state == CXGB4_DCB_STATE_FW_ALLSYNCED) ||
+		(pi->dcb.state == CXGB4_DCB_STATE_HOST));
+}
+#endif /* CONFIG_CHELSIO_T4_DCB */
+
+void t4_os_link_changed(struct adapter *adapter, int port_id, int link_stat)
+{
+	struct net_device *dev = adapter->port[port_id];
+
+	/* Skip changes from disabled ports. */
+	if (netif_running(dev) && link_stat != netif_carrier_ok(dev)) {
+		if (link_stat)
+			netif_carrier_on(dev);
+		else {
+#ifdef CONFIG_CHELSIO_T4_DCB
+			if (cxgb4_dcb_enabled(dev)) {
+				cxgb4_dcb_reset(dev);
+				dcb_tx_queue_prio_enable(dev, false);
+			}
+#endif /* CONFIG_CHELSIO_T4_DCB */
+			netif_carrier_off(dev);
+		}
+
+		link_report(dev);
+	}
+}
+
+void t4_os_portmod_changed(struct adapter *adap, int port_id)
+{
+	static const char *mod_str[] = {
+		NULL, "LR", "SR", "ER", "passive DA", "active DA", "LRM"
+	};
+
+	struct net_device *dev = adap->port[port_id];
+	struct port_info *pi = netdev_priv(dev);
+
+	if (pi->mod_type == FW_PORT_MOD_TYPE_NONE)
+		netdev_info(dev, "port module unplugged\n");
+	else if (pi->mod_type < ARRAY_SIZE(mod_str))
+		netdev_info(dev, "%s module inserted\n", mod_str[pi->mod_type]);
+	else if (pi->mod_type == FW_PORT_MOD_TYPE_NOTSUPPORTED)
+		netdev_info(dev, "%s: unsupported port module inserted\n",
+			    dev->name);
+	else if (pi->mod_type == FW_PORT_MOD_TYPE_UNKNOWN)
+		netdev_info(dev, "%s: unknown port module inserted\n",
+			    dev->name);
+	else if (pi->mod_type == FW_PORT_MOD_TYPE_ERROR)
+		netdev_info(dev, "%s: transceiver module error\n", dev->name);
+	else
+		netdev_info(dev, "%s: unknown module type %d inserted\n",
+			    dev->name, pi->mod_type);
+
+	/* If the interface is running, then we'll need any "sticky" Link
+	 * Parameters redone with a new Transceiver Module.
+	 */
+	pi->link_cfg.redo_l1cfg = netif_running(dev);
+}
+
+int dbfifo_int_thresh = 10; /* 10 == 640 entry threshold */
+module_param(dbfifo_int_thresh, int, 0644);
+MODULE_PARM_DESC(dbfifo_int_thresh, "doorbell fifo interrupt threshold");
+
+/*
+ * usecs to sleep while draining the dbfifo
+ */
+static int dbfifo_drain_delay = 1000;
+module_param(dbfifo_drain_delay, int, 0644);
+MODULE_PARM_DESC(dbfifo_drain_delay,
+		 "usecs to sleep while draining the dbfifo");
+
+static inline int cxgb4_set_addr_hash(struct port_info *pi)
+{
+	struct adapter *adap = pi->adapter;
+	u64 vec = 0;
+	bool ucast = false;
+	struct hash_mac_addr *entry;
+
+	/* Calculate the hash vector for the updated list and program it */
+	list_for_each_entry(entry, &adap->mac_hlist, list) {
+		ucast |= is_unicast_ether_addr(entry->addr);
+		vec |= (1ULL << hash_mac_addr(entry->addr));
+	}
+	return t4_set_addr_hash(adap, adap->mbox, pi->viid, ucast,
+				vec, false);
+}
+
+static int cxgb4_mac_sync(struct net_device *netdev, const u8 *mac_addr)
+{
+	struct port_info *pi = netdev_priv(netdev);
+	struct adapter *adap = pi->adapter;
+	int ret;
+	u64 mhash = 0;
+	u64 uhash = 0;
+	/* idx stores the index of allocated filters,
+	 * its size should be modified based on the number of
+	 * MAC addresses that we allocate filters for
+	 */
+
+	u16 idx[1] = {};
+	bool free = false;
+	bool ucast = is_unicast_ether_addr(mac_addr);
+	const u8 *maclist[1] = {mac_addr};
+	struct hash_mac_addr *new_entry;
+
+	ret = cxgb4_alloc_mac_filt(adap, pi->viid, free, 1, maclist,
+				   idx, ucast ? &uhash : &mhash, false);
+	if (ret < 0)
+		goto out;
+	/* if hash != 0, then add the addr to hash addr list
+	 * so on the end we will calculate the hash for the
+	 * list and program it
+	 */
+	if (uhash || mhash) {
+		new_entry = kzalloc(sizeof(*new_entry), GFP_ATOMIC);
+		if (!new_entry)
+			return -ENOMEM;
+		ether_addr_copy(new_entry->addr, mac_addr);
+		list_add_tail(&new_entry->list, &adap->mac_hlist);
+		ret = cxgb4_set_addr_hash(pi);
+	}
+out:
+	return ret < 0 ? ret : 0;
+}
+
+static int cxgb4_mac_unsync(struct net_device *netdev, const u8 *mac_addr)
+{
+	struct port_info *pi = netdev_priv(netdev);
+	struct adapter *adap = pi->adapter;
+	int ret;
+	const u8 *maclist[1] = {mac_addr};
+	struct hash_mac_addr *entry, *tmp;
+
+	/* If the MAC address to be removed is in the hash addr
+	 * list, delete it from the list and update hash vector
+	 */
+	list_for_each_entry_safe(entry, tmp, &adap->mac_hlist, list) {
+		if (ether_addr_equal(entry->addr, mac_addr)) {
+			list_del(&entry->list);
+			kfree(entry);
+			return cxgb4_set_addr_hash(pi);
+		}
+	}
+
+	ret = cxgb4_free_mac_filt(adap, pi->viid, 1, maclist, false);
+	return ret < 0 ? -EINVAL : 0;
+}
+
+/*
+ * Set Rx properties of a port, such as promiscruity, address filters, and MTU.
+ * If @mtu is -1 it is left unchanged.
+ */
+static int set_rxmode(struct net_device *dev, int mtu, bool sleep_ok)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adapter = pi->adapter;
+
+	__dev_uc_sync(dev, cxgb4_mac_sync, cxgb4_mac_unsync);
+	__dev_mc_sync(dev, cxgb4_mac_sync, cxgb4_mac_unsync);
+
+	return t4_set_rxmode(adapter, adapter->mbox, pi->viid, pi->viid_mirror,
+			     mtu, (dev->flags & IFF_PROMISC) ? 1 : 0,
+			     (dev->flags & IFF_ALLMULTI) ? 1 : 0, 1, -1,
+			     sleep_ok);
+}
+
+/**
+ *	cxgb4_change_mac - Update match filter for a MAC address.
+ *	@pi: the port_info
+ *	@viid: the VI id
+ *	@tcam_idx: TCAM index of existing filter for old value of MAC address,
+ *		   or -1
+ *	@addr: the new MAC address value
+ *	@persist: whether a new MAC allocation should be persistent
+ *	@smt_idx: the destination to store the new SMT index.
+ *
+ *	Modifies an MPS filter and sets it to the new MAC address if
+ *	@tcam_idx >= 0, or adds the MAC address to a new filter if
+ *	@tcam_idx < 0. In the latter case the address is added persistently
+ *	if @persist is %true.
+ *	Addresses are programmed to hash region, if tcam runs out of entries.
+ *
+ */
+int cxgb4_change_mac(struct port_info *pi, unsigned int viid,
+		     int *tcam_idx, const u8 *addr, bool persist,
+		     u8 *smt_idx)
+{
+	struct adapter *adapter = pi->adapter;
+	struct hash_mac_addr *entry, *new_entry;
+	int ret;
+
+	ret = t4_change_mac(adapter, adapter->mbox, viid,
+			    *tcam_idx, addr, persist, smt_idx);
+	/* We ran out of TCAM entries. try programming hash region. */
+	if (ret == -ENOMEM) {
+		/* If the MAC address to be updated is in the hash addr
+		 * list, update it from the list
+		 */
+		list_for_each_entry(entry, &adapter->mac_hlist, list) {
+			if (entry->iface_mac) {
+				ether_addr_copy(entry->addr, addr);
+				goto set_hash;
+			}
+		}
+		new_entry = kzalloc(sizeof(*new_entry), GFP_KERNEL);
+		if (!new_entry)
+			return -ENOMEM;
+		ether_addr_copy(new_entry->addr, addr);
+		new_entry->iface_mac = true;
+		list_add_tail(&new_entry->list, &adapter->mac_hlist);
+set_hash:
+		ret = cxgb4_set_addr_hash(pi);
+	} else if (ret >= 0) {
+		*tcam_idx = ret;
+		ret = 0;
+	}
+
+	return ret;
+}
+
+/*
+ *	link_start - enable a port
+ *	@dev: the port to enable
+ *
+ *	Performs the MAC and PHY actions needed to enable a port.
+ */
+static int link_start(struct net_device *dev)
+{
+	struct port_info *pi = netdev_priv(dev);
+	unsigned int mb = pi->adapter->mbox;
+	int ret;
+
+	/*
+	 * We do not set address filters and promiscuity here, the stack does
+	 * that step explicitly.
+	 */
+	ret = t4_set_rxmode(pi->adapter, mb, pi->viid, pi->viid_mirror,
+			    dev->mtu, -1, -1, -1,
+			    !!(dev->features & NETIF_F_HW_VLAN_CTAG_RX), true);
+	if (ret == 0)
+		ret = cxgb4_update_mac_filt(pi, pi->viid, &pi->xact_addr_filt,
+					    dev->dev_addr, true, &pi->smt_idx);
+	if (ret == 0)
+		ret = t4_link_l1cfg(pi->adapter, mb, pi->tx_chan,
+				    &pi->link_cfg);
+	if (ret == 0) {
+		local_bh_disable();
+		ret = t4_enable_pi_params(pi->adapter, mb, pi, true,
+					  true, CXGB4_DCB_ENABLED);
+		local_bh_enable();
+	}
+
+	return ret;
+}
+
+#ifdef CONFIG_CHELSIO_T4_DCB
+/* Handle a Data Center Bridging update message from the firmware. */
+static void dcb_rpl(struct adapter *adap, const struct fw_port_cmd *pcmd)
+{
+	int port = FW_PORT_CMD_PORTID_G(ntohl(pcmd->op_to_portid));
+	struct net_device *dev = adap->port[adap->chan_map[port]];
+	int old_dcb_enabled = cxgb4_dcb_enabled(dev);
+	int new_dcb_enabled;
+
+	cxgb4_dcb_handle_fw_update(adap, pcmd);
+	new_dcb_enabled = cxgb4_dcb_enabled(dev);
+
+	/* If the DCB has become enabled or disabled on the port then we're
+	 * going to need to set up/tear down DCB Priority parameters for the
+	 * TX Queues associated with the port.
+	 */
+	if (new_dcb_enabled != old_dcb_enabled)
+		dcb_tx_queue_prio_enable(dev, new_dcb_enabled);
+}
+#endif /* CONFIG_CHELSIO_T4_DCB */
+
+/* Response queue handler for the FW event queue.
+ */
+static int fwevtq_handler(struct sge_rspq *q, const __be64 *rsp,
+			  const struct pkt_gl *gl)
+{
+	u8 opcode = ((const struct rss_header *)rsp)->opcode;
+
+	rsp++;                                          /* skip RSS header */
+
+	/* FW can send EGR_UPDATEs encapsulated in a CPL_FW4_MSG.
+	 */
+	if (unlikely(opcode == CPL_FW4_MSG &&
+	   ((const struct cpl_fw4_msg *)rsp)->type == FW_TYPE_RSSCPL)) {
+		rsp++;
+		opcode = ((const struct rss_header *)rsp)->opcode;
+		rsp++;
+		if (opcode != CPL_SGE_EGR_UPDATE) {
+			dev_err(q->adap->pdev_dev, "unexpected FW4/CPL %#x on FW event queue\n"
+				, opcode);
+			goto out;
+		}
+	}
+
+	if (likely(opcode == CPL_SGE_EGR_UPDATE)) {
+		const struct cpl_sge_egr_update *p = (void *)rsp;
+		unsigned int qid = EGR_QID_G(ntohl(p->opcode_qid));
+		struct sge_txq *txq;
+
+		txq = q->adap->sge.egr_map[qid - q->adap->sge.egr_start];
+		txq->restarts++;
+		if (txq->q_type == CXGB4_TXQ_ETH) {
+			struct sge_eth_txq *eq;
+
+			eq = container_of(txq, struct sge_eth_txq, q);
+			t4_sge_eth_txq_egress_update(q->adap, eq, -1);
+		} else {
+			struct sge_uld_txq *oq;
+
+			oq = container_of(txq, struct sge_uld_txq, q);
+			tasklet_schedule(&oq->qresume_tsk);
+		}
+	} else if (opcode == CPL_FW6_MSG || opcode == CPL_FW4_MSG) {
+		const struct cpl_fw6_msg *p = (void *)rsp;
+
+#ifdef CONFIG_CHELSIO_T4_DCB
+		const struct fw_port_cmd *pcmd = (const void *)p->data;
+		unsigned int cmd = FW_CMD_OP_G(ntohl(pcmd->op_to_portid));
+		unsigned int action =
+			FW_PORT_CMD_ACTION_G(ntohl(pcmd->action_to_len16));
+
+		if (cmd == FW_PORT_CMD &&
+		    (action == FW_PORT_ACTION_GET_PORT_INFO ||
+		     action == FW_PORT_ACTION_GET_PORT_INFO32)) {
+			int port = FW_PORT_CMD_PORTID_G(
+					be32_to_cpu(pcmd->op_to_portid));
+			struct net_device *dev;
+			int dcbxdis, state_input;
+
+			dev = q->adap->port[q->adap->chan_map[port]];
+			dcbxdis = (action == FW_PORT_ACTION_GET_PORT_INFO
+			  ? !!(pcmd->u.info.dcbxdis_pkd & FW_PORT_CMD_DCBXDIS_F)
+			  : !!(be32_to_cpu(pcmd->u.info32.lstatus32_to_cbllen32)
+			       & FW_PORT_CMD_DCBXDIS32_F));
+			state_input = (dcbxdis
+				       ? CXGB4_DCB_INPUT_FW_DISABLED
+				       : CXGB4_DCB_INPUT_FW_ENABLED);
+
+			cxgb4_dcb_state_fsm(dev, state_input);
+		}
+
+		if (cmd == FW_PORT_CMD &&
+		    action == FW_PORT_ACTION_L2_DCB_CFG)
+			dcb_rpl(q->adap, pcmd);
+		else
+#endif
+			if (p->type == 0)
+				t4_handle_fw_rpl(q->adap, p->data);
+	} else if (opcode == CPL_L2T_WRITE_RPL) {
+		const struct cpl_l2t_write_rpl *p = (void *)rsp;
+
+		do_l2t_write_rpl(q->adap, p);
+	} else if (opcode == CPL_SMT_WRITE_RPL) {
+		const struct cpl_smt_write_rpl *p = (void *)rsp;
+
+		do_smt_write_rpl(q->adap, p);
+	} else if (opcode == CPL_SET_TCB_RPL) {
+		const struct cpl_set_tcb_rpl *p = (void *)rsp;
+
+		filter_rpl(q->adap, p);
+	} else if (opcode == CPL_ACT_OPEN_RPL) {
+		const struct cpl_act_open_rpl *p = (void *)rsp;
+
+		hash_filter_rpl(q->adap, p);
+	} else if (opcode == CPL_ABORT_RPL_RSS) {
+		const struct cpl_abort_rpl_rss *p = (void *)rsp;
+
+		hash_del_filter_rpl(q->adap, p);
+	} else if (opcode == CPL_SRQ_TABLE_RPL) {
+		const struct cpl_srq_table_rpl *p = (void *)rsp;
+
+		do_srq_table_rpl(q->adap, p);
+	} else
+		dev_err(q->adap->pdev_dev,
+			"unexpected CPL %#x on FW event queue\n", opcode);
+out:
+	return 0;
+}
+
+static void disable_msi(struct adapter *adapter)
+{
+	if (adapter->flags & CXGB4_USING_MSIX) {
+		pci_disable_msix(adapter->pdev);
+		adapter->flags &= ~CXGB4_USING_MSIX;
+	} else if (adapter->flags & CXGB4_USING_MSI) {
+		pci_disable_msi(adapter->pdev);
+		adapter->flags &= ~CXGB4_USING_MSI;
+	}
+}
+
+/*
+ * Interrupt handler for non-data events used with MSI-X.
+ */
+static irqreturn_t t4_nondata_intr(int irq, void *cookie)
+{
+	struct adapter *adap = cookie;
+	u32 v = t4_read_reg(adap, MYPF_REG(PL_PF_INT_CAUSE_A));
+
+	if (v & PFSW_F) {
+		adap->swintr = 1;
+		t4_write_reg(adap, MYPF_REG(PL_PF_INT_CAUSE_A), v);
+	}
+	if (adap->flags & CXGB4_MASTER_PF)
+		t4_slow_intr_handler(adap);
+	return IRQ_HANDLED;
+}
+
+int cxgb4_set_msix_aff(struct adapter *adap, unsigned short vec,
+		       cpumask_var_t *aff_mask, int idx)
+{
+	int rv;
+
+	if (!zalloc_cpumask_var(aff_mask, GFP_KERNEL)) {
+		dev_err(adap->pdev_dev, "alloc_cpumask_var failed\n");
+		return -ENOMEM;
+	}
+
+	cpumask_set_cpu(cpumask_local_spread(idx, dev_to_node(adap->pdev_dev)),
+			*aff_mask);
+
+	rv = irq_set_affinity_hint(vec, *aff_mask);
+	if (rv)
+		dev_warn(adap->pdev_dev,
+			 "irq_set_affinity_hint %u failed %d\n",
+			 vec, rv);
+
+	return 0;
+}
+
+void cxgb4_clear_msix_aff(unsigned short vec, cpumask_var_t aff_mask)
+{
+	irq_set_affinity_hint(vec, NULL);
+	free_cpumask_var(aff_mask);
+}
+
+static int request_msix_queue_irqs(struct adapter *adap)
+{
+	struct sge *s = &adap->sge;
+	struct msix_info *minfo;
+	int err, ethqidx;
+
+	if (s->fwevtq_msix_idx < 0)
+		return -ENOMEM;
+
+	err = request_irq(adap->msix_info[s->fwevtq_msix_idx].vec,
+			  t4_sge_intr_msix, 0,
+			  adap->msix_info[s->fwevtq_msix_idx].desc,
+			  &s->fw_evtq);
+	if (err)
+		return err;
+
+	for_each_ethrxq(s, ethqidx) {
+		minfo = s->ethrxq[ethqidx].msix;
+		err = request_irq(minfo->vec,
+				  t4_sge_intr_msix, 0,
+				  minfo->desc,
+				  &s->ethrxq[ethqidx].rspq);
+		if (err)
+			goto unwind;
+
+		cxgb4_set_msix_aff(adap, minfo->vec,
+				   &minfo->aff_mask, ethqidx);
+	}
+	return 0;
+
+unwind:
+	while (--ethqidx >= 0) {
+		minfo = s->ethrxq[ethqidx].msix;
+		cxgb4_clear_msix_aff(minfo->vec, minfo->aff_mask);
+		free_irq(minfo->vec, &s->ethrxq[ethqidx].rspq);
+	}
+	free_irq(adap->msix_info[s->fwevtq_msix_idx].vec, &s->fw_evtq);
+	return err;
+}
+
+static void free_msix_queue_irqs(struct adapter *adap)
+{
+	struct sge *s = &adap->sge;
+	struct msix_info *minfo;
+	int i;
+
+	free_irq(adap->msix_info[s->fwevtq_msix_idx].vec, &s->fw_evtq);
+	for_each_ethrxq(s, i) {
+		minfo = s->ethrxq[i].msix;
+		cxgb4_clear_msix_aff(minfo->vec, minfo->aff_mask);
+		free_irq(minfo->vec, &s->ethrxq[i].rspq);
+	}
+}
+
+static int setup_ppod_edram(struct adapter *adap)
+{
+	unsigned int param, val;
+	int ret;
+
+	/* Driver sends FW_PARAMS_PARAM_DEV_PPOD_EDRAM read command to check
+	 * if firmware supports ppod edram feature or not. If firmware
+	 * returns 1, then driver can enable this feature by sending
+	 * FW_PARAMS_PARAM_DEV_PPOD_EDRAM write command with value 1 to
+	 * enable ppod edram feature.
+	 */
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_PPOD_EDRAM));
+
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 1, &param, &val);
+	if (ret < 0) {
+		dev_warn(adap->pdev_dev,
+			 "querying PPOD_EDRAM support failed: %d\n",
+			 ret);
+		return -1;
+	}
+
+	if (val != 1)
+		return -1;
+
+	ret = t4_set_params(adap, adap->mbox, adap->pf, 0, 1, &param, &val);
+	if (ret < 0) {
+		dev_err(adap->pdev_dev,
+			"setting PPOD_EDRAM failed: %d\n", ret);
+		return -1;
+	}
+	return 0;
+}
+
+static void adap_config_hpfilter(struct adapter *adapter)
+{
+	u32 param, val = 0;
+	int ret;
+
+	/* Enable HP filter region. Older fw will fail this request and
+	 * it is fine.
+	 */
+	param = FW_PARAM_DEV(HPFILTER_REGION_SUPPORT);
+	ret = t4_set_params(adapter, adapter->mbox, adapter->pf, 0,
+			    1, &param, &val);
+
+	/* An error means FW doesn't know about HP filter support,
+	 * it's not a problem, don't return an error.
+	 */
+	if (ret < 0)
+		dev_err(adapter->pdev_dev,
+			"HP filter region isn't supported by FW\n");
+}
+
+static int cxgb4_config_rss(const struct port_info *pi, u16 *rss,
+			    u16 rss_size, u16 viid)
+{
+	struct adapter *adap = pi->adapter;
+	int ret;
+
+	ret = t4_config_rss_range(adap, adap->mbox, viid, 0, rss_size, rss,
+				  rss_size);
+	if (ret)
+		return ret;
+
+	/* If Tunnel All Lookup isn't specified in the global RSS
+	 * Configuration, then we need to specify a default Ingress
+	 * Queue for any ingress packets which aren't hashed.  We'll
+	 * use our first ingress queue ...
+	 */
+	return t4_config_vi_rss(adap, adap->mbox, viid,
+				FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN_F |
+				FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN_F |
+				FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN_F |
+				FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN_F |
+				FW_RSS_VI_CONFIG_CMD_UDPEN_F,
+				rss[0]);
+}
+
+/**
+ *	cxgb4_write_rss - write the RSS table for a given port
+ *	@pi: the port
+ *	@queues: array of queue indices for RSS
+ *
+ *	Sets up the portion of the HW RSS table for the port's VI to distribute
+ *	packets to the Rx queues in @queues.
+ *	Should never be called before setting up sge eth rx queues
+ */
+int cxgb4_write_rss(const struct port_info *pi, const u16 *queues)
+{
+	struct adapter *adapter = pi->adapter;
+	const struct sge_eth_rxq *rxq;
+	int i, err;
+	u16 *rss;
+
+	rxq = &adapter->sge.ethrxq[pi->first_qset];
+	rss = kmalloc_array(pi->rss_size, sizeof(u16), GFP_KERNEL);
+	if (!rss)
+		return -ENOMEM;
+
+	/* map the queue indices to queue ids */
+	for (i = 0; i < pi->rss_size; i++, queues++)
+		rss[i] = rxq[*queues].rspq.abs_id;
+
+	err = cxgb4_config_rss(pi, rss, pi->rss_size, pi->viid);
+	kfree(rss);
+	return err;
+}
+
+/**
+ *	setup_rss - configure RSS
+ *	@adap: the adapter
+ *
+ *	Sets up RSS for each port.
+ */
+static int setup_rss(struct adapter *adap)
+{
+	int i, j, err;
+
+	for_each_port(adap, i) {
+		const struct port_info *pi = adap2pinfo(adap, i);
+
+		/* Fill default values with equal distribution */
+		for (j = 0; j < pi->rss_size; j++)
+			pi->rss[j] = j % pi->nqsets;
+
+		err = cxgb4_write_rss(pi, pi->rss);
+		if (err)
+			return err;
+	}
+	return 0;
+}
+
+/*
+ * Return the channel of the ingress queue with the given qid.
+ */
+static unsigned int rxq_to_chan(const struct sge *p, unsigned int qid)
+{
+	qid -= p->ingr_start;
+	return netdev2pinfo(p->ingr_map[qid]->netdev)->tx_chan;
+}
+
+void cxgb4_quiesce_rx(struct sge_rspq *q)
+{
+	if (q->handler)
+		napi_disable(&q->napi);
+}
+
+/*
+ * Wait until all NAPI handlers are descheduled.
+ */
+static void quiesce_rx(struct adapter *adap)
+{
+	int i;
+
+	for (i = 0; i < adap->sge.ingr_sz; i++) {
+		struct sge_rspq *q = adap->sge.ingr_map[i];
+
+		if (!q)
+			continue;
+
+		cxgb4_quiesce_rx(q);
+	}
+}
+
+/* Disable interrupt and napi handler */
+static void disable_interrupts(struct adapter *adap)
+{
+	struct sge *s = &adap->sge;
+
+	if (adap->flags & CXGB4_FULL_INIT_DONE) {
+		t4_intr_disable(adap);
+		if (adap->flags & CXGB4_USING_MSIX) {
+			free_msix_queue_irqs(adap);
+			free_irq(adap->msix_info[s->nd_msix_idx].vec,
+				 adap);
+		} else {
+			free_irq(adap->pdev->irq, adap);
+		}
+		quiesce_rx(adap);
+	}
+}
+
+void cxgb4_enable_rx(struct adapter *adap, struct sge_rspq *q)
+{
+	if (q->handler)
+		napi_enable(&q->napi);
+
+	/* 0-increment GTS to start the timer and enable interrupts */
+	t4_write_reg(adap, MYPF_REG(SGE_PF_GTS_A),
+		     SEINTARM_V(q->intr_params) |
+		     INGRESSQID_V(q->cntxt_id));
+}
+
+/*
+ * Enable NAPI scheduling and interrupt generation for all Rx queues.
+ */
+static void enable_rx(struct adapter *adap)
+{
+	int i;
+
+	for (i = 0; i < adap->sge.ingr_sz; i++) {
+		struct sge_rspq *q = adap->sge.ingr_map[i];
+
+		if (!q)
+			continue;
+
+		cxgb4_enable_rx(adap, q);
+	}
+}
+
+static int setup_non_data_intr(struct adapter *adap)
+{
+	int msix;
+
+	adap->sge.nd_msix_idx = -1;
+	if (!(adap->flags & CXGB4_USING_MSIX))
+		return 0;
+
+	/* Request MSI-X vector for non-data interrupt */
+	msix = cxgb4_get_msix_idx_from_bmap(adap);
+	if (msix < 0)
+		return -ENOMEM;
+
+	snprintf(adap->msix_info[msix].desc,
+		 sizeof(adap->msix_info[msix].desc),
+		 "%s", adap->port[0]->name);
+
+	adap->sge.nd_msix_idx = msix;
+	return 0;
+}
+
+static int setup_fw_sge_queues(struct adapter *adap)
+{
+	struct sge *s = &adap->sge;
+	int msix, err = 0;
+
+	bitmap_zero(s->starving_fl, s->egr_sz);
+	bitmap_zero(s->txq_maperr, s->egr_sz);
+
+	if (adap->flags & CXGB4_USING_MSIX) {
+		s->fwevtq_msix_idx = -1;
+		msix = cxgb4_get_msix_idx_from_bmap(adap);
+		if (msix < 0)
+			return -ENOMEM;
+
+		snprintf(adap->msix_info[msix].desc,
+			 sizeof(adap->msix_info[msix].desc),
+			 "%s-FWeventq", adap->port[0]->name);
+	} else {
+		err = t4_sge_alloc_rxq(adap, &s->intrq, false, adap->port[0], 0,
+				       NULL, NULL, NULL, -1);
+		if (err)
+			return err;
+		msix = -((int)s->intrq.abs_id + 1);
+	}
+
+	err = t4_sge_alloc_rxq(adap, &s->fw_evtq, true, adap->port[0],
+			       msix, NULL, fwevtq_handler, NULL, -1);
+	if (err && msix >= 0)
+		cxgb4_free_msix_idx_in_bmap(adap, msix);
+
+	s->fwevtq_msix_idx = msix;
+	return err;
+}
+
+/**
+ *	setup_sge_queues - configure SGE Tx/Rx/response queues
+ *	@adap: the adapter
+ *
+ *	Determines how many sets of SGE queues to use and initializes them.
+ *	We support multiple queue sets per port if we have MSI-X, otherwise
+ *	just one queue set per port.
+ */
+static int setup_sge_queues(struct adapter *adap)
+{
+	struct sge_uld_rxq_info *rxq_info = NULL;
+	struct sge *s = &adap->sge;
+	unsigned int cmplqid = 0;
+	int err, i, j, msix = 0;
+
+	if (is_uld(adap))
+		rxq_info = s->uld_rxq_info[CXGB4_ULD_RDMA];
+
+	if (!(adap->flags & CXGB4_USING_MSIX))
+		msix = -((int)s->intrq.abs_id + 1);
+
+	for_each_port(adap, i) {
+		struct net_device *dev = adap->port[i];
+		struct port_info *pi = netdev_priv(dev);
+		struct sge_eth_rxq *q = &s->ethrxq[pi->first_qset];
+		struct sge_eth_txq *t = &s->ethtxq[pi->first_qset];
+
+		for (j = 0; j < pi->nqsets; j++, q++) {
+			if (msix >= 0) {
+				msix = cxgb4_get_msix_idx_from_bmap(adap);
+				if (msix < 0) {
+					err = msix;
+					goto freeout;
+				}
+
+				snprintf(adap->msix_info[msix].desc,
+					 sizeof(adap->msix_info[msix].desc),
+					 "%s-Rx%d", dev->name, j);
+				q->msix = &adap->msix_info[msix];
+			}
+
+			err = t4_sge_alloc_rxq(adap, &q->rspq, false, dev,
+					       msix, &q->fl,
+					       t4_ethrx_handler,
+					       NULL,
+					       t4_get_tp_ch_map(adap,
+								pi->tx_chan));
+			if (err)
+				goto freeout;
+			q->rspq.idx = j;
+			memset(&q->stats, 0, sizeof(q->stats));
+		}
+
+		q = &s->ethrxq[pi->first_qset];
+		for (j = 0; j < pi->nqsets; j++, t++, q++) {
+			err = t4_sge_alloc_eth_txq(adap, t, dev,
+					netdev_get_tx_queue(dev, j),
+					q->rspq.cntxt_id,
+					!!(adap->flags & CXGB4_SGE_DBQ_TIMER));
+			if (err)
+				goto freeout;
+		}
+	}
+
+	for_each_port(adap, i) {
+		/* Note that cmplqid below is 0 if we don't
+		 * have RDMA queues, and that's the right value.
+		 */
+		if (rxq_info)
+			cmplqid	= rxq_info->uldrxq[i].rspq.cntxt_id;
+
+		err = t4_sge_alloc_ctrl_txq(adap, &s->ctrlq[i], adap->port[i],
+					    s->fw_evtq.cntxt_id, cmplqid);
+		if (err)
+			goto freeout;
+	}
+
+	if (!is_t4(adap->params.chip)) {
+		err = t4_sge_alloc_eth_txq(adap, &s->ptptxq, adap->port[0],
+					   netdev_get_tx_queue(adap->port[0], 0)
+					   , s->fw_evtq.cntxt_id, false);
+		if (err)
+			goto freeout;
+	}
+
+	t4_write_reg(adap, is_t4(adap->params.chip) ?
+				MPS_TRC_RSS_CONTROL_A :
+				MPS_T5_TRC_RSS_CONTROL_A,
+		     RSSCONTROL_V(netdev2pinfo(adap->port[0])->tx_chan) |
+		     QUEUENUMBER_V(s->ethrxq[0].rspq.abs_id));
+	return 0;
+freeout:
+	dev_err(adap->pdev_dev, "Can't allocate queues, err=%d\n", -err);
+	t4_free_sge_resources(adap);
+	return err;
+}
+
+static u16 cxgb_select_queue(struct net_device *dev, struct sk_buff *skb,
+			     struct net_device *sb_dev)
+{
+	int txq;
+
+#ifdef CONFIG_CHELSIO_T4_DCB
+	/* If a Data Center Bridging has been successfully negotiated on this
+	 * link then we'll use the skb's priority to map it to a TX Queue.
+	 * The skb's priority is determined via the VLAN Tag Priority Code
+	 * Point field.
+	 */
+	if (cxgb4_dcb_enabled(dev) && !is_kdump_kernel()) {
+		u16 vlan_tci;
+		int err;
+
+		err = vlan_get_tag(skb, &vlan_tci);
+		if (unlikely(err)) {
+			if (net_ratelimit())
+				netdev_warn(dev,
+					    "TX Packet without VLAN Tag on DCB Link\n");
+			txq = 0;
+		} else {
+			txq = (vlan_tci & VLAN_PRIO_MASK) >> VLAN_PRIO_SHIFT;
+#ifdef CONFIG_CHELSIO_T4_FCOE
+			if (skb->protocol == htons(ETH_P_FCOE))
+				txq = skb->priority & 0x7;
+#endif /* CONFIG_CHELSIO_T4_FCOE */
+		}
+		return txq;
+	}
+#endif /* CONFIG_CHELSIO_T4_DCB */
+
+	if (dev->num_tc) {
+		struct port_info *pi = netdev2pinfo(dev);
+		u8 ver, proto;
+
+		ver = ip_hdr(skb)->version;
+		proto = (ver == 6) ? ipv6_hdr(skb)->nexthdr :
+				     ip_hdr(skb)->protocol;
+
+		/* Send unsupported traffic pattern to normal NIC queues. */
+		txq = netdev_pick_tx(dev, skb, sb_dev);
+		if (xfrm_offload(skb) || is_ptp_enabled(skb, dev) ||
+		    skb->encapsulation ||
+		    cxgb4_is_ktls_skb(skb) ||
+		    (proto != IPPROTO_TCP && proto != IPPROTO_UDP))
+			txq = txq % pi->nqsets;
+
+		return txq;
+	}
+
+	if (select_queue) {
+		txq = (skb_rx_queue_recorded(skb)
+			? skb_get_rx_queue(skb)
+			: smp_processor_id());
+
+		while (unlikely(txq >= dev->real_num_tx_queues))
+			txq -= dev->real_num_tx_queues;
+
+		return txq;
+	}
+
+	return netdev_pick_tx(dev, skb, NULL) % dev->real_num_tx_queues;
+}
+
+static int closest_timer(const struct sge *s, int time)
+{
+	int i, delta, match = 0, min_delta = INT_MAX;
+
+	for (i = 0; i < ARRAY_SIZE(s->timer_val); i++) {
+		delta = time - s->timer_val[i];
+		if (delta < 0)
+			delta = -delta;
+		if (delta < min_delta) {
+			min_delta = delta;
+			match = i;
+		}
+	}
+	return match;
+}
+
+static int closest_thres(const struct sge *s, int thres)
+{
+	int i, delta, match = 0, min_delta = INT_MAX;
+
+	for (i = 0; i < ARRAY_SIZE(s->counter_val); i++) {
+		delta = thres - s->counter_val[i];
+		if (delta < 0)
+			delta = -delta;
+		if (delta < min_delta) {
+			min_delta = delta;
+			match = i;
+		}
+	}
+	return match;
+}
+
+/**
+ *	cxgb4_set_rspq_intr_params - set a queue's interrupt holdoff parameters
+ *	@q: the Rx queue
+ *	@us: the hold-off time in us, or 0 to disable timer
+ *	@cnt: the hold-off packet count, or 0 to disable counter
+ *
+ *	Sets an Rx queue's interrupt hold-off time and packet count.  At least
+ *	one of the two needs to be enabled for the queue to generate interrupts.
+ */
+int cxgb4_set_rspq_intr_params(struct sge_rspq *q,
+			       unsigned int us, unsigned int cnt)
+{
+	struct adapter *adap = q->adap;
+
+	if ((us | cnt) == 0)
+		cnt = 1;
+
+	if (cnt) {
+		int err;
+		u32 v, new_idx;
+
+		new_idx = closest_thres(&adap->sge, cnt);
+		if (q->desc && q->pktcnt_idx != new_idx) {
+			/* the queue has already been created, update it */
+			v = FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DMAQ) |
+			    FW_PARAMS_PARAM_X_V(
+					FW_PARAMS_PARAM_DMAQ_IQ_INTCNTTHRESH) |
+			    FW_PARAMS_PARAM_YZ_V(q->cntxt_id);
+			err = t4_set_params(adap, adap->mbox, adap->pf, 0, 1,
+					    &v, &new_idx);
+			if (err)
+				return err;
+		}
+		q->pktcnt_idx = new_idx;
+	}
+
+	us = us == 0 ? 6 : closest_timer(&adap->sge, us);
+	q->intr_params = QINTR_TIMER_IDX_V(us) | QINTR_CNT_EN_V(cnt > 0);
+	return 0;
+}
+
+static int cxgb_set_features(struct net_device *dev, netdev_features_t features)
+{
+	netdev_features_t changed = dev->features ^ features;
+	const struct port_info *pi = netdev_priv(dev);
+	int err;
+
+	if (!(changed & NETIF_F_HW_VLAN_CTAG_RX))
+		return 0;
+
+	err = t4_set_rxmode(pi->adapter, pi->adapter->mbox, pi->viid,
+			    pi->viid_mirror, -1, -1, -1, -1,
+			    !!(features & NETIF_F_HW_VLAN_CTAG_RX), true);
+	if (unlikely(err))
+		dev->features = features ^ NETIF_F_HW_VLAN_CTAG_RX;
+	return err;
+}
+
+static int setup_debugfs(struct adapter *adap)
+{
+	if (IS_ERR_OR_NULL(adap->debugfs_root))
+		return -1;
+
+#ifdef CONFIG_DEBUG_FS
+	t4_setup_debugfs(adap);
+#endif
+	return 0;
+}
+
+static void cxgb4_port_mirror_free_rxq(struct adapter *adap,
+				       struct sge_eth_rxq *mirror_rxq)
+{
+	if ((adap->flags & CXGB4_FULL_INIT_DONE) &&
+	    !(adap->flags & CXGB4_SHUTTING_DOWN))
+		cxgb4_quiesce_rx(&mirror_rxq->rspq);
+
+	if (adap->flags & CXGB4_USING_MSIX) {
+		cxgb4_clear_msix_aff(mirror_rxq->msix->vec,
+				     mirror_rxq->msix->aff_mask);
+		free_irq(mirror_rxq->msix->vec, &mirror_rxq->rspq);
+		cxgb4_free_msix_idx_in_bmap(adap, mirror_rxq->msix->idx);
+	}
+
+	free_rspq_fl(adap, &mirror_rxq->rspq, &mirror_rxq->fl);
+}
+
+static int cxgb4_port_mirror_alloc_queues(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct sge_eth_rxq *mirror_rxq;
+	struct sge *s = &adap->sge;
+	int ret = 0, msix = 0;
+	u16 i, rxqid;
+	u16 *rss;
+
+	if (!pi->vi_mirror_count)
+		return 0;
+
+	if (s->mirror_rxq[pi->port_id])
+		return 0;
+
+	mirror_rxq = kcalloc(pi->nmirrorqsets, sizeof(*mirror_rxq), GFP_KERNEL);
+	if (!mirror_rxq)
+		return -ENOMEM;
+
+	s->mirror_rxq[pi->port_id] = mirror_rxq;
+
+	if (!(adap->flags & CXGB4_USING_MSIX))
+		msix = -((int)adap->sge.intrq.abs_id + 1);
+
+	for (i = 0, rxqid = 0; i < pi->nmirrorqsets; i++, rxqid++) {
+		mirror_rxq = &s->mirror_rxq[pi->port_id][i];
+
+		/* Allocate Mirror Rxqs */
+		if (msix >= 0) {
+			msix = cxgb4_get_msix_idx_from_bmap(adap);
+			if (msix < 0) {
+				ret = msix;
+				goto out_free_queues;
+			}
+
+			mirror_rxq->msix = &adap->msix_info[msix];
+			snprintf(mirror_rxq->msix->desc,
+				 sizeof(mirror_rxq->msix->desc),
+				 "%s-mirrorrxq%d", dev->name, i);
+		}
+
+		init_rspq(adap, &mirror_rxq->rspq,
+			  CXGB4_MIRROR_RXQ_DEFAULT_INTR_USEC,
+			  CXGB4_MIRROR_RXQ_DEFAULT_PKT_CNT,
+			  CXGB4_MIRROR_RXQ_DEFAULT_DESC_NUM,
+			  CXGB4_MIRROR_RXQ_DEFAULT_DESC_SIZE);
+
+		mirror_rxq->fl.size = CXGB4_MIRROR_FLQ_DEFAULT_DESC_NUM;
+
+		ret = t4_sge_alloc_rxq(adap, &mirror_rxq->rspq, false,
+				       dev, msix, &mirror_rxq->fl,
+				       t4_ethrx_handler, NULL, 0);
+		if (ret)
+			goto out_free_msix_idx;
+
+		/* Setup MSI-X vectors for Mirror Rxqs */
+		if (adap->flags & CXGB4_USING_MSIX) {
+			ret = request_irq(mirror_rxq->msix->vec,
+					  t4_sge_intr_msix, 0,
+					  mirror_rxq->msix->desc,
+					  &mirror_rxq->rspq);
+			if (ret)
+				goto out_free_rxq;
+
+			cxgb4_set_msix_aff(adap, mirror_rxq->msix->vec,
+					   &mirror_rxq->msix->aff_mask, i);
+		}
+
+		/* Start NAPI for Mirror Rxqs */
+		cxgb4_enable_rx(adap, &mirror_rxq->rspq);
+	}
+
+	/* Setup RSS for Mirror Rxqs */
+	rss = kcalloc(pi->rss_size, sizeof(u16), GFP_KERNEL);
+	if (!rss) {
+		ret = -ENOMEM;
+		goto out_free_queues;
+	}
+
+	mirror_rxq = &s->mirror_rxq[pi->port_id][0];
+	for (i = 0; i < pi->rss_size; i++)
+		rss[i] = mirror_rxq[i % pi->nmirrorqsets].rspq.abs_id;
+
+	ret = cxgb4_config_rss(pi, rss, pi->rss_size, pi->viid_mirror);
+	kfree(rss);
+	if (ret)
+		goto out_free_queues;
+
+	return 0;
+
+out_free_rxq:
+	free_rspq_fl(adap, &mirror_rxq->rspq, &mirror_rxq->fl);
+
+out_free_msix_idx:
+	cxgb4_free_msix_idx_in_bmap(adap, mirror_rxq->msix->idx);
+
+out_free_queues:
+	while (rxqid-- > 0)
+		cxgb4_port_mirror_free_rxq(adap,
+					   &s->mirror_rxq[pi->port_id][rxqid]);
+
+	kfree(s->mirror_rxq[pi->port_id]);
+	s->mirror_rxq[pi->port_id] = NULL;
+	return ret;
+}
+
+static void cxgb4_port_mirror_free_queues(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct sge *s = &adap->sge;
+	u16 i;
+
+	if (!pi->vi_mirror_count)
+		return;
+
+	if (!s->mirror_rxq[pi->port_id])
+		return;
+
+	for (i = 0; i < pi->nmirrorqsets; i++)
+		cxgb4_port_mirror_free_rxq(adap,
+					   &s->mirror_rxq[pi->port_id][i]);
+
+	kfree(s->mirror_rxq[pi->port_id]);
+	s->mirror_rxq[pi->port_id] = NULL;
+}
+
+static int cxgb4_port_mirror_start(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	int ret, idx = -1;
+
+	if (!pi->vi_mirror_count)
+		return 0;
+
+	/* Mirror VIs can be created dynamically after stack had
+	 * already setup Rx modes like MTU, promisc, allmulti, etc.
+	 * on main VI. So, parse what the stack had setup on the
+	 * main VI and update the same on the mirror VI.
+	 */
+	ret = t4_set_rxmode(adap, adap->mbox, pi->viid, pi->viid_mirror,
+			    dev->mtu, (dev->flags & IFF_PROMISC) ? 1 : 0,
+			    (dev->flags & IFF_ALLMULTI) ? 1 : 0, 1,
+			    !!(dev->features & NETIF_F_HW_VLAN_CTAG_RX), true);
+	if (ret) {
+		dev_err(adap->pdev_dev,
+			"Failed start up Rx mode for Mirror VI 0x%x, ret: %d\n",
+			pi->viid_mirror, ret);
+		return ret;
+	}
+
+	/* Enable replication bit for the device's MAC address
+	 * in MPS TCAM, so that the packets for the main VI are
+	 * replicated to mirror VI.
+	 */
+	ret = cxgb4_update_mac_filt(pi, pi->viid_mirror, &idx,
+				    dev->dev_addr, true, NULL);
+	if (ret) {
+		dev_err(adap->pdev_dev,
+			"Failed updating MAC filter for Mirror VI 0x%x, ret: %d\n",
+			pi->viid_mirror, ret);
+		return ret;
+	}
+
+	/* Enabling a Virtual Interface can result in an interrupt
+	 * during the processing of the VI Enable command and, in some
+	 * paths, result in an attempt to issue another command in the
+	 * interrupt context. Thus, we disable interrupts during the
+	 * course of the VI Enable command ...
+	 */
+	local_bh_disable();
+	ret = t4_enable_vi_params(adap, adap->mbox, pi->viid_mirror, true, true,
+				  false);
+	local_bh_enable();
+	if (ret)
+		dev_err(adap->pdev_dev,
+			"Failed starting Mirror VI 0x%x, ret: %d\n",
+			pi->viid_mirror, ret);
+
+	return ret;
+}
+
+static void cxgb4_port_mirror_stop(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+
+	if (!pi->vi_mirror_count)
+		return;
+
+	t4_enable_vi_params(adap, adap->mbox, pi->viid_mirror, false, false,
+			    false);
+}
+
+int cxgb4_port_mirror_alloc(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	int ret = 0;
+
+	if (!pi->nmirrorqsets)
+		return -EOPNOTSUPP;
+
+	mutex_lock(&pi->vi_mirror_mutex);
+	if (pi->viid_mirror) {
+		pi->vi_mirror_count++;
+		goto out_unlock;
+	}
+
+	ret = t4_init_port_mirror(pi, adap->mbox, pi->port_id, adap->pf, 0,
+				  &pi->viid_mirror);
+	if (ret)
+		goto out_unlock;
+
+	pi->vi_mirror_count = 1;
+
+	if (adap->flags & CXGB4_FULL_INIT_DONE) {
+		ret = cxgb4_port_mirror_alloc_queues(dev);
+		if (ret)
+			goto out_free_vi;
+
+		ret = cxgb4_port_mirror_start(dev);
+		if (ret)
+			goto out_free_queues;
+	}
+
+	mutex_unlock(&pi->vi_mirror_mutex);
+	return 0;
+
+out_free_queues:
+	cxgb4_port_mirror_free_queues(dev);
+
+out_free_vi:
+	pi->vi_mirror_count = 0;
+	t4_free_vi(adap, adap->mbox, adap->pf, 0, pi->viid_mirror);
+	pi->viid_mirror = 0;
+
+out_unlock:
+	mutex_unlock(&pi->vi_mirror_mutex);
+	return ret;
+}
+
+void cxgb4_port_mirror_free(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+
+	mutex_lock(&pi->vi_mirror_mutex);
+	if (!pi->viid_mirror)
+		goto out_unlock;
+
+	if (pi->vi_mirror_count > 1) {
+		pi->vi_mirror_count--;
+		goto out_unlock;
+	}
+
+	cxgb4_port_mirror_stop(dev);
+	cxgb4_port_mirror_free_queues(dev);
+
+	pi->vi_mirror_count = 0;
+	t4_free_vi(adap, adap->mbox, adap->pf, 0, pi->viid_mirror);
+	pi->viid_mirror = 0;
+
+out_unlock:
+	mutex_unlock(&pi->vi_mirror_mutex);
+}
+
+/*
+ * upper-layer driver support
+ */
+
+/*
+ * Allocate an active-open TID and set it to the supplied value.
+ */
+int cxgb4_alloc_atid(struct tid_info *t, void *data)
+{
+	int atid = -1;
+
+	spin_lock_bh(&t->atid_lock);
+	if (t->afree) {
+		union aopen_entry *p = t->afree;
+
+		atid = (p - t->atid_tab) + t->atid_base;
+		t->afree = p->next;
+		p->data = data;
+		t->atids_in_use++;
+	}
+	spin_unlock_bh(&t->atid_lock);
+	return atid;
+}
+EXPORT_SYMBOL(cxgb4_alloc_atid);
+
+/*
+ * Release an active-open TID.
+ */
+void cxgb4_free_atid(struct tid_info *t, unsigned int atid)
+{
+	union aopen_entry *p = &t->atid_tab[atid - t->atid_base];
+
+	spin_lock_bh(&t->atid_lock);
+	p->next = t->afree;
+	t->afree = p;
+	t->atids_in_use--;
+	spin_unlock_bh(&t->atid_lock);
+}
+EXPORT_SYMBOL(cxgb4_free_atid);
+
+/*
+ * Allocate a server TID and set it to the supplied value.
+ */
+int cxgb4_alloc_stid(struct tid_info *t, int family, void *data)
+{
+	int stid;
+
+	spin_lock_bh(&t->stid_lock);
+	if (family == PF_INET) {
+		stid = find_first_zero_bit(t->stid_bmap, t->nstids);
+		if (stid < t->nstids)
+			__set_bit(stid, t->stid_bmap);
+		else
+			stid = -1;
+	} else {
+		stid = bitmap_find_free_region(t->stid_bmap, t->nstids, 1);
+		if (stid < 0)
+			stid = -1;
+	}
+	if (stid >= 0) {
+		t->stid_tab[stid].data = data;
+		stid += t->stid_base;
+		/* IPv6 requires max of 520 bits or 16 cells in TCAM
+		 * This is equivalent to 4 TIDs. With CLIP enabled it
+		 * needs 2 TIDs.
+		 */
+		if (family == PF_INET6) {
+			t->stids_in_use += 2;
+			t->v6_stids_in_use += 2;
+		} else {
+			t->stids_in_use++;
+		}
+	}
+	spin_unlock_bh(&t->stid_lock);
+	return stid;
+}
+EXPORT_SYMBOL(cxgb4_alloc_stid);
+
+/* Allocate a server filter TID and set it to the supplied value.
+ */
+int cxgb4_alloc_sftid(struct tid_info *t, int family, void *data)
+{
+	int stid;
+
+	spin_lock_bh(&t->stid_lock);
+	if (family == PF_INET) {
+		stid = find_next_zero_bit(t->stid_bmap,
+				t->nstids + t->nsftids, t->nstids);
+		if (stid < (t->nstids + t->nsftids))
+			__set_bit(stid, t->stid_bmap);
+		else
+			stid = -1;
+	} else {
+		stid = -1;
+	}
+	if (stid >= 0) {
+		t->stid_tab[stid].data = data;
+		stid -= t->nstids;
+		stid += t->sftid_base;
+		t->sftids_in_use++;
+	}
+	spin_unlock_bh(&t->stid_lock);
+	return stid;
+}
+EXPORT_SYMBOL(cxgb4_alloc_sftid);
+
+/* Release a server TID.
+ */
+void cxgb4_free_stid(struct tid_info *t, unsigned int stid, int family)
+{
+	/* Is it a server filter TID? */
+	if (t->nsftids && (stid >= t->sftid_base)) {
+		stid -= t->sftid_base;
+		stid += t->nstids;
+	} else {
+		stid -= t->stid_base;
+	}
+
+	spin_lock_bh(&t->stid_lock);
+	if (family == PF_INET)
+		__clear_bit(stid, t->stid_bmap);
+	else
+		bitmap_release_region(t->stid_bmap, stid, 1);
+	t->stid_tab[stid].data = NULL;
+	if (stid < t->nstids) {
+		if (family == PF_INET6) {
+			t->stids_in_use -= 2;
+			t->v6_stids_in_use -= 2;
+		} else {
+			t->stids_in_use--;
+		}
+	} else {
+		t->sftids_in_use--;
+	}
+
+	spin_unlock_bh(&t->stid_lock);
+}
+EXPORT_SYMBOL(cxgb4_free_stid);
+
+/*
+ * Populate a TID_RELEASE WR.  Caller must properly size the skb.
+ */
+static void mk_tid_release(struct sk_buff *skb, unsigned int chan,
+			   unsigned int tid)
+{
+	struct cpl_tid_release *req;
+
+	set_wr_txq(skb, CPL_PRIORITY_SETUP, chan);
+	req = __skb_put(skb, sizeof(*req));
+	INIT_TP_WR(req, tid);
+	OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_TID_RELEASE, tid));
+}
+
+/*
+ * Queue a TID release request and if necessary schedule a work queue to
+ * process it.
+ */
+static void cxgb4_queue_tid_release(struct tid_info *t, unsigned int chan,
+				    unsigned int tid)
+{
+	struct adapter *adap = container_of(t, struct adapter, tids);
+	void **p = &t->tid_tab[tid - t->tid_base];
+
+	spin_lock_bh(&adap->tid_release_lock);
+	*p = adap->tid_release_head;
+	/* Low 2 bits encode the Tx channel number */
+	adap->tid_release_head = (void **)((uintptr_t)p | chan);
+	if (!adap->tid_release_task_busy) {
+		adap->tid_release_task_busy = true;
+		queue_work(adap->workq, &adap->tid_release_task);
+	}
+	spin_unlock_bh(&adap->tid_release_lock);
+}
+
+/*
+ * Process the list of pending TID release requests.
+ */
+static void process_tid_release_list(struct work_struct *work)
+{
+	struct sk_buff *skb;
+	struct adapter *adap;
+
+	adap = container_of(work, struct adapter, tid_release_task);
+
+	spin_lock_bh(&adap->tid_release_lock);
+	while (adap->tid_release_head) {
+		void **p = adap->tid_release_head;
+		unsigned int chan = (uintptr_t)p & 3;
+		p = (void *)p - chan;
+
+		adap->tid_release_head = *p;
+		*p = NULL;
+		spin_unlock_bh(&adap->tid_release_lock);
+
+		while (!(skb = alloc_skb(sizeof(struct cpl_tid_release),
+					 GFP_KERNEL)))
+			schedule_timeout_uninterruptible(1);
+
+		mk_tid_release(skb, chan, p - adap->tids.tid_tab);
+		t4_ofld_send(adap, skb);
+		spin_lock_bh(&adap->tid_release_lock);
+	}
+	adap->tid_release_task_busy = false;
+	spin_unlock_bh(&adap->tid_release_lock);
+}
+
+/*
+ * Release a TID and inform HW.  If we are unable to allocate the release
+ * message we defer to a work queue.
+ */
+void cxgb4_remove_tid(struct tid_info *t, unsigned int chan, unsigned int tid,
+		      unsigned short family)
+{
+	struct adapter *adap = container_of(t, struct adapter, tids);
+	struct sk_buff *skb;
+
+	WARN_ON(tid_out_of_range(&adap->tids, tid));
+
+	if (t->tid_tab[tid - adap->tids.tid_base]) {
+		t->tid_tab[tid - adap->tids.tid_base] = NULL;
+		atomic_dec(&t->conns_in_use);
+		if (t->hash_base && (tid >= t->hash_base)) {
+			if (family == AF_INET6)
+				atomic_sub(2, &t->hash_tids_in_use);
+			else
+				atomic_dec(&t->hash_tids_in_use);
+		} else {
+			if (family == AF_INET6)
+				atomic_sub(2, &t->tids_in_use);
+			else
+				atomic_dec(&t->tids_in_use);
+		}
+	}
+
+	skb = alloc_skb(sizeof(struct cpl_tid_release), GFP_ATOMIC);
+	if (likely(skb)) {
+		mk_tid_release(skb, chan, tid);
+		t4_ofld_send(adap, skb);
+	} else
+		cxgb4_queue_tid_release(t, chan, tid);
+}
+EXPORT_SYMBOL(cxgb4_remove_tid);
+
+/*
+ * Allocate and initialize the TID tables.  Returns 0 on success.
+ */
+static int tid_init(struct tid_info *t)
+{
+	struct adapter *adap = container_of(t, struct adapter, tids);
+	unsigned int max_ftids = t->nftids + t->nsftids;
+	unsigned int natids = t->natids;
+	unsigned int hpftid_bmap_size;
+	unsigned int eotid_bmap_size;
+	unsigned int stid_bmap_size;
+	unsigned int ftid_bmap_size;
+	size_t size;
+
+	stid_bmap_size = BITS_TO_LONGS(t->nstids + t->nsftids);
+	ftid_bmap_size = BITS_TO_LONGS(t->nftids);
+	hpftid_bmap_size = BITS_TO_LONGS(t->nhpftids);
+	eotid_bmap_size = BITS_TO_LONGS(t->neotids);
+	size = t->ntids * sizeof(*t->tid_tab) +
+	       natids * sizeof(*t->atid_tab) +
+	       t->nstids * sizeof(*t->stid_tab) +
+	       t->nsftids * sizeof(*t->stid_tab) +
+	       stid_bmap_size * sizeof(long) +
+	       t->nhpftids * sizeof(*t->hpftid_tab) +
+	       hpftid_bmap_size * sizeof(long) +
+	       max_ftids * sizeof(*t->ftid_tab) +
+	       ftid_bmap_size * sizeof(long) +
+	       t->neotids * sizeof(*t->eotid_tab) +
+	       eotid_bmap_size * sizeof(long);
+
+	t->tid_tab = kvzalloc(size, GFP_KERNEL);
+	if (!t->tid_tab)
+		return -ENOMEM;
+
+	t->atid_tab = (union aopen_entry *)&t->tid_tab[t->ntids];
+	t->stid_tab = (struct serv_entry *)&t->atid_tab[natids];
+	t->stid_bmap = (unsigned long *)&t->stid_tab[t->nstids + t->nsftids];
+	t->hpftid_tab = (struct filter_entry *)&t->stid_bmap[stid_bmap_size];
+	t->hpftid_bmap = (unsigned long *)&t->hpftid_tab[t->nhpftids];
+	t->ftid_tab = (struct filter_entry *)&t->hpftid_bmap[hpftid_bmap_size];
+	t->ftid_bmap = (unsigned long *)&t->ftid_tab[max_ftids];
+	t->eotid_tab = (struct eotid_entry *)&t->ftid_bmap[ftid_bmap_size];
+	t->eotid_bmap = (unsigned long *)&t->eotid_tab[t->neotids];
+	spin_lock_init(&t->stid_lock);
+	spin_lock_init(&t->atid_lock);
+	spin_lock_init(&t->ftid_lock);
+
+	t->stids_in_use = 0;
+	t->v6_stids_in_use = 0;
+	t->sftids_in_use = 0;
+	t->afree = NULL;
+	t->atids_in_use = 0;
+	atomic_set(&t->tids_in_use, 0);
+	atomic_set(&t->conns_in_use, 0);
+	atomic_set(&t->hash_tids_in_use, 0);
+	atomic_set(&t->eotids_in_use, 0);
+
+	/* Setup the free list for atid_tab and clear the stid bitmap. */
+	if (natids) {
+		while (--natids)
+			t->atid_tab[natids - 1].next = &t->atid_tab[natids];
+		t->afree = t->atid_tab;
+	}
+
+	if (is_offload(adap)) {
+		bitmap_zero(t->stid_bmap, t->nstids + t->nsftids);
+		/* Reserve stid 0 for T4/T5 adapters */
+		if (!t->stid_base &&
+		    CHELSIO_CHIP_VERSION(adap->params.chip) <= CHELSIO_T5)
+			__set_bit(0, t->stid_bmap);
+
+		if (t->neotids)
+			bitmap_zero(t->eotid_bmap, t->neotids);
+	}
+
+	if (t->nhpftids)
+		bitmap_zero(t->hpftid_bmap, t->nhpftids);
+	bitmap_zero(t->ftid_bmap, t->nftids);
+	return 0;
+}
+
+/**
+ *	cxgb4_create_server - create an IP server
+ *	@dev: the device
+ *	@stid: the server TID
+ *	@sip: local IP address to bind server to
+ *	@sport: the server's TCP port
+ *	@vlan: the VLAN header information
+ *	@queue: queue to direct messages from this server to
+ *
+ *	Create an IP server for the given port and address.
+ *	Returns <0 on error and one of the %NET_XMIT_* values on success.
+ */
+int cxgb4_create_server(const struct net_device *dev, unsigned int stid,
+			__be32 sip, __be16 sport, __be16 vlan,
+			unsigned int queue)
+{
+	unsigned int chan;
+	struct sk_buff *skb;
+	struct adapter *adap;
+	struct cpl_pass_open_req *req;
+	int ret;
+
+	skb = alloc_skb(sizeof(*req), GFP_KERNEL);
+	if (!skb)
+		return -ENOMEM;
+
+	adap = netdev2adap(dev);
+	req = __skb_put(skb, sizeof(*req));
+	INIT_TP_WR(req, 0);
+	OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_PASS_OPEN_REQ, stid));
+	req->local_port = sport;
+	req->peer_port = htons(0);
+	req->local_ip = sip;
+	req->peer_ip = htonl(0);
+	chan = rxq_to_chan(&adap->sge, queue);
+	req->opt0 = cpu_to_be64(TX_CHAN_V(chan));
+	req->opt1 = cpu_to_be64(CONN_POLICY_V(CPL_CONN_POLICY_ASK) |
+				SYN_RSS_ENABLE_F | SYN_RSS_QUEUE_V(queue));
+	ret = t4_mgmt_tx(adap, skb);
+	return net_xmit_eval(ret);
+}
+EXPORT_SYMBOL(cxgb4_create_server);
+
+/*	cxgb4_create_server6 - create an IPv6 server
+ *	@dev: the device
+ *	@stid: the server TID
+ *	@sip: local IPv6 address to bind server to
+ *	@sport: the server's TCP port
+ *	@queue: queue to direct messages from this server to
+ *
+ *	Create an IPv6 server for the given port and address.
+ *	Returns <0 on error and one of the %NET_XMIT_* values on success.
+ */
+int cxgb4_create_server6(const struct net_device *dev, unsigned int stid,
+			 const struct in6_addr *sip, __be16 sport,
+			 unsigned int queue)
+{
+	unsigned int chan;
+	struct sk_buff *skb;
+	struct adapter *adap;
+	struct cpl_pass_open_req6 *req;
+	int ret;
+
+	skb = alloc_skb(sizeof(*req), GFP_KERNEL);
+	if (!skb)
+		return -ENOMEM;
+
+	adap = netdev2adap(dev);
+	req = __skb_put(skb, sizeof(*req));
+	INIT_TP_WR(req, 0);
+	OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_PASS_OPEN_REQ6, stid));
+	req->local_port = sport;
+	req->peer_port = htons(0);
+	req->local_ip_hi = *(__be64 *)(sip->s6_addr);
+	req->local_ip_lo = *(__be64 *)(sip->s6_addr + 8);
+	req->peer_ip_hi = cpu_to_be64(0);
+	req->peer_ip_lo = cpu_to_be64(0);
+	chan = rxq_to_chan(&adap->sge, queue);
+	req->opt0 = cpu_to_be64(TX_CHAN_V(chan));
+	req->opt1 = cpu_to_be64(CONN_POLICY_V(CPL_CONN_POLICY_ASK) |
+				SYN_RSS_ENABLE_F | SYN_RSS_QUEUE_V(queue));
+	ret = t4_mgmt_tx(adap, skb);
+	return net_xmit_eval(ret);
+}
+EXPORT_SYMBOL(cxgb4_create_server6);
+
+int cxgb4_remove_server(const struct net_device *dev, unsigned int stid,
+			unsigned int queue, bool ipv6)
+{
+	struct sk_buff *skb;
+	struct adapter *adap;
+	struct cpl_close_listsvr_req *req;
+	int ret;
+
+	adap = netdev2adap(dev);
+
+	skb = alloc_skb(sizeof(*req), GFP_KERNEL);
+	if (!skb)
+		return -ENOMEM;
+
+	req = __skb_put(skb, sizeof(*req));
+	INIT_TP_WR(req, 0);
+	OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_CLOSE_LISTSRV_REQ, stid));
+	req->reply_ctrl = htons(NO_REPLY_V(0) | (ipv6 ? LISTSVR_IPV6_V(1) :
+				LISTSVR_IPV6_V(0)) | QUEUENO_V(queue));
+	ret = t4_mgmt_tx(adap, skb);
+	return net_xmit_eval(ret);
+}
+EXPORT_SYMBOL(cxgb4_remove_server);
+
+/**
+ *	cxgb4_best_mtu - find the entry in the MTU table closest to an MTU
+ *	@mtus: the HW MTU table
+ *	@mtu: the target MTU
+ *	@idx: index of selected entry in the MTU table
+ *
+ *	Returns the index and the value in the HW MTU table that is closest to
+ *	but does not exceed @mtu, unless @mtu is smaller than any value in the
+ *	table, in which case that smallest available value is selected.
+ */
+unsigned int cxgb4_best_mtu(const unsigned short *mtus, unsigned short mtu,
+			    unsigned int *idx)
+{
+	unsigned int i = 0;
+
+	while (i < NMTUS - 1 && mtus[i + 1] <= mtu)
+		++i;
+	if (idx)
+		*idx = i;
+	return mtus[i];
+}
+EXPORT_SYMBOL(cxgb4_best_mtu);
+
+/**
+ *     cxgb4_best_aligned_mtu - find best MTU, [hopefully] data size aligned
+ *     @mtus: the HW MTU table
+ *     @header_size: Header Size
+ *     @data_size_max: maximum Data Segment Size
+ *     @data_size_align: desired Data Segment Size Alignment (2^N)
+ *     @mtu_idxp: HW MTU Table Index return value pointer (possibly NULL)
+ *
+ *     Similar to cxgb4_best_mtu() but instead of searching the Hardware
+ *     MTU Table based solely on a Maximum MTU parameter, we break that
+ *     parameter up into a Header Size and Maximum Data Segment Size, and
+ *     provide a desired Data Segment Size Alignment.  If we find an MTU in
+ *     the Hardware MTU Table which will result in a Data Segment Size with
+ *     the requested alignment _and_ that MTU isn't "too far" from the
+ *     closest MTU, then we'll return that rather than the closest MTU.
+ */
+unsigned int cxgb4_best_aligned_mtu(const unsigned short *mtus,
+				    unsigned short header_size,
+				    unsigned short data_size_max,
+				    unsigned short data_size_align,
+				    unsigned int *mtu_idxp)
+{
+	unsigned short max_mtu = header_size + data_size_max;
+	unsigned short data_size_align_mask = data_size_align - 1;
+	int mtu_idx, aligned_mtu_idx;
+
+	/* Scan the MTU Table till we find an MTU which is larger than our
+	 * Maximum MTU or we reach the end of the table.  Along the way,
+	 * record the last MTU found, if any, which will result in a Data
+	 * Segment Length matching the requested alignment.
+	 */
+	for (mtu_idx = 0, aligned_mtu_idx = -1; mtu_idx < NMTUS; mtu_idx++) {
+		unsigned short data_size = mtus[mtu_idx] - header_size;
+
+		/* If this MTU minus the Header Size would result in a
+		 * Data Segment Size of the desired alignment, remember it.
+		 */
+		if ((data_size & data_size_align_mask) == 0)
+			aligned_mtu_idx = mtu_idx;
+
+		/* If we're not at the end of the Hardware MTU Table and the
+		 * next element is larger than our Maximum MTU, drop out of
+		 * the loop.
+		 */
+		if (mtu_idx+1 < NMTUS && mtus[mtu_idx+1] > max_mtu)
+			break;
+	}
+
+	/* If we fell out of the loop because we ran to the end of the table,
+	 * then we just have to use the last [largest] entry.
+	 */
+	if (mtu_idx == NMTUS)
+		mtu_idx--;
+
+	/* If we found an MTU which resulted in the requested Data Segment
+	 * Length alignment and that's "not far" from the largest MTU which is
+	 * less than or equal to the maximum MTU, then use that.
+	 */
+	if (aligned_mtu_idx >= 0 &&
+	    mtu_idx - aligned_mtu_idx <= 1)
+		mtu_idx = aligned_mtu_idx;
+
+	/* If the caller has passed in an MTU Index pointer, pass the
+	 * MTU Index back.  Return the MTU value.
+	 */
+	if (mtu_idxp)
+		*mtu_idxp = mtu_idx;
+	return mtus[mtu_idx];
+}
+EXPORT_SYMBOL(cxgb4_best_aligned_mtu);
+
+/**
+ *	cxgb4_port_chan - get the HW channel of a port
+ *	@dev: the net device for the port
+ *
+ *	Return the HW Tx channel of the given port.
+ */
+unsigned int cxgb4_port_chan(const struct net_device *dev)
+{
+	return netdev2pinfo(dev)->tx_chan;
+}
+EXPORT_SYMBOL(cxgb4_port_chan);
+
+/**
+ *      cxgb4_port_e2cchan - get the HW c-channel of a port
+ *      @dev: the net device for the port
+ *
+ *      Return the HW RX c-channel of the given port.
+ */
+unsigned int cxgb4_port_e2cchan(const struct net_device *dev)
+{
+	return netdev2pinfo(dev)->rx_cchan;
+}
+EXPORT_SYMBOL(cxgb4_port_e2cchan);
+
+unsigned int cxgb4_dbfifo_count(const struct net_device *dev, int lpfifo)
+{
+	struct adapter *adap = netdev2adap(dev);
+	u32 v1, v2, lp_count, hp_count;
+
+	v1 = t4_read_reg(adap, SGE_DBFIFO_STATUS_A);
+	v2 = t4_read_reg(adap, SGE_DBFIFO_STATUS2_A);
+	if (is_t4(adap->params.chip)) {
+		lp_count = LP_COUNT_G(v1);
+		hp_count = HP_COUNT_G(v1);
+	} else {
+		lp_count = LP_COUNT_T5_G(v1);
+		hp_count = HP_COUNT_T5_G(v2);
+	}
+	return lpfifo ? lp_count : hp_count;
+}
+EXPORT_SYMBOL(cxgb4_dbfifo_count);
+
+/**
+ *	cxgb4_port_viid - get the VI id of a port
+ *	@dev: the net device for the port
+ *
+ *	Return the VI id of the given port.
+ */
+unsigned int cxgb4_port_viid(const struct net_device *dev)
+{
+	return netdev2pinfo(dev)->viid;
+}
+EXPORT_SYMBOL(cxgb4_port_viid);
+
+/**
+ *	cxgb4_port_idx - get the index of a port
+ *	@dev: the net device for the port
+ *
+ *	Return the index of the given port.
+ */
+unsigned int cxgb4_port_idx(const struct net_device *dev)
+{
+	return netdev2pinfo(dev)->port_id;
+}
+EXPORT_SYMBOL(cxgb4_port_idx);
+
+void cxgb4_get_tcp_stats(struct pci_dev *pdev, struct tp_tcp_stats *v4,
+			 struct tp_tcp_stats *v6)
+{
+	struct adapter *adap = pci_get_drvdata(pdev);
+
+	spin_lock(&adap->stats_lock);
+	t4_tp_get_tcp_stats(adap, v4, v6, false);
+	spin_unlock(&adap->stats_lock);
+}
+EXPORT_SYMBOL(cxgb4_get_tcp_stats);
+
+void cxgb4_iscsi_init(struct net_device *dev, unsigned int tag_mask,
+		      const unsigned int *pgsz_order)
+{
+	struct adapter *adap = netdev2adap(dev);
+
+	t4_write_reg(adap, ULP_RX_ISCSI_TAGMASK_A, tag_mask);
+	t4_write_reg(adap, ULP_RX_ISCSI_PSZ_A, HPZ0_V(pgsz_order[0]) |
+		     HPZ1_V(pgsz_order[1]) | HPZ2_V(pgsz_order[2]) |
+		     HPZ3_V(pgsz_order[3]));
+}
+EXPORT_SYMBOL(cxgb4_iscsi_init);
+
+int cxgb4_flush_eq_cache(struct net_device *dev)
+{
+	struct adapter *adap = netdev2adap(dev);
+
+	return t4_sge_ctxt_flush(adap, adap->mbox, CTXT_EGRESS);
+}
+EXPORT_SYMBOL(cxgb4_flush_eq_cache);
+
+static int read_eq_indices(struct adapter *adap, u16 qid, u16 *pidx, u16 *cidx)
+{
+	u32 addr = t4_read_reg(adap, SGE_DBQ_CTXT_BADDR_A) + 24 * qid + 8;
+	__be64 indices;
+	int ret;
+
+	spin_lock(&adap->win0_lock);
+	ret = t4_memory_rw(adap, 0, MEM_EDC0, addr,
+			   sizeof(indices), (__be32 *)&indices,
+			   T4_MEMORY_READ);
+	spin_unlock(&adap->win0_lock);
+	if (!ret) {
+		*cidx = (be64_to_cpu(indices) >> 25) & 0xffff;
+		*pidx = (be64_to_cpu(indices) >> 9) & 0xffff;
+	}
+	return ret;
+}
+
+int cxgb4_sync_txq_pidx(struct net_device *dev, u16 qid, u16 pidx,
+			u16 size)
+{
+	struct adapter *adap = netdev2adap(dev);
+	u16 hw_pidx, hw_cidx;
+	int ret;
+
+	ret = read_eq_indices(adap, qid, &hw_pidx, &hw_cidx);
+	if (ret)
+		goto out;
+
+	if (pidx != hw_pidx) {
+		u16 delta;
+		u32 val;
+
+		if (pidx >= hw_pidx)
+			delta = pidx - hw_pidx;
+		else
+			delta = size - hw_pidx + pidx;
+
+		if (is_t4(adap->params.chip))
+			val = PIDX_V(delta);
+		else
+			val = PIDX_T5_V(delta);
+		wmb();
+		t4_write_reg(adap, MYPF_REG(SGE_PF_KDOORBELL_A),
+			     QID_V(qid) | val);
+	}
+out:
+	return ret;
+}
+EXPORT_SYMBOL(cxgb4_sync_txq_pidx);
+
+int cxgb4_read_tpte(struct net_device *dev, u32 stag, __be32 *tpte)
+{
+	u32 edc0_size, edc1_size, mc0_size, mc1_size, size;
+	u32 edc0_end, edc1_end, mc0_end, mc1_end;
+	u32 offset, memtype, memaddr;
+	struct adapter *adap;
+	u32 hma_size = 0;
+	int ret;
+
+	adap = netdev2adap(dev);
+
+	offset = ((stag >> 8) * 32) + adap->vres.stag.start;
+
+	/* Figure out where the offset lands in the Memory Type/Address scheme.
+	 * This code assumes that the memory is laid out starting at offset 0
+	 * with no breaks as: EDC0, EDC1, MC0, MC1. All cards have both EDC0
+	 * and EDC1.  Some cards will have neither MC0 nor MC1, most cards have
+	 * MC0, and some have both MC0 and MC1.
+	 */
+	size = t4_read_reg(adap, MA_EDRAM0_BAR_A);
+	edc0_size = EDRAM0_SIZE_G(size) << 20;
+	size = t4_read_reg(adap, MA_EDRAM1_BAR_A);
+	edc1_size = EDRAM1_SIZE_G(size) << 20;
+	size = t4_read_reg(adap, MA_EXT_MEMORY0_BAR_A);
+	mc0_size = EXT_MEM0_SIZE_G(size) << 20;
+
+	if (t4_read_reg(adap, MA_TARGET_MEM_ENABLE_A) & HMA_MUX_F) {
+		size = t4_read_reg(adap, MA_EXT_MEMORY1_BAR_A);
+		hma_size = EXT_MEM1_SIZE_G(size) << 20;
+	}
+	edc0_end = edc0_size;
+	edc1_end = edc0_end + edc1_size;
+	mc0_end = edc1_end + mc0_size;
+
+	if (offset < edc0_end) {
+		memtype = MEM_EDC0;
+		memaddr = offset;
+	} else if (offset < edc1_end) {
+		memtype = MEM_EDC1;
+		memaddr = offset - edc0_end;
+	} else {
+		if (hma_size && (offset < (edc1_end + hma_size))) {
+			memtype = MEM_HMA;
+			memaddr = offset - edc1_end;
+		} else if (offset < mc0_end) {
+			memtype = MEM_MC0;
+			memaddr = offset - edc1_end;
+		} else if (is_t5(adap->params.chip)) {
+			size = t4_read_reg(adap, MA_EXT_MEMORY1_BAR_A);
+			mc1_size = EXT_MEM1_SIZE_G(size) << 20;
+			mc1_end = mc0_end + mc1_size;
+			if (offset < mc1_end) {
+				memtype = MEM_MC1;
+				memaddr = offset - mc0_end;
+			} else {
+				/* offset beyond the end of any memory */
+				goto err;
+			}
+		} else {
+			/* T4/T6 only has a single memory channel */
+			goto err;
+		}
+	}
+
+	spin_lock(&adap->win0_lock);
+	ret = t4_memory_rw(adap, 0, memtype, memaddr, 32, tpte, T4_MEMORY_READ);
+	spin_unlock(&adap->win0_lock);
+	return ret;
+
+err:
+	dev_err(adap->pdev_dev, "stag %#x, offset %#x out of range\n",
+		stag, offset);
+	return -EINVAL;
+}
+EXPORT_SYMBOL(cxgb4_read_tpte);
+
+u64 cxgb4_read_sge_timestamp(struct net_device *dev)
+{
+	u32 hi, lo;
+	struct adapter *adap;
+
+	adap = netdev2adap(dev);
+	lo = t4_read_reg(adap, SGE_TIMESTAMP_LO_A);
+	hi = TSVAL_G(t4_read_reg(adap, SGE_TIMESTAMP_HI_A));
+
+	return ((u64)hi << 32) | (u64)lo;
+}
+EXPORT_SYMBOL(cxgb4_read_sge_timestamp);
+
+int cxgb4_bar2_sge_qregs(struct net_device *dev,
+			 unsigned int qid,
+			 enum cxgb4_bar2_qtype qtype,
+			 int user,
+			 u64 *pbar2_qoffset,
+			 unsigned int *pbar2_qid)
+{
+	return t4_bar2_sge_qregs(netdev2adap(dev),
+				 qid,
+				 (qtype == CXGB4_BAR2_QTYPE_EGRESS
+				  ? T4_BAR2_QTYPE_EGRESS
+				  : T4_BAR2_QTYPE_INGRESS),
+				 user,
+				 pbar2_qoffset,
+				 pbar2_qid);
+}
+EXPORT_SYMBOL(cxgb4_bar2_sge_qregs);
+
+static struct pci_driver cxgb4_driver;
+
+static void check_neigh_update(struct neighbour *neigh)
+{
+	const struct device *parent;
+	const struct net_device *netdev = neigh->dev;
+
+	if (is_vlan_dev(netdev))
+		netdev = vlan_dev_real_dev(netdev);
+	parent = netdev->dev.parent;
+	if (parent && parent->driver == &cxgb4_driver.driver)
+		t4_l2t_update(dev_get_drvdata(parent), neigh);
+}
+
+static int netevent_cb(struct notifier_block *nb, unsigned long event,
+		       void *data)
+{
+	switch (event) {
+	case NETEVENT_NEIGH_UPDATE:
+		check_neigh_update(data);
+		break;
+	case NETEVENT_REDIRECT:
+	default:
+		break;
+	}
+	return 0;
+}
+
+static bool netevent_registered;
+static struct notifier_block cxgb4_netevent_nb = {
+	.notifier_call = netevent_cb
+};
+
+static void drain_db_fifo(struct adapter *adap, int usecs)
+{
+	u32 v1, v2, lp_count, hp_count;
+
+	do {
+		v1 = t4_read_reg(adap, SGE_DBFIFO_STATUS_A);
+		v2 = t4_read_reg(adap, SGE_DBFIFO_STATUS2_A);
+		if (is_t4(adap->params.chip)) {
+			lp_count = LP_COUNT_G(v1);
+			hp_count = HP_COUNT_G(v1);
+		} else {
+			lp_count = LP_COUNT_T5_G(v1);
+			hp_count = HP_COUNT_T5_G(v2);
+		}
+
+		if (lp_count == 0 && hp_count == 0)
+			break;
+		set_current_state(TASK_UNINTERRUPTIBLE);
+		schedule_timeout(usecs_to_jiffies(usecs));
+	} while (1);
+}
+
+static void disable_txq_db(struct sge_txq *q)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&q->db_lock, flags);
+	q->db_disabled = 1;
+	spin_unlock_irqrestore(&q->db_lock, flags);
+}
+
+static void enable_txq_db(struct adapter *adap, struct sge_txq *q)
+{
+	spin_lock_irq(&q->db_lock);
+	if (q->db_pidx_inc) {
+		/* Make sure that all writes to the TX descriptors
+		 * are committed before we tell HW about them.
+		 */
+		wmb();
+		t4_write_reg(adap, MYPF_REG(SGE_PF_KDOORBELL_A),
+			     QID_V(q->cntxt_id) | PIDX_V(q->db_pidx_inc));
+		q->db_pidx_inc = 0;
+	}
+	q->db_disabled = 0;
+	spin_unlock_irq(&q->db_lock);
+}
+
+static void disable_dbs(struct adapter *adap)
+{
+	int i;
+
+	for_each_ethrxq(&adap->sge, i)
+		disable_txq_db(&adap->sge.ethtxq[i].q);
+	if (is_offload(adap)) {
+		struct sge_uld_txq_info *txq_info =
+			adap->sge.uld_txq_info[CXGB4_TX_OFLD];
+
+		if (txq_info) {
+			for_each_ofldtxq(&adap->sge, i) {
+				struct sge_uld_txq *txq = &txq_info->uldtxq[i];
+
+				disable_txq_db(&txq->q);
+			}
+		}
+	}
+	for_each_port(adap, i)
+		disable_txq_db(&adap->sge.ctrlq[i].q);
+}
+
+static void enable_dbs(struct adapter *adap)
+{
+	int i;
+
+	for_each_ethrxq(&adap->sge, i)
+		enable_txq_db(adap, &adap->sge.ethtxq[i].q);
+	if (is_offload(adap)) {
+		struct sge_uld_txq_info *txq_info =
+			adap->sge.uld_txq_info[CXGB4_TX_OFLD];
+
+		if (txq_info) {
+			for_each_ofldtxq(&adap->sge, i) {
+				struct sge_uld_txq *txq = &txq_info->uldtxq[i];
+
+				enable_txq_db(adap, &txq->q);
+			}
+		}
+	}
+	for_each_port(adap, i)
+		enable_txq_db(adap, &adap->sge.ctrlq[i].q);
+}
+
+static void notify_rdma_uld(struct adapter *adap, enum cxgb4_control cmd)
+{
+	enum cxgb4_uld type = CXGB4_ULD_RDMA;
+
+	if (adap->uld && adap->uld[type].handle)
+		adap->uld[type].control(adap->uld[type].handle, cmd);
+}
+
+static void process_db_full(struct work_struct *work)
+{
+	struct adapter *adap;
+
+	adap = container_of(work, struct adapter, db_full_task);
+
+	drain_db_fifo(adap, dbfifo_drain_delay);
+	enable_dbs(adap);
+	notify_rdma_uld(adap, CXGB4_CONTROL_DB_EMPTY);
+	if (CHELSIO_CHIP_VERSION(adap->params.chip) <= CHELSIO_T5)
+		t4_set_reg_field(adap, SGE_INT_ENABLE3_A,
+				 DBFIFO_HP_INT_F | DBFIFO_LP_INT_F,
+				 DBFIFO_HP_INT_F | DBFIFO_LP_INT_F);
+	else
+		t4_set_reg_field(adap, SGE_INT_ENABLE3_A,
+				 DBFIFO_LP_INT_F, DBFIFO_LP_INT_F);
+}
+
+static void sync_txq_pidx(struct adapter *adap, struct sge_txq *q)
+{
+	u16 hw_pidx, hw_cidx;
+	int ret;
+
+	spin_lock_irq(&q->db_lock);
+	ret = read_eq_indices(adap, (u16)q->cntxt_id, &hw_pidx, &hw_cidx);
+	if (ret)
+		goto out;
+	if (q->db_pidx != hw_pidx) {
+		u16 delta;
+		u32 val;
+
+		if (q->db_pidx >= hw_pidx)
+			delta = q->db_pidx - hw_pidx;
+		else
+			delta = q->size - hw_pidx + q->db_pidx;
+
+		if (is_t4(adap->params.chip))
+			val = PIDX_V(delta);
+		else
+			val = PIDX_T5_V(delta);
+		wmb();
+		t4_write_reg(adap, MYPF_REG(SGE_PF_KDOORBELL_A),
+			     QID_V(q->cntxt_id) | val);
+	}
+out:
+	q->db_disabled = 0;
+	q->db_pidx_inc = 0;
+	spin_unlock_irq(&q->db_lock);
+	if (ret)
+		CH_WARN(adap, "DB drop recovery failed.\n");
+}
+
+static void recover_all_queues(struct adapter *adap)
+{
+	int i;
+
+	for_each_ethrxq(&adap->sge, i)
+		sync_txq_pidx(adap, &adap->sge.ethtxq[i].q);
+	if (is_offload(adap)) {
+		struct sge_uld_txq_info *txq_info =
+			adap->sge.uld_txq_info[CXGB4_TX_OFLD];
+		if (txq_info) {
+			for_each_ofldtxq(&adap->sge, i) {
+				struct sge_uld_txq *txq = &txq_info->uldtxq[i];
+
+				sync_txq_pidx(adap, &txq->q);
+			}
+		}
+	}
+	for_each_port(adap, i)
+		sync_txq_pidx(adap, &adap->sge.ctrlq[i].q);
+}
+
+static void process_db_drop(struct work_struct *work)
+{
+	struct adapter *adap;
+
+	adap = container_of(work, struct adapter, db_drop_task);
+
+	if (is_t4(adap->params.chip)) {
+		drain_db_fifo(adap, dbfifo_drain_delay);
+		notify_rdma_uld(adap, CXGB4_CONTROL_DB_DROP);
+		drain_db_fifo(adap, dbfifo_drain_delay);
+		recover_all_queues(adap);
+		drain_db_fifo(adap, dbfifo_drain_delay);
+		enable_dbs(adap);
+		notify_rdma_uld(adap, CXGB4_CONTROL_DB_EMPTY);
+	} else if (is_t5(adap->params.chip)) {
+		u32 dropped_db = t4_read_reg(adap, 0x010ac);
+		u16 qid = (dropped_db >> 15) & 0x1ffff;
+		u16 pidx_inc = dropped_db & 0x1fff;
+		u64 bar2_qoffset;
+		unsigned int bar2_qid;
+		int ret;
+
+		ret = t4_bar2_sge_qregs(adap, qid, T4_BAR2_QTYPE_EGRESS,
+					0, &bar2_qoffset, &bar2_qid);
+		if (ret)
+			dev_err(adap->pdev_dev, "doorbell drop recovery: "
+				"qid=%d, pidx_inc=%d\n", qid, pidx_inc);
+		else
+			writel(PIDX_T5_V(pidx_inc) | QID_V(bar2_qid),
+			       adap->bar2 + bar2_qoffset + SGE_UDB_KDOORBELL);
+
+		/* Re-enable BAR2 WC */
+		t4_set_reg_field(adap, 0x10b0, 1<<15, 1<<15);
+	}
+
+	if (CHELSIO_CHIP_VERSION(adap->params.chip) <= CHELSIO_T5)
+		t4_set_reg_field(adap, SGE_DOORBELL_CONTROL_A, DROPPED_DB_F, 0);
+}
+
+void t4_db_full(struct adapter *adap)
+{
+	if (is_t4(adap->params.chip)) {
+		disable_dbs(adap);
+		notify_rdma_uld(adap, CXGB4_CONTROL_DB_FULL);
+		t4_set_reg_field(adap, SGE_INT_ENABLE3_A,
+				 DBFIFO_HP_INT_F | DBFIFO_LP_INT_F, 0);
+		queue_work(adap->workq, &adap->db_full_task);
+	}
+}
+
+void t4_db_dropped(struct adapter *adap)
+{
+	if (is_t4(adap->params.chip)) {
+		disable_dbs(adap);
+		notify_rdma_uld(adap, CXGB4_CONTROL_DB_FULL);
+	}
+	queue_work(adap->workq, &adap->db_drop_task);
+}
+
+void t4_register_netevent_notifier(void)
+{
+	if (!netevent_registered) {
+		register_netevent_notifier(&cxgb4_netevent_nb);
+		netevent_registered = true;
+	}
+}
+
+static void detach_ulds(struct adapter *adap)
+{
+	unsigned int i;
+
+	if (!is_uld(adap))
+		return;
+
+	mutex_lock(&uld_mutex);
+	list_del(&adap->list_node);
+
+	for (i = 0; i < CXGB4_ULD_MAX; i++)
+		if (adap->uld && adap->uld[i].handle)
+			adap->uld[i].state_change(adap->uld[i].handle,
+					     CXGB4_STATE_DETACH);
+
+	if (netevent_registered && list_empty(&adapter_list)) {
+		unregister_netevent_notifier(&cxgb4_netevent_nb);
+		netevent_registered = false;
+	}
+	mutex_unlock(&uld_mutex);
+}
+
+static void notify_ulds(struct adapter *adap, enum cxgb4_state new_state)
+{
+	unsigned int i;
+
+	mutex_lock(&uld_mutex);
+	for (i = 0; i < CXGB4_ULD_MAX; i++)
+		if (adap->uld && adap->uld[i].handle)
+			adap->uld[i].state_change(adap->uld[i].handle,
+						  new_state);
+	mutex_unlock(&uld_mutex);
+}
+
+#if IS_ENABLED(CONFIG_IPV6)
+static int cxgb4_inet6addr_handler(struct notifier_block *this,
+				   unsigned long event, void *data)
+{
+	struct inet6_ifaddr *ifa = data;
+	struct net_device *event_dev = ifa->idev->dev;
+	const struct device *parent = NULL;
+#if IS_ENABLED(CONFIG_BONDING)
+	struct adapter *adap;
+#endif
+	if (is_vlan_dev(event_dev))
+		event_dev = vlan_dev_real_dev(event_dev);
+#if IS_ENABLED(CONFIG_BONDING)
+	if (event_dev->flags & IFF_MASTER) {
+		list_for_each_entry(adap, &adapter_list, list_node) {
+			switch (event) {
+			case NETDEV_UP:
+				cxgb4_clip_get(adap->port[0],
+					       (const u32 *)ifa, 1);
+				break;
+			case NETDEV_DOWN:
+				cxgb4_clip_release(adap->port[0],
+						   (const u32 *)ifa, 1);
+				break;
+			default:
+				break;
+			}
+		}
+		return NOTIFY_OK;
+	}
+#endif
+
+	if (event_dev)
+		parent = event_dev->dev.parent;
+
+	if (parent && parent->driver == &cxgb4_driver.driver) {
+		switch (event) {
+		case NETDEV_UP:
+			cxgb4_clip_get(event_dev, (const u32 *)ifa, 1);
+			break;
+		case NETDEV_DOWN:
+			cxgb4_clip_release(event_dev, (const u32 *)ifa, 1);
+			break;
+		default:
+			break;
+		}
+	}
+	return NOTIFY_OK;
+}
+
+static bool inet6addr_registered;
+static struct notifier_block cxgb4_inet6addr_notifier = {
+	.notifier_call = cxgb4_inet6addr_handler
+};
+
+static void update_clip(const struct adapter *adap)
+{
+	int i;
+	struct net_device *dev;
+	int ret;
+
+	rcu_read_lock();
+
+	for (i = 0; i < MAX_NPORTS; i++) {
+		dev = adap->port[i];
+		ret = 0;
+
+		if (dev)
+			ret = cxgb4_update_root_dev_clip(dev);
+
+		if (ret < 0)
+			break;
+	}
+	rcu_read_unlock();
+}
+#endif /* IS_ENABLED(CONFIG_IPV6) */
+
+/**
+ *	cxgb_up - enable the adapter
+ *	@adap: adapter being enabled
+ *
+ *	Called when the first port is enabled, this function performs the
+ *	actions necessary to make an adapter operational, such as completing
+ *	the initialization of HW modules, and enabling interrupts.
+ *
+ *	Must be called with the rtnl lock held.
+ */
+static int cxgb_up(struct adapter *adap)
+{
+	struct sge *s = &adap->sge;
+	int err;
+
+	mutex_lock(&uld_mutex);
+	err = setup_sge_queues(adap);
+	if (err)
+		goto rel_lock;
+	err = setup_rss(adap);
+	if (err)
+		goto freeq;
+
+	if (adap->flags & CXGB4_USING_MSIX) {
+		if (s->nd_msix_idx < 0) {
+			err = -ENOMEM;
+			goto irq_err;
+		}
+
+		err = request_irq(adap->msix_info[s->nd_msix_idx].vec,
+				  t4_nondata_intr, 0,
+				  adap->msix_info[s->nd_msix_idx].desc, adap);
+		if (err)
+			goto irq_err;
+
+		err = request_msix_queue_irqs(adap);
+		if (err)
+			goto irq_err_free_nd_msix;
+	} else {
+		err = request_irq(adap->pdev->irq, t4_intr_handler(adap),
+				  (adap->flags & CXGB4_USING_MSI) ? 0
+								  : IRQF_SHARED,
+				  adap->port[0]->name, adap);
+		if (err)
+			goto irq_err;
+	}
+
+	enable_rx(adap);
+	t4_sge_start(adap);
+	t4_intr_enable(adap);
+	adap->flags |= CXGB4_FULL_INIT_DONE;
+	mutex_unlock(&uld_mutex);
+
+	notify_ulds(adap, CXGB4_STATE_UP);
+#if IS_ENABLED(CONFIG_IPV6)
+	update_clip(adap);
+#endif
+	return err;
+
+irq_err_free_nd_msix:
+	free_irq(adap->msix_info[s->nd_msix_idx].vec, adap);
+irq_err:
+	dev_err(adap->pdev_dev, "request_irq failed, err %d\n", err);
+freeq:
+	t4_free_sge_resources(adap);
+rel_lock:
+	mutex_unlock(&uld_mutex);
+	return err;
+}
+
+static void cxgb_down(struct adapter *adapter)
+{
+	cancel_work_sync(&adapter->tid_release_task);
+	cancel_work_sync(&adapter->db_full_task);
+	cancel_work_sync(&adapter->db_drop_task);
+	adapter->tid_release_task_busy = false;
+	adapter->tid_release_head = NULL;
+
+	t4_sge_stop(adapter);
+	t4_free_sge_resources(adapter);
+
+	adapter->flags &= ~CXGB4_FULL_INIT_DONE;
+}
+
+/*
+ * net_device operations
+ */
+static int cxgb_open(struct net_device *dev)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adapter = pi->adapter;
+	int err;
+
+	netif_carrier_off(dev);
+
+	if (!(adapter->flags & CXGB4_FULL_INIT_DONE)) {
+		err = cxgb_up(adapter);
+		if (err < 0)
+			return err;
+	}
+
+	/* It's possible that the basic port information could have
+	 * changed since we first read it.
+	 */
+	err = t4_update_port_info(pi);
+	if (err < 0)
+		return err;
+
+	err = link_start(dev);
+	if (err)
+		return err;
+
+	if (pi->nmirrorqsets) {
+		mutex_lock(&pi->vi_mirror_mutex);
+		err = cxgb4_port_mirror_alloc_queues(dev);
+		if (err)
+			goto out_unlock;
+
+		err = cxgb4_port_mirror_start(dev);
+		if (err)
+			goto out_free_queues;
+		mutex_unlock(&pi->vi_mirror_mutex);
+	}
+
+	netif_tx_start_all_queues(dev);
+	return 0;
+
+out_free_queues:
+	cxgb4_port_mirror_free_queues(dev);
+
+out_unlock:
+	mutex_unlock(&pi->vi_mirror_mutex);
+	return err;
+}
+
+static int cxgb_close(struct net_device *dev)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adapter = pi->adapter;
+	int ret;
+
+	netif_tx_stop_all_queues(dev);
+	netif_carrier_off(dev);
+	ret = t4_enable_pi_params(adapter, adapter->pf, pi,
+				  false, false, false);
+#ifdef CONFIG_CHELSIO_T4_DCB
+	cxgb4_dcb_reset(dev);
+	dcb_tx_queue_prio_enable(dev, false);
+#endif
+	if (ret)
+		return ret;
+
+	if (pi->nmirrorqsets) {
+		mutex_lock(&pi->vi_mirror_mutex);
+		cxgb4_port_mirror_stop(dev);
+		cxgb4_port_mirror_free_queues(dev);
+		mutex_unlock(&pi->vi_mirror_mutex);
+	}
+
+	return 0;
+}
+
+int cxgb4_create_server_filter(const struct net_device *dev, unsigned int stid,
+		__be32 sip, __be16 sport, __be16 vlan,
+		unsigned int queue, unsigned char port, unsigned char mask)
+{
+	int ret;
+	struct filter_entry *f;
+	struct adapter *adap;
+	int i;
+	u8 *val;
+
+	adap = netdev2adap(dev);
+
+	/* Adjust stid to correct filter index */
+	stid -= adap->tids.sftid_base;
+	stid += adap->tids.nftids;
+
+	/* Check to make sure the filter requested is writable ...
+	 */
+	f = &adap->tids.ftid_tab[stid];
+	ret = writable_filter(f);
+	if (ret)
+		return ret;
+
+	/* Clear out any old resources being used by the filter before
+	 * we start constructing the new filter.
+	 */
+	if (f->valid)
+		clear_filter(adap, f);
+
+	/* Clear out filter specifications */
+	memset(&f->fs, 0, sizeof(struct ch_filter_specification));
+	f->fs.val.lport = be16_to_cpu(sport);
+	f->fs.mask.lport  = ~0;
+	val = (u8 *)&sip;
+	if ((val[0] | val[1] | val[2] | val[3]) != 0) {
+		for (i = 0; i < 4; i++) {
+			f->fs.val.lip[i] = val[i];
+			f->fs.mask.lip[i] = ~0;
+		}
+		if (adap->params.tp.vlan_pri_map & PORT_F) {
+			f->fs.val.iport = port;
+			f->fs.mask.iport = mask;
+		}
+	}
+
+	if (adap->params.tp.vlan_pri_map & PROTOCOL_F) {
+		f->fs.val.proto = IPPROTO_TCP;
+		f->fs.mask.proto = ~0;
+	}
+
+	f->fs.dirsteer = 1;
+	f->fs.iq = queue;
+	/* Mark filter as locked */
+	f->locked = 1;
+	f->fs.rpttid = 1;
+
+	/* Save the actual tid. We need this to get the corresponding
+	 * filter entry structure in filter_rpl.
+	 */
+	f->tid = stid + adap->tids.ftid_base;
+	ret = set_filter_wr(adap, stid);
+	if (ret) {
+		clear_filter(adap, f);
+		return ret;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(cxgb4_create_server_filter);
+
+int cxgb4_remove_server_filter(const struct net_device *dev, unsigned int stid,
+		unsigned int queue, bool ipv6)
+{
+	struct filter_entry *f;
+	struct adapter *adap;
+
+	adap = netdev2adap(dev);
+
+	/* Adjust stid to correct filter index */
+	stid -= adap->tids.sftid_base;
+	stid += adap->tids.nftids;
+
+	f = &adap->tids.ftid_tab[stid];
+	/* Unlock the filter */
+	f->locked = 0;
+
+	return delete_filter(adap, stid);
+}
+EXPORT_SYMBOL(cxgb4_remove_server_filter);
+
+static void cxgb_get_stats(struct net_device *dev,
+			   struct rtnl_link_stats64 *ns)
+{
+	struct port_stats stats;
+	struct port_info *p = netdev_priv(dev);
+	struct adapter *adapter = p->adapter;
+
+	/* Block retrieving statistics during EEH error
+	 * recovery. Otherwise, the recovery might fail
+	 * and the PCI device will be removed permanently
+	 */
+	spin_lock(&adapter->stats_lock);
+	if (!netif_device_present(dev)) {
+		spin_unlock(&adapter->stats_lock);
+		return;
+	}
+	t4_get_port_stats_offset(adapter, p->tx_chan, &stats,
+				 &p->stats_base);
+	spin_unlock(&adapter->stats_lock);
+
+	ns->tx_bytes   = stats.tx_octets;
+	ns->tx_packets = stats.tx_frames;
+	ns->rx_bytes   = stats.rx_octets;
+	ns->rx_packets = stats.rx_frames;
+	ns->multicast  = stats.rx_mcast_frames;
+
+	/* detailed rx_errors */
+	ns->rx_length_errors = stats.rx_jabber + stats.rx_too_long +
+			       stats.rx_runt;
+	ns->rx_over_errors   = 0;
+	ns->rx_crc_errors    = stats.rx_fcs_err;
+	ns->rx_frame_errors  = stats.rx_symbol_err;
+	ns->rx_dropped	     = stats.rx_ovflow0 + stats.rx_ovflow1 +
+			       stats.rx_ovflow2 + stats.rx_ovflow3 +
+			       stats.rx_trunc0 + stats.rx_trunc1 +
+			       stats.rx_trunc2 + stats.rx_trunc3;
+	ns->rx_missed_errors = 0;
+
+	/* detailed tx_errors */
+	ns->tx_aborted_errors   = 0;
+	ns->tx_carrier_errors   = 0;
+	ns->tx_fifo_errors      = 0;
+	ns->tx_heartbeat_errors = 0;
+	ns->tx_window_errors    = 0;
+
+	ns->tx_errors = stats.tx_error_frames;
+	ns->rx_errors = stats.rx_symbol_err + stats.rx_fcs_err +
+		ns->rx_length_errors + stats.rx_len_err + ns->rx_fifo_errors;
+}
+
+static int cxgb_ioctl(struct net_device *dev, struct ifreq *req, int cmd)
+{
+	unsigned int mbox;
+	int ret = 0, prtad, devad;
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adapter = pi->adapter;
+	struct mii_ioctl_data *data = (struct mii_ioctl_data *)&req->ifr_data;
+
+	switch (cmd) {
+	case SIOCGMIIPHY:
+		if (pi->mdio_addr < 0)
+			return -EOPNOTSUPP;
+		data->phy_id = pi->mdio_addr;
+		break;
+	case SIOCGMIIREG:
+	case SIOCSMIIREG:
+		if (mdio_phy_id_is_c45(data->phy_id)) {
+			prtad = mdio_phy_id_prtad(data->phy_id);
+			devad = mdio_phy_id_devad(data->phy_id);
+		} else if (data->phy_id < 32) {
+			prtad = data->phy_id;
+			devad = 0;
+			data->reg_num &= 0x1f;
+		} else
+			return -EINVAL;
+
+		mbox = pi->adapter->pf;
+		if (cmd == SIOCGMIIREG)
+			ret = t4_mdio_rd(pi->adapter, mbox, prtad, devad,
+					 data->reg_num, &data->val_out);
+		else
+			ret = t4_mdio_wr(pi->adapter, mbox, prtad, devad,
+					 data->reg_num, data->val_in);
+		break;
+	case SIOCGHWTSTAMP:
+		return copy_to_user(req->ifr_data, &pi->tstamp_config,
+				    sizeof(pi->tstamp_config)) ?
+			-EFAULT : 0;
+	case SIOCSHWTSTAMP:
+		if (copy_from_user(&pi->tstamp_config, req->ifr_data,
+				   sizeof(pi->tstamp_config)))
+			return -EFAULT;
+
+		if (!is_t4(adapter->params.chip)) {
+			switch (pi->tstamp_config.tx_type) {
+			case HWTSTAMP_TX_OFF:
+			case HWTSTAMP_TX_ON:
+				break;
+			default:
+				return -ERANGE;
+			}
+
+			switch (pi->tstamp_config.rx_filter) {
+			case HWTSTAMP_FILTER_NONE:
+				pi->rxtstamp = false;
+				break;
+			case HWTSTAMP_FILTER_PTP_V1_L4_EVENT:
+			case HWTSTAMP_FILTER_PTP_V2_L4_EVENT:
+				cxgb4_ptprx_timestamping(pi, pi->port_id,
+							 PTP_TS_L4);
+				break;
+			case HWTSTAMP_FILTER_PTP_V2_EVENT:
+				cxgb4_ptprx_timestamping(pi, pi->port_id,
+							 PTP_TS_L2_L4);
+				break;
+			case HWTSTAMP_FILTER_ALL:
+			case HWTSTAMP_FILTER_PTP_V1_L4_SYNC:
+			case HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ:
+			case HWTSTAMP_FILTER_PTP_V2_L4_SYNC:
+			case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ:
+				pi->rxtstamp = true;
+				break;
+			default:
+				pi->tstamp_config.rx_filter =
+					HWTSTAMP_FILTER_NONE;
+				return -ERANGE;
+			}
+
+			if ((pi->tstamp_config.tx_type == HWTSTAMP_TX_OFF) &&
+			    (pi->tstamp_config.rx_filter ==
+				HWTSTAMP_FILTER_NONE)) {
+				if (cxgb4_ptp_txtype(adapter, pi->port_id) >= 0)
+					pi->ptp_enable = false;
+			}
+
+			if (pi->tstamp_config.rx_filter !=
+				HWTSTAMP_FILTER_NONE) {
+				if (cxgb4_ptp_redirect_rx_packet(adapter,
+								 pi) >= 0)
+					pi->ptp_enable = true;
+			}
+		} else {
+			/* For T4 Adapters */
+			switch (pi->tstamp_config.rx_filter) {
+			case HWTSTAMP_FILTER_NONE:
+			pi->rxtstamp = false;
+			break;
+			case HWTSTAMP_FILTER_ALL:
+			pi->rxtstamp = true;
+			break;
+			default:
+			pi->tstamp_config.rx_filter =
+			HWTSTAMP_FILTER_NONE;
+			return -ERANGE;
+			}
+		}
+		return copy_to_user(req->ifr_data, &pi->tstamp_config,
+				    sizeof(pi->tstamp_config)) ?
+			-EFAULT : 0;
+	default:
+		return -EOPNOTSUPP;
+	}
+	return ret;
+}
+
+static void cxgb_set_rxmode(struct net_device *dev)
+{
+	/* unfortunately we can't return errors to the stack */
+	set_rxmode(dev, -1, false);
+}
+
+static int cxgb_change_mtu(struct net_device *dev, int new_mtu)
+{
+	struct port_info *pi = netdev_priv(dev);
+	int ret;
+
+	ret = t4_set_rxmode(pi->adapter, pi->adapter->mbox, pi->viid,
+			    pi->viid_mirror, new_mtu, -1, -1, -1, -1, true);
+	if (!ret)
+		dev->mtu = new_mtu;
+	return ret;
+}
+
+#ifdef CONFIG_PCI_IOV
+static int cxgb4_mgmt_open(struct net_device *dev)
+{
+	/* Turn carrier off since we don't have to transmit anything on this
+	 * interface.
+	 */
+	netif_carrier_off(dev);
+	return 0;
+}
+
+/* Fill MAC address that will be assigned by the FW */
+static void cxgb4_mgmt_fill_vf_station_mac_addr(struct adapter *adap)
+{
+	u8 hw_addr[ETH_ALEN], macaddr[ETH_ALEN];
+	unsigned int i, vf, nvfs;
+	u16 a, b;
+	int err;
+	u8 *na;
+
+	err = t4_get_raw_vpd_params(adap, &adap->params.vpd);
+	if (err)
+		return;
+
+	na = adap->params.vpd.na;
+	for (i = 0; i < ETH_ALEN; i++)
+		hw_addr[i] = (hex2val(na[2 * i + 0]) * 16 +
+			      hex2val(na[2 * i + 1]));
+
+	a = (hw_addr[0] << 8) | hw_addr[1];
+	b = (hw_addr[1] << 8) | hw_addr[2];
+	a ^= b;
+	a |= 0x0200;    /* locally assigned Ethernet MAC address */
+	a &= ~0x0100;   /* not a multicast Ethernet MAC address */
+	macaddr[0] = a >> 8;
+	macaddr[1] = a & 0xff;
+
+	for (i = 2; i < 5; i++)
+		macaddr[i] = hw_addr[i + 1];
+
+	for (vf = 0, nvfs = pci_sriov_get_totalvfs(adap->pdev);
+		vf < nvfs; vf++) {
+		macaddr[5] = adap->pf * nvfs + vf;
+		ether_addr_copy(adap->vfinfo[vf].vf_mac_addr, macaddr);
+	}
+}
+
+static int cxgb4_mgmt_set_vf_mac(struct net_device *dev, int vf, u8 *mac)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	int ret;
+
+	/* verify MAC addr is valid */
+	if (!is_valid_ether_addr(mac)) {
+		dev_err(pi->adapter->pdev_dev,
+			"Invalid Ethernet address %pM for VF %d\n",
+			mac, vf);
+		return -EINVAL;
+	}
+
+	dev_info(pi->adapter->pdev_dev,
+		 "Setting MAC %pM on VF %d\n", mac, vf);
+	ret = t4_set_vf_mac_acl(adap, vf + 1, 1, mac);
+	if (!ret)
+		ether_addr_copy(adap->vfinfo[vf].vf_mac_addr, mac);
+	return ret;
+}
+
+static int cxgb4_mgmt_get_vf_config(struct net_device *dev,
+				    int vf, struct ifla_vf_info *ivi)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	struct vf_info *vfinfo;
+
+	if (vf >= adap->num_vfs)
+		return -EINVAL;
+	vfinfo = &adap->vfinfo[vf];
+
+	ivi->vf = vf;
+	ivi->max_tx_rate = vfinfo->tx_rate;
+	ivi->min_tx_rate = 0;
+	ether_addr_copy(ivi->mac, vfinfo->vf_mac_addr);
+	ivi->vlan = vfinfo->vlan;
+	ivi->linkstate = vfinfo->link_state;
+	return 0;
+}
+
+static int cxgb4_mgmt_get_phys_port_id(struct net_device *dev,
+				       struct netdev_phys_item_id *ppid)
+{
+	struct port_info *pi = netdev_priv(dev);
+	unsigned int phy_port_id;
+
+	phy_port_id = pi->adapter->adap_idx * 10 + pi->port_id;
+	ppid->id_len = sizeof(phy_port_id);
+	memcpy(ppid->id, &phy_port_id, ppid->id_len);
+	return 0;
+}
+
+static int cxgb4_mgmt_set_vf_rate(struct net_device *dev, int vf,
+				  int min_tx_rate, int max_tx_rate)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	unsigned int link_ok, speed, mtu;
+	u32 fw_pfvf, fw_class;
+	int class_id = vf;
+	int ret;
+	u16 pktsize;
+
+	if (vf >= adap->num_vfs)
+		return -EINVAL;
+
+	if (min_tx_rate) {
+		dev_err(adap->pdev_dev,
+			"Min tx rate (%d) (> 0) for VF %d is Invalid.\n",
+			min_tx_rate, vf);
+		return -EINVAL;
+	}
+
+	if (max_tx_rate == 0) {
+		/* unbind VF to to any Traffic Class */
+		fw_pfvf =
+		    (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_PFVF) |
+		     FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_PFVF_SCHEDCLASS_ETH));
+		fw_class = 0xffffffff;
+		ret = t4_set_params(adap, adap->mbox, adap->pf, vf + 1, 1,
+				    &fw_pfvf, &fw_class);
+		if (ret) {
+			dev_err(adap->pdev_dev,
+				"Err %d in unbinding PF %d VF %d from TX Rate Limiting\n",
+				ret, adap->pf, vf);
+			return -EINVAL;
+		}
+		dev_info(adap->pdev_dev,
+			 "PF %d VF %d is unbound from TX Rate Limiting\n",
+			 adap->pf, vf);
+		adap->vfinfo[vf].tx_rate = 0;
+		return 0;
+	}
+
+	ret = t4_get_link_params(pi, &link_ok, &speed, &mtu);
+	if (ret != FW_SUCCESS) {
+		dev_err(adap->pdev_dev,
+			"Failed to get link information for VF %d\n", vf);
+		return -EINVAL;
+	}
+
+	if (!link_ok) {
+		dev_err(adap->pdev_dev, "Link down for VF %d\n", vf);
+		return -EINVAL;
+	}
+
+	if (max_tx_rate > speed) {
+		dev_err(adap->pdev_dev,
+			"Max tx rate %d for VF %d can't be > link-speed %u",
+			max_tx_rate, vf, speed);
+		return -EINVAL;
+	}
+
+	pktsize = mtu;
+	/* subtract ethhdr size and 4 bytes crc since, f/w appends it */
+	pktsize = pktsize - sizeof(struct ethhdr) - 4;
+	/* subtract ipv4 hdr size, tcp hdr size to get typical IPv4 MSS size */
+	pktsize = pktsize - sizeof(struct iphdr) - sizeof(struct tcphdr);
+	/* configure Traffic Class for rate-limiting */
+	ret = t4_sched_params(adap, SCHED_CLASS_TYPE_PACKET,
+			      SCHED_CLASS_LEVEL_CL_RL,
+			      SCHED_CLASS_MODE_CLASS,
+			      SCHED_CLASS_RATEUNIT_BITS,
+			      SCHED_CLASS_RATEMODE_ABS,
+			      pi->tx_chan, class_id, 0,
+			      max_tx_rate * 1000, 0, pktsize, 0);
+	if (ret) {
+		dev_err(adap->pdev_dev, "Err %d for Traffic Class config\n",
+			ret);
+		return -EINVAL;
+	}
+	dev_info(adap->pdev_dev,
+		 "Class %d with MSS %u configured with rate %u\n",
+		 class_id, pktsize, max_tx_rate);
+
+	/* bind VF to configured Traffic Class */
+	fw_pfvf = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_PFVF) |
+		   FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_PFVF_SCHEDCLASS_ETH));
+	fw_class = class_id;
+	ret = t4_set_params(adap, adap->mbox, adap->pf, vf + 1, 1, &fw_pfvf,
+			    &fw_class);
+	if (ret) {
+		dev_err(adap->pdev_dev,
+			"Err %d in binding PF %d VF %d to Traffic Class %d\n",
+			ret, adap->pf, vf, class_id);
+		return -EINVAL;
+	}
+	dev_info(adap->pdev_dev, "PF %d VF %d is bound to Class %d\n",
+		 adap->pf, vf, class_id);
+	adap->vfinfo[vf].tx_rate = max_tx_rate;
+	return 0;
+}
+
+static int cxgb4_mgmt_set_vf_vlan(struct net_device *dev, int vf,
+				  u16 vlan, u8 qos, __be16 vlan_proto)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	int ret;
+
+	if (vf >= adap->num_vfs || vlan > 4095 || qos > 7)
+		return -EINVAL;
+
+	if (vlan_proto != htons(ETH_P_8021Q) || qos != 0)
+		return -EPROTONOSUPPORT;
+
+	ret = t4_set_vlan_acl(adap, adap->mbox, vf + 1, vlan);
+	if (!ret) {
+		adap->vfinfo[vf].vlan = vlan;
+		return 0;
+	}
+
+	dev_err(adap->pdev_dev, "Err %d %s VLAN ACL for PF/VF %d/%d\n",
+		ret, (vlan ? "setting" : "clearing"), adap->pf, vf);
+	return ret;
+}
+
+static int cxgb4_mgmt_set_vf_link_state(struct net_device *dev, int vf,
+					int link)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	u32 param, val;
+	int ret = 0;
+
+	if (vf >= adap->num_vfs)
+		return -EINVAL;
+
+	switch (link) {
+	case IFLA_VF_LINK_STATE_AUTO:
+		val = FW_VF_LINK_STATE_AUTO;
+		break;
+
+	case IFLA_VF_LINK_STATE_ENABLE:
+		val = FW_VF_LINK_STATE_ENABLE;
+		break;
+
+	case IFLA_VF_LINK_STATE_DISABLE:
+		val = FW_VF_LINK_STATE_DISABLE;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_PFVF) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_PFVF_LINK_STATE));
+	ret = t4_set_params(adap, adap->mbox, adap->pf, vf + 1, 1,
+			    &param, &val);
+	if (ret) {
+		dev_err(adap->pdev_dev,
+			"Error %d in setting PF %d VF %d link state\n",
+			ret, adap->pf, vf);
+		return -EINVAL;
+	}
+
+	adap->vfinfo[vf].link_state = link;
+	return ret;
+}
+#endif /* CONFIG_PCI_IOV */
+
+static int cxgb_set_mac_addr(struct net_device *dev, void *p)
+{
+	int ret;
+	struct sockaddr *addr = p;
+	struct port_info *pi = netdev_priv(dev);
+
+	if (!is_valid_ether_addr(addr->sa_data))
+		return -EADDRNOTAVAIL;
+
+	ret = cxgb4_update_mac_filt(pi, pi->viid, &pi->xact_addr_filt,
+				    addr->sa_data, true, &pi->smt_idx);
+	if (ret < 0)
+		return ret;
+
+	eth_hw_addr_set(dev, addr->sa_data);
+	return 0;
+}
+
+#ifdef CONFIG_NET_POLL_CONTROLLER
+static void cxgb_netpoll(struct net_device *dev)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+
+	if (adap->flags & CXGB4_USING_MSIX) {
+		int i;
+		struct sge_eth_rxq *rx = &adap->sge.ethrxq[pi->first_qset];
+
+		for (i = pi->nqsets; i; i--, rx++)
+			t4_sge_intr_msix(0, &rx->rspq);
+	} else
+		t4_intr_handler(adap)(0, adap);
+}
+#endif
+
+static int cxgb_set_tx_maxrate(struct net_device *dev, int index, u32 rate)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adap = pi->adapter;
+	struct ch_sched_queue qe = { 0 };
+	struct ch_sched_params p = { 0 };
+	struct sched_class *e;
+	u32 req_rate;
+	int err = 0;
+
+	if (!can_sched(dev))
+		return -ENOTSUPP;
+
+	if (index < 0 || index > pi->nqsets - 1)
+		return -EINVAL;
+
+	if (!(adap->flags & CXGB4_FULL_INIT_DONE)) {
+		dev_err(adap->pdev_dev,
+			"Failed to rate limit on queue %d. Link Down?\n",
+			index);
+		return -EINVAL;
+	}
+
+	qe.queue = index;
+	e = cxgb4_sched_queue_lookup(dev, &qe);
+	if (e && e->info.u.params.level != SCHED_CLASS_LEVEL_CL_RL) {
+		dev_err(adap->pdev_dev,
+			"Queue %u already bound to class %u of type: %u\n",
+			index, e->idx, e->info.u.params.level);
+		return -EBUSY;
+	}
+
+	/* Convert from Mbps to Kbps */
+	req_rate = rate * 1000;
+
+	/* Max rate is 100 Gbps */
+	if (req_rate > SCHED_MAX_RATE_KBPS) {
+		dev_err(adap->pdev_dev,
+			"Invalid rate %u Mbps, Max rate is %u Mbps\n",
+			rate, SCHED_MAX_RATE_KBPS / 1000);
+		return -ERANGE;
+	}
+
+	/* First unbind the queue from any existing class */
+	memset(&qe, 0, sizeof(qe));
+	qe.queue = index;
+	qe.class = SCHED_CLS_NONE;
+
+	err = cxgb4_sched_class_unbind(dev, (void *)(&qe), SCHED_QUEUE);
+	if (err) {
+		dev_err(adap->pdev_dev,
+			"Unbinding Queue %d on port %d fail. Err: %d\n",
+			index, pi->port_id, err);
+		return err;
+	}
+
+	/* Queue already unbound */
+	if (!req_rate)
+		return 0;
+
+	/* Fetch any available unused or matching scheduling class */
+	p.type = SCHED_CLASS_TYPE_PACKET;
+	p.u.params.level    = SCHED_CLASS_LEVEL_CL_RL;
+	p.u.params.mode     = SCHED_CLASS_MODE_CLASS;
+	p.u.params.rateunit = SCHED_CLASS_RATEUNIT_BITS;
+	p.u.params.ratemode = SCHED_CLASS_RATEMODE_ABS;
+	p.u.params.channel  = pi->tx_chan;
+	p.u.params.class    = SCHED_CLS_NONE;
+	p.u.params.minrate  = 0;
+	p.u.params.maxrate  = req_rate;
+	p.u.params.weight   = 0;
+	p.u.params.pktsize  = dev->mtu;
+
+	e = cxgb4_sched_class_alloc(dev, &p);
+	if (!e)
+		return -ENOMEM;
+
+	/* Bind the queue to a scheduling class */
+	memset(&qe, 0, sizeof(qe));
+	qe.queue = index;
+	qe.class = e->idx;
+
+	err = cxgb4_sched_class_bind(dev, (void *)(&qe), SCHED_QUEUE);
+	if (err)
+		dev_err(adap->pdev_dev,
+			"Queue rate limiting failed. Err: %d\n", err);
+	return err;
+}
+
+static int cxgb_setup_tc_flower(struct net_device *dev,
+				struct flow_cls_offload *cls_flower)
+{
+	switch (cls_flower->command) {
+	case FLOW_CLS_REPLACE:
+		return cxgb4_tc_flower_replace(dev, cls_flower);
+	case FLOW_CLS_DESTROY:
+		return cxgb4_tc_flower_destroy(dev, cls_flower);
+	case FLOW_CLS_STATS:
+		return cxgb4_tc_flower_stats(dev, cls_flower);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static int cxgb_setup_tc_cls_u32(struct net_device *dev,
+				 struct tc_cls_u32_offload *cls_u32)
+{
+	switch (cls_u32->command) {
+	case TC_CLSU32_NEW_KNODE:
+	case TC_CLSU32_REPLACE_KNODE:
+		return cxgb4_config_knode(dev, cls_u32);
+	case TC_CLSU32_DELETE_KNODE:
+		return cxgb4_delete_knode(dev, cls_u32);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static int cxgb_setup_tc_matchall(struct net_device *dev,
+				  struct tc_cls_matchall_offload *cls_matchall,
+				  bool ingress)
+{
+	struct adapter *adap = netdev2adap(dev);
+
+	if (!adap->tc_matchall)
+		return -ENOMEM;
+
+	switch (cls_matchall->command) {
+	case TC_CLSMATCHALL_REPLACE:
+		return cxgb4_tc_matchall_replace(dev, cls_matchall, ingress);
+	case TC_CLSMATCHALL_DESTROY:
+		return cxgb4_tc_matchall_destroy(dev, cls_matchall, ingress);
+	case TC_CLSMATCHALL_STATS:
+		if (ingress)
+			return cxgb4_tc_matchall_stats(dev, cls_matchall);
+		break;
+	default:
+		break;
+	}
+
+	return -EOPNOTSUPP;
+}
+
+static int cxgb_setup_tc_block_ingress_cb(enum tc_setup_type type,
+					  void *type_data, void *cb_priv)
+{
+	struct net_device *dev = cb_priv;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+
+	if (!(adap->flags & CXGB4_FULL_INIT_DONE)) {
+		dev_err(adap->pdev_dev,
+			"Failed to setup tc on port %d. Link Down?\n",
+			pi->port_id);
+		return -EINVAL;
+	}
+
+	if (!tc_cls_can_offload_and_chain0(dev, type_data))
+		return -EOPNOTSUPP;
+
+	switch (type) {
+	case TC_SETUP_CLSU32:
+		return cxgb_setup_tc_cls_u32(dev, type_data);
+	case TC_SETUP_CLSFLOWER:
+		return cxgb_setup_tc_flower(dev, type_data);
+	case TC_SETUP_CLSMATCHALL:
+		return cxgb_setup_tc_matchall(dev, type_data, true);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static int cxgb_setup_tc_block_egress_cb(enum tc_setup_type type,
+					 void *type_data, void *cb_priv)
+{
+	struct net_device *dev = cb_priv;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+
+	if (!(adap->flags & CXGB4_FULL_INIT_DONE)) {
+		dev_err(adap->pdev_dev,
+			"Failed to setup tc on port %d. Link Down?\n",
+			pi->port_id);
+		return -EINVAL;
+	}
+
+	if (!tc_cls_can_offload_and_chain0(dev, type_data))
+		return -EOPNOTSUPP;
+
+	switch (type) {
+	case TC_SETUP_CLSMATCHALL:
+		return cxgb_setup_tc_matchall(dev, type_data, false);
+	default:
+		break;
+	}
+
+	return -EOPNOTSUPP;
+}
+
+static int cxgb_setup_tc_mqprio(struct net_device *dev,
+				struct tc_mqprio_qopt_offload *mqprio)
+{
+	struct adapter *adap = netdev2adap(dev);
+
+	if (!is_ethofld(adap) || !adap->tc_mqprio)
+		return -ENOMEM;
+
+	return cxgb4_setup_tc_mqprio(dev, mqprio);
+}
+
+static LIST_HEAD(cxgb_block_cb_list);
+
+static int cxgb_setup_tc_block(struct net_device *dev,
+			       struct flow_block_offload *f)
+{
+	struct port_info *pi = netdev_priv(dev);
+	flow_setup_cb_t *cb;
+	bool ingress_only;
+
+	pi->tc_block_shared = f->block_shared;
+	if (f->binder_type == FLOW_BLOCK_BINDER_TYPE_CLSACT_EGRESS) {
+		cb = cxgb_setup_tc_block_egress_cb;
+		ingress_only = false;
+	} else {
+		cb = cxgb_setup_tc_block_ingress_cb;
+		ingress_only = true;
+	}
+
+	return flow_block_cb_setup_simple(f, &cxgb_block_cb_list,
+					  cb, pi, dev, ingress_only);
+}
+
+static int cxgb_setup_tc(struct net_device *dev, enum tc_setup_type type,
+			 void *type_data)
+{
+	switch (type) {
+	case TC_SETUP_QDISC_MQPRIO:
+		return cxgb_setup_tc_mqprio(dev, type_data);
+	case TC_SETUP_BLOCK:
+		return cxgb_setup_tc_block(dev, type_data);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static int cxgb_udp_tunnel_unset_port(struct net_device *netdev,
+				      unsigned int table, unsigned int entry,
+				      struct udp_tunnel_info *ti)
+{
+	struct port_info *pi = netdev_priv(netdev);
+	struct adapter *adapter = pi->adapter;
+	u8 match_all_mac[] = { 0, 0, 0, 0, 0, 0 };
+	int ret = 0, i;
+
+	switch (ti->type) {
+	case UDP_TUNNEL_TYPE_VXLAN:
+		adapter->vxlan_port = 0;
+		t4_write_reg(adapter, MPS_RX_VXLAN_TYPE_A, 0);
+		break;
+	case UDP_TUNNEL_TYPE_GENEVE:
+		adapter->geneve_port = 0;
+		t4_write_reg(adapter, MPS_RX_GENEVE_TYPE_A, 0);
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	/* Matchall mac entries can be deleted only after all tunnel ports
+	 * are brought down or removed.
+	 */
+	if (!adapter->rawf_cnt)
+		return 0;
+	for_each_port(adapter, i) {
+		pi = adap2pinfo(adapter, i);
+		ret = t4_free_raw_mac_filt(adapter, pi->viid,
+					   match_all_mac, match_all_mac,
+					   adapter->rawf_start + pi->port_id,
+					   1, pi->port_id, false);
+		if (ret < 0) {
+			netdev_info(netdev, "Failed to free mac filter entry, for port %d\n",
+				    i);
+			return ret;
+		}
+	}
+
+	return 0;
+}
+
+static int cxgb_udp_tunnel_set_port(struct net_device *netdev,
+				    unsigned int table, unsigned int entry,
+				    struct udp_tunnel_info *ti)
+{
+	struct port_info *pi = netdev_priv(netdev);
+	struct adapter *adapter = pi->adapter;
+	u8 match_all_mac[] = { 0, 0, 0, 0, 0, 0 };
+	int i, ret;
+
+	switch (ti->type) {
+	case UDP_TUNNEL_TYPE_VXLAN:
+		adapter->vxlan_port = ti->port;
+		t4_write_reg(adapter, MPS_RX_VXLAN_TYPE_A,
+			     VXLAN_V(be16_to_cpu(ti->port)) | VXLAN_EN_F);
+		break;
+	case UDP_TUNNEL_TYPE_GENEVE:
+		adapter->geneve_port = ti->port;
+		t4_write_reg(adapter, MPS_RX_GENEVE_TYPE_A,
+			     GENEVE_V(be16_to_cpu(ti->port)) | GENEVE_EN_F);
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	/* Create a 'match all' mac filter entry for inner mac,
+	 * if raw mac interface is supported. Once the linux kernel provides
+	 * driver entry points for adding/deleting the inner mac addresses,
+	 * we will remove this 'match all' entry and fallback to adding
+	 * exact match filters.
+	 */
+	for_each_port(adapter, i) {
+		pi = adap2pinfo(adapter, i);
+
+		ret = t4_alloc_raw_mac_filt(adapter, pi->viid,
+					    match_all_mac,
+					    match_all_mac,
+					    adapter->rawf_start + pi->port_id,
+					    1, pi->port_id, false);
+		if (ret < 0) {
+			netdev_info(netdev, "Failed to allocate a mac filter entry, not adding port %d\n",
+				    be16_to_cpu(ti->port));
+			return ret;
+		}
+	}
+
+	return 0;
+}
+
+static const struct udp_tunnel_nic_info cxgb_udp_tunnels = {
+	.set_port	= cxgb_udp_tunnel_set_port,
+	.unset_port	= cxgb_udp_tunnel_unset_port,
+	.tables		= {
+		{ .n_entries = 1, .tunnel_types = UDP_TUNNEL_TYPE_VXLAN,  },
+		{ .n_entries = 1, .tunnel_types = UDP_TUNNEL_TYPE_GENEVE, },
+	},
+};
+
+static netdev_features_t cxgb_features_check(struct sk_buff *skb,
+					     struct net_device *dev,
+					     netdev_features_t features)
+{
+	struct port_info *pi = netdev_priv(dev);
+	struct adapter *adapter = pi->adapter;
+
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) < CHELSIO_T6)
+		return features;
+
+	/* Check if hw supports offload for this packet */
+	if (!skb->encapsulation || cxgb_encap_offload_supported(skb))
+		return features;
+
+	/* Offload is not supported for this encapsulated packet */
+	return features & ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK);
+}
+
+static netdev_features_t cxgb_fix_features(struct net_device *dev,
+					   netdev_features_t features)
+{
+	/* Disable GRO, if RX_CSUM is disabled */
+	if (!(features & NETIF_F_RXCSUM))
+		features &= ~NETIF_F_GRO;
+
+	return features;
+}
+
+static const struct net_device_ops cxgb4_netdev_ops = {
+	.ndo_open             = cxgb_open,
+	.ndo_stop             = cxgb_close,
+	.ndo_start_xmit       = t4_start_xmit,
+	.ndo_select_queue     =	cxgb_select_queue,
+	.ndo_get_stats64      = cxgb_get_stats,
+	.ndo_set_rx_mode      = cxgb_set_rxmode,
+	.ndo_set_mac_address  = cxgb_set_mac_addr,
+	.ndo_set_features     = cxgb_set_features,
+	.ndo_validate_addr    = eth_validate_addr,
+	.ndo_eth_ioctl         = cxgb_ioctl,
+	.ndo_change_mtu       = cxgb_change_mtu,
+#ifdef CONFIG_NET_POLL_CONTROLLER
+	.ndo_poll_controller  = cxgb_netpoll,
+#endif
+#ifdef CONFIG_CHELSIO_T4_FCOE
+	.ndo_fcoe_enable      = cxgb_fcoe_enable,
+	.ndo_fcoe_disable     = cxgb_fcoe_disable,
+#endif /* CONFIG_CHELSIO_T4_FCOE */
+	.ndo_set_tx_maxrate   = cxgb_set_tx_maxrate,
+	.ndo_setup_tc         = cxgb_setup_tc,
+	.ndo_features_check   = cxgb_features_check,
+	.ndo_fix_features     = cxgb_fix_features,
+};
+
+#ifdef CONFIG_PCI_IOV
+static const struct net_device_ops cxgb4_mgmt_netdev_ops = {
+	.ndo_open               = cxgb4_mgmt_open,
+	.ndo_set_vf_mac         = cxgb4_mgmt_set_vf_mac,
+	.ndo_get_vf_config      = cxgb4_mgmt_get_vf_config,
+	.ndo_set_vf_rate        = cxgb4_mgmt_set_vf_rate,
+	.ndo_get_phys_port_id   = cxgb4_mgmt_get_phys_port_id,
+	.ndo_set_vf_vlan        = cxgb4_mgmt_set_vf_vlan,
+	.ndo_set_vf_link_state	= cxgb4_mgmt_set_vf_link_state,
+};
+
+static void cxgb4_mgmt_get_drvinfo(struct net_device *dev,
+				   struct ethtool_drvinfo *info)
+{
+	struct adapter *adapter = netdev2adap(dev);
+
+	strscpy(info->driver, cxgb4_driver_name, sizeof(info->driver));
+	strscpy(info->bus_info, pci_name(adapter->pdev),
+		sizeof(info->bus_info));
+}
+
+static const struct ethtool_ops cxgb4_mgmt_ethtool_ops = {
+	.get_drvinfo       = cxgb4_mgmt_get_drvinfo,
+};
+#endif
+
+static void notify_fatal_err(struct work_struct *work)
+{
+	struct adapter *adap;
+
+	adap = container_of(work, struct adapter, fatal_err_notify_task);
+	notify_ulds(adap, CXGB4_STATE_FATAL_ERROR);
+}
+
+void t4_fatal_err(struct adapter *adap)
+{
+	int port;
+
+	if (pci_channel_offline(adap->pdev))
+		return;
+
+	/* Disable the SGE since ULDs are going to free resources that
+	 * could be exposed to the adapter.  RDMA MWs for example...
+	 */
+	t4_shutdown_adapter(adap);
+	for_each_port(adap, port) {
+		struct net_device *dev = adap->port[port];
+
+		/* If we get here in very early initialization the network
+		 * devices may not have been set up yet.
+		 */
+		if (!dev)
+			continue;
+
+		netif_tx_stop_all_queues(dev);
+		netif_carrier_off(dev);
+	}
+	dev_alert(adap->pdev_dev, "encountered fatal error, adapter stopped\n");
+	queue_work(adap->workq, &adap->fatal_err_notify_task);
+}
+
+static void setup_memwin(struct adapter *adap)
+{
+	u32 nic_win_base = t4_get_util_window(adap);
+
+	t4_setup_memwin(adap, nic_win_base, MEMWIN_NIC);
+}
+
+static void setup_memwin_rdma(struct adapter *adap)
+{
+	if (adap->vres.ocq.size) {
+		u32 start;
+		unsigned int sz_kb;
+
+		start = t4_read_pcie_cfg4(adap, PCI_BASE_ADDRESS_2);
+		start &= PCI_BASE_ADDRESS_MEM_MASK;
+		start += OCQ_WIN_OFFSET(adap->pdev, &adap->vres);
+		sz_kb = roundup_pow_of_two(adap->vres.ocq.size) >> 10;
+		t4_write_reg(adap,
+			     PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN_A, 3),
+			     start | BIR_V(1) | WINDOW_V(ilog2(sz_kb)));
+		t4_write_reg(adap,
+			     PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_OFFSET_A, 3),
+			     adap->vres.ocq.start);
+		t4_read_reg(adap,
+			    PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_OFFSET_A, 3));
+	}
+}
+
+/* HMA Definitions */
+
+/* The maximum number of address that can be send in a single FW cmd */
+#define HMA_MAX_ADDR_IN_CMD	5
+
+#define HMA_PAGE_SIZE		PAGE_SIZE
+
+#define HMA_MAX_NO_FW_ADDRESS	(16 << 10)  /* FW supports 16K addresses */
+
+#define HMA_PAGE_ORDER					\
+	((HMA_PAGE_SIZE < HMA_MAX_NO_FW_ADDRESS) ?	\
+	ilog2(HMA_MAX_NO_FW_ADDRESS / HMA_PAGE_SIZE) : 0)
+
+/* The minimum and maximum possible HMA sizes that can be specified in the FW
+ * configuration(in units of MB).
+ */
+#define HMA_MIN_TOTAL_SIZE	1
+#define HMA_MAX_TOTAL_SIZE				\
+	(((HMA_PAGE_SIZE << HMA_PAGE_ORDER) *		\
+	  HMA_MAX_NO_FW_ADDRESS) >> 20)
+
+static void adap_free_hma_mem(struct adapter *adapter)
+{
+	struct scatterlist *iter;
+	struct page *page;
+	int i;
+
+	if (!adapter->hma.sgt)
+		return;
+
+	if (adapter->hma.flags & HMA_DMA_MAPPED_FLAG) {
+		dma_unmap_sg(adapter->pdev_dev, adapter->hma.sgt->sgl,
+			     adapter->hma.sgt->nents, DMA_BIDIRECTIONAL);
+		adapter->hma.flags &= ~HMA_DMA_MAPPED_FLAG;
+	}
+
+	for_each_sg(adapter->hma.sgt->sgl, iter,
+		    adapter->hma.sgt->orig_nents, i) {
+		page = sg_page(iter);
+		if (page)
+			__free_pages(page, HMA_PAGE_ORDER);
+	}
+
+	kfree(adapter->hma.phy_addr);
+	sg_free_table(adapter->hma.sgt);
+	kfree(adapter->hma.sgt);
+	adapter->hma.sgt = NULL;
+}
+
+static int adap_config_hma(struct adapter *adapter)
+{
+	struct scatterlist *sgl, *iter;
+	struct sg_table *sgt;
+	struct page *newpage;
+	unsigned int i, j, k;
+	u32 param, hma_size;
+	unsigned int ncmds;
+	size_t page_size;
+	u32 page_order;
+	int node, ret;
+
+	/* HMA is supported only for T6+ cards.
+	 * Avoid initializing HMA in kdump kernels.
+	 */
+	if (is_kdump_kernel() ||
+	    CHELSIO_CHIP_VERSION(adapter->params.chip) < CHELSIO_T6)
+		return 0;
+
+	/* Get the HMA region size required by fw */
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_HMA_SIZE));
+	ret = t4_query_params(adapter, adapter->mbox, adapter->pf, 0,
+			      1, &param, &hma_size);
+	/* An error means card has its own memory or HMA is not supported by
+	 * the firmware. Return without any errors.
+	 */
+	if (ret || !hma_size)
+		return 0;
+
+	if (hma_size < HMA_MIN_TOTAL_SIZE ||
+	    hma_size > HMA_MAX_TOTAL_SIZE) {
+		dev_err(adapter->pdev_dev,
+			"HMA size %uMB beyond bounds(%u-%lu)MB\n",
+			hma_size, HMA_MIN_TOTAL_SIZE, HMA_MAX_TOTAL_SIZE);
+		return -EINVAL;
+	}
+
+	page_size = HMA_PAGE_SIZE;
+	page_order = HMA_PAGE_ORDER;
+	adapter->hma.sgt = kzalloc(sizeof(*adapter->hma.sgt), GFP_KERNEL);
+	if (unlikely(!adapter->hma.sgt)) {
+		dev_err(adapter->pdev_dev, "HMA SG table allocation failed\n");
+		return -ENOMEM;
+	}
+	sgt = adapter->hma.sgt;
+	/* FW returned value will be in MB's
+	 */
+	sgt->orig_nents = (hma_size << 20) / (page_size << page_order);
+	if (sg_alloc_table(sgt, sgt->orig_nents, GFP_KERNEL)) {
+		dev_err(adapter->pdev_dev, "HMA SGL allocation failed\n");
+		kfree(adapter->hma.sgt);
+		adapter->hma.sgt = NULL;
+		return -ENOMEM;
+	}
+
+	sgl = adapter->hma.sgt->sgl;
+	node = dev_to_node(adapter->pdev_dev);
+	for_each_sg(sgl, iter, sgt->orig_nents, i) {
+		newpage = alloc_pages_node(node, __GFP_NOWARN | GFP_KERNEL |
+					   __GFP_ZERO, page_order);
+		if (!newpage) {
+			dev_err(adapter->pdev_dev,
+				"Not enough memory for HMA page allocation\n");
+			ret = -ENOMEM;
+			goto free_hma;
+		}
+		sg_set_page(iter, newpage, page_size << page_order, 0);
+	}
+
+	sgt->nents = dma_map_sg(adapter->pdev_dev, sgl, sgt->orig_nents,
+				DMA_BIDIRECTIONAL);
+	if (!sgt->nents) {
+		dev_err(adapter->pdev_dev,
+			"Not enough memory for HMA DMA mapping");
+		ret = -ENOMEM;
+		goto free_hma;
+	}
+	adapter->hma.flags |= HMA_DMA_MAPPED_FLAG;
+
+	adapter->hma.phy_addr = kcalloc(sgt->nents, sizeof(dma_addr_t),
+					GFP_KERNEL);
+	if (unlikely(!adapter->hma.phy_addr))
+		goto free_hma;
+
+	for_each_sg(sgl, iter, sgt->nents, i) {
+		newpage = sg_page(iter);
+		adapter->hma.phy_addr[i] = sg_dma_address(iter);
+	}
+
+	ncmds = DIV_ROUND_UP(sgt->nents, HMA_MAX_ADDR_IN_CMD);
+	/* Pass on the addresses to firmware */
+	for (i = 0, k = 0; i < ncmds; i++, k += HMA_MAX_ADDR_IN_CMD) {
+		struct fw_hma_cmd hma_cmd;
+		u8 naddr = HMA_MAX_ADDR_IN_CMD;
+		u8 soc = 0, eoc = 0;
+		u8 hma_mode = 1; /* Presently we support only Page table mode */
+
+		soc = (i == 0) ? 1 : 0;
+		eoc = (i == ncmds - 1) ? 1 : 0;
+
+		/* For last cmd, set naddr corresponding to remaining
+		 * addresses
+		 */
+		if (i == ncmds - 1) {
+			naddr = sgt->nents % HMA_MAX_ADDR_IN_CMD;
+			naddr = naddr ? naddr : HMA_MAX_ADDR_IN_CMD;
+		}
+		memset(&hma_cmd, 0, sizeof(hma_cmd));
+		hma_cmd.op_pkd = htonl(FW_CMD_OP_V(FW_HMA_CMD) |
+				       FW_CMD_REQUEST_F | FW_CMD_WRITE_F);
+		hma_cmd.retval_len16 = htonl(FW_LEN16(hma_cmd));
+
+		hma_cmd.mode_to_pcie_params =
+			htonl(FW_HMA_CMD_MODE_V(hma_mode) |
+			      FW_HMA_CMD_SOC_V(soc) | FW_HMA_CMD_EOC_V(eoc));
+
+		/* HMA cmd size specified in MB's */
+		hma_cmd.naddr_size =
+			htonl(FW_HMA_CMD_SIZE_V(hma_size) |
+			      FW_HMA_CMD_NADDR_V(naddr));
+
+		/* Total Page size specified in units of 4K */
+		hma_cmd.addr_size_pkd =
+			htonl(FW_HMA_CMD_ADDR_SIZE_V
+				((page_size << page_order) >> 12));
+
+		/* Fill the 5 addresses */
+		for (j = 0; j < naddr; j++) {
+			hma_cmd.phy_address[j] =
+				cpu_to_be64(adapter->hma.phy_addr[j + k]);
+		}
+		ret = t4_wr_mbox(adapter, adapter->mbox, &hma_cmd,
+				 sizeof(hma_cmd), &hma_cmd);
+		if (ret) {
+			dev_err(adapter->pdev_dev,
+				"HMA FW command failed with err %d\n", ret);
+			goto free_hma;
+		}
+	}
+
+	if (!ret)
+		dev_info(adapter->pdev_dev,
+			 "Reserved %uMB host memory for HMA\n", hma_size);
+	return ret;
+
+free_hma:
+	adap_free_hma_mem(adapter);
+	return ret;
+}
+
+static int adap_init1(struct adapter *adap, struct fw_caps_config_cmd *c)
+{
+	u32 v;
+	int ret;
+
+	/* Now that we've successfully configured and initialized the adapter
+	 * can ask the Firmware what resources it has provisioned for us.
+	 */
+	ret = t4_get_pfres(adap);
+	if (ret) {
+		dev_err(adap->pdev_dev,
+			"Unable to retrieve resource provisioning information\n");
+		return ret;
+	}
+
+	/* get device capabilities */
+	memset(c, 0, sizeof(*c));
+	c->op_to_write = htonl(FW_CMD_OP_V(FW_CAPS_CONFIG_CMD) |
+			       FW_CMD_REQUEST_F | FW_CMD_READ_F);
+	c->cfvalid_to_len16 = htonl(FW_LEN16(*c));
+	ret = t4_wr_mbox(adap, adap->mbox, c, sizeof(*c), c);
+	if (ret < 0)
+		return ret;
+
+	c->op_to_write = htonl(FW_CMD_OP_V(FW_CAPS_CONFIG_CMD) |
+			       FW_CMD_REQUEST_F | FW_CMD_WRITE_F);
+	ret = t4_wr_mbox(adap, adap->mbox, c, sizeof(*c), NULL);
+	if (ret < 0)
+		return ret;
+
+	ret = t4_config_glbl_rss(adap, adap->pf,
+				 FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL,
+				 FW_RSS_GLB_CONFIG_CMD_TNLMAPEN_F |
+				 FW_RSS_GLB_CONFIG_CMD_TNLALLLKP_F);
+	if (ret < 0)
+		return ret;
+
+	ret = t4_cfg_pfvf(adap, adap->mbox, adap->pf, 0, adap->sge.egr_sz, 64,
+			  MAX_INGQ, 0, 0, 4, 0xf, 0xf, 16, FW_CMD_CAP_PF,
+			  FW_CMD_CAP_PF);
+	if (ret < 0)
+		return ret;
+
+	t4_sge_init(adap);
+
+	/* tweak some settings */
+	t4_write_reg(adap, TP_SHIFT_CNT_A, 0x64f8849);
+	t4_write_reg(adap, ULP_RX_TDDP_PSZ_A, HPZ0_V(PAGE_SHIFT - 12));
+	t4_write_reg(adap, TP_PIO_ADDR_A, TP_INGRESS_CONFIG_A);
+	v = t4_read_reg(adap, TP_PIO_DATA_A);
+	t4_write_reg(adap, TP_PIO_DATA_A, v & ~CSUM_HAS_PSEUDO_HDR_F);
+
+	/* first 4 Tx modulation queues point to consecutive Tx channels */
+	adap->params.tp.tx_modq_map = 0xE4;
+	t4_write_reg(adap, TP_TX_MOD_QUEUE_REQ_MAP_A,
+		     TX_MOD_QUEUE_REQ_MAP_V(adap->params.tp.tx_modq_map));
+
+	/* associate each Tx modulation queue with consecutive Tx channels */
+	v = 0x84218421;
+	t4_write_indirect(adap, TP_PIO_ADDR_A, TP_PIO_DATA_A,
+			  &v, 1, TP_TX_SCHED_HDR_A);
+	t4_write_indirect(adap, TP_PIO_ADDR_A, TP_PIO_DATA_A,
+			  &v, 1, TP_TX_SCHED_FIFO_A);
+	t4_write_indirect(adap, TP_PIO_ADDR_A, TP_PIO_DATA_A,
+			  &v, 1, TP_TX_SCHED_PCMD_A);
+
+#define T4_TX_MODQ_10G_WEIGHT_DEFAULT 16 /* in KB units */
+	if (is_offload(adap)) {
+		t4_write_reg(adap, TP_TX_MOD_QUEUE_WEIGHT0_A,
+			     TX_MODQ_WEIGHT0_V(T4_TX_MODQ_10G_WEIGHT_DEFAULT) |
+			     TX_MODQ_WEIGHT1_V(T4_TX_MODQ_10G_WEIGHT_DEFAULT) |
+			     TX_MODQ_WEIGHT2_V(T4_TX_MODQ_10G_WEIGHT_DEFAULT) |
+			     TX_MODQ_WEIGHT3_V(T4_TX_MODQ_10G_WEIGHT_DEFAULT));
+		t4_write_reg(adap, TP_TX_MOD_CHANNEL_WEIGHT_A,
+			     TX_MODQ_WEIGHT0_V(T4_TX_MODQ_10G_WEIGHT_DEFAULT) |
+			     TX_MODQ_WEIGHT1_V(T4_TX_MODQ_10G_WEIGHT_DEFAULT) |
+			     TX_MODQ_WEIGHT2_V(T4_TX_MODQ_10G_WEIGHT_DEFAULT) |
+			     TX_MODQ_WEIGHT3_V(T4_TX_MODQ_10G_WEIGHT_DEFAULT));
+	}
+
+	/* get basic stuff going */
+	return t4_early_init(adap, adap->pf);
+}
+
+/*
+ * Max # of ATIDs.  The absolute HW max is 16K but we keep it lower.
+ */
+#define MAX_ATIDS 8192U
+
+/*
+ * Phase 0 of initialization: contact FW, obtain config, perform basic init.
+ *
+ * If the firmware we're dealing with has Configuration File support, then
+ * we use that to perform all configuration
+ */
+
+/*
+ * Tweak configuration based on module parameters, etc.  Most of these have
+ * defaults assigned to them by Firmware Configuration Files (if we're using
+ * them) but need to be explicitly set if we're using hard-coded
+ * initialization.  But even in the case of using Firmware Configuration
+ * Files, we'd like to expose the ability to change these via module
+ * parameters so these are essentially common tweaks/settings for
+ * Configuration Files and hard-coded initialization ...
+ */
+static int adap_init0_tweaks(struct adapter *adapter)
+{
+	/*
+	 * Fix up various Host-Dependent Parameters like Page Size, Cache
+	 * Line Size, etc.  The firmware default is for a 4KB Page Size and
+	 * 64B Cache Line Size ...
+	 */
+	t4_fixup_host_params(adapter, PAGE_SIZE, L1_CACHE_BYTES);
+
+	/*
+	 * Process module parameters which affect early initialization.
+	 */
+	if (rx_dma_offset != 2 && rx_dma_offset != 0) {
+		dev_err(&adapter->pdev->dev,
+			"Ignoring illegal rx_dma_offset=%d, using 2\n",
+			rx_dma_offset);
+		rx_dma_offset = 2;
+	}
+	t4_set_reg_field(adapter, SGE_CONTROL_A,
+			 PKTSHIFT_V(PKTSHIFT_M),
+			 PKTSHIFT_V(rx_dma_offset));
+
+	/*
+	 * Don't include the "IP Pseudo Header" in CPL_RX_PKT checksums: Linux
+	 * adds the pseudo header itself.
+	 */
+	t4_tp_wr_bits_indirect(adapter, TP_INGRESS_CONFIG_A,
+			       CSUM_HAS_PSEUDO_HDR_F, 0);
+
+	return 0;
+}
+
+/* 10Gb/s-BT PHY Support. chip-external 10Gb/s-BT PHYs are complex chips
+ * unto themselves and they contain their own firmware to perform their
+ * tasks ...
+ */
+static int phy_aq1202_version(const u8 *phy_fw_data,
+			      size_t phy_fw_size)
+{
+	int offset;
+
+	/* At offset 0x8 you're looking for the primary image's
+	 * starting offset which is 3 Bytes wide
+	 *
+	 * At offset 0xa of the primary image, you look for the offset
+	 * of the DRAM segment which is 3 Bytes wide.
+	 *
+	 * The FW version is at offset 0x27e of the DRAM and is 2 Bytes
+	 * wide
+	 */
+	#define be16(__p) (((__p)[0] << 8) | (__p)[1])
+	#define le16(__p) ((__p)[0] | ((__p)[1] << 8))
+	#define le24(__p) (le16(__p) | ((__p)[2] << 16))
+
+	offset = le24(phy_fw_data + 0x8) << 12;
+	offset = le24(phy_fw_data + offset + 0xa);
+	return be16(phy_fw_data + offset + 0x27e);
+
+	#undef be16
+	#undef le16
+	#undef le24
+}
+
+static struct info_10gbt_phy_fw {
+	unsigned int phy_fw_id;		/* PCI Device ID */
+	char *phy_fw_file;		/* /lib/firmware/ PHY Firmware file */
+	int (*phy_fw_version)(const u8 *phy_fw_data, size_t phy_fw_size);
+	int phy_flash;			/* Has FLASH for PHY Firmware */
+} phy_info_array[] = {
+	{
+		PHY_AQ1202_DEVICEID,
+		PHY_AQ1202_FIRMWARE,
+		phy_aq1202_version,
+		1,
+	},
+	{
+		PHY_BCM84834_DEVICEID,
+		PHY_BCM84834_FIRMWARE,
+		NULL,
+		0,
+	},
+	{ 0, NULL, NULL },
+};
+
+static struct info_10gbt_phy_fw *find_phy_info(int devid)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(phy_info_array); i++) {
+		if (phy_info_array[i].phy_fw_id == devid)
+			return &phy_info_array[i];
+	}
+	return NULL;
+}
+
+/* Handle updating of chip-external 10Gb/s-BT PHY firmware.  This needs to
+ * happen after the FW_RESET_CMD but before the FW_INITIALIZE_CMD.  On error
+ * we return a negative error number.  If we transfer new firmware we return 1
+ * (from t4_load_phy_fw()).  If we don't do anything we return 0.
+ */
+static int adap_init0_phy(struct adapter *adap)
+{
+	const struct firmware *phyf;
+	int ret;
+	struct info_10gbt_phy_fw *phy_info;
+
+	/* Use the device ID to determine which PHY file to flash.
+	 */
+	phy_info = find_phy_info(adap->pdev->device);
+	if (!phy_info) {
+		dev_warn(adap->pdev_dev,
+			 "No PHY Firmware file found for this PHY\n");
+		return -EOPNOTSUPP;
+	}
+
+	/* If we have a T4 PHY firmware file under /lib/firmware/cxgb4/, then
+	 * use that. The adapter firmware provides us with a memory buffer
+	 * where we can load a PHY firmware file from the host if we want to
+	 * override the PHY firmware File in flash.
+	 */
+	ret = request_firmware_direct(&phyf, phy_info->phy_fw_file,
+				      adap->pdev_dev);
+	if (ret < 0) {
+		/* For adapters without FLASH attached to PHY for their
+		 * firmware, it's obviously a fatal error if we can't get the
+		 * firmware to the adapter.  For adapters with PHY firmware
+		 * FLASH storage, it's worth a warning if we can't find the
+		 * PHY Firmware but we'll neuter the error ...
+		 */
+		dev_err(adap->pdev_dev, "unable to find PHY Firmware image "
+			"/lib/firmware/%s, error %d\n",
+			phy_info->phy_fw_file, -ret);
+		if (phy_info->phy_flash) {
+			int cur_phy_fw_ver = 0;
+
+			t4_phy_fw_ver(adap, &cur_phy_fw_ver);
+			dev_warn(adap->pdev_dev, "continuing with, on-adapter "
+				 "FLASH copy, version %#x\n", cur_phy_fw_ver);
+			ret = 0;
+		}
+
+		return ret;
+	}
+
+	/* Load PHY Firmware onto adapter.
+	 */
+	ret = t4_load_phy_fw(adap, MEMWIN_NIC, phy_info->phy_fw_version,
+			     (u8 *)phyf->data, phyf->size);
+	if (ret < 0)
+		dev_err(adap->pdev_dev, "PHY Firmware transfer error %d\n",
+			-ret);
+	else if (ret > 0) {
+		int new_phy_fw_ver = 0;
+
+		if (phy_info->phy_fw_version)
+			new_phy_fw_ver = phy_info->phy_fw_version(phyf->data,
+								  phyf->size);
+		dev_info(adap->pdev_dev, "Successfully transferred PHY "
+			 "Firmware /lib/firmware/%s, version %#x\n",
+			 phy_info->phy_fw_file, new_phy_fw_ver);
+	}
+
+	release_firmware(phyf);
+
+	return ret;
+}
+
+/*
+ * Attempt to initialize the adapter via a Firmware Configuration File.
+ */
+static int adap_init0_config(struct adapter *adapter, int reset)
+{
+	char *fw_config_file, fw_config_file_path[256];
+	u32 finiver, finicsum, cfcsum, param, val;
+	struct fw_caps_config_cmd caps_cmd;
+	unsigned long mtype = 0, maddr = 0;
+	const struct firmware *cf;
+	char *config_name = NULL;
+	int config_issued = 0;
+	int ret;
+
+	/*
+	 * Reset device if necessary.
+	 */
+	if (reset) {
+		ret = t4_fw_reset(adapter, adapter->mbox,
+				  PIORSTMODE_F | PIORST_F);
+		if (ret < 0)
+			goto bye;
+	}
+
+	/* If this is a 10Gb/s-BT adapter make sure the chip-external
+	 * 10Gb/s-BT PHYs have up-to-date firmware.  Note that this step needs
+	 * to be performed after any global adapter RESET above since some
+	 * PHYs only have local RAM copies of the PHY firmware.
+	 */
+	if (is_10gbt_device(adapter->pdev->device)) {
+		ret = adap_init0_phy(adapter);
+		if (ret < 0)
+			goto bye;
+	}
+	/*
+	 * If we have a T4 configuration file under /lib/firmware/cxgb4/,
+	 * then use that.  Otherwise, use the configuration file stored
+	 * in the adapter flash ...
+	 */
+	switch (CHELSIO_CHIP_VERSION(adapter->params.chip)) {
+	case CHELSIO_T4:
+		fw_config_file = FW4_CFNAME;
+		break;
+	case CHELSIO_T5:
+		fw_config_file = FW5_CFNAME;
+		break;
+	case CHELSIO_T6:
+		fw_config_file = FW6_CFNAME;
+		break;
+	default:
+		dev_err(adapter->pdev_dev, "Device %d is not supported\n",
+		       adapter->pdev->device);
+		ret = -EINVAL;
+		goto bye;
+	}
+
+	ret = request_firmware(&cf, fw_config_file, adapter->pdev_dev);
+	if (ret < 0) {
+		config_name = "On FLASH";
+		mtype = FW_MEMTYPE_CF_FLASH;
+		maddr = t4_flash_cfg_addr(adapter);
+	} else {
+		u32 params[7], val[7];
+
+		sprintf(fw_config_file_path,
+			"/lib/firmware/%s", fw_config_file);
+		config_name = fw_config_file_path;
+
+		if (cf->size >= FLASH_CFG_MAX_SIZE)
+			ret = -ENOMEM;
+		else {
+			params[0] = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+			     FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_CF));
+			ret = t4_query_params(adapter, adapter->mbox,
+					      adapter->pf, 0, 1, params, val);
+			if (ret == 0) {
+				/*
+				 * For t4_memory_rw() below addresses and
+				 * sizes have to be in terms of multiples of 4
+				 * bytes.  So, if the Configuration File isn't
+				 * a multiple of 4 bytes in length we'll have
+				 * to write that out separately since we can't
+				 * guarantee that the bytes following the
+				 * residual byte in the buffer returned by
+				 * request_firmware() are zeroed out ...
+				 */
+				size_t resid = cf->size & 0x3;
+				size_t size = cf->size & ~0x3;
+				__be32 *data = (__be32 *)cf->data;
+
+				mtype = FW_PARAMS_PARAM_Y_G(val[0]);
+				maddr = FW_PARAMS_PARAM_Z_G(val[0]) << 16;
+
+				spin_lock(&adapter->win0_lock);
+				ret = t4_memory_rw(adapter, 0, mtype, maddr,
+						   size, data, T4_MEMORY_WRITE);
+				if (ret == 0 && resid != 0) {
+					union {
+						__be32 word;
+						char buf[4];
+					} last;
+					int i;
+
+					last.word = data[size >> 2];
+					for (i = resid; i < 4; i++)
+						last.buf[i] = 0;
+					ret = t4_memory_rw(adapter, 0, mtype,
+							   maddr + size,
+							   4, &last.word,
+							   T4_MEMORY_WRITE);
+				}
+				spin_unlock(&adapter->win0_lock);
+			}
+		}
+
+		release_firmware(cf);
+		if (ret)
+			goto bye;
+	}
+
+	val = 0;
+
+	/* Ofld + Hash filter is supported. Older fw will fail this request and
+	 * it is fine.
+	 */
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_HASHFILTER_WITH_OFLD));
+	ret = t4_set_params(adapter, adapter->mbox, adapter->pf, 0,
+			    1, &param, &val);
+
+	/* FW doesn't know about Hash filter + ofld support,
+	 * it's not a problem, don't return an error.
+	 */
+	if (ret < 0) {
+		dev_warn(adapter->pdev_dev,
+			 "Hash filter with ofld is not supported by FW\n");
+	}
+
+	/*
+	 * Issue a Capability Configuration command to the firmware to get it
+	 * to parse the Configuration File.  We don't use t4_fw_config_file()
+	 * because we want the ability to modify various features after we've
+	 * processed the configuration file ...
+	 */
+	memset(&caps_cmd, 0, sizeof(caps_cmd));
+	caps_cmd.op_to_write =
+		htonl(FW_CMD_OP_V(FW_CAPS_CONFIG_CMD) |
+		      FW_CMD_REQUEST_F |
+		      FW_CMD_READ_F);
+	caps_cmd.cfvalid_to_len16 =
+		htonl(FW_CAPS_CONFIG_CMD_CFVALID_F |
+		      FW_CAPS_CONFIG_CMD_MEMTYPE_CF_V(mtype) |
+		      FW_CAPS_CONFIG_CMD_MEMADDR64K_CF_V(maddr >> 16) |
+		      FW_LEN16(caps_cmd));
+	ret = t4_wr_mbox(adapter, adapter->mbox, &caps_cmd, sizeof(caps_cmd),
+			 &caps_cmd);
+
+	/* If the CAPS_CONFIG failed with an ENOENT (for a Firmware
+	 * Configuration File in FLASH), our last gasp effort is to use the
+	 * Firmware Configuration File which is embedded in the firmware.  A
+	 * very few early versions of the firmware didn't have one embedded
+	 * but we can ignore those.
+	 */
+	if (ret == -ENOENT) {
+		memset(&caps_cmd, 0, sizeof(caps_cmd));
+		caps_cmd.op_to_write =
+			htonl(FW_CMD_OP_V(FW_CAPS_CONFIG_CMD) |
+					FW_CMD_REQUEST_F |
+					FW_CMD_READ_F);
+		caps_cmd.cfvalid_to_len16 = htonl(FW_LEN16(caps_cmd));
+		ret = t4_wr_mbox(adapter, adapter->mbox, &caps_cmd,
+				sizeof(caps_cmd), &caps_cmd);
+		config_name = "Firmware Default";
+	}
+
+	config_issued = 1;
+	if (ret < 0)
+		goto bye;
+
+	finiver = ntohl(caps_cmd.finiver);
+	finicsum = ntohl(caps_cmd.finicsum);
+	cfcsum = ntohl(caps_cmd.cfcsum);
+	if (finicsum != cfcsum)
+		dev_warn(adapter->pdev_dev, "Configuration File checksum "\
+			 "mismatch: [fini] csum=%#x, computed csum=%#x\n",
+			 finicsum, cfcsum);
+
+	/*
+	 * And now tell the firmware to use the configuration we just loaded.
+	 */
+	caps_cmd.op_to_write =
+		htonl(FW_CMD_OP_V(FW_CAPS_CONFIG_CMD) |
+		      FW_CMD_REQUEST_F |
+		      FW_CMD_WRITE_F);
+	caps_cmd.cfvalid_to_len16 = htonl(FW_LEN16(caps_cmd));
+	ret = t4_wr_mbox(adapter, adapter->mbox, &caps_cmd, sizeof(caps_cmd),
+			 NULL);
+	if (ret < 0)
+		goto bye;
+
+	/*
+	 * Tweak configuration based on system architecture, module
+	 * parameters, etc.
+	 */
+	ret = adap_init0_tweaks(adapter);
+	if (ret < 0)
+		goto bye;
+
+	/* We will proceed even if HMA init fails. */
+	ret = adap_config_hma(adapter);
+	if (ret)
+		dev_err(adapter->pdev_dev,
+			"HMA configuration failed with error %d\n", ret);
+
+	if (is_t6(adapter->params.chip)) {
+		adap_config_hpfilter(adapter);
+		ret = setup_ppod_edram(adapter);
+		if (!ret)
+			dev_info(adapter->pdev_dev, "Successfully enabled "
+				 "ppod edram feature\n");
+	}
+
+	/*
+	 * And finally tell the firmware to initialize itself using the
+	 * parameters from the Configuration File.
+	 */
+	ret = t4_fw_initialize(adapter, adapter->mbox);
+	if (ret < 0)
+		goto bye;
+
+	/* Emit Firmware Configuration File information and return
+	 * successfully.
+	 */
+	dev_info(adapter->pdev_dev, "Successfully configured using Firmware "\
+		 "Configuration File \"%s\", version %#x, computed checksum %#x\n",
+		 config_name, finiver, cfcsum);
+	return 0;
+
+	/*
+	 * Something bad happened.  Return the error ...  (If the "error"
+	 * is that there's no Configuration File on the adapter we don't
+	 * want to issue a warning since this is fairly common.)
+	 */
+bye:
+	if (config_issued && ret != -ENOENT)
+		dev_warn(adapter->pdev_dev, "\"%s\" configuration file error %d\n",
+			 config_name, -ret);
+	return ret;
+}
+
+static struct fw_info fw_info_array[] = {
+	{
+		.chip = CHELSIO_T4,
+		.fs_name = FW4_CFNAME,
+		.fw_mod_name = FW4_FNAME,
+		.fw_hdr = {
+			.chip = FW_HDR_CHIP_T4,
+			.fw_ver = __cpu_to_be32(FW_VERSION(T4)),
+			.intfver_nic = FW_INTFVER(T4, NIC),
+			.intfver_vnic = FW_INTFVER(T4, VNIC),
+			.intfver_ri = FW_INTFVER(T4, RI),
+			.intfver_iscsi = FW_INTFVER(T4, ISCSI),
+			.intfver_fcoe = FW_INTFVER(T4, FCOE),
+		},
+	}, {
+		.chip = CHELSIO_T5,
+		.fs_name = FW5_CFNAME,
+		.fw_mod_name = FW5_FNAME,
+		.fw_hdr = {
+			.chip = FW_HDR_CHIP_T5,
+			.fw_ver = __cpu_to_be32(FW_VERSION(T5)),
+			.intfver_nic = FW_INTFVER(T5, NIC),
+			.intfver_vnic = FW_INTFVER(T5, VNIC),
+			.intfver_ri = FW_INTFVER(T5, RI),
+			.intfver_iscsi = FW_INTFVER(T5, ISCSI),
+			.intfver_fcoe = FW_INTFVER(T5, FCOE),
+		},
+	}, {
+		.chip = CHELSIO_T6,
+		.fs_name = FW6_CFNAME,
+		.fw_mod_name = FW6_FNAME,
+		.fw_hdr = {
+			.chip = FW_HDR_CHIP_T6,
+			.fw_ver = __cpu_to_be32(FW_VERSION(T6)),
+			.intfver_nic = FW_INTFVER(T6, NIC),
+			.intfver_vnic = FW_INTFVER(T6, VNIC),
+			.intfver_ofld = FW_INTFVER(T6, OFLD),
+			.intfver_ri = FW_INTFVER(T6, RI),
+			.intfver_iscsipdu = FW_INTFVER(T6, ISCSIPDU),
+			.intfver_iscsi = FW_INTFVER(T6, ISCSI),
+			.intfver_fcoepdu = FW_INTFVER(T6, FCOEPDU),
+			.intfver_fcoe = FW_INTFVER(T6, FCOE),
+		},
+	}
+
+};
+
+static struct fw_info *find_fw_info(int chip)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(fw_info_array); i++) {
+		if (fw_info_array[i].chip == chip)
+			return &fw_info_array[i];
+	}
+	return NULL;
+}
+
+/*
+ * Phase 0 of initialization: contact FW, obtain config, perform basic init.
+ */
+static int adap_init0(struct adapter *adap, int vpd_skip)
+{
+	struct fw_caps_config_cmd caps_cmd;
+	u32 params[7], val[7];
+	enum dev_state state;
+	u32 v, port_vec;
+	int reset = 1;
+	int ret;
+
+	/* Grab Firmware Device Log parameters as early as possible so we have
+	 * access to it for debugging, etc.
+	 */
+	ret = t4_init_devlog_params(adap);
+	if (ret < 0)
+		return ret;
+
+	/* Contact FW, advertising Master capability */
+	ret = t4_fw_hello(adap, adap->mbox, adap->mbox,
+			  is_kdump_kernel() ? MASTER_MUST : MASTER_MAY, &state);
+	if (ret < 0) {
+		dev_err(adap->pdev_dev, "could not connect to FW, error %d\n",
+			ret);
+		return ret;
+	}
+	if (ret == adap->mbox)
+		adap->flags |= CXGB4_MASTER_PF;
+
+	/*
+	 * If we're the Master PF Driver and the device is uninitialized,
+	 * then let's consider upgrading the firmware ...  (We always want
+	 * to check the firmware version number in order to A. get it for
+	 * later reporting and B. to warn if the currently loaded firmware
+	 * is excessively mismatched relative to the driver.)
+	 */
+
+	t4_get_version_info(adap);
+	ret = t4_check_fw_version(adap);
+	/* If firmware is too old (not supported by driver) force an update. */
+	if (ret)
+		state = DEV_STATE_UNINIT;
+	if ((adap->flags & CXGB4_MASTER_PF) && state != DEV_STATE_INIT) {
+		struct fw_info *fw_info;
+		struct fw_hdr *card_fw;
+		const struct firmware *fw;
+		const u8 *fw_data = NULL;
+		unsigned int fw_size = 0;
+
+		/* This is the firmware whose headers the driver was compiled
+		 * against
+		 */
+		fw_info = find_fw_info(CHELSIO_CHIP_VERSION(adap->params.chip));
+		if (fw_info == NULL) {
+			dev_err(adap->pdev_dev,
+				"unable to get firmware info for chip %d.\n",
+				CHELSIO_CHIP_VERSION(adap->params.chip));
+			return -EINVAL;
+		}
+
+		/* allocate memory to read the header of the firmware on the
+		 * card
+		 */
+		card_fw = kvzalloc(sizeof(*card_fw), GFP_KERNEL);
+		if (!card_fw) {
+			ret = -ENOMEM;
+			goto bye;
+		}
+
+		/* Get FW from from /lib/firmware/ */
+		ret = request_firmware(&fw, fw_info->fw_mod_name,
+				       adap->pdev_dev);
+		if (ret < 0) {
+			dev_err(adap->pdev_dev,
+				"unable to load firmware image %s, error %d\n",
+				fw_info->fw_mod_name, ret);
+		} else {
+			fw_data = fw->data;
+			fw_size = fw->size;
+		}
+
+		/* upgrade FW logic */
+		ret = t4_prep_fw(adap, fw_info, fw_data, fw_size, card_fw,
+				 state, &reset);
+
+		/* Cleaning up */
+		release_firmware(fw);
+		kvfree(card_fw);
+
+		if (ret < 0)
+			goto bye;
+	}
+
+	/* If the firmware is initialized already, emit a simply note to that
+	 * effect. Otherwise, it's time to try initializing the adapter.
+	 */
+	if (state == DEV_STATE_INIT) {
+		ret = adap_config_hma(adap);
+		if (ret)
+			dev_err(adap->pdev_dev,
+				"HMA configuration failed with error %d\n",
+				ret);
+		dev_info(adap->pdev_dev, "Coming up as %s: "\
+			 "Adapter already initialized\n",
+			 adap->flags & CXGB4_MASTER_PF ? "MASTER" : "SLAVE");
+	} else {
+		dev_info(adap->pdev_dev, "Coming up as MASTER: "\
+			 "Initializing adapter\n");
+
+		/* Find out whether we're dealing with a version of the
+		 * firmware which has configuration file support.
+		 */
+		params[0] = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+			     FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_CF));
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 1,
+				      params, val);
+
+		/* If the firmware doesn't support Configuration Files,
+		 * return an error.
+		 */
+		if (ret < 0) {
+			dev_err(adap->pdev_dev, "firmware doesn't support "
+				"Firmware Configuration Files\n");
+			goto bye;
+		}
+
+		/* The firmware provides us with a memory buffer where we can
+		 * load a Configuration File from the host if we want to
+		 * override the Configuration File in flash.
+		 */
+		ret = adap_init0_config(adap, reset);
+		if (ret == -ENOENT) {
+			dev_err(adap->pdev_dev, "no Configuration File "
+				"present on adapter.\n");
+			goto bye;
+		}
+		if (ret < 0) {
+			dev_err(adap->pdev_dev, "could not initialize "
+				"adapter, error %d\n", -ret);
+			goto bye;
+		}
+	}
+
+	/* Now that we've successfully configured and initialized the adapter
+	 * (or found it already initialized), we can ask the Firmware what
+	 * resources it has provisioned for us.
+	 */
+	ret = t4_get_pfres(adap);
+	if (ret) {
+		dev_err(adap->pdev_dev,
+			"Unable to retrieve resource provisioning information\n");
+		goto bye;
+	}
+
+	/* Grab VPD parameters.  This should be done after we establish a
+	 * connection to the firmware since some of the VPD parameters
+	 * (notably the Core Clock frequency) are retrieved via requests to
+	 * the firmware.  On the other hand, we need these fairly early on
+	 * so we do this right after getting ahold of the firmware.
+	 *
+	 * We need to do this after initializing the adapter because someone
+	 * could have FLASHed a new VPD which won't be read by the firmware
+	 * until we do the RESET ...
+	 */
+	if (!vpd_skip) {
+		ret = t4_get_vpd_params(adap, &adap->params.vpd);
+		if (ret < 0)
+			goto bye;
+	}
+
+	/* Find out what ports are available to us.  Note that we need to do
+	 * this before calling adap_init0_no_config() since it needs nports
+	 * and portvec ...
+	 */
+	v =
+	    FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+	    FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_PORTVEC);
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 1, &v, &port_vec);
+	if (ret < 0)
+		goto bye;
+
+	adap->params.nports = hweight32(port_vec);
+	adap->params.portvec = port_vec;
+
+	/* Give the SGE code a chance to pull in anything that it needs ...
+	 * Note that this must be called after we retrieve our VPD parameters
+	 * in order to know how to convert core ticks to seconds, etc.
+	 */
+	ret = t4_sge_init(adap);
+	if (ret < 0)
+		goto bye;
+
+	/* Grab the SGE Doorbell Queue Timer values.  If successful, that
+	 * indicates that the Firmware and Hardware support this.
+	 */
+	params[0] = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		    FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_DBQ_TIMERTICK));
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0,
+			      1, params, val);
+
+	if (!ret) {
+		adap->sge.dbqtimer_tick = val[0];
+		ret = t4_read_sge_dbqtimers(adap,
+					    ARRAY_SIZE(adap->sge.dbqtimer_val),
+					    adap->sge.dbqtimer_val);
+	}
+
+	if (!ret)
+		adap->flags |= CXGB4_SGE_DBQ_TIMER;
+
+	if (is_bypass_device(adap->pdev->device))
+		adap->params.bypass = 1;
+
+	/*
+	 * Grab some of our basic fundamental operating parameters.
+	 */
+	params[0] = FW_PARAM_PFVF(EQ_START);
+	params[1] = FW_PARAM_PFVF(L2T_START);
+	params[2] = FW_PARAM_PFVF(L2T_END);
+	params[3] = FW_PARAM_PFVF(FILTER_START);
+	params[4] = FW_PARAM_PFVF(FILTER_END);
+	params[5] = FW_PARAM_PFVF(IQFLINT_START);
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 6, params, val);
+	if (ret < 0)
+		goto bye;
+	adap->sge.egr_start = val[0];
+	adap->l2t_start = val[1];
+	adap->l2t_end = val[2];
+	adap->tids.ftid_base = val[3];
+	adap->tids.nftids = val[4] - val[3] + 1;
+	adap->sge.ingr_start = val[5];
+
+	if (CHELSIO_CHIP_VERSION(adap->params.chip) > CHELSIO_T5) {
+		params[0] = FW_PARAM_PFVF(HPFILTER_START);
+		params[1] = FW_PARAM_PFVF(HPFILTER_END);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2,
+				      params, val);
+		if (ret < 0)
+			goto bye;
+
+		adap->tids.hpftid_base = val[0];
+		adap->tids.nhpftids = val[1] - val[0] + 1;
+
+		/* Read the raw mps entries. In T6, the last 2 tcam entries
+		 * are reserved for raw mac addresses (rawf = 2, one per port).
+		 */
+		params[0] = FW_PARAM_PFVF(RAWF_START);
+		params[1] = FW_PARAM_PFVF(RAWF_END);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2,
+				      params, val);
+		if (ret == 0) {
+			adap->rawf_start = val[0];
+			adap->rawf_cnt = val[1] - val[0] + 1;
+		}
+
+		adap->tids.tid_base =
+			t4_read_reg(adap, LE_DB_ACTIVE_TABLE_START_INDEX_A);
+	}
+
+	/* qids (ingress/egress) returned from firmware can be anywhere
+	 * in the range from EQ(IQFLINT)_START to EQ(IQFLINT)_END.
+	 * Hence driver needs to allocate memory for this range to
+	 * store the queue info. Get the highest IQFLINT/EQ index returned
+	 * in FW_EQ_*_CMD.alloc command.
+	 */
+	params[0] = FW_PARAM_PFVF(EQ_END);
+	params[1] = FW_PARAM_PFVF(IQFLINT_END);
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2, params, val);
+	if (ret < 0)
+		goto bye;
+	adap->sge.egr_sz = val[0] - adap->sge.egr_start + 1;
+	adap->sge.ingr_sz = val[1] - adap->sge.ingr_start + 1;
+
+	adap->sge.egr_map = kcalloc(adap->sge.egr_sz,
+				    sizeof(*adap->sge.egr_map), GFP_KERNEL);
+	if (!adap->sge.egr_map) {
+		ret = -ENOMEM;
+		goto bye;
+	}
+
+	adap->sge.ingr_map = kcalloc(adap->sge.ingr_sz,
+				     sizeof(*adap->sge.ingr_map), GFP_KERNEL);
+	if (!adap->sge.ingr_map) {
+		ret = -ENOMEM;
+		goto bye;
+	}
+
+	/* Allocate the memory for the vaious egress queue bitmaps
+	 * ie starving_fl, txq_maperr and blocked_fl.
+	 */
+	adap->sge.starving_fl = bitmap_zalloc(adap->sge.egr_sz, GFP_KERNEL);
+	if (!adap->sge.starving_fl) {
+		ret = -ENOMEM;
+		goto bye;
+	}
+
+	adap->sge.txq_maperr = bitmap_zalloc(adap->sge.egr_sz, GFP_KERNEL);
+	if (!adap->sge.txq_maperr) {
+		ret = -ENOMEM;
+		goto bye;
+	}
+
+#ifdef CONFIG_DEBUG_FS
+	adap->sge.blocked_fl = bitmap_zalloc(adap->sge.egr_sz, GFP_KERNEL);
+	if (!adap->sge.blocked_fl) {
+		ret = -ENOMEM;
+		goto bye;
+	}
+#endif
+
+	params[0] = FW_PARAM_PFVF(CLIP_START);
+	params[1] = FW_PARAM_PFVF(CLIP_END);
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2, params, val);
+	if (ret < 0)
+		goto bye;
+	adap->clipt_start = val[0];
+	adap->clipt_end = val[1];
+
+	/* Get the supported number of traffic classes */
+	params[0] = FW_PARAM_DEV(NUM_TM_CLASS);
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 1, params, val);
+	if (ret < 0) {
+		/* We couldn't retrieve the number of Traffic Classes
+		 * supported by the hardware/firmware. So we hard
+		 * code it here.
+		 */
+		adap->params.nsched_cls = is_t4(adap->params.chip) ? 15 : 16;
+	} else {
+		adap->params.nsched_cls = val[0];
+	}
+
+	/* query params related to active filter region */
+	params[0] = FW_PARAM_PFVF(ACTIVE_FILTER_START);
+	params[1] = FW_PARAM_PFVF(ACTIVE_FILTER_END);
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2, params, val);
+	/* If Active filter size is set we enable establishing
+	 * offload connection through firmware work request
+	 */
+	if ((val[0] != val[1]) && (ret >= 0)) {
+		adap->flags |= CXGB4_FW_OFLD_CONN;
+		adap->tids.aftid_base = val[0];
+		adap->tids.aftid_end = val[1];
+	}
+
+	/* If we're running on newer firmware, let it know that we're
+	 * prepared to deal with encapsulated CPL messages.  Older
+	 * firmware won't understand this and we'll just get
+	 * unencapsulated messages ...
+	 */
+	params[0] = FW_PARAM_PFVF(CPLFW4MSG_ENCAP);
+	val[0] = 1;
+	(void)t4_set_params(adap, adap->mbox, adap->pf, 0, 1, params, val);
+
+	/*
+	 * Find out whether we're allowed to use the T5+ ULPTX MEMWRITE DSGL
+	 * capability.  Earlier versions of the firmware didn't have the
+	 * ULPTX_MEMWRITE_DSGL so we'll interpret a query failure as no
+	 * permission to use ULPTX MEMWRITE DSGL.
+	 */
+	if (is_t4(adap->params.chip)) {
+		adap->params.ulptx_memwrite_dsgl = false;
+	} else {
+		params[0] = FW_PARAM_DEV(ULPTX_MEMWRITE_DSGL);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0,
+				      1, params, val);
+		adap->params.ulptx_memwrite_dsgl = (ret == 0 && val[0] != 0);
+	}
+
+	/* See if FW supports FW_RI_FR_NSMR_TPTE_WR work request */
+	params[0] = FW_PARAM_DEV(RI_FR_NSMR_TPTE_WR);
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0,
+			      1, params, val);
+	adap->params.fr_nsmr_tpte_wr_support = (ret == 0 && val[0] != 0);
+
+	/* See if FW supports FW_FILTER2 work request */
+	if (is_t4(adap->params.chip)) {
+		adap->params.filter2_wr_support = false;
+	} else {
+		params[0] = FW_PARAM_DEV(FILTER2_WR);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0,
+				      1, params, val);
+		adap->params.filter2_wr_support = (ret == 0 && val[0] != 0);
+	}
+
+	/* Check if FW supports returning vin and smt index.
+	 * If this is not supported, driver will interpret
+	 * these values from viid.
+	 */
+	params[0] = FW_PARAM_DEV(OPAQUE_VIID_SMT_EXTN);
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0,
+			      1, params, val);
+	adap->params.viid_smt_extn_support = (ret == 0 && val[0] != 0);
+
+	/*
+	 * Get device capabilities so we can determine what resources we need
+	 * to manage.
+	 */
+	memset(&caps_cmd, 0, sizeof(caps_cmd));
+	caps_cmd.op_to_write = htonl(FW_CMD_OP_V(FW_CAPS_CONFIG_CMD) |
+				     FW_CMD_REQUEST_F | FW_CMD_READ_F);
+	caps_cmd.cfvalid_to_len16 = htonl(FW_LEN16(caps_cmd));
+	ret = t4_wr_mbox(adap, adap->mbox, &caps_cmd, sizeof(caps_cmd),
+			 &caps_cmd);
+	if (ret < 0)
+		goto bye;
+
+	/* hash filter has some mandatory register settings to be tested and for
+	 * that it needs to test whether offload is enabled or not, hence
+	 * checking and setting it here.
+	 */
+	if (caps_cmd.ofldcaps)
+		adap->params.offload = 1;
+
+	if (caps_cmd.ofldcaps ||
+	    (caps_cmd.niccaps & htons(FW_CAPS_CONFIG_NIC_HASHFILTER)) ||
+	    (caps_cmd.niccaps & htons(FW_CAPS_CONFIG_NIC_ETHOFLD))) {
+		/* query offload-related parameters */
+		params[0] = FW_PARAM_DEV(NTID);
+		params[1] = FW_PARAM_PFVF(SERVER_START);
+		params[2] = FW_PARAM_PFVF(SERVER_END);
+		params[3] = FW_PARAM_PFVF(TDDP_START);
+		params[4] = FW_PARAM_PFVF(TDDP_END);
+		params[5] = FW_PARAM_DEV(FLOWC_BUFFIFO_SZ);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 6,
+				      params, val);
+		if (ret < 0)
+			goto bye;
+		adap->tids.ntids = val[0];
+		adap->tids.natids = min(adap->tids.ntids / 2, MAX_ATIDS);
+		adap->tids.stid_base = val[1];
+		adap->tids.nstids = val[2] - val[1] + 1;
+		/*
+		 * Setup server filter region. Divide the available filter
+		 * region into two parts. Regular filters get 1/3rd and server
+		 * filters get 2/3rd part. This is only enabled if workarond
+		 * path is enabled.
+		 * 1. For regular filters.
+		 * 2. Server filter: This are special filters which are used
+		 * to redirect SYN packets to offload queue.
+		 */
+		if (adap->flags & CXGB4_FW_OFLD_CONN && !is_bypass(adap)) {
+			adap->tids.sftid_base = adap->tids.ftid_base +
+					DIV_ROUND_UP(adap->tids.nftids, 3);
+			adap->tids.nsftids = adap->tids.nftids -
+					 DIV_ROUND_UP(adap->tids.nftids, 3);
+			adap->tids.nftids = adap->tids.sftid_base -
+						adap->tids.ftid_base;
+		}
+		adap->vres.ddp.start = val[3];
+		adap->vres.ddp.size = val[4] - val[3] + 1;
+		adap->params.ofldq_wr_cred = val[5];
+
+		if (caps_cmd.niccaps & htons(FW_CAPS_CONFIG_NIC_HASHFILTER)) {
+			init_hash_filter(adap);
+		} else {
+			adap->num_ofld_uld += 1;
+		}
+
+		if (caps_cmd.niccaps & htons(FW_CAPS_CONFIG_NIC_ETHOFLD)) {
+			params[0] = FW_PARAM_PFVF(ETHOFLD_START);
+			params[1] = FW_PARAM_PFVF(ETHOFLD_END);
+			ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2,
+					      params, val);
+			if (!ret) {
+				adap->tids.eotid_base = val[0];
+				adap->tids.neotids = min_t(u32, MAX_ATIDS,
+							   val[1] - val[0] + 1);
+				adap->params.ethofld = 1;
+			}
+		}
+	}
+	if (caps_cmd.rdmacaps) {
+		params[0] = FW_PARAM_PFVF(STAG_START);
+		params[1] = FW_PARAM_PFVF(STAG_END);
+		params[2] = FW_PARAM_PFVF(RQ_START);
+		params[3] = FW_PARAM_PFVF(RQ_END);
+		params[4] = FW_PARAM_PFVF(PBL_START);
+		params[5] = FW_PARAM_PFVF(PBL_END);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 6,
+				      params, val);
+		if (ret < 0)
+			goto bye;
+		adap->vres.stag.start = val[0];
+		adap->vres.stag.size = val[1] - val[0] + 1;
+		adap->vres.rq.start = val[2];
+		adap->vres.rq.size = val[3] - val[2] + 1;
+		adap->vres.pbl.start = val[4];
+		adap->vres.pbl.size = val[5] - val[4] + 1;
+
+		params[0] = FW_PARAM_PFVF(SRQ_START);
+		params[1] = FW_PARAM_PFVF(SRQ_END);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2,
+				      params, val);
+		if (!ret) {
+			adap->vres.srq.start = val[0];
+			adap->vres.srq.size = val[1] - val[0] + 1;
+		}
+		if (adap->vres.srq.size) {
+			adap->srq = t4_init_srq(adap->vres.srq.size);
+			if (!adap->srq)
+				dev_warn(&adap->pdev->dev, "could not allocate SRQ, continuing\n");
+		}
+
+		params[0] = FW_PARAM_PFVF(SQRQ_START);
+		params[1] = FW_PARAM_PFVF(SQRQ_END);
+		params[2] = FW_PARAM_PFVF(CQ_START);
+		params[3] = FW_PARAM_PFVF(CQ_END);
+		params[4] = FW_PARAM_PFVF(OCQ_START);
+		params[5] = FW_PARAM_PFVF(OCQ_END);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 6, params,
+				      val);
+		if (ret < 0)
+			goto bye;
+		adap->vres.qp.start = val[0];
+		adap->vres.qp.size = val[1] - val[0] + 1;
+		adap->vres.cq.start = val[2];
+		adap->vres.cq.size = val[3] - val[2] + 1;
+		adap->vres.ocq.start = val[4];
+		adap->vres.ocq.size = val[5] - val[4] + 1;
+
+		params[0] = FW_PARAM_DEV(MAXORDIRD_QP);
+		params[1] = FW_PARAM_DEV(MAXIRD_ADAPTER);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2, params,
+				      val);
+		if (ret < 0) {
+			adap->params.max_ordird_qp = 8;
+			adap->params.max_ird_adapter = 32 * adap->tids.ntids;
+			ret = 0;
+		} else {
+			adap->params.max_ordird_qp = val[0];
+			adap->params.max_ird_adapter = val[1];
+		}
+		dev_info(adap->pdev_dev,
+			 "max_ordird_qp %d max_ird_adapter %d\n",
+			 adap->params.max_ordird_qp,
+			 adap->params.max_ird_adapter);
+
+		/* Enable write_with_immediate if FW supports it */
+		params[0] = FW_PARAM_DEV(RDMA_WRITE_WITH_IMM);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 1, params,
+				      val);
+		adap->params.write_w_imm_support = (ret == 0 && val[0] != 0);
+
+		/* Enable write_cmpl if FW supports it */
+		params[0] = FW_PARAM_DEV(RI_WRITE_CMPL_WR);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 1, params,
+				      val);
+		adap->params.write_cmpl_support = (ret == 0 && val[0] != 0);
+		adap->num_ofld_uld += 2;
+	}
+	if (caps_cmd.iscsicaps) {
+		params[0] = FW_PARAM_PFVF(ISCSI_START);
+		params[1] = FW_PARAM_PFVF(ISCSI_END);
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2,
+				      params, val);
+		if (ret < 0)
+			goto bye;
+		adap->vres.iscsi.start = val[0];
+		adap->vres.iscsi.size = val[1] - val[0] + 1;
+		if (is_t6(adap->params.chip)) {
+			params[0] = FW_PARAM_PFVF(PPOD_EDRAM_START);
+			params[1] = FW_PARAM_PFVF(PPOD_EDRAM_END);
+			ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2,
+					      params, val);
+			if (!ret) {
+				adap->vres.ppod_edram.start = val[0];
+				adap->vres.ppod_edram.size =
+					val[1] - val[0] + 1;
+
+				dev_info(adap->pdev_dev,
+					 "ppod edram start 0x%x end 0x%x size 0x%x\n",
+					 val[0], val[1],
+					 adap->vres.ppod_edram.size);
+			}
+		}
+		/* LIO target and cxgb4i initiaitor */
+		adap->num_ofld_uld += 2;
+	}
+	if (caps_cmd.cryptocaps) {
+		if (ntohs(caps_cmd.cryptocaps) &
+		    FW_CAPS_CONFIG_CRYPTO_LOOKASIDE) {
+			params[0] = FW_PARAM_PFVF(NCRYPTO_LOOKASIDE);
+			ret = t4_query_params(adap, adap->mbox, adap->pf, 0,
+					      2, params, val);
+			if (ret < 0) {
+				if (ret != -EINVAL)
+					goto bye;
+			} else {
+				adap->vres.ncrypto_fc = val[0];
+			}
+			adap->num_ofld_uld += 1;
+		}
+		if (ntohs(caps_cmd.cryptocaps) &
+		    FW_CAPS_CONFIG_TLS_INLINE) {
+			params[0] = FW_PARAM_PFVF(TLS_START);
+			params[1] = FW_PARAM_PFVF(TLS_END);
+			ret = t4_query_params(adap, adap->mbox, adap->pf, 0,
+					      2, params, val);
+			if (ret < 0)
+				goto bye;
+			adap->vres.key.start = val[0];
+			adap->vres.key.size = val[1] - val[0] + 1;
+			adap->num_uld += 1;
+		}
+		adap->params.crypto = ntohs(caps_cmd.cryptocaps);
+	}
+
+	/* The MTU/MSS Table is initialized by now, so load their values.  If
+	 * we're initializing the adapter, then we'll make any modifications
+	 * we want to the MTU/MSS Table and also initialize the congestion
+	 * parameters.
+	 */
+	t4_read_mtu_tbl(adap, adap->params.mtus, NULL);
+	if (state != DEV_STATE_INIT) {
+		int i;
+
+		/* The default MTU Table contains values 1492 and 1500.
+		 * However, for TCP, it's better to have two values which are
+		 * a multiple of 8 +/- 4 bytes apart near this popular MTU.
+		 * This allows us to have a TCP Data Payload which is a
+		 * multiple of 8 regardless of what combination of TCP Options
+		 * are in use (always a multiple of 4 bytes) which is
+		 * important for performance reasons.  For instance, if no
+		 * options are in use, then we have a 20-byte IP header and a
+		 * 20-byte TCP header.  In this case, a 1500-byte MSS would
+		 * result in a TCP Data Payload of 1500 - 40 == 1460 bytes
+		 * which is not a multiple of 8.  So using an MSS of 1488 in
+		 * this case results in a TCP Data Payload of 1448 bytes which
+		 * is a multiple of 8.  On the other hand, if 12-byte TCP Time
+		 * Stamps have been negotiated, then an MTU of 1500 bytes
+		 * results in a TCP Data Payload of 1448 bytes which, as
+		 * above, is a multiple of 8 bytes ...
+		 */
+		for (i = 0; i < NMTUS; i++)
+			if (adap->params.mtus[i] == 1492) {
+				adap->params.mtus[i] = 1488;
+				break;
+			}
+
+		t4_load_mtus(adap, adap->params.mtus, adap->params.a_wnd,
+			     adap->params.b_wnd);
+	}
+	t4_init_sge_params(adap);
+	adap->flags |= CXGB4_FW_OK;
+	t4_init_tp_params(adap, true);
+	return 0;
+
+	/*
+	 * Something bad happened.  If a command timed out or failed with EIO
+	 * FW does not operate within its spec or something catastrophic
+	 * happened to HW/FW, stop issuing commands.
+	 */
+bye:
+	adap_free_hma_mem(adap);
+	kfree(adap->sge.egr_map);
+	kfree(adap->sge.ingr_map);
+	bitmap_free(adap->sge.starving_fl);
+	bitmap_free(adap->sge.txq_maperr);
+#ifdef CONFIG_DEBUG_FS
+	bitmap_free(adap->sge.blocked_fl);
+#endif
+	if (ret != -ETIMEDOUT && ret != -EIO)
+		t4_fw_bye(adap, adap->mbox);
+	return ret;
+}
+
+/* EEH callbacks */
+
+static pci_ers_result_t eeh_err_detected(struct pci_dev *pdev,
+					 pci_channel_state_t state)
+{
+	int i;
+	struct adapter *adap = pci_get_drvdata(pdev);
+
+	if (!adap)
+		goto out;
+
+	rtnl_lock();
+	adap->flags &= ~CXGB4_FW_OK;
+	notify_ulds(adap, CXGB4_STATE_START_RECOVERY);
+	spin_lock(&adap->stats_lock);
+	for_each_port(adap, i) {
+		struct net_device *dev = adap->port[i];
+		if (dev) {
+			netif_device_detach(dev);
+			netif_carrier_off(dev);
+		}
+	}
+	spin_unlock(&adap->stats_lock);
+	disable_interrupts(adap);
+	if (adap->flags & CXGB4_FULL_INIT_DONE)
+		cxgb_down(adap);
+	rtnl_unlock();
+	if ((adap->flags & CXGB4_DEV_ENABLED)) {
+		pci_disable_device(pdev);
+		adap->flags &= ~CXGB4_DEV_ENABLED;
+	}
+out:	return state == pci_channel_io_perm_failure ?
+		PCI_ERS_RESULT_DISCONNECT : PCI_ERS_RESULT_NEED_RESET;
+}
+
+static pci_ers_result_t eeh_slot_reset(struct pci_dev *pdev)
+{
+	int i, ret;
+	struct fw_caps_config_cmd c;
+	struct adapter *adap = pci_get_drvdata(pdev);
+
+	if (!adap) {
+		pci_restore_state(pdev);
+		pci_save_state(pdev);
+		return PCI_ERS_RESULT_RECOVERED;
+	}
+
+	if (!(adap->flags & CXGB4_DEV_ENABLED)) {
+		if (pci_enable_device(pdev)) {
+			dev_err(&pdev->dev, "Cannot reenable PCI "
+					    "device after reset\n");
+			return PCI_ERS_RESULT_DISCONNECT;
+		}
+		adap->flags |= CXGB4_DEV_ENABLED;
+	}
+
+	pci_set_master(pdev);
+	pci_restore_state(pdev);
+	pci_save_state(pdev);
+
+	if (t4_wait_dev_ready(adap->regs) < 0)
+		return PCI_ERS_RESULT_DISCONNECT;
+	if (t4_fw_hello(adap, adap->mbox, adap->pf, MASTER_MUST, NULL) < 0)
+		return PCI_ERS_RESULT_DISCONNECT;
+	adap->flags |= CXGB4_FW_OK;
+	if (adap_init1(adap, &c))
+		return PCI_ERS_RESULT_DISCONNECT;
+
+	for_each_port(adap, i) {
+		struct port_info *pi = adap2pinfo(adap, i);
+		u8 vivld = 0, vin = 0;
+
+		ret = t4_alloc_vi(adap, adap->mbox, pi->tx_chan, adap->pf, 0, 1,
+				  NULL, NULL, &vivld, &vin);
+		if (ret < 0)
+			return PCI_ERS_RESULT_DISCONNECT;
+		pi->viid = ret;
+		pi->xact_addr_filt = -1;
+		/* If fw supports returning the VIN as part of FW_VI_CMD,
+		 * save the returned values.
+		 */
+		if (adap->params.viid_smt_extn_support) {
+			pi->vivld = vivld;
+			pi->vin = vin;
+		} else {
+			/* Retrieve the values from VIID */
+			pi->vivld = FW_VIID_VIVLD_G(pi->viid);
+			pi->vin = FW_VIID_VIN_G(pi->viid);
+		}
+	}
+
+	t4_load_mtus(adap, adap->params.mtus, adap->params.a_wnd,
+		     adap->params.b_wnd);
+	setup_memwin(adap);
+	if (cxgb_up(adap))
+		return PCI_ERS_RESULT_DISCONNECT;
+	return PCI_ERS_RESULT_RECOVERED;
+}
+
+static void eeh_resume(struct pci_dev *pdev)
+{
+	int i;
+	struct adapter *adap = pci_get_drvdata(pdev);
+
+	if (!adap)
+		return;
+
+	rtnl_lock();
+	for_each_port(adap, i) {
+		struct net_device *dev = adap->port[i];
+		if (dev) {
+			if (netif_running(dev)) {
+				link_start(dev);
+				cxgb_set_rxmode(dev);
+			}
+			netif_device_attach(dev);
+		}
+	}
+	rtnl_unlock();
+}
+
+static void eeh_reset_prepare(struct pci_dev *pdev)
+{
+	struct adapter *adapter = pci_get_drvdata(pdev);
+	int i;
+
+	if (adapter->pf != 4)
+		return;
+
+	adapter->flags &= ~CXGB4_FW_OK;
+
+	notify_ulds(adapter, CXGB4_STATE_DOWN);
+
+	for_each_port(adapter, i)
+		if (adapter->port[i]->reg_state == NETREG_REGISTERED)
+			cxgb_close(adapter->port[i]);
+
+	disable_interrupts(adapter);
+	cxgb4_free_mps_ref_entries(adapter);
+
+	adap_free_hma_mem(adapter);
+
+	if (adapter->flags & CXGB4_FULL_INIT_DONE)
+		cxgb_down(adapter);
+}
+
+static void eeh_reset_done(struct pci_dev *pdev)
+{
+	struct adapter *adapter = pci_get_drvdata(pdev);
+	int err, i;
+
+	if (adapter->pf != 4)
+		return;
+
+	err = t4_wait_dev_ready(adapter->regs);
+	if (err < 0) {
+		dev_err(adapter->pdev_dev,
+			"Device not ready, err %d", err);
+		return;
+	}
+
+	setup_memwin(adapter);
+
+	err = adap_init0(adapter, 1);
+	if (err) {
+		dev_err(adapter->pdev_dev,
+			"Adapter init failed, err %d", err);
+		return;
+	}
+
+	setup_memwin_rdma(adapter);
+
+	if (adapter->flags & CXGB4_FW_OK) {
+		err = t4_port_init(adapter, adapter->pf, adapter->pf, 0);
+		if (err) {
+			dev_err(adapter->pdev_dev,
+				"Port init failed, err %d", err);
+			return;
+		}
+	}
+
+	err = cfg_queues(adapter);
+	if (err) {
+		dev_err(adapter->pdev_dev,
+			"Config queues failed, err %d", err);
+		return;
+	}
+
+	cxgb4_init_mps_ref_entries(adapter);
+
+	err = setup_fw_sge_queues(adapter);
+	if (err) {
+		dev_err(adapter->pdev_dev,
+			"FW sge queue allocation failed, err %d", err);
+		return;
+	}
+
+	for_each_port(adapter, i)
+		if (adapter->port[i]->reg_state == NETREG_REGISTERED)
+			cxgb_open(adapter->port[i]);
+}
+
+static const struct pci_error_handlers cxgb4_eeh = {
+	.error_detected = eeh_err_detected,
+	.slot_reset     = eeh_slot_reset,
+	.resume         = eeh_resume,
+	.reset_prepare  = eeh_reset_prepare,
+	.reset_done     = eeh_reset_done,
+};
+
+/* Return true if the Link Configuration supports "High Speeds" (those greater
+ * than 1Gb/s).
+ */
+static inline bool is_x_10g_port(const struct link_config *lc)
+{
+	unsigned int speeds, high_speeds;
+
+	speeds = FW_PORT_CAP32_SPEED_V(FW_PORT_CAP32_SPEED_G(lc->pcaps));
+	high_speeds = speeds &
+			~(FW_PORT_CAP32_SPEED_100M | FW_PORT_CAP32_SPEED_1G);
+
+	return high_speeds != 0;
+}
+
+/* Perform default configuration of DMA queues depending on the number and type
+ * of ports we found and the number of available CPUs.  Most settings can be
+ * modified by the admin prior to actual use.
+ */
+static int cfg_queues(struct adapter *adap)
+{
+	u32 avail_qsets, avail_eth_qsets, avail_uld_qsets;
+	u32 ncpus = num_online_cpus();
+	u32 niqflint, neq, num_ulds;
+	struct sge *s = &adap->sge;
+	u32 i, n10g = 0, qidx = 0;
+	u32 q10g = 0, q1g;
+
+	/* Reduce memory usage in kdump environment, disable all offload. */
+	if (is_kdump_kernel() || (is_uld(adap) && t4_uld_mem_alloc(adap))) {
+		adap->params.offload = 0;
+		adap->params.crypto = 0;
+		adap->params.ethofld = 0;
+	}
+
+	/* Calculate the number of Ethernet Queue Sets available based on
+	 * resources provisioned for us.  We always have an Asynchronous
+	 * Firmware Event Ingress Queue.  If we're operating in MSI or Legacy
+	 * IRQ Pin Interrupt mode, then we'll also have a Forwarded Interrupt
+	 * Ingress Queue.  Meanwhile, we need two Egress Queues for each
+	 * Queue Set: one for the Free List and one for the Ethernet TX Queue.
+	 *
+	 * Note that we should also take into account all of the various
+	 * Offload Queues.  But, in any situation where we're operating in
+	 * a Resource Constrained Provisioning environment, doing any Offload
+	 * at all is problematic ...
+	 */
+	niqflint = adap->params.pfres.niqflint - 1;
+	if (!(adap->flags & CXGB4_USING_MSIX))
+		niqflint--;
+	neq = adap->params.pfres.neq / 2;
+	avail_qsets = min(niqflint, neq);
+
+	if (avail_qsets < adap->params.nports) {
+		dev_err(adap->pdev_dev, "avail_eth_qsets=%d < nports=%d\n",
+			avail_qsets, adap->params.nports);
+		return -ENOMEM;
+	}
+
+	/* Count the number of 10Gb/s or better ports */
+	for_each_port(adap, i)
+		n10g += is_x_10g_port(&adap2pinfo(adap, i)->link_cfg);
+
+	avail_eth_qsets = min_t(u32, avail_qsets, MAX_ETH_QSETS);
+
+	/* We default to 1 queue per non-10G port and up to # of cores queues
+	 * per 10G port.
+	 */
+	if (n10g)
+		q10g = (avail_eth_qsets - (adap->params.nports - n10g)) / n10g;
+
+#ifdef CONFIG_CHELSIO_T4_DCB
+	/* For Data Center Bridging support we need to be able to support up
+	 * to 8 Traffic Priorities; each of which will be assigned to its
+	 * own TX Queue in order to prevent Head-Of-Line Blocking.
+	 */
+	q1g = 8;
+	if (adap->params.nports * 8 > avail_eth_qsets) {
+		dev_err(adap->pdev_dev, "DCB avail_eth_qsets=%d < %d!\n",
+			avail_eth_qsets, adap->params.nports * 8);
+		return -ENOMEM;
+	}
+
+	if (adap->params.nports * ncpus < avail_eth_qsets)
+		q10g = max(8U, ncpus);
+	else
+		q10g = max(8U, q10g);
+
+	while ((q10g * n10g) >
+	       (avail_eth_qsets - (adap->params.nports - n10g) * q1g))
+		q10g--;
+
+#else /* !CONFIG_CHELSIO_T4_DCB */
+	q1g = 1;
+	q10g = min(q10g, ncpus);
+#endif /* !CONFIG_CHELSIO_T4_DCB */
+	if (is_kdump_kernel()) {
+		q10g = 1;
+		q1g = 1;
+	}
+
+	for_each_port(adap, i) {
+		struct port_info *pi = adap2pinfo(adap, i);
+
+		pi->first_qset = qidx;
+		pi->nqsets = is_x_10g_port(&pi->link_cfg) ? q10g : q1g;
+		qidx += pi->nqsets;
+	}
+
+	s->ethqsets = qidx;
+	s->max_ethqsets = qidx;   /* MSI-X may lower it later */
+	avail_qsets -= qidx;
+
+	if (is_uld(adap)) {
+		/* For offload we use 1 queue/channel if all ports are up to 1G,
+		 * otherwise we divide all available queues amongst the channels
+		 * capped by the number of available cores.
+		 */
+		num_ulds = adap->num_uld + adap->num_ofld_uld;
+		i = min_t(u32, MAX_OFLD_QSETS, ncpus);
+		avail_uld_qsets = roundup(i, adap->params.nports);
+		if (avail_qsets < num_ulds * adap->params.nports) {
+			adap->params.offload = 0;
+			adap->params.crypto = 0;
+			s->ofldqsets = 0;
+		} else if (avail_qsets < num_ulds * avail_uld_qsets || !n10g) {
+			s->ofldqsets = adap->params.nports;
+		} else {
+			s->ofldqsets = avail_uld_qsets;
+		}
+
+		avail_qsets -= num_ulds * s->ofldqsets;
+	}
+
+	/* ETHOFLD Queues used for QoS offload should follow same
+	 * allocation scheme as normal Ethernet Queues.
+	 */
+	if (is_ethofld(adap)) {
+		if (avail_qsets < s->max_ethqsets) {
+			adap->params.ethofld = 0;
+			s->eoqsets = 0;
+		} else {
+			s->eoqsets = s->max_ethqsets;
+		}
+		avail_qsets -= s->eoqsets;
+	}
+
+	/* Mirror queues must follow same scheme as normal Ethernet
+	 * Queues, when there are enough queues available. Otherwise,
+	 * allocate at least 1 queue per port. If even 1 queue is not
+	 * available, then disable mirror queues support.
+	 */
+	if (avail_qsets >= s->max_ethqsets)
+		s->mirrorqsets = s->max_ethqsets;
+	else if (avail_qsets >= adap->params.nports)
+		s->mirrorqsets = adap->params.nports;
+	else
+		s->mirrorqsets = 0;
+	avail_qsets -= s->mirrorqsets;
+
+	for (i = 0; i < ARRAY_SIZE(s->ethrxq); i++) {
+		struct sge_eth_rxq *r = &s->ethrxq[i];
+
+		init_rspq(adap, &r->rspq, 5, 10, 1024, 64);
+		r->fl.size = 72;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(s->ethtxq); i++)
+		s->ethtxq[i].q.size = 1024;
+
+	for (i = 0; i < ARRAY_SIZE(s->ctrlq); i++)
+		s->ctrlq[i].q.size = 512;
+
+	if (!is_t4(adap->params.chip))
+		s->ptptxq.q.size = 8;
+
+	init_rspq(adap, &s->fw_evtq, 0, 1, 1024, 64);
+	init_rspq(adap, &s->intrq, 0, 1, 512, 64);
+
+	return 0;
+}
+
+/*
+ * Reduce the number of Ethernet queues across all ports to at most n.
+ * n provides at least one queue per port.
+ */
+static void reduce_ethqs(struct adapter *adap, int n)
+{
+	int i;
+	struct port_info *pi;
+
+	while (n < adap->sge.ethqsets)
+		for_each_port(adap, i) {
+			pi = adap2pinfo(adap, i);
+			if (pi->nqsets > 1) {
+				pi->nqsets--;
+				adap->sge.ethqsets--;
+				if (adap->sge.ethqsets <= n)
+					break;
+			}
+		}
+
+	n = 0;
+	for_each_port(adap, i) {
+		pi = adap2pinfo(adap, i);
+		pi->first_qset = n;
+		n += pi->nqsets;
+	}
+}
+
+static int alloc_msix_info(struct adapter *adap, u32 num_vec)
+{
+	struct msix_info *msix_info;
+
+	msix_info = kcalloc(num_vec, sizeof(*msix_info), GFP_KERNEL);
+	if (!msix_info)
+		return -ENOMEM;
+
+	adap->msix_bmap.msix_bmap = bitmap_zalloc(num_vec, GFP_KERNEL);
+	if (!adap->msix_bmap.msix_bmap) {
+		kfree(msix_info);
+		return -ENOMEM;
+	}
+
+	spin_lock_init(&adap->msix_bmap.lock);
+	adap->msix_bmap.mapsize = num_vec;
+
+	adap->msix_info = msix_info;
+	return 0;
+}
+
+static void free_msix_info(struct adapter *adap)
+{
+	bitmap_free(adap->msix_bmap.msix_bmap);
+	kfree(adap->msix_info);
+}
+
+int cxgb4_get_msix_idx_from_bmap(struct adapter *adap)
+{
+	struct msix_bmap *bmap = &adap->msix_bmap;
+	unsigned int msix_idx;
+	unsigned long flags;
+
+	spin_lock_irqsave(&bmap->lock, flags);
+	msix_idx = find_first_zero_bit(bmap->msix_bmap, bmap->mapsize);
+	if (msix_idx < bmap->mapsize) {
+		__set_bit(msix_idx, bmap->msix_bmap);
+	} else {
+		spin_unlock_irqrestore(&bmap->lock, flags);
+		return -ENOSPC;
+	}
+
+	spin_unlock_irqrestore(&bmap->lock, flags);
+	return msix_idx;
+}
+
+void cxgb4_free_msix_idx_in_bmap(struct adapter *adap,
+				 unsigned int msix_idx)
+{
+	struct msix_bmap *bmap = &adap->msix_bmap;
+	unsigned long flags;
+
+	spin_lock_irqsave(&bmap->lock, flags);
+	__clear_bit(msix_idx, bmap->msix_bmap);
+	spin_unlock_irqrestore(&bmap->lock, flags);
+}
+
+/* 2 MSI-X vectors needed for the FW queue and non-data interrupts */
+#define EXTRA_VECS 2
+
+static int enable_msix(struct adapter *adap)
+{
+	u32 eth_need, uld_need = 0, ethofld_need = 0, mirror_need = 0;
+	u32 ethqsets = 0, ofldqsets = 0, eoqsets = 0, mirrorqsets = 0;
+	u8 num_uld = 0, nchan = adap->params.nports;
+	u32 i, want, need, num_vec;
+	struct sge *s = &adap->sge;
+	struct msix_entry *entries;
+	struct port_info *pi;
+	int allocated, ret;
+
+	want = s->max_ethqsets;
+#ifdef CONFIG_CHELSIO_T4_DCB
+	/* For Data Center Bridging we need 8 Ethernet TX Priority Queues for
+	 * each port.
+	 */
+	need = 8 * nchan;
+#else
+	need = nchan;
+#endif
+	eth_need = need;
+	if (is_uld(adap)) {
+		num_uld = adap->num_ofld_uld + adap->num_uld;
+		want += num_uld * s->ofldqsets;
+		uld_need = num_uld * nchan;
+		need += uld_need;
+	}
+
+	if (is_ethofld(adap)) {
+		want += s->eoqsets;
+		ethofld_need = eth_need;
+		need += ethofld_need;
+	}
+
+	if (s->mirrorqsets) {
+		want += s->mirrorqsets;
+		mirror_need = nchan;
+		need += mirror_need;
+	}
+
+	want += EXTRA_VECS;
+	need += EXTRA_VECS;
+
+	entries = kmalloc_array(want, sizeof(*entries), GFP_KERNEL);
+	if (!entries)
+		return -ENOMEM;
+
+	for (i = 0; i < want; i++)
+		entries[i].entry = i;
+
+	allocated = pci_enable_msix_range(adap->pdev, entries, need, want);
+	if (allocated < 0) {
+		/* Disable offload and attempt to get vectors for NIC
+		 * only mode.
+		 */
+		want = s->max_ethqsets + EXTRA_VECS;
+		need = eth_need + EXTRA_VECS;
+		allocated = pci_enable_msix_range(adap->pdev, entries,
+						  need, want);
+		if (allocated < 0) {
+			dev_info(adap->pdev_dev,
+				 "Disabling MSI-X due to insufficient MSI-X vectors\n");
+			ret = allocated;
+			goto out_free;
+		}
+
+		dev_info(adap->pdev_dev,
+			 "Disabling offload due to insufficient MSI-X vectors\n");
+		adap->params.offload = 0;
+		adap->params.crypto = 0;
+		adap->params.ethofld = 0;
+		s->ofldqsets = 0;
+		s->eoqsets = 0;
+		s->mirrorqsets = 0;
+		uld_need = 0;
+		ethofld_need = 0;
+		mirror_need = 0;
+	}
+
+	num_vec = allocated;
+	if (num_vec < want) {
+		/* Distribute available vectors to the various queue groups.
+		 * Every group gets its minimum requirement and NIC gets top
+		 * priority for leftovers.
+		 */
+		ethqsets = eth_need;
+		if (is_uld(adap))
+			ofldqsets = nchan;
+		if (is_ethofld(adap))
+			eoqsets = ethofld_need;
+		if (s->mirrorqsets)
+			mirrorqsets = mirror_need;
+
+		num_vec -= need;
+		while (num_vec) {
+			if (num_vec < eth_need + ethofld_need ||
+			    ethqsets > s->max_ethqsets)
+				break;
+
+			for_each_port(adap, i) {
+				pi = adap2pinfo(adap, i);
+				if (pi->nqsets < 2)
+					continue;
+
+				ethqsets++;
+				num_vec--;
+				if (ethofld_need) {
+					eoqsets++;
+					num_vec--;
+				}
+			}
+		}
+
+		if (is_uld(adap)) {
+			while (num_vec) {
+				if (num_vec < uld_need ||
+				    ofldqsets > s->ofldqsets)
+					break;
+
+				ofldqsets++;
+				num_vec -= uld_need;
+			}
+		}
+
+		if (s->mirrorqsets) {
+			while (num_vec) {
+				if (num_vec < mirror_need ||
+				    mirrorqsets > s->mirrorqsets)
+					break;
+
+				mirrorqsets++;
+				num_vec -= mirror_need;
+			}
+		}
+	} else {
+		ethqsets = s->max_ethqsets;
+		if (is_uld(adap))
+			ofldqsets = s->ofldqsets;
+		if (is_ethofld(adap))
+			eoqsets = s->eoqsets;
+		if (s->mirrorqsets)
+			mirrorqsets = s->mirrorqsets;
+	}
+
+	if (ethqsets < s->max_ethqsets) {
+		s->max_ethqsets = ethqsets;
+		reduce_ethqs(adap, ethqsets);
+	}
+
+	if (is_uld(adap)) {
+		s->ofldqsets = ofldqsets;
+		s->nqs_per_uld = s->ofldqsets;
+	}
+
+	if (is_ethofld(adap))
+		s->eoqsets = eoqsets;
+
+	if (s->mirrorqsets) {
+		s->mirrorqsets = mirrorqsets;
+		for_each_port(adap, i) {
+			pi = adap2pinfo(adap, i);
+			pi->nmirrorqsets = s->mirrorqsets / nchan;
+			mutex_init(&pi->vi_mirror_mutex);
+		}
+	}
+
+	/* map for msix */
+	ret = alloc_msix_info(adap, allocated);
+	if (ret)
+		goto out_disable_msix;
+
+	for (i = 0; i < allocated; i++) {
+		adap->msix_info[i].vec = entries[i].vector;
+		adap->msix_info[i].idx = i;
+	}
+
+	dev_info(adap->pdev_dev,
+		 "%d MSI-X vectors allocated, nic %d eoqsets %d per uld %d mirrorqsets %d\n",
+		 allocated, s->max_ethqsets, s->eoqsets, s->nqs_per_uld,
+		 s->mirrorqsets);
+
+	kfree(entries);
+	return 0;
+
+out_disable_msix:
+	pci_disable_msix(adap->pdev);
+
+out_free:
+	kfree(entries);
+	return ret;
+}
+
+#undef EXTRA_VECS
+
+static int init_rss(struct adapter *adap)
+{
+	unsigned int i;
+	int err;
+
+	err = t4_init_rss_mode(adap, adap->mbox);
+	if (err)
+		return err;
+
+	for_each_port(adap, i) {
+		struct port_info *pi = adap2pinfo(adap, i);
+
+		pi->rss = kcalloc(pi->rss_size, sizeof(u16), GFP_KERNEL);
+		if (!pi->rss)
+			return -ENOMEM;
+	}
+	return 0;
+}
+
+/* Dump basic information about the adapter */
+static void print_adapter_info(struct adapter *adapter)
+{
+	/* Hardware/Firmware/etc. Version/Revision IDs */
+	t4_dump_version_info(adapter);
+
+	/* Software/Hardware configuration */
+	dev_info(adapter->pdev_dev, "Configuration: %sNIC %s, %s capable\n",
+		 is_offload(adapter) ? "R" : "",
+		 ((adapter->flags & CXGB4_USING_MSIX) ? "MSI-X" :
+		  (adapter->flags & CXGB4_USING_MSI) ? "MSI" : ""),
+		 is_offload(adapter) ? "Offload" : "non-Offload");
+}
+
+static void print_port_info(const struct net_device *dev)
+{
+	char buf[80];
+	char *bufp = buf;
+	const struct port_info *pi = netdev_priv(dev);
+	const struct adapter *adap = pi->adapter;
+
+	if (pi->link_cfg.pcaps & FW_PORT_CAP32_SPEED_100M)
+		bufp += sprintf(bufp, "100M/");
+	if (pi->link_cfg.pcaps & FW_PORT_CAP32_SPEED_1G)
+		bufp += sprintf(bufp, "1G/");
+	if (pi->link_cfg.pcaps & FW_PORT_CAP32_SPEED_10G)
+		bufp += sprintf(bufp, "10G/");
+	if (pi->link_cfg.pcaps & FW_PORT_CAP32_SPEED_25G)
+		bufp += sprintf(bufp, "25G/");
+	if (pi->link_cfg.pcaps & FW_PORT_CAP32_SPEED_40G)
+		bufp += sprintf(bufp, "40G/");
+	if (pi->link_cfg.pcaps & FW_PORT_CAP32_SPEED_50G)
+		bufp += sprintf(bufp, "50G/");
+	if (pi->link_cfg.pcaps & FW_PORT_CAP32_SPEED_100G)
+		bufp += sprintf(bufp, "100G/");
+	if (pi->link_cfg.pcaps & FW_PORT_CAP32_SPEED_200G)
+		bufp += sprintf(bufp, "200G/");
+	if (pi->link_cfg.pcaps & FW_PORT_CAP32_SPEED_400G)
+		bufp += sprintf(bufp, "400G/");
+	if (bufp != buf)
+		--bufp;
+	sprintf(bufp, "BASE-%s", t4_get_port_type_description(pi->port_type));
+
+	netdev_info(dev, "Chelsio %s %s\n", adap->params.vpd.id, buf);
+}
+
+/*
+ * Free the following resources:
+ * - memory used for tables
+ * - MSI/MSI-X
+ * - net devices
+ * - resources FW is holding for us
+ */
+static void free_some_resources(struct adapter *adapter)
+{
+	unsigned int i;
+
+	kvfree(adapter->smt);
+	kvfree(adapter->l2t);
+	kvfree(adapter->srq);
+	t4_cleanup_sched(adapter);
+	kvfree(adapter->tids.tid_tab);
+	cxgb4_cleanup_tc_matchall(adapter);
+	cxgb4_cleanup_tc_mqprio(adapter);
+	cxgb4_cleanup_tc_flower(adapter);
+	cxgb4_cleanup_tc_u32(adapter);
+	cxgb4_cleanup_ethtool_filters(adapter);
+	kfree(adapter->sge.egr_map);
+	kfree(adapter->sge.ingr_map);
+	bitmap_free(adapter->sge.starving_fl);
+	bitmap_free(adapter->sge.txq_maperr);
+#ifdef CONFIG_DEBUG_FS
+	bitmap_free(adapter->sge.blocked_fl);
+#endif
+	disable_msi(adapter);
+
+	for_each_port(adapter, i)
+		if (adapter->port[i]) {
+			struct port_info *pi = adap2pinfo(adapter, i);
+
+			if (pi->viid != 0)
+				t4_free_vi(adapter, adapter->mbox, adapter->pf,
+					   0, pi->viid);
+			kfree(adap2pinfo(adapter, i)->rss);
+			free_netdev(adapter->port[i]);
+		}
+	if (adapter->flags & CXGB4_FW_OK)
+		t4_fw_bye(adapter, adapter->pf);
+}
+
+#define TSO_FLAGS (NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_TSO_ECN | \
+		   NETIF_F_GSO_UDP_L4)
+#define VLAN_FEAT (NETIF_F_SG | NETIF_F_IP_CSUM | TSO_FLAGS | \
+		   NETIF_F_GRO | NETIF_F_IPV6_CSUM | NETIF_F_HIGHDMA)
+#define SEGMENT_SIZE 128
+
+static int t4_get_chip_type(struct adapter *adap, int ver)
+{
+	u32 pl_rev = REV_G(t4_read_reg(adap, PL_REV_A));
+
+	switch (ver) {
+	case CHELSIO_T4:
+		return CHELSIO_CHIP_CODE(CHELSIO_T4, pl_rev);
+	case CHELSIO_T5:
+		return CHELSIO_CHIP_CODE(CHELSIO_T5, pl_rev);
+	case CHELSIO_T6:
+		return CHELSIO_CHIP_CODE(CHELSIO_T6, pl_rev);
+	default:
+		break;
+	}
+	return -EINVAL;
+}
+
+#ifdef CONFIG_PCI_IOV
+static void cxgb4_mgmt_setup(struct net_device *dev)
+{
+	dev->type = ARPHRD_NONE;
+	dev->mtu = 0;
+	dev->hard_header_len = 0;
+	dev->addr_len = 0;
+	dev->tx_queue_len = 0;
+	dev->flags |= IFF_NOARP;
+	dev->priv_flags |= IFF_NO_QUEUE;
+
+	/* Initialize the device structure. */
+	dev->netdev_ops = &cxgb4_mgmt_netdev_ops;
+	dev->ethtool_ops = &cxgb4_mgmt_ethtool_ops;
+}
+
+static int cxgb4_iov_configure(struct pci_dev *pdev, int num_vfs)
+{
+	struct adapter *adap = pci_get_drvdata(pdev);
+	int err = 0;
+	int current_vfs = pci_num_vf(pdev);
+	u32 pcie_fw;
+
+	pcie_fw = readl(adap->regs + PCIE_FW_A);
+	/* Check if fw is initialized */
+	if (!(pcie_fw & PCIE_FW_INIT_F)) {
+		dev_warn(&pdev->dev, "Device not initialized\n");
+		return -EOPNOTSUPP;
+	}
+
+	/* If any of the VF's is already assigned to Guest OS, then
+	 * SRIOV for the same cannot be modified
+	 */
+	if (current_vfs && pci_vfs_assigned(pdev)) {
+		dev_err(&pdev->dev,
+			"Cannot modify SR-IOV while VFs are assigned\n");
+		return current_vfs;
+	}
+	/* Note that the upper-level code ensures that we're never called with
+	 * a non-zero "num_vfs" when we already have VFs instantiated.  But
+	 * it never hurts to code defensively.
+	 */
+	if (num_vfs != 0 && current_vfs != 0)
+		return -EBUSY;
+
+	/* Nothing to do for no change. */
+	if (num_vfs == current_vfs)
+		return num_vfs;
+
+	/* Disable SRIOV when zero is passed. */
+	if (!num_vfs) {
+		pci_disable_sriov(pdev);
+		/* free VF Management Interface */
+		unregister_netdev(adap->port[0]);
+		free_netdev(adap->port[0]);
+		adap->port[0] = NULL;
+
+		/* free VF resources */
+		adap->num_vfs = 0;
+		kfree(adap->vfinfo);
+		adap->vfinfo = NULL;
+		return 0;
+	}
+
+	if (!current_vfs) {
+		struct fw_pfvf_cmd port_cmd, port_rpl;
+		struct net_device *netdev;
+		unsigned int pmask, port;
+		struct pci_dev *pbridge;
+		struct port_info *pi;
+		char name[IFNAMSIZ];
+		u32 devcap2;
+		u16 flags;
+
+		/* If we want to instantiate Virtual Functions, then our
+		 * parent bridge's PCI-E needs to support Alternative Routing
+		 * ID (ARI) because our VFs will show up at function offset 8
+		 * and above.
+		 */
+		pbridge = pdev->bus->self;
+		pcie_capability_read_word(pbridge, PCI_EXP_FLAGS, &flags);
+		pcie_capability_read_dword(pbridge, PCI_EXP_DEVCAP2, &devcap2);
+
+		if ((flags & PCI_EXP_FLAGS_VERS) < 2 ||
+		    !(devcap2 & PCI_EXP_DEVCAP2_ARI)) {
+			/* Our parent bridge does not support ARI so issue a
+			 * warning and skip instantiating the VFs.  They
+			 * won't be reachable.
+			 */
+			dev_warn(&pdev->dev, "Parent bridge %02x:%02x.%x doesn't support ARI; can't instantiate Virtual Functions\n",
+				 pbridge->bus->number, PCI_SLOT(pbridge->devfn),
+				 PCI_FUNC(pbridge->devfn));
+			return -ENOTSUPP;
+		}
+		memset(&port_cmd, 0, sizeof(port_cmd));
+		port_cmd.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_PFVF_CMD) |
+						 FW_CMD_REQUEST_F |
+						 FW_CMD_READ_F |
+						 FW_PFVF_CMD_PFN_V(adap->pf) |
+						 FW_PFVF_CMD_VFN_V(0));
+		port_cmd.retval_len16 = cpu_to_be32(FW_LEN16(port_cmd));
+		err = t4_wr_mbox(adap, adap->mbox, &port_cmd, sizeof(port_cmd),
+				 &port_rpl);
+		if (err)
+			return err;
+		pmask = FW_PFVF_CMD_PMASK_G(be32_to_cpu(port_rpl.type_to_neq));
+		port = ffs(pmask) - 1;
+		/* Allocate VF Management Interface. */
+		snprintf(name, IFNAMSIZ, "mgmtpf%d,%d", adap->adap_idx,
+			 adap->pf);
+		netdev = alloc_netdev(sizeof(struct port_info),
+				      name, NET_NAME_UNKNOWN, cxgb4_mgmt_setup);
+		if (!netdev)
+			return -ENOMEM;
+
+		pi = netdev_priv(netdev);
+		pi->adapter = adap;
+		pi->lport = port;
+		pi->tx_chan = port;
+		SET_NETDEV_DEV(netdev, &pdev->dev);
+
+		adap->port[0] = netdev;
+		pi->port_id = 0;
+
+		err = register_netdev(adap->port[0]);
+		if (err) {
+			pr_info("Unable to register VF mgmt netdev %s\n", name);
+			free_netdev(adap->port[0]);
+			adap->port[0] = NULL;
+			return err;
+		}
+		/* Allocate and set up VF Information. */
+		adap->vfinfo = kcalloc(pci_sriov_get_totalvfs(pdev),
+				       sizeof(struct vf_info), GFP_KERNEL);
+		if (!adap->vfinfo) {
+			unregister_netdev(adap->port[0]);
+			free_netdev(adap->port[0]);
+			adap->port[0] = NULL;
+			return -ENOMEM;
+		}
+		cxgb4_mgmt_fill_vf_station_mac_addr(adap);
+	}
+	/* Instantiate the requested number of VFs. */
+	err = pci_enable_sriov(pdev, num_vfs);
+	if (err) {
+		pr_info("Unable to instantiate %d VFs\n", num_vfs);
+		if (!current_vfs) {
+			unregister_netdev(adap->port[0]);
+			free_netdev(adap->port[0]);
+			adap->port[0] = NULL;
+			kfree(adap->vfinfo);
+			adap->vfinfo = NULL;
+		}
+		return err;
+	}
+
+	adap->num_vfs = num_vfs;
+	return num_vfs;
+}
+#endif /* CONFIG_PCI_IOV */
+
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE) || IS_ENABLED(CONFIG_CHELSIO_IPSEC_INLINE)
+
+static int chcr_offload_state(struct adapter *adap,
+			      enum cxgb4_netdev_tls_ops op_val)
+{
+	switch (op_val) {
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+	case CXGB4_TLSDEV_OPS:
+		if (!adap->uld[CXGB4_ULD_KTLS].handle) {
+			dev_dbg(adap->pdev_dev, "ch_ktls driver is not loaded\n");
+			return -EOPNOTSUPP;
+		}
+		if (!adap->uld[CXGB4_ULD_KTLS].tlsdev_ops) {
+			dev_dbg(adap->pdev_dev,
+				"ch_ktls driver has no registered tlsdev_ops\n");
+			return -EOPNOTSUPP;
+		}
+		break;
+#endif /* CONFIG_CHELSIO_TLS_DEVICE */
+#if IS_ENABLED(CONFIG_CHELSIO_IPSEC_INLINE)
+	case CXGB4_XFRMDEV_OPS:
+		if (!adap->uld[CXGB4_ULD_IPSEC].handle) {
+			dev_dbg(adap->pdev_dev, "chipsec driver is not loaded\n");
+			return -EOPNOTSUPP;
+		}
+		if (!adap->uld[CXGB4_ULD_IPSEC].xfrmdev_ops) {
+			dev_dbg(adap->pdev_dev,
+				"chipsec driver has no registered xfrmdev_ops\n");
+			return -EOPNOTSUPP;
+		}
+		break;
+#endif /* CONFIG_CHELSIO_IPSEC_INLINE */
+	default:
+		dev_dbg(adap->pdev_dev,
+			"driver has no support for offload %d\n", op_val);
+		return -EOPNOTSUPP;
+	}
+
+	return 0;
+}
+
+#endif /* CONFIG_CHELSIO_TLS_DEVICE || CONFIG_CHELSIO_IPSEC_INLINE */
+
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+
+static int cxgb4_ktls_dev_add(struct net_device *netdev, struct sock *sk,
+			      enum tls_offload_ctx_dir direction,
+			      struct tls_crypto_info *crypto_info,
+			      u32 tcp_sn)
+{
+	struct adapter *adap = netdev2adap(netdev);
+	int ret;
+
+	mutex_lock(&uld_mutex);
+	ret = chcr_offload_state(adap, CXGB4_TLSDEV_OPS);
+	if (ret)
+		goto out_unlock;
+
+	ret = cxgb4_set_ktls_feature(adap, FW_PARAMS_PARAM_DEV_KTLS_HW_ENABLE);
+	if (ret)
+		goto out_unlock;
+
+	ret = adap->uld[CXGB4_ULD_KTLS].tlsdev_ops->tls_dev_add(netdev, sk,
+								direction,
+								crypto_info,
+								tcp_sn);
+	/* if there is a failure, clear the refcount */
+	if (ret)
+		cxgb4_set_ktls_feature(adap,
+				       FW_PARAMS_PARAM_DEV_KTLS_HW_DISABLE);
+out_unlock:
+	mutex_unlock(&uld_mutex);
+	return ret;
+}
+
+static void cxgb4_ktls_dev_del(struct net_device *netdev,
+			       struct tls_context *tls_ctx,
+			       enum tls_offload_ctx_dir direction)
+{
+	struct adapter *adap = netdev2adap(netdev);
+
+	mutex_lock(&uld_mutex);
+	if (chcr_offload_state(adap, CXGB4_TLSDEV_OPS))
+		goto out_unlock;
+
+	adap->uld[CXGB4_ULD_KTLS].tlsdev_ops->tls_dev_del(netdev, tls_ctx,
+							  direction);
+
+out_unlock:
+	cxgb4_set_ktls_feature(adap, FW_PARAMS_PARAM_DEV_KTLS_HW_DISABLE);
+	mutex_unlock(&uld_mutex);
+}
+
+static const struct tlsdev_ops cxgb4_ktls_ops = {
+	.tls_dev_add = cxgb4_ktls_dev_add,
+	.tls_dev_del = cxgb4_ktls_dev_del,
+};
+#endif /* CONFIG_CHELSIO_TLS_DEVICE */
+
+#if IS_ENABLED(CONFIG_CHELSIO_IPSEC_INLINE)
+
+static int cxgb4_xfrm_add_state(struct xfrm_state *x)
+{
+	struct adapter *adap = netdev2adap(x->xso.dev);
+	int ret;
+
+	if (!mutex_trylock(&uld_mutex)) {
+		dev_dbg(adap->pdev_dev,
+			"crypto uld critical resource is under use\n");
+		return -EBUSY;
+	}
+	ret = chcr_offload_state(adap, CXGB4_XFRMDEV_OPS);
+	if (ret)
+		goto out_unlock;
+
+	ret = adap->uld[CXGB4_ULD_IPSEC].xfrmdev_ops->xdo_dev_state_add(x);
+
+out_unlock:
+	mutex_unlock(&uld_mutex);
+
+	return ret;
+}
+
+static void cxgb4_xfrm_del_state(struct xfrm_state *x)
+{
+	struct adapter *adap = netdev2adap(x->xso.dev);
+
+	if (!mutex_trylock(&uld_mutex)) {
+		dev_dbg(adap->pdev_dev,
+			"crypto uld critical resource is under use\n");
+		return;
+	}
+	if (chcr_offload_state(adap, CXGB4_XFRMDEV_OPS))
+		goto out_unlock;
+
+	adap->uld[CXGB4_ULD_IPSEC].xfrmdev_ops->xdo_dev_state_delete(x);
+
+out_unlock:
+	mutex_unlock(&uld_mutex);
+}
+
+static void cxgb4_xfrm_free_state(struct xfrm_state *x)
+{
+	struct adapter *adap = netdev2adap(x->xso.dev);
+
+	if (!mutex_trylock(&uld_mutex)) {
+		dev_dbg(adap->pdev_dev,
+			"crypto uld critical resource is under use\n");
+		return;
+	}
+	if (chcr_offload_state(adap, CXGB4_XFRMDEV_OPS))
+		goto out_unlock;
+
+	adap->uld[CXGB4_ULD_IPSEC].xfrmdev_ops->xdo_dev_state_free(x);
+
+out_unlock:
+	mutex_unlock(&uld_mutex);
+}
+
+static bool cxgb4_ipsec_offload_ok(struct sk_buff *skb, struct xfrm_state *x)
+{
+	struct adapter *adap = netdev2adap(x->xso.dev);
+	bool ret = false;
+
+	if (!mutex_trylock(&uld_mutex)) {
+		dev_dbg(adap->pdev_dev,
+			"crypto uld critical resource is under use\n");
+		return ret;
+	}
+	if (chcr_offload_state(adap, CXGB4_XFRMDEV_OPS))
+		goto out_unlock;
+
+	ret = adap->uld[CXGB4_ULD_IPSEC].xfrmdev_ops->xdo_dev_offload_ok(skb, x);
+
+out_unlock:
+	mutex_unlock(&uld_mutex);
+	return ret;
+}
+
+static void cxgb4_advance_esn_state(struct xfrm_state *x)
+{
+	struct adapter *adap = netdev2adap(x->xso.dev);
+
+	if (!mutex_trylock(&uld_mutex)) {
+		dev_dbg(adap->pdev_dev,
+			"crypto uld critical resource is under use\n");
+		return;
+	}
+	if (chcr_offload_state(adap, CXGB4_XFRMDEV_OPS))
+		goto out_unlock;
+
+	adap->uld[CXGB4_ULD_IPSEC].xfrmdev_ops->xdo_dev_state_advance_esn(x);
+
+out_unlock:
+	mutex_unlock(&uld_mutex);
+}
+
+static const struct xfrmdev_ops cxgb4_xfrmdev_ops = {
+	.xdo_dev_state_add      = cxgb4_xfrm_add_state,
+	.xdo_dev_state_delete   = cxgb4_xfrm_del_state,
+	.xdo_dev_state_free     = cxgb4_xfrm_free_state,
+	.xdo_dev_offload_ok     = cxgb4_ipsec_offload_ok,
+	.xdo_dev_state_advance_esn = cxgb4_advance_esn_state,
+};
+
+#endif /* CONFIG_CHELSIO_IPSEC_INLINE */
+
+static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	struct net_device *netdev;
+	struct adapter *adapter;
+	static int adap_idx = 1;
+	int s_qpp, qpp, num_seg;
+	struct port_info *pi;
+	enum chip_type chip;
+	void __iomem *regs;
+	int func, chip_ver;
+	u16 device_id;
+	int i, err;
+	u32 whoami;
+
+	err = pci_request_regions(pdev, KBUILD_MODNAME);
+	if (err) {
+		/* Just info, some other driver may have claimed the device. */
+		dev_info(&pdev->dev, "cannot obtain PCI resources\n");
+		return err;
+	}
+
+	err = pci_enable_device(pdev);
+	if (err) {
+		dev_err(&pdev->dev, "cannot enable PCI device\n");
+		goto out_release_regions;
+	}
+
+	regs = pci_ioremap_bar(pdev, 0);
+	if (!regs) {
+		dev_err(&pdev->dev, "cannot map device registers\n");
+		err = -ENOMEM;
+		goto out_disable_device;
+	}
+
+	adapter = kzalloc(sizeof(*adapter), GFP_KERNEL);
+	if (!adapter) {
+		err = -ENOMEM;
+		goto out_unmap_bar0;
+	}
+
+	adapter->regs = regs;
+	err = t4_wait_dev_ready(regs);
+	if (err < 0)
+		goto out_free_adapter;
+
+	/* We control everything through one PF */
+	whoami = t4_read_reg(adapter, PL_WHOAMI_A);
+	pci_read_config_word(pdev, PCI_DEVICE_ID, &device_id);
+	chip = t4_get_chip_type(adapter, CHELSIO_PCI_ID_VER(device_id));
+	if ((int)chip < 0) {
+		dev_err(&pdev->dev, "Device %d is not supported\n", device_id);
+		err = chip;
+		goto out_free_adapter;
+	}
+	chip_ver = CHELSIO_CHIP_VERSION(chip);
+	func = chip_ver <= CHELSIO_T5 ?
+	       SOURCEPF_G(whoami) : T6_SOURCEPF_G(whoami);
+
+	adapter->pdev = pdev;
+	adapter->pdev_dev = &pdev->dev;
+	adapter->name = pci_name(pdev);
+	adapter->mbox = func;
+	adapter->pf = func;
+	adapter->params.chip = chip;
+	adapter->adap_idx = adap_idx;
+	adapter->msg_enable = DFLT_MSG_ENABLE;
+	adapter->mbox_log = kzalloc(sizeof(*adapter->mbox_log) +
+				    (sizeof(struct mbox_cmd) *
+				     T4_OS_LOG_MBOX_CMDS),
+				    GFP_KERNEL);
+	if (!adapter->mbox_log) {
+		err = -ENOMEM;
+		goto out_free_adapter;
+	}
+	spin_lock_init(&adapter->mbox_lock);
+	INIT_LIST_HEAD(&adapter->mlist.list);
+	adapter->mbox_log->size = T4_OS_LOG_MBOX_CMDS;
+	pci_set_drvdata(pdev, adapter);
+
+	if (func != ent->driver_data) {
+		pci_disable_device(pdev);
+		pci_save_state(pdev);        /* to restore SR-IOV later */
+		return 0;
+	}
+
+	err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
+	if (err) {
+		dev_err(&pdev->dev, "no usable DMA configuration\n");
+		goto out_free_adapter;
+	}
+
+	pci_enable_pcie_error_reporting(pdev);
+	pci_set_master(pdev);
+	pci_save_state(pdev);
+	adap_idx++;
+	adapter->workq = create_singlethread_workqueue("cxgb4");
+	if (!adapter->workq) {
+		err = -ENOMEM;
+		goto out_free_adapter;
+	}
+
+	/* PCI device has been enabled */
+	adapter->flags |= CXGB4_DEV_ENABLED;
+	memset(adapter->chan_map, 0xff, sizeof(adapter->chan_map));
+
+	/* If possible, we use PCIe Relaxed Ordering Attribute to deliver
+	 * Ingress Packet Data to Free List Buffers in order to allow for
+	 * chipset performance optimizations between the Root Complex and
+	 * Memory Controllers.  (Messages to the associated Ingress Queue
+	 * notifying new Packet Placement in the Free Lists Buffers will be
+	 * send without the Relaxed Ordering Attribute thus guaranteeing that
+	 * all preceding PCIe Transaction Layer Packets will be processed
+	 * first.)  But some Root Complexes have various issues with Upstream
+	 * Transaction Layer Packets with the Relaxed Ordering Attribute set.
+	 * The PCIe devices which under the Root Complexes will be cleared the
+	 * Relaxed Ordering bit in the configuration space, So we check our
+	 * PCIe configuration space to see if it's flagged with advice against
+	 * using Relaxed Ordering.
+	 */
+	if (!pcie_relaxed_ordering_enabled(pdev))
+		adapter->flags |= CXGB4_ROOT_NO_RELAXED_ORDERING;
+
+	spin_lock_init(&adapter->stats_lock);
+	spin_lock_init(&adapter->tid_release_lock);
+	spin_lock_init(&adapter->win0_lock);
+
+	INIT_WORK(&adapter->tid_release_task, process_tid_release_list);
+	INIT_WORK(&adapter->db_full_task, process_db_full);
+	INIT_WORK(&adapter->db_drop_task, process_db_drop);
+	INIT_WORK(&adapter->fatal_err_notify_task, notify_fatal_err);
+
+	err = t4_prep_adapter(adapter);
+	if (err)
+		goto out_free_adapter;
+
+	if (is_kdump_kernel()) {
+		/* Collect hardware state and append to /proc/vmcore */
+		err = cxgb4_cudbg_vmcore_add_dump(adapter);
+		if (err) {
+			dev_warn(adapter->pdev_dev,
+				 "Fail collecting vmcore device dump, err: %d. Continuing\n",
+				 err);
+			err = 0;
+		}
+	}
+
+	if (!is_t4(adapter->params.chip)) {
+		s_qpp = (QUEUESPERPAGEPF0_S +
+			(QUEUESPERPAGEPF1_S - QUEUESPERPAGEPF0_S) *
+			adapter->pf);
+		qpp = 1 << QUEUESPERPAGEPF0_G(t4_read_reg(adapter,
+		      SGE_EGRESS_QUEUES_PER_PAGE_PF_A) >> s_qpp);
+		num_seg = PAGE_SIZE / SEGMENT_SIZE;
+
+		/* Each segment size is 128B. Write coalescing is enabled only
+		 * when SGE_EGRESS_QUEUES_PER_PAGE_PF reg value for the
+		 * queue is less no of segments that can be accommodated in
+		 * a page size.
+		 */
+		if (qpp > num_seg) {
+			dev_err(&pdev->dev,
+				"Incorrect number of egress queues per page\n");
+			err = -EINVAL;
+			goto out_free_adapter;
+		}
+		adapter->bar2 = ioremap_wc(pci_resource_start(pdev, 2),
+		pci_resource_len(pdev, 2));
+		if (!adapter->bar2) {
+			dev_err(&pdev->dev, "cannot map device bar2 region\n");
+			err = -ENOMEM;
+			goto out_free_adapter;
+		}
+	}
+
+	setup_memwin(adapter);
+	err = adap_init0(adapter, 0);
+	if (err)
+		goto out_unmap_bar;
+
+	setup_memwin_rdma(adapter);
+
+	/* configure SGE_STAT_CFG_A to read WC stats */
+	if (!is_t4(adapter->params.chip))
+		t4_write_reg(adapter, SGE_STAT_CFG_A, STATSOURCE_T5_V(7) |
+			     (is_t5(adapter->params.chip) ? STATMODE_V(0) :
+			      T6_STATMODE_V(0)));
+
+	/* Initialize hash mac addr list */
+	INIT_LIST_HEAD(&adapter->mac_hlist);
+
+	for_each_port(adapter, i) {
+		/* For supporting MQPRIO Offload, need some extra
+		 * queues for each ETHOFLD TIDs. Keep it equal to
+		 * MAX_ATIDs for now. Once we connect to firmware
+		 * later and query the EOTID params, we'll come to
+		 * know the actual # of EOTIDs supported.
+		 */
+		netdev = alloc_etherdev_mq(sizeof(struct port_info),
+					   MAX_ETH_QSETS + MAX_ATIDS);
+		if (!netdev) {
+			err = -ENOMEM;
+			goto out_free_dev;
+		}
+
+		SET_NETDEV_DEV(netdev, &pdev->dev);
+
+		adapter->port[i] = netdev;
+		pi = netdev_priv(netdev);
+		pi->adapter = adapter;
+		pi->xact_addr_filt = -1;
+		pi->port_id = i;
+		netdev->irq = pdev->irq;
+
+		netdev->hw_features = NETIF_F_SG | TSO_FLAGS |
+			NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
+			NETIF_F_RXCSUM | NETIF_F_RXHASH | NETIF_F_GRO |
+			NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_CTAG_RX |
+			NETIF_F_HW_TC | NETIF_F_NTUPLE | NETIF_F_HIGHDMA;
+
+		if (chip_ver > CHELSIO_T5) {
+			netdev->hw_enc_features |= NETIF_F_IP_CSUM |
+						   NETIF_F_IPV6_CSUM |
+						   NETIF_F_RXCSUM |
+						   NETIF_F_GSO_UDP_TUNNEL |
+						   NETIF_F_GSO_UDP_TUNNEL_CSUM |
+						   NETIF_F_TSO | NETIF_F_TSO6;
+
+			netdev->hw_features |= NETIF_F_GSO_UDP_TUNNEL |
+					       NETIF_F_GSO_UDP_TUNNEL_CSUM |
+					       NETIF_F_HW_TLS_RECORD;
+
+			if (adapter->rawf_cnt)
+				netdev->udp_tunnel_nic_info = &cxgb_udp_tunnels;
+		}
+
+		netdev->features |= netdev->hw_features;
+		netdev->vlan_features = netdev->features & VLAN_FEAT;
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+		if (pi->adapter->params.crypto & FW_CAPS_CONFIG_TLS_HW) {
+			netdev->hw_features |= NETIF_F_HW_TLS_TX;
+			netdev->tlsdev_ops = &cxgb4_ktls_ops;
+			/* initialize the refcount */
+			refcount_set(&pi->adapter->chcr_ktls.ktls_refcount, 0);
+		}
+#endif /* CONFIG_CHELSIO_TLS_DEVICE */
+#if IS_ENABLED(CONFIG_CHELSIO_IPSEC_INLINE)
+		if (pi->adapter->params.crypto & FW_CAPS_CONFIG_IPSEC_INLINE) {
+			netdev->hw_enc_features |= NETIF_F_HW_ESP;
+			netdev->features |= NETIF_F_HW_ESP;
+			netdev->xfrmdev_ops = &cxgb4_xfrmdev_ops;
+		}
+#endif /* CONFIG_CHELSIO_IPSEC_INLINE */
+
+		netdev->priv_flags |= IFF_UNICAST_FLT;
+
+		/* MTU range: 81 - 9600 */
+		netdev->min_mtu = 81;              /* accommodate SACK */
+		netdev->max_mtu = MAX_MTU;
+
+		netdev->netdev_ops = &cxgb4_netdev_ops;
+#ifdef CONFIG_CHELSIO_T4_DCB
+		netdev->dcbnl_ops = &cxgb4_dcb_ops;
+		cxgb4_dcb_state_init(netdev);
+		cxgb4_dcb_version_init(netdev);
+#endif
+		cxgb4_set_ethtool_ops(netdev);
+	}
+
+	cxgb4_init_ethtool_dump(adapter);
+
+	pci_set_drvdata(pdev, adapter);
+
+	if (adapter->flags & CXGB4_FW_OK) {
+		err = t4_port_init(adapter, func, func, 0);
+		if (err)
+			goto out_free_dev;
+	} else if (adapter->params.nports == 1) {
+		/* If we don't have a connection to the firmware -- possibly
+		 * because of an error -- grab the raw VPD parameters so we
+		 * can set the proper MAC Address on the debug network
+		 * interface that we've created.
+		 */
+		u8 hw_addr[ETH_ALEN];
+		u8 *na = adapter->params.vpd.na;
+
+		err = t4_get_raw_vpd_params(adapter, &adapter->params.vpd);
+		if (!err) {
+			for (i = 0; i < ETH_ALEN; i++)
+				hw_addr[i] = (hex2val(na[2 * i + 0]) * 16 +
+					      hex2val(na[2 * i + 1]));
+			t4_set_hw_addr(adapter, 0, hw_addr);
+		}
+	}
+
+	if (!(adapter->flags & CXGB4_FW_OK))
+		goto fw_attach_fail;
+
+	/* Configure queues and allocate tables now, they can be needed as
+	 * soon as the first register_netdev completes.
+	 */
+	err = cfg_queues(adapter);
+	if (err)
+		goto out_free_dev;
+
+	adapter->smt = t4_init_smt();
+	if (!adapter->smt) {
+		/* We tolerate a lack of SMT, giving up some functionality */
+		dev_warn(&pdev->dev, "could not allocate SMT, continuing\n");
+	}
+
+	adapter->l2t = t4_init_l2t(adapter->l2t_start, adapter->l2t_end);
+	if (!adapter->l2t) {
+		/* We tolerate a lack of L2T, giving up some functionality */
+		dev_warn(&pdev->dev, "could not allocate L2T, continuing\n");
+		adapter->params.offload = 0;
+	}
+
+#if IS_ENABLED(CONFIG_IPV6)
+	if (chip_ver <= CHELSIO_T5 &&
+	    (!(t4_read_reg(adapter, LE_DB_CONFIG_A) & ASLIPCOMPEN_F))) {
+		/* CLIP functionality is not present in hardware,
+		 * hence disable all offload features
+		 */
+		dev_warn(&pdev->dev,
+			 "CLIP not enabled in hardware, continuing\n");
+		adapter->params.offload = 0;
+	} else {
+		adapter->clipt = t4_init_clip_tbl(adapter->clipt_start,
+						  adapter->clipt_end);
+		if (!adapter->clipt) {
+			/* We tolerate a lack of clip_table, giving up
+			 * some functionality
+			 */
+			dev_warn(&pdev->dev,
+				 "could not allocate Clip table, continuing\n");
+			adapter->params.offload = 0;
+		}
+	}
+#endif
+
+	for_each_port(adapter, i) {
+		pi = adap2pinfo(adapter, i);
+		pi->sched_tbl = t4_init_sched(adapter->params.nsched_cls);
+		if (!pi->sched_tbl)
+			dev_warn(&pdev->dev,
+				 "could not activate scheduling on port %d\n",
+				 i);
+	}
+
+	if (is_offload(adapter) || is_hashfilter(adapter)) {
+		if (t4_read_reg(adapter, LE_DB_CONFIG_A) & HASHEN_F) {
+			u32 v;
+
+			v = t4_read_reg(adapter, LE_DB_HASH_CONFIG_A);
+			if (chip_ver <= CHELSIO_T5) {
+				adapter->tids.nhash = 1 << HASHTIDSIZE_G(v);
+				v = t4_read_reg(adapter, LE_DB_TID_HASHBASE_A);
+				adapter->tids.hash_base = v / 4;
+			} else {
+				adapter->tids.nhash = HASHTBLSIZE_G(v) << 3;
+				v = t4_read_reg(adapter,
+						T6_LE_DB_HASH_TID_BASE_A);
+				adapter->tids.hash_base = v;
+			}
+		}
+	}
+
+	if (tid_init(&adapter->tids) < 0) {
+		dev_warn(&pdev->dev, "could not allocate TID table, "
+			 "continuing\n");
+		adapter->params.offload = 0;
+	} else {
+		adapter->tc_u32 = cxgb4_init_tc_u32(adapter);
+		if (!adapter->tc_u32)
+			dev_warn(&pdev->dev,
+				 "could not offload tc u32, continuing\n");
+
+		if (cxgb4_init_tc_flower(adapter))
+			dev_warn(&pdev->dev,
+				 "could not offload tc flower, continuing\n");
+
+		if (cxgb4_init_tc_mqprio(adapter))
+			dev_warn(&pdev->dev,
+				 "could not offload tc mqprio, continuing\n");
+
+		if (cxgb4_init_tc_matchall(adapter))
+			dev_warn(&pdev->dev,
+				 "could not offload tc matchall, continuing\n");
+		if (cxgb4_init_ethtool_filters(adapter))
+			dev_warn(&pdev->dev,
+				 "could not initialize ethtool filters, continuing\n");
+	}
+
+	/* See what interrupts we'll be using */
+	if (msi > 1 && enable_msix(adapter) == 0)
+		adapter->flags |= CXGB4_USING_MSIX;
+	else if (msi > 0 && pci_enable_msi(pdev) == 0) {
+		adapter->flags |= CXGB4_USING_MSI;
+		if (msi > 1)
+			free_msix_info(adapter);
+	}
+
+	/* check for PCI Express bandwidth capabiltites */
+	pcie_print_link_status(pdev);
+
+	cxgb4_init_mps_ref_entries(adapter);
+
+	err = init_rss(adapter);
+	if (err)
+		goto out_free_dev;
+
+	err = setup_non_data_intr(adapter);
+	if (err) {
+		dev_err(adapter->pdev_dev,
+			"Non Data interrupt allocation failed, err: %d\n", err);
+		goto out_free_dev;
+	}
+
+	err = setup_fw_sge_queues(adapter);
+	if (err) {
+		dev_err(adapter->pdev_dev,
+			"FW sge queue allocation failed, err %d", err);
+		goto out_free_dev;
+	}
+
+fw_attach_fail:
+	/*
+	 * The card is now ready to go.  If any errors occur during device
+	 * registration we do not fail the whole card but rather proceed only
+	 * with the ports we manage to register successfully.  However we must
+	 * register at least one net device.
+	 */
+	for_each_port(adapter, i) {
+		pi = adap2pinfo(adapter, i);
+		adapter->port[i]->dev_port = pi->lport;
+		netif_set_real_num_tx_queues(adapter->port[i], pi->nqsets);
+		netif_set_real_num_rx_queues(adapter->port[i], pi->nqsets);
+
+		netif_carrier_off(adapter->port[i]);
+
+		err = register_netdev(adapter->port[i]);
+		if (err)
+			break;
+		adapter->chan_map[pi->tx_chan] = i;
+		print_port_info(adapter->port[i]);
+	}
+	if (i == 0) {
+		dev_err(&pdev->dev, "could not register any net devices\n");
+		goto out_free_dev;
+	}
+	if (err) {
+		dev_warn(&pdev->dev, "only %d net devices registered\n", i);
+		err = 0;
+	}
+
+	if (cxgb4_debugfs_root) {
+		adapter->debugfs_root = debugfs_create_dir(pci_name(pdev),
+							   cxgb4_debugfs_root);
+		setup_debugfs(adapter);
+	}
+
+	/* PCIe EEH recovery on powerpc platforms needs fundamental reset */
+	pdev->needs_freset = 1;
+
+	if (is_uld(adapter))
+		cxgb4_uld_enable(adapter);
+
+	if (!is_t4(adapter->params.chip))
+		cxgb4_ptp_init(adapter);
+
+	if (IS_REACHABLE(CONFIG_THERMAL) &&
+	    !is_t4(adapter->params.chip) && (adapter->flags & CXGB4_FW_OK))
+		cxgb4_thermal_init(adapter);
+
+	print_adapter_info(adapter);
+	return 0;
+
+ out_free_dev:
+	t4_free_sge_resources(adapter);
+	free_some_resources(adapter);
+	if (adapter->flags & CXGB4_USING_MSIX)
+		free_msix_info(adapter);
+	if (adapter->num_uld || adapter->num_ofld_uld)
+		t4_uld_mem_free(adapter);
+ out_unmap_bar:
+	if (!is_t4(adapter->params.chip))
+		iounmap(adapter->bar2);
+ out_free_adapter:
+	if (adapter->workq)
+		destroy_workqueue(adapter->workq);
+
+	kfree(adapter->mbox_log);
+	kfree(adapter);
+ out_unmap_bar0:
+	iounmap(regs);
+ out_disable_device:
+	pci_disable_pcie_error_reporting(pdev);
+	pci_disable_device(pdev);
+ out_release_regions:
+	pci_release_regions(pdev);
+	return err;
+}
+
+static void remove_one(struct pci_dev *pdev)
+{
+	struct adapter *adapter = pci_get_drvdata(pdev);
+	struct hash_mac_addr *entry, *tmp;
+
+	if (!adapter) {
+		pci_release_regions(pdev);
+		return;
+	}
+
+	/* If we allocated filters, free up state associated with any
+	 * valid filters ...
+	 */
+	clear_all_filters(adapter);
+
+	adapter->flags |= CXGB4_SHUTTING_DOWN;
+
+	if (adapter->pf == 4) {
+		int i;
+
+		/* Tear down per-adapter Work Queue first since it can contain
+		 * references to our adapter data structure.
+		 */
+		destroy_workqueue(adapter->workq);
+
+		detach_ulds(adapter);
+
+		for_each_port(adapter, i)
+			if (adapter->port[i]->reg_state == NETREG_REGISTERED)
+				unregister_netdev(adapter->port[i]);
+
+		t4_uld_clean_up(adapter);
+
+		adap_free_hma_mem(adapter);
+
+		disable_interrupts(adapter);
+
+		cxgb4_free_mps_ref_entries(adapter);
+
+		debugfs_remove_recursive(adapter->debugfs_root);
+
+		if (!is_t4(adapter->params.chip))
+			cxgb4_ptp_stop(adapter);
+		if (IS_REACHABLE(CONFIG_THERMAL))
+			cxgb4_thermal_remove(adapter);
+
+		if (adapter->flags & CXGB4_FULL_INIT_DONE)
+			cxgb_down(adapter);
+
+		if (adapter->flags & CXGB4_USING_MSIX)
+			free_msix_info(adapter);
+		if (adapter->num_uld || adapter->num_ofld_uld)
+			t4_uld_mem_free(adapter);
+		free_some_resources(adapter);
+		list_for_each_entry_safe(entry, tmp, &adapter->mac_hlist,
+					 list) {
+			list_del(&entry->list);
+			kfree(entry);
+		}
+
+#if IS_ENABLED(CONFIG_IPV6)
+		t4_cleanup_clip_tbl(adapter);
+#endif
+		if (!is_t4(adapter->params.chip))
+			iounmap(adapter->bar2);
+	}
+#ifdef CONFIG_PCI_IOV
+	else {
+		cxgb4_iov_configure(adapter->pdev, 0);
+	}
+#endif
+	iounmap(adapter->regs);
+	pci_disable_pcie_error_reporting(pdev);
+	if ((adapter->flags & CXGB4_DEV_ENABLED)) {
+		pci_disable_device(pdev);
+		adapter->flags &= ~CXGB4_DEV_ENABLED;
+	}
+	pci_release_regions(pdev);
+	kfree(adapter->mbox_log);
+	synchronize_rcu();
+	kfree(adapter);
+}
+
+/* "Shutdown" quiesces the device, stopping Ingress Packet and Interrupt
+ * delivery.  This is essentially a stripped down version of the PCI remove()
+ * function where we do the minimal amount of work necessary to shutdown any
+ * further activity.
+ */
+static void shutdown_one(struct pci_dev *pdev)
+{
+	struct adapter *adapter = pci_get_drvdata(pdev);
+
+	/* As with remove_one() above (see extended comment), we only want do
+	 * do cleanup on PCI Devices which went all the way through init_one()
+	 * ...
+	 */
+	if (!adapter) {
+		pci_release_regions(pdev);
+		return;
+	}
+
+	adapter->flags |= CXGB4_SHUTTING_DOWN;
+
+	if (adapter->pf == 4) {
+		int i;
+
+		for_each_port(adapter, i)
+			if (adapter->port[i]->reg_state == NETREG_REGISTERED)
+				cxgb_close(adapter->port[i]);
+
+		rtnl_lock();
+		cxgb4_mqprio_stop_offload(adapter);
+		rtnl_unlock();
+
+		if (is_uld(adapter)) {
+			detach_ulds(adapter);
+			t4_uld_clean_up(adapter);
+		}
+
+		disable_interrupts(adapter);
+		disable_msi(adapter);
+
+		t4_sge_stop(adapter);
+		if (adapter->flags & CXGB4_FW_OK)
+			t4_fw_bye(adapter, adapter->mbox);
+	}
+}
+
+static struct pci_driver cxgb4_driver = {
+	.name     = KBUILD_MODNAME,
+	.id_table = cxgb4_pci_tbl,
+	.probe    = init_one,
+	.remove   = remove_one,
+	.shutdown = shutdown_one,
+#ifdef CONFIG_PCI_IOV
+	.sriov_configure = cxgb4_iov_configure,
+#endif
+	.err_handler = &cxgb4_eeh,
+};
+
+static int __init cxgb4_init_module(void)
+{
+	int ret;
+
+	cxgb4_debugfs_root = debugfs_create_dir(KBUILD_MODNAME, NULL);
+
+	ret = pci_register_driver(&cxgb4_driver);
+	if (ret < 0)
+		goto err_pci;
+
+#if IS_ENABLED(CONFIG_IPV6)
+	if (!inet6addr_registered) {
+		ret = register_inet6addr_notifier(&cxgb4_inet6addr_notifier);
+		if (ret)
+			pci_unregister_driver(&cxgb4_driver);
+		else
+			inet6addr_registered = true;
+	}
+#endif
+
+	if (ret == 0)
+		return ret;
+
+err_pci:
+	debugfs_remove(cxgb4_debugfs_root);
+
+	return ret;
+}
+
+static void __exit cxgb4_cleanup_module(void)
+{
+#if IS_ENABLED(CONFIG_IPV6)
+	if (inet6addr_registered) {
+		unregister_inet6addr_notifier(&cxgb4_inet6addr_notifier);
+		inet6addr_registered = false;
+	}
+#endif
+	pci_unregister_driver(&cxgb4_driver);
+	debugfs_remove(cxgb4_debugfs_root);  /* NULL ok */
+}
+
+module_init(cxgb4_init_module);
+module_exit(cxgb4_cleanup_module);
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_mps.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_mps.c
new file mode 100644
index 000000000..a020e8490
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_mps.c
@@ -0,0 +1,241 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2019 Chelsio Communications, Inc. All rights reserved. */
+
+#include "cxgb4.h"
+
+static int cxgb4_mps_ref_dec_by_mac(struct adapter *adap,
+				    const u8 *addr, const u8 *mask)
+{
+	u8 bitmask[] = { 0xff, 0xff, 0xff, 0xff, 0xff, 0xff };
+	struct mps_entries_ref *mps_entry, *tmp;
+	int ret = -EINVAL;
+
+	spin_lock_bh(&adap->mps_ref_lock);
+	list_for_each_entry_safe(mps_entry, tmp, &adap->mps_ref, list) {
+		if (ether_addr_equal(mps_entry->addr, addr) &&
+		    ether_addr_equal(mps_entry->mask, mask ? mask : bitmask)) {
+			if (!refcount_dec_and_test(&mps_entry->refcnt)) {
+				spin_unlock_bh(&adap->mps_ref_lock);
+				return -EBUSY;
+			}
+			list_del(&mps_entry->list);
+			kfree(mps_entry);
+			ret = 0;
+			break;
+		}
+	}
+	spin_unlock_bh(&adap->mps_ref_lock);
+	return ret;
+}
+
+static int cxgb4_mps_ref_dec(struct adapter *adap, u16 idx)
+{
+	struct mps_entries_ref *mps_entry, *tmp;
+	int ret = -EINVAL;
+
+	spin_lock(&adap->mps_ref_lock);
+	list_for_each_entry_safe(mps_entry, tmp, &adap->mps_ref, list) {
+		if (mps_entry->idx == idx) {
+			if (!refcount_dec_and_test(&mps_entry->refcnt)) {
+				spin_unlock(&adap->mps_ref_lock);
+				return -EBUSY;
+			}
+			list_del(&mps_entry->list);
+			kfree(mps_entry);
+			ret = 0;
+			break;
+		}
+	}
+	spin_unlock(&adap->mps_ref_lock);
+	return ret;
+}
+
+static int cxgb4_mps_ref_inc(struct adapter *adap, const u8 *mac_addr,
+			     u16 idx, const u8 *mask)
+{
+	u8 bitmask[] = { 0xff, 0xff, 0xff, 0xff, 0xff, 0xff };
+	struct mps_entries_ref *mps_entry;
+	int ret = 0;
+
+	spin_lock_bh(&adap->mps_ref_lock);
+	list_for_each_entry(mps_entry, &adap->mps_ref, list) {
+		if (mps_entry->idx == idx) {
+			refcount_inc(&mps_entry->refcnt);
+			goto unlock;
+		}
+	}
+	mps_entry = kzalloc(sizeof(*mps_entry), GFP_ATOMIC);
+	if (!mps_entry) {
+		ret = -ENOMEM;
+		goto unlock;
+	}
+	ether_addr_copy(mps_entry->mask, mask ? mask : bitmask);
+	ether_addr_copy(mps_entry->addr, mac_addr);
+	mps_entry->idx = idx;
+	refcount_set(&mps_entry->refcnt, 1);
+	list_add_tail(&mps_entry->list, &adap->mps_ref);
+unlock:
+	spin_unlock_bh(&adap->mps_ref_lock);
+	return ret;
+}
+
+int cxgb4_free_mac_filt(struct adapter *adap, unsigned int viid,
+			unsigned int naddr, const u8 **addr, bool sleep_ok)
+{
+	int ret, i;
+
+	for (i = 0; i < naddr; i++) {
+		if (!cxgb4_mps_ref_dec_by_mac(adap, addr[i], NULL)) {
+			ret = t4_free_mac_filt(adap, adap->mbox, viid,
+					       1, &addr[i], sleep_ok);
+			if (ret < 0)
+				return ret;
+		}
+	}
+
+	/* return number of filters freed */
+	return naddr;
+}
+
+int cxgb4_alloc_mac_filt(struct adapter *adap, unsigned int viid,
+			 bool free, unsigned int naddr, const u8 **addr,
+			 u16 *idx, u64 *hash, bool sleep_ok)
+{
+	int ret, i;
+
+	ret = t4_alloc_mac_filt(adap, adap->mbox, viid, free,
+				naddr, addr, idx, hash, sleep_ok);
+	if (ret < 0)
+		return ret;
+
+	for (i = 0; i < naddr; i++) {
+		if (idx[i] != 0xffff) {
+			if (cxgb4_mps_ref_inc(adap, addr[i], idx[i], NULL)) {
+				ret = -ENOMEM;
+				goto error;
+			}
+		}
+	}
+
+	goto out;
+error:
+	cxgb4_free_mac_filt(adap, viid, naddr, addr, sleep_ok);
+
+out:
+	/* Returns a negative error number or the number of filters allocated */
+	return ret;
+}
+
+int cxgb4_update_mac_filt(struct port_info *pi, unsigned int viid,
+			  int *tcam_idx, const u8 *addr,
+			  bool persistent, u8 *smt_idx)
+{
+	int ret;
+
+	ret = cxgb4_change_mac(pi, viid, tcam_idx,
+			       addr, persistent, smt_idx);
+	if (ret < 0)
+		return ret;
+
+	cxgb4_mps_ref_inc(pi->adapter, addr, *tcam_idx, NULL);
+	return ret;
+}
+
+int cxgb4_free_raw_mac_filt(struct adapter *adap,
+			    unsigned int viid,
+			    const u8 *addr,
+			    const u8 *mask,
+			    unsigned int idx,
+			    u8 lookup_type,
+			    u8 port_id,
+			    bool sleep_ok)
+{
+	int ret = 0;
+
+	if (!cxgb4_mps_ref_dec(adap, idx))
+		ret = t4_free_raw_mac_filt(adap, viid, addr,
+					   mask, idx, lookup_type,
+					   port_id, sleep_ok);
+
+	return ret;
+}
+
+int cxgb4_alloc_raw_mac_filt(struct adapter *adap,
+			     unsigned int viid,
+			     const u8 *addr,
+			     const u8 *mask,
+			     unsigned int idx,
+			     u8 lookup_type,
+			     u8 port_id,
+			     bool sleep_ok)
+{
+	int ret;
+
+	ret = t4_alloc_raw_mac_filt(adap, viid, addr,
+				    mask, idx, lookup_type,
+				    port_id, sleep_ok);
+	if (ret < 0)
+		return ret;
+
+	if (cxgb4_mps_ref_inc(adap, addr, ret, mask)) {
+		ret = -ENOMEM;
+		t4_free_raw_mac_filt(adap, viid, addr,
+				     mask, idx, lookup_type,
+				     port_id, sleep_ok);
+	}
+
+	return ret;
+}
+
+int cxgb4_free_encap_mac_filt(struct adapter *adap, unsigned int viid,
+			      int idx, bool sleep_ok)
+{
+	int ret = 0;
+
+	if (!cxgb4_mps_ref_dec(adap, idx))
+		ret = t4_free_encap_mac_filt(adap, viid, idx, sleep_ok);
+
+	return ret;
+}
+
+int cxgb4_alloc_encap_mac_filt(struct adapter *adap, unsigned int viid,
+			       const u8 *addr, const u8 *mask,
+			       unsigned int vni, unsigned int vni_mask,
+			       u8 dip_hit, u8 lookup_type, bool sleep_ok)
+{
+	int ret;
+
+	ret = t4_alloc_encap_mac_filt(adap, viid, addr, mask, vni, vni_mask,
+				      dip_hit, lookup_type, sleep_ok);
+	if (ret < 0)
+		return ret;
+
+	if (cxgb4_mps_ref_inc(adap, addr, ret, mask)) {
+		ret = -ENOMEM;
+		t4_free_encap_mac_filt(adap, viid, ret, sleep_ok);
+	}
+	return ret;
+}
+
+int cxgb4_init_mps_ref_entries(struct adapter *adap)
+{
+	spin_lock_init(&adap->mps_ref_lock);
+	INIT_LIST_HEAD(&adap->mps_ref);
+
+	return 0;
+}
+
+void cxgb4_free_mps_ref_entries(struct adapter *adap)
+{
+	struct mps_entries_ref *mps_entry, *tmp;
+
+	if (list_empty(&adap->mps_ref))
+		return;
+
+	spin_lock(&adap->mps_ref_lock);
+	list_for_each_entry_safe(mps_entry, tmp, &adap->mps_ref, list) {
+		list_del(&mps_entry->list);
+		kfree(mps_entry);
+	}
+	spin_unlock(&adap->mps_ref_lock);
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ptp.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ptp.c
new file mode 100644
index 000000000..5bf117d21
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ptp.c
@@ -0,0 +1,465 @@
+/*
+ * cxgb4_ptp.c:Chelsio PTP support for T5/T6
+ *
+ * Copyright (c) 2003-2017 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ * Written by: Atul Gupta (atul.gupta@chelsio.com)
+ */
+
+#include <linux/module.h>
+#include <linux/net_tstamp.h>
+#include <linux/skbuff.h>
+#include <linux/netdevice.h>
+#include <linux/pps_kernel.h>
+#include <linux/ptp_clock_kernel.h>
+#include <linux/ptp_classify.h>
+#include <linux/udp.h>
+
+#include "cxgb4.h"
+#include "t4_hw.h"
+#include "t4_regs.h"
+#include "t4_msg.h"
+#include "t4fw_api.h"
+#include "cxgb4_ptp.h"
+
+/**
+ * cxgb4_ptp_is_ptp_tx - determine whether TX packet is PTP or not
+ * @skb: skb of outgoing ptp request
+ *
+ */
+bool cxgb4_ptp_is_ptp_tx(struct sk_buff *skb)
+{
+	struct udphdr *uh;
+
+	uh = udp_hdr(skb);
+	return skb->len >= PTP_MIN_LENGTH &&
+		skb->len <= PTP_IN_TRANSMIT_PACKET_MAXNUM &&
+		likely(skb->protocol == htons(ETH_P_IP)) &&
+		ip_hdr(skb)->protocol == IPPROTO_UDP &&
+		uh->dest == htons(PTP_EVENT_PORT);
+}
+
+bool is_ptp_enabled(struct sk_buff *skb, struct net_device *dev)
+{
+	struct port_info *pi;
+
+	pi = netdev_priv(dev);
+	return (pi->ptp_enable && cxgb4_xmit_with_hwtstamp(skb) &&
+		cxgb4_ptp_is_ptp_tx(skb));
+}
+
+/**
+ * cxgb4_ptp_is_ptp_rx - determine whether RX packet is PTP or not
+ * @skb: skb of incoming ptp request
+ *
+ */
+bool cxgb4_ptp_is_ptp_rx(struct sk_buff *skb)
+{
+	struct udphdr *uh = (struct udphdr *)(skb->data + ETH_HLEN +
+					      IPV4_HLEN(skb->data));
+
+	return  uh->dest == htons(PTP_EVENT_PORT) &&
+		uh->source == htons(PTP_EVENT_PORT);
+}
+
+/**
+ * cxgb4_ptp_read_hwstamp - read timestamp for TX event PTP message
+ * @adapter: board private structure
+ * @pi: port private structure
+ *
+ */
+void cxgb4_ptp_read_hwstamp(struct adapter *adapter, struct port_info *pi)
+{
+	struct skb_shared_hwtstamps *skb_ts = NULL;
+	u64 tx_ts;
+
+	skb_ts = skb_hwtstamps(adapter->ptp_tx_skb);
+
+	tx_ts = t4_read_reg(adapter,
+			    T5_PORT_REG(pi->port_id, MAC_PORT_TX_TS_VAL_LO));
+
+	tx_ts |= (u64)t4_read_reg(adapter,
+				  T5_PORT_REG(pi->port_id,
+					      MAC_PORT_TX_TS_VAL_HI)) << 32;
+	skb_ts->hwtstamp = ns_to_ktime(tx_ts);
+	skb_tstamp_tx(adapter->ptp_tx_skb, skb_ts);
+	dev_kfree_skb_any(adapter->ptp_tx_skb);
+	spin_lock(&adapter->ptp_lock);
+	adapter->ptp_tx_skb = NULL;
+	spin_unlock(&adapter->ptp_lock);
+}
+
+/**
+ * cxgb4_ptprx_timestamping - Enable Timestamp for RX PTP event message
+ * @pi: port private structure
+ * @port: pot number
+ * @mode: RX mode
+ *
+ */
+int cxgb4_ptprx_timestamping(struct port_info *pi, u8 port, u16 mode)
+{
+	struct adapter *adapter = pi->adapter;
+	struct fw_ptp_cmd c;
+	int err;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PTP_CMD) |
+				     FW_CMD_REQUEST_F |
+				     FW_CMD_WRITE_F |
+				     FW_PTP_CMD_PORTID_V(port));
+	c.retval_len16 = cpu_to_be32(FW_CMD_LEN16_V(sizeof(c) / 16));
+	c.u.init.sc = FW_PTP_SC_RXTIME_STAMP;
+	c.u.init.mode = cpu_to_be16(mode);
+
+	err = t4_wr_mbox(adapter, adapter->mbox, &c, sizeof(c), NULL);
+	if (err < 0)
+		dev_err(adapter->pdev_dev,
+			"PTP: %s error %d\n", __func__, -err);
+	return err;
+}
+
+int cxgb4_ptp_txtype(struct adapter *adapter, u8 port)
+{
+	struct fw_ptp_cmd c;
+	int err;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PTP_CMD) |
+				     FW_CMD_REQUEST_F |
+				     FW_CMD_WRITE_F |
+				     FW_PTP_CMD_PORTID_V(port));
+	c.retval_len16 = cpu_to_be32(FW_CMD_LEN16_V(sizeof(c) / 16));
+	c.u.init.sc = FW_PTP_SC_TX_TYPE;
+	c.u.init.mode = cpu_to_be16(PTP_TS_NONE);
+
+	err = t4_wr_mbox(adapter, adapter->mbox, &c, sizeof(c), NULL);
+	if (err < 0)
+		dev_err(adapter->pdev_dev,
+			"PTP: %s error %d\n", __func__, -err);
+
+	return err;
+}
+
+int cxgb4_ptp_redirect_rx_packet(struct adapter *adapter, struct port_info *pi)
+{
+	struct sge *s = &adapter->sge;
+	struct sge_eth_rxq *receive_q =  &s->ethrxq[pi->first_qset];
+	struct fw_ptp_cmd c;
+	int err;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PTP_CMD) |
+				     FW_CMD_REQUEST_F |
+				     FW_CMD_WRITE_F |
+				     FW_PTP_CMD_PORTID_V(pi->port_id));
+
+	c.retval_len16 = cpu_to_be32(FW_CMD_LEN16_V(sizeof(c) / 16));
+	c.u.init.sc = FW_PTP_SC_RDRX_TYPE;
+	c.u.init.txchan = pi->tx_chan;
+	c.u.init.absid = cpu_to_be16(receive_q->rspq.abs_id);
+
+	err = t4_wr_mbox(adapter, adapter->mbox, &c, sizeof(c), NULL);
+	if (err < 0)
+		dev_err(adapter->pdev_dev,
+			"PTP: %s error %d\n", __func__, -err);
+	return err;
+}
+
+/**
+ * cxgb4_ptp_adjfreq - Adjust frequency of PHC cycle counter
+ * @ptp: ptp clock structure
+ * @ppb: Desired frequency change in parts per billion
+ *
+ * Adjust the frequency of the PHC cycle counter by the indicated ppb from
+ * the base frequency.
+ */
+static int cxgb4_ptp_adjfreq(struct ptp_clock_info *ptp, s32 ppb)
+{
+	struct adapter *adapter = (struct adapter *)container_of(ptp,
+				   struct adapter, ptp_clock_info);
+	struct fw_ptp_cmd c;
+	int err;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PTP_CMD) |
+				     FW_CMD_REQUEST_F |
+				     FW_CMD_WRITE_F |
+				     FW_PTP_CMD_PORTID_V(0));
+	c.retval_len16 = cpu_to_be32(FW_CMD_LEN16_V(sizeof(c) / 16));
+	c.u.ts.sc = FW_PTP_SC_ADJ_FREQ;
+	c.u.ts.sign = (ppb < 0) ? 1 : 0;
+	if (ppb < 0)
+		ppb = -ppb;
+	c.u.ts.ppb = cpu_to_be32(ppb);
+
+	err = t4_wr_mbox(adapter, adapter->mbox, &c, sizeof(c), NULL);
+	if (err < 0)
+		dev_err(adapter->pdev_dev,
+			"PTP: %s error %d\n", __func__, -err);
+
+	return err;
+}
+
+/**
+ * cxgb4_ptp_fineadjtime - Shift the time of the hardware clock
+ * @adapter: board private structure
+ * @delta: Desired change in nanoseconds
+ *
+ * Adjust the timer by resetting the timecounter structure.
+ */
+static int  cxgb4_ptp_fineadjtime(struct adapter *adapter, s64 delta)
+{
+	struct fw_ptp_cmd c;
+	int err;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PTP_CMD) |
+			     FW_CMD_REQUEST_F |
+			     FW_CMD_WRITE_F |
+			     FW_PTP_CMD_PORTID_V(0));
+	c.retval_len16 = cpu_to_be32(FW_CMD_LEN16_V(sizeof(c) / 16));
+	c.u.ts.sc = FW_PTP_SC_ADJ_FTIME;
+	c.u.ts.sign = (delta < 0) ? 1 : 0;
+	if (delta < 0)
+		delta = -delta;
+	c.u.ts.tm = cpu_to_be64(delta);
+
+	err = t4_wr_mbox(adapter, adapter->mbox, &c, sizeof(c), NULL);
+	if (err < 0)
+		dev_err(adapter->pdev_dev,
+			"PTP: %s error %d\n", __func__, -err);
+	return err;
+}
+
+/**
+ * cxgb4_ptp_adjtime - Shift the time of the hardware clock
+ * @ptp: ptp clock structure
+ * @delta: Desired change in nanoseconds
+ *
+ * Adjust the timer by resetting the timecounter structure.
+ */
+static int cxgb4_ptp_adjtime(struct ptp_clock_info *ptp, s64 delta)
+{
+	struct adapter *adapter =
+		(struct adapter *)container_of(ptp, struct adapter,
+					       ptp_clock_info);
+	struct fw_ptp_cmd c;
+	s64 sign = 1;
+	int err;
+
+	if (delta < 0)
+		sign = -1;
+
+	if (delta * sign > PTP_CLOCK_MAX_ADJTIME) {
+		memset(&c, 0, sizeof(c));
+		c.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PTP_CMD) |
+					     FW_CMD_REQUEST_F |
+					     FW_CMD_WRITE_F |
+					     FW_PTP_CMD_PORTID_V(0));
+		c.retval_len16 = cpu_to_be32(FW_CMD_LEN16_V(sizeof(c) / 16));
+		c.u.ts.sc = FW_PTP_SC_ADJ_TIME;
+		c.u.ts.sign = (delta < 0) ? 1 : 0;
+		if (delta < 0)
+			delta = -delta;
+		c.u.ts.tm = cpu_to_be64(delta);
+
+		err = t4_wr_mbox(adapter, adapter->mbox, &c, sizeof(c), NULL);
+		if (err < 0)
+			dev_err(adapter->pdev_dev,
+				"PTP: %s error %d\n", __func__, -err);
+	} else {
+		err = cxgb4_ptp_fineadjtime(adapter, delta);
+	}
+
+	return err;
+}
+
+/**
+ * cxgb4_ptp_gettime - Reads the current time from the hardware clock
+ * @ptp: ptp clock structure
+ * @ts: timespec structure to hold the current time value
+ *
+ * Read the timecounter and return the correct value in ns after converting
+ * it into a struct timespec.
+ */
+static int cxgb4_ptp_gettime(struct ptp_clock_info *ptp, struct timespec64 *ts)
+{
+	struct adapter *adapter = container_of(ptp, struct adapter,
+					       ptp_clock_info);
+	u64 ns;
+
+	ns = t4_read_reg(adapter, T5_PORT_REG(0, MAC_PORT_PTP_SUM_LO_A));
+	ns |= (u64)t4_read_reg(adapter,
+			       T5_PORT_REG(0, MAC_PORT_PTP_SUM_HI_A)) << 32;
+
+	/* convert to timespec*/
+	*ts = ns_to_timespec64(ns);
+	return 0;
+}
+
+/**
+ *  cxgb4_ptp_settime - Set the current time on the hardware clock
+ *  @ptp: ptp clock structure
+ *  @ts: timespec containing the new time for the cycle counter
+ *
+ *  Reset value to new base value instead of the kernel
+ *  wall timer value.
+ */
+static int cxgb4_ptp_settime(struct ptp_clock_info *ptp,
+			     const struct timespec64 *ts)
+{
+	struct adapter *adapter = (struct adapter *)container_of(ptp,
+				   struct adapter, ptp_clock_info);
+	struct fw_ptp_cmd c;
+	u64 ns;
+	int err;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PTP_CMD) |
+				     FW_CMD_REQUEST_F |
+				     FW_CMD_WRITE_F |
+				     FW_PTP_CMD_PORTID_V(0));
+	c.retval_len16 = cpu_to_be32(FW_CMD_LEN16_V(sizeof(c) / 16));
+	c.u.ts.sc = FW_PTP_SC_SET_TIME;
+
+	ns = timespec64_to_ns(ts);
+	c.u.ts.tm = cpu_to_be64(ns);
+
+	err =  t4_wr_mbox(adapter, adapter->mbox, &c, sizeof(c), NULL);
+	if (err < 0)
+		dev_err(adapter->pdev_dev,
+			"PTP: %s error %d\n", __func__, -err);
+
+	return err;
+}
+
+static void cxgb4_init_ptp_timer(struct adapter *adapter)
+{
+	struct fw_ptp_cmd c;
+	int err;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PTP_CMD) |
+				     FW_CMD_REQUEST_F |
+				     FW_CMD_WRITE_F |
+				     FW_PTP_CMD_PORTID_V(0));
+	c.retval_len16 = cpu_to_be32(FW_CMD_LEN16_V(sizeof(c) / 16));
+	c.u.scmd.sc = FW_PTP_SC_INIT_TIMER;
+
+	err = t4_wr_mbox(adapter, adapter->mbox, &c, sizeof(c), NULL);
+	if (err < 0)
+		dev_err(adapter->pdev_dev,
+			"PTP: %s error %d\n", __func__, -err);
+}
+
+/**
+ * cxgb4_ptp_enable - enable or disable an ancillary feature
+ * @ptp: ptp clock structure
+ * @request: Desired resource to enable or disable
+ * @on: Caller passes one to enable or zero to disable
+ *
+ * Enable (or disable) ancillary features of the PHC subsystem.
+ * Currently, no ancillary features are supported.
+ */
+static int cxgb4_ptp_enable(struct ptp_clock_info __always_unused *ptp,
+			    struct ptp_clock_request __always_unused *request,
+			    int __always_unused on)
+{
+	return -ENOTSUPP;
+}
+
+static const struct ptp_clock_info cxgb4_ptp_clock_info = {
+	.owner          = THIS_MODULE,
+	.name           = "cxgb4_clock",
+	.max_adj        = MAX_PTP_FREQ_ADJ,
+	.n_alarm        = 0,
+	.n_ext_ts       = 0,
+	.n_per_out      = 0,
+	.pps            = 0,
+	.adjfreq        = cxgb4_ptp_adjfreq,
+	.adjtime        = cxgb4_ptp_adjtime,
+	.gettime64      = cxgb4_ptp_gettime,
+	.settime64      = cxgb4_ptp_settime,
+	.enable         = cxgb4_ptp_enable,
+};
+
+/**
+ * cxgb4_ptp_init - initialize PTP for devices which support it
+ * @adapter: board private structure
+ *
+ * This function performs the required steps for enabling PTP support.
+ */
+void cxgb4_ptp_init(struct adapter *adapter)
+{
+	struct timespec64 now;
+	 /* no need to create a clock device if we already have one */
+	if (!IS_ERR_OR_NULL(adapter->ptp_clock))
+		return;
+
+	adapter->ptp_tx_skb = NULL;
+	adapter->ptp_clock_info = cxgb4_ptp_clock_info;
+	spin_lock_init(&adapter->ptp_lock);
+
+	adapter->ptp_clock = ptp_clock_register(&adapter->ptp_clock_info,
+						&adapter->pdev->dev);
+	if (IS_ERR_OR_NULL(adapter->ptp_clock)) {
+		adapter->ptp_clock = NULL;
+		dev_err(adapter->pdev_dev,
+			"PTP %s Clock registration has failed\n", __func__);
+		return;
+	}
+
+	now = ktime_to_timespec64(ktime_get_real());
+	cxgb4_init_ptp_timer(adapter);
+	if (cxgb4_ptp_settime(&adapter->ptp_clock_info, &now) < 0) {
+		ptp_clock_unregister(adapter->ptp_clock);
+		adapter->ptp_clock = NULL;
+	}
+}
+
+/**
+ * cxgb4_ptp_stop - disable PTP device and stop the overflow check
+ * @adapter: board private structure
+ *
+ * Stop the PTP support.
+ */
+void cxgb4_ptp_stop(struct adapter *adapter)
+{
+	if (adapter->ptp_tx_skb) {
+		dev_kfree_skb_any(adapter->ptp_tx_skb);
+		adapter->ptp_tx_skb = NULL;
+	}
+
+	if (adapter->ptp_clock) {
+		ptp_clock_unregister(adapter->ptp_clock);
+		adapter->ptp_clock = NULL;
+	}
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ptp.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ptp.h
new file mode 100644
index 000000000..cccfae84b
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ptp.h
@@ -0,0 +1,74 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2017 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_PTP_H__
+#define __CXGB4_PTP_H__
+
+/* Maximum parts-per-billion adjustment that is acceptable */
+#define MAX_PTP_FREQ_ADJ		1000000
+#define PTP_CLOCK_MAX_ADJTIME		10000000 /* 10 ms */
+
+#define PTP_MIN_LENGTH			63
+#define PTP_IN_TRANSMIT_PACKET_MAXNUM	240
+#define PTP_EVENT_PORT			319
+
+enum ptp_rx_filter_mode {
+	PTP_TS_NONE = 0,
+	PTP_TS_L2,
+	PTP_TS_L4,
+	PTP_TS_L2_L4
+};
+
+struct port_info;
+
+static inline bool cxgb4_xmit_with_hwtstamp(struct sk_buff *skb)
+{
+	return skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP;
+}
+
+static inline void cxgb4_xmit_hwtstamp_pending(struct sk_buff *skb)
+{
+	skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
+}
+
+void cxgb4_ptp_init(struct adapter *adap);
+void cxgb4_ptp_stop(struct adapter *adap);
+bool cxgb4_ptp_is_ptp_tx(struct sk_buff *skb);
+bool cxgb4_ptp_is_ptp_rx(struct sk_buff *skb);
+int cxgb4_ptprx_timestamping(struct port_info *pi, u8 port, u16 mode);
+int cxgb4_ptp_redirect_rx_packet(struct adapter *adap, struct port_info *pi);
+int cxgb4_ptp_txtype(struct adapter *adap, u8 port_id);
+void cxgb4_ptp_read_hwstamp(struct adapter *adap, struct port_info *pi);
+bool is_ptp_enabled(struct sk_buff *skb, struct net_device *dev);
+#endif /* __CXGB4_PTP_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
new file mode 100644
index 000000000..dd9be2298
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
@@ -0,0 +1,1142 @@
+/*
+ * This file is part of the Chelsio T4/T5/T6 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2017 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <net/tc_act/tc_mirred.h>
+#include <net/tc_act/tc_pedit.h>
+#include <net/tc_act/tc_gact.h>
+#include <net/tc_act/tc_vlan.h>
+
+#include "cxgb4.h"
+#include "cxgb4_filter.h"
+#include "cxgb4_tc_flower.h"
+
+#define STATS_CHECK_PERIOD (HZ / 2)
+
+static struct ch_tc_pedit_fields pedits[] = {
+	PEDIT_FIELDS(ETH_, DMAC_31_0, 4, dmac, 0),
+	PEDIT_FIELDS(ETH_, DMAC_47_32, 2, dmac, 4),
+	PEDIT_FIELDS(ETH_, SMAC_15_0, 2, smac, 0),
+	PEDIT_FIELDS(ETH_, SMAC_47_16, 4, smac, 2),
+	PEDIT_FIELDS(IP4_, SRC, 4, nat_fip, 0),
+	PEDIT_FIELDS(IP4_, DST, 4, nat_lip, 0),
+	PEDIT_FIELDS(IP6_, SRC_31_0, 4, nat_fip, 0),
+	PEDIT_FIELDS(IP6_, SRC_63_32, 4, nat_fip, 4),
+	PEDIT_FIELDS(IP6_, SRC_95_64, 4, nat_fip, 8),
+	PEDIT_FIELDS(IP6_, SRC_127_96, 4, nat_fip, 12),
+	PEDIT_FIELDS(IP6_, DST_31_0, 4, nat_lip, 0),
+	PEDIT_FIELDS(IP6_, DST_63_32, 4, nat_lip, 4),
+	PEDIT_FIELDS(IP6_, DST_95_64, 4, nat_lip, 8),
+	PEDIT_FIELDS(IP6_, DST_127_96, 4, nat_lip, 12),
+};
+
+static const struct cxgb4_natmode_config cxgb4_natmode_config_array[] = {
+	/* Default supported NAT modes */
+	{
+		.chip = CHELSIO_T5,
+		.flags = CXGB4_ACTION_NATMODE_NONE,
+		.natmode = NAT_MODE_NONE,
+	},
+	{
+		.chip = CHELSIO_T5,
+		.flags = CXGB4_ACTION_NATMODE_DIP,
+		.natmode = NAT_MODE_DIP,
+	},
+	{
+		.chip = CHELSIO_T5,
+		.flags = CXGB4_ACTION_NATMODE_DIP | CXGB4_ACTION_NATMODE_DPORT,
+		.natmode = NAT_MODE_DIP_DP,
+	},
+	{
+		.chip = CHELSIO_T5,
+		.flags = CXGB4_ACTION_NATMODE_DIP | CXGB4_ACTION_NATMODE_DPORT |
+			 CXGB4_ACTION_NATMODE_SIP,
+		.natmode = NAT_MODE_DIP_DP_SIP,
+	},
+	{
+		.chip = CHELSIO_T5,
+		.flags = CXGB4_ACTION_NATMODE_DIP | CXGB4_ACTION_NATMODE_DPORT |
+			 CXGB4_ACTION_NATMODE_SPORT,
+		.natmode = NAT_MODE_DIP_DP_SP,
+	},
+	{
+		.chip = CHELSIO_T5,
+		.flags = CXGB4_ACTION_NATMODE_SIP | CXGB4_ACTION_NATMODE_SPORT,
+		.natmode = NAT_MODE_SIP_SP,
+	},
+	{
+		.chip = CHELSIO_T5,
+		.flags = CXGB4_ACTION_NATMODE_DIP | CXGB4_ACTION_NATMODE_SIP |
+			 CXGB4_ACTION_NATMODE_SPORT,
+		.natmode = NAT_MODE_DIP_SIP_SP,
+	},
+	{
+		.chip = CHELSIO_T5,
+		.flags = CXGB4_ACTION_NATMODE_DIP | CXGB4_ACTION_NATMODE_SIP |
+			 CXGB4_ACTION_NATMODE_DPORT |
+			 CXGB4_ACTION_NATMODE_SPORT,
+		.natmode = NAT_MODE_ALL,
+	},
+	/* T6+ can ignore L4 ports when they're disabled. */
+	{
+		.chip = CHELSIO_T6,
+		.flags = CXGB4_ACTION_NATMODE_SIP,
+		.natmode = NAT_MODE_SIP_SP,
+	},
+	{
+		.chip = CHELSIO_T6,
+		.flags = CXGB4_ACTION_NATMODE_DIP | CXGB4_ACTION_NATMODE_SPORT,
+		.natmode = NAT_MODE_DIP_DP_SP,
+	},
+	{
+		.chip = CHELSIO_T6,
+		.flags = CXGB4_ACTION_NATMODE_DIP | CXGB4_ACTION_NATMODE_SIP,
+		.natmode = NAT_MODE_ALL,
+	},
+};
+
+static void cxgb4_action_natmode_tweak(struct ch_filter_specification *fs,
+				       u8 natmode_flags)
+{
+	u8 i = 0;
+
+	/* Translate the enabled NAT 4-tuple fields to one of the
+	 * hardware supported NAT mode configurations. This ensures
+	 * that we pick a valid combination, where the disabled fields
+	 * do not get overwritten to 0.
+	 */
+	for (i = 0; i < ARRAY_SIZE(cxgb4_natmode_config_array); i++) {
+		if (cxgb4_natmode_config_array[i].flags == natmode_flags) {
+			fs->nat_mode = cxgb4_natmode_config_array[i].natmode;
+			return;
+		}
+	}
+}
+
+static struct ch_tc_flower_entry *allocate_flower_entry(void)
+{
+	struct ch_tc_flower_entry *new = kzalloc(sizeof(*new), GFP_KERNEL);
+	if (new)
+		spin_lock_init(&new->lock);
+	return new;
+}
+
+/* Must be called with either RTNL or rcu_read_lock */
+static struct ch_tc_flower_entry *ch_flower_lookup(struct adapter *adap,
+						   unsigned long flower_cookie)
+{
+	return rhashtable_lookup_fast(&adap->flower_tbl, &flower_cookie,
+				      adap->flower_ht_params);
+}
+
+static void cxgb4_process_flow_match(struct net_device *dev,
+				     struct flow_rule *rule,
+				     struct ch_filter_specification *fs)
+{
+	u16 addr_type = 0;
+
+	if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_CONTROL)) {
+		struct flow_match_control match;
+
+		flow_rule_match_control(rule, &match);
+		addr_type = match.key->addr_type;
+	} else if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_IPV4_ADDRS)) {
+		addr_type = FLOW_DISSECTOR_KEY_IPV4_ADDRS;
+	} else if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_IPV6_ADDRS)) {
+		addr_type = FLOW_DISSECTOR_KEY_IPV6_ADDRS;
+	}
+
+	if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_BASIC)) {
+		struct flow_match_basic match;
+		u16 ethtype_key, ethtype_mask;
+
+		flow_rule_match_basic(rule, &match);
+		ethtype_key = ntohs(match.key->n_proto);
+		ethtype_mask = ntohs(match.mask->n_proto);
+
+		if (ethtype_key == ETH_P_ALL) {
+			ethtype_key = 0;
+			ethtype_mask = 0;
+		}
+
+		if (ethtype_key == ETH_P_IPV6)
+			fs->type = 1;
+
+		fs->val.ethtype = ethtype_key;
+		fs->mask.ethtype = ethtype_mask;
+		fs->val.proto = match.key->ip_proto;
+		fs->mask.proto = match.mask->ip_proto;
+	}
+
+	if (addr_type == FLOW_DISSECTOR_KEY_IPV4_ADDRS) {
+		struct flow_match_ipv4_addrs match;
+
+		flow_rule_match_ipv4_addrs(rule, &match);
+		fs->type = 0;
+		memcpy(&fs->val.lip[0], &match.key->dst, sizeof(match.key->dst));
+		memcpy(&fs->val.fip[0], &match.key->src, sizeof(match.key->src));
+		memcpy(&fs->mask.lip[0], &match.mask->dst, sizeof(match.mask->dst));
+		memcpy(&fs->mask.fip[0], &match.mask->src, sizeof(match.mask->src));
+
+		/* also initialize nat_lip/fip to same values */
+		memcpy(&fs->nat_lip[0], &match.key->dst, sizeof(match.key->dst));
+		memcpy(&fs->nat_fip[0], &match.key->src, sizeof(match.key->src));
+	}
+
+	if (addr_type == FLOW_DISSECTOR_KEY_IPV6_ADDRS) {
+		struct flow_match_ipv6_addrs match;
+
+		flow_rule_match_ipv6_addrs(rule, &match);
+		fs->type = 1;
+		memcpy(&fs->val.lip[0], match.key->dst.s6_addr,
+		       sizeof(match.key->dst));
+		memcpy(&fs->val.fip[0], match.key->src.s6_addr,
+		       sizeof(match.key->src));
+		memcpy(&fs->mask.lip[0], match.mask->dst.s6_addr,
+		       sizeof(match.mask->dst));
+		memcpy(&fs->mask.fip[0], match.mask->src.s6_addr,
+		       sizeof(match.mask->src));
+
+		/* also initialize nat_lip/fip to same values */
+		memcpy(&fs->nat_lip[0], match.key->dst.s6_addr,
+		       sizeof(match.key->dst));
+		memcpy(&fs->nat_fip[0], match.key->src.s6_addr,
+		       sizeof(match.key->src));
+	}
+
+	if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_PORTS)) {
+		struct flow_match_ports match;
+
+		flow_rule_match_ports(rule, &match);
+		fs->val.lport = be16_to_cpu(match.key->dst);
+		fs->mask.lport = be16_to_cpu(match.mask->dst);
+		fs->val.fport = be16_to_cpu(match.key->src);
+		fs->mask.fport = be16_to_cpu(match.mask->src);
+
+		/* also initialize nat_lport/fport to same values */
+		fs->nat_lport = fs->val.lport;
+		fs->nat_fport = fs->val.fport;
+	}
+
+	if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_IP)) {
+		struct flow_match_ip match;
+
+		flow_rule_match_ip(rule, &match);
+		fs->val.tos = match.key->tos;
+		fs->mask.tos = match.mask->tos;
+	}
+
+	if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_KEYID)) {
+		struct flow_match_enc_keyid match;
+
+		flow_rule_match_enc_keyid(rule, &match);
+		fs->val.vni = be32_to_cpu(match.key->keyid);
+		fs->mask.vni = be32_to_cpu(match.mask->keyid);
+		if (fs->mask.vni) {
+			fs->val.encap_vld = 1;
+			fs->mask.encap_vld = 1;
+		}
+	}
+
+	if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_VLAN)) {
+		struct flow_match_vlan match;
+		u16 vlan_tci, vlan_tci_mask;
+
+		flow_rule_match_vlan(rule, &match);
+		vlan_tci = match.key->vlan_id | (match.key->vlan_priority <<
+					       VLAN_PRIO_SHIFT);
+		vlan_tci_mask = match.mask->vlan_id | (match.mask->vlan_priority <<
+						     VLAN_PRIO_SHIFT);
+		fs->val.ivlan = vlan_tci;
+		fs->mask.ivlan = vlan_tci_mask;
+
+		fs->val.ivlan_vld = 1;
+		fs->mask.ivlan_vld = 1;
+
+		/* Chelsio adapters use ivlan_vld bit to match vlan packets
+		 * as 802.1Q. Also, when vlan tag is present in packets,
+		 * ethtype match is used then to match on ethtype of inner
+		 * header ie. the header following the vlan header.
+		 * So, set the ivlan_vld based on ethtype info supplied by
+		 * TC for vlan packets if its 802.1Q. And then reset the
+		 * ethtype value else, hw will try to match the supplied
+		 * ethtype value with ethtype of inner header.
+		 */
+		if (fs->val.ethtype == ETH_P_8021Q) {
+			fs->val.ethtype = 0;
+			fs->mask.ethtype = 0;
+		}
+	}
+
+	/* Match only packets coming from the ingress port where this
+	 * filter will be created.
+	 */
+	fs->val.iport = netdev2pinfo(dev)->port_id;
+	fs->mask.iport = ~0;
+}
+
+static int cxgb4_validate_flow_match(struct net_device *dev,
+				     struct flow_rule *rule)
+{
+	struct flow_dissector *dissector = rule->match.dissector;
+	u16 ethtype_mask = 0;
+	u16 ethtype_key = 0;
+
+	if (dissector->used_keys &
+	    ~(BIT(FLOW_DISSECTOR_KEY_CONTROL) |
+	      BIT(FLOW_DISSECTOR_KEY_BASIC) |
+	      BIT(FLOW_DISSECTOR_KEY_IPV4_ADDRS) |
+	      BIT(FLOW_DISSECTOR_KEY_IPV6_ADDRS) |
+	      BIT(FLOW_DISSECTOR_KEY_PORTS) |
+	      BIT(FLOW_DISSECTOR_KEY_ENC_KEYID) |
+	      BIT(FLOW_DISSECTOR_KEY_VLAN) |
+	      BIT(FLOW_DISSECTOR_KEY_IP))) {
+		netdev_warn(dev, "Unsupported key used: 0x%x\n",
+			    dissector->used_keys);
+		return -EOPNOTSUPP;
+	}
+
+	if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_BASIC)) {
+		struct flow_match_basic match;
+
+		flow_rule_match_basic(rule, &match);
+		ethtype_key = ntohs(match.key->n_proto);
+		ethtype_mask = ntohs(match.mask->n_proto);
+	}
+
+	if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_IP)) {
+		u16 eth_ip_type = ethtype_key & ethtype_mask;
+		struct flow_match_ip match;
+
+		if (eth_ip_type != ETH_P_IP && eth_ip_type != ETH_P_IPV6) {
+			netdev_err(dev, "IP Key supported only with IPv4/v6");
+			return -EINVAL;
+		}
+
+		flow_rule_match_ip(rule, &match);
+		if (match.mask->ttl) {
+			netdev_warn(dev, "ttl match unsupported for offload");
+			return -EOPNOTSUPP;
+		}
+	}
+
+	return 0;
+}
+
+static void offload_pedit(struct ch_filter_specification *fs, u32 val, u32 mask,
+			  u8 field)
+{
+	u32 set_val = val & ~mask;
+	u32 offset = 0;
+	u8 size = 1;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(pedits); i++) {
+		if (pedits[i].field == field) {
+			offset = pedits[i].offset;
+			size = pedits[i].size;
+			break;
+		}
+	}
+	memcpy((u8 *)fs + offset, &set_val, size);
+}
+
+static void process_pedit_field(struct ch_filter_specification *fs, u32 val,
+				u32 mask, u32 offset, u8 htype,
+				u8 *natmode_flags)
+{
+	switch (htype) {
+	case FLOW_ACT_MANGLE_HDR_TYPE_ETH:
+		switch (offset) {
+		case PEDIT_ETH_DMAC_31_0:
+			fs->newdmac = 1;
+			offload_pedit(fs, val, mask, ETH_DMAC_31_0);
+			break;
+		case PEDIT_ETH_DMAC_47_32_SMAC_15_0:
+			if (~mask & PEDIT_ETH_DMAC_MASK)
+				offload_pedit(fs, val, mask, ETH_DMAC_47_32);
+			else
+				offload_pedit(fs, val >> 16, mask >> 16,
+					      ETH_SMAC_15_0);
+			break;
+		case PEDIT_ETH_SMAC_47_16:
+			fs->newsmac = 1;
+			offload_pedit(fs, val, mask, ETH_SMAC_47_16);
+		}
+		break;
+	case FLOW_ACT_MANGLE_HDR_TYPE_IP4:
+		switch (offset) {
+		case PEDIT_IP4_SRC:
+			offload_pedit(fs, val, mask, IP4_SRC);
+			*natmode_flags |= CXGB4_ACTION_NATMODE_SIP;
+			break;
+		case PEDIT_IP4_DST:
+			offload_pedit(fs, val, mask, IP4_DST);
+			*natmode_flags |= CXGB4_ACTION_NATMODE_DIP;
+		}
+		break;
+	case FLOW_ACT_MANGLE_HDR_TYPE_IP6:
+		switch (offset) {
+		case PEDIT_IP6_SRC_31_0:
+			offload_pedit(fs, val, mask, IP6_SRC_31_0);
+			*natmode_flags |= CXGB4_ACTION_NATMODE_SIP;
+			break;
+		case PEDIT_IP6_SRC_63_32:
+			offload_pedit(fs, val, mask, IP6_SRC_63_32);
+			*natmode_flags |=  CXGB4_ACTION_NATMODE_SIP;
+			break;
+		case PEDIT_IP6_SRC_95_64:
+			offload_pedit(fs, val, mask, IP6_SRC_95_64);
+			*natmode_flags |= CXGB4_ACTION_NATMODE_SIP;
+			break;
+		case PEDIT_IP6_SRC_127_96:
+			offload_pedit(fs, val, mask, IP6_SRC_127_96);
+			*natmode_flags |= CXGB4_ACTION_NATMODE_SIP;
+			break;
+		case PEDIT_IP6_DST_31_0:
+			offload_pedit(fs, val, mask, IP6_DST_31_0);
+			*natmode_flags |= CXGB4_ACTION_NATMODE_DIP;
+			break;
+		case PEDIT_IP6_DST_63_32:
+			offload_pedit(fs, val, mask, IP6_DST_63_32);
+			*natmode_flags |= CXGB4_ACTION_NATMODE_DIP;
+			break;
+		case PEDIT_IP6_DST_95_64:
+			offload_pedit(fs, val, mask, IP6_DST_95_64);
+			*natmode_flags |= CXGB4_ACTION_NATMODE_DIP;
+			break;
+		case PEDIT_IP6_DST_127_96:
+			offload_pedit(fs, val, mask, IP6_DST_127_96);
+			*natmode_flags |= CXGB4_ACTION_NATMODE_DIP;
+		}
+		break;
+	case FLOW_ACT_MANGLE_HDR_TYPE_TCP:
+		switch (offset) {
+		case PEDIT_TCP_SPORT_DPORT:
+			if (~mask & PEDIT_TCP_UDP_SPORT_MASK) {
+				fs->nat_fport = val;
+				*natmode_flags |= CXGB4_ACTION_NATMODE_SPORT;
+			} else {
+				fs->nat_lport = val >> 16;
+				*natmode_flags |= CXGB4_ACTION_NATMODE_DPORT;
+			}
+		}
+		break;
+	case FLOW_ACT_MANGLE_HDR_TYPE_UDP:
+		switch (offset) {
+		case PEDIT_UDP_SPORT_DPORT:
+			if (~mask & PEDIT_TCP_UDP_SPORT_MASK) {
+				fs->nat_fport = val;
+				*natmode_flags |= CXGB4_ACTION_NATMODE_SPORT;
+			} else {
+				fs->nat_lport = val >> 16;
+				*natmode_flags |= CXGB4_ACTION_NATMODE_DPORT;
+			}
+		}
+		break;
+	}
+}
+
+static int cxgb4_action_natmode_validate(struct adapter *adap, u8 natmode_flags,
+					 struct netlink_ext_ack *extack)
+{
+	u8 i = 0;
+
+	/* Extract the NAT mode to enable based on what 4-tuple fields
+	 * are enabled to be overwritten. This ensures that the
+	 * disabled fields don't get overwritten to 0.
+	 */
+	for (i = 0; i < ARRAY_SIZE(cxgb4_natmode_config_array); i++) {
+		const struct cxgb4_natmode_config *c;
+
+		c = &cxgb4_natmode_config_array[i];
+		if (CHELSIO_CHIP_VERSION(adap->params.chip) >= c->chip &&
+		    natmode_flags == c->flags)
+			return 0;
+	}
+	NL_SET_ERR_MSG_MOD(extack, "Unsupported NAT mode 4-tuple combination");
+	return -EOPNOTSUPP;
+}
+
+void cxgb4_process_flow_actions(struct net_device *in,
+				struct flow_action *actions,
+				struct ch_filter_specification *fs)
+{
+	struct flow_action_entry *act;
+	u8 natmode_flags = 0;
+	int i;
+
+	flow_action_for_each(i, act, actions) {
+		switch (act->id) {
+		case FLOW_ACTION_ACCEPT:
+			fs->action = FILTER_PASS;
+			break;
+		case FLOW_ACTION_DROP:
+			fs->action = FILTER_DROP;
+			break;
+		case FLOW_ACTION_MIRRED:
+		case FLOW_ACTION_REDIRECT: {
+			struct net_device *out = act->dev;
+			struct port_info *pi = netdev_priv(out);
+
+			fs->action = FILTER_SWITCH;
+			fs->eport = pi->port_id;
+			}
+			break;
+		case FLOW_ACTION_VLAN_POP:
+		case FLOW_ACTION_VLAN_PUSH:
+		case FLOW_ACTION_VLAN_MANGLE: {
+			u8 prio = act->vlan.prio;
+			u16 vid = act->vlan.vid;
+			u16 vlan_tci = (prio << VLAN_PRIO_SHIFT) | vid;
+			switch (act->id) {
+			case FLOW_ACTION_VLAN_POP:
+				fs->newvlan |= VLAN_REMOVE;
+				break;
+			case FLOW_ACTION_VLAN_PUSH:
+				fs->newvlan |= VLAN_INSERT;
+				fs->vlan = vlan_tci;
+				break;
+			case FLOW_ACTION_VLAN_MANGLE:
+				fs->newvlan |= VLAN_REWRITE;
+				fs->vlan = vlan_tci;
+				break;
+			default:
+				break;
+			}
+			}
+			break;
+		case FLOW_ACTION_MANGLE: {
+			u32 mask, val, offset;
+			u8 htype;
+
+			htype = act->mangle.htype;
+			mask = act->mangle.mask;
+			val = act->mangle.val;
+			offset = act->mangle.offset;
+
+			process_pedit_field(fs, val, mask, offset, htype,
+					    &natmode_flags);
+			}
+			break;
+		case FLOW_ACTION_QUEUE:
+			fs->action = FILTER_PASS;
+			fs->dirsteer = 1;
+			fs->iq = act->queue.index;
+			break;
+		default:
+			break;
+		}
+	}
+	if (natmode_flags)
+		cxgb4_action_natmode_tweak(fs, natmode_flags);
+
+}
+
+static bool valid_l4_mask(u32 mask)
+{
+	u16 hi, lo;
+
+	/* Either the upper 16-bits (SPORT) OR the lower
+	 * 16-bits (DPORT) can be set, but NOT BOTH.
+	 */
+	hi = (mask >> 16) & 0xFFFF;
+	lo = mask & 0xFFFF;
+
+	return hi && lo ? false : true;
+}
+
+static bool valid_pedit_action(struct net_device *dev,
+			       const struct flow_action_entry *act,
+			       u8 *natmode_flags)
+{
+	u32 mask, offset;
+	u8 htype;
+
+	htype = act->mangle.htype;
+	mask = act->mangle.mask;
+	offset = act->mangle.offset;
+
+	switch (htype) {
+	case FLOW_ACT_MANGLE_HDR_TYPE_ETH:
+		switch (offset) {
+		case PEDIT_ETH_DMAC_31_0:
+		case PEDIT_ETH_DMAC_47_32_SMAC_15_0:
+		case PEDIT_ETH_SMAC_47_16:
+			break;
+		default:
+			netdev_err(dev, "%s: Unsupported pedit field\n",
+				   __func__);
+			return false;
+		}
+		break;
+	case FLOW_ACT_MANGLE_HDR_TYPE_IP4:
+		switch (offset) {
+		case PEDIT_IP4_SRC:
+			*natmode_flags |= CXGB4_ACTION_NATMODE_SIP;
+			break;
+		case PEDIT_IP4_DST:
+			*natmode_flags |= CXGB4_ACTION_NATMODE_DIP;
+			break;
+		default:
+			netdev_err(dev, "%s: Unsupported pedit field\n",
+				   __func__);
+			return false;
+		}
+		break;
+	case FLOW_ACT_MANGLE_HDR_TYPE_IP6:
+		switch (offset) {
+		case PEDIT_IP6_SRC_31_0:
+		case PEDIT_IP6_SRC_63_32:
+		case PEDIT_IP6_SRC_95_64:
+		case PEDIT_IP6_SRC_127_96:
+			*natmode_flags |= CXGB4_ACTION_NATMODE_SIP;
+			break;
+		case PEDIT_IP6_DST_31_0:
+		case PEDIT_IP6_DST_63_32:
+		case PEDIT_IP6_DST_95_64:
+		case PEDIT_IP6_DST_127_96:
+			*natmode_flags |= CXGB4_ACTION_NATMODE_DIP;
+			break;
+		default:
+			netdev_err(dev, "%s: Unsupported pedit field\n",
+				   __func__);
+			return false;
+		}
+		break;
+	case FLOW_ACT_MANGLE_HDR_TYPE_TCP:
+		switch (offset) {
+		case PEDIT_TCP_SPORT_DPORT:
+			if (!valid_l4_mask(~mask)) {
+				netdev_err(dev, "%s: Unsupported mask for TCP L4 ports\n",
+					   __func__);
+				return false;
+			}
+			if (~mask & PEDIT_TCP_UDP_SPORT_MASK)
+				*natmode_flags |= CXGB4_ACTION_NATMODE_SPORT;
+			else
+				*natmode_flags |= CXGB4_ACTION_NATMODE_DPORT;
+			break;
+		default:
+			netdev_err(dev, "%s: Unsupported pedit field\n",
+				   __func__);
+			return false;
+		}
+		break;
+	case FLOW_ACT_MANGLE_HDR_TYPE_UDP:
+		switch (offset) {
+		case PEDIT_UDP_SPORT_DPORT:
+			if (!valid_l4_mask(~mask)) {
+				netdev_err(dev, "%s: Unsupported mask for UDP L4 ports\n",
+					   __func__);
+				return false;
+			}
+			if (~mask & PEDIT_TCP_UDP_SPORT_MASK)
+				*natmode_flags |= CXGB4_ACTION_NATMODE_SPORT;
+			else
+				*natmode_flags |= CXGB4_ACTION_NATMODE_DPORT;
+			break;
+		default:
+			netdev_err(dev, "%s: Unsupported pedit field\n",
+				   __func__);
+			return false;
+		}
+		break;
+	default:
+		netdev_err(dev, "%s: Unsupported pedit type\n", __func__);
+		return false;
+	}
+	return true;
+}
+
+int cxgb4_validate_flow_actions(struct net_device *dev,
+				struct flow_action *actions,
+				struct netlink_ext_ack *extack,
+				u8 matchall_filter)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct flow_action_entry *act;
+	bool act_redir = false;
+	bool act_pedit = false;
+	bool act_vlan = false;
+	u8 natmode_flags = 0;
+	int i;
+
+	if (!flow_action_basic_hw_stats_check(actions, extack))
+		return -EOPNOTSUPP;
+
+	flow_action_for_each(i, act, actions) {
+		switch (act->id) {
+		case FLOW_ACTION_ACCEPT:
+		case FLOW_ACTION_DROP:
+			/* Do nothing */
+			break;
+		case FLOW_ACTION_MIRRED:
+		case FLOW_ACTION_REDIRECT: {
+			struct net_device *n_dev, *target_dev;
+			bool found = false;
+			unsigned int i;
+
+			if (act->id == FLOW_ACTION_MIRRED &&
+			    !matchall_filter) {
+				NL_SET_ERR_MSG_MOD(extack,
+						   "Egress mirror action is only supported for tc-matchall");
+				return -EOPNOTSUPP;
+			}
+
+			target_dev = act->dev;
+			for_each_port(adap, i) {
+				n_dev = adap->port[i];
+				if (target_dev == n_dev) {
+					found = true;
+					break;
+				}
+			}
+
+			/* If interface doesn't belong to our hw, then
+			 * the provided output port is not valid
+			 */
+			if (!found) {
+				netdev_err(dev, "%s: Out port invalid\n",
+					   __func__);
+				return -EINVAL;
+			}
+			act_redir = true;
+			}
+			break;
+		case FLOW_ACTION_VLAN_POP:
+		case FLOW_ACTION_VLAN_PUSH:
+		case FLOW_ACTION_VLAN_MANGLE: {
+			u16 proto = be16_to_cpu(act->vlan.proto);
+
+			switch (act->id) {
+			case FLOW_ACTION_VLAN_POP:
+				break;
+			case FLOW_ACTION_VLAN_PUSH:
+			case FLOW_ACTION_VLAN_MANGLE:
+				if (proto != ETH_P_8021Q) {
+					netdev_err(dev, "%s: Unsupported vlan proto\n",
+						   __func__);
+					return -EOPNOTSUPP;
+				}
+				break;
+			default:
+				netdev_err(dev, "%s: Unsupported vlan action\n",
+					   __func__);
+				return -EOPNOTSUPP;
+			}
+			act_vlan = true;
+			}
+			break;
+		case FLOW_ACTION_MANGLE: {
+			bool pedit_valid = valid_pedit_action(dev, act,
+							      &natmode_flags);
+
+			if (!pedit_valid)
+				return -EOPNOTSUPP;
+			act_pedit = true;
+			}
+			break;
+		case FLOW_ACTION_QUEUE:
+			/* Do nothing. cxgb4_set_filter will validate */
+			break;
+		default:
+			netdev_err(dev, "%s: Unsupported action\n", __func__);
+			return -EOPNOTSUPP;
+		}
+	}
+
+	if ((act_pedit || act_vlan) && !act_redir) {
+		netdev_err(dev, "%s: pedit/vlan rewrite invalid without egress redirect\n",
+			   __func__);
+		return -EINVAL;
+	}
+
+	if (act_pedit) {
+		int ret;
+
+		ret = cxgb4_action_natmode_validate(adap, natmode_flags,
+						    extack);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static void cxgb4_tc_flower_hash_prio_add(struct adapter *adap, u32 tc_prio)
+{
+	spin_lock_bh(&adap->tids.ftid_lock);
+	if (adap->tids.tc_hash_tids_max_prio < tc_prio)
+		adap->tids.tc_hash_tids_max_prio = tc_prio;
+	spin_unlock_bh(&adap->tids.ftid_lock);
+}
+
+static void cxgb4_tc_flower_hash_prio_del(struct adapter *adap, u32 tc_prio)
+{
+	struct tid_info *t = &adap->tids;
+	struct ch_tc_flower_entry *fe;
+	struct rhashtable_iter iter;
+	u32 found = 0;
+
+	spin_lock_bh(&t->ftid_lock);
+	/* Bail if the current rule is not the one with the max
+	 * prio.
+	 */
+	if (t->tc_hash_tids_max_prio != tc_prio)
+		goto out_unlock;
+
+	/* Search for the next rule having the same or next lower
+	 * max prio.
+	 */
+	rhashtable_walk_enter(&adap->flower_tbl, &iter);
+	do {
+		rhashtable_walk_start(&iter);
+
+		fe = rhashtable_walk_next(&iter);
+		while (!IS_ERR_OR_NULL(fe)) {
+			if (fe->fs.hash &&
+			    fe->fs.tc_prio <= t->tc_hash_tids_max_prio) {
+				t->tc_hash_tids_max_prio = fe->fs.tc_prio;
+				found++;
+
+				/* Bail if we found another rule
+				 * having the same prio as the
+				 * current max one.
+				 */
+				if (fe->fs.tc_prio == tc_prio)
+					break;
+			}
+
+			fe = rhashtable_walk_next(&iter);
+		}
+
+		rhashtable_walk_stop(&iter);
+	} while (fe == ERR_PTR(-EAGAIN));
+	rhashtable_walk_exit(&iter);
+
+	if (!found)
+		t->tc_hash_tids_max_prio = 0;
+
+out_unlock:
+	spin_unlock_bh(&t->ftid_lock);
+}
+
+int cxgb4_flow_rule_replace(struct net_device *dev, struct flow_rule *rule,
+			    u32 tc_prio, struct netlink_ext_ack *extack,
+			    struct ch_filter_specification *fs, u32 *tid)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct filter_ctx ctx;
+	u8 inet_family;
+	int fidx, ret;
+
+	if (cxgb4_validate_flow_actions(dev, &rule->action, extack, 0))
+		return -EOPNOTSUPP;
+
+	if (cxgb4_validate_flow_match(dev, rule))
+		return -EOPNOTSUPP;
+
+	cxgb4_process_flow_match(dev, rule, fs);
+	cxgb4_process_flow_actions(dev, &rule->action, fs);
+
+	fs->hash = is_filter_exact_match(adap, fs);
+	inet_family = fs->type ? PF_INET6 : PF_INET;
+
+	/* Get a free filter entry TID, where we can insert this new
+	 * rule. Only insert rule if its prio doesn't conflict with
+	 * existing rules.
+	 */
+	fidx = cxgb4_get_free_ftid(dev, inet_family, fs->hash,
+				   tc_prio);
+	if (fidx < 0) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "No free LETCAM index available");
+		return -ENOMEM;
+	}
+
+	if (fidx < adap->tids.nhpftids) {
+		fs->prio = 1;
+		fs->hash = 0;
+	}
+
+	/* If the rule can be inserted into HASH region, then ignore
+	 * the index to normal FILTER region.
+	 */
+	if (fs->hash)
+		fidx = 0;
+
+	fs->tc_prio = tc_prio;
+
+	init_completion(&ctx.completion);
+	ret = __cxgb4_set_filter(dev, fidx, fs, &ctx);
+	if (ret) {
+		netdev_err(dev, "%s: filter creation err %d\n",
+			   __func__, ret);
+		return ret;
+	}
+
+	/* Wait for reply */
+	ret = wait_for_completion_timeout(&ctx.completion, 10 * HZ);
+	if (!ret)
+		return -ETIMEDOUT;
+
+	/* Check if hw returned error for filter creation */
+	if (ctx.result)
+		return ctx.result;
+
+	*tid = ctx.tid;
+
+	if (fs->hash)
+		cxgb4_tc_flower_hash_prio_add(adap, tc_prio);
+
+	return 0;
+}
+
+int cxgb4_tc_flower_replace(struct net_device *dev,
+			    struct flow_cls_offload *cls)
+{
+	struct flow_rule *rule = flow_cls_offload_flow_rule(cls);
+	struct netlink_ext_ack *extack = cls->common.extack;
+	struct adapter *adap = netdev2adap(dev);
+	struct ch_tc_flower_entry *ch_flower;
+	struct ch_filter_specification *fs;
+	int ret;
+
+	ch_flower = allocate_flower_entry();
+	if (!ch_flower) {
+		netdev_err(dev, "%s: ch_flower alloc failed.\n", __func__);
+		return -ENOMEM;
+	}
+
+	fs = &ch_flower->fs;
+	fs->hitcnts = 1;
+	fs->tc_cookie = cls->cookie;
+
+	ret = cxgb4_flow_rule_replace(dev, rule, cls->common.prio, extack, fs,
+				      &ch_flower->filter_id);
+	if (ret)
+		goto free_entry;
+
+	ch_flower->tc_flower_cookie = cls->cookie;
+	ret = rhashtable_insert_fast(&adap->flower_tbl, &ch_flower->node,
+				     adap->flower_ht_params);
+	if (ret)
+		goto del_filter;
+
+	return 0;
+
+del_filter:
+	if (fs->hash)
+		cxgb4_tc_flower_hash_prio_del(adap, cls->common.prio);
+
+	cxgb4_del_filter(dev, ch_flower->filter_id, &ch_flower->fs);
+
+free_entry:
+	kfree(ch_flower);
+	return ret;
+}
+
+int cxgb4_flow_rule_destroy(struct net_device *dev, u32 tc_prio,
+			    struct ch_filter_specification *fs, int tid)
+{
+	struct adapter *adap = netdev2adap(dev);
+	u8 hash;
+	int ret;
+
+	hash = fs->hash;
+
+	ret = cxgb4_del_filter(dev, tid, fs);
+	if (ret)
+		return ret;
+
+	if (hash)
+		cxgb4_tc_flower_hash_prio_del(adap, tc_prio);
+
+	return ret;
+}
+
+int cxgb4_tc_flower_destroy(struct net_device *dev,
+			    struct flow_cls_offload *cls)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct ch_tc_flower_entry *ch_flower;
+	int ret;
+
+	ch_flower = ch_flower_lookup(adap, cls->cookie);
+	if (!ch_flower)
+		return -ENOENT;
+
+	rhashtable_remove_fast(&adap->flower_tbl, &ch_flower->node,
+			       adap->flower_ht_params);
+
+	ret = cxgb4_flow_rule_destroy(dev, ch_flower->fs.tc_prio,
+				      &ch_flower->fs, ch_flower->filter_id);
+	if (ret)
+		netdev_err(dev, "Flow rule destroy failed for tid: %u, ret: %d",
+			   ch_flower->filter_id, ret);
+
+	kfree_rcu(ch_flower, rcu);
+	return ret;
+}
+
+static void ch_flower_stats_handler(struct work_struct *work)
+{
+	struct adapter *adap = container_of(work, struct adapter,
+					    flower_stats_work);
+	struct ch_tc_flower_entry *flower_entry;
+	struct ch_tc_flower_stats *ofld_stats;
+	struct rhashtable_iter iter;
+	u64 packets;
+	u64 bytes;
+	int ret;
+
+	rhashtable_walk_enter(&adap->flower_tbl, &iter);
+	do {
+		rhashtable_walk_start(&iter);
+
+		while ((flower_entry = rhashtable_walk_next(&iter)) &&
+		       !IS_ERR(flower_entry)) {
+			ret = cxgb4_get_filter_counters(adap->port[0],
+							flower_entry->filter_id,
+							&packets, &bytes,
+							flower_entry->fs.hash);
+			if (!ret) {
+				spin_lock(&flower_entry->lock);
+				ofld_stats = &flower_entry->stats;
+
+				if (ofld_stats->prev_packet_count != packets) {
+					ofld_stats->prev_packet_count = packets;
+					ofld_stats->last_used = jiffies;
+				}
+				spin_unlock(&flower_entry->lock);
+			}
+		}
+
+		rhashtable_walk_stop(&iter);
+
+	} while (flower_entry == ERR_PTR(-EAGAIN));
+	rhashtable_walk_exit(&iter);
+	mod_timer(&adap->flower_stats_timer, jiffies + STATS_CHECK_PERIOD);
+}
+
+static void ch_flower_stats_cb(struct timer_list *t)
+{
+	struct adapter *adap = from_timer(adap, t, flower_stats_timer);
+
+	schedule_work(&adap->flower_stats_work);
+}
+
+int cxgb4_tc_flower_stats(struct net_device *dev,
+			  struct flow_cls_offload *cls)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct ch_tc_flower_stats *ofld_stats;
+	struct ch_tc_flower_entry *ch_flower;
+	u64 packets;
+	u64 bytes;
+	int ret;
+
+	ch_flower = ch_flower_lookup(adap, cls->cookie);
+	if (!ch_flower) {
+		ret = -ENOENT;
+		goto err;
+	}
+
+	ret = cxgb4_get_filter_counters(dev, ch_flower->filter_id,
+					&packets, &bytes,
+					ch_flower->fs.hash);
+	if (ret < 0)
+		goto err;
+
+	spin_lock_bh(&ch_flower->lock);
+	ofld_stats = &ch_flower->stats;
+	if (ofld_stats->packet_count != packets) {
+		if (ofld_stats->prev_packet_count != packets)
+			ofld_stats->last_used = jiffies;
+		flow_stats_update(&cls->stats, bytes - ofld_stats->byte_count,
+				  packets - ofld_stats->packet_count, 0,
+				  ofld_stats->last_used,
+				  FLOW_ACTION_HW_STATS_IMMEDIATE);
+
+		ofld_stats->packet_count = packets;
+		ofld_stats->byte_count = bytes;
+		ofld_stats->prev_packet_count = packets;
+	}
+	spin_unlock_bh(&ch_flower->lock);
+	return 0;
+
+err:
+	return ret;
+}
+
+static const struct rhashtable_params cxgb4_tc_flower_ht_params = {
+	.nelem_hint = 384,
+	.head_offset = offsetof(struct ch_tc_flower_entry, node),
+	.key_offset = offsetof(struct ch_tc_flower_entry, tc_flower_cookie),
+	.key_len = sizeof(((struct ch_tc_flower_entry *)0)->tc_flower_cookie),
+	.max_size = 524288,
+	.min_size = 512,
+	.automatic_shrinking = true
+};
+
+int cxgb4_init_tc_flower(struct adapter *adap)
+{
+	int ret;
+
+	if (adap->tc_flower_initialized)
+		return -EEXIST;
+
+	adap->flower_ht_params = cxgb4_tc_flower_ht_params;
+	ret = rhashtable_init(&adap->flower_tbl, &adap->flower_ht_params);
+	if (ret)
+		return ret;
+
+	INIT_WORK(&adap->flower_stats_work, ch_flower_stats_handler);
+	timer_setup(&adap->flower_stats_timer, ch_flower_stats_cb, 0);
+	mod_timer(&adap->flower_stats_timer, jiffies + STATS_CHECK_PERIOD);
+	adap->tc_flower_initialized = true;
+	return 0;
+}
+
+void cxgb4_cleanup_tc_flower(struct adapter *adap)
+{
+	if (!adap->tc_flower_initialized)
+		return;
+
+	if (adap->flower_stats_timer.function)
+		del_timer_sync(&adap->flower_stats_timer);
+	cancel_work_sync(&adap->flower_stats_work);
+	rhashtable_destroy(&adap->flower_tbl);
+	adap->tc_flower_initialized = false;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
new file mode 100644
index 000000000..3a2fa00c8
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
@@ -0,0 +1,148 @@
+/*
+ * This file is part of the Chelsio T4/T5/T6 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2017 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_TC_FLOWER_H
+#define __CXGB4_TC_FLOWER_H
+
+#include <net/pkt_cls.h>
+
+struct ch_tc_flower_stats {
+	u64 prev_packet_count;
+	u64 packet_count;
+	u64 byte_count;
+	u64 last_used;
+};
+
+struct ch_tc_flower_entry {
+	struct ch_filter_specification fs;
+	struct ch_tc_flower_stats stats;
+	unsigned long tc_flower_cookie;
+	struct rhash_head node;
+	struct rcu_head rcu;
+	spinlock_t lock; /* lock for stats */
+	u32 filter_id;
+};
+
+enum {
+	ETH_DMAC_31_0,	/* dmac bits 0.. 31 */
+	ETH_DMAC_47_32,	/* dmac bits 32..47 */
+	ETH_SMAC_15_0,	/* smac bits 0.. 15 */
+	ETH_SMAC_47_16,	/* smac bits 16..47 */
+
+	IP4_SRC,	/* 32-bit IPv4 src  */
+	IP4_DST,	/* 32-bit IPv4 dst  */
+
+	IP6_SRC_31_0,	/* src bits 0..  31 */
+	IP6_SRC_63_32,	/* src bits 63.. 32 */
+	IP6_SRC_95_64,	/* src bits 95.. 64 */
+	IP6_SRC_127_96,	/* src bits 127..96 */
+
+	IP6_DST_31_0,	/* dst bits 0..  31 */
+	IP6_DST_63_32,	/* dst bits 63.. 32 */
+	IP6_DST_95_64,	/* dst bits 95.. 64 */
+	IP6_DST_127_96,	/* dst bits 127..96 */
+
+	TCP_SPORT,	/* 16-bit TCP sport */
+	TCP_DPORT,	/* 16-bit TCP dport */
+
+	UDP_SPORT,	/* 16-bit UDP sport */
+	UDP_DPORT,	/* 16-bit UDP dport */
+};
+
+struct ch_tc_pedit_fields {
+	u8 field;
+	u8 size;
+	u32 offset;
+};
+
+#define PEDIT_FIELDS(type, field, size, fs_field, offset) \
+	{ type## field, size, \
+		offsetof(struct ch_filter_specification, fs_field) + (offset) }
+
+#define PEDIT_ETH_DMAC_MASK		0xffff
+#define PEDIT_TCP_UDP_SPORT_MASK	0xffff
+#define PEDIT_ETH_DMAC_31_0		0x0
+#define PEDIT_ETH_DMAC_47_32_SMAC_15_0	0x4
+#define PEDIT_ETH_SMAC_47_16		0x8
+#define PEDIT_IP4_SRC			0xC
+#define PEDIT_IP4_DST			0x10
+#define PEDIT_IP6_SRC_31_0		0x8
+#define PEDIT_IP6_SRC_63_32		0xC
+#define PEDIT_IP6_SRC_95_64		0x10
+#define PEDIT_IP6_SRC_127_96		0x14
+#define PEDIT_IP6_DST_31_0		0x18
+#define PEDIT_IP6_DST_63_32		0x1C
+#define PEDIT_IP6_DST_95_64		0x20
+#define PEDIT_IP6_DST_127_96		0x24
+#define PEDIT_TCP_SPORT_DPORT		0x0
+#define PEDIT_UDP_SPORT_DPORT		0x0
+
+enum cxgb4_action_natmode_flags {
+	CXGB4_ACTION_NATMODE_NONE = 0,
+	CXGB4_ACTION_NATMODE_DIP = (1 << 0),
+	CXGB4_ACTION_NATMODE_SIP = (1 << 1),
+	CXGB4_ACTION_NATMODE_DPORT = (1 << 2),
+	CXGB4_ACTION_NATMODE_SPORT = (1 << 3),
+};
+
+/* TC PEDIT action to NATMODE translation entry */
+struct cxgb4_natmode_config {
+	enum chip_type chip;
+	u8 flags;
+	u8 natmode;
+};
+
+void cxgb4_process_flow_actions(struct net_device *in,
+				struct flow_action *actions,
+				struct ch_filter_specification *fs);
+int cxgb4_validate_flow_actions(struct net_device *dev,
+				struct flow_action *actions,
+				struct netlink_ext_ack *extack,
+				u8 matchall_filter);
+
+int cxgb4_tc_flower_replace(struct net_device *dev,
+			    struct flow_cls_offload *cls);
+int cxgb4_tc_flower_destroy(struct net_device *dev,
+			    struct flow_cls_offload *cls);
+int cxgb4_tc_flower_stats(struct net_device *dev,
+			  struct flow_cls_offload *cls);
+int cxgb4_flow_rule_replace(struct net_device *dev, struct flow_rule *rule,
+			    u32 tc_prio, struct netlink_ext_ack *extack,
+			    struct ch_filter_specification *fs, u32 *tid);
+int cxgb4_flow_rule_destroy(struct net_device *dev, u32 tc_prio,
+			    struct ch_filter_specification *fs, int tid);
+
+int cxgb4_init_tc_flower(struct adapter *adap);
+void cxgb4_cleanup_tc_flower(struct adapter *adap);
+#endif /* __CXGB4_TC_FLOWER_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.c
new file mode 100644
index 000000000..1672d3afe
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.c
@@ -0,0 +1,571 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2019 Chelsio Communications.  All rights reserved. */
+
+#include "cxgb4.h"
+#include "cxgb4_tc_matchall.h"
+#include "sched.h"
+#include "cxgb4_uld.h"
+#include "cxgb4_filter.h"
+#include "cxgb4_tc_flower.h"
+
+static int cxgb4_policer_validate(const struct flow_action *action,
+				  const struct flow_action_entry *act,
+				  struct netlink_ext_ack *extack)
+{
+	if (act->police.exceed.act_id != FLOW_ACTION_DROP) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Offload not supported when exceed action is not drop");
+		return -EOPNOTSUPP;
+	}
+
+	if (act->police.notexceed.act_id != FLOW_ACTION_PIPE &&
+	    act->police.notexceed.act_id != FLOW_ACTION_ACCEPT) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Offload not supported when conform action is not pipe or ok");
+		return -EOPNOTSUPP;
+	}
+
+	if (act->police.notexceed.act_id == FLOW_ACTION_ACCEPT &&
+	    !flow_action_is_last_entry(action, act)) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Offload not supported when conform action is ok, but action is not last");
+		return -EOPNOTSUPP;
+	}
+
+	if (act->police.peakrate_bytes_ps ||
+	    act->police.avrate || act->police.overhead) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Offload not supported when peakrate/avrate/overhead is configured");
+		return -EOPNOTSUPP;
+	}
+
+	if (act->police.rate_pkt_ps) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "QoS offload not support packets per second");
+		return -EOPNOTSUPP;
+	}
+
+	return 0;
+}
+
+static int cxgb4_matchall_egress_validate(struct net_device *dev,
+					  struct tc_cls_matchall_offload *cls)
+{
+	struct netlink_ext_ack *extack = cls->common.extack;
+	struct flow_action *actions = &cls->rule->action;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct flow_action_entry *entry;
+	struct ch_sched_queue qe;
+	struct sched_class *e;
+	u64 max_link_rate;
+	u32 i, speed;
+	int ret;
+
+	if (!flow_action_has_entries(actions)) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Egress MATCHALL offload needs at least 1 policing action");
+		return -EINVAL;
+	} else if (!flow_offload_has_one_action(actions)) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Egress MATCHALL offload only supports 1 policing action");
+		return -EINVAL;
+	} else if (pi->tc_block_shared) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Egress MATCHALL offload not supported with shared blocks");
+		return -EINVAL;
+	}
+
+	ret = t4_get_link_params(pi, NULL, &speed, NULL);
+	if (ret) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Failed to get max speed supported by the link");
+		return -EINVAL;
+	}
+
+	/* Convert from Mbps to bps */
+	max_link_rate = (u64)speed * 1000 * 1000;
+
+	flow_action_for_each(i, entry, actions) {
+		switch (entry->id) {
+		case FLOW_ACTION_POLICE:
+			ret = cxgb4_policer_validate(actions, entry, extack);
+			if (ret)
+				return ret;
+
+			/* Convert bytes per second to bits per second */
+			if (entry->police.rate_bytes_ps * 8 > max_link_rate) {
+				NL_SET_ERR_MSG_MOD(extack,
+						   "Specified policing max rate is larger than underlying link speed");
+				return -ERANGE;
+			}
+			break;
+		default:
+			NL_SET_ERR_MSG_MOD(extack,
+					   "Only policing action supported with Egress MATCHALL offload");
+			return -EOPNOTSUPP;
+		}
+	}
+
+	for (i = 0; i < pi->nqsets; i++) {
+		memset(&qe, 0, sizeof(qe));
+		qe.queue = i;
+
+		e = cxgb4_sched_queue_lookup(dev, &qe);
+		if (e && e->info.u.params.level != SCHED_CLASS_LEVEL_CH_RL) {
+			NL_SET_ERR_MSG_MOD(extack,
+					   "Some queues are already bound to different class");
+			return -EBUSY;
+		}
+	}
+
+	return 0;
+}
+
+static int cxgb4_matchall_tc_bind_queues(struct net_device *dev, u32 tc)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct ch_sched_queue qe;
+	int ret;
+	u32 i;
+
+	for (i = 0; i < pi->nqsets; i++) {
+		qe.queue = i;
+		qe.class = tc;
+		ret = cxgb4_sched_class_bind(dev, &qe, SCHED_QUEUE);
+		if (ret)
+			goto out_free;
+	}
+
+	return 0;
+
+out_free:
+	while (i--) {
+		qe.queue = i;
+		qe.class = SCHED_CLS_NONE;
+		cxgb4_sched_class_unbind(dev, &qe, SCHED_QUEUE);
+	}
+
+	return ret;
+}
+
+static void cxgb4_matchall_tc_unbind_queues(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct ch_sched_queue qe;
+	u32 i;
+
+	for (i = 0; i < pi->nqsets; i++) {
+		qe.queue = i;
+		qe.class = SCHED_CLS_NONE;
+		cxgb4_sched_class_unbind(dev, &qe, SCHED_QUEUE);
+	}
+}
+
+static int cxgb4_matchall_alloc_tc(struct net_device *dev,
+				   struct tc_cls_matchall_offload *cls)
+{
+	struct ch_sched_params p = {
+		.type = SCHED_CLASS_TYPE_PACKET,
+		.u.params.level = SCHED_CLASS_LEVEL_CH_RL,
+		.u.params.mode = SCHED_CLASS_MODE_CLASS,
+		.u.params.rateunit = SCHED_CLASS_RATEUNIT_BITS,
+		.u.params.ratemode = SCHED_CLASS_RATEMODE_ABS,
+		.u.params.class = SCHED_CLS_NONE,
+		.u.params.minrate = 0,
+		.u.params.weight = 0,
+		.u.params.pktsize = dev->mtu,
+	};
+	struct netlink_ext_ack *extack = cls->common.extack;
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct flow_action_entry *entry;
+	struct sched_class *e;
+	int ret;
+	u32 i;
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+
+	flow_action_for_each(i, entry, &cls->rule->action)
+		if (entry->id == FLOW_ACTION_POLICE)
+			break;
+
+	ret = cxgb4_policer_validate(&cls->rule->action, entry, extack);
+	if (ret)
+		return ret;
+
+	/* Convert from bytes per second to Kbps */
+	p.u.params.maxrate = div_u64(entry->police.rate_bytes_ps * 8, 1000);
+	p.u.params.channel = pi->tx_chan;
+	e = cxgb4_sched_class_alloc(dev, &p);
+	if (!e) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "No free traffic class available for policing action");
+		return -ENOMEM;
+	}
+
+	ret = cxgb4_matchall_tc_bind_queues(dev, e->idx);
+	if (ret) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Could not bind queues to traffic class");
+		goto out_free;
+	}
+
+	tc_port_matchall->egress.hwtc = e->idx;
+	tc_port_matchall->egress.cookie = cls->cookie;
+	tc_port_matchall->egress.state = CXGB4_MATCHALL_STATE_ENABLED;
+	return 0;
+
+out_free:
+	cxgb4_sched_class_free(dev, e->idx);
+	return ret;
+}
+
+static void cxgb4_matchall_free_tc(struct net_device *dev)
+{
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+	cxgb4_matchall_tc_unbind_queues(dev);
+	cxgb4_sched_class_free(dev, tc_port_matchall->egress.hwtc);
+
+	tc_port_matchall->egress.hwtc = SCHED_CLS_NONE;
+	tc_port_matchall->egress.cookie = 0;
+	tc_port_matchall->egress.state = CXGB4_MATCHALL_STATE_DISABLED;
+}
+
+static int cxgb4_matchall_mirror_alloc(struct net_device *dev,
+				       struct tc_cls_matchall_offload *cls)
+{
+	struct netlink_ext_ack *extack = cls->common.extack;
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct flow_action_entry *act;
+	int ret;
+	u32 i;
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+	flow_action_for_each(i, act, &cls->rule->action) {
+		if (act->id == FLOW_ACTION_MIRRED) {
+			ret = cxgb4_port_mirror_alloc(dev);
+			if (ret) {
+				NL_SET_ERR_MSG_MOD(extack,
+						   "Couldn't allocate mirror");
+				return ret;
+			}
+
+			tc_port_matchall->ingress.viid_mirror = pi->viid_mirror;
+			break;
+		}
+	}
+
+	return 0;
+}
+
+static void cxgb4_matchall_mirror_free(struct net_device *dev)
+{
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+	if (!tc_port_matchall->ingress.viid_mirror)
+		return;
+
+	cxgb4_port_mirror_free(dev);
+	tc_port_matchall->ingress.viid_mirror = 0;
+}
+
+static int cxgb4_matchall_del_filter(struct net_device *dev, u8 filter_type)
+{
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	int ret;
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+	ret = cxgb4_del_filter(dev, tc_port_matchall->ingress.tid[filter_type],
+			       &tc_port_matchall->ingress.fs[filter_type]);
+	if (ret)
+		return ret;
+
+	tc_port_matchall->ingress.tid[filter_type] = 0;
+	return 0;
+}
+
+static int cxgb4_matchall_add_filter(struct net_device *dev,
+				     struct tc_cls_matchall_offload *cls,
+				     u8 filter_type)
+{
+	struct netlink_ext_ack *extack = cls->common.extack;
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct ch_filter_specification *fs;
+	int ret, fidx;
+
+	/* Get a free filter entry TID, where we can insert this new
+	 * rule. Only insert rule if its prio doesn't conflict with
+	 * existing rules.
+	 */
+	fidx = cxgb4_get_free_ftid(dev, filter_type ? PF_INET6 : PF_INET,
+				   false, cls->common.prio);
+	if (fidx < 0) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "No free LETCAM index available");
+		return -ENOMEM;
+	}
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+	fs = &tc_port_matchall->ingress.fs[filter_type];
+	memset(fs, 0, sizeof(*fs));
+
+	if (fidx < adap->tids.nhpftids)
+		fs->prio = 1;
+	fs->tc_prio = cls->common.prio;
+	fs->tc_cookie = cls->cookie;
+	fs->type = filter_type;
+	fs->hitcnts = 1;
+
+	fs->val.pfvf_vld = 1;
+	fs->val.pf = adap->pf;
+	fs->val.vf = pi->vin;
+
+	cxgb4_process_flow_actions(dev, &cls->rule->action, fs);
+
+	ret = cxgb4_set_filter(dev, fidx, fs);
+	if (ret)
+		return ret;
+
+	tc_port_matchall->ingress.tid[filter_type] = fidx;
+	return 0;
+}
+
+static int cxgb4_matchall_alloc_filter(struct net_device *dev,
+				       struct tc_cls_matchall_offload *cls)
+{
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	int ret, i;
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+
+	ret = cxgb4_matchall_mirror_alloc(dev, cls);
+	if (ret)
+		return ret;
+
+	for (i = 0; i < CXGB4_FILTER_TYPE_MAX; i++) {
+		ret = cxgb4_matchall_add_filter(dev, cls, i);
+		if (ret)
+			goto out_free;
+	}
+
+	tc_port_matchall->ingress.state = CXGB4_MATCHALL_STATE_ENABLED;
+	return 0;
+
+out_free:
+	while (i-- > 0)
+		cxgb4_matchall_del_filter(dev, i);
+
+	cxgb4_matchall_mirror_free(dev);
+	return ret;
+}
+
+static int cxgb4_matchall_free_filter(struct net_device *dev)
+{
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	int ret;
+	u8 i;
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+
+	for (i = 0; i < CXGB4_FILTER_TYPE_MAX; i++) {
+		ret = cxgb4_matchall_del_filter(dev, i);
+		if (ret)
+			return ret;
+	}
+
+	cxgb4_matchall_mirror_free(dev);
+
+	tc_port_matchall->ingress.packets = 0;
+	tc_port_matchall->ingress.bytes = 0;
+	tc_port_matchall->ingress.last_used = 0;
+	tc_port_matchall->ingress.state = CXGB4_MATCHALL_STATE_DISABLED;
+	return 0;
+}
+
+int cxgb4_tc_matchall_replace(struct net_device *dev,
+			      struct tc_cls_matchall_offload *cls_matchall,
+			      bool ingress)
+{
+	struct netlink_ext_ack *extack = cls_matchall->common.extack;
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	int ret;
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+	if (ingress) {
+		if (tc_port_matchall->ingress.state ==
+		    CXGB4_MATCHALL_STATE_ENABLED) {
+			NL_SET_ERR_MSG_MOD(extack,
+					   "Only 1 Ingress MATCHALL can be offloaded");
+			return -ENOMEM;
+		}
+
+		ret = cxgb4_validate_flow_actions(dev,
+						  &cls_matchall->rule->action,
+						  extack, 1);
+		if (ret)
+			return ret;
+
+		return cxgb4_matchall_alloc_filter(dev, cls_matchall);
+	}
+
+	if (tc_port_matchall->egress.state == CXGB4_MATCHALL_STATE_ENABLED) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "Only 1 Egress MATCHALL can be offloaded");
+		return -ENOMEM;
+	}
+
+	ret = cxgb4_matchall_egress_validate(dev, cls_matchall);
+	if (ret)
+		return ret;
+
+	return cxgb4_matchall_alloc_tc(dev, cls_matchall);
+}
+
+int cxgb4_tc_matchall_destroy(struct net_device *dev,
+			      struct tc_cls_matchall_offload *cls_matchall,
+			      bool ingress)
+{
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+	if (ingress) {
+		/* All the filter types of this matchall rule save the
+		 * same cookie. So, checking for the first one is
+		 * enough.
+		 */
+		if (cls_matchall->cookie !=
+		    tc_port_matchall->ingress.fs[0].tc_cookie)
+			return -ENOENT;
+
+		return cxgb4_matchall_free_filter(dev);
+	}
+
+	if (cls_matchall->cookie != tc_port_matchall->egress.cookie)
+		return -ENOENT;
+
+	cxgb4_matchall_free_tc(dev);
+	return 0;
+}
+
+int cxgb4_tc_matchall_stats(struct net_device *dev,
+			    struct tc_cls_matchall_offload *cls_matchall)
+{
+	u64 tmp_packets, tmp_bytes, packets = 0, bytes = 0;
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct cxgb4_matchall_ingress_entry *ingress;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	int ret;
+	u8 i;
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+	if (tc_port_matchall->ingress.state == CXGB4_MATCHALL_STATE_DISABLED)
+		return -ENOENT;
+
+	ingress = &tc_port_matchall->ingress;
+	for (i = 0; i < CXGB4_FILTER_TYPE_MAX; i++) {
+		ret = cxgb4_get_filter_counters(dev, ingress->tid[i],
+						&tmp_packets, &tmp_bytes,
+						ingress->fs[i].hash);
+		if (ret)
+			return ret;
+
+		packets += tmp_packets;
+		bytes += tmp_bytes;
+	}
+
+	if (tc_port_matchall->ingress.packets != packets) {
+		flow_stats_update(&cls_matchall->stats,
+				  bytes - tc_port_matchall->ingress.bytes,
+				  packets - tc_port_matchall->ingress.packets,
+				  0, tc_port_matchall->ingress.last_used,
+				  FLOW_ACTION_HW_STATS_IMMEDIATE);
+
+		tc_port_matchall->ingress.packets = packets;
+		tc_port_matchall->ingress.bytes = bytes;
+		tc_port_matchall->ingress.last_used = jiffies;
+	}
+
+	return 0;
+}
+
+static void cxgb4_matchall_disable_offload(struct net_device *dev)
+{
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+
+	tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id];
+	if (tc_port_matchall->egress.state == CXGB4_MATCHALL_STATE_ENABLED)
+		cxgb4_matchall_free_tc(dev);
+
+	if (tc_port_matchall->ingress.state == CXGB4_MATCHALL_STATE_ENABLED)
+		cxgb4_matchall_free_filter(dev);
+}
+
+int cxgb4_init_tc_matchall(struct adapter *adap)
+{
+	struct cxgb4_tc_port_matchall *tc_port_matchall;
+	struct cxgb4_tc_matchall *tc_matchall;
+	int ret;
+
+	tc_matchall = kzalloc(sizeof(*tc_matchall), GFP_KERNEL);
+	if (!tc_matchall)
+		return -ENOMEM;
+
+	tc_port_matchall = kcalloc(adap->params.nports,
+				   sizeof(*tc_port_matchall),
+				   GFP_KERNEL);
+	if (!tc_port_matchall) {
+		ret = -ENOMEM;
+		goto out_free_matchall;
+	}
+
+	tc_matchall->port_matchall = tc_port_matchall;
+	adap->tc_matchall = tc_matchall;
+	return 0;
+
+out_free_matchall:
+	kfree(tc_matchall);
+	return ret;
+}
+
+void cxgb4_cleanup_tc_matchall(struct adapter *adap)
+{
+	u8 i;
+
+	if (adap->tc_matchall) {
+		if (adap->tc_matchall->port_matchall) {
+			for (i = 0; i < adap->params.nports; i++) {
+				struct net_device *dev = adap->port[i];
+
+				if (dev)
+					cxgb4_matchall_disable_offload(dev);
+			}
+			kfree(adap->tc_matchall->port_matchall);
+		}
+		kfree(adap->tc_matchall);
+	}
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.h
new file mode 100644
index 000000000..fe7ec423a
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.h
@@ -0,0 +1,51 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright (C) 2019 Chelsio Communications.  All rights reserved. */
+
+#ifndef __CXGB4_TC_MATCHALL_H__
+#define __CXGB4_TC_MATCHALL_H__
+
+#include <net/pkt_cls.h>
+
+enum cxgb4_matchall_state {
+	CXGB4_MATCHALL_STATE_DISABLED = 0,
+	CXGB4_MATCHALL_STATE_ENABLED,
+};
+
+struct cxgb4_matchall_egress_entry {
+	enum cxgb4_matchall_state state; /* Current MATCHALL offload state */
+	u8 hwtc; /* Traffic class bound to port */
+	u64 cookie; /* Used to identify the MATCHALL rule offloaded */
+};
+
+struct cxgb4_matchall_ingress_entry {
+	enum cxgb4_matchall_state state; /* Current MATCHALL offload state */
+	u32 tid[CXGB4_FILTER_TYPE_MAX]; /* Index to hardware filter entries */
+	/* Filter entries */
+	struct ch_filter_specification fs[CXGB4_FILTER_TYPE_MAX];
+	u16 viid_mirror; /* Identifier for allocated Mirror VI */
+	u64 bytes; /* # of bytes hitting the filter */
+	u64 packets; /* # of packets hitting the filter */
+	u64 last_used; /* Last updated jiffies time */
+};
+
+struct cxgb4_tc_port_matchall {
+	struct cxgb4_matchall_egress_entry egress; /* Egress offload info */
+	struct cxgb4_matchall_ingress_entry ingress; /* Ingress offload info */
+};
+
+struct cxgb4_tc_matchall {
+	struct cxgb4_tc_port_matchall *port_matchall; /* Per port entry */
+};
+
+int cxgb4_tc_matchall_replace(struct net_device *dev,
+			      struct tc_cls_matchall_offload *cls_matchall,
+			      bool ingress);
+int cxgb4_tc_matchall_destroy(struct net_device *dev,
+			      struct tc_cls_matchall_offload *cls_matchall,
+			      bool ingress);
+int cxgb4_tc_matchall_stats(struct net_device *dev,
+			    struct tc_cls_matchall_offload *cls_matchall);
+
+int cxgb4_init_tc_matchall(struct adapter *adap);
+void cxgb4_cleanup_tc_matchall(struct adapter *adap);
+#endif /* __CXGB4_TC_MATCHALL_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c
new file mode 100644
index 000000000..338b04f33
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c
@@ -0,0 +1,722 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2019 Chelsio Communications.  All rights reserved. */
+
+#include "cxgb4.h"
+#include "cxgb4_tc_mqprio.h"
+#include "sched.h"
+
+static int cxgb4_mqprio_validate(struct net_device *dev,
+				 struct tc_mqprio_qopt_offload *mqprio)
+{
+	u64 min_rate = 0, max_rate = 0, max_link_rate;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	u32 speed, qcount = 0, qoffset = 0;
+	u32 start_a, start_b, end_a, end_b;
+	int ret;
+	u8 i, j;
+
+	if (!mqprio->qopt.num_tc)
+		return 0;
+
+	if (mqprio->qopt.hw != TC_MQPRIO_HW_OFFLOAD_TCS) {
+		netdev_err(dev, "Only full TC hardware offload is supported\n");
+		return -EINVAL;
+	} else if (mqprio->mode != TC_MQPRIO_MODE_CHANNEL) {
+		netdev_err(dev, "Only channel mode offload is supported\n");
+		return -EINVAL;
+	} else if (mqprio->shaper != TC_MQPRIO_SHAPER_BW_RATE) {
+		netdev_err(dev,	"Only bandwidth rate shaper supported\n");
+		return -EINVAL;
+	} else if (mqprio->qopt.num_tc > adap->params.nsched_cls) {
+		netdev_err(dev,
+			   "Only %u traffic classes supported by hardware\n",
+			   adap->params.nsched_cls);
+		return -ERANGE;
+	}
+
+	ret = t4_get_link_params(pi, NULL, &speed, NULL);
+	if (ret) {
+		netdev_err(dev, "Failed to get link speed, ret: %d\n", ret);
+		return -EINVAL;
+	}
+
+	/* Convert from Mbps to bps */
+	max_link_rate = (u64)speed * 1000 * 1000;
+
+	for (i = 0; i < mqprio->qopt.num_tc; i++) {
+		qoffset = max_t(u16, mqprio->qopt.offset[i], qoffset);
+		qcount += mqprio->qopt.count[i];
+
+		start_a = mqprio->qopt.offset[i];
+		end_a = start_a + mqprio->qopt.count[i] - 1;
+		for (j = i + 1; j < mqprio->qopt.num_tc; j++) {
+			start_b = mqprio->qopt.offset[j];
+			end_b = start_b + mqprio->qopt.count[j] - 1;
+
+			/* If queue count is 0, then the traffic
+			 * belonging to this class will not use
+			 * ETHOFLD queues. So, no need to validate
+			 * further.
+			 */
+			if (!mqprio->qopt.count[i])
+				break;
+
+			if (!mqprio->qopt.count[j])
+				continue;
+
+			if (max_t(u32, start_a, start_b) <=
+			    min_t(u32, end_a, end_b)) {
+				netdev_err(dev,
+					   "Queues can't overlap across tc\n");
+				return -EINVAL;
+			}
+		}
+
+		/* Convert byte per second to bits per second */
+		min_rate += (mqprio->min_rate[i] * 8);
+		max_rate += (mqprio->max_rate[i] * 8);
+	}
+
+	if (qoffset >= adap->tids.neotids || qcount > adap->tids.neotids)
+		return -ENOMEM;
+
+	if (min_rate > max_link_rate || max_rate > max_link_rate) {
+		netdev_err(dev,
+			   "Total Min/Max (%llu/%llu) Rate > supported (%llu)\n",
+			   min_rate, max_rate, max_link_rate);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int cxgb4_init_eosw_txq(struct net_device *dev,
+			       struct sge_eosw_txq *eosw_txq,
+			       u32 eotid, u32 hwqid)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct tx_sw_desc *ring;
+
+	memset(eosw_txq, 0, sizeof(*eosw_txq));
+
+	ring = kcalloc(CXGB4_EOSW_TXQ_DEFAULT_DESC_NUM,
+		       sizeof(*ring), GFP_KERNEL);
+	if (!ring)
+		return -ENOMEM;
+
+	eosw_txq->desc = ring;
+	eosw_txq->ndesc = CXGB4_EOSW_TXQ_DEFAULT_DESC_NUM;
+	spin_lock_init(&eosw_txq->lock);
+	eosw_txq->state = CXGB4_EO_STATE_CLOSED;
+	eosw_txq->eotid = eotid;
+	eosw_txq->hwtid = adap->tids.eotid_base + eosw_txq->eotid;
+	eosw_txq->cred = adap->params.ofldq_wr_cred;
+	eosw_txq->hwqid = hwqid;
+	eosw_txq->netdev = dev;
+	tasklet_setup(&eosw_txq->qresume_tsk, cxgb4_ethofld_restart);
+	return 0;
+}
+
+static void cxgb4_clean_eosw_txq(struct net_device *dev,
+				 struct sge_eosw_txq *eosw_txq)
+{
+	struct adapter *adap = netdev2adap(dev);
+
+	cxgb4_eosw_txq_free_desc(adap, eosw_txq, eosw_txq->ndesc);
+	eosw_txq->pidx = 0;
+	eosw_txq->last_pidx = 0;
+	eosw_txq->cidx = 0;
+	eosw_txq->last_cidx = 0;
+	eosw_txq->flowc_idx = 0;
+	eosw_txq->inuse = 0;
+	eosw_txq->cred = adap->params.ofldq_wr_cred;
+	eosw_txq->ncompl = 0;
+	eosw_txq->last_compl = 0;
+	eosw_txq->state = CXGB4_EO_STATE_CLOSED;
+}
+
+static void cxgb4_free_eosw_txq(struct net_device *dev,
+				struct sge_eosw_txq *eosw_txq)
+{
+	spin_lock_bh(&eosw_txq->lock);
+	cxgb4_clean_eosw_txq(dev, eosw_txq);
+	kfree(eosw_txq->desc);
+	spin_unlock_bh(&eosw_txq->lock);
+	tasklet_kill(&eosw_txq->qresume_tsk);
+}
+
+static int cxgb4_mqprio_alloc_hw_resources(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct sge_ofld_rxq *eorxq;
+	struct sge_eohw_txq *eotxq;
+	int ret, msix = 0;
+	u32 i;
+
+	/* Allocate ETHOFLD hardware queue structures if not done already */
+	if (!refcount_read(&adap->tc_mqprio->refcnt)) {
+		adap->sge.eohw_rxq = kcalloc(adap->sge.eoqsets,
+					     sizeof(struct sge_ofld_rxq),
+					     GFP_KERNEL);
+		if (!adap->sge.eohw_rxq)
+			return -ENOMEM;
+
+		adap->sge.eohw_txq = kcalloc(adap->sge.eoqsets,
+					     sizeof(struct sge_eohw_txq),
+					     GFP_KERNEL);
+		if (!adap->sge.eohw_txq) {
+			kfree(adap->sge.eohw_rxq);
+			return -ENOMEM;
+		}
+
+		refcount_set(&adap->tc_mqprio->refcnt, 1);
+	} else {
+		refcount_inc(&adap->tc_mqprio->refcnt);
+	}
+
+	if (!(adap->flags & CXGB4_USING_MSIX))
+		msix = -((int)adap->sge.intrq.abs_id + 1);
+
+	for (i = 0; i < pi->nqsets; i++) {
+		eorxq = &adap->sge.eohw_rxq[pi->first_qset + i];
+		eotxq = &adap->sge.eohw_txq[pi->first_qset + i];
+
+		/* Allocate Rxqs for receiving ETHOFLD Tx completions */
+		if (msix >= 0) {
+			msix = cxgb4_get_msix_idx_from_bmap(adap);
+			if (msix < 0) {
+				ret = msix;
+				goto out_free_queues;
+			}
+
+			eorxq->msix = &adap->msix_info[msix];
+			snprintf(eorxq->msix->desc,
+				 sizeof(eorxq->msix->desc),
+				 "%s-eorxq%d", dev->name, i);
+		}
+
+		init_rspq(adap, &eorxq->rspq,
+			  CXGB4_EOHW_RXQ_DEFAULT_INTR_USEC,
+			  CXGB4_EOHW_RXQ_DEFAULT_PKT_CNT,
+			  CXGB4_EOHW_RXQ_DEFAULT_DESC_NUM,
+			  CXGB4_EOHW_RXQ_DEFAULT_DESC_SIZE);
+
+		eorxq->fl.size = CXGB4_EOHW_FLQ_DEFAULT_DESC_NUM;
+
+		ret = t4_sge_alloc_rxq(adap, &eorxq->rspq, false,
+				       dev, msix, &eorxq->fl,
+				       cxgb4_ethofld_rx_handler,
+				       NULL, 0);
+		if (ret)
+			goto out_free_queues;
+
+		/* Allocate ETHOFLD hardware Txqs */
+		eotxq->q.size = CXGB4_EOHW_TXQ_DEFAULT_DESC_NUM;
+		ret = t4_sge_alloc_ethofld_txq(adap, eotxq, dev,
+					       eorxq->rspq.cntxt_id);
+		if (ret)
+			goto out_free_queues;
+
+		/* Allocate IRQs, set IRQ affinity, and start Rx */
+		if (adap->flags & CXGB4_USING_MSIX) {
+			ret = request_irq(eorxq->msix->vec, t4_sge_intr_msix, 0,
+					  eorxq->msix->desc, &eorxq->rspq);
+			if (ret)
+				goto out_free_msix;
+
+			cxgb4_set_msix_aff(adap, eorxq->msix->vec,
+					   &eorxq->msix->aff_mask, i);
+		}
+
+		if (adap->flags & CXGB4_FULL_INIT_DONE)
+			cxgb4_enable_rx(adap, &eorxq->rspq);
+	}
+
+	return 0;
+
+out_free_msix:
+	while (i-- > 0) {
+		eorxq = &adap->sge.eohw_rxq[pi->first_qset + i];
+
+		if (adap->flags & CXGB4_FULL_INIT_DONE)
+			cxgb4_quiesce_rx(&eorxq->rspq);
+
+		if (adap->flags & CXGB4_USING_MSIX) {
+			cxgb4_clear_msix_aff(eorxq->msix->vec,
+					     eorxq->msix->aff_mask);
+			free_irq(eorxq->msix->vec, &eorxq->rspq);
+		}
+	}
+
+out_free_queues:
+	for (i = 0; i < pi->nqsets; i++) {
+		eorxq = &adap->sge.eohw_rxq[pi->first_qset + i];
+		eotxq = &adap->sge.eohw_txq[pi->first_qset + i];
+
+		if (eorxq->rspq.desc)
+			free_rspq_fl(adap, &eorxq->rspq, &eorxq->fl);
+		if (eorxq->msix)
+			cxgb4_free_msix_idx_in_bmap(adap, eorxq->msix->idx);
+		t4_sge_free_ethofld_txq(adap, eotxq);
+	}
+
+	if (refcount_dec_and_test(&adap->tc_mqprio->refcnt)) {
+		kfree(adap->sge.eohw_txq);
+		kfree(adap->sge.eohw_rxq);
+	}
+	return ret;
+}
+
+static void cxgb4_mqprio_free_hw_resources(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct sge_ofld_rxq *eorxq;
+	struct sge_eohw_txq *eotxq;
+	u32 i;
+
+	/* Return if no ETHOFLD structures have been allocated yet */
+	if (!refcount_read(&adap->tc_mqprio->refcnt))
+		return;
+
+	/* Return if no hardware queues have been allocated */
+	if (!adap->sge.eohw_rxq[pi->first_qset].rspq.desc)
+		return;
+
+	for (i = 0; i < pi->nqsets; i++) {
+		eorxq = &adap->sge.eohw_rxq[pi->first_qset + i];
+		eotxq = &adap->sge.eohw_txq[pi->first_qset + i];
+
+		/* Device removal path will already disable NAPI
+		 * before unregistering netdevice. So, only disable
+		 * NAPI if we're not in device removal path
+		 */
+		if (!(adap->flags & CXGB4_SHUTTING_DOWN))
+			cxgb4_quiesce_rx(&eorxq->rspq);
+
+		if (adap->flags & CXGB4_USING_MSIX) {
+			cxgb4_clear_msix_aff(eorxq->msix->vec,
+					     eorxq->msix->aff_mask);
+			free_irq(eorxq->msix->vec, &eorxq->rspq);
+			cxgb4_free_msix_idx_in_bmap(adap, eorxq->msix->idx);
+		}
+
+		free_rspq_fl(adap, &eorxq->rspq, &eorxq->fl);
+		t4_sge_free_ethofld_txq(adap, eotxq);
+	}
+
+	/* Free up ETHOFLD structures if there are no users */
+	if (refcount_dec_and_test(&adap->tc_mqprio->refcnt)) {
+		kfree(adap->sge.eohw_txq);
+		kfree(adap->sge.eohw_rxq);
+	}
+}
+
+static int cxgb4_mqprio_alloc_tc(struct net_device *dev,
+				 struct tc_mqprio_qopt_offload *mqprio)
+{
+	struct ch_sched_params p = {
+		.type = SCHED_CLASS_TYPE_PACKET,
+		.u.params.level = SCHED_CLASS_LEVEL_CL_RL,
+		.u.params.mode = SCHED_CLASS_MODE_FLOW,
+		.u.params.rateunit = SCHED_CLASS_RATEUNIT_BITS,
+		.u.params.ratemode = SCHED_CLASS_RATEMODE_ABS,
+		.u.params.class = SCHED_CLS_NONE,
+		.u.params.weight = 0,
+		.u.params.pktsize = dev->mtu,
+	};
+	struct cxgb4_tc_port_mqprio *tc_port_mqprio;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct sched_class *e;
+	int ret;
+	u8 i;
+
+	tc_port_mqprio = &adap->tc_mqprio->port_mqprio[pi->port_id];
+	p.u.params.channel = pi->tx_chan;
+	for (i = 0; i < mqprio->qopt.num_tc; i++) {
+		/* Convert from bytes per second to Kbps */
+		p.u.params.minrate = div_u64(mqprio->min_rate[i] * 8, 1000);
+		p.u.params.maxrate = div_u64(mqprio->max_rate[i] * 8, 1000);
+
+		/* Request larger burst buffer for smaller MTU, so
+		 * that hardware can work on more data per burst
+		 * cycle.
+		 */
+		if (dev->mtu <= ETH_DATA_LEN)
+			p.u.params.burstsize = 8 * dev->mtu;
+
+		e = cxgb4_sched_class_alloc(dev, &p);
+		if (!e) {
+			ret = -ENOMEM;
+			goto out_err;
+		}
+
+		tc_port_mqprio->tc_hwtc_map[i] = e->idx;
+	}
+
+	return 0;
+
+out_err:
+	while (i--)
+		cxgb4_sched_class_free(dev, tc_port_mqprio->tc_hwtc_map[i]);
+
+	return ret;
+}
+
+static void cxgb4_mqprio_free_tc(struct net_device *dev)
+{
+	struct cxgb4_tc_port_mqprio *tc_port_mqprio;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	u8 i;
+
+	tc_port_mqprio = &adap->tc_mqprio->port_mqprio[pi->port_id];
+	for (i = 0; i < tc_port_mqprio->mqprio.qopt.num_tc; i++)
+		cxgb4_sched_class_free(dev, tc_port_mqprio->tc_hwtc_map[i]);
+}
+
+static int cxgb4_mqprio_class_bind(struct net_device *dev,
+				   struct sge_eosw_txq *eosw_txq,
+				   u8 tc)
+{
+	struct ch_sched_flowc fe;
+	int ret;
+
+	init_completion(&eosw_txq->completion);
+
+	fe.tid = eosw_txq->eotid;
+	fe.class = tc;
+
+	ret = cxgb4_sched_class_bind(dev, &fe, SCHED_FLOWC);
+	if (ret)
+		return ret;
+
+	ret = wait_for_completion_timeout(&eosw_txq->completion,
+					  CXGB4_FLOWC_WAIT_TIMEOUT);
+	if (!ret)
+		return -ETIMEDOUT;
+
+	return 0;
+}
+
+static void cxgb4_mqprio_class_unbind(struct net_device *dev,
+				      struct sge_eosw_txq *eosw_txq,
+				      u8 tc)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct ch_sched_flowc fe;
+
+	/* If we're shutting down, interrupts are disabled and no completions
+	 * come back. So, skip waiting for completions in this scenario.
+	 */
+	if (!(adap->flags & CXGB4_SHUTTING_DOWN))
+		init_completion(&eosw_txq->completion);
+
+	fe.tid = eosw_txq->eotid;
+	fe.class = tc;
+	cxgb4_sched_class_unbind(dev, &fe, SCHED_FLOWC);
+
+	if (!(adap->flags & CXGB4_SHUTTING_DOWN))
+		wait_for_completion_timeout(&eosw_txq->completion,
+					    CXGB4_FLOWC_WAIT_TIMEOUT);
+}
+
+static int cxgb4_mqprio_enable_offload(struct net_device *dev,
+				       struct tc_mqprio_qopt_offload *mqprio)
+{
+	struct cxgb4_tc_port_mqprio *tc_port_mqprio;
+	u32 qoffset, qcount, tot_qcount, qid, hwqid;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct sge_eosw_txq *eosw_txq;
+	int eotid, ret;
+	u16 i, j;
+	u8 hwtc;
+
+	ret = cxgb4_mqprio_alloc_hw_resources(dev);
+	if (ret)
+		return -ENOMEM;
+
+	tc_port_mqprio = &adap->tc_mqprio->port_mqprio[pi->port_id];
+	for (i = 0; i < mqprio->qopt.num_tc; i++) {
+		qoffset = mqprio->qopt.offset[i];
+		qcount = mqprio->qopt.count[i];
+		for (j = 0; j < qcount; j++) {
+			eotid = cxgb4_get_free_eotid(&adap->tids);
+			if (eotid < 0) {
+				ret = -ENOMEM;
+				goto out_free_eotids;
+			}
+
+			qid = qoffset + j;
+			hwqid = pi->first_qset + (eotid % pi->nqsets);
+			eosw_txq = &tc_port_mqprio->eosw_txq[qid];
+			ret = cxgb4_init_eosw_txq(dev, eosw_txq,
+						  eotid, hwqid);
+			if (ret)
+				goto out_free_eotids;
+
+			cxgb4_alloc_eotid(&adap->tids, eotid, eosw_txq);
+
+			hwtc = tc_port_mqprio->tc_hwtc_map[i];
+			ret = cxgb4_mqprio_class_bind(dev, eosw_txq, hwtc);
+			if (ret)
+				goto out_free_eotids;
+		}
+	}
+
+	memcpy(&tc_port_mqprio->mqprio, mqprio,
+	       sizeof(struct tc_mqprio_qopt_offload));
+
+	/* Inform the stack about the configured tc params.
+	 *
+	 * Set the correct queue map. If no queue count has been
+	 * specified, then send the traffic through default NIC
+	 * queues; instead of ETHOFLD queues.
+	 */
+	ret = netdev_set_num_tc(dev, mqprio->qopt.num_tc);
+	if (ret)
+		goto out_free_eotids;
+
+	tot_qcount = pi->nqsets;
+	for (i = 0; i < mqprio->qopt.num_tc; i++) {
+		qcount = mqprio->qopt.count[i];
+		if (qcount) {
+			qoffset = mqprio->qopt.offset[i] + pi->nqsets;
+		} else {
+			qcount = pi->nqsets;
+			qoffset = 0;
+		}
+
+		ret = netdev_set_tc_queue(dev, i, qcount, qoffset);
+		if (ret)
+			goto out_reset_tc;
+
+		tot_qcount += mqprio->qopt.count[i];
+	}
+
+	ret = netif_set_real_num_tx_queues(dev, tot_qcount);
+	if (ret)
+		goto out_reset_tc;
+
+	tc_port_mqprio->state = CXGB4_MQPRIO_STATE_ACTIVE;
+	return 0;
+
+out_reset_tc:
+	netdev_reset_tc(dev);
+	i = mqprio->qopt.num_tc;
+
+out_free_eotids:
+	while (i-- > 0) {
+		qoffset = mqprio->qopt.offset[i];
+		qcount = mqprio->qopt.count[i];
+		for (j = 0; j < qcount; j++) {
+			eosw_txq = &tc_port_mqprio->eosw_txq[qoffset + j];
+
+			hwtc = tc_port_mqprio->tc_hwtc_map[i];
+			cxgb4_mqprio_class_unbind(dev, eosw_txq, hwtc);
+
+			cxgb4_free_eotid(&adap->tids, eosw_txq->eotid);
+			cxgb4_free_eosw_txq(dev, eosw_txq);
+		}
+	}
+
+	cxgb4_mqprio_free_hw_resources(dev);
+	return ret;
+}
+
+static void cxgb4_mqprio_disable_offload(struct net_device *dev)
+{
+	struct cxgb4_tc_port_mqprio *tc_port_mqprio;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct sge_eosw_txq *eosw_txq;
+	u32 qoffset, qcount;
+	u16 i, j;
+	u8 hwtc;
+
+	tc_port_mqprio = &adap->tc_mqprio->port_mqprio[pi->port_id];
+	if (tc_port_mqprio->state != CXGB4_MQPRIO_STATE_ACTIVE)
+		return;
+
+	netdev_reset_tc(dev);
+	netif_set_real_num_tx_queues(dev, pi->nqsets);
+
+	for (i = 0; i < tc_port_mqprio->mqprio.qopt.num_tc; i++) {
+		qoffset = tc_port_mqprio->mqprio.qopt.offset[i];
+		qcount = tc_port_mqprio->mqprio.qopt.count[i];
+		for (j = 0; j < qcount; j++) {
+			eosw_txq = &tc_port_mqprio->eosw_txq[qoffset + j];
+
+			hwtc = tc_port_mqprio->tc_hwtc_map[i];
+			cxgb4_mqprio_class_unbind(dev, eosw_txq, hwtc);
+
+			cxgb4_free_eotid(&adap->tids, eosw_txq->eotid);
+			cxgb4_free_eosw_txq(dev, eosw_txq);
+		}
+	}
+
+	cxgb4_mqprio_free_hw_resources(dev);
+
+	/* Free up the traffic classes */
+	cxgb4_mqprio_free_tc(dev);
+
+	memset(&tc_port_mqprio->mqprio, 0,
+	       sizeof(struct tc_mqprio_qopt_offload));
+
+	tc_port_mqprio->state = CXGB4_MQPRIO_STATE_DISABLED;
+}
+
+int cxgb4_setup_tc_mqprio(struct net_device *dev,
+			  struct tc_mqprio_qopt_offload *mqprio)
+{
+	struct adapter *adap = netdev2adap(dev);
+	bool needs_bring_up = false;
+	int ret;
+
+	ret = cxgb4_mqprio_validate(dev, mqprio);
+	if (ret)
+		return ret;
+
+	mutex_lock(&adap->tc_mqprio->mqprio_mutex);
+
+	/* To configure tc params, the current allocated EOTIDs must
+	 * be freed up. However, they can't be freed up if there's
+	 * traffic running on the interface. So, ensure interface is
+	 * down before configuring tc params.
+	 */
+	if (netif_running(dev)) {
+		netif_tx_stop_all_queues(dev);
+		netif_carrier_off(dev);
+		needs_bring_up = true;
+	}
+
+	cxgb4_mqprio_disable_offload(dev);
+
+	/* If requested for clear, then just return since resources are
+	 * already freed up by now.
+	 */
+	if (!mqprio->qopt.num_tc)
+		goto out;
+
+	/* Allocate free available traffic classes and configure
+	 * their rate parameters.
+	 */
+	ret = cxgb4_mqprio_alloc_tc(dev, mqprio);
+	if (ret)
+		goto out;
+
+	ret = cxgb4_mqprio_enable_offload(dev, mqprio);
+	if (ret) {
+		cxgb4_mqprio_free_tc(dev);
+		goto out;
+	}
+
+out:
+	if (needs_bring_up) {
+		netif_tx_start_all_queues(dev);
+		netif_carrier_on(dev);
+	}
+
+	mutex_unlock(&adap->tc_mqprio->mqprio_mutex);
+	return ret;
+}
+
+void cxgb4_mqprio_stop_offload(struct adapter *adap)
+{
+	struct cxgb4_tc_port_mqprio *tc_port_mqprio;
+	struct net_device *dev;
+	u8 i;
+
+	if (!adap->tc_mqprio || !adap->tc_mqprio->port_mqprio)
+		return;
+
+	mutex_lock(&adap->tc_mqprio->mqprio_mutex);
+	for_each_port(adap, i) {
+		dev = adap->port[i];
+		if (!dev)
+			continue;
+
+		tc_port_mqprio = &adap->tc_mqprio->port_mqprio[i];
+		if (!tc_port_mqprio->mqprio.qopt.num_tc)
+			continue;
+
+		cxgb4_mqprio_disable_offload(dev);
+	}
+	mutex_unlock(&adap->tc_mqprio->mqprio_mutex);
+}
+
+int cxgb4_init_tc_mqprio(struct adapter *adap)
+{
+	struct cxgb4_tc_port_mqprio *tc_port_mqprio, *port_mqprio;
+	struct cxgb4_tc_mqprio *tc_mqprio;
+	struct sge_eosw_txq *eosw_txq;
+	int ret = 0;
+	u8 i;
+
+	tc_mqprio = kzalloc(sizeof(*tc_mqprio), GFP_KERNEL);
+	if (!tc_mqprio)
+		return -ENOMEM;
+
+	tc_port_mqprio = kcalloc(adap->params.nports, sizeof(*tc_port_mqprio),
+				 GFP_KERNEL);
+	if (!tc_port_mqprio) {
+		ret = -ENOMEM;
+		goto out_free_mqprio;
+	}
+
+	mutex_init(&tc_mqprio->mqprio_mutex);
+
+	tc_mqprio->port_mqprio = tc_port_mqprio;
+	for (i = 0; i < adap->params.nports; i++) {
+		port_mqprio = &tc_mqprio->port_mqprio[i];
+		eosw_txq = kcalloc(adap->tids.neotids, sizeof(*eosw_txq),
+				   GFP_KERNEL);
+		if (!eosw_txq) {
+			ret = -ENOMEM;
+			goto out_free_ports;
+		}
+		port_mqprio->eosw_txq = eosw_txq;
+	}
+
+	adap->tc_mqprio = tc_mqprio;
+	refcount_set(&adap->tc_mqprio->refcnt, 0);
+	return 0;
+
+out_free_ports:
+	for (i = 0; i < adap->params.nports; i++) {
+		port_mqprio = &tc_mqprio->port_mqprio[i];
+		kfree(port_mqprio->eosw_txq);
+	}
+	kfree(tc_port_mqprio);
+
+out_free_mqprio:
+	kfree(tc_mqprio);
+	return ret;
+}
+
+void cxgb4_cleanup_tc_mqprio(struct adapter *adap)
+{
+	struct cxgb4_tc_port_mqprio *port_mqprio;
+	u8 i;
+
+	if (adap->tc_mqprio) {
+		mutex_lock(&adap->tc_mqprio->mqprio_mutex);
+		if (adap->tc_mqprio->port_mqprio) {
+			for (i = 0; i < adap->params.nports; i++) {
+				struct net_device *dev = adap->port[i];
+
+				if (dev)
+					cxgb4_mqprio_disable_offload(dev);
+				port_mqprio = &adap->tc_mqprio->port_mqprio[i];
+				kfree(port_mqprio->eosw_txq);
+			}
+			kfree(adap->tc_mqprio->port_mqprio);
+		}
+		mutex_unlock(&adap->tc_mqprio->mqprio_mutex);
+		kfree(adap->tc_mqprio);
+	}
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h
new file mode 100644
index 000000000..be96f1dc0
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h
@@ -0,0 +1,45 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright (C) 2019 Chelsio Communications.  All rights reserved. */
+
+#ifndef __CXGB4_TC_MQPRIO_H__
+#define __CXGB4_TC_MQPRIO_H__
+
+#include <net/pkt_cls.h>
+
+#define CXGB4_EOSW_TXQ_DEFAULT_DESC_NUM 128
+
+#define CXGB4_EOHW_TXQ_DEFAULT_DESC_NUM 1024
+
+#define CXGB4_EOHW_RXQ_DEFAULT_DESC_NUM 1024
+#define CXGB4_EOHW_RXQ_DEFAULT_DESC_SIZE 64
+#define CXGB4_EOHW_RXQ_DEFAULT_INTR_USEC 5
+#define CXGB4_EOHW_RXQ_DEFAULT_PKT_CNT 8
+
+#define CXGB4_EOHW_FLQ_DEFAULT_DESC_NUM 72
+
+#define CXGB4_FLOWC_WAIT_TIMEOUT (5 * HZ)
+
+enum cxgb4_mqprio_state {
+	CXGB4_MQPRIO_STATE_DISABLED = 0,
+	CXGB4_MQPRIO_STATE_ACTIVE,
+};
+
+struct cxgb4_tc_port_mqprio {
+	enum cxgb4_mqprio_state state; /* Current MQPRIO offload state */
+	struct tc_mqprio_qopt_offload mqprio; /* MQPRIO offload params */
+	struct sge_eosw_txq *eosw_txq; /* Netdev SW Tx queue array */
+	u8 tc_hwtc_map[TC_QOPT_MAX_QUEUE]; /* MQPRIO tc to hardware tc map */
+};
+
+struct cxgb4_tc_mqprio {
+	refcount_t refcnt; /* Refcount for adapter-wide resources */
+	struct mutex mqprio_mutex; /* Lock for accessing MQPRIO info */
+	struct cxgb4_tc_port_mqprio *port_mqprio; /* Per port MQPRIO info */
+};
+
+int cxgb4_setup_tc_mqprio(struct net_device *dev,
+			  struct tc_mqprio_qopt_offload *mqprio);
+void cxgb4_mqprio_stop_offload(struct adapter *adap);
+int cxgb4_init_tc_mqprio(struct adapter *adap);
+void cxgb4_cleanup_tc_mqprio(struct adapter *adap);
+#endif /* __CXGB4_TC_MQPRIO_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.c
new file mode 100644
index 000000000..a5d2f84dc
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.c
@@ -0,0 +1,532 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <net/tc_act/tc_gact.h>
+#include <net/tc_act/tc_mirred.h>
+
+#include "cxgb4.h"
+#include "cxgb4_filter.h"
+#include "cxgb4_tc_u32_parse.h"
+#include "cxgb4_tc_u32.h"
+
+/* Fill ch_filter_specification with parsed match value/mask pair. */
+static int fill_match_fields(struct adapter *adap,
+			     struct ch_filter_specification *fs,
+			     struct tc_cls_u32_offload *cls,
+			     const struct cxgb4_match_field *entry,
+			     bool next_header)
+{
+	unsigned int i, j;
+	__be32 val, mask;
+	int off, err;
+	bool found;
+
+	for (i = 0; i < cls->knode.sel->nkeys; i++) {
+		off = cls->knode.sel->keys[i].off;
+		val = cls->knode.sel->keys[i].val;
+		mask = cls->knode.sel->keys[i].mask;
+
+		if (next_header) {
+			/* For next headers, parse only keys with offmask */
+			if (!cls->knode.sel->keys[i].offmask)
+				continue;
+		} else {
+			/* For the remaining, parse only keys without offmask */
+			if (cls->knode.sel->keys[i].offmask)
+				continue;
+		}
+
+		found = false;
+
+		for (j = 0; entry[j].val; j++) {
+			if (off == entry[j].off) {
+				found = true;
+				err = entry[j].val(fs, val, mask);
+				if (err)
+					return err;
+				break;
+			}
+		}
+
+		if (!found)
+			return -EINVAL;
+	}
+
+	return 0;
+}
+
+/* Fill ch_filter_specification with parsed action. */
+static int fill_action_fields(struct adapter *adap,
+			      struct ch_filter_specification *fs,
+			      struct tc_cls_u32_offload *cls)
+{
+	unsigned int num_actions = 0;
+	const struct tc_action *a;
+	struct tcf_exts *exts;
+	int i;
+
+	exts = cls->knode.exts;
+	if (!tcf_exts_has_actions(exts))
+		return -EINVAL;
+
+	tcf_exts_for_each_action(i, a, exts) {
+		/* Don't allow more than one action per rule. */
+		if (num_actions)
+			return -EINVAL;
+
+		/* Drop in hardware. */
+		if (is_tcf_gact_shot(a)) {
+			fs->action = FILTER_DROP;
+			num_actions++;
+			continue;
+		}
+
+		/* Re-direct to specified port in hardware. */
+		if (is_tcf_mirred_egress_redirect(a)) {
+			struct net_device *n_dev, *target_dev;
+			bool found = false;
+			unsigned int i;
+
+			target_dev = tcf_mirred_dev(a);
+			for_each_port(adap, i) {
+				n_dev = adap->port[i];
+				if (target_dev == n_dev) {
+					fs->action = FILTER_SWITCH;
+					fs->eport = i;
+					found = true;
+					break;
+				}
+			}
+
+			/* Interface doesn't belong to any port of
+			 * the underlying hardware.
+			 */
+			if (!found)
+				return -EINVAL;
+
+			num_actions++;
+			continue;
+		}
+
+		/* Un-supported action. */
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+int cxgb4_config_knode(struct net_device *dev, struct tc_cls_u32_offload *cls)
+{
+	const struct cxgb4_match_field *start, *link_start = NULL;
+	struct netlink_ext_ack *extack = cls->common.extack;
+	struct adapter *adapter = netdev2adap(dev);
+	__be16 protocol = cls->common.protocol;
+	struct ch_filter_specification fs;
+	struct cxgb4_tc_u32_table *t;
+	struct cxgb4_link *link;
+	u32 uhtid, link_uhtid;
+	bool is_ipv6 = false;
+	u8 inet_family;
+	int filter_id;
+	int ret;
+
+	if (!can_tc_u32_offload(dev))
+		return -EOPNOTSUPP;
+
+	if (protocol != htons(ETH_P_IP) && protocol != htons(ETH_P_IPV6))
+		return -EOPNOTSUPP;
+
+	inet_family = (protocol == htons(ETH_P_IPV6)) ? PF_INET6 : PF_INET;
+
+	/* Get a free filter entry TID, where we can insert this new
+	 * rule. Only insert rule if its prio doesn't conflict with
+	 * existing rules.
+	 */
+	filter_id = cxgb4_get_free_ftid(dev, inet_family, false,
+					TC_U32_NODE(cls->knode.handle));
+	if (filter_id < 0) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "No free LETCAM index available");
+		return -ENOMEM;
+	}
+
+	t = adapter->tc_u32;
+	uhtid = TC_U32_USERHTID(cls->knode.handle);
+	link_uhtid = TC_U32_USERHTID(cls->knode.link_handle);
+
+	/* Ensure that uhtid is either root u32 (i.e. 0x800)
+	 * or a a valid linked bucket.
+	 */
+	if (uhtid != 0x800 && uhtid >= t->size)
+		return -EINVAL;
+
+	/* Ensure link handle uhtid is sane, if specified. */
+	if (link_uhtid >= t->size)
+		return -EINVAL;
+
+	memset(&fs, 0, sizeof(fs));
+
+	if (filter_id < adapter->tids.nhpftids)
+		fs.prio = 1;
+	fs.tc_prio = cls->common.prio;
+	fs.tc_cookie = cls->knode.handle;
+
+	if (protocol == htons(ETH_P_IPV6)) {
+		start = cxgb4_ipv6_fields;
+		is_ipv6 = true;
+	} else {
+		start = cxgb4_ipv4_fields;
+		is_ipv6 = false;
+	}
+
+	if (uhtid != 0x800) {
+		/* Link must exist from root node before insertion. */
+		if (!t->table[uhtid - 1].link_handle)
+			return -EINVAL;
+
+		/* Link must have a valid supported next header. */
+		link_start = t->table[uhtid - 1].match_field;
+		if (!link_start)
+			return -EINVAL;
+	}
+
+	/* Parse links and record them for subsequent jumps to valid
+	 * next headers.
+	 */
+	if (link_uhtid) {
+		const struct cxgb4_next_header *next;
+		bool found = false;
+		unsigned int i, j;
+		__be32 val, mask;
+		int off;
+
+		if (t->table[link_uhtid - 1].link_handle) {
+			dev_err(adapter->pdev_dev,
+				"Link handle exists for: 0x%x\n",
+				link_uhtid);
+			return -EINVAL;
+		}
+
+		next = is_ipv6 ? cxgb4_ipv6_jumps : cxgb4_ipv4_jumps;
+
+		/* Try to find matches that allow jumps to next header. */
+		for (i = 0; next[i].jump; i++) {
+			if (next[i].sel.offoff != cls->knode.sel->offoff ||
+			    next[i].sel.offshift != cls->knode.sel->offshift ||
+			    next[i].sel.offmask != cls->knode.sel->offmask ||
+			    next[i].sel.off != cls->knode.sel->off)
+				continue;
+
+			/* Found a possible candidate.  Find a key that
+			 * matches the corresponding offset, value, and
+			 * mask to jump to next header.
+			 */
+			for (j = 0; j < cls->knode.sel->nkeys; j++) {
+				off = cls->knode.sel->keys[j].off;
+				val = cls->knode.sel->keys[j].val;
+				mask = cls->knode.sel->keys[j].mask;
+
+				if (next[i].key.off == off &&
+				    next[i].key.val == val &&
+				    next[i].key.mask == mask) {
+					found = true;
+					break;
+				}
+			}
+
+			if (!found)
+				continue; /* Try next candidate. */
+
+			/* Candidate to jump to next header found.
+			 * Translate all keys to internal specification
+			 * and store them in jump table. This spec is copied
+			 * later to set the actual filters.
+			 */
+			ret = fill_match_fields(adapter, &fs, cls,
+						start, false);
+			if (ret)
+				goto out;
+
+			link = &t->table[link_uhtid - 1];
+			link->match_field = next[i].jump;
+			link->link_handle = cls->knode.handle;
+			memcpy(&link->fs, &fs, sizeof(fs));
+			break;
+		}
+
+		/* No candidate found to jump to next header. */
+		if (!found)
+			return -EINVAL;
+
+		return 0;
+	}
+
+	/* Fill ch_filter_specification match fields to be shipped to hardware.
+	 * Copy the linked spec (if any) first.  And then update the spec as
+	 * needed.
+	 */
+	if (uhtid != 0x800 && t->table[uhtid - 1].link_handle) {
+		/* Copy linked ch_filter_specification */
+		memcpy(&fs, &t->table[uhtid - 1].fs, sizeof(fs));
+		ret = fill_match_fields(adapter, &fs, cls,
+					link_start, true);
+		if (ret)
+			goto out;
+	}
+
+	ret = fill_match_fields(adapter, &fs, cls, start, false);
+	if (ret)
+		goto out;
+
+	/* Fill ch_filter_specification action fields to be shipped to
+	 * hardware.
+	 */
+	ret = fill_action_fields(adapter, &fs, cls);
+	if (ret)
+		goto out;
+
+	/* The filter spec has been completely built from the info
+	 * provided from u32.  We now set some default fields in the
+	 * spec for sanity.
+	 */
+
+	/* Match only packets coming from the ingress port where this
+	 * filter will be created.
+	 */
+	fs.val.iport = netdev2pinfo(dev)->port_id;
+	fs.mask.iport = ~0;
+
+	/* Enable filter hit counts. */
+	fs.hitcnts = 1;
+
+	/* Set type of filter - IPv6 or IPv4 */
+	fs.type = is_ipv6 ? 1 : 0;
+
+	/* Set the filter */
+	ret = cxgb4_set_filter(dev, filter_id, &fs);
+	if (ret)
+		goto out;
+
+	/* If this is a linked bucket, then set the corresponding
+	 * entry in the bitmap to mark it as belonging to this linked
+	 * bucket.
+	 */
+	if (uhtid != 0x800 && t->table[uhtid - 1].link_handle)
+		set_bit(filter_id, t->table[uhtid - 1].tid_map);
+
+out:
+	return ret;
+}
+
+int cxgb4_delete_knode(struct net_device *dev, struct tc_cls_u32_offload *cls)
+{
+	struct adapter *adapter = netdev2adap(dev);
+	unsigned int filter_id, max_tids, i, j;
+	struct cxgb4_link *link = NULL;
+	struct cxgb4_tc_u32_table *t;
+	struct filter_entry *f;
+	bool found = false;
+	u32 handle, uhtid;
+	u8 nslots;
+	int ret;
+
+	if (!can_tc_u32_offload(dev))
+		return -EOPNOTSUPP;
+
+	/* Fetch the location to delete the filter. */
+	max_tids = adapter->tids.nhpftids + adapter->tids.nftids;
+
+	spin_lock_bh(&adapter->tids.ftid_lock);
+	filter_id = 0;
+	while (filter_id < max_tids) {
+		if (filter_id < adapter->tids.nhpftids) {
+			i = filter_id;
+			f = &adapter->tids.hpftid_tab[i];
+			if (f->valid && f->fs.tc_cookie == cls->knode.handle) {
+				found = true;
+				break;
+			}
+
+			i = find_next_bit(adapter->tids.hpftid_bmap,
+					  adapter->tids.nhpftids, i + 1);
+			if (i >= adapter->tids.nhpftids) {
+				filter_id = adapter->tids.nhpftids;
+				continue;
+			}
+
+			filter_id = i;
+		} else {
+			i = filter_id - adapter->tids.nhpftids;
+			f = &adapter->tids.ftid_tab[i];
+			if (f->valid && f->fs.tc_cookie == cls->knode.handle) {
+				found = true;
+				break;
+			}
+
+			i = find_next_bit(adapter->tids.ftid_bmap,
+					  adapter->tids.nftids, i + 1);
+			if (i >= adapter->tids.nftids)
+				break;
+
+			filter_id = i + adapter->tids.nhpftids;
+		}
+
+		nslots = 0;
+		if (f->fs.type) {
+			nslots++;
+			if (CHELSIO_CHIP_VERSION(adapter->params.chip) <
+			    CHELSIO_T6)
+				nslots += 2;
+		}
+
+		filter_id += nslots;
+	}
+	spin_unlock_bh(&adapter->tids.ftid_lock);
+
+	if (!found)
+		return -ERANGE;
+
+	t = adapter->tc_u32;
+	handle = cls->knode.handle;
+	uhtid = TC_U32_USERHTID(cls->knode.handle);
+
+	/* Ensure that uhtid is either root u32 (i.e. 0x800)
+	 * or a a valid linked bucket.
+	 */
+	if (uhtid != 0x800 && uhtid >= t->size)
+		return -EINVAL;
+
+	/* Delete the specified filter */
+	if (uhtid != 0x800) {
+		link = &t->table[uhtid - 1];
+		if (!link->link_handle)
+			return -EINVAL;
+
+		if (!test_bit(filter_id, link->tid_map))
+			return -EINVAL;
+	}
+
+	ret = cxgb4_del_filter(dev, filter_id, NULL);
+	if (ret)
+		goto out;
+
+	if (link)
+		clear_bit(filter_id, link->tid_map);
+
+	/* If a link is being deleted, then delete all filters
+	 * associated with the link.
+	 */
+	for (i = 0; i < t->size; i++) {
+		link = &t->table[i];
+
+		if (link->link_handle == handle) {
+			for (j = 0; j < max_tids; j++) {
+				if (!test_bit(j, link->tid_map))
+					continue;
+
+				ret = __cxgb4_del_filter(dev, j, NULL, NULL);
+				if (ret)
+					goto out;
+
+				clear_bit(j, link->tid_map);
+			}
+
+			/* Clear the link state */
+			link->match_field = NULL;
+			link->link_handle = 0;
+			memset(&link->fs, 0, sizeof(link->fs));
+			break;
+		}
+	}
+
+out:
+	return ret;
+}
+
+void cxgb4_cleanup_tc_u32(struct adapter *adap)
+{
+	struct cxgb4_tc_u32_table *t;
+	unsigned int i;
+
+	if (!adap->tc_u32)
+		return;
+
+	/* Free up all allocated memory. */
+	t = adap->tc_u32;
+	for (i = 0; i < t->size; i++) {
+		struct cxgb4_link *link = &t->table[i];
+
+		kvfree(link->tid_map);
+	}
+	kvfree(adap->tc_u32);
+}
+
+struct cxgb4_tc_u32_table *cxgb4_init_tc_u32(struct adapter *adap)
+{
+	unsigned int max_tids = adap->tids.nftids + adap->tids.nhpftids;
+	struct cxgb4_tc_u32_table *t;
+	unsigned int i;
+
+	if (!max_tids)
+		return NULL;
+
+	t = kvzalloc(struct_size(t, table, max_tids), GFP_KERNEL);
+	if (!t)
+		return NULL;
+
+	t->size = max_tids;
+
+	for (i = 0; i < t->size; i++) {
+		struct cxgb4_link *link = &t->table[i];
+		unsigned int bmap_size;
+
+		bmap_size = BITS_TO_LONGS(max_tids);
+		link->tid_map = kvcalloc(bmap_size, sizeof(unsigned long),
+					 GFP_KERNEL);
+		if (!link->tid_map)
+			goto out_no_mem;
+		bitmap_zero(link->tid_map, max_tids);
+	}
+
+	return t;
+
+out_no_mem:
+	for (i = 0; i < t->size; i++) {
+		struct cxgb4_link *link = &t->table[i];
+		kvfree(link->tid_map);
+	}
+	kvfree(t);
+
+	return NULL;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.h
new file mode 100644
index 000000000..70a07b7cc
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.h
@@ -0,0 +1,52 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_TC_U32_H
+#define __CXGB4_TC_U32_H
+
+#include <net/pkt_cls.h>
+
+static inline bool can_tc_u32_offload(struct net_device *dev)
+{
+	struct adapter *adap = netdev2adap(dev);
+
+	return (dev->features & NETIF_F_HW_TC) && adap->tc_u32 ? true : false;
+}
+
+int cxgb4_config_knode(struct net_device *dev, struct tc_cls_u32_offload *cls);
+int cxgb4_delete_knode(struct net_device *dev, struct tc_cls_u32_offload *cls);
+
+void cxgb4_cleanup_tc_u32(struct adapter *adapter);
+struct cxgb4_tc_u32_table *cxgb4_init_tc_u32(struct adapter *adap);
+#endif /* __CXGB4_TC_U32_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32_parse.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32_parse.h
new file mode 100644
index 000000000..f59dd4b2a
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32_parse.h
@@ -0,0 +1,336 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_TC_U32_PARSE_H
+#define __CXGB4_TC_U32_PARSE_H
+
+struct cxgb4_match_field {
+	int off; /* Offset from the beginning of the header to match */
+	/* Fill the value/mask pair in the spec if matched */
+	int (*val)(struct ch_filter_specification *f, __be32 val, __be32 mask);
+};
+
+/* IPv4 match fields */
+static inline int cxgb4_fill_ipv4_tos(struct ch_filter_specification *f,
+				      __be32 val, __be32 mask)
+{
+	f->val.tos  = (ntohl(val)  >> 16) & 0x000000FF;
+	f->mask.tos = (ntohl(mask) >> 16) & 0x000000FF;
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv4_frag(struct ch_filter_specification *f,
+				       __be32 val, __be32 mask)
+{
+	u32 mask_val;
+	u8 frag_val;
+
+	frag_val = (ntohl(val) >> 13) & 0x00000007;
+	mask_val = ntohl(mask) & 0x0000FFFF;
+
+	if (frag_val == 0x1 && mask_val != 0x3FFF) { /* MF set */
+		f->val.frag = 1;
+		f->mask.frag = 1;
+	} else if (frag_val == 0x2 && mask_val != 0x3FFF) { /* DF set */
+		f->val.frag = 0;
+		f->mask.frag = 1;
+	} else {
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv4_proto(struct ch_filter_specification *f,
+					__be32 val, __be32 mask)
+{
+	f->val.proto  = (ntohl(val)  >> 16) & 0x000000FF;
+	f->mask.proto = (ntohl(mask) >> 16) & 0x000000FF;
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv4_src_ip(struct ch_filter_specification *f,
+					 __be32 val, __be32 mask)
+{
+	memcpy(&f->val.fip[0],  &val,  sizeof(u32));
+	memcpy(&f->mask.fip[0], &mask, sizeof(u32));
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv4_dst_ip(struct ch_filter_specification *f,
+					 __be32 val, __be32 mask)
+{
+	memcpy(&f->val.lip[0],  &val,  sizeof(u32));
+	memcpy(&f->mask.lip[0], &mask, sizeof(u32));
+
+	return 0;
+}
+
+static const struct cxgb4_match_field cxgb4_ipv4_fields[] = {
+	{ .off = 0,  .val = cxgb4_fill_ipv4_tos },
+	{ .off = 4,  .val = cxgb4_fill_ipv4_frag },
+	{ .off = 8,  .val = cxgb4_fill_ipv4_proto },
+	{ .off = 12, .val = cxgb4_fill_ipv4_src_ip },
+	{ .off = 16, .val = cxgb4_fill_ipv4_dst_ip },
+	{ .val = NULL }
+};
+
+/* IPv6 match fields */
+static inline int cxgb4_fill_ipv6_tos(struct ch_filter_specification *f,
+				      __be32 val, __be32 mask)
+{
+	f->val.tos  = (ntohl(val)  >> 20) & 0x000000FF;
+	f->mask.tos = (ntohl(mask) >> 20) & 0x000000FF;
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv6_proto(struct ch_filter_specification *f,
+					__be32 val, __be32 mask)
+{
+	f->val.proto  = (ntohl(val)  >> 8) & 0x000000FF;
+	f->mask.proto = (ntohl(mask) >> 8) & 0x000000FF;
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv6_src_ip0(struct ch_filter_specification *f,
+					  __be32 val, __be32 mask)
+{
+	memcpy(&f->val.fip[0],  &val,  sizeof(u32));
+	memcpy(&f->mask.fip[0], &mask, sizeof(u32));
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv6_src_ip1(struct ch_filter_specification *f,
+					  __be32 val, __be32 mask)
+{
+	memcpy(&f->val.fip[4],  &val,  sizeof(u32));
+	memcpy(&f->mask.fip[4], &mask, sizeof(u32));
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv6_src_ip2(struct ch_filter_specification *f,
+					  __be32 val, __be32 mask)
+{
+	memcpy(&f->val.fip[8],  &val,  sizeof(u32));
+	memcpy(&f->mask.fip[8], &mask, sizeof(u32));
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv6_src_ip3(struct ch_filter_specification *f,
+					  __be32 val, __be32 mask)
+{
+	memcpy(&f->val.fip[12],  &val,  sizeof(u32));
+	memcpy(&f->mask.fip[12], &mask, sizeof(u32));
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv6_dst_ip0(struct ch_filter_specification *f,
+					  __be32 val, __be32 mask)
+{
+	memcpy(&f->val.lip[0],  &val,  sizeof(u32));
+	memcpy(&f->mask.lip[0], &mask, sizeof(u32));
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv6_dst_ip1(struct ch_filter_specification *f,
+					  __be32 val, __be32 mask)
+{
+	memcpy(&f->val.lip[4],  &val,  sizeof(u32));
+	memcpy(&f->mask.lip[4], &mask, sizeof(u32));
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv6_dst_ip2(struct ch_filter_specification *f,
+					  __be32 val, __be32 mask)
+{
+	memcpy(&f->val.lip[8],  &val,  sizeof(u32));
+	memcpy(&f->mask.lip[8], &mask, sizeof(u32));
+
+	return 0;
+}
+
+static inline int cxgb4_fill_ipv6_dst_ip3(struct ch_filter_specification *f,
+					  __be32 val, __be32 mask)
+{
+	memcpy(&f->val.lip[12],  &val,  sizeof(u32));
+	memcpy(&f->mask.lip[12], &mask, sizeof(u32));
+
+	return 0;
+}
+
+static const struct cxgb4_match_field cxgb4_ipv6_fields[] = {
+	{ .off = 0,  .val = cxgb4_fill_ipv6_tos },
+	{ .off = 4,  .val = cxgb4_fill_ipv6_proto },
+	{ .off = 8,  .val = cxgb4_fill_ipv6_src_ip0 },
+	{ .off = 12, .val = cxgb4_fill_ipv6_src_ip1 },
+	{ .off = 16, .val = cxgb4_fill_ipv6_src_ip2 },
+	{ .off = 20, .val = cxgb4_fill_ipv6_src_ip3 },
+	{ .off = 24, .val = cxgb4_fill_ipv6_dst_ip0 },
+	{ .off = 28, .val = cxgb4_fill_ipv6_dst_ip1 },
+	{ .off = 32, .val = cxgb4_fill_ipv6_dst_ip2 },
+	{ .off = 36, .val = cxgb4_fill_ipv6_dst_ip3 },
+	{ .val = NULL }
+};
+
+/* TCP/UDP match */
+static inline int cxgb4_fill_l4_ports(struct ch_filter_specification *f,
+				      __be32 val, __be32 mask)
+{
+	f->val.fport  = ntohl(val)  >> 16;
+	f->mask.fport = ntohl(mask) >> 16;
+	f->val.lport  = ntohl(val)  & 0x0000FFFF;
+	f->mask.lport = ntohl(mask) & 0x0000FFFF;
+
+	return 0;
+};
+
+static const struct cxgb4_match_field cxgb4_tcp_fields[] = {
+	{ .off = 0, .val = cxgb4_fill_l4_ports },
+	{ .val = NULL }
+};
+
+static const struct cxgb4_match_field cxgb4_udp_fields[] = {
+	{ .off = 0, .val = cxgb4_fill_l4_ports },
+	{ .val = NULL }
+};
+
+struct cxgb4_next_header {
+	/* Offset, shift, and mask added to beginning of the header
+	 * to get to next header.  Useful when using a header
+	 * field's value to jump to next header such as IHL field
+	 * in IPv4 header.
+	 */
+	struct tc_u32_sel sel;
+	struct tc_u32_key key;
+	/* location of jump to make */
+	const struct cxgb4_match_field *jump;
+};
+
+/* Accept a rule with a jump to transport layer header based on IHL field in
+ * IPv4 header.
+ */
+static const struct cxgb4_next_header cxgb4_ipv4_jumps[] = {
+	{
+		/* TCP Jump */
+		.sel = {
+			.off = 0,
+			.offoff = 0,
+			.offshift = 6,
+			.offmask = cpu_to_be16(0x0f00),
+		},
+		.key = {
+			.off = 8,
+			.val = cpu_to_be32(0x00060000),
+			.mask = cpu_to_be32(0x00ff0000),
+		},
+		.jump = cxgb4_tcp_fields,
+	},
+	{
+		/* UDP Jump */
+		.sel = {
+			.off = 0,
+			.offoff = 0,
+			.offshift = 6,
+			.offmask = cpu_to_be16(0x0f00),
+		},
+		.key = {
+			.off = 8,
+			.val = cpu_to_be32(0x00110000),
+			.mask = cpu_to_be32(0x00ff0000),
+		},
+		.jump = cxgb4_udp_fields,
+	},
+	{ .jump = NULL },
+};
+
+/* Accept a rule with a jump directly past the 40 Bytes of IPv6 fixed header
+ * to get to transport layer header.
+ */
+static const struct cxgb4_next_header cxgb4_ipv6_jumps[] = {
+	{
+		/* TCP Jump */
+		.sel = {
+			.off = 40,
+			.offoff = 0,
+			.offshift = 0,
+			.offmask = 0,
+		},
+		.key = {
+			.off = 4,
+			.val = cpu_to_be32(0x00000600),
+			.mask = cpu_to_be32(0x0000ff00),
+		},
+		.jump = cxgb4_tcp_fields,
+	},
+	{
+		/* UDP Jump */
+		.sel = {
+			.off = 40,
+			.offoff = 0,
+			.offshift = 0,
+			.offmask = 0,
+		},
+		.key = {
+			.off = 4,
+			.val = cpu_to_be32(0x00001100),
+			.mask = cpu_to_be32(0x0000ff00),
+		},
+		.jump = cxgb4_udp_fields,
+	},
+	{ .jump = NULL },
+};
+
+struct cxgb4_link {
+	const struct cxgb4_match_field *match_field;  /* Next header */
+	struct ch_filter_specification fs; /* Match spec associated with link */
+	u32 link_handle;         /* Knode handle associated with the link */
+	unsigned long *tid_map;  /* Bitmap for filter tids */
+};
+
+struct cxgb4_tc_u32_table {
+	unsigned int size;          /* number of entries in table */
+	struct cxgb4_link table[]; /* Jump table */
+};
+#endif /* __CXGB4_TC_U32_PARSE_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_thermal.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_thermal.c
new file mode 100644
index 000000000..9a6d65243
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_thermal.c
@@ -0,0 +1,115 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  Copyright (C) 2017 Chelsio Communications.  All rights reserved.
+ *
+ *  Written by: Ganesh Goudar (ganeshgr@chelsio.com)
+ */
+
+#include "cxgb4.h"
+
+#define CXGB4_NUM_TRIPS 1
+
+static int cxgb4_thermal_get_temp(struct thermal_zone_device *tzdev,
+				  int *temp)
+{
+	struct adapter *adap = tzdev->devdata;
+	u32 param, val;
+	int ret;
+
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_DIAG) |
+		 FW_PARAMS_PARAM_Y_V(FW_PARAM_DEV_DIAG_TMP));
+
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 1,
+			      &param, &val);
+	if (ret < 0 || val == 0)
+		return -1;
+
+	*temp = val * 1000;
+	return 0;
+}
+
+static int cxgb4_thermal_get_trip_type(struct thermal_zone_device *tzdev,
+				       int trip, enum thermal_trip_type *type)
+{
+	struct adapter *adap = tzdev->devdata;
+
+	if (!adap->ch_thermal.trip_temp)
+		return -EINVAL;
+
+	*type = adap->ch_thermal.trip_type;
+	return 0;
+}
+
+static int cxgb4_thermal_get_trip_temp(struct thermal_zone_device *tzdev,
+				       int trip, int *temp)
+{
+	struct adapter *adap = tzdev->devdata;
+
+	if (!adap->ch_thermal.trip_temp)
+		return -EINVAL;
+
+	*temp = adap->ch_thermal.trip_temp;
+	return 0;
+}
+
+static struct thermal_zone_device_ops cxgb4_thermal_ops = {
+	.get_temp = cxgb4_thermal_get_temp,
+	.get_trip_type = cxgb4_thermal_get_trip_type,
+	.get_trip_temp = cxgb4_thermal_get_trip_temp,
+};
+
+int cxgb4_thermal_init(struct adapter *adap)
+{
+	struct ch_thermal *ch_thermal = &adap->ch_thermal;
+	char ch_tz_name[THERMAL_NAME_LENGTH];
+	int num_trip = CXGB4_NUM_TRIPS;
+	u32 param, val;
+	int ret;
+
+	/* on older firmwares we may not get the trip temperature,
+	 * set the num of trips to 0.
+	 */
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_DIAG) |
+		 FW_PARAMS_PARAM_Y_V(FW_PARAM_DEV_DIAG_MAXTMPTHRESH));
+
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 1,
+			      &param, &val);
+	if (ret < 0) {
+		num_trip = 0; /* could not get trip temperature */
+	} else {
+		ch_thermal->trip_temp = val * 1000;
+		ch_thermal->trip_type = THERMAL_TRIP_CRITICAL;
+	}
+
+	snprintf(ch_tz_name, sizeof(ch_tz_name), "cxgb4_%s", adap->name);
+	ch_thermal->tzdev = thermal_zone_device_register(ch_tz_name, num_trip,
+							 0, adap,
+							 &cxgb4_thermal_ops,
+							 NULL, 0, 0);
+	if (IS_ERR(ch_thermal->tzdev)) {
+		ret = PTR_ERR(ch_thermal->tzdev);
+		dev_err(adap->pdev_dev, "Failed to register thermal zone\n");
+		ch_thermal->tzdev = NULL;
+		return ret;
+	}
+
+	ret = thermal_zone_device_enable(ch_thermal->tzdev);
+	if (ret) {
+		dev_err(adap->pdev_dev, "Failed to enable thermal zone\n");
+		thermal_zone_device_unregister(adap->ch_thermal.tzdev);
+		return ret;
+	}
+
+	return 0;
+}
+
+int cxgb4_thermal_remove(struct adapter *adap)
+{
+	if (adap->ch_thermal.tzdev) {
+		thermal_zone_device_unregister(adap->ch_thermal.tzdev);
+		adap->ch_thermal.tzdev = NULL;
+	}
+	return 0;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.c
new file mode 100644
index 000000000..17faac715
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.c
@@ -0,0 +1,859 @@
+/*
+ * cxgb4_uld.c:Chelsio Upper Layer Driver Interface for T4/T5/T6 SGE management
+ *
+ * Copyright (c) 2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ *  Written by: Atul Gupta (atul.gupta@chelsio.com)
+ *  Written by: Hariprasad Shenai (hariprasad@chelsio.com)
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/debugfs.h>
+#include <linux/export.h>
+#include <linux/list.h>
+#include <linux/skbuff.h>
+#include <linux/pci.h>
+
+#include "cxgb4.h"
+#include "cxgb4_uld.h"
+#include "t4_regs.h"
+#include "t4fw_api.h"
+#include "t4_msg.h"
+
+#define for_each_uldrxq(m, i) for (i = 0; i < ((m)->nrxq + (m)->nciq); i++)
+
+/* Flush the aggregated lro sessions */
+static void uldrx_flush_handler(struct sge_rspq *q)
+{
+	struct adapter *adap = q->adap;
+
+	if (adap->uld[q->uld].lro_flush)
+		adap->uld[q->uld].lro_flush(&q->lro_mgr);
+}
+
+/**
+ *	uldrx_handler - response queue handler for ULD queues
+ *	@q: the response queue that received the packet
+ *	@rsp: the response queue descriptor holding the offload message
+ *	@gl: the gather list of packet fragments
+ *
+ *	Deliver an ingress offload packet to a ULD.  All processing is done by
+ *	the ULD, we just maintain statistics.
+ */
+static int uldrx_handler(struct sge_rspq *q, const __be64 *rsp,
+			 const struct pkt_gl *gl)
+{
+	struct adapter *adap = q->adap;
+	struct sge_ofld_rxq *rxq = container_of(q, struct sge_ofld_rxq, rspq);
+	int ret;
+
+	/* FW can send CPLs encapsulated in a CPL_FW4_MSG */
+	if (((const struct rss_header *)rsp)->opcode == CPL_FW4_MSG &&
+	    ((const struct cpl_fw4_msg *)(rsp + 1))->type == FW_TYPE_RSSCPL)
+		rsp += 2;
+
+	if (q->flush_handler)
+		ret = adap->uld[q->uld].lro_rx_handler(adap->uld[q->uld].handle,
+				rsp, gl, &q->lro_mgr,
+				&q->napi);
+	else
+		ret = adap->uld[q->uld].rx_handler(adap->uld[q->uld].handle,
+				rsp, gl);
+
+	if (ret) {
+		rxq->stats.nomem++;
+		return -1;
+	}
+
+	if (!gl)
+		rxq->stats.imm++;
+	else if (gl == CXGB4_MSG_AN)
+		rxq->stats.an++;
+	else
+		rxq->stats.pkts++;
+	return 0;
+}
+
+static int alloc_uld_rxqs(struct adapter *adap,
+			  struct sge_uld_rxq_info *rxq_info, bool lro)
+{
+	unsigned int nq = rxq_info->nrxq + rxq_info->nciq;
+	struct sge_ofld_rxq *q = rxq_info->uldrxq;
+	unsigned short *ids = rxq_info->rspq_id;
+	int i, err, msi_idx, que_idx = 0;
+	struct sge *s = &adap->sge;
+	unsigned int per_chan;
+
+	per_chan = rxq_info->nrxq / adap->params.nports;
+
+	if (adap->flags & CXGB4_USING_MSIX)
+		msi_idx = 1;
+	else
+		msi_idx = -((int)s->intrq.abs_id + 1);
+
+	for (i = 0; i < nq; i++, q++) {
+		if (i == rxq_info->nrxq) {
+			/* start allocation of concentrator queues */
+			per_chan = rxq_info->nciq / adap->params.nports;
+			que_idx = 0;
+		}
+
+		if (msi_idx >= 0) {
+			msi_idx = cxgb4_get_msix_idx_from_bmap(adap);
+			if (msi_idx < 0) {
+				err = -ENOSPC;
+				goto freeout;
+			}
+
+			snprintf(adap->msix_info[msi_idx].desc,
+				 sizeof(adap->msix_info[msi_idx].desc),
+				 "%s-%s%d",
+				 adap->port[0]->name, rxq_info->name, i);
+
+			q->msix = &adap->msix_info[msi_idx];
+		}
+		err = t4_sge_alloc_rxq(adap, &q->rspq, false,
+				       adap->port[que_idx++ / per_chan],
+				       msi_idx,
+				       q->fl.size ? &q->fl : NULL,
+				       uldrx_handler,
+				       lro ? uldrx_flush_handler : NULL,
+				       0);
+		if (err)
+			goto freeout;
+
+		memset(&q->stats, 0, sizeof(q->stats));
+		if (ids)
+			ids[i] = q->rspq.abs_id;
+	}
+	return 0;
+freeout:
+	q = rxq_info->uldrxq;
+	for ( ; i; i--, q++) {
+		if (q->rspq.desc)
+			free_rspq_fl(adap, &q->rspq,
+				     q->fl.size ? &q->fl : NULL);
+		if (q->msix)
+			cxgb4_free_msix_idx_in_bmap(adap, q->msix->idx);
+	}
+	return err;
+}
+
+static int
+setup_sge_queues_uld(struct adapter *adap, unsigned int uld_type, bool lro)
+{
+	struct sge_uld_rxq_info *rxq_info = adap->sge.uld_rxq_info[uld_type];
+	int i, ret;
+
+	ret = alloc_uld_rxqs(adap, rxq_info, lro);
+	if (ret)
+		return ret;
+
+	/* Tell uP to route control queue completions to rdma rspq */
+	if (adap->flags & CXGB4_FULL_INIT_DONE && uld_type == CXGB4_ULD_RDMA) {
+		struct sge *s = &adap->sge;
+		unsigned int cmplqid;
+		u32 param, cmdop;
+
+		cmdop = FW_PARAMS_PARAM_DMAQ_EQ_CMPLIQID_CTRL;
+		for_each_port(adap, i) {
+			cmplqid = rxq_info->uldrxq[i].rspq.cntxt_id;
+			param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DMAQ) |
+				 FW_PARAMS_PARAM_X_V(cmdop) |
+				 FW_PARAMS_PARAM_YZ_V(s->ctrlq[i].q.cntxt_id));
+			ret = t4_set_params(adap, adap->mbox, adap->pf,
+					    0, 1, &param, &cmplqid);
+		}
+	}
+	return ret;
+}
+
+static void t4_free_uld_rxqs(struct adapter *adap, int n,
+			     struct sge_ofld_rxq *q)
+{
+	for ( ; n; n--, q++) {
+		if (q->rspq.desc)
+			free_rspq_fl(adap, &q->rspq,
+				     q->fl.size ? &q->fl : NULL);
+	}
+}
+
+static void free_sge_queues_uld(struct adapter *adap, unsigned int uld_type)
+{
+	struct sge_uld_rxq_info *rxq_info = adap->sge.uld_rxq_info[uld_type];
+
+	if (adap->flags & CXGB4_FULL_INIT_DONE && uld_type == CXGB4_ULD_RDMA) {
+		struct sge *s = &adap->sge;
+		u32 param, cmdop, cmplqid = 0;
+		int i;
+
+		cmdop = FW_PARAMS_PARAM_DMAQ_EQ_CMPLIQID_CTRL;
+		for_each_port(adap, i) {
+			param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DMAQ) |
+				 FW_PARAMS_PARAM_X_V(cmdop) |
+				 FW_PARAMS_PARAM_YZ_V(s->ctrlq[i].q.cntxt_id));
+			t4_set_params(adap, adap->mbox, adap->pf,
+				      0, 1, &param, &cmplqid);
+		}
+	}
+
+	if (rxq_info->nciq)
+		t4_free_uld_rxqs(adap, rxq_info->nciq,
+				 rxq_info->uldrxq + rxq_info->nrxq);
+	t4_free_uld_rxqs(adap, rxq_info->nrxq, rxq_info->uldrxq);
+}
+
+static int cfg_queues_uld(struct adapter *adap, unsigned int uld_type,
+			  const struct cxgb4_uld_info *uld_info)
+{
+	struct sge *s = &adap->sge;
+	struct sge_uld_rxq_info *rxq_info;
+	int i, nrxq, ciq_size;
+
+	rxq_info = kzalloc(sizeof(*rxq_info), GFP_KERNEL);
+	if (!rxq_info)
+		return -ENOMEM;
+
+	if (adap->flags & CXGB4_USING_MSIX && uld_info->nrxq > s->nqs_per_uld) {
+		i = s->nqs_per_uld;
+		rxq_info->nrxq = roundup(i, adap->params.nports);
+	} else {
+		i = min_t(int, uld_info->nrxq,
+			  num_online_cpus());
+		rxq_info->nrxq = roundup(i, adap->params.nports);
+	}
+	if (!uld_info->ciq) {
+		rxq_info->nciq = 0;
+	} else  {
+		if (adap->flags & CXGB4_USING_MSIX)
+			rxq_info->nciq = min_t(int, s->nqs_per_uld,
+					       num_online_cpus());
+		else
+			rxq_info->nciq = min_t(int, MAX_OFLD_QSETS,
+					       num_online_cpus());
+		rxq_info->nciq = ((rxq_info->nciq / adap->params.nports) *
+				  adap->params.nports);
+		rxq_info->nciq = max_t(int, rxq_info->nciq,
+				       adap->params.nports);
+	}
+
+	nrxq = rxq_info->nrxq + rxq_info->nciq; /* total rxq's */
+	rxq_info->uldrxq = kcalloc(nrxq, sizeof(struct sge_ofld_rxq),
+				   GFP_KERNEL);
+	if (!rxq_info->uldrxq) {
+		kfree(rxq_info);
+		return -ENOMEM;
+	}
+
+	rxq_info->rspq_id = kcalloc(nrxq, sizeof(unsigned short), GFP_KERNEL);
+	if (!rxq_info->rspq_id) {
+		kfree(rxq_info->uldrxq);
+		kfree(rxq_info);
+		return -ENOMEM;
+	}
+
+	for (i = 0; i < rxq_info->nrxq; i++) {
+		struct sge_ofld_rxq *r = &rxq_info->uldrxq[i];
+
+		init_rspq(adap, &r->rspq, 5, 1, uld_info->rxq_size, 64);
+		r->rspq.uld = uld_type;
+		r->fl.size = 72;
+	}
+
+	ciq_size = 64 + adap->vres.cq.size + adap->tids.nftids;
+	if (ciq_size > SGE_MAX_IQ_SIZE) {
+		dev_warn(adap->pdev_dev, "CIQ size too small for available IQs\n");
+		ciq_size = SGE_MAX_IQ_SIZE;
+	}
+
+	for (i = rxq_info->nrxq; i < nrxq; i++) {
+		struct sge_ofld_rxq *r = &rxq_info->uldrxq[i];
+
+		init_rspq(adap, &r->rspq, 5, 1, ciq_size, 64);
+		r->rspq.uld = uld_type;
+	}
+
+	memcpy(rxq_info->name, uld_info->name, IFNAMSIZ);
+	adap->sge.uld_rxq_info[uld_type] = rxq_info;
+
+	return 0;
+}
+
+static void free_queues_uld(struct adapter *adap, unsigned int uld_type)
+{
+	struct sge_uld_rxq_info *rxq_info = adap->sge.uld_rxq_info[uld_type];
+
+	adap->sge.uld_rxq_info[uld_type] = NULL;
+	kfree(rxq_info->rspq_id);
+	kfree(rxq_info->uldrxq);
+	kfree(rxq_info);
+}
+
+static int
+request_msix_queue_irqs_uld(struct adapter *adap, unsigned int uld_type)
+{
+	struct sge_uld_rxq_info *rxq_info = adap->sge.uld_rxq_info[uld_type];
+	struct msix_info *minfo;
+	unsigned int idx;
+	int err = 0;
+
+	for_each_uldrxq(rxq_info, idx) {
+		minfo = rxq_info->uldrxq[idx].msix;
+		err = request_irq(minfo->vec,
+				  t4_sge_intr_msix, 0,
+				  minfo->desc,
+				  &rxq_info->uldrxq[idx].rspq);
+		if (err)
+			goto unwind;
+
+		cxgb4_set_msix_aff(adap, minfo->vec,
+				   &minfo->aff_mask, idx);
+	}
+	return 0;
+
+unwind:
+	while (idx-- > 0) {
+		minfo = rxq_info->uldrxq[idx].msix;
+		cxgb4_clear_msix_aff(minfo->vec, minfo->aff_mask);
+		cxgb4_free_msix_idx_in_bmap(adap, minfo->idx);
+		free_irq(minfo->vec, &rxq_info->uldrxq[idx].rspq);
+	}
+	return err;
+}
+
+static void
+free_msix_queue_irqs_uld(struct adapter *adap, unsigned int uld_type)
+{
+	struct sge_uld_rxq_info *rxq_info = adap->sge.uld_rxq_info[uld_type];
+	struct msix_info *minfo;
+	unsigned int idx;
+
+	for_each_uldrxq(rxq_info, idx) {
+		minfo = rxq_info->uldrxq[idx].msix;
+		cxgb4_clear_msix_aff(minfo->vec, minfo->aff_mask);
+		cxgb4_free_msix_idx_in_bmap(adap, minfo->idx);
+		free_irq(minfo->vec, &rxq_info->uldrxq[idx].rspq);
+	}
+}
+
+static void enable_rx_uld(struct adapter *adap, unsigned int uld_type)
+{
+	struct sge_uld_rxq_info *rxq_info = adap->sge.uld_rxq_info[uld_type];
+	int idx;
+
+	for_each_uldrxq(rxq_info, idx) {
+		struct sge_rspq *q = &rxq_info->uldrxq[idx].rspq;
+
+		if (!q)
+			continue;
+
+		cxgb4_enable_rx(adap, q);
+	}
+}
+
+static void quiesce_rx_uld(struct adapter *adap, unsigned int uld_type)
+{
+	struct sge_uld_rxq_info *rxq_info = adap->sge.uld_rxq_info[uld_type];
+	int idx;
+
+	for_each_uldrxq(rxq_info, idx) {
+		struct sge_rspq *q = &rxq_info->uldrxq[idx].rspq;
+
+		if (!q)
+			continue;
+
+		cxgb4_quiesce_rx(q);
+	}
+}
+
+static void
+free_sge_txq_uld(struct adapter *adap, struct sge_uld_txq_info *txq_info)
+{
+	int nq = txq_info->ntxq;
+	int i;
+
+	for (i = 0; i < nq; i++) {
+		struct sge_uld_txq *txq = &txq_info->uldtxq[i];
+
+		if (txq && txq->q.desc) {
+			tasklet_kill(&txq->qresume_tsk);
+			t4_ofld_eq_free(adap, adap->mbox, adap->pf, 0,
+					txq->q.cntxt_id);
+			free_tx_desc(adap, &txq->q, txq->q.in_use, false);
+			kfree(txq->q.sdesc);
+			__skb_queue_purge(&txq->sendq);
+			free_txq(adap, &txq->q);
+		}
+	}
+}
+
+static int
+alloc_sge_txq_uld(struct adapter *adap, struct sge_uld_txq_info *txq_info,
+		  unsigned int uld_type)
+{
+	struct sge *s = &adap->sge;
+	int nq = txq_info->ntxq;
+	int i, j, err;
+
+	j = nq / adap->params.nports;
+	for (i = 0; i < nq; i++) {
+		struct sge_uld_txq *txq = &txq_info->uldtxq[i];
+
+		txq->q.size = 1024;
+		err = t4_sge_alloc_uld_txq(adap, txq, adap->port[i / j],
+					   s->fw_evtq.cntxt_id, uld_type);
+		if (err)
+			goto freeout;
+	}
+	return 0;
+freeout:
+	free_sge_txq_uld(adap, txq_info);
+	return err;
+}
+
+static void
+release_sge_txq_uld(struct adapter *adap, unsigned int uld_type)
+{
+	struct sge_uld_txq_info *txq_info = NULL;
+	int tx_uld_type = TX_ULD(uld_type);
+
+	txq_info = adap->sge.uld_txq_info[tx_uld_type];
+
+	if (txq_info && atomic_dec_and_test(&txq_info->users)) {
+		free_sge_txq_uld(adap, txq_info);
+		kfree(txq_info->uldtxq);
+		kfree(txq_info);
+		adap->sge.uld_txq_info[tx_uld_type] = NULL;
+	}
+}
+
+static int
+setup_sge_txq_uld(struct adapter *adap, unsigned int uld_type,
+		  const struct cxgb4_uld_info *uld_info)
+{
+	struct sge_uld_txq_info *txq_info = NULL;
+	int tx_uld_type, i;
+
+	tx_uld_type = TX_ULD(uld_type);
+	txq_info = adap->sge.uld_txq_info[tx_uld_type];
+
+	if ((tx_uld_type == CXGB4_TX_OFLD) && txq_info &&
+	    (atomic_inc_return(&txq_info->users) > 1))
+		return 0;
+
+	txq_info = kzalloc(sizeof(*txq_info), GFP_KERNEL);
+	if (!txq_info)
+		return -ENOMEM;
+	if (uld_type == CXGB4_ULD_CRYPTO) {
+		i = min_t(int, adap->vres.ncrypto_fc,
+			  num_online_cpus());
+		txq_info->ntxq = rounddown(i, adap->params.nports);
+		if (txq_info->ntxq <= 0) {
+			dev_warn(adap->pdev_dev, "Crypto Tx Queues can't be zero\n");
+			kfree(txq_info);
+			return -EINVAL;
+		}
+
+	} else {
+		i = min_t(int, uld_info->ntxq, num_online_cpus());
+		txq_info->ntxq = roundup(i, adap->params.nports);
+	}
+	txq_info->uldtxq = kcalloc(txq_info->ntxq, sizeof(struct sge_uld_txq),
+				   GFP_KERNEL);
+	if (!txq_info->uldtxq) {
+		kfree(txq_info);
+		return -ENOMEM;
+	}
+
+	if (alloc_sge_txq_uld(adap, txq_info, tx_uld_type)) {
+		kfree(txq_info->uldtxq);
+		kfree(txq_info);
+		return -ENOMEM;
+	}
+
+	atomic_inc(&txq_info->users);
+	adap->sge.uld_txq_info[tx_uld_type] = txq_info;
+	return 0;
+}
+
+static void uld_queue_init(struct adapter *adap, unsigned int uld_type,
+			   struct cxgb4_lld_info *lli)
+{
+	struct sge_uld_rxq_info *rxq_info = adap->sge.uld_rxq_info[uld_type];
+	int tx_uld_type = TX_ULD(uld_type);
+	struct sge_uld_txq_info *txq_info = adap->sge.uld_txq_info[tx_uld_type];
+
+	lli->rxq_ids = rxq_info->rspq_id;
+	lli->nrxq = rxq_info->nrxq;
+	lli->ciq_ids = rxq_info->rspq_id + rxq_info->nrxq;
+	lli->nciq = rxq_info->nciq;
+	lli->ntxq = txq_info->ntxq;
+}
+
+int t4_uld_mem_alloc(struct adapter *adap)
+{
+	struct sge *s = &adap->sge;
+
+	adap->uld = kcalloc(CXGB4_ULD_MAX, sizeof(*adap->uld), GFP_KERNEL);
+	if (!adap->uld)
+		return -ENOMEM;
+
+	s->uld_rxq_info = kcalloc(CXGB4_ULD_MAX,
+				  sizeof(struct sge_uld_rxq_info *),
+				  GFP_KERNEL);
+	if (!s->uld_rxq_info)
+		goto err_uld;
+
+	s->uld_txq_info = kcalloc(CXGB4_TX_MAX,
+				  sizeof(struct sge_uld_txq_info *),
+				  GFP_KERNEL);
+	if (!s->uld_txq_info)
+		goto err_uld_rx;
+	return 0;
+
+err_uld_rx:
+	kfree(s->uld_rxq_info);
+err_uld:
+	kfree(adap->uld);
+	return -ENOMEM;
+}
+
+void t4_uld_mem_free(struct adapter *adap)
+{
+	struct sge *s = &adap->sge;
+
+	kfree(s->uld_txq_info);
+	kfree(s->uld_rxq_info);
+	kfree(adap->uld);
+}
+
+/* This function should be called with uld_mutex taken. */
+static void cxgb4_shutdown_uld_adapter(struct adapter *adap, enum cxgb4_uld type)
+{
+	if (adap->uld[type].handle) {
+		adap->uld[type].handle = NULL;
+		adap->uld[type].add = NULL;
+		release_sge_txq_uld(adap, type);
+
+		if (adap->flags & CXGB4_FULL_INIT_DONE)
+			quiesce_rx_uld(adap, type);
+
+		if (adap->flags & CXGB4_USING_MSIX)
+			free_msix_queue_irqs_uld(adap, type);
+
+		free_sge_queues_uld(adap, type);
+		free_queues_uld(adap, type);
+	}
+}
+
+void t4_uld_clean_up(struct adapter *adap)
+{
+	unsigned int i;
+
+	if (!is_uld(adap))
+		return;
+
+	mutex_lock(&uld_mutex);
+	for (i = 0; i < CXGB4_ULD_MAX; i++) {
+		if (!adap->uld[i].handle)
+			continue;
+
+		cxgb4_shutdown_uld_adapter(adap, i);
+	}
+	mutex_unlock(&uld_mutex);
+}
+
+static void uld_init(struct adapter *adap, struct cxgb4_lld_info *lld)
+{
+	int i;
+
+	lld->pdev = adap->pdev;
+	lld->pf = adap->pf;
+	lld->l2t = adap->l2t;
+	lld->tids = &adap->tids;
+	lld->ports = adap->port;
+	lld->vr = &adap->vres;
+	lld->mtus = adap->params.mtus;
+	lld->nchan = adap->params.nports;
+	lld->nports = adap->params.nports;
+	lld->wr_cred = adap->params.ofldq_wr_cred;
+	lld->crypto = adap->params.crypto;
+	lld->iscsi_iolen = MAXRXDATA_G(t4_read_reg(adap, TP_PARA_REG2_A));
+	lld->iscsi_tagmask = t4_read_reg(adap, ULP_RX_ISCSI_TAGMASK_A);
+	lld->iscsi_pgsz_order = t4_read_reg(adap, ULP_RX_ISCSI_PSZ_A);
+	lld->iscsi_llimit = t4_read_reg(adap, ULP_RX_ISCSI_LLIMIT_A);
+	lld->iscsi_ppm = &adap->iscsi_ppm;
+	lld->adapter_type = adap->params.chip;
+	lld->cclk_ps = 1000000000 / adap->params.vpd.cclk;
+	lld->udb_density = 1 << adap->params.sge.eq_qpp;
+	lld->ucq_density = 1 << adap->params.sge.iq_qpp;
+	lld->sge_host_page_size = 1 << (adap->params.sge.hps + 10);
+	lld->filt_mode = adap->params.tp.vlan_pri_map;
+	/* MODQ_REQ_MAP sets queues 0-3 to chan 0-3 */
+	for (i = 0; i < NCHAN; i++)
+		lld->tx_modq[i] = i;
+	lld->gts_reg = adap->regs + MYPF_REG(SGE_PF_GTS_A);
+	lld->db_reg = adap->regs + MYPF_REG(SGE_PF_KDOORBELL_A);
+	lld->fw_vers = adap->params.fw_vers;
+	lld->dbfifo_int_thresh = dbfifo_int_thresh;
+	lld->sge_ingpadboundary = adap->sge.fl_align;
+	lld->sge_egrstatuspagesize = adap->sge.stat_len;
+	lld->sge_pktshift = adap->sge.pktshift;
+	lld->ulp_crypto = adap->params.crypto;
+	lld->enable_fw_ofld_conn = adap->flags & CXGB4_FW_OFLD_CONN;
+	lld->max_ordird_qp = adap->params.max_ordird_qp;
+	lld->max_ird_adapter = adap->params.max_ird_adapter;
+	lld->ulptx_memwrite_dsgl = adap->params.ulptx_memwrite_dsgl;
+	lld->nodeid = dev_to_node(adap->pdev_dev);
+	lld->fr_nsmr_tpte_wr_support = adap->params.fr_nsmr_tpte_wr_support;
+	lld->write_w_imm_support = adap->params.write_w_imm_support;
+	lld->write_cmpl_support = adap->params.write_cmpl_support;
+}
+
+static int uld_attach(struct adapter *adap, unsigned int uld)
+{
+	struct cxgb4_lld_info lli;
+	void *handle;
+
+	uld_init(adap, &lli);
+	uld_queue_init(adap, uld, &lli);
+
+	handle = adap->uld[uld].add(&lli);
+	if (IS_ERR(handle)) {
+		dev_warn(adap->pdev_dev,
+			 "could not attach to the %s driver, error %ld\n",
+			 adap->uld[uld].name, PTR_ERR(handle));
+		return PTR_ERR(handle);
+	}
+
+	adap->uld[uld].handle = handle;
+	t4_register_netevent_notifier();
+
+	if (adap->flags & CXGB4_FULL_INIT_DONE)
+		adap->uld[uld].state_change(handle, CXGB4_STATE_UP);
+
+	return 0;
+}
+
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+static bool cxgb4_uld_in_use(struct adapter *adap)
+{
+	const struct tid_info *t = &adap->tids;
+
+	return (atomic_read(&t->conns_in_use) || t->stids_in_use);
+}
+
+/* cxgb4_set_ktls_feature: request FW to enable/disable ktls settings.
+ * @adap: adapter info
+ * @enable: 1 to enable / 0 to disable ktls settings.
+ */
+int cxgb4_set_ktls_feature(struct adapter *adap, bool enable)
+{
+	int ret = 0;
+	u32 params =
+		FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_KTLS_HW) |
+		FW_PARAMS_PARAM_Y_V(enable) |
+		FW_PARAMS_PARAM_Z_V(FW_PARAMS_PARAM_DEV_KTLS_HW_USER_ENABLE);
+
+	if (enable) {
+		if (!refcount_read(&adap->chcr_ktls.ktls_refcount)) {
+			/* At this moment if ULD connection are up means, other
+			 * ULD is/are already active, return failure.
+			 */
+			if (cxgb4_uld_in_use(adap)) {
+				dev_dbg(adap->pdev_dev,
+					"ULD connections (tid/stid) active. Can't enable kTLS\n");
+				return -EINVAL;
+			}
+			ret = t4_set_params(adap, adap->mbox, adap->pf,
+					    0, 1, &params, &params);
+			if (ret)
+				return ret;
+			refcount_set(&adap->chcr_ktls.ktls_refcount, 1);
+			pr_debug("kTLS has been enabled. Restrictions placed on ULD support\n");
+		} else {
+			/* ktls settings already up, just increment refcount. */
+			refcount_inc(&adap->chcr_ktls.ktls_refcount);
+		}
+	} else {
+		/* return failure if refcount is already 0. */
+		if (!refcount_read(&adap->chcr_ktls.ktls_refcount))
+			return -EINVAL;
+		/* decrement refcount and test, if 0, disable ktls feature,
+		 * else return command success.
+		 */
+		if (refcount_dec_and_test(&adap->chcr_ktls.ktls_refcount)) {
+			ret = t4_set_params(adap, adap->mbox, adap->pf,
+					    0, 1, &params, &params);
+			if (ret)
+				return ret;
+			pr_debug("kTLS is disabled. Restrictions on ULD support removed\n");
+		}
+	}
+
+	return ret;
+}
+#endif
+
+static void cxgb4_uld_alloc_resources(struct adapter *adap,
+				      enum cxgb4_uld type,
+				      const struct cxgb4_uld_info *p)
+{
+	int ret = 0;
+
+	if ((type == CXGB4_ULD_CRYPTO && !is_pci_uld(adap)) ||
+	    (type != CXGB4_ULD_CRYPTO && !is_offload(adap)))
+		return;
+	if (type == CXGB4_ULD_ISCSIT && is_t4(adap->params.chip))
+		return;
+	ret = cfg_queues_uld(adap, type, p);
+	if (ret)
+		goto out;
+	ret = setup_sge_queues_uld(adap, type, p->lro);
+	if (ret)
+		goto free_queues;
+	if (adap->flags & CXGB4_USING_MSIX) {
+		ret = request_msix_queue_irqs_uld(adap, type);
+		if (ret)
+			goto free_rxq;
+	}
+	if (adap->flags & CXGB4_FULL_INIT_DONE)
+		enable_rx_uld(adap, type);
+	if (adap->uld[type].add)
+		goto free_irq;
+	ret = setup_sge_txq_uld(adap, type, p);
+	if (ret)
+		goto free_irq;
+	adap->uld[type] = *p;
+	ret = uld_attach(adap, type);
+	if (ret)
+		goto free_txq;
+	return;
+free_txq:
+	release_sge_txq_uld(adap, type);
+free_irq:
+	if (adap->flags & CXGB4_FULL_INIT_DONE)
+		quiesce_rx_uld(adap, type);
+	if (adap->flags & CXGB4_USING_MSIX)
+		free_msix_queue_irqs_uld(adap, type);
+free_rxq:
+	free_sge_queues_uld(adap, type);
+free_queues:
+	free_queues_uld(adap, type);
+out:
+	dev_warn(adap->pdev_dev,
+		 "ULD registration failed for uld type %d\n", type);
+}
+
+void cxgb4_uld_enable(struct adapter *adap)
+{
+	struct cxgb4_uld_list *uld_entry;
+
+	mutex_lock(&uld_mutex);
+	list_add_tail(&adap->list_node, &adapter_list);
+	list_for_each_entry(uld_entry, &uld_list, list_node)
+		cxgb4_uld_alloc_resources(adap, uld_entry->uld_type,
+					  &uld_entry->uld_info);
+	mutex_unlock(&uld_mutex);
+}
+
+/* cxgb4_register_uld - register an upper-layer driver
+ * @type: the ULD type
+ * @p: the ULD methods
+ *
+ * Registers an upper-layer driver with this driver and notifies the ULD
+ * about any presently available devices that support its type.
+ */
+void cxgb4_register_uld(enum cxgb4_uld type,
+			const struct cxgb4_uld_info *p)
+{
+	struct cxgb4_uld_list *uld_entry;
+	struct adapter *adap;
+
+	if (type >= CXGB4_ULD_MAX)
+		return;
+
+	uld_entry = kzalloc(sizeof(*uld_entry), GFP_KERNEL);
+	if (!uld_entry)
+		return;
+
+	memcpy(&uld_entry->uld_info, p, sizeof(struct cxgb4_uld_info));
+	mutex_lock(&uld_mutex);
+	list_for_each_entry(adap, &adapter_list, list_node)
+		cxgb4_uld_alloc_resources(adap, type, p);
+
+	uld_entry->uld_type = type;
+	list_add_tail(&uld_entry->list_node, &uld_list);
+	mutex_unlock(&uld_mutex);
+	return;
+}
+EXPORT_SYMBOL(cxgb4_register_uld);
+
+/**
+ *	cxgb4_unregister_uld - unregister an upper-layer driver
+ *	@type: the ULD type
+ *
+ *	Unregisters an existing upper-layer driver.
+ */
+int cxgb4_unregister_uld(enum cxgb4_uld type)
+{
+	struct cxgb4_uld_list *uld_entry, *tmp;
+	struct adapter *adap;
+
+	if (type >= CXGB4_ULD_MAX)
+		return -EINVAL;
+
+	mutex_lock(&uld_mutex);
+	list_for_each_entry(adap, &adapter_list, list_node) {
+		if ((type == CXGB4_ULD_CRYPTO && !is_pci_uld(adap)) ||
+		    (type != CXGB4_ULD_CRYPTO && !is_offload(adap)))
+			continue;
+		if (type == CXGB4_ULD_ISCSIT && is_t4(adap->params.chip))
+			continue;
+
+		cxgb4_shutdown_uld_adapter(adap, type);
+	}
+
+	list_for_each_entry_safe(uld_entry, tmp, &uld_list, list_node) {
+		if (uld_entry->uld_type == type) {
+			list_del(&uld_entry->list_node);
+			kfree(uld_entry);
+		}
+	}
+	mutex_unlock(&uld_mutex);
+
+	return 0;
+}
+EXPORT_SYMBOL(cxgb4_unregister_uld);
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h
new file mode 100644
index 000000000..34546f531
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h
@@ -0,0 +1,544 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_ULD_H
+#define __CXGB4_ULD_H
+
+#include <linux/cache.h>
+#include <linux/spinlock.h>
+#include <linux/skbuff.h>
+#include <linux/inetdevice.h>
+#include <linux/atomic.h>
+#include <net/tls.h>
+#include "cxgb4.h"
+
+#define MAX_ULD_QSETS 16
+#define MAX_ULD_NPORTS 4
+
+/* ulp_mem_io + ulptx_idata + payload + padding */
+#define MAX_IMM_ULPTX_WR_LEN (32 + 8 + 256 + 8)
+
+/* CPL message priority levels */
+enum {
+	CPL_PRIORITY_DATA     = 0,  /* data messages */
+	CPL_PRIORITY_SETUP    = 1,  /* connection setup messages */
+	CPL_PRIORITY_TEARDOWN = 0,  /* connection teardown messages */
+	CPL_PRIORITY_LISTEN   = 1,  /* listen start/stop messages */
+	CPL_PRIORITY_ACK      = 1,  /* RX ACK messages */
+	CPL_PRIORITY_CONTROL  = 1   /* control messages */
+};
+
+#define INIT_TP_WR(w, tid) do { \
+	(w)->wr.wr_hi = htonl(FW_WR_OP_V(FW_TP_WR) | \
+			      FW_WR_IMMDLEN_V(sizeof(*w) - sizeof(w->wr))); \
+	(w)->wr.wr_mid = htonl(FW_WR_LEN16_V(DIV_ROUND_UP(sizeof(*w), 16)) | \
+			       FW_WR_FLOWID_V(tid)); \
+	(w)->wr.wr_lo = cpu_to_be64(0); \
+} while (0)
+
+#define INIT_TP_WR_CPL(w, cpl, tid) do { \
+	INIT_TP_WR(w, tid); \
+	OPCODE_TID(w) = htonl(MK_OPCODE_TID(cpl, tid)); \
+} while (0)
+
+#define INIT_ULPTX_WR(w, wrlen, atomic, tid) do { \
+	(w)->wr.wr_hi = htonl(FW_WR_OP_V(FW_ULPTX_WR) | \
+			      FW_WR_ATOMIC_V(atomic)); \
+	(w)->wr.wr_mid = htonl(FW_WR_LEN16_V(DIV_ROUND_UP(wrlen, 16)) | \
+			       FW_WR_FLOWID_V(tid)); \
+	(w)->wr.wr_lo = cpu_to_be64(0); \
+} while (0)
+
+/* Special asynchronous notification message */
+#define CXGB4_MSG_AN ((void *)1)
+#define TX_ULD(uld)(((uld) != CXGB4_ULD_CRYPTO) ? CXGB4_TX_OFLD :\
+		      CXGB4_TX_CRYPTO)
+
+struct serv_entry {
+	void *data;
+};
+
+union aopen_entry {
+	void *data;
+	union aopen_entry *next;
+};
+
+struct eotid_entry {
+	void *data;
+};
+
+/*
+ * Holds the size, base address, free list start, etc of the TID, server TID,
+ * and active-open TID tables.  The tables themselves are allocated dynamically.
+ */
+struct tid_info {
+	void **tid_tab;
+	unsigned int tid_base;
+	unsigned int ntids;
+
+	struct serv_entry *stid_tab;
+	unsigned long *stid_bmap;
+	unsigned int nstids;
+	unsigned int stid_base;
+
+	unsigned int nhash;
+	unsigned int hash_base;
+
+	union aopen_entry *atid_tab;
+	unsigned int natids;
+	unsigned int atid_base;
+
+	struct filter_entry *hpftid_tab;
+	unsigned long *hpftid_bmap;
+	unsigned int nhpftids;
+	unsigned int hpftid_base;
+
+	struct filter_entry *ftid_tab;
+	unsigned long *ftid_bmap;
+	unsigned int nftids;
+	unsigned int ftid_base;
+	unsigned int aftid_base;
+	unsigned int aftid_end;
+	/* Server filter region */
+	unsigned int sftid_base;
+	unsigned int nsftids;
+
+	spinlock_t atid_lock ____cacheline_aligned_in_smp;
+	union aopen_entry *afree;
+	unsigned int atids_in_use;
+
+	spinlock_t stid_lock;
+	unsigned int stids_in_use;
+	unsigned int v6_stids_in_use;
+	unsigned int sftids_in_use;
+
+	/* ETHOFLD range */
+	struct eotid_entry *eotid_tab;
+	unsigned long *eotid_bmap;
+	unsigned int eotid_base;
+	unsigned int neotids;
+
+	/* TIDs in the TCAM */
+	atomic_t tids_in_use;
+	/* TIDs in the HASH */
+	atomic_t hash_tids_in_use;
+	atomic_t conns_in_use;
+	/* ETHOFLD TIDs used for rate limiting */
+	atomic_t eotids_in_use;
+
+	/* lock for setting/clearing filter bitmap */
+	spinlock_t ftid_lock;
+
+	unsigned int tc_hash_tids_max_prio;
+};
+
+static inline void *lookup_tid(const struct tid_info *t, unsigned int tid)
+{
+	tid -= t->tid_base;
+	return tid < t->ntids ? t->tid_tab[tid] : NULL;
+}
+
+static inline bool tid_out_of_range(const struct tid_info *t, unsigned int tid)
+{
+	return ((tid - t->tid_base) >= t->ntids);
+}
+
+static inline void *lookup_atid(const struct tid_info *t, unsigned int atid)
+{
+	return atid < t->natids ? t->atid_tab[atid].data : NULL;
+}
+
+static inline void *lookup_stid(const struct tid_info *t, unsigned int stid)
+{
+	/* Is it a server filter TID? */
+	if (t->nsftids && (stid >= t->sftid_base)) {
+		stid -= t->sftid_base;
+		stid += t->nstids;
+	} else {
+		stid -= t->stid_base;
+	}
+
+	return stid < (t->nstids + t->nsftids) ? t->stid_tab[stid].data : NULL;
+}
+
+static inline void cxgb4_insert_tid(struct tid_info *t, void *data,
+				    unsigned int tid, unsigned short family)
+{
+	t->tid_tab[tid - t->tid_base] = data;
+	if (t->hash_base && (tid >= t->hash_base)) {
+		if (family == AF_INET6)
+			atomic_add(2, &t->hash_tids_in_use);
+		else
+			atomic_inc(&t->hash_tids_in_use);
+	} else {
+		if (family == AF_INET6)
+			atomic_add(2, &t->tids_in_use);
+		else
+			atomic_inc(&t->tids_in_use);
+	}
+	atomic_inc(&t->conns_in_use);
+}
+
+static inline struct eotid_entry *cxgb4_lookup_eotid(struct tid_info *t,
+						     u32 eotid)
+{
+	return eotid < t->neotids ? &t->eotid_tab[eotid] : NULL;
+}
+
+static inline int cxgb4_get_free_eotid(struct tid_info *t)
+{
+	int eotid;
+
+	eotid = find_first_zero_bit(t->eotid_bmap, t->neotids);
+	if (eotid >= t->neotids)
+		eotid = -1;
+
+	return eotid;
+}
+
+static inline void cxgb4_alloc_eotid(struct tid_info *t, u32 eotid, void *data)
+{
+	set_bit(eotid, t->eotid_bmap);
+	t->eotid_tab[eotid].data = data;
+	atomic_inc(&t->eotids_in_use);
+}
+
+static inline void cxgb4_free_eotid(struct tid_info *t, u32 eotid)
+{
+	clear_bit(eotid, t->eotid_bmap);
+	t->eotid_tab[eotid].data = NULL;
+	atomic_dec(&t->eotids_in_use);
+}
+
+int cxgb4_alloc_atid(struct tid_info *t, void *data);
+int cxgb4_alloc_stid(struct tid_info *t, int family, void *data);
+int cxgb4_alloc_sftid(struct tid_info *t, int family, void *data);
+void cxgb4_free_atid(struct tid_info *t, unsigned int atid);
+void cxgb4_free_stid(struct tid_info *t, unsigned int stid, int family);
+void cxgb4_remove_tid(struct tid_info *t, unsigned int qid, unsigned int tid,
+		      unsigned short family);
+struct in6_addr;
+
+int cxgb4_create_server(const struct net_device *dev, unsigned int stid,
+			__be32 sip, __be16 sport, __be16 vlan,
+			unsigned int queue);
+int cxgb4_create_server6(const struct net_device *dev, unsigned int stid,
+			 const struct in6_addr *sip, __be16 sport,
+			 unsigned int queue);
+int cxgb4_remove_server(const struct net_device *dev, unsigned int stid,
+			unsigned int queue, bool ipv6);
+int cxgb4_create_server_filter(const struct net_device *dev, unsigned int stid,
+			       __be32 sip, __be16 sport, __be16 vlan,
+			       unsigned int queue,
+			       unsigned char port, unsigned char mask);
+int cxgb4_remove_server_filter(const struct net_device *dev, unsigned int stid,
+			       unsigned int queue, bool ipv6);
+
+/* Filter operation context to allow callers of cxgb4_set_filter() and
+ * cxgb4_del_filter() to wait for an asynchronous completion.
+ */
+struct filter_ctx {
+	struct completion completion;	/* completion rendezvous */
+	void *closure;			/* caller's opaque information */
+	int result;			/* result of operation */
+	u32 tid;			/* to store tid */
+};
+
+struct chcr_ktls {
+	refcount_t ktls_refcount;
+};
+
+struct ch_filter_specification;
+
+int cxgb4_get_free_ftid(struct net_device *dev, u8 family, bool hash_en,
+			u32 tc_prio);
+int __cxgb4_set_filter(struct net_device *dev, int filter_id,
+		       struct ch_filter_specification *fs,
+		       struct filter_ctx *ctx);
+int __cxgb4_del_filter(struct net_device *dev, int filter_id,
+		       struct ch_filter_specification *fs,
+		       struct filter_ctx *ctx);
+int cxgb4_set_filter(struct net_device *dev, int filter_id,
+		     struct ch_filter_specification *fs);
+int cxgb4_del_filter(struct net_device *dev, int filter_id,
+		     struct ch_filter_specification *fs);
+int cxgb4_get_filter_counters(struct net_device *dev, unsigned int fidx,
+			      u64 *hitcnt, u64 *bytecnt, bool hash);
+
+static inline void set_wr_txq(struct sk_buff *skb, int prio, int queue)
+{
+	skb_set_queue_mapping(skb, (queue << 1) | prio);
+}
+
+enum cxgb4_uld {
+	CXGB4_ULD_INIT,
+	CXGB4_ULD_RDMA,
+	CXGB4_ULD_ISCSI,
+	CXGB4_ULD_ISCSIT,
+	CXGB4_ULD_CRYPTO,
+	CXGB4_ULD_IPSEC,
+	CXGB4_ULD_TLS,
+	CXGB4_ULD_KTLS,
+	CXGB4_ULD_MAX
+};
+
+enum cxgb4_tx_uld {
+	CXGB4_TX_OFLD,
+	CXGB4_TX_CRYPTO,
+	CXGB4_TX_MAX
+};
+
+enum cxgb4_txq_type {
+	CXGB4_TXQ_ETH,
+	CXGB4_TXQ_ULD,
+	CXGB4_TXQ_CTRL,
+	CXGB4_TXQ_MAX
+};
+
+enum cxgb4_state {
+	CXGB4_STATE_UP,
+	CXGB4_STATE_START_RECOVERY,
+	CXGB4_STATE_DOWN,
+	CXGB4_STATE_DETACH,
+	CXGB4_STATE_FATAL_ERROR
+};
+
+enum cxgb4_control {
+	CXGB4_CONTROL_DB_FULL,
+	CXGB4_CONTROL_DB_EMPTY,
+	CXGB4_CONTROL_DB_DROP,
+};
+
+struct adapter;
+struct pci_dev;
+struct l2t_data;
+struct net_device;
+struct pkt_gl;
+struct tp_tcp_stats;
+struct t4_lro_mgr;
+
+struct cxgb4_range {
+	unsigned int start;
+	unsigned int size;
+};
+
+struct cxgb4_virt_res {                      /* virtualized HW resources */
+	struct cxgb4_range ddp;
+	struct cxgb4_range iscsi;
+	struct cxgb4_range stag;
+	struct cxgb4_range rq;
+	struct cxgb4_range srq;
+	struct cxgb4_range pbl;
+	struct cxgb4_range qp;
+	struct cxgb4_range cq;
+	struct cxgb4_range ocq;
+	struct cxgb4_range key;
+	unsigned int ncrypto_fc;
+	struct cxgb4_range ppod_edram;
+};
+
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+struct ch_ktls_port_stats_debug {
+	atomic64_t ktls_tx_connection_open;
+	atomic64_t ktls_tx_connection_fail;
+	atomic64_t ktls_tx_connection_close;
+	atomic64_t ktls_tx_encrypted_packets;
+	atomic64_t ktls_tx_encrypted_bytes;
+	atomic64_t ktls_tx_ctx;
+	atomic64_t ktls_tx_ooo;
+	atomic64_t ktls_tx_skip_no_sync_data;
+	atomic64_t ktls_tx_drop_no_sync_data;
+	atomic64_t ktls_tx_drop_bypass_req;
+};
+
+struct ch_ktls_stats_debug {
+	struct ch_ktls_port_stats_debug ktls_port[MAX_ULD_NPORTS];
+	atomic64_t ktls_tx_send_records;
+	atomic64_t ktls_tx_end_pkts;
+	atomic64_t ktls_tx_start_pkts;
+	atomic64_t ktls_tx_middle_pkts;
+	atomic64_t ktls_tx_retransmit_pkts;
+	atomic64_t ktls_tx_complete_pkts;
+	atomic64_t ktls_tx_trimmed_pkts;
+	atomic64_t ktls_tx_fallback;
+};
+#endif
+
+struct chcr_stats_debug {
+	atomic_t cipher_rqst;
+	atomic_t digest_rqst;
+	atomic_t aead_rqst;
+	atomic_t complete;
+	atomic_t error;
+	atomic_t fallback;
+	atomic_t tls_pdu_tx;
+	atomic_t tls_pdu_rx;
+	atomic_t tls_key;
+};
+
+#if IS_ENABLED(CONFIG_CHELSIO_IPSEC_INLINE)
+struct ch_ipsec_stats_debug {
+	atomic_t ipsec_cnt;
+};
+#endif
+
+#define OCQ_WIN_OFFSET(pdev, vres) \
+	(pci_resource_len((pdev), 2) - roundup_pow_of_two((vres)->ocq.size))
+
+/*
+ * Block of information the LLD provides to ULDs attaching to a device.
+ */
+struct cxgb4_lld_info {
+	struct pci_dev *pdev;                /* associated PCI device */
+	struct l2t_data *l2t;                /* L2 table */
+	struct tid_info *tids;               /* TID table */
+	struct net_device **ports;           /* device ports */
+	const struct cxgb4_virt_res *vr;     /* assorted HW resources */
+	const unsigned short *mtus;          /* MTU table */
+	const unsigned short *rxq_ids;       /* the ULD's Rx queue ids */
+	const unsigned short *ciq_ids;       /* the ULD's concentrator IQ ids */
+	unsigned short nrxq;                 /* # of Rx queues */
+	unsigned short ntxq;                 /* # of Tx queues */
+	unsigned short nciq;		     /* # of concentrator IQ */
+	unsigned char nchan:4;               /* # of channels */
+	unsigned char nports:4;              /* # of ports */
+	unsigned char wr_cred;               /* WR 16-byte credits */
+	unsigned char adapter_type;          /* type of adapter */
+	unsigned char fw_api_ver;            /* FW API version */
+	unsigned char crypto;                /* crypto support */
+	unsigned int fw_vers;                /* FW version */
+	unsigned int iscsi_iolen;            /* iSCSI max I/O length */
+	unsigned int cclk_ps;                /* Core clock period in psec */
+	unsigned short udb_density;          /* # of user DB/page */
+	unsigned short ucq_density;          /* # of user CQs/page */
+	unsigned int sge_host_page_size;     /* SGE host page size */
+	unsigned short filt_mode;            /* filter optional components */
+	unsigned short tx_modq[NCHAN];       /* maps each tx channel to a */
+					     /* scheduler queue */
+	void __iomem *gts_reg;               /* address of GTS register */
+	void __iomem *db_reg;                /* address of kernel doorbell */
+	int dbfifo_int_thresh;		     /* doorbell fifo int threshold */
+	unsigned int sge_ingpadboundary;     /* SGE ingress padding boundary */
+	unsigned int sge_egrstatuspagesize;  /* SGE egress status page size */
+	unsigned int sge_pktshift;           /* Padding between CPL and */
+					     /*	packet data */
+	unsigned int pf;		     /* Physical Function we're using */
+	bool enable_fw_ofld_conn;            /* Enable connection through fw */
+					     /* WR */
+	unsigned int max_ordird_qp;          /* Max ORD/IRD depth per RDMA QP */
+	unsigned int max_ird_adapter;        /* Max IRD memory per adapter */
+	bool ulptx_memwrite_dsgl;            /* use of T5 DSGL allowed */
+	unsigned int iscsi_tagmask;	     /* iscsi ddp tag mask */
+	unsigned int iscsi_pgsz_order;	     /* iscsi ddp page size orders */
+	unsigned int iscsi_llimit;	     /* chip's iscsi region llimit */
+	unsigned int ulp_crypto;             /* crypto lookaside support */
+	void **iscsi_ppm;		     /* iscsi page pod manager */
+	int nodeid;			     /* device numa node id */
+	bool fr_nsmr_tpte_wr_support;	     /* FW supports FR_NSMR_TPTE_WR */
+	bool write_w_imm_support;         /* FW supports WRITE_WITH_IMMEDIATE */
+	bool write_cmpl_support;             /* FW supports WRITE_CMPL WR */
+};
+
+struct cxgb4_uld_info {
+	char name[IFNAMSIZ];
+	void *handle;
+	unsigned int nrxq;
+	unsigned int rxq_size;
+	unsigned int ntxq;
+	bool ciq;
+	bool lro;
+	void *(*add)(const struct cxgb4_lld_info *p);
+	int (*rx_handler)(void *handle, const __be64 *rsp,
+			  const struct pkt_gl *gl);
+	int (*state_change)(void *handle, enum cxgb4_state new_state);
+	int (*control)(void *handle, enum cxgb4_control control, ...);
+	int (*lro_rx_handler)(void *handle, const __be64 *rsp,
+			      const struct pkt_gl *gl,
+			      struct t4_lro_mgr *lro_mgr,
+			      struct napi_struct *napi);
+	void (*lro_flush)(struct t4_lro_mgr *);
+	int (*tx_handler)(struct sk_buff *skb, struct net_device *dev);
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+	const struct tlsdev_ops *tlsdev_ops;
+#endif
+#if IS_ENABLED(CONFIG_XFRM_OFFLOAD)
+	const struct xfrmdev_ops *xfrmdev_ops;
+#endif
+};
+
+static inline bool cxgb4_is_ktls_skb(struct sk_buff *skb)
+{
+	return skb->sk && tls_is_sk_tx_device_offloaded(skb->sk);
+}
+
+void cxgb4_uld_enable(struct adapter *adap);
+void cxgb4_register_uld(enum cxgb4_uld type, const struct cxgb4_uld_info *p);
+int cxgb4_unregister_uld(enum cxgb4_uld type);
+int cxgb4_ofld_send(struct net_device *dev, struct sk_buff *skb);
+int cxgb4_immdata_send(struct net_device *dev, unsigned int idx,
+		       const void *src, unsigned int len);
+int cxgb4_crypto_send(struct net_device *dev, struct sk_buff *skb);
+unsigned int cxgb4_dbfifo_count(const struct net_device *dev, int lpfifo);
+unsigned int cxgb4_port_chan(const struct net_device *dev);
+unsigned int cxgb4_port_e2cchan(const struct net_device *dev);
+unsigned int cxgb4_port_viid(const struct net_device *dev);
+unsigned int cxgb4_tp_smt_idx(enum chip_type chip, unsigned int viid);
+unsigned int cxgb4_port_idx(const struct net_device *dev);
+unsigned int cxgb4_best_mtu(const unsigned short *mtus, unsigned short mtu,
+			    unsigned int *idx);
+unsigned int cxgb4_best_aligned_mtu(const unsigned short *mtus,
+				    unsigned short header_size,
+				    unsigned short data_size_max,
+				    unsigned short data_size_align,
+				    unsigned int *mtu_idxp);
+void cxgb4_get_tcp_stats(struct pci_dev *pdev, struct tp_tcp_stats *v4,
+			 struct tp_tcp_stats *v6);
+void cxgb4_iscsi_init(struct net_device *dev, unsigned int tag_mask,
+		      const unsigned int *pgsz_order);
+struct sk_buff *cxgb4_pktgl_to_skb(const struct pkt_gl *gl,
+				   unsigned int skb_len, unsigned int pull_len);
+int cxgb4_sync_txq_pidx(struct net_device *dev, u16 qid, u16 pidx, u16 size);
+int cxgb4_flush_eq_cache(struct net_device *dev);
+int cxgb4_read_tpte(struct net_device *dev, u32 stag, __be32 *tpte);
+u64 cxgb4_read_sge_timestamp(struct net_device *dev);
+
+enum cxgb4_bar2_qtype { CXGB4_BAR2_QTYPE_EGRESS, CXGB4_BAR2_QTYPE_INGRESS };
+int cxgb4_bar2_sge_qregs(struct net_device *dev,
+			 unsigned int qid,
+			 enum cxgb4_bar2_qtype qtype,
+			 int user,
+			 u64 *pbar2_qoffset,
+			 unsigned int *pbar2_qid);
+
+#endif  /* !__CXGB4_ULD_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/l2t.c b/drivers/net/ethernet/chelsio/cxgb4/l2t.c
new file mode 100644
index 000000000..a10a6862a
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/l2t.c
@@ -0,0 +1,762 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/skbuff.h>
+#include <linux/netdevice.h>
+#include <linux/if.h>
+#include <linux/if_vlan.h>
+#include <linux/jhash.h>
+#include <linux/module.h>
+#include <linux/debugfs.h>
+#include <linux/seq_file.h>
+#include <net/neighbour.h>
+#include "cxgb4.h"
+#include "l2t.h"
+#include "t4_msg.h"
+#include "t4fw_api.h"
+#include "t4_regs.h"
+#include "t4_values.h"
+
+/* identifies sync vs async L2T_WRITE_REQs */
+#define SYNC_WR_S    12
+#define SYNC_WR_V(x) ((x) << SYNC_WR_S)
+#define SYNC_WR_F    SYNC_WR_V(1)
+
+struct l2t_data {
+	unsigned int l2t_start;     /* start index of our piece of the L2T */
+	unsigned int l2t_size;      /* number of entries in l2tab */
+	rwlock_t lock;
+	atomic_t nfree;             /* number of free entries */
+	struct l2t_entry *rover;    /* starting point for next allocation */
+	struct l2t_entry l2tab[];  /* MUST BE LAST */
+};
+
+static inline unsigned int vlan_prio(const struct l2t_entry *e)
+{
+	return e->vlan >> VLAN_PRIO_SHIFT;
+}
+
+static inline void l2t_hold(struct l2t_data *d, struct l2t_entry *e)
+{
+	if (atomic_add_return(1, &e->refcnt) == 1)  /* 0 -> 1 transition */
+		atomic_dec(&d->nfree);
+}
+
+/*
+ * To avoid having to check address families we do not allow v4 and v6
+ * neighbors to be on the same hash chain.  We keep v4 entries in the first
+ * half of available hash buckets and v6 in the second.  We need at least two
+ * entries in our L2T for this scheme to work.
+ */
+enum {
+	L2T_MIN_HASH_BUCKETS = 2,
+};
+
+static inline unsigned int arp_hash(struct l2t_data *d, const u32 *key,
+				    int ifindex)
+{
+	unsigned int l2t_size_half = d->l2t_size / 2;
+
+	return jhash_2words(*key, ifindex, 0) % l2t_size_half;
+}
+
+static inline unsigned int ipv6_hash(struct l2t_data *d, const u32 *key,
+				     int ifindex)
+{
+	unsigned int l2t_size_half = d->l2t_size / 2;
+	u32 xor = key[0] ^ key[1] ^ key[2] ^ key[3];
+
+	return (l2t_size_half +
+		(jhash_2words(xor, ifindex, 0) % l2t_size_half));
+}
+
+static unsigned int addr_hash(struct l2t_data *d, const u32 *addr,
+			      int addr_len, int ifindex)
+{
+	return addr_len == 4 ? arp_hash(d, addr, ifindex) :
+			       ipv6_hash(d, addr, ifindex);
+}
+
+/*
+ * Checks if an L2T entry is for the given IP/IPv6 address.  It does not check
+ * whether the L2T entry and the address are of the same address family.
+ * Callers ensure an address is only checked against L2T entries of the same
+ * family, something made trivial by the separation of IP and IPv6 hash chains
+ * mentioned above.  Returns 0 if there's a match,
+ */
+static int addreq(const struct l2t_entry *e, const u32 *addr)
+{
+	if (e->v6)
+		return (e->addr[0] ^ addr[0]) | (e->addr[1] ^ addr[1]) |
+		       (e->addr[2] ^ addr[2]) | (e->addr[3] ^ addr[3]);
+	return e->addr[0] ^ addr[0];
+}
+
+static void neigh_replace(struct l2t_entry *e, struct neighbour *n)
+{
+	neigh_hold(n);
+	if (e->neigh)
+		neigh_release(e->neigh);
+	e->neigh = n;
+}
+
+/*
+ * Write an L2T entry.  Must be called with the entry locked.
+ * The write may be synchronous or asynchronous.
+ */
+static int write_l2e(struct adapter *adap, struct l2t_entry *e, int sync)
+{
+	struct l2t_data *d = adap->l2t;
+	unsigned int l2t_idx = e->idx + d->l2t_start;
+	struct sk_buff *skb;
+	struct cpl_l2t_write_req *req;
+
+	skb = alloc_skb(sizeof(*req), GFP_ATOMIC);
+	if (!skb)
+		return -ENOMEM;
+
+	req = __skb_put(skb, sizeof(*req));
+	INIT_TP_WR(req, 0);
+
+	OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_L2T_WRITE_REQ,
+					l2t_idx | (sync ? SYNC_WR_F : 0) |
+					TID_QID_V(adap->sge.fw_evtq.abs_id)));
+	req->params = htons(L2T_W_PORT_V(e->lport) | L2T_W_NOREPLY_V(!sync));
+	req->l2t_idx = htons(l2t_idx);
+	req->vlan = htons(e->vlan);
+	if (e->neigh && !(e->neigh->dev->flags & IFF_LOOPBACK))
+		memcpy(e->dmac, e->neigh->ha, sizeof(e->dmac));
+	memcpy(req->dst_mac, e->dmac, sizeof(req->dst_mac));
+
+	t4_mgmt_tx(adap, skb);
+
+	if (sync && e->state != L2T_STATE_SWITCHING)
+		e->state = L2T_STATE_SYNC_WRITE;
+	return 0;
+}
+
+/*
+ * Send packets waiting in an L2T entry's ARP queue.  Must be called with the
+ * entry locked.
+ */
+static void send_pending(struct adapter *adap, struct l2t_entry *e)
+{
+	struct sk_buff *skb;
+
+	while ((skb = __skb_dequeue(&e->arpq)) != NULL)
+		t4_ofld_send(adap, skb);
+}
+
+/*
+ * Process a CPL_L2T_WRITE_RPL.  Wake up the ARP queue if it completes a
+ * synchronous L2T_WRITE.  Note that the TID in the reply is really the L2T
+ * index it refers to.
+ */
+void do_l2t_write_rpl(struct adapter *adap, const struct cpl_l2t_write_rpl *rpl)
+{
+	struct l2t_data *d = adap->l2t;
+	unsigned int tid = GET_TID(rpl);
+	unsigned int l2t_idx = tid % L2T_SIZE;
+
+	if (unlikely(rpl->status != CPL_ERR_NONE)) {
+		dev_err(adap->pdev_dev,
+			"Unexpected L2T_WRITE_RPL status %u for entry %u\n",
+			rpl->status, l2t_idx);
+		return;
+	}
+
+	if (tid & SYNC_WR_F) {
+		struct l2t_entry *e = &d->l2tab[l2t_idx - d->l2t_start];
+
+		spin_lock(&e->lock);
+		if (e->state != L2T_STATE_SWITCHING) {
+			send_pending(adap, e);
+			e->state = (e->neigh->nud_state & NUD_STALE) ?
+					L2T_STATE_STALE : L2T_STATE_VALID;
+		}
+		spin_unlock(&e->lock);
+	}
+}
+
+/*
+ * Add a packet to an L2T entry's queue of packets awaiting resolution.
+ * Must be called with the entry's lock held.
+ */
+static inline void arpq_enqueue(struct l2t_entry *e, struct sk_buff *skb)
+{
+	__skb_queue_tail(&e->arpq, skb);
+}
+
+int cxgb4_l2t_send(struct net_device *dev, struct sk_buff *skb,
+		   struct l2t_entry *e)
+{
+	struct adapter *adap = netdev2adap(dev);
+
+again:
+	switch (e->state) {
+	case L2T_STATE_STALE:     /* entry is stale, kick off revalidation */
+		neigh_event_send(e->neigh, NULL);
+		spin_lock_bh(&e->lock);
+		if (e->state == L2T_STATE_STALE)
+			e->state = L2T_STATE_VALID;
+		spin_unlock_bh(&e->lock);
+		fallthrough;
+	case L2T_STATE_VALID:     /* fast-path, send the packet on */
+		return t4_ofld_send(adap, skb);
+	case L2T_STATE_RESOLVING:
+	case L2T_STATE_SYNC_WRITE:
+		spin_lock_bh(&e->lock);
+		if (e->state != L2T_STATE_SYNC_WRITE &&
+		    e->state != L2T_STATE_RESOLVING) {
+			spin_unlock_bh(&e->lock);
+			goto again;
+		}
+		arpq_enqueue(e, skb);
+		spin_unlock_bh(&e->lock);
+
+		if (e->state == L2T_STATE_RESOLVING &&
+		    !neigh_event_send(e->neigh, NULL)) {
+			spin_lock_bh(&e->lock);
+			if (e->state == L2T_STATE_RESOLVING &&
+			    !skb_queue_empty(&e->arpq))
+				write_l2e(adap, e, 1);
+			spin_unlock_bh(&e->lock);
+		}
+	}
+	return 0;
+}
+EXPORT_SYMBOL(cxgb4_l2t_send);
+
+/*
+ * Allocate a free L2T entry.  Must be called with l2t_data.lock held.
+ */
+static struct l2t_entry *alloc_l2e(struct l2t_data *d)
+{
+	struct l2t_entry *end, *e, **p;
+
+	if (!atomic_read(&d->nfree))
+		return NULL;
+
+	/* there's definitely a free entry */
+	for (e = d->rover, end = &d->l2tab[d->l2t_size]; e != end; ++e)
+		if (atomic_read(&e->refcnt) == 0)
+			goto found;
+
+	for (e = d->l2tab; atomic_read(&e->refcnt); ++e)
+		;
+found:
+	d->rover = e + 1;
+	atomic_dec(&d->nfree);
+
+	/*
+	 * The entry we found may be an inactive entry that is
+	 * presently in the hash table.  We need to remove it.
+	 */
+	if (e->state < L2T_STATE_SWITCHING)
+		for (p = &d->l2tab[e->hash].first; *p; p = &(*p)->next)
+			if (*p == e) {
+				*p = e->next;
+				e->next = NULL;
+				break;
+			}
+
+	e->state = L2T_STATE_UNUSED;
+	return e;
+}
+
+static struct l2t_entry *find_or_alloc_l2e(struct l2t_data *d, u16 vlan,
+					   u8 port, u8 *dmac)
+{
+	struct l2t_entry *end, *e, **p;
+	struct l2t_entry *first_free = NULL;
+
+	for (e = &d->l2tab[0], end = &d->l2tab[d->l2t_size]; e != end; ++e) {
+		if (atomic_read(&e->refcnt) == 0) {
+			if (!first_free)
+				first_free = e;
+		} else {
+			if (e->state == L2T_STATE_SWITCHING) {
+				if (ether_addr_equal(e->dmac, dmac) &&
+				    (e->vlan == vlan) && (e->lport == port))
+					goto exists;
+			}
+		}
+	}
+
+	if (first_free) {
+		e = first_free;
+		goto found;
+	}
+
+	return NULL;
+
+found:
+	/* The entry we found may be an inactive entry that is
+	 * presently in the hash table.  We need to remove it.
+	 */
+	if (e->state < L2T_STATE_SWITCHING)
+		for (p = &d->l2tab[e->hash].first; *p; p = &(*p)->next)
+			if (*p == e) {
+				*p = e->next;
+				e->next = NULL;
+				break;
+			}
+	e->state = L2T_STATE_UNUSED;
+
+exists:
+	return e;
+}
+
+/* Called when an L2T entry has no more users.  The entry is left in the hash
+ * table since it is likely to be reused but we also bump nfree to indicate
+ * that the entry can be reallocated for a different neighbor.  We also drop
+ * the existing neighbor reference in case the neighbor is going away and is
+ * waiting on our reference.
+ *
+ * Because entries can be reallocated to other neighbors once their ref count
+ * drops to 0 we need to take the entry's lock to avoid races with a new
+ * incarnation.
+ */
+static void _t4_l2e_free(struct l2t_entry *e)
+{
+	struct l2t_data *d;
+
+	if (atomic_read(&e->refcnt) == 0) {  /* hasn't been recycled */
+		if (e->neigh) {
+			neigh_release(e->neigh);
+			e->neigh = NULL;
+		}
+		__skb_queue_purge(&e->arpq);
+	}
+
+	d = container_of(e, struct l2t_data, l2tab[e->idx]);
+	atomic_inc(&d->nfree);
+}
+
+/* Locked version of _t4_l2e_free */
+static void t4_l2e_free(struct l2t_entry *e)
+{
+	struct l2t_data *d;
+
+	spin_lock_bh(&e->lock);
+	if (atomic_read(&e->refcnt) == 0) {  /* hasn't been recycled */
+		if (e->neigh) {
+			neigh_release(e->neigh);
+			e->neigh = NULL;
+		}
+		__skb_queue_purge(&e->arpq);
+	}
+	spin_unlock_bh(&e->lock);
+
+	d = container_of(e, struct l2t_data, l2tab[e->idx]);
+	atomic_inc(&d->nfree);
+}
+
+void cxgb4_l2t_release(struct l2t_entry *e)
+{
+	if (atomic_dec_and_test(&e->refcnt))
+		t4_l2e_free(e);
+}
+EXPORT_SYMBOL(cxgb4_l2t_release);
+
+/*
+ * Update an L2T entry that was previously used for the same next hop as neigh.
+ * Must be called with softirqs disabled.
+ */
+static void reuse_entry(struct l2t_entry *e, struct neighbour *neigh)
+{
+	unsigned int nud_state;
+
+	spin_lock(&e->lock);                /* avoid race with t4_l2t_free */
+	if (neigh != e->neigh)
+		neigh_replace(e, neigh);
+	nud_state = neigh->nud_state;
+	if (memcmp(e->dmac, neigh->ha, sizeof(e->dmac)) ||
+	    !(nud_state & NUD_VALID))
+		e->state = L2T_STATE_RESOLVING;
+	else if (nud_state & NUD_CONNECTED)
+		e->state = L2T_STATE_VALID;
+	else
+		e->state = L2T_STATE_STALE;
+	spin_unlock(&e->lock);
+}
+
+struct l2t_entry *cxgb4_l2t_get(struct l2t_data *d, struct neighbour *neigh,
+				const struct net_device *physdev,
+				unsigned int priority)
+{
+	u8 lport;
+	u16 vlan;
+	struct l2t_entry *e;
+	unsigned int addr_len = neigh->tbl->key_len;
+	u32 *addr = (u32 *)neigh->primary_key;
+	int ifidx = neigh->dev->ifindex;
+	int hash = addr_hash(d, addr, addr_len, ifidx);
+
+	if (neigh->dev->flags & IFF_LOOPBACK)
+		lport = netdev2pinfo(physdev)->tx_chan + 4;
+	else
+		lport = netdev2pinfo(physdev)->lport;
+
+	if (is_vlan_dev(neigh->dev)) {
+		vlan = vlan_dev_vlan_id(neigh->dev);
+		vlan |= vlan_dev_get_egress_qos_mask(neigh->dev, priority);
+	} else {
+		vlan = VLAN_NONE;
+	}
+
+	write_lock_bh(&d->lock);
+	for (e = d->l2tab[hash].first; e; e = e->next)
+		if (!addreq(e, addr) && e->ifindex == ifidx &&
+		    e->vlan == vlan && e->lport == lport) {
+			l2t_hold(d, e);
+			if (atomic_read(&e->refcnt) == 1)
+				reuse_entry(e, neigh);
+			goto done;
+		}
+
+	/* Need to allocate a new entry */
+	e = alloc_l2e(d);
+	if (e) {
+		spin_lock(&e->lock);          /* avoid race with t4_l2t_free */
+		e->state = L2T_STATE_RESOLVING;
+		if (neigh->dev->flags & IFF_LOOPBACK)
+			memcpy(e->dmac, physdev->dev_addr, sizeof(e->dmac));
+		memcpy(e->addr, addr, addr_len);
+		e->ifindex = ifidx;
+		e->hash = hash;
+		e->lport = lport;
+		e->v6 = addr_len == 16;
+		atomic_set(&e->refcnt, 1);
+		neigh_replace(e, neigh);
+		e->vlan = vlan;
+		e->next = d->l2tab[hash].first;
+		d->l2tab[hash].first = e;
+		spin_unlock(&e->lock);
+	}
+done:
+	write_unlock_bh(&d->lock);
+	return e;
+}
+EXPORT_SYMBOL(cxgb4_l2t_get);
+
+u64 cxgb4_select_ntuple(struct net_device *dev,
+			const struct l2t_entry *l2t)
+{
+	struct adapter *adap = netdev2adap(dev);
+	struct tp_params *tp = &adap->params.tp;
+	u64 ntuple = 0;
+
+	/* Initialize each of the fields which we care about which are present
+	 * in the Compressed Filter Tuple.
+	 */
+	if (tp->vlan_shift >= 0 && l2t->vlan != VLAN_NONE)
+		ntuple |= (u64)(FT_VLAN_VLD_F | l2t->vlan) << tp->vlan_shift;
+
+	if (tp->port_shift >= 0)
+		ntuple |= (u64)l2t->lport << tp->port_shift;
+
+	if (tp->protocol_shift >= 0)
+		ntuple |= (u64)IPPROTO_TCP << tp->protocol_shift;
+
+	if (tp->vnic_shift >= 0 && (tp->ingress_config & VNIC_F)) {
+		struct port_info *pi = (struct port_info *)netdev_priv(dev);
+
+		ntuple |= (u64)(FT_VNID_ID_VF_V(pi->vin) |
+				FT_VNID_ID_PF_V(adap->pf) |
+				FT_VNID_ID_VLD_V(pi->vivld)) << tp->vnic_shift;
+	}
+
+	return ntuple;
+}
+EXPORT_SYMBOL(cxgb4_select_ntuple);
+
+/*
+ * Called when the host's neighbor layer makes a change to some entry that is
+ * loaded into the HW L2 table.
+ */
+void t4_l2t_update(struct adapter *adap, struct neighbour *neigh)
+{
+	unsigned int addr_len = neigh->tbl->key_len;
+	u32 *addr = (u32 *) neigh->primary_key;
+	int hash, ifidx = neigh->dev->ifindex;
+	struct sk_buff_head *arpq = NULL;
+	struct l2t_data *d = adap->l2t;
+	struct l2t_entry *e;
+
+	hash = addr_hash(d, addr, addr_len, ifidx);
+	read_lock_bh(&d->lock);
+	for (e = d->l2tab[hash].first; e; e = e->next)
+		if (!addreq(e, addr) && e->ifindex == ifidx) {
+			spin_lock(&e->lock);
+			if (atomic_read(&e->refcnt))
+				goto found;
+			spin_unlock(&e->lock);
+			break;
+		}
+	read_unlock_bh(&d->lock);
+	return;
+
+ found:
+	read_unlock(&d->lock);
+
+	if (neigh != e->neigh)
+		neigh_replace(e, neigh);
+
+	if (e->state == L2T_STATE_RESOLVING) {
+		if (neigh->nud_state & NUD_FAILED) {
+			arpq = &e->arpq;
+		} else if ((neigh->nud_state & (NUD_CONNECTED | NUD_STALE)) &&
+			   !skb_queue_empty(&e->arpq)) {
+			write_l2e(adap, e, 1);
+		}
+	} else {
+		e->state = neigh->nud_state & NUD_CONNECTED ?
+			L2T_STATE_VALID : L2T_STATE_STALE;
+		if (memcmp(e->dmac, neigh->ha, sizeof(e->dmac)))
+			write_l2e(adap, e, 0);
+	}
+
+	if (arpq) {
+		struct sk_buff *skb;
+
+		/* Called when address resolution fails for an L2T
+		 * entry to handle packets on the arpq head. If a
+		 * packet specifies a failure handler it is invoked,
+		 * otherwise the packet is sent to the device.
+		 */
+		while ((skb = __skb_dequeue(&e->arpq)) != NULL) {
+			const struct l2t_skb_cb *cb = L2T_SKB_CB(skb);
+
+			spin_unlock(&e->lock);
+			if (cb->arp_err_handler)
+				cb->arp_err_handler(cb->handle, skb);
+			else
+				t4_ofld_send(adap, skb);
+			spin_lock(&e->lock);
+		}
+	}
+	spin_unlock_bh(&e->lock);
+}
+
+/* Allocate an L2T entry for use by a switching rule.  Such need to be
+ * explicitly freed and while busy they are not on any hash chain, so normal
+ * address resolution updates do not see them.
+ */
+struct l2t_entry *t4_l2t_alloc_switching(struct adapter *adap, u16 vlan,
+					 u8 port, u8 *eth_addr)
+{
+	struct l2t_data *d = adap->l2t;
+	struct l2t_entry *e;
+	int ret;
+
+	write_lock_bh(&d->lock);
+	e = find_or_alloc_l2e(d, vlan, port, eth_addr);
+	if (e) {
+		spin_lock(&e->lock);          /* avoid race with t4_l2t_free */
+		if (!atomic_read(&e->refcnt)) {
+			e->state = L2T_STATE_SWITCHING;
+			e->vlan = vlan;
+			e->lport = port;
+			ether_addr_copy(e->dmac, eth_addr);
+			atomic_set(&e->refcnt, 1);
+			ret = write_l2e(adap, e, 0);
+			if (ret < 0) {
+				_t4_l2e_free(e);
+				spin_unlock(&e->lock);
+				write_unlock_bh(&d->lock);
+				return NULL;
+			}
+		} else {
+			atomic_inc(&e->refcnt);
+		}
+
+		spin_unlock(&e->lock);
+	}
+	write_unlock_bh(&d->lock);
+	return e;
+}
+
+/**
+ * cxgb4_l2t_alloc_switching - Allocates an L2T entry for switch filters
+ * @dev: net_device pointer
+ * @vlan: VLAN Id
+ * @port: Associated port
+ * @dmac: Destination MAC address to add to L2T
+ * Returns pointer to the allocated l2t entry
+ *
+ * Allocates an L2T entry for use by switching rule of a filter
+ */
+struct l2t_entry *cxgb4_l2t_alloc_switching(struct net_device *dev, u16 vlan,
+					    u8 port, u8 *dmac)
+{
+	struct adapter *adap = netdev2adap(dev);
+
+	return t4_l2t_alloc_switching(adap, vlan, port, dmac);
+}
+EXPORT_SYMBOL(cxgb4_l2t_alloc_switching);
+
+struct l2t_data *t4_init_l2t(unsigned int l2t_start, unsigned int l2t_end)
+{
+	unsigned int l2t_size;
+	int i;
+	struct l2t_data *d;
+
+	if (l2t_start >= l2t_end || l2t_end >= L2T_SIZE)
+		return NULL;
+	l2t_size = l2t_end - l2t_start + 1;
+	if (l2t_size < L2T_MIN_HASH_BUCKETS)
+		return NULL;
+
+	d = kvzalloc(struct_size(d, l2tab, l2t_size), GFP_KERNEL);
+	if (!d)
+		return NULL;
+
+	d->l2t_start = l2t_start;
+	d->l2t_size = l2t_size;
+
+	d->rover = d->l2tab;
+	atomic_set(&d->nfree, l2t_size);
+	rwlock_init(&d->lock);
+
+	for (i = 0; i < d->l2t_size; ++i) {
+		d->l2tab[i].idx = i;
+		d->l2tab[i].state = L2T_STATE_UNUSED;
+		spin_lock_init(&d->l2tab[i].lock);
+		atomic_set(&d->l2tab[i].refcnt, 0);
+		skb_queue_head_init(&d->l2tab[i].arpq);
+	}
+	return d;
+}
+
+static inline void *l2t_get_idx(struct seq_file *seq, loff_t pos)
+{
+	struct l2t_data *d = seq->private;
+
+	return pos >= d->l2t_size ? NULL : &d->l2tab[pos];
+}
+
+static void *l2t_seq_start(struct seq_file *seq, loff_t *pos)
+{
+	return *pos ? l2t_get_idx(seq, *pos - 1) : SEQ_START_TOKEN;
+}
+
+static void *l2t_seq_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+	v = l2t_get_idx(seq, *pos);
+	++(*pos);
+	return v;
+}
+
+static void l2t_seq_stop(struct seq_file *seq, void *v)
+{
+}
+
+static char l2e_state(const struct l2t_entry *e)
+{
+	switch (e->state) {
+	case L2T_STATE_VALID: return 'V';
+	case L2T_STATE_STALE: return 'S';
+	case L2T_STATE_SYNC_WRITE: return 'W';
+	case L2T_STATE_RESOLVING:
+		return skb_queue_empty(&e->arpq) ? 'R' : 'A';
+	case L2T_STATE_SWITCHING: return 'X';
+	default:
+		return 'U';
+	}
+}
+
+bool cxgb4_check_l2t_valid(struct l2t_entry *e)
+{
+	bool valid;
+
+	spin_lock(&e->lock);
+	valid = (e->state == L2T_STATE_VALID);
+	spin_unlock(&e->lock);
+	return valid;
+}
+EXPORT_SYMBOL(cxgb4_check_l2t_valid);
+
+static int l2t_seq_show(struct seq_file *seq, void *v)
+{
+	if (v == SEQ_START_TOKEN)
+		seq_puts(seq, " Idx IP address                "
+			 "Ethernet address  VLAN/P LP State Users Port\n");
+	else {
+		char ip[60];
+		struct l2t_data *d = seq->private;
+		struct l2t_entry *e = v;
+
+		spin_lock_bh(&e->lock);
+		if (e->state == L2T_STATE_SWITCHING)
+			ip[0] = '\0';
+		else
+			sprintf(ip, e->v6 ? "%pI6c" : "%pI4", e->addr);
+		seq_printf(seq, "%4u %-25s %17pM %4d %u %2u   %c   %5u %s\n",
+			   e->idx + d->l2t_start, ip, e->dmac,
+			   e->vlan & VLAN_VID_MASK, vlan_prio(e), e->lport,
+			   l2e_state(e), atomic_read(&e->refcnt),
+			   e->neigh ? e->neigh->dev->name : "");
+		spin_unlock_bh(&e->lock);
+	}
+	return 0;
+}
+
+static const struct seq_operations l2t_seq_ops = {
+	.start = l2t_seq_start,
+	.next = l2t_seq_next,
+	.stop = l2t_seq_stop,
+	.show = l2t_seq_show
+};
+
+static int l2t_seq_open(struct inode *inode, struct file *file)
+{
+	int rc = seq_open(file, &l2t_seq_ops);
+
+	if (!rc) {
+		struct adapter *adap = inode->i_private;
+		struct seq_file *seq = file->private_data;
+
+		seq->private = adap->l2t;
+	}
+	return rc;
+}
+
+const struct file_operations t4_l2t_fops = {
+	.owner = THIS_MODULE,
+	.open = l2t_seq_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = seq_release,
+};
diff --git a/drivers/net/ethernet/chelsio/cxgb4/l2t.h b/drivers/net/ethernet/chelsio/cxgb4/l2t.h
new file mode 100644
index 000000000..340fecb28
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/l2t.h
@@ -0,0 +1,128 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_L2T_H
+#define __CXGB4_L2T_H
+
+#include <linux/spinlock.h>
+#include <linux/if_ether.h>
+#include <linux/atomic.h>
+
+#define VLAN_NONE 0xfff
+
+enum { L2T_SIZE = 4096 };     /* # of L2T entries */
+
+enum {
+	L2T_STATE_VALID,      /* entry is up to date */
+	L2T_STATE_STALE,      /* entry may be used but needs revalidation */
+	L2T_STATE_RESOLVING,  /* entry needs address resolution */
+	L2T_STATE_SYNC_WRITE, /* synchronous write of entry underway */
+	L2T_STATE_NOARP,      /* Netdev down or removed*/
+
+	/* when state is one of the below the entry is not hashed */
+	L2T_STATE_SWITCHING,  /* entry is being used by a switching filter */
+	L2T_STATE_UNUSED      /* entry not in use */
+};
+
+struct adapter;
+struct l2t_data;
+struct neighbour;
+struct net_device;
+struct file_operations;
+struct cpl_l2t_write_rpl;
+
+/*
+ * Each L2T entry plays multiple roles.  First of all, it keeps state for the
+ * corresponding entry of the HW L2 table and maintains a queue of offload
+ * packets awaiting address resolution.  Second, it is a node of a hash table
+ * chain, where the nodes of the chain are linked together through their next
+ * pointer.  Finally, each node is a bucket of a hash table, pointing to the
+ * first element in its chain through its first pointer.
+ */
+struct l2t_entry {
+	u16 state;                  /* entry state */
+	u16 idx;                    /* entry index within in-memory table */
+	u32 addr[4];                /* next hop IP or IPv6 address */
+	int ifindex;                /* neighbor's net_device's ifindex */
+	struct neighbour *neigh;    /* associated neighbour */
+	struct l2t_entry *first;    /* start of hash chain */
+	struct l2t_entry *next;     /* next l2t_entry on chain */
+	struct sk_buff_head arpq;   /* packet queue awaiting resolution */
+	spinlock_t lock;
+	atomic_t refcnt;            /* entry reference count */
+	u16 hash;                   /* hash bucket the entry is on */
+	u16 vlan;                   /* VLAN TCI (id: bits 0-11, prio: 13-15 */
+	u8 v6;                      /* whether entry is for IPv6 */
+	u8 lport;                   /* associated offload logical interface */
+	u8 dmac[ETH_ALEN];          /* neighbour's MAC address */
+};
+
+typedef void (*arp_err_handler_t)(void *handle, struct sk_buff *skb);
+
+/*
+ * Callback stored in an skb to handle address resolution failure.
+ */
+struct l2t_skb_cb {
+	void *handle;
+	arp_err_handler_t arp_err_handler;
+};
+
+#define L2T_SKB_CB(skb) ((struct l2t_skb_cb *)(skb)->cb)
+
+static inline void t4_set_arp_err_handler(struct sk_buff *skb, void *handle,
+					  arp_err_handler_t handler)
+{
+	L2T_SKB_CB(skb)->handle = handle;
+	L2T_SKB_CB(skb)->arp_err_handler = handler;
+}
+
+void cxgb4_l2t_release(struct l2t_entry *e);
+int cxgb4_l2t_send(struct net_device *dev, struct sk_buff *skb,
+		   struct l2t_entry *e);
+struct l2t_entry *cxgb4_l2t_get(struct l2t_data *d, struct neighbour *neigh,
+				const struct net_device *physdev,
+				unsigned int priority);
+u64 cxgb4_select_ntuple(struct net_device *dev,
+			const struct l2t_entry *l2t);
+struct l2t_entry *cxgb4_l2t_alloc_switching(struct net_device *dev, u16 vlan,
+					    u8 port, u8 *dmac);
+void t4_l2t_update(struct adapter *adap, struct neighbour *neigh);
+struct l2t_entry *t4_l2t_alloc_switching(struct adapter *adap, u16 vlan,
+					 u8 port, u8 *dmac);
+struct l2t_data *t4_init_l2t(unsigned int l2t_start, unsigned int l2t_end);
+void do_l2t_write_rpl(struct adapter *p, const struct cpl_l2t_write_rpl *rpl);
+bool cxgb4_check_l2t_valid(struct l2t_entry *e);
+
+extern const struct file_operations t4_l2t_fops;
+#endif  /* __CXGB4_L2T_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/sched.c b/drivers/net/ethernet/chelsio/cxgb4/sched.c
new file mode 100644
index 000000000..a1b14468d
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/sched.c
@@ -0,0 +1,693 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/module.h>
+#include <linux/netdevice.h>
+
+#include "cxgb4.h"
+#include "sched.h"
+
+static int t4_sched_class_fw_cmd(struct port_info *pi,
+				 struct ch_sched_params *p,
+				 enum sched_fw_ops op)
+{
+	struct adapter *adap = pi->adapter;
+	struct sched_table *s = pi->sched_tbl;
+	struct sched_class *e;
+	int err = 0;
+
+	e = &s->tab[p->u.params.class];
+	switch (op) {
+	case SCHED_FW_OP_ADD:
+	case SCHED_FW_OP_DEL:
+		err = t4_sched_params(adap, p->type,
+				      p->u.params.level, p->u.params.mode,
+				      p->u.params.rateunit,
+				      p->u.params.ratemode,
+				      p->u.params.channel, e->idx,
+				      p->u.params.minrate, p->u.params.maxrate,
+				      p->u.params.weight, p->u.params.pktsize,
+				      p->u.params.burstsize);
+		break;
+	default:
+		err = -ENOTSUPP;
+		break;
+	}
+
+	return err;
+}
+
+static int t4_sched_bind_unbind_op(struct port_info *pi, void *arg,
+				   enum sched_bind_type type, bool bind)
+{
+	struct adapter *adap = pi->adapter;
+	u32 fw_mnem, fw_class, fw_param;
+	unsigned int pf = adap->pf;
+	unsigned int vf = 0;
+	int err = 0;
+
+	switch (type) {
+	case SCHED_QUEUE: {
+		struct sched_queue_entry *qe;
+
+		qe = (struct sched_queue_entry *)arg;
+
+		/* Create a template for the FW_PARAMS_CMD mnemonic and
+		 * value (TX Scheduling Class in this case).
+		 */
+		fw_mnem = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DMAQ) |
+			   FW_PARAMS_PARAM_X_V(
+				   FW_PARAMS_PARAM_DMAQ_EQ_SCHEDCLASS_ETH));
+		fw_class = bind ? qe->param.class : FW_SCHED_CLS_NONE;
+		fw_param = (fw_mnem | FW_PARAMS_PARAM_YZ_V(qe->cntxt_id));
+
+		pf = adap->pf;
+		vf = 0;
+
+		err = t4_set_params(adap, adap->mbox, pf, vf, 1,
+				    &fw_param, &fw_class);
+		break;
+	}
+	case SCHED_FLOWC: {
+		struct sched_flowc_entry *fe;
+
+		fe = (struct sched_flowc_entry *)arg;
+
+		fw_class = bind ? fe->param.class : FW_SCHED_CLS_NONE;
+		err = cxgb4_ethofld_send_flowc(adap->port[pi->port_id],
+					       fe->param.tid, fw_class);
+		break;
+	}
+	default:
+		err = -ENOTSUPP;
+		break;
+	}
+
+	return err;
+}
+
+static void *t4_sched_entry_lookup(struct port_info *pi,
+				   enum sched_bind_type type,
+				   const u32 val)
+{
+	struct sched_table *s = pi->sched_tbl;
+	struct sched_class *e, *end;
+	void *found = NULL;
+
+	/* Look for an entry with matching @val */
+	end = &s->tab[s->sched_size];
+	for (e = &s->tab[0]; e != end; ++e) {
+		if (e->state == SCHED_STATE_UNUSED ||
+		    e->bind_type != type)
+			continue;
+
+		switch (type) {
+		case SCHED_QUEUE: {
+			struct sched_queue_entry *qe;
+
+			list_for_each_entry(qe, &e->entry_list, list) {
+				if (qe->cntxt_id == val) {
+					found = qe;
+					break;
+				}
+			}
+			break;
+		}
+		case SCHED_FLOWC: {
+			struct sched_flowc_entry *fe;
+
+			list_for_each_entry(fe, &e->entry_list, list) {
+				if (fe->param.tid == val) {
+					found = fe;
+					break;
+				}
+			}
+			break;
+		}
+		default:
+			return NULL;
+		}
+
+		if (found)
+			break;
+	}
+
+	return found;
+}
+
+struct sched_class *cxgb4_sched_queue_lookup(struct net_device *dev,
+					     struct ch_sched_queue *p)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct sched_queue_entry *qe = NULL;
+	struct adapter *adap = pi->adapter;
+	struct sge_eth_txq *txq;
+
+	if (p->queue < 0 || p->queue >= pi->nqsets)
+		return NULL;
+
+	txq = &adap->sge.ethtxq[pi->first_qset + p->queue];
+	qe = t4_sched_entry_lookup(pi, SCHED_QUEUE, txq->q.cntxt_id);
+	return qe ? &pi->sched_tbl->tab[qe->param.class] : NULL;
+}
+
+static int t4_sched_queue_unbind(struct port_info *pi, struct ch_sched_queue *p)
+{
+	struct sched_queue_entry *qe = NULL;
+	struct adapter *adap = pi->adapter;
+	struct sge_eth_txq *txq;
+	struct sched_class *e;
+	int err = 0;
+
+	if (p->queue < 0 || p->queue >= pi->nqsets)
+		return -ERANGE;
+
+	txq = &adap->sge.ethtxq[pi->first_qset + p->queue];
+
+	/* Find the existing entry that the queue is bound to */
+	qe = t4_sched_entry_lookup(pi, SCHED_QUEUE, txq->q.cntxt_id);
+	if (qe) {
+		err = t4_sched_bind_unbind_op(pi, (void *)qe, SCHED_QUEUE,
+					      false);
+		if (err)
+			return err;
+
+		e = &pi->sched_tbl->tab[qe->param.class];
+		list_del(&qe->list);
+		kvfree(qe);
+		if (atomic_dec_and_test(&e->refcnt))
+			cxgb4_sched_class_free(adap->port[pi->port_id], e->idx);
+	}
+	return err;
+}
+
+static int t4_sched_queue_bind(struct port_info *pi, struct ch_sched_queue *p)
+{
+	struct sched_table *s = pi->sched_tbl;
+	struct sched_queue_entry *qe = NULL;
+	struct adapter *adap = pi->adapter;
+	struct sge_eth_txq *txq;
+	struct sched_class *e;
+	unsigned int qid;
+	int err = 0;
+
+	if (p->queue < 0 || p->queue >= pi->nqsets)
+		return -ERANGE;
+
+	qe = kvzalloc(sizeof(struct sched_queue_entry), GFP_KERNEL);
+	if (!qe)
+		return -ENOMEM;
+
+	txq = &adap->sge.ethtxq[pi->first_qset + p->queue];
+	qid = txq->q.cntxt_id;
+
+	/* Unbind queue from any existing class */
+	err = t4_sched_queue_unbind(pi, p);
+	if (err)
+		goto out_err;
+
+	/* Bind queue to specified class */
+	qe->cntxt_id = qid;
+	memcpy(&qe->param, p, sizeof(qe->param));
+
+	e = &s->tab[qe->param.class];
+	err = t4_sched_bind_unbind_op(pi, (void *)qe, SCHED_QUEUE, true);
+	if (err)
+		goto out_err;
+
+	list_add_tail(&qe->list, &e->entry_list);
+	e->bind_type = SCHED_QUEUE;
+	atomic_inc(&e->refcnt);
+	return err;
+
+out_err:
+	kvfree(qe);
+	return err;
+}
+
+static int t4_sched_flowc_unbind(struct port_info *pi, struct ch_sched_flowc *p)
+{
+	struct sched_flowc_entry *fe = NULL;
+	struct adapter *adap = pi->adapter;
+	struct sched_class *e;
+	int err = 0;
+
+	if (p->tid < 0 || p->tid >= adap->tids.neotids)
+		return -ERANGE;
+
+	/* Find the existing entry that the flowc is bound to */
+	fe = t4_sched_entry_lookup(pi, SCHED_FLOWC, p->tid);
+	if (fe) {
+		err = t4_sched_bind_unbind_op(pi, (void *)fe, SCHED_FLOWC,
+					      false);
+		if (err)
+			return err;
+
+		e = &pi->sched_tbl->tab[fe->param.class];
+		list_del(&fe->list);
+		kvfree(fe);
+		if (atomic_dec_and_test(&e->refcnt))
+			cxgb4_sched_class_free(adap->port[pi->port_id], e->idx);
+	}
+	return err;
+}
+
+static int t4_sched_flowc_bind(struct port_info *pi, struct ch_sched_flowc *p)
+{
+	struct sched_table *s = pi->sched_tbl;
+	struct sched_flowc_entry *fe = NULL;
+	struct adapter *adap = pi->adapter;
+	struct sched_class *e;
+	int err = 0;
+
+	if (p->tid < 0 || p->tid >= adap->tids.neotids)
+		return -ERANGE;
+
+	fe = kvzalloc(sizeof(*fe), GFP_KERNEL);
+	if (!fe)
+		return -ENOMEM;
+
+	/* Unbind flowc from any existing class */
+	err = t4_sched_flowc_unbind(pi, p);
+	if (err)
+		goto out_err;
+
+	/* Bind flowc to specified class */
+	memcpy(&fe->param, p, sizeof(fe->param));
+
+	e = &s->tab[fe->param.class];
+	err = t4_sched_bind_unbind_op(pi, (void *)fe, SCHED_FLOWC, true);
+	if (err)
+		goto out_err;
+
+	list_add_tail(&fe->list, &e->entry_list);
+	e->bind_type = SCHED_FLOWC;
+	atomic_inc(&e->refcnt);
+	return err;
+
+out_err:
+	kvfree(fe);
+	return err;
+}
+
+static void t4_sched_class_unbind_all(struct port_info *pi,
+				      struct sched_class *e,
+				      enum sched_bind_type type)
+{
+	if (!e)
+		return;
+
+	switch (type) {
+	case SCHED_QUEUE: {
+		struct sched_queue_entry *qe;
+
+		list_for_each_entry(qe, &e->entry_list, list)
+			t4_sched_queue_unbind(pi, &qe->param);
+		break;
+	}
+	case SCHED_FLOWC: {
+		struct sched_flowc_entry *fe;
+
+		list_for_each_entry(fe, &e->entry_list, list)
+			t4_sched_flowc_unbind(pi, &fe->param);
+		break;
+	}
+	default:
+		break;
+	}
+}
+
+static int t4_sched_class_bind_unbind_op(struct port_info *pi, void *arg,
+					 enum sched_bind_type type, bool bind)
+{
+	int err = 0;
+
+	if (!arg)
+		return -EINVAL;
+
+	switch (type) {
+	case SCHED_QUEUE: {
+		struct ch_sched_queue *qe = (struct ch_sched_queue *)arg;
+
+		if (bind)
+			err = t4_sched_queue_bind(pi, qe);
+		else
+			err = t4_sched_queue_unbind(pi, qe);
+		break;
+	}
+	case SCHED_FLOWC: {
+		struct ch_sched_flowc *fe = (struct ch_sched_flowc *)arg;
+
+		if (bind)
+			err = t4_sched_flowc_bind(pi, fe);
+		else
+			err = t4_sched_flowc_unbind(pi, fe);
+		break;
+	}
+	default:
+		err = -ENOTSUPP;
+		break;
+	}
+
+	return err;
+}
+
+/**
+ * cxgb4_sched_class_bind - Bind an entity to a scheduling class
+ * @dev: net_device pointer
+ * @arg: Entity opaque data
+ * @type: Entity type (Queue)
+ *
+ * Binds an entity (queue) to a scheduling class.  If the entity
+ * is bound to another class, it will be unbound from the other class
+ * and bound to the class specified in @arg.
+ */
+int cxgb4_sched_class_bind(struct net_device *dev, void *arg,
+			   enum sched_bind_type type)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	u8 class_id;
+
+	if (!can_sched(dev))
+		return -ENOTSUPP;
+
+	if (!arg)
+		return -EINVAL;
+
+	switch (type) {
+	case SCHED_QUEUE: {
+		struct ch_sched_queue *qe = (struct ch_sched_queue *)arg;
+
+		class_id = qe->class;
+		break;
+	}
+	case SCHED_FLOWC: {
+		struct ch_sched_flowc *fe = (struct ch_sched_flowc *)arg;
+
+		class_id = fe->class;
+		break;
+	}
+	default:
+		return -ENOTSUPP;
+	}
+
+	if (!valid_class_id(dev, class_id))
+		return -EINVAL;
+
+	if (class_id == SCHED_CLS_NONE)
+		return -ENOTSUPP;
+
+	return t4_sched_class_bind_unbind_op(pi, arg, type, true);
+
+}
+
+/**
+ * cxgb4_sched_class_unbind - Unbind an entity from a scheduling class
+ * @dev: net_device pointer
+ * @arg: Entity opaque data
+ * @type: Entity type (Queue)
+ *
+ * Unbinds an entity (queue) from a scheduling class.
+ */
+int cxgb4_sched_class_unbind(struct net_device *dev, void *arg,
+			     enum sched_bind_type type)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	u8 class_id;
+
+	if (!can_sched(dev))
+		return -ENOTSUPP;
+
+	if (!arg)
+		return -EINVAL;
+
+	switch (type) {
+	case SCHED_QUEUE: {
+		struct ch_sched_queue *qe = (struct ch_sched_queue *)arg;
+
+		class_id = qe->class;
+		break;
+	}
+	case SCHED_FLOWC: {
+		struct ch_sched_flowc *fe = (struct ch_sched_flowc *)arg;
+
+		class_id = fe->class;
+		break;
+	}
+	default:
+		return -ENOTSUPP;
+	}
+
+	if (!valid_class_id(dev, class_id))
+		return -EINVAL;
+
+	return t4_sched_class_bind_unbind_op(pi, arg, type, false);
+}
+
+/* If @p is NULL, fetch any available unused class */
+static struct sched_class *t4_sched_class_lookup(struct port_info *pi,
+						const struct ch_sched_params *p)
+{
+	struct sched_table *s = pi->sched_tbl;
+	struct sched_class *found = NULL;
+	struct sched_class *e, *end;
+
+	if (!p) {
+		/* Get any available unused class */
+		end = &s->tab[s->sched_size];
+		for (e = &s->tab[0]; e != end; ++e) {
+			if (e->state == SCHED_STATE_UNUSED) {
+				found = e;
+				break;
+			}
+		}
+	} else {
+		/* Look for a class with matching scheduling parameters */
+		struct ch_sched_params info;
+		struct ch_sched_params tp;
+
+		memcpy(&tp, p, sizeof(tp));
+		/* Don't try to match class parameter */
+		tp.u.params.class = SCHED_CLS_NONE;
+
+		end = &s->tab[s->sched_size];
+		for (e = &s->tab[0]; e != end; ++e) {
+			if (e->state == SCHED_STATE_UNUSED)
+				continue;
+
+			memcpy(&info, &e->info, sizeof(info));
+			/* Don't try to match class parameter */
+			info.u.params.class = SCHED_CLS_NONE;
+
+			if ((info.type == tp.type) &&
+			    (!memcmp(&info.u.params, &tp.u.params,
+				     sizeof(info.u.params)))) {
+				found = e;
+				break;
+			}
+		}
+	}
+
+	return found;
+}
+
+static struct sched_class *t4_sched_class_alloc(struct port_info *pi,
+						struct ch_sched_params *p)
+{
+	struct sched_class *e = NULL;
+	u8 class_id;
+	int err;
+
+	if (!p)
+		return NULL;
+
+	class_id = p->u.params.class;
+
+	/* Only accept search for existing class with matching params
+	 * or allocation of new class with specified params
+	 */
+	if (class_id != SCHED_CLS_NONE)
+		return NULL;
+
+	/* See if there's an exisiting class with same requested sched
+	 * params. Classes can only be shared among FLOWC types. For
+	 * other types, always request a new class.
+	 */
+	if (p->u.params.mode == SCHED_CLASS_MODE_FLOW)
+		e = t4_sched_class_lookup(pi, p);
+
+	if (!e) {
+		struct ch_sched_params np;
+
+		/* Fetch any available unused class */
+		e = t4_sched_class_lookup(pi, NULL);
+		if (!e)
+			return NULL;
+
+		memcpy(&np, p, sizeof(np));
+		np.u.params.class = e->idx;
+		/* New class */
+		err = t4_sched_class_fw_cmd(pi, &np, SCHED_FW_OP_ADD);
+		if (err)
+			return NULL;
+		memcpy(&e->info, &np, sizeof(e->info));
+		atomic_set(&e->refcnt, 0);
+		e->state = SCHED_STATE_ACTIVE;
+	}
+
+	return e;
+}
+
+/**
+ * cxgb4_sched_class_alloc - allocate a scheduling class
+ * @dev: net_device pointer
+ * @p: new scheduling class to create.
+ *
+ * Returns pointer to the scheduling class created.  If @p is NULL, then
+ * it allocates and returns any available unused scheduling class. If a
+ * scheduling class with matching @p is found, then the matching class is
+ * returned.
+ */
+struct sched_class *cxgb4_sched_class_alloc(struct net_device *dev,
+					    struct ch_sched_params *p)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	u8 class_id;
+
+	if (!can_sched(dev))
+		return NULL;
+
+	class_id = p->u.params.class;
+	if (!valid_class_id(dev, class_id))
+		return NULL;
+
+	return t4_sched_class_alloc(pi, p);
+}
+
+/**
+ * cxgb4_sched_class_free - free a scheduling class
+ * @dev: net_device pointer
+ * @classid: scheduling class id to free
+ *
+ * Frees a scheduling class if there are no users.
+ */
+void cxgb4_sched_class_free(struct net_device *dev, u8 classid)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct sched_table *s = pi->sched_tbl;
+	struct ch_sched_params p;
+	struct sched_class *e;
+	u32 speed;
+	int ret;
+
+	e = &s->tab[classid];
+	if (!atomic_read(&e->refcnt) && e->state != SCHED_STATE_UNUSED) {
+		/* Port based rate limiting needs explicit reset back
+		 * to max rate. But, we'll do explicit reset for all
+		 * types, instead of just port based type, to be on
+		 * the safer side.
+		 */
+		memcpy(&p, &e->info, sizeof(p));
+		/* Always reset mode to 0. Otherwise, FLOWC mode will
+		 * still be enabled even after resetting the traffic
+		 * class.
+		 */
+		p.u.params.mode = 0;
+		p.u.params.minrate = 0;
+		p.u.params.pktsize = 0;
+
+		ret = t4_get_link_params(pi, NULL, &speed, NULL);
+		if (!ret)
+			p.u.params.maxrate = speed * 1000; /* Mbps to Kbps */
+		else
+			p.u.params.maxrate = SCHED_MAX_RATE_KBPS;
+
+		t4_sched_class_fw_cmd(pi, &p, SCHED_FW_OP_DEL);
+
+		e->state = SCHED_STATE_UNUSED;
+		memset(&e->info, 0, sizeof(e->info));
+	}
+}
+
+static void t4_sched_class_free(struct net_device *dev, struct sched_class *e)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+
+	t4_sched_class_unbind_all(pi, e, e->bind_type);
+	cxgb4_sched_class_free(dev, e->idx);
+}
+
+struct sched_table *t4_init_sched(unsigned int sched_size)
+{
+	struct sched_table *s;
+	unsigned int i;
+
+	s = kvzalloc(struct_size(s, tab, sched_size), GFP_KERNEL);
+	if (!s)
+		return NULL;
+
+	s->sched_size = sched_size;
+
+	for (i = 0; i < s->sched_size; i++) {
+		memset(&s->tab[i], 0, sizeof(struct sched_class));
+		s->tab[i].idx = i;
+		s->tab[i].state = SCHED_STATE_UNUSED;
+		INIT_LIST_HEAD(&s->tab[i].entry_list);
+		atomic_set(&s->tab[i].refcnt, 0);
+	}
+	return s;
+}
+
+void t4_cleanup_sched(struct adapter *adap)
+{
+	struct sched_table *s;
+	unsigned int j, i;
+
+	for_each_port(adap, j) {
+		struct port_info *pi = netdev2pinfo(adap->port[j]);
+
+		s = pi->sched_tbl;
+		if (!s)
+			continue;
+
+		for (i = 0; i < s->sched_size; i++) {
+			struct sched_class *e;
+
+			e = &s->tab[i];
+			if (e->state == SCHED_STATE_ACTIVE)
+				t4_sched_class_free(adap->port[j], e);
+		}
+		kvfree(s);
+	}
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/sched.h b/drivers/net/ethernet/chelsio/cxgb4/sched.h
new file mode 100644
index 000000000..5f8b871d7
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/sched.h
@@ -0,0 +1,119 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_SCHED_H
+#define __CXGB4_SCHED_H
+
+#include <linux/spinlock.h>
+#include <linux/atomic.h>
+
+#define SCHED_CLS_NONE 0xff
+
+#define FW_SCHED_CLS_NONE 0xffffffff
+
+/* Max rate that can be set to a scheduling class is 100 Gbps */
+#define SCHED_MAX_RATE_KBPS 100000000U
+
+enum {
+	SCHED_STATE_ACTIVE,
+	SCHED_STATE_UNUSED,
+};
+
+enum sched_fw_ops {
+	SCHED_FW_OP_ADD,
+	SCHED_FW_OP_DEL,
+};
+
+enum sched_bind_type {
+	SCHED_QUEUE,
+	SCHED_FLOWC,
+};
+
+struct sched_queue_entry {
+	struct list_head list;
+	unsigned int cntxt_id;
+	struct ch_sched_queue param;
+};
+
+struct sched_flowc_entry {
+	struct list_head list;
+	struct ch_sched_flowc param;
+};
+
+struct sched_class {
+	u8 state;
+	u8 idx;
+	struct ch_sched_params info;
+	enum sched_bind_type bind_type;
+	struct list_head entry_list;
+	atomic_t refcnt;
+};
+
+struct sched_table {      /* per port scheduling table */
+	u8 sched_size;
+	struct sched_class tab[];
+};
+
+static inline bool can_sched(struct net_device *dev)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+
+	return !pi->sched_tbl ? false : true;
+}
+
+static inline bool valid_class_id(struct net_device *dev, u8 class_id)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+
+	if ((class_id > pi->sched_tbl->sched_size - 1) &&
+	    (class_id != SCHED_CLS_NONE))
+		return false;
+
+	return true;
+}
+
+struct sched_class *cxgb4_sched_queue_lookup(struct net_device *dev,
+					     struct ch_sched_queue *p);
+int cxgb4_sched_class_bind(struct net_device *dev, void *arg,
+			   enum sched_bind_type type);
+int cxgb4_sched_class_unbind(struct net_device *dev, void *arg,
+			     enum sched_bind_type type);
+
+struct sched_class *cxgb4_sched_class_alloc(struct net_device *dev,
+					    struct ch_sched_params *p);
+void cxgb4_sched_class_free(struct net_device *dev, u8 classid);
+
+struct sched_table *t4_init_sched(unsigned int size);
+void t4_cleanup_sched(struct adapter *adap);
+#endif  /* __CXGB4_SCHED_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/sge.c b/drivers/net/ethernet/chelsio/cxgb4/sge.c
new file mode 100644
index 000000000..46809e2d9
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/sge.c
@@ -0,0 +1,5234 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/skbuff.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/if_vlan.h>
+#include <linux/ip.h>
+#include <linux/dma-mapping.h>
+#include <linux/jiffies.h>
+#include <linux/prefetch.h>
+#include <linux/export.h>
+#include <net/xfrm.h>
+#include <net/ipv6.h>
+#include <net/tcp.h>
+#include <net/busy_poll.h>
+#ifdef CONFIG_CHELSIO_T4_FCOE
+#include <scsi/fc/fc_fcoe.h>
+#endif /* CONFIG_CHELSIO_T4_FCOE */
+#include "cxgb4.h"
+#include "t4_regs.h"
+#include "t4_values.h"
+#include "t4_msg.h"
+#include "t4fw_api.h"
+#include "cxgb4_ptp.h"
+#include "cxgb4_uld.h"
+#include "cxgb4_tc_mqprio.h"
+#include "sched.h"
+
+/*
+ * Rx buffer size.  We use largish buffers if possible but settle for single
+ * pages under memory shortage.
+ */
+#if PAGE_SHIFT >= 16
+# define FL_PG_ORDER 0
+#else
+# define FL_PG_ORDER (16 - PAGE_SHIFT)
+#endif
+
+/* RX_PULL_LEN should be <= RX_COPY_THRES */
+#define RX_COPY_THRES    256
+#define RX_PULL_LEN      128
+
+/*
+ * Main body length for sk_buffs used for Rx Ethernet packets with fragments.
+ * Should be >= RX_PULL_LEN but possibly bigger to give pskb_may_pull some room.
+ */
+#define RX_PKT_SKB_LEN   512
+
+/*
+ * Max number of Tx descriptors we clean up at a time.  Should be modest as
+ * freeing skbs isn't cheap and it happens while holding locks.  We just need
+ * to free packets faster than they arrive, we eventually catch up and keep
+ * the amortized cost reasonable.  Must be >= 2 * TXQ_STOP_THRES.  It should
+ * also match the CIDX Flush Threshold.
+ */
+#define MAX_TX_RECLAIM 32
+
+/*
+ * Max number of Rx buffers we replenish at a time.  Again keep this modest,
+ * allocating buffers isn't cheap either.
+ */
+#define MAX_RX_REFILL 16U
+
+/*
+ * Period of the Rx queue check timer.  This timer is infrequent as it has
+ * something to do only when the system experiences severe memory shortage.
+ */
+#define RX_QCHECK_PERIOD (HZ / 2)
+
+/*
+ * Period of the Tx queue check timer.
+ */
+#define TX_QCHECK_PERIOD (HZ / 2)
+
+/*
+ * Max number of Tx descriptors to be reclaimed by the Tx timer.
+ */
+#define MAX_TIMER_TX_RECLAIM 100
+
+/*
+ * Timer index used when backing off due to memory shortage.
+ */
+#define NOMEM_TMR_IDX (SGE_NTIMERS - 1)
+
+/*
+ * Suspension threshold for non-Ethernet Tx queues.  We require enough room
+ * for a full sized WR.
+ */
+#define TXQ_STOP_THRES (SGE_MAX_WR_LEN / sizeof(struct tx_desc))
+
+/*
+ * Max Tx descriptor space we allow for an Ethernet packet to be inlined
+ * into a WR.
+ */
+#define MAX_IMM_TX_PKT_LEN 256
+
+/*
+ * Max size of a WR sent through a control Tx queue.
+ */
+#define MAX_CTRL_WR_LEN SGE_MAX_WR_LEN
+
+struct rx_sw_desc {                /* SW state per Rx descriptor */
+	struct page *page;
+	dma_addr_t dma_addr;
+};
+
+/*
+ * Rx buffer sizes for "useskbs" Free List buffers (one ingress packet pe skb
+ * buffer).  We currently only support two sizes for 1500- and 9000-byte MTUs.
+ * We could easily support more but there doesn't seem to be much need for
+ * that ...
+ */
+#define FL_MTU_SMALL 1500
+#define FL_MTU_LARGE 9000
+
+static inline unsigned int fl_mtu_bufsize(struct adapter *adapter,
+					  unsigned int mtu)
+{
+	struct sge *s = &adapter->sge;
+
+	return ALIGN(s->pktshift + ETH_HLEN + VLAN_HLEN + mtu, s->fl_align);
+}
+
+#define FL_MTU_SMALL_BUFSIZE(adapter) fl_mtu_bufsize(adapter, FL_MTU_SMALL)
+#define FL_MTU_LARGE_BUFSIZE(adapter) fl_mtu_bufsize(adapter, FL_MTU_LARGE)
+
+/*
+ * Bits 0..3 of rx_sw_desc.dma_addr have special meaning.  The hardware uses
+ * these to specify the buffer size as an index into the SGE Free List Buffer
+ * Size register array.  We also use bit 4, when the buffer has been unmapped
+ * for DMA, but this is of course never sent to the hardware and is only used
+ * to prevent double unmappings.  All of the above requires that the Free List
+ * Buffers which we allocate have the bottom 5 bits free (0) -- i.e. are
+ * 32-byte or or a power of 2 greater in alignment.  Since the SGE's minimal
+ * Free List Buffer alignment is 32 bytes, this works out for us ...
+ */
+enum {
+	RX_BUF_FLAGS     = 0x1f,   /* bottom five bits are special */
+	RX_BUF_SIZE      = 0x0f,   /* bottom three bits are for buf sizes */
+	RX_UNMAPPED_BUF  = 0x10,   /* buffer is not mapped */
+
+	/*
+	 * XXX We shouldn't depend on being able to use these indices.
+	 * XXX Especially when some other Master PF has initialized the
+	 * XXX adapter or we use the Firmware Configuration File.  We
+	 * XXX should really search through the Host Buffer Size register
+	 * XXX array for the appropriately sized buffer indices.
+	 */
+	RX_SMALL_PG_BUF  = 0x0,   /* small (PAGE_SIZE) page buffer */
+	RX_LARGE_PG_BUF  = 0x1,   /* buffer large (FL_PG_ORDER) page buffer */
+
+	RX_SMALL_MTU_BUF = 0x2,   /* small MTU buffer */
+	RX_LARGE_MTU_BUF = 0x3,   /* large MTU buffer */
+};
+
+static int timer_pkt_quota[] = {1, 1, 2, 3, 4, 5};
+#define MIN_NAPI_WORK  1
+
+static inline dma_addr_t get_buf_addr(const struct rx_sw_desc *d)
+{
+	return d->dma_addr & ~(dma_addr_t)RX_BUF_FLAGS;
+}
+
+static inline bool is_buf_mapped(const struct rx_sw_desc *d)
+{
+	return !(d->dma_addr & RX_UNMAPPED_BUF);
+}
+
+/**
+ *	txq_avail - return the number of available slots in a Tx queue
+ *	@q: the Tx queue
+ *
+ *	Returns the number of descriptors in a Tx queue available to write new
+ *	packets.
+ */
+static inline unsigned int txq_avail(const struct sge_txq *q)
+{
+	return q->size - 1 - q->in_use;
+}
+
+/**
+ *	fl_cap - return the capacity of a free-buffer list
+ *	@fl: the FL
+ *
+ *	Returns the capacity of a free-buffer list.  The capacity is less than
+ *	the size because one descriptor needs to be left unpopulated, otherwise
+ *	HW will think the FL is empty.
+ */
+static inline unsigned int fl_cap(const struct sge_fl *fl)
+{
+	return fl->size - 8;   /* 1 descriptor = 8 buffers */
+}
+
+/**
+ *	fl_starving - return whether a Free List is starving.
+ *	@adapter: pointer to the adapter
+ *	@fl: the Free List
+ *
+ *	Tests specified Free List to see whether the number of buffers
+ *	available to the hardware has falled below our "starvation"
+ *	threshold.
+ */
+static inline bool fl_starving(const struct adapter *adapter,
+			       const struct sge_fl *fl)
+{
+	const struct sge *s = &adapter->sge;
+
+	return fl->avail - fl->pend_cred <= s->fl_starve_thres;
+}
+
+int cxgb4_map_skb(struct device *dev, const struct sk_buff *skb,
+		  dma_addr_t *addr)
+{
+	const skb_frag_t *fp, *end;
+	const struct skb_shared_info *si;
+
+	*addr = dma_map_single(dev, skb->data, skb_headlen(skb), DMA_TO_DEVICE);
+	if (dma_mapping_error(dev, *addr))
+		goto out_err;
+
+	si = skb_shinfo(skb);
+	end = &si->frags[si->nr_frags];
+
+	for (fp = si->frags; fp < end; fp++) {
+		*++addr = skb_frag_dma_map(dev, fp, 0, skb_frag_size(fp),
+					   DMA_TO_DEVICE);
+		if (dma_mapping_error(dev, *addr))
+			goto unwind;
+	}
+	return 0;
+
+unwind:
+	while (fp-- > si->frags)
+		dma_unmap_page(dev, *--addr, skb_frag_size(fp), DMA_TO_DEVICE);
+
+	dma_unmap_single(dev, addr[-1], skb_headlen(skb), DMA_TO_DEVICE);
+out_err:
+	return -ENOMEM;
+}
+EXPORT_SYMBOL(cxgb4_map_skb);
+
+static void unmap_skb(struct device *dev, const struct sk_buff *skb,
+		      const dma_addr_t *addr)
+{
+	const skb_frag_t *fp, *end;
+	const struct skb_shared_info *si;
+
+	dma_unmap_single(dev, *addr++, skb_headlen(skb), DMA_TO_DEVICE);
+
+	si = skb_shinfo(skb);
+	end = &si->frags[si->nr_frags];
+	for (fp = si->frags; fp < end; fp++)
+		dma_unmap_page(dev, *addr++, skb_frag_size(fp), DMA_TO_DEVICE);
+}
+
+#ifdef CONFIG_NEED_DMA_MAP_STATE
+/**
+ *	deferred_unmap_destructor - unmap a packet when it is freed
+ *	@skb: the packet
+ *
+ *	This is the packet destructor used for Tx packets that need to remain
+ *	mapped until they are freed rather than until their Tx descriptors are
+ *	freed.
+ */
+static void deferred_unmap_destructor(struct sk_buff *skb)
+{
+	unmap_skb(skb->dev->dev.parent, skb, (dma_addr_t *)skb->head);
+}
+#endif
+
+/**
+ *	free_tx_desc - reclaims Tx descriptors and their buffers
+ *	@adap: the adapter
+ *	@q: the Tx queue to reclaim descriptors from
+ *	@n: the number of descriptors to reclaim
+ *	@unmap: whether the buffers should be unmapped for DMA
+ *
+ *	Reclaims Tx descriptors from an SGE Tx queue and frees the associated
+ *	Tx buffers.  Called with the Tx queue lock held.
+ */
+void free_tx_desc(struct adapter *adap, struct sge_txq *q,
+		  unsigned int n, bool unmap)
+{
+	unsigned int cidx = q->cidx;
+	struct tx_sw_desc *d;
+
+	d = &q->sdesc[cidx];
+	while (n--) {
+		if (d->skb) {                       /* an SGL is present */
+			if (unmap && d->addr[0]) {
+				unmap_skb(adap->pdev_dev, d->skb, d->addr);
+				memset(d->addr, 0, sizeof(d->addr));
+			}
+			dev_consume_skb_any(d->skb);
+			d->skb = NULL;
+		}
+		++d;
+		if (++cidx == q->size) {
+			cidx = 0;
+			d = q->sdesc;
+		}
+	}
+	q->cidx = cidx;
+}
+
+/*
+ * Return the number of reclaimable descriptors in a Tx queue.
+ */
+static inline int reclaimable(const struct sge_txq *q)
+{
+	int hw_cidx = ntohs(READ_ONCE(q->stat->cidx));
+	hw_cidx -= q->cidx;
+	return hw_cidx < 0 ? hw_cidx + q->size : hw_cidx;
+}
+
+/**
+ *	reclaim_completed_tx - reclaims completed TX Descriptors
+ *	@adap: the adapter
+ *	@q: the Tx queue to reclaim completed descriptors from
+ *	@maxreclaim: the maximum number of TX Descriptors to reclaim or -1
+ *	@unmap: whether the buffers should be unmapped for DMA
+ *
+ *	Reclaims Tx Descriptors that the SGE has indicated it has processed,
+ *	and frees the associated buffers if possible.  If @max == -1, then
+ *	we'll use a defaiult maximum.  Called with the TX Queue locked.
+ */
+static inline int reclaim_completed_tx(struct adapter *adap, struct sge_txq *q,
+				       int maxreclaim, bool unmap)
+{
+	int reclaim = reclaimable(q);
+
+	if (reclaim) {
+		/*
+		 * Limit the amount of clean up work we do at a time to keep
+		 * the Tx lock hold time O(1).
+		 */
+		if (maxreclaim < 0)
+			maxreclaim = MAX_TX_RECLAIM;
+		if (reclaim > maxreclaim)
+			reclaim = maxreclaim;
+
+		free_tx_desc(adap, q, reclaim, unmap);
+		q->in_use -= reclaim;
+	}
+
+	return reclaim;
+}
+
+/**
+ *	cxgb4_reclaim_completed_tx - reclaims completed Tx descriptors
+ *	@adap: the adapter
+ *	@q: the Tx queue to reclaim completed descriptors from
+ *	@unmap: whether the buffers should be unmapped for DMA
+ *
+ *	Reclaims Tx descriptors that the SGE has indicated it has processed,
+ *	and frees the associated buffers if possible.  Called with the Tx
+ *	queue locked.
+ */
+void cxgb4_reclaim_completed_tx(struct adapter *adap, struct sge_txq *q,
+				bool unmap)
+{
+	(void)reclaim_completed_tx(adap, q, -1, unmap);
+}
+EXPORT_SYMBOL(cxgb4_reclaim_completed_tx);
+
+static inline int get_buf_size(struct adapter *adapter,
+			       const struct rx_sw_desc *d)
+{
+	struct sge *s = &adapter->sge;
+	unsigned int rx_buf_size_idx = d->dma_addr & RX_BUF_SIZE;
+	int buf_size;
+
+	switch (rx_buf_size_idx) {
+	case RX_SMALL_PG_BUF:
+		buf_size = PAGE_SIZE;
+		break;
+
+	case RX_LARGE_PG_BUF:
+		buf_size = PAGE_SIZE << s->fl_pg_order;
+		break;
+
+	case RX_SMALL_MTU_BUF:
+		buf_size = FL_MTU_SMALL_BUFSIZE(adapter);
+		break;
+
+	case RX_LARGE_MTU_BUF:
+		buf_size = FL_MTU_LARGE_BUFSIZE(adapter);
+		break;
+
+	default:
+		BUG();
+	}
+
+	return buf_size;
+}
+
+/**
+ *	free_rx_bufs - free the Rx buffers on an SGE free list
+ *	@adap: the adapter
+ *	@q: the SGE free list to free buffers from
+ *	@n: how many buffers to free
+ *
+ *	Release the next @n buffers on an SGE free-buffer Rx queue.   The
+ *	buffers must be made inaccessible to HW before calling this function.
+ */
+static void free_rx_bufs(struct adapter *adap, struct sge_fl *q, int n)
+{
+	while (n--) {
+		struct rx_sw_desc *d = &q->sdesc[q->cidx];
+
+		if (is_buf_mapped(d))
+			dma_unmap_page(adap->pdev_dev, get_buf_addr(d),
+				       get_buf_size(adap, d),
+				       DMA_FROM_DEVICE);
+		put_page(d->page);
+		d->page = NULL;
+		if (++q->cidx == q->size)
+			q->cidx = 0;
+		q->avail--;
+	}
+}
+
+/**
+ *	unmap_rx_buf - unmap the current Rx buffer on an SGE free list
+ *	@adap: the adapter
+ *	@q: the SGE free list
+ *
+ *	Unmap the current buffer on an SGE free-buffer Rx queue.   The
+ *	buffer must be made inaccessible to HW before calling this function.
+ *
+ *	This is similar to @free_rx_bufs above but does not free the buffer.
+ *	Do note that the FL still loses any further access to the buffer.
+ */
+static void unmap_rx_buf(struct adapter *adap, struct sge_fl *q)
+{
+	struct rx_sw_desc *d = &q->sdesc[q->cidx];
+
+	if (is_buf_mapped(d))
+		dma_unmap_page(adap->pdev_dev, get_buf_addr(d),
+			       get_buf_size(adap, d), DMA_FROM_DEVICE);
+	d->page = NULL;
+	if (++q->cidx == q->size)
+		q->cidx = 0;
+	q->avail--;
+}
+
+static inline void ring_fl_db(struct adapter *adap, struct sge_fl *q)
+{
+	if (q->pend_cred >= 8) {
+		u32 val = adap->params.arch.sge_fl_db;
+
+		if (is_t4(adap->params.chip))
+			val |= PIDX_V(q->pend_cred / 8);
+		else
+			val |= PIDX_T5_V(q->pend_cred / 8);
+
+		/* Make sure all memory writes to the Free List queue are
+		 * committed before we tell the hardware about them.
+		 */
+		wmb();
+
+		/* If we don't have access to the new User Doorbell (T5+), use
+		 * the old doorbell mechanism; otherwise use the new BAR2
+		 * mechanism.
+		 */
+		if (unlikely(q->bar2_addr == NULL)) {
+			t4_write_reg(adap, MYPF_REG(SGE_PF_KDOORBELL_A),
+				     val | QID_V(q->cntxt_id));
+		} else {
+			writel(val | QID_V(q->bar2_qid),
+			       q->bar2_addr + SGE_UDB_KDOORBELL);
+
+			/* This Write memory Barrier will force the write to
+			 * the User Doorbell area to be flushed.
+			 */
+			wmb();
+		}
+		q->pend_cred &= 7;
+	}
+}
+
+static inline void set_rx_sw_desc(struct rx_sw_desc *sd, struct page *pg,
+				  dma_addr_t mapping)
+{
+	sd->page = pg;
+	sd->dma_addr = mapping;      /* includes size low bits */
+}
+
+/**
+ *	refill_fl - refill an SGE Rx buffer ring
+ *	@adap: the adapter
+ *	@q: the ring to refill
+ *	@n: the number of new buffers to allocate
+ *	@gfp: the gfp flags for the allocations
+ *
+ *	(Re)populate an SGE free-buffer queue with up to @n new packet buffers,
+ *	allocated with the supplied gfp flags.  The caller must assure that
+ *	@n does not exceed the queue's capacity.  If afterwards the queue is
+ *	found critically low mark it as starving in the bitmap of starving FLs.
+ *
+ *	Returns the number of buffers allocated.
+ */
+static unsigned int refill_fl(struct adapter *adap, struct sge_fl *q, int n,
+			      gfp_t gfp)
+{
+	struct sge *s = &adap->sge;
+	struct page *pg;
+	dma_addr_t mapping;
+	unsigned int cred = q->avail;
+	__be64 *d = &q->desc[q->pidx];
+	struct rx_sw_desc *sd = &q->sdesc[q->pidx];
+	int node;
+
+#ifdef CONFIG_DEBUG_FS
+	if (test_bit(q->cntxt_id - adap->sge.egr_start, adap->sge.blocked_fl))
+		goto out;
+#endif
+
+	gfp |= __GFP_NOWARN;
+	node = dev_to_node(adap->pdev_dev);
+
+	if (s->fl_pg_order == 0)
+		goto alloc_small_pages;
+
+	/*
+	 * Prefer large buffers
+	 */
+	while (n) {
+		pg = alloc_pages_node(node, gfp | __GFP_COMP, s->fl_pg_order);
+		if (unlikely(!pg)) {
+			q->large_alloc_failed++;
+			break;       /* fall back to single pages */
+		}
+
+		mapping = dma_map_page(adap->pdev_dev, pg, 0,
+				       PAGE_SIZE << s->fl_pg_order,
+				       DMA_FROM_DEVICE);
+		if (unlikely(dma_mapping_error(adap->pdev_dev, mapping))) {
+			__free_pages(pg, s->fl_pg_order);
+			q->mapping_err++;
+			goto out;   /* do not try small pages for this error */
+		}
+		mapping |= RX_LARGE_PG_BUF;
+		*d++ = cpu_to_be64(mapping);
+
+		set_rx_sw_desc(sd, pg, mapping);
+		sd++;
+
+		q->avail++;
+		if (++q->pidx == q->size) {
+			q->pidx = 0;
+			sd = q->sdesc;
+			d = q->desc;
+		}
+		n--;
+	}
+
+alloc_small_pages:
+	while (n--) {
+		pg = alloc_pages_node(node, gfp, 0);
+		if (unlikely(!pg)) {
+			q->alloc_failed++;
+			break;
+		}
+
+		mapping = dma_map_page(adap->pdev_dev, pg, 0, PAGE_SIZE,
+				       DMA_FROM_DEVICE);
+		if (unlikely(dma_mapping_error(adap->pdev_dev, mapping))) {
+			put_page(pg);
+			q->mapping_err++;
+			goto out;
+		}
+		*d++ = cpu_to_be64(mapping);
+
+		set_rx_sw_desc(sd, pg, mapping);
+		sd++;
+
+		q->avail++;
+		if (++q->pidx == q->size) {
+			q->pidx = 0;
+			sd = q->sdesc;
+			d = q->desc;
+		}
+	}
+
+out:	cred = q->avail - cred;
+	q->pend_cred += cred;
+	ring_fl_db(adap, q);
+
+	if (unlikely(fl_starving(adap, q))) {
+		smp_wmb();
+		q->low++;
+		set_bit(q->cntxt_id - adap->sge.egr_start,
+			adap->sge.starving_fl);
+	}
+
+	return cred;
+}
+
+static inline void __refill_fl(struct adapter *adap, struct sge_fl *fl)
+{
+	refill_fl(adap, fl, min(MAX_RX_REFILL, fl_cap(fl) - fl->avail),
+		  GFP_ATOMIC);
+}
+
+/**
+ *	alloc_ring - allocate resources for an SGE descriptor ring
+ *	@dev: the PCI device's core device
+ *	@nelem: the number of descriptors
+ *	@elem_size: the size of each descriptor
+ *	@sw_size: the size of the SW state associated with each ring element
+ *	@phys: the physical address of the allocated ring
+ *	@metadata: address of the array holding the SW state for the ring
+ *	@stat_size: extra space in HW ring for status information
+ *	@node: preferred node for memory allocations
+ *
+ *	Allocates resources for an SGE descriptor ring, such as Tx queues,
+ *	free buffer lists, or response queues.  Each SGE ring requires
+ *	space for its HW descriptors plus, optionally, space for the SW state
+ *	associated with each HW entry (the metadata).  The function returns
+ *	three values: the virtual address for the HW ring (the return value
+ *	of the function), the bus address of the HW ring, and the address
+ *	of the SW ring.
+ */
+static void *alloc_ring(struct device *dev, size_t nelem, size_t elem_size,
+			size_t sw_size, dma_addr_t *phys, void *metadata,
+			size_t stat_size, int node)
+{
+	size_t len = nelem * elem_size + stat_size;
+	void *s = NULL;
+	void *p = dma_alloc_coherent(dev, len, phys, GFP_KERNEL);
+
+	if (!p)
+		return NULL;
+	if (sw_size) {
+		s = kcalloc_node(sw_size, nelem, GFP_KERNEL, node);
+
+		if (!s) {
+			dma_free_coherent(dev, len, p, *phys);
+			return NULL;
+		}
+	}
+	if (metadata)
+		*(void **)metadata = s;
+	return p;
+}
+
+/**
+ *	sgl_len - calculates the size of an SGL of the given capacity
+ *	@n: the number of SGL entries
+ *
+ *	Calculates the number of flits needed for a scatter/gather list that
+ *	can hold the given number of entries.
+ */
+static inline unsigned int sgl_len(unsigned int n)
+{
+	/* A Direct Scatter Gather List uses 32-bit lengths and 64-bit PCI DMA
+	 * addresses.  The DSGL Work Request starts off with a 32-bit DSGL
+	 * ULPTX header, then Length0, then Address0, then, for 1 <= i <= N,
+	 * repeated sequences of { Length[i], Length[i+1], Address[i],
+	 * Address[i+1] } (this ensures that all addresses are on 64-bit
+	 * boundaries).  If N is even, then Length[N+1] should be set to 0 and
+	 * Address[N+1] is omitted.
+	 *
+	 * The following calculation incorporates all of the above.  It's
+	 * somewhat hard to follow but, briefly: the "+2" accounts for the
+	 * first two flits which include the DSGL header, Length0 and
+	 * Address0; the "(3*(n-1))/2" covers the main body of list entries (3
+	 * flits for every pair of the remaining N) +1 if (n-1) is odd; and
+	 * finally the "+((n-1)&1)" adds the one remaining flit needed if
+	 * (n-1) is odd ...
+	 */
+	n--;
+	return (3 * n) / 2 + (n & 1) + 2;
+}
+
+/**
+ *	flits_to_desc - returns the num of Tx descriptors for the given flits
+ *	@n: the number of flits
+ *
+ *	Returns the number of Tx descriptors needed for the supplied number
+ *	of flits.
+ */
+static inline unsigned int flits_to_desc(unsigned int n)
+{
+	BUG_ON(n > SGE_MAX_WR_LEN / 8);
+	return DIV_ROUND_UP(n, 8);
+}
+
+/**
+ *	is_eth_imm - can an Ethernet packet be sent as immediate data?
+ *	@skb: the packet
+ *	@chip_ver: chip version
+ *
+ *	Returns whether an Ethernet packet is small enough to fit as
+ *	immediate data. Return value corresponds to headroom required.
+ */
+static inline int is_eth_imm(const struct sk_buff *skb, unsigned int chip_ver)
+{
+	int hdrlen = 0;
+
+	if (skb->encapsulation && skb_shinfo(skb)->gso_size &&
+	    chip_ver > CHELSIO_T5) {
+		hdrlen = sizeof(struct cpl_tx_tnl_lso);
+		hdrlen += sizeof(struct cpl_tx_pkt_core);
+	} else if (skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4) {
+		return 0;
+	} else {
+		hdrlen = skb_shinfo(skb)->gso_size ?
+			 sizeof(struct cpl_tx_pkt_lso_core) : 0;
+		hdrlen += sizeof(struct cpl_tx_pkt);
+	}
+	if (skb->len <= MAX_IMM_TX_PKT_LEN - hdrlen)
+		return hdrlen;
+	return 0;
+}
+
+/**
+ *	calc_tx_flits - calculate the number of flits for a packet Tx WR
+ *	@skb: the packet
+ *	@chip_ver: chip version
+ *
+ *	Returns the number of flits needed for a Tx WR for the given Ethernet
+ *	packet, including the needed WR and CPL headers.
+ */
+static inline unsigned int calc_tx_flits(const struct sk_buff *skb,
+					 unsigned int chip_ver)
+{
+	unsigned int flits;
+	int hdrlen = is_eth_imm(skb, chip_ver);
+
+	/* If the skb is small enough, we can pump it out as a work request
+	 * with only immediate data.  In that case we just have to have the
+	 * TX Packet header plus the skb data in the Work Request.
+	 */
+
+	if (hdrlen)
+		return DIV_ROUND_UP(skb->len + hdrlen, sizeof(__be64));
+
+	/* Otherwise, we're going to have to construct a Scatter gather list
+	 * of the skb body and fragments.  We also include the flits necessary
+	 * for the TX Packet Work Request and CPL.  We always have a firmware
+	 * Write Header (incorporated as part of the cpl_tx_pkt_lso and
+	 * cpl_tx_pkt structures), followed by either a TX Packet Write CPL
+	 * message or, if we're doing a Large Send Offload, an LSO CPL message
+	 * with an embedded TX Packet Write CPL message.
+	 */
+	flits = sgl_len(skb_shinfo(skb)->nr_frags + 1);
+	if (skb_shinfo(skb)->gso_size) {
+		if (skb->encapsulation && chip_ver > CHELSIO_T5) {
+			hdrlen = sizeof(struct fw_eth_tx_pkt_wr) +
+				 sizeof(struct cpl_tx_tnl_lso);
+		} else if (skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4) {
+			u32 pkt_hdrlen;
+
+			pkt_hdrlen = eth_get_headlen(skb->dev, skb->data,
+						     skb_headlen(skb));
+			hdrlen = sizeof(struct fw_eth_tx_eo_wr) +
+				 round_up(pkt_hdrlen, 16);
+		} else {
+			hdrlen = sizeof(struct fw_eth_tx_pkt_wr) +
+				 sizeof(struct cpl_tx_pkt_lso_core);
+		}
+
+		hdrlen += sizeof(struct cpl_tx_pkt_core);
+		flits += (hdrlen / sizeof(__be64));
+	} else {
+		flits += (sizeof(struct fw_eth_tx_pkt_wr) +
+			  sizeof(struct cpl_tx_pkt_core)) / sizeof(__be64);
+	}
+	return flits;
+}
+
+/**
+ *	calc_tx_descs - calculate the number of Tx descriptors for a packet
+ *	@skb: the packet
+ *	@chip_ver: chip version
+ *
+ *	Returns the number of Tx descriptors needed for the given Ethernet
+ *	packet, including the needed WR and CPL headers.
+ */
+static inline unsigned int calc_tx_descs(const struct sk_buff *skb,
+					 unsigned int chip_ver)
+{
+	return flits_to_desc(calc_tx_flits(skb, chip_ver));
+}
+
+/**
+ *	cxgb4_write_sgl - populate a scatter/gather list for a packet
+ *	@skb: the packet
+ *	@q: the Tx queue we are writing into
+ *	@sgl: starting location for writing the SGL
+ *	@end: points right after the end of the SGL
+ *	@start: start offset into skb main-body data to include in the SGL
+ *	@addr: the list of bus addresses for the SGL elements
+ *
+ *	Generates a gather list for the buffers that make up a packet.
+ *	The caller must provide adequate space for the SGL that will be written.
+ *	The SGL includes all of the packet's page fragments and the data in its
+ *	main body except for the first @start bytes.  @sgl must be 16-byte
+ *	aligned and within a Tx descriptor with available space.  @end points
+ *	right after the end of the SGL but does not account for any potential
+ *	wrap around, i.e., @end > @sgl.
+ */
+void cxgb4_write_sgl(const struct sk_buff *skb, struct sge_txq *q,
+		     struct ulptx_sgl *sgl, u64 *end, unsigned int start,
+		     const dma_addr_t *addr)
+{
+	unsigned int i, len;
+	struct ulptx_sge_pair *to;
+	const struct skb_shared_info *si = skb_shinfo(skb);
+	unsigned int nfrags = si->nr_frags;
+	struct ulptx_sge_pair buf[MAX_SKB_FRAGS / 2 + 1];
+
+	len = skb_headlen(skb) - start;
+	if (likely(len)) {
+		sgl->len0 = htonl(len);
+		sgl->addr0 = cpu_to_be64(addr[0] + start);
+		nfrags++;
+	} else {
+		sgl->len0 = htonl(skb_frag_size(&si->frags[0]));
+		sgl->addr0 = cpu_to_be64(addr[1]);
+	}
+
+	sgl->cmd_nsge = htonl(ULPTX_CMD_V(ULP_TX_SC_DSGL) |
+			      ULPTX_NSGE_V(nfrags));
+	if (likely(--nfrags == 0))
+		return;
+	/*
+	 * Most of the complexity below deals with the possibility we hit the
+	 * end of the queue in the middle of writing the SGL.  For this case
+	 * only we create the SGL in a temporary buffer and then copy it.
+	 */
+	to = (u8 *)end > (u8 *)q->stat ? buf : sgl->sge;
+
+	for (i = (nfrags != si->nr_frags); nfrags >= 2; nfrags -= 2, to++) {
+		to->len[0] = cpu_to_be32(skb_frag_size(&si->frags[i]));
+		to->len[1] = cpu_to_be32(skb_frag_size(&si->frags[++i]));
+		to->addr[0] = cpu_to_be64(addr[i]);
+		to->addr[1] = cpu_to_be64(addr[++i]);
+	}
+	if (nfrags) {
+		to->len[0] = cpu_to_be32(skb_frag_size(&si->frags[i]));
+		to->len[1] = cpu_to_be32(0);
+		to->addr[0] = cpu_to_be64(addr[i + 1]);
+	}
+	if (unlikely((u8 *)end > (u8 *)q->stat)) {
+		unsigned int part0 = (u8 *)q->stat - (u8 *)sgl->sge, part1;
+
+		if (likely(part0))
+			memcpy(sgl->sge, buf, part0);
+		part1 = (u8 *)end - (u8 *)q->stat;
+		memcpy(q->desc, (u8 *)buf + part0, part1);
+		end = (void *)q->desc + part1;
+	}
+	if ((uintptr_t)end & 8)           /* 0-pad to multiple of 16 */
+		*end = 0;
+}
+EXPORT_SYMBOL(cxgb4_write_sgl);
+
+/*	cxgb4_write_partial_sgl - populate SGL for partial packet
+ *	@skb: the packet
+ *	@q: the Tx queue we are writing into
+ *	@sgl: starting location for writing the SGL
+ *	@end: points right after the end of the SGL
+ *	@addr: the list of bus addresses for the SGL elements
+ *	@start: start offset in the SKB where partial data starts
+ *	@len: length of data from @start to send out
+ *
+ *	This API will handle sending out partial data of a skb if required.
+ *	Unlike cxgb4_write_sgl, @start can be any offset into the skb data,
+ *	and @len will decide how much data after @start offset to send out.
+ */
+void cxgb4_write_partial_sgl(const struct sk_buff *skb, struct sge_txq *q,
+			     struct ulptx_sgl *sgl, u64 *end,
+			     const dma_addr_t *addr, u32 start, u32 len)
+{
+	struct ulptx_sge_pair buf[MAX_SKB_FRAGS / 2 + 1] = {0}, *to;
+	u32 frag_size, skb_linear_data_len = skb_headlen(skb);
+	struct skb_shared_info *si = skb_shinfo(skb);
+	u8 i = 0, frag_idx = 0, nfrags = 0;
+	skb_frag_t *frag;
+
+	/* Fill the first SGL either from linear data or from partial
+	 * frag based on @start.
+	 */
+	if (unlikely(start < skb_linear_data_len)) {
+		frag_size = min(len, skb_linear_data_len - start);
+		sgl->len0 = htonl(frag_size);
+		sgl->addr0 = cpu_to_be64(addr[0] + start);
+		len -= frag_size;
+		nfrags++;
+	} else {
+		start -= skb_linear_data_len;
+		frag = &si->frags[frag_idx];
+		frag_size = skb_frag_size(frag);
+		/* find the first frag */
+		while (start >= frag_size) {
+			start -= frag_size;
+			frag_idx++;
+			frag = &si->frags[frag_idx];
+			frag_size = skb_frag_size(frag);
+		}
+
+		frag_size = min(len, skb_frag_size(frag) - start);
+		sgl->len0 = cpu_to_be32(frag_size);
+		sgl->addr0 = cpu_to_be64(addr[frag_idx + 1] + start);
+		len -= frag_size;
+		nfrags++;
+		frag_idx++;
+	}
+
+	/* If the entire partial data fit in one SGL, then send it out
+	 * now.
+	 */
+	if (!len)
+		goto done;
+
+	/* Most of the complexity below deals with the possibility we hit the
+	 * end of the queue in the middle of writing the SGL.  For this case
+	 * only we create the SGL in a temporary buffer and then copy it.
+	 */
+	to = (u8 *)end > (u8 *)q->stat ? buf : sgl->sge;
+
+	/* If the skb couldn't fit in first SGL completely, fill the
+	 * rest of the frags in subsequent SGLs. Note that each SGL
+	 * pair can store 2 frags.
+	 */
+	while (len) {
+		frag_size = min(len, skb_frag_size(&si->frags[frag_idx]));
+		to->len[i & 1] = cpu_to_be32(frag_size);
+		to->addr[i & 1] = cpu_to_be64(addr[frag_idx + 1]);
+		if (i && (i & 1))
+			to++;
+		nfrags++;
+		frag_idx++;
+		i++;
+		len -= frag_size;
+	}
+
+	/* If we ended in an odd boundary, then set the second SGL's
+	 * length in the pair to 0.
+	 */
+	if (i & 1)
+		to->len[1] = cpu_to_be32(0);
+
+	/* Copy from temporary buffer to Tx ring, in case we hit the
+	 * end of the queue in the middle of writing the SGL.
+	 */
+	if (unlikely((u8 *)end > (u8 *)q->stat)) {
+		u32 part0 = (u8 *)q->stat - (u8 *)sgl->sge, part1;
+
+		if (likely(part0))
+			memcpy(sgl->sge, buf, part0);
+		part1 = (u8 *)end - (u8 *)q->stat;
+		memcpy(q->desc, (u8 *)buf + part0, part1);
+		end = (void *)q->desc + part1;
+	}
+
+	/* 0-pad to multiple of 16 */
+	if ((uintptr_t)end & 8)
+		*end = 0;
+done:
+	sgl->cmd_nsge = htonl(ULPTX_CMD_V(ULP_TX_SC_DSGL) |
+			ULPTX_NSGE_V(nfrags));
+}
+EXPORT_SYMBOL(cxgb4_write_partial_sgl);
+
+/* This function copies 64 byte coalesced work request to
+ * memory mapped BAR2 space. For coalesced WR SGE fetches
+ * data from the FIFO instead of from Host.
+ */
+static void cxgb_pio_copy(u64 __iomem *dst, u64 *src)
+{
+	int count = 8;
+
+	while (count) {
+		writeq(*src, dst);
+		src++;
+		dst++;
+		count--;
+	}
+}
+
+/**
+ *	cxgb4_ring_tx_db - check and potentially ring a Tx queue's doorbell
+ *	@adap: the adapter
+ *	@q: the Tx queue
+ *	@n: number of new descriptors to give to HW
+ *
+ *	Ring the doorbel for a Tx queue.
+ */
+inline void cxgb4_ring_tx_db(struct adapter *adap, struct sge_txq *q, int n)
+{
+	/* Make sure that all writes to the TX Descriptors are committed
+	 * before we tell the hardware about them.
+	 */
+	wmb();
+
+	/* If we don't have access to the new User Doorbell (T5+), use the old
+	 * doorbell mechanism; otherwise use the new BAR2 mechanism.
+	 */
+	if (unlikely(q->bar2_addr == NULL)) {
+		u32 val = PIDX_V(n);
+		unsigned long flags;
+
+		/* For T4 we need to participate in the Doorbell Recovery
+		 * mechanism.
+		 */
+		spin_lock_irqsave(&q->db_lock, flags);
+		if (!q->db_disabled)
+			t4_write_reg(adap, MYPF_REG(SGE_PF_KDOORBELL_A),
+				     QID_V(q->cntxt_id) | val);
+		else
+			q->db_pidx_inc += n;
+		q->db_pidx = q->pidx;
+		spin_unlock_irqrestore(&q->db_lock, flags);
+	} else {
+		u32 val = PIDX_T5_V(n);
+
+		/* T4 and later chips share the same PIDX field offset within
+		 * the doorbell, but T5 and later shrank the field in order to
+		 * gain a bit for Doorbell Priority.  The field was absurdly
+		 * large in the first place (14 bits) so we just use the T5
+		 * and later limits and warn if a Queue ID is too large.
+		 */
+		WARN_ON(val & DBPRIO_F);
+
+		/* If we're only writing a single TX Descriptor and we can use
+		 * Inferred QID registers, we can use the Write Combining
+		 * Gather Buffer; otherwise we use the simple doorbell.
+		 */
+		if (n == 1 && q->bar2_qid == 0) {
+			int index = (q->pidx
+				     ? (q->pidx - 1)
+				     : (q->size - 1));
+			u64 *wr = (u64 *)&q->desc[index];
+
+			cxgb_pio_copy((u64 __iomem *)
+				      (q->bar2_addr + SGE_UDB_WCDOORBELL),
+				      wr);
+		} else {
+			writel(val | QID_V(q->bar2_qid),
+			       q->bar2_addr + SGE_UDB_KDOORBELL);
+		}
+
+		/* This Write Memory Barrier will force the write to the User
+		 * Doorbell area to be flushed.  This is needed to prevent
+		 * writes on different CPUs for the same queue from hitting
+		 * the adapter out of order.  This is required when some Work
+		 * Requests take the Write Combine Gather Buffer path (user
+		 * doorbell area offset [SGE_UDB_WCDOORBELL..+63]) and some
+		 * take the traditional path where we simply increment the
+		 * PIDX (User Doorbell area SGE_UDB_KDOORBELL) and have the
+		 * hardware DMA read the actual Work Request.
+		 */
+		wmb();
+	}
+}
+EXPORT_SYMBOL(cxgb4_ring_tx_db);
+
+/**
+ *	cxgb4_inline_tx_skb - inline a packet's data into Tx descriptors
+ *	@skb: the packet
+ *	@q: the Tx queue where the packet will be inlined
+ *	@pos: starting position in the Tx queue where to inline the packet
+ *
+ *	Inline a packet's contents directly into Tx descriptors, starting at
+ *	the given position within the Tx DMA ring.
+ *	Most of the complexity of this operation is dealing with wrap arounds
+ *	in the middle of the packet we want to inline.
+ */
+void cxgb4_inline_tx_skb(const struct sk_buff *skb,
+			 const struct sge_txq *q, void *pos)
+{
+	int left = (void *)q->stat - pos;
+	u64 *p;
+
+	if (likely(skb->len <= left)) {
+		if (likely(!skb->data_len))
+			skb_copy_from_linear_data(skb, pos, skb->len);
+		else
+			skb_copy_bits(skb, 0, pos, skb->len);
+		pos += skb->len;
+	} else {
+		skb_copy_bits(skb, 0, pos, left);
+		skb_copy_bits(skb, left, q->desc, skb->len - left);
+		pos = (void *)q->desc + (skb->len - left);
+	}
+
+	/* 0-pad to multiple of 16 */
+	p = PTR_ALIGN(pos, 8);
+	if ((uintptr_t)p & 8)
+		*p = 0;
+}
+EXPORT_SYMBOL(cxgb4_inline_tx_skb);
+
+static void *inline_tx_skb_header(const struct sk_buff *skb,
+				  const struct sge_txq *q,  void *pos,
+				  int length)
+{
+	u64 *p;
+	int left = (void *)q->stat - pos;
+
+	if (likely(length <= left)) {
+		memcpy(pos, skb->data, length);
+		pos += length;
+	} else {
+		memcpy(pos, skb->data, left);
+		memcpy(q->desc, skb->data + left, length - left);
+		pos = (void *)q->desc + (length - left);
+	}
+	/* 0-pad to multiple of 16 */
+	p = PTR_ALIGN(pos, 8);
+	if ((uintptr_t)p & 8) {
+		*p = 0;
+		return p + 1;
+	}
+	return p;
+}
+
+/*
+ * Figure out what HW csum a packet wants and return the appropriate control
+ * bits.
+ */
+static u64 hwcsum(enum chip_type chip, const struct sk_buff *skb)
+{
+	int csum_type;
+	bool inner_hdr_csum = false;
+	u16 proto, ver;
+
+	if (skb->encapsulation &&
+	    (CHELSIO_CHIP_VERSION(chip) > CHELSIO_T5))
+		inner_hdr_csum = true;
+
+	if (inner_hdr_csum) {
+		ver = inner_ip_hdr(skb)->version;
+		proto = (ver == 4) ? inner_ip_hdr(skb)->protocol :
+			inner_ipv6_hdr(skb)->nexthdr;
+	} else {
+		ver = ip_hdr(skb)->version;
+		proto = (ver == 4) ? ip_hdr(skb)->protocol :
+			ipv6_hdr(skb)->nexthdr;
+	}
+
+	if (ver == 4) {
+		if (proto == IPPROTO_TCP)
+			csum_type = TX_CSUM_TCPIP;
+		else if (proto == IPPROTO_UDP)
+			csum_type = TX_CSUM_UDPIP;
+		else {
+nocsum:			/*
+			 * unknown protocol, disable HW csum
+			 * and hope a bad packet is detected
+			 */
+			return TXPKT_L4CSUM_DIS_F;
+		}
+	} else {
+		/*
+		 * this doesn't work with extension headers
+		 */
+		if (proto == IPPROTO_TCP)
+			csum_type = TX_CSUM_TCPIP6;
+		else if (proto == IPPROTO_UDP)
+			csum_type = TX_CSUM_UDPIP6;
+		else
+			goto nocsum;
+	}
+
+	if (likely(csum_type >= TX_CSUM_TCPIP)) {
+		int eth_hdr_len, l4_len;
+		u64 hdr_len;
+
+		if (inner_hdr_csum) {
+			/* This allows checksum offload for all encapsulated
+			 * packets like GRE etc..
+			 */
+			l4_len = skb_inner_network_header_len(skb);
+			eth_hdr_len = skb_inner_network_offset(skb) - ETH_HLEN;
+		} else {
+			l4_len = skb_network_header_len(skb);
+			eth_hdr_len = skb_network_offset(skb) - ETH_HLEN;
+		}
+		hdr_len = TXPKT_IPHDR_LEN_V(l4_len);
+
+		if (CHELSIO_CHIP_VERSION(chip) <= CHELSIO_T5)
+			hdr_len |= TXPKT_ETHHDR_LEN_V(eth_hdr_len);
+		else
+			hdr_len |= T6_TXPKT_ETHHDR_LEN_V(eth_hdr_len);
+		return TXPKT_CSUM_TYPE_V(csum_type) | hdr_len;
+	} else {
+		int start = skb_transport_offset(skb);
+
+		return TXPKT_CSUM_TYPE_V(csum_type) |
+			TXPKT_CSUM_START_V(start) |
+			TXPKT_CSUM_LOC_V(start + skb->csum_offset);
+	}
+}
+
+static void eth_txq_stop(struct sge_eth_txq *q)
+{
+	netif_tx_stop_queue(q->txq);
+	q->q.stops++;
+}
+
+static inline void txq_advance(struct sge_txq *q, unsigned int n)
+{
+	q->in_use += n;
+	q->pidx += n;
+	if (q->pidx >= q->size)
+		q->pidx -= q->size;
+}
+
+#ifdef CONFIG_CHELSIO_T4_FCOE
+static inline int
+cxgb_fcoe_offload(struct sk_buff *skb, struct adapter *adap,
+		  const struct port_info *pi, u64 *cntrl)
+{
+	const struct cxgb_fcoe *fcoe = &pi->fcoe;
+
+	if (!(fcoe->flags & CXGB_FCOE_ENABLED))
+		return 0;
+
+	if (skb->protocol != htons(ETH_P_FCOE))
+		return 0;
+
+	skb_reset_mac_header(skb);
+	skb->mac_len = sizeof(struct ethhdr);
+
+	skb_set_network_header(skb, skb->mac_len);
+	skb_set_transport_header(skb, skb->mac_len + sizeof(struct fcoe_hdr));
+
+	if (!cxgb_fcoe_sof_eof_supported(adap, skb))
+		return -ENOTSUPP;
+
+	/* FC CRC offload */
+	*cntrl = TXPKT_CSUM_TYPE_V(TX_CSUM_FCOE) |
+		     TXPKT_L4CSUM_DIS_F | TXPKT_IPCSUM_DIS_F |
+		     TXPKT_CSUM_START_V(CXGB_FCOE_TXPKT_CSUM_START) |
+		     TXPKT_CSUM_END_V(CXGB_FCOE_TXPKT_CSUM_END) |
+		     TXPKT_CSUM_LOC_V(CXGB_FCOE_TXPKT_CSUM_END);
+	return 0;
+}
+#endif /* CONFIG_CHELSIO_T4_FCOE */
+
+/* Returns tunnel type if hardware supports offloading of the same.
+ * It is called only for T5 and onwards.
+ */
+enum cpl_tx_tnl_lso_type cxgb_encap_offload_supported(struct sk_buff *skb)
+{
+	u8 l4_hdr = 0;
+	enum cpl_tx_tnl_lso_type tnl_type = TX_TNL_TYPE_OPAQUE;
+	struct port_info *pi = netdev_priv(skb->dev);
+	struct adapter *adapter = pi->adapter;
+
+	if (skb->inner_protocol_type != ENCAP_TYPE_ETHER ||
+	    skb->inner_protocol != htons(ETH_P_TEB))
+		return tnl_type;
+
+	switch (vlan_get_protocol(skb)) {
+	case htons(ETH_P_IP):
+		l4_hdr = ip_hdr(skb)->protocol;
+		break;
+	case htons(ETH_P_IPV6):
+		l4_hdr = ipv6_hdr(skb)->nexthdr;
+		break;
+	default:
+		return tnl_type;
+	}
+
+	switch (l4_hdr) {
+	case IPPROTO_UDP:
+		if (adapter->vxlan_port == udp_hdr(skb)->dest)
+			tnl_type = TX_TNL_TYPE_VXLAN;
+		else if (adapter->geneve_port == udp_hdr(skb)->dest)
+			tnl_type = TX_TNL_TYPE_GENEVE;
+		break;
+	default:
+		return tnl_type;
+	}
+
+	return tnl_type;
+}
+
+static inline void t6_fill_tnl_lso(struct sk_buff *skb,
+				   struct cpl_tx_tnl_lso *tnl_lso,
+				   enum cpl_tx_tnl_lso_type tnl_type)
+{
+	u32 val;
+	int in_eth_xtra_len;
+	int l3hdr_len = skb_network_header_len(skb);
+	int eth_xtra_len = skb_network_offset(skb) - ETH_HLEN;
+	const struct skb_shared_info *ssi = skb_shinfo(skb);
+	bool v6 = (ip_hdr(skb)->version == 6);
+
+	val = CPL_TX_TNL_LSO_OPCODE_V(CPL_TX_TNL_LSO) |
+	      CPL_TX_TNL_LSO_FIRST_F |
+	      CPL_TX_TNL_LSO_LAST_F |
+	      (v6 ? CPL_TX_TNL_LSO_IPV6OUT_F : 0) |
+	      CPL_TX_TNL_LSO_ETHHDRLENOUT_V(eth_xtra_len / 4) |
+	      CPL_TX_TNL_LSO_IPHDRLENOUT_V(l3hdr_len / 4) |
+	      (v6 ? 0 : CPL_TX_TNL_LSO_IPHDRCHKOUT_F) |
+	      CPL_TX_TNL_LSO_IPLENSETOUT_F |
+	      (v6 ? 0 : CPL_TX_TNL_LSO_IPIDINCOUT_F);
+	tnl_lso->op_to_IpIdSplitOut = htonl(val);
+
+	tnl_lso->IpIdOffsetOut = 0;
+
+	/* Get the tunnel header length */
+	val = skb_inner_mac_header(skb) - skb_mac_header(skb);
+	in_eth_xtra_len = skb_inner_network_header(skb) -
+			  skb_inner_mac_header(skb) - ETH_HLEN;
+
+	switch (tnl_type) {
+	case TX_TNL_TYPE_VXLAN:
+	case TX_TNL_TYPE_GENEVE:
+		tnl_lso->UdpLenSetOut_to_TnlHdrLen =
+			htons(CPL_TX_TNL_LSO_UDPCHKCLROUT_F |
+			CPL_TX_TNL_LSO_UDPLENSETOUT_F);
+		break;
+	default:
+		tnl_lso->UdpLenSetOut_to_TnlHdrLen = 0;
+		break;
+	}
+
+	tnl_lso->UdpLenSetOut_to_TnlHdrLen |=
+		 htons(CPL_TX_TNL_LSO_TNLHDRLEN_V(val) |
+		       CPL_TX_TNL_LSO_TNLTYPE_V(tnl_type));
+
+	tnl_lso->r1 = 0;
+
+	val = CPL_TX_TNL_LSO_ETHHDRLEN_V(in_eth_xtra_len / 4) |
+	      CPL_TX_TNL_LSO_IPV6_V(inner_ip_hdr(skb)->version == 6) |
+	      CPL_TX_TNL_LSO_IPHDRLEN_V(skb_inner_network_header_len(skb) / 4) |
+	      CPL_TX_TNL_LSO_TCPHDRLEN_V(inner_tcp_hdrlen(skb) / 4);
+	tnl_lso->Flow_to_TcpHdrLen = htonl(val);
+
+	tnl_lso->IpIdOffset = htons(0);
+
+	tnl_lso->IpIdSplit_to_Mss = htons(CPL_TX_TNL_LSO_MSS_V(ssi->gso_size));
+	tnl_lso->TCPSeqOffset = htonl(0);
+	tnl_lso->EthLenOffset_Size = htonl(CPL_TX_TNL_LSO_SIZE_V(skb->len));
+}
+
+static inline void *write_tso_wr(struct adapter *adap, struct sk_buff *skb,
+				 struct cpl_tx_pkt_lso_core *lso)
+{
+	int eth_xtra_len = skb_network_offset(skb) - ETH_HLEN;
+	int l3hdr_len = skb_network_header_len(skb);
+	const struct skb_shared_info *ssi;
+	bool ipv6 = false;
+
+	ssi = skb_shinfo(skb);
+	if (ssi->gso_type & SKB_GSO_TCPV6)
+		ipv6 = true;
+
+	lso->lso_ctrl = htonl(LSO_OPCODE_V(CPL_TX_PKT_LSO) |
+			      LSO_FIRST_SLICE_F | LSO_LAST_SLICE_F |
+			      LSO_IPV6_V(ipv6) |
+			      LSO_ETHHDR_LEN_V(eth_xtra_len / 4) |
+			      LSO_IPHDR_LEN_V(l3hdr_len / 4) |
+			      LSO_TCPHDR_LEN_V(tcp_hdr(skb)->doff));
+	lso->ipid_ofst = htons(0);
+	lso->mss = htons(ssi->gso_size);
+	lso->seqno_offset = htonl(0);
+	if (is_t4(adap->params.chip))
+		lso->len = htonl(skb->len);
+	else
+		lso->len = htonl(LSO_T5_XFER_SIZE_V(skb->len));
+
+	return (void *)(lso + 1);
+}
+
+/**
+ *	t4_sge_eth_txq_egress_update - handle Ethernet TX Queue update
+ *	@adap: the adapter
+ *	@eq: the Ethernet TX Queue
+ *	@maxreclaim: the maximum number of TX Descriptors to reclaim or -1
+ *
+ *	We're typically called here to update the state of an Ethernet TX
+ *	Queue with respect to the hardware's progress in consuming the TX
+ *	Work Requests that we've put on that Egress Queue.  This happens
+ *	when we get Egress Queue Update messages and also prophylactically
+ *	in regular timer-based Ethernet TX Queue maintenance.
+ */
+int t4_sge_eth_txq_egress_update(struct adapter *adap, struct sge_eth_txq *eq,
+				 int maxreclaim)
+{
+	unsigned int reclaimed, hw_cidx;
+	struct sge_txq *q = &eq->q;
+	int hw_in_use;
+
+	if (!q->in_use || !__netif_tx_trylock(eq->txq))
+		return 0;
+
+	/* Reclaim pending completed TX Descriptors. */
+	reclaimed = reclaim_completed_tx(adap, &eq->q, maxreclaim, true);
+
+	hw_cidx = ntohs(READ_ONCE(q->stat->cidx));
+	hw_in_use = q->pidx - hw_cidx;
+	if (hw_in_use < 0)
+		hw_in_use += q->size;
+
+	/* If the TX Queue is currently stopped and there's now more than half
+	 * the queue available, restart it.  Otherwise bail out since the rest
+	 * of what we want do here is with the possibility of shipping any
+	 * currently buffered Coalesced TX Work Request.
+	 */
+	if (netif_tx_queue_stopped(eq->txq) && hw_in_use < (q->size / 2)) {
+		netif_tx_wake_queue(eq->txq);
+		eq->q.restarts++;
+	}
+
+	__netif_tx_unlock(eq->txq);
+	return reclaimed;
+}
+
+static inline int cxgb4_validate_skb(struct sk_buff *skb,
+				     struct net_device *dev,
+				     u32 min_pkt_len)
+{
+	u32 max_pkt_len;
+
+	/* The chip min packet length is 10 octets but some firmware
+	 * commands have a minimum packet length requirement. So, play
+	 * safe and reject anything shorter than @min_pkt_len.
+	 */
+	if (unlikely(skb->len < min_pkt_len))
+		return -EINVAL;
+
+	/* Discard the packet if the length is greater than mtu */
+	max_pkt_len = ETH_HLEN + dev->mtu;
+
+	if (skb_vlan_tagged(skb))
+		max_pkt_len += VLAN_HLEN;
+
+	if (!skb_shinfo(skb)->gso_size && (unlikely(skb->len > max_pkt_len)))
+		return -EINVAL;
+
+	return 0;
+}
+
+static void *write_eo_udp_wr(struct sk_buff *skb, struct fw_eth_tx_eo_wr *wr,
+			     u32 hdr_len)
+{
+	wr->u.udpseg.type = FW_ETH_TX_EO_TYPE_UDPSEG;
+	wr->u.udpseg.ethlen = skb_network_offset(skb);
+	wr->u.udpseg.iplen = cpu_to_be16(skb_network_header_len(skb));
+	wr->u.udpseg.udplen = sizeof(struct udphdr);
+	wr->u.udpseg.rtplen = 0;
+	wr->u.udpseg.r4 = 0;
+	if (skb_shinfo(skb)->gso_size)
+		wr->u.udpseg.mss = cpu_to_be16(skb_shinfo(skb)->gso_size);
+	else
+		wr->u.udpseg.mss = cpu_to_be16(skb->len - hdr_len);
+	wr->u.udpseg.schedpktsize = wr->u.udpseg.mss;
+	wr->u.udpseg.plen = cpu_to_be32(skb->len - hdr_len);
+
+	return (void *)(wr + 1);
+}
+
+/**
+ *	cxgb4_eth_xmit - add a packet to an Ethernet Tx queue
+ *	@skb: the packet
+ *	@dev: the egress net device
+ *
+ *	Add a packet to an SGE Ethernet Tx queue.  Runs with softirqs disabled.
+ */
+static netdev_tx_t cxgb4_eth_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	enum cpl_tx_tnl_lso_type tnl_type = TX_TNL_TYPE_OPAQUE;
+	bool ptp_enabled = is_ptp_enabled(skb, dev);
+	unsigned int last_desc, flits, ndesc;
+	u32 wr_mid, ctrl0, op, sgl_off = 0;
+	const struct skb_shared_info *ssi;
+	int len, qidx, credits, ret, left;
+	struct tx_sw_desc *sgl_sdesc;
+	struct fw_eth_tx_eo_wr *eowr;
+	struct fw_eth_tx_pkt_wr *wr;
+	struct cpl_tx_pkt_core *cpl;
+	const struct port_info *pi;
+	bool immediate = false;
+	u64 cntrl, *end, *sgl;
+	struct sge_eth_txq *q;
+	unsigned int chip_ver;
+	struct adapter *adap;
+
+	ret = cxgb4_validate_skb(skb, dev, ETH_HLEN);
+	if (ret)
+		goto out_free;
+
+	pi = netdev_priv(dev);
+	adap = pi->adapter;
+	ssi = skb_shinfo(skb);
+#if IS_ENABLED(CONFIG_CHELSIO_IPSEC_INLINE)
+	if (xfrm_offload(skb) && !ssi->gso_size)
+		return adap->uld[CXGB4_ULD_IPSEC].tx_handler(skb, dev);
+#endif /* CHELSIO_IPSEC_INLINE */
+
+#if IS_ENABLED(CONFIG_CHELSIO_TLS_DEVICE)
+	if (cxgb4_is_ktls_skb(skb) &&
+	    (skb->len - skb_tcp_all_headers(skb)))
+		return adap->uld[CXGB4_ULD_KTLS].tx_handler(skb, dev);
+#endif /* CHELSIO_TLS_DEVICE */
+
+	qidx = skb_get_queue_mapping(skb);
+	if (ptp_enabled) {
+		if (!(adap->ptp_tx_skb)) {
+			skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
+			adap->ptp_tx_skb = skb_get(skb);
+		} else {
+			goto out_free;
+		}
+		q = &adap->sge.ptptxq;
+	} else {
+		q = &adap->sge.ethtxq[qidx + pi->first_qset];
+	}
+	skb_tx_timestamp(skb);
+
+	reclaim_completed_tx(adap, &q->q, -1, true);
+	cntrl = TXPKT_L4CSUM_DIS_F | TXPKT_IPCSUM_DIS_F;
+
+#ifdef CONFIG_CHELSIO_T4_FCOE
+	ret = cxgb_fcoe_offload(skb, adap, pi, &cntrl);
+	if (unlikely(ret == -EOPNOTSUPP))
+		goto out_free;
+#endif /* CONFIG_CHELSIO_T4_FCOE */
+
+	chip_ver = CHELSIO_CHIP_VERSION(adap->params.chip);
+	flits = calc_tx_flits(skb, chip_ver);
+	ndesc = flits_to_desc(flits);
+	credits = txq_avail(&q->q) - ndesc;
+
+	if (unlikely(credits < 0)) {
+		eth_txq_stop(q);
+		dev_err(adap->pdev_dev,
+			"%s: Tx ring %u full while queue awake!\n",
+			dev->name, qidx);
+		return NETDEV_TX_BUSY;
+	}
+
+	if (is_eth_imm(skb, chip_ver))
+		immediate = true;
+
+	if (skb->encapsulation && chip_ver > CHELSIO_T5)
+		tnl_type = cxgb_encap_offload_supported(skb);
+
+	last_desc = q->q.pidx + ndesc - 1;
+	if (last_desc >= q->q.size)
+		last_desc -= q->q.size;
+	sgl_sdesc = &q->q.sdesc[last_desc];
+
+	if (!immediate &&
+	    unlikely(cxgb4_map_skb(adap->pdev_dev, skb, sgl_sdesc->addr) < 0)) {
+		memset(sgl_sdesc->addr, 0, sizeof(sgl_sdesc->addr));
+		q->mapping_err++;
+		goto out_free;
+	}
+
+	wr_mid = FW_WR_LEN16_V(DIV_ROUND_UP(flits, 2));
+	if (unlikely(credits < ETHTXQ_STOP_THRES)) {
+		/* After we're done injecting the Work Request for this
+		 * packet, we'll be below our "stop threshold" so stop the TX
+		 * Queue now and schedule a request for an SGE Egress Queue
+		 * Update message. The queue will get started later on when
+		 * the firmware processes this Work Request and sends us an
+		 * Egress Queue Status Update message indicating that space
+		 * has opened up.
+		 */
+		eth_txq_stop(q);
+		if (chip_ver > CHELSIO_T5)
+			wr_mid |= FW_WR_EQUEQ_F | FW_WR_EQUIQ_F;
+	}
+
+	wr = (void *)&q->q.desc[q->q.pidx];
+	eowr = (void *)&q->q.desc[q->q.pidx];
+	wr->equiq_to_len16 = htonl(wr_mid);
+	wr->r3 = cpu_to_be64(0);
+	if (skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4)
+		end = (u64 *)eowr + flits;
+	else
+		end = (u64 *)wr + flits;
+
+	len = immediate ? skb->len : 0;
+	len += sizeof(*cpl);
+	if (ssi->gso_size && !(ssi->gso_type & SKB_GSO_UDP_L4)) {
+		struct cpl_tx_pkt_lso_core *lso = (void *)(wr + 1);
+		struct cpl_tx_tnl_lso *tnl_lso = (void *)(wr + 1);
+
+		if (tnl_type)
+			len += sizeof(*tnl_lso);
+		else
+			len += sizeof(*lso);
+
+		wr->op_immdlen = htonl(FW_WR_OP_V(FW_ETH_TX_PKT_WR) |
+				       FW_WR_IMMDLEN_V(len));
+		if (tnl_type) {
+			struct iphdr *iph = ip_hdr(skb);
+
+			t6_fill_tnl_lso(skb, tnl_lso, tnl_type);
+			cpl = (void *)(tnl_lso + 1);
+			/* Driver is expected to compute partial checksum that
+			 * does not include the IP Total Length.
+			 */
+			if (iph->version == 4) {
+				iph->check = 0;
+				iph->tot_len = 0;
+				iph->check = ~ip_fast_csum((u8 *)iph, iph->ihl);
+			}
+			if (skb->ip_summed == CHECKSUM_PARTIAL)
+				cntrl = hwcsum(adap->params.chip, skb);
+		} else {
+			cpl = write_tso_wr(adap, skb, lso);
+			cntrl = hwcsum(adap->params.chip, skb);
+		}
+		sgl = (u64 *)(cpl + 1); /* sgl start here */
+		q->tso++;
+		q->tx_cso += ssi->gso_segs;
+	} else if (ssi->gso_size) {
+		u64 *start;
+		u32 hdrlen;
+
+		hdrlen = eth_get_headlen(dev, skb->data, skb_headlen(skb));
+		len += hdrlen;
+		wr->op_immdlen = cpu_to_be32(FW_WR_OP_V(FW_ETH_TX_EO_WR) |
+					     FW_ETH_TX_EO_WR_IMMDLEN_V(len));
+		cpl = write_eo_udp_wr(skb, eowr, hdrlen);
+		cntrl = hwcsum(adap->params.chip, skb);
+
+		start = (u64 *)(cpl + 1);
+		sgl = (u64 *)inline_tx_skb_header(skb, &q->q, (void *)start,
+						  hdrlen);
+		if (unlikely(start > sgl)) {
+			left = (u8 *)end - (u8 *)q->q.stat;
+			end = (void *)q->q.desc + left;
+		}
+		sgl_off = hdrlen;
+		q->uso++;
+		q->tx_cso += ssi->gso_segs;
+	} else {
+		if (ptp_enabled)
+			op = FW_PTP_TX_PKT_WR;
+		else
+			op = FW_ETH_TX_PKT_WR;
+		wr->op_immdlen = htonl(FW_WR_OP_V(op) |
+				       FW_WR_IMMDLEN_V(len));
+		cpl = (void *)(wr + 1);
+		sgl = (u64 *)(cpl + 1);
+		if (skb->ip_summed == CHECKSUM_PARTIAL) {
+			cntrl = hwcsum(adap->params.chip, skb) |
+				TXPKT_IPCSUM_DIS_F;
+			q->tx_cso++;
+		}
+	}
+
+	if (unlikely((u8 *)sgl >= (u8 *)q->q.stat)) {
+		/* If current position is already at the end of the
+		 * txq, reset the current to point to start of the queue
+		 * and update the end ptr as well.
+		 */
+		left = (u8 *)end - (u8 *)q->q.stat;
+		end = (void *)q->q.desc + left;
+		sgl = (void *)q->q.desc;
+	}
+
+	if (skb_vlan_tag_present(skb)) {
+		q->vlan_ins++;
+		cntrl |= TXPKT_VLAN_VLD_F | TXPKT_VLAN_V(skb_vlan_tag_get(skb));
+#ifdef CONFIG_CHELSIO_T4_FCOE
+		if (skb->protocol == htons(ETH_P_FCOE))
+			cntrl |= TXPKT_VLAN_V(
+				 ((skb->priority & 0x7) << VLAN_PRIO_SHIFT));
+#endif /* CONFIG_CHELSIO_T4_FCOE */
+	}
+
+	ctrl0 = TXPKT_OPCODE_V(CPL_TX_PKT_XT) | TXPKT_INTF_V(pi->tx_chan) |
+		TXPKT_PF_V(adap->pf);
+	if (ptp_enabled)
+		ctrl0 |= TXPKT_TSTAMP_F;
+#ifdef CONFIG_CHELSIO_T4_DCB
+	if (is_t4(adap->params.chip))
+		ctrl0 |= TXPKT_OVLAN_IDX_V(q->dcb_prio);
+	else
+		ctrl0 |= TXPKT_T5_OVLAN_IDX_V(q->dcb_prio);
+#endif
+	cpl->ctrl0 = htonl(ctrl0);
+	cpl->pack = htons(0);
+	cpl->len = htons(skb->len);
+	cpl->ctrl1 = cpu_to_be64(cntrl);
+
+	if (immediate) {
+		cxgb4_inline_tx_skb(skb, &q->q, sgl);
+		dev_consume_skb_any(skb);
+	} else {
+		cxgb4_write_sgl(skb, &q->q, (void *)sgl, end, sgl_off,
+				sgl_sdesc->addr);
+		skb_orphan(skb);
+		sgl_sdesc->skb = skb;
+	}
+
+	txq_advance(&q->q, ndesc);
+
+	cxgb4_ring_tx_db(adap, &q->q, ndesc);
+	return NETDEV_TX_OK;
+
+out_free:
+	dev_kfree_skb_any(skb);
+	return NETDEV_TX_OK;
+}
+
+/* Constants ... */
+enum {
+	/* Egress Queue sizes, producer and consumer indices are all in units
+	 * of Egress Context Units bytes.  Note that as far as the hardware is
+	 * concerned, the free list is an Egress Queue (the host produces free
+	 * buffers which the hardware consumes) and free list entries are
+	 * 64-bit PCI DMA addresses.
+	 */
+	EQ_UNIT = SGE_EQ_IDXSIZE,
+	FL_PER_EQ_UNIT = EQ_UNIT / sizeof(__be64),
+	TXD_PER_EQ_UNIT = EQ_UNIT / sizeof(__be64),
+
+	T4VF_ETHTXQ_MAX_HDR = (sizeof(struct fw_eth_tx_pkt_vm_wr) +
+			       sizeof(struct cpl_tx_pkt_lso_core) +
+			       sizeof(struct cpl_tx_pkt_core)) / sizeof(__be64),
+};
+
+/**
+ *	t4vf_is_eth_imm - can an Ethernet packet be sent as immediate data?
+ *	@skb: the packet
+ *
+ *	Returns whether an Ethernet packet is small enough to fit completely as
+ *	immediate data.
+ */
+static inline int t4vf_is_eth_imm(const struct sk_buff *skb)
+{
+	/* The VF Driver uses the FW_ETH_TX_PKT_VM_WR firmware Work Request
+	 * which does not accommodate immediate data.  We could dike out all
+	 * of the support code for immediate data but that would tie our hands
+	 * too much if we ever want to enhace the firmware.  It would also
+	 * create more differences between the PF and VF Drivers.
+	 */
+	return false;
+}
+
+/**
+ *	t4vf_calc_tx_flits - calculate the number of flits for a packet TX WR
+ *	@skb: the packet
+ *
+ *	Returns the number of flits needed for a TX Work Request for the
+ *	given Ethernet packet, including the needed WR and CPL headers.
+ */
+static inline unsigned int t4vf_calc_tx_flits(const struct sk_buff *skb)
+{
+	unsigned int flits;
+
+	/* If the skb is small enough, we can pump it out as a work request
+	 * with only immediate data.  In that case we just have to have the
+	 * TX Packet header plus the skb data in the Work Request.
+	 */
+	if (t4vf_is_eth_imm(skb))
+		return DIV_ROUND_UP(skb->len + sizeof(struct cpl_tx_pkt),
+				    sizeof(__be64));
+
+	/* Otherwise, we're going to have to construct a Scatter gather list
+	 * of the skb body and fragments.  We also include the flits necessary
+	 * for the TX Packet Work Request and CPL.  We always have a firmware
+	 * Write Header (incorporated as part of the cpl_tx_pkt_lso and
+	 * cpl_tx_pkt structures), followed by either a TX Packet Write CPL
+	 * message or, if we're doing a Large Send Offload, an LSO CPL message
+	 * with an embedded TX Packet Write CPL message.
+	 */
+	flits = sgl_len(skb_shinfo(skb)->nr_frags + 1);
+	if (skb_shinfo(skb)->gso_size)
+		flits += (sizeof(struct fw_eth_tx_pkt_vm_wr) +
+			  sizeof(struct cpl_tx_pkt_lso_core) +
+			  sizeof(struct cpl_tx_pkt_core)) / sizeof(__be64);
+	else
+		flits += (sizeof(struct fw_eth_tx_pkt_vm_wr) +
+			  sizeof(struct cpl_tx_pkt_core)) / sizeof(__be64);
+	return flits;
+}
+
+/**
+ *	cxgb4_vf_eth_xmit - add a packet to an Ethernet TX queue
+ *	@skb: the packet
+ *	@dev: the egress net device
+ *
+ *	Add a packet to an SGE Ethernet TX queue.  Runs with softirqs disabled.
+ */
+static netdev_tx_t cxgb4_vf_eth_xmit(struct sk_buff *skb,
+				     struct net_device *dev)
+{
+	unsigned int last_desc, flits, ndesc;
+	const struct skb_shared_info *ssi;
+	struct fw_eth_tx_pkt_vm_wr *wr;
+	struct tx_sw_desc *sgl_sdesc;
+	struct cpl_tx_pkt_core *cpl;
+	const struct port_info *pi;
+	struct sge_eth_txq *txq;
+	struct adapter *adapter;
+	int qidx, credits, ret;
+	size_t fw_hdr_copy_len;
+	unsigned int chip_ver;
+	u64 cntrl, *end;
+	u32 wr_mid;
+
+	/* The chip minimum packet length is 10 octets but the firmware
+	 * command that we are using requires that we copy the Ethernet header
+	 * (including the VLAN tag) into the header so we reject anything
+	 * smaller than that ...
+	 */
+	BUILD_BUG_ON(sizeof(wr->firmware) !=
+		     (sizeof(wr->ethmacdst) + sizeof(wr->ethmacsrc) +
+		      sizeof(wr->ethtype) + sizeof(wr->vlantci)));
+	fw_hdr_copy_len = sizeof(wr->firmware);
+	ret = cxgb4_validate_skb(skb, dev, fw_hdr_copy_len);
+	if (ret)
+		goto out_free;
+
+	/* Figure out which TX Queue we're going to use. */
+	pi = netdev_priv(dev);
+	adapter = pi->adapter;
+	qidx = skb_get_queue_mapping(skb);
+	WARN_ON(qidx >= pi->nqsets);
+	txq = &adapter->sge.ethtxq[pi->first_qset + qidx];
+
+	/* Take this opportunity to reclaim any TX Descriptors whose DMA
+	 * transfers have completed.
+	 */
+	reclaim_completed_tx(adapter, &txq->q, -1, true);
+
+	/* Calculate the number of flits and TX Descriptors we're going to
+	 * need along with how many TX Descriptors will be left over after
+	 * we inject our Work Request.
+	 */
+	flits = t4vf_calc_tx_flits(skb);
+	ndesc = flits_to_desc(flits);
+	credits = txq_avail(&txq->q) - ndesc;
+
+	if (unlikely(credits < 0)) {
+		/* Not enough room for this packet's Work Request.  Stop the
+		 * TX Queue and return a "busy" condition.  The queue will get
+		 * started later on when the firmware informs us that space
+		 * has opened up.
+		 */
+		eth_txq_stop(txq);
+		dev_err(adapter->pdev_dev,
+			"%s: TX ring %u full while queue awake!\n",
+			dev->name, qidx);
+		return NETDEV_TX_BUSY;
+	}
+
+	last_desc = txq->q.pidx + ndesc - 1;
+	if (last_desc >= txq->q.size)
+		last_desc -= txq->q.size;
+	sgl_sdesc = &txq->q.sdesc[last_desc];
+
+	if (!t4vf_is_eth_imm(skb) &&
+	    unlikely(cxgb4_map_skb(adapter->pdev_dev, skb,
+				   sgl_sdesc->addr) < 0)) {
+		/* We need to map the skb into PCI DMA space (because it can't
+		 * be in-lined directly into the Work Request) and the mapping
+		 * operation failed.  Record the error and drop the packet.
+		 */
+		memset(sgl_sdesc->addr, 0, sizeof(sgl_sdesc->addr));
+		txq->mapping_err++;
+		goto out_free;
+	}
+
+	chip_ver = CHELSIO_CHIP_VERSION(adapter->params.chip);
+	wr_mid = FW_WR_LEN16_V(DIV_ROUND_UP(flits, 2));
+	if (unlikely(credits < ETHTXQ_STOP_THRES)) {
+		/* After we're done injecting the Work Request for this
+		 * packet, we'll be below our "stop threshold" so stop the TX
+		 * Queue now and schedule a request for an SGE Egress Queue
+		 * Update message.  The queue will get started later on when
+		 * the firmware processes this Work Request and sends us an
+		 * Egress Queue Status Update message indicating that space
+		 * has opened up.
+		 */
+		eth_txq_stop(txq);
+		if (chip_ver > CHELSIO_T5)
+			wr_mid |= FW_WR_EQUEQ_F | FW_WR_EQUIQ_F;
+	}
+
+	/* Start filling in our Work Request.  Note that we do _not_ handle
+	 * the WR Header wrapping around the TX Descriptor Ring.  If our
+	 * maximum header size ever exceeds one TX Descriptor, we'll need to
+	 * do something else here.
+	 */
+	WARN_ON(DIV_ROUND_UP(T4VF_ETHTXQ_MAX_HDR, TXD_PER_EQ_UNIT) > 1);
+	wr = (void *)&txq->q.desc[txq->q.pidx];
+	wr->equiq_to_len16 = cpu_to_be32(wr_mid);
+	wr->r3[0] = cpu_to_be32(0);
+	wr->r3[1] = cpu_to_be32(0);
+	skb_copy_from_linear_data(skb, &wr->firmware, fw_hdr_copy_len);
+	end = (u64 *)wr + flits;
+
+	/* If this is a Large Send Offload packet we'll put in an LSO CPL
+	 * message with an encapsulated TX Packet CPL message.  Otherwise we
+	 * just use a TX Packet CPL message.
+	 */
+	ssi = skb_shinfo(skb);
+	if (ssi->gso_size) {
+		struct cpl_tx_pkt_lso_core *lso = (void *)(wr + 1);
+		bool v6 = (ssi->gso_type & SKB_GSO_TCPV6) != 0;
+		int l3hdr_len = skb_network_header_len(skb);
+		int eth_xtra_len = skb_network_offset(skb) - ETH_HLEN;
+
+		wr->op_immdlen =
+			cpu_to_be32(FW_WR_OP_V(FW_ETH_TX_PKT_VM_WR) |
+				    FW_WR_IMMDLEN_V(sizeof(*lso) +
+						    sizeof(*cpl)));
+		 /* Fill in the LSO CPL message. */
+		lso->lso_ctrl =
+			cpu_to_be32(LSO_OPCODE_V(CPL_TX_PKT_LSO) |
+				    LSO_FIRST_SLICE_F |
+				    LSO_LAST_SLICE_F |
+				    LSO_IPV6_V(v6) |
+				    LSO_ETHHDR_LEN_V(eth_xtra_len / 4) |
+				    LSO_IPHDR_LEN_V(l3hdr_len / 4) |
+				    LSO_TCPHDR_LEN_V(tcp_hdr(skb)->doff));
+		lso->ipid_ofst = cpu_to_be16(0);
+		lso->mss = cpu_to_be16(ssi->gso_size);
+		lso->seqno_offset = cpu_to_be32(0);
+		if (is_t4(adapter->params.chip))
+			lso->len = cpu_to_be32(skb->len);
+		else
+			lso->len = cpu_to_be32(LSO_T5_XFER_SIZE_V(skb->len));
+
+		/* Set up TX Packet CPL pointer, control word and perform
+		 * accounting.
+		 */
+		cpl = (void *)(lso + 1);
+
+		if (chip_ver <= CHELSIO_T5)
+			cntrl = TXPKT_ETHHDR_LEN_V(eth_xtra_len);
+		else
+			cntrl = T6_TXPKT_ETHHDR_LEN_V(eth_xtra_len);
+
+		cntrl |= TXPKT_CSUM_TYPE_V(v6 ?
+					   TX_CSUM_TCPIP6 : TX_CSUM_TCPIP) |
+			 TXPKT_IPHDR_LEN_V(l3hdr_len);
+		txq->tso++;
+		txq->tx_cso += ssi->gso_segs;
+	} else {
+		int len;
+
+		len = (t4vf_is_eth_imm(skb)
+		       ? skb->len + sizeof(*cpl)
+		       : sizeof(*cpl));
+		wr->op_immdlen =
+			cpu_to_be32(FW_WR_OP_V(FW_ETH_TX_PKT_VM_WR) |
+				    FW_WR_IMMDLEN_V(len));
+
+		/* Set up TX Packet CPL pointer, control word and perform
+		 * accounting.
+		 */
+		cpl = (void *)(wr + 1);
+		if (skb->ip_summed == CHECKSUM_PARTIAL) {
+			cntrl = hwcsum(adapter->params.chip, skb) |
+				TXPKT_IPCSUM_DIS_F;
+			txq->tx_cso++;
+		} else {
+			cntrl = TXPKT_L4CSUM_DIS_F | TXPKT_IPCSUM_DIS_F;
+		}
+	}
+
+	/* If there's a VLAN tag present, add that to the list of things to
+	 * do in this Work Request.
+	 */
+	if (skb_vlan_tag_present(skb)) {
+		txq->vlan_ins++;
+		cntrl |= TXPKT_VLAN_VLD_F | TXPKT_VLAN_V(skb_vlan_tag_get(skb));
+	}
+
+	 /* Fill in the TX Packet CPL message header. */
+	cpl->ctrl0 = cpu_to_be32(TXPKT_OPCODE_V(CPL_TX_PKT_XT) |
+				 TXPKT_INTF_V(pi->port_id) |
+				 TXPKT_PF_V(0));
+	cpl->pack = cpu_to_be16(0);
+	cpl->len = cpu_to_be16(skb->len);
+	cpl->ctrl1 = cpu_to_be64(cntrl);
+
+	/* Fill in the body of the TX Packet CPL message with either in-lined
+	 * data or a Scatter/Gather List.
+	 */
+	if (t4vf_is_eth_imm(skb)) {
+		/* In-line the packet's data and free the skb since we don't
+		 * need it any longer.
+		 */
+		cxgb4_inline_tx_skb(skb, &txq->q, cpl + 1);
+		dev_consume_skb_any(skb);
+	} else {
+		/* Write the skb's Scatter/Gather list into the TX Packet CPL
+		 * message and retain a pointer to the skb so we can free it
+		 * later when its DMA completes.  (We store the skb pointer
+		 * in the Software Descriptor corresponding to the last TX
+		 * Descriptor used by the Work Request.)
+		 *
+		 * The retained skb will be freed when the corresponding TX
+		 * Descriptors are reclaimed after their DMAs complete.
+		 * However, this could take quite a while since, in general,
+		 * the hardware is set up to be lazy about sending DMA
+		 * completion notifications to us and we mostly perform TX
+		 * reclaims in the transmit routine.
+		 *
+		 * This is good for performamce but means that we rely on new
+		 * TX packets arriving to run the destructors of completed
+		 * packets, which open up space in their sockets' send queues.
+		 * Sometimes we do not get such new packets causing TX to
+		 * stall.  A single UDP transmitter is a good example of this
+		 * situation.  We have a clean up timer that periodically
+		 * reclaims completed packets but it doesn't run often enough
+		 * (nor do we want it to) to prevent lengthy stalls.  A
+		 * solution to this problem is to run the destructor early,
+		 * after the packet is queued but before it's DMAd.  A con is
+		 * that we lie to socket memory accounting, but the amount of
+		 * extra memory is reasonable (limited by the number of TX
+		 * descriptors), the packets do actually get freed quickly by
+		 * new packets almost always, and for protocols like TCP that
+		 * wait for acks to really free up the data the extra memory
+		 * is even less.  On the positive side we run the destructors
+		 * on the sending CPU rather than on a potentially different
+		 * completing CPU, usually a good thing.
+		 *
+		 * Run the destructor before telling the DMA engine about the
+		 * packet to make sure it doesn't complete and get freed
+		 * prematurely.
+		 */
+		struct ulptx_sgl *sgl = (struct ulptx_sgl *)(cpl + 1);
+		struct sge_txq *tq = &txq->q;
+
+		/* If the Work Request header was an exact multiple of our TX
+		 * Descriptor length, then it's possible that the starting SGL
+		 * pointer lines up exactly with the end of our TX Descriptor
+		 * ring.  If that's the case, wrap around to the beginning
+		 * here ...
+		 */
+		if (unlikely((void *)sgl == (void *)tq->stat)) {
+			sgl = (void *)tq->desc;
+			end = (void *)((void *)tq->desc +
+				       ((void *)end - (void *)tq->stat));
+		}
+
+		cxgb4_write_sgl(skb, tq, sgl, end, 0, sgl_sdesc->addr);
+		skb_orphan(skb);
+		sgl_sdesc->skb = skb;
+	}
+
+	/* Advance our internal TX Queue state, tell the hardware about
+	 * the new TX descriptors and return success.
+	 */
+	txq_advance(&txq->q, ndesc);
+
+	cxgb4_ring_tx_db(adapter, &txq->q, ndesc);
+	return NETDEV_TX_OK;
+
+out_free:
+	/* An error of some sort happened.  Free the TX skb and tell the
+	 * OS that we've "dealt" with the packet ...
+	 */
+	dev_kfree_skb_any(skb);
+	return NETDEV_TX_OK;
+}
+
+/**
+ * reclaim_completed_tx_imm - reclaim completed control-queue Tx descs
+ * @q: the SGE control Tx queue
+ *
+ * This is a variant of cxgb4_reclaim_completed_tx() that is used
+ * for Tx queues that send only immediate data (presently just
+ * the control queues) and	thus do not have any sk_buffs to release.
+ */
+static inline void reclaim_completed_tx_imm(struct sge_txq *q)
+{
+	int hw_cidx = ntohs(READ_ONCE(q->stat->cidx));
+	int reclaim = hw_cidx - q->cidx;
+
+	if (reclaim < 0)
+		reclaim += q->size;
+
+	q->in_use -= reclaim;
+	q->cidx = hw_cidx;
+}
+
+static inline void eosw_txq_advance_index(u32 *idx, u32 n, u32 max)
+{
+	u32 val = *idx + n;
+
+	if (val >= max)
+		val -= max;
+
+	*idx = val;
+}
+
+void cxgb4_eosw_txq_free_desc(struct adapter *adap,
+			      struct sge_eosw_txq *eosw_txq, u32 ndesc)
+{
+	struct tx_sw_desc *d;
+
+	d = &eosw_txq->desc[eosw_txq->last_cidx];
+	while (ndesc--) {
+		if (d->skb) {
+			if (d->addr[0]) {
+				unmap_skb(adap->pdev_dev, d->skb, d->addr);
+				memset(d->addr, 0, sizeof(d->addr));
+			}
+			dev_consume_skb_any(d->skb);
+			d->skb = NULL;
+		}
+		eosw_txq_advance_index(&eosw_txq->last_cidx, 1,
+				       eosw_txq->ndesc);
+		d = &eosw_txq->desc[eosw_txq->last_cidx];
+	}
+}
+
+static inline void eosw_txq_advance(struct sge_eosw_txq *eosw_txq, u32 n)
+{
+	eosw_txq_advance_index(&eosw_txq->pidx, n, eosw_txq->ndesc);
+	eosw_txq->inuse += n;
+}
+
+static inline int eosw_txq_enqueue(struct sge_eosw_txq *eosw_txq,
+				   struct sk_buff *skb)
+{
+	if (eosw_txq->inuse == eosw_txq->ndesc)
+		return -ENOMEM;
+
+	eosw_txq->desc[eosw_txq->pidx].skb = skb;
+	return 0;
+}
+
+static inline struct sk_buff *eosw_txq_peek(struct sge_eosw_txq *eosw_txq)
+{
+	return eosw_txq->desc[eosw_txq->last_pidx].skb;
+}
+
+static inline u8 ethofld_calc_tx_flits(struct adapter *adap,
+				       struct sk_buff *skb, u32 hdr_len)
+{
+	u8 flits, nsgl = 0;
+	u32 wrlen;
+
+	wrlen = sizeof(struct fw_eth_tx_eo_wr) + sizeof(struct cpl_tx_pkt_core);
+	if (skb_shinfo(skb)->gso_size &&
+	    !(skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4))
+		wrlen += sizeof(struct cpl_tx_pkt_lso_core);
+
+	wrlen += roundup(hdr_len, 16);
+
+	/* Packet headers + WR + CPLs */
+	flits = DIV_ROUND_UP(wrlen, 8);
+
+	if (skb_shinfo(skb)->nr_frags > 0) {
+		if (skb_headlen(skb) - hdr_len)
+			nsgl = sgl_len(skb_shinfo(skb)->nr_frags + 1);
+		else
+			nsgl = sgl_len(skb_shinfo(skb)->nr_frags);
+	} else if (skb->len - hdr_len) {
+		nsgl = sgl_len(1);
+	}
+
+	return flits + nsgl;
+}
+
+static void *write_eo_wr(struct adapter *adap, struct sge_eosw_txq *eosw_txq,
+			 struct sk_buff *skb, struct fw_eth_tx_eo_wr *wr,
+			 u32 hdr_len, u32 wrlen)
+{
+	const struct skb_shared_info *ssi = skb_shinfo(skb);
+	struct cpl_tx_pkt_core *cpl;
+	u32 immd_len, wrlen16;
+	bool compl = false;
+	u8 ver, proto;
+
+	ver = ip_hdr(skb)->version;
+	proto = (ver == 6) ? ipv6_hdr(skb)->nexthdr : ip_hdr(skb)->protocol;
+
+	wrlen16 = DIV_ROUND_UP(wrlen, 16);
+	immd_len = sizeof(struct cpl_tx_pkt_core);
+	if (skb_shinfo(skb)->gso_size &&
+	    !(skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4))
+		immd_len += sizeof(struct cpl_tx_pkt_lso_core);
+	immd_len += hdr_len;
+
+	if (!eosw_txq->ncompl ||
+	    (eosw_txq->last_compl + wrlen16) >=
+	    (adap->params.ofldq_wr_cred / 2)) {
+		compl = true;
+		eosw_txq->ncompl++;
+		eosw_txq->last_compl = 0;
+	}
+
+	wr->op_immdlen = cpu_to_be32(FW_WR_OP_V(FW_ETH_TX_EO_WR) |
+				     FW_ETH_TX_EO_WR_IMMDLEN_V(immd_len) |
+				     FW_WR_COMPL_V(compl));
+	wr->equiq_to_len16 = cpu_to_be32(FW_WR_LEN16_V(wrlen16) |
+					 FW_WR_FLOWID_V(eosw_txq->hwtid));
+	wr->r3 = 0;
+	if (proto == IPPROTO_UDP) {
+		cpl = write_eo_udp_wr(skb, wr, hdr_len);
+	} else {
+		wr->u.tcpseg.type = FW_ETH_TX_EO_TYPE_TCPSEG;
+		wr->u.tcpseg.ethlen = skb_network_offset(skb);
+		wr->u.tcpseg.iplen = cpu_to_be16(skb_network_header_len(skb));
+		wr->u.tcpseg.tcplen = tcp_hdrlen(skb);
+		wr->u.tcpseg.tsclk_tsoff = 0;
+		wr->u.tcpseg.r4 = 0;
+		wr->u.tcpseg.r5 = 0;
+		wr->u.tcpseg.plen = cpu_to_be32(skb->len - hdr_len);
+
+		if (ssi->gso_size) {
+			struct cpl_tx_pkt_lso_core *lso = (void *)(wr + 1);
+
+			wr->u.tcpseg.mss = cpu_to_be16(ssi->gso_size);
+			cpl = write_tso_wr(adap, skb, lso);
+		} else {
+			wr->u.tcpseg.mss = cpu_to_be16(0xffff);
+			cpl = (void *)(wr + 1);
+		}
+	}
+
+	eosw_txq->cred -= wrlen16;
+	eosw_txq->last_compl += wrlen16;
+	return cpl;
+}
+
+static int ethofld_hard_xmit(struct net_device *dev,
+			     struct sge_eosw_txq *eosw_txq)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	u32 wrlen, wrlen16, hdr_len, data_len;
+	enum sge_eosw_state next_state;
+	u64 cntrl, *start, *end, *sgl;
+	struct sge_eohw_txq *eohw_txq;
+	struct cpl_tx_pkt_core *cpl;
+	struct fw_eth_tx_eo_wr *wr;
+	bool skip_eotx_wr = false;
+	struct tx_sw_desc *d;
+	struct sk_buff *skb;
+	int left, ret = 0;
+	u8 flits, ndesc;
+
+	eohw_txq = &adap->sge.eohw_txq[eosw_txq->hwqid];
+	spin_lock(&eohw_txq->lock);
+	reclaim_completed_tx_imm(&eohw_txq->q);
+
+	d = &eosw_txq->desc[eosw_txq->last_pidx];
+	skb = d->skb;
+	skb_tx_timestamp(skb);
+
+	wr = (struct fw_eth_tx_eo_wr *)&eohw_txq->q.desc[eohw_txq->q.pidx];
+	if (unlikely(eosw_txq->state != CXGB4_EO_STATE_ACTIVE &&
+		     eosw_txq->last_pidx == eosw_txq->flowc_idx)) {
+		hdr_len = skb->len;
+		data_len = 0;
+		flits = DIV_ROUND_UP(hdr_len, 8);
+		if (eosw_txq->state == CXGB4_EO_STATE_FLOWC_OPEN_SEND)
+			next_state = CXGB4_EO_STATE_FLOWC_OPEN_REPLY;
+		else
+			next_state = CXGB4_EO_STATE_FLOWC_CLOSE_REPLY;
+		skip_eotx_wr = true;
+	} else {
+		hdr_len = eth_get_headlen(dev, skb->data, skb_headlen(skb));
+		data_len = skb->len - hdr_len;
+		flits = ethofld_calc_tx_flits(adap, skb, hdr_len);
+	}
+	ndesc = flits_to_desc(flits);
+	wrlen = flits * 8;
+	wrlen16 = DIV_ROUND_UP(wrlen, 16);
+
+	left = txq_avail(&eohw_txq->q) - ndesc;
+
+	/* If there are no descriptors left in hardware queues or no
+	 * CPL credits left in software queues, then wait for them
+	 * to come back and retry again. Note that we always request
+	 * for credits update via interrupt for every half credits
+	 * consumed. So, the interrupt will eventually restore the
+	 * credits and invoke the Tx path again.
+	 */
+	if (unlikely(left < 0 || wrlen16 > eosw_txq->cred)) {
+		ret = -ENOMEM;
+		goto out_unlock;
+	}
+
+	if (unlikely(skip_eotx_wr)) {
+		start = (u64 *)wr;
+		eosw_txq->state = next_state;
+		eosw_txq->cred -= wrlen16;
+		eosw_txq->ncompl++;
+		eosw_txq->last_compl = 0;
+		goto write_wr_headers;
+	}
+
+	cpl = write_eo_wr(adap, eosw_txq, skb, wr, hdr_len, wrlen);
+	cntrl = hwcsum(adap->params.chip, skb);
+	if (skb_vlan_tag_present(skb))
+		cntrl |= TXPKT_VLAN_VLD_F | TXPKT_VLAN_V(skb_vlan_tag_get(skb));
+
+	cpl->ctrl0 = cpu_to_be32(TXPKT_OPCODE_V(CPL_TX_PKT_XT) |
+				 TXPKT_INTF_V(pi->tx_chan) |
+				 TXPKT_PF_V(adap->pf));
+	cpl->pack = 0;
+	cpl->len = cpu_to_be16(skb->len);
+	cpl->ctrl1 = cpu_to_be64(cntrl);
+
+	start = (u64 *)(cpl + 1);
+
+write_wr_headers:
+	sgl = (u64 *)inline_tx_skb_header(skb, &eohw_txq->q, (void *)start,
+					  hdr_len);
+	if (data_len) {
+		ret = cxgb4_map_skb(adap->pdev_dev, skb, d->addr);
+		if (unlikely(ret)) {
+			memset(d->addr, 0, sizeof(d->addr));
+			eohw_txq->mapping_err++;
+			goto out_unlock;
+		}
+
+		end = (u64 *)wr + flits;
+		if (unlikely(start > sgl)) {
+			left = (u8 *)end - (u8 *)eohw_txq->q.stat;
+			end = (void *)eohw_txq->q.desc + left;
+		}
+
+		if (unlikely((u8 *)sgl >= (u8 *)eohw_txq->q.stat)) {
+			/* If current position is already at the end of the
+			 * txq, reset the current to point to start of the queue
+			 * and update the end ptr as well.
+			 */
+			left = (u8 *)end - (u8 *)eohw_txq->q.stat;
+
+			end = (void *)eohw_txq->q.desc + left;
+			sgl = (void *)eohw_txq->q.desc;
+		}
+
+		cxgb4_write_sgl(skb, &eohw_txq->q, (void *)sgl, end, hdr_len,
+				d->addr);
+	}
+
+	if (skb_shinfo(skb)->gso_size) {
+		if (skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4)
+			eohw_txq->uso++;
+		else
+			eohw_txq->tso++;
+		eohw_txq->tx_cso += skb_shinfo(skb)->gso_segs;
+	} else if (skb->ip_summed == CHECKSUM_PARTIAL) {
+		eohw_txq->tx_cso++;
+	}
+
+	if (skb_vlan_tag_present(skb))
+		eohw_txq->vlan_ins++;
+
+	txq_advance(&eohw_txq->q, ndesc);
+	cxgb4_ring_tx_db(adap, &eohw_txq->q, ndesc);
+	eosw_txq_advance_index(&eosw_txq->last_pidx, 1, eosw_txq->ndesc);
+
+out_unlock:
+	spin_unlock(&eohw_txq->lock);
+	return ret;
+}
+
+static void ethofld_xmit(struct net_device *dev, struct sge_eosw_txq *eosw_txq)
+{
+	struct sk_buff *skb;
+	int pktcount, ret;
+
+	switch (eosw_txq->state) {
+	case CXGB4_EO_STATE_ACTIVE:
+	case CXGB4_EO_STATE_FLOWC_OPEN_SEND:
+	case CXGB4_EO_STATE_FLOWC_CLOSE_SEND:
+		pktcount = eosw_txq->pidx - eosw_txq->last_pidx;
+		if (pktcount < 0)
+			pktcount += eosw_txq->ndesc;
+		break;
+	case CXGB4_EO_STATE_FLOWC_OPEN_REPLY:
+	case CXGB4_EO_STATE_FLOWC_CLOSE_REPLY:
+	case CXGB4_EO_STATE_CLOSED:
+	default:
+		return;
+	}
+
+	while (pktcount--) {
+		skb = eosw_txq_peek(eosw_txq);
+		if (!skb) {
+			eosw_txq_advance_index(&eosw_txq->last_pidx, 1,
+					       eosw_txq->ndesc);
+			continue;
+		}
+
+		ret = ethofld_hard_xmit(dev, eosw_txq);
+		if (ret)
+			break;
+	}
+}
+
+static netdev_tx_t cxgb4_ethofld_xmit(struct sk_buff *skb,
+				      struct net_device *dev)
+{
+	struct cxgb4_tc_port_mqprio *tc_port_mqprio;
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	struct sge_eosw_txq *eosw_txq;
+	u32 qid;
+	int ret;
+
+	ret = cxgb4_validate_skb(skb, dev, ETH_HLEN);
+	if (ret)
+		goto out_free;
+
+	tc_port_mqprio = &adap->tc_mqprio->port_mqprio[pi->port_id];
+	qid = skb_get_queue_mapping(skb) - pi->nqsets;
+	eosw_txq = &tc_port_mqprio->eosw_txq[qid];
+	spin_lock_bh(&eosw_txq->lock);
+	if (eosw_txq->state != CXGB4_EO_STATE_ACTIVE)
+		goto out_unlock;
+
+	ret = eosw_txq_enqueue(eosw_txq, skb);
+	if (ret)
+		goto out_unlock;
+
+	/* SKB is queued for processing until credits are available.
+	 * So, call the destructor now and we'll free the skb later
+	 * after it has been successfully transmitted.
+	 */
+	skb_orphan(skb);
+
+	eosw_txq_advance(eosw_txq, 1);
+	ethofld_xmit(dev, eosw_txq);
+	spin_unlock_bh(&eosw_txq->lock);
+	return NETDEV_TX_OK;
+
+out_unlock:
+	spin_unlock_bh(&eosw_txq->lock);
+out_free:
+	dev_kfree_skb_any(skb);
+	return NETDEV_TX_OK;
+}
+
+netdev_tx_t t4_start_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	struct port_info *pi = netdev_priv(dev);
+	u16 qid = skb_get_queue_mapping(skb);
+
+	if (unlikely(pi->eth_flags & PRIV_FLAG_PORT_TX_VM))
+		return cxgb4_vf_eth_xmit(skb, dev);
+
+	if (unlikely(qid >= pi->nqsets))
+		return cxgb4_ethofld_xmit(skb, dev);
+
+	if (is_ptp_enabled(skb, dev)) {
+		struct adapter *adap = netdev2adap(dev);
+		netdev_tx_t ret;
+
+		spin_lock(&adap->ptp_lock);
+		ret = cxgb4_eth_xmit(skb, dev);
+		spin_unlock(&adap->ptp_lock);
+		return ret;
+	}
+
+	return cxgb4_eth_xmit(skb, dev);
+}
+
+static void eosw_txq_flush_pending_skbs(struct sge_eosw_txq *eosw_txq)
+{
+	int pktcount = eosw_txq->pidx - eosw_txq->last_pidx;
+	int pidx = eosw_txq->pidx;
+	struct sk_buff *skb;
+
+	if (!pktcount)
+		return;
+
+	if (pktcount < 0)
+		pktcount += eosw_txq->ndesc;
+
+	while (pktcount--) {
+		pidx--;
+		if (pidx < 0)
+			pidx += eosw_txq->ndesc;
+
+		skb = eosw_txq->desc[pidx].skb;
+		if (skb) {
+			dev_consume_skb_any(skb);
+			eosw_txq->desc[pidx].skb = NULL;
+			eosw_txq->inuse--;
+		}
+	}
+
+	eosw_txq->pidx = eosw_txq->last_pidx + 1;
+}
+
+/**
+ * cxgb4_ethofld_send_flowc - Send ETHOFLD flowc request to bind eotid to tc.
+ * @dev: netdevice
+ * @eotid: ETHOFLD tid to bind/unbind
+ * @tc: traffic class. If set to FW_SCHED_CLS_NONE, then unbinds the @eotid
+ *
+ * Send a FLOWC work request to bind an ETHOFLD TID to a traffic class.
+ * If @tc is set to FW_SCHED_CLS_NONE, then the @eotid is unbound from
+ * a traffic class.
+ */
+int cxgb4_ethofld_send_flowc(struct net_device *dev, u32 eotid, u32 tc)
+{
+	struct port_info *pi = netdev2pinfo(dev);
+	struct adapter *adap = netdev2adap(dev);
+	enum sge_eosw_state next_state;
+	struct sge_eosw_txq *eosw_txq;
+	u32 len, len16, nparams = 6;
+	struct fw_flowc_wr *flowc;
+	struct eotid_entry *entry;
+	struct sge_ofld_rxq *rxq;
+	struct sk_buff *skb;
+	int ret = 0;
+
+	len = struct_size(flowc, mnemval, nparams);
+	len16 = DIV_ROUND_UP(len, 16);
+
+	entry = cxgb4_lookup_eotid(&adap->tids, eotid);
+	if (!entry)
+		return -ENOMEM;
+
+	eosw_txq = (struct sge_eosw_txq *)entry->data;
+	if (!eosw_txq)
+		return -ENOMEM;
+
+	if (!(adap->flags & CXGB4_FW_OK)) {
+		/* Don't stall caller when access to FW is lost */
+		complete(&eosw_txq->completion);
+		return -EIO;
+	}
+
+	skb = alloc_skb(len, GFP_KERNEL);
+	if (!skb)
+		return -ENOMEM;
+
+	spin_lock_bh(&eosw_txq->lock);
+	if (tc != FW_SCHED_CLS_NONE) {
+		if (eosw_txq->state != CXGB4_EO_STATE_CLOSED)
+			goto out_free_skb;
+
+		next_state = CXGB4_EO_STATE_FLOWC_OPEN_SEND;
+	} else {
+		if (eosw_txq->state != CXGB4_EO_STATE_ACTIVE)
+			goto out_free_skb;
+
+		next_state = CXGB4_EO_STATE_FLOWC_CLOSE_SEND;
+	}
+
+	flowc = __skb_put(skb, len);
+	memset(flowc, 0, len);
+
+	rxq = &adap->sge.eohw_rxq[eosw_txq->hwqid];
+	flowc->flowid_len16 = cpu_to_be32(FW_WR_LEN16_V(len16) |
+					  FW_WR_FLOWID_V(eosw_txq->hwtid));
+	flowc->op_to_nparams = cpu_to_be32(FW_WR_OP_V(FW_FLOWC_WR) |
+					   FW_FLOWC_WR_NPARAMS_V(nparams) |
+					   FW_WR_COMPL_V(1));
+	flowc->mnemval[0].mnemonic = FW_FLOWC_MNEM_PFNVFN;
+	flowc->mnemval[0].val = cpu_to_be32(FW_PFVF_CMD_PFN_V(adap->pf));
+	flowc->mnemval[1].mnemonic = FW_FLOWC_MNEM_CH;
+	flowc->mnemval[1].val = cpu_to_be32(pi->tx_chan);
+	flowc->mnemval[2].mnemonic = FW_FLOWC_MNEM_PORT;
+	flowc->mnemval[2].val = cpu_to_be32(pi->tx_chan);
+	flowc->mnemval[3].mnemonic = FW_FLOWC_MNEM_IQID;
+	flowc->mnemval[3].val = cpu_to_be32(rxq->rspq.abs_id);
+	flowc->mnemval[4].mnemonic = FW_FLOWC_MNEM_SCHEDCLASS;
+	flowc->mnemval[4].val = cpu_to_be32(tc);
+	flowc->mnemval[5].mnemonic = FW_FLOWC_MNEM_EOSTATE;
+	flowc->mnemval[5].val = cpu_to_be32(tc == FW_SCHED_CLS_NONE ?
+					    FW_FLOWC_MNEM_EOSTATE_CLOSING :
+					    FW_FLOWC_MNEM_EOSTATE_ESTABLISHED);
+
+	/* Free up any pending skbs to ensure there's room for
+	 * termination FLOWC.
+	 */
+	if (tc == FW_SCHED_CLS_NONE)
+		eosw_txq_flush_pending_skbs(eosw_txq);
+
+	ret = eosw_txq_enqueue(eosw_txq, skb);
+	if (ret)
+		goto out_free_skb;
+
+	eosw_txq->state = next_state;
+	eosw_txq->flowc_idx = eosw_txq->pidx;
+	eosw_txq_advance(eosw_txq, 1);
+	ethofld_xmit(dev, eosw_txq);
+
+	spin_unlock_bh(&eosw_txq->lock);
+	return 0;
+
+out_free_skb:
+	dev_consume_skb_any(skb);
+	spin_unlock_bh(&eosw_txq->lock);
+	return ret;
+}
+
+/**
+ *	is_imm - check whether a packet can be sent as immediate data
+ *	@skb: the packet
+ *
+ *	Returns true if a packet can be sent as a WR with immediate data.
+ */
+static inline int is_imm(const struct sk_buff *skb)
+{
+	return skb->len <= MAX_CTRL_WR_LEN;
+}
+
+/**
+ *	ctrlq_check_stop - check if a control queue is full and should stop
+ *	@q: the queue
+ *	@wr: most recent WR written to the queue
+ *
+ *	Check if a control queue has become full and should be stopped.
+ *	We clean up control queue descriptors very lazily, only when we are out.
+ *	If the queue is still full after reclaiming any completed descriptors
+ *	we suspend it and have the last WR wake it up.
+ */
+static void ctrlq_check_stop(struct sge_ctrl_txq *q, struct fw_wr_hdr *wr)
+{
+	reclaim_completed_tx_imm(&q->q);
+	if (unlikely(txq_avail(&q->q) < TXQ_STOP_THRES)) {
+		wr->lo |= htonl(FW_WR_EQUEQ_F | FW_WR_EQUIQ_F);
+		q->q.stops++;
+		q->full = 1;
+	}
+}
+
+#define CXGB4_SELFTEST_LB_STR "CHELSIO_SELFTEST"
+
+int cxgb4_selftest_lb_pkt(struct net_device *netdev)
+{
+	struct port_info *pi = netdev_priv(netdev);
+	struct adapter *adap = pi->adapter;
+	struct cxgb4_ethtool_lb_test *lb;
+	int ret, i = 0, pkt_len, credits;
+	struct fw_eth_tx_pkt_wr *wr;
+	struct cpl_tx_pkt_core *cpl;
+	u32 ctrl0, ndesc, flits;
+	struct sge_eth_txq *q;
+	u8 *sgl;
+
+	pkt_len = ETH_HLEN + sizeof(CXGB4_SELFTEST_LB_STR);
+
+	flits = DIV_ROUND_UP(pkt_len + sizeof(*cpl) + sizeof(*wr),
+			     sizeof(__be64));
+	ndesc = flits_to_desc(flits);
+
+	lb = &pi->ethtool_lb;
+	lb->loopback = 1;
+
+	q = &adap->sge.ethtxq[pi->first_qset];
+	__netif_tx_lock(q->txq, smp_processor_id());
+
+	reclaim_completed_tx(adap, &q->q, -1, true);
+	credits = txq_avail(&q->q) - ndesc;
+	if (unlikely(credits < 0)) {
+		__netif_tx_unlock(q->txq);
+		return -ENOMEM;
+	}
+
+	wr = (void *)&q->q.desc[q->q.pidx];
+	memset(wr, 0, sizeof(struct tx_desc));
+
+	wr->op_immdlen = htonl(FW_WR_OP_V(FW_ETH_TX_PKT_WR) |
+			       FW_WR_IMMDLEN_V(pkt_len +
+			       sizeof(*cpl)));
+	wr->equiq_to_len16 = htonl(FW_WR_LEN16_V(DIV_ROUND_UP(flits, 2)));
+	wr->r3 = cpu_to_be64(0);
+
+	cpl = (void *)(wr + 1);
+	sgl = (u8 *)(cpl + 1);
+
+	ctrl0 = TXPKT_OPCODE_V(CPL_TX_PKT_XT) | TXPKT_PF_V(adap->pf) |
+		TXPKT_INTF_V(pi->tx_chan + 4);
+
+	cpl->ctrl0 = htonl(ctrl0);
+	cpl->pack = htons(0);
+	cpl->len = htons(pkt_len);
+	cpl->ctrl1 = cpu_to_be64(TXPKT_L4CSUM_DIS_F | TXPKT_IPCSUM_DIS_F);
+
+	eth_broadcast_addr(sgl);
+	i += ETH_ALEN;
+	ether_addr_copy(&sgl[i], netdev->dev_addr);
+	i += ETH_ALEN;
+
+	snprintf(&sgl[i], sizeof(CXGB4_SELFTEST_LB_STR), "%s",
+		 CXGB4_SELFTEST_LB_STR);
+
+	init_completion(&lb->completion);
+	txq_advance(&q->q, ndesc);
+	cxgb4_ring_tx_db(adap, &q->q, ndesc);
+	__netif_tx_unlock(q->txq);
+
+	/* wait for the pkt to return */
+	ret = wait_for_completion_timeout(&lb->completion, 10 * HZ);
+	if (!ret)
+		ret = -ETIMEDOUT;
+	else
+		ret = lb->result;
+
+	lb->loopback = 0;
+
+	return ret;
+}
+
+/**
+ *	ctrl_xmit - send a packet through an SGE control Tx queue
+ *	@q: the control queue
+ *	@skb: the packet
+ *
+ *	Send a packet through an SGE control Tx queue.  Packets sent through
+ *	a control queue must fit entirely as immediate data.
+ */
+static int ctrl_xmit(struct sge_ctrl_txq *q, struct sk_buff *skb)
+{
+	unsigned int ndesc;
+	struct fw_wr_hdr *wr;
+
+	if (unlikely(!is_imm(skb))) {
+		WARN_ON(1);
+		dev_kfree_skb(skb);
+		return NET_XMIT_DROP;
+	}
+
+	ndesc = DIV_ROUND_UP(skb->len, sizeof(struct tx_desc));
+	spin_lock(&q->sendq.lock);
+
+	if (unlikely(q->full)) {
+		skb->priority = ndesc;                  /* save for restart */
+		__skb_queue_tail(&q->sendq, skb);
+		spin_unlock(&q->sendq.lock);
+		return NET_XMIT_CN;
+	}
+
+	wr = (struct fw_wr_hdr *)&q->q.desc[q->q.pidx];
+	cxgb4_inline_tx_skb(skb, &q->q, wr);
+
+	txq_advance(&q->q, ndesc);
+	if (unlikely(txq_avail(&q->q) < TXQ_STOP_THRES))
+		ctrlq_check_stop(q, wr);
+
+	cxgb4_ring_tx_db(q->adap, &q->q, ndesc);
+	spin_unlock(&q->sendq.lock);
+
+	kfree_skb(skb);
+	return NET_XMIT_SUCCESS;
+}
+
+/**
+ *	restart_ctrlq - restart a suspended control queue
+ *	@t: pointer to the tasklet associated with this handler
+ *
+ *	Resumes transmission on a suspended Tx control queue.
+ */
+static void restart_ctrlq(struct tasklet_struct *t)
+{
+	struct sk_buff *skb;
+	unsigned int written = 0;
+	struct sge_ctrl_txq *q = from_tasklet(q, t, qresume_tsk);
+
+	spin_lock(&q->sendq.lock);
+	reclaim_completed_tx_imm(&q->q);
+	BUG_ON(txq_avail(&q->q) < TXQ_STOP_THRES);  /* q should be empty */
+
+	while ((skb = __skb_dequeue(&q->sendq)) != NULL) {
+		struct fw_wr_hdr *wr;
+		unsigned int ndesc = skb->priority;     /* previously saved */
+
+		written += ndesc;
+		/* Write descriptors and free skbs outside the lock to limit
+		 * wait times.  q->full is still set so new skbs will be queued.
+		 */
+		wr = (struct fw_wr_hdr *)&q->q.desc[q->q.pidx];
+		txq_advance(&q->q, ndesc);
+		spin_unlock(&q->sendq.lock);
+
+		cxgb4_inline_tx_skb(skb, &q->q, wr);
+		kfree_skb(skb);
+
+		if (unlikely(txq_avail(&q->q) < TXQ_STOP_THRES)) {
+			unsigned long old = q->q.stops;
+
+			ctrlq_check_stop(q, wr);
+			if (q->q.stops != old) {          /* suspended anew */
+				spin_lock(&q->sendq.lock);
+				goto ringdb;
+			}
+		}
+		if (written > 16) {
+			cxgb4_ring_tx_db(q->adap, &q->q, written);
+			written = 0;
+		}
+		spin_lock(&q->sendq.lock);
+	}
+	q->full = 0;
+ringdb:
+	if (written)
+		cxgb4_ring_tx_db(q->adap, &q->q, written);
+	spin_unlock(&q->sendq.lock);
+}
+
+/**
+ *	t4_mgmt_tx - send a management message
+ *	@adap: the adapter
+ *	@skb: the packet containing the management message
+ *
+ *	Send a management message through control queue 0.
+ */
+int t4_mgmt_tx(struct adapter *adap, struct sk_buff *skb)
+{
+	int ret;
+
+	local_bh_disable();
+	ret = ctrl_xmit(&adap->sge.ctrlq[0], skb);
+	local_bh_enable();
+	return ret;
+}
+
+/**
+ *	is_ofld_imm - check whether a packet can be sent as immediate data
+ *	@skb: the packet
+ *
+ *	Returns true if a packet can be sent as an offload WR with immediate
+ *	data.
+ *	FW_OFLD_TX_DATA_WR limits the payload to 255 bytes due to 8-bit field.
+ *      However, FW_ULPTX_WR commands have a 256 byte immediate only
+ *      payload limit.
+ */
+static inline int is_ofld_imm(const struct sk_buff *skb)
+{
+	struct work_request_hdr *req = (struct work_request_hdr *)skb->data;
+	unsigned long opcode = FW_WR_OP_G(ntohl(req->wr_hi));
+
+	if (unlikely(opcode == FW_ULPTX_WR))
+		return skb->len <= MAX_IMM_ULPTX_WR_LEN;
+	else if (opcode == FW_CRYPTO_LOOKASIDE_WR)
+		return skb->len <= SGE_MAX_WR_LEN;
+	else
+		return skb->len <= MAX_IMM_OFLD_TX_DATA_WR_LEN;
+}
+
+/**
+ *	calc_tx_flits_ofld - calculate # of flits for an offload packet
+ *	@skb: the packet
+ *
+ *	Returns the number of flits needed for the given offload packet.
+ *	These packets are already fully constructed and no additional headers
+ *	will be added.
+ */
+static inline unsigned int calc_tx_flits_ofld(const struct sk_buff *skb)
+{
+	unsigned int flits, cnt;
+
+	if (is_ofld_imm(skb))
+		return DIV_ROUND_UP(skb->len, 8);
+
+	flits = skb_transport_offset(skb) / 8U;   /* headers */
+	cnt = skb_shinfo(skb)->nr_frags;
+	if (skb_tail_pointer(skb) != skb_transport_header(skb))
+		cnt++;
+	return flits + sgl_len(cnt);
+}
+
+/**
+ *	txq_stop_maperr - stop a Tx queue due to I/O MMU exhaustion
+ *	@q: the queue to stop
+ *
+ *	Mark a Tx queue stopped due to I/O MMU exhaustion and resulting
+ *	inability to map packets.  A periodic timer attempts to restart
+ *	queues so marked.
+ */
+static void txq_stop_maperr(struct sge_uld_txq *q)
+{
+	q->mapping_err++;
+	q->q.stops++;
+	set_bit(q->q.cntxt_id - q->adap->sge.egr_start,
+		q->adap->sge.txq_maperr);
+}
+
+/**
+ *	ofldtxq_stop - stop an offload Tx queue that has become full
+ *	@q: the queue to stop
+ *	@wr: the Work Request causing the queue to become full
+ *
+ *	Stops an offload Tx queue that has become full and modifies the packet
+ *	being written to request a wakeup.
+ */
+static void ofldtxq_stop(struct sge_uld_txq *q, struct fw_wr_hdr *wr)
+{
+	wr->lo |= htonl(FW_WR_EQUEQ_F | FW_WR_EQUIQ_F);
+	q->q.stops++;
+	q->full = 1;
+}
+
+/**
+ *	service_ofldq - service/restart a suspended offload queue
+ *	@q: the offload queue
+ *
+ *	Services an offload Tx queue by moving packets from its Pending Send
+ *	Queue to the Hardware TX ring.  The function starts and ends with the
+ *	Send Queue locked, but drops the lock while putting the skb at the
+ *	head of the Send Queue onto the Hardware TX Ring.  Dropping the lock
+ *	allows more skbs to be added to the Send Queue by other threads.
+ *	The packet being processed at the head of the Pending Send Queue is
+ *	left on the queue in case we experience DMA Mapping errors, etc.
+ *	and need to give up and restart later.
+ *
+ *	service_ofldq() can be thought of as a task which opportunistically
+ *	uses other threads execution contexts.  We use the Offload Queue
+ *	boolean "service_ofldq_running" to make sure that only one instance
+ *	is ever running at a time ...
+ */
+static void service_ofldq(struct sge_uld_txq *q)
+	__must_hold(&q->sendq.lock)
+{
+	u64 *pos, *before, *end;
+	int credits;
+	struct sk_buff *skb;
+	struct sge_txq *txq;
+	unsigned int left;
+	unsigned int written = 0;
+	unsigned int flits, ndesc;
+
+	/* If another thread is currently in service_ofldq() processing the
+	 * Pending Send Queue then there's nothing to do. Otherwise, flag
+	 * that we're doing the work and continue.  Examining/modifying
+	 * the Offload Queue boolean "service_ofldq_running" must be done
+	 * while holding the Pending Send Queue Lock.
+	 */
+	if (q->service_ofldq_running)
+		return;
+	q->service_ofldq_running = true;
+
+	while ((skb = skb_peek(&q->sendq)) != NULL && !q->full) {
+		/* We drop the lock while we're working with the skb at the
+		 * head of the Pending Send Queue.  This allows more skbs to
+		 * be added to the Pending Send Queue while we're working on
+		 * this one.  We don't need to lock to guard the TX Ring
+		 * updates because only one thread of execution is ever
+		 * allowed into service_ofldq() at a time.
+		 */
+		spin_unlock(&q->sendq.lock);
+
+		cxgb4_reclaim_completed_tx(q->adap, &q->q, false);
+
+		flits = skb->priority;                /* previously saved */
+		ndesc = flits_to_desc(flits);
+		credits = txq_avail(&q->q) - ndesc;
+		BUG_ON(credits < 0);
+		if (unlikely(credits < TXQ_STOP_THRES))
+			ofldtxq_stop(q, (struct fw_wr_hdr *)skb->data);
+
+		pos = (u64 *)&q->q.desc[q->q.pidx];
+		if (is_ofld_imm(skb))
+			cxgb4_inline_tx_skb(skb, &q->q, pos);
+		else if (cxgb4_map_skb(q->adap->pdev_dev, skb,
+				       (dma_addr_t *)skb->head)) {
+			txq_stop_maperr(q);
+			spin_lock(&q->sendq.lock);
+			break;
+		} else {
+			int last_desc, hdr_len = skb_transport_offset(skb);
+
+			/* The WR headers  may not fit within one descriptor.
+			 * So we need to deal with wrap-around here.
+			 */
+			before = (u64 *)pos;
+			end = (u64 *)pos + flits;
+			txq = &q->q;
+			pos = (void *)inline_tx_skb_header(skb, &q->q,
+							   (void *)pos,
+							   hdr_len);
+			if (before > (u64 *)pos) {
+				left = (u8 *)end - (u8 *)txq->stat;
+				end = (void *)txq->desc + left;
+			}
+
+			/* If current position is already at the end of the
+			 * ofld queue, reset the current to point to
+			 * start of the queue and update the end ptr as well.
+			 */
+			if (pos == (u64 *)txq->stat) {
+				left = (u8 *)end - (u8 *)txq->stat;
+				end = (void *)txq->desc + left;
+				pos = (void *)txq->desc;
+			}
+
+			cxgb4_write_sgl(skb, &q->q, (void *)pos,
+					end, hdr_len,
+					(dma_addr_t *)skb->head);
+#ifdef CONFIG_NEED_DMA_MAP_STATE
+			skb->dev = q->adap->port[0];
+			skb->destructor = deferred_unmap_destructor;
+#endif
+			last_desc = q->q.pidx + ndesc - 1;
+			if (last_desc >= q->q.size)
+				last_desc -= q->q.size;
+			q->q.sdesc[last_desc].skb = skb;
+		}
+
+		txq_advance(&q->q, ndesc);
+		written += ndesc;
+		if (unlikely(written > 32)) {
+			cxgb4_ring_tx_db(q->adap, &q->q, written);
+			written = 0;
+		}
+
+		/* Reacquire the Pending Send Queue Lock so we can unlink the
+		 * skb we've just successfully transferred to the TX Ring and
+		 * loop for the next skb which may be at the head of the
+		 * Pending Send Queue.
+		 */
+		spin_lock(&q->sendq.lock);
+		__skb_unlink(skb, &q->sendq);
+		if (is_ofld_imm(skb))
+			kfree_skb(skb);
+	}
+	if (likely(written))
+		cxgb4_ring_tx_db(q->adap, &q->q, written);
+
+	/*Indicate that no thread is processing the Pending Send Queue
+	 * currently.
+	 */
+	q->service_ofldq_running = false;
+}
+
+/**
+ *	ofld_xmit - send a packet through an offload queue
+ *	@q: the Tx offload queue
+ *	@skb: the packet
+ *
+ *	Send an offload packet through an SGE offload queue.
+ */
+static int ofld_xmit(struct sge_uld_txq *q, struct sk_buff *skb)
+{
+	skb->priority = calc_tx_flits_ofld(skb);       /* save for restart */
+	spin_lock(&q->sendq.lock);
+
+	/* Queue the new skb onto the Offload Queue's Pending Send Queue.  If
+	 * that results in this new skb being the only one on the queue, start
+	 * servicing it.  If there are other skbs already on the list, then
+	 * either the queue is currently being processed or it's been stopped
+	 * for some reason and it'll be restarted at a later time.  Restart
+	 * paths are triggered by events like experiencing a DMA Mapping Error
+	 * or filling the Hardware TX Ring.
+	 */
+	__skb_queue_tail(&q->sendq, skb);
+	if (q->sendq.qlen == 1)
+		service_ofldq(q);
+
+	spin_unlock(&q->sendq.lock);
+	return NET_XMIT_SUCCESS;
+}
+
+/**
+ *	restart_ofldq - restart a suspended offload queue
+ *	@t: pointer to the tasklet associated with this handler
+ *
+ *	Resumes transmission on a suspended Tx offload queue.
+ */
+static void restart_ofldq(struct tasklet_struct *t)
+{
+	struct sge_uld_txq *q = from_tasklet(q, t, qresume_tsk);
+
+	spin_lock(&q->sendq.lock);
+	q->full = 0;            /* the queue actually is completely empty now */
+	service_ofldq(q);
+	spin_unlock(&q->sendq.lock);
+}
+
+/**
+ *	skb_txq - return the Tx queue an offload packet should use
+ *	@skb: the packet
+ *
+ *	Returns the Tx queue an offload packet should use as indicated by bits
+ *	1-15 in the packet's queue_mapping.
+ */
+static inline unsigned int skb_txq(const struct sk_buff *skb)
+{
+	return skb->queue_mapping >> 1;
+}
+
+/**
+ *	is_ctrl_pkt - return whether an offload packet is a control packet
+ *	@skb: the packet
+ *
+ *	Returns whether an offload packet should use an OFLD or a CTRL
+ *	Tx queue as indicated by bit 0 in the packet's queue_mapping.
+ */
+static inline unsigned int is_ctrl_pkt(const struct sk_buff *skb)
+{
+	return skb->queue_mapping & 1;
+}
+
+static inline int uld_send(struct adapter *adap, struct sk_buff *skb,
+			   unsigned int tx_uld_type)
+{
+	struct sge_uld_txq_info *txq_info;
+	struct sge_uld_txq *txq;
+	unsigned int idx = skb_txq(skb);
+
+	if (unlikely(is_ctrl_pkt(skb))) {
+		/* Single ctrl queue is a requirement for LE workaround path */
+		if (adap->tids.nsftids)
+			idx = 0;
+		return ctrl_xmit(&adap->sge.ctrlq[idx], skb);
+	}
+
+	txq_info = adap->sge.uld_txq_info[tx_uld_type];
+	if (unlikely(!txq_info)) {
+		WARN_ON(true);
+		kfree_skb(skb);
+		return NET_XMIT_DROP;
+	}
+
+	txq = &txq_info->uldtxq[idx];
+	return ofld_xmit(txq, skb);
+}
+
+/**
+ *	t4_ofld_send - send an offload packet
+ *	@adap: the adapter
+ *	@skb: the packet
+ *
+ *	Sends an offload packet.  We use the packet queue_mapping to select the
+ *	appropriate Tx queue as follows: bit 0 indicates whether the packet
+ *	should be sent as regular or control, bits 1-15 select the queue.
+ */
+int t4_ofld_send(struct adapter *adap, struct sk_buff *skb)
+{
+	int ret;
+
+	local_bh_disable();
+	ret = uld_send(adap, skb, CXGB4_TX_OFLD);
+	local_bh_enable();
+	return ret;
+}
+
+/**
+ *	cxgb4_ofld_send - send an offload packet
+ *	@dev: the net device
+ *	@skb: the packet
+ *
+ *	Sends an offload packet.  This is an exported version of @t4_ofld_send,
+ *	intended for ULDs.
+ */
+int cxgb4_ofld_send(struct net_device *dev, struct sk_buff *skb)
+{
+	return t4_ofld_send(netdev2adap(dev), skb);
+}
+EXPORT_SYMBOL(cxgb4_ofld_send);
+
+static void *inline_tx_header(const void *src,
+			      const struct sge_txq *q,
+			      void *pos, int length)
+{
+	int left = (void *)q->stat - pos;
+	u64 *p;
+
+	if (likely(length <= left)) {
+		memcpy(pos, src, length);
+		pos += length;
+	} else {
+		memcpy(pos, src, left);
+		memcpy(q->desc, src + left, length - left);
+		pos = (void *)q->desc + (length - left);
+	}
+	/* 0-pad to multiple of 16 */
+	p = PTR_ALIGN(pos, 8);
+	if ((uintptr_t)p & 8) {
+		*p = 0;
+		return p + 1;
+	}
+	return p;
+}
+
+/**
+ *      ofld_xmit_direct - copy a WR into offload queue
+ *      @q: the Tx offload queue
+ *      @src: location of WR
+ *      @len: WR length
+ *
+ *      Copy an immediate WR into an uncontended SGE offload queue.
+ */
+static int ofld_xmit_direct(struct sge_uld_txq *q, const void *src,
+			    unsigned int len)
+{
+	unsigned int ndesc;
+	int credits;
+	u64 *pos;
+
+	/* Use the lower limit as the cut-off */
+	if (len > MAX_IMM_OFLD_TX_DATA_WR_LEN) {
+		WARN_ON(1);
+		return NET_XMIT_DROP;
+	}
+
+	/* Don't return NET_XMIT_CN here as the current
+	 * implementation doesn't queue the request
+	 * using an skb when the following conditions not met
+	 */
+	if (!spin_trylock(&q->sendq.lock))
+		return NET_XMIT_DROP;
+
+	if (q->full || !skb_queue_empty(&q->sendq) ||
+	    q->service_ofldq_running) {
+		spin_unlock(&q->sendq.lock);
+		return NET_XMIT_DROP;
+	}
+	ndesc = flits_to_desc(DIV_ROUND_UP(len, 8));
+	credits = txq_avail(&q->q) - ndesc;
+	pos = (u64 *)&q->q.desc[q->q.pidx];
+
+	/* ofldtxq_stop modifies WR header in-situ */
+	inline_tx_header(src, &q->q, pos, len);
+	if (unlikely(credits < TXQ_STOP_THRES))
+		ofldtxq_stop(q, (struct fw_wr_hdr *)pos);
+	txq_advance(&q->q, ndesc);
+	cxgb4_ring_tx_db(q->adap, &q->q, ndesc);
+
+	spin_unlock(&q->sendq.lock);
+	return NET_XMIT_SUCCESS;
+}
+
+int cxgb4_immdata_send(struct net_device *dev, unsigned int idx,
+		       const void *src, unsigned int len)
+{
+	struct sge_uld_txq_info *txq_info;
+	struct sge_uld_txq *txq;
+	struct adapter *adap;
+	int ret;
+
+	adap = netdev2adap(dev);
+
+	local_bh_disable();
+	txq_info = adap->sge.uld_txq_info[CXGB4_TX_OFLD];
+	if (unlikely(!txq_info)) {
+		WARN_ON(true);
+		local_bh_enable();
+		return NET_XMIT_DROP;
+	}
+	txq = &txq_info->uldtxq[idx];
+
+	ret = ofld_xmit_direct(txq, src, len);
+	local_bh_enable();
+	return net_xmit_eval(ret);
+}
+EXPORT_SYMBOL(cxgb4_immdata_send);
+
+/**
+ *	t4_crypto_send - send crypto packet
+ *	@adap: the adapter
+ *	@skb: the packet
+ *
+ *	Sends crypto packet.  We use the packet queue_mapping to select the
+ *	appropriate Tx queue as follows: bit 0 indicates whether the packet
+ *	should be sent as regular or control, bits 1-15 select the queue.
+ */
+static int t4_crypto_send(struct adapter *adap, struct sk_buff *skb)
+{
+	int ret;
+
+	local_bh_disable();
+	ret = uld_send(adap, skb, CXGB4_TX_CRYPTO);
+	local_bh_enable();
+	return ret;
+}
+
+/**
+ *	cxgb4_crypto_send - send crypto packet
+ *	@dev: the net device
+ *	@skb: the packet
+ *
+ *	Sends crypto packet.  This is an exported version of @t4_crypto_send,
+ *	intended for ULDs.
+ */
+int cxgb4_crypto_send(struct net_device *dev, struct sk_buff *skb)
+{
+	return t4_crypto_send(netdev2adap(dev), skb);
+}
+EXPORT_SYMBOL(cxgb4_crypto_send);
+
+static inline void copy_frags(struct sk_buff *skb,
+			      const struct pkt_gl *gl, unsigned int offset)
+{
+	int i;
+
+	/* usually there's just one frag */
+	__skb_fill_page_desc(skb, 0, gl->frags[0].page,
+			     gl->frags[0].offset + offset,
+			     gl->frags[0].size - offset);
+	skb_shinfo(skb)->nr_frags = gl->nfrags;
+	for (i = 1; i < gl->nfrags; i++)
+		__skb_fill_page_desc(skb, i, gl->frags[i].page,
+				     gl->frags[i].offset,
+				     gl->frags[i].size);
+
+	/* get a reference to the last page, we don't own it */
+	get_page(gl->frags[gl->nfrags - 1].page);
+}
+
+/**
+ *	cxgb4_pktgl_to_skb - build an sk_buff from a packet gather list
+ *	@gl: the gather list
+ *	@skb_len: size of sk_buff main body if it carries fragments
+ *	@pull_len: amount of data to move to the sk_buff's main body
+ *
+ *	Builds an sk_buff from the given packet gather list.  Returns the
+ *	sk_buff or %NULL if sk_buff allocation failed.
+ */
+struct sk_buff *cxgb4_pktgl_to_skb(const struct pkt_gl *gl,
+				   unsigned int skb_len, unsigned int pull_len)
+{
+	struct sk_buff *skb;
+
+	/*
+	 * Below we rely on RX_COPY_THRES being less than the smallest Rx buffer
+	 * size, which is expected since buffers are at least PAGE_SIZEd.
+	 * In this case packets up to RX_COPY_THRES have only one fragment.
+	 */
+	if (gl->tot_len <= RX_COPY_THRES) {
+		skb = dev_alloc_skb(gl->tot_len);
+		if (unlikely(!skb))
+			goto out;
+		__skb_put(skb, gl->tot_len);
+		skb_copy_to_linear_data(skb, gl->va, gl->tot_len);
+	} else {
+		skb = dev_alloc_skb(skb_len);
+		if (unlikely(!skb))
+			goto out;
+		__skb_put(skb, pull_len);
+		skb_copy_to_linear_data(skb, gl->va, pull_len);
+
+		copy_frags(skb, gl, pull_len);
+		skb->len = gl->tot_len;
+		skb->data_len = skb->len - pull_len;
+		skb->truesize += skb->data_len;
+	}
+out:	return skb;
+}
+EXPORT_SYMBOL(cxgb4_pktgl_to_skb);
+
+/**
+ *	t4_pktgl_free - free a packet gather list
+ *	@gl: the gather list
+ *
+ *	Releases the pages of a packet gather list.  We do not own the last
+ *	page on the list and do not free it.
+ */
+static void t4_pktgl_free(const struct pkt_gl *gl)
+{
+	int n;
+	const struct page_frag *p;
+
+	for (p = gl->frags, n = gl->nfrags - 1; n--; p++)
+		put_page(p->page);
+}
+
+/*
+ * Process an MPS trace packet.  Give it an unused protocol number so it won't
+ * be delivered to anyone and send it to the stack for capture.
+ */
+static noinline int handle_trace_pkt(struct adapter *adap,
+				     const struct pkt_gl *gl)
+{
+	struct sk_buff *skb;
+
+	skb = cxgb4_pktgl_to_skb(gl, RX_PULL_LEN, RX_PULL_LEN);
+	if (unlikely(!skb)) {
+		t4_pktgl_free(gl);
+		return 0;
+	}
+
+	if (is_t4(adap->params.chip))
+		__skb_pull(skb, sizeof(struct cpl_trace_pkt));
+	else
+		__skb_pull(skb, sizeof(struct cpl_t5_trace_pkt));
+
+	skb_reset_mac_header(skb);
+	skb->protocol = htons(0xffff);
+	skb->dev = adap->port[0];
+	netif_receive_skb(skb);
+	return 0;
+}
+
+/**
+ * cxgb4_sgetim_to_hwtstamp - convert sge time stamp to hw time stamp
+ * @adap: the adapter
+ * @hwtstamps: time stamp structure to update
+ * @sgetstamp: 60bit iqe timestamp
+ *
+ * Every ingress queue entry has the 60-bit timestamp, convert that timestamp
+ * which is in Core Clock ticks into ktime_t and assign it
+ **/
+static void cxgb4_sgetim_to_hwtstamp(struct adapter *adap,
+				     struct skb_shared_hwtstamps *hwtstamps,
+				     u64 sgetstamp)
+{
+	u64 ns;
+	u64 tmp = (sgetstamp * 1000 * 1000 + adap->params.vpd.cclk / 2);
+
+	ns = div_u64(tmp, adap->params.vpd.cclk);
+
+	memset(hwtstamps, 0, sizeof(*hwtstamps));
+	hwtstamps->hwtstamp = ns_to_ktime(ns);
+}
+
+static void do_gro(struct sge_eth_rxq *rxq, const struct pkt_gl *gl,
+		   const struct cpl_rx_pkt *pkt, unsigned long tnl_hdr_len)
+{
+	struct adapter *adapter = rxq->rspq.adap;
+	struct sge *s = &adapter->sge;
+	struct port_info *pi;
+	int ret;
+	struct sk_buff *skb;
+
+	skb = napi_get_frags(&rxq->rspq.napi);
+	if (unlikely(!skb)) {
+		t4_pktgl_free(gl);
+		rxq->stats.rx_drops++;
+		return;
+	}
+
+	copy_frags(skb, gl, s->pktshift);
+	if (tnl_hdr_len)
+		skb->csum_level = 1;
+	skb->len = gl->tot_len - s->pktshift;
+	skb->data_len = skb->len;
+	skb->truesize += skb->data_len;
+	skb->ip_summed = CHECKSUM_UNNECESSARY;
+	skb_record_rx_queue(skb, rxq->rspq.idx);
+	pi = netdev_priv(skb->dev);
+	if (pi->rxtstamp)
+		cxgb4_sgetim_to_hwtstamp(adapter, skb_hwtstamps(skb),
+					 gl->sgetstamp);
+	if (rxq->rspq.netdev->features & NETIF_F_RXHASH)
+		skb_set_hash(skb, (__force u32)pkt->rsshdr.hash_val,
+			     PKT_HASH_TYPE_L3);
+
+	if (unlikely(pkt->vlan_ex)) {
+		__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), ntohs(pkt->vlan));
+		rxq->stats.vlan_ex++;
+	}
+	ret = napi_gro_frags(&rxq->rspq.napi);
+	if (ret == GRO_HELD)
+		rxq->stats.lro_pkts++;
+	else if (ret == GRO_MERGED || ret == GRO_MERGED_FREE)
+		rxq->stats.lro_merged++;
+	rxq->stats.pkts++;
+	rxq->stats.rx_cso++;
+}
+
+enum {
+	RX_NON_PTP_PKT = 0,
+	RX_PTP_PKT_SUC = 1,
+	RX_PTP_PKT_ERR = 2
+};
+
+/**
+ *     t4_systim_to_hwstamp - read hardware time stamp
+ *     @adapter: the adapter
+ *     @skb: the packet
+ *
+ *     Read Time Stamp from MPS packet and insert in skb which
+ *     is forwarded to PTP application
+ */
+static noinline int t4_systim_to_hwstamp(struct adapter *adapter,
+					 struct sk_buff *skb)
+{
+	struct skb_shared_hwtstamps *hwtstamps;
+	struct cpl_rx_mps_pkt *cpl = NULL;
+	unsigned char *data;
+	int offset;
+
+	cpl = (struct cpl_rx_mps_pkt *)skb->data;
+	if (!(CPL_RX_MPS_PKT_TYPE_G(ntohl(cpl->op_to_r1_hi)) &
+	     X_CPL_RX_MPS_PKT_TYPE_PTP))
+		return RX_PTP_PKT_ERR;
+
+	data = skb->data + sizeof(*cpl);
+	skb_pull(skb, 2 * sizeof(u64) + sizeof(struct cpl_rx_mps_pkt));
+	offset = ETH_HLEN + IPV4_HLEN(skb->data) + UDP_HLEN;
+	if (skb->len < offset + OFF_PTP_SEQUENCE_ID + sizeof(short))
+		return RX_PTP_PKT_ERR;
+
+	hwtstamps = skb_hwtstamps(skb);
+	memset(hwtstamps, 0, sizeof(*hwtstamps));
+	hwtstamps->hwtstamp = ns_to_ktime(get_unaligned_be64(data));
+
+	return RX_PTP_PKT_SUC;
+}
+
+/**
+ *     t4_rx_hststamp - Recv PTP Event Message
+ *     @adapter: the adapter
+ *     @rsp: the response queue descriptor holding the RX_PKT message
+ *     @rxq: the response queue holding the RX_PKT message
+ *     @skb: the packet
+ *
+ *     PTP enabled and MPS packet, read HW timestamp
+ */
+static int t4_rx_hststamp(struct adapter *adapter, const __be64 *rsp,
+			  struct sge_eth_rxq *rxq, struct sk_buff *skb)
+{
+	int ret;
+
+	if (unlikely((*(u8 *)rsp == CPL_RX_MPS_PKT) &&
+		     !is_t4(adapter->params.chip))) {
+		ret = t4_systim_to_hwstamp(adapter, skb);
+		if (ret == RX_PTP_PKT_ERR) {
+			kfree_skb(skb);
+			rxq->stats.rx_drops++;
+		}
+		return ret;
+	}
+	return RX_NON_PTP_PKT;
+}
+
+/**
+ *      t4_tx_hststamp - Loopback PTP Transmit Event Message
+ *      @adapter: the adapter
+ *      @skb: the packet
+ *      @dev: the ingress net device
+ *
+ *      Read hardware timestamp for the loopback PTP Tx event message
+ */
+static int t4_tx_hststamp(struct adapter *adapter, struct sk_buff *skb,
+			  struct net_device *dev)
+{
+	struct port_info *pi = netdev_priv(dev);
+
+	if (!is_t4(adapter->params.chip) && adapter->ptp_tx_skb) {
+		cxgb4_ptp_read_hwstamp(adapter, pi);
+		kfree_skb(skb);
+		return 0;
+	}
+	return 1;
+}
+
+/**
+ *	t4_tx_completion_handler - handle CPL_SGE_EGR_UPDATE messages
+ *	@rspq: Ethernet RX Response Queue associated with Ethernet TX Queue
+ *	@rsp: Response Entry pointer into Response Queue
+ *	@gl: Gather List pointer
+ *
+ *	For adapters which support the SGE Doorbell Queue Timer facility,
+ *	we configure the Ethernet TX Queues to send CIDX Updates to the
+ *	Associated Ethernet RX Response Queue with CPL_SGE_EGR_UPDATE
+ *	messages.  This adds a small load to PCIe Link RX bandwidth and,
+ *	potentially, higher CPU Interrupt load, but allows us to respond
+ *	much more quickly to the CIDX Updates.  This is important for
+ *	Upper Layer Software which isn't willing to have a large amount
+ *	of TX Data outstanding before receiving DMA Completions.
+ */
+static void t4_tx_completion_handler(struct sge_rspq *rspq,
+				     const __be64 *rsp,
+				     const struct pkt_gl *gl)
+{
+	u8 opcode = ((const struct rss_header *)rsp)->opcode;
+	struct port_info *pi = netdev_priv(rspq->netdev);
+	struct adapter *adapter = rspq->adap;
+	struct sge *s = &adapter->sge;
+	struct sge_eth_txq *txq;
+
+	/* skip RSS header */
+	rsp++;
+
+	/* FW can send EGR_UPDATEs encapsulated in a CPL_FW4_MSG.
+	 */
+	if (unlikely(opcode == CPL_FW4_MSG &&
+		     ((const struct cpl_fw4_msg *)rsp)->type ==
+							FW_TYPE_RSSCPL)) {
+		rsp++;
+		opcode = ((const struct rss_header *)rsp)->opcode;
+		rsp++;
+	}
+
+	if (unlikely(opcode != CPL_SGE_EGR_UPDATE)) {
+		pr_info("%s: unexpected FW4/CPL %#x on Rx queue\n",
+			__func__, opcode);
+		return;
+	}
+
+	txq = &s->ethtxq[pi->first_qset + rspq->idx];
+
+	/* We've got the Hardware Consumer Index Update in the Egress Update
+	 * message. These Egress Update messages will be our sole CIDX Updates
+	 * we get since we don't want to chew up PCIe bandwidth for both Ingress
+	 * Messages and Status Page writes.  However, The code which manages
+	 * reclaiming successfully DMA'ed TX Work Requests uses the CIDX value
+	 * stored in the Status Page at the end of the TX Queue.  It's easiest
+	 * to simply copy the CIDX Update value from the Egress Update message
+	 * to the Status Page.  Also note that no Endian issues need to be
+	 * considered here since both are Big Endian and we're just copying
+	 * bytes consistently ...
+	 */
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) <= CHELSIO_T5) {
+		struct cpl_sge_egr_update *egr;
+
+		egr = (struct cpl_sge_egr_update *)rsp;
+		WRITE_ONCE(txq->q.stat->cidx, egr->cidx);
+	}
+
+	t4_sge_eth_txq_egress_update(adapter, txq, -1);
+}
+
+static int cxgb4_validate_lb_pkt(struct port_info *pi, const struct pkt_gl *si)
+{
+	struct adapter *adap = pi->adapter;
+	struct cxgb4_ethtool_lb_test *lb;
+	struct sge *s = &adap->sge;
+	struct net_device *netdev;
+	u8 *data;
+	int i;
+
+	netdev = adap->port[pi->port_id];
+	lb = &pi->ethtool_lb;
+	data = si->va + s->pktshift;
+
+	i = ETH_ALEN;
+	if (!ether_addr_equal(data + i, netdev->dev_addr))
+		return -1;
+
+	i += ETH_ALEN;
+	if (strcmp(&data[i], CXGB4_SELFTEST_LB_STR))
+		lb->result = -EIO;
+
+	complete(&lb->completion);
+	return 0;
+}
+
+/**
+ *	t4_ethrx_handler - process an ingress ethernet packet
+ *	@q: the response queue that received the packet
+ *	@rsp: the response queue descriptor holding the RX_PKT message
+ *	@si: the gather list of packet fragments
+ *
+ *	Process an ingress ethernet packet and deliver it to the stack.
+ */
+int t4_ethrx_handler(struct sge_rspq *q, const __be64 *rsp,
+		     const struct pkt_gl *si)
+{
+	bool csum_ok;
+	struct sk_buff *skb;
+	const struct cpl_rx_pkt *pkt;
+	struct sge_eth_rxq *rxq = container_of(q, struct sge_eth_rxq, rspq);
+	struct adapter *adapter = q->adap;
+	struct sge *s = &q->adap->sge;
+	int cpl_trace_pkt = is_t4(q->adap->params.chip) ?
+			    CPL_TRACE_PKT : CPL_TRACE_PKT_T5;
+	u16 err_vec, tnl_hdr_len = 0;
+	struct port_info *pi;
+	int ret = 0;
+
+	pi = netdev_priv(q->netdev);
+	/* If we're looking at TX Queue CIDX Update, handle that separately
+	 * and return.
+	 */
+	if (unlikely((*(u8 *)rsp == CPL_FW4_MSG) ||
+		     (*(u8 *)rsp == CPL_SGE_EGR_UPDATE))) {
+		t4_tx_completion_handler(q, rsp, si);
+		return 0;
+	}
+
+	if (unlikely(*(u8 *)rsp == cpl_trace_pkt))
+		return handle_trace_pkt(q->adap, si);
+
+	pkt = (const struct cpl_rx_pkt *)rsp;
+	/* Compressed error vector is enabled for T6 only */
+	if (q->adap->params.tp.rx_pkt_encap) {
+		err_vec = T6_COMPR_RXERR_VEC_G(be16_to_cpu(pkt->err_vec));
+		tnl_hdr_len = T6_RX_TNLHDR_LEN_G(ntohs(pkt->err_vec));
+	} else {
+		err_vec = be16_to_cpu(pkt->err_vec);
+	}
+
+	csum_ok = pkt->csum_calc && !err_vec &&
+		  (q->netdev->features & NETIF_F_RXCSUM);
+
+	if (err_vec)
+		rxq->stats.bad_rx_pkts++;
+
+	if (unlikely(pi->ethtool_lb.loopback && pkt->iff >= NCHAN)) {
+		ret = cxgb4_validate_lb_pkt(pi, si);
+		if (!ret)
+			return 0;
+	}
+
+	if (((pkt->l2info & htonl(RXF_TCP_F)) ||
+	     tnl_hdr_len) &&
+	    (q->netdev->features & NETIF_F_GRO) && csum_ok && !pkt->ip_frag) {
+		do_gro(rxq, si, pkt, tnl_hdr_len);
+		return 0;
+	}
+
+	skb = cxgb4_pktgl_to_skb(si, RX_PKT_SKB_LEN, RX_PULL_LEN);
+	if (unlikely(!skb)) {
+		t4_pktgl_free(si);
+		rxq->stats.rx_drops++;
+		return 0;
+	}
+
+	/* Handle PTP Event Rx packet */
+	if (unlikely(pi->ptp_enable)) {
+		ret = t4_rx_hststamp(adapter, rsp, rxq, skb);
+		if (ret == RX_PTP_PKT_ERR)
+			return 0;
+	}
+	if (likely(!ret))
+		__skb_pull(skb, s->pktshift); /* remove ethernet header pad */
+
+	/* Handle the PTP Event Tx Loopback packet */
+	if (unlikely(pi->ptp_enable && !ret &&
+		     (pkt->l2info & htonl(RXF_UDP_F)) &&
+		     cxgb4_ptp_is_ptp_rx(skb))) {
+		if (!t4_tx_hststamp(adapter, skb, q->netdev))
+			return 0;
+	}
+
+	skb->protocol = eth_type_trans(skb, q->netdev);
+	skb_record_rx_queue(skb, q->idx);
+	if (skb->dev->features & NETIF_F_RXHASH)
+		skb_set_hash(skb, (__force u32)pkt->rsshdr.hash_val,
+			     PKT_HASH_TYPE_L3);
+
+	rxq->stats.pkts++;
+
+	if (pi->rxtstamp)
+		cxgb4_sgetim_to_hwtstamp(q->adap, skb_hwtstamps(skb),
+					 si->sgetstamp);
+	if (csum_ok && (pkt->l2info & htonl(RXF_UDP_F | RXF_TCP_F))) {
+		if (!pkt->ip_frag) {
+			skb->ip_summed = CHECKSUM_UNNECESSARY;
+			rxq->stats.rx_cso++;
+		} else if (pkt->l2info & htonl(RXF_IP_F)) {
+			__sum16 c = (__force __sum16)pkt->csum;
+			skb->csum = csum_unfold(c);
+
+			if (tnl_hdr_len) {
+				skb->ip_summed = CHECKSUM_UNNECESSARY;
+				skb->csum_level = 1;
+			} else {
+				skb->ip_summed = CHECKSUM_COMPLETE;
+			}
+			rxq->stats.rx_cso++;
+		}
+	} else {
+		skb_checksum_none_assert(skb);
+#ifdef CONFIG_CHELSIO_T4_FCOE
+#define CPL_RX_PKT_FLAGS (RXF_PSH_F | RXF_SYN_F | RXF_UDP_F | \
+			  RXF_TCP_F | RXF_IP_F | RXF_IP6_F | RXF_LRO_F)
+
+		if (!(pkt->l2info & cpu_to_be32(CPL_RX_PKT_FLAGS))) {
+			if ((pkt->l2info & cpu_to_be32(RXF_FCOE_F)) &&
+			    (pi->fcoe.flags & CXGB_FCOE_ENABLED)) {
+				if (q->adap->params.tp.rx_pkt_encap)
+					csum_ok = err_vec &
+						  T6_COMPR_RXERR_SUM_F;
+				else
+					csum_ok = err_vec & RXERR_CSUM_F;
+				if (!csum_ok)
+					skb->ip_summed = CHECKSUM_UNNECESSARY;
+			}
+		}
+
+#undef CPL_RX_PKT_FLAGS
+#endif /* CONFIG_CHELSIO_T4_FCOE */
+	}
+
+	if (unlikely(pkt->vlan_ex)) {
+		__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), ntohs(pkt->vlan));
+		rxq->stats.vlan_ex++;
+	}
+	skb_mark_napi_id(skb, &q->napi);
+	netif_receive_skb(skb);
+	return 0;
+}
+
+/**
+ *	restore_rx_bufs - put back a packet's Rx buffers
+ *	@si: the packet gather list
+ *	@q: the SGE free list
+ *	@frags: number of FL buffers to restore
+ *
+ *	Puts back on an FL the Rx buffers associated with @si.  The buffers
+ *	have already been unmapped and are left unmapped, we mark them so to
+ *	prevent further unmapping attempts.
+ *
+ *	This function undoes a series of @unmap_rx_buf calls when we find out
+ *	that the current packet can't be processed right away afterall and we
+ *	need to come back to it later.  This is a very rare event and there's
+ *	no effort to make this particularly efficient.
+ */
+static void restore_rx_bufs(const struct pkt_gl *si, struct sge_fl *q,
+			    int frags)
+{
+	struct rx_sw_desc *d;
+
+	while (frags--) {
+		if (q->cidx == 0)
+			q->cidx = q->size - 1;
+		else
+			q->cidx--;
+		d = &q->sdesc[q->cidx];
+		d->page = si->frags[frags].page;
+		d->dma_addr |= RX_UNMAPPED_BUF;
+		q->avail++;
+	}
+}
+
+/**
+ *	is_new_response - check if a response is newly written
+ *	@r: the response descriptor
+ *	@q: the response queue
+ *
+ *	Returns true if a response descriptor contains a yet unprocessed
+ *	response.
+ */
+static inline bool is_new_response(const struct rsp_ctrl *r,
+				   const struct sge_rspq *q)
+{
+	return (r->type_gen >> RSPD_GEN_S) == q->gen;
+}
+
+/**
+ *	rspq_next - advance to the next entry in a response queue
+ *	@q: the queue
+ *
+ *	Updates the state of a response queue to advance it to the next entry.
+ */
+static inline void rspq_next(struct sge_rspq *q)
+{
+	q->cur_desc = (void *)q->cur_desc + q->iqe_len;
+	if (unlikely(++q->cidx == q->size)) {
+		q->cidx = 0;
+		q->gen ^= 1;
+		q->cur_desc = q->desc;
+	}
+}
+
+/**
+ *	process_responses - process responses from an SGE response queue
+ *	@q: the ingress queue to process
+ *	@budget: how many responses can be processed in this round
+ *
+ *	Process responses from an SGE response queue up to the supplied budget.
+ *	Responses include received packets as well as control messages from FW
+ *	or HW.
+ *
+ *	Additionally choose the interrupt holdoff time for the next interrupt
+ *	on this queue.  If the system is under memory shortage use a fairly
+ *	long delay to help recovery.
+ */
+static int process_responses(struct sge_rspq *q, int budget)
+{
+	int ret, rsp_type;
+	int budget_left = budget;
+	const struct rsp_ctrl *rc;
+	struct sge_eth_rxq *rxq = container_of(q, struct sge_eth_rxq, rspq);
+	struct adapter *adapter = q->adap;
+	struct sge *s = &adapter->sge;
+
+	while (likely(budget_left)) {
+		rc = (void *)q->cur_desc + (q->iqe_len - sizeof(*rc));
+		if (!is_new_response(rc, q)) {
+			if (q->flush_handler)
+				q->flush_handler(q);
+			break;
+		}
+
+		dma_rmb();
+		rsp_type = RSPD_TYPE_G(rc->type_gen);
+		if (likely(rsp_type == RSPD_TYPE_FLBUF_X)) {
+			struct page_frag *fp;
+			struct pkt_gl si;
+			const struct rx_sw_desc *rsd;
+			u32 len = ntohl(rc->pldbuflen_qid), bufsz, frags;
+
+			if (len & RSPD_NEWBUF_F) {
+				if (likely(q->offset > 0)) {
+					free_rx_bufs(q->adap, &rxq->fl, 1);
+					q->offset = 0;
+				}
+				len = RSPD_LEN_G(len);
+			}
+			si.tot_len = len;
+
+			/* gather packet fragments */
+			for (frags = 0, fp = si.frags; ; frags++, fp++) {
+				rsd = &rxq->fl.sdesc[rxq->fl.cidx];
+				bufsz = get_buf_size(adapter, rsd);
+				fp->page = rsd->page;
+				fp->offset = q->offset;
+				fp->size = min(bufsz, len);
+				len -= fp->size;
+				if (!len)
+					break;
+				unmap_rx_buf(q->adap, &rxq->fl);
+			}
+
+			si.sgetstamp = SGE_TIMESTAMP_G(
+					be64_to_cpu(rc->last_flit));
+			/*
+			 * Last buffer remains mapped so explicitly make it
+			 * coherent for CPU access.
+			 */
+			dma_sync_single_for_cpu(q->adap->pdev_dev,
+						get_buf_addr(rsd),
+						fp->size, DMA_FROM_DEVICE);
+
+			si.va = page_address(si.frags[0].page) +
+				si.frags[0].offset;
+			prefetch(si.va);
+
+			si.nfrags = frags + 1;
+			ret = q->handler(q, q->cur_desc, &si);
+			if (likely(ret == 0))
+				q->offset += ALIGN(fp->size, s->fl_align);
+			else
+				restore_rx_bufs(&si, &rxq->fl, frags);
+		} else if (likely(rsp_type == RSPD_TYPE_CPL_X)) {
+			ret = q->handler(q, q->cur_desc, NULL);
+		} else {
+			ret = q->handler(q, (const __be64 *)rc, CXGB4_MSG_AN);
+		}
+
+		if (unlikely(ret)) {
+			/* couldn't process descriptor, back off for recovery */
+			q->next_intr_params = QINTR_TIMER_IDX_V(NOMEM_TMR_IDX);
+			break;
+		}
+
+		rspq_next(q);
+		budget_left--;
+	}
+
+	if (q->offset >= 0 && fl_cap(&rxq->fl) - rxq->fl.avail >= 16)
+		__refill_fl(q->adap, &rxq->fl);
+	return budget - budget_left;
+}
+
+/**
+ *	napi_rx_handler - the NAPI handler for Rx processing
+ *	@napi: the napi instance
+ *	@budget: how many packets we can process in this round
+ *
+ *	Handler for new data events when using NAPI.  This does not need any
+ *	locking or protection from interrupts as data interrupts are off at
+ *	this point and other adapter interrupts do not interfere (the latter
+ *	in not a concern at all with MSI-X as non-data interrupts then have
+ *	a separate handler).
+ */
+static int napi_rx_handler(struct napi_struct *napi, int budget)
+{
+	unsigned int params;
+	struct sge_rspq *q = container_of(napi, struct sge_rspq, napi);
+	int work_done;
+	u32 val;
+
+	work_done = process_responses(q, budget);
+	if (likely(work_done < budget)) {
+		int timer_index;
+
+		napi_complete_done(napi, work_done);
+		timer_index = QINTR_TIMER_IDX_G(q->next_intr_params);
+
+		if (q->adaptive_rx) {
+			if (work_done > max(timer_pkt_quota[timer_index],
+					    MIN_NAPI_WORK))
+				timer_index = (timer_index + 1);
+			else
+				timer_index = timer_index - 1;
+
+			timer_index = clamp(timer_index, 0, SGE_TIMERREGS - 1);
+			q->next_intr_params =
+					QINTR_TIMER_IDX_V(timer_index) |
+					QINTR_CNT_EN_V(0);
+			params = q->next_intr_params;
+		} else {
+			params = q->next_intr_params;
+			q->next_intr_params = q->intr_params;
+		}
+	} else
+		params = QINTR_TIMER_IDX_V(7);
+
+	val = CIDXINC_V(work_done) | SEINTARM_V(params);
+
+	/* If we don't have access to the new User GTS (T5+), use the old
+	 * doorbell mechanism; otherwise use the new BAR2 mechanism.
+	 */
+	if (unlikely(q->bar2_addr == NULL)) {
+		t4_write_reg(q->adap, MYPF_REG(SGE_PF_GTS_A),
+			     val | INGRESSQID_V((u32)q->cntxt_id));
+	} else {
+		writel(val | INGRESSQID_V(q->bar2_qid),
+		       q->bar2_addr + SGE_UDB_GTS);
+		wmb();
+	}
+	return work_done;
+}
+
+void cxgb4_ethofld_restart(struct tasklet_struct *t)
+{
+	struct sge_eosw_txq *eosw_txq = from_tasklet(eosw_txq, t,
+						     qresume_tsk);
+	int pktcount;
+
+	spin_lock(&eosw_txq->lock);
+	pktcount = eosw_txq->cidx - eosw_txq->last_cidx;
+	if (pktcount < 0)
+		pktcount += eosw_txq->ndesc;
+
+	if (pktcount) {
+		cxgb4_eosw_txq_free_desc(netdev2adap(eosw_txq->netdev),
+					 eosw_txq, pktcount);
+		eosw_txq->inuse -= pktcount;
+	}
+
+	/* There may be some packets waiting for completions. So,
+	 * attempt to send these packets now.
+	 */
+	ethofld_xmit(eosw_txq->netdev, eosw_txq);
+	spin_unlock(&eosw_txq->lock);
+}
+
+/* cxgb4_ethofld_rx_handler - Process ETHOFLD Tx completions
+ * @q: the response queue that received the packet
+ * @rsp: the response queue descriptor holding the CPL message
+ * @si: the gather list of packet fragments
+ *
+ * Process a ETHOFLD Tx completion. Increment the cidx here, but
+ * free up the descriptors in a tasklet later.
+ */
+int cxgb4_ethofld_rx_handler(struct sge_rspq *q, const __be64 *rsp,
+			     const struct pkt_gl *si)
+{
+	u8 opcode = ((const struct rss_header *)rsp)->opcode;
+
+	/* skip RSS header */
+	rsp++;
+
+	if (opcode == CPL_FW4_ACK) {
+		const struct cpl_fw4_ack *cpl;
+		struct sge_eosw_txq *eosw_txq;
+		struct eotid_entry *entry;
+		struct sk_buff *skb;
+		u32 hdr_len, eotid;
+		u8 flits, wrlen16;
+		int credits;
+
+		cpl = (const struct cpl_fw4_ack *)rsp;
+		eotid = CPL_FW4_ACK_FLOWID_G(ntohl(OPCODE_TID(cpl))) -
+			q->adap->tids.eotid_base;
+		entry = cxgb4_lookup_eotid(&q->adap->tids, eotid);
+		if (!entry)
+			goto out_done;
+
+		eosw_txq = (struct sge_eosw_txq *)entry->data;
+		if (!eosw_txq)
+			goto out_done;
+
+		spin_lock(&eosw_txq->lock);
+		credits = cpl->credits;
+		while (credits > 0) {
+			skb = eosw_txq->desc[eosw_txq->cidx].skb;
+			if (!skb)
+				break;
+
+			if (unlikely((eosw_txq->state ==
+				      CXGB4_EO_STATE_FLOWC_OPEN_REPLY ||
+				      eosw_txq->state ==
+				      CXGB4_EO_STATE_FLOWC_CLOSE_REPLY) &&
+				     eosw_txq->cidx == eosw_txq->flowc_idx)) {
+				flits = DIV_ROUND_UP(skb->len, 8);
+				if (eosw_txq->state ==
+				    CXGB4_EO_STATE_FLOWC_OPEN_REPLY)
+					eosw_txq->state = CXGB4_EO_STATE_ACTIVE;
+				else
+					eosw_txq->state = CXGB4_EO_STATE_CLOSED;
+				complete(&eosw_txq->completion);
+			} else {
+				hdr_len = eth_get_headlen(eosw_txq->netdev,
+							  skb->data,
+							  skb_headlen(skb));
+				flits = ethofld_calc_tx_flits(q->adap, skb,
+							      hdr_len);
+			}
+			eosw_txq_advance_index(&eosw_txq->cidx, 1,
+					       eosw_txq->ndesc);
+			wrlen16 = DIV_ROUND_UP(flits * 8, 16);
+			credits -= wrlen16;
+		}
+
+		eosw_txq->cred += cpl->credits;
+		eosw_txq->ncompl--;
+
+		spin_unlock(&eosw_txq->lock);
+
+		/* Schedule a tasklet to reclaim SKBs and restart ETHOFLD Tx,
+		 * if there were packets waiting for completion.
+		 */
+		tasklet_schedule(&eosw_txq->qresume_tsk);
+	}
+
+out_done:
+	return 0;
+}
+
+/*
+ * The MSI-X interrupt handler for an SGE response queue.
+ */
+irqreturn_t t4_sge_intr_msix(int irq, void *cookie)
+{
+	struct sge_rspq *q = cookie;
+
+	napi_schedule(&q->napi);
+	return IRQ_HANDLED;
+}
+
+/*
+ * Process the indirect interrupt entries in the interrupt queue and kick off
+ * NAPI for each queue that has generated an entry.
+ */
+static unsigned int process_intrq(struct adapter *adap)
+{
+	unsigned int credits;
+	const struct rsp_ctrl *rc;
+	struct sge_rspq *q = &adap->sge.intrq;
+	u32 val;
+
+	spin_lock(&adap->sge.intrq_lock);
+	for (credits = 0; ; credits++) {
+		rc = (void *)q->cur_desc + (q->iqe_len - sizeof(*rc));
+		if (!is_new_response(rc, q))
+			break;
+
+		dma_rmb();
+		if (RSPD_TYPE_G(rc->type_gen) == RSPD_TYPE_INTR_X) {
+			unsigned int qid = ntohl(rc->pldbuflen_qid);
+
+			qid -= adap->sge.ingr_start;
+			napi_schedule(&adap->sge.ingr_map[qid]->napi);
+		}
+
+		rspq_next(q);
+	}
+
+	val =  CIDXINC_V(credits) | SEINTARM_V(q->intr_params);
+
+	/* If we don't have access to the new User GTS (T5+), use the old
+	 * doorbell mechanism; otherwise use the new BAR2 mechanism.
+	 */
+	if (unlikely(q->bar2_addr == NULL)) {
+		t4_write_reg(adap, MYPF_REG(SGE_PF_GTS_A),
+			     val | INGRESSQID_V(q->cntxt_id));
+	} else {
+		writel(val | INGRESSQID_V(q->bar2_qid),
+		       q->bar2_addr + SGE_UDB_GTS);
+		wmb();
+	}
+	spin_unlock(&adap->sge.intrq_lock);
+	return credits;
+}
+
+/*
+ * The MSI interrupt handler, which handles data events from SGE response queues
+ * as well as error and other async events as they all use the same MSI vector.
+ */
+static irqreturn_t t4_intr_msi(int irq, void *cookie)
+{
+	struct adapter *adap = cookie;
+
+	if (adap->flags & CXGB4_MASTER_PF)
+		t4_slow_intr_handler(adap);
+	process_intrq(adap);
+	return IRQ_HANDLED;
+}
+
+/*
+ * Interrupt handler for legacy INTx interrupts.
+ * Handles data events from SGE response queues as well as error and other
+ * async events as they all use the same interrupt line.
+ */
+static irqreturn_t t4_intr_intx(int irq, void *cookie)
+{
+	struct adapter *adap = cookie;
+
+	t4_write_reg(adap, MYPF_REG(PCIE_PF_CLI_A), 0);
+	if (((adap->flags & CXGB4_MASTER_PF) && t4_slow_intr_handler(adap)) |
+	    process_intrq(adap))
+		return IRQ_HANDLED;
+	return IRQ_NONE;             /* probably shared interrupt */
+}
+
+/**
+ *	t4_intr_handler - select the top-level interrupt handler
+ *	@adap: the adapter
+ *
+ *	Selects the top-level interrupt handler based on the type of interrupts
+ *	(MSI-X, MSI, or INTx).
+ */
+irq_handler_t t4_intr_handler(struct adapter *adap)
+{
+	if (adap->flags & CXGB4_USING_MSIX)
+		return t4_sge_intr_msix;
+	if (adap->flags & CXGB4_USING_MSI)
+		return t4_intr_msi;
+	return t4_intr_intx;
+}
+
+static void sge_rx_timer_cb(struct timer_list *t)
+{
+	unsigned long m;
+	unsigned int i;
+	struct adapter *adap = from_timer(adap, t, sge.rx_timer);
+	struct sge *s = &adap->sge;
+
+	for (i = 0; i < BITS_TO_LONGS(s->egr_sz); i++)
+		for (m = s->starving_fl[i]; m; m &= m - 1) {
+			struct sge_eth_rxq *rxq;
+			unsigned int id = __ffs(m) + i * BITS_PER_LONG;
+			struct sge_fl *fl = s->egr_map[id];
+
+			clear_bit(id, s->starving_fl);
+			smp_mb__after_atomic();
+
+			if (fl_starving(adap, fl)) {
+				rxq = container_of(fl, struct sge_eth_rxq, fl);
+				if (napi_reschedule(&rxq->rspq.napi))
+					fl->starving++;
+				else
+					set_bit(id, s->starving_fl);
+			}
+		}
+	/* The remainder of the SGE RX Timer Callback routine is dedicated to
+	 * global Master PF activities like checking for chip ingress stalls,
+	 * etc.
+	 */
+	if (!(adap->flags & CXGB4_MASTER_PF))
+		goto done;
+
+	t4_idma_monitor(adap, &s->idma_monitor, HZ, RX_QCHECK_PERIOD);
+
+done:
+	mod_timer(&s->rx_timer, jiffies + RX_QCHECK_PERIOD);
+}
+
+static void sge_tx_timer_cb(struct timer_list *t)
+{
+	struct adapter *adap = from_timer(adap, t, sge.tx_timer);
+	struct sge *s = &adap->sge;
+	unsigned long m, period;
+	unsigned int i, budget;
+
+	for (i = 0; i < BITS_TO_LONGS(s->egr_sz); i++)
+		for (m = s->txq_maperr[i]; m; m &= m - 1) {
+			unsigned long id = __ffs(m) + i * BITS_PER_LONG;
+			struct sge_uld_txq *txq = s->egr_map[id];
+
+			clear_bit(id, s->txq_maperr);
+			tasklet_schedule(&txq->qresume_tsk);
+		}
+
+	if (!is_t4(adap->params.chip)) {
+		struct sge_eth_txq *q = &s->ptptxq;
+		int avail;
+
+		spin_lock(&adap->ptp_lock);
+		avail = reclaimable(&q->q);
+
+		if (avail) {
+			free_tx_desc(adap, &q->q, avail, false);
+			q->q.in_use -= avail;
+		}
+		spin_unlock(&adap->ptp_lock);
+	}
+
+	budget = MAX_TIMER_TX_RECLAIM;
+	i = s->ethtxq_rover;
+	do {
+		budget -= t4_sge_eth_txq_egress_update(adap, &s->ethtxq[i],
+						       budget);
+		if (!budget)
+			break;
+
+		if (++i >= s->ethqsets)
+			i = 0;
+	} while (i != s->ethtxq_rover);
+	s->ethtxq_rover = i;
+
+	if (budget == 0) {
+		/* If we found too many reclaimable packets schedule a timer
+		 * in the near future to continue where we left off.
+		 */
+		period = 2;
+	} else {
+		/* We reclaimed all reclaimable TX Descriptors, so reschedule
+		 * at the normal period.
+		 */
+		period = TX_QCHECK_PERIOD;
+	}
+
+	mod_timer(&s->tx_timer, jiffies + period);
+}
+
+/**
+ *	bar2_address - return the BAR2 address for an SGE Queue's Registers
+ *	@adapter: the adapter
+ *	@qid: the SGE Queue ID
+ *	@qtype: the SGE Queue Type (Egress or Ingress)
+ *	@pbar2_qid: BAR2 Queue ID or 0 for Queue ID inferred SGE Queues
+ *
+ *	Returns the BAR2 address for the SGE Queue Registers associated with
+ *	@qid.  If BAR2 SGE Registers aren't available, returns NULL.  Also
+ *	returns the BAR2 Queue ID to be used with writes to the BAR2 SGE
+ *	Queue Registers.  If the BAR2 Queue ID is 0, then "Inferred Queue ID"
+ *	Registers are supported (e.g. the Write Combining Doorbell Buffer).
+ */
+static void __iomem *bar2_address(struct adapter *adapter,
+				  unsigned int qid,
+				  enum t4_bar2_qtype qtype,
+				  unsigned int *pbar2_qid)
+{
+	u64 bar2_qoffset;
+	int ret;
+
+	ret = t4_bar2_sge_qregs(adapter, qid, qtype, 0,
+				&bar2_qoffset, pbar2_qid);
+	if (ret)
+		return NULL;
+
+	return adapter->bar2 + bar2_qoffset;
+}
+
+/* @intr_idx: MSI/MSI-X vector if >=0, -(absolute qid + 1) if < 0
+ * @cong: < 0 -> no congestion feedback, >= 0 -> congestion channel map
+ */
+int t4_sge_alloc_rxq(struct adapter *adap, struct sge_rspq *iq, bool fwevtq,
+		     struct net_device *dev, int intr_idx,
+		     struct sge_fl *fl, rspq_handler_t hnd,
+		     rspq_flush_handler_t flush_hnd, int cong)
+{
+	int ret, flsz = 0;
+	struct fw_iq_cmd c;
+	struct sge *s = &adap->sge;
+	struct port_info *pi = netdev_priv(dev);
+	int relaxed = !(adap->flags & CXGB4_ROOT_NO_RELAXED_ORDERING);
+
+	/* Size needs to be multiple of 16, including status entry. */
+	iq->size = roundup(iq->size, 16);
+
+	iq->desc = alloc_ring(adap->pdev_dev, iq->size, iq->iqe_len, 0,
+			      &iq->phys_addr, NULL, 0,
+			      dev_to_node(adap->pdev_dev));
+	if (!iq->desc)
+		return -ENOMEM;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = htonl(FW_CMD_OP_V(FW_IQ_CMD) | FW_CMD_REQUEST_F |
+			    FW_CMD_WRITE_F | FW_CMD_EXEC_F |
+			    FW_IQ_CMD_PFN_V(adap->pf) | FW_IQ_CMD_VFN_V(0));
+	c.alloc_to_len16 = htonl(FW_IQ_CMD_ALLOC_F | FW_IQ_CMD_IQSTART_F |
+				 FW_LEN16(c));
+	c.type_to_iqandstindex = htonl(FW_IQ_CMD_TYPE_V(FW_IQ_TYPE_FL_INT_CAP) |
+		FW_IQ_CMD_IQASYNCH_V(fwevtq) | FW_IQ_CMD_VIID_V(pi->viid) |
+		FW_IQ_CMD_IQANDST_V(intr_idx < 0) |
+		FW_IQ_CMD_IQANUD_V(UPDATEDELIVERY_INTERRUPT_X) |
+		FW_IQ_CMD_IQANDSTINDEX_V(intr_idx >= 0 ? intr_idx :
+							-intr_idx - 1));
+	c.iqdroprss_to_iqesize = htons(FW_IQ_CMD_IQPCIECH_V(pi->tx_chan) |
+		FW_IQ_CMD_IQGTSMODE_F |
+		FW_IQ_CMD_IQINTCNTTHRESH_V(iq->pktcnt_idx) |
+		FW_IQ_CMD_IQESIZE_V(ilog2(iq->iqe_len) - 4));
+	c.iqsize = htons(iq->size);
+	c.iqaddr = cpu_to_be64(iq->phys_addr);
+	if (cong >= 0)
+		c.iqns_to_fl0congen = htonl(FW_IQ_CMD_IQFLINTCONGEN_F |
+				FW_IQ_CMD_IQTYPE_V(cong ? FW_IQ_IQTYPE_NIC
+							:  FW_IQ_IQTYPE_OFLD));
+
+	if (fl) {
+		unsigned int chip_ver =
+			CHELSIO_CHIP_VERSION(adap->params.chip);
+
+		/* Allocate the ring for the hardware free list (with space
+		 * for its status page) along with the associated software
+		 * descriptor ring.  The free list size needs to be a multiple
+		 * of the Egress Queue Unit and at least 2 Egress Units larger
+		 * than the SGE's Egress Congrestion Threshold
+		 * (fl_starve_thres - 1).
+		 */
+		if (fl->size < s->fl_starve_thres - 1 + 2 * 8)
+			fl->size = s->fl_starve_thres - 1 + 2 * 8;
+		fl->size = roundup(fl->size, 8);
+		fl->desc = alloc_ring(adap->pdev_dev, fl->size, sizeof(__be64),
+				      sizeof(struct rx_sw_desc), &fl->addr,
+				      &fl->sdesc, s->stat_len,
+				      dev_to_node(adap->pdev_dev));
+		if (!fl->desc)
+			goto fl_nomem;
+
+		flsz = fl->size / 8 + s->stat_len / sizeof(struct tx_desc);
+		c.iqns_to_fl0congen |= htonl(FW_IQ_CMD_FL0PACKEN_F |
+					     FW_IQ_CMD_FL0FETCHRO_V(relaxed) |
+					     FW_IQ_CMD_FL0DATARO_V(relaxed) |
+					     FW_IQ_CMD_FL0PADEN_F);
+		if (cong >= 0)
+			c.iqns_to_fl0congen |=
+				htonl(FW_IQ_CMD_FL0CNGCHMAP_V(cong) |
+				      FW_IQ_CMD_FL0CONGCIF_F |
+				      FW_IQ_CMD_FL0CONGEN_F);
+		/* In T6, for egress queue type FL there is internal overhead
+		 * of 16B for header going into FLM module.  Hence the maximum
+		 * allowed burst size is 448 bytes.  For T4/T5, the hardware
+		 * doesn't coalesce fetch requests if more than 64 bytes of
+		 * Free List pointers are provided, so we use a 128-byte Fetch
+		 * Burst Minimum there (T6 implements coalescing so we can use
+		 * the smaller 64-byte value there).
+		 */
+		c.fl0dcaen_to_fl0cidxfthresh =
+			htons(FW_IQ_CMD_FL0FBMIN_V(chip_ver <= CHELSIO_T5 ?
+						   FETCHBURSTMIN_128B_X :
+						   FETCHBURSTMIN_64B_T6_X) |
+			      FW_IQ_CMD_FL0FBMAX_V((chip_ver <= CHELSIO_T5) ?
+						   FETCHBURSTMAX_512B_X :
+						   FETCHBURSTMAX_256B_X));
+		c.fl0size = htons(flsz);
+		c.fl0addr = cpu_to_be64(fl->addr);
+	}
+
+	ret = t4_wr_mbox(adap, adap->mbox, &c, sizeof(c), &c);
+	if (ret)
+		goto err;
+
+	netif_napi_add(dev, &iq->napi, napi_rx_handler);
+	iq->cur_desc = iq->desc;
+	iq->cidx = 0;
+	iq->gen = 1;
+	iq->next_intr_params = iq->intr_params;
+	iq->cntxt_id = ntohs(c.iqid);
+	iq->abs_id = ntohs(c.physiqid);
+	iq->bar2_addr = bar2_address(adap,
+				     iq->cntxt_id,
+				     T4_BAR2_QTYPE_INGRESS,
+				     &iq->bar2_qid);
+	iq->size--;                           /* subtract status entry */
+	iq->netdev = dev;
+	iq->handler = hnd;
+	iq->flush_handler = flush_hnd;
+
+	memset(&iq->lro_mgr, 0, sizeof(struct t4_lro_mgr));
+	skb_queue_head_init(&iq->lro_mgr.lroq);
+
+	/* set offset to -1 to distinguish ingress queues without FL */
+	iq->offset = fl ? 0 : -1;
+
+	adap->sge.ingr_map[iq->cntxt_id - adap->sge.ingr_start] = iq;
+
+	if (fl) {
+		fl->cntxt_id = ntohs(c.fl0id);
+		fl->avail = fl->pend_cred = 0;
+		fl->pidx = fl->cidx = 0;
+		fl->alloc_failed = fl->large_alloc_failed = fl->starving = 0;
+		adap->sge.egr_map[fl->cntxt_id - adap->sge.egr_start] = fl;
+
+		/* Note, we must initialize the BAR2 Free List User Doorbell
+		 * information before refilling the Free List!
+		 */
+		fl->bar2_addr = bar2_address(adap,
+					     fl->cntxt_id,
+					     T4_BAR2_QTYPE_EGRESS,
+					     &fl->bar2_qid);
+		refill_fl(adap, fl, fl_cap(fl), GFP_KERNEL);
+	}
+
+	/* For T5 and later we attempt to set up the Congestion Manager values
+	 * of the new RX Ethernet Queue.  This should really be handled by
+	 * firmware because it's more complex than any host driver wants to
+	 * get involved with and it's different per chip and this is almost
+	 * certainly wrong.  Firmware would be wrong as well, but it would be
+	 * a lot easier to fix in one place ...  For now we do something very
+	 * simple (and hopefully less wrong).
+	 */
+	if (!is_t4(adap->params.chip) && cong >= 0) {
+		u32 param, val, ch_map = 0;
+		int i;
+		u16 cng_ch_bits_log = adap->params.arch.cng_ch_bits_log;
+
+		param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DMAQ) |
+			 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DMAQ_CONM_CTXT) |
+			 FW_PARAMS_PARAM_YZ_V(iq->cntxt_id));
+		if (cong == 0) {
+			val = CONMCTXT_CNGTPMODE_V(CONMCTXT_CNGTPMODE_QUEUE_X);
+		} else {
+			val =
+			    CONMCTXT_CNGTPMODE_V(CONMCTXT_CNGTPMODE_CHANNEL_X);
+			for (i = 0; i < 4; i++) {
+				if (cong & (1 << i))
+					ch_map |= 1 << (i << cng_ch_bits_log);
+			}
+			val |= CONMCTXT_CNGCHMAP_V(ch_map);
+		}
+		ret = t4_set_params(adap, adap->mbox, adap->pf, 0, 1,
+				    &param, &val);
+		if (ret)
+			dev_warn(adap->pdev_dev, "Failed to set Congestion"
+				 " Manager Context for Ingress Queue %d: %d\n",
+				 iq->cntxt_id, -ret);
+	}
+
+	return 0;
+
+fl_nomem:
+	ret = -ENOMEM;
+err:
+	if (iq->desc) {
+		dma_free_coherent(adap->pdev_dev, iq->size * iq->iqe_len,
+				  iq->desc, iq->phys_addr);
+		iq->desc = NULL;
+	}
+	if (fl && fl->desc) {
+		kfree(fl->sdesc);
+		fl->sdesc = NULL;
+		dma_free_coherent(adap->pdev_dev, flsz * sizeof(struct tx_desc),
+				  fl->desc, fl->addr);
+		fl->desc = NULL;
+	}
+	return ret;
+}
+
+static void init_txq(struct adapter *adap, struct sge_txq *q, unsigned int id)
+{
+	q->cntxt_id = id;
+	q->bar2_addr = bar2_address(adap,
+				    q->cntxt_id,
+				    T4_BAR2_QTYPE_EGRESS,
+				    &q->bar2_qid);
+	q->in_use = 0;
+	q->cidx = q->pidx = 0;
+	q->stops = q->restarts = 0;
+	q->stat = (void *)&q->desc[q->size];
+	spin_lock_init(&q->db_lock);
+	adap->sge.egr_map[id - adap->sge.egr_start] = q;
+}
+
+/**
+ *	t4_sge_alloc_eth_txq - allocate an Ethernet TX Queue
+ *	@adap: the adapter
+ *	@txq: the SGE Ethernet TX Queue to initialize
+ *	@dev: the Linux Network Device
+ *	@netdevq: the corresponding Linux TX Queue
+ *	@iqid: the Ingress Queue to which to deliver CIDX Update messages
+ *	@dbqt: whether this TX Queue will use the SGE Doorbell Queue Timers
+ */
+int t4_sge_alloc_eth_txq(struct adapter *adap, struct sge_eth_txq *txq,
+			 struct net_device *dev, struct netdev_queue *netdevq,
+			 unsigned int iqid, u8 dbqt)
+{
+	unsigned int chip_ver = CHELSIO_CHIP_VERSION(adap->params.chip);
+	struct port_info *pi = netdev_priv(dev);
+	struct sge *s = &adap->sge;
+	struct fw_eq_eth_cmd c;
+	int ret, nentries;
+
+	/* Add status entries */
+	nentries = txq->q.size + s->stat_len / sizeof(struct tx_desc);
+
+	txq->q.desc = alloc_ring(adap->pdev_dev, txq->q.size,
+			sizeof(struct tx_desc), sizeof(struct tx_sw_desc),
+			&txq->q.phys_addr, &txq->q.sdesc, s->stat_len,
+			netdev_queue_numa_node_read(netdevq));
+	if (!txq->q.desc)
+		return -ENOMEM;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = htonl(FW_CMD_OP_V(FW_EQ_ETH_CMD) | FW_CMD_REQUEST_F |
+			    FW_CMD_WRITE_F | FW_CMD_EXEC_F |
+			    FW_EQ_ETH_CMD_PFN_V(adap->pf) |
+			    FW_EQ_ETH_CMD_VFN_V(0));
+	c.alloc_to_len16 = htonl(FW_EQ_ETH_CMD_ALLOC_F |
+				 FW_EQ_ETH_CMD_EQSTART_F | FW_LEN16(c));
+
+	/* For TX Ethernet Queues using the SGE Doorbell Queue Timer
+	 * mechanism, we use Ingress Queue messages for Hardware Consumer
+	 * Index Updates on the TX Queue.  Otherwise we have the Hardware
+	 * write the CIDX Updates into the Status Page at the end of the
+	 * TX Queue.
+	 */
+	c.autoequiqe_to_viid = htonl(((chip_ver <= CHELSIO_T5) ?
+				      FW_EQ_ETH_CMD_AUTOEQUIQE_F :
+				      FW_EQ_ETH_CMD_AUTOEQUEQE_F) |
+				     FW_EQ_ETH_CMD_VIID_V(pi->viid));
+
+	c.fetchszm_to_iqid =
+		htonl(FW_EQ_ETH_CMD_HOSTFCMODE_V((chip_ver <= CHELSIO_T5) ?
+						 HOSTFCMODE_INGRESS_QUEUE_X :
+						 HOSTFCMODE_STATUS_PAGE_X) |
+		      FW_EQ_ETH_CMD_PCIECHN_V(pi->tx_chan) |
+		      FW_EQ_ETH_CMD_FETCHRO_F | FW_EQ_ETH_CMD_IQID_V(iqid));
+
+	/* Note that the CIDX Flush Threshold should match MAX_TX_RECLAIM. */
+	c.dcaen_to_eqsize =
+		htonl(FW_EQ_ETH_CMD_FBMIN_V(chip_ver <= CHELSIO_T5
+					    ? FETCHBURSTMIN_64B_X
+					    : FETCHBURSTMIN_64B_T6_X) |
+		      FW_EQ_ETH_CMD_FBMAX_V(FETCHBURSTMAX_512B_X) |
+		      FW_EQ_ETH_CMD_CIDXFTHRESH_V(CIDXFLUSHTHRESH_32_X) |
+		      FW_EQ_ETH_CMD_CIDXFTHRESHO_V(chip_ver == CHELSIO_T5) |
+		      FW_EQ_ETH_CMD_EQSIZE_V(nentries));
+
+	c.eqaddr = cpu_to_be64(txq->q.phys_addr);
+
+	/* If we're using the SGE Doorbell Queue Timer mechanism, pass in the
+	 * currently configured Timer Index.  THis can be changed later via an
+	 * ethtool -C tx-usecs {Timer Val} command.  Note that the SGE
+	 * Doorbell Queue mode is currently automatically enabled in the
+	 * Firmware by setting either AUTOEQUEQE or AUTOEQUIQE ...
+	 */
+	if (dbqt)
+		c.timeren_timerix =
+			cpu_to_be32(FW_EQ_ETH_CMD_TIMEREN_F |
+				    FW_EQ_ETH_CMD_TIMERIX_V(txq->dbqtimerix));
+
+	ret = t4_wr_mbox(adap, adap->mbox, &c, sizeof(c), &c);
+	if (ret) {
+		kfree(txq->q.sdesc);
+		txq->q.sdesc = NULL;
+		dma_free_coherent(adap->pdev_dev,
+				  nentries * sizeof(struct tx_desc),
+				  txq->q.desc, txq->q.phys_addr);
+		txq->q.desc = NULL;
+		return ret;
+	}
+
+	txq->q.q_type = CXGB4_TXQ_ETH;
+	init_txq(adap, &txq->q, FW_EQ_ETH_CMD_EQID_G(ntohl(c.eqid_pkd)));
+	txq->txq = netdevq;
+	txq->tso = 0;
+	txq->uso = 0;
+	txq->tx_cso = 0;
+	txq->vlan_ins = 0;
+	txq->mapping_err = 0;
+	txq->dbqt = dbqt;
+
+	return 0;
+}
+
+int t4_sge_alloc_ctrl_txq(struct adapter *adap, struct sge_ctrl_txq *txq,
+			  struct net_device *dev, unsigned int iqid,
+			  unsigned int cmplqid)
+{
+	unsigned int chip_ver = CHELSIO_CHIP_VERSION(adap->params.chip);
+	struct port_info *pi = netdev_priv(dev);
+	struct sge *s = &adap->sge;
+	struct fw_eq_ctrl_cmd c;
+	int ret, nentries;
+
+	/* Add status entries */
+	nentries = txq->q.size + s->stat_len / sizeof(struct tx_desc);
+
+	txq->q.desc = alloc_ring(adap->pdev_dev, nentries,
+				 sizeof(struct tx_desc), 0, &txq->q.phys_addr,
+				 NULL, 0, dev_to_node(adap->pdev_dev));
+	if (!txq->q.desc)
+		return -ENOMEM;
+
+	c.op_to_vfn = htonl(FW_CMD_OP_V(FW_EQ_CTRL_CMD) | FW_CMD_REQUEST_F |
+			    FW_CMD_WRITE_F | FW_CMD_EXEC_F |
+			    FW_EQ_CTRL_CMD_PFN_V(adap->pf) |
+			    FW_EQ_CTRL_CMD_VFN_V(0));
+	c.alloc_to_len16 = htonl(FW_EQ_CTRL_CMD_ALLOC_F |
+				 FW_EQ_CTRL_CMD_EQSTART_F | FW_LEN16(c));
+	c.cmpliqid_eqid = htonl(FW_EQ_CTRL_CMD_CMPLIQID_V(cmplqid));
+	c.physeqid_pkd = htonl(0);
+	c.fetchszm_to_iqid =
+		htonl(FW_EQ_CTRL_CMD_HOSTFCMODE_V(HOSTFCMODE_STATUS_PAGE_X) |
+		      FW_EQ_CTRL_CMD_PCIECHN_V(pi->tx_chan) |
+		      FW_EQ_CTRL_CMD_FETCHRO_F | FW_EQ_CTRL_CMD_IQID_V(iqid));
+	c.dcaen_to_eqsize =
+		htonl(FW_EQ_CTRL_CMD_FBMIN_V(chip_ver <= CHELSIO_T5
+					     ? FETCHBURSTMIN_64B_X
+					     : FETCHBURSTMIN_64B_T6_X) |
+		      FW_EQ_CTRL_CMD_FBMAX_V(FETCHBURSTMAX_512B_X) |
+		      FW_EQ_CTRL_CMD_CIDXFTHRESH_V(CIDXFLUSHTHRESH_32_X) |
+		      FW_EQ_CTRL_CMD_EQSIZE_V(nentries));
+	c.eqaddr = cpu_to_be64(txq->q.phys_addr);
+
+	ret = t4_wr_mbox(adap, adap->mbox, &c, sizeof(c), &c);
+	if (ret) {
+		dma_free_coherent(adap->pdev_dev,
+				  nentries * sizeof(struct tx_desc),
+				  txq->q.desc, txq->q.phys_addr);
+		txq->q.desc = NULL;
+		return ret;
+	}
+
+	txq->q.q_type = CXGB4_TXQ_CTRL;
+	init_txq(adap, &txq->q, FW_EQ_CTRL_CMD_EQID_G(ntohl(c.cmpliqid_eqid)));
+	txq->adap = adap;
+	skb_queue_head_init(&txq->sendq);
+	tasklet_setup(&txq->qresume_tsk, restart_ctrlq);
+	txq->full = 0;
+	return 0;
+}
+
+int t4_sge_mod_ctrl_txq(struct adapter *adap, unsigned int eqid,
+			unsigned int cmplqid)
+{
+	u32 param, val;
+
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DMAQ) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DMAQ_EQ_CMPLIQID_CTRL) |
+		 FW_PARAMS_PARAM_YZ_V(eqid));
+	val = cmplqid;
+	return t4_set_params(adap, adap->mbox, adap->pf, 0, 1, &param, &val);
+}
+
+static int t4_sge_alloc_ofld_txq(struct adapter *adap, struct sge_txq *q,
+				 struct net_device *dev, u32 cmd, u32 iqid)
+{
+	unsigned int chip_ver = CHELSIO_CHIP_VERSION(adap->params.chip);
+	struct port_info *pi = netdev_priv(dev);
+	struct sge *s = &adap->sge;
+	struct fw_eq_ofld_cmd c;
+	u32 fb_min, nentries;
+	int ret;
+
+	/* Add status entries */
+	nentries = q->size + s->stat_len / sizeof(struct tx_desc);
+	q->desc = alloc_ring(adap->pdev_dev, q->size, sizeof(struct tx_desc),
+			     sizeof(struct tx_sw_desc), &q->phys_addr,
+			     &q->sdesc, s->stat_len, NUMA_NO_NODE);
+	if (!q->desc)
+		return -ENOMEM;
+
+	if (chip_ver <= CHELSIO_T5)
+		fb_min = FETCHBURSTMIN_64B_X;
+	else
+		fb_min = FETCHBURSTMIN_64B_T6_X;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = htonl(FW_CMD_OP_V(cmd) | FW_CMD_REQUEST_F |
+			    FW_CMD_WRITE_F | FW_CMD_EXEC_F |
+			    FW_EQ_OFLD_CMD_PFN_V(adap->pf) |
+			    FW_EQ_OFLD_CMD_VFN_V(0));
+	c.alloc_to_len16 = htonl(FW_EQ_OFLD_CMD_ALLOC_F |
+				 FW_EQ_OFLD_CMD_EQSTART_F | FW_LEN16(c));
+	c.fetchszm_to_iqid =
+		htonl(FW_EQ_OFLD_CMD_HOSTFCMODE_V(HOSTFCMODE_STATUS_PAGE_X) |
+		      FW_EQ_OFLD_CMD_PCIECHN_V(pi->tx_chan) |
+		      FW_EQ_OFLD_CMD_FETCHRO_F | FW_EQ_OFLD_CMD_IQID_V(iqid));
+	c.dcaen_to_eqsize =
+		htonl(FW_EQ_OFLD_CMD_FBMIN_V(fb_min) |
+		      FW_EQ_OFLD_CMD_FBMAX_V(FETCHBURSTMAX_512B_X) |
+		      FW_EQ_OFLD_CMD_CIDXFTHRESH_V(CIDXFLUSHTHRESH_32_X) |
+		      FW_EQ_OFLD_CMD_EQSIZE_V(nentries));
+	c.eqaddr = cpu_to_be64(q->phys_addr);
+
+	ret = t4_wr_mbox(adap, adap->mbox, &c, sizeof(c), &c);
+	if (ret) {
+		kfree(q->sdesc);
+		q->sdesc = NULL;
+		dma_free_coherent(adap->pdev_dev,
+				  nentries * sizeof(struct tx_desc),
+				  q->desc, q->phys_addr);
+		q->desc = NULL;
+		return ret;
+	}
+
+	init_txq(adap, q, FW_EQ_OFLD_CMD_EQID_G(ntohl(c.eqid_pkd)));
+	return 0;
+}
+
+int t4_sge_alloc_uld_txq(struct adapter *adap, struct sge_uld_txq *txq,
+			 struct net_device *dev, unsigned int iqid,
+			 unsigned int uld_type)
+{
+	u32 cmd = FW_EQ_OFLD_CMD;
+	int ret;
+
+	if (unlikely(uld_type == CXGB4_TX_CRYPTO))
+		cmd = FW_EQ_CTRL_CMD;
+
+	ret = t4_sge_alloc_ofld_txq(adap, &txq->q, dev, cmd, iqid);
+	if (ret)
+		return ret;
+
+	txq->q.q_type = CXGB4_TXQ_ULD;
+	txq->adap = adap;
+	skb_queue_head_init(&txq->sendq);
+	tasklet_setup(&txq->qresume_tsk, restart_ofldq);
+	txq->full = 0;
+	txq->mapping_err = 0;
+	return 0;
+}
+
+int t4_sge_alloc_ethofld_txq(struct adapter *adap, struct sge_eohw_txq *txq,
+			     struct net_device *dev, u32 iqid)
+{
+	int ret;
+
+	ret = t4_sge_alloc_ofld_txq(adap, &txq->q, dev, FW_EQ_OFLD_CMD, iqid);
+	if (ret)
+		return ret;
+
+	txq->q.q_type = CXGB4_TXQ_ULD;
+	spin_lock_init(&txq->lock);
+	txq->adap = adap;
+	txq->tso = 0;
+	txq->uso = 0;
+	txq->tx_cso = 0;
+	txq->vlan_ins = 0;
+	txq->mapping_err = 0;
+	return 0;
+}
+
+void free_txq(struct adapter *adap, struct sge_txq *q)
+{
+	struct sge *s = &adap->sge;
+
+	dma_free_coherent(adap->pdev_dev,
+			  q->size * sizeof(struct tx_desc) + s->stat_len,
+			  q->desc, q->phys_addr);
+	q->cntxt_id = 0;
+	q->sdesc = NULL;
+	q->desc = NULL;
+}
+
+void free_rspq_fl(struct adapter *adap, struct sge_rspq *rq,
+		  struct sge_fl *fl)
+{
+	struct sge *s = &adap->sge;
+	unsigned int fl_id = fl ? fl->cntxt_id : 0xffff;
+
+	adap->sge.ingr_map[rq->cntxt_id - adap->sge.ingr_start] = NULL;
+	t4_iq_free(adap, adap->mbox, adap->pf, 0, FW_IQ_TYPE_FL_INT_CAP,
+		   rq->cntxt_id, fl_id, 0xffff);
+	dma_free_coherent(adap->pdev_dev, (rq->size + 1) * rq->iqe_len,
+			  rq->desc, rq->phys_addr);
+	netif_napi_del(&rq->napi);
+	rq->netdev = NULL;
+	rq->cntxt_id = rq->abs_id = 0;
+	rq->desc = NULL;
+
+	if (fl) {
+		free_rx_bufs(adap, fl, fl->avail);
+		dma_free_coherent(adap->pdev_dev, fl->size * 8 + s->stat_len,
+				  fl->desc, fl->addr);
+		kfree(fl->sdesc);
+		fl->sdesc = NULL;
+		fl->cntxt_id = 0;
+		fl->desc = NULL;
+	}
+}
+
+/**
+ *      t4_free_ofld_rxqs - free a block of consecutive Rx queues
+ *      @adap: the adapter
+ *      @n: number of queues
+ *      @q: pointer to first queue
+ *
+ *      Release the resources of a consecutive block of offload Rx queues.
+ */
+void t4_free_ofld_rxqs(struct adapter *adap, int n, struct sge_ofld_rxq *q)
+{
+	for ( ; n; n--, q++)
+		if (q->rspq.desc)
+			free_rspq_fl(adap, &q->rspq,
+				     q->fl.size ? &q->fl : NULL);
+}
+
+void t4_sge_free_ethofld_txq(struct adapter *adap, struct sge_eohw_txq *txq)
+{
+	if (txq->q.desc) {
+		t4_ofld_eq_free(adap, adap->mbox, adap->pf, 0,
+				txq->q.cntxt_id);
+		free_tx_desc(adap, &txq->q, txq->q.in_use, false);
+		kfree(txq->q.sdesc);
+		free_txq(adap, &txq->q);
+	}
+}
+
+/**
+ *	t4_free_sge_resources - free SGE resources
+ *	@adap: the adapter
+ *
+ *	Frees resources used by the SGE queue sets.
+ */
+void t4_free_sge_resources(struct adapter *adap)
+{
+	int i;
+	struct sge_eth_rxq *eq;
+	struct sge_eth_txq *etq;
+
+	/* stop all Rx queues in order to start them draining */
+	for (i = 0; i < adap->sge.ethqsets; i++) {
+		eq = &adap->sge.ethrxq[i];
+		if (eq->rspq.desc)
+			t4_iq_stop(adap, adap->mbox, adap->pf, 0,
+				   FW_IQ_TYPE_FL_INT_CAP,
+				   eq->rspq.cntxt_id,
+				   eq->fl.size ? eq->fl.cntxt_id : 0xffff,
+				   0xffff);
+	}
+
+	/* clean up Ethernet Tx/Rx queues */
+	for (i = 0; i < adap->sge.ethqsets; i++) {
+		eq = &adap->sge.ethrxq[i];
+		if (eq->rspq.desc)
+			free_rspq_fl(adap, &eq->rspq,
+				     eq->fl.size ? &eq->fl : NULL);
+		if (eq->msix) {
+			cxgb4_free_msix_idx_in_bmap(adap, eq->msix->idx);
+			eq->msix = NULL;
+		}
+
+		etq = &adap->sge.ethtxq[i];
+		if (etq->q.desc) {
+			t4_eth_eq_free(adap, adap->mbox, adap->pf, 0,
+				       etq->q.cntxt_id);
+			__netif_tx_lock_bh(etq->txq);
+			free_tx_desc(adap, &etq->q, etq->q.in_use, true);
+			__netif_tx_unlock_bh(etq->txq);
+			kfree(etq->q.sdesc);
+			free_txq(adap, &etq->q);
+		}
+	}
+
+	/* clean up control Tx queues */
+	for (i = 0; i < ARRAY_SIZE(adap->sge.ctrlq); i++) {
+		struct sge_ctrl_txq *cq = &adap->sge.ctrlq[i];
+
+		if (cq->q.desc) {
+			tasklet_kill(&cq->qresume_tsk);
+			t4_ctrl_eq_free(adap, adap->mbox, adap->pf, 0,
+					cq->q.cntxt_id);
+			__skb_queue_purge(&cq->sendq);
+			free_txq(adap, &cq->q);
+		}
+	}
+
+	if (adap->sge.fw_evtq.desc) {
+		free_rspq_fl(adap, &adap->sge.fw_evtq, NULL);
+		if (adap->sge.fwevtq_msix_idx >= 0)
+			cxgb4_free_msix_idx_in_bmap(adap,
+						    adap->sge.fwevtq_msix_idx);
+	}
+
+	if (adap->sge.nd_msix_idx >= 0)
+		cxgb4_free_msix_idx_in_bmap(adap, adap->sge.nd_msix_idx);
+
+	if (adap->sge.intrq.desc)
+		free_rspq_fl(adap, &adap->sge.intrq, NULL);
+
+	if (!is_t4(adap->params.chip)) {
+		etq = &adap->sge.ptptxq;
+		if (etq->q.desc) {
+			t4_eth_eq_free(adap, adap->mbox, adap->pf, 0,
+				       etq->q.cntxt_id);
+			spin_lock_bh(&adap->ptp_lock);
+			free_tx_desc(adap, &etq->q, etq->q.in_use, true);
+			spin_unlock_bh(&adap->ptp_lock);
+			kfree(etq->q.sdesc);
+			free_txq(adap, &etq->q);
+		}
+	}
+
+	/* clear the reverse egress queue map */
+	memset(adap->sge.egr_map, 0,
+	       adap->sge.egr_sz * sizeof(*adap->sge.egr_map));
+}
+
+void t4_sge_start(struct adapter *adap)
+{
+	adap->sge.ethtxq_rover = 0;
+	mod_timer(&adap->sge.rx_timer, jiffies + RX_QCHECK_PERIOD);
+	mod_timer(&adap->sge.tx_timer, jiffies + TX_QCHECK_PERIOD);
+}
+
+/**
+ *	t4_sge_stop - disable SGE operation
+ *	@adap: the adapter
+ *
+ *	Stop tasklets and timers associated with the DMA engine.  Note that
+ *	this is effective only if measures have been taken to disable any HW
+ *	events that may restart them.
+ */
+void t4_sge_stop(struct adapter *adap)
+{
+	int i;
+	struct sge *s = &adap->sge;
+
+	if (s->rx_timer.function)
+		del_timer_sync(&s->rx_timer);
+	if (s->tx_timer.function)
+		del_timer_sync(&s->tx_timer);
+
+	if (is_offload(adap)) {
+		struct sge_uld_txq_info *txq_info;
+
+		txq_info = adap->sge.uld_txq_info[CXGB4_TX_OFLD];
+		if (txq_info) {
+			struct sge_uld_txq *txq = txq_info->uldtxq;
+
+			for_each_ofldtxq(&adap->sge, i) {
+				if (txq->q.desc)
+					tasklet_kill(&txq->qresume_tsk);
+			}
+		}
+	}
+
+	if (is_pci_uld(adap)) {
+		struct sge_uld_txq_info *txq_info;
+
+		txq_info = adap->sge.uld_txq_info[CXGB4_TX_CRYPTO];
+		if (txq_info) {
+			struct sge_uld_txq *txq = txq_info->uldtxq;
+
+			for_each_ofldtxq(&adap->sge, i) {
+				if (txq->q.desc)
+					tasklet_kill(&txq->qresume_tsk);
+			}
+		}
+	}
+
+	for (i = 0; i < ARRAY_SIZE(s->ctrlq); i++) {
+		struct sge_ctrl_txq *cq = &s->ctrlq[i];
+
+		if (cq->q.desc)
+			tasklet_kill(&cq->qresume_tsk);
+	}
+}
+
+/**
+ *	t4_sge_init_soft - grab core SGE values needed by SGE code
+ *	@adap: the adapter
+ *
+ *	We need to grab the SGE operating parameters that we need to have
+ *	in order to do our job and make sure we can live with them.
+ */
+
+static int t4_sge_init_soft(struct adapter *adap)
+{
+	struct sge *s = &adap->sge;
+	u32 fl_small_pg, fl_large_pg, fl_small_mtu, fl_large_mtu;
+	u32 timer_value_0_and_1, timer_value_2_and_3, timer_value_4_and_5;
+	u32 ingress_rx_threshold;
+
+	/*
+	 * Verify that CPL messages are going to the Ingress Queue for
+	 * process_responses() and that only packet data is going to the
+	 * Free Lists.
+	 */
+	if ((t4_read_reg(adap, SGE_CONTROL_A) & RXPKTCPLMODE_F) !=
+	    RXPKTCPLMODE_V(RXPKTCPLMODE_SPLIT_X)) {
+		dev_err(adap->pdev_dev, "bad SGE CPL MODE\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * Validate the Host Buffer Register Array indices that we want to
+	 * use ...
+	 *
+	 * XXX Note that we should really read through the Host Buffer Size
+	 * XXX register array and find the indices of the Buffer Sizes which
+	 * XXX meet our needs!
+	 */
+	#define READ_FL_BUF(x) \
+		t4_read_reg(adap, SGE_FL_BUFFER_SIZE0_A+(x)*sizeof(u32))
+
+	fl_small_pg = READ_FL_BUF(RX_SMALL_PG_BUF);
+	fl_large_pg = READ_FL_BUF(RX_LARGE_PG_BUF);
+	fl_small_mtu = READ_FL_BUF(RX_SMALL_MTU_BUF);
+	fl_large_mtu = READ_FL_BUF(RX_LARGE_MTU_BUF);
+
+	/* We only bother using the Large Page logic if the Large Page Buffer
+	 * is larger than our Page Size Buffer.
+	 */
+	if (fl_large_pg <= fl_small_pg)
+		fl_large_pg = 0;
+
+	#undef READ_FL_BUF
+
+	/* The Page Size Buffer must be exactly equal to our Page Size and the
+	 * Large Page Size Buffer should be 0 (per above) or a power of 2.
+	 */
+	if (fl_small_pg != PAGE_SIZE ||
+	    (fl_large_pg & (fl_large_pg-1)) != 0) {
+		dev_err(adap->pdev_dev, "bad SGE FL page buffer sizes [%d, %d]\n",
+			fl_small_pg, fl_large_pg);
+		return -EINVAL;
+	}
+	if (fl_large_pg)
+		s->fl_pg_order = ilog2(fl_large_pg) - PAGE_SHIFT;
+
+	if (fl_small_mtu < FL_MTU_SMALL_BUFSIZE(adap) ||
+	    fl_large_mtu < FL_MTU_LARGE_BUFSIZE(adap)) {
+		dev_err(adap->pdev_dev, "bad SGE FL MTU sizes [%d, %d]\n",
+			fl_small_mtu, fl_large_mtu);
+		return -EINVAL;
+	}
+
+	/*
+	 * Retrieve our RX interrupt holdoff timer values and counter
+	 * threshold values from the SGE parameters.
+	 */
+	timer_value_0_and_1 = t4_read_reg(adap, SGE_TIMER_VALUE_0_AND_1_A);
+	timer_value_2_and_3 = t4_read_reg(adap, SGE_TIMER_VALUE_2_AND_3_A);
+	timer_value_4_and_5 = t4_read_reg(adap, SGE_TIMER_VALUE_4_AND_5_A);
+	s->timer_val[0] = core_ticks_to_us(adap,
+		TIMERVALUE0_G(timer_value_0_and_1));
+	s->timer_val[1] = core_ticks_to_us(adap,
+		TIMERVALUE1_G(timer_value_0_and_1));
+	s->timer_val[2] = core_ticks_to_us(adap,
+		TIMERVALUE2_G(timer_value_2_and_3));
+	s->timer_val[3] = core_ticks_to_us(adap,
+		TIMERVALUE3_G(timer_value_2_and_3));
+	s->timer_val[4] = core_ticks_to_us(adap,
+		TIMERVALUE4_G(timer_value_4_and_5));
+	s->timer_val[5] = core_ticks_to_us(adap,
+		TIMERVALUE5_G(timer_value_4_and_5));
+
+	ingress_rx_threshold = t4_read_reg(adap, SGE_INGRESS_RX_THRESHOLD_A);
+	s->counter_val[0] = THRESHOLD_0_G(ingress_rx_threshold);
+	s->counter_val[1] = THRESHOLD_1_G(ingress_rx_threshold);
+	s->counter_val[2] = THRESHOLD_2_G(ingress_rx_threshold);
+	s->counter_val[3] = THRESHOLD_3_G(ingress_rx_threshold);
+
+	return 0;
+}
+
+/**
+ *     t4_sge_init - initialize SGE
+ *     @adap: the adapter
+ *
+ *     Perform low-level SGE code initialization needed every time after a
+ *     chip reset.
+ */
+int t4_sge_init(struct adapter *adap)
+{
+	struct sge *s = &adap->sge;
+	u32 sge_control, sge_conm_ctrl;
+	int ret, egress_threshold;
+
+	/*
+	 * Ingress Padding Boundary and Egress Status Page Size are set up by
+	 * t4_fixup_host_params().
+	 */
+	sge_control = t4_read_reg(adap, SGE_CONTROL_A);
+	s->pktshift = PKTSHIFT_G(sge_control);
+	s->stat_len = (sge_control & EGRSTATUSPAGESIZE_F) ? 128 : 64;
+
+	s->fl_align = t4_fl_pkt_align(adap);
+	ret = t4_sge_init_soft(adap);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * A FL with <= fl_starve_thres buffers is starving and a periodic
+	 * timer will attempt to refill it.  This needs to be larger than the
+	 * SGE's Egress Congestion Threshold.  If it isn't, then we can get
+	 * stuck waiting for new packets while the SGE is waiting for us to
+	 * give it more Free List entries.  (Note that the SGE's Egress
+	 * Congestion Threshold is in units of 2 Free List pointers.) For T4,
+	 * there was only a single field to control this.  For T5 there's the
+	 * original field which now only applies to Unpacked Mode Free List
+	 * buffers and a new field which only applies to Packed Mode Free List
+	 * buffers.
+	 */
+	sge_conm_ctrl = t4_read_reg(adap, SGE_CONM_CTRL_A);
+	switch (CHELSIO_CHIP_VERSION(adap->params.chip)) {
+	case CHELSIO_T4:
+		egress_threshold = EGRTHRESHOLD_G(sge_conm_ctrl);
+		break;
+	case CHELSIO_T5:
+		egress_threshold = EGRTHRESHOLDPACKING_G(sge_conm_ctrl);
+		break;
+	case CHELSIO_T6:
+		egress_threshold = T6_EGRTHRESHOLDPACKING_G(sge_conm_ctrl);
+		break;
+	default:
+		dev_err(adap->pdev_dev, "Unsupported Chip version %d\n",
+			CHELSIO_CHIP_VERSION(adap->params.chip));
+		return -EINVAL;
+	}
+	s->fl_starve_thres = 2*egress_threshold + 1;
+
+	t4_idma_monitor_init(adap, &s->idma_monitor);
+
+	/* Set up timers used for recuring callbacks to process RX and TX
+	 * administrative tasks.
+	 */
+	timer_setup(&s->rx_timer, sge_rx_timer_cb, 0);
+	timer_setup(&s->tx_timer, sge_tx_timer_cb, 0);
+
+	spin_lock_init(&s->intrq_lock);
+
+	return 0;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/smt.c b/drivers/net/ethernet/chelsio/cxgb4/smt.c
new file mode 100644
index 000000000..e617e4aab
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/smt.c
@@ -0,0 +1,248 @@
+/*
+ * This file is part of the Chelsio T4/T5/T6 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2017 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include "cxgb4.h"
+#include "smt.h"
+#include "t4_msg.h"
+#include "t4fw_api.h"
+#include "t4_regs.h"
+#include "t4_values.h"
+
+struct smt_data *t4_init_smt(void)
+{
+	unsigned int smt_size;
+	struct smt_data *s;
+	int i;
+
+	smt_size = SMT_SIZE;
+
+	s = kvzalloc(struct_size(s, smtab, smt_size), GFP_KERNEL);
+	if (!s)
+		return NULL;
+	s->smt_size = smt_size;
+	rwlock_init(&s->lock);
+	for (i = 0; i < s->smt_size; ++i) {
+		s->smtab[i].idx = i;
+		s->smtab[i].state = SMT_STATE_UNUSED;
+		eth_zero_addr(s->smtab[i].src_mac);
+		spin_lock_init(&s->smtab[i].lock);
+		s->smtab[i].refcnt = 0;
+	}
+	return s;
+}
+
+static struct smt_entry *find_or_alloc_smte(struct smt_data *s, u8 *smac)
+{
+	struct smt_entry *first_free = NULL;
+	struct smt_entry *e, *end;
+
+	for (e = &s->smtab[0], end = &s->smtab[s->smt_size]; e != end; ++e) {
+		if (e->refcnt == 0) {
+			if (!first_free)
+				first_free = e;
+		} else {
+			if (e->state == SMT_STATE_SWITCHING) {
+				/* This entry is actually in use. See if we can
+				 * re-use it ?
+				 */
+				if (memcmp(e->src_mac, smac, ETH_ALEN) == 0)
+					goto found_reuse;
+			}
+		}
+	}
+
+	if (first_free) {
+		e = first_free;
+		goto found;
+	}
+	return NULL;
+
+found:
+	e->state = SMT_STATE_UNUSED;
+
+found_reuse:
+	return e;
+}
+
+static void t4_smte_free(struct smt_entry *e)
+{
+	if (e->refcnt == 0) {  /* hasn't been recycled */
+		e->state = SMT_STATE_UNUSED;
+	}
+}
+
+/**
+ * cxgb4_smt_release - Release SMT entry
+ * @e: smt entry to release
+ *
+ * Releases ref count and frees up an smt entry from SMT table
+ */
+void cxgb4_smt_release(struct smt_entry *e)
+{
+	spin_lock_bh(&e->lock);
+	if ((--e->refcnt) == 0)
+		t4_smte_free(e);
+	spin_unlock_bh(&e->lock);
+}
+EXPORT_SYMBOL(cxgb4_smt_release);
+
+void do_smt_write_rpl(struct adapter *adap, const struct cpl_smt_write_rpl *rpl)
+{
+	unsigned int smtidx = TID_TID_G(GET_TID(rpl));
+	struct smt_data *s = adap->smt;
+
+	if (unlikely(rpl->status != CPL_ERR_NONE)) {
+		struct smt_entry *e = &s->smtab[smtidx];
+
+		dev_err(adap->pdev_dev,
+			"Unexpected SMT_WRITE_RPL status %u for entry %u\n",
+			rpl->status, smtidx);
+		spin_lock(&e->lock);
+		e->state = SMT_STATE_ERROR;
+		spin_unlock(&e->lock);
+		return;
+	}
+}
+
+static int write_smt_entry(struct adapter *adapter, struct smt_entry *e)
+{
+	struct cpl_t6_smt_write_req *t6req;
+	struct smt_data *s = adapter->smt;
+	struct cpl_smt_write_req *req;
+	struct sk_buff *skb;
+	int size;
+	u8 row;
+
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) <= CHELSIO_T5) {
+		size = sizeof(*req);
+		skb = alloc_skb(size, GFP_ATOMIC);
+		if (!skb)
+			return -ENOMEM;
+		/* Source MAC Table (SMT) contains 256 SMAC entries
+		 * organized in 128 rows of 2 entries each.
+		 */
+		req = (struct cpl_smt_write_req *)__skb_put(skb, size);
+		INIT_TP_WR(req, 0);
+
+		/* Each row contains an SMAC pair.
+		 * LSB selects the SMAC entry within a row
+		 */
+		row = (e->idx >> 1);
+		if (e->idx & 1) {
+			req->pfvf1 = 0x0;
+			memcpy(req->src_mac1, e->src_mac, ETH_ALEN);
+
+			/* fill pfvf0/src_mac0 with entry
+			 * at prev index from smt-tab.
+			 */
+			req->pfvf0 = 0x0;
+			memcpy(req->src_mac0, s->smtab[e->idx - 1].src_mac,
+			       ETH_ALEN);
+		} else {
+			req->pfvf0 = 0x0;
+			memcpy(req->src_mac0, e->src_mac, ETH_ALEN);
+
+			/* fill pfvf1/src_mac1 with entry
+			 * at next index from smt-tab
+			 */
+			req->pfvf1 = 0x0;
+			memcpy(req->src_mac1, s->smtab[e->idx + 1].src_mac,
+			       ETH_ALEN);
+		}
+	} else {
+		size = sizeof(*t6req);
+		skb = alloc_skb(size, GFP_ATOMIC);
+		if (!skb)
+			return -ENOMEM;
+		/* Source MAC Table (SMT) contains 256 SMAC entries */
+		t6req = (struct cpl_t6_smt_write_req *)__skb_put(skb, size);
+		INIT_TP_WR(t6req, 0);
+		req = (struct cpl_smt_write_req *)t6req;
+
+		/* fill pfvf0/src_mac0 from smt-tab */
+		req->pfvf0 = 0x0;
+		memcpy(req->src_mac0, s->smtab[e->idx].src_mac, ETH_ALEN);
+		row = e->idx;
+	}
+
+	OPCODE_TID(req) =
+		htonl(MK_OPCODE_TID(CPL_SMT_WRITE_REQ, e->idx |
+				    TID_QID_V(adapter->sge.fw_evtq.abs_id)));
+	req->params = htonl(SMTW_NORPL_V(0) |
+			    SMTW_IDX_V(row) |
+			    SMTW_OVLAN_IDX_V(0));
+	t4_mgmt_tx(adapter, skb);
+	return 0;
+}
+
+static struct smt_entry *t4_smt_alloc_switching(struct adapter *adap, u16 pfvf,
+						u8 *smac)
+{
+	struct smt_data *s = adap->smt;
+	struct smt_entry *e;
+
+	write_lock_bh(&s->lock);
+	e = find_or_alloc_smte(s, smac);
+	if (e) {
+		spin_lock(&e->lock);
+		if (!e->refcnt) {
+			e->refcnt = 1;
+			e->state = SMT_STATE_SWITCHING;
+			e->pfvf = pfvf;
+			memcpy(e->src_mac, smac, ETH_ALEN);
+			write_smt_entry(adap, e);
+		} else {
+			++e->refcnt;
+		}
+		spin_unlock(&e->lock);
+	}
+	write_unlock_bh(&s->lock);
+	return e;
+}
+
+/**
+ * cxgb4_smt_alloc_switching - Allocates an SMT entry for switch filters.
+ * @dev: net_device pointer
+ * @smac: MAC address to add to SMT
+ * Returns pointer to the SMT entry created
+ *
+ * Allocates an SMT entry to be used by switching rule of a filter.
+ */
+struct smt_entry *cxgb4_smt_alloc_switching(struct net_device *dev, u8 *smac)
+{
+	struct adapter *adap = netdev2adap(dev);
+
+	return t4_smt_alloc_switching(adap, 0x0, smac);
+}
+EXPORT_SYMBOL(cxgb4_smt_alloc_switching);
diff --git a/drivers/net/ethernet/chelsio/cxgb4/smt.h b/drivers/net/ethernet/chelsio/cxgb4/smt.h
new file mode 100644
index 000000000..541249d78
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/smt.h
@@ -0,0 +1,76 @@
+/*
+ * This file is part of the Chelsio T4/T5/T6 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2017 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_SMT_H
+#define __CXGB4_SMT_H
+
+#include <linux/spinlock.h>
+#include <linux/if_ether.h>
+#include <linux/atomic.h>
+
+struct adapter;
+struct cpl_smt_write_rpl;
+
+/* SMT related handling. Heavily adapted based on l2t ops in l2t.h/l2t.c
+ */
+enum {
+	SMT_STATE_SWITCHING,
+	SMT_STATE_UNUSED,
+	SMT_STATE_ERROR
+};
+
+enum {
+	SMT_SIZE = 256
+};
+
+struct smt_entry {
+	u16 state;
+	u16 idx;
+	u16 pfvf;
+	u8 src_mac[ETH_ALEN];
+	int refcnt;
+	spinlock_t lock;	/* protect smt entry add,removal */
+};
+
+struct smt_data {
+	unsigned int smt_size;
+	rwlock_t lock;
+	struct smt_entry smtab[];
+};
+
+struct smt_data *t4_init_smt(void);
+struct smt_entry *cxgb4_smt_alloc_switching(struct net_device *dev, u8 *smac);
+void cxgb4_smt_release(struct smt_entry *e);
+void do_smt_write_rpl(struct adapter *p, const struct cpl_smt_write_rpl *rpl);
+#endif /* __CXGB4_SMT_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/srq.c b/drivers/net/ethernet/chelsio/cxgb4/srq.c
new file mode 100644
index 000000000..9a54302bb
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/srq.c
@@ -0,0 +1,137 @@
+/*
+ * This file is part of the Chelsio T6 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2017-2018 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include "cxgb4.h"
+#include "t4_msg.h"
+#include "srq.h"
+
+struct srq_data *t4_init_srq(int srq_size)
+{
+	struct srq_data *s;
+
+	s = kvzalloc(sizeof(*s), GFP_KERNEL);
+	if (!s)
+		return NULL;
+
+	s->srq_size = srq_size;
+	init_completion(&s->comp);
+	mutex_init(&s->lock);
+
+	return s;
+}
+
+/* cxgb4_get_srq_entry: read the SRQ table entry
+ * @dev: Pointer to the net_device
+ * @idx: Index to the srq
+ * @entryp: pointer to the srq entry
+ *
+ * Sends CPL_SRQ_TABLE_REQ message for the given index.
+ * Contents will be returned in CPL_SRQ_TABLE_RPL message.
+ *
+ * Returns zero if the read is successful, else a error
+ * number will be returned. Caller should not use the srq
+ * entry if the return value is non-zero.
+ *
+ *
+ */
+int cxgb4_get_srq_entry(struct net_device *dev,
+			int srq_idx, struct srq_entry *entryp)
+{
+	struct cpl_srq_table_req *req;
+	struct adapter *adap;
+	struct sk_buff *skb;
+	struct srq_data *s;
+	int rc = -ENODEV;
+
+	adap = netdev2adap(dev);
+	s = adap->srq;
+
+	if (!(adap->flags & CXGB4_FULL_INIT_DONE) || !s)
+		goto out;
+
+	skb = alloc_skb(sizeof(*req), GFP_KERNEL);
+	if (!skb)
+		return -ENOMEM;
+	req = (struct cpl_srq_table_req *)
+		__skb_put_zero(skb, sizeof(*req));
+	INIT_TP_WR(req, 0);
+	OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_SRQ_TABLE_REQ,
+					      TID_TID_V(srq_idx) |
+				TID_QID_V(adap->sge.fw_evtq.abs_id)));
+	req->idx = srq_idx;
+
+	mutex_lock(&s->lock);
+
+	s->entryp = entryp;
+	t4_mgmt_tx(adap, skb);
+
+	rc = wait_for_completion_timeout(&s->comp, SRQ_WAIT_TO);
+	if (rc)
+		rc = 0;
+	else /* !rc means we timed out */
+		rc = -ETIMEDOUT;
+
+	WARN_ON_ONCE(entryp->idx != srq_idx);
+	mutex_unlock(&s->lock);
+out:
+	return rc;
+}
+EXPORT_SYMBOL(cxgb4_get_srq_entry);
+
+void do_srq_table_rpl(struct adapter *adap,
+		      const struct cpl_srq_table_rpl *rpl)
+{
+	unsigned int idx = TID_TID_G(GET_TID(rpl));
+	struct srq_data *s = adap->srq;
+	struct srq_entry *e;
+
+	if (unlikely(rpl->status != CPL_CONTAINS_READ_RPL)) {
+		dev_err(adap->pdev_dev,
+			"Unexpected SRQ_TABLE_RPL status %u for entry %u\n",
+				rpl->status, idx);
+		goto out;
+	}
+
+	/* Store the read entry */
+	e = s->entryp;
+	e->valid = 1;
+	e->idx = idx;
+	e->pdid = SRQT_PDID_G(be64_to_cpu(rpl->rsvd_pdid));
+	e->qlen = SRQT_QLEN_G(be32_to_cpu(rpl->qlen_qbase));
+	e->qbase = SRQT_QBASE_G(be32_to_cpu(rpl->qlen_qbase));
+	e->cur_msn = be16_to_cpu(rpl->cur_msn);
+	e->max_msn = be16_to_cpu(rpl->max_msn);
+out:
+	complete(&s->comp);
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/srq.h b/drivers/net/ethernet/chelsio/cxgb4/srq.h
new file mode 100644
index 000000000..ec85cf938
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/srq.h
@@ -0,0 +1,65 @@
+/*
+ * This file is part of the Chelsio T6 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2017-2018 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_SRQ_H
+#define __CXGB4_SRQ_H
+
+struct adapter;
+struct cpl_srq_table_rpl;
+
+#define SRQ_WAIT_TO	(HZ * 5)
+
+struct srq_entry {
+	u8 valid;
+	u8 idx;
+	u8 qlen;
+	u16 pdid;
+	u16 cur_msn;
+	u16 max_msn;
+	u32 qbase;
+};
+
+struct srq_data {
+	unsigned int srq_size;
+	struct srq_entry *entryp;
+	struct completion comp;
+	struct mutex lock; /* generic mutex for srq data */
+};
+
+struct srq_data *t4_init_srq(int srq_size);
+int cxgb4_get_srq_entry(struct net_device *dev,
+			int srq_idx, struct srq_entry *entryp);
+void do_srq_table_rpl(struct adapter *adap,
+		      const struct cpl_srq_table_rpl *rpl);
+#endif  /* __CXGB4_SRQ_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_chip_type.h b/drivers/net/ethernet/chelsio/cxgb4/t4_chip_type.h
new file mode 100644
index 000000000..721c77577
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_chip_type.h
@@ -0,0 +1,87 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2015 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef __T4_CHIP_TYPE_H__
+#define __T4_CHIP_TYPE_H__
+
+#define CHELSIO_PCI_ID_VER(__DeviceID)  ((__DeviceID) >> 12)
+
+#define CHELSIO_T4		0x4
+#define CHELSIO_T5		0x5
+#define CHELSIO_T6		0x6
+
+/* We code the Chelsio T4 Family "Chip Code" as a tuple:
+ *
+ *     (Chip Version, Chip Revision)
+ *
+ * where:
+ *
+ *     Chip Version: is T4, T5, etc.
+ *     Chip Revision: is the FAB "spin" of the Chip Version.
+ */
+#define CHELSIO_CHIP_CODE(version, revision) (((version) << 4) | (revision))
+#define CHELSIO_CHIP_VERSION(code) (((code) >> 4) & 0xf)
+#define CHELSIO_CHIP_RELEASE(code) ((code) & 0xf)
+
+enum chip_type {
+	T4_A1 = CHELSIO_CHIP_CODE(CHELSIO_T4, 1),
+	T4_A2 = CHELSIO_CHIP_CODE(CHELSIO_T4, 2),
+	T4_FIRST_REV	= T4_A1,
+	T4_LAST_REV	= T4_A2,
+
+	T5_A0 = CHELSIO_CHIP_CODE(CHELSIO_T5, 0),
+	T5_A1 = CHELSIO_CHIP_CODE(CHELSIO_T5, 1),
+	T5_FIRST_REV	= T5_A0,
+	T5_LAST_REV	= T5_A1,
+
+	T6_A0 = CHELSIO_CHIP_CODE(CHELSIO_T6, 0),
+	T6_FIRST_REV	= T6_A0,
+	T6_LAST_REV	= T6_A0,
+};
+
+static inline int is_t4(enum chip_type chip)
+{
+	return (CHELSIO_CHIP_VERSION(chip) == CHELSIO_T4);
+}
+
+static inline int is_t5(enum chip_type chip)
+{
+	return (CHELSIO_CHIP_VERSION(chip) == CHELSIO_T5);
+}
+
+static inline int is_t6(enum chip_type chip)
+{
+	return (CHELSIO_CHIP_VERSION(chip) == CHELSIO_T6);
+}
+
+#endif /* __T4_CHIP_TYPE_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c
new file mode 100644
index 000000000..76de55306
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c
@@ -0,0 +1,10773 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/delay.h>
+#include "cxgb4.h"
+#include "t4_regs.h"
+#include "t4_values.h"
+#include "t4fw_api.h"
+#include "t4fw_version.h"
+
+/**
+ *	t4_wait_op_done_val - wait until an operation is completed
+ *	@adapter: the adapter performing the operation
+ *	@reg: the register to check for completion
+ *	@mask: a single-bit field within @reg that indicates completion
+ *	@polarity: the value of the field when the operation is completed
+ *	@attempts: number of check iterations
+ *	@delay: delay in usecs between iterations
+ *	@valp: where to store the value of the register at completion time
+ *
+ *	Wait until an operation is completed by checking a bit in a register
+ *	up to @attempts times.  If @valp is not NULL the value of the register
+ *	at the time it indicated completion is stored there.  Returns 0 if the
+ *	operation completes and	-EAGAIN	otherwise.
+ */
+static int t4_wait_op_done_val(struct adapter *adapter, int reg, u32 mask,
+			       int polarity, int attempts, int delay, u32 *valp)
+{
+	while (1) {
+		u32 val = t4_read_reg(adapter, reg);
+
+		if (!!(val & mask) == polarity) {
+			if (valp)
+				*valp = val;
+			return 0;
+		}
+		if (--attempts == 0)
+			return -EAGAIN;
+		if (delay)
+			udelay(delay);
+	}
+}
+
+static inline int t4_wait_op_done(struct adapter *adapter, int reg, u32 mask,
+				  int polarity, int attempts, int delay)
+{
+	return t4_wait_op_done_val(adapter, reg, mask, polarity, attempts,
+				   delay, NULL);
+}
+
+/**
+ *	t4_set_reg_field - set a register field to a value
+ *	@adapter: the adapter to program
+ *	@addr: the register address
+ *	@mask: specifies the portion of the register to modify
+ *	@val: the new value for the register field
+ *
+ *	Sets a register field specified by the supplied mask to the
+ *	given value.
+ */
+void t4_set_reg_field(struct adapter *adapter, unsigned int addr, u32 mask,
+		      u32 val)
+{
+	u32 v = t4_read_reg(adapter, addr) & ~mask;
+
+	t4_write_reg(adapter, addr, v | val);
+	(void) t4_read_reg(adapter, addr);      /* flush */
+}
+
+/**
+ *	t4_read_indirect - read indirectly addressed registers
+ *	@adap: the adapter
+ *	@addr_reg: register holding the indirect address
+ *	@data_reg: register holding the value of the indirect register
+ *	@vals: where the read register values are stored
+ *	@nregs: how many indirect registers to read
+ *	@start_idx: index of first indirect register to read
+ *
+ *	Reads registers that are accessed indirectly through an address/data
+ *	register pair.
+ */
+void t4_read_indirect(struct adapter *adap, unsigned int addr_reg,
+			     unsigned int data_reg, u32 *vals,
+			     unsigned int nregs, unsigned int start_idx)
+{
+	while (nregs--) {
+		t4_write_reg(adap, addr_reg, start_idx);
+		*vals++ = t4_read_reg(adap, data_reg);
+		start_idx++;
+	}
+}
+
+/**
+ *	t4_write_indirect - write indirectly addressed registers
+ *	@adap: the adapter
+ *	@addr_reg: register holding the indirect addresses
+ *	@data_reg: register holding the value for the indirect registers
+ *	@vals: values to write
+ *	@nregs: how many indirect registers to write
+ *	@start_idx: address of first indirect register to write
+ *
+ *	Writes a sequential block of registers that are accessed indirectly
+ *	through an address/data register pair.
+ */
+void t4_write_indirect(struct adapter *adap, unsigned int addr_reg,
+		       unsigned int data_reg, const u32 *vals,
+		       unsigned int nregs, unsigned int start_idx)
+{
+	while (nregs--) {
+		t4_write_reg(adap, addr_reg, start_idx++);
+		t4_write_reg(adap, data_reg, *vals++);
+	}
+}
+
+/*
+ * Read a 32-bit PCI Configuration Space register via the PCI-E backdoor
+ * mechanism.  This guarantees that we get the real value even if we're
+ * operating within a Virtual Machine and the Hypervisor is trapping our
+ * Configuration Space accesses.
+ */
+void t4_hw_pci_read_cfg4(struct adapter *adap, int reg, u32 *val)
+{
+	u32 req = FUNCTION_V(adap->pf) | REGISTER_V(reg);
+
+	if (CHELSIO_CHIP_VERSION(adap->params.chip) <= CHELSIO_T5)
+		req |= ENABLE_F;
+	else
+		req |= T6_ENABLE_F;
+
+	if (is_t4(adap->params.chip))
+		req |= LOCALCFG_F;
+
+	t4_write_reg(adap, PCIE_CFG_SPACE_REQ_A, req);
+	*val = t4_read_reg(adap, PCIE_CFG_SPACE_DATA_A);
+
+	/* Reset ENABLE to 0 so reads of PCIE_CFG_SPACE_DATA won't cause a
+	 * Configuration Space read.  (None of the other fields matter when
+	 * ENABLE is 0 so a simple register write is easier than a
+	 * read-modify-write via t4_set_reg_field().)
+	 */
+	t4_write_reg(adap, PCIE_CFG_SPACE_REQ_A, 0);
+}
+
+/*
+ * t4_report_fw_error - report firmware error
+ * @adap: the adapter
+ *
+ * The adapter firmware can indicate error conditions to the host.
+ * If the firmware has indicated an error, print out the reason for
+ * the firmware error.
+ */
+static void t4_report_fw_error(struct adapter *adap)
+{
+	static const char *const reason[] = {
+		"Crash",                        /* PCIE_FW_EVAL_CRASH */
+		"During Device Preparation",    /* PCIE_FW_EVAL_PREP */
+		"During Device Configuration",  /* PCIE_FW_EVAL_CONF */
+		"During Device Initialization", /* PCIE_FW_EVAL_INIT */
+		"Unexpected Event",             /* PCIE_FW_EVAL_UNEXPECTEDEVENT */
+		"Insufficient Airflow",         /* PCIE_FW_EVAL_OVERHEAT */
+		"Device Shutdown",              /* PCIE_FW_EVAL_DEVICESHUTDOWN */
+		"Reserved",                     /* reserved */
+	};
+	u32 pcie_fw;
+
+	pcie_fw = t4_read_reg(adap, PCIE_FW_A);
+	if (pcie_fw & PCIE_FW_ERR_F) {
+		dev_err(adap->pdev_dev, "Firmware reports adapter error: %s\n",
+			reason[PCIE_FW_EVAL_G(pcie_fw)]);
+		adap->flags &= ~CXGB4_FW_OK;
+	}
+}
+
+/*
+ * Get the reply to a mailbox command and store it in @rpl in big-endian order.
+ */
+static void get_mbox_rpl(struct adapter *adap, __be64 *rpl, int nflit,
+			 u32 mbox_addr)
+{
+	for ( ; nflit; nflit--, mbox_addr += 8)
+		*rpl++ = cpu_to_be64(t4_read_reg64(adap, mbox_addr));
+}
+
+/*
+ * Handle a FW assertion reported in a mailbox.
+ */
+static void fw_asrt(struct adapter *adap, u32 mbox_addr)
+{
+	struct fw_debug_cmd asrt;
+
+	get_mbox_rpl(adap, (__be64 *)&asrt, sizeof(asrt) / 8, mbox_addr);
+	dev_alert(adap->pdev_dev,
+		  "FW assertion at %.16s:%u, val0 %#x, val1 %#x\n",
+		  asrt.u.assert.filename_0_7, be32_to_cpu(asrt.u.assert.line),
+		  be32_to_cpu(asrt.u.assert.x), be32_to_cpu(asrt.u.assert.y));
+}
+
+/**
+ *	t4_record_mbox - record a Firmware Mailbox Command/Reply in the log
+ *	@adapter: the adapter
+ *	@cmd: the Firmware Mailbox Command or Reply
+ *	@size: command length in bytes
+ *	@access: the time (ms) needed to access the Firmware Mailbox
+ *	@execute: the time (ms) the command spent being executed
+ */
+static void t4_record_mbox(struct adapter *adapter,
+			   const __be64 *cmd, unsigned int size,
+			   int access, int execute)
+{
+	struct mbox_cmd_log *log = adapter->mbox_log;
+	struct mbox_cmd *entry;
+	int i;
+
+	entry = mbox_cmd_log_entry(log, log->cursor++);
+	if (log->cursor == log->size)
+		log->cursor = 0;
+
+	for (i = 0; i < size / 8; i++)
+		entry->cmd[i] = be64_to_cpu(cmd[i]);
+	while (i < MBOX_LEN / 8)
+		entry->cmd[i++] = 0;
+	entry->timestamp = jiffies;
+	entry->seqno = log->seqno++;
+	entry->access = access;
+	entry->execute = execute;
+}
+
+/**
+ *	t4_wr_mbox_meat_timeout - send a command to FW through the given mailbox
+ *	@adap: the adapter
+ *	@mbox: index of the mailbox to use
+ *	@cmd: the command to write
+ *	@size: command length in bytes
+ *	@rpl: where to optionally store the reply
+ *	@sleep_ok: if true we may sleep while awaiting command completion
+ *	@timeout: time to wait for command to finish before timing out
+ *
+ *	Sends the given command to FW through the selected mailbox and waits
+ *	for the FW to execute the command.  If @rpl is not %NULL it is used to
+ *	store the FW's reply to the command.  The command and its optional
+ *	reply are of the same length.  FW can take up to %FW_CMD_MAX_TIMEOUT ms
+ *	to respond.  @sleep_ok determines whether we may sleep while awaiting
+ *	the response.  If sleeping is allowed we use progressive backoff
+ *	otherwise we spin.
+ *
+ *	The return value is 0 on success or a negative errno on failure.  A
+ *	failure can happen either because we are not able to execute the
+ *	command or FW executes it but signals an error.  In the latter case
+ *	the return value is the error code indicated by FW (negated).
+ */
+int t4_wr_mbox_meat_timeout(struct adapter *adap, int mbox, const void *cmd,
+			    int size, void *rpl, bool sleep_ok, int timeout)
+{
+	static const int delay[] = {
+		1, 1, 3, 5, 10, 10, 20, 50, 100, 200
+	};
+
+	struct mbox_list entry;
+	u16 access = 0;
+	u16 execute = 0;
+	u32 v;
+	u64 res;
+	int i, ms, delay_idx, ret;
+	const __be64 *p = cmd;
+	u32 data_reg = PF_REG(mbox, CIM_PF_MAILBOX_DATA_A);
+	u32 ctl_reg = PF_REG(mbox, CIM_PF_MAILBOX_CTRL_A);
+	__be64 cmd_rpl[MBOX_LEN / 8];
+	u32 pcie_fw;
+
+	if ((size & 15) || size > MBOX_LEN)
+		return -EINVAL;
+
+	/*
+	 * If the device is off-line, as in EEH, commands will time out.
+	 * Fail them early so we don't waste time waiting.
+	 */
+	if (adap->pdev->error_state != pci_channel_io_normal)
+		return -EIO;
+
+	/* If we have a negative timeout, that implies that we can't sleep. */
+	if (timeout < 0) {
+		sleep_ok = false;
+		timeout = -timeout;
+	}
+
+	/* Queue ourselves onto the mailbox access list.  When our entry is at
+	 * the front of the list, we have rights to access the mailbox.  So we
+	 * wait [for a while] till we're at the front [or bail out with an
+	 * EBUSY] ...
+	 */
+	spin_lock_bh(&adap->mbox_lock);
+	list_add_tail(&entry.list, &adap->mlist.list);
+	spin_unlock_bh(&adap->mbox_lock);
+
+	delay_idx = 0;
+	ms = delay[0];
+
+	for (i = 0; ; i += ms) {
+		/* If we've waited too long, return a busy indication.  This
+		 * really ought to be based on our initial position in the
+		 * mailbox access list but this is a start.  We very rarely
+		 * contend on access to the mailbox ...
+		 */
+		pcie_fw = t4_read_reg(adap, PCIE_FW_A);
+		if (i > FW_CMD_MAX_TIMEOUT || (pcie_fw & PCIE_FW_ERR_F)) {
+			spin_lock_bh(&adap->mbox_lock);
+			list_del(&entry.list);
+			spin_unlock_bh(&adap->mbox_lock);
+			ret = (pcie_fw & PCIE_FW_ERR_F) ? -ENXIO : -EBUSY;
+			t4_record_mbox(adap, cmd, size, access, ret);
+			return ret;
+		}
+
+		/* If we're at the head, break out and start the mailbox
+		 * protocol.
+		 */
+		if (list_first_entry(&adap->mlist.list, struct mbox_list,
+				     list) == &entry)
+			break;
+
+		/* Delay for a bit before checking again ... */
+		if (sleep_ok) {
+			ms = delay[delay_idx];  /* last element may repeat */
+			if (delay_idx < ARRAY_SIZE(delay) - 1)
+				delay_idx++;
+			msleep(ms);
+		} else {
+			mdelay(ms);
+		}
+	}
+
+	/* Loop trying to get ownership of the mailbox.  Return an error
+	 * if we can't gain ownership.
+	 */
+	v = MBOWNER_G(t4_read_reg(adap, ctl_reg));
+	for (i = 0; v == MBOX_OWNER_NONE && i < 3; i++)
+		v = MBOWNER_G(t4_read_reg(adap, ctl_reg));
+	if (v != MBOX_OWNER_DRV) {
+		spin_lock_bh(&adap->mbox_lock);
+		list_del(&entry.list);
+		spin_unlock_bh(&adap->mbox_lock);
+		ret = (v == MBOX_OWNER_FW) ? -EBUSY : -ETIMEDOUT;
+		t4_record_mbox(adap, cmd, size, access, ret);
+		return ret;
+	}
+
+	/* Copy in the new mailbox command and send it on its way ... */
+	t4_record_mbox(adap, cmd, size, access, 0);
+	for (i = 0; i < size; i += 8)
+		t4_write_reg64(adap, data_reg + i, be64_to_cpu(*p++));
+
+	t4_write_reg(adap, ctl_reg, MBMSGVALID_F | MBOWNER_V(MBOX_OWNER_FW));
+	t4_read_reg(adap, ctl_reg);          /* flush write */
+
+	delay_idx = 0;
+	ms = delay[0];
+
+	for (i = 0;
+	     !((pcie_fw = t4_read_reg(adap, PCIE_FW_A)) & PCIE_FW_ERR_F) &&
+	     i < timeout;
+	     i += ms) {
+		if (sleep_ok) {
+			ms = delay[delay_idx];  /* last element may repeat */
+			if (delay_idx < ARRAY_SIZE(delay) - 1)
+				delay_idx++;
+			msleep(ms);
+		} else
+			mdelay(ms);
+
+		v = t4_read_reg(adap, ctl_reg);
+		if (MBOWNER_G(v) == MBOX_OWNER_DRV) {
+			if (!(v & MBMSGVALID_F)) {
+				t4_write_reg(adap, ctl_reg, 0);
+				continue;
+			}
+
+			get_mbox_rpl(adap, cmd_rpl, MBOX_LEN / 8, data_reg);
+			res = be64_to_cpu(cmd_rpl[0]);
+
+			if (FW_CMD_OP_G(res >> 32) == FW_DEBUG_CMD) {
+				fw_asrt(adap, data_reg);
+				res = FW_CMD_RETVAL_V(EIO);
+			} else if (rpl) {
+				memcpy(rpl, cmd_rpl, size);
+			}
+
+			t4_write_reg(adap, ctl_reg, 0);
+
+			execute = i + ms;
+			t4_record_mbox(adap, cmd_rpl,
+				       MBOX_LEN, access, execute);
+			spin_lock_bh(&adap->mbox_lock);
+			list_del(&entry.list);
+			spin_unlock_bh(&adap->mbox_lock);
+			return -FW_CMD_RETVAL_G((int)res);
+		}
+	}
+
+	ret = (pcie_fw & PCIE_FW_ERR_F) ? -ENXIO : -ETIMEDOUT;
+	t4_record_mbox(adap, cmd, size, access, ret);
+	dev_err(adap->pdev_dev, "command %#x in mailbox %d timed out\n",
+		*(const u8 *)cmd, mbox);
+	t4_report_fw_error(adap);
+	spin_lock_bh(&adap->mbox_lock);
+	list_del(&entry.list);
+	spin_unlock_bh(&adap->mbox_lock);
+	t4_fatal_err(adap);
+	return ret;
+}
+
+int t4_wr_mbox_meat(struct adapter *adap, int mbox, const void *cmd, int size,
+		    void *rpl, bool sleep_ok)
+{
+	return t4_wr_mbox_meat_timeout(adap, mbox, cmd, size, rpl, sleep_ok,
+				       FW_CMD_MAX_TIMEOUT);
+}
+
+static int t4_edc_err_read(struct adapter *adap, int idx)
+{
+	u32 edc_ecc_err_addr_reg;
+	u32 rdata_reg;
+
+	if (is_t4(adap->params.chip)) {
+		CH_WARN(adap, "%s: T4 NOT supported.\n", __func__);
+		return 0;
+	}
+	if (idx != 0 && idx != 1) {
+		CH_WARN(adap, "%s: idx %d NOT supported.\n", __func__, idx);
+		return 0;
+	}
+
+	edc_ecc_err_addr_reg = EDC_T5_REG(EDC_H_ECC_ERR_ADDR_A, idx);
+	rdata_reg = EDC_T5_REG(EDC_H_BIST_STATUS_RDATA_A, idx);
+
+	CH_WARN(adap,
+		"edc%d err addr 0x%x: 0x%x.\n",
+		idx, edc_ecc_err_addr_reg,
+		t4_read_reg(adap, edc_ecc_err_addr_reg));
+	CH_WARN(adap,
+		"bist: 0x%x, status %llx %llx %llx %llx %llx %llx %llx %llx %llx.\n",
+		rdata_reg,
+		(unsigned long long)t4_read_reg64(adap, rdata_reg),
+		(unsigned long long)t4_read_reg64(adap, rdata_reg + 8),
+		(unsigned long long)t4_read_reg64(adap, rdata_reg + 16),
+		(unsigned long long)t4_read_reg64(adap, rdata_reg + 24),
+		(unsigned long long)t4_read_reg64(adap, rdata_reg + 32),
+		(unsigned long long)t4_read_reg64(adap, rdata_reg + 40),
+		(unsigned long long)t4_read_reg64(adap, rdata_reg + 48),
+		(unsigned long long)t4_read_reg64(adap, rdata_reg + 56),
+		(unsigned long long)t4_read_reg64(adap, rdata_reg + 64));
+
+	return 0;
+}
+
+/**
+ * t4_memory_rw_init - Get memory window relative offset, base, and size.
+ * @adap: the adapter
+ * @win: PCI-E Memory Window to use
+ * @mtype: memory type: MEM_EDC0, MEM_EDC1, MEM_HMA or MEM_MC
+ * @mem_off: memory relative offset with respect to @mtype.
+ * @mem_base: configured memory base address.
+ * @mem_aperture: configured memory window aperture.
+ *
+ * Get the configured memory window's relative offset, base, and size.
+ */
+int t4_memory_rw_init(struct adapter *adap, int win, int mtype, u32 *mem_off,
+		      u32 *mem_base, u32 *mem_aperture)
+{
+	u32 edc_size, mc_size, mem_reg;
+
+	/* Offset into the region of memory which is being accessed
+	 * MEM_EDC0 = 0
+	 * MEM_EDC1 = 1
+	 * MEM_MC   = 2 -- MEM_MC for chips with only 1 memory controller
+	 * MEM_MC1  = 3 -- for chips with 2 memory controllers (e.g. T5)
+	 * MEM_HMA  = 4
+	 */
+	edc_size  = EDRAM0_SIZE_G(t4_read_reg(adap, MA_EDRAM0_BAR_A));
+	if (mtype == MEM_HMA) {
+		*mem_off = 2 * (edc_size * 1024 * 1024);
+	} else if (mtype != MEM_MC1) {
+		*mem_off = (mtype * (edc_size * 1024 * 1024));
+	} else {
+		mc_size = EXT_MEM0_SIZE_G(t4_read_reg(adap,
+						      MA_EXT_MEMORY0_BAR_A));
+		*mem_off = (MEM_MC0 * edc_size + mc_size) * 1024 * 1024;
+	}
+
+	/* Each PCI-E Memory Window is programmed with a window size -- or
+	 * "aperture" -- which controls the granularity of its mapping onto
+	 * adapter memory.  We need to grab that aperture in order to know
+	 * how to use the specified window.  The window is also programmed
+	 * with the base address of the Memory Window in BAR0's address
+	 * space.  For T4 this is an absolute PCI-E Bus Address.  For T5
+	 * the address is relative to BAR0.
+	 */
+	mem_reg = t4_read_reg(adap,
+			      PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN_A,
+						  win));
+	/* a dead adapter will return 0xffffffff for PIO reads */
+	if (mem_reg == 0xffffffff)
+		return -ENXIO;
+
+	*mem_aperture = 1 << (WINDOW_G(mem_reg) + WINDOW_SHIFT_X);
+	*mem_base = PCIEOFST_G(mem_reg) << PCIEOFST_SHIFT_X;
+	if (is_t4(adap->params.chip))
+		*mem_base -= adap->t4_bar0;
+
+	return 0;
+}
+
+/**
+ * t4_memory_update_win - Move memory window to specified address.
+ * @adap: the adapter
+ * @win: PCI-E Memory Window to use
+ * @addr: location to move.
+ *
+ * Move memory window to specified address.
+ */
+void t4_memory_update_win(struct adapter *adap, int win, u32 addr)
+{
+	t4_write_reg(adap,
+		     PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_OFFSET_A, win),
+		     addr);
+	/* Read it back to ensure that changes propagate before we
+	 * attempt to use the new value.
+	 */
+	t4_read_reg(adap,
+		    PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_OFFSET_A, win));
+}
+
+/**
+ * t4_memory_rw_residual - Read/Write residual data.
+ * @adap: the adapter
+ * @off: relative offset within residual to start read/write.
+ * @addr: address within indicated memory type.
+ * @buf: host memory buffer
+ * @dir: direction of transfer T4_MEMORY_READ (1) or T4_MEMORY_WRITE (0)
+ *
+ * Read/Write residual data less than 32-bits.
+ */
+void t4_memory_rw_residual(struct adapter *adap, u32 off, u32 addr, u8 *buf,
+			   int dir)
+{
+	union {
+		u32 word;
+		char byte[4];
+	} last;
+	unsigned char *bp;
+	int i;
+
+	if (dir == T4_MEMORY_READ) {
+		last.word = le32_to_cpu((__force __le32)
+					t4_read_reg(adap, addr));
+		for (bp = (unsigned char *)buf, i = off; i < 4; i++)
+			bp[i] = last.byte[i];
+	} else {
+		last.word = *buf;
+		for (i = off; i < 4; i++)
+			last.byte[i] = 0;
+		t4_write_reg(adap, addr,
+			     (__force u32)cpu_to_le32(last.word));
+	}
+}
+
+/**
+ *	t4_memory_rw - read/write EDC 0, EDC 1 or MC via PCIE memory window
+ *	@adap: the adapter
+ *	@win: PCI-E Memory Window to use
+ *	@mtype: memory type: MEM_EDC0, MEM_EDC1 or MEM_MC
+ *	@addr: address within indicated memory type
+ *	@len: amount of memory to transfer
+ *	@hbuf: host memory buffer
+ *	@dir: direction of transfer T4_MEMORY_READ (1) or T4_MEMORY_WRITE (0)
+ *
+ *	Reads/writes an [almost] arbitrary memory region in the firmware: the
+ *	firmware memory address and host buffer must be aligned on 32-bit
+ *	boundaries; the length may be arbitrary.  The memory is transferred as
+ *	a raw byte sequence from/to the firmware's memory.  If this memory
+ *	contains data structures which contain multi-byte integers, it's the
+ *	caller's responsibility to perform appropriate byte order conversions.
+ */
+int t4_memory_rw(struct adapter *adap, int win, int mtype, u32 addr,
+		 u32 len, void *hbuf, int dir)
+{
+	u32 pos, offset, resid, memoffset;
+	u32 win_pf, mem_aperture, mem_base;
+	u32 *buf;
+	int ret;
+
+	/* Argument sanity checks ...
+	 */
+	if (addr & 0x3 || (uintptr_t)hbuf & 0x3)
+		return -EINVAL;
+	buf = (u32 *)hbuf;
+
+	/* It's convenient to be able to handle lengths which aren't a
+	 * multiple of 32-bits because we often end up transferring files to
+	 * the firmware.  So we'll handle that by normalizing the length here
+	 * and then handling any residual transfer at the end.
+	 */
+	resid = len & 0x3;
+	len -= resid;
+
+	ret = t4_memory_rw_init(adap, win, mtype, &memoffset, &mem_base,
+				&mem_aperture);
+	if (ret)
+		return ret;
+
+	/* Determine the PCIE_MEM_ACCESS_OFFSET */
+	addr = addr + memoffset;
+
+	win_pf = is_t4(adap->params.chip) ? 0 : PFNUM_V(adap->pf);
+
+	/* Calculate our initial PCI-E Memory Window Position and Offset into
+	 * that Window.
+	 */
+	pos = addr & ~(mem_aperture - 1);
+	offset = addr - pos;
+
+	/* Set up initial PCI-E Memory Window to cover the start of our
+	 * transfer.
+	 */
+	t4_memory_update_win(adap, win, pos | win_pf);
+
+	/* Transfer data to/from the adapter as long as there's an integral
+	 * number of 32-bit transfers to complete.
+	 *
+	 * A note on Endianness issues:
+	 *
+	 * The "register" reads and writes below from/to the PCI-E Memory
+	 * Window invoke the standard adapter Big-Endian to PCI-E Link
+	 * Little-Endian "swizzel."  As a result, if we have the following
+	 * data in adapter memory:
+	 *
+	 *     Memory:  ... | b0 | b1 | b2 | b3 | ...
+	 *     Address:      i+0  i+1  i+2  i+3
+	 *
+	 * Then a read of the adapter memory via the PCI-E Memory Window
+	 * will yield:
+	 *
+	 *     x = readl(i)
+	 *         31                  0
+	 *         [ b3 | b2 | b1 | b0 ]
+	 *
+	 * If this value is stored into local memory on a Little-Endian system
+	 * it will show up correctly in local memory as:
+	 *
+	 *     ( ..., b0, b1, b2, b3, ... )
+	 *
+	 * But on a Big-Endian system, the store will show up in memory
+	 * incorrectly swizzled as:
+	 *
+	 *     ( ..., b3, b2, b1, b0, ... )
+	 *
+	 * So we need to account for this in the reads and writes to the
+	 * PCI-E Memory Window below by undoing the register read/write
+	 * swizzels.
+	 */
+	while (len > 0) {
+		if (dir == T4_MEMORY_READ)
+			*buf++ = le32_to_cpu((__force __le32)t4_read_reg(adap,
+						mem_base + offset));
+		else
+			t4_write_reg(adap, mem_base + offset,
+				     (__force u32)cpu_to_le32(*buf++));
+		offset += sizeof(__be32);
+		len -= sizeof(__be32);
+
+		/* If we've reached the end of our current window aperture,
+		 * move the PCI-E Memory Window on to the next.  Note that
+		 * doing this here after "len" may be 0 allows us to set up
+		 * the PCI-E Memory Window for a possible final residual
+		 * transfer below ...
+		 */
+		if (offset == mem_aperture) {
+			pos += mem_aperture;
+			offset = 0;
+			t4_memory_update_win(adap, win, pos | win_pf);
+		}
+	}
+
+	/* If the original transfer had a length which wasn't a multiple of
+	 * 32-bits, now's where we need to finish off the transfer of the
+	 * residual amount.  The PCI-E Memory Window has already been moved
+	 * above (if necessary) to cover this final transfer.
+	 */
+	if (resid)
+		t4_memory_rw_residual(adap, resid, mem_base + offset,
+				      (u8 *)buf, dir);
+
+	return 0;
+}
+
+/* Return the specified PCI-E Configuration Space register from our Physical
+ * Function.  We try first via a Firmware LDST Command since we prefer to let
+ * the firmware own all of these registers, but if that fails we go for it
+ * directly ourselves.
+ */
+u32 t4_read_pcie_cfg4(struct adapter *adap, int reg)
+{
+	u32 val, ldst_addrspace;
+
+	/* If fw_attach != 0, construct and send the Firmware LDST Command to
+	 * retrieve the specified PCI-E Configuration Space register.
+	 */
+	struct fw_ldst_cmd ldst_cmd;
+	int ret;
+
+	memset(&ldst_cmd, 0, sizeof(ldst_cmd));
+	ldst_addrspace = FW_LDST_CMD_ADDRSPACE_V(FW_LDST_ADDRSPC_FUNC_PCIE);
+	ldst_cmd.op_to_addrspace = cpu_to_be32(FW_CMD_OP_V(FW_LDST_CMD) |
+					       FW_CMD_REQUEST_F |
+					       FW_CMD_READ_F |
+					       ldst_addrspace);
+	ldst_cmd.cycles_to_len16 = cpu_to_be32(FW_LEN16(ldst_cmd));
+	ldst_cmd.u.pcie.select_naccess = FW_LDST_CMD_NACCESS_V(1);
+	ldst_cmd.u.pcie.ctrl_to_fn =
+		(FW_LDST_CMD_LC_F | FW_LDST_CMD_FN_V(adap->pf));
+	ldst_cmd.u.pcie.r = reg;
+
+	/* If the LDST Command succeeds, return the result, otherwise
+	 * fall through to reading it directly ourselves ...
+	 */
+	ret = t4_wr_mbox(adap, adap->mbox, &ldst_cmd, sizeof(ldst_cmd),
+			 &ldst_cmd);
+	if (ret == 0)
+		val = be32_to_cpu(ldst_cmd.u.pcie.data[0]);
+	else
+		/* Read the desired Configuration Space register via the PCI-E
+		 * Backdoor mechanism.
+		 */
+		t4_hw_pci_read_cfg4(adap, reg, &val);
+	return val;
+}
+
+/* Get the window based on base passed to it.
+ * Window aperture is currently unhandled, but there is no use case for it
+ * right now
+ */
+static u32 t4_get_window(struct adapter *adap, u32 pci_base, u64 pci_mask,
+			 u32 memwin_base)
+{
+	u32 ret;
+
+	if (is_t4(adap->params.chip)) {
+		u32 bar0;
+
+		/* Truncation intentional: we only read the bottom 32-bits of
+		 * the 64-bit BAR0/BAR1 ...  We use the hardware backdoor
+		 * mechanism to read BAR0 instead of using
+		 * pci_resource_start() because we could be operating from
+		 * within a Virtual Machine which is trapping our accesses to
+		 * our Configuration Space and we need to set up the PCI-E
+		 * Memory Window decoders with the actual addresses which will
+		 * be coming across the PCI-E link.
+		 */
+		bar0 = t4_read_pcie_cfg4(adap, pci_base);
+		bar0 &= pci_mask;
+		adap->t4_bar0 = bar0;
+
+		ret = bar0 + memwin_base;
+	} else {
+		/* For T5, only relative offset inside the PCIe BAR is passed */
+		ret = memwin_base;
+	}
+	return ret;
+}
+
+/* Get the default utility window (win0) used by everyone */
+u32 t4_get_util_window(struct adapter *adap)
+{
+	return t4_get_window(adap, PCI_BASE_ADDRESS_0,
+			     PCI_BASE_ADDRESS_MEM_MASK, MEMWIN0_BASE);
+}
+
+/* Set up memory window for accessing adapter memory ranges.  (Read
+ * back MA register to ensure that changes propagate before we attempt
+ * to use the new values.)
+ */
+void t4_setup_memwin(struct adapter *adap, u32 memwin_base, u32 window)
+{
+	t4_write_reg(adap,
+		     PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN_A, window),
+		     memwin_base | BIR_V(0) |
+		     WINDOW_V(ilog2(MEMWIN0_APERTURE) - WINDOW_SHIFT_X));
+	t4_read_reg(adap,
+		    PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN_A, window));
+}
+
+/**
+ *	t4_get_regs_len - return the size of the chips register set
+ *	@adapter: the adapter
+ *
+ *	Returns the size of the chip's BAR0 register space.
+ */
+unsigned int t4_get_regs_len(struct adapter *adapter)
+{
+	unsigned int chip_version = CHELSIO_CHIP_VERSION(adapter->params.chip);
+
+	switch (chip_version) {
+	case CHELSIO_T4:
+		return T4_REGMAP_SIZE;
+
+	case CHELSIO_T5:
+	case CHELSIO_T6:
+		return T5_REGMAP_SIZE;
+	}
+
+	dev_err(adapter->pdev_dev,
+		"Unsupported chip version %d\n", chip_version);
+	return 0;
+}
+
+/**
+ *	t4_get_regs - read chip registers into provided buffer
+ *	@adap: the adapter
+ *	@buf: register buffer
+ *	@buf_size: size (in bytes) of register buffer
+ *
+ *	If the provided register buffer isn't large enough for the chip's
+ *	full register range, the register dump will be truncated to the
+ *	register buffer's size.
+ */
+void t4_get_regs(struct adapter *adap, void *buf, size_t buf_size)
+{
+	static const unsigned int t4_reg_ranges[] = {
+		0x1008, 0x1108,
+		0x1180, 0x1184,
+		0x1190, 0x1194,
+		0x11a0, 0x11a4,
+		0x11b0, 0x11b4,
+		0x11fc, 0x123c,
+		0x1300, 0x173c,
+		0x1800, 0x18fc,
+		0x3000, 0x30d8,
+		0x30e0, 0x30e4,
+		0x30ec, 0x5910,
+		0x5920, 0x5924,
+		0x5960, 0x5960,
+		0x5968, 0x5968,
+		0x5970, 0x5970,
+		0x5978, 0x5978,
+		0x5980, 0x5980,
+		0x5988, 0x5988,
+		0x5990, 0x5990,
+		0x5998, 0x5998,
+		0x59a0, 0x59d4,
+		0x5a00, 0x5ae0,
+		0x5ae8, 0x5ae8,
+		0x5af0, 0x5af0,
+		0x5af8, 0x5af8,
+		0x6000, 0x6098,
+		0x6100, 0x6150,
+		0x6200, 0x6208,
+		0x6240, 0x6248,
+		0x6280, 0x62b0,
+		0x62c0, 0x6338,
+		0x6370, 0x638c,
+		0x6400, 0x643c,
+		0x6500, 0x6524,
+		0x6a00, 0x6a04,
+		0x6a14, 0x6a38,
+		0x6a60, 0x6a70,
+		0x6a78, 0x6a78,
+		0x6b00, 0x6b0c,
+		0x6b1c, 0x6b84,
+		0x6bf0, 0x6bf8,
+		0x6c00, 0x6c0c,
+		0x6c1c, 0x6c84,
+		0x6cf0, 0x6cf8,
+		0x6d00, 0x6d0c,
+		0x6d1c, 0x6d84,
+		0x6df0, 0x6df8,
+		0x6e00, 0x6e0c,
+		0x6e1c, 0x6e84,
+		0x6ef0, 0x6ef8,
+		0x6f00, 0x6f0c,
+		0x6f1c, 0x6f84,
+		0x6ff0, 0x6ff8,
+		0x7000, 0x700c,
+		0x701c, 0x7084,
+		0x70f0, 0x70f8,
+		0x7100, 0x710c,
+		0x711c, 0x7184,
+		0x71f0, 0x71f8,
+		0x7200, 0x720c,
+		0x721c, 0x7284,
+		0x72f0, 0x72f8,
+		0x7300, 0x730c,
+		0x731c, 0x7384,
+		0x73f0, 0x73f8,
+		0x7400, 0x7450,
+		0x7500, 0x7530,
+		0x7600, 0x760c,
+		0x7614, 0x761c,
+		0x7680, 0x76cc,
+		0x7700, 0x7798,
+		0x77c0, 0x77fc,
+		0x7900, 0x79fc,
+		0x7b00, 0x7b58,
+		0x7b60, 0x7b84,
+		0x7b8c, 0x7c38,
+		0x7d00, 0x7d38,
+		0x7d40, 0x7d80,
+		0x7d8c, 0x7ddc,
+		0x7de4, 0x7e04,
+		0x7e10, 0x7e1c,
+		0x7e24, 0x7e38,
+		0x7e40, 0x7e44,
+		0x7e4c, 0x7e78,
+		0x7e80, 0x7ea4,
+		0x7eac, 0x7edc,
+		0x7ee8, 0x7efc,
+		0x8dc0, 0x8e04,
+		0x8e10, 0x8e1c,
+		0x8e30, 0x8e78,
+		0x8ea0, 0x8eb8,
+		0x8ec0, 0x8f6c,
+		0x8fc0, 0x9008,
+		0x9010, 0x9058,
+		0x9060, 0x9060,
+		0x9068, 0x9074,
+		0x90fc, 0x90fc,
+		0x9400, 0x9408,
+		0x9410, 0x9458,
+		0x9600, 0x9600,
+		0x9608, 0x9638,
+		0x9640, 0x96bc,
+		0x9800, 0x9808,
+		0x9820, 0x983c,
+		0x9850, 0x9864,
+		0x9c00, 0x9c6c,
+		0x9c80, 0x9cec,
+		0x9d00, 0x9d6c,
+		0x9d80, 0x9dec,
+		0x9e00, 0x9e6c,
+		0x9e80, 0x9eec,
+		0x9f00, 0x9f6c,
+		0x9f80, 0x9fec,
+		0xd004, 0xd004,
+		0xd010, 0xd03c,
+		0xdfc0, 0xdfe0,
+		0xe000, 0xea7c,
+		0xf000, 0x11110,
+		0x11118, 0x11190,
+		0x19040, 0x1906c,
+		0x19078, 0x19080,
+		0x1908c, 0x190e4,
+		0x190f0, 0x190f8,
+		0x19100, 0x19110,
+		0x19120, 0x19124,
+		0x19150, 0x19194,
+		0x1919c, 0x191b0,
+		0x191d0, 0x191e8,
+		0x19238, 0x1924c,
+		0x193f8, 0x1943c,
+		0x1944c, 0x19474,
+		0x19490, 0x194e0,
+		0x194f0, 0x194f8,
+		0x19800, 0x19c08,
+		0x19c10, 0x19c90,
+		0x19ca0, 0x19ce4,
+		0x19cf0, 0x19d40,
+		0x19d50, 0x19d94,
+		0x19da0, 0x19de8,
+		0x19df0, 0x19e40,
+		0x19e50, 0x19e90,
+		0x19ea0, 0x19f4c,
+		0x1a000, 0x1a004,
+		0x1a010, 0x1a06c,
+		0x1a0b0, 0x1a0e4,
+		0x1a0ec, 0x1a0f4,
+		0x1a100, 0x1a108,
+		0x1a114, 0x1a120,
+		0x1a128, 0x1a130,
+		0x1a138, 0x1a138,
+		0x1a190, 0x1a1c4,
+		0x1a1fc, 0x1a1fc,
+		0x1e040, 0x1e04c,
+		0x1e284, 0x1e28c,
+		0x1e2c0, 0x1e2c0,
+		0x1e2e0, 0x1e2e0,
+		0x1e300, 0x1e384,
+		0x1e3c0, 0x1e3c8,
+		0x1e440, 0x1e44c,
+		0x1e684, 0x1e68c,
+		0x1e6c0, 0x1e6c0,
+		0x1e6e0, 0x1e6e0,
+		0x1e700, 0x1e784,
+		0x1e7c0, 0x1e7c8,
+		0x1e840, 0x1e84c,
+		0x1ea84, 0x1ea8c,
+		0x1eac0, 0x1eac0,
+		0x1eae0, 0x1eae0,
+		0x1eb00, 0x1eb84,
+		0x1ebc0, 0x1ebc8,
+		0x1ec40, 0x1ec4c,
+		0x1ee84, 0x1ee8c,
+		0x1eec0, 0x1eec0,
+		0x1eee0, 0x1eee0,
+		0x1ef00, 0x1ef84,
+		0x1efc0, 0x1efc8,
+		0x1f040, 0x1f04c,
+		0x1f284, 0x1f28c,
+		0x1f2c0, 0x1f2c0,
+		0x1f2e0, 0x1f2e0,
+		0x1f300, 0x1f384,
+		0x1f3c0, 0x1f3c8,
+		0x1f440, 0x1f44c,
+		0x1f684, 0x1f68c,
+		0x1f6c0, 0x1f6c0,
+		0x1f6e0, 0x1f6e0,
+		0x1f700, 0x1f784,
+		0x1f7c0, 0x1f7c8,
+		0x1f840, 0x1f84c,
+		0x1fa84, 0x1fa8c,
+		0x1fac0, 0x1fac0,
+		0x1fae0, 0x1fae0,
+		0x1fb00, 0x1fb84,
+		0x1fbc0, 0x1fbc8,
+		0x1fc40, 0x1fc4c,
+		0x1fe84, 0x1fe8c,
+		0x1fec0, 0x1fec0,
+		0x1fee0, 0x1fee0,
+		0x1ff00, 0x1ff84,
+		0x1ffc0, 0x1ffc8,
+		0x20000, 0x2002c,
+		0x20100, 0x2013c,
+		0x20190, 0x201a0,
+		0x201a8, 0x201b8,
+		0x201c4, 0x201c8,
+		0x20200, 0x20318,
+		0x20400, 0x204b4,
+		0x204c0, 0x20528,
+		0x20540, 0x20614,
+		0x21000, 0x21040,
+		0x2104c, 0x21060,
+		0x210c0, 0x210ec,
+		0x21200, 0x21268,
+		0x21270, 0x21284,
+		0x212fc, 0x21388,
+		0x21400, 0x21404,
+		0x21500, 0x21500,
+		0x21510, 0x21518,
+		0x2152c, 0x21530,
+		0x2153c, 0x2153c,
+		0x21550, 0x21554,
+		0x21600, 0x21600,
+		0x21608, 0x2161c,
+		0x21624, 0x21628,
+		0x21630, 0x21634,
+		0x2163c, 0x2163c,
+		0x21700, 0x2171c,
+		0x21780, 0x2178c,
+		0x21800, 0x21818,
+		0x21820, 0x21828,
+		0x21830, 0x21848,
+		0x21850, 0x21854,
+		0x21860, 0x21868,
+		0x21870, 0x21870,
+		0x21878, 0x21898,
+		0x218a0, 0x218a8,
+		0x218b0, 0x218c8,
+		0x218d0, 0x218d4,
+		0x218e0, 0x218e8,
+		0x218f0, 0x218f0,
+		0x218f8, 0x21a18,
+		0x21a20, 0x21a28,
+		0x21a30, 0x21a48,
+		0x21a50, 0x21a54,
+		0x21a60, 0x21a68,
+		0x21a70, 0x21a70,
+		0x21a78, 0x21a98,
+		0x21aa0, 0x21aa8,
+		0x21ab0, 0x21ac8,
+		0x21ad0, 0x21ad4,
+		0x21ae0, 0x21ae8,
+		0x21af0, 0x21af0,
+		0x21af8, 0x21c18,
+		0x21c20, 0x21c20,
+		0x21c28, 0x21c30,
+		0x21c38, 0x21c38,
+		0x21c80, 0x21c98,
+		0x21ca0, 0x21ca8,
+		0x21cb0, 0x21cc8,
+		0x21cd0, 0x21cd4,
+		0x21ce0, 0x21ce8,
+		0x21cf0, 0x21cf0,
+		0x21cf8, 0x21d7c,
+		0x21e00, 0x21e04,
+		0x22000, 0x2202c,
+		0x22100, 0x2213c,
+		0x22190, 0x221a0,
+		0x221a8, 0x221b8,
+		0x221c4, 0x221c8,
+		0x22200, 0x22318,
+		0x22400, 0x224b4,
+		0x224c0, 0x22528,
+		0x22540, 0x22614,
+		0x23000, 0x23040,
+		0x2304c, 0x23060,
+		0x230c0, 0x230ec,
+		0x23200, 0x23268,
+		0x23270, 0x23284,
+		0x232fc, 0x23388,
+		0x23400, 0x23404,
+		0x23500, 0x23500,
+		0x23510, 0x23518,
+		0x2352c, 0x23530,
+		0x2353c, 0x2353c,
+		0x23550, 0x23554,
+		0x23600, 0x23600,
+		0x23608, 0x2361c,
+		0x23624, 0x23628,
+		0x23630, 0x23634,
+		0x2363c, 0x2363c,
+		0x23700, 0x2371c,
+		0x23780, 0x2378c,
+		0x23800, 0x23818,
+		0x23820, 0x23828,
+		0x23830, 0x23848,
+		0x23850, 0x23854,
+		0x23860, 0x23868,
+		0x23870, 0x23870,
+		0x23878, 0x23898,
+		0x238a0, 0x238a8,
+		0x238b0, 0x238c8,
+		0x238d0, 0x238d4,
+		0x238e0, 0x238e8,
+		0x238f0, 0x238f0,
+		0x238f8, 0x23a18,
+		0x23a20, 0x23a28,
+		0x23a30, 0x23a48,
+		0x23a50, 0x23a54,
+		0x23a60, 0x23a68,
+		0x23a70, 0x23a70,
+		0x23a78, 0x23a98,
+		0x23aa0, 0x23aa8,
+		0x23ab0, 0x23ac8,
+		0x23ad0, 0x23ad4,
+		0x23ae0, 0x23ae8,
+		0x23af0, 0x23af0,
+		0x23af8, 0x23c18,
+		0x23c20, 0x23c20,
+		0x23c28, 0x23c30,
+		0x23c38, 0x23c38,
+		0x23c80, 0x23c98,
+		0x23ca0, 0x23ca8,
+		0x23cb0, 0x23cc8,
+		0x23cd0, 0x23cd4,
+		0x23ce0, 0x23ce8,
+		0x23cf0, 0x23cf0,
+		0x23cf8, 0x23d7c,
+		0x23e00, 0x23e04,
+		0x24000, 0x2402c,
+		0x24100, 0x2413c,
+		0x24190, 0x241a0,
+		0x241a8, 0x241b8,
+		0x241c4, 0x241c8,
+		0x24200, 0x24318,
+		0x24400, 0x244b4,
+		0x244c0, 0x24528,
+		0x24540, 0x24614,
+		0x25000, 0x25040,
+		0x2504c, 0x25060,
+		0x250c0, 0x250ec,
+		0x25200, 0x25268,
+		0x25270, 0x25284,
+		0x252fc, 0x25388,
+		0x25400, 0x25404,
+		0x25500, 0x25500,
+		0x25510, 0x25518,
+		0x2552c, 0x25530,
+		0x2553c, 0x2553c,
+		0x25550, 0x25554,
+		0x25600, 0x25600,
+		0x25608, 0x2561c,
+		0x25624, 0x25628,
+		0x25630, 0x25634,
+		0x2563c, 0x2563c,
+		0x25700, 0x2571c,
+		0x25780, 0x2578c,
+		0x25800, 0x25818,
+		0x25820, 0x25828,
+		0x25830, 0x25848,
+		0x25850, 0x25854,
+		0x25860, 0x25868,
+		0x25870, 0x25870,
+		0x25878, 0x25898,
+		0x258a0, 0x258a8,
+		0x258b0, 0x258c8,
+		0x258d0, 0x258d4,
+		0x258e0, 0x258e8,
+		0x258f0, 0x258f0,
+		0x258f8, 0x25a18,
+		0x25a20, 0x25a28,
+		0x25a30, 0x25a48,
+		0x25a50, 0x25a54,
+		0x25a60, 0x25a68,
+		0x25a70, 0x25a70,
+		0x25a78, 0x25a98,
+		0x25aa0, 0x25aa8,
+		0x25ab0, 0x25ac8,
+		0x25ad0, 0x25ad4,
+		0x25ae0, 0x25ae8,
+		0x25af0, 0x25af0,
+		0x25af8, 0x25c18,
+		0x25c20, 0x25c20,
+		0x25c28, 0x25c30,
+		0x25c38, 0x25c38,
+		0x25c80, 0x25c98,
+		0x25ca0, 0x25ca8,
+		0x25cb0, 0x25cc8,
+		0x25cd0, 0x25cd4,
+		0x25ce0, 0x25ce8,
+		0x25cf0, 0x25cf0,
+		0x25cf8, 0x25d7c,
+		0x25e00, 0x25e04,
+		0x26000, 0x2602c,
+		0x26100, 0x2613c,
+		0x26190, 0x261a0,
+		0x261a8, 0x261b8,
+		0x261c4, 0x261c8,
+		0x26200, 0x26318,
+		0x26400, 0x264b4,
+		0x264c0, 0x26528,
+		0x26540, 0x26614,
+		0x27000, 0x27040,
+		0x2704c, 0x27060,
+		0x270c0, 0x270ec,
+		0x27200, 0x27268,
+		0x27270, 0x27284,
+		0x272fc, 0x27388,
+		0x27400, 0x27404,
+		0x27500, 0x27500,
+		0x27510, 0x27518,
+		0x2752c, 0x27530,
+		0x2753c, 0x2753c,
+		0x27550, 0x27554,
+		0x27600, 0x27600,
+		0x27608, 0x2761c,
+		0x27624, 0x27628,
+		0x27630, 0x27634,
+		0x2763c, 0x2763c,
+		0x27700, 0x2771c,
+		0x27780, 0x2778c,
+		0x27800, 0x27818,
+		0x27820, 0x27828,
+		0x27830, 0x27848,
+		0x27850, 0x27854,
+		0x27860, 0x27868,
+		0x27870, 0x27870,
+		0x27878, 0x27898,
+		0x278a0, 0x278a8,
+		0x278b0, 0x278c8,
+		0x278d0, 0x278d4,
+		0x278e0, 0x278e8,
+		0x278f0, 0x278f0,
+		0x278f8, 0x27a18,
+		0x27a20, 0x27a28,
+		0x27a30, 0x27a48,
+		0x27a50, 0x27a54,
+		0x27a60, 0x27a68,
+		0x27a70, 0x27a70,
+		0x27a78, 0x27a98,
+		0x27aa0, 0x27aa8,
+		0x27ab0, 0x27ac8,
+		0x27ad0, 0x27ad4,
+		0x27ae0, 0x27ae8,
+		0x27af0, 0x27af0,
+		0x27af8, 0x27c18,
+		0x27c20, 0x27c20,
+		0x27c28, 0x27c30,
+		0x27c38, 0x27c38,
+		0x27c80, 0x27c98,
+		0x27ca0, 0x27ca8,
+		0x27cb0, 0x27cc8,
+		0x27cd0, 0x27cd4,
+		0x27ce0, 0x27ce8,
+		0x27cf0, 0x27cf0,
+		0x27cf8, 0x27d7c,
+		0x27e00, 0x27e04,
+	};
+
+	static const unsigned int t5_reg_ranges[] = {
+		0x1008, 0x10c0,
+		0x10cc, 0x10f8,
+		0x1100, 0x1100,
+		0x110c, 0x1148,
+		0x1180, 0x1184,
+		0x1190, 0x1194,
+		0x11a0, 0x11a4,
+		0x11b0, 0x11b4,
+		0x11fc, 0x123c,
+		0x1280, 0x173c,
+		0x1800, 0x18fc,
+		0x3000, 0x3028,
+		0x3060, 0x30b0,
+		0x30b8, 0x30d8,
+		0x30e0, 0x30fc,
+		0x3140, 0x357c,
+		0x35a8, 0x35cc,
+		0x35ec, 0x35ec,
+		0x3600, 0x5624,
+		0x56cc, 0x56ec,
+		0x56f4, 0x5720,
+		0x5728, 0x575c,
+		0x580c, 0x5814,
+		0x5890, 0x589c,
+		0x58a4, 0x58ac,
+		0x58b8, 0x58bc,
+		0x5940, 0x59c8,
+		0x59d0, 0x59dc,
+		0x59fc, 0x5a18,
+		0x5a60, 0x5a70,
+		0x5a80, 0x5a9c,
+		0x5b94, 0x5bfc,
+		0x6000, 0x6020,
+		0x6028, 0x6040,
+		0x6058, 0x609c,
+		0x60a8, 0x614c,
+		0x7700, 0x7798,
+		0x77c0, 0x78fc,
+		0x7b00, 0x7b58,
+		0x7b60, 0x7b84,
+		0x7b8c, 0x7c54,
+		0x7d00, 0x7d38,
+		0x7d40, 0x7d80,
+		0x7d8c, 0x7ddc,
+		0x7de4, 0x7e04,
+		0x7e10, 0x7e1c,
+		0x7e24, 0x7e38,
+		0x7e40, 0x7e44,
+		0x7e4c, 0x7e78,
+		0x7e80, 0x7edc,
+		0x7ee8, 0x7efc,
+		0x8dc0, 0x8de0,
+		0x8df8, 0x8e04,
+		0x8e10, 0x8e84,
+		0x8ea0, 0x8f84,
+		0x8fc0, 0x9058,
+		0x9060, 0x9060,
+		0x9068, 0x90f8,
+		0x9400, 0x9408,
+		0x9410, 0x9470,
+		0x9600, 0x9600,
+		0x9608, 0x9638,
+		0x9640, 0x96f4,
+		0x9800, 0x9808,
+		0x9810, 0x9864,
+		0x9c00, 0x9c6c,
+		0x9c80, 0x9cec,
+		0x9d00, 0x9d6c,
+		0x9d80, 0x9dec,
+		0x9e00, 0x9e6c,
+		0x9e80, 0x9eec,
+		0x9f00, 0x9f6c,
+		0x9f80, 0xa020,
+		0xd000, 0xd004,
+		0xd010, 0xd03c,
+		0xdfc0, 0xdfe0,
+		0xe000, 0x1106c,
+		0x11074, 0x11088,
+		0x1109c, 0x1117c,
+		0x11190, 0x11204,
+		0x19040, 0x1906c,
+		0x19078, 0x19080,
+		0x1908c, 0x190e8,
+		0x190f0, 0x190f8,
+		0x19100, 0x19110,
+		0x19120, 0x19124,
+		0x19150, 0x19194,
+		0x1919c, 0x191b0,
+		0x191d0, 0x191e8,
+		0x19238, 0x19290,
+		0x193f8, 0x19428,
+		0x19430, 0x19444,
+		0x1944c, 0x1946c,
+		0x19474, 0x19474,
+		0x19490, 0x194cc,
+		0x194f0, 0x194f8,
+		0x19c00, 0x19c08,
+		0x19c10, 0x19c60,
+		0x19c94, 0x19ce4,
+		0x19cf0, 0x19d40,
+		0x19d50, 0x19d94,
+		0x19da0, 0x19de8,
+		0x19df0, 0x19e10,
+		0x19e50, 0x19e90,
+		0x19ea0, 0x19f24,
+		0x19f34, 0x19f34,
+		0x19f40, 0x19f50,
+		0x19f90, 0x19fb4,
+		0x19fc4, 0x19fe4,
+		0x1a000, 0x1a004,
+		0x1a010, 0x1a06c,
+		0x1a0b0, 0x1a0e4,
+		0x1a0ec, 0x1a0f8,
+		0x1a100, 0x1a108,
+		0x1a114, 0x1a130,
+		0x1a138, 0x1a1c4,
+		0x1a1fc, 0x1a1fc,
+		0x1e008, 0x1e00c,
+		0x1e040, 0x1e044,
+		0x1e04c, 0x1e04c,
+		0x1e284, 0x1e290,
+		0x1e2c0, 0x1e2c0,
+		0x1e2e0, 0x1e2e0,
+		0x1e300, 0x1e384,
+		0x1e3c0, 0x1e3c8,
+		0x1e408, 0x1e40c,
+		0x1e440, 0x1e444,
+		0x1e44c, 0x1e44c,
+		0x1e684, 0x1e690,
+		0x1e6c0, 0x1e6c0,
+		0x1e6e0, 0x1e6e0,
+		0x1e700, 0x1e784,
+		0x1e7c0, 0x1e7c8,
+		0x1e808, 0x1e80c,
+		0x1e840, 0x1e844,
+		0x1e84c, 0x1e84c,
+		0x1ea84, 0x1ea90,
+		0x1eac0, 0x1eac0,
+		0x1eae0, 0x1eae0,
+		0x1eb00, 0x1eb84,
+		0x1ebc0, 0x1ebc8,
+		0x1ec08, 0x1ec0c,
+		0x1ec40, 0x1ec44,
+		0x1ec4c, 0x1ec4c,
+		0x1ee84, 0x1ee90,
+		0x1eec0, 0x1eec0,
+		0x1eee0, 0x1eee0,
+		0x1ef00, 0x1ef84,
+		0x1efc0, 0x1efc8,
+		0x1f008, 0x1f00c,
+		0x1f040, 0x1f044,
+		0x1f04c, 0x1f04c,
+		0x1f284, 0x1f290,
+		0x1f2c0, 0x1f2c0,
+		0x1f2e0, 0x1f2e0,
+		0x1f300, 0x1f384,
+		0x1f3c0, 0x1f3c8,
+		0x1f408, 0x1f40c,
+		0x1f440, 0x1f444,
+		0x1f44c, 0x1f44c,
+		0x1f684, 0x1f690,
+		0x1f6c0, 0x1f6c0,
+		0x1f6e0, 0x1f6e0,
+		0x1f700, 0x1f784,
+		0x1f7c0, 0x1f7c8,
+		0x1f808, 0x1f80c,
+		0x1f840, 0x1f844,
+		0x1f84c, 0x1f84c,
+		0x1fa84, 0x1fa90,
+		0x1fac0, 0x1fac0,
+		0x1fae0, 0x1fae0,
+		0x1fb00, 0x1fb84,
+		0x1fbc0, 0x1fbc8,
+		0x1fc08, 0x1fc0c,
+		0x1fc40, 0x1fc44,
+		0x1fc4c, 0x1fc4c,
+		0x1fe84, 0x1fe90,
+		0x1fec0, 0x1fec0,
+		0x1fee0, 0x1fee0,
+		0x1ff00, 0x1ff84,
+		0x1ffc0, 0x1ffc8,
+		0x30000, 0x30030,
+		0x30100, 0x30144,
+		0x30190, 0x301a0,
+		0x301a8, 0x301b8,
+		0x301c4, 0x301c8,
+		0x301d0, 0x301d0,
+		0x30200, 0x30318,
+		0x30400, 0x304b4,
+		0x304c0, 0x3052c,
+		0x30540, 0x3061c,
+		0x30800, 0x30828,
+		0x30834, 0x30834,
+		0x308c0, 0x30908,
+		0x30910, 0x309ac,
+		0x30a00, 0x30a14,
+		0x30a1c, 0x30a2c,
+		0x30a44, 0x30a50,
+		0x30a74, 0x30a74,
+		0x30a7c, 0x30afc,
+		0x30b08, 0x30c24,
+		0x30d00, 0x30d00,
+		0x30d08, 0x30d14,
+		0x30d1c, 0x30d20,
+		0x30d3c, 0x30d3c,
+		0x30d48, 0x30d50,
+		0x31200, 0x3120c,
+		0x31220, 0x31220,
+		0x31240, 0x31240,
+		0x31600, 0x3160c,
+		0x31a00, 0x31a1c,
+		0x31e00, 0x31e20,
+		0x31e38, 0x31e3c,
+		0x31e80, 0x31e80,
+		0x31e88, 0x31ea8,
+		0x31eb0, 0x31eb4,
+		0x31ec8, 0x31ed4,
+		0x31fb8, 0x32004,
+		0x32200, 0x32200,
+		0x32208, 0x32240,
+		0x32248, 0x32280,
+		0x32288, 0x322c0,
+		0x322c8, 0x322fc,
+		0x32600, 0x32630,
+		0x32a00, 0x32abc,
+		0x32b00, 0x32b10,
+		0x32b20, 0x32b30,
+		0x32b40, 0x32b50,
+		0x32b60, 0x32b70,
+		0x33000, 0x33028,
+		0x33030, 0x33048,
+		0x33060, 0x33068,
+		0x33070, 0x3309c,
+		0x330f0, 0x33128,
+		0x33130, 0x33148,
+		0x33160, 0x33168,
+		0x33170, 0x3319c,
+		0x331f0, 0x33238,
+		0x33240, 0x33240,
+		0x33248, 0x33250,
+		0x3325c, 0x33264,
+		0x33270, 0x332b8,
+		0x332c0, 0x332e4,
+		0x332f8, 0x33338,
+		0x33340, 0x33340,
+		0x33348, 0x33350,
+		0x3335c, 0x33364,
+		0x33370, 0x333b8,
+		0x333c0, 0x333e4,
+		0x333f8, 0x33428,
+		0x33430, 0x33448,
+		0x33460, 0x33468,
+		0x33470, 0x3349c,
+		0x334f0, 0x33528,
+		0x33530, 0x33548,
+		0x33560, 0x33568,
+		0x33570, 0x3359c,
+		0x335f0, 0x33638,
+		0x33640, 0x33640,
+		0x33648, 0x33650,
+		0x3365c, 0x33664,
+		0x33670, 0x336b8,
+		0x336c0, 0x336e4,
+		0x336f8, 0x33738,
+		0x33740, 0x33740,
+		0x33748, 0x33750,
+		0x3375c, 0x33764,
+		0x33770, 0x337b8,
+		0x337c0, 0x337e4,
+		0x337f8, 0x337fc,
+		0x33814, 0x33814,
+		0x3382c, 0x3382c,
+		0x33880, 0x3388c,
+		0x338e8, 0x338ec,
+		0x33900, 0x33928,
+		0x33930, 0x33948,
+		0x33960, 0x33968,
+		0x33970, 0x3399c,
+		0x339f0, 0x33a38,
+		0x33a40, 0x33a40,
+		0x33a48, 0x33a50,
+		0x33a5c, 0x33a64,
+		0x33a70, 0x33ab8,
+		0x33ac0, 0x33ae4,
+		0x33af8, 0x33b10,
+		0x33b28, 0x33b28,
+		0x33b3c, 0x33b50,
+		0x33bf0, 0x33c10,
+		0x33c28, 0x33c28,
+		0x33c3c, 0x33c50,
+		0x33cf0, 0x33cfc,
+		0x34000, 0x34030,
+		0x34100, 0x34144,
+		0x34190, 0x341a0,
+		0x341a8, 0x341b8,
+		0x341c4, 0x341c8,
+		0x341d0, 0x341d0,
+		0x34200, 0x34318,
+		0x34400, 0x344b4,
+		0x344c0, 0x3452c,
+		0x34540, 0x3461c,
+		0x34800, 0x34828,
+		0x34834, 0x34834,
+		0x348c0, 0x34908,
+		0x34910, 0x349ac,
+		0x34a00, 0x34a14,
+		0x34a1c, 0x34a2c,
+		0x34a44, 0x34a50,
+		0x34a74, 0x34a74,
+		0x34a7c, 0x34afc,
+		0x34b08, 0x34c24,
+		0x34d00, 0x34d00,
+		0x34d08, 0x34d14,
+		0x34d1c, 0x34d20,
+		0x34d3c, 0x34d3c,
+		0x34d48, 0x34d50,
+		0x35200, 0x3520c,
+		0x35220, 0x35220,
+		0x35240, 0x35240,
+		0x35600, 0x3560c,
+		0x35a00, 0x35a1c,
+		0x35e00, 0x35e20,
+		0x35e38, 0x35e3c,
+		0x35e80, 0x35e80,
+		0x35e88, 0x35ea8,
+		0x35eb0, 0x35eb4,
+		0x35ec8, 0x35ed4,
+		0x35fb8, 0x36004,
+		0x36200, 0x36200,
+		0x36208, 0x36240,
+		0x36248, 0x36280,
+		0x36288, 0x362c0,
+		0x362c8, 0x362fc,
+		0x36600, 0x36630,
+		0x36a00, 0x36abc,
+		0x36b00, 0x36b10,
+		0x36b20, 0x36b30,
+		0x36b40, 0x36b50,
+		0x36b60, 0x36b70,
+		0x37000, 0x37028,
+		0x37030, 0x37048,
+		0x37060, 0x37068,
+		0x37070, 0x3709c,
+		0x370f0, 0x37128,
+		0x37130, 0x37148,
+		0x37160, 0x37168,
+		0x37170, 0x3719c,
+		0x371f0, 0x37238,
+		0x37240, 0x37240,
+		0x37248, 0x37250,
+		0x3725c, 0x37264,
+		0x37270, 0x372b8,
+		0x372c0, 0x372e4,
+		0x372f8, 0x37338,
+		0x37340, 0x37340,
+		0x37348, 0x37350,
+		0x3735c, 0x37364,
+		0x37370, 0x373b8,
+		0x373c0, 0x373e4,
+		0x373f8, 0x37428,
+		0x37430, 0x37448,
+		0x37460, 0x37468,
+		0x37470, 0x3749c,
+		0x374f0, 0x37528,
+		0x37530, 0x37548,
+		0x37560, 0x37568,
+		0x37570, 0x3759c,
+		0x375f0, 0x37638,
+		0x37640, 0x37640,
+		0x37648, 0x37650,
+		0x3765c, 0x37664,
+		0x37670, 0x376b8,
+		0x376c0, 0x376e4,
+		0x376f8, 0x37738,
+		0x37740, 0x37740,
+		0x37748, 0x37750,
+		0x3775c, 0x37764,
+		0x37770, 0x377b8,
+		0x377c0, 0x377e4,
+		0x377f8, 0x377fc,
+		0x37814, 0x37814,
+		0x3782c, 0x3782c,
+		0x37880, 0x3788c,
+		0x378e8, 0x378ec,
+		0x37900, 0x37928,
+		0x37930, 0x37948,
+		0x37960, 0x37968,
+		0x37970, 0x3799c,
+		0x379f0, 0x37a38,
+		0x37a40, 0x37a40,
+		0x37a48, 0x37a50,
+		0x37a5c, 0x37a64,
+		0x37a70, 0x37ab8,
+		0x37ac0, 0x37ae4,
+		0x37af8, 0x37b10,
+		0x37b28, 0x37b28,
+		0x37b3c, 0x37b50,
+		0x37bf0, 0x37c10,
+		0x37c28, 0x37c28,
+		0x37c3c, 0x37c50,
+		0x37cf0, 0x37cfc,
+		0x38000, 0x38030,
+		0x38100, 0x38144,
+		0x38190, 0x381a0,
+		0x381a8, 0x381b8,
+		0x381c4, 0x381c8,
+		0x381d0, 0x381d0,
+		0x38200, 0x38318,
+		0x38400, 0x384b4,
+		0x384c0, 0x3852c,
+		0x38540, 0x3861c,
+		0x38800, 0x38828,
+		0x38834, 0x38834,
+		0x388c0, 0x38908,
+		0x38910, 0x389ac,
+		0x38a00, 0x38a14,
+		0x38a1c, 0x38a2c,
+		0x38a44, 0x38a50,
+		0x38a74, 0x38a74,
+		0x38a7c, 0x38afc,
+		0x38b08, 0x38c24,
+		0x38d00, 0x38d00,
+		0x38d08, 0x38d14,
+		0x38d1c, 0x38d20,
+		0x38d3c, 0x38d3c,
+		0x38d48, 0x38d50,
+		0x39200, 0x3920c,
+		0x39220, 0x39220,
+		0x39240, 0x39240,
+		0x39600, 0x3960c,
+		0x39a00, 0x39a1c,
+		0x39e00, 0x39e20,
+		0x39e38, 0x39e3c,
+		0x39e80, 0x39e80,
+		0x39e88, 0x39ea8,
+		0x39eb0, 0x39eb4,
+		0x39ec8, 0x39ed4,
+		0x39fb8, 0x3a004,
+		0x3a200, 0x3a200,
+		0x3a208, 0x3a240,
+		0x3a248, 0x3a280,
+		0x3a288, 0x3a2c0,
+		0x3a2c8, 0x3a2fc,
+		0x3a600, 0x3a630,
+		0x3aa00, 0x3aabc,
+		0x3ab00, 0x3ab10,
+		0x3ab20, 0x3ab30,
+		0x3ab40, 0x3ab50,
+		0x3ab60, 0x3ab70,
+		0x3b000, 0x3b028,
+		0x3b030, 0x3b048,
+		0x3b060, 0x3b068,
+		0x3b070, 0x3b09c,
+		0x3b0f0, 0x3b128,
+		0x3b130, 0x3b148,
+		0x3b160, 0x3b168,
+		0x3b170, 0x3b19c,
+		0x3b1f0, 0x3b238,
+		0x3b240, 0x3b240,
+		0x3b248, 0x3b250,
+		0x3b25c, 0x3b264,
+		0x3b270, 0x3b2b8,
+		0x3b2c0, 0x3b2e4,
+		0x3b2f8, 0x3b338,
+		0x3b340, 0x3b340,
+		0x3b348, 0x3b350,
+		0x3b35c, 0x3b364,
+		0x3b370, 0x3b3b8,
+		0x3b3c0, 0x3b3e4,
+		0x3b3f8, 0x3b428,
+		0x3b430, 0x3b448,
+		0x3b460, 0x3b468,
+		0x3b470, 0x3b49c,
+		0x3b4f0, 0x3b528,
+		0x3b530, 0x3b548,
+		0x3b560, 0x3b568,
+		0x3b570, 0x3b59c,
+		0x3b5f0, 0x3b638,
+		0x3b640, 0x3b640,
+		0x3b648, 0x3b650,
+		0x3b65c, 0x3b664,
+		0x3b670, 0x3b6b8,
+		0x3b6c0, 0x3b6e4,
+		0x3b6f8, 0x3b738,
+		0x3b740, 0x3b740,
+		0x3b748, 0x3b750,
+		0x3b75c, 0x3b764,
+		0x3b770, 0x3b7b8,
+		0x3b7c0, 0x3b7e4,
+		0x3b7f8, 0x3b7fc,
+		0x3b814, 0x3b814,
+		0x3b82c, 0x3b82c,
+		0x3b880, 0x3b88c,
+		0x3b8e8, 0x3b8ec,
+		0x3b900, 0x3b928,
+		0x3b930, 0x3b948,
+		0x3b960, 0x3b968,
+		0x3b970, 0x3b99c,
+		0x3b9f0, 0x3ba38,
+		0x3ba40, 0x3ba40,
+		0x3ba48, 0x3ba50,
+		0x3ba5c, 0x3ba64,
+		0x3ba70, 0x3bab8,
+		0x3bac0, 0x3bae4,
+		0x3baf8, 0x3bb10,
+		0x3bb28, 0x3bb28,
+		0x3bb3c, 0x3bb50,
+		0x3bbf0, 0x3bc10,
+		0x3bc28, 0x3bc28,
+		0x3bc3c, 0x3bc50,
+		0x3bcf0, 0x3bcfc,
+		0x3c000, 0x3c030,
+		0x3c100, 0x3c144,
+		0x3c190, 0x3c1a0,
+		0x3c1a8, 0x3c1b8,
+		0x3c1c4, 0x3c1c8,
+		0x3c1d0, 0x3c1d0,
+		0x3c200, 0x3c318,
+		0x3c400, 0x3c4b4,
+		0x3c4c0, 0x3c52c,
+		0x3c540, 0x3c61c,
+		0x3c800, 0x3c828,
+		0x3c834, 0x3c834,
+		0x3c8c0, 0x3c908,
+		0x3c910, 0x3c9ac,
+		0x3ca00, 0x3ca14,
+		0x3ca1c, 0x3ca2c,
+		0x3ca44, 0x3ca50,
+		0x3ca74, 0x3ca74,
+		0x3ca7c, 0x3cafc,
+		0x3cb08, 0x3cc24,
+		0x3cd00, 0x3cd00,
+		0x3cd08, 0x3cd14,
+		0x3cd1c, 0x3cd20,
+		0x3cd3c, 0x3cd3c,
+		0x3cd48, 0x3cd50,
+		0x3d200, 0x3d20c,
+		0x3d220, 0x3d220,
+		0x3d240, 0x3d240,
+		0x3d600, 0x3d60c,
+		0x3da00, 0x3da1c,
+		0x3de00, 0x3de20,
+		0x3de38, 0x3de3c,
+		0x3de80, 0x3de80,
+		0x3de88, 0x3dea8,
+		0x3deb0, 0x3deb4,
+		0x3dec8, 0x3ded4,
+		0x3dfb8, 0x3e004,
+		0x3e200, 0x3e200,
+		0x3e208, 0x3e240,
+		0x3e248, 0x3e280,
+		0x3e288, 0x3e2c0,
+		0x3e2c8, 0x3e2fc,
+		0x3e600, 0x3e630,
+		0x3ea00, 0x3eabc,
+		0x3eb00, 0x3eb10,
+		0x3eb20, 0x3eb30,
+		0x3eb40, 0x3eb50,
+		0x3eb60, 0x3eb70,
+		0x3f000, 0x3f028,
+		0x3f030, 0x3f048,
+		0x3f060, 0x3f068,
+		0x3f070, 0x3f09c,
+		0x3f0f0, 0x3f128,
+		0x3f130, 0x3f148,
+		0x3f160, 0x3f168,
+		0x3f170, 0x3f19c,
+		0x3f1f0, 0x3f238,
+		0x3f240, 0x3f240,
+		0x3f248, 0x3f250,
+		0x3f25c, 0x3f264,
+		0x3f270, 0x3f2b8,
+		0x3f2c0, 0x3f2e4,
+		0x3f2f8, 0x3f338,
+		0x3f340, 0x3f340,
+		0x3f348, 0x3f350,
+		0x3f35c, 0x3f364,
+		0x3f370, 0x3f3b8,
+		0x3f3c0, 0x3f3e4,
+		0x3f3f8, 0x3f428,
+		0x3f430, 0x3f448,
+		0x3f460, 0x3f468,
+		0x3f470, 0x3f49c,
+		0x3f4f0, 0x3f528,
+		0x3f530, 0x3f548,
+		0x3f560, 0x3f568,
+		0x3f570, 0x3f59c,
+		0x3f5f0, 0x3f638,
+		0x3f640, 0x3f640,
+		0x3f648, 0x3f650,
+		0x3f65c, 0x3f664,
+		0x3f670, 0x3f6b8,
+		0x3f6c0, 0x3f6e4,
+		0x3f6f8, 0x3f738,
+		0x3f740, 0x3f740,
+		0x3f748, 0x3f750,
+		0x3f75c, 0x3f764,
+		0x3f770, 0x3f7b8,
+		0x3f7c0, 0x3f7e4,
+		0x3f7f8, 0x3f7fc,
+		0x3f814, 0x3f814,
+		0x3f82c, 0x3f82c,
+		0x3f880, 0x3f88c,
+		0x3f8e8, 0x3f8ec,
+		0x3f900, 0x3f928,
+		0x3f930, 0x3f948,
+		0x3f960, 0x3f968,
+		0x3f970, 0x3f99c,
+		0x3f9f0, 0x3fa38,
+		0x3fa40, 0x3fa40,
+		0x3fa48, 0x3fa50,
+		0x3fa5c, 0x3fa64,
+		0x3fa70, 0x3fab8,
+		0x3fac0, 0x3fae4,
+		0x3faf8, 0x3fb10,
+		0x3fb28, 0x3fb28,
+		0x3fb3c, 0x3fb50,
+		0x3fbf0, 0x3fc10,
+		0x3fc28, 0x3fc28,
+		0x3fc3c, 0x3fc50,
+		0x3fcf0, 0x3fcfc,
+		0x40000, 0x4000c,
+		0x40040, 0x40050,
+		0x40060, 0x40068,
+		0x4007c, 0x4008c,
+		0x40094, 0x400b0,
+		0x400c0, 0x40144,
+		0x40180, 0x4018c,
+		0x40200, 0x40254,
+		0x40260, 0x40264,
+		0x40270, 0x40288,
+		0x40290, 0x40298,
+		0x402ac, 0x402c8,
+		0x402d0, 0x402e0,
+		0x402f0, 0x402f0,
+		0x40300, 0x4033c,
+		0x403f8, 0x403fc,
+		0x41304, 0x413c4,
+		0x41400, 0x4140c,
+		0x41414, 0x4141c,
+		0x41480, 0x414d0,
+		0x44000, 0x44054,
+		0x4405c, 0x44078,
+		0x440c0, 0x44174,
+		0x44180, 0x441ac,
+		0x441b4, 0x441b8,
+		0x441c0, 0x44254,
+		0x4425c, 0x44278,
+		0x442c0, 0x44374,
+		0x44380, 0x443ac,
+		0x443b4, 0x443b8,
+		0x443c0, 0x44454,
+		0x4445c, 0x44478,
+		0x444c0, 0x44574,
+		0x44580, 0x445ac,
+		0x445b4, 0x445b8,
+		0x445c0, 0x44654,
+		0x4465c, 0x44678,
+		0x446c0, 0x44774,
+		0x44780, 0x447ac,
+		0x447b4, 0x447b8,
+		0x447c0, 0x44854,
+		0x4485c, 0x44878,
+		0x448c0, 0x44974,
+		0x44980, 0x449ac,
+		0x449b4, 0x449b8,
+		0x449c0, 0x449fc,
+		0x45000, 0x45004,
+		0x45010, 0x45030,
+		0x45040, 0x45060,
+		0x45068, 0x45068,
+		0x45080, 0x45084,
+		0x450a0, 0x450b0,
+		0x45200, 0x45204,
+		0x45210, 0x45230,
+		0x45240, 0x45260,
+		0x45268, 0x45268,
+		0x45280, 0x45284,
+		0x452a0, 0x452b0,
+		0x460c0, 0x460e4,
+		0x47000, 0x4703c,
+		0x47044, 0x4708c,
+		0x47200, 0x47250,
+		0x47400, 0x47408,
+		0x47414, 0x47420,
+		0x47600, 0x47618,
+		0x47800, 0x47814,
+		0x48000, 0x4800c,
+		0x48040, 0x48050,
+		0x48060, 0x48068,
+		0x4807c, 0x4808c,
+		0x48094, 0x480b0,
+		0x480c0, 0x48144,
+		0x48180, 0x4818c,
+		0x48200, 0x48254,
+		0x48260, 0x48264,
+		0x48270, 0x48288,
+		0x48290, 0x48298,
+		0x482ac, 0x482c8,
+		0x482d0, 0x482e0,
+		0x482f0, 0x482f0,
+		0x48300, 0x4833c,
+		0x483f8, 0x483fc,
+		0x49304, 0x493c4,
+		0x49400, 0x4940c,
+		0x49414, 0x4941c,
+		0x49480, 0x494d0,
+		0x4c000, 0x4c054,
+		0x4c05c, 0x4c078,
+		0x4c0c0, 0x4c174,
+		0x4c180, 0x4c1ac,
+		0x4c1b4, 0x4c1b8,
+		0x4c1c0, 0x4c254,
+		0x4c25c, 0x4c278,
+		0x4c2c0, 0x4c374,
+		0x4c380, 0x4c3ac,
+		0x4c3b4, 0x4c3b8,
+		0x4c3c0, 0x4c454,
+		0x4c45c, 0x4c478,
+		0x4c4c0, 0x4c574,
+		0x4c580, 0x4c5ac,
+		0x4c5b4, 0x4c5b8,
+		0x4c5c0, 0x4c654,
+		0x4c65c, 0x4c678,
+		0x4c6c0, 0x4c774,
+		0x4c780, 0x4c7ac,
+		0x4c7b4, 0x4c7b8,
+		0x4c7c0, 0x4c854,
+		0x4c85c, 0x4c878,
+		0x4c8c0, 0x4c974,
+		0x4c980, 0x4c9ac,
+		0x4c9b4, 0x4c9b8,
+		0x4c9c0, 0x4c9fc,
+		0x4d000, 0x4d004,
+		0x4d010, 0x4d030,
+		0x4d040, 0x4d060,
+		0x4d068, 0x4d068,
+		0x4d080, 0x4d084,
+		0x4d0a0, 0x4d0b0,
+		0x4d200, 0x4d204,
+		0x4d210, 0x4d230,
+		0x4d240, 0x4d260,
+		0x4d268, 0x4d268,
+		0x4d280, 0x4d284,
+		0x4d2a0, 0x4d2b0,
+		0x4e0c0, 0x4e0e4,
+		0x4f000, 0x4f03c,
+		0x4f044, 0x4f08c,
+		0x4f200, 0x4f250,
+		0x4f400, 0x4f408,
+		0x4f414, 0x4f420,
+		0x4f600, 0x4f618,
+		0x4f800, 0x4f814,
+		0x50000, 0x50084,
+		0x50090, 0x500cc,
+		0x50400, 0x50400,
+		0x50800, 0x50884,
+		0x50890, 0x508cc,
+		0x50c00, 0x50c00,
+		0x51000, 0x5101c,
+		0x51300, 0x51308,
+	};
+
+	static const unsigned int t6_reg_ranges[] = {
+		0x1008, 0x101c,
+		0x1024, 0x10a8,
+		0x10b4, 0x10f8,
+		0x1100, 0x1114,
+		0x111c, 0x112c,
+		0x1138, 0x113c,
+		0x1144, 0x114c,
+		0x1180, 0x1184,
+		0x1190, 0x1194,
+		0x11a0, 0x11a4,
+		0x11b0, 0x11b4,
+		0x11fc, 0x123c,
+		0x1254, 0x1274,
+		0x1280, 0x133c,
+		0x1800, 0x18fc,
+		0x3000, 0x302c,
+		0x3060, 0x30b0,
+		0x30b8, 0x30d8,
+		0x30e0, 0x30fc,
+		0x3140, 0x357c,
+		0x35a8, 0x35cc,
+		0x35ec, 0x35ec,
+		0x3600, 0x5624,
+		0x56cc, 0x56ec,
+		0x56f4, 0x5720,
+		0x5728, 0x575c,
+		0x580c, 0x5814,
+		0x5890, 0x589c,
+		0x58a4, 0x58ac,
+		0x58b8, 0x58bc,
+		0x5940, 0x595c,
+		0x5980, 0x598c,
+		0x59b0, 0x59c8,
+		0x59d0, 0x59dc,
+		0x59fc, 0x5a18,
+		0x5a60, 0x5a6c,
+		0x5a80, 0x5a8c,
+		0x5a94, 0x5a9c,
+		0x5b94, 0x5bfc,
+		0x5c10, 0x5e48,
+		0x5e50, 0x5e94,
+		0x5ea0, 0x5eb0,
+		0x5ec0, 0x5ec0,
+		0x5ec8, 0x5ed0,
+		0x5ee0, 0x5ee0,
+		0x5ef0, 0x5ef0,
+		0x5f00, 0x5f00,
+		0x6000, 0x6020,
+		0x6028, 0x6040,
+		0x6058, 0x609c,
+		0x60a8, 0x619c,
+		0x7700, 0x7798,
+		0x77c0, 0x7880,
+		0x78cc, 0x78fc,
+		0x7b00, 0x7b58,
+		0x7b60, 0x7b84,
+		0x7b8c, 0x7c54,
+		0x7d00, 0x7d38,
+		0x7d40, 0x7d84,
+		0x7d8c, 0x7ddc,
+		0x7de4, 0x7e04,
+		0x7e10, 0x7e1c,
+		0x7e24, 0x7e38,
+		0x7e40, 0x7e44,
+		0x7e4c, 0x7e78,
+		0x7e80, 0x7edc,
+		0x7ee8, 0x7efc,
+		0x8dc0, 0x8de4,
+		0x8df8, 0x8e04,
+		0x8e10, 0x8e84,
+		0x8ea0, 0x8f88,
+		0x8fb8, 0x9058,
+		0x9060, 0x9060,
+		0x9068, 0x90f8,
+		0x9100, 0x9124,
+		0x9400, 0x9470,
+		0x9600, 0x9600,
+		0x9608, 0x9638,
+		0x9640, 0x9704,
+		0x9710, 0x971c,
+		0x9800, 0x9808,
+		0x9810, 0x9864,
+		0x9c00, 0x9c6c,
+		0x9c80, 0x9cec,
+		0x9d00, 0x9d6c,
+		0x9d80, 0x9dec,
+		0x9e00, 0x9e6c,
+		0x9e80, 0x9eec,
+		0x9f00, 0x9f6c,
+		0x9f80, 0xa020,
+		0xd000, 0xd03c,
+		0xd100, 0xd118,
+		0xd200, 0xd214,
+		0xd220, 0xd234,
+		0xd240, 0xd254,
+		0xd260, 0xd274,
+		0xd280, 0xd294,
+		0xd2a0, 0xd2b4,
+		0xd2c0, 0xd2d4,
+		0xd2e0, 0xd2f4,
+		0xd300, 0xd31c,
+		0xdfc0, 0xdfe0,
+		0xe000, 0xf008,
+		0xf010, 0xf018,
+		0xf020, 0xf028,
+		0x11000, 0x11014,
+		0x11048, 0x1106c,
+		0x11074, 0x11088,
+		0x11098, 0x11120,
+		0x1112c, 0x1117c,
+		0x11190, 0x112e0,
+		0x11300, 0x1130c,
+		0x12000, 0x1206c,
+		0x19040, 0x1906c,
+		0x19078, 0x19080,
+		0x1908c, 0x190e8,
+		0x190f0, 0x190f8,
+		0x19100, 0x19110,
+		0x19120, 0x19124,
+		0x19150, 0x19194,
+		0x1919c, 0x191b0,
+		0x191d0, 0x191e8,
+		0x19238, 0x19290,
+		0x192a4, 0x192b0,
+		0x192bc, 0x192bc,
+		0x19348, 0x1934c,
+		0x193f8, 0x19418,
+		0x19420, 0x19428,
+		0x19430, 0x19444,
+		0x1944c, 0x1946c,
+		0x19474, 0x19474,
+		0x19490, 0x194cc,
+		0x194f0, 0x194f8,
+		0x19c00, 0x19c48,
+		0x19c50, 0x19c80,
+		0x19c94, 0x19c98,
+		0x19ca0, 0x19cbc,
+		0x19ce4, 0x19ce4,
+		0x19cf0, 0x19cf8,
+		0x19d00, 0x19d28,
+		0x19d50, 0x19d78,
+		0x19d94, 0x19d98,
+		0x19da0, 0x19dc8,
+		0x19df0, 0x19e10,
+		0x19e50, 0x19e6c,
+		0x19ea0, 0x19ebc,
+		0x19ec4, 0x19ef4,
+		0x19f04, 0x19f2c,
+		0x19f34, 0x19f34,
+		0x19f40, 0x19f50,
+		0x19f90, 0x19fac,
+		0x19fc4, 0x19fc8,
+		0x19fd0, 0x19fe4,
+		0x1a000, 0x1a004,
+		0x1a010, 0x1a06c,
+		0x1a0b0, 0x1a0e4,
+		0x1a0ec, 0x1a0f8,
+		0x1a100, 0x1a108,
+		0x1a114, 0x1a130,
+		0x1a138, 0x1a1c4,
+		0x1a1fc, 0x1a1fc,
+		0x1e008, 0x1e00c,
+		0x1e040, 0x1e044,
+		0x1e04c, 0x1e04c,
+		0x1e284, 0x1e290,
+		0x1e2c0, 0x1e2c0,
+		0x1e2e0, 0x1e2e0,
+		0x1e300, 0x1e384,
+		0x1e3c0, 0x1e3c8,
+		0x1e408, 0x1e40c,
+		0x1e440, 0x1e444,
+		0x1e44c, 0x1e44c,
+		0x1e684, 0x1e690,
+		0x1e6c0, 0x1e6c0,
+		0x1e6e0, 0x1e6e0,
+		0x1e700, 0x1e784,
+		0x1e7c0, 0x1e7c8,
+		0x1e808, 0x1e80c,
+		0x1e840, 0x1e844,
+		0x1e84c, 0x1e84c,
+		0x1ea84, 0x1ea90,
+		0x1eac0, 0x1eac0,
+		0x1eae0, 0x1eae0,
+		0x1eb00, 0x1eb84,
+		0x1ebc0, 0x1ebc8,
+		0x1ec08, 0x1ec0c,
+		0x1ec40, 0x1ec44,
+		0x1ec4c, 0x1ec4c,
+		0x1ee84, 0x1ee90,
+		0x1eec0, 0x1eec0,
+		0x1eee0, 0x1eee0,
+		0x1ef00, 0x1ef84,
+		0x1efc0, 0x1efc8,
+		0x1f008, 0x1f00c,
+		0x1f040, 0x1f044,
+		0x1f04c, 0x1f04c,
+		0x1f284, 0x1f290,
+		0x1f2c0, 0x1f2c0,
+		0x1f2e0, 0x1f2e0,
+		0x1f300, 0x1f384,
+		0x1f3c0, 0x1f3c8,
+		0x1f408, 0x1f40c,
+		0x1f440, 0x1f444,
+		0x1f44c, 0x1f44c,
+		0x1f684, 0x1f690,
+		0x1f6c0, 0x1f6c0,
+		0x1f6e0, 0x1f6e0,
+		0x1f700, 0x1f784,
+		0x1f7c0, 0x1f7c8,
+		0x1f808, 0x1f80c,
+		0x1f840, 0x1f844,
+		0x1f84c, 0x1f84c,
+		0x1fa84, 0x1fa90,
+		0x1fac0, 0x1fac0,
+		0x1fae0, 0x1fae0,
+		0x1fb00, 0x1fb84,
+		0x1fbc0, 0x1fbc8,
+		0x1fc08, 0x1fc0c,
+		0x1fc40, 0x1fc44,
+		0x1fc4c, 0x1fc4c,
+		0x1fe84, 0x1fe90,
+		0x1fec0, 0x1fec0,
+		0x1fee0, 0x1fee0,
+		0x1ff00, 0x1ff84,
+		0x1ffc0, 0x1ffc8,
+		0x30000, 0x30030,
+		0x30100, 0x30168,
+		0x30190, 0x301a0,
+		0x301a8, 0x301b8,
+		0x301c4, 0x301c8,
+		0x301d0, 0x301d0,
+		0x30200, 0x30320,
+		0x30400, 0x304b4,
+		0x304c0, 0x3052c,
+		0x30540, 0x3061c,
+		0x30800, 0x308a0,
+		0x308c0, 0x30908,
+		0x30910, 0x309b8,
+		0x30a00, 0x30a04,
+		0x30a0c, 0x30a14,
+		0x30a1c, 0x30a2c,
+		0x30a44, 0x30a50,
+		0x30a74, 0x30a74,
+		0x30a7c, 0x30afc,
+		0x30b08, 0x30c24,
+		0x30d00, 0x30d14,
+		0x30d1c, 0x30d3c,
+		0x30d44, 0x30d4c,
+		0x30d54, 0x30d74,
+		0x30d7c, 0x30d7c,
+		0x30de0, 0x30de0,
+		0x30e00, 0x30ed4,
+		0x30f00, 0x30fa4,
+		0x30fc0, 0x30fc4,
+		0x31000, 0x31004,
+		0x31080, 0x310fc,
+		0x31208, 0x31220,
+		0x3123c, 0x31254,
+		0x31300, 0x31300,
+		0x31308, 0x3131c,
+		0x31338, 0x3133c,
+		0x31380, 0x31380,
+		0x31388, 0x313a8,
+		0x313b4, 0x313b4,
+		0x31400, 0x31420,
+		0x31438, 0x3143c,
+		0x31480, 0x31480,
+		0x314a8, 0x314a8,
+		0x314b0, 0x314b4,
+		0x314c8, 0x314d4,
+		0x31a40, 0x31a4c,
+		0x31af0, 0x31b20,
+		0x31b38, 0x31b3c,
+		0x31b80, 0x31b80,
+		0x31ba8, 0x31ba8,
+		0x31bb0, 0x31bb4,
+		0x31bc8, 0x31bd4,
+		0x32140, 0x3218c,
+		0x321f0, 0x321f4,
+		0x32200, 0x32200,
+		0x32218, 0x32218,
+		0x32400, 0x32400,
+		0x32408, 0x3241c,
+		0x32618, 0x32620,
+		0x32664, 0x32664,
+		0x326a8, 0x326a8,
+		0x326ec, 0x326ec,
+		0x32a00, 0x32abc,
+		0x32b00, 0x32b18,
+		0x32b20, 0x32b38,
+		0x32b40, 0x32b58,
+		0x32b60, 0x32b78,
+		0x32c00, 0x32c00,
+		0x32c08, 0x32c3c,
+		0x33000, 0x3302c,
+		0x33034, 0x33050,
+		0x33058, 0x33058,
+		0x33060, 0x3308c,
+		0x3309c, 0x330ac,
+		0x330c0, 0x330c0,
+		0x330c8, 0x330d0,
+		0x330d8, 0x330e0,
+		0x330ec, 0x3312c,
+		0x33134, 0x33150,
+		0x33158, 0x33158,
+		0x33160, 0x3318c,
+		0x3319c, 0x331ac,
+		0x331c0, 0x331c0,
+		0x331c8, 0x331d0,
+		0x331d8, 0x331e0,
+		0x331ec, 0x33290,
+		0x33298, 0x332c4,
+		0x332e4, 0x33390,
+		0x33398, 0x333c4,
+		0x333e4, 0x3342c,
+		0x33434, 0x33450,
+		0x33458, 0x33458,
+		0x33460, 0x3348c,
+		0x3349c, 0x334ac,
+		0x334c0, 0x334c0,
+		0x334c8, 0x334d0,
+		0x334d8, 0x334e0,
+		0x334ec, 0x3352c,
+		0x33534, 0x33550,
+		0x33558, 0x33558,
+		0x33560, 0x3358c,
+		0x3359c, 0x335ac,
+		0x335c0, 0x335c0,
+		0x335c8, 0x335d0,
+		0x335d8, 0x335e0,
+		0x335ec, 0x33690,
+		0x33698, 0x336c4,
+		0x336e4, 0x33790,
+		0x33798, 0x337c4,
+		0x337e4, 0x337fc,
+		0x33814, 0x33814,
+		0x33854, 0x33868,
+		0x33880, 0x3388c,
+		0x338c0, 0x338d0,
+		0x338e8, 0x338ec,
+		0x33900, 0x3392c,
+		0x33934, 0x33950,
+		0x33958, 0x33958,
+		0x33960, 0x3398c,
+		0x3399c, 0x339ac,
+		0x339c0, 0x339c0,
+		0x339c8, 0x339d0,
+		0x339d8, 0x339e0,
+		0x339ec, 0x33a90,
+		0x33a98, 0x33ac4,
+		0x33ae4, 0x33b10,
+		0x33b24, 0x33b28,
+		0x33b38, 0x33b50,
+		0x33bf0, 0x33c10,
+		0x33c24, 0x33c28,
+		0x33c38, 0x33c50,
+		0x33cf0, 0x33cfc,
+		0x34000, 0x34030,
+		0x34100, 0x34168,
+		0x34190, 0x341a0,
+		0x341a8, 0x341b8,
+		0x341c4, 0x341c8,
+		0x341d0, 0x341d0,
+		0x34200, 0x34320,
+		0x34400, 0x344b4,
+		0x344c0, 0x3452c,
+		0x34540, 0x3461c,
+		0x34800, 0x348a0,
+		0x348c0, 0x34908,
+		0x34910, 0x349b8,
+		0x34a00, 0x34a04,
+		0x34a0c, 0x34a14,
+		0x34a1c, 0x34a2c,
+		0x34a44, 0x34a50,
+		0x34a74, 0x34a74,
+		0x34a7c, 0x34afc,
+		0x34b08, 0x34c24,
+		0x34d00, 0x34d14,
+		0x34d1c, 0x34d3c,
+		0x34d44, 0x34d4c,
+		0x34d54, 0x34d74,
+		0x34d7c, 0x34d7c,
+		0x34de0, 0x34de0,
+		0x34e00, 0x34ed4,
+		0x34f00, 0x34fa4,
+		0x34fc0, 0x34fc4,
+		0x35000, 0x35004,
+		0x35080, 0x350fc,
+		0x35208, 0x35220,
+		0x3523c, 0x35254,
+		0x35300, 0x35300,
+		0x35308, 0x3531c,
+		0x35338, 0x3533c,
+		0x35380, 0x35380,
+		0x35388, 0x353a8,
+		0x353b4, 0x353b4,
+		0x35400, 0x35420,
+		0x35438, 0x3543c,
+		0x35480, 0x35480,
+		0x354a8, 0x354a8,
+		0x354b0, 0x354b4,
+		0x354c8, 0x354d4,
+		0x35a40, 0x35a4c,
+		0x35af0, 0x35b20,
+		0x35b38, 0x35b3c,
+		0x35b80, 0x35b80,
+		0x35ba8, 0x35ba8,
+		0x35bb0, 0x35bb4,
+		0x35bc8, 0x35bd4,
+		0x36140, 0x3618c,
+		0x361f0, 0x361f4,
+		0x36200, 0x36200,
+		0x36218, 0x36218,
+		0x36400, 0x36400,
+		0x36408, 0x3641c,
+		0x36618, 0x36620,
+		0x36664, 0x36664,
+		0x366a8, 0x366a8,
+		0x366ec, 0x366ec,
+		0x36a00, 0x36abc,
+		0x36b00, 0x36b18,
+		0x36b20, 0x36b38,
+		0x36b40, 0x36b58,
+		0x36b60, 0x36b78,
+		0x36c00, 0x36c00,
+		0x36c08, 0x36c3c,
+		0x37000, 0x3702c,
+		0x37034, 0x37050,
+		0x37058, 0x37058,
+		0x37060, 0x3708c,
+		0x3709c, 0x370ac,
+		0x370c0, 0x370c0,
+		0x370c8, 0x370d0,
+		0x370d8, 0x370e0,
+		0x370ec, 0x3712c,
+		0x37134, 0x37150,
+		0x37158, 0x37158,
+		0x37160, 0x3718c,
+		0x3719c, 0x371ac,
+		0x371c0, 0x371c0,
+		0x371c8, 0x371d0,
+		0x371d8, 0x371e0,
+		0x371ec, 0x37290,
+		0x37298, 0x372c4,
+		0x372e4, 0x37390,
+		0x37398, 0x373c4,
+		0x373e4, 0x3742c,
+		0x37434, 0x37450,
+		0x37458, 0x37458,
+		0x37460, 0x3748c,
+		0x3749c, 0x374ac,
+		0x374c0, 0x374c0,
+		0x374c8, 0x374d0,
+		0x374d8, 0x374e0,
+		0x374ec, 0x3752c,
+		0x37534, 0x37550,
+		0x37558, 0x37558,
+		0x37560, 0x3758c,
+		0x3759c, 0x375ac,
+		0x375c0, 0x375c0,
+		0x375c8, 0x375d0,
+		0x375d8, 0x375e0,
+		0x375ec, 0x37690,
+		0x37698, 0x376c4,
+		0x376e4, 0x37790,
+		0x37798, 0x377c4,
+		0x377e4, 0x377fc,
+		0x37814, 0x37814,
+		0x37854, 0x37868,
+		0x37880, 0x3788c,
+		0x378c0, 0x378d0,
+		0x378e8, 0x378ec,
+		0x37900, 0x3792c,
+		0x37934, 0x37950,
+		0x37958, 0x37958,
+		0x37960, 0x3798c,
+		0x3799c, 0x379ac,
+		0x379c0, 0x379c0,
+		0x379c8, 0x379d0,
+		0x379d8, 0x379e0,
+		0x379ec, 0x37a90,
+		0x37a98, 0x37ac4,
+		0x37ae4, 0x37b10,
+		0x37b24, 0x37b28,
+		0x37b38, 0x37b50,
+		0x37bf0, 0x37c10,
+		0x37c24, 0x37c28,
+		0x37c38, 0x37c50,
+		0x37cf0, 0x37cfc,
+		0x40040, 0x40040,
+		0x40080, 0x40084,
+		0x40100, 0x40100,
+		0x40140, 0x401bc,
+		0x40200, 0x40214,
+		0x40228, 0x40228,
+		0x40240, 0x40258,
+		0x40280, 0x40280,
+		0x40304, 0x40304,
+		0x40330, 0x4033c,
+		0x41304, 0x413c8,
+		0x413d0, 0x413dc,
+		0x413f0, 0x413f0,
+		0x41400, 0x4140c,
+		0x41414, 0x4141c,
+		0x41480, 0x414d0,
+		0x44000, 0x4407c,
+		0x440c0, 0x441ac,
+		0x441b4, 0x4427c,
+		0x442c0, 0x443ac,
+		0x443b4, 0x4447c,
+		0x444c0, 0x445ac,
+		0x445b4, 0x4467c,
+		0x446c0, 0x447ac,
+		0x447b4, 0x4487c,
+		0x448c0, 0x449ac,
+		0x449b4, 0x44a7c,
+		0x44ac0, 0x44bac,
+		0x44bb4, 0x44c7c,
+		0x44cc0, 0x44dac,
+		0x44db4, 0x44e7c,
+		0x44ec0, 0x44fac,
+		0x44fb4, 0x4507c,
+		0x450c0, 0x451ac,
+		0x451b4, 0x451fc,
+		0x45800, 0x45804,
+		0x45810, 0x45830,
+		0x45840, 0x45860,
+		0x45868, 0x45868,
+		0x45880, 0x45884,
+		0x458a0, 0x458b0,
+		0x45a00, 0x45a04,
+		0x45a10, 0x45a30,
+		0x45a40, 0x45a60,
+		0x45a68, 0x45a68,
+		0x45a80, 0x45a84,
+		0x45aa0, 0x45ab0,
+		0x460c0, 0x460e4,
+		0x47000, 0x4703c,
+		0x47044, 0x4708c,
+		0x47200, 0x47250,
+		0x47400, 0x47408,
+		0x47414, 0x47420,
+		0x47600, 0x47618,
+		0x47800, 0x47814,
+		0x47820, 0x4782c,
+		0x50000, 0x50084,
+		0x50090, 0x500cc,
+		0x50300, 0x50384,
+		0x50400, 0x50400,
+		0x50800, 0x50884,
+		0x50890, 0x508cc,
+		0x50b00, 0x50b84,
+		0x50c00, 0x50c00,
+		0x51000, 0x51020,
+		0x51028, 0x510b0,
+		0x51300, 0x51324,
+	};
+
+	u32 *buf_end = (u32 *)((char *)buf + buf_size);
+	const unsigned int *reg_ranges;
+	int reg_ranges_size, range;
+	unsigned int chip_version = CHELSIO_CHIP_VERSION(adap->params.chip);
+
+	/* Select the right set of register ranges to dump depending on the
+	 * adapter chip type.
+	 */
+	switch (chip_version) {
+	case CHELSIO_T4:
+		reg_ranges = t4_reg_ranges;
+		reg_ranges_size = ARRAY_SIZE(t4_reg_ranges);
+		break;
+
+	case CHELSIO_T5:
+		reg_ranges = t5_reg_ranges;
+		reg_ranges_size = ARRAY_SIZE(t5_reg_ranges);
+		break;
+
+	case CHELSIO_T6:
+		reg_ranges = t6_reg_ranges;
+		reg_ranges_size = ARRAY_SIZE(t6_reg_ranges);
+		break;
+
+	default:
+		dev_err(adap->pdev_dev,
+			"Unsupported chip version %d\n", chip_version);
+		return;
+	}
+
+	/* Clear the register buffer and insert the appropriate register
+	 * values selected by the above register ranges.
+	 */
+	memset(buf, 0, buf_size);
+	for (range = 0; range < reg_ranges_size; range += 2) {
+		unsigned int reg = reg_ranges[range];
+		unsigned int last_reg = reg_ranges[range + 1];
+		u32 *bufp = (u32 *)((char *)buf + reg);
+
+		/* Iterate across the register range filling in the register
+		 * buffer but don't write past the end of the register buffer.
+		 */
+		while (reg <= last_reg && bufp < buf_end) {
+			*bufp++ = t4_read_reg(adap, reg);
+			reg += sizeof(u32);
+		}
+	}
+}
+
+#define EEPROM_STAT_ADDR   0x7bfc
+#define VPD_BASE           0x400
+#define VPD_BASE_OLD       0
+#define VPD_LEN            1024
+
+/**
+ * t4_eeprom_ptov - translate a physical EEPROM address to virtual
+ * @phys_addr: the physical EEPROM address
+ * @fn: the PCI function number
+ * @sz: size of function-specific area
+ *
+ * Translate a physical EEPROM address to virtual.  The first 1K is
+ * accessed through virtual addresses starting at 31K, the rest is
+ * accessed through virtual addresses starting at 0.
+ *
+ * The mapping is as follows:
+ * [0..1K) -> [31K..32K)
+ * [1K..1K+A) -> [31K-A..31K)
+ * [1K+A..ES) -> [0..ES-A-1K)
+ *
+ * where A = @fn * @sz, and ES = EEPROM size.
+ */
+int t4_eeprom_ptov(unsigned int phys_addr, unsigned int fn, unsigned int sz)
+{
+	fn *= sz;
+	if (phys_addr < 1024)
+		return phys_addr + (31 << 10);
+	if (phys_addr < 1024 + fn)
+		return 31744 - fn + phys_addr - 1024;
+	if (phys_addr < EEPROMSIZE)
+		return phys_addr - 1024 - fn;
+	return -EINVAL;
+}
+
+/**
+ *	t4_seeprom_wp - enable/disable EEPROM write protection
+ *	@adapter: the adapter
+ *	@enable: whether to enable or disable write protection
+ *
+ *	Enables or disables write protection on the serial EEPROM.
+ */
+int t4_seeprom_wp(struct adapter *adapter, bool enable)
+{
+	unsigned int v = enable ? 0xc : 0;
+	int ret = pci_write_vpd(adapter->pdev, EEPROM_STAT_ADDR, 4, &v);
+	return ret < 0 ? ret : 0;
+}
+
+/**
+ *	t4_get_raw_vpd_params - read VPD parameters from VPD EEPROM
+ *	@adapter: adapter to read
+ *	@p: where to store the parameters
+ *
+ *	Reads card parameters stored in VPD EEPROM.
+ */
+int t4_get_raw_vpd_params(struct adapter *adapter, struct vpd_params *p)
+{
+	unsigned int id_len, pn_len, sn_len, na_len;
+	int id, sn, pn, na, addr, ret = 0;
+	u8 *vpd, base_val = 0;
+
+	vpd = vmalloc(VPD_LEN);
+	if (!vpd)
+		return -ENOMEM;
+
+	/* Card information normally starts at VPD_BASE but early cards had
+	 * it at 0.
+	 */
+	ret = pci_read_vpd(adapter->pdev, VPD_BASE, 1, &base_val);
+	if (ret < 0)
+		goto out;
+
+	addr = base_val == PCI_VPD_LRDT_ID_STRING ? VPD_BASE : VPD_BASE_OLD;
+
+	ret = pci_read_vpd(adapter->pdev, addr, VPD_LEN, vpd);
+	if (ret < 0)
+		goto out;
+
+	ret = pci_vpd_find_id_string(vpd, VPD_LEN, &id_len);
+	if (ret < 0)
+		goto out;
+	id = ret;
+
+	ret = pci_vpd_check_csum(vpd, VPD_LEN);
+	if (ret) {
+		dev_err(adapter->pdev_dev, "VPD checksum incorrect or missing\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	ret = pci_vpd_find_ro_info_keyword(vpd, VPD_LEN,
+					   PCI_VPD_RO_KEYWORD_SERIALNO, &sn_len);
+	if (ret < 0)
+		goto out;
+	sn = ret;
+
+	ret = pci_vpd_find_ro_info_keyword(vpd, VPD_LEN,
+					   PCI_VPD_RO_KEYWORD_PARTNO, &pn_len);
+	if (ret < 0)
+		goto out;
+	pn = ret;
+
+	ret = pci_vpd_find_ro_info_keyword(vpd, VPD_LEN, "NA", &na_len);
+	if (ret < 0)
+		goto out;
+	na = ret;
+
+	memcpy(p->id, vpd + id, min_t(unsigned int, id_len, ID_LEN));
+	strim(p->id);
+	memcpy(p->sn, vpd + sn, min_t(unsigned int, sn_len, SERNUM_LEN));
+	strim(p->sn);
+	memcpy(p->pn, vpd + pn, min_t(unsigned int, pn_len, PN_LEN));
+	strim(p->pn);
+	memcpy(p->na, vpd + na, min_t(unsigned int, na_len, MACADDR_LEN));
+	strim(p->na);
+
+out:
+	vfree(vpd);
+	if (ret < 0) {
+		dev_err(adapter->pdev_dev, "error reading VPD\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+/**
+ *	t4_get_vpd_params - read VPD parameters & retrieve Core Clock
+ *	@adapter: adapter to read
+ *	@p: where to store the parameters
+ *
+ *	Reads card parameters stored in VPD EEPROM and retrieves the Core
+ *	Clock.  This can only be called after a connection to the firmware
+ *	is established.
+ */
+int t4_get_vpd_params(struct adapter *adapter, struct vpd_params *p)
+{
+	u32 cclk_param, cclk_val;
+	int ret;
+
+	/* Grab the raw VPD parameters.
+	 */
+	ret = t4_get_raw_vpd_params(adapter, p);
+	if (ret)
+		return ret;
+
+	/* Ask firmware for the Core Clock since it knows how to translate the
+	 * Reference Clock ('V2') VPD field into a Core Clock value ...
+	 */
+	cclk_param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		      FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_CCLK));
+	ret = t4_query_params(adapter, adapter->mbox, adapter->pf, 0,
+			      1, &cclk_param, &cclk_val);
+
+	if (ret)
+		return ret;
+	p->cclk = cclk_val;
+
+	return 0;
+}
+
+/**
+ *	t4_get_pfres - retrieve VF resource limits
+ *	@adapter: the adapter
+ *
+ *	Retrieves configured resource limits and capabilities for a physical
+ *	function.  The results are stored in @adapter->pfres.
+ */
+int t4_get_pfres(struct adapter *adapter)
+{
+	struct pf_resources *pfres = &adapter->params.pfres;
+	struct fw_pfvf_cmd cmd, rpl;
+	int v;
+	u32 word;
+
+	/* Execute PFVF Read command to get VF resource limits; bail out early
+	 * with error on command failure.
+	 */
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_PFVF_CMD) |
+				    FW_CMD_REQUEST_F |
+				    FW_CMD_READ_F |
+				    FW_PFVF_CMD_PFN_V(adapter->pf) |
+				    FW_PFVF_CMD_VFN_V(0));
+	cmd.retval_len16 = cpu_to_be32(FW_LEN16(cmd));
+	v = t4_wr_mbox(adapter, adapter->mbox, &cmd, sizeof(cmd), &rpl);
+	if (v != FW_SUCCESS)
+		return v;
+
+	/* Extract PF resource limits and return success.
+	 */
+	word = be32_to_cpu(rpl.niqflint_niq);
+	pfres->niqflint = FW_PFVF_CMD_NIQFLINT_G(word);
+	pfres->niq = FW_PFVF_CMD_NIQ_G(word);
+
+	word = be32_to_cpu(rpl.type_to_neq);
+	pfres->neq = FW_PFVF_CMD_NEQ_G(word);
+	pfres->pmask = FW_PFVF_CMD_PMASK_G(word);
+
+	word = be32_to_cpu(rpl.tc_to_nexactf);
+	pfres->tc = FW_PFVF_CMD_TC_G(word);
+	pfres->nvi = FW_PFVF_CMD_NVI_G(word);
+	pfres->nexactf = FW_PFVF_CMD_NEXACTF_G(word);
+
+	word = be32_to_cpu(rpl.r_caps_to_nethctrl);
+	pfres->r_caps = FW_PFVF_CMD_R_CAPS_G(word);
+	pfres->wx_caps = FW_PFVF_CMD_WX_CAPS_G(word);
+	pfres->nethctrl = FW_PFVF_CMD_NETHCTRL_G(word);
+
+	return 0;
+}
+
+/* serial flash and firmware constants */
+enum {
+	SF_ATTEMPTS = 10,             /* max retries for SF operations */
+
+	/* flash command opcodes */
+	SF_PROG_PAGE    = 2,          /* program page */
+	SF_WR_DISABLE   = 4,          /* disable writes */
+	SF_RD_STATUS    = 5,          /* read status register */
+	SF_WR_ENABLE    = 6,          /* enable writes */
+	SF_RD_DATA_FAST = 0xb,        /* read flash */
+	SF_RD_ID        = 0x9f,       /* read ID */
+	SF_ERASE_SECTOR = 0xd8,       /* erase sector */
+};
+
+/**
+ *	sf1_read - read data from the serial flash
+ *	@adapter: the adapter
+ *	@byte_cnt: number of bytes to read
+ *	@cont: whether another operation will be chained
+ *	@lock: whether to lock SF for PL access only
+ *	@valp: where to store the read data
+ *
+ *	Reads up to 4 bytes of data from the serial flash.  The location of
+ *	the read needs to be specified prior to calling this by issuing the
+ *	appropriate commands to the serial flash.
+ */
+static int sf1_read(struct adapter *adapter, unsigned int byte_cnt, int cont,
+		    int lock, u32 *valp)
+{
+	int ret;
+
+	if (!byte_cnt || byte_cnt > 4)
+		return -EINVAL;
+	if (t4_read_reg(adapter, SF_OP_A) & SF_BUSY_F)
+		return -EBUSY;
+	t4_write_reg(adapter, SF_OP_A, SF_LOCK_V(lock) |
+		     SF_CONT_V(cont) | BYTECNT_V(byte_cnt - 1));
+	ret = t4_wait_op_done(adapter, SF_OP_A, SF_BUSY_F, 0, SF_ATTEMPTS, 5);
+	if (!ret)
+		*valp = t4_read_reg(adapter, SF_DATA_A);
+	return ret;
+}
+
+/**
+ *	sf1_write - write data to the serial flash
+ *	@adapter: the adapter
+ *	@byte_cnt: number of bytes to write
+ *	@cont: whether another operation will be chained
+ *	@lock: whether to lock SF for PL access only
+ *	@val: value to write
+ *
+ *	Writes up to 4 bytes of data to the serial flash.  The location of
+ *	the write needs to be specified prior to calling this by issuing the
+ *	appropriate commands to the serial flash.
+ */
+static int sf1_write(struct adapter *adapter, unsigned int byte_cnt, int cont,
+		     int lock, u32 val)
+{
+	if (!byte_cnt || byte_cnt > 4)
+		return -EINVAL;
+	if (t4_read_reg(adapter, SF_OP_A) & SF_BUSY_F)
+		return -EBUSY;
+	t4_write_reg(adapter, SF_DATA_A, val);
+	t4_write_reg(adapter, SF_OP_A, SF_LOCK_V(lock) |
+		     SF_CONT_V(cont) | BYTECNT_V(byte_cnt - 1) | OP_V(1));
+	return t4_wait_op_done(adapter, SF_OP_A, SF_BUSY_F, 0, SF_ATTEMPTS, 5);
+}
+
+/**
+ *	flash_wait_op - wait for a flash operation to complete
+ *	@adapter: the adapter
+ *	@attempts: max number of polls of the status register
+ *	@delay: delay between polls in ms
+ *
+ *	Wait for a flash operation to complete by polling the status register.
+ */
+static int flash_wait_op(struct adapter *adapter, int attempts, int delay)
+{
+	int ret;
+	u32 status;
+
+	while (1) {
+		if ((ret = sf1_write(adapter, 1, 1, 1, SF_RD_STATUS)) != 0 ||
+		    (ret = sf1_read(adapter, 1, 0, 1, &status)) != 0)
+			return ret;
+		if (!(status & 1))
+			return 0;
+		if (--attempts == 0)
+			return -EAGAIN;
+		if (delay)
+			msleep(delay);
+	}
+}
+
+/**
+ *	t4_read_flash - read words from serial flash
+ *	@adapter: the adapter
+ *	@addr: the start address for the read
+ *	@nwords: how many 32-bit words to read
+ *	@data: where to store the read data
+ *	@byte_oriented: whether to store data as bytes or as words
+ *
+ *	Read the specified number of 32-bit words from the serial flash.
+ *	If @byte_oriented is set the read data is stored as a byte array
+ *	(i.e., big-endian), otherwise as 32-bit words in the platform's
+ *	natural endianness.
+ */
+int t4_read_flash(struct adapter *adapter, unsigned int addr,
+		  unsigned int nwords, u32 *data, int byte_oriented)
+{
+	int ret;
+
+	if (addr + nwords * sizeof(u32) > adapter->params.sf_size || (addr & 3))
+		return -EINVAL;
+
+	addr = swab32(addr) | SF_RD_DATA_FAST;
+
+	if ((ret = sf1_write(adapter, 4, 1, 0, addr)) != 0 ||
+	    (ret = sf1_read(adapter, 1, 1, 0, data)) != 0)
+		return ret;
+
+	for ( ; nwords; nwords--, data++) {
+		ret = sf1_read(adapter, 4, nwords > 1, nwords == 1, data);
+		if (nwords == 1)
+			t4_write_reg(adapter, SF_OP_A, 0);    /* unlock SF */
+		if (ret)
+			return ret;
+		if (byte_oriented)
+			*data = (__force __u32)(cpu_to_be32(*data));
+	}
+	return 0;
+}
+
+/**
+ *	t4_write_flash - write up to a page of data to the serial flash
+ *	@adapter: the adapter
+ *	@addr: the start address to write
+ *	@n: length of data to write in bytes
+ *	@data: the data to write
+ *	@byte_oriented: whether to store data as bytes or as words
+ *
+ *	Writes up to a page of data (256 bytes) to the serial flash starting
+ *	at the given address.  All the data must be written to the same page.
+ *	If @byte_oriented is set the write data is stored as byte stream
+ *	(i.e. matches what on disk), otherwise in big-endian.
+ */
+static int t4_write_flash(struct adapter *adapter, unsigned int addr,
+			  unsigned int n, const u8 *data, bool byte_oriented)
+{
+	unsigned int i, c, left, val, offset = addr & 0xff;
+	u32 buf[64];
+	int ret;
+
+	if (addr >= adapter->params.sf_size || offset + n > SF_PAGE_SIZE)
+		return -EINVAL;
+
+	val = swab32(addr) | SF_PROG_PAGE;
+
+	if ((ret = sf1_write(adapter, 1, 0, 1, SF_WR_ENABLE)) != 0 ||
+	    (ret = sf1_write(adapter, 4, 1, 1, val)) != 0)
+		goto unlock;
+
+	for (left = n; left; left -= c, data += c) {
+		c = min(left, 4U);
+		for (val = 0, i = 0; i < c; ++i) {
+			if (byte_oriented)
+				val = (val << 8) + data[i];
+			else
+				val = (val << 8) + data[c - i - 1];
+		}
+
+		ret = sf1_write(adapter, c, c != left, 1, val);
+		if (ret)
+			goto unlock;
+	}
+	ret = flash_wait_op(adapter, 8, 1);
+	if (ret)
+		goto unlock;
+
+	t4_write_reg(adapter, SF_OP_A, 0);    /* unlock SF */
+
+	/* Read the page to verify the write succeeded */
+	ret = t4_read_flash(adapter, addr & ~0xff, ARRAY_SIZE(buf), buf,
+			    byte_oriented);
+	if (ret)
+		return ret;
+
+	if (memcmp(data - n, (u8 *)buf + offset, n)) {
+		dev_err(adapter->pdev_dev,
+			"failed to correctly write the flash page at %#x\n",
+			addr);
+		return -EIO;
+	}
+	return 0;
+
+unlock:
+	t4_write_reg(adapter, SF_OP_A, 0);    /* unlock SF */
+	return ret;
+}
+
+/**
+ *	t4_get_fw_version - read the firmware version
+ *	@adapter: the adapter
+ *	@vers: where to place the version
+ *
+ *	Reads the FW version from flash.
+ */
+int t4_get_fw_version(struct adapter *adapter, u32 *vers)
+{
+	return t4_read_flash(adapter, FLASH_FW_START +
+			     offsetof(struct fw_hdr, fw_ver), 1,
+			     vers, 0);
+}
+
+/**
+ *	t4_get_bs_version - read the firmware bootstrap version
+ *	@adapter: the adapter
+ *	@vers: where to place the version
+ *
+ *	Reads the FW Bootstrap version from flash.
+ */
+int t4_get_bs_version(struct adapter *adapter, u32 *vers)
+{
+	return t4_read_flash(adapter, FLASH_FWBOOTSTRAP_START +
+			     offsetof(struct fw_hdr, fw_ver), 1,
+			     vers, 0);
+}
+
+/**
+ *	t4_get_tp_version - read the TP microcode version
+ *	@adapter: the adapter
+ *	@vers: where to place the version
+ *
+ *	Reads the TP microcode version from flash.
+ */
+int t4_get_tp_version(struct adapter *adapter, u32 *vers)
+{
+	return t4_read_flash(adapter, FLASH_FW_START +
+			     offsetof(struct fw_hdr, tp_microcode_ver),
+			     1, vers, 0);
+}
+
+/**
+ *	t4_get_exprom_version - return the Expansion ROM version (if any)
+ *	@adap: the adapter
+ *	@vers: where to place the version
+ *
+ *	Reads the Expansion ROM header from FLASH and returns the version
+ *	number (if present) through the @vers return value pointer.  We return
+ *	this in the Firmware Version Format since it's convenient.  Return
+ *	0 on success, -ENOENT if no Expansion ROM is present.
+ */
+int t4_get_exprom_version(struct adapter *adap, u32 *vers)
+{
+	struct exprom_header {
+		unsigned char hdr_arr[16];	/* must start with 0x55aa */
+		unsigned char hdr_ver[4];	/* Expansion ROM version */
+	} *hdr;
+	u32 exprom_header_buf[DIV_ROUND_UP(sizeof(struct exprom_header),
+					   sizeof(u32))];
+	int ret;
+
+	ret = t4_read_flash(adap, FLASH_EXP_ROM_START,
+			    ARRAY_SIZE(exprom_header_buf), exprom_header_buf,
+			    0);
+	if (ret)
+		return ret;
+
+	hdr = (struct exprom_header *)exprom_header_buf;
+	if (hdr->hdr_arr[0] != 0x55 || hdr->hdr_arr[1] != 0xaa)
+		return -ENOENT;
+
+	*vers = (FW_HDR_FW_VER_MAJOR_V(hdr->hdr_ver[0]) |
+		 FW_HDR_FW_VER_MINOR_V(hdr->hdr_ver[1]) |
+		 FW_HDR_FW_VER_MICRO_V(hdr->hdr_ver[2]) |
+		 FW_HDR_FW_VER_BUILD_V(hdr->hdr_ver[3]));
+	return 0;
+}
+
+/**
+ *      t4_get_vpd_version - return the VPD version
+ *      @adapter: the adapter
+ *      @vers: where to place the version
+ *
+ *      Reads the VPD via the Firmware interface (thus this can only be called
+ *      once we're ready to issue Firmware commands).  The format of the
+ *      VPD version is adapter specific.  Returns 0 on success, an error on
+ *      failure.
+ *
+ *      Note that early versions of the Firmware didn't include the ability
+ *      to retrieve the VPD version, so we zero-out the return-value parameter
+ *      in that case to avoid leaving it with garbage in it.
+ *
+ *      Also note that the Firmware will return its cached copy of the VPD
+ *      Revision ID, not the actual Revision ID as written in the Serial
+ *      EEPROM.  This is only an issue if a new VPD has been written and the
+ *      Firmware/Chip haven't yet gone through a RESET sequence.  So it's best
+ *      to defer calling this routine till after a FW_RESET_CMD has been issued
+ *      if the Host Driver will be performing a full adapter initialization.
+ */
+int t4_get_vpd_version(struct adapter *adapter, u32 *vers)
+{
+	u32 vpdrev_param;
+	int ret;
+
+	vpdrev_param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+			FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_VPDREV));
+	ret = t4_query_params(adapter, adapter->mbox, adapter->pf, 0,
+			      1, &vpdrev_param, vers);
+	if (ret)
+		*vers = 0;
+	return ret;
+}
+
+/**
+ *      t4_get_scfg_version - return the Serial Configuration version
+ *      @adapter: the adapter
+ *      @vers: where to place the version
+ *
+ *      Reads the Serial Configuration Version via the Firmware interface
+ *      (thus this can only be called once we're ready to issue Firmware
+ *      commands).  The format of the Serial Configuration version is
+ *      adapter specific.  Returns 0 on success, an error on failure.
+ *
+ *      Note that early versions of the Firmware didn't include the ability
+ *      to retrieve the Serial Configuration version, so we zero-out the
+ *      return-value parameter in that case to avoid leaving it with
+ *      garbage in it.
+ *
+ *      Also note that the Firmware will return its cached copy of the Serial
+ *      Initialization Revision ID, not the actual Revision ID as written in
+ *      the Serial EEPROM.  This is only an issue if a new VPD has been written
+ *      and the Firmware/Chip haven't yet gone through a RESET sequence.  So
+ *      it's best to defer calling this routine till after a FW_RESET_CMD has
+ *      been issued if the Host Driver will be performing a full adapter
+ *      initialization.
+ */
+int t4_get_scfg_version(struct adapter *adapter, u32 *vers)
+{
+	u32 scfgrev_param;
+	int ret;
+
+	scfgrev_param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+			 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_SCFGREV));
+	ret = t4_query_params(adapter, adapter->mbox, adapter->pf, 0,
+			      1, &scfgrev_param, vers);
+	if (ret)
+		*vers = 0;
+	return ret;
+}
+
+/**
+ *      t4_get_version_info - extract various chip/firmware version information
+ *      @adapter: the adapter
+ *
+ *      Reads various chip/firmware version numbers and stores them into the
+ *      adapter Adapter Parameters structure.  If any of the efforts fails
+ *      the first failure will be returned, but all of the version numbers
+ *      will be read.
+ */
+int t4_get_version_info(struct adapter *adapter)
+{
+	int ret = 0;
+
+	#define FIRST_RET(__getvinfo) \
+	do { \
+		int __ret = __getvinfo; \
+		if (__ret && !ret) \
+			ret = __ret; \
+	} while (0)
+
+	FIRST_RET(t4_get_fw_version(adapter, &adapter->params.fw_vers));
+	FIRST_RET(t4_get_bs_version(adapter, &adapter->params.bs_vers));
+	FIRST_RET(t4_get_tp_version(adapter, &adapter->params.tp_vers));
+	FIRST_RET(t4_get_exprom_version(adapter, &adapter->params.er_vers));
+	FIRST_RET(t4_get_scfg_version(adapter, &adapter->params.scfg_vers));
+	FIRST_RET(t4_get_vpd_version(adapter, &adapter->params.vpd_vers));
+
+	#undef FIRST_RET
+	return ret;
+}
+
+/**
+ *      t4_dump_version_info - dump all of the adapter configuration IDs
+ *      @adapter: the adapter
+ *
+ *      Dumps all of the various bits of adapter configuration version/revision
+ *      IDs information.  This is typically called at some point after
+ *      t4_get_version_info() has been called.
+ */
+void t4_dump_version_info(struct adapter *adapter)
+{
+	/* Device information */
+	dev_info(adapter->pdev_dev, "Chelsio %s rev %d\n",
+		 adapter->params.vpd.id,
+		 CHELSIO_CHIP_RELEASE(adapter->params.chip));
+	dev_info(adapter->pdev_dev, "S/N: %s, P/N: %s\n",
+		 adapter->params.vpd.sn, adapter->params.vpd.pn);
+
+	/* Firmware Version */
+	if (!adapter->params.fw_vers)
+		dev_warn(adapter->pdev_dev, "No firmware loaded\n");
+	else
+		dev_info(adapter->pdev_dev, "Firmware version: %u.%u.%u.%u\n",
+			 FW_HDR_FW_VER_MAJOR_G(adapter->params.fw_vers),
+			 FW_HDR_FW_VER_MINOR_G(adapter->params.fw_vers),
+			 FW_HDR_FW_VER_MICRO_G(adapter->params.fw_vers),
+			 FW_HDR_FW_VER_BUILD_G(adapter->params.fw_vers));
+
+	/* Bootstrap Firmware Version. (Some adapters don't have Bootstrap
+	 * Firmware, so dev_info() is more appropriate here.)
+	 */
+	if (!adapter->params.bs_vers)
+		dev_info(adapter->pdev_dev, "No bootstrap loaded\n");
+	else
+		dev_info(adapter->pdev_dev, "Bootstrap version: %u.%u.%u.%u\n",
+			 FW_HDR_FW_VER_MAJOR_G(adapter->params.bs_vers),
+			 FW_HDR_FW_VER_MINOR_G(adapter->params.bs_vers),
+			 FW_HDR_FW_VER_MICRO_G(adapter->params.bs_vers),
+			 FW_HDR_FW_VER_BUILD_G(adapter->params.bs_vers));
+
+	/* TP Microcode Version */
+	if (!adapter->params.tp_vers)
+		dev_warn(adapter->pdev_dev, "No TP Microcode loaded\n");
+	else
+		dev_info(adapter->pdev_dev,
+			 "TP Microcode version: %u.%u.%u.%u\n",
+			 FW_HDR_FW_VER_MAJOR_G(adapter->params.tp_vers),
+			 FW_HDR_FW_VER_MINOR_G(adapter->params.tp_vers),
+			 FW_HDR_FW_VER_MICRO_G(adapter->params.tp_vers),
+			 FW_HDR_FW_VER_BUILD_G(adapter->params.tp_vers));
+
+	/* Expansion ROM version */
+	if (!adapter->params.er_vers)
+		dev_info(adapter->pdev_dev, "No Expansion ROM loaded\n");
+	else
+		dev_info(adapter->pdev_dev,
+			 "Expansion ROM version: %u.%u.%u.%u\n",
+			 FW_HDR_FW_VER_MAJOR_G(adapter->params.er_vers),
+			 FW_HDR_FW_VER_MINOR_G(adapter->params.er_vers),
+			 FW_HDR_FW_VER_MICRO_G(adapter->params.er_vers),
+			 FW_HDR_FW_VER_BUILD_G(adapter->params.er_vers));
+
+	/* Serial Configuration version */
+	dev_info(adapter->pdev_dev, "Serial Configuration version: %#x\n",
+		 adapter->params.scfg_vers);
+
+	/* VPD Version */
+	dev_info(adapter->pdev_dev, "VPD version: %#x\n",
+		 adapter->params.vpd_vers);
+}
+
+/**
+ *	t4_check_fw_version - check if the FW is supported with this driver
+ *	@adap: the adapter
+ *
+ *	Checks if an adapter's FW is compatible with the driver.  Returns 0
+ *	if there's exact match, a negative error if the version could not be
+ *	read or there's a major version mismatch
+ */
+int t4_check_fw_version(struct adapter *adap)
+{
+	int i, ret, major, minor, micro;
+	int exp_major, exp_minor, exp_micro;
+	unsigned int chip_version = CHELSIO_CHIP_VERSION(adap->params.chip);
+
+	ret = t4_get_fw_version(adap, &adap->params.fw_vers);
+	/* Try multiple times before returning error */
+	for (i = 0; (ret == -EBUSY || ret == -EAGAIN) && i < 3; i++)
+		ret = t4_get_fw_version(adap, &adap->params.fw_vers);
+
+	if (ret)
+		return ret;
+
+	major = FW_HDR_FW_VER_MAJOR_G(adap->params.fw_vers);
+	minor = FW_HDR_FW_VER_MINOR_G(adap->params.fw_vers);
+	micro = FW_HDR_FW_VER_MICRO_G(adap->params.fw_vers);
+
+	switch (chip_version) {
+	case CHELSIO_T4:
+		exp_major = T4FW_MIN_VERSION_MAJOR;
+		exp_minor = T4FW_MIN_VERSION_MINOR;
+		exp_micro = T4FW_MIN_VERSION_MICRO;
+		break;
+	case CHELSIO_T5:
+		exp_major = T5FW_MIN_VERSION_MAJOR;
+		exp_minor = T5FW_MIN_VERSION_MINOR;
+		exp_micro = T5FW_MIN_VERSION_MICRO;
+		break;
+	case CHELSIO_T6:
+		exp_major = T6FW_MIN_VERSION_MAJOR;
+		exp_minor = T6FW_MIN_VERSION_MINOR;
+		exp_micro = T6FW_MIN_VERSION_MICRO;
+		break;
+	default:
+		dev_err(adap->pdev_dev, "Unsupported chip type, %x\n",
+			adap->chip);
+		return -EINVAL;
+	}
+
+	if (major < exp_major || (major == exp_major && minor < exp_minor) ||
+	    (major == exp_major && minor == exp_minor && micro < exp_micro)) {
+		dev_err(adap->pdev_dev,
+			"Card has firmware version %u.%u.%u, minimum "
+			"supported firmware is %u.%u.%u.\n", major, minor,
+			micro, exp_major, exp_minor, exp_micro);
+		return -EFAULT;
+	}
+	return 0;
+}
+
+/* Is the given firmware API compatible with the one the driver was compiled
+ * with?
+ */
+static int fw_compatible(const struct fw_hdr *hdr1, const struct fw_hdr *hdr2)
+{
+
+	/* short circuit if it's the exact same firmware version */
+	if (hdr1->chip == hdr2->chip && hdr1->fw_ver == hdr2->fw_ver)
+		return 1;
+
+#define SAME_INTF(x) (hdr1->intfver_##x == hdr2->intfver_##x)
+	if (hdr1->chip == hdr2->chip && SAME_INTF(nic) && SAME_INTF(vnic) &&
+	    SAME_INTF(ri) && SAME_INTF(iscsi) && SAME_INTF(fcoe))
+		return 1;
+#undef SAME_INTF
+
+	return 0;
+}
+
+/* The firmware in the filesystem is usable, but should it be installed?
+ * This routine explains itself in detail if it indicates the filesystem
+ * firmware should be installed.
+ */
+static int should_install_fs_fw(struct adapter *adap, int card_fw_usable,
+				int k, int c)
+{
+	const char *reason;
+
+	if (!card_fw_usable) {
+		reason = "incompatible or unusable";
+		goto install;
+	}
+
+	if (k > c) {
+		reason = "older than the version supported with this driver";
+		goto install;
+	}
+
+	return 0;
+
+install:
+	dev_err(adap->pdev_dev, "firmware on card (%u.%u.%u.%u) is %s, "
+		"installing firmware %u.%u.%u.%u on card.\n",
+		FW_HDR_FW_VER_MAJOR_G(c), FW_HDR_FW_VER_MINOR_G(c),
+		FW_HDR_FW_VER_MICRO_G(c), FW_HDR_FW_VER_BUILD_G(c), reason,
+		FW_HDR_FW_VER_MAJOR_G(k), FW_HDR_FW_VER_MINOR_G(k),
+		FW_HDR_FW_VER_MICRO_G(k), FW_HDR_FW_VER_BUILD_G(k));
+
+	return 1;
+}
+
+int t4_prep_fw(struct adapter *adap, struct fw_info *fw_info,
+	       const u8 *fw_data, unsigned int fw_size,
+	       struct fw_hdr *card_fw, enum dev_state state,
+	       int *reset)
+{
+	int ret, card_fw_usable, fs_fw_usable;
+	const struct fw_hdr *fs_fw;
+	const struct fw_hdr *drv_fw;
+
+	drv_fw = &fw_info->fw_hdr;
+
+	/* Read the header of the firmware on the card */
+	ret = t4_read_flash(adap, FLASH_FW_START,
+			    sizeof(*card_fw) / sizeof(uint32_t),
+			    (uint32_t *)card_fw, 1);
+	if (ret == 0) {
+		card_fw_usable = fw_compatible(drv_fw, (const void *)card_fw);
+	} else {
+		dev_err(adap->pdev_dev,
+			"Unable to read card's firmware header: %d\n", ret);
+		card_fw_usable = 0;
+	}
+
+	if (fw_data != NULL) {
+		fs_fw = (const void *)fw_data;
+		fs_fw_usable = fw_compatible(drv_fw, fs_fw);
+	} else {
+		fs_fw = NULL;
+		fs_fw_usable = 0;
+	}
+
+	if (card_fw_usable && card_fw->fw_ver == drv_fw->fw_ver &&
+	    (!fs_fw_usable || fs_fw->fw_ver == drv_fw->fw_ver)) {
+		/* Common case: the firmware on the card is an exact match and
+		 * the filesystem one is an exact match too, or the filesystem
+		 * one is absent/incompatible.
+		 */
+	} else if (fs_fw_usable && state == DEV_STATE_UNINIT &&
+		   should_install_fs_fw(adap, card_fw_usable,
+					be32_to_cpu(fs_fw->fw_ver),
+					be32_to_cpu(card_fw->fw_ver))) {
+		ret = t4_fw_upgrade(adap, adap->mbox, fw_data,
+				    fw_size, 0);
+		if (ret != 0) {
+			dev_err(adap->pdev_dev,
+				"failed to install firmware: %d\n", ret);
+			goto bye;
+		}
+
+		/* Installed successfully, update the cached header too. */
+		*card_fw = *fs_fw;
+		card_fw_usable = 1;
+		*reset = 0;	/* already reset as part of load_fw */
+	}
+
+	if (!card_fw_usable) {
+		uint32_t d, c, k;
+
+		d = be32_to_cpu(drv_fw->fw_ver);
+		c = be32_to_cpu(card_fw->fw_ver);
+		k = fs_fw ? be32_to_cpu(fs_fw->fw_ver) : 0;
+
+		dev_err(adap->pdev_dev, "Cannot find a usable firmware: "
+			"chip state %d, "
+			"driver compiled with %d.%d.%d.%d, "
+			"card has %d.%d.%d.%d, filesystem has %d.%d.%d.%d\n",
+			state,
+			FW_HDR_FW_VER_MAJOR_G(d), FW_HDR_FW_VER_MINOR_G(d),
+			FW_HDR_FW_VER_MICRO_G(d), FW_HDR_FW_VER_BUILD_G(d),
+			FW_HDR_FW_VER_MAJOR_G(c), FW_HDR_FW_VER_MINOR_G(c),
+			FW_HDR_FW_VER_MICRO_G(c), FW_HDR_FW_VER_BUILD_G(c),
+			FW_HDR_FW_VER_MAJOR_G(k), FW_HDR_FW_VER_MINOR_G(k),
+			FW_HDR_FW_VER_MICRO_G(k), FW_HDR_FW_VER_BUILD_G(k));
+		ret = -EINVAL;
+		goto bye;
+	}
+
+	/* We're using whatever's on the card and it's known to be good. */
+	adap->params.fw_vers = be32_to_cpu(card_fw->fw_ver);
+	adap->params.tp_vers = be32_to_cpu(card_fw->tp_microcode_ver);
+
+bye:
+	return ret;
+}
+
+/**
+ *	t4_flash_erase_sectors - erase a range of flash sectors
+ *	@adapter: the adapter
+ *	@start: the first sector to erase
+ *	@end: the last sector to erase
+ *
+ *	Erases the sectors in the given inclusive range.
+ */
+static int t4_flash_erase_sectors(struct adapter *adapter, int start, int end)
+{
+	int ret = 0;
+
+	if (end >= adapter->params.sf_nsec)
+		return -EINVAL;
+
+	while (start <= end) {
+		if ((ret = sf1_write(adapter, 1, 0, 1, SF_WR_ENABLE)) != 0 ||
+		    (ret = sf1_write(adapter, 4, 0, 1,
+				     SF_ERASE_SECTOR | (start << 8))) != 0 ||
+		    (ret = flash_wait_op(adapter, 14, 500)) != 0) {
+			dev_err(adapter->pdev_dev,
+				"erase of flash sector %d failed, error %d\n",
+				start, ret);
+			break;
+		}
+		start++;
+	}
+	t4_write_reg(adapter, SF_OP_A, 0);    /* unlock SF */
+	return ret;
+}
+
+/**
+ *	t4_flash_cfg_addr - return the address of the flash configuration file
+ *	@adapter: the adapter
+ *
+ *	Return the address within the flash where the Firmware Configuration
+ *	File is stored.
+ */
+unsigned int t4_flash_cfg_addr(struct adapter *adapter)
+{
+	if (adapter->params.sf_size == 0x100000)
+		return FLASH_FPGA_CFG_START;
+	else
+		return FLASH_CFG_START;
+}
+
+/* Return TRUE if the specified firmware matches the adapter.  I.e. T4
+ * firmware for T4 adapters, T5 firmware for T5 adapters, etc.  We go ahead
+ * and emit an error message for mismatched firmware to save our caller the
+ * effort ...
+ */
+static bool t4_fw_matches_chip(const struct adapter *adap,
+			       const struct fw_hdr *hdr)
+{
+	/* The expression below will return FALSE for any unsupported adapter
+	 * which will keep us "honest" in the future ...
+	 */
+	if ((is_t4(adap->params.chip) && hdr->chip == FW_HDR_CHIP_T4) ||
+	    (is_t5(adap->params.chip) && hdr->chip == FW_HDR_CHIP_T5) ||
+	    (is_t6(adap->params.chip) && hdr->chip == FW_HDR_CHIP_T6))
+		return true;
+
+	dev_err(adap->pdev_dev,
+		"FW image (%d) is not suitable for this adapter (%d)\n",
+		hdr->chip, CHELSIO_CHIP_VERSION(adap->params.chip));
+	return false;
+}
+
+/**
+ *	t4_load_fw - download firmware
+ *	@adap: the adapter
+ *	@fw_data: the firmware image to write
+ *	@size: image size
+ *
+ *	Write the supplied firmware image to the card's serial flash.
+ */
+int t4_load_fw(struct adapter *adap, const u8 *fw_data, unsigned int size)
+{
+	u32 csum;
+	int ret, addr;
+	unsigned int i;
+	u8 first_page[SF_PAGE_SIZE];
+	const __be32 *p = (const __be32 *)fw_data;
+	const struct fw_hdr *hdr = (const struct fw_hdr *)fw_data;
+	unsigned int sf_sec_size = adap->params.sf_size / adap->params.sf_nsec;
+	unsigned int fw_start_sec = FLASH_FW_START_SEC;
+	unsigned int fw_size = FLASH_FW_MAX_SIZE;
+	unsigned int fw_start = FLASH_FW_START;
+
+	if (!size) {
+		dev_err(adap->pdev_dev, "FW image has no data\n");
+		return -EINVAL;
+	}
+	if (size & 511) {
+		dev_err(adap->pdev_dev,
+			"FW image size not multiple of 512 bytes\n");
+		return -EINVAL;
+	}
+	if ((unsigned int)be16_to_cpu(hdr->len512) * 512 != size) {
+		dev_err(adap->pdev_dev,
+			"FW image size differs from size in FW header\n");
+		return -EINVAL;
+	}
+	if (size > fw_size) {
+		dev_err(adap->pdev_dev, "FW image too large, max is %u bytes\n",
+			fw_size);
+		return -EFBIG;
+	}
+	if (!t4_fw_matches_chip(adap, hdr))
+		return -EINVAL;
+
+	for (csum = 0, i = 0; i < size / sizeof(csum); i++)
+		csum += be32_to_cpu(p[i]);
+
+	if (csum != 0xffffffff) {
+		dev_err(adap->pdev_dev,
+			"corrupted firmware image, checksum %#x\n", csum);
+		return -EINVAL;
+	}
+
+	i = DIV_ROUND_UP(size, sf_sec_size);        /* # of sectors spanned */
+	ret = t4_flash_erase_sectors(adap, fw_start_sec, fw_start_sec + i - 1);
+	if (ret)
+		goto out;
+
+	/*
+	 * We write the correct version at the end so the driver can see a bad
+	 * version if the FW write fails.  Start by writing a copy of the
+	 * first page with a bad version.
+	 */
+	memcpy(first_page, fw_data, SF_PAGE_SIZE);
+	((struct fw_hdr *)first_page)->fw_ver = cpu_to_be32(0xffffffff);
+	ret = t4_write_flash(adap, fw_start, SF_PAGE_SIZE, first_page, true);
+	if (ret)
+		goto out;
+
+	addr = fw_start;
+	for (size -= SF_PAGE_SIZE; size; size -= SF_PAGE_SIZE) {
+		addr += SF_PAGE_SIZE;
+		fw_data += SF_PAGE_SIZE;
+		ret = t4_write_flash(adap, addr, SF_PAGE_SIZE, fw_data, true);
+		if (ret)
+			goto out;
+	}
+
+	ret = t4_write_flash(adap, fw_start + offsetof(struct fw_hdr, fw_ver),
+			     sizeof(hdr->fw_ver), (const u8 *)&hdr->fw_ver,
+			     true);
+out:
+	if (ret)
+		dev_err(adap->pdev_dev, "firmware download failed, error %d\n",
+			ret);
+	else
+		ret = t4_get_fw_version(adap, &adap->params.fw_vers);
+	return ret;
+}
+
+/**
+ *	t4_phy_fw_ver - return current PHY firmware version
+ *	@adap: the adapter
+ *	@phy_fw_ver: return value buffer for PHY firmware version
+ *
+ *	Returns the current version of external PHY firmware on the
+ *	adapter.
+ */
+int t4_phy_fw_ver(struct adapter *adap, int *phy_fw_ver)
+{
+	u32 param, val;
+	int ret;
+
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_PHYFW) |
+		 FW_PARAMS_PARAM_Y_V(adap->params.portvec) |
+		 FW_PARAMS_PARAM_Z_V(FW_PARAMS_PARAM_DEV_PHYFW_VERSION));
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 1,
+			      &param, &val);
+	if (ret)
+		return ret;
+	*phy_fw_ver = val;
+	return 0;
+}
+
+/**
+ *	t4_load_phy_fw - download port PHY firmware
+ *	@adap: the adapter
+ *	@win: the PCI-E Memory Window index to use for t4_memory_rw()
+ *	@phy_fw_version: function to check PHY firmware versions
+ *	@phy_fw_data: the PHY firmware image to write
+ *	@phy_fw_size: image size
+ *
+ *	Transfer the specified PHY firmware to the adapter.  If a non-NULL
+ *	@phy_fw_version is supplied, then it will be used to determine if
+ *	it's necessary to perform the transfer by comparing the version
+ *	of any existing adapter PHY firmware with that of the passed in
+ *	PHY firmware image.
+ *
+ *	A negative error number will be returned if an error occurs.  If
+ *	version number support is available and there's no need to upgrade
+ *	the firmware, 0 will be returned.  If firmware is successfully
+ *	transferred to the adapter, 1 will be returned.
+ *
+ *	NOTE: some adapters only have local RAM to store the PHY firmware.  As
+ *	a result, a RESET of the adapter would cause that RAM to lose its
+ *	contents.  Thus, loading PHY firmware on such adapters must happen
+ *	after any FW_RESET_CMDs ...
+ */
+int t4_load_phy_fw(struct adapter *adap, int win,
+		   int (*phy_fw_version)(const u8 *, size_t),
+		   const u8 *phy_fw_data, size_t phy_fw_size)
+{
+	int cur_phy_fw_ver = 0, new_phy_fw_vers = 0;
+	unsigned long mtype = 0, maddr = 0;
+	u32 param, val;
+	int ret;
+
+	/* If we have version number support, then check to see if the adapter
+	 * already has up-to-date PHY firmware loaded.
+	 */
+	if (phy_fw_version) {
+		new_phy_fw_vers = phy_fw_version(phy_fw_data, phy_fw_size);
+		ret = t4_phy_fw_ver(adap, &cur_phy_fw_ver);
+		if (ret < 0)
+			return ret;
+
+		if (cur_phy_fw_ver >= new_phy_fw_vers) {
+			CH_WARN(adap, "PHY Firmware already up-to-date, "
+				"version %#x\n", cur_phy_fw_ver);
+			return 0;
+		}
+	}
+
+	/* Ask the firmware where it wants us to copy the PHY firmware image.
+	 * The size of the file requires a special version of the READ command
+	 * which will pass the file size via the values field in PARAMS_CMD and
+	 * retrieve the return value from firmware and place it in the same
+	 * buffer values
+	 */
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_PHYFW) |
+		 FW_PARAMS_PARAM_Y_V(adap->params.portvec) |
+		 FW_PARAMS_PARAM_Z_V(FW_PARAMS_PARAM_DEV_PHYFW_DOWNLOAD));
+	val = phy_fw_size;
+	ret = t4_query_params_rw(adap, adap->mbox, adap->pf, 0, 1,
+				 &param, &val, 1, true);
+	if (ret < 0)
+		return ret;
+	mtype = val >> 8;
+	maddr = (val & 0xff) << 16;
+
+	/* Copy the supplied PHY Firmware image to the adapter memory location
+	 * allocated by the adapter firmware.
+	 */
+	spin_lock_bh(&adap->win0_lock);
+	ret = t4_memory_rw(adap, win, mtype, maddr,
+			   phy_fw_size, (__be32 *)phy_fw_data,
+			   T4_MEMORY_WRITE);
+	spin_unlock_bh(&adap->win0_lock);
+	if (ret)
+		return ret;
+
+	/* Tell the firmware that the PHY firmware image has been written to
+	 * RAM and it can now start copying it over to the PHYs.  The chip
+	 * firmware will RESET the affected PHYs as part of this operation
+	 * leaving them running the new PHY firmware image.
+	 */
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_PHYFW) |
+		 FW_PARAMS_PARAM_Y_V(adap->params.portvec) |
+		 FW_PARAMS_PARAM_Z_V(FW_PARAMS_PARAM_DEV_PHYFW_DOWNLOAD));
+	ret = t4_set_params_timeout(adap, adap->mbox, adap->pf, 0, 1,
+				    &param, &val, 30000);
+	if (ret)
+		return ret;
+
+	/* If we have version number support, then check to see that the new
+	 * firmware got loaded properly.
+	 */
+	if (phy_fw_version) {
+		ret = t4_phy_fw_ver(adap, &cur_phy_fw_ver);
+		if (ret < 0)
+			return ret;
+
+		if (cur_phy_fw_ver != new_phy_fw_vers) {
+			CH_WARN(adap, "PHY Firmware did not update: "
+				"version on adapter %#x, "
+				"version flashed %#x\n",
+				cur_phy_fw_ver, new_phy_fw_vers);
+			return -ENXIO;
+		}
+	}
+
+	return 1;
+}
+
+/**
+ *	t4_fwcache - firmware cache operation
+ *	@adap: the adapter
+ *	@op  : the operation (flush or flush and invalidate)
+ */
+int t4_fwcache(struct adapter *adap, enum fw_params_param_dev_fwcache op)
+{
+	struct fw_params_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn =
+		cpu_to_be32(FW_CMD_OP_V(FW_PARAMS_CMD) |
+			    FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+			    FW_PARAMS_CMD_PFN_V(adap->pf) |
+			    FW_PARAMS_CMD_VFN_V(0));
+	c.retval_len16 = cpu_to_be32(FW_LEN16(c));
+	c.param[0].mnem =
+		cpu_to_be32(FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+			    FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_FWCACHE));
+	c.param[0].val = cpu_to_be32(op);
+
+	return t4_wr_mbox(adap, adap->mbox, &c, sizeof(c), NULL);
+}
+
+void t4_cim_read_pif_la(struct adapter *adap, u32 *pif_req, u32 *pif_rsp,
+			unsigned int *pif_req_wrptr,
+			unsigned int *pif_rsp_wrptr)
+{
+	int i, j;
+	u32 cfg, val, req, rsp;
+
+	cfg = t4_read_reg(adap, CIM_DEBUGCFG_A);
+	if (cfg & LADBGEN_F)
+		t4_write_reg(adap, CIM_DEBUGCFG_A, cfg ^ LADBGEN_F);
+
+	val = t4_read_reg(adap, CIM_DEBUGSTS_A);
+	req = POLADBGWRPTR_G(val);
+	rsp = PILADBGWRPTR_G(val);
+	if (pif_req_wrptr)
+		*pif_req_wrptr = req;
+	if (pif_rsp_wrptr)
+		*pif_rsp_wrptr = rsp;
+
+	for (i = 0; i < CIM_PIFLA_SIZE; i++) {
+		for (j = 0; j < 6; j++) {
+			t4_write_reg(adap, CIM_DEBUGCFG_A, POLADBGRDPTR_V(req) |
+				     PILADBGRDPTR_V(rsp));
+			*pif_req++ = t4_read_reg(adap, CIM_PO_LA_DEBUGDATA_A);
+			*pif_rsp++ = t4_read_reg(adap, CIM_PI_LA_DEBUGDATA_A);
+			req++;
+			rsp++;
+		}
+		req = (req + 2) & POLADBGRDPTR_M;
+		rsp = (rsp + 2) & PILADBGRDPTR_M;
+	}
+	t4_write_reg(adap, CIM_DEBUGCFG_A, cfg);
+}
+
+void t4_cim_read_ma_la(struct adapter *adap, u32 *ma_req, u32 *ma_rsp)
+{
+	u32 cfg;
+	int i, j, idx;
+
+	cfg = t4_read_reg(adap, CIM_DEBUGCFG_A);
+	if (cfg & LADBGEN_F)
+		t4_write_reg(adap, CIM_DEBUGCFG_A, cfg ^ LADBGEN_F);
+
+	for (i = 0; i < CIM_MALA_SIZE; i++) {
+		for (j = 0; j < 5; j++) {
+			idx = 8 * i + j;
+			t4_write_reg(adap, CIM_DEBUGCFG_A, POLADBGRDPTR_V(idx) |
+				     PILADBGRDPTR_V(idx));
+			*ma_req++ = t4_read_reg(adap, CIM_PO_LA_MADEBUGDATA_A);
+			*ma_rsp++ = t4_read_reg(adap, CIM_PI_LA_MADEBUGDATA_A);
+		}
+	}
+	t4_write_reg(adap, CIM_DEBUGCFG_A, cfg);
+}
+
+void t4_ulprx_read_la(struct adapter *adap, u32 *la_buf)
+{
+	unsigned int i, j;
+
+	for (i = 0; i < 8; i++) {
+		u32 *p = la_buf + i;
+
+		t4_write_reg(adap, ULP_RX_LA_CTL_A, i);
+		j = t4_read_reg(adap, ULP_RX_LA_WRPTR_A);
+		t4_write_reg(adap, ULP_RX_LA_RDPTR_A, j);
+		for (j = 0; j < ULPRX_LA_SIZE; j++, p += 8)
+			*p = t4_read_reg(adap, ULP_RX_LA_RDDATA_A);
+	}
+}
+
+/* The ADVERT_MASK is used to mask out all of the Advertised Firmware Port
+ * Capabilities which we control with separate controls -- see, for instance,
+ * Pause Frames and Forward Error Correction.  In order to determine what the
+ * full set of Advertised Port Capabilities are, the base Advertised Port
+ * Capabilities (masked by ADVERT_MASK) must be combined with the Advertised
+ * Port Capabilities associated with those other controls.  See
+ * t4_link_acaps() for how this is done.
+ */
+#define ADVERT_MASK (FW_PORT_CAP32_SPEED_V(FW_PORT_CAP32_SPEED_M) | \
+		     FW_PORT_CAP32_ANEG)
+
+/**
+ *	fwcaps16_to_caps32 - convert 16-bit Port Capabilities to 32-bits
+ *	@caps16: a 16-bit Port Capabilities value
+ *
+ *	Returns the equivalent 32-bit Port Capabilities value.
+ */
+static fw_port_cap32_t fwcaps16_to_caps32(fw_port_cap16_t caps16)
+{
+	fw_port_cap32_t caps32 = 0;
+
+	#define CAP16_TO_CAP32(__cap) \
+		do { \
+			if (caps16 & FW_PORT_CAP_##__cap) \
+				caps32 |= FW_PORT_CAP32_##__cap; \
+		} while (0)
+
+	CAP16_TO_CAP32(SPEED_100M);
+	CAP16_TO_CAP32(SPEED_1G);
+	CAP16_TO_CAP32(SPEED_25G);
+	CAP16_TO_CAP32(SPEED_10G);
+	CAP16_TO_CAP32(SPEED_40G);
+	CAP16_TO_CAP32(SPEED_100G);
+	CAP16_TO_CAP32(FC_RX);
+	CAP16_TO_CAP32(FC_TX);
+	CAP16_TO_CAP32(ANEG);
+	CAP16_TO_CAP32(FORCE_PAUSE);
+	CAP16_TO_CAP32(MDIAUTO);
+	CAP16_TO_CAP32(MDISTRAIGHT);
+	CAP16_TO_CAP32(FEC_RS);
+	CAP16_TO_CAP32(FEC_BASER_RS);
+	CAP16_TO_CAP32(802_3_PAUSE);
+	CAP16_TO_CAP32(802_3_ASM_DIR);
+
+	#undef CAP16_TO_CAP32
+
+	return caps32;
+}
+
+/**
+ *	fwcaps32_to_caps16 - convert 32-bit Port Capabilities to 16-bits
+ *	@caps32: a 32-bit Port Capabilities value
+ *
+ *	Returns the equivalent 16-bit Port Capabilities value.  Note that
+ *	not all 32-bit Port Capabilities can be represented in the 16-bit
+ *	Port Capabilities and some fields/values may not make it.
+ */
+static fw_port_cap16_t fwcaps32_to_caps16(fw_port_cap32_t caps32)
+{
+	fw_port_cap16_t caps16 = 0;
+
+	#define CAP32_TO_CAP16(__cap) \
+		do { \
+			if (caps32 & FW_PORT_CAP32_##__cap) \
+				caps16 |= FW_PORT_CAP_##__cap; \
+		} while (0)
+
+	CAP32_TO_CAP16(SPEED_100M);
+	CAP32_TO_CAP16(SPEED_1G);
+	CAP32_TO_CAP16(SPEED_10G);
+	CAP32_TO_CAP16(SPEED_25G);
+	CAP32_TO_CAP16(SPEED_40G);
+	CAP32_TO_CAP16(SPEED_100G);
+	CAP32_TO_CAP16(FC_RX);
+	CAP32_TO_CAP16(FC_TX);
+	CAP32_TO_CAP16(802_3_PAUSE);
+	CAP32_TO_CAP16(802_3_ASM_DIR);
+	CAP32_TO_CAP16(ANEG);
+	CAP32_TO_CAP16(FORCE_PAUSE);
+	CAP32_TO_CAP16(MDIAUTO);
+	CAP32_TO_CAP16(MDISTRAIGHT);
+	CAP32_TO_CAP16(FEC_RS);
+	CAP32_TO_CAP16(FEC_BASER_RS);
+
+	#undef CAP32_TO_CAP16
+
+	return caps16;
+}
+
+/* Translate Firmware Port Capabilities Pause specification to Common Code */
+static inline enum cc_pause fwcap_to_cc_pause(fw_port_cap32_t fw_pause)
+{
+	enum cc_pause cc_pause = 0;
+
+	if (fw_pause & FW_PORT_CAP32_FC_RX)
+		cc_pause |= PAUSE_RX;
+	if (fw_pause & FW_PORT_CAP32_FC_TX)
+		cc_pause |= PAUSE_TX;
+
+	return cc_pause;
+}
+
+/* Translate Common Code Pause specification into Firmware Port Capabilities */
+static inline fw_port_cap32_t cc_to_fwcap_pause(enum cc_pause cc_pause)
+{
+	/* Translate orthogonal RX/TX Pause Controls for L1 Configure
+	 * commands, etc.
+	 */
+	fw_port_cap32_t fw_pause = 0;
+
+	if (cc_pause & PAUSE_RX)
+		fw_pause |= FW_PORT_CAP32_FC_RX;
+	if (cc_pause & PAUSE_TX)
+		fw_pause |= FW_PORT_CAP32_FC_TX;
+	if (!(cc_pause & PAUSE_AUTONEG))
+		fw_pause |= FW_PORT_CAP32_FORCE_PAUSE;
+
+	/* Translate orthogonal Pause controls into IEEE 802.3 Pause,
+	 * Asymmetrical Pause for use in reporting to upper layer OS code, etc.
+	 * Note that these bits are ignored in L1 Configure commands.
+	 */
+	if (cc_pause & PAUSE_RX) {
+		if (cc_pause & PAUSE_TX)
+			fw_pause |= FW_PORT_CAP32_802_3_PAUSE;
+		else
+			fw_pause |= FW_PORT_CAP32_802_3_ASM_DIR |
+				    FW_PORT_CAP32_802_3_PAUSE;
+	} else if (cc_pause & PAUSE_TX) {
+		fw_pause |= FW_PORT_CAP32_802_3_ASM_DIR;
+	}
+
+	return fw_pause;
+}
+
+/* Translate Firmware Forward Error Correction specification to Common Code */
+static inline enum cc_fec fwcap_to_cc_fec(fw_port_cap32_t fw_fec)
+{
+	enum cc_fec cc_fec = 0;
+
+	if (fw_fec & FW_PORT_CAP32_FEC_RS)
+		cc_fec |= FEC_RS;
+	if (fw_fec & FW_PORT_CAP32_FEC_BASER_RS)
+		cc_fec |= FEC_BASER_RS;
+
+	return cc_fec;
+}
+
+/* Translate Common Code Forward Error Correction specification to Firmware */
+static inline fw_port_cap32_t cc_to_fwcap_fec(enum cc_fec cc_fec)
+{
+	fw_port_cap32_t fw_fec = 0;
+
+	if (cc_fec & FEC_RS)
+		fw_fec |= FW_PORT_CAP32_FEC_RS;
+	if (cc_fec & FEC_BASER_RS)
+		fw_fec |= FW_PORT_CAP32_FEC_BASER_RS;
+
+	return fw_fec;
+}
+
+/**
+ *	t4_link_acaps - compute Link Advertised Port Capabilities
+ *	@adapter: the adapter
+ *	@port: the Port ID
+ *	@lc: the Port's Link Configuration
+ *
+ *	Synthesize the Advertised Port Capabilities we'll be using based on
+ *	the base Advertised Port Capabilities (which have been filtered by
+ *	ADVERT_MASK) plus the individual controls for things like Pause
+ *	Frames, Forward Error Correction, MDI, etc.
+ */
+fw_port_cap32_t t4_link_acaps(struct adapter *adapter, unsigned int port,
+			      struct link_config *lc)
+{
+	fw_port_cap32_t fw_fc, fw_fec, acaps;
+	unsigned int fw_mdi;
+	char cc_fec;
+
+	fw_mdi = (FW_PORT_CAP32_MDI_V(FW_PORT_CAP32_MDI_AUTO) & lc->pcaps);
+
+	/* Convert driver coding of Pause Frame Flow Control settings into the
+	 * Firmware's API.
+	 */
+	fw_fc = cc_to_fwcap_pause(lc->requested_fc);
+
+	/* Convert Common Code Forward Error Control settings into the
+	 * Firmware's API.  If the current Requested FEC has "Automatic"
+	 * (IEEE 802.3) specified, then we use whatever the Firmware
+	 * sent us as part of its IEEE 802.3-based interpretation of
+	 * the Transceiver Module EPROM FEC parameters.  Otherwise we
+	 * use whatever is in the current Requested FEC settings.
+	 */
+	if (lc->requested_fec & FEC_AUTO)
+		cc_fec = fwcap_to_cc_fec(lc->def_acaps);
+	else
+		cc_fec = lc->requested_fec;
+	fw_fec = cc_to_fwcap_fec(cc_fec);
+
+	/* Figure out what our Requested Port Capabilities are going to be.
+	 * Note parallel structure in t4_handle_get_port_info() and
+	 * init_link_config().
+	 */
+	if (!(lc->pcaps & FW_PORT_CAP32_ANEG)) {
+		acaps = lc->acaps | fw_fc | fw_fec;
+		lc->fc = lc->requested_fc & ~PAUSE_AUTONEG;
+		lc->fec = cc_fec;
+	} else if (lc->autoneg == AUTONEG_DISABLE) {
+		acaps = lc->speed_caps | fw_fc | fw_fec | fw_mdi;
+		lc->fc = lc->requested_fc & ~PAUSE_AUTONEG;
+		lc->fec = cc_fec;
+	} else {
+		acaps = lc->acaps | fw_fc | fw_fec | fw_mdi;
+	}
+
+	/* Some Requested Port Capabilities are trivially wrong if they exceed
+	 * the Physical Port Capabilities.  We can check that here and provide
+	 * moderately useful feedback in the system log.
+	 *
+	 * Note that older Firmware doesn't have FW_PORT_CAP32_FORCE_PAUSE, so
+	 * we need to exclude this from this check in order to maintain
+	 * compatibility ...
+	 */
+	if ((acaps & ~lc->pcaps) & ~FW_PORT_CAP32_FORCE_PAUSE) {
+		dev_err(adapter->pdev_dev, "Requested Port Capabilities %#x exceed Physical Port Capabilities %#x\n",
+			acaps, lc->pcaps);
+		return -EINVAL;
+	}
+
+	return acaps;
+}
+
+/**
+ *	t4_link_l1cfg_core - apply link configuration to MAC/PHY
+ *	@adapter: the adapter
+ *	@mbox: the Firmware Mailbox to use
+ *	@port: the Port ID
+ *	@lc: the Port's Link Configuration
+ *	@sleep_ok: if true we may sleep while awaiting command completion
+ *	@timeout: time to wait for command to finish before timing out
+ *		(negative implies @sleep_ok=false)
+ *
+ *	Set up a port's MAC and PHY according to a desired link configuration.
+ *	- If the PHY can auto-negotiate first decide what to advertise, then
+ *	  enable/disable auto-negotiation as desired, and reset.
+ *	- If the PHY does not auto-negotiate just reset it.
+ *	- If auto-negotiation is off set the MAC to the proper speed/duplex/FC,
+ *	  otherwise do it later based on the outcome of auto-negotiation.
+ */
+int t4_link_l1cfg_core(struct adapter *adapter, unsigned int mbox,
+		       unsigned int port, struct link_config *lc,
+		       u8 sleep_ok, int timeout)
+{
+	unsigned int fw_caps = adapter->params.fw_caps_support;
+	struct fw_port_cmd cmd;
+	fw_port_cap32_t rcap;
+	int ret;
+
+	if (!(lc->pcaps & FW_PORT_CAP32_ANEG) &&
+	    lc->autoneg == AUTONEG_ENABLE) {
+		return -EINVAL;
+	}
+
+	/* Compute our Requested Port Capabilities and send that on to the
+	 * Firmware.
+	 */
+	rcap = t4_link_acaps(adapter, port, lc);
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PORT_CMD) |
+				       FW_CMD_REQUEST_F | FW_CMD_EXEC_F |
+				       FW_PORT_CMD_PORTID_V(port));
+	cmd.action_to_len16 =
+		cpu_to_be32(FW_PORT_CMD_ACTION_V(fw_caps == FW_CAPS16
+						 ? FW_PORT_ACTION_L1_CFG
+						 : FW_PORT_ACTION_L1_CFG32) |
+						 FW_LEN16(cmd));
+	if (fw_caps == FW_CAPS16)
+		cmd.u.l1cfg.rcap = cpu_to_be32(fwcaps32_to_caps16(rcap));
+	else
+		cmd.u.l1cfg32.rcap32 = cpu_to_be32(rcap);
+
+	ret = t4_wr_mbox_meat_timeout(adapter, mbox, &cmd, sizeof(cmd), NULL,
+				      sleep_ok, timeout);
+
+	/* Unfortunately, even if the Requested Port Capabilities "fit" within
+	 * the Physical Port Capabilities, some combinations of features may
+	 * still not be legal.  For example, 40Gb/s and Reed-Solomon Forward
+	 * Error Correction.  So if the Firmware rejects the L1 Configure
+	 * request, flag that here.
+	 */
+	if (ret) {
+		dev_err(adapter->pdev_dev,
+			"Requested Port Capabilities %#x rejected, error %d\n",
+			rcap, -ret);
+		return ret;
+	}
+	return 0;
+}
+
+/**
+ *	t4_restart_aneg - restart autonegotiation
+ *	@adap: the adapter
+ *	@mbox: mbox to use for the FW command
+ *	@port: the port id
+ *
+ *	Restarts autonegotiation for the selected port.
+ */
+int t4_restart_aneg(struct adapter *adap, unsigned int mbox, unsigned int port)
+{
+	unsigned int fw_caps = adap->params.fw_caps_support;
+	struct fw_port_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PORT_CMD) |
+				     FW_CMD_REQUEST_F | FW_CMD_EXEC_F |
+				     FW_PORT_CMD_PORTID_V(port));
+	c.action_to_len16 =
+		cpu_to_be32(FW_PORT_CMD_ACTION_V(fw_caps == FW_CAPS16
+						 ? FW_PORT_ACTION_L1_CFG
+						 : FW_PORT_ACTION_L1_CFG32) |
+			    FW_LEN16(c));
+	if (fw_caps == FW_CAPS16)
+		c.u.l1cfg.rcap = cpu_to_be32(FW_PORT_CAP_ANEG);
+	else
+		c.u.l1cfg32.rcap32 = cpu_to_be32(FW_PORT_CAP32_ANEG);
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+typedef void (*int_handler_t)(struct adapter *adap);
+
+struct intr_info {
+	unsigned int mask;       /* bits to check in interrupt status */
+	const char *msg;         /* message to print or NULL */
+	short stat_idx;          /* stat counter to increment or -1 */
+	unsigned short fatal;    /* whether the condition reported is fatal */
+	int_handler_t int_handler; /* platform-specific int handler */
+};
+
+/**
+ *	t4_handle_intr_status - table driven interrupt handler
+ *	@adapter: the adapter that generated the interrupt
+ *	@reg: the interrupt status register to process
+ *	@acts: table of interrupt actions
+ *
+ *	A table driven interrupt handler that applies a set of masks to an
+ *	interrupt status word and performs the corresponding actions if the
+ *	interrupts described by the mask have occurred.  The actions include
+ *	optionally emitting a warning or alert message.  The table is terminated
+ *	by an entry specifying mask 0.  Returns the number of fatal interrupt
+ *	conditions.
+ */
+static int t4_handle_intr_status(struct adapter *adapter, unsigned int reg,
+				 const struct intr_info *acts)
+{
+	int fatal = 0;
+	unsigned int mask = 0;
+	unsigned int status = t4_read_reg(adapter, reg);
+
+	for ( ; acts->mask; ++acts) {
+		if (!(status & acts->mask))
+			continue;
+		if (acts->fatal) {
+			fatal++;
+			dev_alert(adapter->pdev_dev, "%s (0x%x)\n", acts->msg,
+				  status & acts->mask);
+		} else if (acts->msg && printk_ratelimit())
+			dev_warn(adapter->pdev_dev, "%s (0x%x)\n", acts->msg,
+				 status & acts->mask);
+		if (acts->int_handler)
+			acts->int_handler(adapter);
+		mask |= acts->mask;
+	}
+	status &= mask;
+	if (status)                           /* clear processed interrupts */
+		t4_write_reg(adapter, reg, status);
+	return fatal;
+}
+
+/*
+ * Interrupt handler for the PCIE module.
+ */
+static void pcie_intr_handler(struct adapter *adapter)
+{
+	static const struct intr_info sysbus_intr_info[] = {
+		{ RNPP_F, "RXNP array parity error", -1, 1 },
+		{ RPCP_F, "RXPC array parity error", -1, 1 },
+		{ RCIP_F, "RXCIF array parity error", -1, 1 },
+		{ RCCP_F, "Rx completions control array parity error", -1, 1 },
+		{ RFTP_F, "RXFT array parity error", -1, 1 },
+		{ 0 }
+	};
+	static const struct intr_info pcie_port_intr_info[] = {
+		{ TPCP_F, "TXPC array parity error", -1, 1 },
+		{ TNPP_F, "TXNP array parity error", -1, 1 },
+		{ TFTP_F, "TXFT array parity error", -1, 1 },
+		{ TCAP_F, "TXCA array parity error", -1, 1 },
+		{ TCIP_F, "TXCIF array parity error", -1, 1 },
+		{ RCAP_F, "RXCA array parity error", -1, 1 },
+		{ OTDD_F, "outbound request TLP discarded", -1, 1 },
+		{ RDPE_F, "Rx data parity error", -1, 1 },
+		{ TDUE_F, "Tx uncorrectable data error", -1, 1 },
+		{ 0 }
+	};
+	static const struct intr_info pcie_intr_info[] = {
+		{ MSIADDRLPERR_F, "MSI AddrL parity error", -1, 1 },
+		{ MSIADDRHPERR_F, "MSI AddrH parity error", -1, 1 },
+		{ MSIDATAPERR_F, "MSI data parity error", -1, 1 },
+		{ MSIXADDRLPERR_F, "MSI-X AddrL parity error", -1, 1 },
+		{ MSIXADDRHPERR_F, "MSI-X AddrH parity error", -1, 1 },
+		{ MSIXDATAPERR_F, "MSI-X data parity error", -1, 1 },
+		{ MSIXDIPERR_F, "MSI-X DI parity error", -1, 1 },
+		{ PIOCPLPERR_F, "PCI PIO completion FIFO parity error", -1, 1 },
+		{ PIOREQPERR_F, "PCI PIO request FIFO parity error", -1, 1 },
+		{ TARTAGPERR_F, "PCI PCI target tag FIFO parity error", -1, 1 },
+		{ CCNTPERR_F, "PCI CMD channel count parity error", -1, 1 },
+		{ CREQPERR_F, "PCI CMD channel request parity error", -1, 1 },
+		{ CRSPPERR_F, "PCI CMD channel response parity error", -1, 1 },
+		{ DCNTPERR_F, "PCI DMA channel count parity error", -1, 1 },
+		{ DREQPERR_F, "PCI DMA channel request parity error", -1, 1 },
+		{ DRSPPERR_F, "PCI DMA channel response parity error", -1, 1 },
+		{ HCNTPERR_F, "PCI HMA channel count parity error", -1, 1 },
+		{ HREQPERR_F, "PCI HMA channel request parity error", -1, 1 },
+		{ HRSPPERR_F, "PCI HMA channel response parity error", -1, 1 },
+		{ CFGSNPPERR_F, "PCI config snoop FIFO parity error", -1, 1 },
+		{ FIDPERR_F, "PCI FID parity error", -1, 1 },
+		{ INTXCLRPERR_F, "PCI INTx clear parity error", -1, 1 },
+		{ MATAGPERR_F, "PCI MA tag parity error", -1, 1 },
+		{ PIOTAGPERR_F, "PCI PIO tag parity error", -1, 1 },
+		{ RXCPLPERR_F, "PCI Rx completion parity error", -1, 1 },
+		{ RXWRPERR_F, "PCI Rx write parity error", -1, 1 },
+		{ RPLPERR_F, "PCI replay buffer parity error", -1, 1 },
+		{ PCIESINT_F, "PCI core secondary fault", -1, 1 },
+		{ PCIEPINT_F, "PCI core primary fault", -1, 1 },
+		{ UNXSPLCPLERR_F, "PCI unexpected split completion error",
+		  -1, 0 },
+		{ 0 }
+	};
+
+	static struct intr_info t5_pcie_intr_info[] = {
+		{ MSTGRPPERR_F, "Master Response Read Queue parity error",
+		  -1, 1 },
+		{ MSTTIMEOUTPERR_F, "Master Timeout FIFO parity error", -1, 1 },
+		{ MSIXSTIPERR_F, "MSI-X STI SRAM parity error", -1, 1 },
+		{ MSIXADDRLPERR_F, "MSI-X AddrL parity error", -1, 1 },
+		{ MSIXADDRHPERR_F, "MSI-X AddrH parity error", -1, 1 },
+		{ MSIXDATAPERR_F, "MSI-X data parity error", -1, 1 },
+		{ MSIXDIPERR_F, "MSI-X DI parity error", -1, 1 },
+		{ PIOCPLGRPPERR_F, "PCI PIO completion Group FIFO parity error",
+		  -1, 1 },
+		{ PIOREQGRPPERR_F, "PCI PIO request Group FIFO parity error",
+		  -1, 1 },
+		{ TARTAGPERR_F, "PCI PCI target tag FIFO parity error", -1, 1 },
+		{ MSTTAGQPERR_F, "PCI master tag queue parity error", -1, 1 },
+		{ CREQPERR_F, "PCI CMD channel request parity error", -1, 1 },
+		{ CRSPPERR_F, "PCI CMD channel response parity error", -1, 1 },
+		{ DREQWRPERR_F, "PCI DMA channel write request parity error",
+		  -1, 1 },
+		{ DREQPERR_F, "PCI DMA channel request parity error", -1, 1 },
+		{ DRSPPERR_F, "PCI DMA channel response parity error", -1, 1 },
+		{ HREQWRPERR_F, "PCI HMA channel count parity error", -1, 1 },
+		{ HREQPERR_F, "PCI HMA channel request parity error", -1, 1 },
+		{ HRSPPERR_F, "PCI HMA channel response parity error", -1, 1 },
+		{ CFGSNPPERR_F, "PCI config snoop FIFO parity error", -1, 1 },
+		{ FIDPERR_F, "PCI FID parity error", -1, 1 },
+		{ VFIDPERR_F, "PCI INTx clear parity error", -1, 1 },
+		{ MAGRPPERR_F, "PCI MA group FIFO parity error", -1, 1 },
+		{ PIOTAGPERR_F, "PCI PIO tag parity error", -1, 1 },
+		{ IPRXHDRGRPPERR_F, "PCI IP Rx header group parity error",
+		  -1, 1 },
+		{ IPRXDATAGRPPERR_F, "PCI IP Rx data group parity error",
+		  -1, 1 },
+		{ RPLPERR_F, "PCI IP replay buffer parity error", -1, 1 },
+		{ IPSOTPERR_F, "PCI IP SOT buffer parity error", -1, 1 },
+		{ TRGT1GRPPERR_F, "PCI TRGT1 group FIFOs parity error", -1, 1 },
+		{ READRSPERR_F, "Outbound read error", -1, 0 },
+		{ 0 }
+	};
+
+	int fat;
+
+	if (is_t4(adapter->params.chip))
+		fat = t4_handle_intr_status(adapter,
+				PCIE_CORE_UTL_SYSTEM_BUS_AGENT_STATUS_A,
+				sysbus_intr_info) +
+			t4_handle_intr_status(adapter,
+					PCIE_CORE_UTL_PCI_EXPRESS_PORT_STATUS_A,
+					pcie_port_intr_info) +
+			t4_handle_intr_status(adapter, PCIE_INT_CAUSE_A,
+					      pcie_intr_info);
+	else
+		fat = t4_handle_intr_status(adapter, PCIE_INT_CAUSE_A,
+					    t5_pcie_intr_info);
+
+	if (fat)
+		t4_fatal_err(adapter);
+}
+
+/*
+ * TP interrupt handler.
+ */
+static void tp_intr_handler(struct adapter *adapter)
+{
+	static const struct intr_info tp_intr_info[] = {
+		{ 0x3fffffff, "TP parity error", -1, 1 },
+		{ FLMTXFLSTEMPTY_F, "TP out of Tx pages", -1, 1 },
+		{ 0 }
+	};
+
+	if (t4_handle_intr_status(adapter, TP_INT_CAUSE_A, tp_intr_info))
+		t4_fatal_err(adapter);
+}
+
+/*
+ * SGE interrupt handler.
+ */
+static void sge_intr_handler(struct adapter *adapter)
+{
+	u32 v = 0, perr;
+	u32 err;
+
+	static const struct intr_info sge_intr_info[] = {
+		{ ERR_CPL_EXCEED_IQE_SIZE_F,
+		  "SGE received CPL exceeding IQE size", -1, 1 },
+		{ ERR_INVALID_CIDX_INC_F,
+		  "SGE GTS CIDX increment too large", -1, 0 },
+		{ ERR_CPL_OPCODE_0_F, "SGE received 0-length CPL", -1, 0 },
+		{ DBFIFO_LP_INT_F, NULL, -1, 0, t4_db_full },
+		{ ERR_DATA_CPL_ON_HIGH_QID1_F | ERR_DATA_CPL_ON_HIGH_QID0_F,
+		  "SGE IQID > 1023 received CPL for FL", -1, 0 },
+		{ ERR_BAD_DB_PIDX3_F, "SGE DBP 3 pidx increment too large", -1,
+		  0 },
+		{ ERR_BAD_DB_PIDX2_F, "SGE DBP 2 pidx increment too large", -1,
+		  0 },
+		{ ERR_BAD_DB_PIDX1_F, "SGE DBP 1 pidx increment too large", -1,
+		  0 },
+		{ ERR_BAD_DB_PIDX0_F, "SGE DBP 0 pidx increment too large", -1,
+		  0 },
+		{ ERR_ING_CTXT_PRIO_F,
+		  "SGE too many priority ingress contexts", -1, 0 },
+		{ INGRESS_SIZE_ERR_F, "SGE illegal ingress QID", -1, 0 },
+		{ EGRESS_SIZE_ERR_F, "SGE illegal egress QID", -1, 0 },
+		{ 0 }
+	};
+
+	static struct intr_info t4t5_sge_intr_info[] = {
+		{ ERR_DROPPED_DB_F, NULL, -1, 0, t4_db_dropped },
+		{ DBFIFO_HP_INT_F, NULL, -1, 0, t4_db_full },
+		{ ERR_EGR_CTXT_PRIO_F,
+		  "SGE too many priority egress contexts", -1, 0 },
+		{ 0 }
+	};
+
+	perr = t4_read_reg(adapter, SGE_INT_CAUSE1_A);
+	if (perr) {
+		v |= perr;
+		dev_alert(adapter->pdev_dev, "SGE Cause1 Parity Error %#x\n",
+			  perr);
+	}
+
+	perr = t4_read_reg(adapter, SGE_INT_CAUSE2_A);
+	if (perr) {
+		v |= perr;
+		dev_alert(adapter->pdev_dev, "SGE Cause2 Parity Error %#x\n",
+			  perr);
+	}
+
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) >= CHELSIO_T5) {
+		perr = t4_read_reg(adapter, SGE_INT_CAUSE5_A);
+		/* Parity error (CRC) for err_T_RxCRC is trivial, ignore it */
+		perr &= ~ERR_T_RXCRC_F;
+		if (perr) {
+			v |= perr;
+			dev_alert(adapter->pdev_dev,
+				  "SGE Cause5 Parity Error %#x\n", perr);
+		}
+	}
+
+	v |= t4_handle_intr_status(adapter, SGE_INT_CAUSE3_A, sge_intr_info);
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) <= CHELSIO_T5)
+		v |= t4_handle_intr_status(adapter, SGE_INT_CAUSE3_A,
+					   t4t5_sge_intr_info);
+
+	err = t4_read_reg(adapter, SGE_ERROR_STATS_A);
+	if (err & ERROR_QID_VALID_F) {
+		dev_err(adapter->pdev_dev, "SGE error for queue %u\n",
+			ERROR_QID_G(err));
+		if (err & UNCAPTURED_ERROR_F)
+			dev_err(adapter->pdev_dev,
+				"SGE UNCAPTURED_ERROR set (clearing)\n");
+		t4_write_reg(adapter, SGE_ERROR_STATS_A, ERROR_QID_VALID_F |
+			     UNCAPTURED_ERROR_F);
+	}
+
+	if (v != 0)
+		t4_fatal_err(adapter);
+}
+
+#define CIM_OBQ_INTR (OBQULP0PARERR_F | OBQULP1PARERR_F | OBQULP2PARERR_F |\
+		      OBQULP3PARERR_F | OBQSGEPARERR_F | OBQNCSIPARERR_F)
+#define CIM_IBQ_INTR (IBQTP0PARERR_F | IBQTP1PARERR_F | IBQULPPARERR_F |\
+		      IBQSGEHIPARERR_F | IBQSGELOPARERR_F | IBQNCSIPARERR_F)
+
+/*
+ * CIM interrupt handler.
+ */
+static void cim_intr_handler(struct adapter *adapter)
+{
+	static const struct intr_info cim_intr_info[] = {
+		{ PREFDROPINT_F, "CIM control register prefetch drop", -1, 1 },
+		{ CIM_OBQ_INTR, "CIM OBQ parity error", -1, 1 },
+		{ CIM_IBQ_INTR, "CIM IBQ parity error", -1, 1 },
+		{ MBUPPARERR_F, "CIM mailbox uP parity error", -1, 1 },
+		{ MBHOSTPARERR_F, "CIM mailbox host parity error", -1, 1 },
+		{ TIEQINPARERRINT_F, "CIM TIEQ outgoing parity error", -1, 1 },
+		{ TIEQOUTPARERRINT_F, "CIM TIEQ incoming parity error", -1, 1 },
+		{ TIMER0INT_F, "CIM TIMER0 interrupt", -1, 1 },
+		{ 0 }
+	};
+	static const struct intr_info cim_upintr_info[] = {
+		{ RSVDSPACEINT_F, "CIM reserved space access", -1, 1 },
+		{ ILLTRANSINT_F, "CIM illegal transaction", -1, 1 },
+		{ ILLWRINT_F, "CIM illegal write", -1, 1 },
+		{ ILLRDINT_F, "CIM illegal read", -1, 1 },
+		{ ILLRDBEINT_F, "CIM illegal read BE", -1, 1 },
+		{ ILLWRBEINT_F, "CIM illegal write BE", -1, 1 },
+		{ SGLRDBOOTINT_F, "CIM single read from boot space", -1, 1 },
+		{ SGLWRBOOTINT_F, "CIM single write to boot space", -1, 1 },
+		{ BLKWRBOOTINT_F, "CIM block write to boot space", -1, 1 },
+		{ SGLRDFLASHINT_F, "CIM single read from flash space", -1, 1 },
+		{ SGLWRFLASHINT_F, "CIM single write to flash space", -1, 1 },
+		{ BLKWRFLASHINT_F, "CIM block write to flash space", -1, 1 },
+		{ SGLRDEEPROMINT_F, "CIM single EEPROM read", -1, 1 },
+		{ SGLWREEPROMINT_F, "CIM single EEPROM write", -1, 1 },
+		{ BLKRDEEPROMINT_F, "CIM block EEPROM read", -1, 1 },
+		{ BLKWREEPROMINT_F, "CIM block EEPROM write", -1, 1 },
+		{ SGLRDCTLINT_F, "CIM single read from CTL space", -1, 1 },
+		{ SGLWRCTLINT_F, "CIM single write to CTL space", -1, 1 },
+		{ BLKRDCTLINT_F, "CIM block read from CTL space", -1, 1 },
+		{ BLKWRCTLINT_F, "CIM block write to CTL space", -1, 1 },
+		{ SGLRDPLINT_F, "CIM single read from PL space", -1, 1 },
+		{ SGLWRPLINT_F, "CIM single write to PL space", -1, 1 },
+		{ BLKRDPLINT_F, "CIM block read from PL space", -1, 1 },
+		{ BLKWRPLINT_F, "CIM block write to PL space", -1, 1 },
+		{ REQOVRLOOKUPINT_F, "CIM request FIFO overwrite", -1, 1 },
+		{ RSPOVRLOOKUPINT_F, "CIM response FIFO overwrite", -1, 1 },
+		{ TIMEOUTINT_F, "CIM PIF timeout", -1, 1 },
+		{ TIMEOUTMAINT_F, "CIM PIF MA timeout", -1, 1 },
+		{ 0 }
+	};
+
+	u32 val, fw_err;
+	int fat;
+
+	fw_err = t4_read_reg(adapter, PCIE_FW_A);
+	if (fw_err & PCIE_FW_ERR_F)
+		t4_report_fw_error(adapter);
+
+	/* When the Firmware detects an internal error which normally
+	 * wouldn't raise a Host Interrupt, it forces a CIM Timer0 interrupt
+	 * in order to make sure the Host sees the Firmware Crash.  So
+	 * if we have a Timer0 interrupt and don't see a Firmware Crash,
+	 * ignore the Timer0 interrupt.
+	 */
+
+	val = t4_read_reg(adapter, CIM_HOST_INT_CAUSE_A);
+	if (val & TIMER0INT_F)
+		if (!(fw_err & PCIE_FW_ERR_F) ||
+		    (PCIE_FW_EVAL_G(fw_err) != PCIE_FW_EVAL_CRASH))
+			t4_write_reg(adapter, CIM_HOST_INT_CAUSE_A,
+				     TIMER0INT_F);
+
+	fat = t4_handle_intr_status(adapter, CIM_HOST_INT_CAUSE_A,
+				    cim_intr_info) +
+	      t4_handle_intr_status(adapter, CIM_HOST_UPACC_INT_CAUSE_A,
+				    cim_upintr_info);
+	if (fat)
+		t4_fatal_err(adapter);
+}
+
+/*
+ * ULP RX interrupt handler.
+ */
+static void ulprx_intr_handler(struct adapter *adapter)
+{
+	static const struct intr_info ulprx_intr_info[] = {
+		{ 0x1800000, "ULPRX context error", -1, 1 },
+		{ 0x7fffff, "ULPRX parity error", -1, 1 },
+		{ 0 }
+	};
+
+	if (t4_handle_intr_status(adapter, ULP_RX_INT_CAUSE_A, ulprx_intr_info))
+		t4_fatal_err(adapter);
+}
+
+/*
+ * ULP TX interrupt handler.
+ */
+static void ulptx_intr_handler(struct adapter *adapter)
+{
+	static const struct intr_info ulptx_intr_info[] = {
+		{ PBL_BOUND_ERR_CH3_F, "ULPTX channel 3 PBL out of bounds", -1,
+		  0 },
+		{ PBL_BOUND_ERR_CH2_F, "ULPTX channel 2 PBL out of bounds", -1,
+		  0 },
+		{ PBL_BOUND_ERR_CH1_F, "ULPTX channel 1 PBL out of bounds", -1,
+		  0 },
+		{ PBL_BOUND_ERR_CH0_F, "ULPTX channel 0 PBL out of bounds", -1,
+		  0 },
+		{ 0xfffffff, "ULPTX parity error", -1, 1 },
+		{ 0 }
+	};
+
+	if (t4_handle_intr_status(adapter, ULP_TX_INT_CAUSE_A, ulptx_intr_info))
+		t4_fatal_err(adapter);
+}
+
+/*
+ * PM TX interrupt handler.
+ */
+static void pmtx_intr_handler(struct adapter *adapter)
+{
+	static const struct intr_info pmtx_intr_info[] = {
+		{ PCMD_LEN_OVFL0_F, "PMTX channel 0 pcmd too large", -1, 1 },
+		{ PCMD_LEN_OVFL1_F, "PMTX channel 1 pcmd too large", -1, 1 },
+		{ PCMD_LEN_OVFL2_F, "PMTX channel 2 pcmd too large", -1, 1 },
+		{ ZERO_C_CMD_ERROR_F, "PMTX 0-length pcmd", -1, 1 },
+		{ PMTX_FRAMING_ERROR_F, "PMTX framing error", -1, 1 },
+		{ OESPI_PAR_ERROR_F, "PMTX oespi parity error", -1, 1 },
+		{ DB_OPTIONS_PAR_ERROR_F, "PMTX db_options parity error",
+		  -1, 1 },
+		{ ICSPI_PAR_ERROR_F, "PMTX icspi parity error", -1, 1 },
+		{ PMTX_C_PCMD_PAR_ERROR_F, "PMTX c_pcmd parity error", -1, 1},
+		{ 0 }
+	};
+
+	if (t4_handle_intr_status(adapter, PM_TX_INT_CAUSE_A, pmtx_intr_info))
+		t4_fatal_err(adapter);
+}
+
+/*
+ * PM RX interrupt handler.
+ */
+static void pmrx_intr_handler(struct adapter *adapter)
+{
+	static const struct intr_info pmrx_intr_info[] = {
+		{ ZERO_E_CMD_ERROR_F, "PMRX 0-length pcmd", -1, 1 },
+		{ PMRX_FRAMING_ERROR_F, "PMRX framing error", -1, 1 },
+		{ OCSPI_PAR_ERROR_F, "PMRX ocspi parity error", -1, 1 },
+		{ DB_OPTIONS_PAR_ERROR_F, "PMRX db_options parity error",
+		  -1, 1 },
+		{ IESPI_PAR_ERROR_F, "PMRX iespi parity error", -1, 1 },
+		{ PMRX_E_PCMD_PAR_ERROR_F, "PMRX e_pcmd parity error", -1, 1},
+		{ 0 }
+	};
+
+	if (t4_handle_intr_status(adapter, PM_RX_INT_CAUSE_A, pmrx_intr_info))
+		t4_fatal_err(adapter);
+}
+
+/*
+ * CPL switch interrupt handler.
+ */
+static void cplsw_intr_handler(struct adapter *adapter)
+{
+	static const struct intr_info cplsw_intr_info[] = {
+		{ CIM_OP_MAP_PERR_F, "CPLSW CIM op_map parity error", -1, 1 },
+		{ CIM_OVFL_ERROR_F, "CPLSW CIM overflow", -1, 1 },
+		{ TP_FRAMING_ERROR_F, "CPLSW TP framing error", -1, 1 },
+		{ SGE_FRAMING_ERROR_F, "CPLSW SGE framing error", -1, 1 },
+		{ CIM_FRAMING_ERROR_F, "CPLSW CIM framing error", -1, 1 },
+		{ ZERO_SWITCH_ERROR_F, "CPLSW no-switch error", -1, 1 },
+		{ 0 }
+	};
+
+	if (t4_handle_intr_status(adapter, CPL_INTR_CAUSE_A, cplsw_intr_info))
+		t4_fatal_err(adapter);
+}
+
+/*
+ * LE interrupt handler.
+ */
+static void le_intr_handler(struct adapter *adap)
+{
+	enum chip_type chip = CHELSIO_CHIP_VERSION(adap->params.chip);
+	static const struct intr_info le_intr_info[] = {
+		{ LIPMISS_F, "LE LIP miss", -1, 0 },
+		{ LIP0_F, "LE 0 LIP error", -1, 0 },
+		{ PARITYERR_F, "LE parity error", -1, 1 },
+		{ UNKNOWNCMD_F, "LE unknown command", -1, 1 },
+		{ REQQPARERR_F, "LE request queue parity error", -1, 1 },
+		{ 0 }
+	};
+
+	static struct intr_info t6_le_intr_info[] = {
+		{ T6_LIPMISS_F, "LE LIP miss", -1, 0 },
+		{ T6_LIP0_F, "LE 0 LIP error", -1, 0 },
+		{ CMDTIDERR_F, "LE cmd tid error", -1, 1 },
+		{ TCAMINTPERR_F, "LE parity error", -1, 1 },
+		{ T6_UNKNOWNCMD_F, "LE unknown command", -1, 1 },
+		{ SSRAMINTPERR_F, "LE request queue parity error", -1, 1 },
+		{ HASHTBLMEMCRCERR_F, "LE hash table mem crc error", -1, 0 },
+		{ 0 }
+	};
+
+	if (t4_handle_intr_status(adap, LE_DB_INT_CAUSE_A,
+				  (chip <= CHELSIO_T5) ?
+				  le_intr_info : t6_le_intr_info))
+		t4_fatal_err(adap);
+}
+
+/*
+ * MPS interrupt handler.
+ */
+static void mps_intr_handler(struct adapter *adapter)
+{
+	static const struct intr_info mps_rx_intr_info[] = {
+		{ 0xffffff, "MPS Rx parity error", -1, 1 },
+		{ 0 }
+	};
+	static const struct intr_info mps_tx_intr_info[] = {
+		{ TPFIFO_V(TPFIFO_M), "MPS Tx TP FIFO parity error", -1, 1 },
+		{ NCSIFIFO_F, "MPS Tx NC-SI FIFO parity error", -1, 1 },
+		{ TXDATAFIFO_V(TXDATAFIFO_M), "MPS Tx data FIFO parity error",
+		  -1, 1 },
+		{ TXDESCFIFO_V(TXDESCFIFO_M), "MPS Tx desc FIFO parity error",
+		  -1, 1 },
+		{ BUBBLE_F, "MPS Tx underflow", -1, 1 },
+		{ SECNTERR_F, "MPS Tx SOP/EOP error", -1, 1 },
+		{ FRMERR_F, "MPS Tx framing error", -1, 1 },
+		{ 0 }
+	};
+	static const struct intr_info t6_mps_tx_intr_info[] = {
+		{ TPFIFO_V(TPFIFO_M), "MPS Tx TP FIFO parity error", -1, 1 },
+		{ NCSIFIFO_F, "MPS Tx NC-SI FIFO parity error", -1, 1 },
+		{ TXDATAFIFO_V(TXDATAFIFO_M), "MPS Tx data FIFO parity error",
+		  -1, 1 },
+		{ TXDESCFIFO_V(TXDESCFIFO_M), "MPS Tx desc FIFO parity error",
+		  -1, 1 },
+		/* MPS Tx Bubble is normal for T6 */
+		{ SECNTERR_F, "MPS Tx SOP/EOP error", -1, 1 },
+		{ FRMERR_F, "MPS Tx framing error", -1, 1 },
+		{ 0 }
+	};
+	static const struct intr_info mps_trc_intr_info[] = {
+		{ FILTMEM_V(FILTMEM_M), "MPS TRC filter parity error", -1, 1 },
+		{ PKTFIFO_V(PKTFIFO_M), "MPS TRC packet FIFO parity error",
+		  -1, 1 },
+		{ MISCPERR_F, "MPS TRC misc parity error", -1, 1 },
+		{ 0 }
+	};
+	static const struct intr_info mps_stat_sram_intr_info[] = {
+		{ 0x1fffff, "MPS statistics SRAM parity error", -1, 1 },
+		{ 0 }
+	};
+	static const struct intr_info mps_stat_tx_intr_info[] = {
+		{ 0xfffff, "MPS statistics Tx FIFO parity error", -1, 1 },
+		{ 0 }
+	};
+	static const struct intr_info mps_stat_rx_intr_info[] = {
+		{ 0xffffff, "MPS statistics Rx FIFO parity error", -1, 1 },
+		{ 0 }
+	};
+	static const struct intr_info mps_cls_intr_info[] = {
+		{ MATCHSRAM_F, "MPS match SRAM parity error", -1, 1 },
+		{ MATCHTCAM_F, "MPS match TCAM parity error", -1, 1 },
+		{ HASHSRAM_F, "MPS hash SRAM parity error", -1, 1 },
+		{ 0 }
+	};
+
+	int fat;
+
+	fat = t4_handle_intr_status(adapter, MPS_RX_PERR_INT_CAUSE_A,
+				    mps_rx_intr_info) +
+	      t4_handle_intr_status(adapter, MPS_TX_INT_CAUSE_A,
+				    is_t6(adapter->params.chip)
+				    ? t6_mps_tx_intr_info
+				    : mps_tx_intr_info) +
+	      t4_handle_intr_status(adapter, MPS_TRC_INT_CAUSE_A,
+				    mps_trc_intr_info) +
+	      t4_handle_intr_status(adapter, MPS_STAT_PERR_INT_CAUSE_SRAM_A,
+				    mps_stat_sram_intr_info) +
+	      t4_handle_intr_status(adapter, MPS_STAT_PERR_INT_CAUSE_TX_FIFO_A,
+				    mps_stat_tx_intr_info) +
+	      t4_handle_intr_status(adapter, MPS_STAT_PERR_INT_CAUSE_RX_FIFO_A,
+				    mps_stat_rx_intr_info) +
+	      t4_handle_intr_status(adapter, MPS_CLS_INT_CAUSE_A,
+				    mps_cls_intr_info);
+
+	t4_write_reg(adapter, MPS_INT_CAUSE_A, 0);
+	t4_read_reg(adapter, MPS_INT_CAUSE_A);                    /* flush */
+	if (fat)
+		t4_fatal_err(adapter);
+}
+
+#define MEM_INT_MASK (PERR_INT_CAUSE_F | ECC_CE_INT_CAUSE_F | \
+		      ECC_UE_INT_CAUSE_F)
+
+/*
+ * EDC/MC interrupt handler.
+ */
+static void mem_intr_handler(struct adapter *adapter, int idx)
+{
+	static const char name[4][7] = { "EDC0", "EDC1", "MC/MC0", "MC1" };
+
+	unsigned int addr, cnt_addr, v;
+
+	if (idx <= MEM_EDC1) {
+		addr = EDC_REG(EDC_INT_CAUSE_A, idx);
+		cnt_addr = EDC_REG(EDC_ECC_STATUS_A, idx);
+	} else if (idx == MEM_MC) {
+		if (is_t4(adapter->params.chip)) {
+			addr = MC_INT_CAUSE_A;
+			cnt_addr = MC_ECC_STATUS_A;
+		} else {
+			addr = MC_P_INT_CAUSE_A;
+			cnt_addr = MC_P_ECC_STATUS_A;
+		}
+	} else {
+		addr = MC_REG(MC_P_INT_CAUSE_A, 1);
+		cnt_addr = MC_REG(MC_P_ECC_STATUS_A, 1);
+	}
+
+	v = t4_read_reg(adapter, addr) & MEM_INT_MASK;
+	if (v & PERR_INT_CAUSE_F)
+		dev_alert(adapter->pdev_dev, "%s FIFO parity error\n",
+			  name[idx]);
+	if (v & ECC_CE_INT_CAUSE_F) {
+		u32 cnt = ECC_CECNT_G(t4_read_reg(adapter, cnt_addr));
+
+		t4_edc_err_read(adapter, idx);
+
+		t4_write_reg(adapter, cnt_addr, ECC_CECNT_V(ECC_CECNT_M));
+		if (printk_ratelimit())
+			dev_warn(adapter->pdev_dev,
+				 "%u %s correctable ECC data error%s\n",
+				 cnt, name[idx], cnt > 1 ? "s" : "");
+	}
+	if (v & ECC_UE_INT_CAUSE_F)
+		dev_alert(adapter->pdev_dev,
+			  "%s uncorrectable ECC data error\n", name[idx]);
+
+	t4_write_reg(adapter, addr, v);
+	if (v & (PERR_INT_CAUSE_F | ECC_UE_INT_CAUSE_F))
+		t4_fatal_err(adapter);
+}
+
+/*
+ * MA interrupt handler.
+ */
+static void ma_intr_handler(struct adapter *adap)
+{
+	u32 v, status = t4_read_reg(adap, MA_INT_CAUSE_A);
+
+	if (status & MEM_PERR_INT_CAUSE_F) {
+		dev_alert(adap->pdev_dev,
+			  "MA parity error, parity status %#x\n",
+			  t4_read_reg(adap, MA_PARITY_ERROR_STATUS1_A));
+		if (is_t5(adap->params.chip))
+			dev_alert(adap->pdev_dev,
+				  "MA parity error, parity status %#x\n",
+				  t4_read_reg(adap,
+					      MA_PARITY_ERROR_STATUS2_A));
+	}
+	if (status & MEM_WRAP_INT_CAUSE_F) {
+		v = t4_read_reg(adap, MA_INT_WRAP_STATUS_A);
+		dev_alert(adap->pdev_dev, "MA address wrap-around error by "
+			  "client %u to address %#x\n",
+			  MEM_WRAP_CLIENT_NUM_G(v),
+			  MEM_WRAP_ADDRESS_G(v) << 4);
+	}
+	t4_write_reg(adap, MA_INT_CAUSE_A, status);
+	t4_fatal_err(adap);
+}
+
+/*
+ * SMB interrupt handler.
+ */
+static void smb_intr_handler(struct adapter *adap)
+{
+	static const struct intr_info smb_intr_info[] = {
+		{ MSTTXFIFOPARINT_F, "SMB master Tx FIFO parity error", -1, 1 },
+		{ MSTRXFIFOPARINT_F, "SMB master Rx FIFO parity error", -1, 1 },
+		{ SLVFIFOPARINT_F, "SMB slave FIFO parity error", -1, 1 },
+		{ 0 }
+	};
+
+	if (t4_handle_intr_status(adap, SMB_INT_CAUSE_A, smb_intr_info))
+		t4_fatal_err(adap);
+}
+
+/*
+ * NC-SI interrupt handler.
+ */
+static void ncsi_intr_handler(struct adapter *adap)
+{
+	static const struct intr_info ncsi_intr_info[] = {
+		{ CIM_DM_PRTY_ERR_F, "NC-SI CIM parity error", -1, 1 },
+		{ MPS_DM_PRTY_ERR_F, "NC-SI MPS parity error", -1, 1 },
+		{ TXFIFO_PRTY_ERR_F, "NC-SI Tx FIFO parity error", -1, 1 },
+		{ RXFIFO_PRTY_ERR_F, "NC-SI Rx FIFO parity error", -1, 1 },
+		{ 0 }
+	};
+
+	if (t4_handle_intr_status(adap, NCSI_INT_CAUSE_A, ncsi_intr_info))
+		t4_fatal_err(adap);
+}
+
+/*
+ * XGMAC interrupt handler.
+ */
+static void xgmac_intr_handler(struct adapter *adap, int port)
+{
+	u32 v, int_cause_reg;
+
+	if (is_t4(adap->params.chip))
+		int_cause_reg = PORT_REG(port, XGMAC_PORT_INT_CAUSE_A);
+	else
+		int_cause_reg = T5_PORT_REG(port, MAC_PORT_INT_CAUSE_A);
+
+	v = t4_read_reg(adap, int_cause_reg);
+
+	v &= TXFIFO_PRTY_ERR_F | RXFIFO_PRTY_ERR_F;
+	if (!v)
+		return;
+
+	if (v & TXFIFO_PRTY_ERR_F)
+		dev_alert(adap->pdev_dev, "XGMAC %d Tx FIFO parity error\n",
+			  port);
+	if (v & RXFIFO_PRTY_ERR_F)
+		dev_alert(adap->pdev_dev, "XGMAC %d Rx FIFO parity error\n",
+			  port);
+	t4_write_reg(adap, PORT_REG(port, XGMAC_PORT_INT_CAUSE_A), v);
+	t4_fatal_err(adap);
+}
+
+/*
+ * PL interrupt handler.
+ */
+static void pl_intr_handler(struct adapter *adap)
+{
+	static const struct intr_info pl_intr_info[] = {
+		{ FATALPERR_F, "T4 fatal parity error", -1, 1 },
+		{ PERRVFID_F, "PL VFID_MAP parity error", -1, 1 },
+		{ 0 }
+	};
+
+	if (t4_handle_intr_status(adap, PL_PL_INT_CAUSE_A, pl_intr_info))
+		t4_fatal_err(adap);
+}
+
+#define PF_INTR_MASK (PFSW_F)
+#define GLBL_INTR_MASK (CIM_F | MPS_F | PL_F | PCIE_F | MC_F | EDC0_F | \
+		EDC1_F | LE_F | TP_F | MA_F | PM_TX_F | PM_RX_F | ULP_RX_F | \
+		CPL_SWITCH_F | SGE_F | ULP_TX_F | SF_F)
+
+/**
+ *	t4_slow_intr_handler - control path interrupt handler
+ *	@adapter: the adapter
+ *
+ *	T4 interrupt handler for non-data global interrupt events, e.g., errors.
+ *	The designation 'slow' is because it involves register reads, while
+ *	data interrupts typically don't involve any MMIOs.
+ */
+int t4_slow_intr_handler(struct adapter *adapter)
+{
+	/* There are rare cases where a PL_INT_CAUSE bit may end up getting
+	 * set when the corresponding PL_INT_ENABLE bit isn't set.  It's
+	 * easiest just to mask that case here.
+	 */
+	u32 raw_cause = t4_read_reg(adapter, PL_INT_CAUSE_A);
+	u32 enable = t4_read_reg(adapter, PL_INT_ENABLE_A);
+	u32 cause = raw_cause & enable;
+
+	if (!(cause & GLBL_INTR_MASK))
+		return 0;
+	if (cause & CIM_F)
+		cim_intr_handler(adapter);
+	if (cause & MPS_F)
+		mps_intr_handler(adapter);
+	if (cause & NCSI_F)
+		ncsi_intr_handler(adapter);
+	if (cause & PL_F)
+		pl_intr_handler(adapter);
+	if (cause & SMB_F)
+		smb_intr_handler(adapter);
+	if (cause & XGMAC0_F)
+		xgmac_intr_handler(adapter, 0);
+	if (cause & XGMAC1_F)
+		xgmac_intr_handler(adapter, 1);
+	if (cause & XGMAC_KR0_F)
+		xgmac_intr_handler(adapter, 2);
+	if (cause & XGMAC_KR1_F)
+		xgmac_intr_handler(adapter, 3);
+	if (cause & PCIE_F)
+		pcie_intr_handler(adapter);
+	if (cause & MC_F)
+		mem_intr_handler(adapter, MEM_MC);
+	if (is_t5(adapter->params.chip) && (cause & MC1_F))
+		mem_intr_handler(adapter, MEM_MC1);
+	if (cause & EDC0_F)
+		mem_intr_handler(adapter, MEM_EDC0);
+	if (cause & EDC1_F)
+		mem_intr_handler(adapter, MEM_EDC1);
+	if (cause & LE_F)
+		le_intr_handler(adapter);
+	if (cause & TP_F)
+		tp_intr_handler(adapter);
+	if (cause & MA_F)
+		ma_intr_handler(adapter);
+	if (cause & PM_TX_F)
+		pmtx_intr_handler(adapter);
+	if (cause & PM_RX_F)
+		pmrx_intr_handler(adapter);
+	if (cause & ULP_RX_F)
+		ulprx_intr_handler(adapter);
+	if (cause & CPL_SWITCH_F)
+		cplsw_intr_handler(adapter);
+	if (cause & SGE_F)
+		sge_intr_handler(adapter);
+	if (cause & ULP_TX_F)
+		ulptx_intr_handler(adapter);
+
+	/* Clear the interrupts just processed for which we are the master. */
+	t4_write_reg(adapter, PL_INT_CAUSE_A, raw_cause & GLBL_INTR_MASK);
+	(void)t4_read_reg(adapter, PL_INT_CAUSE_A); /* flush */
+	return 1;
+}
+
+/**
+ *	t4_intr_enable - enable interrupts
+ *	@adapter: the adapter whose interrupts should be enabled
+ *
+ *	Enable PF-specific interrupts for the calling function and the top-level
+ *	interrupt concentrator for global interrupts.  Interrupts are already
+ *	enabled at each module,	here we just enable the roots of the interrupt
+ *	hierarchies.
+ *
+ *	Note: this function should be called only when the driver manages
+ *	non PF-specific interrupts from the various HW modules.  Only one PCI
+ *	function at a time should be doing this.
+ */
+void t4_intr_enable(struct adapter *adapter)
+{
+	u32 val = 0;
+	u32 whoami = t4_read_reg(adapter, PL_WHOAMI_A);
+	u32 pf = CHELSIO_CHIP_VERSION(adapter->params.chip) <= CHELSIO_T5 ?
+			SOURCEPF_G(whoami) : T6_SOURCEPF_G(whoami);
+
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) <= CHELSIO_T5)
+		val = ERR_DROPPED_DB_F | ERR_EGR_CTXT_PRIO_F | DBFIFO_HP_INT_F;
+	t4_write_reg(adapter, SGE_INT_ENABLE3_A, ERR_CPL_EXCEED_IQE_SIZE_F |
+		     ERR_INVALID_CIDX_INC_F | ERR_CPL_OPCODE_0_F |
+		     ERR_DATA_CPL_ON_HIGH_QID1_F | INGRESS_SIZE_ERR_F |
+		     ERR_DATA_CPL_ON_HIGH_QID0_F | ERR_BAD_DB_PIDX3_F |
+		     ERR_BAD_DB_PIDX2_F | ERR_BAD_DB_PIDX1_F |
+		     ERR_BAD_DB_PIDX0_F | ERR_ING_CTXT_PRIO_F |
+		     DBFIFO_LP_INT_F | EGRESS_SIZE_ERR_F | val);
+	t4_write_reg(adapter, MYPF_REG(PL_PF_INT_ENABLE_A), PF_INTR_MASK);
+	t4_set_reg_field(adapter, PL_INT_MAP0_A, 0, 1 << pf);
+}
+
+/**
+ *	t4_intr_disable - disable interrupts
+ *	@adapter: the adapter whose interrupts should be disabled
+ *
+ *	Disable interrupts.  We only disable the top-level interrupt
+ *	concentrators.  The caller must be a PCI function managing global
+ *	interrupts.
+ */
+void t4_intr_disable(struct adapter *adapter)
+{
+	u32 whoami, pf;
+
+	if (pci_channel_offline(adapter->pdev))
+		return;
+
+	whoami = t4_read_reg(adapter, PL_WHOAMI_A);
+	pf = CHELSIO_CHIP_VERSION(adapter->params.chip) <= CHELSIO_T5 ?
+			SOURCEPF_G(whoami) : T6_SOURCEPF_G(whoami);
+
+	t4_write_reg(adapter, MYPF_REG(PL_PF_INT_ENABLE_A), 0);
+	t4_set_reg_field(adapter, PL_INT_MAP0_A, 1 << pf, 0);
+}
+
+unsigned int t4_chip_rss_size(struct adapter *adap)
+{
+	if (CHELSIO_CHIP_VERSION(adap->params.chip) <= CHELSIO_T5)
+		return RSS_NENTRIES;
+	else
+		return T6_RSS_NENTRIES;
+}
+
+/**
+ *	t4_config_rss_range - configure a portion of the RSS mapping table
+ *	@adapter: the adapter
+ *	@mbox: mbox to use for the FW command
+ *	@viid: virtual interface whose RSS subtable is to be written
+ *	@start: start entry in the table to write
+ *	@n: how many table entries to write
+ *	@rspq: values for the response queue lookup table
+ *	@nrspq: number of values in @rspq
+ *
+ *	Programs the selected part of the VI's RSS mapping table with the
+ *	provided values.  If @nrspq < @n the supplied values are used repeatedly
+ *	until the full table range is populated.
+ *
+ *	The caller must ensure the values in @rspq are in the range allowed for
+ *	@viid.
+ */
+int t4_config_rss_range(struct adapter *adapter, int mbox, unsigned int viid,
+			int start, int n, const u16 *rspq, unsigned int nrspq)
+{
+	int ret;
+	const u16 *rsp = rspq;
+	const u16 *rsp_end = rspq + nrspq;
+	struct fw_rss_ind_tbl_cmd cmd;
+
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_RSS_IND_TBL_CMD) |
+			       FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+			       FW_RSS_IND_TBL_CMD_VIID_V(viid));
+	cmd.retval_len16 = cpu_to_be32(FW_LEN16(cmd));
+
+	/* each fw_rss_ind_tbl_cmd takes up to 32 entries */
+	while (n > 0) {
+		int nq = min(n, 32);
+		__be32 *qp = &cmd.iq0_to_iq2;
+
+		cmd.niqid = cpu_to_be16(nq);
+		cmd.startidx = cpu_to_be16(start);
+
+		start += nq;
+		n -= nq;
+
+		while (nq > 0) {
+			unsigned int v;
+
+			v = FW_RSS_IND_TBL_CMD_IQ0_V(*rsp);
+			if (++rsp >= rsp_end)
+				rsp = rspq;
+			v |= FW_RSS_IND_TBL_CMD_IQ1_V(*rsp);
+			if (++rsp >= rsp_end)
+				rsp = rspq;
+			v |= FW_RSS_IND_TBL_CMD_IQ2_V(*rsp);
+			if (++rsp >= rsp_end)
+				rsp = rspq;
+
+			*qp++ = cpu_to_be32(v);
+			nq -= 3;
+		}
+
+		ret = t4_wr_mbox(adapter, mbox, &cmd, sizeof(cmd), NULL);
+		if (ret)
+			return ret;
+	}
+	return 0;
+}
+
+/**
+ *	t4_config_glbl_rss - configure the global RSS mode
+ *	@adapter: the adapter
+ *	@mbox: mbox to use for the FW command
+ *	@mode: global RSS mode
+ *	@flags: mode-specific flags
+ *
+ *	Sets the global RSS mode.
+ */
+int t4_config_glbl_rss(struct adapter *adapter, int mbox, unsigned int mode,
+		       unsigned int flags)
+{
+	struct fw_rss_glb_config_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_write = cpu_to_be32(FW_CMD_OP_V(FW_RSS_GLB_CONFIG_CMD) |
+				    FW_CMD_REQUEST_F | FW_CMD_WRITE_F);
+	c.retval_len16 = cpu_to_be32(FW_LEN16(c));
+	if (mode == FW_RSS_GLB_CONFIG_CMD_MODE_MANUAL) {
+		c.u.manual.mode_pkd =
+			cpu_to_be32(FW_RSS_GLB_CONFIG_CMD_MODE_V(mode));
+	} else if (mode == FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL) {
+		c.u.basicvirtual.mode_pkd =
+			cpu_to_be32(FW_RSS_GLB_CONFIG_CMD_MODE_V(mode));
+		c.u.basicvirtual.synmapen_to_hashtoeplitz = cpu_to_be32(flags);
+	} else
+		return -EINVAL;
+	return t4_wr_mbox(adapter, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_config_vi_rss - configure per VI RSS settings
+ *	@adapter: the adapter
+ *	@mbox: mbox to use for the FW command
+ *	@viid: the VI id
+ *	@flags: RSS flags
+ *	@defq: id of the default RSS queue for the VI.
+ *
+ *	Configures VI-specific RSS properties.
+ */
+int t4_config_vi_rss(struct adapter *adapter, int mbox, unsigned int viid,
+		     unsigned int flags, unsigned int defq)
+{
+	struct fw_rss_vi_config_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_RSS_VI_CONFIG_CMD) |
+				   FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+				   FW_RSS_VI_CONFIG_CMD_VIID_V(viid));
+	c.retval_len16 = cpu_to_be32(FW_LEN16(c));
+	c.u.basicvirtual.defaultq_to_udpen = cpu_to_be32(flags |
+					FW_RSS_VI_CONFIG_CMD_DEFAULTQ_V(defq));
+	return t4_wr_mbox(adapter, mbox, &c, sizeof(c), NULL);
+}
+
+/* Read an RSS table row */
+static int rd_rss_row(struct adapter *adap, int row, u32 *val)
+{
+	t4_write_reg(adap, TP_RSS_LKP_TABLE_A, 0xfff00000 | row);
+	return t4_wait_op_done_val(adap, TP_RSS_LKP_TABLE_A, LKPTBLROWVLD_F, 1,
+				   5, 0, val);
+}
+
+/**
+ *	t4_read_rss - read the contents of the RSS mapping table
+ *	@adapter: the adapter
+ *	@map: holds the contents of the RSS mapping table
+ *
+ *	Reads the contents of the RSS hash->queue mapping table.
+ */
+int t4_read_rss(struct adapter *adapter, u16 *map)
+{
+	int i, ret, nentries;
+	u32 val;
+
+	nentries = t4_chip_rss_size(adapter);
+	for (i = 0; i < nentries / 2; ++i) {
+		ret = rd_rss_row(adapter, i, &val);
+		if (ret)
+			return ret;
+		*map++ = LKPTBLQUEUE0_G(val);
+		*map++ = LKPTBLQUEUE1_G(val);
+	}
+	return 0;
+}
+
+static unsigned int t4_use_ldst(struct adapter *adap)
+{
+	return (adap->flags & CXGB4_FW_OK) && !adap->use_bd;
+}
+
+/**
+ * t4_tp_fw_ldst_rw - Access TP indirect register through LDST
+ * @adap: the adapter
+ * @cmd: TP fw ldst address space type
+ * @vals: where the indirect register values are stored/written
+ * @nregs: how many indirect registers to read/write
+ * @start_index: index of first indirect register to read/write
+ * @rw: Read (1) or Write (0)
+ * @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ * Access TP indirect registers through LDST
+ */
+static int t4_tp_fw_ldst_rw(struct adapter *adap, int cmd, u32 *vals,
+			    unsigned int nregs, unsigned int start_index,
+			    unsigned int rw, bool sleep_ok)
+{
+	int ret = 0;
+	unsigned int i;
+	struct fw_ldst_cmd c;
+
+	for (i = 0; i < nregs; i++) {
+		memset(&c, 0, sizeof(c));
+		c.op_to_addrspace = cpu_to_be32(FW_CMD_OP_V(FW_LDST_CMD) |
+						FW_CMD_REQUEST_F |
+						(rw ? FW_CMD_READ_F :
+						      FW_CMD_WRITE_F) |
+						FW_LDST_CMD_ADDRSPACE_V(cmd));
+		c.cycles_to_len16 = cpu_to_be32(FW_LEN16(c));
+
+		c.u.addrval.addr = cpu_to_be32(start_index + i);
+		c.u.addrval.val  = rw ? 0 : cpu_to_be32(vals[i]);
+		ret = t4_wr_mbox_meat(adap, adap->mbox, &c, sizeof(c), &c,
+				      sleep_ok);
+		if (ret)
+			return ret;
+
+		if (rw)
+			vals[i] = be32_to_cpu(c.u.addrval.val);
+	}
+	return 0;
+}
+
+/**
+ * t4_tp_indirect_rw - Read/Write TP indirect register through LDST or backdoor
+ * @adap: the adapter
+ * @reg_addr: Address Register
+ * @reg_data: Data register
+ * @buff: where the indirect register values are stored/written
+ * @nregs: how many indirect registers to read/write
+ * @start_index: index of first indirect register to read/write
+ * @rw: READ(1) or WRITE(0)
+ * @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ * Read/Write TP indirect registers through LDST if possible.
+ * Else, use backdoor access
+ **/
+static void t4_tp_indirect_rw(struct adapter *adap, u32 reg_addr, u32 reg_data,
+			      u32 *buff, u32 nregs, u32 start_index, int rw,
+			      bool sleep_ok)
+{
+	int rc = -EINVAL;
+	int cmd;
+
+	switch (reg_addr) {
+	case TP_PIO_ADDR_A:
+		cmd = FW_LDST_ADDRSPC_TP_PIO;
+		break;
+	case TP_TM_PIO_ADDR_A:
+		cmd = FW_LDST_ADDRSPC_TP_TM_PIO;
+		break;
+	case TP_MIB_INDEX_A:
+		cmd = FW_LDST_ADDRSPC_TP_MIB;
+		break;
+	default:
+		goto indirect_access;
+	}
+
+	if (t4_use_ldst(adap))
+		rc = t4_tp_fw_ldst_rw(adap, cmd, buff, nregs, start_index, rw,
+				      sleep_ok);
+
+indirect_access:
+
+	if (rc) {
+		if (rw)
+			t4_read_indirect(adap, reg_addr, reg_data, buff, nregs,
+					 start_index);
+		else
+			t4_write_indirect(adap, reg_addr, reg_data, buff, nregs,
+					  start_index);
+	}
+}
+
+/**
+ * t4_tp_pio_read - Read TP PIO registers
+ * @adap: the adapter
+ * @buff: where the indirect register values are written
+ * @nregs: how many indirect registers to read
+ * @start_index: index of first indirect register to read
+ * @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ * Read TP PIO Registers
+ **/
+void t4_tp_pio_read(struct adapter *adap, u32 *buff, u32 nregs,
+		    u32 start_index, bool sleep_ok)
+{
+	t4_tp_indirect_rw(adap, TP_PIO_ADDR_A, TP_PIO_DATA_A, buff, nregs,
+			  start_index, 1, sleep_ok);
+}
+
+/**
+ * t4_tp_pio_write - Write TP PIO registers
+ * @adap: the adapter
+ * @buff: where the indirect register values are stored
+ * @nregs: how many indirect registers to write
+ * @start_index: index of first indirect register to write
+ * @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ * Write TP PIO Registers
+ **/
+static void t4_tp_pio_write(struct adapter *adap, u32 *buff, u32 nregs,
+			    u32 start_index, bool sleep_ok)
+{
+	t4_tp_indirect_rw(adap, TP_PIO_ADDR_A, TP_PIO_DATA_A, buff, nregs,
+			  start_index, 0, sleep_ok);
+}
+
+/**
+ * t4_tp_tm_pio_read - Read TP TM PIO registers
+ * @adap: the adapter
+ * @buff: where the indirect register values are written
+ * @nregs: how many indirect registers to read
+ * @start_index: index of first indirect register to read
+ * @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ * Read TP TM PIO Registers
+ **/
+void t4_tp_tm_pio_read(struct adapter *adap, u32 *buff, u32 nregs,
+		       u32 start_index, bool sleep_ok)
+{
+	t4_tp_indirect_rw(adap, TP_TM_PIO_ADDR_A, TP_TM_PIO_DATA_A, buff,
+			  nregs, start_index, 1, sleep_ok);
+}
+
+/**
+ * t4_tp_mib_read - Read TP MIB registers
+ * @adap: the adapter
+ * @buff: where the indirect register values are written
+ * @nregs: how many indirect registers to read
+ * @start_index: index of first indirect register to read
+ * @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ * Read TP MIB Registers
+ **/
+void t4_tp_mib_read(struct adapter *adap, u32 *buff, u32 nregs, u32 start_index,
+		    bool sleep_ok)
+{
+	t4_tp_indirect_rw(adap, TP_MIB_INDEX_A, TP_MIB_DATA_A, buff, nregs,
+			  start_index, 1, sleep_ok);
+}
+
+/**
+ *	t4_read_rss_key - read the global RSS key
+ *	@adap: the adapter
+ *	@key: 10-entry array holding the 320-bit RSS key
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Reads the global 320-bit RSS key.
+ */
+void t4_read_rss_key(struct adapter *adap, u32 *key, bool sleep_ok)
+{
+	t4_tp_pio_read(adap, key, 10, TP_RSS_SECRET_KEY0_A, sleep_ok);
+}
+
+/**
+ *	t4_write_rss_key - program one of the RSS keys
+ *	@adap: the adapter
+ *	@key: 10-entry array holding the 320-bit RSS key
+ *	@idx: which RSS key to write
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Writes one of the RSS keys with the given 320-bit value.  If @idx is
+ *	0..15 the corresponding entry in the RSS key table is written,
+ *	otherwise the global RSS key is written.
+ */
+void t4_write_rss_key(struct adapter *adap, const u32 *key, int idx,
+		      bool sleep_ok)
+{
+	u8 rss_key_addr_cnt = 16;
+	u32 vrt = t4_read_reg(adap, TP_RSS_CONFIG_VRT_A);
+
+	/* T6 and later: for KeyMode 3 (per-vf and per-vf scramble),
+	 * allows access to key addresses 16-63 by using KeyWrAddrX
+	 * as index[5:4](upper 2) into key table
+	 */
+	if ((CHELSIO_CHIP_VERSION(adap->params.chip) > CHELSIO_T5) &&
+	    (vrt & KEYEXTEND_F) && (KEYMODE_G(vrt) == 3))
+		rss_key_addr_cnt = 32;
+
+	t4_tp_pio_write(adap, (void *)key, 10, TP_RSS_SECRET_KEY0_A, sleep_ok);
+
+	if (idx >= 0 && idx < rss_key_addr_cnt) {
+		if (rss_key_addr_cnt > 16)
+			t4_write_reg(adap, TP_RSS_CONFIG_VRT_A,
+				     KEYWRADDRX_V(idx >> 4) |
+				     T6_VFWRADDR_V(idx) | KEYWREN_F);
+		else
+			t4_write_reg(adap, TP_RSS_CONFIG_VRT_A,
+				     KEYWRADDR_V(idx) | KEYWREN_F);
+	}
+}
+
+/**
+ *	t4_read_rss_pf_config - read PF RSS Configuration Table
+ *	@adapter: the adapter
+ *	@index: the entry in the PF RSS table to read
+ *	@valp: where to store the returned value
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Reads the PF RSS Configuration Table at the specified index and returns
+ *	the value found there.
+ */
+void t4_read_rss_pf_config(struct adapter *adapter, unsigned int index,
+			   u32 *valp, bool sleep_ok)
+{
+	t4_tp_pio_read(adapter, valp, 1, TP_RSS_PF0_CONFIG_A + index, sleep_ok);
+}
+
+/**
+ *	t4_read_rss_vf_config - read VF RSS Configuration Table
+ *	@adapter: the adapter
+ *	@index: the entry in the VF RSS table to read
+ *	@vfl: where to store the returned VFL
+ *	@vfh: where to store the returned VFH
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Reads the VF RSS Configuration Table at the specified index and returns
+ *	the (VFL, VFH) values found there.
+ */
+void t4_read_rss_vf_config(struct adapter *adapter, unsigned int index,
+			   u32 *vfl, u32 *vfh, bool sleep_ok)
+{
+	u32 vrt, mask, data;
+
+	if (CHELSIO_CHIP_VERSION(adapter->params.chip) <= CHELSIO_T5) {
+		mask = VFWRADDR_V(VFWRADDR_M);
+		data = VFWRADDR_V(index);
+	} else {
+		 mask =  T6_VFWRADDR_V(T6_VFWRADDR_M);
+		 data = T6_VFWRADDR_V(index);
+	}
+
+	/* Request that the index'th VF Table values be read into VFL/VFH.
+	 */
+	vrt = t4_read_reg(adapter, TP_RSS_CONFIG_VRT_A);
+	vrt &= ~(VFRDRG_F | VFWREN_F | KEYWREN_F | mask);
+	vrt |= data | VFRDEN_F;
+	t4_write_reg(adapter, TP_RSS_CONFIG_VRT_A, vrt);
+
+	/* Grab the VFL/VFH values ...
+	 */
+	t4_tp_pio_read(adapter, vfl, 1, TP_RSS_VFL_CONFIG_A, sleep_ok);
+	t4_tp_pio_read(adapter, vfh, 1, TP_RSS_VFH_CONFIG_A, sleep_ok);
+}
+
+/**
+ *	t4_read_rss_pf_map - read PF RSS Map
+ *	@adapter: the adapter
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Reads the PF RSS Map register and returns its value.
+ */
+u32 t4_read_rss_pf_map(struct adapter *adapter, bool sleep_ok)
+{
+	u32 pfmap;
+
+	t4_tp_pio_read(adapter, &pfmap, 1, TP_RSS_PF_MAP_A, sleep_ok);
+	return pfmap;
+}
+
+/**
+ *	t4_read_rss_pf_mask - read PF RSS Mask
+ *	@adapter: the adapter
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Reads the PF RSS Mask register and returns its value.
+ */
+u32 t4_read_rss_pf_mask(struct adapter *adapter, bool sleep_ok)
+{
+	u32 pfmask;
+
+	t4_tp_pio_read(adapter, &pfmask, 1, TP_RSS_PF_MSK_A, sleep_ok);
+	return pfmask;
+}
+
+/**
+ *	t4_tp_get_tcp_stats - read TP's TCP MIB counters
+ *	@adap: the adapter
+ *	@v4: holds the TCP/IP counter values
+ *	@v6: holds the TCP/IPv6 counter values
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Returns the values of TP's TCP/IP and TCP/IPv6 MIB counters.
+ *	Either @v4 or @v6 may be %NULL to skip the corresponding stats.
+ */
+void t4_tp_get_tcp_stats(struct adapter *adap, struct tp_tcp_stats *v4,
+			 struct tp_tcp_stats *v6, bool sleep_ok)
+{
+	u32 val[TP_MIB_TCP_RXT_SEG_LO_A - TP_MIB_TCP_OUT_RST_A + 1];
+
+#define STAT_IDX(x) ((TP_MIB_TCP_##x##_A) - TP_MIB_TCP_OUT_RST_A)
+#define STAT(x)     val[STAT_IDX(x)]
+#define STAT64(x)   (((u64)STAT(x##_HI) << 32) | STAT(x##_LO))
+
+	if (v4) {
+		t4_tp_mib_read(adap, val, ARRAY_SIZE(val),
+			       TP_MIB_TCP_OUT_RST_A, sleep_ok);
+		v4->tcp_out_rsts = STAT(OUT_RST);
+		v4->tcp_in_segs  = STAT64(IN_SEG);
+		v4->tcp_out_segs = STAT64(OUT_SEG);
+		v4->tcp_retrans_segs = STAT64(RXT_SEG);
+	}
+	if (v6) {
+		t4_tp_mib_read(adap, val, ARRAY_SIZE(val),
+			       TP_MIB_TCP_V6OUT_RST_A, sleep_ok);
+		v6->tcp_out_rsts = STAT(OUT_RST);
+		v6->tcp_in_segs  = STAT64(IN_SEG);
+		v6->tcp_out_segs = STAT64(OUT_SEG);
+		v6->tcp_retrans_segs = STAT64(RXT_SEG);
+	}
+#undef STAT64
+#undef STAT
+#undef STAT_IDX
+}
+
+/**
+ *	t4_tp_get_err_stats - read TP's error MIB counters
+ *	@adap: the adapter
+ *	@st: holds the counter values
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Returns the values of TP's error counters.
+ */
+void t4_tp_get_err_stats(struct adapter *adap, struct tp_err_stats *st,
+			 bool sleep_ok)
+{
+	int nchan = adap->params.arch.nchan;
+
+	t4_tp_mib_read(adap, st->mac_in_errs, nchan, TP_MIB_MAC_IN_ERR_0_A,
+		       sleep_ok);
+	t4_tp_mib_read(adap, st->hdr_in_errs, nchan, TP_MIB_HDR_IN_ERR_0_A,
+		       sleep_ok);
+	t4_tp_mib_read(adap, st->tcp_in_errs, nchan, TP_MIB_TCP_IN_ERR_0_A,
+		       sleep_ok);
+	t4_tp_mib_read(adap, st->tnl_cong_drops, nchan,
+		       TP_MIB_TNL_CNG_DROP_0_A, sleep_ok);
+	t4_tp_mib_read(adap, st->ofld_chan_drops, nchan,
+		       TP_MIB_OFD_CHN_DROP_0_A, sleep_ok);
+	t4_tp_mib_read(adap, st->tnl_tx_drops, nchan, TP_MIB_TNL_DROP_0_A,
+		       sleep_ok);
+	t4_tp_mib_read(adap, st->ofld_vlan_drops, nchan,
+		       TP_MIB_OFD_VLN_DROP_0_A, sleep_ok);
+	t4_tp_mib_read(adap, st->tcp6_in_errs, nchan,
+		       TP_MIB_TCP_V6IN_ERR_0_A, sleep_ok);
+	t4_tp_mib_read(adap, &st->ofld_no_neigh, 2, TP_MIB_OFD_ARP_DROP_A,
+		       sleep_ok);
+}
+
+/**
+ *	t4_tp_get_cpl_stats - read TP's CPL MIB counters
+ *	@adap: the adapter
+ *	@st: holds the counter values
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Returns the values of TP's CPL counters.
+ */
+void t4_tp_get_cpl_stats(struct adapter *adap, struct tp_cpl_stats *st,
+			 bool sleep_ok)
+{
+	int nchan = adap->params.arch.nchan;
+
+	t4_tp_mib_read(adap, st->req, nchan, TP_MIB_CPL_IN_REQ_0_A, sleep_ok);
+
+	t4_tp_mib_read(adap, st->rsp, nchan, TP_MIB_CPL_OUT_RSP_0_A, sleep_ok);
+}
+
+/**
+ *	t4_tp_get_rdma_stats - read TP's RDMA MIB counters
+ *	@adap: the adapter
+ *	@st: holds the counter values
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Returns the values of TP's RDMA counters.
+ */
+void t4_tp_get_rdma_stats(struct adapter *adap, struct tp_rdma_stats *st,
+			  bool sleep_ok)
+{
+	t4_tp_mib_read(adap, &st->rqe_dfr_pkt, 2, TP_MIB_RQE_DFR_PKT_A,
+		       sleep_ok);
+}
+
+/**
+ *	t4_get_fcoe_stats - read TP's FCoE MIB counters for a port
+ *	@adap: the adapter
+ *	@idx: the port index
+ *	@st: holds the counter values
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Returns the values of TP's FCoE counters for the selected port.
+ */
+void t4_get_fcoe_stats(struct adapter *adap, unsigned int idx,
+		       struct tp_fcoe_stats *st, bool sleep_ok)
+{
+	u32 val[2];
+
+	t4_tp_mib_read(adap, &st->frames_ddp, 1, TP_MIB_FCOE_DDP_0_A + idx,
+		       sleep_ok);
+
+	t4_tp_mib_read(adap, &st->frames_drop, 1,
+		       TP_MIB_FCOE_DROP_0_A + idx, sleep_ok);
+
+	t4_tp_mib_read(adap, val, 2, TP_MIB_FCOE_BYTE_0_HI_A + 2 * idx,
+		       sleep_ok);
+
+	st->octets_ddp = ((u64)val[0] << 32) | val[1];
+}
+
+/**
+ *	t4_get_usm_stats - read TP's non-TCP DDP MIB counters
+ *	@adap: the adapter
+ *	@st: holds the counter values
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Returns the values of TP's counters for non-TCP directly-placed packets.
+ */
+void t4_get_usm_stats(struct adapter *adap, struct tp_usm_stats *st,
+		      bool sleep_ok)
+{
+	u32 val[4];
+
+	t4_tp_mib_read(adap, val, 4, TP_MIB_USM_PKTS_A, sleep_ok);
+	st->frames = val[0];
+	st->drops = val[1];
+	st->octets = ((u64)val[2] << 32) | val[3];
+}
+
+/**
+ *	t4_read_mtu_tbl - returns the values in the HW path MTU table
+ *	@adap: the adapter
+ *	@mtus: where to store the MTU values
+ *	@mtu_log: where to store the MTU base-2 log (may be %NULL)
+ *
+ *	Reads the HW path MTU table.
+ */
+void t4_read_mtu_tbl(struct adapter *adap, u16 *mtus, u8 *mtu_log)
+{
+	u32 v;
+	int i;
+
+	for (i = 0; i < NMTUS; ++i) {
+		t4_write_reg(adap, TP_MTU_TABLE_A,
+			     MTUINDEX_V(0xff) | MTUVALUE_V(i));
+		v = t4_read_reg(adap, TP_MTU_TABLE_A);
+		mtus[i] = MTUVALUE_G(v);
+		if (mtu_log)
+			mtu_log[i] = MTUWIDTH_G(v);
+	}
+}
+
+/**
+ *	t4_read_cong_tbl - reads the congestion control table
+ *	@adap: the adapter
+ *	@incr: where to store the alpha values
+ *
+ *	Reads the additive increments programmed into the HW congestion
+ *	control table.
+ */
+void t4_read_cong_tbl(struct adapter *adap, u16 incr[NMTUS][NCCTRL_WIN])
+{
+	unsigned int mtu, w;
+
+	for (mtu = 0; mtu < NMTUS; ++mtu)
+		for (w = 0; w < NCCTRL_WIN; ++w) {
+			t4_write_reg(adap, TP_CCTRL_TABLE_A,
+				     ROWINDEX_V(0xffff) | (mtu << 5) | w);
+			incr[mtu][w] = (u16)t4_read_reg(adap,
+						TP_CCTRL_TABLE_A) & 0x1fff;
+		}
+}
+
+/**
+ *	t4_tp_wr_bits_indirect - set/clear bits in an indirect TP register
+ *	@adap: the adapter
+ *	@addr: the indirect TP register address
+ *	@mask: specifies the field within the register to modify
+ *	@val: new value for the field
+ *
+ *	Sets a field of an indirect TP register to the given value.
+ */
+void t4_tp_wr_bits_indirect(struct adapter *adap, unsigned int addr,
+			    unsigned int mask, unsigned int val)
+{
+	t4_write_reg(adap, TP_PIO_ADDR_A, addr);
+	val |= t4_read_reg(adap, TP_PIO_DATA_A) & ~mask;
+	t4_write_reg(adap, TP_PIO_DATA_A, val);
+}
+
+/**
+ *	init_cong_ctrl - initialize congestion control parameters
+ *	@a: the alpha values for congestion control
+ *	@b: the beta values for congestion control
+ *
+ *	Initialize the congestion control parameters.
+ */
+static void init_cong_ctrl(unsigned short *a, unsigned short *b)
+{
+	a[0] = a[1] = a[2] = a[3] = a[4] = a[5] = a[6] = a[7] = a[8] = 1;
+	a[9] = 2;
+	a[10] = 3;
+	a[11] = 4;
+	a[12] = 5;
+	a[13] = 6;
+	a[14] = 7;
+	a[15] = 8;
+	a[16] = 9;
+	a[17] = 10;
+	a[18] = 14;
+	a[19] = 17;
+	a[20] = 21;
+	a[21] = 25;
+	a[22] = 30;
+	a[23] = 35;
+	a[24] = 45;
+	a[25] = 60;
+	a[26] = 80;
+	a[27] = 100;
+	a[28] = 200;
+	a[29] = 300;
+	a[30] = 400;
+	a[31] = 500;
+
+	b[0] = b[1] = b[2] = b[3] = b[4] = b[5] = b[6] = b[7] = b[8] = 0;
+	b[9] = b[10] = 1;
+	b[11] = b[12] = 2;
+	b[13] = b[14] = b[15] = b[16] = 3;
+	b[17] = b[18] = b[19] = b[20] = b[21] = 4;
+	b[22] = b[23] = b[24] = b[25] = b[26] = b[27] = 5;
+	b[28] = b[29] = 6;
+	b[30] = b[31] = 7;
+}
+
+/* The minimum additive increment value for the congestion control table */
+#define CC_MIN_INCR 2U
+
+/**
+ *	t4_load_mtus - write the MTU and congestion control HW tables
+ *	@adap: the adapter
+ *	@mtus: the values for the MTU table
+ *	@alpha: the values for the congestion control alpha parameter
+ *	@beta: the values for the congestion control beta parameter
+ *
+ *	Write the HW MTU table with the supplied MTUs and the high-speed
+ *	congestion control table with the supplied alpha, beta, and MTUs.
+ *	We write the two tables together because the additive increments
+ *	depend on the MTUs.
+ */
+void t4_load_mtus(struct adapter *adap, const unsigned short *mtus,
+		  const unsigned short *alpha, const unsigned short *beta)
+{
+	static const unsigned int avg_pkts[NCCTRL_WIN] = {
+		2, 6, 10, 14, 20, 28, 40, 56, 80, 112, 160, 224, 320, 448, 640,
+		896, 1281, 1792, 2560, 3584, 5120, 7168, 10240, 14336, 20480,
+		28672, 40960, 57344, 81920, 114688, 163840, 229376
+	};
+
+	unsigned int i, w;
+
+	for (i = 0; i < NMTUS; ++i) {
+		unsigned int mtu = mtus[i];
+		unsigned int log2 = fls(mtu);
+
+		if (!(mtu & ((1 << log2) >> 2)))     /* round */
+			log2--;
+		t4_write_reg(adap, TP_MTU_TABLE_A, MTUINDEX_V(i) |
+			     MTUWIDTH_V(log2) | MTUVALUE_V(mtu));
+
+		for (w = 0; w < NCCTRL_WIN; ++w) {
+			unsigned int inc;
+
+			inc = max(((mtu - 40) * alpha[w]) / avg_pkts[w],
+				  CC_MIN_INCR);
+
+			t4_write_reg(adap, TP_CCTRL_TABLE_A, (i << 21) |
+				     (w << 16) | (beta[w] << 13) | inc);
+		}
+	}
+}
+
+/* Calculates a rate in bytes/s given the number of 256-byte units per 4K core
+ * clocks.  The formula is
+ *
+ * bytes/s = bytes256 * 256 * ClkFreq / 4096
+ *
+ * which is equivalent to
+ *
+ * bytes/s = 62.5 * bytes256 * ClkFreq_ms
+ */
+static u64 chan_rate(struct adapter *adap, unsigned int bytes256)
+{
+	u64 v = bytes256 * adap->params.vpd.cclk;
+
+	return v * 62 + v / 2;
+}
+
+/**
+ *	t4_get_chan_txrate - get the current per channel Tx rates
+ *	@adap: the adapter
+ *	@nic_rate: rates for NIC traffic
+ *	@ofld_rate: rates for offloaded traffic
+ *
+ *	Return the current Tx rates in bytes/s for NIC and offloaded traffic
+ *	for each channel.
+ */
+void t4_get_chan_txrate(struct adapter *adap, u64 *nic_rate, u64 *ofld_rate)
+{
+	u32 v;
+
+	v = t4_read_reg(adap, TP_TX_TRATE_A);
+	nic_rate[0] = chan_rate(adap, TNLRATE0_G(v));
+	nic_rate[1] = chan_rate(adap, TNLRATE1_G(v));
+	if (adap->params.arch.nchan == NCHAN) {
+		nic_rate[2] = chan_rate(adap, TNLRATE2_G(v));
+		nic_rate[3] = chan_rate(adap, TNLRATE3_G(v));
+	}
+
+	v = t4_read_reg(adap, TP_TX_ORATE_A);
+	ofld_rate[0] = chan_rate(adap, OFDRATE0_G(v));
+	ofld_rate[1] = chan_rate(adap, OFDRATE1_G(v));
+	if (adap->params.arch.nchan == NCHAN) {
+		ofld_rate[2] = chan_rate(adap, OFDRATE2_G(v));
+		ofld_rate[3] = chan_rate(adap, OFDRATE3_G(v));
+	}
+}
+
+/**
+ *	t4_set_trace_filter - configure one of the tracing filters
+ *	@adap: the adapter
+ *	@tp: the desired trace filter parameters
+ *	@idx: which filter to configure
+ *	@enable: whether to enable or disable the filter
+ *
+ *	Configures one of the tracing filters available in HW.  If @enable is
+ *	%0 @tp is not examined and may be %NULL. The user is responsible to
+ *	set the single/multiple trace mode by writing to MPS_TRC_CFG_A register
+ */
+int t4_set_trace_filter(struct adapter *adap, const struct trace_params *tp,
+			int idx, int enable)
+{
+	int i, ofst = idx * 4;
+	u32 data_reg, mask_reg, cfg;
+
+	if (!enable) {
+		t4_write_reg(adap, MPS_TRC_FILTER_MATCH_CTL_A_A + ofst, 0);
+		return 0;
+	}
+
+	cfg = t4_read_reg(adap, MPS_TRC_CFG_A);
+	if (cfg & TRCMULTIFILTER_F) {
+		/* If multiple tracers are enabled, then maximum
+		 * capture size is 2.5KB (FIFO size of a single channel)
+		 * minus 2 flits for CPL_TRACE_PKT header.
+		 */
+		if (tp->snap_len > ((10 * 1024 / 4) - (2 * 8)))
+			return -EINVAL;
+	} else {
+		/* If multiple tracers are disabled, to avoid deadlocks
+		 * maximum packet capture size of 9600 bytes is recommended.
+		 * Also in this mode, only trace0 can be enabled and running.
+		 */
+		if (tp->snap_len > 9600 || idx)
+			return -EINVAL;
+	}
+
+	if (tp->port > (is_t4(adap->params.chip) ? 11 : 19) || tp->invert > 1 ||
+	    tp->skip_len > TFLENGTH_M || tp->skip_ofst > TFOFFSET_M ||
+	    tp->min_len > TFMINPKTSIZE_M)
+		return -EINVAL;
+
+	/* stop the tracer we'll be changing */
+	t4_write_reg(adap, MPS_TRC_FILTER_MATCH_CTL_A_A + ofst, 0);
+
+	idx *= (MPS_TRC_FILTER1_MATCH_A - MPS_TRC_FILTER0_MATCH_A);
+	data_reg = MPS_TRC_FILTER0_MATCH_A + idx;
+	mask_reg = MPS_TRC_FILTER0_DONT_CARE_A + idx;
+
+	for (i = 0; i < TRACE_LEN / 4; i++, data_reg += 4, mask_reg += 4) {
+		t4_write_reg(adap, data_reg, tp->data[i]);
+		t4_write_reg(adap, mask_reg, ~tp->mask[i]);
+	}
+	t4_write_reg(adap, MPS_TRC_FILTER_MATCH_CTL_B_A + ofst,
+		     TFCAPTUREMAX_V(tp->snap_len) |
+		     TFMINPKTSIZE_V(tp->min_len));
+	t4_write_reg(adap, MPS_TRC_FILTER_MATCH_CTL_A_A + ofst,
+		     TFOFFSET_V(tp->skip_ofst) | TFLENGTH_V(tp->skip_len) |
+		     (is_t4(adap->params.chip) ?
+		     TFPORT_V(tp->port) | TFEN_F | TFINVERTMATCH_V(tp->invert) :
+		     T5_TFPORT_V(tp->port) | T5_TFEN_F |
+		     T5_TFINVERTMATCH_V(tp->invert)));
+
+	return 0;
+}
+
+/**
+ *	t4_get_trace_filter - query one of the tracing filters
+ *	@adap: the adapter
+ *	@tp: the current trace filter parameters
+ *	@idx: which trace filter to query
+ *	@enabled: non-zero if the filter is enabled
+ *
+ *	Returns the current settings of one of the HW tracing filters.
+ */
+void t4_get_trace_filter(struct adapter *adap, struct trace_params *tp, int idx,
+			 int *enabled)
+{
+	u32 ctla, ctlb;
+	int i, ofst = idx * 4;
+	u32 data_reg, mask_reg;
+
+	ctla = t4_read_reg(adap, MPS_TRC_FILTER_MATCH_CTL_A_A + ofst);
+	ctlb = t4_read_reg(adap, MPS_TRC_FILTER_MATCH_CTL_B_A + ofst);
+
+	if (is_t4(adap->params.chip)) {
+		*enabled = !!(ctla & TFEN_F);
+		tp->port =  TFPORT_G(ctla);
+		tp->invert = !!(ctla & TFINVERTMATCH_F);
+	} else {
+		*enabled = !!(ctla & T5_TFEN_F);
+		tp->port = T5_TFPORT_G(ctla);
+		tp->invert = !!(ctla & T5_TFINVERTMATCH_F);
+	}
+	tp->snap_len = TFCAPTUREMAX_G(ctlb);
+	tp->min_len = TFMINPKTSIZE_G(ctlb);
+	tp->skip_ofst = TFOFFSET_G(ctla);
+	tp->skip_len = TFLENGTH_G(ctla);
+
+	ofst = (MPS_TRC_FILTER1_MATCH_A - MPS_TRC_FILTER0_MATCH_A) * idx;
+	data_reg = MPS_TRC_FILTER0_MATCH_A + ofst;
+	mask_reg = MPS_TRC_FILTER0_DONT_CARE_A + ofst;
+
+	for (i = 0; i < TRACE_LEN / 4; i++, data_reg += 4, mask_reg += 4) {
+		tp->mask[i] = ~t4_read_reg(adap, mask_reg);
+		tp->data[i] = t4_read_reg(adap, data_reg) & tp->mask[i];
+	}
+}
+
+/**
+ *	t4_pmtx_get_stats - returns the HW stats from PMTX
+ *	@adap: the adapter
+ *	@cnt: where to store the count statistics
+ *	@cycles: where to store the cycle statistics
+ *
+ *	Returns performance statistics from PMTX.
+ */
+void t4_pmtx_get_stats(struct adapter *adap, u32 cnt[], u64 cycles[])
+{
+	int i;
+	u32 data[2];
+
+	for (i = 0; i < adap->params.arch.pm_stats_cnt; i++) {
+		t4_write_reg(adap, PM_TX_STAT_CONFIG_A, i + 1);
+		cnt[i] = t4_read_reg(adap, PM_TX_STAT_COUNT_A);
+		if (is_t4(adap->params.chip)) {
+			cycles[i] = t4_read_reg64(adap, PM_TX_STAT_LSB_A);
+		} else {
+			t4_read_indirect(adap, PM_TX_DBG_CTRL_A,
+					 PM_TX_DBG_DATA_A, data, 2,
+					 PM_TX_DBG_STAT_MSB_A);
+			cycles[i] = (((u64)data[0] << 32) | data[1]);
+		}
+	}
+}
+
+/**
+ *	t4_pmrx_get_stats - returns the HW stats from PMRX
+ *	@adap: the adapter
+ *	@cnt: where to store the count statistics
+ *	@cycles: where to store the cycle statistics
+ *
+ *	Returns performance statistics from PMRX.
+ */
+void t4_pmrx_get_stats(struct adapter *adap, u32 cnt[], u64 cycles[])
+{
+	int i;
+	u32 data[2];
+
+	for (i = 0; i < adap->params.arch.pm_stats_cnt; i++) {
+		t4_write_reg(adap, PM_RX_STAT_CONFIG_A, i + 1);
+		cnt[i] = t4_read_reg(adap, PM_RX_STAT_COUNT_A);
+		if (is_t4(adap->params.chip)) {
+			cycles[i] = t4_read_reg64(adap, PM_RX_STAT_LSB_A);
+		} else {
+			t4_read_indirect(adap, PM_RX_DBG_CTRL_A,
+					 PM_RX_DBG_DATA_A, data, 2,
+					 PM_RX_DBG_STAT_MSB_A);
+			cycles[i] = (((u64)data[0] << 32) | data[1]);
+		}
+	}
+}
+
+/**
+ *	compute_mps_bg_map - compute the MPS Buffer Group Map for a Port
+ *	@adapter: the adapter
+ *	@pidx: the port index
+ *
+ *	Computes and returns a bitmap indicating which MPS buffer groups are
+ *	associated with the given Port.  Bit i is set if buffer group i is
+ *	used by the Port.
+ */
+static inline unsigned int compute_mps_bg_map(struct adapter *adapter,
+					      int pidx)
+{
+	unsigned int chip_version, nports;
+
+	chip_version = CHELSIO_CHIP_VERSION(adapter->params.chip);
+	nports = 1 << NUMPORTS_G(t4_read_reg(adapter, MPS_CMN_CTL_A));
+
+	switch (chip_version) {
+	case CHELSIO_T4:
+	case CHELSIO_T5:
+		switch (nports) {
+		case 1: return 0xf;
+		case 2: return 3 << (2 * pidx);
+		case 4: return 1 << pidx;
+		}
+		break;
+
+	case CHELSIO_T6:
+		switch (nports) {
+		case 2: return 1 << (2 * pidx);
+		}
+		break;
+	}
+
+	dev_err(adapter->pdev_dev, "Need MPS Buffer Group Map for Chip %0x, Nports %d\n",
+		chip_version, nports);
+
+	return 0;
+}
+
+/**
+ *	t4_get_mps_bg_map - return the buffer groups associated with a port
+ *	@adapter: the adapter
+ *	@pidx: the port index
+ *
+ *	Returns a bitmap indicating which MPS buffer groups are associated
+ *	with the given Port.  Bit i is set if buffer group i is used by the
+ *	Port.
+ */
+unsigned int t4_get_mps_bg_map(struct adapter *adapter, int pidx)
+{
+	u8 *mps_bg_map;
+	unsigned int nports;
+
+	nports = 1 << NUMPORTS_G(t4_read_reg(adapter, MPS_CMN_CTL_A));
+	if (pidx >= nports) {
+		CH_WARN(adapter, "MPS Port Index %d >= Nports %d\n",
+			pidx, nports);
+		return 0;
+	}
+
+	/* If we've already retrieved/computed this, just return the result.
+	 */
+	mps_bg_map = adapter->params.mps_bg_map;
+	if (mps_bg_map[pidx])
+		return mps_bg_map[pidx];
+
+	/* Newer Firmware can tell us what the MPS Buffer Group Map is.
+	 * If we're talking to such Firmware, let it tell us.  If the new
+	 * API isn't supported, revert back to old hardcoded way.  The value
+	 * obtained from Firmware is encoded in below format:
+	 *
+	 * val = (( MPSBGMAP[Port 3] << 24 ) |
+	 *        ( MPSBGMAP[Port 2] << 16 ) |
+	 *        ( MPSBGMAP[Port 1] <<  8 ) |
+	 *        ( MPSBGMAP[Port 0] <<  0 ))
+	 */
+	if (adapter->flags & CXGB4_FW_OK) {
+		u32 param, val;
+		int ret;
+
+		param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+			 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_MPSBGMAP));
+		ret = t4_query_params_ns(adapter, adapter->mbox, adapter->pf,
+					 0, 1, &param, &val);
+		if (!ret) {
+			int p;
+
+			/* Store the BG Map for all of the Ports in order to
+			 * avoid more calls to the Firmware in the future.
+			 */
+			for (p = 0; p < MAX_NPORTS; p++, val >>= 8)
+				mps_bg_map[p] = val & 0xff;
+
+			return mps_bg_map[pidx];
+		}
+	}
+
+	/* Either we're not talking to the Firmware or we're dealing with
+	 * older Firmware which doesn't support the new API to get the MPS
+	 * Buffer Group Map.  Fall back to computing it ourselves.
+	 */
+	mps_bg_map[pidx] = compute_mps_bg_map(adapter, pidx);
+	return mps_bg_map[pidx];
+}
+
+/**
+ *      t4_get_tp_e2c_map - return the E2C channel map associated with a port
+ *      @adapter: the adapter
+ *      @pidx: the port index
+ */
+static unsigned int t4_get_tp_e2c_map(struct adapter *adapter, int pidx)
+{
+	unsigned int nports;
+	u32 param, val = 0;
+	int ret;
+
+	nports = 1 << NUMPORTS_G(t4_read_reg(adapter, MPS_CMN_CTL_A));
+	if (pidx >= nports) {
+		CH_WARN(adapter, "TP E2C Channel Port Index %d >= Nports %d\n",
+			pidx, nports);
+		return 0;
+	}
+
+	/* FW version >= 1.16.44.0 can determine E2C channel map using
+	 * FW_PARAMS_PARAM_DEV_TPCHMAP API.
+	 */
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_TPCHMAP));
+	ret = t4_query_params_ns(adapter, adapter->mbox, adapter->pf,
+				 0, 1, &param, &val);
+	if (!ret)
+		return (val >> (8 * pidx)) & 0xff;
+
+	return 0;
+}
+
+/**
+ *	t4_get_tp_ch_map - return TP ingress channels associated with a port
+ *	@adap: the adapter
+ *	@pidx: the port index
+ *
+ *	Returns a bitmap indicating which TP Ingress Channels are associated
+ *	with a given Port.  Bit i is set if TP Ingress Channel i is used by
+ *	the Port.
+ */
+unsigned int t4_get_tp_ch_map(struct adapter *adap, int pidx)
+{
+	unsigned int chip_version = CHELSIO_CHIP_VERSION(adap->params.chip);
+	unsigned int nports = 1 << NUMPORTS_G(t4_read_reg(adap, MPS_CMN_CTL_A));
+
+	if (pidx >= nports) {
+		dev_warn(adap->pdev_dev, "TP Port Index %d >= Nports %d\n",
+			 pidx, nports);
+		return 0;
+	}
+
+	switch (chip_version) {
+	case CHELSIO_T4:
+	case CHELSIO_T5:
+		/* Note that this happens to be the same values as the MPS
+		 * Buffer Group Map for these Chips.  But we replicate the code
+		 * here because they're really separate concepts.
+		 */
+		switch (nports) {
+		case 1: return 0xf;
+		case 2: return 3 << (2 * pidx);
+		case 4: return 1 << pidx;
+		}
+		break;
+
+	case CHELSIO_T6:
+		switch (nports) {
+		case 1:
+		case 2: return 1 << pidx;
+		}
+		break;
+	}
+
+	dev_err(adap->pdev_dev, "Need TP Channel Map for Chip %0x, Nports %d\n",
+		chip_version, nports);
+	return 0;
+}
+
+/**
+ *      t4_get_port_type_description - return Port Type string description
+ *      @port_type: firmware Port Type enumeration
+ */
+const char *t4_get_port_type_description(enum fw_port_type port_type)
+{
+	static const char *const port_type_description[] = {
+		"Fiber_XFI",
+		"Fiber_XAUI",
+		"BT_SGMII",
+		"BT_XFI",
+		"BT_XAUI",
+		"KX4",
+		"CX4",
+		"KX",
+		"KR",
+		"SFP",
+		"BP_AP",
+		"BP4_AP",
+		"QSFP_10G",
+		"QSA",
+		"QSFP",
+		"BP40_BA",
+		"KR4_100G",
+		"CR4_QSFP",
+		"CR_QSFP",
+		"CR2_QSFP",
+		"SFP28",
+		"KR_SFP28",
+		"KR_XLAUI"
+	};
+
+	if (port_type < ARRAY_SIZE(port_type_description))
+		return port_type_description[port_type];
+	return "UNKNOWN";
+}
+
+/**
+ *      t4_get_port_stats_offset - collect port stats relative to a previous
+ *                                 snapshot
+ *      @adap: The adapter
+ *      @idx: The port
+ *      @stats: Current stats to fill
+ *      @offset: Previous stats snapshot
+ */
+void t4_get_port_stats_offset(struct adapter *adap, int idx,
+			      struct port_stats *stats,
+			      struct port_stats *offset)
+{
+	u64 *s, *o;
+	int i;
+
+	t4_get_port_stats(adap, idx, stats);
+	for (i = 0, s = (u64 *)stats, o = (u64 *)offset;
+			i < (sizeof(struct port_stats) / sizeof(u64));
+			i++, s++, o++)
+		*s -= *o;
+}
+
+/**
+ *	t4_get_port_stats - collect port statistics
+ *	@adap: the adapter
+ *	@idx: the port index
+ *	@p: the stats structure to fill
+ *
+ *	Collect statistics related to the given port from HW.
+ */
+void t4_get_port_stats(struct adapter *adap, int idx, struct port_stats *p)
+{
+	u32 bgmap = t4_get_mps_bg_map(adap, idx);
+	u32 stat_ctl = t4_read_reg(adap, MPS_STAT_CTL_A);
+
+#define GET_STAT(name) \
+	t4_read_reg64(adap, \
+	(is_t4(adap->params.chip) ? PORT_REG(idx, MPS_PORT_STAT_##name##_L) : \
+	T5_PORT_REG(idx, MPS_PORT_STAT_##name##_L)))
+#define GET_STAT_COM(name) t4_read_reg64(adap, MPS_STAT_##name##_L)
+
+	p->tx_octets           = GET_STAT(TX_PORT_BYTES);
+	p->tx_frames           = GET_STAT(TX_PORT_FRAMES);
+	p->tx_bcast_frames     = GET_STAT(TX_PORT_BCAST);
+	p->tx_mcast_frames     = GET_STAT(TX_PORT_MCAST);
+	p->tx_ucast_frames     = GET_STAT(TX_PORT_UCAST);
+	p->tx_error_frames     = GET_STAT(TX_PORT_ERROR);
+	p->tx_frames_64        = GET_STAT(TX_PORT_64B);
+	p->tx_frames_65_127    = GET_STAT(TX_PORT_65B_127B);
+	p->tx_frames_128_255   = GET_STAT(TX_PORT_128B_255B);
+	p->tx_frames_256_511   = GET_STAT(TX_PORT_256B_511B);
+	p->tx_frames_512_1023  = GET_STAT(TX_PORT_512B_1023B);
+	p->tx_frames_1024_1518 = GET_STAT(TX_PORT_1024B_1518B);
+	p->tx_frames_1519_max  = GET_STAT(TX_PORT_1519B_MAX);
+	p->tx_drop             = GET_STAT(TX_PORT_DROP);
+	p->tx_pause            = GET_STAT(TX_PORT_PAUSE);
+	p->tx_ppp0             = GET_STAT(TX_PORT_PPP0);
+	p->tx_ppp1             = GET_STAT(TX_PORT_PPP1);
+	p->tx_ppp2             = GET_STAT(TX_PORT_PPP2);
+	p->tx_ppp3             = GET_STAT(TX_PORT_PPP3);
+	p->tx_ppp4             = GET_STAT(TX_PORT_PPP4);
+	p->tx_ppp5             = GET_STAT(TX_PORT_PPP5);
+	p->tx_ppp6             = GET_STAT(TX_PORT_PPP6);
+	p->tx_ppp7             = GET_STAT(TX_PORT_PPP7);
+
+	if (CHELSIO_CHIP_VERSION(adap->params.chip) >= CHELSIO_T5) {
+		if (stat_ctl & COUNTPAUSESTATTX_F)
+			p->tx_frames_64 -= p->tx_pause;
+		if (stat_ctl & COUNTPAUSEMCTX_F)
+			p->tx_mcast_frames -= p->tx_pause;
+	}
+	p->rx_octets           = GET_STAT(RX_PORT_BYTES);
+	p->rx_frames           = GET_STAT(RX_PORT_FRAMES);
+	p->rx_bcast_frames     = GET_STAT(RX_PORT_BCAST);
+	p->rx_mcast_frames     = GET_STAT(RX_PORT_MCAST);
+	p->rx_ucast_frames     = GET_STAT(RX_PORT_UCAST);
+	p->rx_too_long         = GET_STAT(RX_PORT_MTU_ERROR);
+	p->rx_jabber           = GET_STAT(RX_PORT_MTU_CRC_ERROR);
+	p->rx_fcs_err          = GET_STAT(RX_PORT_CRC_ERROR);
+	p->rx_len_err          = GET_STAT(RX_PORT_LEN_ERROR);
+	p->rx_symbol_err       = GET_STAT(RX_PORT_SYM_ERROR);
+	p->rx_runt             = GET_STAT(RX_PORT_LESS_64B);
+	p->rx_frames_64        = GET_STAT(RX_PORT_64B);
+	p->rx_frames_65_127    = GET_STAT(RX_PORT_65B_127B);
+	p->rx_frames_128_255   = GET_STAT(RX_PORT_128B_255B);
+	p->rx_frames_256_511   = GET_STAT(RX_PORT_256B_511B);
+	p->rx_frames_512_1023  = GET_STAT(RX_PORT_512B_1023B);
+	p->rx_frames_1024_1518 = GET_STAT(RX_PORT_1024B_1518B);
+	p->rx_frames_1519_max  = GET_STAT(RX_PORT_1519B_MAX);
+	p->rx_pause            = GET_STAT(RX_PORT_PAUSE);
+	p->rx_ppp0             = GET_STAT(RX_PORT_PPP0);
+	p->rx_ppp1             = GET_STAT(RX_PORT_PPP1);
+	p->rx_ppp2             = GET_STAT(RX_PORT_PPP2);
+	p->rx_ppp3             = GET_STAT(RX_PORT_PPP3);
+	p->rx_ppp4             = GET_STAT(RX_PORT_PPP4);
+	p->rx_ppp5             = GET_STAT(RX_PORT_PPP5);
+	p->rx_ppp6             = GET_STAT(RX_PORT_PPP6);
+	p->rx_ppp7             = GET_STAT(RX_PORT_PPP7);
+
+	if (CHELSIO_CHIP_VERSION(adap->params.chip) >= CHELSIO_T5) {
+		if (stat_ctl & COUNTPAUSESTATRX_F)
+			p->rx_frames_64 -= p->rx_pause;
+		if (stat_ctl & COUNTPAUSEMCRX_F)
+			p->rx_mcast_frames -= p->rx_pause;
+	}
+
+	p->rx_ovflow0 = (bgmap & 1) ? GET_STAT_COM(RX_BG_0_MAC_DROP_FRAME) : 0;
+	p->rx_ovflow1 = (bgmap & 2) ? GET_STAT_COM(RX_BG_1_MAC_DROP_FRAME) : 0;
+	p->rx_ovflow2 = (bgmap & 4) ? GET_STAT_COM(RX_BG_2_MAC_DROP_FRAME) : 0;
+	p->rx_ovflow3 = (bgmap & 8) ? GET_STAT_COM(RX_BG_3_MAC_DROP_FRAME) : 0;
+	p->rx_trunc0 = (bgmap & 1) ? GET_STAT_COM(RX_BG_0_MAC_TRUNC_FRAME) : 0;
+	p->rx_trunc1 = (bgmap & 2) ? GET_STAT_COM(RX_BG_1_MAC_TRUNC_FRAME) : 0;
+	p->rx_trunc2 = (bgmap & 4) ? GET_STAT_COM(RX_BG_2_MAC_TRUNC_FRAME) : 0;
+	p->rx_trunc3 = (bgmap & 8) ? GET_STAT_COM(RX_BG_3_MAC_TRUNC_FRAME) : 0;
+
+#undef GET_STAT
+#undef GET_STAT_COM
+}
+
+/**
+ *	t4_get_lb_stats - collect loopback port statistics
+ *	@adap: the adapter
+ *	@idx: the loopback port index
+ *	@p: the stats structure to fill
+ *
+ *	Return HW statistics for the given loopback port.
+ */
+void t4_get_lb_stats(struct adapter *adap, int idx, struct lb_port_stats *p)
+{
+	u32 bgmap = t4_get_mps_bg_map(adap, idx);
+
+#define GET_STAT(name) \
+	t4_read_reg64(adap, \
+	(is_t4(adap->params.chip) ? \
+	PORT_REG(idx, MPS_PORT_STAT_LB_PORT_##name##_L) : \
+	T5_PORT_REG(idx, MPS_PORT_STAT_LB_PORT_##name##_L)))
+#define GET_STAT_COM(name) t4_read_reg64(adap, MPS_STAT_##name##_L)
+
+	p->octets           = GET_STAT(BYTES);
+	p->frames           = GET_STAT(FRAMES);
+	p->bcast_frames     = GET_STAT(BCAST);
+	p->mcast_frames     = GET_STAT(MCAST);
+	p->ucast_frames     = GET_STAT(UCAST);
+	p->error_frames     = GET_STAT(ERROR);
+
+	p->frames_64        = GET_STAT(64B);
+	p->frames_65_127    = GET_STAT(65B_127B);
+	p->frames_128_255   = GET_STAT(128B_255B);
+	p->frames_256_511   = GET_STAT(256B_511B);
+	p->frames_512_1023  = GET_STAT(512B_1023B);
+	p->frames_1024_1518 = GET_STAT(1024B_1518B);
+	p->frames_1519_max  = GET_STAT(1519B_MAX);
+	p->drop             = GET_STAT(DROP_FRAMES);
+
+	p->ovflow0 = (bgmap & 1) ? GET_STAT_COM(RX_BG_0_LB_DROP_FRAME) : 0;
+	p->ovflow1 = (bgmap & 2) ? GET_STAT_COM(RX_BG_1_LB_DROP_FRAME) : 0;
+	p->ovflow2 = (bgmap & 4) ? GET_STAT_COM(RX_BG_2_LB_DROP_FRAME) : 0;
+	p->ovflow3 = (bgmap & 8) ? GET_STAT_COM(RX_BG_3_LB_DROP_FRAME) : 0;
+	p->trunc0 = (bgmap & 1) ? GET_STAT_COM(RX_BG_0_LB_TRUNC_FRAME) : 0;
+	p->trunc1 = (bgmap & 2) ? GET_STAT_COM(RX_BG_1_LB_TRUNC_FRAME) : 0;
+	p->trunc2 = (bgmap & 4) ? GET_STAT_COM(RX_BG_2_LB_TRUNC_FRAME) : 0;
+	p->trunc3 = (bgmap & 8) ? GET_STAT_COM(RX_BG_3_LB_TRUNC_FRAME) : 0;
+
+#undef GET_STAT
+#undef GET_STAT_COM
+}
+
+/*     t4_mk_filtdelwr - create a delete filter WR
+ *     @ftid: the filter ID
+ *     @wr: the filter work request to populate
+ *     @qid: ingress queue to receive the delete notification
+ *
+ *     Creates a filter work request to delete the supplied filter.  If @qid is
+ *     negative the delete notification is suppressed.
+ */
+void t4_mk_filtdelwr(unsigned int ftid, struct fw_filter_wr *wr, int qid)
+{
+	memset(wr, 0, sizeof(*wr));
+	wr->op_pkd = cpu_to_be32(FW_WR_OP_V(FW_FILTER_WR));
+	wr->len16_pkd = cpu_to_be32(FW_WR_LEN16_V(sizeof(*wr) / 16));
+	wr->tid_to_iq = cpu_to_be32(FW_FILTER_WR_TID_V(ftid) |
+				    FW_FILTER_WR_NOREPLY_V(qid < 0));
+	wr->del_filter_to_l2tix = cpu_to_be32(FW_FILTER_WR_DEL_FILTER_F);
+	if (qid >= 0)
+		wr->rx_chan_rx_rpl_iq =
+			cpu_to_be16(FW_FILTER_WR_RX_RPL_IQ_V(qid));
+}
+
+#define INIT_CMD(var, cmd, rd_wr) do { \
+	(var).op_to_write = cpu_to_be32(FW_CMD_OP_V(FW_##cmd##_CMD) | \
+					FW_CMD_REQUEST_F | \
+					FW_CMD_##rd_wr##_F); \
+	(var).retval_len16 = cpu_to_be32(FW_LEN16(var)); \
+} while (0)
+
+int t4_fwaddrspace_write(struct adapter *adap, unsigned int mbox,
+			  u32 addr, u32 val)
+{
+	u32 ldst_addrspace;
+	struct fw_ldst_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	ldst_addrspace = FW_LDST_CMD_ADDRSPACE_V(FW_LDST_ADDRSPC_FIRMWARE);
+	c.op_to_addrspace = cpu_to_be32(FW_CMD_OP_V(FW_LDST_CMD) |
+					FW_CMD_REQUEST_F |
+					FW_CMD_WRITE_F |
+					ldst_addrspace);
+	c.cycles_to_len16 = cpu_to_be32(FW_LEN16(c));
+	c.u.addrval.addr = cpu_to_be32(addr);
+	c.u.addrval.val = cpu_to_be32(val);
+
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_mdio_rd - read a PHY register through MDIO
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@phy_addr: the PHY address
+ *	@mmd: the PHY MMD to access (0 for clause 22 PHYs)
+ *	@reg: the register to read
+ *	@valp: where to store the value
+ *
+ *	Issues a FW command through the given mailbox to read a PHY register.
+ */
+int t4_mdio_rd(struct adapter *adap, unsigned int mbox, unsigned int phy_addr,
+	       unsigned int mmd, unsigned int reg, u16 *valp)
+{
+	int ret;
+	u32 ldst_addrspace;
+	struct fw_ldst_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	ldst_addrspace = FW_LDST_CMD_ADDRSPACE_V(FW_LDST_ADDRSPC_MDIO);
+	c.op_to_addrspace = cpu_to_be32(FW_CMD_OP_V(FW_LDST_CMD) |
+					FW_CMD_REQUEST_F | FW_CMD_READ_F |
+					ldst_addrspace);
+	c.cycles_to_len16 = cpu_to_be32(FW_LEN16(c));
+	c.u.mdio.paddr_mmd = cpu_to_be16(FW_LDST_CMD_PADDR_V(phy_addr) |
+					 FW_LDST_CMD_MMD_V(mmd));
+	c.u.mdio.raddr = cpu_to_be16(reg);
+
+	ret = t4_wr_mbox(adap, mbox, &c, sizeof(c), &c);
+	if (ret == 0)
+		*valp = be16_to_cpu(c.u.mdio.rval);
+	return ret;
+}
+
+/**
+ *	t4_mdio_wr - write a PHY register through MDIO
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@phy_addr: the PHY address
+ *	@mmd: the PHY MMD to access (0 for clause 22 PHYs)
+ *	@reg: the register to write
+ *	@val: value to write
+ *
+ *	Issues a FW command through the given mailbox to write a PHY register.
+ */
+int t4_mdio_wr(struct adapter *adap, unsigned int mbox, unsigned int phy_addr,
+	       unsigned int mmd, unsigned int reg, u16 val)
+{
+	u32 ldst_addrspace;
+	struct fw_ldst_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	ldst_addrspace = FW_LDST_CMD_ADDRSPACE_V(FW_LDST_ADDRSPC_MDIO);
+	c.op_to_addrspace = cpu_to_be32(FW_CMD_OP_V(FW_LDST_CMD) |
+					FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+					ldst_addrspace);
+	c.cycles_to_len16 = cpu_to_be32(FW_LEN16(c));
+	c.u.mdio.paddr_mmd = cpu_to_be16(FW_LDST_CMD_PADDR_V(phy_addr) |
+					 FW_LDST_CMD_MMD_V(mmd));
+	c.u.mdio.raddr = cpu_to_be16(reg);
+	c.u.mdio.rval = cpu_to_be16(val);
+
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_sge_decode_idma_state - decode the idma state
+ *	@adapter: the adapter
+ *	@state: the state idma is stuck in
+ */
+void t4_sge_decode_idma_state(struct adapter *adapter, int state)
+{
+	static const char * const t4_decode[] = {
+		"IDMA_IDLE",
+		"IDMA_PUSH_MORE_CPL_FIFO",
+		"IDMA_PUSH_CPL_MSG_HEADER_TO_FIFO",
+		"Not used",
+		"IDMA_PHYSADDR_SEND_PCIEHDR",
+		"IDMA_PHYSADDR_SEND_PAYLOAD_FIRST",
+		"IDMA_PHYSADDR_SEND_PAYLOAD",
+		"IDMA_SEND_FIFO_TO_IMSG",
+		"IDMA_FL_REQ_DATA_FL_PREP",
+		"IDMA_FL_REQ_DATA_FL",
+		"IDMA_FL_DROP",
+		"IDMA_FL_H_REQ_HEADER_FL",
+		"IDMA_FL_H_SEND_PCIEHDR",
+		"IDMA_FL_H_PUSH_CPL_FIFO",
+		"IDMA_FL_H_SEND_CPL",
+		"IDMA_FL_H_SEND_IP_HDR_FIRST",
+		"IDMA_FL_H_SEND_IP_HDR",
+		"IDMA_FL_H_REQ_NEXT_HEADER_FL",
+		"IDMA_FL_H_SEND_NEXT_PCIEHDR",
+		"IDMA_FL_H_SEND_IP_HDR_PADDING",
+		"IDMA_FL_D_SEND_PCIEHDR",
+		"IDMA_FL_D_SEND_CPL_AND_IP_HDR",
+		"IDMA_FL_D_REQ_NEXT_DATA_FL",
+		"IDMA_FL_SEND_PCIEHDR",
+		"IDMA_FL_PUSH_CPL_FIFO",
+		"IDMA_FL_SEND_CPL",
+		"IDMA_FL_SEND_PAYLOAD_FIRST",
+		"IDMA_FL_SEND_PAYLOAD",
+		"IDMA_FL_REQ_NEXT_DATA_FL",
+		"IDMA_FL_SEND_NEXT_PCIEHDR",
+		"IDMA_FL_SEND_PADDING",
+		"IDMA_FL_SEND_COMPLETION_TO_IMSG",
+		"IDMA_FL_SEND_FIFO_TO_IMSG",
+		"IDMA_FL_REQ_DATAFL_DONE",
+		"IDMA_FL_REQ_HEADERFL_DONE",
+	};
+	static const char * const t5_decode[] = {
+		"IDMA_IDLE",
+		"IDMA_ALMOST_IDLE",
+		"IDMA_PUSH_MORE_CPL_FIFO",
+		"IDMA_PUSH_CPL_MSG_HEADER_TO_FIFO",
+		"IDMA_SGEFLRFLUSH_SEND_PCIEHDR",
+		"IDMA_PHYSADDR_SEND_PCIEHDR",
+		"IDMA_PHYSADDR_SEND_PAYLOAD_FIRST",
+		"IDMA_PHYSADDR_SEND_PAYLOAD",
+		"IDMA_SEND_FIFO_TO_IMSG",
+		"IDMA_FL_REQ_DATA_FL",
+		"IDMA_FL_DROP",
+		"IDMA_FL_DROP_SEND_INC",
+		"IDMA_FL_H_REQ_HEADER_FL",
+		"IDMA_FL_H_SEND_PCIEHDR",
+		"IDMA_FL_H_PUSH_CPL_FIFO",
+		"IDMA_FL_H_SEND_CPL",
+		"IDMA_FL_H_SEND_IP_HDR_FIRST",
+		"IDMA_FL_H_SEND_IP_HDR",
+		"IDMA_FL_H_REQ_NEXT_HEADER_FL",
+		"IDMA_FL_H_SEND_NEXT_PCIEHDR",
+		"IDMA_FL_H_SEND_IP_HDR_PADDING",
+		"IDMA_FL_D_SEND_PCIEHDR",
+		"IDMA_FL_D_SEND_CPL_AND_IP_HDR",
+		"IDMA_FL_D_REQ_NEXT_DATA_FL",
+		"IDMA_FL_SEND_PCIEHDR",
+		"IDMA_FL_PUSH_CPL_FIFO",
+		"IDMA_FL_SEND_CPL",
+		"IDMA_FL_SEND_PAYLOAD_FIRST",
+		"IDMA_FL_SEND_PAYLOAD",
+		"IDMA_FL_REQ_NEXT_DATA_FL",
+		"IDMA_FL_SEND_NEXT_PCIEHDR",
+		"IDMA_FL_SEND_PADDING",
+		"IDMA_FL_SEND_COMPLETION_TO_IMSG",
+	};
+	static const char * const t6_decode[] = {
+		"IDMA_IDLE",
+		"IDMA_PUSH_MORE_CPL_FIFO",
+		"IDMA_PUSH_CPL_MSG_HEADER_TO_FIFO",
+		"IDMA_SGEFLRFLUSH_SEND_PCIEHDR",
+		"IDMA_PHYSADDR_SEND_PCIEHDR",
+		"IDMA_PHYSADDR_SEND_PAYLOAD_FIRST",
+		"IDMA_PHYSADDR_SEND_PAYLOAD",
+		"IDMA_FL_REQ_DATA_FL",
+		"IDMA_FL_DROP",
+		"IDMA_FL_DROP_SEND_INC",
+		"IDMA_FL_H_REQ_HEADER_FL",
+		"IDMA_FL_H_SEND_PCIEHDR",
+		"IDMA_FL_H_PUSH_CPL_FIFO",
+		"IDMA_FL_H_SEND_CPL",
+		"IDMA_FL_H_SEND_IP_HDR_FIRST",
+		"IDMA_FL_H_SEND_IP_HDR",
+		"IDMA_FL_H_REQ_NEXT_HEADER_FL",
+		"IDMA_FL_H_SEND_NEXT_PCIEHDR",
+		"IDMA_FL_H_SEND_IP_HDR_PADDING",
+		"IDMA_FL_D_SEND_PCIEHDR",
+		"IDMA_FL_D_SEND_CPL_AND_IP_HDR",
+		"IDMA_FL_D_REQ_NEXT_DATA_FL",
+		"IDMA_FL_SEND_PCIEHDR",
+		"IDMA_FL_PUSH_CPL_FIFO",
+		"IDMA_FL_SEND_CPL",
+		"IDMA_FL_SEND_PAYLOAD_FIRST",
+		"IDMA_FL_SEND_PAYLOAD",
+		"IDMA_FL_REQ_NEXT_DATA_FL",
+		"IDMA_FL_SEND_NEXT_PCIEHDR",
+		"IDMA_FL_SEND_PADDING",
+		"IDMA_FL_SEND_COMPLETION_TO_IMSG",
+	};
+	static const u32 sge_regs[] = {
+		SGE_DEBUG_DATA_LOW_INDEX_2_A,
+		SGE_DEBUG_DATA_LOW_INDEX_3_A,
+		SGE_DEBUG_DATA_HIGH_INDEX_10_A,
+	};
+	const char **sge_idma_decode;
+	int sge_idma_decode_nstates;
+	int i;
+	unsigned int chip_version = CHELSIO_CHIP_VERSION(adapter->params.chip);
+
+	/* Select the right set of decode strings to dump depending on the
+	 * adapter chip type.
+	 */
+	switch (chip_version) {
+	case CHELSIO_T4:
+		sge_idma_decode = (const char **)t4_decode;
+		sge_idma_decode_nstates = ARRAY_SIZE(t4_decode);
+		break;
+
+	case CHELSIO_T5:
+		sge_idma_decode = (const char **)t5_decode;
+		sge_idma_decode_nstates = ARRAY_SIZE(t5_decode);
+		break;
+
+	case CHELSIO_T6:
+		sge_idma_decode = (const char **)t6_decode;
+		sge_idma_decode_nstates = ARRAY_SIZE(t6_decode);
+		break;
+
+	default:
+		dev_err(adapter->pdev_dev,
+			"Unsupported chip version %d\n", chip_version);
+		return;
+	}
+
+	if (is_t4(adapter->params.chip)) {
+		sge_idma_decode = (const char **)t4_decode;
+		sge_idma_decode_nstates = ARRAY_SIZE(t4_decode);
+	} else {
+		sge_idma_decode = (const char **)t5_decode;
+		sge_idma_decode_nstates = ARRAY_SIZE(t5_decode);
+	}
+
+	if (state < sge_idma_decode_nstates)
+		CH_WARN(adapter, "idma state %s\n", sge_idma_decode[state]);
+	else
+		CH_WARN(adapter, "idma state %d unknown\n", state);
+
+	for (i = 0; i < ARRAY_SIZE(sge_regs); i++)
+		CH_WARN(adapter, "SGE register %#x value %#x\n",
+			sge_regs[i], t4_read_reg(adapter, sge_regs[i]));
+}
+
+/**
+ *      t4_sge_ctxt_flush - flush the SGE context cache
+ *      @adap: the adapter
+ *      @mbox: mailbox to use for the FW command
+ *      @ctxt_type: Egress or Ingress
+ *
+ *      Issues a FW command through the given mailbox to flush the
+ *      SGE context cache.
+ */
+int t4_sge_ctxt_flush(struct adapter *adap, unsigned int mbox, int ctxt_type)
+{
+	int ret;
+	u32 ldst_addrspace;
+	struct fw_ldst_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	ldst_addrspace = FW_LDST_CMD_ADDRSPACE_V(ctxt_type == CTXT_EGRESS ?
+						 FW_LDST_ADDRSPC_SGE_EGRC :
+						 FW_LDST_ADDRSPC_SGE_INGC);
+	c.op_to_addrspace = cpu_to_be32(FW_CMD_OP_V(FW_LDST_CMD) |
+					FW_CMD_REQUEST_F | FW_CMD_READ_F |
+					ldst_addrspace);
+	c.cycles_to_len16 = cpu_to_be32(FW_LEN16(c));
+	c.u.idctxt.msg_ctxtflush = cpu_to_be32(FW_LDST_CMD_CTXTFLUSH_F);
+
+	ret = t4_wr_mbox(adap, mbox, &c, sizeof(c), &c);
+	return ret;
+}
+
+/**
+ *	t4_read_sge_dbqtimers - read SGE Doorbell Queue Timer values
+ *	@adap: the adapter
+ *	@ndbqtimers: size of the provided SGE Doorbell Queue Timer table
+ *	@dbqtimers: SGE Doorbell Queue Timer table
+ *
+ *	Reads the SGE Doorbell Queue Timer values into the provided table.
+ *	Returns 0 on success (Firmware and Hardware support this feature),
+ *	an error on failure.
+ */
+int t4_read_sge_dbqtimers(struct adapter *adap, unsigned int ndbqtimers,
+			  u16 *dbqtimers)
+{
+	int ret, dbqtimerix;
+
+	ret = 0;
+	dbqtimerix = 0;
+	while (dbqtimerix < ndbqtimers) {
+		int nparams, param;
+		u32 params[7], vals[7];
+
+		nparams = ndbqtimers - dbqtimerix;
+		if (nparams > ARRAY_SIZE(params))
+			nparams = ARRAY_SIZE(params);
+
+		for (param = 0; param < nparams; param++)
+			params[param] =
+			  (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+			   FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_DBQ_TIMER) |
+			   FW_PARAMS_PARAM_Y_V(dbqtimerix + param));
+		ret = t4_query_params(adap, adap->mbox, adap->pf, 0,
+				      nparams, params, vals);
+		if (ret)
+			break;
+
+		for (param = 0; param < nparams; param++)
+			dbqtimers[dbqtimerix++] = vals[param];
+	}
+	return ret;
+}
+
+/**
+ *      t4_fw_hello - establish communication with FW
+ *      @adap: the adapter
+ *      @mbox: mailbox to use for the FW command
+ *      @evt_mbox: mailbox to receive async FW events
+ *      @master: specifies the caller's willingness to be the device master
+ *	@state: returns the current device state (if non-NULL)
+ *
+ *	Issues a command to establish communication with FW.  Returns either
+ *	an error (negative integer) or the mailbox of the Master PF.
+ */
+int t4_fw_hello(struct adapter *adap, unsigned int mbox, unsigned int evt_mbox,
+		enum dev_master master, enum dev_state *state)
+{
+	int ret;
+	struct fw_hello_cmd c;
+	u32 v;
+	unsigned int master_mbox;
+	int retries = FW_CMD_HELLO_RETRIES;
+
+retry:
+	memset(&c, 0, sizeof(c));
+	INIT_CMD(c, HELLO, WRITE);
+	c.err_to_clearinit = cpu_to_be32(
+		FW_HELLO_CMD_MASTERDIS_V(master == MASTER_CANT) |
+		FW_HELLO_CMD_MASTERFORCE_V(master == MASTER_MUST) |
+		FW_HELLO_CMD_MBMASTER_V(master == MASTER_MUST ?
+					mbox : FW_HELLO_CMD_MBMASTER_M) |
+		FW_HELLO_CMD_MBASYNCNOT_V(evt_mbox) |
+		FW_HELLO_CMD_STAGE_V(fw_hello_cmd_stage_os) |
+		FW_HELLO_CMD_CLEARINIT_F);
+
+	/*
+	 * Issue the HELLO command to the firmware.  If it's not successful
+	 * but indicates that we got a "busy" or "timeout" condition, retry
+	 * the HELLO until we exhaust our retry limit.  If we do exceed our
+	 * retry limit, check to see if the firmware left us any error
+	 * information and report that if so.
+	 */
+	ret = t4_wr_mbox(adap, mbox, &c, sizeof(c), &c);
+	if (ret < 0) {
+		if ((ret == -EBUSY || ret == -ETIMEDOUT) && retries-- > 0)
+			goto retry;
+		if (t4_read_reg(adap, PCIE_FW_A) & PCIE_FW_ERR_F)
+			t4_report_fw_error(adap);
+		return ret;
+	}
+
+	v = be32_to_cpu(c.err_to_clearinit);
+	master_mbox = FW_HELLO_CMD_MBMASTER_G(v);
+	if (state) {
+		if (v & FW_HELLO_CMD_ERR_F)
+			*state = DEV_STATE_ERR;
+		else if (v & FW_HELLO_CMD_INIT_F)
+			*state = DEV_STATE_INIT;
+		else
+			*state = DEV_STATE_UNINIT;
+	}
+
+	/*
+	 * If we're not the Master PF then we need to wait around for the
+	 * Master PF Driver to finish setting up the adapter.
+	 *
+	 * Note that we also do this wait if we're a non-Master-capable PF and
+	 * there is no current Master PF; a Master PF may show up momentarily
+	 * and we wouldn't want to fail pointlessly.  (This can happen when an
+	 * OS loads lots of different drivers rapidly at the same time).  In
+	 * this case, the Master PF returned by the firmware will be
+	 * PCIE_FW_MASTER_M so the test below will work ...
+	 */
+	if ((v & (FW_HELLO_CMD_ERR_F|FW_HELLO_CMD_INIT_F)) == 0 &&
+	    master_mbox != mbox) {
+		int waiting = FW_CMD_HELLO_TIMEOUT;
+
+		/*
+		 * Wait for the firmware to either indicate an error or
+		 * initialized state.  If we see either of these we bail out
+		 * and report the issue to the caller.  If we exhaust the
+		 * "hello timeout" and we haven't exhausted our retries, try
+		 * again.  Otherwise bail with a timeout error.
+		 */
+		for (;;) {
+			u32 pcie_fw;
+
+			msleep(50);
+			waiting -= 50;
+
+			/*
+			 * If neither Error nor Initialized are indicated
+			 * by the firmware keep waiting till we exhaust our
+			 * timeout ... and then retry if we haven't exhausted
+			 * our retries ...
+			 */
+			pcie_fw = t4_read_reg(adap, PCIE_FW_A);
+			if (!(pcie_fw & (PCIE_FW_ERR_F|PCIE_FW_INIT_F))) {
+				if (waiting <= 0) {
+					if (retries-- > 0)
+						goto retry;
+
+					return -ETIMEDOUT;
+				}
+				continue;
+			}
+
+			/*
+			 * We either have an Error or Initialized condition
+			 * report errors preferentially.
+			 */
+			if (state) {
+				if (pcie_fw & PCIE_FW_ERR_F)
+					*state = DEV_STATE_ERR;
+				else if (pcie_fw & PCIE_FW_INIT_F)
+					*state = DEV_STATE_INIT;
+			}
+
+			/*
+			 * If we arrived before a Master PF was selected and
+			 * there's not a valid Master PF, grab its identity
+			 * for our caller.
+			 */
+			if (master_mbox == PCIE_FW_MASTER_M &&
+			    (pcie_fw & PCIE_FW_MASTER_VLD_F))
+				master_mbox = PCIE_FW_MASTER_G(pcie_fw);
+			break;
+		}
+	}
+
+	return master_mbox;
+}
+
+/**
+ *	t4_fw_bye - end communication with FW
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *
+ *	Issues a command to terminate communication with FW.
+ */
+int t4_fw_bye(struct adapter *adap, unsigned int mbox)
+{
+	struct fw_bye_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	INIT_CMD(c, BYE, WRITE);
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_early_init - ask FW to initialize the device
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *
+ *	Issues a command to FW to partially initialize the device.  This
+ *	performs initialization that generally doesn't depend on user input.
+ */
+int t4_early_init(struct adapter *adap, unsigned int mbox)
+{
+	struct fw_initialize_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	INIT_CMD(c, INITIALIZE, WRITE);
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_fw_reset - issue a reset to FW
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@reset: specifies the type of reset to perform
+ *
+ *	Issues a reset command of the specified type to FW.
+ */
+int t4_fw_reset(struct adapter *adap, unsigned int mbox, int reset)
+{
+	struct fw_reset_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	INIT_CMD(c, RESET, WRITE);
+	c.val = cpu_to_be32(reset);
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_fw_halt - issue a reset/halt to FW and put uP into RESET
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW RESET command (if desired)
+ *	@force: force uP into RESET even if FW RESET command fails
+ *
+ *	Issues a RESET command to firmware (if desired) with a HALT indication
+ *	and then puts the microprocessor into RESET state.  The RESET command
+ *	will only be issued if a legitimate mailbox is provided (mbox <=
+ *	PCIE_FW_MASTER_M).
+ *
+ *	This is generally used in order for the host to safely manipulate the
+ *	adapter without fear of conflicting with whatever the firmware might
+ *	be doing.  The only way out of this state is to RESTART the firmware
+ *	...
+ */
+static int t4_fw_halt(struct adapter *adap, unsigned int mbox, int force)
+{
+	int ret = 0;
+
+	/*
+	 * If a legitimate mailbox is provided, issue a RESET command
+	 * with a HALT indication.
+	 */
+	if (mbox <= PCIE_FW_MASTER_M) {
+		struct fw_reset_cmd c;
+
+		memset(&c, 0, sizeof(c));
+		INIT_CMD(c, RESET, WRITE);
+		c.val = cpu_to_be32(PIORST_F | PIORSTMODE_F);
+		c.halt_pkd = cpu_to_be32(FW_RESET_CMD_HALT_F);
+		ret = t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+	}
+
+	/*
+	 * Normally we won't complete the operation if the firmware RESET
+	 * command fails but if our caller insists we'll go ahead and put the
+	 * uP into RESET.  This can be useful if the firmware is hung or even
+	 * missing ...  We'll have to take the risk of putting the uP into
+	 * RESET without the cooperation of firmware in that case.
+	 *
+	 * We also force the firmware's HALT flag to be on in case we bypassed
+	 * the firmware RESET command above or we're dealing with old firmware
+	 * which doesn't have the HALT capability.  This will serve as a flag
+	 * for the incoming firmware to know that it's coming out of a HALT
+	 * rather than a RESET ... if it's new enough to understand that ...
+	 */
+	if (ret == 0 || force) {
+		t4_set_reg_field(adap, CIM_BOOT_CFG_A, UPCRST_F, UPCRST_F);
+		t4_set_reg_field(adap, PCIE_FW_A, PCIE_FW_HALT_F,
+				 PCIE_FW_HALT_F);
+	}
+
+	/*
+	 * And we always return the result of the firmware RESET command
+	 * even when we force the uP into RESET ...
+	 */
+	return ret;
+}
+
+/**
+ *	t4_fw_restart - restart the firmware by taking the uP out of RESET
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@reset: if we want to do a RESET to restart things
+ *
+ *	Restart firmware previously halted by t4_fw_halt().  On successful
+ *	return the previous PF Master remains as the new PF Master and there
+ *	is no need to issue a new HELLO command, etc.
+ *
+ *	We do this in two ways:
+ *
+ *	 1. If we're dealing with newer firmware we'll simply want to take
+ *	    the chip's microprocessor out of RESET.  This will cause the
+ *	    firmware to start up from its start vector.  And then we'll loop
+ *	    until the firmware indicates it's started again (PCIE_FW.HALT
+ *	    reset to 0) or we timeout.
+ *
+ *	 2. If we're dealing with older firmware then we'll need to RESET
+ *	    the chip since older firmware won't recognize the PCIE_FW.HALT
+ *	    flag and automatically RESET itself on startup.
+ */
+static int t4_fw_restart(struct adapter *adap, unsigned int mbox, int reset)
+{
+	if (reset) {
+		/*
+		 * Since we're directing the RESET instead of the firmware
+		 * doing it automatically, we need to clear the PCIE_FW.HALT
+		 * bit.
+		 */
+		t4_set_reg_field(adap, PCIE_FW_A, PCIE_FW_HALT_F, 0);
+
+		/*
+		 * If we've been given a valid mailbox, first try to get the
+		 * firmware to do the RESET.  If that works, great and we can
+		 * return success.  Otherwise, if we haven't been given a
+		 * valid mailbox or the RESET command failed, fall back to
+		 * hitting the chip with a hammer.
+		 */
+		if (mbox <= PCIE_FW_MASTER_M) {
+			t4_set_reg_field(adap, CIM_BOOT_CFG_A, UPCRST_F, 0);
+			msleep(100);
+			if (t4_fw_reset(adap, mbox,
+					PIORST_F | PIORSTMODE_F) == 0)
+				return 0;
+		}
+
+		t4_write_reg(adap, PL_RST_A, PIORST_F | PIORSTMODE_F);
+		msleep(2000);
+	} else {
+		int ms;
+
+		t4_set_reg_field(adap, CIM_BOOT_CFG_A, UPCRST_F, 0);
+		for (ms = 0; ms < FW_CMD_MAX_TIMEOUT; ) {
+			if (!(t4_read_reg(adap, PCIE_FW_A) & PCIE_FW_HALT_F))
+				return 0;
+			msleep(100);
+			ms += 100;
+		}
+		return -ETIMEDOUT;
+	}
+	return 0;
+}
+
+/**
+ *	t4_fw_upgrade - perform all of the steps necessary to upgrade FW
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW RESET command (if desired)
+ *	@fw_data: the firmware image to write
+ *	@size: image size
+ *	@force: force upgrade even if firmware doesn't cooperate
+ *
+ *	Perform all of the steps necessary for upgrading an adapter's
+ *	firmware image.  Normally this requires the cooperation of the
+ *	existing firmware in order to halt all existing activities
+ *	but if an invalid mailbox token is passed in we skip that step
+ *	(though we'll still put the adapter microprocessor into RESET in
+ *	that case).
+ *
+ *	On successful return the new firmware will have been loaded and
+ *	the adapter will have been fully RESET losing all previous setup
+ *	state.  On unsuccessful return the adapter may be completely hosed ...
+ *	positive errno indicates that the adapter is ~probably~ intact, a
+ *	negative errno indicates that things are looking bad ...
+ */
+int t4_fw_upgrade(struct adapter *adap, unsigned int mbox,
+		  const u8 *fw_data, unsigned int size, int force)
+{
+	const struct fw_hdr *fw_hdr = (const struct fw_hdr *)fw_data;
+	int reset, ret;
+
+	if (!t4_fw_matches_chip(adap, fw_hdr))
+		return -EINVAL;
+
+	/* Disable CXGB4_FW_OK flag so that mbox commands with CXGB4_FW_OK flag
+	 * set wont be sent when we are flashing FW.
+	 */
+	adap->flags &= ~CXGB4_FW_OK;
+
+	ret = t4_fw_halt(adap, mbox, force);
+	if (ret < 0 && !force)
+		goto out;
+
+	ret = t4_load_fw(adap, fw_data, size);
+	if (ret < 0)
+		goto out;
+
+	/*
+	 * If there was a Firmware Configuration File stored in FLASH,
+	 * there's a good chance that it won't be compatible with the new
+	 * Firmware.  In order to prevent difficult to diagnose adapter
+	 * initialization issues, we clear out the Firmware Configuration File
+	 * portion of the FLASH .  The user will need to re-FLASH a new
+	 * Firmware Configuration File which is compatible with the new
+	 * Firmware if that's desired.
+	 */
+	(void)t4_load_cfg(adap, NULL, 0);
+
+	/*
+	 * Older versions of the firmware don't understand the new
+	 * PCIE_FW.HALT flag and so won't know to perform a RESET when they
+	 * restart.  So for newly loaded older firmware we'll have to do the
+	 * RESET for it so it starts up on a clean slate.  We can tell if
+	 * the newly loaded firmware will handle this right by checking
+	 * its header flags to see if it advertises the capability.
+	 */
+	reset = ((be32_to_cpu(fw_hdr->flags) & FW_HDR_FLAGS_RESET_HALT) == 0);
+	ret = t4_fw_restart(adap, mbox, reset);
+
+	/* Grab potentially new Firmware Device Log parameters so we can see
+	 * how healthy the new Firmware is.  It's okay to contact the new
+	 * Firmware for these parameters even though, as far as it's
+	 * concerned, we've never said "HELLO" to it ...
+	 */
+	(void)t4_init_devlog_params(adap);
+out:
+	adap->flags |= CXGB4_FW_OK;
+	return ret;
+}
+
+/**
+ *	t4_fl_pkt_align - return the fl packet alignment
+ *	@adap: the adapter
+ *
+ *	T4 has a single field to specify the packing and padding boundary.
+ *	T5 onwards has separate fields for this and hence the alignment for
+ *	next packet offset is maximum of these two.
+ *
+ */
+int t4_fl_pkt_align(struct adapter *adap)
+{
+	u32 sge_control, sge_control2;
+	unsigned int ingpadboundary, ingpackboundary, fl_align, ingpad_shift;
+
+	sge_control = t4_read_reg(adap, SGE_CONTROL_A);
+
+	/* T4 uses a single control field to specify both the PCIe Padding and
+	 * Packing Boundary.  T5 introduced the ability to specify these
+	 * separately.  The actual Ingress Packet Data alignment boundary
+	 * within Packed Buffer Mode is the maximum of these two
+	 * specifications.  (Note that it makes no real practical sense to
+	 * have the Padding Boundary be larger than the Packing Boundary but you
+	 * could set the chip up that way and, in fact, legacy T4 code would
+	 * end doing this because it would initialize the Padding Boundary and
+	 * leave the Packing Boundary initialized to 0 (16 bytes).)
+	 * Padding Boundary values in T6 starts from 8B,
+	 * where as it is 32B for T4 and T5.
+	 */
+	if (CHELSIO_CHIP_VERSION(adap->params.chip) <= CHELSIO_T5)
+		ingpad_shift = INGPADBOUNDARY_SHIFT_X;
+	else
+		ingpad_shift = T6_INGPADBOUNDARY_SHIFT_X;
+
+	ingpadboundary = 1 << (INGPADBOUNDARY_G(sge_control) + ingpad_shift);
+
+	fl_align = ingpadboundary;
+	if (!is_t4(adap->params.chip)) {
+		/* T5 has a weird interpretation of one of the PCIe Packing
+		 * Boundary values.  No idea why ...
+		 */
+		sge_control2 = t4_read_reg(adap, SGE_CONTROL2_A);
+		ingpackboundary = INGPACKBOUNDARY_G(sge_control2);
+		if (ingpackboundary == INGPACKBOUNDARY_16B_X)
+			ingpackboundary = 16;
+		else
+			ingpackboundary = 1 << (ingpackboundary +
+						INGPACKBOUNDARY_SHIFT_X);
+
+		fl_align = max(ingpadboundary, ingpackboundary);
+	}
+	return fl_align;
+}
+
+/**
+ *	t4_fixup_host_params - fix up host-dependent parameters
+ *	@adap: the adapter
+ *	@page_size: the host's Base Page Size
+ *	@cache_line_size: the host's Cache Line Size
+ *
+ *	Various registers in T4 contain values which are dependent on the
+ *	host's Base Page and Cache Line Sizes.  This function will fix all of
+ *	those registers with the appropriate values as passed in ...
+ */
+int t4_fixup_host_params(struct adapter *adap, unsigned int page_size,
+			 unsigned int cache_line_size)
+{
+	unsigned int page_shift = fls(page_size) - 1;
+	unsigned int sge_hps = page_shift - 10;
+	unsigned int stat_len = cache_line_size > 64 ? 128 : 64;
+	unsigned int fl_align = cache_line_size < 32 ? 32 : cache_line_size;
+	unsigned int fl_align_log = fls(fl_align) - 1;
+
+	t4_write_reg(adap, SGE_HOST_PAGE_SIZE_A,
+		     HOSTPAGESIZEPF0_V(sge_hps) |
+		     HOSTPAGESIZEPF1_V(sge_hps) |
+		     HOSTPAGESIZEPF2_V(sge_hps) |
+		     HOSTPAGESIZEPF3_V(sge_hps) |
+		     HOSTPAGESIZEPF4_V(sge_hps) |
+		     HOSTPAGESIZEPF5_V(sge_hps) |
+		     HOSTPAGESIZEPF6_V(sge_hps) |
+		     HOSTPAGESIZEPF7_V(sge_hps));
+
+	if (is_t4(adap->params.chip)) {
+		t4_set_reg_field(adap, SGE_CONTROL_A,
+				 INGPADBOUNDARY_V(INGPADBOUNDARY_M) |
+				 EGRSTATUSPAGESIZE_F,
+				 INGPADBOUNDARY_V(fl_align_log -
+						  INGPADBOUNDARY_SHIFT_X) |
+				 EGRSTATUSPAGESIZE_V(stat_len != 64));
+	} else {
+		unsigned int pack_align;
+		unsigned int ingpad, ingpack;
+
+		/* T5 introduced the separation of the Free List Padding and
+		 * Packing Boundaries.  Thus, we can select a smaller Padding
+		 * Boundary to avoid uselessly chewing up PCIe Link and Memory
+		 * Bandwidth, and use a Packing Boundary which is large enough
+		 * to avoid false sharing between CPUs, etc.
+		 *
+		 * For the PCI Link, the smaller the Padding Boundary the
+		 * better.  For the Memory Controller, a smaller Padding
+		 * Boundary is better until we cross under the Memory Line
+		 * Size (the minimum unit of transfer to/from Memory).  If we
+		 * have a Padding Boundary which is smaller than the Memory
+		 * Line Size, that'll involve a Read-Modify-Write cycle on the
+		 * Memory Controller which is never good.
+		 */
+
+		/* We want the Packing Boundary to be based on the Cache Line
+		 * Size in order to help avoid False Sharing performance
+		 * issues between CPUs, etc.  We also want the Packing
+		 * Boundary to incorporate the PCI-E Maximum Payload Size.  We
+		 * get best performance when the Packing Boundary is a
+		 * multiple of the Maximum Payload Size.
+		 */
+		pack_align = fl_align;
+		if (pci_is_pcie(adap->pdev)) {
+			unsigned int mps, mps_log;
+			u16 devctl;
+
+			/* The PCIe Device Control Maximum Payload Size field
+			 * [bits 7:5] encodes sizes as powers of 2 starting at
+			 * 128 bytes.
+			 */
+			pcie_capability_read_word(adap->pdev, PCI_EXP_DEVCTL,
+						  &devctl);
+			mps_log = ((devctl & PCI_EXP_DEVCTL_PAYLOAD) >> 5) + 7;
+			mps = 1 << mps_log;
+			if (mps > pack_align)
+				pack_align = mps;
+		}
+
+		/* N.B. T5/T6 have a crazy special interpretation of the "0"
+		 * value for the Packing Boundary.  This corresponds to 16
+		 * bytes instead of the expected 32 bytes.  So if we want 32
+		 * bytes, the best we can really do is 64 bytes ...
+		 */
+		if (pack_align <= 16) {
+			ingpack = INGPACKBOUNDARY_16B_X;
+			fl_align = 16;
+		} else if (pack_align == 32) {
+			ingpack = INGPACKBOUNDARY_64B_X;
+			fl_align = 64;
+		} else {
+			unsigned int pack_align_log = fls(pack_align) - 1;
+
+			ingpack = pack_align_log - INGPACKBOUNDARY_SHIFT_X;
+			fl_align = pack_align;
+		}
+
+		/* Use the smallest Ingress Padding which isn't smaller than
+		 * the Memory Controller Read/Write Size.  We'll take that as
+		 * being 8 bytes since we don't know of any system with a
+		 * wider Memory Controller Bus Width.
+		 */
+		if (is_t5(adap->params.chip))
+			ingpad = INGPADBOUNDARY_32B_X;
+		else
+			ingpad = T6_INGPADBOUNDARY_8B_X;
+
+		t4_set_reg_field(adap, SGE_CONTROL_A,
+				 INGPADBOUNDARY_V(INGPADBOUNDARY_M) |
+				 EGRSTATUSPAGESIZE_F,
+				 INGPADBOUNDARY_V(ingpad) |
+				 EGRSTATUSPAGESIZE_V(stat_len != 64));
+		t4_set_reg_field(adap, SGE_CONTROL2_A,
+				 INGPACKBOUNDARY_V(INGPACKBOUNDARY_M),
+				 INGPACKBOUNDARY_V(ingpack));
+	}
+	/*
+	 * Adjust various SGE Free List Host Buffer Sizes.
+	 *
+	 * This is something of a crock since we're using fixed indices into
+	 * the array which are also known by the sge.c code and the T4
+	 * Firmware Configuration File.  We need to come up with a much better
+	 * approach to managing this array.  For now, the first four entries
+	 * are:
+	 *
+	 *   0: Host Page Size
+	 *   1: 64KB
+	 *   2: Buffer size corresponding to 1500 byte MTU (unpacked mode)
+	 *   3: Buffer size corresponding to 9000 byte MTU (unpacked mode)
+	 *
+	 * For the single-MTU buffers in unpacked mode we need to include
+	 * space for the SGE Control Packet Shift, 14 byte Ethernet header,
+	 * possible 4 byte VLAN tag, all rounded up to the next Ingress Packet
+	 * Padding boundary.  All of these are accommodated in the Factory
+	 * Default Firmware Configuration File but we need to adjust it for
+	 * this host's cache line size.
+	 */
+	t4_write_reg(adap, SGE_FL_BUFFER_SIZE0_A, page_size);
+	t4_write_reg(adap, SGE_FL_BUFFER_SIZE2_A,
+		     (t4_read_reg(adap, SGE_FL_BUFFER_SIZE2_A) + fl_align-1)
+		     & ~(fl_align-1));
+	t4_write_reg(adap, SGE_FL_BUFFER_SIZE3_A,
+		     (t4_read_reg(adap, SGE_FL_BUFFER_SIZE3_A) + fl_align-1)
+		     & ~(fl_align-1));
+
+	t4_write_reg(adap, ULP_RX_TDDP_PSZ_A, HPZ0_V(page_shift - 12));
+
+	return 0;
+}
+
+/**
+ *	t4_fw_initialize - ask FW to initialize the device
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *
+ *	Issues a command to FW to partially initialize the device.  This
+ *	performs initialization that generally doesn't depend on user input.
+ */
+int t4_fw_initialize(struct adapter *adap, unsigned int mbox)
+{
+	struct fw_initialize_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	INIT_CMD(c, INITIALIZE, WRITE);
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_query_params_rw - query FW or device parameters
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@pf: the PF
+ *	@vf: the VF
+ *	@nparams: the number of parameters
+ *	@params: the parameter names
+ *	@val: the parameter values
+ *	@rw: Write and read flag
+ *	@sleep_ok: if true, we may sleep awaiting mbox cmd completion
+ *
+ *	Reads the value of FW or device parameters.  Up to 7 parameters can be
+ *	queried at once.
+ */
+int t4_query_params_rw(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		       unsigned int vf, unsigned int nparams, const u32 *params,
+		       u32 *val, int rw, bool sleep_ok)
+{
+	int i, ret;
+	struct fw_params_cmd c;
+	__be32 *p = &c.param[0].mnem;
+
+	if (nparams > 7)
+		return -EINVAL;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_PARAMS_CMD) |
+				  FW_CMD_REQUEST_F | FW_CMD_READ_F |
+				  FW_PARAMS_CMD_PFN_V(pf) |
+				  FW_PARAMS_CMD_VFN_V(vf));
+	c.retval_len16 = cpu_to_be32(FW_LEN16(c));
+
+	for (i = 0; i < nparams; i++) {
+		*p++ = cpu_to_be32(*params++);
+		if (rw)
+			*p = cpu_to_be32(*(val + i));
+		p++;
+	}
+
+	ret = t4_wr_mbox_meat(adap, mbox, &c, sizeof(c), &c, sleep_ok);
+	if (ret == 0)
+		for (i = 0, p = &c.param[0].val; i < nparams; i++, p += 2)
+			*val++ = be32_to_cpu(*p);
+	return ret;
+}
+
+int t4_query_params(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		    unsigned int vf, unsigned int nparams, const u32 *params,
+		    u32 *val)
+{
+	return t4_query_params_rw(adap, mbox, pf, vf, nparams, params, val, 0,
+				  true);
+}
+
+int t4_query_params_ns(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		       unsigned int vf, unsigned int nparams, const u32 *params,
+		       u32 *val)
+{
+	return t4_query_params_rw(adap, mbox, pf, vf, nparams, params, val, 0,
+				  false);
+}
+
+/**
+ *      t4_set_params_timeout - sets FW or device parameters
+ *      @adap: the adapter
+ *      @mbox: mailbox to use for the FW command
+ *      @pf: the PF
+ *      @vf: the VF
+ *      @nparams: the number of parameters
+ *      @params: the parameter names
+ *      @val: the parameter values
+ *      @timeout: the timeout time
+ *
+ *      Sets the value of FW or device parameters.  Up to 7 parameters can be
+ *      specified at once.
+ */
+int t4_set_params_timeout(struct adapter *adap, unsigned int mbox,
+			  unsigned int pf, unsigned int vf,
+			  unsigned int nparams, const u32 *params,
+			  const u32 *val, int timeout)
+{
+	struct fw_params_cmd c;
+	__be32 *p = &c.param[0].mnem;
+
+	if (nparams > 7)
+		return -EINVAL;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_PARAMS_CMD) |
+				  FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+				  FW_PARAMS_CMD_PFN_V(pf) |
+				  FW_PARAMS_CMD_VFN_V(vf));
+	c.retval_len16 = cpu_to_be32(FW_LEN16(c));
+
+	while (nparams--) {
+		*p++ = cpu_to_be32(*params++);
+		*p++ = cpu_to_be32(*val++);
+	}
+
+	return t4_wr_mbox_timeout(adap, mbox, &c, sizeof(c), NULL, timeout);
+}
+
+/**
+ *	t4_set_params - sets FW or device parameters
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@pf: the PF
+ *	@vf: the VF
+ *	@nparams: the number of parameters
+ *	@params: the parameter names
+ *	@val: the parameter values
+ *
+ *	Sets the value of FW or device parameters.  Up to 7 parameters can be
+ *	specified at once.
+ */
+int t4_set_params(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		  unsigned int vf, unsigned int nparams, const u32 *params,
+		  const u32 *val)
+{
+	return t4_set_params_timeout(adap, mbox, pf, vf, nparams, params, val,
+				     FW_CMD_MAX_TIMEOUT);
+}
+
+/**
+ *	t4_cfg_pfvf - configure PF/VF resource limits
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@pf: the PF being configured
+ *	@vf: the VF being configured
+ *	@txq: the max number of egress queues
+ *	@txq_eth_ctrl: the max number of egress Ethernet or control queues
+ *	@rxqi: the max number of interrupt-capable ingress queues
+ *	@rxq: the max number of interruptless ingress queues
+ *	@tc: the PCI traffic class
+ *	@vi: the max number of virtual interfaces
+ *	@cmask: the channel access rights mask for the PF/VF
+ *	@pmask: the port access rights mask for the PF/VF
+ *	@nexact: the maximum number of exact MPS filters
+ *	@rcaps: read capabilities
+ *	@wxcaps: write/execute capabilities
+ *
+ *	Configures resource limits and capabilities for a physical or virtual
+ *	function.
+ */
+int t4_cfg_pfvf(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		unsigned int vf, unsigned int txq, unsigned int txq_eth_ctrl,
+		unsigned int rxqi, unsigned int rxq, unsigned int tc,
+		unsigned int vi, unsigned int cmask, unsigned int pmask,
+		unsigned int nexact, unsigned int rcaps, unsigned int wxcaps)
+{
+	struct fw_pfvf_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_PFVF_CMD) | FW_CMD_REQUEST_F |
+				  FW_CMD_WRITE_F | FW_PFVF_CMD_PFN_V(pf) |
+				  FW_PFVF_CMD_VFN_V(vf));
+	c.retval_len16 = cpu_to_be32(FW_LEN16(c));
+	c.niqflint_niq = cpu_to_be32(FW_PFVF_CMD_NIQFLINT_V(rxqi) |
+				     FW_PFVF_CMD_NIQ_V(rxq));
+	c.type_to_neq = cpu_to_be32(FW_PFVF_CMD_CMASK_V(cmask) |
+				    FW_PFVF_CMD_PMASK_V(pmask) |
+				    FW_PFVF_CMD_NEQ_V(txq));
+	c.tc_to_nexactf = cpu_to_be32(FW_PFVF_CMD_TC_V(tc) |
+				      FW_PFVF_CMD_NVI_V(vi) |
+				      FW_PFVF_CMD_NEXACTF_V(nexact));
+	c.r_caps_to_nethctrl = cpu_to_be32(FW_PFVF_CMD_R_CAPS_V(rcaps) |
+					FW_PFVF_CMD_WX_CAPS_V(wxcaps) |
+					FW_PFVF_CMD_NETHCTRL_V(txq_eth_ctrl));
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_alloc_vi - allocate a virtual interface
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@port: physical port associated with the VI
+ *	@pf: the PF owning the VI
+ *	@vf: the VF owning the VI
+ *	@nmac: number of MAC addresses needed (1 to 5)
+ *	@mac: the MAC addresses of the VI
+ *	@rss_size: size of RSS table slice associated with this VI
+ *	@vivld: the destination to store the VI Valid value.
+ *	@vin: the destination to store the VIN value.
+ *
+ *	Allocates a virtual interface for the given physical port.  If @mac is
+ *	not %NULL it contains the MAC addresses of the VI as assigned by FW.
+ *	@mac should be large enough to hold @nmac Ethernet addresses, they are
+ *	stored consecutively so the space needed is @nmac * 6 bytes.
+ *	Returns a negative error number or the non-negative VI id.
+ */
+int t4_alloc_vi(struct adapter *adap, unsigned int mbox, unsigned int port,
+		unsigned int pf, unsigned int vf, unsigned int nmac, u8 *mac,
+		unsigned int *rss_size, u8 *vivld, u8 *vin)
+{
+	int ret;
+	struct fw_vi_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_VI_CMD) | FW_CMD_REQUEST_F |
+				  FW_CMD_WRITE_F | FW_CMD_EXEC_F |
+				  FW_VI_CMD_PFN_V(pf) | FW_VI_CMD_VFN_V(vf));
+	c.alloc_to_len16 = cpu_to_be32(FW_VI_CMD_ALLOC_F | FW_LEN16(c));
+	c.portid_pkd = FW_VI_CMD_PORTID_V(port);
+	c.nmac = nmac - 1;
+
+	ret = t4_wr_mbox(adap, mbox, &c, sizeof(c), &c);
+	if (ret)
+		return ret;
+
+	if (mac) {
+		memcpy(mac, c.mac, sizeof(c.mac));
+		switch (nmac) {
+		case 5:
+			memcpy(mac + 24, c.nmac3, sizeof(c.nmac3));
+			fallthrough;
+		case 4:
+			memcpy(mac + 18, c.nmac2, sizeof(c.nmac2));
+			fallthrough;
+		case 3:
+			memcpy(mac + 12, c.nmac1, sizeof(c.nmac1));
+			fallthrough;
+		case 2:
+			memcpy(mac + 6,  c.nmac0, sizeof(c.nmac0));
+		}
+	}
+	if (rss_size)
+		*rss_size = FW_VI_CMD_RSSSIZE_G(be16_to_cpu(c.rsssize_pkd));
+
+	if (vivld)
+		*vivld = FW_VI_CMD_VFVLD_G(be32_to_cpu(c.alloc_to_len16));
+
+	if (vin)
+		*vin = FW_VI_CMD_VIN_G(be32_to_cpu(c.alloc_to_len16));
+
+	return FW_VI_CMD_VIID_G(be16_to_cpu(c.type_viid));
+}
+
+/**
+ *	t4_free_vi - free a virtual interface
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@pf: the PF owning the VI
+ *	@vf: the VF owning the VI
+ *	@viid: virtual interface identifiler
+ *
+ *	Free a previously allocated virtual interface.
+ */
+int t4_free_vi(struct adapter *adap, unsigned int mbox, unsigned int pf,
+	       unsigned int vf, unsigned int viid)
+{
+	struct fw_vi_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_VI_CMD) |
+				  FW_CMD_REQUEST_F |
+				  FW_CMD_EXEC_F |
+				  FW_VI_CMD_PFN_V(pf) |
+				  FW_VI_CMD_VFN_V(vf));
+	c.alloc_to_len16 = cpu_to_be32(FW_VI_CMD_FREE_F | FW_LEN16(c));
+	c.type_viid = cpu_to_be16(FW_VI_CMD_VIID_V(viid));
+
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), &c);
+}
+
+/**
+ *	t4_set_rxmode - set Rx properties of a virtual interface
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@viid: the VI id
+ *	@viid_mirror: the mirror VI id
+ *	@mtu: the new MTU or -1
+ *	@promisc: 1 to enable promiscuous mode, 0 to disable it, -1 no change
+ *	@all_multi: 1 to enable all-multi mode, 0 to disable it, -1 no change
+ *	@bcast: 1 to enable broadcast Rx, 0 to disable it, -1 no change
+ *	@vlanex: 1 to enable HW VLAN extraction, 0 to disable it, -1 no change
+ *	@sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *	Sets Rx properties of a virtual interface.
+ */
+int t4_set_rxmode(struct adapter *adap, unsigned int mbox, unsigned int viid,
+		  unsigned int viid_mirror, int mtu, int promisc, int all_multi,
+		  int bcast, int vlanex, bool sleep_ok)
+{
+	struct fw_vi_rxmode_cmd c, c_mirror;
+	int ret;
+
+	/* convert to FW values */
+	if (mtu < 0)
+		mtu = FW_RXMODE_MTU_NO_CHG;
+	if (promisc < 0)
+		promisc = FW_VI_RXMODE_CMD_PROMISCEN_M;
+	if (all_multi < 0)
+		all_multi = FW_VI_RXMODE_CMD_ALLMULTIEN_M;
+	if (bcast < 0)
+		bcast = FW_VI_RXMODE_CMD_BROADCASTEN_M;
+	if (vlanex < 0)
+		vlanex = FW_VI_RXMODE_CMD_VLANEXEN_M;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_VI_RXMODE_CMD) |
+				   FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+				   FW_VI_RXMODE_CMD_VIID_V(viid));
+	c.retval_len16 = cpu_to_be32(FW_LEN16(c));
+	c.mtu_to_vlanexen =
+		cpu_to_be32(FW_VI_RXMODE_CMD_MTU_V(mtu) |
+			    FW_VI_RXMODE_CMD_PROMISCEN_V(promisc) |
+			    FW_VI_RXMODE_CMD_ALLMULTIEN_V(all_multi) |
+			    FW_VI_RXMODE_CMD_BROADCASTEN_V(bcast) |
+			    FW_VI_RXMODE_CMD_VLANEXEN_V(vlanex));
+
+	if (viid_mirror) {
+		memcpy(&c_mirror, &c, sizeof(c_mirror));
+		c_mirror.op_to_viid =
+			cpu_to_be32(FW_CMD_OP_V(FW_VI_RXMODE_CMD) |
+				    FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+				    FW_VI_RXMODE_CMD_VIID_V(viid_mirror));
+	}
+
+	ret = t4_wr_mbox_meat(adap, mbox, &c, sizeof(c), NULL, sleep_ok);
+	if (ret)
+		return ret;
+
+	if (viid_mirror)
+		ret = t4_wr_mbox_meat(adap, mbox, &c_mirror, sizeof(c_mirror),
+				      NULL, sleep_ok);
+
+	return ret;
+}
+
+/**
+ *      t4_free_encap_mac_filt - frees MPS entry at given index
+ *      @adap: the adapter
+ *      @viid: the VI id
+ *      @idx: index of MPS entry to be freed
+ *      @sleep_ok: call is allowed to sleep
+ *
+ *      Frees the MPS entry at supplied index
+ *
+ *      Returns a negative error number or zero on success
+ */
+int t4_free_encap_mac_filt(struct adapter *adap, unsigned int viid,
+			   int idx, bool sleep_ok)
+{
+	struct fw_vi_mac_exact *p;
+	struct fw_vi_mac_cmd c;
+	int ret = 0;
+	u32 exact;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_VI_MAC_CMD) |
+				   FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+				   FW_CMD_EXEC_V(0) |
+				   FW_VI_MAC_CMD_VIID_V(viid));
+	exact = FW_VI_MAC_CMD_ENTRY_TYPE_V(FW_VI_MAC_TYPE_EXACTMAC);
+	c.freemacs_to_len16 = cpu_to_be32(FW_VI_MAC_CMD_FREEMACS_V(0) |
+					  exact |
+					  FW_CMD_LEN16_V(1));
+	p = c.u.exact;
+	p->valid_to_idx = cpu_to_be16(FW_VI_MAC_CMD_VALID_F |
+				      FW_VI_MAC_CMD_IDX_V(idx));
+	eth_zero_addr(p->macaddr);
+	ret = t4_wr_mbox_meat(adap, adap->mbox, &c, sizeof(c), &c, sleep_ok);
+	return ret;
+}
+
+/**
+ *	t4_free_raw_mac_filt - Frees a raw mac entry in mps tcam
+ *	@adap: the adapter
+ *	@viid: the VI id
+ *	@addr: the MAC address
+ *	@mask: the mask
+ *	@idx: index of the entry in mps tcam
+ *	@lookup_type: MAC address for inner (1) or outer (0) header
+ *	@port_id: the port index
+ *	@sleep_ok: call is allowed to sleep
+ *
+ *	Removes the mac entry at the specified index using raw mac interface.
+ *
+ *	Returns a negative error number on failure.
+ */
+int t4_free_raw_mac_filt(struct adapter *adap, unsigned int viid,
+			 const u8 *addr, const u8 *mask, unsigned int idx,
+			 u8 lookup_type, u8 port_id, bool sleep_ok)
+{
+	struct fw_vi_mac_cmd c;
+	struct fw_vi_mac_raw *p = &c.u.raw;
+	u32 val;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_VI_MAC_CMD) |
+				   FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+				   FW_CMD_EXEC_V(0) |
+				   FW_VI_MAC_CMD_VIID_V(viid));
+	val = FW_CMD_LEN16_V(1) |
+	      FW_VI_MAC_CMD_ENTRY_TYPE_V(FW_VI_MAC_TYPE_RAW);
+	c.freemacs_to_len16 = cpu_to_be32(FW_VI_MAC_CMD_FREEMACS_V(0) |
+					  FW_CMD_LEN16_V(val));
+
+	p->raw_idx_pkd = cpu_to_be32(FW_VI_MAC_CMD_RAW_IDX_V(idx) |
+				     FW_VI_MAC_ID_BASED_FREE);
+
+	/* Lookup Type. Outer header: 0, Inner header: 1 */
+	p->data0_pkd = cpu_to_be32(DATALKPTYPE_V(lookup_type) |
+				   DATAPORTNUM_V(port_id));
+	/* Lookup mask and port mask */
+	p->data0m_pkd = cpu_to_be64(DATALKPTYPE_V(DATALKPTYPE_M) |
+				    DATAPORTNUM_V(DATAPORTNUM_M));
+
+	/* Copy the address and the mask */
+	memcpy((u8 *)&p->data1[0] + 2, addr, ETH_ALEN);
+	memcpy((u8 *)&p->data1m[0] + 2, mask, ETH_ALEN);
+
+	return t4_wr_mbox_meat(adap, adap->mbox, &c, sizeof(c), &c, sleep_ok);
+}
+
+/**
+ *      t4_alloc_encap_mac_filt - Adds a mac entry in mps tcam with VNI support
+ *      @adap: the adapter
+ *      @viid: the VI id
+ *      @addr: the MAC address
+ *      @mask: the mask
+ *      @vni: the VNI id for the tunnel protocol
+ *      @vni_mask: mask for the VNI id
+ *      @dip_hit: to enable DIP match for the MPS entry
+ *      @lookup_type: MAC address for inner (1) or outer (0) header
+ *      @sleep_ok: call is allowed to sleep
+ *
+ *      Allocates an MPS entry with specified MAC address and VNI value.
+ *
+ *      Returns a negative error number or the allocated index for this mac.
+ */
+int t4_alloc_encap_mac_filt(struct adapter *adap, unsigned int viid,
+			    const u8 *addr, const u8 *mask, unsigned int vni,
+			    unsigned int vni_mask, u8 dip_hit, u8 lookup_type,
+			    bool sleep_ok)
+{
+	struct fw_vi_mac_cmd c;
+	struct fw_vi_mac_vni *p = c.u.exact_vni;
+	int ret = 0;
+	u32 val;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_VI_MAC_CMD) |
+				   FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+				   FW_VI_MAC_CMD_VIID_V(viid));
+	val = FW_CMD_LEN16_V(1) |
+	      FW_VI_MAC_CMD_ENTRY_TYPE_V(FW_VI_MAC_TYPE_EXACTMAC_VNI);
+	c.freemacs_to_len16 = cpu_to_be32(val);
+	p->valid_to_idx = cpu_to_be16(FW_VI_MAC_CMD_VALID_F |
+				      FW_VI_MAC_CMD_IDX_V(FW_VI_MAC_ADD_MAC));
+	memcpy(p->macaddr, addr, sizeof(p->macaddr));
+	memcpy(p->macaddr_mask, mask, sizeof(p->macaddr_mask));
+
+	p->lookup_type_to_vni =
+		cpu_to_be32(FW_VI_MAC_CMD_VNI_V(vni) |
+			    FW_VI_MAC_CMD_DIP_HIT_V(dip_hit) |
+			    FW_VI_MAC_CMD_LOOKUP_TYPE_V(lookup_type));
+	p->vni_mask_pkd = cpu_to_be32(FW_VI_MAC_CMD_VNI_MASK_V(vni_mask));
+	ret = t4_wr_mbox_meat(adap, adap->mbox, &c, sizeof(c), &c, sleep_ok);
+	if (ret == 0)
+		ret = FW_VI_MAC_CMD_IDX_G(be16_to_cpu(p->valid_to_idx));
+	return ret;
+}
+
+/**
+ *	t4_alloc_raw_mac_filt - Adds a mac entry in mps tcam
+ *	@adap: the adapter
+ *	@viid: the VI id
+ *	@addr: the MAC address
+ *	@mask: the mask
+ *	@idx: index at which to add this entry
+ *	@lookup_type: MAC address for inner (1) or outer (0) header
+ *	@port_id: the port index
+ *	@sleep_ok: call is allowed to sleep
+ *
+ *	Adds the mac entry at the specified index using raw mac interface.
+ *
+ *	Returns a negative error number or the allocated index for this mac.
+ */
+int t4_alloc_raw_mac_filt(struct adapter *adap, unsigned int viid,
+			  const u8 *addr, const u8 *mask, unsigned int idx,
+			  u8 lookup_type, u8 port_id, bool sleep_ok)
+{
+	int ret = 0;
+	struct fw_vi_mac_cmd c;
+	struct fw_vi_mac_raw *p = &c.u.raw;
+	u32 val;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_VI_MAC_CMD) |
+				   FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+				   FW_VI_MAC_CMD_VIID_V(viid));
+	val = FW_CMD_LEN16_V(1) |
+	      FW_VI_MAC_CMD_ENTRY_TYPE_V(FW_VI_MAC_TYPE_RAW);
+	c.freemacs_to_len16 = cpu_to_be32(val);
+
+	/* Specify that this is an inner mac address */
+	p->raw_idx_pkd = cpu_to_be32(FW_VI_MAC_CMD_RAW_IDX_V(idx));
+
+	/* Lookup Type. Outer header: 0, Inner header: 1 */
+	p->data0_pkd = cpu_to_be32(DATALKPTYPE_V(lookup_type) |
+				   DATAPORTNUM_V(port_id));
+	/* Lookup mask and port mask */
+	p->data0m_pkd = cpu_to_be64(DATALKPTYPE_V(DATALKPTYPE_M) |
+				    DATAPORTNUM_V(DATAPORTNUM_M));
+
+	/* Copy the address and the mask */
+	memcpy((u8 *)&p->data1[0] + 2, addr, ETH_ALEN);
+	memcpy((u8 *)&p->data1m[0] + 2, mask, ETH_ALEN);
+
+	ret = t4_wr_mbox_meat(adap, adap->mbox, &c, sizeof(c), &c, sleep_ok);
+	if (ret == 0) {
+		ret = FW_VI_MAC_CMD_RAW_IDX_G(be32_to_cpu(p->raw_idx_pkd));
+		if (ret != idx)
+			ret = -ENOMEM;
+	}
+
+	return ret;
+}
+
+/**
+ *	t4_alloc_mac_filt - allocates exact-match filters for MAC addresses
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@viid: the VI id
+ *	@free: if true any existing filters for this VI id are first removed
+ *	@naddr: the number of MAC addresses to allocate filters for (up to 7)
+ *	@addr: the MAC address(es)
+ *	@idx: where to store the index of each allocated filter
+ *	@hash: pointer to hash address filter bitmap
+ *	@sleep_ok: call is allowed to sleep
+ *
+ *	Allocates an exact-match filter for each of the supplied addresses and
+ *	sets it to the corresponding address.  If @idx is not %NULL it should
+ *	have at least @naddr entries, each of which will be set to the index of
+ *	the filter allocated for the corresponding MAC address.  If a filter
+ *	could not be allocated for an address its index is set to 0xffff.
+ *	If @hash is not %NULL addresses that fail to allocate an exact filter
+ *	are hashed and update the hash filter bitmap pointed at by @hash.
+ *
+ *	Returns a negative error number or the number of filters allocated.
+ */
+int t4_alloc_mac_filt(struct adapter *adap, unsigned int mbox,
+		      unsigned int viid, bool free, unsigned int naddr,
+		      const u8 **addr, u16 *idx, u64 *hash, bool sleep_ok)
+{
+	int offset, ret = 0;
+	struct fw_vi_mac_cmd c;
+	unsigned int nfilters = 0;
+	unsigned int max_naddr = adap->params.arch.mps_tcam_size;
+	unsigned int rem = naddr;
+
+	if (naddr > max_naddr)
+		return -EINVAL;
+
+	for (offset = 0; offset < naddr ; /**/) {
+		unsigned int fw_naddr = (rem < ARRAY_SIZE(c.u.exact) ?
+					 rem : ARRAY_SIZE(c.u.exact));
+		size_t len16 = DIV_ROUND_UP(offsetof(struct fw_vi_mac_cmd,
+						     u.exact[fw_naddr]), 16);
+		struct fw_vi_mac_exact *p;
+		int i;
+
+		memset(&c, 0, sizeof(c));
+		c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_VI_MAC_CMD) |
+					   FW_CMD_REQUEST_F |
+					   FW_CMD_WRITE_F |
+					   FW_CMD_EXEC_V(free) |
+					   FW_VI_MAC_CMD_VIID_V(viid));
+		c.freemacs_to_len16 =
+			cpu_to_be32(FW_VI_MAC_CMD_FREEMACS_V(free) |
+				    FW_CMD_LEN16_V(len16));
+
+		for (i = 0, p = c.u.exact; i < fw_naddr; i++, p++) {
+			p->valid_to_idx =
+				cpu_to_be16(FW_VI_MAC_CMD_VALID_F |
+					    FW_VI_MAC_CMD_IDX_V(
+						    FW_VI_MAC_ADD_MAC));
+			memcpy(p->macaddr, addr[offset + i],
+			       sizeof(p->macaddr));
+		}
+
+		/* It's okay if we run out of space in our MAC address arena.
+		 * Some of the addresses we submit may get stored so we need
+		 * to run through the reply to see what the results were ...
+		 */
+		ret = t4_wr_mbox_meat(adap, mbox, &c, sizeof(c), &c, sleep_ok);
+		if (ret && ret != -FW_ENOMEM)
+			break;
+
+		for (i = 0, p = c.u.exact; i < fw_naddr; i++, p++) {
+			u16 index = FW_VI_MAC_CMD_IDX_G(
+					be16_to_cpu(p->valid_to_idx));
+
+			if (idx)
+				idx[offset + i] = (index >= max_naddr ?
+						   0xffff : index);
+			if (index < max_naddr)
+				nfilters++;
+			else if (hash)
+				*hash |= (1ULL <<
+					  hash_mac_addr(addr[offset + i]));
+		}
+
+		free = false;
+		offset += fw_naddr;
+		rem -= fw_naddr;
+	}
+
+	if (ret == 0 || ret == -FW_ENOMEM)
+		ret = nfilters;
+	return ret;
+}
+
+/**
+ *	t4_free_mac_filt - frees exact-match filters of given MAC addresses
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@viid: the VI id
+ *	@naddr: the number of MAC addresses to allocate filters for (up to 7)
+ *	@addr: the MAC address(es)
+ *	@sleep_ok: call is allowed to sleep
+ *
+ *	Frees the exact-match filter for each of the supplied addresses
+ *
+ *	Returns a negative error number or the number of filters freed.
+ */
+int t4_free_mac_filt(struct adapter *adap, unsigned int mbox,
+		     unsigned int viid, unsigned int naddr,
+		     const u8 **addr, bool sleep_ok)
+{
+	int offset, ret = 0;
+	struct fw_vi_mac_cmd c;
+	unsigned int nfilters = 0;
+	unsigned int max_naddr = is_t4(adap->params.chip) ?
+				       NUM_MPS_CLS_SRAM_L_INSTANCES :
+				       NUM_MPS_T5_CLS_SRAM_L_INSTANCES;
+	unsigned int rem = naddr;
+
+	if (naddr > max_naddr)
+		return -EINVAL;
+
+	for (offset = 0; offset < (int)naddr ; /**/) {
+		unsigned int fw_naddr = (rem < ARRAY_SIZE(c.u.exact)
+					 ? rem
+					 : ARRAY_SIZE(c.u.exact));
+		size_t len16 = DIV_ROUND_UP(offsetof(struct fw_vi_mac_cmd,
+						     u.exact[fw_naddr]), 16);
+		struct fw_vi_mac_exact *p;
+		int i;
+
+		memset(&c, 0, sizeof(c));
+		c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_VI_MAC_CMD) |
+				     FW_CMD_REQUEST_F |
+				     FW_CMD_WRITE_F |
+				     FW_CMD_EXEC_V(0) |
+				     FW_VI_MAC_CMD_VIID_V(viid));
+		c.freemacs_to_len16 =
+				cpu_to_be32(FW_VI_MAC_CMD_FREEMACS_V(0) |
+					    FW_CMD_LEN16_V(len16));
+
+		for (i = 0, p = c.u.exact; i < (int)fw_naddr; i++, p++) {
+			p->valid_to_idx = cpu_to_be16(
+				FW_VI_MAC_CMD_VALID_F |
+				FW_VI_MAC_CMD_IDX_V(FW_VI_MAC_MAC_BASED_FREE));
+			memcpy(p->macaddr, addr[offset+i], sizeof(p->macaddr));
+		}
+
+		ret = t4_wr_mbox_meat(adap, mbox, &c, sizeof(c), &c, sleep_ok);
+		if (ret)
+			break;
+
+		for (i = 0, p = c.u.exact; i < fw_naddr; i++, p++) {
+			u16 index = FW_VI_MAC_CMD_IDX_G(
+						be16_to_cpu(p->valid_to_idx));
+
+			if (index < max_naddr)
+				nfilters++;
+		}
+
+		offset += fw_naddr;
+		rem -= fw_naddr;
+	}
+
+	if (ret == 0)
+		ret = nfilters;
+	return ret;
+}
+
+/**
+ *	t4_change_mac - modifies the exact-match filter for a MAC address
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@viid: the VI id
+ *	@idx: index of existing filter for old value of MAC address, or -1
+ *	@addr: the new MAC address value
+ *	@persist: whether a new MAC allocation should be persistent
+ *	@smt_idx: the destination to store the new SMT index.
+ *
+ *	Modifies an exact-match filter and sets it to the new MAC address.
+ *	Note that in general it is not possible to modify the value of a given
+ *	filter so the generic way to modify an address filter is to free the one
+ *	being used by the old address value and allocate a new filter for the
+ *	new address value.  @idx can be -1 if the address is a new addition.
+ *
+ *	Returns a negative error number or the index of the filter with the new
+ *	MAC value.
+ */
+int t4_change_mac(struct adapter *adap, unsigned int mbox, unsigned int viid,
+		  int idx, const u8 *addr, bool persist, u8 *smt_idx)
+{
+	int ret, mode;
+	struct fw_vi_mac_cmd c;
+	struct fw_vi_mac_exact *p = c.u.exact;
+	unsigned int max_mac_addr = adap->params.arch.mps_tcam_size;
+
+	if (idx < 0)                             /* new allocation */
+		idx = persist ? FW_VI_MAC_ADD_PERSIST_MAC : FW_VI_MAC_ADD_MAC;
+	mode = smt_idx ? FW_VI_MAC_SMT_AND_MPSTCAM : FW_VI_MAC_MPS_TCAM_ENTRY;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_VI_MAC_CMD) |
+				   FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+				   FW_VI_MAC_CMD_VIID_V(viid));
+	c.freemacs_to_len16 = cpu_to_be32(FW_CMD_LEN16_V(1));
+	p->valid_to_idx = cpu_to_be16(FW_VI_MAC_CMD_VALID_F |
+				      FW_VI_MAC_CMD_SMAC_RESULT_V(mode) |
+				      FW_VI_MAC_CMD_IDX_V(idx));
+	memcpy(p->macaddr, addr, sizeof(p->macaddr));
+
+	ret = t4_wr_mbox(adap, mbox, &c, sizeof(c), &c);
+	if (ret == 0) {
+		ret = FW_VI_MAC_CMD_IDX_G(be16_to_cpu(p->valid_to_idx));
+		if (ret >= max_mac_addr)
+			ret = -ENOMEM;
+		if (smt_idx) {
+			if (adap->params.viid_smt_extn_support) {
+				*smt_idx = FW_VI_MAC_CMD_SMTID_G
+						    (be32_to_cpu(c.op_to_viid));
+			} else {
+				/* In T4/T5, SMT contains 256 SMAC entries
+				 * organized in 128 rows of 2 entries each.
+				 * In T6, SMT contains 256 SMAC entries in
+				 * 256 rows.
+				 */
+				if (CHELSIO_CHIP_VERSION(adap->params.chip) <=
+								     CHELSIO_T5)
+					*smt_idx = (viid & FW_VIID_VIN_M) << 1;
+				else
+					*smt_idx = (viid & FW_VIID_VIN_M);
+			}
+		}
+	}
+	return ret;
+}
+
+/**
+ *	t4_set_addr_hash - program the MAC inexact-match hash filter
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@viid: the VI id
+ *	@ucast: whether the hash filter should also match unicast addresses
+ *	@vec: the value to be written to the hash filter
+ *	@sleep_ok: call is allowed to sleep
+ *
+ *	Sets the 64-bit inexact-match hash filter for a virtual interface.
+ */
+int t4_set_addr_hash(struct adapter *adap, unsigned int mbox, unsigned int viid,
+		     bool ucast, u64 vec, bool sleep_ok)
+{
+	struct fw_vi_mac_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_VI_MAC_CMD) |
+				   FW_CMD_REQUEST_F | FW_CMD_WRITE_F |
+				   FW_VI_ENABLE_CMD_VIID_V(viid));
+	c.freemacs_to_len16 = cpu_to_be32(FW_VI_MAC_CMD_HASHVECEN_F |
+					  FW_VI_MAC_CMD_HASHUNIEN_V(ucast) |
+					  FW_CMD_LEN16_V(1));
+	c.u.hash.hashvec = cpu_to_be64(vec);
+	return t4_wr_mbox_meat(adap, mbox, &c, sizeof(c), NULL, sleep_ok);
+}
+
+/**
+ *      t4_enable_vi_params - enable/disable a virtual interface
+ *      @adap: the adapter
+ *      @mbox: mailbox to use for the FW command
+ *      @viid: the VI id
+ *      @rx_en: 1=enable Rx, 0=disable Rx
+ *      @tx_en: 1=enable Tx, 0=disable Tx
+ *      @dcb_en: 1=enable delivery of Data Center Bridging messages.
+ *
+ *      Enables/disables a virtual interface.  Note that setting DCB Enable
+ *      only makes sense when enabling a Virtual Interface ...
+ */
+int t4_enable_vi_params(struct adapter *adap, unsigned int mbox,
+			unsigned int viid, bool rx_en, bool tx_en, bool dcb_en)
+{
+	struct fw_vi_enable_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_VI_ENABLE_CMD) |
+				   FW_CMD_REQUEST_F | FW_CMD_EXEC_F |
+				   FW_VI_ENABLE_CMD_VIID_V(viid));
+	c.ien_to_len16 = cpu_to_be32(FW_VI_ENABLE_CMD_IEN_V(rx_en) |
+				     FW_VI_ENABLE_CMD_EEN_V(tx_en) |
+				     FW_VI_ENABLE_CMD_DCB_INFO_V(dcb_en) |
+				     FW_LEN16(c));
+	return t4_wr_mbox_ns(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_enable_vi - enable/disable a virtual interface
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@viid: the VI id
+ *	@rx_en: 1=enable Rx, 0=disable Rx
+ *	@tx_en: 1=enable Tx, 0=disable Tx
+ *
+ *	Enables/disables a virtual interface.
+ */
+int t4_enable_vi(struct adapter *adap, unsigned int mbox, unsigned int viid,
+		 bool rx_en, bool tx_en)
+{
+	return t4_enable_vi_params(adap, mbox, viid, rx_en, tx_en, 0);
+}
+
+/**
+ *	t4_enable_pi_params - enable/disable a Port's Virtual Interface
+ *      @adap: the adapter
+ *      @mbox: mailbox to use for the FW command
+ *      @pi: the Port Information structure
+ *      @rx_en: 1=enable Rx, 0=disable Rx
+ *      @tx_en: 1=enable Tx, 0=disable Tx
+ *      @dcb_en: 1=enable delivery of Data Center Bridging messages.
+ *
+ *      Enables/disables a Port's Virtual Interface.  Note that setting DCB
+ *	Enable only makes sense when enabling a Virtual Interface ...
+ *	If the Virtual Interface enable/disable operation is successful,
+ *	we notify the OS-specific code of a potential Link Status change
+ *	via the OS Contract API t4_os_link_changed().
+ */
+int t4_enable_pi_params(struct adapter *adap, unsigned int mbox,
+			struct port_info *pi,
+			bool rx_en, bool tx_en, bool dcb_en)
+{
+	int ret = t4_enable_vi_params(adap, mbox, pi->viid,
+				      rx_en, tx_en, dcb_en);
+	if (ret)
+		return ret;
+	t4_os_link_changed(adap, pi->port_id,
+			   rx_en && tx_en && pi->link_cfg.link_ok);
+	return 0;
+}
+
+/**
+ *	t4_identify_port - identify a VI's port by blinking its LED
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@viid: the VI id
+ *	@nblinks: how many times to blink LED at 2.5 Hz
+ *
+ *	Identifies a VI's port by blinking its LED.
+ */
+int t4_identify_port(struct adapter *adap, unsigned int mbox, unsigned int viid,
+		     unsigned int nblinks)
+{
+	struct fw_vi_enable_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_viid = cpu_to_be32(FW_CMD_OP_V(FW_VI_ENABLE_CMD) |
+				   FW_CMD_REQUEST_F | FW_CMD_EXEC_F |
+				   FW_VI_ENABLE_CMD_VIID_V(viid));
+	c.ien_to_len16 = cpu_to_be32(FW_VI_ENABLE_CMD_LED_F | FW_LEN16(c));
+	c.blinkdur = cpu_to_be16(nblinks);
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_iq_stop - stop an ingress queue and its FLs
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@pf: the PF owning the queues
+ *	@vf: the VF owning the queues
+ *	@iqtype: the ingress queue type (FW_IQ_TYPE_FL_INT_CAP, etc.)
+ *	@iqid: ingress queue id
+ *	@fl0id: FL0 queue id or 0xffff if no attached FL0
+ *	@fl1id: FL1 queue id or 0xffff if no attached FL1
+ *
+ *	Stops an ingress queue and its associated FLs, if any.  This causes
+ *	any current or future data/messages destined for these queues to be
+ *	tossed.
+ */
+int t4_iq_stop(struct adapter *adap, unsigned int mbox, unsigned int pf,
+	       unsigned int vf, unsigned int iqtype, unsigned int iqid,
+	       unsigned int fl0id, unsigned int fl1id)
+{
+	struct fw_iq_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_IQ_CMD) | FW_CMD_REQUEST_F |
+				  FW_CMD_EXEC_F | FW_IQ_CMD_PFN_V(pf) |
+				  FW_IQ_CMD_VFN_V(vf));
+	c.alloc_to_len16 = cpu_to_be32(FW_IQ_CMD_IQSTOP_F | FW_LEN16(c));
+	c.type_to_iqandstindex = cpu_to_be32(FW_IQ_CMD_TYPE_V(iqtype));
+	c.iqid = cpu_to_be16(iqid);
+	c.fl0id = cpu_to_be16(fl0id);
+	c.fl1id = cpu_to_be16(fl1id);
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_iq_free - free an ingress queue and its FLs
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@pf: the PF owning the queues
+ *	@vf: the VF owning the queues
+ *	@iqtype: the ingress queue type
+ *	@iqid: ingress queue id
+ *	@fl0id: FL0 queue id or 0xffff if no attached FL0
+ *	@fl1id: FL1 queue id or 0xffff if no attached FL1
+ *
+ *	Frees an ingress queue and its associated FLs, if any.
+ */
+int t4_iq_free(struct adapter *adap, unsigned int mbox, unsigned int pf,
+	       unsigned int vf, unsigned int iqtype, unsigned int iqid,
+	       unsigned int fl0id, unsigned int fl1id)
+{
+	struct fw_iq_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_IQ_CMD) | FW_CMD_REQUEST_F |
+				  FW_CMD_EXEC_F | FW_IQ_CMD_PFN_V(pf) |
+				  FW_IQ_CMD_VFN_V(vf));
+	c.alloc_to_len16 = cpu_to_be32(FW_IQ_CMD_FREE_F | FW_LEN16(c));
+	c.type_to_iqandstindex = cpu_to_be32(FW_IQ_CMD_TYPE_V(iqtype));
+	c.iqid = cpu_to_be16(iqid);
+	c.fl0id = cpu_to_be16(fl0id);
+	c.fl1id = cpu_to_be16(fl1id);
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_eth_eq_free - free an Ethernet egress queue
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@pf: the PF owning the queue
+ *	@vf: the VF owning the queue
+ *	@eqid: egress queue id
+ *
+ *	Frees an Ethernet egress queue.
+ */
+int t4_eth_eq_free(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		   unsigned int vf, unsigned int eqid)
+{
+	struct fw_eq_eth_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_EQ_ETH_CMD) |
+				  FW_CMD_REQUEST_F | FW_CMD_EXEC_F |
+				  FW_EQ_ETH_CMD_PFN_V(pf) |
+				  FW_EQ_ETH_CMD_VFN_V(vf));
+	c.alloc_to_len16 = cpu_to_be32(FW_EQ_ETH_CMD_FREE_F | FW_LEN16(c));
+	c.eqid_pkd = cpu_to_be32(FW_EQ_ETH_CMD_EQID_V(eqid));
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_ctrl_eq_free - free a control egress queue
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@pf: the PF owning the queue
+ *	@vf: the VF owning the queue
+ *	@eqid: egress queue id
+ *
+ *	Frees a control egress queue.
+ */
+int t4_ctrl_eq_free(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		    unsigned int vf, unsigned int eqid)
+{
+	struct fw_eq_ctrl_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_EQ_CTRL_CMD) |
+				  FW_CMD_REQUEST_F | FW_CMD_EXEC_F |
+				  FW_EQ_CTRL_CMD_PFN_V(pf) |
+				  FW_EQ_CTRL_CMD_VFN_V(vf));
+	c.alloc_to_len16 = cpu_to_be32(FW_EQ_CTRL_CMD_FREE_F | FW_LEN16(c));
+	c.cmpliqid_eqid = cpu_to_be32(FW_EQ_CTRL_CMD_EQID_V(eqid));
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_ofld_eq_free - free an offload egress queue
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@pf: the PF owning the queue
+ *	@vf: the VF owning the queue
+ *	@eqid: egress queue id
+ *
+ *	Frees a control egress queue.
+ */
+int t4_ofld_eq_free(struct adapter *adap, unsigned int mbox, unsigned int pf,
+		    unsigned int vf, unsigned int eqid)
+{
+	struct fw_eq_ofld_cmd c;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_EQ_OFLD_CMD) |
+				  FW_CMD_REQUEST_F | FW_CMD_EXEC_F |
+				  FW_EQ_OFLD_CMD_PFN_V(pf) |
+				  FW_EQ_OFLD_CMD_VFN_V(vf));
+	c.alloc_to_len16 = cpu_to_be32(FW_EQ_OFLD_CMD_FREE_F | FW_LEN16(c));
+	c.eqid_pkd = cpu_to_be32(FW_EQ_OFLD_CMD_EQID_V(eqid));
+	return t4_wr_mbox(adap, mbox, &c, sizeof(c), NULL);
+}
+
+/**
+ *	t4_link_down_rc_str - return a string for a Link Down Reason Code
+ *	@link_down_rc: Link Down Reason Code
+ *
+ *	Returns a string representation of the Link Down Reason Code.
+ */
+static const char *t4_link_down_rc_str(unsigned char link_down_rc)
+{
+	static const char * const reason[] = {
+		"Link Down",
+		"Remote Fault",
+		"Auto-negotiation Failure",
+		"Reserved",
+		"Insufficient Airflow",
+		"Unable To Determine Reason",
+		"No RX Signal Detected",
+		"Reserved",
+	};
+
+	if (link_down_rc >= ARRAY_SIZE(reason))
+		return "Bad Reason Code";
+
+	return reason[link_down_rc];
+}
+
+/* Return the highest speed set in the port capabilities, in Mb/s. */
+static unsigned int fwcap_to_speed(fw_port_cap32_t caps)
+{
+	#define TEST_SPEED_RETURN(__caps_speed, __speed) \
+		do { \
+			if (caps & FW_PORT_CAP32_SPEED_##__caps_speed) \
+				return __speed; \
+		} while (0)
+
+	TEST_SPEED_RETURN(400G, 400000);
+	TEST_SPEED_RETURN(200G, 200000);
+	TEST_SPEED_RETURN(100G, 100000);
+	TEST_SPEED_RETURN(50G,   50000);
+	TEST_SPEED_RETURN(40G,   40000);
+	TEST_SPEED_RETURN(25G,   25000);
+	TEST_SPEED_RETURN(10G,   10000);
+	TEST_SPEED_RETURN(1G,     1000);
+	TEST_SPEED_RETURN(100M,    100);
+
+	#undef TEST_SPEED_RETURN
+
+	return 0;
+}
+
+/**
+ *	fwcap_to_fwspeed - return highest speed in Port Capabilities
+ *	@acaps: advertised Port Capabilities
+ *
+ *	Get the highest speed for the port from the advertised Port
+ *	Capabilities.  It will be either the highest speed from the list of
+ *	speeds or whatever user has set using ethtool.
+ */
+static fw_port_cap32_t fwcap_to_fwspeed(fw_port_cap32_t acaps)
+{
+	#define TEST_SPEED_RETURN(__caps_speed) \
+		do { \
+			if (acaps & FW_PORT_CAP32_SPEED_##__caps_speed) \
+				return FW_PORT_CAP32_SPEED_##__caps_speed; \
+		} while (0)
+
+	TEST_SPEED_RETURN(400G);
+	TEST_SPEED_RETURN(200G);
+	TEST_SPEED_RETURN(100G);
+	TEST_SPEED_RETURN(50G);
+	TEST_SPEED_RETURN(40G);
+	TEST_SPEED_RETURN(25G);
+	TEST_SPEED_RETURN(10G);
+	TEST_SPEED_RETURN(1G);
+	TEST_SPEED_RETURN(100M);
+
+	#undef TEST_SPEED_RETURN
+
+	return 0;
+}
+
+/**
+ *	lstatus_to_fwcap - translate old lstatus to 32-bit Port Capabilities
+ *	@lstatus: old FW_PORT_ACTION_GET_PORT_INFO lstatus value
+ *
+ *	Translates old FW_PORT_ACTION_GET_PORT_INFO lstatus field into new
+ *	32-bit Port Capabilities value.
+ */
+static fw_port_cap32_t lstatus_to_fwcap(u32 lstatus)
+{
+	fw_port_cap32_t linkattr = 0;
+
+	/* Unfortunately the format of the Link Status in the old
+	 * 16-bit Port Information message isn't the same as the
+	 * 16-bit Port Capabilities bitfield used everywhere else ...
+	 */
+	if (lstatus & FW_PORT_CMD_RXPAUSE_F)
+		linkattr |= FW_PORT_CAP32_FC_RX;
+	if (lstatus & FW_PORT_CMD_TXPAUSE_F)
+		linkattr |= FW_PORT_CAP32_FC_TX;
+	if (lstatus & FW_PORT_CMD_LSPEED_V(FW_PORT_CAP_SPEED_100M))
+		linkattr |= FW_PORT_CAP32_SPEED_100M;
+	if (lstatus & FW_PORT_CMD_LSPEED_V(FW_PORT_CAP_SPEED_1G))
+		linkattr |= FW_PORT_CAP32_SPEED_1G;
+	if (lstatus & FW_PORT_CMD_LSPEED_V(FW_PORT_CAP_SPEED_10G))
+		linkattr |= FW_PORT_CAP32_SPEED_10G;
+	if (lstatus & FW_PORT_CMD_LSPEED_V(FW_PORT_CAP_SPEED_25G))
+		linkattr |= FW_PORT_CAP32_SPEED_25G;
+	if (lstatus & FW_PORT_CMD_LSPEED_V(FW_PORT_CAP_SPEED_40G))
+		linkattr |= FW_PORT_CAP32_SPEED_40G;
+	if (lstatus & FW_PORT_CMD_LSPEED_V(FW_PORT_CAP_SPEED_100G))
+		linkattr |= FW_PORT_CAP32_SPEED_100G;
+
+	return linkattr;
+}
+
+/**
+ *	t4_handle_get_port_info - process a FW reply message
+ *	@pi: the port info
+ *	@rpl: start of the FW message
+ *
+ *	Processes a GET_PORT_INFO FW reply message.
+ */
+void t4_handle_get_port_info(struct port_info *pi, const __be64 *rpl)
+{
+	const struct fw_port_cmd *cmd = (const void *)rpl;
+	fw_port_cap32_t pcaps, acaps, lpacaps, linkattr;
+	struct link_config *lc = &pi->link_cfg;
+	struct adapter *adapter = pi->adapter;
+	unsigned int speed, fc, fec, adv_fc;
+	enum fw_port_module_type mod_type;
+	int action, link_ok, linkdnrc;
+	enum fw_port_type port_type;
+
+	/* Extract the various fields from the Port Information message.
+	 */
+	action = FW_PORT_CMD_ACTION_G(be32_to_cpu(cmd->action_to_len16));
+	switch (action) {
+	case FW_PORT_ACTION_GET_PORT_INFO: {
+		u32 lstatus = be32_to_cpu(cmd->u.info.lstatus_to_modtype);
+
+		link_ok = (lstatus & FW_PORT_CMD_LSTATUS_F) != 0;
+		linkdnrc = FW_PORT_CMD_LINKDNRC_G(lstatus);
+		port_type = FW_PORT_CMD_PTYPE_G(lstatus);
+		mod_type = FW_PORT_CMD_MODTYPE_G(lstatus);
+		pcaps = fwcaps16_to_caps32(be16_to_cpu(cmd->u.info.pcap));
+		acaps = fwcaps16_to_caps32(be16_to_cpu(cmd->u.info.acap));
+		lpacaps = fwcaps16_to_caps32(be16_to_cpu(cmd->u.info.lpacap));
+		linkattr = lstatus_to_fwcap(lstatus);
+		break;
+	}
+
+	case FW_PORT_ACTION_GET_PORT_INFO32: {
+		u32 lstatus32;
+
+		lstatus32 = be32_to_cpu(cmd->u.info32.lstatus32_to_cbllen32);
+		link_ok = (lstatus32 & FW_PORT_CMD_LSTATUS32_F) != 0;
+		linkdnrc = FW_PORT_CMD_LINKDNRC32_G(lstatus32);
+		port_type = FW_PORT_CMD_PORTTYPE32_G(lstatus32);
+		mod_type = FW_PORT_CMD_MODTYPE32_G(lstatus32);
+		pcaps = be32_to_cpu(cmd->u.info32.pcaps32);
+		acaps = be32_to_cpu(cmd->u.info32.acaps32);
+		lpacaps = be32_to_cpu(cmd->u.info32.lpacaps32);
+		linkattr = be32_to_cpu(cmd->u.info32.linkattr32);
+		break;
+	}
+
+	default:
+		dev_err(adapter->pdev_dev, "Handle Port Information: Bad Command/Action %#x\n",
+			be32_to_cpu(cmd->action_to_len16));
+		return;
+	}
+
+	fec = fwcap_to_cc_fec(acaps);
+	adv_fc = fwcap_to_cc_pause(acaps);
+	fc = fwcap_to_cc_pause(linkattr);
+	speed = fwcap_to_speed(linkattr);
+
+	/* Reset state for communicating new Transceiver Module status and
+	 * whether the OS-dependent layer wants us to redo the current
+	 * "sticky" L1 Configure Link Parameters.
+	 */
+	lc->new_module = false;
+	lc->redo_l1cfg = false;
+
+	if (mod_type != pi->mod_type) {
+		/* With the newer SFP28 and QSFP28 Transceiver Module Types,
+		 * various fundamental Port Capabilities which used to be
+		 * immutable can now change radically.  We can now have
+		 * Speeds, Auto-Negotiation, Forward Error Correction, etc.
+		 * all change based on what Transceiver Module is inserted.
+		 * So we need to record the Physical "Port" Capabilities on
+		 * every Transceiver Module change.
+		 */
+		lc->pcaps = pcaps;
+
+		/* When a new Transceiver Module is inserted, the Firmware
+		 * will examine its i2c EPROM to determine its type and
+		 * general operating parameters including things like Forward
+		 * Error Control, etc.  Various IEEE 802.3 standards dictate
+		 * how to interpret these i2c values to determine default
+		 * "sutomatic" settings.  We record these for future use when
+		 * the user explicitly requests these standards-based values.
+		 */
+		lc->def_acaps = acaps;
+
+		/* Some versions of the early T6 Firmware "cheated" when
+		 * handling different Transceiver Modules by changing the
+		 * underlaying Port Type reported to the Host Drivers.  As
+		 * such we need to capture whatever Port Type the Firmware
+		 * sends us and record it in case it's different from what we
+		 * were told earlier.  Unfortunately, since Firmware is
+		 * forever, we'll need to keep this code here forever, but in
+		 * later T6 Firmware it should just be an assignment of the
+		 * same value already recorded.
+		 */
+		pi->port_type = port_type;
+
+		/* Record new Module Type information.
+		 */
+		pi->mod_type = mod_type;
+
+		/* Let the OS-dependent layer know if we have a new
+		 * Transceiver Module inserted.
+		 */
+		lc->new_module = t4_is_inserted_mod_type(mod_type);
+
+		t4_os_portmod_changed(adapter, pi->port_id);
+	}
+
+	if (link_ok != lc->link_ok || speed != lc->speed ||
+	    fc != lc->fc || adv_fc != lc->advertised_fc ||
+	    fec != lc->fec) {
+		/* something changed */
+		if (!link_ok && lc->link_ok) {
+			lc->link_down_rc = linkdnrc;
+			dev_warn_ratelimited(adapter->pdev_dev,
+					     "Port %d link down, reason: %s\n",
+					     pi->tx_chan,
+					     t4_link_down_rc_str(linkdnrc));
+		}
+		lc->link_ok = link_ok;
+		lc->speed = speed;
+		lc->advertised_fc = adv_fc;
+		lc->fc = fc;
+		lc->fec = fec;
+
+		lc->lpacaps = lpacaps;
+		lc->acaps = acaps & ADVERT_MASK;
+
+		/* If we're not physically capable of Auto-Negotiation, note
+		 * this as Auto-Negotiation disabled.  Otherwise, we track
+		 * what Auto-Negotiation settings we have.  Note parallel
+		 * structure in t4_link_l1cfg_core() and init_link_config().
+		 */
+		if (!(lc->acaps & FW_PORT_CAP32_ANEG)) {
+			lc->autoneg = AUTONEG_DISABLE;
+		} else if (lc->acaps & FW_PORT_CAP32_ANEG) {
+			lc->autoneg = AUTONEG_ENABLE;
+		} else {
+			/* When Autoneg is disabled, user needs to set
+			 * single speed.
+			 * Similar to cxgb4_ethtool.c: set_link_ksettings
+			 */
+			lc->acaps = 0;
+			lc->speed_caps = fwcap_to_fwspeed(acaps);
+			lc->autoneg = AUTONEG_DISABLE;
+		}
+
+		t4_os_link_changed(adapter, pi->port_id, link_ok);
+	}
+
+	/* If we have a new Transceiver Module and the OS-dependent code has
+	 * told us that it wants us to redo whatever "sticky" L1 Configuration
+	 * Link Parameters are set, do that now.
+	 */
+	if (lc->new_module && lc->redo_l1cfg) {
+		struct link_config old_lc;
+		int ret;
+
+		/* Save the current L1 Configuration and restore it if an
+		 * error occurs.  We probably should fix the l1_cfg*()
+		 * routines not to change the link_config when an error
+		 * occurs ...
+		 */
+		old_lc = *lc;
+		ret = t4_link_l1cfg_ns(adapter, adapter->mbox, pi->lport, lc);
+		if (ret) {
+			*lc = old_lc;
+			dev_warn(adapter->pdev_dev,
+				 "Attempt to update new Transceiver Module settings failed\n");
+		}
+	}
+	lc->new_module = false;
+	lc->redo_l1cfg = false;
+}
+
+/**
+ *	t4_update_port_info - retrieve and update port information if changed
+ *	@pi: the port_info
+ *
+ *	We issue a Get Port Information Command to the Firmware and, if
+ *	successful, we check to see if anything is different from what we
+ *	last recorded and update things accordingly.
+ */
+int t4_update_port_info(struct port_info *pi)
+{
+	unsigned int fw_caps = pi->adapter->params.fw_caps_support;
+	struct fw_port_cmd port_cmd;
+	int ret;
+
+	memset(&port_cmd, 0, sizeof(port_cmd));
+	port_cmd.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PORT_CMD) |
+					    FW_CMD_REQUEST_F | FW_CMD_READ_F |
+					    FW_PORT_CMD_PORTID_V(pi->tx_chan));
+	port_cmd.action_to_len16 = cpu_to_be32(
+		FW_PORT_CMD_ACTION_V(fw_caps == FW_CAPS16
+				     ? FW_PORT_ACTION_GET_PORT_INFO
+				     : FW_PORT_ACTION_GET_PORT_INFO32) |
+		FW_LEN16(port_cmd));
+	ret = t4_wr_mbox(pi->adapter, pi->adapter->mbox,
+			 &port_cmd, sizeof(port_cmd), &port_cmd);
+	if (ret)
+		return ret;
+
+	t4_handle_get_port_info(pi, (__be64 *)&port_cmd);
+	return 0;
+}
+
+/**
+ *	t4_get_link_params - retrieve basic link parameters for given port
+ *	@pi: the port
+ *	@link_okp: value return pointer for link up/down
+ *	@speedp: value return pointer for speed (Mb/s)
+ *	@mtup: value return pointer for mtu
+ *
+ *	Retrieves basic link parameters for a port: link up/down, speed (Mb/s),
+ *	and MTU for a specified port.  A negative error is returned on
+ *	failure; 0 on success.
+ */
+int t4_get_link_params(struct port_info *pi, unsigned int *link_okp,
+		       unsigned int *speedp, unsigned int *mtup)
+{
+	unsigned int fw_caps = pi->adapter->params.fw_caps_support;
+	unsigned int action, link_ok, mtu;
+	struct fw_port_cmd port_cmd;
+	fw_port_cap32_t linkattr;
+	int ret;
+
+	memset(&port_cmd, 0, sizeof(port_cmd));
+	port_cmd.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PORT_CMD) |
+					    FW_CMD_REQUEST_F | FW_CMD_READ_F |
+					    FW_PORT_CMD_PORTID_V(pi->tx_chan));
+	action = (fw_caps == FW_CAPS16
+		  ? FW_PORT_ACTION_GET_PORT_INFO
+		  : FW_PORT_ACTION_GET_PORT_INFO32);
+	port_cmd.action_to_len16 = cpu_to_be32(
+		FW_PORT_CMD_ACTION_V(action) |
+		FW_LEN16(port_cmd));
+	ret = t4_wr_mbox(pi->adapter, pi->adapter->mbox,
+			 &port_cmd, sizeof(port_cmd), &port_cmd);
+	if (ret)
+		return ret;
+
+	if (action == FW_PORT_ACTION_GET_PORT_INFO) {
+		u32 lstatus = be32_to_cpu(port_cmd.u.info.lstatus_to_modtype);
+
+		link_ok = !!(lstatus & FW_PORT_CMD_LSTATUS_F);
+		linkattr = lstatus_to_fwcap(lstatus);
+		mtu = be16_to_cpu(port_cmd.u.info.mtu);
+	} else {
+		u32 lstatus32 =
+			   be32_to_cpu(port_cmd.u.info32.lstatus32_to_cbllen32);
+
+		link_ok = !!(lstatus32 & FW_PORT_CMD_LSTATUS32_F);
+		linkattr = be32_to_cpu(port_cmd.u.info32.linkattr32);
+		mtu = FW_PORT_CMD_MTU32_G(
+			be32_to_cpu(port_cmd.u.info32.auxlinfo32_mtu32));
+	}
+
+	if (link_okp)
+		*link_okp = link_ok;
+	if (speedp)
+		*speedp = fwcap_to_speed(linkattr);
+	if (mtup)
+		*mtup = mtu;
+
+	return 0;
+}
+
+/**
+ *      t4_handle_fw_rpl - process a FW reply message
+ *      @adap: the adapter
+ *      @rpl: start of the FW message
+ *
+ *      Processes a FW message, such as link state change messages.
+ */
+int t4_handle_fw_rpl(struct adapter *adap, const __be64 *rpl)
+{
+	u8 opcode = *(const u8 *)rpl;
+
+	/* This might be a port command ... this simplifies the following
+	 * conditionals ...  We can get away with pre-dereferencing
+	 * action_to_len16 because it's in the first 16 bytes and all messages
+	 * will be at least that long.
+	 */
+	const struct fw_port_cmd *p = (const void *)rpl;
+	unsigned int action =
+		FW_PORT_CMD_ACTION_G(be32_to_cpu(p->action_to_len16));
+
+	if (opcode == FW_PORT_CMD &&
+	    (action == FW_PORT_ACTION_GET_PORT_INFO ||
+	     action == FW_PORT_ACTION_GET_PORT_INFO32)) {
+		int i;
+		int chan = FW_PORT_CMD_PORTID_G(be32_to_cpu(p->op_to_portid));
+		struct port_info *pi = NULL;
+
+		for_each_port(adap, i) {
+			pi = adap2pinfo(adap, i);
+			if (pi->tx_chan == chan)
+				break;
+		}
+
+		t4_handle_get_port_info(pi, rpl);
+	} else {
+		dev_warn(adap->pdev_dev, "Unknown firmware reply %d\n",
+			 opcode);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static void get_pci_mode(struct adapter *adapter, struct pci_params *p)
+{
+	u16 val;
+
+	if (pci_is_pcie(adapter->pdev)) {
+		pcie_capability_read_word(adapter->pdev, PCI_EXP_LNKSTA, &val);
+		p->speed = val & PCI_EXP_LNKSTA_CLS;
+		p->width = (val & PCI_EXP_LNKSTA_NLW) >> 4;
+	}
+}
+
+/**
+ *	init_link_config - initialize a link's SW state
+ *	@lc: pointer to structure holding the link state
+ *	@pcaps: link Port Capabilities
+ *	@acaps: link current Advertised Port Capabilities
+ *
+ *	Initializes the SW state maintained for each link, including the link's
+ *	capabilities and default speed/flow-control/autonegotiation settings.
+ */
+static void init_link_config(struct link_config *lc, fw_port_cap32_t pcaps,
+			     fw_port_cap32_t acaps)
+{
+	lc->pcaps = pcaps;
+	lc->def_acaps = acaps;
+	lc->lpacaps = 0;
+	lc->speed_caps = 0;
+	lc->speed = 0;
+	lc->requested_fc = lc->fc = PAUSE_RX | PAUSE_TX;
+
+	/* For Forward Error Control, we default to whatever the Firmware
+	 * tells us the Link is currently advertising.
+	 */
+	lc->requested_fec = FEC_AUTO;
+	lc->fec = fwcap_to_cc_fec(lc->def_acaps);
+
+	/* If the Port is capable of Auto-Negtotiation, initialize it as
+	 * "enabled" and copy over all of the Physical Port Capabilities
+	 * to the Advertised Port Capabilities.  Otherwise mark it as
+	 * Auto-Negotiate disabled and select the highest supported speed
+	 * for the link.  Note parallel structure in t4_link_l1cfg_core()
+	 * and t4_handle_get_port_info().
+	 */
+	if (lc->pcaps & FW_PORT_CAP32_ANEG) {
+		lc->acaps = lc->pcaps & ADVERT_MASK;
+		lc->autoneg = AUTONEG_ENABLE;
+		lc->requested_fc |= PAUSE_AUTONEG;
+	} else {
+		lc->acaps = 0;
+		lc->autoneg = AUTONEG_DISABLE;
+		lc->speed_caps = fwcap_to_fwspeed(acaps);
+	}
+}
+
+#define CIM_PF_NOACCESS 0xeeeeeeee
+
+int t4_wait_dev_ready(void __iomem *regs)
+{
+	u32 whoami;
+
+	whoami = readl(regs + PL_WHOAMI_A);
+	if (whoami != 0xffffffff && whoami != CIM_PF_NOACCESS)
+		return 0;
+
+	msleep(500);
+	whoami = readl(regs + PL_WHOAMI_A);
+	return (whoami != 0xffffffff && whoami != CIM_PF_NOACCESS ? 0 : -EIO);
+}
+
+struct flash_desc {
+	u32 vendor_and_model_id;
+	u32 size_mb;
+};
+
+static int t4_get_flash_params(struct adapter *adap)
+{
+	/* Table for non-Numonix supported flash parts.  Numonix parts are left
+	 * to the preexisting code.  All flash parts have 64KB sectors.
+	 */
+	static struct flash_desc supported_flash[] = {
+		{ 0x150201, 4 << 20 },       /* Spansion 4MB S25FL032P */
+	};
+
+	unsigned int part, manufacturer;
+	unsigned int density, size = 0;
+	u32 flashid = 0;
+	int ret;
+
+	/* Issue a Read ID Command to the Flash part.  We decode supported
+	 * Flash parts and their sizes from this.  There's a newer Query
+	 * Command which can retrieve detailed geometry information but many
+	 * Flash parts don't support it.
+	 */
+
+	ret = sf1_write(adap, 1, 1, 0, SF_RD_ID);
+	if (!ret)
+		ret = sf1_read(adap, 3, 0, 1, &flashid);
+	t4_write_reg(adap, SF_OP_A, 0);                    /* unlock SF */
+	if (ret)
+		return ret;
+
+	/* Check to see if it's one of our non-standard supported Flash parts.
+	 */
+	for (part = 0; part < ARRAY_SIZE(supported_flash); part++)
+		if (supported_flash[part].vendor_and_model_id == flashid) {
+			adap->params.sf_size = supported_flash[part].size_mb;
+			adap->params.sf_nsec =
+				adap->params.sf_size / SF_SEC_SIZE;
+			goto found;
+		}
+
+	/* Decode Flash part size.  The code below looks repetitive with
+	 * common encodings, but that's not guaranteed in the JEDEC
+	 * specification for the Read JEDEC ID command.  The only thing that
+	 * we're guaranteed by the JEDEC specification is where the
+	 * Manufacturer ID is in the returned result.  After that each
+	 * Manufacturer ~could~ encode things completely differently.
+	 * Note, all Flash parts must have 64KB sectors.
+	 */
+	manufacturer = flashid & 0xff;
+	switch (manufacturer) {
+	case 0x20: { /* Micron/Numonix */
+		/* This Density -> Size decoding table is taken from Micron
+		 * Data Sheets.
+		 */
+		density = (flashid >> 16) & 0xff;
+		switch (density) {
+		case 0x14: /* 1MB */
+			size = 1 << 20;
+			break;
+		case 0x15: /* 2MB */
+			size = 1 << 21;
+			break;
+		case 0x16: /* 4MB */
+			size = 1 << 22;
+			break;
+		case 0x17: /* 8MB */
+			size = 1 << 23;
+			break;
+		case 0x18: /* 16MB */
+			size = 1 << 24;
+			break;
+		case 0x19: /* 32MB */
+			size = 1 << 25;
+			break;
+		case 0x20: /* 64MB */
+			size = 1 << 26;
+			break;
+		case 0x21: /* 128MB */
+			size = 1 << 27;
+			break;
+		case 0x22: /* 256MB */
+			size = 1 << 28;
+			break;
+		}
+		break;
+	}
+	case 0x9d: { /* ISSI -- Integrated Silicon Solution, Inc. */
+		/* This Density -> Size decoding table is taken from ISSI
+		 * Data Sheets.
+		 */
+		density = (flashid >> 16) & 0xff;
+		switch (density) {
+		case 0x16: /* 32 MB */
+			size = 1 << 25;
+			break;
+		case 0x17: /* 64MB */
+			size = 1 << 26;
+			break;
+		}
+		break;
+	}
+	case 0xc2: { /* Macronix */
+		/* This Density -> Size decoding table is taken from Macronix
+		 * Data Sheets.
+		 */
+		density = (flashid >> 16) & 0xff;
+		switch (density) {
+		case 0x17: /* 8MB */
+			size = 1 << 23;
+			break;
+		case 0x18: /* 16MB */
+			size = 1 << 24;
+			break;
+		}
+		break;
+	}
+	case 0xef: { /* Winbond */
+		/* This Density -> Size decoding table is taken from Winbond
+		 * Data Sheets.
+		 */
+		density = (flashid >> 16) & 0xff;
+		switch (density) {
+		case 0x17: /* 8MB */
+			size = 1 << 23;
+			break;
+		case 0x18: /* 16MB */
+			size = 1 << 24;
+			break;
+		}
+		break;
+	}
+	}
+
+	/* If we didn't recognize the FLASH part, that's no real issue: the
+	 * Hardware/Software contract says that Hardware will _*ALWAYS*_
+	 * use a FLASH part which is at least 4MB in size and has 64KB
+	 * sectors.  The unrecognized FLASH part is likely to be much larger
+	 * than 4MB, but that's all we really need.
+	 */
+	if (size == 0) {
+		dev_warn(adap->pdev_dev, "Unknown Flash Part, ID = %#x, assuming 4MB\n",
+			 flashid);
+		size = 1 << 22;
+	}
+
+	/* Store decoded Flash size and fall through into vetting code. */
+	adap->params.sf_size = size;
+	adap->params.sf_nsec = size / SF_SEC_SIZE;
+
+found:
+	if (adap->params.sf_size < FLASH_MIN_SIZE)
+		dev_warn(adap->pdev_dev, "WARNING: Flash Part ID %#x, size %#x < %#x\n",
+			 flashid, adap->params.sf_size, FLASH_MIN_SIZE);
+	return 0;
+}
+
+/**
+ *	t4_prep_adapter - prepare SW and HW for operation
+ *	@adapter: the adapter
+ *
+ *	Initialize adapter SW state for the various HW modules, set initial
+ *	values for some adapter tunables, take PHYs out of reset, and
+ *	initialize the MDIO interface.
+ */
+int t4_prep_adapter(struct adapter *adapter)
+{
+	int ret, ver;
+	uint16_t device_id;
+	u32 pl_rev;
+
+	get_pci_mode(adapter, &adapter->params.pci);
+	pl_rev = REV_G(t4_read_reg(adapter, PL_REV_A));
+
+	ret = t4_get_flash_params(adapter);
+	if (ret < 0) {
+		dev_err(adapter->pdev_dev, "error %d identifying flash\n", ret);
+		return ret;
+	}
+
+	/* Retrieve adapter's device ID
+	 */
+	pci_read_config_word(adapter->pdev, PCI_DEVICE_ID, &device_id);
+	ver = device_id >> 12;
+	adapter->params.chip = 0;
+	switch (ver) {
+	case CHELSIO_T4:
+		adapter->params.chip |= CHELSIO_CHIP_CODE(CHELSIO_T4, pl_rev);
+		adapter->params.arch.sge_fl_db = DBPRIO_F;
+		adapter->params.arch.mps_tcam_size =
+				 NUM_MPS_CLS_SRAM_L_INSTANCES;
+		adapter->params.arch.mps_rplc_size = 128;
+		adapter->params.arch.nchan = NCHAN;
+		adapter->params.arch.pm_stats_cnt = PM_NSTATS;
+		adapter->params.arch.vfcount = 128;
+		/* Congestion map is for 4 channels so that
+		 * MPS can have 4 priority per port.
+		 */
+		adapter->params.arch.cng_ch_bits_log = 2;
+		break;
+	case CHELSIO_T5:
+		adapter->params.chip |= CHELSIO_CHIP_CODE(CHELSIO_T5, pl_rev);
+		adapter->params.arch.sge_fl_db = DBPRIO_F | DBTYPE_F;
+		adapter->params.arch.mps_tcam_size =
+				 NUM_MPS_T5_CLS_SRAM_L_INSTANCES;
+		adapter->params.arch.mps_rplc_size = 128;
+		adapter->params.arch.nchan = NCHAN;
+		adapter->params.arch.pm_stats_cnt = PM_NSTATS;
+		adapter->params.arch.vfcount = 128;
+		adapter->params.arch.cng_ch_bits_log = 2;
+		break;
+	case CHELSIO_T6:
+		adapter->params.chip |= CHELSIO_CHIP_CODE(CHELSIO_T6, pl_rev);
+		adapter->params.arch.sge_fl_db = 0;
+		adapter->params.arch.mps_tcam_size =
+				 NUM_MPS_T5_CLS_SRAM_L_INSTANCES;
+		adapter->params.arch.mps_rplc_size = 256;
+		adapter->params.arch.nchan = 2;
+		adapter->params.arch.pm_stats_cnt = T6_PM_NSTATS;
+		adapter->params.arch.vfcount = 256;
+		/* Congestion map will be for 2 channels so that
+		 * MPS can have 8 priority per port.
+		 */
+		adapter->params.arch.cng_ch_bits_log = 3;
+		break;
+	default:
+		dev_err(adapter->pdev_dev, "Device %d is not supported\n",
+			device_id);
+		return -EINVAL;
+	}
+
+	adapter->params.cim_la_size = CIMLA_SIZE;
+	init_cong_ctrl(adapter->params.a_wnd, adapter->params.b_wnd);
+
+	/*
+	 * Default port for debugging in case we can't reach FW.
+	 */
+	adapter->params.nports = 1;
+	adapter->params.portvec = 1;
+	adapter->params.vpd.cclk = 50000;
+
+	/* Set PCIe completion timeout to 4 seconds. */
+	pcie_capability_clear_and_set_word(adapter->pdev, PCI_EXP_DEVCTL2,
+					   PCI_EXP_DEVCTL2_COMP_TIMEOUT, 0xd);
+	return 0;
+}
+
+/**
+ *	t4_shutdown_adapter - shut down adapter, host & wire
+ *	@adapter: the adapter
+ *
+ *	Perform an emergency shutdown of the adapter and stop it from
+ *	continuing any further communication on the ports or DMA to the
+ *	host.  This is typically used when the adapter and/or firmware
+ *	have crashed and we want to prevent any further accidental
+ *	communication with the rest of the world.  This will also force
+ *	the port Link Status to go down -- if register writes work --
+ *	which should help our peers figure out that we're down.
+ */
+int t4_shutdown_adapter(struct adapter *adapter)
+{
+	int port;
+
+	t4_intr_disable(adapter);
+	t4_write_reg(adapter, DBG_GPIO_EN_A, 0);
+	for_each_port(adapter, port) {
+		u32 a_port_cfg = is_t4(adapter->params.chip) ?
+				       PORT_REG(port, XGMAC_PORT_CFG_A) :
+				       T5_PORT_REG(port, MAC_PORT_CFG_A);
+
+		t4_write_reg(adapter, a_port_cfg,
+			     t4_read_reg(adapter, a_port_cfg)
+			     & ~SIGNAL_DET_V(1));
+	}
+	t4_set_reg_field(adapter, SGE_CONTROL_A, GLOBALENABLE_F, 0);
+
+	return 0;
+}
+
+/**
+ *	t4_bar2_sge_qregs - return BAR2 SGE Queue register information
+ *	@adapter: the adapter
+ *	@qid: the Queue ID
+ *	@qtype: the Ingress or Egress type for @qid
+ *	@user: true if this request is for a user mode queue
+ *	@pbar2_qoffset: BAR2 Queue Offset
+ *	@pbar2_qid: BAR2 Queue ID or 0 for Queue ID inferred SGE Queues
+ *
+ *	Returns the BAR2 SGE Queue Registers information associated with the
+ *	indicated Absolute Queue ID.  These are passed back in return value
+ *	pointers.  @qtype should be T4_BAR2_QTYPE_EGRESS for Egress Queue
+ *	and T4_BAR2_QTYPE_INGRESS for Ingress Queues.
+ *
+ *	This may return an error which indicates that BAR2 SGE Queue
+ *	registers aren't available.  If an error is not returned, then the
+ *	following values are returned:
+ *
+ *	  *@pbar2_qoffset: the BAR2 Offset of the @qid Registers
+ *	  *@pbar2_qid: the BAR2 SGE Queue ID or 0 of @qid
+ *
+ *	If the returned BAR2 Queue ID is 0, then BAR2 SGE registers which
+ *	require the "Inferred Queue ID" ability may be used.  E.g. the
+ *	Write Combining Doorbell Buffer. If the BAR2 Queue ID is not 0,
+ *	then these "Inferred Queue ID" register may not be used.
+ */
+int t4_bar2_sge_qregs(struct adapter *adapter,
+		      unsigned int qid,
+		      enum t4_bar2_qtype qtype,
+		      int user,
+		      u64 *pbar2_qoffset,
+		      unsigned int *pbar2_qid)
+{
+	unsigned int page_shift, page_size, qpp_shift, qpp_mask;
+	u64 bar2_page_offset, bar2_qoffset;
+	unsigned int bar2_qid, bar2_qid_offset, bar2_qinferred;
+
+	/* T4 doesn't support BAR2 SGE Queue registers for kernel mode queues */
+	if (!user && is_t4(adapter->params.chip))
+		return -EINVAL;
+
+	/* Get our SGE Page Size parameters.
+	 */
+	page_shift = adapter->params.sge.hps + 10;
+	page_size = 1 << page_shift;
+
+	/* Get the right Queues per Page parameters for our Queue.
+	 */
+	qpp_shift = (qtype == T4_BAR2_QTYPE_EGRESS
+		     ? adapter->params.sge.eq_qpp
+		     : adapter->params.sge.iq_qpp);
+	qpp_mask = (1 << qpp_shift) - 1;
+
+	/*  Calculate the basics of the BAR2 SGE Queue register area:
+	 *  o The BAR2 page the Queue registers will be in.
+	 *  o The BAR2 Queue ID.
+	 *  o The BAR2 Queue ID Offset into the BAR2 page.
+	 */
+	bar2_page_offset = ((u64)(qid >> qpp_shift) << page_shift);
+	bar2_qid = qid & qpp_mask;
+	bar2_qid_offset = bar2_qid * SGE_UDB_SIZE;
+
+	/* If the BAR2 Queue ID Offset is less than the Page Size, then the
+	 * hardware will infer the Absolute Queue ID simply from the writes to
+	 * the BAR2 Queue ID Offset within the BAR2 Page (and we need to use a
+	 * BAR2 Queue ID of 0 for those writes).  Otherwise, we'll simply
+	 * write to the first BAR2 SGE Queue Area within the BAR2 Page with
+	 * the BAR2 Queue ID and the hardware will infer the Absolute Queue ID
+	 * from the BAR2 Page and BAR2 Queue ID.
+	 *
+	 * One important censequence of this is that some BAR2 SGE registers
+	 * have a "Queue ID" field and we can write the BAR2 SGE Queue ID
+	 * there.  But other registers synthesize the SGE Queue ID purely
+	 * from the writes to the registers -- the Write Combined Doorbell
+	 * Buffer is a good example.  These BAR2 SGE Registers are only
+	 * available for those BAR2 SGE Register areas where the SGE Absolute
+	 * Queue ID can be inferred from simple writes.
+	 */
+	bar2_qoffset = bar2_page_offset;
+	bar2_qinferred = (bar2_qid_offset < page_size);
+	if (bar2_qinferred) {
+		bar2_qoffset += bar2_qid_offset;
+		bar2_qid = 0;
+	}
+
+	*pbar2_qoffset = bar2_qoffset;
+	*pbar2_qid = bar2_qid;
+	return 0;
+}
+
+/**
+ *	t4_init_devlog_params - initialize adapter->params.devlog
+ *	@adap: the adapter
+ *
+ *	Initialize various fields of the adapter's Firmware Device Log
+ *	Parameters structure.
+ */
+int t4_init_devlog_params(struct adapter *adap)
+{
+	struct devlog_params *dparams = &adap->params.devlog;
+	u32 pf_dparams;
+	unsigned int devlog_meminfo;
+	struct fw_devlog_cmd devlog_cmd;
+	int ret;
+
+	/* If we're dealing with newer firmware, the Device Log Parameters
+	 * are stored in a designated register which allows us to access the
+	 * Device Log even if we can't talk to the firmware.
+	 */
+	pf_dparams =
+		t4_read_reg(adap, PCIE_FW_REG(PCIE_FW_PF_A, PCIE_FW_PF_DEVLOG));
+	if (pf_dparams) {
+		unsigned int nentries, nentries128;
+
+		dparams->memtype = PCIE_FW_PF_DEVLOG_MEMTYPE_G(pf_dparams);
+		dparams->start = PCIE_FW_PF_DEVLOG_ADDR16_G(pf_dparams) << 4;
+
+		nentries128 = PCIE_FW_PF_DEVLOG_NENTRIES128_G(pf_dparams);
+		nentries = (nentries128 + 1) * 128;
+		dparams->size = nentries * sizeof(struct fw_devlog_e);
+
+		return 0;
+	}
+
+	/* Otherwise, ask the firmware for it's Device Log Parameters.
+	 */
+	memset(&devlog_cmd, 0, sizeof(devlog_cmd));
+	devlog_cmd.op_to_write = cpu_to_be32(FW_CMD_OP_V(FW_DEVLOG_CMD) |
+					     FW_CMD_REQUEST_F | FW_CMD_READ_F);
+	devlog_cmd.retval_len16 = cpu_to_be32(FW_LEN16(devlog_cmd));
+	ret = t4_wr_mbox(adap, adap->mbox, &devlog_cmd, sizeof(devlog_cmd),
+			 &devlog_cmd);
+	if (ret)
+		return ret;
+
+	devlog_meminfo =
+		be32_to_cpu(devlog_cmd.memtype_devlog_memaddr16_devlog);
+	dparams->memtype = FW_DEVLOG_CMD_MEMTYPE_DEVLOG_G(devlog_meminfo);
+	dparams->start = FW_DEVLOG_CMD_MEMADDR16_DEVLOG_G(devlog_meminfo) << 4;
+	dparams->size = be32_to_cpu(devlog_cmd.memsize_devlog);
+
+	return 0;
+}
+
+/**
+ *	t4_init_sge_params - initialize adap->params.sge
+ *	@adapter: the adapter
+ *
+ *	Initialize various fields of the adapter's SGE Parameters structure.
+ */
+int t4_init_sge_params(struct adapter *adapter)
+{
+	struct sge_params *sge_params = &adapter->params.sge;
+	u32 hps, qpp;
+	unsigned int s_hps, s_qpp;
+
+	/* Extract the SGE Page Size for our PF.
+	 */
+	hps = t4_read_reg(adapter, SGE_HOST_PAGE_SIZE_A);
+	s_hps = (HOSTPAGESIZEPF0_S +
+		 (HOSTPAGESIZEPF1_S - HOSTPAGESIZEPF0_S) * adapter->pf);
+	sge_params->hps = ((hps >> s_hps) & HOSTPAGESIZEPF0_M);
+
+	/* Extract the SGE Egress and Ingess Queues Per Page for our PF.
+	 */
+	s_qpp = (QUEUESPERPAGEPF0_S +
+		(QUEUESPERPAGEPF1_S - QUEUESPERPAGEPF0_S) * adapter->pf);
+	qpp = t4_read_reg(adapter, SGE_EGRESS_QUEUES_PER_PAGE_PF_A);
+	sge_params->eq_qpp = ((qpp >> s_qpp) & QUEUESPERPAGEPF0_M);
+	qpp = t4_read_reg(adapter, SGE_INGRESS_QUEUES_PER_PAGE_PF_A);
+	sge_params->iq_qpp = ((qpp >> s_qpp) & QUEUESPERPAGEPF0_M);
+
+	return 0;
+}
+
+/**
+ *      t4_init_tp_params - initialize adap->params.tp
+ *      @adap: the adapter
+ *      @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ *      Initialize various fields of the adapter's TP Parameters structure.
+ */
+int t4_init_tp_params(struct adapter *adap, bool sleep_ok)
+{
+	u32 param, val, v;
+	int chan, ret;
+
+
+	v = t4_read_reg(adap, TP_TIMER_RESOLUTION_A);
+	adap->params.tp.tre = TIMERRESOLUTION_G(v);
+	adap->params.tp.dack_re = DELAYEDACKRESOLUTION_G(v);
+
+	/* MODQ_REQ_MAP defaults to setting queues 0-3 to chan 0-3 */
+	for (chan = 0; chan < NCHAN; chan++)
+		adap->params.tp.tx_modq[chan] = chan;
+
+	/* Cache the adapter's Compressed Filter Mode/Mask and global Ingress
+	 * Configuration.
+	 */
+	param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) |
+		 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_FILTER) |
+		 FW_PARAMS_PARAM_Y_V(FW_PARAM_DEV_FILTER_MODE_MASK));
+
+	/* Read current value */
+	ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 1,
+			      &param, &val);
+	if (ret == 0) {
+		dev_info(adap->pdev_dev,
+			 "Current filter mode/mask 0x%x:0x%x\n",
+			 FW_PARAMS_PARAM_FILTER_MODE_G(val),
+			 FW_PARAMS_PARAM_FILTER_MASK_G(val));
+		adap->params.tp.vlan_pri_map =
+			FW_PARAMS_PARAM_FILTER_MODE_G(val);
+		adap->params.tp.filter_mask =
+			FW_PARAMS_PARAM_FILTER_MASK_G(val);
+	} else {
+		dev_info(adap->pdev_dev,
+			 "Failed to read filter mode/mask via fw api, using indirect-reg-read\n");
+
+		/* Incase of older-fw (which doesn't expose the api
+		 * FW_PARAM_DEV_FILTER_MODE_MASK) and newer-driver (which uses
+		 * the fw api) combination, fall-back to older method of reading
+		 * the filter mode from indirect-register
+		 */
+		t4_tp_pio_read(adap, &adap->params.tp.vlan_pri_map, 1,
+			       TP_VLAN_PRI_MAP_A, sleep_ok);
+
+		/* With the older-fw and newer-driver combination we might run
+		 * into an issue when user wants to use hash filter region but
+		 * the filter_mask is zero, in this case filter_mask validation
+		 * is tough. To avoid that we set the filter_mask same as filter
+		 * mode, which will behave exactly as the older way of ignoring
+		 * the filter mask validation.
+		 */
+		adap->params.tp.filter_mask = adap->params.tp.vlan_pri_map;
+	}
+
+	t4_tp_pio_read(adap, &adap->params.tp.ingress_config, 1,
+		       TP_INGRESS_CONFIG_A, sleep_ok);
+
+	/* For T6, cache the adapter's compressed error vector
+	 * and passing outer header info for encapsulated packets.
+	 */
+	if (CHELSIO_CHIP_VERSION(adap->params.chip) > CHELSIO_T5) {
+		v = t4_read_reg(adap, TP_OUT_CONFIG_A);
+		adap->params.tp.rx_pkt_encap = (v & CRXPKTENC_F) ? 1 : 0;
+	}
+
+	/* Now that we have TP_VLAN_PRI_MAP cached, we can calculate the field
+	 * shift positions of several elements of the Compressed Filter Tuple
+	 * for this adapter which we need frequently ...
+	 */
+	adap->params.tp.fcoe_shift = t4_filter_field_shift(adap, FCOE_F);
+	adap->params.tp.port_shift = t4_filter_field_shift(adap, PORT_F);
+	adap->params.tp.vnic_shift = t4_filter_field_shift(adap, VNIC_ID_F);
+	adap->params.tp.vlan_shift = t4_filter_field_shift(adap, VLAN_F);
+	adap->params.tp.tos_shift = t4_filter_field_shift(adap, TOS_F);
+	adap->params.tp.protocol_shift = t4_filter_field_shift(adap,
+							       PROTOCOL_F);
+	adap->params.tp.ethertype_shift = t4_filter_field_shift(adap,
+								ETHERTYPE_F);
+	adap->params.tp.macmatch_shift = t4_filter_field_shift(adap,
+							       MACMATCH_F);
+	adap->params.tp.matchtype_shift = t4_filter_field_shift(adap,
+								MPSHITTYPE_F);
+	adap->params.tp.frag_shift = t4_filter_field_shift(adap,
+							   FRAGMENTATION_F);
+
+	/* If TP_INGRESS_CONFIG.VNID == 0, then TP_VLAN_PRI_MAP.VNIC_ID
+	 * represents the presence of an Outer VLAN instead of a VNIC ID.
+	 */
+	if ((adap->params.tp.ingress_config & VNIC_F) == 0)
+		adap->params.tp.vnic_shift = -1;
+
+	v = t4_read_reg(adap, LE_3_DB_HASH_MASK_GEN_IPV4_T6_A);
+	adap->params.tp.hash_filter_mask = v;
+	v = t4_read_reg(adap, LE_4_DB_HASH_MASK_GEN_IPV4_T6_A);
+	adap->params.tp.hash_filter_mask |= ((u64)v << 32);
+	return 0;
+}
+
+/**
+ *      t4_filter_field_shift - calculate filter field shift
+ *      @adap: the adapter
+ *      @filter_sel: the desired field (from TP_VLAN_PRI_MAP bits)
+ *
+ *      Return the shift position of a filter field within the Compressed
+ *      Filter Tuple.  The filter field is specified via its selection bit
+ *      within TP_VLAN_PRI_MAL (filter mode).  E.g. F_VLAN.
+ */
+int t4_filter_field_shift(const struct adapter *adap, int filter_sel)
+{
+	unsigned int filter_mode = adap->params.tp.vlan_pri_map;
+	unsigned int sel;
+	int field_shift;
+
+	if ((filter_mode & filter_sel) == 0)
+		return -1;
+
+	for (sel = 1, field_shift = 0; sel < filter_sel; sel <<= 1) {
+		switch (filter_mode & sel) {
+		case FCOE_F:
+			field_shift += FT_FCOE_W;
+			break;
+		case PORT_F:
+			field_shift += FT_PORT_W;
+			break;
+		case VNIC_ID_F:
+			field_shift += FT_VNIC_ID_W;
+			break;
+		case VLAN_F:
+			field_shift += FT_VLAN_W;
+			break;
+		case TOS_F:
+			field_shift += FT_TOS_W;
+			break;
+		case PROTOCOL_F:
+			field_shift += FT_PROTOCOL_W;
+			break;
+		case ETHERTYPE_F:
+			field_shift += FT_ETHERTYPE_W;
+			break;
+		case MACMATCH_F:
+			field_shift += FT_MACMATCH_W;
+			break;
+		case MPSHITTYPE_F:
+			field_shift += FT_MPSHITTYPE_W;
+			break;
+		case FRAGMENTATION_F:
+			field_shift += FT_FRAGMENTATION_W;
+			break;
+		}
+	}
+	return field_shift;
+}
+
+int t4_init_rss_mode(struct adapter *adap, int mbox)
+{
+	int i, ret;
+	struct fw_rss_vi_config_cmd rvc;
+
+	memset(&rvc, 0, sizeof(rvc));
+
+	for_each_port(adap, i) {
+		struct port_info *p = adap2pinfo(adap, i);
+
+		rvc.op_to_viid =
+			cpu_to_be32(FW_CMD_OP_V(FW_RSS_VI_CONFIG_CMD) |
+				    FW_CMD_REQUEST_F | FW_CMD_READ_F |
+				    FW_RSS_VI_CONFIG_CMD_VIID_V(p->viid));
+		rvc.retval_len16 = cpu_to_be32(FW_LEN16(rvc));
+		ret = t4_wr_mbox(adap, mbox, &rvc, sizeof(rvc), &rvc);
+		if (ret)
+			return ret;
+		p->rss_mode = be32_to_cpu(rvc.u.basicvirtual.defaultq_to_udpen);
+	}
+	return 0;
+}
+
+/**
+ *	t4_init_portinfo - allocate a virtual interface and initialize port_info
+ *	@pi: the port_info
+ *	@mbox: mailbox to use for the FW command
+ *	@port: physical port associated with the VI
+ *	@pf: the PF owning the VI
+ *	@vf: the VF owning the VI
+ *	@mac: the MAC address of the VI
+ *
+ *	Allocates a virtual interface for the given physical port.  If @mac is
+ *	not %NULL it contains the MAC address of the VI as assigned by FW.
+ *	@mac should be large enough to hold an Ethernet address.
+ *	Returns < 0 on error.
+ */
+int t4_init_portinfo(struct port_info *pi, int mbox,
+		     int port, int pf, int vf, u8 mac[])
+{
+	struct adapter *adapter = pi->adapter;
+	unsigned int fw_caps = adapter->params.fw_caps_support;
+	struct fw_port_cmd cmd;
+	unsigned int rss_size;
+	enum fw_port_type port_type;
+	int mdio_addr;
+	fw_port_cap32_t pcaps, acaps;
+	u8 vivld = 0, vin = 0;
+	int ret;
+
+	/* If we haven't yet determined whether we're talking to Firmware
+	 * which knows the new 32-bit Port Capabilities, it's time to find
+	 * out now.  This will also tell new Firmware to send us Port Status
+	 * Updates using the new 32-bit Port Capabilities version of the
+	 * Port Information message.
+	 */
+	if (fw_caps == FW_CAPS_UNKNOWN) {
+		u32 param, val;
+
+		param = (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_PFVF) |
+			 FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_PFVF_PORT_CAPS32));
+		val = 1;
+		ret = t4_set_params(adapter, mbox, pf, vf, 1, &param, &val);
+		fw_caps = (ret == 0 ? FW_CAPS32 : FW_CAPS16);
+		adapter->params.fw_caps_support = fw_caps;
+	}
+
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_portid = cpu_to_be32(FW_CMD_OP_V(FW_PORT_CMD) |
+				       FW_CMD_REQUEST_F | FW_CMD_READ_F |
+				       FW_PORT_CMD_PORTID_V(port));
+	cmd.action_to_len16 = cpu_to_be32(
+		FW_PORT_CMD_ACTION_V(fw_caps == FW_CAPS16
+				     ? FW_PORT_ACTION_GET_PORT_INFO
+				     : FW_PORT_ACTION_GET_PORT_INFO32) |
+		FW_LEN16(cmd));
+	ret = t4_wr_mbox(pi->adapter, mbox, &cmd, sizeof(cmd), &cmd);
+	if (ret)
+		return ret;
+
+	/* Extract the various fields from the Port Information message.
+	 */
+	if (fw_caps == FW_CAPS16) {
+		u32 lstatus = be32_to_cpu(cmd.u.info.lstatus_to_modtype);
+
+		port_type = FW_PORT_CMD_PTYPE_G(lstatus);
+		mdio_addr = ((lstatus & FW_PORT_CMD_MDIOCAP_F)
+			     ? FW_PORT_CMD_MDIOADDR_G(lstatus)
+			     : -1);
+		pcaps = fwcaps16_to_caps32(be16_to_cpu(cmd.u.info.pcap));
+		acaps = fwcaps16_to_caps32(be16_to_cpu(cmd.u.info.acap));
+	} else {
+		u32 lstatus32 = be32_to_cpu(cmd.u.info32.lstatus32_to_cbllen32);
+
+		port_type = FW_PORT_CMD_PORTTYPE32_G(lstatus32);
+		mdio_addr = ((lstatus32 & FW_PORT_CMD_MDIOCAP32_F)
+			     ? FW_PORT_CMD_MDIOADDR32_G(lstatus32)
+			     : -1);
+		pcaps = be32_to_cpu(cmd.u.info32.pcaps32);
+		acaps = be32_to_cpu(cmd.u.info32.acaps32);
+	}
+
+	ret = t4_alloc_vi(pi->adapter, mbox, port, pf, vf, 1, mac, &rss_size,
+			  &vivld, &vin);
+	if (ret < 0)
+		return ret;
+
+	pi->viid = ret;
+	pi->tx_chan = port;
+	pi->lport = port;
+	pi->rss_size = rss_size;
+	pi->rx_cchan = t4_get_tp_e2c_map(pi->adapter, port);
+
+	/* If fw supports returning the VIN as part of FW_VI_CMD,
+	 * save the returned values.
+	 */
+	if (adapter->params.viid_smt_extn_support) {
+		pi->vivld = vivld;
+		pi->vin = vin;
+	} else {
+		/* Retrieve the values from VIID */
+		pi->vivld = FW_VIID_VIVLD_G(pi->viid);
+		pi->vin =  FW_VIID_VIN_G(pi->viid);
+	}
+
+	pi->port_type = port_type;
+	pi->mdio_addr = mdio_addr;
+	pi->mod_type = FW_PORT_MOD_TYPE_NA;
+
+	init_link_config(&pi->link_cfg, pcaps, acaps);
+	return 0;
+}
+
+int t4_port_init(struct adapter *adap, int mbox, int pf, int vf)
+{
+	u8 addr[6];
+	int ret, i, j = 0;
+
+	for_each_port(adap, i) {
+		struct port_info *pi = adap2pinfo(adap, i);
+
+		while ((adap->params.portvec & (1 << j)) == 0)
+			j++;
+
+		ret = t4_init_portinfo(pi, mbox, j, pf, vf, addr);
+		if (ret)
+			return ret;
+
+		eth_hw_addr_set(adap->port[i], addr);
+		j++;
+	}
+	return 0;
+}
+
+int t4_init_port_mirror(struct port_info *pi, u8 mbox, u8 port, u8 pf, u8 vf,
+			u16 *mirror_viid)
+{
+	int ret;
+
+	ret = t4_alloc_vi(pi->adapter, mbox, port, pf, vf, 1, NULL, NULL,
+			  NULL, NULL);
+	if (ret < 0)
+		return ret;
+
+	if (mirror_viid)
+		*mirror_viid = ret;
+
+	return 0;
+}
+
+/**
+ *	t4_read_cimq_cfg - read CIM queue configuration
+ *	@adap: the adapter
+ *	@base: holds the queue base addresses in bytes
+ *	@size: holds the queue sizes in bytes
+ *	@thres: holds the queue full thresholds in bytes
+ *
+ *	Returns the current configuration of the CIM queues, starting with
+ *	the IBQs, then the OBQs.
+ */
+void t4_read_cimq_cfg(struct adapter *adap, u16 *base, u16 *size, u16 *thres)
+{
+	unsigned int i, v;
+	int cim_num_obq = is_t4(adap->params.chip) ?
+				CIM_NUM_OBQ : CIM_NUM_OBQ_T5;
+
+	for (i = 0; i < CIM_NUM_IBQ; i++) {
+		t4_write_reg(adap, CIM_QUEUE_CONFIG_REF_A, IBQSELECT_F |
+			     QUENUMSELECT_V(i));
+		v = t4_read_reg(adap, CIM_QUEUE_CONFIG_CTRL_A);
+		/* value is in 256-byte units */
+		*base++ = CIMQBASE_G(v) * 256;
+		*size++ = CIMQSIZE_G(v) * 256;
+		*thres++ = QUEFULLTHRSH_G(v) * 8; /* 8-byte unit */
+	}
+	for (i = 0; i < cim_num_obq; i++) {
+		t4_write_reg(adap, CIM_QUEUE_CONFIG_REF_A, OBQSELECT_F |
+			     QUENUMSELECT_V(i));
+		v = t4_read_reg(adap, CIM_QUEUE_CONFIG_CTRL_A);
+		/* value is in 256-byte units */
+		*base++ = CIMQBASE_G(v) * 256;
+		*size++ = CIMQSIZE_G(v) * 256;
+	}
+}
+
+/**
+ *	t4_read_cim_ibq - read the contents of a CIM inbound queue
+ *	@adap: the adapter
+ *	@qid: the queue index
+ *	@data: where to store the queue contents
+ *	@n: capacity of @data in 32-bit words
+ *
+ *	Reads the contents of the selected CIM queue starting at address 0 up
+ *	to the capacity of @data.  @n must be a multiple of 4.  Returns < 0 on
+ *	error and the number of 32-bit words actually read on success.
+ */
+int t4_read_cim_ibq(struct adapter *adap, unsigned int qid, u32 *data, size_t n)
+{
+	int i, err, attempts;
+	unsigned int addr;
+	const unsigned int nwords = CIM_IBQ_SIZE * 4;
+
+	if (qid > 5 || (n & 3))
+		return -EINVAL;
+
+	addr = qid * nwords;
+	if (n > nwords)
+		n = nwords;
+
+	/* It might take 3-10ms before the IBQ debug read access is allowed.
+	 * Wait for 1 Sec with a delay of 1 usec.
+	 */
+	attempts = 1000000;
+
+	for (i = 0; i < n; i++, addr++) {
+		t4_write_reg(adap, CIM_IBQ_DBG_CFG_A, IBQDBGADDR_V(addr) |
+			     IBQDBGEN_F);
+		err = t4_wait_op_done(adap, CIM_IBQ_DBG_CFG_A, IBQDBGBUSY_F, 0,
+				      attempts, 1);
+		if (err)
+			return err;
+		*data++ = t4_read_reg(adap, CIM_IBQ_DBG_DATA_A);
+	}
+	t4_write_reg(adap, CIM_IBQ_DBG_CFG_A, 0);
+	return i;
+}
+
+/**
+ *	t4_read_cim_obq - read the contents of a CIM outbound queue
+ *	@adap: the adapter
+ *	@qid: the queue index
+ *	@data: where to store the queue contents
+ *	@n: capacity of @data in 32-bit words
+ *
+ *	Reads the contents of the selected CIM queue starting at address 0 up
+ *	to the capacity of @data.  @n must be a multiple of 4.  Returns < 0 on
+ *	error and the number of 32-bit words actually read on success.
+ */
+int t4_read_cim_obq(struct adapter *adap, unsigned int qid, u32 *data, size_t n)
+{
+	int i, err;
+	unsigned int addr, v, nwords;
+	int cim_num_obq = is_t4(adap->params.chip) ?
+				CIM_NUM_OBQ : CIM_NUM_OBQ_T5;
+
+	if ((qid > (cim_num_obq - 1)) || (n & 3))
+		return -EINVAL;
+
+	t4_write_reg(adap, CIM_QUEUE_CONFIG_REF_A, OBQSELECT_F |
+		     QUENUMSELECT_V(qid));
+	v = t4_read_reg(adap, CIM_QUEUE_CONFIG_CTRL_A);
+
+	addr = CIMQBASE_G(v) * 64;    /* muliple of 256 -> muliple of 4 */
+	nwords = CIMQSIZE_G(v) * 64;  /* same */
+	if (n > nwords)
+		n = nwords;
+
+	for (i = 0; i < n; i++, addr++) {
+		t4_write_reg(adap, CIM_OBQ_DBG_CFG_A, OBQDBGADDR_V(addr) |
+			     OBQDBGEN_F);
+		err = t4_wait_op_done(adap, CIM_OBQ_DBG_CFG_A, OBQDBGBUSY_F, 0,
+				      2, 1);
+		if (err)
+			return err;
+		*data++ = t4_read_reg(adap, CIM_OBQ_DBG_DATA_A);
+	}
+	t4_write_reg(adap, CIM_OBQ_DBG_CFG_A, 0);
+	return i;
+}
+
+/**
+ *	t4_cim_read - read a block from CIM internal address space
+ *	@adap: the adapter
+ *	@addr: the start address within the CIM address space
+ *	@n: number of words to read
+ *	@valp: where to store the result
+ *
+ *	Reads a block of 4-byte words from the CIM intenal address space.
+ */
+int t4_cim_read(struct adapter *adap, unsigned int addr, unsigned int n,
+		unsigned int *valp)
+{
+	int ret = 0;
+
+	if (t4_read_reg(adap, CIM_HOST_ACC_CTRL_A) & HOSTBUSY_F)
+		return -EBUSY;
+
+	for ( ; !ret && n--; addr += 4) {
+		t4_write_reg(adap, CIM_HOST_ACC_CTRL_A, addr);
+		ret = t4_wait_op_done(adap, CIM_HOST_ACC_CTRL_A, HOSTBUSY_F,
+				      0, 5, 2);
+		if (!ret)
+			*valp++ = t4_read_reg(adap, CIM_HOST_ACC_DATA_A);
+	}
+	return ret;
+}
+
+/**
+ *	t4_cim_write - write a block into CIM internal address space
+ *	@adap: the adapter
+ *	@addr: the start address within the CIM address space
+ *	@n: number of words to write
+ *	@valp: set of values to write
+ *
+ *	Writes a block of 4-byte words into the CIM intenal address space.
+ */
+int t4_cim_write(struct adapter *adap, unsigned int addr, unsigned int n,
+		 const unsigned int *valp)
+{
+	int ret = 0;
+
+	if (t4_read_reg(adap, CIM_HOST_ACC_CTRL_A) & HOSTBUSY_F)
+		return -EBUSY;
+
+	for ( ; !ret && n--; addr += 4) {
+		t4_write_reg(adap, CIM_HOST_ACC_DATA_A, *valp++);
+		t4_write_reg(adap, CIM_HOST_ACC_CTRL_A, addr | HOSTWRITE_F);
+		ret = t4_wait_op_done(adap, CIM_HOST_ACC_CTRL_A, HOSTBUSY_F,
+				      0, 5, 2);
+	}
+	return ret;
+}
+
+static int t4_cim_write1(struct adapter *adap, unsigned int addr,
+			 unsigned int val)
+{
+	return t4_cim_write(adap, addr, 1, &val);
+}
+
+/**
+ *	t4_cim_read_la - read CIM LA capture buffer
+ *	@adap: the adapter
+ *	@la_buf: where to store the LA data
+ *	@wrptr: the HW write pointer within the capture buffer
+ *
+ *	Reads the contents of the CIM LA buffer with the most recent entry at
+ *	the end	of the returned data and with the entry at @wrptr first.
+ *	We try to leave the LA in the running state we find it in.
+ */
+int t4_cim_read_la(struct adapter *adap, u32 *la_buf, unsigned int *wrptr)
+{
+	int i, ret;
+	unsigned int cfg, val, idx;
+
+	ret = t4_cim_read(adap, UP_UP_DBG_LA_CFG_A, 1, &cfg);
+	if (ret)
+		return ret;
+
+	if (cfg & UPDBGLAEN_F) {	/* LA is running, freeze it */
+		ret = t4_cim_write1(adap, UP_UP_DBG_LA_CFG_A, 0);
+		if (ret)
+			return ret;
+	}
+
+	ret = t4_cim_read(adap, UP_UP_DBG_LA_CFG_A, 1, &val);
+	if (ret)
+		goto restart;
+
+	idx = UPDBGLAWRPTR_G(val);
+	if (wrptr)
+		*wrptr = idx;
+
+	for (i = 0; i < adap->params.cim_la_size; i++) {
+		ret = t4_cim_write1(adap, UP_UP_DBG_LA_CFG_A,
+				    UPDBGLARDPTR_V(idx) | UPDBGLARDEN_F);
+		if (ret)
+			break;
+		ret = t4_cim_read(adap, UP_UP_DBG_LA_CFG_A, 1, &val);
+		if (ret)
+			break;
+		if (val & UPDBGLARDEN_F) {
+			ret = -ETIMEDOUT;
+			break;
+		}
+		ret = t4_cim_read(adap, UP_UP_DBG_LA_DATA_A, 1, &la_buf[i]);
+		if (ret)
+			break;
+
+		/* Bits 0-3 of UpDbgLaRdPtr can be between 0000 to 1001 to
+		 * identify the 32-bit portion of the full 312-bit data
+		 */
+		if (is_t6(adap->params.chip) && (idx & 0xf) >= 9)
+			idx = (idx & 0xff0) + 0x10;
+		else
+			idx++;
+		/* address can't exceed 0xfff */
+		idx &= UPDBGLARDPTR_M;
+	}
+restart:
+	if (cfg & UPDBGLAEN_F) {
+		int r = t4_cim_write1(adap, UP_UP_DBG_LA_CFG_A,
+				      cfg & ~UPDBGLARDEN_F);
+		if (!ret)
+			ret = r;
+	}
+	return ret;
+}
+
+/**
+ *	t4_tp_read_la - read TP LA capture buffer
+ *	@adap: the adapter
+ *	@la_buf: where to store the LA data
+ *	@wrptr: the HW write pointer within the capture buffer
+ *
+ *	Reads the contents of the TP LA buffer with the most recent entry at
+ *	the end	of the returned data and with the entry at @wrptr first.
+ *	We leave the LA in the running state we find it in.
+ */
+void t4_tp_read_la(struct adapter *adap, u64 *la_buf, unsigned int *wrptr)
+{
+	bool last_incomplete;
+	unsigned int i, cfg, val, idx;
+
+	cfg = t4_read_reg(adap, TP_DBG_LA_CONFIG_A) & 0xffff;
+	if (cfg & DBGLAENABLE_F)			/* freeze LA */
+		t4_write_reg(adap, TP_DBG_LA_CONFIG_A,
+			     adap->params.tp.la_mask | (cfg ^ DBGLAENABLE_F));
+
+	val = t4_read_reg(adap, TP_DBG_LA_CONFIG_A);
+	idx = DBGLAWPTR_G(val);
+	last_incomplete = DBGLAMODE_G(val) >= 2 && (val & DBGLAWHLF_F) == 0;
+	if (last_incomplete)
+		idx = (idx + 1) & DBGLARPTR_M;
+	if (wrptr)
+		*wrptr = idx;
+
+	val &= 0xffff;
+	val &= ~DBGLARPTR_V(DBGLARPTR_M);
+	val |= adap->params.tp.la_mask;
+
+	for (i = 0; i < TPLA_SIZE; i++) {
+		t4_write_reg(adap, TP_DBG_LA_CONFIG_A, DBGLARPTR_V(idx) | val);
+		la_buf[i] = t4_read_reg64(adap, TP_DBG_LA_DATAL_A);
+		idx = (idx + 1) & DBGLARPTR_M;
+	}
+
+	/* Wipe out last entry if it isn't valid */
+	if (last_incomplete)
+		la_buf[TPLA_SIZE - 1] = ~0ULL;
+
+	if (cfg & DBGLAENABLE_F)                    /* restore running state */
+		t4_write_reg(adap, TP_DBG_LA_CONFIG_A,
+			     cfg | adap->params.tp.la_mask);
+}
+
+/* SGE Hung Ingress DMA Warning Threshold time and Warning Repeat Rate (in
+ * seconds).  If we find one of the SGE Ingress DMA State Machines in the same
+ * state for more than the Warning Threshold then we'll issue a warning about
+ * a potential hang.  We'll repeat the warning as the SGE Ingress DMA Channel
+ * appears to be hung every Warning Repeat second till the situation clears.
+ * If the situation clears, we'll note that as well.
+ */
+#define SGE_IDMA_WARN_THRESH 1
+#define SGE_IDMA_WARN_REPEAT 300
+
+/**
+ *	t4_idma_monitor_init - initialize SGE Ingress DMA Monitor
+ *	@adapter: the adapter
+ *	@idma: the adapter IDMA Monitor state
+ *
+ *	Initialize the state of an SGE Ingress DMA Monitor.
+ */
+void t4_idma_monitor_init(struct adapter *adapter,
+			  struct sge_idma_monitor_state *idma)
+{
+	/* Initialize the state variables for detecting an SGE Ingress DMA
+	 * hang.  The SGE has internal counters which count up on each clock
+	 * tick whenever the SGE finds its Ingress DMA State Engines in the
+	 * same state they were on the previous clock tick.  The clock used is
+	 * the Core Clock so we have a limit on the maximum "time" they can
+	 * record; typically a very small number of seconds.  For instance,
+	 * with a 600MHz Core Clock, we can only count up to a bit more than
+	 * 7s.  So we'll synthesize a larger counter in order to not run the
+	 * risk of having the "timers" overflow and give us the flexibility to
+	 * maintain a Hung SGE State Machine of our own which operates across
+	 * a longer time frame.
+	 */
+	idma->idma_1s_thresh = core_ticks_per_usec(adapter) * 1000000; /* 1s */
+	idma->idma_stalled[0] = 0;
+	idma->idma_stalled[1] = 0;
+}
+
+/**
+ *	t4_idma_monitor - monitor SGE Ingress DMA state
+ *	@adapter: the adapter
+ *	@idma: the adapter IDMA Monitor state
+ *	@hz: number of ticks/second
+ *	@ticks: number of ticks since the last IDMA Monitor call
+ */
+void t4_idma_monitor(struct adapter *adapter,
+		     struct sge_idma_monitor_state *idma,
+		     int hz, int ticks)
+{
+	int i, idma_same_state_cnt[2];
+
+	 /* Read the SGE Debug Ingress DMA Same State Count registers.  These
+	  * are counters inside the SGE which count up on each clock when the
+	  * SGE finds its Ingress DMA State Engines in the same states they
+	  * were in the previous clock.  The counters will peg out at
+	  * 0xffffffff without wrapping around so once they pass the 1s
+	  * threshold they'll stay above that till the IDMA state changes.
+	  */
+	t4_write_reg(adapter, SGE_DEBUG_INDEX_A, 13);
+	idma_same_state_cnt[0] = t4_read_reg(adapter, SGE_DEBUG_DATA_HIGH_A);
+	idma_same_state_cnt[1] = t4_read_reg(adapter, SGE_DEBUG_DATA_LOW_A);
+
+	for (i = 0; i < 2; i++) {
+		u32 debug0, debug11;
+
+		/* If the Ingress DMA Same State Counter ("timer") is less
+		 * than 1s, then we can reset our synthesized Stall Timer and
+		 * continue.  If we have previously emitted warnings about a
+		 * potential stalled Ingress Queue, issue a note indicating
+		 * that the Ingress Queue has resumed forward progress.
+		 */
+		if (idma_same_state_cnt[i] < idma->idma_1s_thresh) {
+			if (idma->idma_stalled[i] >= SGE_IDMA_WARN_THRESH * hz)
+				dev_warn(adapter->pdev_dev, "SGE idma%d, queue %u, "
+					 "resumed after %d seconds\n",
+					 i, idma->idma_qid[i],
+					 idma->idma_stalled[i] / hz);
+			idma->idma_stalled[i] = 0;
+			continue;
+		}
+
+		/* Synthesize an SGE Ingress DMA Same State Timer in the Hz
+		 * domain.  The first time we get here it'll be because we
+		 * passed the 1s Threshold; each additional time it'll be
+		 * because the RX Timer Callback is being fired on its regular
+		 * schedule.
+		 *
+		 * If the stall is below our Potential Hung Ingress Queue
+		 * Warning Threshold, continue.
+		 */
+		if (idma->idma_stalled[i] == 0) {
+			idma->idma_stalled[i] = hz;
+			idma->idma_warn[i] = 0;
+		} else {
+			idma->idma_stalled[i] += ticks;
+			idma->idma_warn[i] -= ticks;
+		}
+
+		if (idma->idma_stalled[i] < SGE_IDMA_WARN_THRESH * hz)
+			continue;
+
+		/* We'll issue a warning every SGE_IDMA_WARN_REPEAT seconds.
+		 */
+		if (idma->idma_warn[i] > 0)
+			continue;
+		idma->idma_warn[i] = SGE_IDMA_WARN_REPEAT * hz;
+
+		/* Read and save the SGE IDMA State and Queue ID information.
+		 * We do this every time in case it changes across time ...
+		 * can't be too careful ...
+		 */
+		t4_write_reg(adapter, SGE_DEBUG_INDEX_A, 0);
+		debug0 = t4_read_reg(adapter, SGE_DEBUG_DATA_LOW_A);
+		idma->idma_state[i] = (debug0 >> (i * 9)) & 0x3f;
+
+		t4_write_reg(adapter, SGE_DEBUG_INDEX_A, 11);
+		debug11 = t4_read_reg(adapter, SGE_DEBUG_DATA_LOW_A);
+		idma->idma_qid[i] = (debug11 >> (i * 16)) & 0xffff;
+
+		dev_warn(adapter->pdev_dev, "SGE idma%u, queue %u, potentially stuck in "
+			 "state %u for %d seconds (debug0=%#x, debug11=%#x)\n",
+			 i, idma->idma_qid[i], idma->idma_state[i],
+			 idma->idma_stalled[i] / hz,
+			 debug0, debug11);
+		t4_sge_decode_idma_state(adapter, idma->idma_state[i]);
+	}
+}
+
+/**
+ *	t4_load_cfg - download config file
+ *	@adap: the adapter
+ *	@cfg_data: the cfg text file to write
+ *	@size: text file size
+ *
+ *	Write the supplied config text file to the card's serial flash.
+ */
+int t4_load_cfg(struct adapter *adap, const u8 *cfg_data, unsigned int size)
+{
+	int ret, i, n, cfg_addr;
+	unsigned int addr;
+	unsigned int flash_cfg_start_sec;
+	unsigned int sf_sec_size = adap->params.sf_size / adap->params.sf_nsec;
+
+	cfg_addr = t4_flash_cfg_addr(adap);
+	if (cfg_addr < 0)
+		return cfg_addr;
+
+	addr = cfg_addr;
+	flash_cfg_start_sec = addr / SF_SEC_SIZE;
+
+	if (size > FLASH_CFG_MAX_SIZE) {
+		dev_err(adap->pdev_dev, "cfg file too large, max is %u bytes\n",
+			FLASH_CFG_MAX_SIZE);
+		return -EFBIG;
+	}
+
+	i = DIV_ROUND_UP(FLASH_CFG_MAX_SIZE,	/* # of sectors spanned */
+			 sf_sec_size);
+	ret = t4_flash_erase_sectors(adap, flash_cfg_start_sec,
+				     flash_cfg_start_sec + i - 1);
+	/* If size == 0 then we're simply erasing the FLASH sectors associated
+	 * with the on-adapter Firmware Configuration File.
+	 */
+	if (ret || size == 0)
+		goto out;
+
+	/* this will write to the flash up to SF_PAGE_SIZE at a time */
+	for (i = 0; i < size; i += SF_PAGE_SIZE) {
+		if ((size - i) <  SF_PAGE_SIZE)
+			n = size - i;
+		else
+			n = SF_PAGE_SIZE;
+		ret = t4_write_flash(adap, addr, n, cfg_data, true);
+		if (ret)
+			goto out;
+
+		addr += SF_PAGE_SIZE;
+		cfg_data += SF_PAGE_SIZE;
+	}
+
+out:
+	if (ret)
+		dev_err(adap->pdev_dev, "config file %s failed %d\n",
+			(size == 0 ? "clear" : "download"), ret);
+	return ret;
+}
+
+/**
+ *	t4_set_vf_mac_acl - Set MAC address for the specified VF
+ *	@adapter: The adapter
+ *	@vf: one of the VFs instantiated by the specified PF
+ *	@naddr: the number of MAC addresses
+ *	@addr: the MAC address(es) to be set to the specified VF
+ */
+int t4_set_vf_mac_acl(struct adapter *adapter, unsigned int vf,
+		      unsigned int naddr, u8 *addr)
+{
+	struct fw_acl_mac_cmd cmd;
+
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_ACL_MAC_CMD) |
+				    FW_CMD_REQUEST_F |
+				    FW_CMD_WRITE_F |
+				    FW_ACL_MAC_CMD_PFN_V(adapter->pf) |
+				    FW_ACL_MAC_CMD_VFN_V(vf));
+
+	/* Note: Do not enable the ACL */
+	cmd.en_to_len16 = cpu_to_be32((unsigned int)FW_LEN16(cmd));
+	cmd.nmac = naddr;
+
+	switch (adapter->pf) {
+	case 3:
+		memcpy(cmd.macaddr3, addr, sizeof(cmd.macaddr3));
+		break;
+	case 2:
+		memcpy(cmd.macaddr2, addr, sizeof(cmd.macaddr2));
+		break;
+	case 1:
+		memcpy(cmd.macaddr1, addr, sizeof(cmd.macaddr1));
+		break;
+	case 0:
+		memcpy(cmd.macaddr0, addr, sizeof(cmd.macaddr0));
+		break;
+	}
+
+	return t4_wr_mbox(adapter, adapter->mbox, &cmd, sizeof(cmd), &cmd);
+}
+
+/**
+ * t4_read_pace_tbl - read the pace table
+ * @adap: the adapter
+ * @pace_vals: holds the returned values
+ *
+ * Returns the values of TP's pace table in microseconds.
+ */
+void t4_read_pace_tbl(struct adapter *adap, unsigned int pace_vals[NTX_SCHED])
+{
+	unsigned int i, v;
+
+	for (i = 0; i < NTX_SCHED; i++) {
+		t4_write_reg(adap, TP_PACE_TABLE_A, 0xffff0000 + i);
+		v = t4_read_reg(adap, TP_PACE_TABLE_A);
+		pace_vals[i] = dack_ticks_to_usec(adap, v);
+	}
+}
+
+/**
+ * t4_get_tx_sched - get the configuration of a Tx HW traffic scheduler
+ * @adap: the adapter
+ * @sched: the scheduler index
+ * @kbps: the byte rate in Kbps
+ * @ipg: the interpacket delay in tenths of nanoseconds
+ * @sleep_ok: if true we may sleep while awaiting command completion
+ *
+ * Return the current configuration of a HW Tx scheduler.
+ */
+void t4_get_tx_sched(struct adapter *adap, unsigned int sched,
+		     unsigned int *kbps, unsigned int *ipg, bool sleep_ok)
+{
+	unsigned int v, addr, bpt, cpt;
+
+	if (kbps) {
+		addr = TP_TX_MOD_Q1_Q0_RATE_LIMIT_A - sched / 2;
+		t4_tp_tm_pio_read(adap, &v, 1, addr, sleep_ok);
+		if (sched & 1)
+			v >>= 16;
+		bpt = (v >> 8) & 0xff;
+		cpt = v & 0xff;
+		if (!cpt) {
+			*kbps = 0;	/* scheduler disabled */
+		} else {
+			v = (adap->params.vpd.cclk * 1000) / cpt; /* ticks/s */
+			*kbps = (v * bpt) / 125;
+		}
+	}
+	if (ipg) {
+		addr = TP_TX_MOD_Q1_Q0_TIMER_SEPARATOR_A - sched / 2;
+		t4_tp_tm_pio_read(adap, &v, 1, addr, sleep_ok);
+		if (sched & 1)
+			v >>= 16;
+		v &= 0xffff;
+		*ipg = (10000 * v) / core_ticks_per_usec(adap);
+	}
+}
+
+/* t4_sge_ctxt_rd - read an SGE context through FW
+ * @adap: the adapter
+ * @mbox: mailbox to use for the FW command
+ * @cid: the context id
+ * @ctype: the context type
+ * @data: where to store the context data
+ *
+ * Issues a FW command through the given mailbox to read an SGE context.
+ */
+int t4_sge_ctxt_rd(struct adapter *adap, unsigned int mbox, unsigned int cid,
+		   enum ctxt_type ctype, u32 *data)
+{
+	struct fw_ldst_cmd c;
+	int ret;
+
+	if (ctype == CTXT_FLM)
+		ret = FW_LDST_ADDRSPC_SGE_FLMC;
+	else
+		ret = FW_LDST_ADDRSPC_SGE_CONMC;
+
+	memset(&c, 0, sizeof(c));
+	c.op_to_addrspace = cpu_to_be32(FW_CMD_OP_V(FW_LDST_CMD) |
+					FW_CMD_REQUEST_F | FW_CMD_READ_F |
+					FW_LDST_CMD_ADDRSPACE_V(ret));
+	c.cycles_to_len16 = cpu_to_be32(FW_LEN16(c));
+	c.u.idctxt.physid = cpu_to_be32(cid);
+
+	ret = t4_wr_mbox(adap, mbox, &c, sizeof(c), &c);
+	if (ret == 0) {
+		data[0] = be32_to_cpu(c.u.idctxt.ctxt_data0);
+		data[1] = be32_to_cpu(c.u.idctxt.ctxt_data1);
+		data[2] = be32_to_cpu(c.u.idctxt.ctxt_data2);
+		data[3] = be32_to_cpu(c.u.idctxt.ctxt_data3);
+		data[4] = be32_to_cpu(c.u.idctxt.ctxt_data4);
+		data[5] = be32_to_cpu(c.u.idctxt.ctxt_data5);
+	}
+	return ret;
+}
+
+/**
+ * t4_sge_ctxt_rd_bd - read an SGE context bypassing FW
+ * @adap: the adapter
+ * @cid: the context id
+ * @ctype: the context type
+ * @data: where to store the context data
+ *
+ * Reads an SGE context directly, bypassing FW.  This is only for
+ * debugging when FW is unavailable.
+ */
+int t4_sge_ctxt_rd_bd(struct adapter *adap, unsigned int cid,
+		      enum ctxt_type ctype, u32 *data)
+{
+	int i, ret;
+
+	t4_write_reg(adap, SGE_CTXT_CMD_A, CTXTQID_V(cid) | CTXTTYPE_V(ctype));
+	ret = t4_wait_op_done(adap, SGE_CTXT_CMD_A, BUSY_F, 0, 3, 1);
+	if (!ret)
+		for (i = SGE_CTXT_DATA0_A; i <= SGE_CTXT_DATA5_A; i += 4)
+			*data++ = t4_read_reg(adap, i);
+	return ret;
+}
+
+int t4_sched_params(struct adapter *adapter, u8 type, u8 level, u8 mode,
+		    u8 rateunit, u8 ratemode, u8 channel, u8 class,
+		    u32 minrate, u32 maxrate, u16 weight, u16 pktsize,
+		    u16 burstsize)
+{
+	struct fw_sched_cmd cmd;
+
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.op_to_write = cpu_to_be32(FW_CMD_OP_V(FW_SCHED_CMD) |
+				      FW_CMD_REQUEST_F |
+				      FW_CMD_WRITE_F);
+	cmd.retval_len16 = cpu_to_be32(FW_LEN16(cmd));
+
+	cmd.u.params.sc = FW_SCHED_SC_PARAMS;
+	cmd.u.params.type = type;
+	cmd.u.params.level = level;
+	cmd.u.params.mode = mode;
+	cmd.u.params.ch = channel;
+	cmd.u.params.cl = class;
+	cmd.u.params.unit = rateunit;
+	cmd.u.params.rate = ratemode;
+	cmd.u.params.min = cpu_to_be32(minrate);
+	cmd.u.params.max = cpu_to_be32(maxrate);
+	cmd.u.params.weight = cpu_to_be16(weight);
+	cmd.u.params.pktsize = cpu_to_be16(pktsize);
+	cmd.u.params.burstsize = cpu_to_be16(burstsize);
+
+	return t4_wr_mbox_meat(adapter, adapter->mbox, &cmd, sizeof(cmd),
+			       NULL, 1);
+}
+
+/**
+ *	t4_i2c_rd - read I2C data from adapter
+ *	@adap: the adapter
+ *	@mbox: mailbox to use for the FW command
+ *	@port: Port number if per-port device; <0 if not
+ *	@devid: per-port device ID or absolute device ID
+ *	@offset: byte offset into device I2C space
+ *	@len: byte length of I2C space data
+ *	@buf: buffer in which to return I2C data
+ *
+ *	Reads the I2C data from the indicated device and location.
+ */
+int t4_i2c_rd(struct adapter *adap, unsigned int mbox, int port,
+	      unsigned int devid, unsigned int offset,
+	      unsigned int len, u8 *buf)
+{
+	struct fw_ldst_cmd ldst_cmd, ldst_rpl;
+	unsigned int i2c_max = sizeof(ldst_cmd.u.i2c.data);
+	int ret = 0;
+
+	if (len > I2C_PAGE_SIZE)
+		return -EINVAL;
+
+	/* Dont allow reads that spans multiple pages */
+	if (offset < I2C_PAGE_SIZE && offset + len > I2C_PAGE_SIZE)
+		return -EINVAL;
+
+	memset(&ldst_cmd, 0, sizeof(ldst_cmd));
+	ldst_cmd.op_to_addrspace =
+		cpu_to_be32(FW_CMD_OP_V(FW_LDST_CMD) |
+			    FW_CMD_REQUEST_F |
+			    FW_CMD_READ_F |
+			    FW_LDST_CMD_ADDRSPACE_V(FW_LDST_ADDRSPC_I2C));
+	ldst_cmd.cycles_to_len16 = cpu_to_be32(FW_LEN16(ldst_cmd));
+	ldst_cmd.u.i2c.pid = (port < 0 ? 0xff : port);
+	ldst_cmd.u.i2c.did = devid;
+
+	while (len > 0) {
+		unsigned int i2c_len = (len < i2c_max) ? len : i2c_max;
+
+		ldst_cmd.u.i2c.boffset = offset;
+		ldst_cmd.u.i2c.blen = i2c_len;
+
+		ret = t4_wr_mbox(adap, mbox, &ldst_cmd, sizeof(ldst_cmd),
+				 &ldst_rpl);
+		if (ret)
+			break;
+
+		memcpy(buf, ldst_rpl.u.i2c.data, i2c_len);
+		offset += i2c_len;
+		buf += i2c_len;
+		len -= i2c_len;
+	}
+
+	return ret;
+}
+
+/**
+ *      t4_set_vlan_acl - Set a VLAN id for the specified VF
+ *      @adap: the adapter
+ *      @mbox: mailbox to use for the FW command
+ *      @vf: one of the VFs instantiated by the specified PF
+ *      @vlan: The vlanid to be set
+ */
+int t4_set_vlan_acl(struct adapter *adap, unsigned int mbox, unsigned int vf,
+		    u16 vlan)
+{
+	struct fw_acl_vlan_cmd vlan_cmd;
+	unsigned int enable;
+
+	enable = (vlan ? FW_ACL_VLAN_CMD_EN_F : 0);
+	memset(&vlan_cmd, 0, sizeof(vlan_cmd));
+	vlan_cmd.op_to_vfn = cpu_to_be32(FW_CMD_OP_V(FW_ACL_VLAN_CMD) |
+					 FW_CMD_REQUEST_F |
+					 FW_CMD_WRITE_F |
+					 FW_CMD_EXEC_F |
+					 FW_ACL_VLAN_CMD_PFN_V(adap->pf) |
+					 FW_ACL_VLAN_CMD_VFN_V(vf));
+	vlan_cmd.en_to_len16 = cpu_to_be32(enable | FW_LEN16(vlan_cmd));
+	/* Drop all packets that donot match vlan id */
+	vlan_cmd.dropnovlan_fm = (enable
+				  ? (FW_ACL_VLAN_CMD_DROPNOVLAN_F |
+				     FW_ACL_VLAN_CMD_FM_F) : 0);
+	if (enable != 0) {
+		vlan_cmd.nvlan = 1;
+		vlan_cmd.vlanid[0] = cpu_to_be16(vlan);
+	}
+
+	return t4_wr_mbox(adap, adap->mbox, &vlan_cmd, sizeof(vlan_cmd), NULL);
+}
+
+/**
+ *	modify_device_id - Modifies the device ID of the Boot BIOS image
+ *	@device_id: the device ID to write.
+ *	@boot_data: the boot image to modify.
+ *
+ *	Write the supplied device ID to the boot BIOS image.
+ */
+static void modify_device_id(int device_id, u8 *boot_data)
+{
+	struct cxgb4_pcir_data *pcir_header;
+	struct legacy_pci_rom_hdr *header;
+	u8 *cur_header = boot_data;
+	u16 pcir_offset;
+
+	 /* Loop through all chained images and change the device ID's */
+	do {
+		header = (struct legacy_pci_rom_hdr *)cur_header;
+		pcir_offset = le16_to_cpu(header->pcir_offset);
+		pcir_header = (struct cxgb4_pcir_data *)(cur_header +
+			      pcir_offset);
+
+		/**
+		 * Only modify the Device ID if code type is Legacy or HP.
+		 * 0x00: Okay to modify
+		 * 0x01: FCODE. Do not modify
+		 * 0x03: Okay to modify
+		 * 0x04-0xFF: Do not modify
+		 */
+		if (pcir_header->code_type == CXGB4_HDR_CODE1) {
+			u8 csum = 0;
+			int i;
+
+			/**
+			 * Modify Device ID to match current adatper
+			 */
+			pcir_header->device_id = cpu_to_le16(device_id);
+
+			/**
+			 * Set checksum temporarily to 0.
+			 * We will recalculate it later.
+			 */
+			header->cksum = 0x0;
+
+			/**
+			 * Calculate and update checksum
+			 */
+			for (i = 0; i < (header->size512 * 512); i++)
+				csum += cur_header[i];
+
+			/**
+			 * Invert summed value to create the checksum
+			 * Writing new checksum value directly to the boot data
+			 */
+			cur_header[7] = -csum;
+
+		} else if (pcir_header->code_type == CXGB4_HDR_CODE2) {
+			/**
+			 * Modify Device ID to match current adatper
+			 */
+			pcir_header->device_id = cpu_to_le16(device_id);
+		}
+
+		/**
+		 * Move header pointer up to the next image in the ROM.
+		 */
+		cur_header += header->size512 * 512;
+	} while (!(pcir_header->indicator & CXGB4_HDR_INDI));
+}
+
+/**
+ *	t4_load_boot - download boot flash
+ *	@adap: the adapter
+ *	@boot_data: the boot image to write
+ *	@boot_addr: offset in flash to write boot_data
+ *	@size: image size
+ *
+ *	Write the supplied boot image to the card's serial flash.
+ *	The boot image has the following sections: a 28-byte header and the
+ *	boot image.
+ */
+int t4_load_boot(struct adapter *adap, u8 *boot_data,
+		 unsigned int boot_addr, unsigned int size)
+{
+	unsigned int sf_sec_size = adap->params.sf_size / adap->params.sf_nsec;
+	unsigned int boot_sector = (boot_addr * 1024);
+	struct cxgb4_pci_exp_rom_header *header;
+	struct cxgb4_pcir_data *pcir_header;
+	int pcir_offset;
+	unsigned int i;
+	u16 device_id;
+	int ret, addr;
+
+	/**
+	 * Make sure the boot image does not encroach on the firmware region
+	 */
+	if ((boot_sector + size) >> 16 > FLASH_FW_START_SEC) {
+		dev_err(adap->pdev_dev, "boot image encroaching on firmware region\n");
+		return -EFBIG;
+	}
+
+	/* Get boot header */
+	header = (struct cxgb4_pci_exp_rom_header *)boot_data;
+	pcir_offset = le16_to_cpu(header->pcir_offset);
+	/* PCIR Data Structure */
+	pcir_header = (struct cxgb4_pcir_data *)&boot_data[pcir_offset];
+
+	/**
+	 * Perform some primitive sanity testing to avoid accidentally
+	 * writing garbage over the boot sectors.  We ought to check for
+	 * more but it's not worth it for now ...
+	 */
+	if (size < BOOT_MIN_SIZE || size > BOOT_MAX_SIZE) {
+		dev_err(adap->pdev_dev, "boot image too small/large\n");
+		return -EFBIG;
+	}
+
+	if (le16_to_cpu(header->signature) != BOOT_SIGNATURE) {
+		dev_err(adap->pdev_dev, "Boot image missing signature\n");
+		return -EINVAL;
+	}
+
+	/* Check PCI header signature */
+	if (le32_to_cpu(pcir_header->signature) != PCIR_SIGNATURE) {
+		dev_err(adap->pdev_dev, "PCI header missing signature\n");
+		return -EINVAL;
+	}
+
+	/* Check Vendor ID matches Chelsio ID*/
+	if (le16_to_cpu(pcir_header->vendor_id) != PCI_VENDOR_ID_CHELSIO) {
+		dev_err(adap->pdev_dev, "Vendor ID missing signature\n");
+		return -EINVAL;
+	}
+
+	/**
+	 * The boot sector is comprised of the Expansion-ROM boot, iSCSI boot,
+	 * and Boot configuration data sections. These 3 boot sections span
+	 * sectors 0 to 7 in flash and live right before the FW image location.
+	 */
+	i = DIV_ROUND_UP(size ? size : FLASH_FW_START,  sf_sec_size);
+	ret = t4_flash_erase_sectors(adap, boot_sector >> 16,
+				     (boot_sector >> 16) + i - 1);
+
+	/**
+	 * If size == 0 then we're simply erasing the FLASH sectors associated
+	 * with the on-adapter option ROM file
+	 */
+	if (ret || size == 0)
+		goto out;
+	/* Retrieve adapter's device ID */
+	pci_read_config_word(adap->pdev, PCI_DEVICE_ID, &device_id);
+       /* Want to deal with PF 0 so I strip off PF 4 indicator */
+	device_id = device_id & 0xf0ff;
+
+	 /* Check PCIE Device ID */
+	if (le16_to_cpu(pcir_header->device_id) != device_id) {
+		/**
+		 * Change the device ID in the Boot BIOS image to match
+		 * the Device ID of the current adapter.
+		 */
+		modify_device_id(device_id, boot_data);
+	}
+
+	/**
+	 * Skip over the first SF_PAGE_SIZE worth of data and write it after
+	 * we finish copying the rest of the boot image. This will ensure
+	 * that the BIOS boot header will only be written if the boot image
+	 * was written in full.
+	 */
+	addr = boot_sector;
+	for (size -= SF_PAGE_SIZE; size; size -= SF_PAGE_SIZE) {
+		addr += SF_PAGE_SIZE;
+		boot_data += SF_PAGE_SIZE;
+		ret = t4_write_flash(adap, addr, SF_PAGE_SIZE, boot_data,
+				     false);
+		if (ret)
+			goto out;
+	}
+
+	ret = t4_write_flash(adap, boot_sector, SF_PAGE_SIZE,
+			     (const u8 *)header, false);
+
+out:
+	if (ret)
+		dev_err(adap->pdev_dev, "boot image load failed, error %d\n",
+			ret);
+	return ret;
+}
+
+/**
+ *	t4_flash_bootcfg_addr - return the address of the flash
+ *	optionrom configuration
+ *	@adapter: the adapter
+ *
+ *	Return the address within the flash where the OptionROM Configuration
+ *	is stored, or an error if the device FLASH is too small to contain
+ *	a OptionROM Configuration.
+ */
+static int t4_flash_bootcfg_addr(struct adapter *adapter)
+{
+	/**
+	 * If the device FLASH isn't large enough to hold a Firmware
+	 * Configuration File, return an error.
+	 */
+	if (adapter->params.sf_size <
+	    FLASH_BOOTCFG_START + FLASH_BOOTCFG_MAX_SIZE)
+		return -ENOSPC;
+
+	return FLASH_BOOTCFG_START;
+}
+
+int t4_load_bootcfg(struct adapter *adap, const u8 *cfg_data, unsigned int size)
+{
+	unsigned int sf_sec_size = adap->params.sf_size / adap->params.sf_nsec;
+	struct cxgb4_bootcfg_data *header;
+	unsigned int flash_cfg_start_sec;
+	unsigned int addr, npad;
+	int ret, i, n, cfg_addr;
+
+	cfg_addr = t4_flash_bootcfg_addr(adap);
+	if (cfg_addr < 0)
+		return cfg_addr;
+
+	addr = cfg_addr;
+	flash_cfg_start_sec = addr / SF_SEC_SIZE;
+
+	if (size > FLASH_BOOTCFG_MAX_SIZE) {
+		dev_err(adap->pdev_dev, "bootcfg file too large, max is %u bytes\n",
+			FLASH_BOOTCFG_MAX_SIZE);
+		return -EFBIG;
+	}
+
+	header = (struct cxgb4_bootcfg_data *)cfg_data;
+	if (le16_to_cpu(header->signature) != BOOT_CFG_SIG) {
+		dev_err(adap->pdev_dev, "Wrong bootcfg signature\n");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	i = DIV_ROUND_UP(FLASH_BOOTCFG_MAX_SIZE,
+			 sf_sec_size);
+	ret = t4_flash_erase_sectors(adap, flash_cfg_start_sec,
+				     flash_cfg_start_sec + i - 1);
+
+	/**
+	 * If size == 0 then we're simply erasing the FLASH sectors associated
+	 * with the on-adapter OptionROM Configuration File.
+	 */
+	if (ret || size == 0)
+		goto out;
+
+	/* this will write to the flash up to SF_PAGE_SIZE at a time */
+	for (i = 0; i < size; i += SF_PAGE_SIZE) {
+		n = min_t(u32, size - i, SF_PAGE_SIZE);
+
+		ret = t4_write_flash(adap, addr, n, cfg_data, false);
+		if (ret)
+			goto out;
+
+		addr += SF_PAGE_SIZE;
+		cfg_data += SF_PAGE_SIZE;
+	}
+
+	npad = ((size + 4 - 1) & ~3) - size;
+	for (i = 0; i < npad; i++) {
+		u8 data = 0;
+
+		ret = t4_write_flash(adap, cfg_addr + size + i, 1, &data,
+				     false);
+		if (ret)
+			goto out;
+	}
+
+out:
+	if (ret)
+		dev_err(adap->pdev_dev, "boot config data %s failed %d\n",
+			(size == 0 ? "clear" : "download"), ret);
+	return ret;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h
new file mode 100644
index 000000000..63bc956d2
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h
@@ -0,0 +1,303 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __T4_HW_H
+#define __T4_HW_H
+
+#include <linux/types.h>
+
+enum {
+	NCHAN           = 4,    /* # of HW channels */
+	MAX_MTU         = 9600, /* max MAC MTU, excluding header + FCS */
+	EEPROMSIZE      = 17408,/* Serial EEPROM physical size */
+	EEPROMVSIZE     = 32768,/* Serial EEPROM virtual address space size */
+	EEPROMPFSIZE    = 1024, /* EEPROM writable area size for PFn, n>0 */
+	RSS_NENTRIES    = 2048, /* # of entries in RSS mapping table */
+	T6_RSS_NENTRIES = 4096, /* # of entries in RSS mapping table */
+	TCB_SIZE        = 128,  /* TCB size */
+	NMTUS           = 16,   /* size of MTU table */
+	NCCTRL_WIN      = 32,   /* # of congestion control windows */
+	NTX_SCHED       = 8,    /* # of HW Tx scheduling queues */
+	PM_NSTATS       = 5,    /* # of PM stats */
+	T6_PM_NSTATS    = 7,    /* # of PM stats in T6 */
+	MBOX_LEN        = 64,   /* mailbox size in bytes */
+	TRACE_LEN       = 112,  /* length of trace data and mask */
+	FILTER_OPT_LEN  = 36,   /* filter tuple width for optional components */
+};
+
+enum {
+	CIM_NUM_IBQ    = 6,     /* # of CIM IBQs */
+	CIM_NUM_OBQ    = 6,     /* # of CIM OBQs */
+	CIM_NUM_OBQ_T5 = 8,     /* # of CIM OBQs for T5 adapter */
+	CIMLA_SIZE     = 2048,  /* # of 32-bit words in CIM LA */
+	CIM_PIFLA_SIZE = 64,    /* # of 192-bit words in CIM PIF LA */
+	CIM_MALA_SIZE  = 64,    /* # of 160-bit words in CIM MA LA */
+	CIM_IBQ_SIZE   = 128,   /* # of 128-bit words in a CIM IBQ */
+	CIM_OBQ_SIZE   = 128,   /* # of 128-bit words in a CIM OBQ */
+	TPLA_SIZE      = 128,   /* # of 64-bit words in TP LA */
+	ULPRX_LA_SIZE  = 512,   /* # of 256-bit words in ULP_RX LA */
+};
+
+/* SGE context types */
+enum ctxt_type {
+	CTXT_EGRESS,
+	CTXT_INGRESS,
+	CTXT_FLM,
+	CTXT_CNM,
+};
+
+enum {
+	SF_PAGE_SIZE = 256,           /* serial flash page size */
+	SF_SEC_SIZE = 64 * 1024,      /* serial flash sector size */
+};
+
+enum { RSP_TYPE_FLBUF, RSP_TYPE_CPL, RSP_TYPE_INTR }; /* response entry types */
+
+enum { MBOX_OWNER_NONE, MBOX_OWNER_FW, MBOX_OWNER_DRV };    /* mailbox owners */
+
+enum {
+	SGE_MAX_WR_LEN = 512,     /* max WR size in bytes */
+	SGE_CTXT_SIZE = 24,       /* size of SGE context */
+	SGE_NTIMERS = 6,          /* # of interrupt holdoff timer values */
+	SGE_NCOUNTERS = 4,        /* # of interrupt packet counter values */
+	SGE_NDBQTIMERS = 8,       /* # of Doorbell Queue Timer values */
+	SGE_MAX_IQ_SIZE = 65520,
+
+	SGE_TIMER_RSTRT_CNTR = 6, /* restart RX packet threshold counter */
+	SGE_TIMER_UPD_CIDX = 7,   /* update cidx only */
+
+	SGE_EQ_IDXSIZE = 64,      /* egress queue pidx/cidx unit size */
+
+	SGE_INTRDST_PCI = 0,      /* interrupt destination is PCI-E */
+	SGE_INTRDST_IQ = 1,       /*   destination is an ingress queue */
+
+	SGE_UPDATEDEL_NONE = 0,   /* ingress queue pidx update delivery */
+	SGE_UPDATEDEL_INTR = 1,   /*   interrupt */
+	SGE_UPDATEDEL_STPG = 2,   /*   status page */
+	SGE_UPDATEDEL_BOTH = 3,   /*   interrupt and status page */
+
+	SGE_HOSTFCMODE_NONE = 0,  /* egress queue cidx updates */
+	SGE_HOSTFCMODE_IQ = 1,    /*   sent to ingress queue */
+	SGE_HOSTFCMODE_STPG = 2,  /*   sent to status page */
+	SGE_HOSTFCMODE_BOTH = 3,  /*   ingress queue and status page */
+
+	SGE_FETCHBURSTMIN_16B = 0,/* egress queue descriptor fetch minimum */
+	SGE_FETCHBURSTMIN_32B = 1,
+	SGE_FETCHBURSTMIN_64B = 2,
+	SGE_FETCHBURSTMIN_128B = 3,
+
+	SGE_FETCHBURSTMAX_64B = 0,/* egress queue descriptor fetch maximum */
+	SGE_FETCHBURSTMAX_128B = 1,
+	SGE_FETCHBURSTMAX_256B = 2,
+	SGE_FETCHBURSTMAX_512B = 3,
+
+	SGE_CIDXFLUSHTHRESH_1 = 0,/* egress queue cidx flush threshold */
+	SGE_CIDXFLUSHTHRESH_2 = 1,
+	SGE_CIDXFLUSHTHRESH_4 = 2,
+	SGE_CIDXFLUSHTHRESH_8 = 3,
+	SGE_CIDXFLUSHTHRESH_16 = 4,
+	SGE_CIDXFLUSHTHRESH_32 = 5,
+	SGE_CIDXFLUSHTHRESH_64 = 6,
+	SGE_CIDXFLUSHTHRESH_128 = 7,
+
+	SGE_INGPADBOUNDARY_SHIFT = 5,/* ingress queue pad boundary */
+};
+
+/* PCI-e memory window access */
+enum pcie_memwin {
+	MEMWIN_NIC      = 0,
+	MEMWIN_RSVD1    = 1,
+	MEMWIN_RSVD2    = 2,
+	MEMWIN_RDMA     = 3,
+	MEMWIN_RSVD4    = 4,
+	MEMWIN_FOISCSI  = 5,
+	MEMWIN_CSIOSTOR = 6,
+	MEMWIN_RSVD7    = 7,
+};
+
+struct sge_qstat {                /* data written to SGE queue status entries */
+	__be32 qid;
+	__be16 cidx;
+	__be16 pidx;
+};
+
+/*
+ * Structure for last 128 bits of response descriptors
+ */
+struct rsp_ctrl {
+	__be32 hdrbuflen_pidx;
+	__be32 pldbuflen_qid;
+	union {
+		u8 type_gen;
+		__be64 last_flit;
+	};
+};
+
+#define RSPD_NEWBUF_S    31
+#define RSPD_NEWBUF_V(x) ((x) << RSPD_NEWBUF_S)
+#define RSPD_NEWBUF_F    RSPD_NEWBUF_V(1U)
+
+#define RSPD_LEN_S    0
+#define RSPD_LEN_M    0x7fffffff
+#define RSPD_LEN_G(x) (((x) >> RSPD_LEN_S) & RSPD_LEN_M)
+
+#define RSPD_QID_S    RSPD_LEN_S
+#define RSPD_QID_M    RSPD_LEN_M
+#define RSPD_QID_G(x) RSPD_LEN_G(x)
+
+#define RSPD_GEN_S    7
+
+#define RSPD_TYPE_S    4
+#define RSPD_TYPE_M    0x3
+#define RSPD_TYPE_G(x) (((x) >> RSPD_TYPE_S) & RSPD_TYPE_M)
+
+/* Rx queue interrupt deferral fields: counter enable and timer index */
+#define QINTR_CNT_EN_S    0
+#define QINTR_CNT_EN_V(x) ((x) << QINTR_CNT_EN_S)
+#define QINTR_CNT_EN_F    QINTR_CNT_EN_V(1U)
+
+#define QINTR_TIMER_IDX_S    1
+#define QINTR_TIMER_IDX_M    0x7
+#define QINTR_TIMER_IDX_V(x) ((x) << QINTR_TIMER_IDX_S)
+#define QINTR_TIMER_IDX_G(x) (((x) >> QINTR_TIMER_IDX_S) & QINTR_TIMER_IDX_M)
+
+/*
+ * Flash layout.
+ */
+#define FLASH_START(start)	((start) * SF_SEC_SIZE)
+#define FLASH_MAX_SIZE(nsecs)	((nsecs) * SF_SEC_SIZE)
+
+enum {
+	/*
+	 * Various Expansion-ROM boot images, etc.
+	 */
+	FLASH_EXP_ROM_START_SEC = 0,
+	FLASH_EXP_ROM_NSECS = 6,
+	FLASH_EXP_ROM_START = FLASH_START(FLASH_EXP_ROM_START_SEC),
+	FLASH_EXP_ROM_MAX_SIZE = FLASH_MAX_SIZE(FLASH_EXP_ROM_NSECS),
+
+	/*
+	 * iSCSI Boot Firmware Table (iBFT) and other driver-related
+	 * parameters ...
+	 */
+	FLASH_IBFT_START_SEC = 6,
+	FLASH_IBFT_NSECS = 1,
+	FLASH_IBFT_START = FLASH_START(FLASH_IBFT_START_SEC),
+	FLASH_IBFT_MAX_SIZE = FLASH_MAX_SIZE(FLASH_IBFT_NSECS),
+
+	/*
+	 * Boot configuration data.
+	 */
+	FLASH_BOOTCFG_START_SEC = 7,
+	FLASH_BOOTCFG_NSECS = 1,
+	FLASH_BOOTCFG_START = FLASH_START(FLASH_BOOTCFG_START_SEC),
+	FLASH_BOOTCFG_MAX_SIZE = FLASH_MAX_SIZE(FLASH_BOOTCFG_NSECS),
+
+	/*
+	 * Location of firmware image in FLASH.
+	 */
+	FLASH_FW_START_SEC = 8,
+	FLASH_FW_NSECS = 16,
+	FLASH_FW_START = FLASH_START(FLASH_FW_START_SEC),
+	FLASH_FW_MAX_SIZE = FLASH_MAX_SIZE(FLASH_FW_NSECS),
+
+	/* Location of bootstrap firmware image in FLASH.
+	 */
+	FLASH_FWBOOTSTRAP_START_SEC = 27,
+	FLASH_FWBOOTSTRAP_NSECS = 1,
+	FLASH_FWBOOTSTRAP_START = FLASH_START(FLASH_FWBOOTSTRAP_START_SEC),
+	FLASH_FWBOOTSTRAP_MAX_SIZE = FLASH_MAX_SIZE(FLASH_FWBOOTSTRAP_NSECS),
+
+	/*
+	 * iSCSI persistent/crash information.
+	 */
+	FLASH_ISCSI_CRASH_START_SEC = 29,
+	FLASH_ISCSI_CRASH_NSECS = 1,
+	FLASH_ISCSI_CRASH_START = FLASH_START(FLASH_ISCSI_CRASH_START_SEC),
+	FLASH_ISCSI_CRASH_MAX_SIZE = FLASH_MAX_SIZE(FLASH_ISCSI_CRASH_NSECS),
+
+	/*
+	 * FCoE persistent/crash information.
+	 */
+	FLASH_FCOE_CRASH_START_SEC = 30,
+	FLASH_FCOE_CRASH_NSECS = 1,
+	FLASH_FCOE_CRASH_START = FLASH_START(FLASH_FCOE_CRASH_START_SEC),
+	FLASH_FCOE_CRASH_MAX_SIZE = FLASH_MAX_SIZE(FLASH_FCOE_CRASH_NSECS),
+
+	/*
+	 * Location of Firmware Configuration File in FLASH.  Since the FPGA
+	 * "FLASH" is smaller we need to store the Configuration File in a
+	 * different location -- which will overlap the end of the firmware
+	 * image if firmware ever gets that large ...
+	 */
+	FLASH_CFG_START_SEC = 31,
+	FLASH_CFG_NSECS = 1,
+	FLASH_CFG_START = FLASH_START(FLASH_CFG_START_SEC),
+	FLASH_CFG_MAX_SIZE = FLASH_MAX_SIZE(FLASH_CFG_NSECS),
+
+	/* We don't support FLASH devices which can't support the full
+	 * standard set of sections which we need for normal
+	 * operations.
+	 */
+	FLASH_MIN_SIZE = FLASH_CFG_START + FLASH_CFG_MAX_SIZE,
+
+	FLASH_FPGA_CFG_START_SEC = 15,
+	FLASH_FPGA_CFG_START = FLASH_START(FLASH_FPGA_CFG_START_SEC),
+
+	/*
+	 * Sectors 32-63 are reserved for FLASH failover.
+	 */
+};
+
+#undef FLASH_START
+#undef FLASH_MAX_SIZE
+
+#define SGE_TIMESTAMP_S 0
+#define SGE_TIMESTAMP_M 0xfffffffffffffffULL
+#define SGE_TIMESTAMP_V(x) ((__u64)(x) << SGE_TIMESTAMP_S)
+#define SGE_TIMESTAMP_G(x) (((__u64)(x) >> SGE_TIMESTAMP_S) & SGE_TIMESTAMP_M)
+
+#define I2C_DEV_ADDR_A0		0xa0
+#define I2C_DEV_ADDR_A2		0xa2
+#define I2C_PAGE_SIZE		0x100
+#define SFP_DIAG_TYPE_ADDR	0x5c
+#define SFP_DIAG_TYPE_LEN	0x1
+#define SFP_DIAG_ADDRMODE	BIT(2)
+#define SFP_DIAG_IMPLEMENTED	BIT(6)
+#define SFF_8472_COMP_ADDR	0x5e
+#define SFF_8472_COMP_LEN	0x1
+#define SFF_REV_ADDR		0x1
+#define SFF_REV_LEN		0x1
+
+#endif /* __T4_HW_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_msg.h b/drivers/net/ethernet/chelsio/cxgb4/t4_msg.h
new file mode 100644
index 000000000..fed5f93bf
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_msg.h
@@ -0,0 +1,2352 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __T4_MSG_H
+#define __T4_MSG_H
+
+#include <linux/types.h>
+
+enum {
+	CPL_PASS_OPEN_REQ     = 0x1,
+	CPL_PASS_ACCEPT_RPL   = 0x2,
+	CPL_ACT_OPEN_REQ      = 0x3,
+	CPL_SET_TCB_FIELD     = 0x5,
+	CPL_GET_TCB           = 0x6,
+	CPL_CLOSE_CON_REQ     = 0x8,
+	CPL_CLOSE_LISTSRV_REQ = 0x9,
+	CPL_ABORT_REQ         = 0xA,
+	CPL_ABORT_RPL         = 0xB,
+	CPL_TX_DATA           = 0xC,
+	CPL_RX_DATA_ACK       = 0xD,
+	CPL_TX_PKT            = 0xE,
+	CPL_L2T_WRITE_REQ     = 0x12,
+	CPL_SMT_WRITE_REQ     = 0x14,
+	CPL_TID_RELEASE       = 0x1A,
+	CPL_SRQ_TABLE_REQ     = 0x1C,
+	CPL_TX_DATA_ISO	      = 0x1F,
+
+	CPL_CLOSE_LISTSRV_RPL = 0x20,
+	CPL_GET_TCB_RPL       = 0x22,
+	CPL_L2T_WRITE_RPL     = 0x23,
+	CPL_PASS_OPEN_RPL     = 0x24,
+	CPL_ACT_OPEN_RPL      = 0x25,
+	CPL_PEER_CLOSE        = 0x26,
+	CPL_ABORT_REQ_RSS     = 0x2B,
+	CPL_ABORT_RPL_RSS     = 0x2D,
+	CPL_SMT_WRITE_RPL     = 0x2E,
+
+	CPL_RX_PHYS_ADDR      = 0x30,
+	CPL_CLOSE_CON_RPL     = 0x32,
+	CPL_ISCSI_HDR         = 0x33,
+	CPL_RDMA_CQE          = 0x35,
+	CPL_RDMA_CQE_READ_RSP = 0x36,
+	CPL_RDMA_CQE_ERR      = 0x37,
+	CPL_RX_DATA           = 0x39,
+	CPL_SET_TCB_RPL       = 0x3A,
+	CPL_RX_PKT            = 0x3B,
+	CPL_RX_DDP_COMPLETE   = 0x3F,
+
+	CPL_ACT_ESTABLISH     = 0x40,
+	CPL_PASS_ESTABLISH    = 0x41,
+	CPL_RX_DATA_DDP       = 0x42,
+	CPL_PASS_ACCEPT_REQ   = 0x44,
+	CPL_RX_ISCSI_CMP      = 0x45,
+	CPL_TRACE_PKT_T5      = 0x48,
+	CPL_RX_ISCSI_DDP      = 0x49,
+	CPL_RX_TLS_CMP        = 0x4E,
+
+	CPL_RDMA_READ_REQ     = 0x60,
+
+	CPL_PASS_OPEN_REQ6    = 0x81,
+	CPL_ACT_OPEN_REQ6     = 0x83,
+
+	CPL_TX_TLS_PDU        = 0x88,
+	CPL_TX_TLS_SFO        = 0x89,
+	CPL_TX_SEC_PDU        = 0x8A,
+	CPL_TX_TLS_ACK        = 0x8B,
+
+	CPL_RDMA_TERMINATE    = 0xA2,
+	CPL_RDMA_WRITE        = 0xA4,
+	CPL_SGE_EGR_UPDATE    = 0xA5,
+	CPL_RX_MPS_PKT        = 0xAF,
+
+	CPL_TRACE_PKT         = 0xB0,
+	CPL_TLS_DATA          = 0xB1,
+	CPL_ISCSI_DATA	      = 0xB2,
+
+	CPL_FW4_MSG           = 0xC0,
+	CPL_FW4_PLD           = 0xC1,
+	CPL_FW4_ACK           = 0xC3,
+	CPL_SRQ_TABLE_RPL     = 0xCC,
+
+	CPL_RX_PHYS_DSGL      = 0xD0,
+
+	CPL_FW6_MSG           = 0xE0,
+	CPL_FW6_PLD           = 0xE1,
+	CPL_TX_TNL_LSO        = 0xEC,
+	CPL_TX_PKT_LSO        = 0xED,
+	CPL_TX_PKT_XT         = 0xEE,
+
+	NUM_CPL_CMDS
+};
+
+enum CPL_error {
+	CPL_ERR_NONE               = 0,
+	CPL_ERR_TCAM_PARITY        = 1,
+	CPL_ERR_TCAM_MISS          = 2,
+	CPL_ERR_TCAM_FULL          = 3,
+	CPL_ERR_BAD_LENGTH         = 15,
+	CPL_ERR_BAD_ROUTE          = 18,
+	CPL_ERR_CONN_RESET         = 20,
+	CPL_ERR_CONN_EXIST_SYNRECV = 21,
+	CPL_ERR_CONN_EXIST         = 22,
+	CPL_ERR_ARP_MISS           = 23,
+	CPL_ERR_BAD_SYN            = 24,
+	CPL_ERR_CONN_TIMEDOUT      = 30,
+	CPL_ERR_XMIT_TIMEDOUT      = 31,
+	CPL_ERR_PERSIST_TIMEDOUT   = 32,
+	CPL_ERR_FINWAIT2_TIMEDOUT  = 33,
+	CPL_ERR_KEEPALIVE_TIMEDOUT = 34,
+	CPL_ERR_RTX_NEG_ADVICE     = 35,
+	CPL_ERR_PERSIST_NEG_ADVICE = 36,
+	CPL_ERR_KEEPALV_NEG_ADVICE = 37,
+	CPL_ERR_ABORT_FAILED       = 42,
+	CPL_ERR_IWARP_FLM          = 50,
+	CPL_CONTAINS_READ_RPL      = 60,
+	CPL_CONTAINS_WRITE_RPL     = 61,
+};
+
+enum {
+	CPL_CONN_POLICY_AUTO = 0,
+	CPL_CONN_POLICY_ASK  = 1,
+	CPL_CONN_POLICY_FILTER = 2,
+	CPL_CONN_POLICY_DENY = 3
+};
+
+enum {
+	ULP_MODE_NONE          = 0,
+	ULP_MODE_ISCSI         = 2,
+	ULP_MODE_RDMA          = 4,
+	ULP_MODE_TCPDDP	       = 5,
+	ULP_MODE_FCOE          = 6,
+	ULP_MODE_TLS           = 8,
+};
+
+enum {
+	ULP_CRC_HEADER = 1 << 0,
+	ULP_CRC_DATA   = 1 << 1
+};
+
+enum {
+	CPL_ABORT_SEND_RST = 0,
+	CPL_ABORT_NO_RST,
+};
+
+enum {                     /* TX_PKT_XT checksum types */
+	TX_CSUM_TCP    = 0,
+	TX_CSUM_UDP    = 1,
+	TX_CSUM_CRC16  = 4,
+	TX_CSUM_CRC32  = 5,
+	TX_CSUM_CRC32C = 6,
+	TX_CSUM_FCOE   = 7,
+	TX_CSUM_TCPIP  = 8,
+	TX_CSUM_UDPIP  = 9,
+	TX_CSUM_TCPIP6 = 10,
+	TX_CSUM_UDPIP6 = 11,
+	TX_CSUM_IP     = 12,
+};
+
+union opcode_tid {
+	__be32 opcode_tid;
+	u8 opcode;
+};
+
+#define CPL_OPCODE_S    24
+#define CPL_OPCODE_V(x) ((x) << CPL_OPCODE_S)
+#define CPL_OPCODE_G(x) (((x) >> CPL_OPCODE_S) & 0xFF)
+#define TID_G(x)    ((x) & 0xFFFFFF)
+
+/* tid is assumed to be 24-bits */
+#define MK_OPCODE_TID(opcode, tid) (CPL_OPCODE_V(opcode) | (tid))
+
+#define OPCODE_TID(cmd) ((cmd)->ot.opcode_tid)
+
+/* extract the TID from a CPL command */
+#define GET_TID(cmd) (TID_G(be32_to_cpu(OPCODE_TID(cmd))))
+
+/* partitioning of TID fields that also carry a queue id */
+#define TID_TID_S    0
+#define TID_TID_M    0x3fff
+#define TID_TID_V(x) ((x) << TID_TID_S)
+#define TID_TID_G(x) (((x) >> TID_TID_S) & TID_TID_M)
+
+#define TID_QID_S    14
+#define TID_QID_M    0x3ff
+#define TID_QID_V(x) ((x) << TID_QID_S)
+#define TID_QID_G(x) (((x) >> TID_QID_S) & TID_QID_M)
+
+struct rss_header {
+	u8 opcode;
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	u8 channel:2;
+	u8 filter_hit:1;
+	u8 filter_tid:1;
+	u8 hash_type:2;
+	u8 ipv6:1;
+	u8 send2fw:1;
+#else
+	u8 send2fw:1;
+	u8 ipv6:1;
+	u8 hash_type:2;
+	u8 filter_tid:1;
+	u8 filter_hit:1;
+	u8 channel:2;
+#endif
+	__be16 qid;
+	__be32 hash_val;
+};
+
+struct work_request_hdr {
+	__be32 wr_hi;
+	__be32 wr_mid;
+	__be64 wr_lo;
+};
+
+/* wr_hi fields */
+#define WR_OP_S    24
+#define WR_OP_V(x) ((__u64)(x) << WR_OP_S)
+
+#define WR_HDR struct work_request_hdr wr
+
+/* option 0 fields */
+#define TX_CHAN_S    2
+#define TX_CHAN_V(x) ((x) << TX_CHAN_S)
+
+#define ULP_MODE_S    8
+#define ULP_MODE_V(x) ((x) << ULP_MODE_S)
+
+#define RCV_BUFSIZ_S    12
+#define RCV_BUFSIZ_M    0x3FFU
+#define RCV_BUFSIZ_V(x) ((x) << RCV_BUFSIZ_S)
+
+#define SMAC_SEL_S    28
+#define SMAC_SEL_V(x) ((__u64)(x) << SMAC_SEL_S)
+
+#define L2T_IDX_S    36
+#define L2T_IDX_V(x) ((__u64)(x) << L2T_IDX_S)
+
+#define WND_SCALE_S    50
+#define WND_SCALE_V(x) ((__u64)(x) << WND_SCALE_S)
+
+#define KEEP_ALIVE_S    54
+#define KEEP_ALIVE_V(x) ((__u64)(x) << KEEP_ALIVE_S)
+#define KEEP_ALIVE_F    KEEP_ALIVE_V(1ULL)
+
+#define MSS_IDX_S    60
+#define MSS_IDX_M    0xF
+#define MSS_IDX_V(x) ((__u64)(x) << MSS_IDX_S)
+#define MSS_IDX_G(x) (((x) >> MSS_IDX_S) & MSS_IDX_M)
+
+/* option 2 fields */
+#define RSS_QUEUE_S    0
+#define RSS_QUEUE_M    0x3FF
+#define RSS_QUEUE_V(x) ((x) << RSS_QUEUE_S)
+#define RSS_QUEUE_G(x) (((x) >> RSS_QUEUE_S) & RSS_QUEUE_M)
+
+#define RSS_QUEUE_VALID_S    10
+#define RSS_QUEUE_VALID_V(x) ((x) << RSS_QUEUE_VALID_S)
+#define RSS_QUEUE_VALID_F    RSS_QUEUE_VALID_V(1U)
+
+#define RX_FC_DISABLE_S    20
+#define RX_FC_DISABLE_V(x) ((x) << RX_FC_DISABLE_S)
+#define RX_FC_DISABLE_F    RX_FC_DISABLE_V(1U)
+
+#define RX_FC_VALID_S    22
+#define RX_FC_VALID_V(x) ((x) << RX_FC_VALID_S)
+#define RX_FC_VALID_F    RX_FC_VALID_V(1U)
+
+#define RX_CHANNEL_S    26
+#define RX_CHANNEL_V(x) ((x) << RX_CHANNEL_S)
+#define RX_CHANNEL_F	RX_CHANNEL_V(1U)
+
+#define WND_SCALE_EN_S    28
+#define WND_SCALE_EN_V(x) ((x) << WND_SCALE_EN_S)
+#define WND_SCALE_EN_F    WND_SCALE_EN_V(1U)
+
+#define T5_OPT_2_VALID_S    31
+#define T5_OPT_2_VALID_V(x) ((x) << T5_OPT_2_VALID_S)
+#define T5_OPT_2_VALID_F    T5_OPT_2_VALID_V(1U)
+
+struct cpl_pass_open_req {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 local_port;
+	__be16 peer_port;
+	__be32 local_ip;
+	__be32 peer_ip;
+	__be64 opt0;
+	__be64 opt1;
+};
+
+/* option 0 fields */
+#define NO_CONG_S    4
+#define NO_CONG_V(x) ((x) << NO_CONG_S)
+#define NO_CONG_F    NO_CONG_V(1U)
+
+#define DELACK_S    5
+#define DELACK_V(x) ((x) << DELACK_S)
+#define DELACK_F    DELACK_V(1U)
+
+#define NON_OFFLOAD_S		7
+#define NON_OFFLOAD_V(x)	((x) << NON_OFFLOAD_S)
+#define NON_OFFLOAD_F		NON_OFFLOAD_V(1U)
+
+#define DSCP_S    22
+#define DSCP_M    0x3F
+#define DSCP_V(x) ((x) << DSCP_S)
+#define DSCP_G(x) (((x) >> DSCP_S) & DSCP_M)
+
+#define TCAM_BYPASS_S    48
+#define TCAM_BYPASS_V(x) ((__u64)(x) << TCAM_BYPASS_S)
+#define TCAM_BYPASS_F    TCAM_BYPASS_V(1ULL)
+
+#define NAGLE_S    49
+#define NAGLE_V(x) ((__u64)(x) << NAGLE_S)
+#define NAGLE_F    NAGLE_V(1ULL)
+
+/* option 1 fields */
+#define SYN_RSS_ENABLE_S    0
+#define SYN_RSS_ENABLE_V(x) ((x) << SYN_RSS_ENABLE_S)
+#define SYN_RSS_ENABLE_F    SYN_RSS_ENABLE_V(1U)
+
+#define SYN_RSS_QUEUE_S    2
+#define SYN_RSS_QUEUE_V(x) ((x) << SYN_RSS_QUEUE_S)
+
+#define CONN_POLICY_S    22
+#define CONN_POLICY_V(x) ((x) << CONN_POLICY_S)
+
+struct cpl_pass_open_req6 {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 local_port;
+	__be16 peer_port;
+	__be64 local_ip_hi;
+	__be64 local_ip_lo;
+	__be64 peer_ip_hi;
+	__be64 peer_ip_lo;
+	__be64 opt0;
+	__be64 opt1;
+};
+
+struct cpl_pass_open_rpl {
+	union opcode_tid ot;
+	u8 rsvd[3];
+	u8 status;
+};
+
+struct tcp_options {
+	__be16 mss;
+	__u8 wsf;
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	__u8:4;
+	__u8 unknown:1;
+	__u8:1;
+	__u8 sack:1;
+	__u8 tstamp:1;
+#else
+	__u8 tstamp:1;
+	__u8 sack:1;
+	__u8:1;
+	__u8 unknown:1;
+	__u8:4;
+#endif
+};
+
+struct cpl_pass_accept_req {
+	union opcode_tid ot;
+	__be16 rsvd;
+	__be16 len;
+	__be32 hdr_len;
+	__be16 vlan;
+	__be16 l2info;
+	__be32 tos_stid;
+	struct tcp_options tcpopt;
+};
+
+/* cpl_pass_accept_req.hdr_len fields */
+#define SYN_RX_CHAN_S    0
+#define SYN_RX_CHAN_M    0xF
+#define SYN_RX_CHAN_V(x) ((x) << SYN_RX_CHAN_S)
+#define SYN_RX_CHAN_G(x) (((x) >> SYN_RX_CHAN_S) & SYN_RX_CHAN_M)
+
+#define TCP_HDR_LEN_S    10
+#define TCP_HDR_LEN_M    0x3F
+#define TCP_HDR_LEN_V(x) ((x) << TCP_HDR_LEN_S)
+#define TCP_HDR_LEN_G(x) (((x) >> TCP_HDR_LEN_S) & TCP_HDR_LEN_M)
+
+#define IP_HDR_LEN_S    16
+#define IP_HDR_LEN_M    0x3FF
+#define IP_HDR_LEN_V(x) ((x) << IP_HDR_LEN_S)
+#define IP_HDR_LEN_G(x) (((x) >> IP_HDR_LEN_S) & IP_HDR_LEN_M)
+
+#define ETH_HDR_LEN_S    26
+#define ETH_HDR_LEN_M    0x1F
+#define ETH_HDR_LEN_V(x) ((x) << ETH_HDR_LEN_S)
+#define ETH_HDR_LEN_G(x) (((x) >> ETH_HDR_LEN_S) & ETH_HDR_LEN_M)
+
+/* cpl_pass_accept_req.l2info fields */
+#define SYN_MAC_IDX_S    0
+#define SYN_MAC_IDX_M    0x1FF
+#define SYN_MAC_IDX_V(x) ((x) << SYN_MAC_IDX_S)
+#define SYN_MAC_IDX_G(x) (((x) >> SYN_MAC_IDX_S) & SYN_MAC_IDX_M)
+
+#define SYN_XACT_MATCH_S    9
+#define SYN_XACT_MATCH_V(x) ((x) << SYN_XACT_MATCH_S)
+#define SYN_XACT_MATCH_F    SYN_XACT_MATCH_V(1U)
+
+#define SYN_INTF_S    12
+#define SYN_INTF_M    0xF
+#define SYN_INTF_V(x) ((x) << SYN_INTF_S)
+#define SYN_INTF_G(x) (((x) >> SYN_INTF_S) & SYN_INTF_M)
+
+enum {                     /* TCP congestion control algorithms */
+	CONG_ALG_RENO,
+	CONG_ALG_TAHOE,
+	CONG_ALG_NEWRENO,
+	CONG_ALG_HIGHSPEED
+};
+
+#define CONG_CNTRL_S    14
+#define CONG_CNTRL_M    0x3
+#define CONG_CNTRL_V(x) ((x) << CONG_CNTRL_S)
+#define CONG_CNTRL_G(x) (((x) >> CONG_CNTRL_S) & CONG_CNTRL_M)
+
+#define T5_ISS_S    18
+#define T5_ISS_V(x) ((x) << T5_ISS_S)
+#define T5_ISS_F    T5_ISS_V(1U)
+
+struct cpl_pass_accept_rpl {
+	WR_HDR;
+	union opcode_tid ot;
+	__be32 opt2;
+	__be64 opt0;
+};
+
+/* option 2 fields */
+#define RX_COALESCE_VALID_S    11
+#define RX_COALESCE_VALID_V(x) ((x) << RX_COALESCE_VALID_S)
+#define RX_COALESCE_VALID_F    RX_COALESCE_VALID_V(1U)
+
+#define RX_COALESCE_S    12
+#define RX_COALESCE_V(x) ((x) << RX_COALESCE_S)
+
+#define PACE_S    16
+#define PACE_V(x) ((x) << PACE_S)
+
+#define TX_QUEUE_S    23
+#define TX_QUEUE_M    0x7
+#define TX_QUEUE_V(x) ((x) << TX_QUEUE_S)
+#define TX_QUEUE_G(x) (((x) >> TX_QUEUE_S) & TX_QUEUE_M)
+
+#define CCTRL_ECN_S    27
+#define CCTRL_ECN_V(x) ((x) << CCTRL_ECN_S)
+#define CCTRL_ECN_F    CCTRL_ECN_V(1U)
+
+#define TSTAMPS_EN_S    29
+#define TSTAMPS_EN_V(x) ((x) << TSTAMPS_EN_S)
+#define TSTAMPS_EN_F    TSTAMPS_EN_V(1U)
+
+#define SACK_EN_S    30
+#define SACK_EN_V(x) ((x) << SACK_EN_S)
+#define SACK_EN_F    SACK_EN_V(1U)
+
+struct cpl_t5_pass_accept_rpl {
+	WR_HDR;
+	union opcode_tid ot;
+	__be32 opt2;
+	__be64 opt0;
+	__be32 iss;
+	__be32 rsvd;
+};
+
+struct cpl_act_open_req {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 local_port;
+	__be16 peer_port;
+	__be32 local_ip;
+	__be32 peer_ip;
+	__be64 opt0;
+	__be32 params;
+	__be32 opt2;
+};
+
+#define FILTER_TUPLE_S  24
+#define FILTER_TUPLE_M  0xFFFFFFFFFF
+#define FILTER_TUPLE_V(x) ((x) << FILTER_TUPLE_S)
+#define FILTER_TUPLE_G(x) (((x) >> FILTER_TUPLE_S) & FILTER_TUPLE_M)
+struct cpl_t5_act_open_req {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 local_port;
+	__be16 peer_port;
+	__be32 local_ip;
+	__be32 peer_ip;
+	__be64 opt0;
+	__be32 rsvd;
+	__be32 opt2;
+	__be64 params;
+};
+
+struct cpl_t6_act_open_req {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 local_port;
+	__be16 peer_port;
+	__be32 local_ip;
+	__be32 peer_ip;
+	__be64 opt0;
+	__be32 rsvd;
+	__be32 opt2;
+	__be64 params;
+	__be32 rsvd2;
+	__be32 opt3;
+};
+
+struct cpl_act_open_req6 {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 local_port;
+	__be16 peer_port;
+	__be64 local_ip_hi;
+	__be64 local_ip_lo;
+	__be64 peer_ip_hi;
+	__be64 peer_ip_lo;
+	__be64 opt0;
+	__be32 params;
+	__be32 opt2;
+};
+
+struct cpl_t5_act_open_req6 {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 local_port;
+	__be16 peer_port;
+	__be64 local_ip_hi;
+	__be64 local_ip_lo;
+	__be64 peer_ip_hi;
+	__be64 peer_ip_lo;
+	__be64 opt0;
+	__be32 rsvd;
+	__be32 opt2;
+	__be64 params;
+};
+
+struct cpl_t6_act_open_req6 {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 local_port;
+	__be16 peer_port;
+	__be64 local_ip_hi;
+	__be64 local_ip_lo;
+	__be64 peer_ip_hi;
+	__be64 peer_ip_lo;
+	__be64 opt0;
+	__be32 rsvd;
+	__be32 opt2;
+	__be64 params;
+	__be32 rsvd2;
+	__be32 opt3;
+};
+
+struct cpl_act_open_rpl {
+	union opcode_tid ot;
+	__be32 atid_status;
+};
+
+/* cpl_act_open_rpl.atid_status fields */
+#define AOPEN_STATUS_S    0
+#define AOPEN_STATUS_M    0xFF
+#define AOPEN_STATUS_G(x) (((x) >> AOPEN_STATUS_S) & AOPEN_STATUS_M)
+
+#define AOPEN_ATID_S    8
+#define AOPEN_ATID_M    0xFFFFFF
+#define AOPEN_ATID_G(x) (((x) >> AOPEN_ATID_S) & AOPEN_ATID_M)
+
+struct cpl_pass_establish {
+	union opcode_tid ot;
+	__be32 rsvd;
+	__be32 tos_stid;
+	__be16 mac_idx;
+	__be16 tcp_opt;
+	__be32 snd_isn;
+	__be32 rcv_isn;
+};
+
+/* cpl_pass_establish.tos_stid fields */
+#define PASS_OPEN_TID_S    0
+#define PASS_OPEN_TID_M    0xFFFFFF
+#define PASS_OPEN_TID_V(x) ((x) << PASS_OPEN_TID_S)
+#define PASS_OPEN_TID_G(x) (((x) >> PASS_OPEN_TID_S) & PASS_OPEN_TID_M)
+
+#define PASS_OPEN_TOS_S    24
+#define PASS_OPEN_TOS_M    0xFF
+#define PASS_OPEN_TOS_V(x) ((x) << PASS_OPEN_TOS_S)
+#define PASS_OPEN_TOS_G(x) (((x) >> PASS_OPEN_TOS_S) & PASS_OPEN_TOS_M)
+
+/* cpl_pass_establish.tcp_opt fields (also applies to act_open_establish) */
+#define TCPOPT_WSCALE_OK_S	5
+#define TCPOPT_WSCALE_OK_M	0x1
+#define TCPOPT_WSCALE_OK_G(x)	\
+	(((x) >> TCPOPT_WSCALE_OK_S) & TCPOPT_WSCALE_OK_M)
+
+#define TCPOPT_SACK_S		6
+#define TCPOPT_SACK_M		0x1
+#define TCPOPT_SACK_G(x)	(((x) >> TCPOPT_SACK_S) & TCPOPT_SACK_M)
+
+#define TCPOPT_TSTAMP_S		7
+#define TCPOPT_TSTAMP_M		0x1
+#define TCPOPT_TSTAMP_G(x)	(((x) >> TCPOPT_TSTAMP_S) & TCPOPT_TSTAMP_M)
+
+#define TCPOPT_SND_WSCALE_S	8
+#define TCPOPT_SND_WSCALE_M	0xF
+#define TCPOPT_SND_WSCALE_G(x)	\
+	(((x) >> TCPOPT_SND_WSCALE_S) & TCPOPT_SND_WSCALE_M)
+
+#define TCPOPT_MSS_S	12
+#define TCPOPT_MSS_M	0xF
+#define TCPOPT_MSS_G(x)	(((x) >> TCPOPT_MSS_S) & TCPOPT_MSS_M)
+
+#define T6_TCP_HDR_LEN_S   8
+#define T6_TCP_HDR_LEN_V(x) ((x) << T6_TCP_HDR_LEN_S)
+#define T6_TCP_HDR_LEN_G(x) (((x) >> T6_TCP_HDR_LEN_S) & TCP_HDR_LEN_M)
+
+#define T6_IP_HDR_LEN_S    14
+#define T6_IP_HDR_LEN_V(x) ((x) << T6_IP_HDR_LEN_S)
+#define T6_IP_HDR_LEN_G(x) (((x) >> T6_IP_HDR_LEN_S) & IP_HDR_LEN_M)
+
+#define T6_ETH_HDR_LEN_S    24
+#define T6_ETH_HDR_LEN_M    0xFF
+#define T6_ETH_HDR_LEN_V(x) ((x) << T6_ETH_HDR_LEN_S)
+#define T6_ETH_HDR_LEN_G(x) (((x) >> T6_ETH_HDR_LEN_S) & T6_ETH_HDR_LEN_M)
+
+struct cpl_act_establish {
+	union opcode_tid ot;
+	__be32 rsvd;
+	__be32 tos_atid;
+	__be16 mac_idx;
+	__be16 tcp_opt;
+	__be32 snd_isn;
+	__be32 rcv_isn;
+};
+
+struct cpl_get_tcb {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 reply_ctrl;
+	__be16 cookie;
+};
+
+/* cpl_get_tcb.reply_ctrl fields */
+#define QUEUENO_S    0
+#define QUEUENO_V(x) ((x) << QUEUENO_S)
+
+#define REPLY_CHAN_S    14
+#define REPLY_CHAN_V(x) ((x) << REPLY_CHAN_S)
+#define REPLY_CHAN_F    REPLY_CHAN_V(1U)
+
+#define NO_REPLY_S    15
+#define NO_REPLY_V(x) ((x) << NO_REPLY_S)
+#define NO_REPLY_F    NO_REPLY_V(1U)
+
+struct cpl_get_tcb_rpl {
+	union opcode_tid ot;
+	__u8 cookie;
+	__u8 status;
+	__be16 len;
+};
+
+struct cpl_set_tcb_field {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 reply_ctrl;
+	__be16 word_cookie;
+	__be64 mask;
+	__be64 val;
+};
+
+struct cpl_set_tcb_field_core {
+	union opcode_tid ot;
+	__be16 reply_ctrl;
+	__be16 word_cookie;
+	__be64 mask;
+	__be64 val;
+};
+
+/* cpl_set_tcb_field.word_cookie fields */
+#define TCB_WORD_S	0
+#define TCB_WORD_V(x)	((x) << TCB_WORD_S)
+
+#define TCB_COOKIE_S    5
+#define TCB_COOKIE_M    0x7
+#define TCB_COOKIE_V(x) ((x) << TCB_COOKIE_S)
+#define TCB_COOKIE_G(x) (((x) >> TCB_COOKIE_S) & TCB_COOKIE_M)
+
+struct cpl_set_tcb_rpl {
+	union opcode_tid ot;
+	__be16 rsvd;
+	u8 cookie;
+	u8 status;
+	__be64 oldval;
+};
+
+struct cpl_close_con_req {
+	WR_HDR;
+	union opcode_tid ot;
+	__be32 rsvd;
+};
+
+struct cpl_close_con_rpl {
+	union opcode_tid ot;
+	u8 rsvd[3];
+	u8 status;
+	__be32 snd_nxt;
+	__be32 rcv_nxt;
+};
+
+struct cpl_close_listsvr_req {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 reply_ctrl;
+	__be16 rsvd;
+};
+
+/* additional cpl_close_listsvr_req.reply_ctrl field */
+#define LISTSVR_IPV6_S    14
+#define LISTSVR_IPV6_V(x) ((x) << LISTSVR_IPV6_S)
+#define LISTSVR_IPV6_F    LISTSVR_IPV6_V(1U)
+
+struct cpl_close_listsvr_rpl {
+	union opcode_tid ot;
+	u8 rsvd[3];
+	u8 status;
+};
+
+struct cpl_abort_req_rss {
+	union opcode_tid ot;
+	u8 rsvd[3];
+	u8 status;
+};
+
+struct cpl_abort_req_rss6 {
+	union opcode_tid ot;
+	__be32 srqidx_status;
+};
+
+#define ABORT_RSS_STATUS_S    0
+#define ABORT_RSS_STATUS_M    0xff
+#define ABORT_RSS_STATUS_V(x) ((x) << ABORT_RSS_STATUS_S)
+#define ABORT_RSS_STATUS_G(x) (((x) >> ABORT_RSS_STATUS_S) & ABORT_RSS_STATUS_M)
+
+#define ABORT_RSS_SRQIDX_S    8
+#define ABORT_RSS_SRQIDX_M    0xffffff
+#define ABORT_RSS_SRQIDX_V(x) ((x) << ABORT_RSS_SRQIDX_S)
+#define ABORT_RSS_SRQIDX_G(x) (((x) >> ABORT_RSS_SRQIDX_S) & ABORT_RSS_SRQIDX_M)
+
+struct cpl_abort_req {
+	WR_HDR;
+	union opcode_tid ot;
+	__be32 rsvd0;
+	u8 rsvd1;
+	u8 cmd;
+	u8 rsvd2[6];
+};
+
+struct cpl_abort_rpl_rss {
+	union opcode_tid ot;
+	u8 rsvd[3];
+	u8 status;
+};
+
+struct cpl_abort_rpl_rss6 {
+	union opcode_tid ot;
+	__be32 srqidx_status;
+};
+
+struct cpl_abort_rpl {
+	WR_HDR;
+	union opcode_tid ot;
+	__be32 rsvd0;
+	u8 rsvd1;
+	u8 cmd;
+	u8 rsvd2[6];
+};
+
+struct cpl_peer_close {
+	union opcode_tid ot;
+	__be32 rcv_nxt;
+};
+
+struct cpl_tid_release {
+	WR_HDR;
+	union opcode_tid ot;
+	__be32 rsvd;
+};
+
+struct cpl_tx_pkt_core {
+	__be32 ctrl0;
+	__be16 pack;
+	__be16 len;
+	__be64 ctrl1;
+};
+
+struct cpl_tx_pkt {
+	WR_HDR;
+	struct cpl_tx_pkt_core c;
+};
+
+#define cpl_tx_pkt_xt cpl_tx_pkt
+
+/* cpl_tx_pkt_core.ctrl0 fields */
+#define TXPKT_VF_S    0
+#define TXPKT_VF_V(x) ((x) << TXPKT_VF_S)
+
+#define TXPKT_PF_S    8
+#define TXPKT_PF_V(x) ((x) << TXPKT_PF_S)
+
+#define TXPKT_VF_VLD_S    11
+#define TXPKT_VF_VLD_V(x) ((x) << TXPKT_VF_VLD_S)
+#define TXPKT_VF_VLD_F    TXPKT_VF_VLD_V(1U)
+
+#define TXPKT_OVLAN_IDX_S    12
+#define TXPKT_OVLAN_IDX_V(x) ((x) << TXPKT_OVLAN_IDX_S)
+
+#define TXPKT_T5_OVLAN_IDX_S	12
+#define TXPKT_T5_OVLAN_IDX_V(x)	((x) << TXPKT_T5_OVLAN_IDX_S)
+
+#define TXPKT_INTF_S    16
+#define TXPKT_INTF_V(x) ((x) << TXPKT_INTF_S)
+
+#define TXPKT_INS_OVLAN_S    21
+#define TXPKT_INS_OVLAN_V(x) ((x) << TXPKT_INS_OVLAN_S)
+#define TXPKT_INS_OVLAN_F    TXPKT_INS_OVLAN_V(1U)
+
+#define TXPKT_TSTAMP_S    23
+#define TXPKT_TSTAMP_V(x) ((x) << TXPKT_TSTAMP_S)
+#define TXPKT_TSTAMP_F    TXPKT_TSTAMP_V(1ULL)
+
+#define TXPKT_OPCODE_S    24
+#define TXPKT_OPCODE_V(x) ((x) << TXPKT_OPCODE_S)
+
+/* cpl_tx_pkt_core.ctrl1 fields */
+#define TXPKT_CSUM_END_S    12
+#define TXPKT_CSUM_END_V(x) ((x) << TXPKT_CSUM_END_S)
+
+#define TXPKT_CSUM_START_S    20
+#define TXPKT_CSUM_START_V(x) ((x) << TXPKT_CSUM_START_S)
+
+#define TXPKT_IPHDR_LEN_S    20
+#define TXPKT_IPHDR_LEN_V(x) ((__u64)(x) << TXPKT_IPHDR_LEN_S)
+
+#define TXPKT_CSUM_LOC_S    30
+#define TXPKT_CSUM_LOC_V(x) ((__u64)(x) << TXPKT_CSUM_LOC_S)
+
+#define TXPKT_ETHHDR_LEN_S    34
+#define TXPKT_ETHHDR_LEN_V(x) ((__u64)(x) << TXPKT_ETHHDR_LEN_S)
+
+#define T6_TXPKT_ETHHDR_LEN_S    32
+#define T6_TXPKT_ETHHDR_LEN_V(x) ((__u64)(x) << T6_TXPKT_ETHHDR_LEN_S)
+
+#define TXPKT_CSUM_TYPE_S    40
+#define TXPKT_CSUM_TYPE_V(x) ((__u64)(x) << TXPKT_CSUM_TYPE_S)
+
+#define TXPKT_VLAN_S    44
+#define TXPKT_VLAN_V(x) ((__u64)(x) << TXPKT_VLAN_S)
+
+#define TXPKT_VLAN_VLD_S    60
+#define TXPKT_VLAN_VLD_V(x) ((__u64)(x) << TXPKT_VLAN_VLD_S)
+#define TXPKT_VLAN_VLD_F    TXPKT_VLAN_VLD_V(1ULL)
+
+#define TXPKT_IPCSUM_DIS_S    62
+#define TXPKT_IPCSUM_DIS_V(x) ((__u64)(x) << TXPKT_IPCSUM_DIS_S)
+#define TXPKT_IPCSUM_DIS_F    TXPKT_IPCSUM_DIS_V(1ULL)
+
+#define TXPKT_L4CSUM_DIS_S    63
+#define TXPKT_L4CSUM_DIS_V(x) ((__u64)(x) << TXPKT_L4CSUM_DIS_S)
+#define TXPKT_L4CSUM_DIS_F    TXPKT_L4CSUM_DIS_V(1ULL)
+
+struct cpl_tx_pkt_lso_core {
+	__be32 lso_ctrl;
+	__be16 ipid_ofst;
+	__be16 mss;
+	__be32 seqno_offset;
+	__be32 len;
+	/* encapsulated CPL (TX_PKT, TX_PKT_XT or TX_DATA) follows here */
+};
+
+/* cpl_tx_pkt_lso_core.lso_ctrl fields */
+#define LSO_TCPHDR_LEN_S    0
+#define LSO_TCPHDR_LEN_V(x) ((x) << LSO_TCPHDR_LEN_S)
+
+#define LSO_IPHDR_LEN_S    4
+#define LSO_IPHDR_LEN_V(x) ((x) << LSO_IPHDR_LEN_S)
+
+#define LSO_ETHHDR_LEN_S    16
+#define LSO_ETHHDR_LEN_V(x) ((x) << LSO_ETHHDR_LEN_S)
+
+#define LSO_IPV6_S    20
+#define LSO_IPV6_V(x) ((x) << LSO_IPV6_S)
+#define LSO_IPV6_F    LSO_IPV6_V(1U)
+
+#define LSO_LAST_SLICE_S    22
+#define LSO_LAST_SLICE_V(x) ((x) << LSO_LAST_SLICE_S)
+#define LSO_LAST_SLICE_F    LSO_LAST_SLICE_V(1U)
+
+#define LSO_FIRST_SLICE_S    23
+#define LSO_FIRST_SLICE_V(x) ((x) << LSO_FIRST_SLICE_S)
+#define LSO_FIRST_SLICE_F    LSO_FIRST_SLICE_V(1U)
+
+#define LSO_OPCODE_S    24
+#define LSO_OPCODE_V(x) ((x) << LSO_OPCODE_S)
+
+#define LSO_T5_XFER_SIZE_S	   0
+#define LSO_T5_XFER_SIZE_V(x) ((x) << LSO_T5_XFER_SIZE_S)
+
+struct cpl_tx_pkt_lso {
+	WR_HDR;
+	struct cpl_tx_pkt_lso_core c;
+	/* encapsulated CPL (TX_PKT, TX_PKT_XT or TX_DATA) follows here */
+};
+
+struct cpl_iscsi_hdr {
+	union opcode_tid ot;
+	__be16 pdu_len_ddp;
+	__be16 len;
+	__be32 seq;
+	__be16 urg;
+	u8 rsvd;
+	u8 status;
+};
+
+/* cpl_iscsi_hdr.pdu_len_ddp fields */
+#define ISCSI_PDU_LEN_S    0
+#define ISCSI_PDU_LEN_M    0x7FFF
+#define ISCSI_PDU_LEN_V(x) ((x) << ISCSI_PDU_LEN_S)
+#define ISCSI_PDU_LEN_G(x) (((x) >> ISCSI_PDU_LEN_S) & ISCSI_PDU_LEN_M)
+
+#define ISCSI_DDP_S    15
+#define ISCSI_DDP_V(x) ((x) << ISCSI_DDP_S)
+#define ISCSI_DDP_F    ISCSI_DDP_V(1U)
+
+struct cpl_rx_data_ddp {
+	union opcode_tid ot;
+	__be16 urg;
+	__be16 len;
+	__be32 seq;
+	union {
+		__be32 nxt_seq;
+		__be32 ddp_report;
+	};
+	__be32 ulp_crc;
+	__be32 ddpvld;
+};
+
+#define cpl_rx_iscsi_ddp cpl_rx_data_ddp
+
+struct cpl_iscsi_data {
+	union opcode_tid ot;
+	__u8 rsvd0[2];
+	__be16 len;
+	__be32 seq;
+	__be16 urg;
+	__u8 rsvd1;
+	__u8 status;
+};
+
+struct cpl_rx_iscsi_cmp {
+	union opcode_tid ot;
+	__be16 pdu_len_ddp;
+	__be16 len;
+	__be32 seq;
+	__be16 urg;
+	__u8 rsvd;
+	__u8 status;
+	__be32 ulp_crc;
+	__be32 ddpvld;
+};
+
+struct cpl_tx_data_iso {
+	__be32 op_to_scsi;
+	__u8   reserved1;
+	__u8   ahs_len;
+	__be16 mpdu;
+	__be32 burst_size;
+	__be32 len;
+	__be32 reserved2_seglen_offset;
+	__be32 datasn_offset;
+	__be32 buffer_offset;
+	__be32 reserved3;
+
+	/* encapsulated CPL_TX_DATA follows here */
+};
+
+/* cpl_tx_data_iso.op_to_scsi fields */
+#define CPL_TX_DATA_ISO_OP_S	24
+#define CPL_TX_DATA_ISO_OP_M	0xff
+#define CPL_TX_DATA_ISO_OP_V(x)	((x) << CPL_TX_DATA_ISO_OP_S)
+#define CPL_TX_DATA_ISO_OP_G(x)	\
+	(((x) >> CPL_TX_DATA_ISO_OP_S) & CPL_TX_DATA_ISO_OP_M)
+
+#define CPL_TX_DATA_ISO_FIRST_S		23
+#define CPL_TX_DATA_ISO_FIRST_M		0x1
+#define CPL_TX_DATA_ISO_FIRST_V(x)	((x) << CPL_TX_DATA_ISO_FIRST_S)
+#define CPL_TX_DATA_ISO_FIRST_G(x)	\
+	(((x) >> CPL_TX_DATA_ISO_FIRST_S) & CPL_TX_DATA_ISO_FIRST_M)
+#define CPL_TX_DATA_ISO_FIRST_F	CPL_TX_DATA_ISO_FIRST_V(1U)
+
+#define CPL_TX_DATA_ISO_LAST_S		22
+#define CPL_TX_DATA_ISO_LAST_M		0x1
+#define CPL_TX_DATA_ISO_LAST_V(x)	((x) << CPL_TX_DATA_ISO_LAST_S)
+#define CPL_TX_DATA_ISO_LAST_G(x)	\
+	(((x) >> CPL_TX_DATA_ISO_LAST_S) & CPL_TX_DATA_ISO_LAST_M)
+#define CPL_TX_DATA_ISO_LAST_F	CPL_TX_DATA_ISO_LAST_V(1U)
+
+#define CPL_TX_DATA_ISO_CPLHDRLEN_S	21
+#define CPL_TX_DATA_ISO_CPLHDRLEN_M	0x1
+#define CPL_TX_DATA_ISO_CPLHDRLEN_V(x)	((x) << CPL_TX_DATA_ISO_CPLHDRLEN_S)
+#define CPL_TX_DATA_ISO_CPLHDRLEN_G(x)	\
+	(((x) >> CPL_TX_DATA_ISO_CPLHDRLEN_S) & CPL_TX_DATA_ISO_CPLHDRLEN_M)
+#define CPL_TX_DATA_ISO_CPLHDRLEN_F	CPL_TX_DATA_ISO_CPLHDRLEN_V(1U)
+
+#define CPL_TX_DATA_ISO_HDRCRC_S	20
+#define CPL_TX_DATA_ISO_HDRCRC_M	0x1
+#define CPL_TX_DATA_ISO_HDRCRC_V(x)	((x) << CPL_TX_DATA_ISO_HDRCRC_S)
+#define CPL_TX_DATA_ISO_HDRCRC_G(x)	\
+	(((x) >> CPL_TX_DATA_ISO_HDRCRC_S) & CPL_TX_DATA_ISO_HDRCRC_M)
+#define CPL_TX_DATA_ISO_HDRCRC_F	CPL_TX_DATA_ISO_HDRCRC_V(1U)
+
+#define CPL_TX_DATA_ISO_PLDCRC_S	19
+#define CPL_TX_DATA_ISO_PLDCRC_M	0x1
+#define CPL_TX_DATA_ISO_PLDCRC_V(x)	((x) << CPL_TX_DATA_ISO_PLDCRC_S)
+#define CPL_TX_DATA_ISO_PLDCRC_G(x)	\
+	(((x) >> CPL_TX_DATA_ISO_PLDCRC_S) & CPL_TX_DATA_ISO_PLDCRC_M)
+#define CPL_TX_DATA_ISO_PLDCRC_F	CPL_TX_DATA_ISO_PLDCRC_V(1U)
+
+#define CPL_TX_DATA_ISO_IMMEDIATE_S	18
+#define CPL_TX_DATA_ISO_IMMEDIATE_M	0x1
+#define CPL_TX_DATA_ISO_IMMEDIATE_V(x)	((x) << CPL_TX_DATA_ISO_IMMEDIATE_S)
+#define CPL_TX_DATA_ISO_IMMEDIATE_G(x)	\
+	(((x) >> CPL_TX_DATA_ISO_IMMEDIATE_S) & CPL_TX_DATA_ISO_IMMEDIATE_M)
+#define CPL_TX_DATA_ISO_IMMEDIATE_F	CPL_TX_DATA_ISO_IMMEDIATE_V(1U)
+
+#define CPL_TX_DATA_ISO_SCSI_S		16
+#define CPL_TX_DATA_ISO_SCSI_M		0x3
+#define CPL_TX_DATA_ISO_SCSI_V(x)	((x) << CPL_TX_DATA_ISO_SCSI_S)
+#define CPL_TX_DATA_ISO_SCSI_G(x)	\
+	(((x) >> CPL_TX_DATA_ISO_SCSI_S) & CPL_TX_DATA_ISO_SCSI_M)
+
+/* cpl_tx_data_iso.reserved2_seglen_offset fields */
+#define CPL_TX_DATA_ISO_SEGLEN_OFFSET_S		0
+#define CPL_TX_DATA_ISO_SEGLEN_OFFSET_M		0xffffff
+#define CPL_TX_DATA_ISO_SEGLEN_OFFSET_V(x)	\
+	((x) << CPL_TX_DATA_ISO_SEGLEN_OFFSET_S)
+#define CPL_TX_DATA_ISO_SEGLEN_OFFSET_G(x)	\
+	(((x) >> CPL_TX_DATA_ISO_SEGLEN_OFFSET_S) & \
+	 CPL_TX_DATA_ISO_SEGLEN_OFFSET_M)
+
+struct cpl_rx_data {
+	union opcode_tid ot;
+	__be16 rsvd;
+	__be16 len;
+	__be32 seq;
+	__be16 urg;
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	u8 dack_mode:2;
+	u8 psh:1;
+	u8 heartbeat:1;
+	u8 ddp_off:1;
+	u8 :3;
+#else
+	u8 :3;
+	u8 ddp_off:1;
+	u8 heartbeat:1;
+	u8 psh:1;
+	u8 dack_mode:2;
+#endif
+	u8 status;
+};
+
+struct cpl_rx_data_ack {
+	WR_HDR;
+	union opcode_tid ot;
+	__be32 credit_dack;
+};
+
+/* cpl_rx_data_ack.ack_seq fields */
+#define RX_CREDITS_S    0
+#define RX_CREDITS_V(x) ((x) << RX_CREDITS_S)
+
+#define RX_FORCE_ACK_S    28
+#define RX_FORCE_ACK_V(x) ((x) << RX_FORCE_ACK_S)
+#define RX_FORCE_ACK_F    RX_FORCE_ACK_V(1U)
+
+#define RX_DACK_MODE_S    29
+#define RX_DACK_MODE_M    0x3
+#define RX_DACK_MODE_V(x) ((x) << RX_DACK_MODE_S)
+#define RX_DACK_MODE_G(x) (((x) >> RX_DACK_MODE_S) & RX_DACK_MODE_M)
+
+#define RX_DACK_CHANGE_S    31
+#define RX_DACK_CHANGE_V(x) ((x) << RX_DACK_CHANGE_S)
+#define RX_DACK_CHANGE_F    RX_DACK_CHANGE_V(1U)
+
+struct cpl_rx_pkt {
+	struct rss_header rsshdr;
+	u8 opcode;
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	u8 iff:4;
+	u8 csum_calc:1;
+	u8 ipmi_pkt:1;
+	u8 vlan_ex:1;
+	u8 ip_frag:1;
+#else
+	u8 ip_frag:1;
+	u8 vlan_ex:1;
+	u8 ipmi_pkt:1;
+	u8 csum_calc:1;
+	u8 iff:4;
+#endif
+	__be16 csum;
+	__be16 vlan;
+	__be16 len;
+	__be32 l2info;
+	__be16 hdr_len;
+	__be16 err_vec;
+};
+
+#define RX_T6_ETHHDR_LEN_M    0xFF
+#define RX_T6_ETHHDR_LEN_G(x) (((x) >> RX_ETHHDR_LEN_S) & RX_T6_ETHHDR_LEN_M)
+
+#define RXF_PSH_S    20
+#define RXF_PSH_V(x) ((x) << RXF_PSH_S)
+#define RXF_PSH_F    RXF_PSH_V(1U)
+
+#define RXF_SYN_S    21
+#define RXF_SYN_V(x) ((x) << RXF_SYN_S)
+#define RXF_SYN_F    RXF_SYN_V(1U)
+
+#define RXF_UDP_S    22
+#define RXF_UDP_V(x) ((x) << RXF_UDP_S)
+#define RXF_UDP_F    RXF_UDP_V(1U)
+
+#define RXF_TCP_S    23
+#define RXF_TCP_V(x) ((x) << RXF_TCP_S)
+#define RXF_TCP_F    RXF_TCP_V(1U)
+
+#define RXF_IP_S    24
+#define RXF_IP_V(x) ((x) << RXF_IP_S)
+#define RXF_IP_F    RXF_IP_V(1U)
+
+#define RXF_IP6_S    25
+#define RXF_IP6_V(x) ((x) << RXF_IP6_S)
+#define RXF_IP6_F    RXF_IP6_V(1U)
+
+#define RXF_SYN_COOKIE_S    26
+#define RXF_SYN_COOKIE_V(x) ((x) << RXF_SYN_COOKIE_S)
+#define RXF_SYN_COOKIE_F    RXF_SYN_COOKIE_V(1U)
+
+#define RXF_FCOE_S    26
+#define RXF_FCOE_V(x) ((x) << RXF_FCOE_S)
+#define RXF_FCOE_F    RXF_FCOE_V(1U)
+
+#define RXF_LRO_S    27
+#define RXF_LRO_V(x) ((x) << RXF_LRO_S)
+#define RXF_LRO_F    RXF_LRO_V(1U)
+
+/* rx_pkt.l2info fields */
+#define RX_ETHHDR_LEN_S    0
+#define RX_ETHHDR_LEN_M    0x1F
+#define RX_ETHHDR_LEN_V(x) ((x) << RX_ETHHDR_LEN_S)
+#define RX_ETHHDR_LEN_G(x) (((x) >> RX_ETHHDR_LEN_S) & RX_ETHHDR_LEN_M)
+
+#define RX_T5_ETHHDR_LEN_S    0
+#define RX_T5_ETHHDR_LEN_M    0x3F
+#define RX_T5_ETHHDR_LEN_V(x) ((x) << RX_T5_ETHHDR_LEN_S)
+#define RX_T5_ETHHDR_LEN_G(x) (((x) >> RX_T5_ETHHDR_LEN_S) & RX_T5_ETHHDR_LEN_M)
+
+#define RX_MACIDX_S    8
+#define RX_MACIDX_M    0x1FF
+#define RX_MACIDX_V(x) ((x) << RX_MACIDX_S)
+#define RX_MACIDX_G(x) (((x) >> RX_MACIDX_S) & RX_MACIDX_M)
+
+#define RXF_SYN_S    21
+#define RXF_SYN_V(x) ((x) << RXF_SYN_S)
+#define RXF_SYN_F    RXF_SYN_V(1U)
+
+#define RX_CHAN_S    28
+#define RX_CHAN_M    0xF
+#define RX_CHAN_V(x) ((x) << RX_CHAN_S)
+#define RX_CHAN_G(x) (((x) >> RX_CHAN_S) & RX_CHAN_M)
+
+/* rx_pkt.hdr_len fields */
+#define RX_TCPHDR_LEN_S    0
+#define RX_TCPHDR_LEN_M    0x3F
+#define RX_TCPHDR_LEN_V(x) ((x) << RX_TCPHDR_LEN_S)
+#define RX_TCPHDR_LEN_G(x) (((x) >> RX_TCPHDR_LEN_S) & RX_TCPHDR_LEN_M)
+
+#define RX_IPHDR_LEN_S    6
+#define RX_IPHDR_LEN_M    0x3FF
+#define RX_IPHDR_LEN_V(x) ((x) << RX_IPHDR_LEN_S)
+#define RX_IPHDR_LEN_G(x) (((x) >> RX_IPHDR_LEN_S) & RX_IPHDR_LEN_M)
+
+/* rx_pkt.err_vec fields */
+#define RXERR_CSUM_S    13
+#define RXERR_CSUM_V(x) ((x) << RXERR_CSUM_S)
+#define RXERR_CSUM_F    RXERR_CSUM_V(1U)
+
+#define T6_COMPR_RXERR_LEN_S    1
+#define T6_COMPR_RXERR_LEN_V(x) ((x) << T6_COMPR_RXERR_LEN_S)
+#define T6_COMPR_RXERR_LEN_F    T6_COMPR_RXERR_LEN_V(1U)
+
+#define T6_COMPR_RXERR_VEC_S    0
+#define T6_COMPR_RXERR_VEC_M    0x3F
+#define T6_COMPR_RXERR_VEC_V(x) ((x) << T6_COMPR_RXERR_LEN_S)
+#define T6_COMPR_RXERR_VEC_G(x) \
+		(((x) >> T6_COMPR_RXERR_VEC_S) & T6_COMPR_RXERR_VEC_M)
+
+/* Logical OR of RX_ERROR_CSUM, RX_ERROR_CSIP */
+#define T6_COMPR_RXERR_SUM_S    4
+#define T6_COMPR_RXERR_SUM_V(x) ((x) << T6_COMPR_RXERR_SUM_S)
+#define T6_COMPR_RXERR_SUM_F    T6_COMPR_RXERR_SUM_V(1U)
+
+#define T6_RX_TNLHDR_LEN_S    8
+#define T6_RX_TNLHDR_LEN_M    0xFF
+#define T6_RX_TNLHDR_LEN_V(x) ((x) << T6_RX_TNLHDR_LEN_S)
+#define T6_RX_TNLHDR_LEN_G(x) (((x) >> T6_RX_TNLHDR_LEN_S) & T6_RX_TNLHDR_LEN_M)
+
+struct cpl_trace_pkt {
+	u8 opcode;
+	u8 intf;
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	u8 runt:4;
+	u8 filter_hit:4;
+	u8 :6;
+	u8 err:1;
+	u8 trunc:1;
+#else
+	u8 filter_hit:4;
+	u8 runt:4;
+	u8 trunc:1;
+	u8 err:1;
+	u8 :6;
+#endif
+	__be16 rsvd;
+	__be16 len;
+	__be64 tstamp;
+};
+
+struct cpl_t5_trace_pkt {
+	__u8 opcode;
+	__u8 intf;
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	__u8 runt:4;
+	__u8 filter_hit:4;
+	__u8:6;
+	__u8 err:1;
+	__u8 trunc:1;
+#else
+	__u8 filter_hit:4;
+	__u8 runt:4;
+	__u8 trunc:1;
+	__u8 err:1;
+	__u8:6;
+#endif
+	__be16 rsvd;
+	__be16 len;
+	__be64 tstamp;
+	__be64 rsvd1;
+};
+
+struct cpl_l2t_write_req {
+	WR_HDR;
+	union opcode_tid ot;
+	__be16 params;
+	__be16 l2t_idx;
+	__be16 vlan;
+	u8 dst_mac[6];
+};
+
+/* cpl_l2t_write_req.params fields */
+#define L2T_W_INFO_S    2
+#define L2T_W_INFO_V(x) ((x) << L2T_W_INFO_S)
+
+#define L2T_W_PORT_S    8
+#define L2T_W_PORT_V(x) ((x) << L2T_W_PORT_S)
+
+#define L2T_W_NOREPLY_S    15
+#define L2T_W_NOREPLY_V(x) ((x) << L2T_W_NOREPLY_S)
+#define L2T_W_NOREPLY_F    L2T_W_NOREPLY_V(1U)
+
+#define CPL_L2T_VLAN_NONE 0xfff
+
+struct cpl_l2t_write_rpl {
+	union opcode_tid ot;
+	u8 status;
+	u8 rsvd[3];
+};
+
+struct cpl_smt_write_req {
+	WR_HDR;
+	union opcode_tid ot;
+	__be32 params;
+	__be16 pfvf1;
+	u8 src_mac1[6];
+	__be16 pfvf0;
+	u8 src_mac0[6];
+};
+
+struct cpl_t6_smt_write_req {
+	WR_HDR;
+	union opcode_tid ot;
+	__be32 params;
+	__be64 tag;
+	__be16 pfvf0;
+	u8 src_mac0[6];
+	__be32 local_ip;
+	__be32 rsvd;
+};
+
+struct cpl_smt_write_rpl {
+	union opcode_tid ot;
+	u8 status;
+	u8 rsvd[3];
+};
+
+/* cpl_smt_{read,write}_req.params fields */
+#define SMTW_OVLAN_IDX_S	16
+#define SMTW_OVLAN_IDX_V(x)	((x) << SMTW_OVLAN_IDX_S)
+
+#define SMTW_IDX_S	20
+#define SMTW_IDX_V(x)	((x) << SMTW_IDX_S)
+
+#define SMTW_NORPL_S	31
+#define SMTW_NORPL_V(x)	((x) << SMTW_NORPL_S)
+#define SMTW_NORPL_F	SMTW_NORPL_V(1U)
+
+struct cpl_rdma_terminate {
+	union opcode_tid ot;
+	__be16 rsvd;
+	__be16 len;
+};
+
+struct cpl_sge_egr_update {
+	__be32 opcode_qid;
+	__be16 cidx;
+	__be16 pidx;
+};
+
+/* cpl_sge_egr_update.ot fields */
+#define EGR_QID_S    0
+#define EGR_QID_M    0x1FFFF
+#define EGR_QID_G(x) (((x) >> EGR_QID_S) & EGR_QID_M)
+
+/* cpl_fw*.type values */
+enum {
+	FW_TYPE_CMD_RPL = 0,
+	FW_TYPE_WR_RPL = 1,
+	FW_TYPE_CQE = 2,
+	FW_TYPE_OFLD_CONNECTION_WR_RPL = 3,
+	FW_TYPE_RSSCPL = 4,
+};
+
+struct cpl_fw4_pld {
+	u8 opcode;
+	u8 rsvd0[3];
+	u8 type;
+	u8 rsvd1;
+	__be16 len;
+	__be64 data;
+	__be64 rsvd2;
+};
+
+struct cpl_fw6_pld {
+	u8 opcode;
+	u8 rsvd[5];
+	__be16 len;
+	__be64 data[4];
+};
+
+struct cpl_fw4_msg {
+	u8 opcode;
+	u8 type;
+	__be16 rsvd0;
+	__be32 rsvd1;
+	__be64 data[2];
+};
+
+struct cpl_fw4_ack {
+	union opcode_tid ot;
+	u8 credits;
+	u8 rsvd0[2];
+	u8 seq_vld;
+	__be32 snd_nxt;
+	__be32 snd_una;
+	__be64 rsvd1;
+};
+
+enum {
+	CPL_FW4_ACK_FLAGS_SEQVAL	= 0x1,	/* seqn valid */
+	CPL_FW4_ACK_FLAGS_CH		= 0x2,	/* channel change complete */
+	CPL_FW4_ACK_FLAGS_FLOWC		= 0x4,	/* fw_flowc_wr complete */
+};
+
+#define CPL_FW4_ACK_FLOWID_S    0
+#define CPL_FW4_ACK_FLOWID_M    0xffffff
+#define CPL_FW4_ACK_FLOWID_G(x) \
+	(((x) >> CPL_FW4_ACK_FLOWID_S) & CPL_FW4_ACK_FLOWID_M)
+
+struct cpl_fw6_msg {
+	u8 opcode;
+	u8 type;
+	__be16 rsvd0;
+	__be32 rsvd1;
+	__be64 data[4];
+};
+
+/* cpl_fw6_msg.type values */
+enum {
+	FW6_TYPE_CMD_RPL = 0,
+	FW6_TYPE_WR_RPL = 1,
+	FW6_TYPE_CQE = 2,
+	FW6_TYPE_OFLD_CONNECTION_WR_RPL = 3,
+	FW6_TYPE_RSSCPL = FW_TYPE_RSSCPL,
+};
+
+struct cpl_fw6_msg_ofld_connection_wr_rpl {
+	__u64   cookie;
+	__be32  tid;    /* or atid in case of active failure */
+	__u8    t_state;
+	__u8    retval;
+	__u8    rsvd[2];
+};
+
+struct cpl_tx_data {
+	union opcode_tid ot;
+	__be32 len;
+	__be32 rsvd;
+	__be32 flags;
+};
+
+/* cpl_tx_data.flags field */
+#define TX_FORCE_S	13
+#define TX_FORCE_V(x)	((x) << TX_FORCE_S)
+
+#define TX_DATA_MSS_S    16
+#define TX_DATA_MSS_M    0xFFFF
+#define TX_DATA_MSS_V(x) ((x) << TX_DATA_MSS_S)
+#define TX_DATA_MSS_G(x) (((x) >> TX_DATA_MSS_S) & TX_DATA_MSS_M)
+
+#define TX_LENGTH_S    0
+#define TX_LENGTH_M    0xFFFF
+#define TX_LENGTH_V(x) ((x) << TX_LENGTH_S)
+#define TX_LENGTH_G(x) (((x) >> TX_LENGTH_S) & TX_LENGTH_M)
+
+#define T6_TX_FORCE_S		20
+#define T6_TX_FORCE_V(x)	((x) << T6_TX_FORCE_S)
+#define T6_TX_FORCE_F		T6_TX_FORCE_V(1U)
+
+#define TX_URG_S    16
+#define TX_URG_V(x) ((x) << TX_URG_S)
+
+#define TX_SHOVE_S    14
+#define TX_SHOVE_V(x) ((x) << TX_SHOVE_S)
+#define TX_SHOVE_F    TX_SHOVE_V(1U)
+
+#define TX_BYPASS_S    21
+#define TX_BYPASS_V(x) ((x) << TX_BYPASS_S)
+#define TX_BYPASS_F    TX_BYPASS_V(1U)
+
+#define TX_PUSH_S    22
+#define TX_PUSH_V(x) ((x) << TX_PUSH_S)
+#define TX_PUSH_F    TX_PUSH_V(1U)
+
+#define TX_ULP_MODE_S    10
+#define TX_ULP_MODE_M    0x7
+#define TX_ULP_MODE_V(x) ((x) << TX_ULP_MODE_S)
+#define TX_ULP_MODE_G(x) (((x) >> TX_ULP_MODE_S) & TX_ULP_MODE_M)
+
+enum {
+	ULP_TX_MEM_READ = 2,
+	ULP_TX_MEM_WRITE = 3,
+	ULP_TX_PKT = 4
+};
+
+enum {
+	ULP_TX_SC_NOOP = 0x80,
+	ULP_TX_SC_IMM  = 0x81,
+	ULP_TX_SC_DSGL = 0x82,
+	ULP_TX_SC_ISGL = 0x83,
+	ULP_TX_SC_MEMRD = 0x86
+};
+
+#define ULPTX_CMD_S    24
+#define ULPTX_CMD_V(x) ((x) << ULPTX_CMD_S)
+
+#define ULPTX_LEN16_S    0
+#define ULPTX_LEN16_M    0xFF
+#define ULPTX_LEN16_V(x) ((x) << ULPTX_LEN16_S)
+
+#define ULP_TX_SC_MORE_S 23
+#define ULP_TX_SC_MORE_V(x) ((x) << ULP_TX_SC_MORE_S)
+#define ULP_TX_SC_MORE_F  ULP_TX_SC_MORE_V(1U)
+
+struct ulptx_sge_pair {
+	__be32 len[2];
+	__be64 addr[2];
+};
+
+struct ulptx_sgl {
+	__be32 cmd_nsge;
+	__be32 len0;
+	__be64 addr0;
+	struct ulptx_sge_pair sge[];
+};
+
+struct ulptx_idata {
+	__be32 cmd_more;
+	__be32 len;
+};
+
+struct ulp_txpkt {
+	__be32 cmd_dest;
+	__be32 len;
+};
+
+#define ULPTX_CMD_S    24
+#define ULPTX_CMD_M    0xFF
+#define ULPTX_CMD_V(x) ((x) << ULPTX_CMD_S)
+
+#define ULPTX_NSGE_S    0
+#define ULPTX_NSGE_V(x) ((x) << ULPTX_NSGE_S)
+
+#define ULPTX_MORE_S	23
+#define ULPTX_MORE_V(x)	((x) << ULPTX_MORE_S)
+#define ULPTX_MORE_F	ULPTX_MORE_V(1U)
+
+#define ULP_TXPKT_DEST_S    16
+#define ULP_TXPKT_DEST_M    0x3
+#define ULP_TXPKT_DEST_V(x) ((x) << ULP_TXPKT_DEST_S)
+
+#define ULP_TXPKT_FID_S     4
+#define ULP_TXPKT_FID_M     0x7ff
+#define ULP_TXPKT_FID_V(x)  ((x) << ULP_TXPKT_FID_S)
+
+#define ULP_TXPKT_RO_S      3
+#define ULP_TXPKT_RO_V(x) ((x) << ULP_TXPKT_RO_S)
+#define ULP_TXPKT_RO_F ULP_TXPKT_RO_V(1U)
+
+enum cpl_tx_tnl_lso_type {
+	TX_TNL_TYPE_OPAQUE,
+	TX_TNL_TYPE_NVGRE,
+	TX_TNL_TYPE_VXLAN,
+	TX_TNL_TYPE_GENEVE,
+};
+
+struct cpl_tx_tnl_lso {
+	__be32 op_to_IpIdSplitOut;
+	__be16 IpIdOffsetOut;
+	__be16 UdpLenSetOut_to_TnlHdrLen;
+	__be64 r1;
+	__be32 Flow_to_TcpHdrLen;
+	__be16 IpIdOffset;
+	__be16 IpIdSplit_to_Mss;
+	__be32 TCPSeqOffset;
+	__be32 EthLenOffset_Size;
+	/* encapsulated CPL (TX_PKT_XT) follows here */
+};
+
+#define CPL_TX_TNL_LSO_OPCODE_S		24
+#define CPL_TX_TNL_LSO_OPCODE_M		0xff
+#define CPL_TX_TNL_LSO_OPCODE_V(x)      ((x) << CPL_TX_TNL_LSO_OPCODE_S)
+#define CPL_TX_TNL_LSO_OPCODE_G(x)      \
+	(((x) >> CPL_TX_TNL_LSO_OPCODE_S) & CPL_TX_TNL_LSO_OPCODE_M)
+
+#define CPL_TX_TNL_LSO_FIRST_S		23
+#define CPL_TX_TNL_LSO_FIRST_M		0x1
+#define CPL_TX_TNL_LSO_FIRST_V(x)	((x) << CPL_TX_TNL_LSO_FIRST_S)
+#define CPL_TX_TNL_LSO_FIRST_G(x)	\
+	(((x) >> CPL_TX_TNL_LSO_FIRST_S) & CPL_TX_TNL_LSO_FIRST_M)
+#define CPL_TX_TNL_LSO_FIRST_F		CPL_TX_TNL_LSO_FIRST_V(1U)
+
+#define CPL_TX_TNL_LSO_LAST_S		22
+#define CPL_TX_TNL_LSO_LAST_M		0x1
+#define CPL_TX_TNL_LSO_LAST_V(x)	((x) << CPL_TX_TNL_LSO_LAST_S)
+#define CPL_TX_TNL_LSO_LAST_G(x)	\
+	(((x) >> CPL_TX_TNL_LSO_LAST_S) & CPL_TX_TNL_LSO_LAST_M)
+#define CPL_TX_TNL_LSO_LAST_F		CPL_TX_TNL_LSO_LAST_V(1U)
+
+#define CPL_TX_TNL_LSO_ETHHDRLENXOUT_S	21
+#define CPL_TX_TNL_LSO_ETHHDRLENXOUT_M	0x1
+#define CPL_TX_TNL_LSO_ETHHDRLENXOUT_V(x) \
+	((x) << CPL_TX_TNL_LSO_ETHHDRLENXOUT_S)
+#define CPL_TX_TNL_LSO_ETHHDRLENXOUT_G(x) \
+	(((x) >> CPL_TX_TNL_LSO_ETHHDRLENXOUT_S) & \
+	 CPL_TX_TNL_LSO_ETHHDRLENXOUT_M)
+#define CPL_TX_TNL_LSO_ETHHDRLENXOUT_F CPL_TX_TNL_LSO_ETHHDRLENXOUT_V(1U)
+
+#define CPL_TX_TNL_LSO_IPV6OUT_S	20
+#define CPL_TX_TNL_LSO_IPV6OUT_M	0x1
+#define CPL_TX_TNL_LSO_IPV6OUT_V(x)	((x) << CPL_TX_TNL_LSO_IPV6OUT_S)
+#define CPL_TX_TNL_LSO_IPV6OUT_G(x)	\
+	(((x) >> CPL_TX_TNL_LSO_IPV6OUT_S) & CPL_TX_TNL_LSO_IPV6OUT_M)
+#define CPL_TX_TNL_LSO_IPV6OUT_F        CPL_TX_TNL_LSO_IPV6OUT_V(1U)
+
+#define CPL_TX_TNL_LSO_ETHHDRLEN_S	16
+#define CPL_TX_TNL_LSO_ETHHDRLEN_M	0xf
+#define CPL_TX_TNL_LSO_ETHHDRLEN_V(x)	((x) << CPL_TX_TNL_LSO_ETHHDRLEN_S)
+#define CPL_TX_TNL_LSO_ETHHDRLEN_G(x)	\
+	(((x) >> CPL_TX_TNL_LSO_ETHHDRLEN_S) & CPL_TX_TNL_LSO_ETHHDRLEN_M)
+
+#define CPL_TX_TNL_LSO_IPHDRLEN_S	4
+#define CPL_TX_TNL_LSO_IPHDRLEN_M	0xfff
+#define CPL_TX_TNL_LSO_IPHDRLEN_V(x)	((x) << CPL_TX_TNL_LSO_IPHDRLEN_S)
+#define CPL_TX_TNL_LSO_IPHDRLEN_G(x)    \
+	(((x) >> CPL_TX_TNL_LSO_IPHDRLEN_S) & CPL_TX_TNL_LSO_IPHDRLEN_M)
+
+#define CPL_TX_TNL_LSO_TCPHDRLEN_S	0
+#define CPL_TX_TNL_LSO_TCPHDRLEN_M	0xf
+#define CPL_TX_TNL_LSO_TCPHDRLEN_V(x)	((x) << CPL_TX_TNL_LSO_TCPHDRLEN_S)
+#define CPL_TX_TNL_LSO_TCPHDRLEN_G(x)   \
+	(((x) >> CPL_TX_TNL_LSO_TCPHDRLEN_S) & CPL_TX_TNL_LSO_TCPHDRLEN_M)
+
+#define CPL_TX_TNL_LSO_MSS_S            0
+#define CPL_TX_TNL_LSO_MSS_M            0x3fff
+#define CPL_TX_TNL_LSO_MSS_V(x)         ((x) << CPL_TX_TNL_LSO_MSS_S)
+#define CPL_TX_TNL_LSO_MSS_G(x)         \
+	(((x) >> CPL_TX_TNL_LSO_MSS_S) & CPL_TX_TNL_LSO_MSS_M)
+
+#define CPL_TX_TNL_LSO_SIZE_S		0
+#define CPL_TX_TNL_LSO_SIZE_M		0xfffffff
+#define CPL_TX_TNL_LSO_SIZE_V(x)	((x) << CPL_TX_TNL_LSO_SIZE_S)
+#define CPL_TX_TNL_LSO_SIZE_G(x)	\
+	(((x) >> CPL_TX_TNL_LSO_SIZE_S) & CPL_TX_TNL_LSO_SIZE_M)
+
+#define CPL_TX_TNL_LSO_ETHHDRLENOUT_S   16
+#define CPL_TX_TNL_LSO_ETHHDRLENOUT_M   0xf
+#define CPL_TX_TNL_LSO_ETHHDRLENOUT_V(x) \
+	((x) << CPL_TX_TNL_LSO_ETHHDRLENOUT_S)
+#define CPL_TX_TNL_LSO_ETHHDRLENOUT_G(x) \
+	(((x) >> CPL_TX_TNL_LSO_ETHHDRLENOUT_S) & CPL_TX_TNL_LSO_ETHHDRLENOUT_M)
+
+#define CPL_TX_TNL_LSO_IPHDRLENOUT_S    4
+#define CPL_TX_TNL_LSO_IPHDRLENOUT_M    0xfff
+#define CPL_TX_TNL_LSO_IPHDRLENOUT_V(x) ((x) << CPL_TX_TNL_LSO_IPHDRLENOUT_S)
+#define CPL_TX_TNL_LSO_IPHDRLENOUT_G(x) \
+	(((x) >> CPL_TX_TNL_LSO_IPHDRLENOUT_S) & CPL_TX_TNL_LSO_IPHDRLENOUT_M)
+
+#define CPL_TX_TNL_LSO_IPHDRCHKOUT_S    3
+#define CPL_TX_TNL_LSO_IPHDRCHKOUT_M    0x1
+#define CPL_TX_TNL_LSO_IPHDRCHKOUT_V(x) ((x) << CPL_TX_TNL_LSO_IPHDRCHKOUT_S)
+#define CPL_TX_TNL_LSO_IPHDRCHKOUT_G(x) \
+	(((x) >> CPL_TX_TNL_LSO_IPHDRCHKOUT_S) & CPL_TX_TNL_LSO_IPHDRCHKOUT_M)
+#define CPL_TX_TNL_LSO_IPHDRCHKOUT_F    CPL_TX_TNL_LSO_IPHDRCHKOUT_V(1U)
+
+#define CPL_TX_TNL_LSO_IPLENSETOUT_S	2
+#define CPL_TX_TNL_LSO_IPLENSETOUT_M	0x1
+#define CPL_TX_TNL_LSO_IPLENSETOUT_V(x) ((x) << CPL_TX_TNL_LSO_IPLENSETOUT_S)
+#define CPL_TX_TNL_LSO_IPLENSETOUT_G(x) \
+	(((x) >> CPL_TX_TNL_LSO_IPLENSETOUT_S) & CPL_TX_TNL_LSO_IPLENSETOUT_M)
+#define CPL_TX_TNL_LSO_IPLENSETOUT_F	CPL_TX_TNL_LSO_IPLENSETOUT_V(1U)
+
+#define CPL_TX_TNL_LSO_IPIDINCOUT_S	1
+#define CPL_TX_TNL_LSO_IPIDINCOUT_M	0x1
+#define CPL_TX_TNL_LSO_IPIDINCOUT_V(x)  ((x) << CPL_TX_TNL_LSO_IPIDINCOUT_S)
+#define CPL_TX_TNL_LSO_IPIDINCOUT_G(x)  \
+	(((x) >> CPL_TX_TNL_LSO_IPIDINCOUT_S) & CPL_TX_TNL_LSO_IPIDINCOUT_M)
+#define CPL_TX_TNL_LSO_IPIDINCOUT_F     CPL_TX_TNL_LSO_IPIDINCOUT_V(1U)
+
+#define CPL_TX_TNL_LSO_UDPCHKCLROUT_S   14
+#define CPL_TX_TNL_LSO_UDPCHKCLROUT_M   0x1
+#define CPL_TX_TNL_LSO_UDPCHKCLROUT_V(x) \
+	((x) << CPL_TX_TNL_LSO_UDPCHKCLROUT_S)
+#define CPL_TX_TNL_LSO_UDPCHKCLROUT_G(x) \
+	(((x) >> CPL_TX_TNL_LSO_UDPCHKCLROUT_S) & \
+	 CPL_TX_TNL_LSO_UDPCHKCLROUT_M)
+#define CPL_TX_TNL_LSO_UDPCHKCLROUT_F   CPL_TX_TNL_LSO_UDPCHKCLROUT_V(1U)
+
+#define CPL_TX_TNL_LSO_UDPLENSETOUT_S   15
+#define CPL_TX_TNL_LSO_UDPLENSETOUT_M   0x1
+#define CPL_TX_TNL_LSO_UDPLENSETOUT_V(x) \
+	((x) << CPL_TX_TNL_LSO_UDPLENSETOUT_S)
+#define CPL_TX_TNL_LSO_UDPLENSETOUT_G(x) \
+	(((x) >> CPL_TX_TNL_LSO_UDPLENSETOUT_S) & \
+	 CPL_TX_TNL_LSO_UDPLENSETOUT_M)
+#define CPL_TX_TNL_LSO_UDPLENSETOUT_F   CPL_TX_TNL_LSO_UDPLENSETOUT_V(1U)
+
+#define CPL_TX_TNL_LSO_TNLTYPE_S	12
+#define CPL_TX_TNL_LSO_TNLTYPE_M	0x3
+#define CPL_TX_TNL_LSO_TNLTYPE_V(x)	((x) << CPL_TX_TNL_LSO_TNLTYPE_S)
+#define CPL_TX_TNL_LSO_TNLTYPE_G(x)	\
+	(((x) >> CPL_TX_TNL_LSO_TNLTYPE_S) & CPL_TX_TNL_LSO_TNLTYPE_M)
+
+#define S_CPL_TX_TNL_LSO_ETHHDRLEN	16
+#define M_CPL_TX_TNL_LSO_ETHHDRLEN	0xf
+#define V_CPL_TX_TNL_LSO_ETHHDRLEN(x)	((x) << S_CPL_TX_TNL_LSO_ETHHDRLEN)
+#define G_CPL_TX_TNL_LSO_ETHHDRLEN(x)	\
+	(((x) >> S_CPL_TX_TNL_LSO_ETHHDRLEN) & M_CPL_TX_TNL_LSO_ETHHDRLEN)
+
+#define CPL_TX_TNL_LSO_TNLHDRLEN_S      0
+#define CPL_TX_TNL_LSO_TNLHDRLEN_M      0xfff
+#define CPL_TX_TNL_LSO_TNLHDRLEN_V(x)	((x) << CPL_TX_TNL_LSO_TNLHDRLEN_S)
+#define CPL_TX_TNL_LSO_TNLHDRLEN_G(x)   \
+	(((x) >> CPL_TX_TNL_LSO_TNLHDRLEN_S) & CPL_TX_TNL_LSO_TNLHDRLEN_M)
+
+#define CPL_TX_TNL_LSO_IPV6_S		20
+#define CPL_TX_TNL_LSO_IPV6_M		0x1
+#define CPL_TX_TNL_LSO_IPV6_V(x)	((x) << CPL_TX_TNL_LSO_IPV6_S)
+#define CPL_TX_TNL_LSO_IPV6_G(x)	\
+	(((x) >> CPL_TX_TNL_LSO_IPV6_S) & CPL_TX_TNL_LSO_IPV6_M)
+#define CPL_TX_TNL_LSO_IPV6_F		CPL_TX_TNL_LSO_IPV6_V(1U)
+
+#define ULP_TX_SC_MORE_S 23
+#define ULP_TX_SC_MORE_V(x) ((x) << ULP_TX_SC_MORE_S)
+#define ULP_TX_SC_MORE_F  ULP_TX_SC_MORE_V(1U)
+
+struct ulp_mem_io {
+	WR_HDR;
+	__be32 cmd;
+	__be32 len16;             /* command length */
+	__be32 dlen;              /* data length in 32-byte units */
+	__be32 lock_addr;
+};
+
+#define ULP_MEMIO_LOCK_S    31
+#define ULP_MEMIO_LOCK_V(x) ((x) << ULP_MEMIO_LOCK_S)
+#define ULP_MEMIO_LOCK_F    ULP_MEMIO_LOCK_V(1U)
+
+/* additional ulp_mem_io.cmd fields */
+#define ULP_MEMIO_ORDER_S    23
+#define ULP_MEMIO_ORDER_V(x) ((x) << ULP_MEMIO_ORDER_S)
+#define ULP_MEMIO_ORDER_F    ULP_MEMIO_ORDER_V(1U)
+
+#define T5_ULP_MEMIO_IMM_S    23
+#define T5_ULP_MEMIO_IMM_V(x) ((x) << T5_ULP_MEMIO_IMM_S)
+#define T5_ULP_MEMIO_IMM_F    T5_ULP_MEMIO_IMM_V(1U)
+
+#define T5_ULP_MEMIO_ORDER_S    22
+#define T5_ULP_MEMIO_ORDER_V(x) ((x) << T5_ULP_MEMIO_ORDER_S)
+#define T5_ULP_MEMIO_ORDER_F    T5_ULP_MEMIO_ORDER_V(1U)
+
+#define T5_ULP_MEMIO_FID_S	4
+#define T5_ULP_MEMIO_FID_M	0x7ff
+#define T5_ULP_MEMIO_FID_V(x)	((x) << T5_ULP_MEMIO_FID_S)
+
+/* ulp_mem_io.lock_addr fields */
+#define ULP_MEMIO_ADDR_S    0
+#define ULP_MEMIO_ADDR_V(x) ((x) << ULP_MEMIO_ADDR_S)
+
+/* ulp_mem_io.dlen fields */
+#define ULP_MEMIO_DATA_LEN_S    0
+#define ULP_MEMIO_DATA_LEN_V(x) ((x) << ULP_MEMIO_DATA_LEN_S)
+
+#define ULPTX_NSGE_S    0
+#define ULPTX_NSGE_M    0xFFFF
+#define ULPTX_NSGE_V(x) ((x) << ULPTX_NSGE_S)
+#define ULPTX_NSGE_G(x) (((x) >> ULPTX_NSGE_S) & ULPTX_NSGE_M)
+
+struct ulptx_sc_memrd {
+	__be32 cmd_to_len;
+	__be32 addr;
+};
+
+#define ULP_TXPKT_DATAMODIFY_S       23
+#define ULP_TXPKT_DATAMODIFY_M       0x1
+#define ULP_TXPKT_DATAMODIFY_V(x)    ((x) << ULP_TXPKT_DATAMODIFY_S)
+#define ULP_TXPKT_DATAMODIFY_G(x)    \
+	(((x) >> ULP_TXPKT_DATAMODIFY_S) & ULP_TXPKT_DATAMODIFY__M)
+#define ULP_TXPKT_DATAMODIFY_F       ULP_TXPKT_DATAMODIFY_V(1U)
+
+#define ULP_TXPKT_CHANNELID_S        22
+#define ULP_TXPKT_CHANNELID_M        0x1
+#define ULP_TXPKT_CHANNELID_V(x)     ((x) << ULP_TXPKT_CHANNELID_S)
+#define ULP_TXPKT_CHANNELID_G(x)     \
+	(((x) >> ULP_TXPKT_CHANNELID_S) & ULP_TXPKT_CHANNELID_M)
+#define ULP_TXPKT_CHANNELID_F        ULP_TXPKT_CHANNELID_V(1U)
+
+#define SCMD_SEQ_NO_CTRL_S      29
+#define SCMD_SEQ_NO_CTRL_M      0x3
+#define SCMD_SEQ_NO_CTRL_V(x)   ((x) << SCMD_SEQ_NO_CTRL_S)
+#define SCMD_SEQ_NO_CTRL_G(x)   \
+	(((x) >> SCMD_SEQ_NO_CTRL_S) & SCMD_SEQ_NO_CTRL_M)
+
+/* StsFieldPrsnt- Status field at the end of the TLS PDU */
+#define SCMD_STATUS_PRESENT_S   28
+#define SCMD_STATUS_PRESENT_M   0x1
+#define SCMD_STATUS_PRESENT_V(x)    ((x) << SCMD_STATUS_PRESENT_S)
+#define SCMD_STATUS_PRESENT_G(x)    \
+	(((x) >> SCMD_STATUS_PRESENT_S) & SCMD_STATUS_PRESENT_M)
+#define SCMD_STATUS_PRESENT_F   SCMD_STATUS_PRESENT_V(1U)
+
+/* ProtoVersion - Protocol Version 0: 1.2, 1:1.1, 2:DTLS, 3:Generic,
+ * 3-15: Reserved.
+ */
+#define SCMD_PROTO_VERSION_S    24
+#define SCMD_PROTO_VERSION_M    0xf
+#define SCMD_PROTO_VERSION_V(x) ((x) << SCMD_PROTO_VERSION_S)
+#define SCMD_PROTO_VERSION_G(x) \
+	(((x) >> SCMD_PROTO_VERSION_S) & SCMD_PROTO_VERSION_M)
+
+/* EncDecCtrl - Encryption/Decryption Control. 0: Encrypt, 1: Decrypt */
+#define SCMD_ENC_DEC_CTRL_S     23
+#define SCMD_ENC_DEC_CTRL_M     0x1
+#define SCMD_ENC_DEC_CTRL_V(x)  ((x) << SCMD_ENC_DEC_CTRL_S)
+#define SCMD_ENC_DEC_CTRL_G(x)  \
+	(((x) >> SCMD_ENC_DEC_CTRL_S) & SCMD_ENC_DEC_CTRL_M)
+#define SCMD_ENC_DEC_CTRL_F SCMD_ENC_DEC_CTRL_V(1U)
+
+/* CipherAuthSeqCtrl - Cipher Authentication Sequence Control. */
+#define SCMD_CIPH_AUTH_SEQ_CTRL_S       22
+#define SCMD_CIPH_AUTH_SEQ_CTRL_M       0x1
+#define SCMD_CIPH_AUTH_SEQ_CTRL_V(x)    \
+	((x) << SCMD_CIPH_AUTH_SEQ_CTRL_S)
+#define SCMD_CIPH_AUTH_SEQ_CTRL_G(x)    \
+	(((x) >> SCMD_CIPH_AUTH_SEQ_CTRL_S) & SCMD_CIPH_AUTH_SEQ_CTRL_M)
+#define SCMD_CIPH_AUTH_SEQ_CTRL_F   SCMD_CIPH_AUTH_SEQ_CTRL_V(1U)
+
+/* CiphMode -  Cipher Mode. 0: NOP, 1:AES-CBC, 2:AES-GCM, 3:AES-CTR,
+ * 4:Generic-AES, 5-15: Reserved.
+ */
+#define SCMD_CIPH_MODE_S    18
+#define SCMD_CIPH_MODE_M    0xf
+#define SCMD_CIPH_MODE_V(x) ((x) << SCMD_CIPH_MODE_S)
+#define SCMD_CIPH_MODE_G(x) \
+	(((x) >> SCMD_CIPH_MODE_S) & SCMD_CIPH_MODE_M)
+
+/* AuthMode - Auth Mode. 0: NOP, 1:SHA1, 2:SHA2-224, 3:SHA2-256
+ * 4-15: Reserved
+ */
+#define SCMD_AUTH_MODE_S    14
+#define SCMD_AUTH_MODE_M    0xf
+#define SCMD_AUTH_MODE_V(x) ((x) << SCMD_AUTH_MODE_S)
+#define SCMD_AUTH_MODE_G(x) \
+	(((x) >> SCMD_AUTH_MODE_S) & SCMD_AUTH_MODE_M)
+
+/* HmacCtrl - HMAC Control. 0:NOP, 1:No truncation, 2:Support HMAC Truncation
+ * per RFC 4366, 3:IPSec 96 bits, 4-7:Reserved
+ */
+#define SCMD_HMAC_CTRL_S    11
+#define SCMD_HMAC_CTRL_M    0x7
+#define SCMD_HMAC_CTRL_V(x) ((x) << SCMD_HMAC_CTRL_S)
+#define SCMD_HMAC_CTRL_G(x) \
+	(((x) >> SCMD_HMAC_CTRL_S) & SCMD_HMAC_CTRL_M)
+
+/* IvSize - IV size in units of 2 bytes */
+#define SCMD_IV_SIZE_S  7
+#define SCMD_IV_SIZE_M  0xf
+#define SCMD_IV_SIZE_V(x)   ((x) << SCMD_IV_SIZE_S)
+#define SCMD_IV_SIZE_G(x)   \
+	(((x) >> SCMD_IV_SIZE_S) & SCMD_IV_SIZE_M)
+
+/* NumIVs - Number of IVs */
+#define SCMD_NUM_IVS_S  0
+#define SCMD_NUM_IVS_M  0x7f
+#define SCMD_NUM_IVS_V(x)   ((x) << SCMD_NUM_IVS_S)
+#define SCMD_NUM_IVS_G(x)   \
+	(((x) >> SCMD_NUM_IVS_S) & SCMD_NUM_IVS_M)
+
+/* EnbDbgId - If this is enabled upper 20 (63:44) bits if SeqNumber
+ * (below) are used as Cid (connection id for debug status), these
+ * bits are padded to zero for forming the 64 bit
+ * sequence number for TLS
+ */
+#define SCMD_ENB_DBGID_S  31
+#define SCMD_ENB_DBGID_M  0x1
+#define SCMD_ENB_DBGID_V(x)   ((x) << SCMD_ENB_DBGID_S)
+#define SCMD_ENB_DBGID_G(x)   \
+	(((x) >> SCMD_ENB_DBGID_S) & SCMD_ENB_DBGID_M)
+
+/* IV generation in SW. */
+#define SCMD_IV_GEN_CTRL_S      30
+#define SCMD_IV_GEN_CTRL_M      0x1
+#define SCMD_IV_GEN_CTRL_V(x)   ((x) << SCMD_IV_GEN_CTRL_S)
+#define SCMD_IV_GEN_CTRL_G(x)   \
+	(((x) >> SCMD_IV_GEN_CTRL_S) & SCMD_IV_GEN_CTRL_M)
+#define SCMD_IV_GEN_CTRL_F  SCMD_IV_GEN_CTRL_V(1U)
+
+/* More frags */
+#define SCMD_MORE_FRAGS_S   20
+#define SCMD_MORE_FRAGS_M   0x1
+#define SCMD_MORE_FRAGS_V(x)    ((x) << SCMD_MORE_FRAGS_S)
+#define SCMD_MORE_FRAGS_G(x)    (((x) >> SCMD_MORE_FRAGS_S) & SCMD_MORE_FRAGS_M)
+
+/*last frag */
+#define SCMD_LAST_FRAG_S    19
+#define SCMD_LAST_FRAG_M    0x1
+#define SCMD_LAST_FRAG_V(x) ((x) << SCMD_LAST_FRAG_S)
+#define SCMD_LAST_FRAG_G(x) (((x) >> SCMD_LAST_FRAG_S) & SCMD_LAST_FRAG_M)
+
+/* TlsCompPdu */
+#define SCMD_TLS_COMPPDU_S    18
+#define SCMD_TLS_COMPPDU_M    0x1
+#define SCMD_TLS_COMPPDU_V(x) ((x) << SCMD_TLS_COMPPDU_S)
+#define SCMD_TLS_COMPPDU_G(x) (((x) >> SCMD_TLS_COMPPDU_S) & SCMD_TLS_COMPPDU_M)
+
+/* KeyCntxtInline - Key context inline after the scmd  OR PayloadOnly*/
+#define SCMD_KEY_CTX_INLINE_S   17
+#define SCMD_KEY_CTX_INLINE_M   0x1
+#define SCMD_KEY_CTX_INLINE_V(x)    ((x) << SCMD_KEY_CTX_INLINE_S)
+#define SCMD_KEY_CTX_INLINE_G(x)    \
+	(((x) >> SCMD_KEY_CTX_INLINE_S) & SCMD_KEY_CTX_INLINE_M)
+#define SCMD_KEY_CTX_INLINE_F   SCMD_KEY_CTX_INLINE_V(1U)
+
+/* TLSFragEnable - 0: Host created TLS PDUs, 1: TLS Framgmentation in ASIC */
+#define SCMD_TLS_FRAG_ENABLE_S  16
+#define SCMD_TLS_FRAG_ENABLE_M  0x1
+#define SCMD_TLS_FRAG_ENABLE_V(x)   ((x) << SCMD_TLS_FRAG_ENABLE_S)
+#define SCMD_TLS_FRAG_ENABLE_G(x)   \
+	(((x) >> SCMD_TLS_FRAG_ENABLE_S) & SCMD_TLS_FRAG_ENABLE_M)
+#define SCMD_TLS_FRAG_ENABLE_F  SCMD_TLS_FRAG_ENABLE_V(1U)
+
+/* MacOnly - Only send the MAC and discard PDU. This is valid for hash only
+ * modes, in this case TLS_TX  will drop the PDU and only
+ * send back the MAC bytes.
+ */
+#define SCMD_MAC_ONLY_S 15
+#define SCMD_MAC_ONLY_M 0x1
+#define SCMD_MAC_ONLY_V(x)  ((x) << SCMD_MAC_ONLY_S)
+#define SCMD_MAC_ONLY_G(x)  \
+	(((x) >> SCMD_MAC_ONLY_S) & SCMD_MAC_ONLY_M)
+#define SCMD_MAC_ONLY_F SCMD_MAC_ONLY_V(1U)
+
+/* AadIVDrop - Drop the AAD and IV fields. Useful in protocols
+ * which have complex AAD and IV formations Eg:AES-CCM
+ */
+#define SCMD_AADIVDROP_S 14
+#define SCMD_AADIVDROP_M 0x1
+#define SCMD_AADIVDROP_V(x)  ((x) << SCMD_AADIVDROP_S)
+#define SCMD_AADIVDROP_G(x)  \
+	(((x) >> SCMD_AADIVDROP_S) & SCMD_AADIVDROP_M)
+#define SCMD_AADIVDROP_F SCMD_AADIVDROP_V(1U)
+
+/* HdrLength - Length of all headers excluding TLS header
+ * present before start of crypto PDU/payload.
+ */
+#define SCMD_HDR_LEN_S  0
+#define SCMD_HDR_LEN_M  0x3fff
+#define SCMD_HDR_LEN_V(x)   ((x) << SCMD_HDR_LEN_S)
+#define SCMD_HDR_LEN_G(x)   \
+	(((x) >> SCMD_HDR_LEN_S) & SCMD_HDR_LEN_M)
+
+struct cpl_tx_sec_pdu {
+	__be32 op_ivinsrtofst;
+	__be32 pldlen;
+	__be32 aadstart_cipherstop_hi;
+	__be32 cipherstop_lo_authinsert;
+	__be32 seqno_numivs;
+	__be32 ivgen_hdrlen;
+	__be64 scmd1;
+};
+
+#define CPL_TX_SEC_PDU_OPCODE_S     24
+#define CPL_TX_SEC_PDU_OPCODE_M     0xff
+#define CPL_TX_SEC_PDU_OPCODE_V(x)  ((x) << CPL_TX_SEC_PDU_OPCODE_S)
+#define CPL_TX_SEC_PDU_OPCODE_G(x)  \
+	(((x) >> CPL_TX_SEC_PDU_OPCODE_S) & CPL_TX_SEC_PDU_OPCODE_M)
+
+/* RX Channel Id */
+#define CPL_TX_SEC_PDU_RXCHID_S  22
+#define CPL_TX_SEC_PDU_RXCHID_M  0x1
+#define CPL_TX_SEC_PDU_RXCHID_V(x)   ((x) << CPL_TX_SEC_PDU_RXCHID_S)
+#define CPL_TX_SEC_PDU_RXCHID_G(x)   \
+	(((x) >> CPL_TX_SEC_PDU_RXCHID_S) & CPL_TX_SEC_PDU_RXCHID_M)
+#define CPL_TX_SEC_PDU_RXCHID_F  CPL_TX_SEC_PDU_RXCHID_V(1U)
+
+/* Ack Follows */
+#define CPL_TX_SEC_PDU_ACKFOLLOWS_S  21
+#define CPL_TX_SEC_PDU_ACKFOLLOWS_M  0x1
+#define CPL_TX_SEC_PDU_ACKFOLLOWS_V(x)   ((x) << CPL_TX_SEC_PDU_ACKFOLLOWS_S)
+#define CPL_TX_SEC_PDU_ACKFOLLOWS_G(x)   \
+	(((x) >> CPL_TX_SEC_PDU_ACKFOLLOWS_S) & CPL_TX_SEC_PDU_ACKFOLLOWS_M)
+#define CPL_TX_SEC_PDU_ACKFOLLOWS_F  CPL_TX_SEC_PDU_ACKFOLLOWS_V(1U)
+
+/* Loopback bit in cpl_tx_sec_pdu */
+#define CPL_TX_SEC_PDU_ULPTXLPBK_S  20
+#define CPL_TX_SEC_PDU_ULPTXLPBK_M  0x1
+#define CPL_TX_SEC_PDU_ULPTXLPBK_V(x)   ((x) << CPL_TX_SEC_PDU_ULPTXLPBK_S)
+#define CPL_TX_SEC_PDU_ULPTXLPBK_G(x)   \
+	(((x) >> CPL_TX_SEC_PDU_ULPTXLPBK_S) & CPL_TX_SEC_PDU_ULPTXLPBK_M)
+#define CPL_TX_SEC_PDU_ULPTXLPBK_F  CPL_TX_SEC_PDU_ULPTXLPBK_V(1U)
+
+/* Length of cpl header encapsulated */
+#define CPL_TX_SEC_PDU_CPLLEN_S     16
+#define CPL_TX_SEC_PDU_CPLLEN_M     0xf
+#define CPL_TX_SEC_PDU_CPLLEN_V(x)  ((x) << CPL_TX_SEC_PDU_CPLLEN_S)
+#define CPL_TX_SEC_PDU_CPLLEN_G(x)  \
+	(((x) >> CPL_TX_SEC_PDU_CPLLEN_S) & CPL_TX_SEC_PDU_CPLLEN_M)
+
+/* PlaceHolder */
+#define CPL_TX_SEC_PDU_PLACEHOLDER_S    10
+#define CPL_TX_SEC_PDU_PLACEHOLDER_M    0x1
+#define CPL_TX_SEC_PDU_PLACEHOLDER_V(x) ((x) << CPL_TX_SEC_PDU_PLACEHOLDER_S)
+#define CPL_TX_SEC_PDU_PLACEHOLDER_G(x) \
+	(((x) >> CPL_TX_SEC_PDU_PLACEHOLDER_S) & \
+	 CPL_TX_SEC_PDU_PLACEHOLDER_M)
+
+/* IvInsrtOffset: Insertion location for IV */
+#define CPL_TX_SEC_PDU_IVINSRTOFST_S    0
+#define CPL_TX_SEC_PDU_IVINSRTOFST_M    0x3ff
+#define CPL_TX_SEC_PDU_IVINSRTOFST_V(x) ((x) << CPL_TX_SEC_PDU_IVINSRTOFST_S)
+#define CPL_TX_SEC_PDU_IVINSRTOFST_G(x) \
+	(((x) >> CPL_TX_SEC_PDU_IVINSRTOFST_S) & \
+	 CPL_TX_SEC_PDU_IVINSRTOFST_M)
+
+/* AadStartOffset: Offset in bytes for AAD start from
+ * the first byte following the pkt headers (0-255 bytes)
+ */
+#define CPL_TX_SEC_PDU_AADSTART_S   24
+#define CPL_TX_SEC_PDU_AADSTART_M   0xff
+#define CPL_TX_SEC_PDU_AADSTART_V(x)    ((x) << CPL_TX_SEC_PDU_AADSTART_S)
+#define CPL_TX_SEC_PDU_AADSTART_G(x)    \
+	(((x) >> CPL_TX_SEC_PDU_AADSTART_S) & \
+	 CPL_TX_SEC_PDU_AADSTART_M)
+
+/* AadStopOffset: offset in bytes for AAD stop/end from the first byte following
+ * the pkt headers (0-511 bytes)
+ */
+#define CPL_TX_SEC_PDU_AADSTOP_S    15
+#define CPL_TX_SEC_PDU_AADSTOP_M    0x1ff
+#define CPL_TX_SEC_PDU_AADSTOP_V(x) ((x) << CPL_TX_SEC_PDU_AADSTOP_S)
+#define CPL_TX_SEC_PDU_AADSTOP_G(x) \
+	(((x) >> CPL_TX_SEC_PDU_AADSTOP_S) & CPL_TX_SEC_PDU_AADSTOP_M)
+
+/* CipherStartOffset: offset in bytes for encryption/decryption start from the
+ * first byte following the pkt headers (0-1023 bytes)
+ */
+#define CPL_TX_SEC_PDU_CIPHERSTART_S    5
+#define CPL_TX_SEC_PDU_CIPHERSTART_M    0x3ff
+#define CPL_TX_SEC_PDU_CIPHERSTART_V(x) ((x) << CPL_TX_SEC_PDU_CIPHERSTART_S)
+#define CPL_TX_SEC_PDU_CIPHERSTART_G(x) \
+	(((x) >> CPL_TX_SEC_PDU_CIPHERSTART_S) & \
+	 CPL_TX_SEC_PDU_CIPHERSTART_M)
+
+/* CipherStopOffset: offset in bytes for encryption/decryption end
+ * from end of the payload of this command (0-511 bytes)
+ */
+#define CPL_TX_SEC_PDU_CIPHERSTOP_HI_S      0
+#define CPL_TX_SEC_PDU_CIPHERSTOP_HI_M      0x1f
+#define CPL_TX_SEC_PDU_CIPHERSTOP_HI_V(x)   \
+	((x) << CPL_TX_SEC_PDU_CIPHERSTOP_HI_S)
+#define CPL_TX_SEC_PDU_CIPHERSTOP_HI_G(x)   \
+	(((x) >> CPL_TX_SEC_PDU_CIPHERSTOP_HI_S) & \
+	 CPL_TX_SEC_PDU_CIPHERSTOP_HI_M)
+
+#define CPL_TX_SEC_PDU_CIPHERSTOP_LO_S      28
+#define CPL_TX_SEC_PDU_CIPHERSTOP_LO_M      0xf
+#define CPL_TX_SEC_PDU_CIPHERSTOP_LO_V(x)   \
+	((x) << CPL_TX_SEC_PDU_CIPHERSTOP_LO_S)
+#define CPL_TX_SEC_PDU_CIPHERSTOP_LO_G(x)   \
+	(((x) >> CPL_TX_SEC_PDU_CIPHERSTOP_LO_S) & \
+	 CPL_TX_SEC_PDU_CIPHERSTOP_LO_M)
+
+/* AuthStartOffset: offset in bytes for authentication start from
+ * the first byte following the pkt headers (0-1023)
+ */
+#define CPL_TX_SEC_PDU_AUTHSTART_S  18
+#define CPL_TX_SEC_PDU_AUTHSTART_M  0x3ff
+#define CPL_TX_SEC_PDU_AUTHSTART_V(x)   ((x) << CPL_TX_SEC_PDU_AUTHSTART_S)
+#define CPL_TX_SEC_PDU_AUTHSTART_G(x)   \
+	(((x) >> CPL_TX_SEC_PDU_AUTHSTART_S) & \
+	 CPL_TX_SEC_PDU_AUTHSTART_M)
+
+/* AuthStopOffset: offset in bytes for authentication
+ * end from end of the payload of this command (0-511 Bytes)
+ */
+#define CPL_TX_SEC_PDU_AUTHSTOP_S   9
+#define CPL_TX_SEC_PDU_AUTHSTOP_M   0x1ff
+#define CPL_TX_SEC_PDU_AUTHSTOP_V(x)    ((x) << CPL_TX_SEC_PDU_AUTHSTOP_S)
+#define CPL_TX_SEC_PDU_AUTHSTOP_G(x)    \
+	(((x) >> CPL_TX_SEC_PDU_AUTHSTOP_S) & \
+	 CPL_TX_SEC_PDU_AUTHSTOP_M)
+
+/* AuthInsrtOffset: offset in bytes for authentication insertion
+ * from end of the payload of this command (0-511 bytes)
+ */
+#define CPL_TX_SEC_PDU_AUTHINSERT_S 0
+#define CPL_TX_SEC_PDU_AUTHINSERT_M 0x1ff
+#define CPL_TX_SEC_PDU_AUTHINSERT_V(x)  ((x) << CPL_TX_SEC_PDU_AUTHINSERT_S)
+#define CPL_TX_SEC_PDU_AUTHINSERT_G(x)  \
+	(((x) >> CPL_TX_SEC_PDU_AUTHINSERT_S) & \
+	 CPL_TX_SEC_PDU_AUTHINSERT_M)
+
+struct cpl_rx_phys_dsgl {
+	__be32 op_to_tid;
+	__be32 pcirlxorder_to_noofsgentr;
+	struct rss_header rss_hdr_int;
+};
+
+#define CPL_RX_PHYS_DSGL_OPCODE_S       24
+#define CPL_RX_PHYS_DSGL_OPCODE_M       0xff
+#define CPL_RX_PHYS_DSGL_OPCODE_V(x)    ((x) << CPL_RX_PHYS_DSGL_OPCODE_S)
+#define CPL_RX_PHYS_DSGL_OPCODE_G(x)    \
+	(((x) >> CPL_RX_PHYS_DSGL_OPCODE_S) & CPL_RX_PHYS_DSGL_OPCODE_M)
+
+#define CPL_RX_PHYS_DSGL_ISRDMA_S       23
+#define CPL_RX_PHYS_DSGL_ISRDMA_M       0x1
+#define CPL_RX_PHYS_DSGL_ISRDMA_V(x)    ((x) << CPL_RX_PHYS_DSGL_ISRDMA_S)
+#define CPL_RX_PHYS_DSGL_ISRDMA_G(x)    \
+	(((x) >> CPL_RX_PHYS_DSGL_ISRDMA_S) & CPL_RX_PHYS_DSGL_ISRDMA_M)
+#define CPL_RX_PHYS_DSGL_ISRDMA_F       CPL_RX_PHYS_DSGL_ISRDMA_V(1U)
+
+#define CPL_RX_PHYS_DSGL_RSVD1_S        20
+#define CPL_RX_PHYS_DSGL_RSVD1_M        0x7
+#define CPL_RX_PHYS_DSGL_RSVD1_V(x)     ((x) << CPL_RX_PHYS_DSGL_RSVD1_S)
+#define CPL_RX_PHYS_DSGL_RSVD1_G(x)     \
+	(((x) >> CPL_RX_PHYS_DSGL_RSVD1_S) & \
+	 CPL_RX_PHYS_DSGL_RSVD1_M)
+
+#define CPL_RX_PHYS_DSGL_PCIRLXORDER_S          31
+#define CPL_RX_PHYS_DSGL_PCIRLXORDER_M          0x1
+#define CPL_RX_PHYS_DSGL_PCIRLXORDER_V(x)       \
+	((x) << CPL_RX_PHYS_DSGL_PCIRLXORDER_S)
+#define CPL_RX_PHYS_DSGL_PCIRLXORDER_G(x)       \
+	(((x) >> CPL_RX_PHYS_DSGL_PCIRLXORDER_S) & \
+	 CPL_RX_PHYS_DSGL_PCIRLXORDER_M)
+#define CPL_RX_PHYS_DSGL_PCIRLXORDER_F  CPL_RX_PHYS_DSGL_PCIRLXORDER_V(1U)
+
+#define CPL_RX_PHYS_DSGL_PCINOSNOOP_S           30
+#define CPL_RX_PHYS_DSGL_PCINOSNOOP_M           0x1
+#define CPL_RX_PHYS_DSGL_PCINOSNOOP_V(x)        \
+	((x) << CPL_RX_PHYS_DSGL_PCINOSNOOP_S)
+#define CPL_RX_PHYS_DSGL_PCINOSNOOP_G(x)        \
+	(((x) >> CPL_RX_PHYS_DSGL_PCINOSNOOP_S) & \
+	 CPL_RX_PHYS_DSGL_PCINOSNOOP_M)
+
+#define CPL_RX_PHYS_DSGL_PCINOSNOOP_F   CPL_RX_PHYS_DSGL_PCINOSNOOP_V(1U)
+
+#define CPL_RX_PHYS_DSGL_PCITPHNTENB_S          29
+#define CPL_RX_PHYS_DSGL_PCITPHNTENB_M          0x1
+#define CPL_RX_PHYS_DSGL_PCITPHNTENB_V(x)       \
+	((x) << CPL_RX_PHYS_DSGL_PCITPHNTENB_S)
+#define CPL_RX_PHYS_DSGL_PCITPHNTENB_G(x)       \
+	(((x) >> CPL_RX_PHYS_DSGL_PCITPHNTENB_S) & \
+	 CPL_RX_PHYS_DSGL_PCITPHNTENB_M)
+#define CPL_RX_PHYS_DSGL_PCITPHNTENB_F  CPL_RX_PHYS_DSGL_PCITPHNTENB_V(1U)
+
+#define CPL_RX_PHYS_DSGL_PCITPHNT_S     27
+#define CPL_RX_PHYS_DSGL_PCITPHNT_M     0x3
+#define CPL_RX_PHYS_DSGL_PCITPHNT_V(x)  ((x) << CPL_RX_PHYS_DSGL_PCITPHNT_S)
+#define CPL_RX_PHYS_DSGL_PCITPHNT_G(x)  \
+	(((x) >> CPL_RX_PHYS_DSGL_PCITPHNT_S) & \
+	 CPL_RX_PHYS_DSGL_PCITPHNT_M)
+
+#define CPL_RX_PHYS_DSGL_DCAID_S        16
+#define CPL_RX_PHYS_DSGL_DCAID_M        0x7ff
+#define CPL_RX_PHYS_DSGL_DCAID_V(x)     ((x) << CPL_RX_PHYS_DSGL_DCAID_S)
+#define CPL_RX_PHYS_DSGL_DCAID_G(x)     \
+	(((x) >> CPL_RX_PHYS_DSGL_DCAID_S) & \
+	 CPL_RX_PHYS_DSGL_DCAID_M)
+
+#define CPL_RX_PHYS_DSGL_NOOFSGENTR_S           0
+#define CPL_RX_PHYS_DSGL_NOOFSGENTR_M           0xffff
+#define CPL_RX_PHYS_DSGL_NOOFSGENTR_V(x)        \
+	((x) << CPL_RX_PHYS_DSGL_NOOFSGENTR_S)
+#define CPL_RX_PHYS_DSGL_NOOFSGENTR_G(x)        \
+	(((x) >> CPL_RX_PHYS_DSGL_NOOFSGENTR_S) & \
+	 CPL_RX_PHYS_DSGL_NOOFSGENTR_M)
+
+struct cpl_rx_mps_pkt {
+	__be32 op_to_r1_hi;
+	__be32 r1_lo_length;
+};
+
+#define CPL_RX_MPS_PKT_OP_S     24
+#define CPL_RX_MPS_PKT_OP_M     0xff
+#define CPL_RX_MPS_PKT_OP_V(x)  ((x) << CPL_RX_MPS_PKT_OP_S)
+#define CPL_RX_MPS_PKT_OP_G(x)  \
+	(((x) >> CPL_RX_MPS_PKT_OP_S) & CPL_RX_MPS_PKT_OP_M)
+
+#define CPL_RX_MPS_PKT_TYPE_S           20
+#define CPL_RX_MPS_PKT_TYPE_M           0xf
+#define CPL_RX_MPS_PKT_TYPE_V(x)        ((x) << CPL_RX_MPS_PKT_TYPE_S)
+#define CPL_RX_MPS_PKT_TYPE_G(x)        \
+	(((x) >> CPL_RX_MPS_PKT_TYPE_S) & CPL_RX_MPS_PKT_TYPE_M)
+
+enum {
+	X_CPL_RX_MPS_PKT_TYPE_PAUSE = 1 << 0,
+	X_CPL_RX_MPS_PKT_TYPE_PPP   = 1 << 1,
+	X_CPL_RX_MPS_PKT_TYPE_QFC   = 1 << 2,
+	X_CPL_RX_MPS_PKT_TYPE_PTP   = 1 << 3
+};
+
+struct cpl_srq_table_req {
+	WR_HDR;
+	union opcode_tid ot;
+	__u8 status;
+	__u8 rsvd[2];
+	__u8 idx;
+	__be64 rsvd_pdid;
+	__be32 qlen_qbase;
+	__be16 cur_msn;
+	__be16 max_msn;
+};
+
+struct cpl_srq_table_rpl {
+	union opcode_tid ot;
+	__u8 status;
+	__u8 rsvd[2];
+	__u8 idx;
+	__be64 rsvd_pdid;
+	__be32 qlen_qbase;
+	__be16 cur_msn;
+	__be16 max_msn;
+};
+
+/* cpl_srq_table_{req,rpl}.params fields */
+#define SRQT_QLEN_S   28
+#define SRQT_QLEN_M   0xF
+#define SRQT_QLEN_V(x) ((x) << SRQT_QLEN_S)
+#define SRQT_QLEN_G(x) (((x) >> SRQT_QLEN_S) & SRQT_QLEN_M)
+
+#define SRQT_QBASE_S    0
+#define SRQT_QBASE_M   0x3FFFFFF
+#define SRQT_QBASE_V(x) ((x) << SRQT_QBASE_S)
+#define SRQT_QBASE_G(x) (((x) >> SRQT_QBASE_S) & SRQT_QBASE_M)
+
+#define SRQT_PDID_S    0
+#define SRQT_PDID_M   0xFF
+#define SRQT_PDID_V(x) ((x) << SRQT_PDID_S)
+#define SRQT_PDID_G(x) (((x) >> SRQT_PDID_S) & SRQT_PDID_M)
+
+#define SRQT_IDX_S    0
+#define SRQT_IDX_M    0xF
+#define SRQT_IDX_V(x) ((x) << SRQT_IDX_S)
+#define SRQT_IDX_G(x) (((x) >> SRQT_IDX_S) & SRQT_IDX_M)
+
+struct cpl_tx_tls_sfo {
+	__be32 op_to_seg_len;
+	__be32 pld_len;
+	__be32 type_protover;
+	__be32 r1_lo;
+	__be32 seqno_numivs;
+	__be32 ivgen_hdrlen;
+	__be64 scmd1;
+};
+
+/* cpl_tx_tls_sfo macros */
+#define CPL_TX_TLS_SFO_OPCODE_S         24
+#define CPL_TX_TLS_SFO_OPCODE_V(x)      ((x) << CPL_TX_TLS_SFO_OPCODE_S)
+
+#define CPL_TX_TLS_SFO_DATA_TYPE_S      20
+#define CPL_TX_TLS_SFO_DATA_TYPE_V(x)   ((x) << CPL_TX_TLS_SFO_DATA_TYPE_S)
+
+#define CPL_TX_TLS_SFO_CPL_LEN_S        16
+#define CPL_TX_TLS_SFO_CPL_LEN_V(x)     ((x) << CPL_TX_TLS_SFO_CPL_LEN_S)
+
+#define CPL_TX_TLS_SFO_SEG_LEN_S        0
+#define CPL_TX_TLS_SFO_SEG_LEN_M        0xffff
+#define CPL_TX_TLS_SFO_SEG_LEN_V(x)     ((x) << CPL_TX_TLS_SFO_SEG_LEN_S)
+#define CPL_TX_TLS_SFO_SEG_LEN_G(x)     \
+	(((x) >> CPL_TX_TLS_SFO_SEG_LEN_S) & CPL_TX_TLS_SFO_SEG_LEN_M)
+
+#define CPL_TX_TLS_SFO_TYPE_S           24
+#define CPL_TX_TLS_SFO_TYPE_M           0xff
+#define CPL_TX_TLS_SFO_TYPE_V(x)        ((x) << CPL_TX_TLS_SFO_TYPE_S)
+#define CPL_TX_TLS_SFO_TYPE_G(x)        \
+	(((x) >> CPL_TX_TLS_SFO_TYPE_S) & CPL_TX_TLS_SFO_TYPE_M)
+
+#define CPL_TX_TLS_SFO_PROTOVER_S       8
+#define CPL_TX_TLS_SFO_PROTOVER_M       0xffff
+#define CPL_TX_TLS_SFO_PROTOVER_V(x)    ((x) << CPL_TX_TLS_SFO_PROTOVER_S)
+#define CPL_TX_TLS_SFO_PROTOVER_G(x)    \
+	(((x) >> CPL_TX_TLS_SFO_PROTOVER_S) & CPL_TX_TLS_SFO_PROTOVER_M)
+
+struct cpl_tls_data {
+	struct rss_header rsshdr;
+	union opcode_tid ot;
+	__be32 length_pkd;
+	__be32 seq;
+	__be32 r1;
+};
+
+#define CPL_TLS_DATA_OPCODE_S           24
+#define CPL_TLS_DATA_OPCODE_M           0xff
+#define CPL_TLS_DATA_OPCODE_V(x)        ((x) << CPL_TLS_DATA_OPCODE_S)
+#define CPL_TLS_DATA_OPCODE_G(x)        \
+	(((x) >> CPL_TLS_DATA_OPCODE_S) & CPL_TLS_DATA_OPCODE_M)
+
+#define CPL_TLS_DATA_TID_S              0
+#define CPL_TLS_DATA_TID_M              0xffffff
+#define CPL_TLS_DATA_TID_V(x)           ((x) << CPL_TLS_DATA_TID_S)
+#define CPL_TLS_DATA_TID_G(x)           \
+	(((x) >> CPL_TLS_DATA_TID_S) & CPL_TLS_DATA_TID_M)
+
+#define CPL_TLS_DATA_LENGTH_S           0
+#define CPL_TLS_DATA_LENGTH_M           0xffff
+#define CPL_TLS_DATA_LENGTH_V(x)        ((x) << CPL_TLS_DATA_LENGTH_S)
+#define CPL_TLS_DATA_LENGTH_G(x)        \
+	(((x) >> CPL_TLS_DATA_LENGTH_S) & CPL_TLS_DATA_LENGTH_M)
+
+struct cpl_rx_tls_cmp {
+	struct rss_header rsshdr;
+	union opcode_tid ot;
+	__be32 pdulength_length;
+	__be32 seq;
+	__be32 ddp_report;
+	__be32 r;
+	__be32 ddp_valid;
+};
+
+#define CPL_RX_TLS_CMP_OPCODE_S         24
+#define CPL_RX_TLS_CMP_OPCODE_M         0xff
+#define CPL_RX_TLS_CMP_OPCODE_V(x)      ((x) << CPL_RX_TLS_CMP_OPCODE_S)
+#define CPL_RX_TLS_CMP_OPCODE_G(x)      \
+	(((x) >> CPL_RX_TLS_CMP_OPCODE_S) & CPL_RX_TLS_CMP_OPCODE_M)
+
+#define CPL_RX_TLS_CMP_TID_S            0
+#define CPL_RX_TLS_CMP_TID_M            0xffffff
+#define CPL_RX_TLS_CMP_TID_V(x)         ((x) << CPL_RX_TLS_CMP_TID_S)
+#define CPL_RX_TLS_CMP_TID_G(x)         \
+	(((x) >> CPL_RX_TLS_CMP_TID_S) & CPL_RX_TLS_CMP_TID_M)
+
+#define CPL_RX_TLS_CMP_PDULENGTH_S      16
+#define CPL_RX_TLS_CMP_PDULENGTH_M      0xffff
+#define CPL_RX_TLS_CMP_PDULENGTH_V(x)   ((x) << CPL_RX_TLS_CMP_PDULENGTH_S)
+#define CPL_RX_TLS_CMP_PDULENGTH_G(x)   \
+	(((x) >> CPL_RX_TLS_CMP_PDULENGTH_S) & CPL_RX_TLS_CMP_PDULENGTH_M)
+
+#define CPL_RX_TLS_CMP_LENGTH_S         0
+#define CPL_RX_TLS_CMP_LENGTH_M         0xffff
+#define CPL_RX_TLS_CMP_LENGTH_V(x)      ((x) << CPL_RX_TLS_CMP_LENGTH_S)
+#define CPL_RX_TLS_CMP_LENGTH_G(x)      \
+	(((x) >> CPL_RX_TLS_CMP_LENGTH_S) & CPL_RX_TLS_CMP_LENGTH_M)
+#endif  /* __T4_MSG_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_pci_id_tbl.h b/drivers/net/ethernet/chelsio/cxgb4/t4_pci_id_tbl.h
new file mode 100644
index 000000000..0b1b5f9c6
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_pci_id_tbl.h
@@ -0,0 +1,225 @@
+/*
+ * This file is part of the Chelsio T4/T5 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef __T4_PCI_ID_TBL_H__
+#define __T4_PCI_ID_TBL_H__
+
+/* The code can defined cpp macros for creating a PCI Device ID Table. This is
+ * useful because it allows the PCI ID Table to be maintained in a single place.
+ *
+ * The macros are:
+ *
+ * CH_PCI_DEVICE_ID_TABLE_DEFINE_BEGIN
+ *   -- Used to start the definition of the PCI ID Table.
+ *
+ * CH_PCI_DEVICE_ID_FUNCTION
+ *   -- The PCI Function Number to use in the PCI Device ID Table.  "0"
+ *   -- for drivers attaching to PF0-3, "4" for drivers attaching to PF4,
+ *   -- "8" for drivers attaching to SR-IOV Virtual Functions, etc.
+ *
+ * CH_PCI_DEVICE_ID_FUNCTION2 [optional]
+ *   -- If defined, create a PCI Device ID Table with both
+ *   -- CH_PCI_DEVICE_ID_FUNCTION and CH_PCI_DEVICE_ID_FUNCTION2 populated.
+ *
+ * CH_PCI_ID_TABLE_ENTRY(DeviceID)
+ *   -- Used for the individual PCI Device ID entries.  Note that we will
+ *   -- be adding a trailing comma (",") after all of the entries (and
+ *   -- between the pairs of entries if CH_PCI_DEVICE_ID_FUNCTION2 is defined).
+ *
+ * CH_PCI_DEVICE_ID_TABLE_DEFINE_END
+ *   -- Used to finish the definition of the PCI ID Table.  Note that we
+ *   -- will be adding a trailing semi-colon (";") here.
+ */
+#ifndef CH_PCI_DEVICE_ID_FUNCTION
+#error CH_PCI_DEVICE_ID_FUNCTION not defined!
+#endif
+#ifndef CH_PCI_ID_TABLE_ENTRY
+#error CH_PCI_ID_TABLE_ENTRY not defined!
+#endif
+#ifndef CH_PCI_DEVICE_ID_TABLE_DEFINE_END
+#error CH_PCI_DEVICE_ID_TABLE_DEFINE_END not defined!
+#endif
+
+/* T4 and later ASICs use a PCI Device ID scheme of 0xVFPP where:
+ *
+ *   V  = "4" for T4; "5" for T5, etc.
+ *   F  = "0" for PF 0..3; "4".."7" for PF4..7; and "8" for VFs
+ *   PP = adapter product designation
+ *
+ * We use this consistency in order to create the proper PCI Device IDs
+ * for the specified CH_PCI_DEVICE_ID_FUNCTION.
+ */
+#ifndef CH_PCI_DEVICE_ID_FUNCTION2
+#define CH_PCI_ID_TABLE_FENTRY(devid) \
+	CH_PCI_ID_TABLE_ENTRY((devid) | \
+			      ((CH_PCI_DEVICE_ID_FUNCTION) << 8))
+#else
+#define CH_PCI_ID_TABLE_FENTRY(devid) \
+	CH_PCI_ID_TABLE_ENTRY((devid) | \
+			      ((CH_PCI_DEVICE_ID_FUNCTION) << 8)), \
+	CH_PCI_ID_TABLE_ENTRY((devid) | \
+			      ((CH_PCI_DEVICE_ID_FUNCTION2) << 8))
+#endif
+
+CH_PCI_DEVICE_ID_TABLE_DEFINE_BEGIN
+	/* T4 adapters:
+	 */
+	CH_PCI_ID_TABLE_FENTRY(0x4000),	/* T440-dbg */
+	CH_PCI_ID_TABLE_FENTRY(0x4001),	/* T420-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x4002),	/* T422-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x4003),	/* T440-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x4004),	/* T420-bch */
+	CH_PCI_ID_TABLE_FENTRY(0x4005),	/* T440-bch */
+	CH_PCI_ID_TABLE_FENTRY(0x4006),	/* T440-ch */
+	CH_PCI_ID_TABLE_FENTRY(0x4007),	/* T420-so */
+	CH_PCI_ID_TABLE_FENTRY(0x4008),	/* T420-cx */
+	CH_PCI_ID_TABLE_FENTRY(0x4009),	/* T420-bt */
+	CH_PCI_ID_TABLE_FENTRY(0x400a),	/* T404-bt */
+	CH_PCI_ID_TABLE_FENTRY(0x400b),	/* B420-sr */
+	CH_PCI_ID_TABLE_FENTRY(0x400c),	/* B404-bt */
+	CH_PCI_ID_TABLE_FENTRY(0x400d),	/* T480-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x400e),	/* T440-LP-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x4080),	/* Custom T480-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x4081),	/* Custom T440-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x4082),	/* Custom T420-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x4083),	/* Custom T420-xaui */
+	CH_PCI_ID_TABLE_FENTRY(0x4084),	/* Custom T440-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x4085),	/* Custom T420-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x4086),	/* Custom T440-bt */
+	CH_PCI_ID_TABLE_FENTRY(0x4087),	/* Custom T440-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x4088),	/* Custom T440 2-xaui, 2-xfi */
+
+	/* T5 adapters:
+	 */
+	CH_PCI_ID_TABLE_FENTRY(0x5000),	/* T580-dbg */
+	CH_PCI_ID_TABLE_FENTRY(0x5001),	/* T520-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x5002),	/* T522-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x5003),	/* T540-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x5004),	/* T520-bch */
+	CH_PCI_ID_TABLE_FENTRY(0x5005),	/* T540-bch */
+	CH_PCI_ID_TABLE_FENTRY(0x5006),	/* T540-ch */
+	CH_PCI_ID_TABLE_FENTRY(0x5007),	/* T520-so */
+	CH_PCI_ID_TABLE_FENTRY(0x5008),	/* T520-cx */
+	CH_PCI_ID_TABLE_FENTRY(0x5009),	/* T520-bt */
+	CH_PCI_ID_TABLE_FENTRY(0x500a),	/* T504-bt */
+	CH_PCI_ID_TABLE_FENTRY(0x500b),	/* B520-sr */
+	CH_PCI_ID_TABLE_FENTRY(0x500c),	/* B504-bt */
+	CH_PCI_ID_TABLE_FENTRY(0x500d),	/* T580-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x500e),	/* T540-LP-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x5010),	/* T580-LP-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x5011),	/* T520-LL-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x5012),	/* T560-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x5013),	/* T580-chr */
+	CH_PCI_ID_TABLE_FENTRY(0x5014),	/* T580-so */
+	CH_PCI_ID_TABLE_FENTRY(0x5015),	/* T502-bt */
+	CH_PCI_ID_TABLE_FENTRY(0x5016),	/* T580-OCP-SO */
+	CH_PCI_ID_TABLE_FENTRY(0x5017),	/* T520-OCP-SO */
+	CH_PCI_ID_TABLE_FENTRY(0x5018),	/* T540-BT */
+	CH_PCI_ID_TABLE_FENTRY(0x5019),	/* T540-LP-BT */
+	CH_PCI_ID_TABLE_FENTRY(0x501a),	/* T540-SO-BT */
+	CH_PCI_ID_TABLE_FENTRY(0x501b),	/* T540-SO-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5080),	/* Custom T540-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x5081),	/* Custom T540-LL-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x5082),	/* Custom T504-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x5083),	/* Custom T540-LP-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5084),	/* Custom T580-cr */
+	CH_PCI_ID_TABLE_FENTRY(0x5085),	/* Custom 3x T580-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5086),	/* Custom 2x T580-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5087),	/* Custom T580-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5088),	/* Custom T570-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5089),	/* Custom T520-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5090),	/* Custom T540-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5091),	/* Custom T522-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5092),	/* Custom T520-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5093),	/* Custom T580-LP-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5094),	/* Custom T540-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5095),	/* Custom T540-CR-SO */
+	CH_PCI_ID_TABLE_FENTRY(0x5096),	/* Custom T580-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x5097),	/* Custom T520-KR */
+	CH_PCI_ID_TABLE_FENTRY(0x5098),	/* Custom 2x40G QSFP */
+	CH_PCI_ID_TABLE_FENTRY(0x5099),	/* Custom 2x40G QSFP */
+	CH_PCI_ID_TABLE_FENTRY(0x509a),	/* Custom T520-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x509b),	/* Custom T540-CR LOM */
+	CH_PCI_ID_TABLE_FENTRY(0x509c),	/* Custom T520-CR*/
+	CH_PCI_ID_TABLE_FENTRY(0x509d),	/* Custom T540-CR*/
+	CH_PCI_ID_TABLE_FENTRY(0x509e), /* Custom T520-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x509f), /* Custom T540-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x50a0), /* Custom T540-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x50a1), /* Custom T540-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x50a2), /* Custom T540-KR4 */
+	CH_PCI_ID_TABLE_FENTRY(0x50a3), /* Custom T580-KR4 */
+	CH_PCI_ID_TABLE_FENTRY(0x50a4), /* Custom 2x T540-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x50a5), /* Custom T522-BT */
+	CH_PCI_ID_TABLE_FENTRY(0x50a6), /* Custom T522-BT-SO */
+	CH_PCI_ID_TABLE_FENTRY(0x50a7), /* Custom T580-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x50a8), /* Custom T580-KR */
+	CH_PCI_ID_TABLE_FENTRY(0x50a9), /* Custom T580-KR */
+	CH_PCI_ID_TABLE_FENTRY(0x50aa), /* Custom T580-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x50ab), /* Custom T520-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x50ac), /* Custom T540-BT */
+	CH_PCI_ID_TABLE_FENTRY(0x50ad), /* Custom T520-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x50ae), /* Custom T540-XL-SO */
+	CH_PCI_ID_TABLE_FENTRY(0x50af), /* Custom T580-KR-SO */
+	CH_PCI_ID_TABLE_FENTRY(0x50b0), /* Custom T520-CR-LOM */
+
+	/* T6 adapters:
+	 */
+	CH_PCI_ID_TABLE_FENTRY(0x6001),
+	CH_PCI_ID_TABLE_FENTRY(0x6002),
+	CH_PCI_ID_TABLE_FENTRY(0x6003),
+	CH_PCI_ID_TABLE_FENTRY(0x6004),
+	CH_PCI_ID_TABLE_FENTRY(0x6005),
+	CH_PCI_ID_TABLE_FENTRY(0x6006),
+	CH_PCI_ID_TABLE_FENTRY(0x6007),
+	CH_PCI_ID_TABLE_FENTRY(0x6008),
+	CH_PCI_ID_TABLE_FENTRY(0x6009),
+	CH_PCI_ID_TABLE_FENTRY(0x600d),
+	CH_PCI_ID_TABLE_FENTRY(0x6011),
+	CH_PCI_ID_TABLE_FENTRY(0x6014),
+	CH_PCI_ID_TABLE_FENTRY(0x6015),
+	CH_PCI_ID_TABLE_FENTRY(0x6080),
+	CH_PCI_ID_TABLE_FENTRY(0x6081),
+	CH_PCI_ID_TABLE_FENTRY(0x6082), /* Custom T6225-CR SFP28 */
+	CH_PCI_ID_TABLE_FENTRY(0x6083), /* Custom T62100-CR QSFP28 */
+	CH_PCI_ID_TABLE_FENTRY(0x6084), /* Custom T64100-CR QSFP28 */
+	CH_PCI_ID_TABLE_FENTRY(0x6085), /* Custom T6240-SO */
+	CH_PCI_ID_TABLE_FENTRY(0x6086), /* Custom T6225-SO-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x6087), /* Custom T6225-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x6088), /* Custom T62100-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x6089), /* Custom T62100-KR */
+	CH_PCI_ID_TABLE_FENTRY(0x608a), /* Custom T62100-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x608b), /* Custom T6225-CR */
+	CH_PCI_ID_TABLE_FENTRY(0x6092), /* Custom T62100-CR-LOM */
+CH_PCI_DEVICE_ID_TABLE_DEFINE_END;
+
+#endif /* __T4_PCI_ID_TBL_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h b/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
new file mode 100644
index 000000000..695916ba0
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
@@ -0,0 +1,3386 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __T4_REGS_H
+#define __T4_REGS_H
+
+#define MYPF_BASE 0x1b000
+#define MYPF_REG(reg_addr) (MYPF_BASE + (reg_addr))
+
+#define PF0_BASE 0x1e000
+#define PF0_REG(reg_addr) (PF0_BASE + (reg_addr))
+
+#define PF_STRIDE 0x400
+#define PF_BASE(idx) (PF0_BASE + (idx) * PF_STRIDE)
+#define PF_REG(idx, reg) (PF_BASE(idx) + (reg))
+
+#define NUM_CIM_CTL_TSCH_CHANNEL_INSTANCES 4
+#define NUM_CIM_CTL_TSCH_CHANNEL_TSCH_CLASS_INSTANCES 16
+
+#define MYPORT_BASE 0x1c000
+#define MYPORT_REG(reg_addr) (MYPORT_BASE + (reg_addr))
+
+#define PORT0_BASE 0x20000
+#define PORT0_REG(reg_addr) (PORT0_BASE + (reg_addr))
+
+#define PORT_STRIDE 0x2000
+#define PORT_BASE(idx) (PORT0_BASE + (idx) * PORT_STRIDE)
+#define PORT_REG(idx, reg) (PORT_BASE(idx) + (reg))
+
+#define EDC_STRIDE (EDC_1_BASE_ADDR - EDC_0_BASE_ADDR)
+#define EDC_REG(reg, idx) (reg + EDC_STRIDE * idx)
+
+#define PCIE_MEM_ACCESS_REG(reg_addr, idx) ((reg_addr) + (idx) * 8)
+#define PCIE_MAILBOX_REG(reg_addr, idx) ((reg_addr) + (idx) * 8)
+#define MC_BIST_STATUS_REG(reg_addr, idx) ((reg_addr) + (idx) * 4)
+#define EDC_BIST_STATUS_REG(reg_addr, idx) ((reg_addr) + (idx) * 4)
+
+#define PCIE_FW_REG(reg_addr, idx) ((reg_addr) + (idx) * 4)
+
+#define NUM_LE_DB_DBGI_REQ_DATA_INSTANCES 17
+#define NUM_LE_DB_DBGI_RSP_DATA_INSTANCES 17
+
+#define SGE_PF_KDOORBELL_A 0x0
+
+#define QID_S    15
+#define QID_V(x) ((x) << QID_S)
+
+#define DBPRIO_S    14
+#define DBPRIO_V(x) ((x) << DBPRIO_S)
+#define DBPRIO_F    DBPRIO_V(1U)
+
+#define PIDX_S    0
+#define PIDX_V(x) ((x) << PIDX_S)
+
+#define SGE_VF_KDOORBELL_A 0x0
+
+#define DBTYPE_S    13
+#define DBTYPE_V(x) ((x) << DBTYPE_S)
+#define DBTYPE_F    DBTYPE_V(1U)
+
+#define PIDX_T5_S    0
+#define PIDX_T5_M    0x1fffU
+#define PIDX_T5_V(x) ((x) << PIDX_T5_S)
+#define PIDX_T5_G(x) (((x) >> PIDX_T5_S) & PIDX_T5_M)
+
+#define SGE_PF_GTS_A 0x4
+
+#define INGRESSQID_S    16
+#define INGRESSQID_V(x) ((x) << INGRESSQID_S)
+
+#define TIMERREG_S    13
+#define TIMERREG_V(x) ((x) << TIMERREG_S)
+
+#define SEINTARM_S    12
+#define SEINTARM_V(x) ((x) << SEINTARM_S)
+
+#define CIDXINC_S    0
+#define CIDXINC_M    0xfffU
+#define CIDXINC_V(x) ((x) << CIDXINC_S)
+
+#define SGE_CONTROL_A	0x1008
+#define SGE_CONTROL2_A	0x1124
+
+#define RXPKTCPLMODE_S    18
+#define RXPKTCPLMODE_V(x) ((x) << RXPKTCPLMODE_S)
+#define RXPKTCPLMODE_F    RXPKTCPLMODE_V(1U)
+
+#define EGRSTATUSPAGESIZE_S    17
+#define EGRSTATUSPAGESIZE_V(x) ((x) << EGRSTATUSPAGESIZE_S)
+#define EGRSTATUSPAGESIZE_F    EGRSTATUSPAGESIZE_V(1U)
+
+#define PKTSHIFT_S    10
+#define PKTSHIFT_M    0x7U
+#define PKTSHIFT_V(x) ((x) << PKTSHIFT_S)
+#define PKTSHIFT_G(x) (((x) >> PKTSHIFT_S) & PKTSHIFT_M)
+
+#define INGPCIEBOUNDARY_S    7
+#define INGPCIEBOUNDARY_V(x) ((x) << INGPCIEBOUNDARY_S)
+
+#define INGPADBOUNDARY_S    4
+#define INGPADBOUNDARY_M    0x7U
+#define INGPADBOUNDARY_V(x) ((x) << INGPADBOUNDARY_S)
+#define INGPADBOUNDARY_G(x) (((x) >> INGPADBOUNDARY_S) & INGPADBOUNDARY_M)
+
+#define EGRPCIEBOUNDARY_S    1
+#define EGRPCIEBOUNDARY_V(x) ((x) << EGRPCIEBOUNDARY_S)
+
+#define  INGPACKBOUNDARY_S	16
+#define  INGPACKBOUNDARY_M	0x7U
+#define  INGPACKBOUNDARY_V(x)	((x) << INGPACKBOUNDARY_S)
+#define  INGPACKBOUNDARY_G(x)	(((x) >> INGPACKBOUNDARY_S) \
+				 & INGPACKBOUNDARY_M)
+
+#define VFIFO_ENABLE_S    10
+#define VFIFO_ENABLE_V(x) ((x) << VFIFO_ENABLE_S)
+#define VFIFO_ENABLE_F    VFIFO_ENABLE_V(1U)
+
+#define SGE_DBVFIFO_BADDR_A 0x1138
+
+#define DBVFIFO_SIZE_S    6
+#define DBVFIFO_SIZE_M    0xfffU
+#define DBVFIFO_SIZE_G(x) (((x) >> DBVFIFO_SIZE_S) & DBVFIFO_SIZE_M)
+
+#define T6_DBVFIFO_SIZE_S    0
+#define T6_DBVFIFO_SIZE_M    0x1fffU
+#define T6_DBVFIFO_SIZE_G(x) (((x) >> T6_DBVFIFO_SIZE_S) & T6_DBVFIFO_SIZE_M)
+
+#define SGE_CTXT_CMD_A 0x11fc
+
+#define BUSY_S    31
+#define BUSY_V(x) ((x) << BUSY_S)
+#define BUSY_F    BUSY_V(1U)
+
+#define CTXTTYPE_S    24
+#define CTXTTYPE_M    0x3U
+#define CTXTTYPE_V(x) ((x) << CTXTTYPE_S)
+
+#define CTXTQID_S    0
+#define CTXTQID_M    0x1ffffU
+#define CTXTQID_V(x) ((x) << CTXTQID_S)
+
+#define SGE_CTXT_DATA0_A 0x1200
+#define SGE_CTXT_DATA5_A 0x1214
+
+#define GLOBALENABLE_S    0
+#define GLOBALENABLE_V(x) ((x) << GLOBALENABLE_S)
+#define GLOBALENABLE_F    GLOBALENABLE_V(1U)
+
+#define SGE_HOST_PAGE_SIZE_A 0x100c
+
+#define HOSTPAGESIZEPF7_S    28
+#define HOSTPAGESIZEPF7_M    0xfU
+#define HOSTPAGESIZEPF7_V(x) ((x) << HOSTPAGESIZEPF7_S)
+#define HOSTPAGESIZEPF7_G(x) (((x) >> HOSTPAGESIZEPF7_S) & HOSTPAGESIZEPF7_M)
+
+#define HOSTPAGESIZEPF6_S    24
+#define HOSTPAGESIZEPF6_M    0xfU
+#define HOSTPAGESIZEPF6_V(x) ((x) << HOSTPAGESIZEPF6_S)
+#define HOSTPAGESIZEPF6_G(x) (((x) >> HOSTPAGESIZEPF6_S) & HOSTPAGESIZEPF6_M)
+
+#define HOSTPAGESIZEPF5_S    20
+#define HOSTPAGESIZEPF5_M    0xfU
+#define HOSTPAGESIZEPF5_V(x) ((x) << HOSTPAGESIZEPF5_S)
+#define HOSTPAGESIZEPF5_G(x) (((x) >> HOSTPAGESIZEPF5_S) & HOSTPAGESIZEPF5_M)
+
+#define HOSTPAGESIZEPF4_S    16
+#define HOSTPAGESIZEPF4_M    0xfU
+#define HOSTPAGESIZEPF4_V(x) ((x) << HOSTPAGESIZEPF4_S)
+#define HOSTPAGESIZEPF4_G(x) (((x) >> HOSTPAGESIZEPF4_S) & HOSTPAGESIZEPF4_M)
+
+#define HOSTPAGESIZEPF3_S    12
+#define HOSTPAGESIZEPF3_M    0xfU
+#define HOSTPAGESIZEPF3_V(x) ((x) << HOSTPAGESIZEPF3_S)
+#define HOSTPAGESIZEPF3_G(x) (((x) >> HOSTPAGESIZEPF3_S) & HOSTPAGESIZEPF3_M)
+
+#define HOSTPAGESIZEPF2_S    8
+#define HOSTPAGESIZEPF2_M    0xfU
+#define HOSTPAGESIZEPF2_V(x) ((x) << HOSTPAGESIZEPF2_S)
+#define HOSTPAGESIZEPF2_G(x) (((x) >> HOSTPAGESIZEPF2_S) & HOSTPAGESIZEPF2_M)
+
+#define HOSTPAGESIZEPF1_S    4
+#define HOSTPAGESIZEPF1_M    0xfU
+#define HOSTPAGESIZEPF1_V(x) ((x) << HOSTPAGESIZEPF1_S)
+#define HOSTPAGESIZEPF1_G(x) (((x) >> HOSTPAGESIZEPF1_S) & HOSTPAGESIZEPF1_M)
+
+#define HOSTPAGESIZEPF0_S    0
+#define HOSTPAGESIZEPF0_M    0xfU
+#define HOSTPAGESIZEPF0_V(x) ((x) << HOSTPAGESIZEPF0_S)
+#define HOSTPAGESIZEPF0_G(x) (((x) >> HOSTPAGESIZEPF0_S) & HOSTPAGESIZEPF0_M)
+
+#define SGE_EGRESS_QUEUES_PER_PAGE_PF_A 0x1010
+#define SGE_EGRESS_QUEUES_PER_PAGE_VF_A 0x1014
+
+#define QUEUESPERPAGEPF1_S    4
+
+#define QUEUESPERPAGEPF0_S    0
+#define QUEUESPERPAGEPF0_M    0xfU
+#define QUEUESPERPAGEPF0_V(x) ((x) << QUEUESPERPAGEPF0_S)
+#define QUEUESPERPAGEPF0_G(x) (((x) >> QUEUESPERPAGEPF0_S) & QUEUESPERPAGEPF0_M)
+
+#define SGE_INT_CAUSE1_A	0x1024
+#define SGE_INT_CAUSE2_A	0x1030
+#define SGE_INT_CAUSE3_A	0x103c
+
+#define ERR_FLM_DBP_S    31
+#define ERR_FLM_DBP_V(x) ((x) << ERR_FLM_DBP_S)
+#define ERR_FLM_DBP_F    ERR_FLM_DBP_V(1U)
+
+#define ERR_FLM_IDMA1_S    30
+#define ERR_FLM_IDMA1_V(x) ((x) << ERR_FLM_IDMA1_S)
+#define ERR_FLM_IDMA1_F    ERR_FLM_IDMA1_V(1U)
+
+#define ERR_FLM_IDMA0_S    29
+#define ERR_FLM_IDMA0_V(x) ((x) << ERR_FLM_IDMA0_S)
+#define ERR_FLM_IDMA0_F    ERR_FLM_IDMA0_V(1U)
+
+#define ERR_FLM_HINT_S    28
+#define ERR_FLM_HINT_V(x) ((x) << ERR_FLM_HINT_S)
+#define ERR_FLM_HINT_F    ERR_FLM_HINT_V(1U)
+
+#define ERR_PCIE_ERROR3_S    27
+#define ERR_PCIE_ERROR3_V(x) ((x) << ERR_PCIE_ERROR3_S)
+#define ERR_PCIE_ERROR3_F    ERR_PCIE_ERROR3_V(1U)
+
+#define ERR_PCIE_ERROR2_S    26
+#define ERR_PCIE_ERROR2_V(x) ((x) << ERR_PCIE_ERROR2_S)
+#define ERR_PCIE_ERROR2_F    ERR_PCIE_ERROR2_V(1U)
+
+#define ERR_PCIE_ERROR1_S    25
+#define ERR_PCIE_ERROR1_V(x) ((x) << ERR_PCIE_ERROR1_S)
+#define ERR_PCIE_ERROR1_F    ERR_PCIE_ERROR1_V(1U)
+
+#define ERR_PCIE_ERROR0_S    24
+#define ERR_PCIE_ERROR0_V(x) ((x) << ERR_PCIE_ERROR0_S)
+#define ERR_PCIE_ERROR0_F    ERR_PCIE_ERROR0_V(1U)
+
+#define ERR_CPL_EXCEED_IQE_SIZE_S    22
+#define ERR_CPL_EXCEED_IQE_SIZE_V(x) ((x) << ERR_CPL_EXCEED_IQE_SIZE_S)
+#define ERR_CPL_EXCEED_IQE_SIZE_F    ERR_CPL_EXCEED_IQE_SIZE_V(1U)
+
+#define ERR_INVALID_CIDX_INC_S    21
+#define ERR_INVALID_CIDX_INC_V(x) ((x) << ERR_INVALID_CIDX_INC_S)
+#define ERR_INVALID_CIDX_INC_F    ERR_INVALID_CIDX_INC_V(1U)
+
+#define ERR_CPL_OPCODE_0_S    19
+#define ERR_CPL_OPCODE_0_V(x) ((x) << ERR_CPL_OPCODE_0_S)
+#define ERR_CPL_OPCODE_0_F    ERR_CPL_OPCODE_0_V(1U)
+
+#define ERR_DROPPED_DB_S    18
+#define ERR_DROPPED_DB_V(x) ((x) << ERR_DROPPED_DB_S)
+#define ERR_DROPPED_DB_F    ERR_DROPPED_DB_V(1U)
+
+#define ERR_DATA_CPL_ON_HIGH_QID1_S    17
+#define ERR_DATA_CPL_ON_HIGH_QID1_V(x) ((x) << ERR_DATA_CPL_ON_HIGH_QID1_S)
+#define ERR_DATA_CPL_ON_HIGH_QID1_F    ERR_DATA_CPL_ON_HIGH_QID1_V(1U)
+
+#define ERR_DATA_CPL_ON_HIGH_QID0_S    16
+#define ERR_DATA_CPL_ON_HIGH_QID0_V(x) ((x) << ERR_DATA_CPL_ON_HIGH_QID0_S)
+#define ERR_DATA_CPL_ON_HIGH_QID0_F    ERR_DATA_CPL_ON_HIGH_QID0_V(1U)
+
+#define ERR_BAD_DB_PIDX3_S    15
+#define ERR_BAD_DB_PIDX3_V(x) ((x) << ERR_BAD_DB_PIDX3_S)
+#define ERR_BAD_DB_PIDX3_F    ERR_BAD_DB_PIDX3_V(1U)
+
+#define ERR_BAD_DB_PIDX2_S    14
+#define ERR_BAD_DB_PIDX2_V(x) ((x) << ERR_BAD_DB_PIDX2_S)
+#define ERR_BAD_DB_PIDX2_F    ERR_BAD_DB_PIDX2_V(1U)
+
+#define ERR_BAD_DB_PIDX1_S    13
+#define ERR_BAD_DB_PIDX1_V(x) ((x) << ERR_BAD_DB_PIDX1_S)
+#define ERR_BAD_DB_PIDX1_F    ERR_BAD_DB_PIDX1_V(1U)
+
+#define ERR_BAD_DB_PIDX0_S    12
+#define ERR_BAD_DB_PIDX0_V(x) ((x) << ERR_BAD_DB_PIDX0_S)
+#define ERR_BAD_DB_PIDX0_F    ERR_BAD_DB_PIDX0_V(1U)
+
+#define ERR_ING_CTXT_PRIO_S    10
+#define ERR_ING_CTXT_PRIO_V(x) ((x) << ERR_ING_CTXT_PRIO_S)
+#define ERR_ING_CTXT_PRIO_F    ERR_ING_CTXT_PRIO_V(1U)
+
+#define ERR_EGR_CTXT_PRIO_S    9
+#define ERR_EGR_CTXT_PRIO_V(x) ((x) << ERR_EGR_CTXT_PRIO_S)
+#define ERR_EGR_CTXT_PRIO_F    ERR_EGR_CTXT_PRIO_V(1U)
+
+#define DBFIFO_HP_INT_S    8
+#define DBFIFO_HP_INT_V(x) ((x) << DBFIFO_HP_INT_S)
+#define DBFIFO_HP_INT_F    DBFIFO_HP_INT_V(1U)
+
+#define DBFIFO_LP_INT_S    7
+#define DBFIFO_LP_INT_V(x) ((x) << DBFIFO_LP_INT_S)
+#define DBFIFO_LP_INT_F    DBFIFO_LP_INT_V(1U)
+
+#define INGRESS_SIZE_ERR_S    5
+#define INGRESS_SIZE_ERR_V(x) ((x) << INGRESS_SIZE_ERR_S)
+#define INGRESS_SIZE_ERR_F    INGRESS_SIZE_ERR_V(1U)
+
+#define EGRESS_SIZE_ERR_S    4
+#define EGRESS_SIZE_ERR_V(x) ((x) << EGRESS_SIZE_ERR_S)
+#define EGRESS_SIZE_ERR_F    EGRESS_SIZE_ERR_V(1U)
+
+#define SGE_INT_ENABLE3_A 0x1040
+#define SGE_FL_BUFFER_SIZE0_A 0x1044
+#define SGE_FL_BUFFER_SIZE1_A 0x1048
+#define SGE_FL_BUFFER_SIZE2_A 0x104c
+#define SGE_FL_BUFFER_SIZE3_A 0x1050
+#define SGE_FL_BUFFER_SIZE4_A 0x1054
+#define SGE_FL_BUFFER_SIZE5_A 0x1058
+#define SGE_FL_BUFFER_SIZE6_A 0x105c
+#define SGE_FL_BUFFER_SIZE7_A 0x1060
+#define SGE_FL_BUFFER_SIZE8_A 0x1064
+
+#define SGE_IMSG_CTXT_BADDR_A 0x1088
+#define SGE_FLM_CACHE_BADDR_A 0x108c
+#define SGE_FLM_CFG_A 0x1090
+
+#define NOHDR_S    18
+#define NOHDR_V(x) ((x) << NOHDR_S)
+#define NOHDR_F    NOHDR_V(1U)
+
+#define HDRSTARTFLQ_S    11
+#define HDRSTARTFLQ_M    0x7U
+#define HDRSTARTFLQ_G(x) (((x) >> HDRSTARTFLQ_S) & HDRSTARTFLQ_M)
+
+#define SGE_INGRESS_RX_THRESHOLD_A 0x10a0
+
+#define THRESHOLD_0_S    24
+#define THRESHOLD_0_M    0x3fU
+#define THRESHOLD_0_V(x) ((x) << THRESHOLD_0_S)
+#define THRESHOLD_0_G(x) (((x) >> THRESHOLD_0_S) & THRESHOLD_0_M)
+
+#define THRESHOLD_1_S    16
+#define THRESHOLD_1_M    0x3fU
+#define THRESHOLD_1_V(x) ((x) << THRESHOLD_1_S)
+#define THRESHOLD_1_G(x) (((x) >> THRESHOLD_1_S) & THRESHOLD_1_M)
+
+#define THRESHOLD_2_S    8
+#define THRESHOLD_2_M    0x3fU
+#define THRESHOLD_2_V(x) ((x) << THRESHOLD_2_S)
+#define THRESHOLD_2_G(x) (((x) >> THRESHOLD_2_S) & THRESHOLD_2_M)
+
+#define THRESHOLD_3_S    0
+#define THRESHOLD_3_M    0x3fU
+#define THRESHOLD_3_V(x) ((x) << THRESHOLD_3_S)
+#define THRESHOLD_3_G(x) (((x) >> THRESHOLD_3_S) & THRESHOLD_3_M)
+
+#define SGE_CONM_CTRL_A 0x1094
+
+#define EGRTHRESHOLD_S    8
+#define EGRTHRESHOLD_M    0x3fU
+#define EGRTHRESHOLD_V(x) ((x) << EGRTHRESHOLD_S)
+#define EGRTHRESHOLD_G(x) (((x) >> EGRTHRESHOLD_S) & EGRTHRESHOLD_M)
+
+#define EGRTHRESHOLDPACKING_S    14
+#define EGRTHRESHOLDPACKING_M    0x3fU
+#define EGRTHRESHOLDPACKING_V(x) ((x) << EGRTHRESHOLDPACKING_S)
+#define EGRTHRESHOLDPACKING_G(x) \
+	(((x) >> EGRTHRESHOLDPACKING_S) & EGRTHRESHOLDPACKING_M)
+
+#define T6_EGRTHRESHOLDPACKING_S    16
+#define T6_EGRTHRESHOLDPACKING_M    0xffU
+#define T6_EGRTHRESHOLDPACKING_G(x) \
+	(((x) >> T6_EGRTHRESHOLDPACKING_S) & T6_EGRTHRESHOLDPACKING_M)
+
+#define SGE_TIMESTAMP_LO_A 0x1098
+#define SGE_TIMESTAMP_HI_A 0x109c
+
+#define TSOP_S    28
+#define TSOP_M    0x3U
+#define TSOP_V(x) ((x) << TSOP_S)
+#define TSOP_G(x) (((x) >> TSOP_S) & TSOP_M)
+
+#define TSVAL_S    0
+#define TSVAL_M    0xfffffffU
+#define TSVAL_V(x) ((x) << TSVAL_S)
+#define TSVAL_G(x) (((x) >> TSVAL_S) & TSVAL_M)
+
+#define SGE_DBFIFO_STATUS_A 0x10a4
+#define SGE_DBVFIFO_SIZE_A 0x113c
+
+#define HP_INT_THRESH_S    28
+#define HP_INT_THRESH_M    0xfU
+#define HP_INT_THRESH_V(x) ((x) << HP_INT_THRESH_S)
+
+#define LP_INT_THRESH_S    12
+#define LP_INT_THRESH_M    0xfU
+#define LP_INT_THRESH_V(x) ((x) << LP_INT_THRESH_S)
+
+#define SGE_DOORBELL_CONTROL_A 0x10a8
+
+#define NOCOALESCE_S    26
+#define NOCOALESCE_V(x) ((x) << NOCOALESCE_S)
+#define NOCOALESCE_F    NOCOALESCE_V(1U)
+
+#define ENABLE_DROP_S    13
+#define ENABLE_DROP_V(x) ((x) << ENABLE_DROP_S)
+#define ENABLE_DROP_F    ENABLE_DROP_V(1U)
+
+#define SGE_TIMER_VALUE_0_AND_1_A 0x10b8
+
+#define TIMERVALUE0_S    16
+#define TIMERVALUE0_M    0xffffU
+#define TIMERVALUE0_V(x) ((x) << TIMERVALUE0_S)
+#define TIMERVALUE0_G(x) (((x) >> TIMERVALUE0_S) & TIMERVALUE0_M)
+
+#define TIMERVALUE1_S    0
+#define TIMERVALUE1_M    0xffffU
+#define TIMERVALUE1_V(x) ((x) << TIMERVALUE1_S)
+#define TIMERVALUE1_G(x) (((x) >> TIMERVALUE1_S) & TIMERVALUE1_M)
+
+#define SGE_TIMER_VALUE_2_AND_3_A 0x10bc
+
+#define TIMERVALUE2_S    16
+#define TIMERVALUE2_M    0xffffU
+#define TIMERVALUE2_V(x) ((x) << TIMERVALUE2_S)
+#define TIMERVALUE2_G(x) (((x) >> TIMERVALUE2_S) & TIMERVALUE2_M)
+
+#define TIMERVALUE3_S    0
+#define TIMERVALUE3_M    0xffffU
+#define TIMERVALUE3_V(x) ((x) << TIMERVALUE3_S)
+#define TIMERVALUE3_G(x) (((x) >> TIMERVALUE3_S) & TIMERVALUE3_M)
+
+#define SGE_TIMER_VALUE_4_AND_5_A 0x10c0
+
+#define TIMERVALUE4_S    16
+#define TIMERVALUE4_M    0xffffU
+#define TIMERVALUE4_V(x) ((x) << TIMERVALUE4_S)
+#define TIMERVALUE4_G(x) (((x) >> TIMERVALUE4_S) & TIMERVALUE4_M)
+
+#define TIMERVALUE5_S    0
+#define TIMERVALUE5_M    0xffffU
+#define TIMERVALUE5_V(x) ((x) << TIMERVALUE5_S)
+#define TIMERVALUE5_G(x) (((x) >> TIMERVALUE5_S) & TIMERVALUE5_M)
+
+#define SGE_DEBUG_INDEX_A 0x10cc
+#define SGE_DEBUG_DATA_HIGH_A 0x10d0
+#define SGE_DEBUG_DATA_LOW_A 0x10d4
+
+#define SGE_DEBUG_DATA_LOW_INDEX_2_A	0x12c8
+#define SGE_DEBUG_DATA_LOW_INDEX_3_A	0x12cc
+#define SGE_DEBUG_DATA_HIGH_INDEX_10_A	0x12a8
+
+#define SGE_INGRESS_QUEUES_PER_PAGE_PF_A 0x10f4
+#define SGE_INGRESS_QUEUES_PER_PAGE_VF_A 0x10f8
+
+#define SGE_ERROR_STATS_A 0x1100
+
+#define UNCAPTURED_ERROR_S    18
+#define UNCAPTURED_ERROR_V(x) ((x) << UNCAPTURED_ERROR_S)
+#define UNCAPTURED_ERROR_F    UNCAPTURED_ERROR_V(1U)
+
+#define ERROR_QID_VALID_S    17
+#define ERROR_QID_VALID_V(x) ((x) << ERROR_QID_VALID_S)
+#define ERROR_QID_VALID_F    ERROR_QID_VALID_V(1U)
+
+#define ERROR_QID_S    0
+#define ERROR_QID_M    0x1ffffU
+#define ERROR_QID_G(x) (((x) >> ERROR_QID_S) & ERROR_QID_M)
+
+#define SGE_INT_CAUSE5_A        0x110c
+
+#define ERR_T_RXCRC_S    31
+#define ERR_T_RXCRC_V(x) ((x) << ERR_T_RXCRC_S)
+#define ERR_T_RXCRC_F    ERR_T_RXCRC_V(1U)
+
+#define HP_INT_THRESH_S    28
+#define HP_INT_THRESH_M    0xfU
+#define HP_INT_THRESH_V(x) ((x) << HP_INT_THRESH_S)
+
+#define HP_COUNT_S    16
+#define HP_COUNT_M    0x7ffU
+#define HP_COUNT_G(x) (((x) >> HP_COUNT_S) & HP_COUNT_M)
+
+#define LP_INT_THRESH_S    12
+#define LP_INT_THRESH_M    0xfU
+#define LP_INT_THRESH_V(x) ((x) << LP_INT_THRESH_S)
+
+#define LP_COUNT_S    0
+#define LP_COUNT_M    0x7ffU
+#define LP_COUNT_G(x) (((x) >> LP_COUNT_S) & LP_COUNT_M)
+
+#define LP_INT_THRESH_T5_S    18
+#define LP_INT_THRESH_T5_M    0xfffU
+#define LP_INT_THRESH_T5_V(x) ((x) << LP_INT_THRESH_T5_S)
+
+#define LP_COUNT_T5_S    0
+#define LP_COUNT_T5_M    0x3ffffU
+#define LP_COUNT_T5_G(x) (((x) >> LP_COUNT_T5_S) & LP_COUNT_T5_M)
+
+#define SGE_DOORBELL_CONTROL_A 0x10a8
+
+#define SGE_STAT_TOTAL_A	0x10e4
+#define SGE_STAT_MATCH_A	0x10e8
+#define SGE_STAT_CFG_A		0x10ec
+
+#define STATMODE_S    2
+#define STATMODE_V(x) ((x) << STATMODE_S)
+
+#define STATSOURCE_T5_S    9
+#define STATSOURCE_T5_M    0xfU
+#define STATSOURCE_T5_V(x) ((x) << STATSOURCE_T5_S)
+#define STATSOURCE_T5_G(x) (((x) >> STATSOURCE_T5_S) & STATSOURCE_T5_M)
+
+#define T6_STATMODE_S    0
+#define T6_STATMODE_V(x) ((x) << T6_STATMODE_S)
+
+#define SGE_DBFIFO_STATUS2_A 0x1118
+
+#define HP_INT_THRESH_T5_S    10
+#define HP_INT_THRESH_T5_M    0xfU
+#define HP_INT_THRESH_T5_V(x) ((x) << HP_INT_THRESH_T5_S)
+
+#define HP_COUNT_T5_S    0
+#define HP_COUNT_T5_M    0x3ffU
+#define HP_COUNT_T5_G(x) (((x) >> HP_COUNT_T5_S) & HP_COUNT_T5_M)
+
+#define ENABLE_DROP_S    13
+#define ENABLE_DROP_V(x) ((x) << ENABLE_DROP_S)
+#define ENABLE_DROP_F    ENABLE_DROP_V(1U)
+
+#define DROPPED_DB_S    0
+#define DROPPED_DB_V(x) ((x) << DROPPED_DB_S)
+#define DROPPED_DB_F    DROPPED_DB_V(1U)
+
+#define SGE_CTXT_CMD_A 0x11fc
+#define SGE_DBQ_CTXT_BADDR_A 0x1084
+
+/* registers for module PCIE */
+#define PCIE_PF_CFG_A	0x40
+
+#define AIVEC_S    4
+#define AIVEC_M    0x3ffU
+#define AIVEC_V(x) ((x) << AIVEC_S)
+
+#define PCIE_PF_CLI_A	0x44
+
+#define PCIE_PF_EXPROM_OFST_A 0x4c
+#define OFFSET_S    10
+#define OFFSET_M    0x3fffU
+#define OFFSET_G(x) (((x) >> OFFSET_S) & OFFSET_M)
+
+#define PCIE_INT_CAUSE_A	0x3004
+
+#define UNXSPLCPLERR_S    29
+#define UNXSPLCPLERR_V(x) ((x) << UNXSPLCPLERR_S)
+#define UNXSPLCPLERR_F    UNXSPLCPLERR_V(1U)
+
+#define PCIEPINT_S    28
+#define PCIEPINT_V(x) ((x) << PCIEPINT_S)
+#define PCIEPINT_F    PCIEPINT_V(1U)
+
+#define PCIESINT_S    27
+#define PCIESINT_V(x) ((x) << PCIESINT_S)
+#define PCIESINT_F    PCIESINT_V(1U)
+
+#define RPLPERR_S    26
+#define RPLPERR_V(x) ((x) << RPLPERR_S)
+#define RPLPERR_F    RPLPERR_V(1U)
+
+#define RXWRPERR_S    25
+#define RXWRPERR_V(x) ((x) << RXWRPERR_S)
+#define RXWRPERR_F    RXWRPERR_V(1U)
+
+#define RXCPLPERR_S    24
+#define RXCPLPERR_V(x) ((x) << RXCPLPERR_S)
+#define RXCPLPERR_F    RXCPLPERR_V(1U)
+
+#define PIOTAGPERR_S    23
+#define PIOTAGPERR_V(x) ((x) << PIOTAGPERR_S)
+#define PIOTAGPERR_F    PIOTAGPERR_V(1U)
+
+#define MATAGPERR_S    22
+#define MATAGPERR_V(x) ((x) << MATAGPERR_S)
+#define MATAGPERR_F    MATAGPERR_V(1U)
+
+#define INTXCLRPERR_S    21
+#define INTXCLRPERR_V(x) ((x) << INTXCLRPERR_S)
+#define INTXCLRPERR_F    INTXCLRPERR_V(1U)
+
+#define FIDPERR_S    20
+#define FIDPERR_V(x) ((x) << FIDPERR_S)
+#define FIDPERR_F    FIDPERR_V(1U)
+
+#define CFGSNPPERR_S    19
+#define CFGSNPPERR_V(x) ((x) << CFGSNPPERR_S)
+#define CFGSNPPERR_F    CFGSNPPERR_V(1U)
+
+#define HRSPPERR_S    18
+#define HRSPPERR_V(x) ((x) << HRSPPERR_S)
+#define HRSPPERR_F    HRSPPERR_V(1U)
+
+#define HREQPERR_S    17
+#define HREQPERR_V(x) ((x) << HREQPERR_S)
+#define HREQPERR_F    HREQPERR_V(1U)
+
+#define HCNTPERR_S    16
+#define HCNTPERR_V(x) ((x) << HCNTPERR_S)
+#define HCNTPERR_F    HCNTPERR_V(1U)
+
+#define DRSPPERR_S    15
+#define DRSPPERR_V(x) ((x) << DRSPPERR_S)
+#define DRSPPERR_F    DRSPPERR_V(1U)
+
+#define DREQPERR_S    14
+#define DREQPERR_V(x) ((x) << DREQPERR_S)
+#define DREQPERR_F    DREQPERR_V(1U)
+
+#define DCNTPERR_S    13
+#define DCNTPERR_V(x) ((x) << DCNTPERR_S)
+#define DCNTPERR_F    DCNTPERR_V(1U)
+
+#define CRSPPERR_S    12
+#define CRSPPERR_V(x) ((x) << CRSPPERR_S)
+#define CRSPPERR_F    CRSPPERR_V(1U)
+
+#define CREQPERR_S    11
+#define CREQPERR_V(x) ((x) << CREQPERR_S)
+#define CREQPERR_F    CREQPERR_V(1U)
+
+#define CCNTPERR_S    10
+#define CCNTPERR_V(x) ((x) << CCNTPERR_S)
+#define CCNTPERR_F    CCNTPERR_V(1U)
+
+#define TARTAGPERR_S    9
+#define TARTAGPERR_V(x) ((x) << TARTAGPERR_S)
+#define TARTAGPERR_F    TARTAGPERR_V(1U)
+
+#define PIOREQPERR_S    8
+#define PIOREQPERR_V(x) ((x) << PIOREQPERR_S)
+#define PIOREQPERR_F    PIOREQPERR_V(1U)
+
+#define PIOCPLPERR_S    7
+#define PIOCPLPERR_V(x) ((x) << PIOCPLPERR_S)
+#define PIOCPLPERR_F    PIOCPLPERR_V(1U)
+
+#define MSIXDIPERR_S    6
+#define MSIXDIPERR_V(x) ((x) << MSIXDIPERR_S)
+#define MSIXDIPERR_F    MSIXDIPERR_V(1U)
+
+#define MSIXDATAPERR_S    5
+#define MSIXDATAPERR_V(x) ((x) << MSIXDATAPERR_S)
+#define MSIXDATAPERR_F    MSIXDATAPERR_V(1U)
+
+#define MSIXADDRHPERR_S    4
+#define MSIXADDRHPERR_V(x) ((x) << MSIXADDRHPERR_S)
+#define MSIXADDRHPERR_F    MSIXADDRHPERR_V(1U)
+
+#define MSIXADDRLPERR_S    3
+#define MSIXADDRLPERR_V(x) ((x) << MSIXADDRLPERR_S)
+#define MSIXADDRLPERR_F    MSIXADDRLPERR_V(1U)
+
+#define MSIDATAPERR_S    2
+#define MSIDATAPERR_V(x) ((x) << MSIDATAPERR_S)
+#define MSIDATAPERR_F    MSIDATAPERR_V(1U)
+
+#define MSIADDRHPERR_S    1
+#define MSIADDRHPERR_V(x) ((x) << MSIADDRHPERR_S)
+#define MSIADDRHPERR_F    MSIADDRHPERR_V(1U)
+
+#define MSIADDRLPERR_S    0
+#define MSIADDRLPERR_V(x) ((x) << MSIADDRLPERR_S)
+#define MSIADDRLPERR_F    MSIADDRLPERR_V(1U)
+
+#define READRSPERR_S    29
+#define READRSPERR_V(x) ((x) << READRSPERR_S)
+#define READRSPERR_F    READRSPERR_V(1U)
+
+#define TRGT1GRPPERR_S    28
+#define TRGT1GRPPERR_V(x) ((x) << TRGT1GRPPERR_S)
+#define TRGT1GRPPERR_F    TRGT1GRPPERR_V(1U)
+
+#define IPSOTPERR_S    27
+#define IPSOTPERR_V(x) ((x) << IPSOTPERR_S)
+#define IPSOTPERR_F    IPSOTPERR_V(1U)
+
+#define IPRETRYPERR_S    26
+#define IPRETRYPERR_V(x) ((x) << IPRETRYPERR_S)
+#define IPRETRYPERR_F    IPRETRYPERR_V(1U)
+
+#define IPRXDATAGRPPERR_S    25
+#define IPRXDATAGRPPERR_V(x) ((x) << IPRXDATAGRPPERR_S)
+#define IPRXDATAGRPPERR_F    IPRXDATAGRPPERR_V(1U)
+
+#define IPRXHDRGRPPERR_S    24
+#define IPRXHDRGRPPERR_V(x) ((x) << IPRXHDRGRPPERR_S)
+#define IPRXHDRGRPPERR_F    IPRXHDRGRPPERR_V(1U)
+
+#define MAGRPPERR_S    22
+#define MAGRPPERR_V(x) ((x) << MAGRPPERR_S)
+#define MAGRPPERR_F    MAGRPPERR_V(1U)
+
+#define VFIDPERR_S    21
+#define VFIDPERR_V(x) ((x) << VFIDPERR_S)
+#define VFIDPERR_F    VFIDPERR_V(1U)
+
+#define HREQWRPERR_S    16
+#define HREQWRPERR_V(x) ((x) << HREQWRPERR_S)
+#define HREQWRPERR_F    HREQWRPERR_V(1U)
+
+#define DREQWRPERR_S    13
+#define DREQWRPERR_V(x) ((x) << DREQWRPERR_S)
+#define DREQWRPERR_F    DREQWRPERR_V(1U)
+
+#define CREQRDPERR_S    11
+#define CREQRDPERR_V(x) ((x) << CREQRDPERR_S)
+#define CREQRDPERR_F    CREQRDPERR_V(1U)
+
+#define MSTTAGQPERR_S    10
+#define MSTTAGQPERR_V(x) ((x) << MSTTAGQPERR_S)
+#define MSTTAGQPERR_F    MSTTAGQPERR_V(1U)
+
+#define PIOREQGRPPERR_S    8
+#define PIOREQGRPPERR_V(x) ((x) << PIOREQGRPPERR_S)
+#define PIOREQGRPPERR_F    PIOREQGRPPERR_V(1U)
+
+#define PIOCPLGRPPERR_S    7
+#define PIOCPLGRPPERR_V(x) ((x) << PIOCPLGRPPERR_S)
+#define PIOCPLGRPPERR_F    PIOCPLGRPPERR_V(1U)
+
+#define MSIXSTIPERR_S    2
+#define MSIXSTIPERR_V(x) ((x) << MSIXSTIPERR_S)
+#define MSIXSTIPERR_F    MSIXSTIPERR_V(1U)
+
+#define MSTTIMEOUTPERR_S    1
+#define MSTTIMEOUTPERR_V(x) ((x) << MSTTIMEOUTPERR_S)
+#define MSTTIMEOUTPERR_F    MSTTIMEOUTPERR_V(1U)
+
+#define MSTGRPPERR_S    0
+#define MSTGRPPERR_V(x) ((x) << MSTGRPPERR_S)
+#define MSTGRPPERR_F    MSTGRPPERR_V(1U)
+
+#define PCIE_NONFAT_ERR_A	0x3010
+#define PCIE_CFG_SPACE_REQ_A	0x3060
+#define PCIE_CFG_SPACE_DATA_A	0x3064
+#define PCIE_MEM_ACCESS_BASE_WIN_A 0x3068
+
+#define PCIEOFST_S    10
+#define PCIEOFST_M    0x3fffffU
+#define PCIEOFST_G(x) (((x) >> PCIEOFST_S) & PCIEOFST_M)
+
+#define BIR_S    8
+#define BIR_M    0x3U
+#define BIR_V(x) ((x) << BIR_S)
+#define BIR_G(x) (((x) >> BIR_S) & BIR_M)
+
+#define WINDOW_S    0
+#define WINDOW_M    0xffU
+#define WINDOW_V(x) ((x) << WINDOW_S)
+#define WINDOW_G(x) (((x) >> WINDOW_S) & WINDOW_M)
+
+#define PCIE_MEM_ACCESS_OFFSET_A 0x306c
+
+#define ENABLE_S    30
+#define ENABLE_V(x) ((x) << ENABLE_S)
+#define ENABLE_F    ENABLE_V(1U)
+
+#define LOCALCFG_S    28
+#define LOCALCFG_V(x) ((x) << LOCALCFG_S)
+#define LOCALCFG_F    LOCALCFG_V(1U)
+
+#define FUNCTION_S    12
+#define FUNCTION_V(x) ((x) << FUNCTION_S)
+
+#define REGISTER_S    0
+#define REGISTER_V(x) ((x) << REGISTER_S)
+
+#define T6_ENABLE_S    31
+#define T6_ENABLE_V(x) ((x) << T6_ENABLE_S)
+#define T6_ENABLE_F    T6_ENABLE_V(1U)
+
+#define PFNUM_S    0
+#define PFNUM_V(x) ((x) << PFNUM_S)
+
+#define PCIE_FW_A 0x30b8
+#define PCIE_FW_PF_A 0x30bc
+
+#define PCIE_CORE_UTL_SYSTEM_BUS_AGENT_STATUS_A 0x5908
+
+#define RNPP_S    31
+#define RNPP_V(x) ((x) << RNPP_S)
+#define RNPP_F    RNPP_V(1U)
+
+#define RPCP_S    29
+#define RPCP_V(x) ((x) << RPCP_S)
+#define RPCP_F    RPCP_V(1U)
+
+#define RCIP_S    27
+#define RCIP_V(x) ((x) << RCIP_S)
+#define RCIP_F    RCIP_V(1U)
+
+#define RCCP_S    26
+#define RCCP_V(x) ((x) << RCCP_S)
+#define RCCP_F    RCCP_V(1U)
+
+#define RFTP_S    23
+#define RFTP_V(x) ((x) << RFTP_S)
+#define RFTP_F    RFTP_V(1U)
+
+#define PTRP_S    20
+#define PTRP_V(x) ((x) << PTRP_S)
+#define PTRP_F    PTRP_V(1U)
+
+#define PCIE_CORE_UTL_PCI_EXPRESS_PORT_STATUS_A 0x59a4
+
+#define TPCP_S    30
+#define TPCP_V(x) ((x) << TPCP_S)
+#define TPCP_F    TPCP_V(1U)
+
+#define TNPP_S    29
+#define TNPP_V(x) ((x) << TNPP_S)
+#define TNPP_F    TNPP_V(1U)
+
+#define TFTP_S    28
+#define TFTP_V(x) ((x) << TFTP_S)
+#define TFTP_F    TFTP_V(1U)
+
+#define TCAP_S    27
+#define TCAP_V(x) ((x) << TCAP_S)
+#define TCAP_F    TCAP_V(1U)
+
+#define TCIP_S    26
+#define TCIP_V(x) ((x) << TCIP_S)
+#define TCIP_F    TCIP_V(1U)
+
+#define RCAP_S    25
+#define RCAP_V(x) ((x) << RCAP_S)
+#define RCAP_F    RCAP_V(1U)
+
+#define PLUP_S    23
+#define PLUP_V(x) ((x) << PLUP_S)
+#define PLUP_F    PLUP_V(1U)
+
+#define PLDN_S    22
+#define PLDN_V(x) ((x) << PLDN_S)
+#define PLDN_F    PLDN_V(1U)
+
+#define OTDD_S    21
+#define OTDD_V(x) ((x) << OTDD_S)
+#define OTDD_F    OTDD_V(1U)
+
+#define GTRP_S    20
+#define GTRP_V(x) ((x) << GTRP_S)
+#define GTRP_F    GTRP_V(1U)
+
+#define RDPE_S    18
+#define RDPE_V(x) ((x) << RDPE_S)
+#define RDPE_F    RDPE_V(1U)
+
+#define TDCE_S    17
+#define TDCE_V(x) ((x) << TDCE_S)
+#define TDCE_F    TDCE_V(1U)
+
+#define TDUE_S    16
+#define TDUE_V(x) ((x) << TDUE_S)
+#define TDUE_F    TDUE_V(1U)
+
+/* SPARE2 register contains 32-bit value at offset 0x6 in Serial INIT
+ * Configuration flashed on EEPROM. This value corresponds to 32-bit
+ * Serial Configuration Version information.
+ */
+#define PCIE_STATIC_SPARE2_A	0x5bfc
+
+/* registers for module MC */
+#define MC_INT_CAUSE_A		0x7518
+#define MC_P_INT_CAUSE_A	0x41318
+
+#define ECC_UE_INT_CAUSE_S    2
+#define ECC_UE_INT_CAUSE_V(x) ((x) << ECC_UE_INT_CAUSE_S)
+#define ECC_UE_INT_CAUSE_F    ECC_UE_INT_CAUSE_V(1U)
+
+#define ECC_CE_INT_CAUSE_S    1
+#define ECC_CE_INT_CAUSE_V(x) ((x) << ECC_CE_INT_CAUSE_S)
+#define ECC_CE_INT_CAUSE_F    ECC_CE_INT_CAUSE_V(1U)
+
+#define PERR_INT_CAUSE_S    0
+#define PERR_INT_CAUSE_V(x) ((x) << PERR_INT_CAUSE_S)
+#define PERR_INT_CAUSE_F    PERR_INT_CAUSE_V(1U)
+
+#define DBG_GPIO_EN_A		0x6010
+#define XGMAC_PORT_CFG_A	0x1000
+#define MAC_PORT_CFG_A		0x800
+
+#define SIGNAL_DET_S    14
+#define SIGNAL_DET_V(x) ((x) << SIGNAL_DET_S)
+#define SIGNAL_DET_F    SIGNAL_DET_V(1U)
+
+#define MC_ECC_STATUS_A		0x751c
+#define MC_P_ECC_STATUS_A	0x4131c
+
+#define ECC_CECNT_S    16
+#define ECC_CECNT_M    0xffffU
+#define ECC_CECNT_V(x) ((x) << ECC_CECNT_S)
+#define ECC_CECNT_G(x) (((x) >> ECC_CECNT_S) & ECC_CECNT_M)
+
+#define ECC_UECNT_S    0
+#define ECC_UECNT_M    0xffffU
+#define ECC_UECNT_V(x) ((x) << ECC_UECNT_S)
+#define ECC_UECNT_G(x) (((x) >> ECC_UECNT_S) & ECC_UECNT_M)
+
+#define MC_BIST_CMD_A 0x7600
+
+#define START_BIST_S    31
+#define START_BIST_V(x) ((x) << START_BIST_S)
+#define START_BIST_F    START_BIST_V(1U)
+
+#define BIST_CMD_GAP_S    8
+#define BIST_CMD_GAP_V(x) ((x) << BIST_CMD_GAP_S)
+
+#define BIST_OPCODE_S    0
+#define BIST_OPCODE_V(x) ((x) << BIST_OPCODE_S)
+
+#define MC_BIST_CMD_ADDR_A 0x7604
+#define MC_BIST_CMD_LEN_A 0x7608
+#define MC_BIST_DATA_PATTERN_A 0x760c
+
+#define MC_BIST_STATUS_RDATA_A 0x7688
+
+/* registers for module MA */
+#define MA_EDRAM0_BAR_A 0x77c0
+
+#define EDRAM0_BASE_S    16
+#define EDRAM0_BASE_M    0xfffU
+#define EDRAM0_BASE_G(x) (((x) >> EDRAM0_BASE_S) & EDRAM0_BASE_M)
+
+#define EDRAM0_SIZE_S    0
+#define EDRAM0_SIZE_M    0xfffU
+#define EDRAM0_SIZE_V(x) ((x) << EDRAM0_SIZE_S)
+#define EDRAM0_SIZE_G(x) (((x) >> EDRAM0_SIZE_S) & EDRAM0_SIZE_M)
+
+#define MA_EDRAM1_BAR_A 0x77c4
+
+#define EDRAM1_BASE_S    16
+#define EDRAM1_BASE_M    0xfffU
+#define EDRAM1_BASE_G(x) (((x) >> EDRAM1_BASE_S) & EDRAM1_BASE_M)
+
+#define EDRAM1_SIZE_S    0
+#define EDRAM1_SIZE_M    0xfffU
+#define EDRAM1_SIZE_V(x) ((x) << EDRAM1_SIZE_S)
+#define EDRAM1_SIZE_G(x) (((x) >> EDRAM1_SIZE_S) & EDRAM1_SIZE_M)
+
+#define MA_EXT_MEMORY_BAR_A 0x77c8
+
+#define EXT_MEM_BASE_S    16
+#define EXT_MEM_BASE_M    0xfffU
+#define EXT_MEM_BASE_V(x) ((x) << EXT_MEM_BASE_S)
+#define EXT_MEM_BASE_G(x) (((x) >> EXT_MEM_BASE_S) & EXT_MEM_BASE_M)
+
+#define EXT_MEM_SIZE_S    0
+#define EXT_MEM_SIZE_M    0xfffU
+#define EXT_MEM_SIZE_V(x) ((x) << EXT_MEM_SIZE_S)
+#define EXT_MEM_SIZE_G(x) (((x) >> EXT_MEM_SIZE_S) & EXT_MEM_SIZE_M)
+
+#define MA_EXT_MEMORY1_BAR_A 0x7808
+
+#define HMA_MUX_S    5
+#define HMA_MUX_V(x) ((x) << HMA_MUX_S)
+#define HMA_MUX_F    HMA_MUX_V(1U)
+
+#define EXT_MEM1_BASE_S    16
+#define EXT_MEM1_BASE_M    0xfffU
+#define EXT_MEM1_BASE_G(x) (((x) >> EXT_MEM1_BASE_S) & EXT_MEM1_BASE_M)
+
+#define EXT_MEM1_SIZE_S    0
+#define EXT_MEM1_SIZE_M    0xfffU
+#define EXT_MEM1_SIZE_V(x) ((x) << EXT_MEM1_SIZE_S)
+#define EXT_MEM1_SIZE_G(x) (((x) >> EXT_MEM1_SIZE_S) & EXT_MEM1_SIZE_M)
+
+#define MA_EXT_MEMORY0_BAR_A 0x77c8
+
+#define EXT_MEM0_BASE_S    16
+#define EXT_MEM0_BASE_M    0xfffU
+#define EXT_MEM0_BASE_G(x) (((x) >> EXT_MEM0_BASE_S) & EXT_MEM0_BASE_M)
+
+#define EXT_MEM0_SIZE_S    0
+#define EXT_MEM0_SIZE_M    0xfffU
+#define EXT_MEM0_SIZE_V(x) ((x) << EXT_MEM0_SIZE_S)
+#define EXT_MEM0_SIZE_G(x) (((x) >> EXT_MEM0_SIZE_S) & EXT_MEM0_SIZE_M)
+
+#define MA_TARGET_MEM_ENABLE_A 0x77d8
+
+#define EXT_MEM_ENABLE_S    2
+#define EXT_MEM_ENABLE_V(x) ((x) << EXT_MEM_ENABLE_S)
+#define EXT_MEM_ENABLE_F    EXT_MEM_ENABLE_V(1U)
+
+#define EDRAM1_ENABLE_S    1
+#define EDRAM1_ENABLE_V(x) ((x) << EDRAM1_ENABLE_S)
+#define EDRAM1_ENABLE_F    EDRAM1_ENABLE_V(1U)
+
+#define EDRAM0_ENABLE_S    0
+#define EDRAM0_ENABLE_V(x) ((x) << EDRAM0_ENABLE_S)
+#define EDRAM0_ENABLE_F    EDRAM0_ENABLE_V(1U)
+
+#define EXT_MEM1_ENABLE_S    4
+#define EXT_MEM1_ENABLE_V(x) ((x) << EXT_MEM1_ENABLE_S)
+#define EXT_MEM1_ENABLE_F    EXT_MEM1_ENABLE_V(1U)
+
+#define EXT_MEM0_ENABLE_S    2
+#define EXT_MEM0_ENABLE_V(x) ((x) << EXT_MEM0_ENABLE_S)
+#define EXT_MEM0_ENABLE_F    EXT_MEM0_ENABLE_V(1U)
+
+#define MA_INT_CAUSE_A	0x77e0
+
+#define MEM_PERR_INT_CAUSE_S    1
+#define MEM_PERR_INT_CAUSE_V(x) ((x) << MEM_PERR_INT_CAUSE_S)
+#define MEM_PERR_INT_CAUSE_F    MEM_PERR_INT_CAUSE_V(1U)
+
+#define MEM_WRAP_INT_CAUSE_S    0
+#define MEM_WRAP_INT_CAUSE_V(x) ((x) << MEM_WRAP_INT_CAUSE_S)
+#define MEM_WRAP_INT_CAUSE_F    MEM_WRAP_INT_CAUSE_V(1U)
+
+#define MA_INT_WRAP_STATUS_A	0x77e4
+
+#define MEM_WRAP_ADDRESS_S    4
+#define MEM_WRAP_ADDRESS_M    0xfffffffU
+#define MEM_WRAP_ADDRESS_G(x) (((x) >> MEM_WRAP_ADDRESS_S) & MEM_WRAP_ADDRESS_M)
+
+#define MEM_WRAP_CLIENT_NUM_S    0
+#define MEM_WRAP_CLIENT_NUM_M    0xfU
+#define MEM_WRAP_CLIENT_NUM_G(x) \
+	(((x) >> MEM_WRAP_CLIENT_NUM_S) & MEM_WRAP_CLIENT_NUM_M)
+
+#define MA_PARITY_ERROR_STATUS_A	0x77f4
+#define MA_PARITY_ERROR_STATUS1_A	0x77f4
+#define MA_PARITY_ERROR_STATUS2_A	0x7804
+
+/* registers for module EDC_0 */
+#define EDC_0_BASE_ADDR		0x7900
+
+#define EDC_BIST_CMD_A		0x7904
+#define EDC_BIST_CMD_ADDR_A	0x7908
+#define EDC_BIST_CMD_LEN_A	0x790c
+#define EDC_BIST_DATA_PATTERN_A 0x7910
+#define EDC_BIST_STATUS_RDATA_A	0x7928
+#define EDC_INT_CAUSE_A		0x7978
+
+#define ECC_UE_PAR_S    5
+#define ECC_UE_PAR_V(x) ((x) << ECC_UE_PAR_S)
+#define ECC_UE_PAR_F    ECC_UE_PAR_V(1U)
+
+#define ECC_CE_PAR_S    4
+#define ECC_CE_PAR_V(x) ((x) << ECC_CE_PAR_S)
+#define ECC_CE_PAR_F    ECC_CE_PAR_V(1U)
+
+#define PERR_PAR_CAUSE_S    3
+#define PERR_PAR_CAUSE_V(x) ((x) << PERR_PAR_CAUSE_S)
+#define PERR_PAR_CAUSE_F    PERR_PAR_CAUSE_V(1U)
+
+#define EDC_ECC_STATUS_A	0x797c
+
+/* registers for module EDC_1 */
+#define EDC_1_BASE_ADDR	0x7980
+
+/* registers for module CIM */
+#define CIM_BOOT_CFG_A 0x7b00
+#define CIM_SDRAM_BASE_ADDR_A 0x7b14
+#define CIM_SDRAM_ADDR_SIZE_A 0x7b18
+#define CIM_EXTMEM2_BASE_ADDR_A 0x7b1c
+#define CIM_EXTMEM2_ADDR_SIZE_A 0x7b20
+#define CIM_PF_MAILBOX_CTRL_SHADOW_COPY_A 0x290
+
+#define  BOOTADDR_M	0xffffff00U
+
+#define UPCRST_S    0
+#define UPCRST_V(x) ((x) << UPCRST_S)
+#define UPCRST_F    UPCRST_V(1U)
+
+#define CIM_PF_MAILBOX_DATA_A 0x240
+#define CIM_PF_MAILBOX_CTRL_A 0x280
+
+#define MBMSGVALID_S    3
+#define MBMSGVALID_V(x) ((x) << MBMSGVALID_S)
+#define MBMSGVALID_F    MBMSGVALID_V(1U)
+
+#define MBINTREQ_S    2
+#define MBINTREQ_V(x) ((x) << MBINTREQ_S)
+#define MBINTREQ_F    MBINTREQ_V(1U)
+
+#define MBOWNER_S    0
+#define MBOWNER_M    0x3U
+#define MBOWNER_V(x) ((x) << MBOWNER_S)
+#define MBOWNER_G(x) (((x) >> MBOWNER_S) & MBOWNER_M)
+
+#define CIM_PF_HOST_INT_ENABLE_A 0x288
+
+#define MBMSGRDYINTEN_S    19
+#define MBMSGRDYINTEN_V(x) ((x) << MBMSGRDYINTEN_S)
+#define MBMSGRDYINTEN_F    MBMSGRDYINTEN_V(1U)
+
+#define CIM_PF_HOST_INT_CAUSE_A 0x28c
+
+#define MBMSGRDYINT_S    19
+#define MBMSGRDYINT_V(x) ((x) << MBMSGRDYINT_S)
+#define MBMSGRDYINT_F    MBMSGRDYINT_V(1U)
+
+#define CIM_HOST_INT_CAUSE_A 0x7b2c
+
+#define TIEQOUTPARERRINT_S    20
+#define TIEQOUTPARERRINT_V(x) ((x) << TIEQOUTPARERRINT_S)
+#define TIEQOUTPARERRINT_F    TIEQOUTPARERRINT_V(1U)
+
+#define TIEQINPARERRINT_S    19
+#define TIEQINPARERRINT_V(x) ((x) << TIEQINPARERRINT_S)
+#define TIEQINPARERRINT_F    TIEQINPARERRINT_V(1U)
+
+#define TIMER0INT_S    2
+#define TIMER0INT_V(x) ((x) << TIMER0INT_S)
+#define TIMER0INT_F    TIMER0INT_V(1U)
+
+#define PREFDROPINT_S    1
+#define PREFDROPINT_V(x) ((x) << PREFDROPINT_S)
+#define PREFDROPINT_F    PREFDROPINT_V(1U)
+
+#define UPACCNONZERO_S    0
+#define UPACCNONZERO_V(x) ((x) << UPACCNONZERO_S)
+#define UPACCNONZERO_F    UPACCNONZERO_V(1U)
+
+#define MBHOSTPARERR_S    18
+#define MBHOSTPARERR_V(x) ((x) << MBHOSTPARERR_S)
+#define MBHOSTPARERR_F    MBHOSTPARERR_V(1U)
+
+#define MBUPPARERR_S    17
+#define MBUPPARERR_V(x) ((x) << MBUPPARERR_S)
+#define MBUPPARERR_F    MBUPPARERR_V(1U)
+
+#define IBQTP0PARERR_S    16
+#define IBQTP0PARERR_V(x) ((x) << IBQTP0PARERR_S)
+#define IBQTP0PARERR_F    IBQTP0PARERR_V(1U)
+
+#define IBQTP1PARERR_S    15
+#define IBQTP1PARERR_V(x) ((x) << IBQTP1PARERR_S)
+#define IBQTP1PARERR_F    IBQTP1PARERR_V(1U)
+
+#define IBQULPPARERR_S    14
+#define IBQULPPARERR_V(x) ((x) << IBQULPPARERR_S)
+#define IBQULPPARERR_F    IBQULPPARERR_V(1U)
+
+#define IBQSGELOPARERR_S    13
+#define IBQSGELOPARERR_V(x) ((x) << IBQSGELOPARERR_S)
+#define IBQSGELOPARERR_F    IBQSGELOPARERR_V(1U)
+
+#define IBQSGEHIPARERR_S    12
+#define IBQSGEHIPARERR_V(x) ((x) << IBQSGEHIPARERR_S)
+#define IBQSGEHIPARERR_F    IBQSGEHIPARERR_V(1U)
+
+#define IBQNCSIPARERR_S    11
+#define IBQNCSIPARERR_V(x) ((x) << IBQNCSIPARERR_S)
+#define IBQNCSIPARERR_F    IBQNCSIPARERR_V(1U)
+
+#define OBQULP0PARERR_S    10
+#define OBQULP0PARERR_V(x) ((x) << OBQULP0PARERR_S)
+#define OBQULP0PARERR_F    OBQULP0PARERR_V(1U)
+
+#define OBQULP1PARERR_S    9
+#define OBQULP1PARERR_V(x) ((x) << OBQULP1PARERR_S)
+#define OBQULP1PARERR_F    OBQULP1PARERR_V(1U)
+
+#define OBQULP2PARERR_S    8
+#define OBQULP2PARERR_V(x) ((x) << OBQULP2PARERR_S)
+#define OBQULP2PARERR_F    OBQULP2PARERR_V(1U)
+
+#define OBQULP3PARERR_S    7
+#define OBQULP3PARERR_V(x) ((x) << OBQULP3PARERR_S)
+#define OBQULP3PARERR_F    OBQULP3PARERR_V(1U)
+
+#define OBQSGEPARERR_S    6
+#define OBQSGEPARERR_V(x) ((x) << OBQSGEPARERR_S)
+#define OBQSGEPARERR_F    OBQSGEPARERR_V(1U)
+
+#define OBQNCSIPARERR_S    5
+#define OBQNCSIPARERR_V(x) ((x) << OBQNCSIPARERR_S)
+#define OBQNCSIPARERR_F    OBQNCSIPARERR_V(1U)
+
+#define CIM_HOST_UPACC_INT_CAUSE_A 0x7b34
+
+#define EEPROMWRINT_S    30
+#define EEPROMWRINT_V(x) ((x) << EEPROMWRINT_S)
+#define EEPROMWRINT_F    EEPROMWRINT_V(1U)
+
+#define TIMEOUTMAINT_S    29
+#define TIMEOUTMAINT_V(x) ((x) << TIMEOUTMAINT_S)
+#define TIMEOUTMAINT_F    TIMEOUTMAINT_V(1U)
+
+#define TIMEOUTINT_S    28
+#define TIMEOUTINT_V(x) ((x) << TIMEOUTINT_S)
+#define TIMEOUTINT_F    TIMEOUTINT_V(1U)
+
+#define RSPOVRLOOKUPINT_S    27
+#define RSPOVRLOOKUPINT_V(x) ((x) << RSPOVRLOOKUPINT_S)
+#define RSPOVRLOOKUPINT_F    RSPOVRLOOKUPINT_V(1U)
+
+#define REQOVRLOOKUPINT_S    26
+#define REQOVRLOOKUPINT_V(x) ((x) << REQOVRLOOKUPINT_S)
+#define REQOVRLOOKUPINT_F    REQOVRLOOKUPINT_V(1U)
+
+#define BLKWRPLINT_S    25
+#define BLKWRPLINT_V(x) ((x) << BLKWRPLINT_S)
+#define BLKWRPLINT_F    BLKWRPLINT_V(1U)
+
+#define BLKRDPLINT_S    24
+#define BLKRDPLINT_V(x) ((x) << BLKRDPLINT_S)
+#define BLKRDPLINT_F    BLKRDPLINT_V(1U)
+
+#define SGLWRPLINT_S    23
+#define SGLWRPLINT_V(x) ((x) << SGLWRPLINT_S)
+#define SGLWRPLINT_F    SGLWRPLINT_V(1U)
+
+#define SGLRDPLINT_S    22
+#define SGLRDPLINT_V(x) ((x) << SGLRDPLINT_S)
+#define SGLRDPLINT_F    SGLRDPLINT_V(1U)
+
+#define BLKWRCTLINT_S    21
+#define BLKWRCTLINT_V(x) ((x) << BLKWRCTLINT_S)
+#define BLKWRCTLINT_F    BLKWRCTLINT_V(1U)
+
+#define BLKRDCTLINT_S    20
+#define BLKRDCTLINT_V(x) ((x) << BLKRDCTLINT_S)
+#define BLKRDCTLINT_F    BLKRDCTLINT_V(1U)
+
+#define SGLWRCTLINT_S    19
+#define SGLWRCTLINT_V(x) ((x) << SGLWRCTLINT_S)
+#define SGLWRCTLINT_F    SGLWRCTLINT_V(1U)
+
+#define SGLRDCTLINT_S    18
+#define SGLRDCTLINT_V(x) ((x) << SGLRDCTLINT_S)
+#define SGLRDCTLINT_F    SGLRDCTLINT_V(1U)
+
+#define BLKWREEPROMINT_S    17
+#define BLKWREEPROMINT_V(x) ((x) << BLKWREEPROMINT_S)
+#define BLKWREEPROMINT_F    BLKWREEPROMINT_V(1U)
+
+#define BLKRDEEPROMINT_S    16
+#define BLKRDEEPROMINT_V(x) ((x) << BLKRDEEPROMINT_S)
+#define BLKRDEEPROMINT_F    BLKRDEEPROMINT_V(1U)
+
+#define SGLWREEPROMINT_S    15
+#define SGLWREEPROMINT_V(x) ((x) << SGLWREEPROMINT_S)
+#define SGLWREEPROMINT_F    SGLWREEPROMINT_V(1U)
+
+#define SGLRDEEPROMINT_S    14
+#define SGLRDEEPROMINT_V(x) ((x) << SGLRDEEPROMINT_S)
+#define SGLRDEEPROMINT_F    SGLRDEEPROMINT_V(1U)
+
+#define BLKWRFLASHINT_S    13
+#define BLKWRFLASHINT_V(x) ((x) << BLKWRFLASHINT_S)
+#define BLKWRFLASHINT_F    BLKWRFLASHINT_V(1U)
+
+#define BLKRDFLASHINT_S    12
+#define BLKRDFLASHINT_V(x) ((x) << BLKRDFLASHINT_S)
+#define BLKRDFLASHINT_F    BLKRDFLASHINT_V(1U)
+
+#define SGLWRFLASHINT_S    11
+#define SGLWRFLASHINT_V(x) ((x) << SGLWRFLASHINT_S)
+#define SGLWRFLASHINT_F    SGLWRFLASHINT_V(1U)
+
+#define SGLRDFLASHINT_S    10
+#define SGLRDFLASHINT_V(x) ((x) << SGLRDFLASHINT_S)
+#define SGLRDFLASHINT_F    SGLRDFLASHINT_V(1U)
+
+#define BLKWRBOOTINT_S    9
+#define BLKWRBOOTINT_V(x) ((x) << BLKWRBOOTINT_S)
+#define BLKWRBOOTINT_F    BLKWRBOOTINT_V(1U)
+
+#define BLKRDBOOTINT_S    8
+#define BLKRDBOOTINT_V(x) ((x) << BLKRDBOOTINT_S)
+#define BLKRDBOOTINT_F    BLKRDBOOTINT_V(1U)
+
+#define SGLWRBOOTINT_S    7
+#define SGLWRBOOTINT_V(x) ((x) << SGLWRBOOTINT_S)
+#define SGLWRBOOTINT_F    SGLWRBOOTINT_V(1U)
+
+#define SGLRDBOOTINT_S    6
+#define SGLRDBOOTINT_V(x) ((x) << SGLRDBOOTINT_S)
+#define SGLRDBOOTINT_F    SGLRDBOOTINT_V(1U)
+
+#define ILLWRBEINT_S    5
+#define ILLWRBEINT_V(x) ((x) << ILLWRBEINT_S)
+#define ILLWRBEINT_F    ILLWRBEINT_V(1U)
+
+#define ILLRDBEINT_S    4
+#define ILLRDBEINT_V(x) ((x) << ILLRDBEINT_S)
+#define ILLRDBEINT_F    ILLRDBEINT_V(1U)
+
+#define ILLRDINT_S    3
+#define ILLRDINT_V(x) ((x) << ILLRDINT_S)
+#define ILLRDINT_F    ILLRDINT_V(1U)
+
+#define ILLWRINT_S    2
+#define ILLWRINT_V(x) ((x) << ILLWRINT_S)
+#define ILLWRINT_F    ILLWRINT_V(1U)
+
+#define ILLTRANSINT_S    1
+#define ILLTRANSINT_V(x) ((x) << ILLTRANSINT_S)
+#define ILLTRANSINT_F    ILLTRANSINT_V(1U)
+
+#define RSVDSPACEINT_S    0
+#define RSVDSPACEINT_V(x) ((x) << RSVDSPACEINT_S)
+#define RSVDSPACEINT_F    RSVDSPACEINT_V(1U)
+
+/* registers for module TP */
+#define DBGLAWHLF_S    23
+#define DBGLAWHLF_V(x) ((x) << DBGLAWHLF_S)
+#define DBGLAWHLF_F    DBGLAWHLF_V(1U)
+
+#define DBGLAWPTR_S    16
+#define DBGLAWPTR_M    0x7fU
+#define DBGLAWPTR_G(x) (((x) >> DBGLAWPTR_S) & DBGLAWPTR_M)
+
+#define DBGLAENABLE_S    12
+#define DBGLAENABLE_V(x) ((x) << DBGLAENABLE_S)
+#define DBGLAENABLE_F    DBGLAENABLE_V(1U)
+
+#define DBGLARPTR_S    0
+#define DBGLARPTR_M    0x7fU
+#define DBGLARPTR_V(x) ((x) << DBGLARPTR_S)
+
+#define CRXPKTENC_S    3
+#define CRXPKTENC_V(x) ((x) << CRXPKTENC_S)
+#define CRXPKTENC_F    CRXPKTENC_V(1U)
+
+#define TP_DBG_LA_DATAL_A	0x7ed8
+#define TP_DBG_LA_CONFIG_A	0x7ed4
+#define TP_OUT_CONFIG_A		0x7d04
+#define TP_GLOBAL_CONFIG_A	0x7d08
+
+#define ACTIVEFILTERCOUNTS_S    22
+#define ACTIVEFILTERCOUNTS_V(x) ((x) << ACTIVEFILTERCOUNTS_S)
+#define ACTIVEFILTERCOUNTS_F    ACTIVEFILTERCOUNTS_V(1U)
+
+#define TP_CMM_TCB_BASE_A 0x7d10
+#define TP_CMM_MM_BASE_A 0x7d14
+#define TP_CMM_TIMER_BASE_A 0x7d18
+#define TP_PMM_TX_BASE_A 0x7d20
+#define TP_PMM_RX_BASE_A 0x7d28
+#define TP_PMM_RX_PAGE_SIZE_A 0x7d2c
+#define TP_PMM_RX_MAX_PAGE_A 0x7d30
+#define TP_PMM_TX_PAGE_SIZE_A 0x7d34
+#define TP_PMM_TX_MAX_PAGE_A 0x7d38
+#define TP_CMM_MM_MAX_PSTRUCT_A 0x7e6c
+
+#define PMRXNUMCHN_S    31
+#define PMRXNUMCHN_V(x) ((x) << PMRXNUMCHN_S)
+#define PMRXNUMCHN_F    PMRXNUMCHN_V(1U)
+
+#define PMTXNUMCHN_S    30
+#define PMTXNUMCHN_M    0x3U
+#define PMTXNUMCHN_G(x) (((x) >> PMTXNUMCHN_S) & PMTXNUMCHN_M)
+
+#define PMTXMAXPAGE_S    0
+#define PMTXMAXPAGE_M    0x1fffffU
+#define PMTXMAXPAGE_G(x) (((x) >> PMTXMAXPAGE_S) & PMTXMAXPAGE_M)
+
+#define PMRXMAXPAGE_S    0
+#define PMRXMAXPAGE_M    0x1fffffU
+#define PMRXMAXPAGE_G(x) (((x) >> PMRXMAXPAGE_S) & PMRXMAXPAGE_M)
+
+#define DBGLAMODE_S	14
+#define DBGLAMODE_M	0x3U
+#define DBGLAMODE_G(x)	(((x) >> DBGLAMODE_S) & DBGLAMODE_M)
+
+#define FIVETUPLELOOKUP_S    17
+#define FIVETUPLELOOKUP_M    0x3U
+#define FIVETUPLELOOKUP_V(x) ((x) << FIVETUPLELOOKUP_S)
+#define FIVETUPLELOOKUP_G(x) (((x) >> FIVETUPLELOOKUP_S) & FIVETUPLELOOKUP_M)
+
+#define TP_PARA_REG2_A 0x7d68
+
+#define MAXRXDATA_S    16
+#define MAXRXDATA_M    0xffffU
+#define MAXRXDATA_G(x) (((x) >> MAXRXDATA_S) & MAXRXDATA_M)
+
+#define TP_TIMER_RESOLUTION_A 0x7d90
+
+#define TIMERRESOLUTION_S    16
+#define TIMERRESOLUTION_M    0xffU
+#define TIMERRESOLUTION_G(x) (((x) >> TIMERRESOLUTION_S) & TIMERRESOLUTION_M)
+
+#define TIMESTAMPRESOLUTION_S    8
+#define TIMESTAMPRESOLUTION_M    0xffU
+#define TIMESTAMPRESOLUTION_G(x) \
+	(((x) >> TIMESTAMPRESOLUTION_S) & TIMESTAMPRESOLUTION_M)
+
+#define DELAYEDACKRESOLUTION_S    0
+#define DELAYEDACKRESOLUTION_M    0xffU
+#define DELAYEDACKRESOLUTION_G(x) \
+	(((x) >> DELAYEDACKRESOLUTION_S) & DELAYEDACKRESOLUTION_M)
+
+#define TP_SHIFT_CNT_A 0x7dc0
+#define TP_RXT_MIN_A 0x7d98
+#define TP_RXT_MAX_A 0x7d9c
+#define TP_PERS_MIN_A 0x7da0
+#define TP_PERS_MAX_A 0x7da4
+#define TP_KEEP_IDLE_A 0x7da8
+#define TP_KEEP_INTVL_A 0x7dac
+#define TP_INIT_SRTT_A 0x7db0
+#define TP_DACK_TIMER_A 0x7db4
+#define TP_FINWAIT2_TIMER_A 0x7db8
+
+#define INITSRTT_S    0
+#define INITSRTT_M    0xffffU
+#define INITSRTT_G(x) (((x) >> INITSRTT_S) & INITSRTT_M)
+
+#define PERSMAX_S    0
+#define PERSMAX_M    0x3fffffffU
+#define PERSMAX_V(x) ((x) << PERSMAX_S)
+#define PERSMAX_G(x) (((x) >> PERSMAX_S) & PERSMAX_M)
+
+#define SYNSHIFTMAX_S    24
+#define SYNSHIFTMAX_M    0xffU
+#define SYNSHIFTMAX_V(x) ((x) << SYNSHIFTMAX_S)
+#define SYNSHIFTMAX_G(x) (((x) >> SYNSHIFTMAX_S) & SYNSHIFTMAX_M)
+
+#define RXTSHIFTMAXR1_S    20
+#define RXTSHIFTMAXR1_M    0xfU
+#define RXTSHIFTMAXR1_V(x) ((x) << RXTSHIFTMAXR1_S)
+#define RXTSHIFTMAXR1_G(x) (((x) >> RXTSHIFTMAXR1_S) & RXTSHIFTMAXR1_M)
+
+#define RXTSHIFTMAXR2_S    16
+#define RXTSHIFTMAXR2_M    0xfU
+#define RXTSHIFTMAXR2_V(x) ((x) << RXTSHIFTMAXR2_S)
+#define RXTSHIFTMAXR2_G(x) (((x) >> RXTSHIFTMAXR2_S) & RXTSHIFTMAXR2_M)
+
+#define PERSHIFTBACKOFFMAX_S    12
+#define PERSHIFTBACKOFFMAX_M    0xfU
+#define PERSHIFTBACKOFFMAX_V(x) ((x) << PERSHIFTBACKOFFMAX_S)
+#define PERSHIFTBACKOFFMAX_G(x) \
+	(((x) >> PERSHIFTBACKOFFMAX_S) & PERSHIFTBACKOFFMAX_M)
+
+#define PERSHIFTMAX_S    8
+#define PERSHIFTMAX_M    0xfU
+#define PERSHIFTMAX_V(x) ((x) << PERSHIFTMAX_S)
+#define PERSHIFTMAX_G(x) (((x) >> PERSHIFTMAX_S) & PERSHIFTMAX_M)
+
+#define KEEPALIVEMAXR1_S    4
+#define KEEPALIVEMAXR1_M    0xfU
+#define KEEPALIVEMAXR1_V(x) ((x) << KEEPALIVEMAXR1_S)
+#define KEEPALIVEMAXR1_G(x) (((x) >> KEEPALIVEMAXR1_S) & KEEPALIVEMAXR1_M)
+
+#define KEEPALIVEMAXR2_S    0
+#define KEEPALIVEMAXR2_M    0xfU
+#define KEEPALIVEMAXR2_V(x) ((x) << KEEPALIVEMAXR2_S)
+#define KEEPALIVEMAXR2_G(x) (((x) >> KEEPALIVEMAXR2_S) & KEEPALIVEMAXR2_M)
+
+#define ROWINDEX_S    16
+#define ROWINDEX_V(x) ((x) << ROWINDEX_S)
+
+#define TP_CCTRL_TABLE_A	0x7ddc
+#define TP_PACE_TABLE_A 0x7dd8
+#define TP_MTU_TABLE_A		0x7de4
+
+#define MTUINDEX_S    24
+#define MTUINDEX_V(x) ((x) << MTUINDEX_S)
+
+#define MTUWIDTH_S    16
+#define MTUWIDTH_M    0xfU
+#define MTUWIDTH_V(x) ((x) << MTUWIDTH_S)
+#define MTUWIDTH_G(x) (((x) >> MTUWIDTH_S) & MTUWIDTH_M)
+
+#define MTUVALUE_S    0
+#define MTUVALUE_M    0x3fffU
+#define MTUVALUE_V(x) ((x) << MTUVALUE_S)
+#define MTUVALUE_G(x) (((x) >> MTUVALUE_S) & MTUVALUE_M)
+
+#define TP_RSS_LKP_TABLE_A	0x7dec
+#define TP_CMM_MM_RX_FLST_BASE_A 0x7e60
+#define TP_CMM_MM_TX_FLST_BASE_A 0x7e64
+#define TP_CMM_MM_PS_FLST_BASE_A 0x7e68
+
+#define LKPTBLROWVLD_S    31
+#define LKPTBLROWVLD_V(x) ((x) << LKPTBLROWVLD_S)
+#define LKPTBLROWVLD_F    LKPTBLROWVLD_V(1U)
+
+#define LKPTBLQUEUE1_S    10
+#define LKPTBLQUEUE1_M    0x3ffU
+#define LKPTBLQUEUE1_G(x) (((x) >> LKPTBLQUEUE1_S) & LKPTBLQUEUE1_M)
+
+#define LKPTBLQUEUE0_S    0
+#define LKPTBLQUEUE0_M    0x3ffU
+#define LKPTBLQUEUE0_G(x) (((x) >> LKPTBLQUEUE0_S) & LKPTBLQUEUE0_M)
+
+#define TP_TM_PIO_ADDR_A 0x7e18
+#define TP_TM_PIO_DATA_A 0x7e1c
+#define TP_MOD_CONFIG_A 0x7e24
+
+#define TIMERMODE_S    8
+#define TIMERMODE_M    0xffU
+#define TIMERMODE_G(x) (((x) >> TIMERMODE_S) & TIMERMODE_M)
+
+#define TP_TX_MOD_Q1_Q0_TIMER_SEPARATOR_A 0x3
+#define TP_TX_MOD_Q1_Q0_RATE_LIMIT_A 0x8
+
+#define TP_PIO_ADDR_A	0x7e40
+#define TP_PIO_DATA_A	0x7e44
+#define TP_MIB_INDEX_A	0x7e50
+#define TP_MIB_DATA_A	0x7e54
+#define TP_INT_CAUSE_A	0x7e74
+
+#define TP_FLM_FREE_PS_CNT_A 0x7e80
+#define TP_FLM_FREE_RX_CNT_A 0x7e84
+
+#define FREEPSTRUCTCOUNT_S    0
+#define FREEPSTRUCTCOUNT_M    0x1fffffU
+#define FREEPSTRUCTCOUNT_G(x) (((x) >> FREEPSTRUCTCOUNT_S) & FREEPSTRUCTCOUNT_M)
+
+#define FREERXPAGECOUNT_S    0
+#define FREERXPAGECOUNT_M    0x1fffffU
+#define FREERXPAGECOUNT_V(x) ((x) << FREERXPAGECOUNT_S)
+#define FREERXPAGECOUNT_G(x) (((x) >> FREERXPAGECOUNT_S) & FREERXPAGECOUNT_M)
+
+#define TP_FLM_FREE_TX_CNT_A 0x7e88
+
+#define FREETXPAGECOUNT_S    0
+#define FREETXPAGECOUNT_M    0x1fffffU
+#define FREETXPAGECOUNT_V(x) ((x) << FREETXPAGECOUNT_S)
+#define FREETXPAGECOUNT_G(x) (((x) >> FREETXPAGECOUNT_S) & FREETXPAGECOUNT_M)
+
+#define FLMTXFLSTEMPTY_S    30
+#define FLMTXFLSTEMPTY_V(x) ((x) << FLMTXFLSTEMPTY_S)
+#define FLMTXFLSTEMPTY_F    FLMTXFLSTEMPTY_V(1U)
+
+#define TP_TX_ORATE_A 0x7ebc
+
+#define OFDRATE3_S    24
+#define OFDRATE3_M    0xffU
+#define OFDRATE3_G(x) (((x) >> OFDRATE3_S) & OFDRATE3_M)
+
+#define OFDRATE2_S    16
+#define OFDRATE2_M    0xffU
+#define OFDRATE2_G(x) (((x) >> OFDRATE2_S) & OFDRATE2_M)
+
+#define OFDRATE1_S    8
+#define OFDRATE1_M    0xffU
+#define OFDRATE1_G(x) (((x) >> OFDRATE1_S) & OFDRATE1_M)
+
+#define OFDRATE0_S    0
+#define OFDRATE0_M    0xffU
+#define OFDRATE0_G(x) (((x) >> OFDRATE0_S) & OFDRATE0_M)
+
+#define TP_TX_TRATE_A 0x7ed0
+
+#define TNLRATE3_S    24
+#define TNLRATE3_M    0xffU
+#define TNLRATE3_G(x) (((x) >> TNLRATE3_S) & TNLRATE3_M)
+
+#define TNLRATE2_S    16
+#define TNLRATE2_M    0xffU
+#define TNLRATE2_G(x) (((x) >> TNLRATE2_S) & TNLRATE2_M)
+
+#define TNLRATE1_S    8
+#define TNLRATE1_M    0xffU
+#define TNLRATE1_G(x) (((x) >> TNLRATE1_S) & TNLRATE1_M)
+
+#define TNLRATE0_S    0
+#define TNLRATE0_M    0xffU
+#define TNLRATE0_G(x) (((x) >> TNLRATE0_S) & TNLRATE0_M)
+
+#define TP_VLAN_PRI_MAP_A 0x140
+
+#define FRAGMENTATION_S    9
+#define FRAGMENTATION_V(x) ((x) << FRAGMENTATION_S)
+#define FRAGMENTATION_F    FRAGMENTATION_V(1U)
+
+#define MPSHITTYPE_S    8
+#define MPSHITTYPE_V(x) ((x) << MPSHITTYPE_S)
+#define MPSHITTYPE_F    MPSHITTYPE_V(1U)
+
+#define MACMATCH_S    7
+#define MACMATCH_V(x) ((x) << MACMATCH_S)
+#define MACMATCH_F    MACMATCH_V(1U)
+
+#define ETHERTYPE_S    6
+#define ETHERTYPE_V(x) ((x) << ETHERTYPE_S)
+#define ETHERTYPE_F    ETHERTYPE_V(1U)
+
+#define PROTOCOL_S    5
+#define PROTOCOL_V(x) ((x) << PROTOCOL_S)
+#define PROTOCOL_F    PROTOCOL_V(1U)
+
+#define TOS_S    4
+#define TOS_V(x) ((x) << TOS_S)
+#define TOS_F    TOS_V(1U)
+
+#define VLAN_S    3
+#define VLAN_V(x) ((x) << VLAN_S)
+#define VLAN_F    VLAN_V(1U)
+
+#define VNIC_ID_S    2
+#define VNIC_ID_V(x) ((x) << VNIC_ID_S)
+#define VNIC_ID_F    VNIC_ID_V(1U)
+
+#define PORT_S    1
+#define PORT_V(x) ((x) << PORT_S)
+#define PORT_F    PORT_V(1U)
+
+#define FCOE_S    0
+#define FCOE_V(x) ((x) << FCOE_S)
+#define FCOE_F    FCOE_V(1U)
+
+#define FILTERMODE_S    15
+#define FILTERMODE_V(x) ((x) << FILTERMODE_S)
+#define FILTERMODE_F    FILTERMODE_V(1U)
+
+#define FCOEMASK_S    14
+#define FCOEMASK_V(x) ((x) << FCOEMASK_S)
+#define FCOEMASK_F    FCOEMASK_V(1U)
+
+#define TP_INGRESS_CONFIG_A	0x141
+
+#define VNIC_S    11
+#define VNIC_V(x) ((x) << VNIC_S)
+#define VNIC_F    VNIC_V(1U)
+
+#define USE_ENC_IDX_S		13
+#define USE_ENC_IDX_V(x)	((x) << USE_ENC_IDX_S)
+#define USE_ENC_IDX_F		USE_ENC_IDX_V(1U)
+
+#define CSUM_HAS_PSEUDO_HDR_S    10
+#define CSUM_HAS_PSEUDO_HDR_V(x) ((x) << CSUM_HAS_PSEUDO_HDR_S)
+#define CSUM_HAS_PSEUDO_HDR_F    CSUM_HAS_PSEUDO_HDR_V(1U)
+
+#define TP_MIB_MAC_IN_ERR_0_A	0x0
+#define TP_MIB_HDR_IN_ERR_0_A	0x4
+#define TP_MIB_TCP_IN_ERR_0_A	0x8
+#define TP_MIB_TCP_OUT_RST_A	0xc
+#define TP_MIB_TCP_IN_SEG_HI_A	0x10
+#define TP_MIB_TCP_IN_SEG_LO_A	0x11
+#define TP_MIB_TCP_OUT_SEG_HI_A	0x12
+#define TP_MIB_TCP_OUT_SEG_LO_A 0x13
+#define TP_MIB_TCP_RXT_SEG_HI_A	0x14
+#define TP_MIB_TCP_RXT_SEG_LO_A	0x15
+#define TP_MIB_TNL_CNG_DROP_0_A 0x18
+#define TP_MIB_OFD_CHN_DROP_0_A 0x1c
+#define TP_MIB_TCP_V6IN_ERR_0_A 0x28
+#define TP_MIB_TCP_V6OUT_RST_A	0x2c
+#define TP_MIB_OFD_ARP_DROP_A	0x36
+#define TP_MIB_CPL_IN_REQ_0_A	0x38
+#define TP_MIB_CPL_OUT_RSP_0_A	0x3c
+#define TP_MIB_TNL_DROP_0_A	0x44
+#define TP_MIB_FCOE_DDP_0_A	0x48
+#define TP_MIB_FCOE_DROP_0_A	0x4c
+#define TP_MIB_FCOE_BYTE_0_HI_A	0x50
+#define TP_MIB_OFD_VLN_DROP_0_A	0x58
+#define TP_MIB_USM_PKTS_A	0x5c
+#define TP_MIB_RQE_DFR_PKT_A	0x64
+
+#define ULP_TX_INT_CAUSE_A	0x8dcc
+#define ULP_TX_TPT_LLIMIT_A	0x8dd4
+#define ULP_TX_TPT_ULIMIT_A	0x8dd8
+#define ULP_TX_PBL_LLIMIT_A	0x8ddc
+#define ULP_TX_PBL_ULIMIT_A	0x8de0
+#define ULP_TX_ERR_TABLE_BASE_A 0x8e04
+
+#define PBL_BOUND_ERR_CH3_S    31
+#define PBL_BOUND_ERR_CH3_V(x) ((x) << PBL_BOUND_ERR_CH3_S)
+#define PBL_BOUND_ERR_CH3_F    PBL_BOUND_ERR_CH3_V(1U)
+
+#define PBL_BOUND_ERR_CH2_S    30
+#define PBL_BOUND_ERR_CH2_V(x) ((x) << PBL_BOUND_ERR_CH2_S)
+#define PBL_BOUND_ERR_CH2_F    PBL_BOUND_ERR_CH2_V(1U)
+
+#define PBL_BOUND_ERR_CH1_S    29
+#define PBL_BOUND_ERR_CH1_V(x) ((x) << PBL_BOUND_ERR_CH1_S)
+#define PBL_BOUND_ERR_CH1_F    PBL_BOUND_ERR_CH1_V(1U)
+
+#define PBL_BOUND_ERR_CH0_S    28
+#define PBL_BOUND_ERR_CH0_V(x) ((x) << PBL_BOUND_ERR_CH0_S)
+#define PBL_BOUND_ERR_CH0_F    PBL_BOUND_ERR_CH0_V(1U)
+
+#define PM_RX_INT_CAUSE_A	0x8fdc
+#define PM_RX_STAT_CONFIG_A 0x8fc8
+#define PM_RX_STAT_COUNT_A 0x8fcc
+#define PM_RX_STAT_LSB_A 0x8fd0
+#define PM_RX_DBG_CTRL_A 0x8fd0
+#define PM_RX_DBG_DATA_A 0x8fd4
+#define PM_RX_DBG_STAT_MSB_A 0x10013
+
+#define PMRX_FRAMING_ERROR_F	0x003ffff0U
+
+#define ZERO_E_CMD_ERROR_S    22
+#define ZERO_E_CMD_ERROR_V(x) ((x) << ZERO_E_CMD_ERROR_S)
+#define ZERO_E_CMD_ERROR_F    ZERO_E_CMD_ERROR_V(1U)
+
+#define OCSPI_PAR_ERROR_S    3
+#define OCSPI_PAR_ERROR_V(x) ((x) << OCSPI_PAR_ERROR_S)
+#define OCSPI_PAR_ERROR_F    OCSPI_PAR_ERROR_V(1U)
+
+#define DB_OPTIONS_PAR_ERROR_S    2
+#define DB_OPTIONS_PAR_ERROR_V(x) ((x) << DB_OPTIONS_PAR_ERROR_S)
+#define DB_OPTIONS_PAR_ERROR_F    DB_OPTIONS_PAR_ERROR_V(1U)
+
+#define IESPI_PAR_ERROR_S    1
+#define IESPI_PAR_ERROR_V(x) ((x) << IESPI_PAR_ERROR_S)
+#define IESPI_PAR_ERROR_F    IESPI_PAR_ERROR_V(1U)
+
+#define ULP_TX_LA_RDPTR_0_A 0x8ec0
+#define ULP_TX_LA_RDDATA_0_A 0x8ec4
+#define ULP_TX_LA_WRPTR_0_A 0x8ec8
+#define ULP_TX_ASIC_DEBUG_CTRL_A 0x8f70
+
+#define ULP_TX_ASIC_DEBUG_0_A 0x8f74
+#define ULP_TX_ASIC_DEBUG_1_A 0x8f78
+#define ULP_TX_ASIC_DEBUG_2_A 0x8f7c
+#define ULP_TX_ASIC_DEBUG_3_A 0x8f80
+#define ULP_TX_ASIC_DEBUG_4_A 0x8f84
+
+/* registers for module PM_RX */
+#define PM_RX_BASE_ADDR 0x8fc0
+
+#define PMRX_E_PCMD_PAR_ERROR_S    0
+#define PMRX_E_PCMD_PAR_ERROR_V(x) ((x) << PMRX_E_PCMD_PAR_ERROR_S)
+#define PMRX_E_PCMD_PAR_ERROR_F    PMRX_E_PCMD_PAR_ERROR_V(1U)
+
+#define PM_TX_INT_CAUSE_A	0x8ffc
+#define PM_TX_STAT_CONFIG_A 0x8fe8
+#define PM_TX_STAT_COUNT_A 0x8fec
+#define PM_TX_STAT_LSB_A 0x8ff0
+#define PM_TX_DBG_CTRL_A 0x8ff0
+#define PM_TX_DBG_DATA_A 0x8ff4
+#define PM_TX_DBG_STAT_MSB_A 0x1001a
+
+#define PCMD_LEN_OVFL0_S    31
+#define PCMD_LEN_OVFL0_V(x) ((x) << PCMD_LEN_OVFL0_S)
+#define PCMD_LEN_OVFL0_F    PCMD_LEN_OVFL0_V(1U)
+
+#define PCMD_LEN_OVFL1_S    30
+#define PCMD_LEN_OVFL1_V(x) ((x) << PCMD_LEN_OVFL1_S)
+#define PCMD_LEN_OVFL1_F    PCMD_LEN_OVFL1_V(1U)
+
+#define PCMD_LEN_OVFL2_S    29
+#define PCMD_LEN_OVFL2_V(x) ((x) << PCMD_LEN_OVFL2_S)
+#define PCMD_LEN_OVFL2_F    PCMD_LEN_OVFL2_V(1U)
+
+#define ZERO_C_CMD_ERROR_S    28
+#define ZERO_C_CMD_ERROR_V(x) ((x) << ZERO_C_CMD_ERROR_S)
+#define ZERO_C_CMD_ERROR_F    ZERO_C_CMD_ERROR_V(1U)
+
+#define  PMTX_FRAMING_ERROR_F 0x0ffffff0U
+
+#define OESPI_PAR_ERROR_S    3
+#define OESPI_PAR_ERROR_V(x) ((x) << OESPI_PAR_ERROR_S)
+#define OESPI_PAR_ERROR_F    OESPI_PAR_ERROR_V(1U)
+
+#define ICSPI_PAR_ERROR_S    1
+#define ICSPI_PAR_ERROR_V(x) ((x) << ICSPI_PAR_ERROR_S)
+#define ICSPI_PAR_ERROR_F    ICSPI_PAR_ERROR_V(1U)
+
+#define PMTX_C_PCMD_PAR_ERROR_S    0
+#define PMTX_C_PCMD_PAR_ERROR_V(x) ((x) << PMTX_C_PCMD_PAR_ERROR_S)
+#define PMTX_C_PCMD_PAR_ERROR_F    PMTX_C_PCMD_PAR_ERROR_V(1U)
+
+#define MPS_PORT_STAT_TX_PORT_BYTES_L 0x400
+#define MPS_PORT_STAT_TX_PORT_BYTES_H 0x404
+#define MPS_PORT_STAT_TX_PORT_FRAMES_L 0x408
+#define MPS_PORT_STAT_TX_PORT_FRAMES_H 0x40c
+#define MPS_PORT_STAT_TX_PORT_BCAST_L 0x410
+#define MPS_PORT_STAT_TX_PORT_BCAST_H 0x414
+#define MPS_PORT_STAT_TX_PORT_MCAST_L 0x418
+#define MPS_PORT_STAT_TX_PORT_MCAST_H 0x41c
+#define MPS_PORT_STAT_TX_PORT_UCAST_L 0x420
+#define MPS_PORT_STAT_TX_PORT_UCAST_H 0x424
+#define MPS_PORT_STAT_TX_PORT_ERROR_L 0x428
+#define MPS_PORT_STAT_TX_PORT_ERROR_H 0x42c
+#define MPS_PORT_STAT_TX_PORT_64B_L 0x430
+#define MPS_PORT_STAT_TX_PORT_64B_H 0x434
+#define MPS_PORT_STAT_TX_PORT_65B_127B_L 0x438
+#define MPS_PORT_STAT_TX_PORT_65B_127B_H 0x43c
+#define MPS_PORT_STAT_TX_PORT_128B_255B_L 0x440
+#define MPS_PORT_STAT_TX_PORT_128B_255B_H 0x444
+#define MPS_PORT_STAT_TX_PORT_256B_511B_L 0x448
+#define MPS_PORT_STAT_TX_PORT_256B_511B_H 0x44c
+#define MPS_PORT_STAT_TX_PORT_512B_1023B_L 0x450
+#define MPS_PORT_STAT_TX_PORT_512B_1023B_H 0x454
+#define MPS_PORT_STAT_TX_PORT_1024B_1518B_L 0x458
+#define MPS_PORT_STAT_TX_PORT_1024B_1518B_H 0x45c
+#define MPS_PORT_STAT_TX_PORT_1519B_MAX_L 0x460
+#define MPS_PORT_STAT_TX_PORT_1519B_MAX_H 0x464
+#define MPS_PORT_STAT_TX_PORT_DROP_L 0x468
+#define MPS_PORT_STAT_TX_PORT_DROP_H 0x46c
+#define MPS_PORT_STAT_TX_PORT_PAUSE_L 0x470
+#define MPS_PORT_STAT_TX_PORT_PAUSE_H 0x474
+#define MPS_PORT_STAT_TX_PORT_PPP0_L 0x478
+#define MPS_PORT_STAT_TX_PORT_PPP0_H 0x47c
+#define MPS_PORT_STAT_TX_PORT_PPP1_L 0x480
+#define MPS_PORT_STAT_TX_PORT_PPP1_H 0x484
+#define MPS_PORT_STAT_TX_PORT_PPP2_L 0x488
+#define MPS_PORT_STAT_TX_PORT_PPP2_H 0x48c
+#define MPS_PORT_STAT_TX_PORT_PPP3_L 0x490
+#define MPS_PORT_STAT_TX_PORT_PPP3_H 0x494
+#define MPS_PORT_STAT_TX_PORT_PPP4_L 0x498
+#define MPS_PORT_STAT_TX_PORT_PPP4_H 0x49c
+#define MPS_PORT_STAT_TX_PORT_PPP5_L 0x4a0
+#define MPS_PORT_STAT_TX_PORT_PPP5_H 0x4a4
+#define MPS_PORT_STAT_TX_PORT_PPP6_L 0x4a8
+#define MPS_PORT_STAT_TX_PORT_PPP6_H 0x4ac
+#define MPS_PORT_STAT_TX_PORT_PPP7_L 0x4b0
+#define MPS_PORT_STAT_TX_PORT_PPP7_H 0x4b4
+#define MPS_PORT_STAT_LB_PORT_BYTES_L 0x4c0
+#define MPS_PORT_STAT_LB_PORT_BYTES_H 0x4c4
+#define MPS_PORT_STAT_LB_PORT_FRAMES_L 0x4c8
+#define MPS_PORT_STAT_LB_PORT_FRAMES_H 0x4cc
+#define MPS_PORT_STAT_LB_PORT_BCAST_L 0x4d0
+#define MPS_PORT_STAT_LB_PORT_BCAST_H 0x4d4
+#define MPS_PORT_STAT_LB_PORT_MCAST_L 0x4d8
+#define MPS_PORT_STAT_LB_PORT_MCAST_H 0x4dc
+#define MPS_PORT_STAT_LB_PORT_UCAST_L 0x4e0
+#define MPS_PORT_STAT_LB_PORT_UCAST_H 0x4e4
+#define MPS_PORT_STAT_LB_PORT_ERROR_L 0x4e8
+#define MPS_PORT_STAT_LB_PORT_ERROR_H 0x4ec
+#define MPS_PORT_STAT_LB_PORT_64B_L 0x4f0
+#define MPS_PORT_STAT_LB_PORT_64B_H 0x4f4
+#define MPS_PORT_STAT_LB_PORT_65B_127B_L 0x4f8
+#define MPS_PORT_STAT_LB_PORT_65B_127B_H 0x4fc
+#define MPS_PORT_STAT_LB_PORT_128B_255B_L 0x500
+#define MPS_PORT_STAT_LB_PORT_128B_255B_H 0x504
+#define MPS_PORT_STAT_LB_PORT_256B_511B_L 0x508
+#define MPS_PORT_STAT_LB_PORT_256B_511B_H 0x50c
+#define MPS_PORT_STAT_LB_PORT_512B_1023B_L 0x510
+#define MPS_PORT_STAT_LB_PORT_512B_1023B_H 0x514
+#define MPS_PORT_STAT_LB_PORT_1024B_1518B_L 0x518
+#define MPS_PORT_STAT_LB_PORT_1024B_1518B_H 0x51c
+#define MPS_PORT_STAT_LB_PORT_1519B_MAX_L 0x520
+#define MPS_PORT_STAT_LB_PORT_1519B_MAX_H 0x524
+#define MPS_PORT_STAT_LB_PORT_DROP_FRAMES 0x528
+#define MPS_PORT_STAT_LB_PORT_DROP_FRAMES_L 0x528
+#define MPS_PORT_STAT_RX_PORT_BYTES_L 0x540
+#define MPS_PORT_STAT_RX_PORT_BYTES_H 0x544
+#define MPS_PORT_STAT_RX_PORT_FRAMES_L 0x548
+#define MPS_PORT_STAT_RX_PORT_FRAMES_H 0x54c
+#define MPS_PORT_STAT_RX_PORT_BCAST_L 0x550
+#define MPS_PORT_STAT_RX_PORT_BCAST_H 0x554
+#define MPS_PORT_STAT_RX_PORT_MCAST_L 0x558
+#define MPS_PORT_STAT_RX_PORT_MCAST_H 0x55c
+#define MPS_PORT_STAT_RX_PORT_UCAST_L 0x560
+#define MPS_PORT_STAT_RX_PORT_UCAST_H 0x564
+#define MPS_PORT_STAT_RX_PORT_MTU_ERROR_L 0x568
+#define MPS_PORT_STAT_RX_PORT_MTU_ERROR_H 0x56c
+#define MPS_PORT_STAT_RX_PORT_MTU_CRC_ERROR_L 0x570
+#define MPS_PORT_STAT_RX_PORT_MTU_CRC_ERROR_H 0x574
+#define MPS_PORT_STAT_RX_PORT_CRC_ERROR_L 0x578
+#define MPS_PORT_STAT_RX_PORT_CRC_ERROR_H 0x57c
+#define MPS_PORT_STAT_RX_PORT_LEN_ERROR_L 0x580
+#define MPS_PORT_STAT_RX_PORT_LEN_ERROR_H 0x584
+#define MPS_PORT_STAT_RX_PORT_SYM_ERROR_L 0x588
+#define MPS_PORT_STAT_RX_PORT_SYM_ERROR_H 0x58c
+#define MPS_PORT_STAT_RX_PORT_64B_L 0x590
+#define MPS_PORT_STAT_RX_PORT_64B_H 0x594
+#define MPS_PORT_STAT_RX_PORT_65B_127B_L 0x598
+#define MPS_PORT_STAT_RX_PORT_65B_127B_H 0x59c
+#define MPS_PORT_STAT_RX_PORT_128B_255B_L 0x5a0
+#define MPS_PORT_STAT_RX_PORT_128B_255B_H 0x5a4
+#define MPS_PORT_STAT_RX_PORT_256B_511B_L 0x5a8
+#define MPS_PORT_STAT_RX_PORT_256B_511B_H 0x5ac
+#define MPS_PORT_STAT_RX_PORT_512B_1023B_L 0x5b0
+#define MPS_PORT_STAT_RX_PORT_512B_1023B_H 0x5b4
+#define MPS_PORT_STAT_RX_PORT_1024B_1518B_L 0x5b8
+#define MPS_PORT_STAT_RX_PORT_1024B_1518B_H 0x5bc
+#define MPS_PORT_STAT_RX_PORT_1519B_MAX_L 0x5c0
+#define MPS_PORT_STAT_RX_PORT_1519B_MAX_H 0x5c4
+#define MPS_PORT_STAT_RX_PORT_PAUSE_L 0x5c8
+#define MPS_PORT_STAT_RX_PORT_PAUSE_H 0x5cc
+#define MPS_PORT_STAT_RX_PORT_PPP0_L 0x5d0
+#define MPS_PORT_STAT_RX_PORT_PPP0_H 0x5d4
+#define MPS_PORT_STAT_RX_PORT_PPP1_L 0x5d8
+#define MPS_PORT_STAT_RX_PORT_PPP1_H 0x5dc
+#define MPS_PORT_STAT_RX_PORT_PPP2_L 0x5e0
+#define MPS_PORT_STAT_RX_PORT_PPP2_H 0x5e4
+#define MPS_PORT_STAT_RX_PORT_PPP3_L 0x5e8
+#define MPS_PORT_STAT_RX_PORT_PPP3_H 0x5ec
+#define MPS_PORT_STAT_RX_PORT_PPP4_L 0x5f0
+#define MPS_PORT_STAT_RX_PORT_PPP4_H 0x5f4
+#define MPS_PORT_STAT_RX_PORT_PPP5_L 0x5f8
+#define MPS_PORT_STAT_RX_PORT_PPP5_H 0x5fc
+#define MPS_PORT_STAT_RX_PORT_PPP6_L 0x600
+#define MPS_PORT_STAT_RX_PORT_PPP6_H 0x604
+#define MPS_PORT_STAT_RX_PORT_PPP7_L 0x608
+#define MPS_PORT_STAT_RX_PORT_PPP7_H 0x60c
+#define MPS_PORT_STAT_RX_PORT_LESS_64B_L 0x610
+#define MPS_PORT_STAT_RX_PORT_LESS_64B_H 0x614
+#define MAC_PORT_MAGIC_MACID_LO 0x824
+#define MAC_PORT_MAGIC_MACID_HI 0x828
+#define MAC_PORT_TX_TS_VAL_LO   0x928
+#define MAC_PORT_TX_TS_VAL_HI   0x92c
+
+#define MAC_PORT_EPIO_DATA0_A 0x8c0
+#define MAC_PORT_EPIO_DATA1_A 0x8c4
+#define MAC_PORT_EPIO_DATA2_A 0x8c8
+#define MAC_PORT_EPIO_DATA3_A 0x8cc
+#define MAC_PORT_EPIO_OP_A 0x8d0
+
+#define MAC_PORT_CFG2_A 0x818
+
+#define MAC_PORT_PTP_SUM_LO_A 0x990
+#define MAC_PORT_PTP_SUM_HI_A 0x994
+
+#define MPS_CMN_CTL_A	0x9000
+
+#define COUNTPAUSEMCRX_S    5
+#define COUNTPAUSEMCRX_V(x) ((x) << COUNTPAUSEMCRX_S)
+#define COUNTPAUSEMCRX_F    COUNTPAUSEMCRX_V(1U)
+
+#define COUNTPAUSESTATRX_S    4
+#define COUNTPAUSESTATRX_V(x) ((x) << COUNTPAUSESTATRX_S)
+#define COUNTPAUSESTATRX_F    COUNTPAUSESTATRX_V(1U)
+
+#define COUNTPAUSEMCTX_S    3
+#define COUNTPAUSEMCTX_V(x) ((x) << COUNTPAUSEMCTX_S)
+#define COUNTPAUSEMCTX_F    COUNTPAUSEMCTX_V(1U)
+
+#define COUNTPAUSESTATTX_S    2
+#define COUNTPAUSESTATTX_V(x) ((x) << COUNTPAUSESTATTX_S)
+#define COUNTPAUSESTATTX_F    COUNTPAUSESTATTX_V(1U)
+
+#define NUMPORTS_S    0
+#define NUMPORTS_M    0x3U
+#define NUMPORTS_G(x) (((x) >> NUMPORTS_S) & NUMPORTS_M)
+
+#define MPS_INT_CAUSE_A 0x9008
+#define MPS_TX_INT_CAUSE_A 0x9408
+#define MPS_STAT_CTL_A 0x9600
+
+#define FRMERR_S    15
+#define FRMERR_V(x) ((x) << FRMERR_S)
+#define FRMERR_F    FRMERR_V(1U)
+
+#define SECNTERR_S    14
+#define SECNTERR_V(x) ((x) << SECNTERR_S)
+#define SECNTERR_F    SECNTERR_V(1U)
+
+#define BUBBLE_S    13
+#define BUBBLE_V(x) ((x) << BUBBLE_S)
+#define BUBBLE_F    BUBBLE_V(1U)
+
+#define TXDESCFIFO_S    9
+#define TXDESCFIFO_M    0xfU
+#define TXDESCFIFO_V(x) ((x) << TXDESCFIFO_S)
+
+#define TXDATAFIFO_S    5
+#define TXDATAFIFO_M    0xfU
+#define TXDATAFIFO_V(x) ((x) << TXDATAFIFO_S)
+
+#define NCSIFIFO_S    4
+#define NCSIFIFO_V(x) ((x) << NCSIFIFO_S)
+#define NCSIFIFO_F    NCSIFIFO_V(1U)
+
+#define TPFIFO_S    0
+#define TPFIFO_M    0xfU
+#define TPFIFO_V(x) ((x) << TPFIFO_S)
+
+#define MPS_STAT_PERR_INT_CAUSE_SRAM_A		0x9614
+#define MPS_STAT_PERR_INT_CAUSE_TX_FIFO_A	0x9620
+#define MPS_STAT_PERR_INT_CAUSE_RX_FIFO_A	0x962c
+
+#define MPS_STAT_RX_BG_0_MAC_DROP_FRAME_L 0x9640
+#define MPS_STAT_RX_BG_0_MAC_DROP_FRAME_H 0x9644
+#define MPS_STAT_RX_BG_1_MAC_DROP_FRAME_L 0x9648
+#define MPS_STAT_RX_BG_1_MAC_DROP_FRAME_H 0x964c
+#define MPS_STAT_RX_BG_2_MAC_DROP_FRAME_L 0x9650
+#define MPS_STAT_RX_BG_2_MAC_DROP_FRAME_H 0x9654
+#define MPS_STAT_RX_BG_3_MAC_DROP_FRAME_L 0x9658
+#define MPS_STAT_RX_BG_3_MAC_DROP_FRAME_H 0x965c
+#define MPS_STAT_RX_BG_0_LB_DROP_FRAME_L 0x9660
+#define MPS_STAT_RX_BG_0_LB_DROP_FRAME_H 0x9664
+#define MPS_STAT_RX_BG_1_LB_DROP_FRAME_L 0x9668
+#define MPS_STAT_RX_BG_1_LB_DROP_FRAME_H 0x966c
+#define MPS_STAT_RX_BG_2_LB_DROP_FRAME_L 0x9670
+#define MPS_STAT_RX_BG_2_LB_DROP_FRAME_H 0x9674
+#define MPS_STAT_RX_BG_3_LB_DROP_FRAME_L 0x9678
+#define MPS_STAT_RX_BG_3_LB_DROP_FRAME_H 0x967c
+#define MPS_STAT_RX_BG_0_MAC_TRUNC_FRAME_L 0x9680
+#define MPS_STAT_RX_BG_0_MAC_TRUNC_FRAME_H 0x9684
+#define MPS_STAT_RX_BG_1_MAC_TRUNC_FRAME_L 0x9688
+#define MPS_STAT_RX_BG_1_MAC_TRUNC_FRAME_H 0x968c
+#define MPS_STAT_RX_BG_2_MAC_TRUNC_FRAME_L 0x9690
+#define MPS_STAT_RX_BG_2_MAC_TRUNC_FRAME_H 0x9694
+#define MPS_STAT_RX_BG_3_MAC_TRUNC_FRAME_L 0x9698
+#define MPS_STAT_RX_BG_3_MAC_TRUNC_FRAME_H 0x969c
+#define MPS_STAT_RX_BG_0_LB_TRUNC_FRAME_L 0x96a0
+#define MPS_STAT_RX_BG_0_LB_TRUNC_FRAME_H 0x96a4
+#define MPS_STAT_RX_BG_1_LB_TRUNC_FRAME_L 0x96a8
+#define MPS_STAT_RX_BG_1_LB_TRUNC_FRAME_H 0x96ac
+#define MPS_STAT_RX_BG_2_LB_TRUNC_FRAME_L 0x96b0
+#define MPS_STAT_RX_BG_2_LB_TRUNC_FRAME_H 0x96b4
+#define MPS_STAT_RX_BG_3_LB_TRUNC_FRAME_L 0x96b8
+#define MPS_STAT_RX_BG_3_LB_TRUNC_FRAME_H 0x96bc
+
+#define MPS_TRC_CFG_A 0x9800
+
+#define TRCFIFOEMPTY_S    4
+#define TRCFIFOEMPTY_V(x) ((x) << TRCFIFOEMPTY_S)
+#define TRCFIFOEMPTY_F    TRCFIFOEMPTY_V(1U)
+
+#define TRCIGNOREDROPINPUT_S    3
+#define TRCIGNOREDROPINPUT_V(x) ((x) << TRCIGNOREDROPINPUT_S)
+#define TRCIGNOREDROPINPUT_F    TRCIGNOREDROPINPUT_V(1U)
+
+#define TRCKEEPDUPLICATES_S    2
+#define TRCKEEPDUPLICATES_V(x) ((x) << TRCKEEPDUPLICATES_S)
+#define TRCKEEPDUPLICATES_F    TRCKEEPDUPLICATES_V(1U)
+
+#define TRCEN_S    1
+#define TRCEN_V(x) ((x) << TRCEN_S)
+#define TRCEN_F    TRCEN_V(1U)
+
+#define TRCMULTIFILTER_S    0
+#define TRCMULTIFILTER_V(x) ((x) << TRCMULTIFILTER_S)
+#define TRCMULTIFILTER_F    TRCMULTIFILTER_V(1U)
+
+#define MPS_TRC_RSS_CONTROL_A		0x9808
+#define MPS_TRC_FILTER1_RSS_CONTROL_A	0x9ff4
+#define MPS_TRC_FILTER2_RSS_CONTROL_A	0x9ffc
+#define MPS_TRC_FILTER3_RSS_CONTROL_A	0xa004
+#define MPS_T5_TRC_RSS_CONTROL_A	0xa00c
+
+#define RSSCONTROL_S    16
+#define RSSCONTROL_V(x) ((x) << RSSCONTROL_S)
+
+#define QUEUENUMBER_S    0
+#define QUEUENUMBER_V(x) ((x) << QUEUENUMBER_S)
+
+#define TFINVERTMATCH_S    24
+#define TFINVERTMATCH_V(x) ((x) << TFINVERTMATCH_S)
+#define TFINVERTMATCH_F    TFINVERTMATCH_V(1U)
+
+#define TFEN_S    22
+#define TFEN_V(x) ((x) << TFEN_S)
+#define TFEN_F    TFEN_V(1U)
+
+#define TFPORT_S    18
+#define TFPORT_M    0xfU
+#define TFPORT_V(x) ((x) << TFPORT_S)
+#define TFPORT_G(x) (((x) >> TFPORT_S) & TFPORT_M)
+
+#define TFLENGTH_S    8
+#define TFLENGTH_M    0x1fU
+#define TFLENGTH_V(x) ((x) << TFLENGTH_S)
+#define TFLENGTH_G(x) (((x) >> TFLENGTH_S) & TFLENGTH_M)
+
+#define TFOFFSET_S    0
+#define TFOFFSET_M    0x1fU
+#define TFOFFSET_V(x) ((x) << TFOFFSET_S)
+#define TFOFFSET_G(x) (((x) >> TFOFFSET_S) & TFOFFSET_M)
+
+#define T5_TFINVERTMATCH_S    25
+#define T5_TFINVERTMATCH_V(x) ((x) << T5_TFINVERTMATCH_S)
+#define T5_TFINVERTMATCH_F    T5_TFINVERTMATCH_V(1U)
+
+#define T5_TFEN_S    23
+#define T5_TFEN_V(x) ((x) << T5_TFEN_S)
+#define T5_TFEN_F    T5_TFEN_V(1U)
+
+#define T5_TFPORT_S    18
+#define T5_TFPORT_M    0x1fU
+#define T5_TFPORT_V(x) ((x) << T5_TFPORT_S)
+#define T5_TFPORT_G(x) (((x) >> T5_TFPORT_S) & T5_TFPORT_M)
+
+#define MPS_TRC_FILTER_MATCH_CTL_A_A 0x9810
+#define MPS_TRC_FILTER_MATCH_CTL_B_A 0x9820
+
+#define TFMINPKTSIZE_S    16
+#define TFMINPKTSIZE_M    0x1ffU
+#define TFMINPKTSIZE_V(x) ((x) << TFMINPKTSIZE_S)
+#define TFMINPKTSIZE_G(x) (((x) >> TFMINPKTSIZE_S) & TFMINPKTSIZE_M)
+
+#define TFCAPTUREMAX_S    0
+#define TFCAPTUREMAX_M    0x3fffU
+#define TFCAPTUREMAX_V(x) ((x) << TFCAPTUREMAX_S)
+#define TFCAPTUREMAX_G(x) (((x) >> TFCAPTUREMAX_S) & TFCAPTUREMAX_M)
+
+#define MPS_TRC_FILTER0_MATCH_A 0x9c00
+#define MPS_TRC_FILTER0_DONT_CARE_A 0x9c80
+#define MPS_TRC_FILTER1_MATCH_A 0x9d00
+
+#define TP_RSS_CONFIG_A 0x7df0
+
+#define TNL4TUPENIPV6_S    31
+#define TNL4TUPENIPV6_V(x) ((x) << TNL4TUPENIPV6_S)
+#define TNL4TUPENIPV6_F    TNL4TUPENIPV6_V(1U)
+
+#define TNL2TUPENIPV6_S    30
+#define TNL2TUPENIPV6_V(x) ((x) << TNL2TUPENIPV6_S)
+#define TNL2TUPENIPV6_F    TNL2TUPENIPV6_V(1U)
+
+#define TNL4TUPENIPV4_S    29
+#define TNL4TUPENIPV4_V(x) ((x) << TNL4TUPENIPV4_S)
+#define TNL4TUPENIPV4_F    TNL4TUPENIPV4_V(1U)
+
+#define TNL2TUPENIPV4_S    28
+#define TNL2TUPENIPV4_V(x) ((x) << TNL2TUPENIPV4_S)
+#define TNL2TUPENIPV4_F    TNL2TUPENIPV4_V(1U)
+
+#define TNLTCPSEL_S    27
+#define TNLTCPSEL_V(x) ((x) << TNLTCPSEL_S)
+#define TNLTCPSEL_F    TNLTCPSEL_V(1U)
+
+#define TNLIP6SEL_S    26
+#define TNLIP6SEL_V(x) ((x) << TNLIP6SEL_S)
+#define TNLIP6SEL_F    TNLIP6SEL_V(1U)
+
+#define TNLVRTSEL_S    25
+#define TNLVRTSEL_V(x) ((x) << TNLVRTSEL_S)
+#define TNLVRTSEL_F    TNLVRTSEL_V(1U)
+
+#define TNLMAPEN_S    24
+#define TNLMAPEN_V(x) ((x) << TNLMAPEN_S)
+#define TNLMAPEN_F    TNLMAPEN_V(1U)
+
+#define OFDHASHSAVE_S    19
+#define OFDHASHSAVE_V(x) ((x) << OFDHASHSAVE_S)
+#define OFDHASHSAVE_F    OFDHASHSAVE_V(1U)
+
+#define OFDVRTSEL_S    18
+#define OFDVRTSEL_V(x) ((x) << OFDVRTSEL_S)
+#define OFDVRTSEL_F    OFDVRTSEL_V(1U)
+
+#define OFDMAPEN_S    17
+#define OFDMAPEN_V(x) ((x) << OFDMAPEN_S)
+#define OFDMAPEN_F    OFDMAPEN_V(1U)
+
+#define OFDLKPEN_S    16
+#define OFDLKPEN_V(x) ((x) << OFDLKPEN_S)
+#define OFDLKPEN_F    OFDLKPEN_V(1U)
+
+#define SYN4TUPENIPV6_S    15
+#define SYN4TUPENIPV6_V(x) ((x) << SYN4TUPENIPV6_S)
+#define SYN4TUPENIPV6_F    SYN4TUPENIPV6_V(1U)
+
+#define SYN2TUPENIPV6_S    14
+#define SYN2TUPENIPV6_V(x) ((x) << SYN2TUPENIPV6_S)
+#define SYN2TUPENIPV6_F    SYN2TUPENIPV6_V(1U)
+
+#define SYN4TUPENIPV4_S    13
+#define SYN4TUPENIPV4_V(x) ((x) << SYN4TUPENIPV4_S)
+#define SYN4TUPENIPV4_F    SYN4TUPENIPV4_V(1U)
+
+#define SYN2TUPENIPV4_S    12
+#define SYN2TUPENIPV4_V(x) ((x) << SYN2TUPENIPV4_S)
+#define SYN2TUPENIPV4_F    SYN2TUPENIPV4_V(1U)
+
+#define SYNIP6SEL_S    11
+#define SYNIP6SEL_V(x) ((x) << SYNIP6SEL_S)
+#define SYNIP6SEL_F    SYNIP6SEL_V(1U)
+
+#define SYNVRTSEL_S    10
+#define SYNVRTSEL_V(x) ((x) << SYNVRTSEL_S)
+#define SYNVRTSEL_F    SYNVRTSEL_V(1U)
+
+#define SYNMAPEN_S    9
+#define SYNMAPEN_V(x) ((x) << SYNMAPEN_S)
+#define SYNMAPEN_F    SYNMAPEN_V(1U)
+
+#define SYNLKPEN_S    8
+#define SYNLKPEN_V(x) ((x) << SYNLKPEN_S)
+#define SYNLKPEN_F    SYNLKPEN_V(1U)
+
+#define CHANNELENABLE_S    7
+#define CHANNELENABLE_V(x) ((x) << CHANNELENABLE_S)
+#define CHANNELENABLE_F    CHANNELENABLE_V(1U)
+
+#define PORTENABLE_S    6
+#define PORTENABLE_V(x) ((x) << PORTENABLE_S)
+#define PORTENABLE_F    PORTENABLE_V(1U)
+
+#define TNLALLLOOKUP_S    5
+#define TNLALLLOOKUP_V(x) ((x) << TNLALLLOOKUP_S)
+#define TNLALLLOOKUP_F    TNLALLLOOKUP_V(1U)
+
+#define VIRTENABLE_S    4
+#define VIRTENABLE_V(x) ((x) << VIRTENABLE_S)
+#define VIRTENABLE_F    VIRTENABLE_V(1U)
+
+#define CONGESTIONENABLE_S    3
+#define CONGESTIONENABLE_V(x) ((x) << CONGESTIONENABLE_S)
+#define CONGESTIONENABLE_F    CONGESTIONENABLE_V(1U)
+
+#define HASHTOEPLITZ_S    2
+#define HASHTOEPLITZ_V(x) ((x) << HASHTOEPLITZ_S)
+#define HASHTOEPLITZ_F    HASHTOEPLITZ_V(1U)
+
+#define UDPENABLE_S    1
+#define UDPENABLE_V(x) ((x) << UDPENABLE_S)
+#define UDPENABLE_F    UDPENABLE_V(1U)
+
+#define DISABLE_S    0
+#define DISABLE_V(x) ((x) << DISABLE_S)
+#define DISABLE_F    DISABLE_V(1U)
+
+#define TP_RSS_CONFIG_TNL_A 0x7df4
+
+#define MASKSIZE_S    28
+#define MASKSIZE_M    0xfU
+#define MASKSIZE_V(x) ((x) << MASKSIZE_S)
+#define MASKSIZE_G(x) (((x) >> MASKSIZE_S) & MASKSIZE_M)
+
+#define MASKFILTER_S    16
+#define MASKFILTER_M    0x7ffU
+#define MASKFILTER_V(x) ((x) << MASKFILTER_S)
+#define MASKFILTER_G(x) (((x) >> MASKFILTER_S) & MASKFILTER_M)
+
+#define USEWIRECH_S    0
+#define USEWIRECH_V(x) ((x) << USEWIRECH_S)
+#define USEWIRECH_F    USEWIRECH_V(1U)
+
+#define HASHALL_S    2
+#define HASHALL_V(x) ((x) << HASHALL_S)
+#define HASHALL_F    HASHALL_V(1U)
+
+#define HASHETH_S    1
+#define HASHETH_V(x) ((x) << HASHETH_S)
+#define HASHETH_F    HASHETH_V(1U)
+
+#define TP_RSS_CONFIG_OFD_A 0x7df8
+
+#define RRCPLMAPEN_S    20
+#define RRCPLMAPEN_V(x) ((x) << RRCPLMAPEN_S)
+#define RRCPLMAPEN_F    RRCPLMAPEN_V(1U)
+
+#define RRCPLQUEWIDTH_S    16
+#define RRCPLQUEWIDTH_M    0xfU
+#define RRCPLQUEWIDTH_V(x) ((x) << RRCPLQUEWIDTH_S)
+#define RRCPLQUEWIDTH_G(x) (((x) >> RRCPLQUEWIDTH_S) & RRCPLQUEWIDTH_M)
+
+#define TP_RSS_CONFIG_SYN_A 0x7dfc
+#define TP_RSS_CONFIG_VRT_A 0x7e00
+
+#define VFRDRG_S    25
+#define VFRDRG_V(x) ((x) << VFRDRG_S)
+#define VFRDRG_F    VFRDRG_V(1U)
+
+#define VFRDEN_S    24
+#define VFRDEN_V(x) ((x) << VFRDEN_S)
+#define VFRDEN_F    VFRDEN_V(1U)
+
+#define VFPERREN_S    23
+#define VFPERREN_V(x) ((x) << VFPERREN_S)
+#define VFPERREN_F    VFPERREN_V(1U)
+
+#define KEYPERREN_S    22
+#define KEYPERREN_V(x) ((x) << KEYPERREN_S)
+#define KEYPERREN_F    KEYPERREN_V(1U)
+
+#define DISABLEVLAN_S    21
+#define DISABLEVLAN_V(x) ((x) << DISABLEVLAN_S)
+#define DISABLEVLAN_F    DISABLEVLAN_V(1U)
+
+#define ENABLEUP0_S    20
+#define ENABLEUP0_V(x) ((x) << ENABLEUP0_S)
+#define ENABLEUP0_F    ENABLEUP0_V(1U)
+
+#define HASHDELAY_S    16
+#define HASHDELAY_M    0xfU
+#define HASHDELAY_V(x) ((x) << HASHDELAY_S)
+#define HASHDELAY_G(x) (((x) >> HASHDELAY_S) & HASHDELAY_M)
+
+#define VFWRADDR_S    8
+#define VFWRADDR_M    0x7fU
+#define VFWRADDR_V(x) ((x) << VFWRADDR_S)
+#define VFWRADDR_G(x) (((x) >> VFWRADDR_S) & VFWRADDR_M)
+
+#define KEYMODE_S    6
+#define KEYMODE_M    0x3U
+#define KEYMODE_V(x) ((x) << KEYMODE_S)
+#define KEYMODE_G(x) (((x) >> KEYMODE_S) & KEYMODE_M)
+
+#define VFWREN_S    5
+#define VFWREN_V(x) ((x) << VFWREN_S)
+#define VFWREN_F    VFWREN_V(1U)
+
+#define KEYWREN_S    4
+#define KEYWREN_V(x) ((x) << KEYWREN_S)
+#define KEYWREN_F    KEYWREN_V(1U)
+
+#define KEYWRADDR_S    0
+#define KEYWRADDR_M    0xfU
+#define KEYWRADDR_V(x) ((x) << KEYWRADDR_S)
+#define KEYWRADDR_G(x) (((x) >> KEYWRADDR_S) & KEYWRADDR_M)
+
+#define KEYWRADDRX_S    30
+#define KEYWRADDRX_M    0x3U
+#define KEYWRADDRX_V(x) ((x) << KEYWRADDRX_S)
+#define KEYWRADDRX_G(x) (((x) >> KEYWRADDRX_S) & KEYWRADDRX_M)
+
+#define KEYEXTEND_S    26
+#define KEYEXTEND_V(x) ((x) << KEYEXTEND_S)
+#define KEYEXTEND_F    KEYEXTEND_V(1U)
+
+#define LKPIDXSIZE_S    24
+#define LKPIDXSIZE_M    0x3U
+#define LKPIDXSIZE_V(x) ((x) << LKPIDXSIZE_S)
+#define LKPIDXSIZE_G(x) (((x) >> LKPIDXSIZE_S) & LKPIDXSIZE_M)
+
+#define TP_RSS_VFL_CONFIG_A 0x3a
+#define TP_RSS_VFH_CONFIG_A 0x3b
+
+#define ENABLEUDPHASH_S    31
+#define ENABLEUDPHASH_V(x) ((x) << ENABLEUDPHASH_S)
+#define ENABLEUDPHASH_F    ENABLEUDPHASH_V(1U)
+
+#define VFUPEN_S    30
+#define VFUPEN_V(x) ((x) << VFUPEN_S)
+#define VFUPEN_F    VFUPEN_V(1U)
+
+#define VFVLNEX_S    28
+#define VFVLNEX_V(x) ((x) << VFVLNEX_S)
+#define VFVLNEX_F    VFVLNEX_V(1U)
+
+#define VFPRTEN_S    27
+#define VFPRTEN_V(x) ((x) << VFPRTEN_S)
+#define VFPRTEN_F    VFPRTEN_V(1U)
+
+#define VFCHNEN_S    26
+#define VFCHNEN_V(x) ((x) << VFCHNEN_S)
+#define VFCHNEN_F    VFCHNEN_V(1U)
+
+#define DEFAULTQUEUE_S    16
+#define DEFAULTQUEUE_M    0x3ffU
+#define DEFAULTQUEUE_G(x) (((x) >> DEFAULTQUEUE_S) & DEFAULTQUEUE_M)
+
+#define VFIP6TWOTUPEN_S    6
+#define VFIP6TWOTUPEN_V(x) ((x) << VFIP6TWOTUPEN_S)
+#define VFIP6TWOTUPEN_F    VFIP6TWOTUPEN_V(1U)
+
+#define VFIP4FOURTUPEN_S    5
+#define VFIP4FOURTUPEN_V(x) ((x) << VFIP4FOURTUPEN_S)
+#define VFIP4FOURTUPEN_F    VFIP4FOURTUPEN_V(1U)
+
+#define VFIP4TWOTUPEN_S    4
+#define VFIP4TWOTUPEN_V(x) ((x) << VFIP4TWOTUPEN_S)
+#define VFIP4TWOTUPEN_F    VFIP4TWOTUPEN_V(1U)
+
+#define KEYINDEX_S    0
+#define KEYINDEX_M    0xfU
+#define KEYINDEX_G(x) (((x) >> KEYINDEX_S) & KEYINDEX_M)
+
+#define MAPENABLE_S    31
+#define MAPENABLE_V(x) ((x) << MAPENABLE_S)
+#define MAPENABLE_F    MAPENABLE_V(1U)
+
+#define CHNENABLE_S    30
+#define CHNENABLE_V(x) ((x) << CHNENABLE_S)
+#define CHNENABLE_F    CHNENABLE_V(1U)
+
+#define LE_DB_DBGI_CONFIG_A 0x19cf0
+
+#define DBGICMDBUSY_S    3
+#define DBGICMDBUSY_V(x) ((x) << DBGICMDBUSY_S)
+#define DBGICMDBUSY_F    DBGICMDBUSY_V(1U)
+
+#define DBGICMDSTRT_S    2
+#define DBGICMDSTRT_V(x) ((x) << DBGICMDSTRT_S)
+#define DBGICMDSTRT_F    DBGICMDSTRT_V(1U)
+
+#define DBGICMDMODE_S    0
+#define DBGICMDMODE_M    0x3U
+#define DBGICMDMODE_V(x) ((x) << DBGICMDMODE_S)
+
+#define LE_DB_DBGI_REQ_TCAM_CMD_A 0x19cf4
+
+#define DBGICMD_S    20
+#define DBGICMD_M    0xfU
+#define DBGICMD_V(x) ((x) << DBGICMD_S)
+
+#define DBGITID_S    0
+#define DBGITID_M    0xfffffU
+#define DBGITID_V(x) ((x) << DBGITID_S)
+
+#define LE_DB_DBGI_REQ_DATA_A 0x19d00
+#define LE_DB_DBGI_RSP_STATUS_A 0x19d94
+
+#define LE_DB_DBGI_RSP_DATA_A 0x19da0
+
+#define PRTENABLE_S    29
+#define PRTENABLE_V(x) ((x) << PRTENABLE_S)
+#define PRTENABLE_F    PRTENABLE_V(1U)
+
+#define UDPFOURTUPEN_S    28
+#define UDPFOURTUPEN_V(x) ((x) << UDPFOURTUPEN_S)
+#define UDPFOURTUPEN_F    UDPFOURTUPEN_V(1U)
+
+#define IP6FOURTUPEN_S    27
+#define IP6FOURTUPEN_V(x) ((x) << IP6FOURTUPEN_S)
+#define IP6FOURTUPEN_F    IP6FOURTUPEN_V(1U)
+
+#define IP6TWOTUPEN_S    26
+#define IP6TWOTUPEN_V(x) ((x) << IP6TWOTUPEN_S)
+#define IP6TWOTUPEN_F    IP6TWOTUPEN_V(1U)
+
+#define IP4FOURTUPEN_S    25
+#define IP4FOURTUPEN_V(x) ((x) << IP4FOURTUPEN_S)
+#define IP4FOURTUPEN_F    IP4FOURTUPEN_V(1U)
+
+#define IP4TWOTUPEN_S    24
+#define IP4TWOTUPEN_V(x) ((x) << IP4TWOTUPEN_S)
+#define IP4TWOTUPEN_F    IP4TWOTUPEN_V(1U)
+
+#define IVFWIDTH_S    20
+#define IVFWIDTH_M    0xfU
+#define IVFWIDTH_V(x) ((x) << IVFWIDTH_S)
+#define IVFWIDTH_G(x) (((x) >> IVFWIDTH_S) & IVFWIDTH_M)
+
+#define CH1DEFAULTQUEUE_S    10
+#define CH1DEFAULTQUEUE_M    0x3ffU
+#define CH1DEFAULTQUEUE_V(x) ((x) << CH1DEFAULTQUEUE_S)
+#define CH1DEFAULTQUEUE_G(x) (((x) >> CH1DEFAULTQUEUE_S) & CH1DEFAULTQUEUE_M)
+
+#define CH0DEFAULTQUEUE_S    0
+#define CH0DEFAULTQUEUE_M    0x3ffU
+#define CH0DEFAULTQUEUE_V(x) ((x) << CH0DEFAULTQUEUE_S)
+#define CH0DEFAULTQUEUE_G(x) (((x) >> CH0DEFAULTQUEUE_S) & CH0DEFAULTQUEUE_M)
+
+#define VFLKPIDX_S    8
+#define VFLKPIDX_M    0xffU
+#define VFLKPIDX_G(x) (((x) >> VFLKPIDX_S) & VFLKPIDX_M)
+
+#define T6_VFWRADDR_S    8
+#define T6_VFWRADDR_M    0xffU
+#define T6_VFWRADDR_V(x) ((x) << T6_VFWRADDR_S)
+#define T6_VFWRADDR_G(x) (((x) >> T6_VFWRADDR_S) & T6_VFWRADDR_M)
+
+#define TP_RSS_CONFIG_CNG_A 0x7e04
+#define TP_RSS_SECRET_KEY0_A 0x40
+#define TP_RSS_PF0_CONFIG_A 0x30
+#define TP_RSS_PF_MAP_A 0x38
+#define TP_RSS_PF_MSK_A 0x39
+
+#define PF1LKPIDX_S    3
+
+#define PF0LKPIDX_M    0x7U
+
+#define PF1MSKSIZE_S    4
+#define PF1MSKSIZE_M    0xfU
+
+#define CHNCOUNT3_S    31
+#define CHNCOUNT3_V(x) ((x) << CHNCOUNT3_S)
+#define CHNCOUNT3_F    CHNCOUNT3_V(1U)
+
+#define CHNCOUNT2_S    30
+#define CHNCOUNT2_V(x) ((x) << CHNCOUNT2_S)
+#define CHNCOUNT2_F    CHNCOUNT2_V(1U)
+
+#define CHNCOUNT1_S    29
+#define CHNCOUNT1_V(x) ((x) << CHNCOUNT1_S)
+#define CHNCOUNT1_F    CHNCOUNT1_V(1U)
+
+#define CHNCOUNT0_S    28
+#define CHNCOUNT0_V(x) ((x) << CHNCOUNT0_S)
+#define CHNCOUNT0_F    CHNCOUNT0_V(1U)
+
+#define CHNUNDFLOW3_S    27
+#define CHNUNDFLOW3_V(x) ((x) << CHNUNDFLOW3_S)
+#define CHNUNDFLOW3_F    CHNUNDFLOW3_V(1U)
+
+#define CHNUNDFLOW2_S    26
+#define CHNUNDFLOW2_V(x) ((x) << CHNUNDFLOW2_S)
+#define CHNUNDFLOW2_F    CHNUNDFLOW2_V(1U)
+
+#define CHNUNDFLOW1_S    25
+#define CHNUNDFLOW1_V(x) ((x) << CHNUNDFLOW1_S)
+#define CHNUNDFLOW1_F    CHNUNDFLOW1_V(1U)
+
+#define CHNUNDFLOW0_S    24
+#define CHNUNDFLOW0_V(x) ((x) << CHNUNDFLOW0_S)
+#define CHNUNDFLOW0_F    CHNUNDFLOW0_V(1U)
+
+#define RSTCHN3_S    19
+#define RSTCHN3_V(x) ((x) << RSTCHN3_S)
+#define RSTCHN3_F    RSTCHN3_V(1U)
+
+#define RSTCHN2_S    18
+#define RSTCHN2_V(x) ((x) << RSTCHN2_S)
+#define RSTCHN2_F    RSTCHN2_V(1U)
+
+#define RSTCHN1_S    17
+#define RSTCHN1_V(x) ((x) << RSTCHN1_S)
+#define RSTCHN1_F    RSTCHN1_V(1U)
+
+#define RSTCHN0_S    16
+#define RSTCHN0_V(x) ((x) << RSTCHN0_S)
+#define RSTCHN0_F    RSTCHN0_V(1U)
+
+#define UPDVLD_S    15
+#define UPDVLD_V(x) ((x) << UPDVLD_S)
+#define UPDVLD_F    UPDVLD_V(1U)
+
+#define XOFF_S    14
+#define XOFF_V(x) ((x) << XOFF_S)
+#define XOFF_F    XOFF_V(1U)
+
+#define UPDCHN3_S    13
+#define UPDCHN3_V(x) ((x) << UPDCHN3_S)
+#define UPDCHN3_F    UPDCHN3_V(1U)
+
+#define UPDCHN2_S    12
+#define UPDCHN2_V(x) ((x) << UPDCHN2_S)
+#define UPDCHN2_F    UPDCHN2_V(1U)
+
+#define UPDCHN1_S    11
+#define UPDCHN1_V(x) ((x) << UPDCHN1_S)
+#define UPDCHN1_F    UPDCHN1_V(1U)
+
+#define UPDCHN0_S    10
+#define UPDCHN0_V(x) ((x) << UPDCHN0_S)
+#define UPDCHN0_F    UPDCHN0_V(1U)
+
+#define QUEUE_S    0
+#define QUEUE_M    0x3ffU
+#define QUEUE_V(x) ((x) << QUEUE_S)
+#define QUEUE_G(x) (((x) >> QUEUE_S) & QUEUE_M)
+
+#define MPS_TRC_INT_CAUSE_A	0x985c
+
+#define MISCPERR_S    8
+#define MISCPERR_V(x) ((x) << MISCPERR_S)
+#define MISCPERR_F    MISCPERR_V(1U)
+
+#define PKTFIFO_S    4
+#define PKTFIFO_M    0xfU
+#define PKTFIFO_V(x) ((x) << PKTFIFO_S)
+
+#define FILTMEM_S    0
+#define FILTMEM_M    0xfU
+#define FILTMEM_V(x) ((x) << FILTMEM_S)
+
+#define MPS_CLS_INT_CAUSE_A 0xd028
+
+#define HASHSRAM_S    2
+#define HASHSRAM_V(x) ((x) << HASHSRAM_S)
+#define HASHSRAM_F    HASHSRAM_V(1U)
+
+#define MATCHTCAM_S    1
+#define MATCHTCAM_V(x) ((x) << MATCHTCAM_S)
+#define MATCHTCAM_F    MATCHTCAM_V(1U)
+
+#define MATCHSRAM_S    0
+#define MATCHSRAM_V(x) ((x) << MATCHSRAM_S)
+#define MATCHSRAM_F    MATCHSRAM_V(1U)
+
+#define MPS_RX_PG_RSV0_A 0x11010
+#define MPS_RX_PG_RSV4_A 0x11020
+#define MPS_RX_PERR_INT_CAUSE_A 0x11074
+#define MPS_RX_MAC_BG_PG_CNT0_A 0x11208
+#define MPS_RX_LPBK_BG_PG_CNT0_A 0x11218
+
+#define MPS_RX_VXLAN_TYPE_A 0x11234
+
+#define VXLAN_EN_S    16
+#define VXLAN_EN_V(x) ((x) << VXLAN_EN_S)
+#define VXLAN_EN_F    VXLAN_EN_V(1U)
+
+#define VXLAN_S    0
+#define VXLAN_M    0xffffU
+#define VXLAN_V(x) ((x) << VXLAN_S)
+#define VXLAN_G(x) (((x) >> VXLAN_S) & VXLAN_M)
+
+#define MPS_RX_GENEVE_TYPE_A 0x11238
+
+#define GENEVE_EN_S    16
+#define GENEVE_EN_V(x) ((x) << GENEVE_EN_S)
+#define GENEVE_EN_F    GENEVE_EN_V(1U)
+
+#define GENEVE_S    0
+#define GENEVE_M    0xffffU
+#define GENEVE_V(x) ((x) << GENEVE_S)
+#define GENEVE_G(x) (((x) >> GENEVE_S) & GENEVE_M)
+
+#define MPS_CLS_TCAM_Y_L_A 0xf000
+#define MPS_CLS_TCAM_DATA0_A 0xf000
+#define MPS_CLS_TCAM_DATA1_A 0xf004
+
+#define CTLREQID_S    30
+#define CTLREQID_V(x) ((x) << CTLREQID_S)
+
+#define MPS_VF_RPLCT_MAP0_A 0x1111c
+#define MPS_VF_RPLCT_MAP1_A 0x11120
+#define MPS_VF_RPLCT_MAP2_A 0x11124
+#define MPS_VF_RPLCT_MAP3_A 0x11128
+#define MPS_VF_RPLCT_MAP4_A 0x11300
+#define MPS_VF_RPLCT_MAP5_A 0x11304
+#define MPS_VF_RPLCT_MAP6_A 0x11308
+#define MPS_VF_RPLCT_MAP7_A 0x1130c
+
+#define VIDL_S    16
+#define VIDL_M    0xffffU
+#define VIDL_G(x) (((x) >> VIDL_S) & VIDL_M)
+
+#define DATALKPTYPE_S    10
+#define DATALKPTYPE_M    0x3U
+#define DATALKPTYPE_G(x) (((x) >> DATALKPTYPE_S) & DATALKPTYPE_M)
+
+#define DATAPORTNUM_S    12
+#define DATAPORTNUM_M    0xfU
+#define DATAPORTNUM_V(x) ((x) << DATAPORTNUM_S)
+#define DATAPORTNUM_G(x) (((x) >> DATAPORTNUM_S) & DATAPORTNUM_M)
+
+#define DATALKPTYPE_S    10
+#define DATALKPTYPE_M    0x3U
+#define DATALKPTYPE_V(x) ((x) << DATALKPTYPE_S)
+#define DATALKPTYPE_G(x) (((x) >> DATALKPTYPE_S) & DATALKPTYPE_M)
+
+#define DATADIPHIT_S    8
+#define DATADIPHIT_V(x) ((x) << DATADIPHIT_S)
+#define DATADIPHIT_F    DATADIPHIT_V(1U)
+
+#define DATAVIDH2_S    7
+#define DATAVIDH2_V(x) ((x) << DATAVIDH2_S)
+#define DATAVIDH2_F    DATAVIDH2_V(1U)
+
+#define DATAVIDH1_S    0
+#define DATAVIDH1_M    0x7fU
+#define DATAVIDH1_G(x) (((x) >> DATAVIDH1_S) & DATAVIDH1_M)
+
+#define MPS_CLS_TCAM_RDATA0_REQ_ID1_A 0xf020
+#define MPS_CLS_TCAM_RDATA1_REQ_ID1_A 0xf024
+#define MPS_CLS_TCAM_RDATA2_REQ_ID1_A 0xf028
+
+#define USED_S    16
+#define USED_M    0x7ffU
+#define USED_G(x) (((x) >> USED_S) & USED_M)
+
+#define ALLOC_S    0
+#define ALLOC_M    0x7ffU
+#define ALLOC_G(x) (((x) >> ALLOC_S) & ALLOC_M)
+
+#define T5_USED_S    16
+#define T5_USED_M    0xfffU
+#define T5_USED_G(x) (((x) >> T5_USED_S) & T5_USED_M)
+
+#define T5_ALLOC_S    0
+#define T5_ALLOC_M    0xfffU
+#define T5_ALLOC_G(x) (((x) >> T5_ALLOC_S) & T5_ALLOC_M)
+
+#define DMACH_S    0
+#define DMACH_M    0xffffU
+#define DMACH_G(x) (((x) >> DMACH_S) & DMACH_M)
+
+#define MPS_CLS_TCAM_X_L_A 0xf008
+#define MPS_CLS_TCAM_DATA2_CTL_A 0xf008
+
+#define CTLCMDTYPE_S    31
+#define CTLCMDTYPE_V(x) ((x) << CTLCMDTYPE_S)
+#define CTLCMDTYPE_F    CTLCMDTYPE_V(1U)
+
+#define CTLTCAMSEL_S    25
+#define CTLTCAMSEL_V(x) ((x) << CTLTCAMSEL_S)
+
+#define CTLTCAMINDEX_S    17
+#define CTLTCAMINDEX_V(x) ((x) << CTLTCAMINDEX_S)
+
+#define CTLXYBITSEL_S    16
+#define CTLXYBITSEL_V(x) ((x) << CTLXYBITSEL_S)
+
+#define MPS_CLS_TCAM_Y_L(idx) (MPS_CLS_TCAM_Y_L_A + (idx) * 16)
+#define NUM_MPS_CLS_TCAM_Y_L_INSTANCES 512
+
+#define MPS_CLS_TCAM_X_L(idx) (MPS_CLS_TCAM_X_L_A + (idx) * 16)
+#define NUM_MPS_CLS_TCAM_X_L_INSTANCES 512
+
+#define MPS_CLS_SRAM_L_A 0xe000
+
+#define T6_MULTILISTEN0_S    26
+
+#define T6_SRAM_PRIO3_S    23
+#define T6_SRAM_PRIO3_M    0x7U
+#define T6_SRAM_PRIO3_G(x) (((x) >> T6_SRAM_PRIO3_S) & T6_SRAM_PRIO3_M)
+
+#define T6_SRAM_PRIO2_S    20
+#define T6_SRAM_PRIO2_M    0x7U
+#define T6_SRAM_PRIO2_G(x) (((x) >> T6_SRAM_PRIO2_S) & T6_SRAM_PRIO2_M)
+
+#define T6_SRAM_PRIO1_S    17
+#define T6_SRAM_PRIO1_M    0x7U
+#define T6_SRAM_PRIO1_G(x) (((x) >> T6_SRAM_PRIO1_S) & T6_SRAM_PRIO1_M)
+
+#define T6_SRAM_PRIO0_S    14
+#define T6_SRAM_PRIO0_M    0x7U
+#define T6_SRAM_PRIO0_G(x) (((x) >> T6_SRAM_PRIO0_S) & T6_SRAM_PRIO0_M)
+
+#define T6_SRAM_VLD_S    13
+#define T6_SRAM_VLD_V(x) ((x) << T6_SRAM_VLD_S)
+#define T6_SRAM_VLD_F    T6_SRAM_VLD_V(1U)
+
+#define T6_REPLICATE_S    12
+#define T6_REPLICATE_V(x) ((x) << T6_REPLICATE_S)
+#define T6_REPLICATE_F    T6_REPLICATE_V(1U)
+
+#define T6_PF_S    9
+#define T6_PF_M    0x7U
+#define T6_PF_G(x) (((x) >> T6_PF_S) & T6_PF_M)
+
+#define T6_VF_VALID_S    8
+#define T6_VF_VALID_V(x) ((x) << T6_VF_VALID_S)
+#define T6_VF_VALID_F    T6_VF_VALID_V(1U)
+
+#define T6_VF_S    0
+#define T6_VF_M    0xffU
+#define T6_VF_G(x) (((x) >> T6_VF_S) & T6_VF_M)
+
+#define MPS_CLS_SRAM_H_A 0xe004
+
+#define MPS_CLS_SRAM_L(idx) (MPS_CLS_SRAM_L_A + (idx) * 8)
+#define NUM_MPS_CLS_SRAM_L_INSTANCES 336
+
+#define MPS_CLS_SRAM_H(idx) (MPS_CLS_SRAM_H_A + (idx) * 8)
+#define NUM_MPS_CLS_SRAM_H_INSTANCES 336
+
+#define MULTILISTEN0_S    25
+
+#define REPLICATE_S    11
+#define REPLICATE_V(x) ((x) << REPLICATE_S)
+#define REPLICATE_F    REPLICATE_V(1U)
+
+#define PF_S    8
+#define PF_M    0x7U
+#define PF_G(x) (((x) >> PF_S) & PF_M)
+
+#define VF_VALID_S    7
+#define VF_VALID_V(x) ((x) << VF_VALID_S)
+#define VF_VALID_F    VF_VALID_V(1U)
+
+#define VF_S    0
+#define VF_M    0x7fU
+#define VF_G(x) (((x) >> VF_S) & VF_M)
+
+#define SRAM_PRIO3_S    22
+#define SRAM_PRIO3_M    0x7U
+#define SRAM_PRIO3_G(x) (((x) >> SRAM_PRIO3_S) & SRAM_PRIO3_M)
+
+#define SRAM_PRIO2_S    19
+#define SRAM_PRIO2_M    0x7U
+#define SRAM_PRIO2_G(x) (((x) >> SRAM_PRIO2_S) & SRAM_PRIO2_M)
+
+#define SRAM_PRIO1_S    16
+#define SRAM_PRIO1_M    0x7U
+#define SRAM_PRIO1_G(x) (((x) >> SRAM_PRIO1_S) & SRAM_PRIO1_M)
+
+#define SRAM_PRIO0_S    13
+#define SRAM_PRIO0_M    0x7U
+#define SRAM_PRIO0_G(x) (((x) >> SRAM_PRIO0_S) & SRAM_PRIO0_M)
+
+#define SRAM_VLD_S    12
+#define SRAM_VLD_V(x) ((x) << SRAM_VLD_S)
+#define SRAM_VLD_F    SRAM_VLD_V(1U)
+
+#define PORTMAP_S    0
+#define PORTMAP_M    0xfU
+#define PORTMAP_G(x) (((x) >> PORTMAP_S) & PORTMAP_M)
+
+#define CPL_INTR_CAUSE_A 0x19054
+
+#define CIM_OP_MAP_PERR_S    5
+#define CIM_OP_MAP_PERR_V(x) ((x) << CIM_OP_MAP_PERR_S)
+#define CIM_OP_MAP_PERR_F    CIM_OP_MAP_PERR_V(1U)
+
+#define CIM_OVFL_ERROR_S    4
+#define CIM_OVFL_ERROR_V(x) ((x) << CIM_OVFL_ERROR_S)
+#define CIM_OVFL_ERROR_F    CIM_OVFL_ERROR_V(1U)
+
+#define TP_FRAMING_ERROR_S    3
+#define TP_FRAMING_ERROR_V(x) ((x) << TP_FRAMING_ERROR_S)
+#define TP_FRAMING_ERROR_F    TP_FRAMING_ERROR_V(1U)
+
+#define SGE_FRAMING_ERROR_S    2
+#define SGE_FRAMING_ERROR_V(x) ((x) << SGE_FRAMING_ERROR_S)
+#define SGE_FRAMING_ERROR_F    SGE_FRAMING_ERROR_V(1U)
+
+#define CIM_FRAMING_ERROR_S    1
+#define CIM_FRAMING_ERROR_V(x) ((x) << CIM_FRAMING_ERROR_S)
+#define CIM_FRAMING_ERROR_F    CIM_FRAMING_ERROR_V(1U)
+
+#define ZERO_SWITCH_ERROR_S    0
+#define ZERO_SWITCH_ERROR_V(x) ((x) << ZERO_SWITCH_ERROR_S)
+#define ZERO_SWITCH_ERROR_F    ZERO_SWITCH_ERROR_V(1U)
+
+#define SMB_INT_CAUSE_A 0x19090
+
+#define MSTTXFIFOPARINT_S    21
+#define MSTTXFIFOPARINT_V(x) ((x) << MSTTXFIFOPARINT_S)
+#define MSTTXFIFOPARINT_F    MSTTXFIFOPARINT_V(1U)
+
+#define MSTRXFIFOPARINT_S    20
+#define MSTRXFIFOPARINT_V(x) ((x) << MSTRXFIFOPARINT_S)
+#define MSTRXFIFOPARINT_F    MSTRXFIFOPARINT_V(1U)
+
+#define SLVFIFOPARINT_S    19
+#define SLVFIFOPARINT_V(x) ((x) << SLVFIFOPARINT_S)
+#define SLVFIFOPARINT_F    SLVFIFOPARINT_V(1U)
+
+#define ULP_RX_INT_CAUSE_A 0x19158
+#define ULP_RX_ISCSI_LLIMIT_A 0x1915c
+#define ULP_RX_ISCSI_ULIMIT_A 0x19160
+#define ULP_RX_ISCSI_TAGMASK_A 0x19164
+#define ULP_RX_ISCSI_PSZ_A 0x19168
+#define ULP_RX_TDDP_LLIMIT_A 0x1916c
+#define ULP_RX_TDDP_ULIMIT_A 0x19170
+#define ULP_RX_STAG_LLIMIT_A 0x1917c
+#define ULP_RX_STAG_ULIMIT_A 0x19180
+#define ULP_RX_RQ_LLIMIT_A 0x19184
+#define ULP_RX_RQ_ULIMIT_A 0x19188
+#define ULP_RX_PBL_LLIMIT_A 0x1918c
+#define ULP_RX_PBL_ULIMIT_A 0x19190
+#define ULP_RX_CTX_BASE_A 0x19194
+#define ULP_RX_RQUDP_LLIMIT_A 0x191a4
+#define ULP_RX_RQUDP_ULIMIT_A 0x191a8
+#define ULP_RX_LA_CTL_A 0x1923c
+#define ULP_RX_LA_RDPTR_A 0x19240
+#define ULP_RX_LA_RDDATA_A 0x19244
+#define ULP_RX_LA_WRPTR_A 0x19248
+#define ULP_RX_TLS_KEY_LLIMIT_A 0x192ac
+#define ULP_RX_TLS_KEY_ULIMIT_A 0x192b0
+
+#define HPZ3_S    24
+#define HPZ3_V(x) ((x) << HPZ3_S)
+
+#define HPZ2_S    16
+#define HPZ2_V(x) ((x) << HPZ2_S)
+
+#define HPZ1_S    8
+#define HPZ1_V(x) ((x) << HPZ1_S)
+
+#define HPZ0_S    0
+#define HPZ0_V(x) ((x) << HPZ0_S)
+
+#define ULP_RX_TDDP_PSZ_A 0x19178
+
+/* registers for module SF */
+#define SF_DATA_A 0x193f8
+#define SF_OP_A 0x193fc
+
+#define SF_BUSY_S    31
+#define SF_BUSY_V(x) ((x) << SF_BUSY_S)
+#define SF_BUSY_F    SF_BUSY_V(1U)
+
+#define SF_LOCK_S    4
+#define SF_LOCK_V(x) ((x) << SF_LOCK_S)
+#define SF_LOCK_F    SF_LOCK_V(1U)
+
+#define SF_CONT_S    3
+#define SF_CONT_V(x) ((x) << SF_CONT_S)
+#define SF_CONT_F    SF_CONT_V(1U)
+
+#define BYTECNT_S    1
+#define BYTECNT_V(x) ((x) << BYTECNT_S)
+
+#define OP_S    0
+#define OP_V(x) ((x) << OP_S)
+#define OP_F    OP_V(1U)
+
+#define PL_PF_INT_CAUSE_A 0x3c0
+
+#define PFSW_S    3
+#define PFSW_V(x) ((x) << PFSW_S)
+#define PFSW_F    PFSW_V(1U)
+
+#define PFCIM_S    1
+#define PFCIM_V(x) ((x) << PFCIM_S)
+#define PFCIM_F    PFCIM_V(1U)
+
+#define PL_PF_INT_ENABLE_A 0x3c4
+#define PL_PF_CTL_A 0x3c8
+
+#define PL_WHOAMI_A 0x19400
+
+#define SOURCEPF_S    8
+#define SOURCEPF_M    0x7U
+#define SOURCEPF_G(x) (((x) >> SOURCEPF_S) & SOURCEPF_M)
+
+#define T6_SOURCEPF_S    9
+#define T6_SOURCEPF_M    0x7U
+#define T6_SOURCEPF_G(x) (((x) >> T6_SOURCEPF_S) & T6_SOURCEPF_M)
+
+#define PL_INT_CAUSE_A 0x1940c
+
+#define ULP_TX_S    27
+#define ULP_TX_V(x) ((x) << ULP_TX_S)
+#define ULP_TX_F    ULP_TX_V(1U)
+
+#define SGE_S    26
+#define SGE_V(x) ((x) << SGE_S)
+#define SGE_F    SGE_V(1U)
+
+#define CPL_SWITCH_S    24
+#define CPL_SWITCH_V(x) ((x) << CPL_SWITCH_S)
+#define CPL_SWITCH_F    CPL_SWITCH_V(1U)
+
+#define ULP_RX_S    23
+#define ULP_RX_V(x) ((x) << ULP_RX_S)
+#define ULP_RX_F    ULP_RX_V(1U)
+
+#define PM_RX_S    22
+#define PM_RX_V(x) ((x) << PM_RX_S)
+#define PM_RX_F    PM_RX_V(1U)
+
+#define PM_TX_S    21
+#define PM_TX_V(x) ((x) << PM_TX_S)
+#define PM_TX_F    PM_TX_V(1U)
+
+#define MA_S    20
+#define MA_V(x) ((x) << MA_S)
+#define MA_F    MA_V(1U)
+
+#define TP_S    19
+#define TP_V(x) ((x) << TP_S)
+#define TP_F    TP_V(1U)
+
+#define LE_S    18
+#define LE_V(x) ((x) << LE_S)
+#define LE_F    LE_V(1U)
+
+#define EDC1_S    17
+#define EDC1_V(x) ((x) << EDC1_S)
+#define EDC1_F    EDC1_V(1U)
+
+#define EDC0_S    16
+#define EDC0_V(x) ((x) << EDC0_S)
+#define EDC0_F    EDC0_V(1U)
+
+#define MC_S    15
+#define MC_V(x) ((x) << MC_S)
+#define MC_F    MC_V(1U)
+
+#define PCIE_S    14
+#define PCIE_V(x) ((x) << PCIE_S)
+#define PCIE_F    PCIE_V(1U)
+
+#define XGMAC_KR1_S    12
+#define XGMAC_KR1_V(x) ((x) << XGMAC_KR1_S)
+#define XGMAC_KR1_F    XGMAC_KR1_V(1U)
+
+#define XGMAC_KR0_S    11
+#define XGMAC_KR0_V(x) ((x) << XGMAC_KR0_S)
+#define XGMAC_KR0_F    XGMAC_KR0_V(1U)
+
+#define XGMAC1_S    10
+#define XGMAC1_V(x) ((x) << XGMAC1_S)
+#define XGMAC1_F    XGMAC1_V(1U)
+
+#define XGMAC0_S    9
+#define XGMAC0_V(x) ((x) << XGMAC0_S)
+#define XGMAC0_F    XGMAC0_V(1U)
+
+#define SMB_S    8
+#define SMB_V(x) ((x) << SMB_S)
+#define SMB_F    SMB_V(1U)
+
+#define SF_S    7
+#define SF_V(x) ((x) << SF_S)
+#define SF_F    SF_V(1U)
+
+#define PL_S    6
+#define PL_V(x) ((x) << PL_S)
+#define PL_F    PL_V(1U)
+
+#define NCSI_S    5
+#define NCSI_V(x) ((x) << NCSI_S)
+#define NCSI_F    NCSI_V(1U)
+
+#define MPS_S    4
+#define MPS_V(x) ((x) << MPS_S)
+#define MPS_F    MPS_V(1U)
+
+#define CIM_S    0
+#define CIM_V(x) ((x) << CIM_S)
+#define CIM_F    CIM_V(1U)
+
+#define MC1_S    31
+#define MC1_V(x) ((x) << MC1_S)
+#define MC1_F    MC1_V(1U)
+
+#define PL_INT_ENABLE_A 0x19410
+#define PL_INT_MAP0_A 0x19414
+#define PL_RST_A 0x19428
+
+#define PIORST_S    1
+#define PIORST_V(x) ((x) << PIORST_S)
+#define PIORST_F    PIORST_V(1U)
+
+#define PIORSTMODE_S    0
+#define PIORSTMODE_V(x) ((x) << PIORSTMODE_S)
+#define PIORSTMODE_F    PIORSTMODE_V(1U)
+
+#define PL_PL_INT_CAUSE_A 0x19430
+
+#define FATALPERR_S    4
+#define FATALPERR_V(x) ((x) << FATALPERR_S)
+#define FATALPERR_F    FATALPERR_V(1U)
+
+#define PERRVFID_S    0
+#define PERRVFID_V(x) ((x) << PERRVFID_S)
+#define PERRVFID_F    PERRVFID_V(1U)
+
+#define PL_REV_A 0x1943c
+
+#define REV_S    0
+#define REV_M    0xfU
+#define REV_V(x) ((x) << REV_S)
+#define REV_G(x) (((x) >> REV_S) & REV_M)
+
+#define HASHTBLMEMCRCERR_S    27
+#define HASHTBLMEMCRCERR_V(x) ((x) << HASHTBLMEMCRCERR_S)
+#define HASHTBLMEMCRCERR_F    HASHTBLMEMCRCERR_V(1U)
+
+#define CMDTIDERR_S    22
+#define CMDTIDERR_V(x) ((x) << CMDTIDERR_S)
+#define CMDTIDERR_F    CMDTIDERR_V(1U)
+
+#define T6_UNKNOWNCMD_S    3
+#define T6_UNKNOWNCMD_V(x) ((x) << T6_UNKNOWNCMD_S)
+#define T6_UNKNOWNCMD_F    T6_UNKNOWNCMD_V(1U)
+
+#define T6_LIP0_S    2
+#define T6_LIP0_V(x) ((x) << T6_LIP0_S)
+#define T6_LIP0_F    T6_LIP0_V(1U)
+
+#define T6_LIPMISS_S    1
+#define T6_LIPMISS_V(x) ((x) << T6_LIPMISS_S)
+#define T6_LIPMISS_F    T6_LIPMISS_V(1U)
+
+#define LE_DB_CONFIG_A 0x19c04
+#define LE_DB_ROUTING_TABLE_INDEX_A 0x19c10
+#define LE_DB_ACTIVE_TABLE_START_INDEX_A 0x19c10
+#define LE_DB_FILTER_TABLE_INDEX_A 0x19c14
+#define LE_DB_SERVER_INDEX_A 0x19c18
+#define LE_DB_SRVR_START_INDEX_A 0x19c18
+#define LE_DB_CLIP_TABLE_INDEX_A 0x19c1c
+#define LE_DB_ACT_CNT_IPV4_A 0x19c20
+#define LE_DB_ACT_CNT_IPV6_A 0x19c24
+#define LE_DB_HASH_CONFIG_A 0x19c28
+
+#define HASHTIDSIZE_S    16
+#define HASHTIDSIZE_M    0x3fU
+#define HASHTIDSIZE_G(x) (((x) >> HASHTIDSIZE_S) & HASHTIDSIZE_M)
+
+#define HASHTBLSIZE_S    3
+#define HASHTBLSIZE_M    0x1ffffU
+#define HASHTBLSIZE_G(x) (((x) >> HASHTBLSIZE_S) & HASHTBLSIZE_M)
+
+#define LE_DB_HASH_TID_BASE_A 0x19c30
+#define LE_DB_HASH_TBL_BASE_ADDR_A 0x19c30
+#define LE_DB_INT_CAUSE_A 0x19c3c
+#define LE_DB_CLCAM_TID_BASE_A 0x19df4
+#define LE_DB_TID_HASHBASE_A 0x19df8
+#define T6_LE_DB_HASH_TID_BASE_A 0x19df8
+
+#define HASHEN_S    20
+#define HASHEN_V(x) ((x) << HASHEN_S)
+#define HASHEN_F    HASHEN_V(1U)
+
+#define ASLIPCOMPEN_S    17
+#define ASLIPCOMPEN_V(x) ((x) << ASLIPCOMPEN_S)
+#define ASLIPCOMPEN_F    ASLIPCOMPEN_V(1U)
+
+#define REQQPARERR_S    16
+#define REQQPARERR_V(x) ((x) << REQQPARERR_S)
+#define REQQPARERR_F    REQQPARERR_V(1U)
+
+#define UNKNOWNCMD_S    15
+#define UNKNOWNCMD_V(x) ((x) << UNKNOWNCMD_S)
+#define UNKNOWNCMD_F    UNKNOWNCMD_V(1U)
+
+#define PARITYERR_S    6
+#define PARITYERR_V(x) ((x) << PARITYERR_S)
+#define PARITYERR_F    PARITYERR_V(1U)
+
+#define LIPMISS_S    5
+#define LIPMISS_V(x) ((x) << LIPMISS_S)
+#define LIPMISS_F    LIPMISS_V(1U)
+
+#define LIP0_S    4
+#define LIP0_V(x) ((x) << LIP0_S)
+#define LIP0_F    LIP0_V(1U)
+
+#define BASEADDR_S    3
+#define BASEADDR_M    0x1fffffffU
+#define BASEADDR_G(x) (((x) >> BASEADDR_S) & BASEADDR_M)
+
+#define TCAMINTPERR_S    13
+#define TCAMINTPERR_V(x) ((x) << TCAMINTPERR_S)
+#define TCAMINTPERR_F    TCAMINTPERR_V(1U)
+
+#define SSRAMINTPERR_S    10
+#define SSRAMINTPERR_V(x) ((x) << SSRAMINTPERR_S)
+#define SSRAMINTPERR_F    SSRAMINTPERR_V(1U)
+
+#define LE_DB_RSP_CODE_0_A	0x19c74
+
+#define TCAM_ACTV_HIT_S		0
+#define TCAM_ACTV_HIT_M		0x1fU
+#define TCAM_ACTV_HIT_V(x)	((x) << TCAM_ACTV_HIT_S)
+#define TCAM_ACTV_HIT_G(x)	(((x) >> TCAM_ACTV_HIT_S) & TCAM_ACTV_HIT_M)
+
+#define LE_DB_RSP_CODE_1_A     0x19c78
+
+#define HASH_ACTV_HIT_S		25
+#define HASH_ACTV_HIT_M		0x1fU
+#define HASH_ACTV_HIT_V(x)	((x) << HASH_ACTV_HIT_S)
+#define HASH_ACTV_HIT_G(x)	(((x) >> HASH_ACTV_HIT_S) & HASH_ACTV_HIT_M)
+
+#define LE_3_DB_HASH_MASK_GEN_IPV4_T6_A	0x19eac
+#define LE_4_DB_HASH_MASK_GEN_IPV4_T6_A	0x19eb0
+
+#define NCSI_INT_CAUSE_A 0x1a0d8
+
+#define CIM_DM_PRTY_ERR_S    8
+#define CIM_DM_PRTY_ERR_V(x) ((x) << CIM_DM_PRTY_ERR_S)
+#define CIM_DM_PRTY_ERR_F    CIM_DM_PRTY_ERR_V(1U)
+
+#define MPS_DM_PRTY_ERR_S    7
+#define MPS_DM_PRTY_ERR_V(x) ((x) << MPS_DM_PRTY_ERR_S)
+#define MPS_DM_PRTY_ERR_F    MPS_DM_PRTY_ERR_V(1U)
+
+#define TXFIFO_PRTY_ERR_S    1
+#define TXFIFO_PRTY_ERR_V(x) ((x) << TXFIFO_PRTY_ERR_S)
+#define TXFIFO_PRTY_ERR_F    TXFIFO_PRTY_ERR_V(1U)
+
+#define RXFIFO_PRTY_ERR_S    0
+#define RXFIFO_PRTY_ERR_V(x) ((x) << RXFIFO_PRTY_ERR_S)
+#define RXFIFO_PRTY_ERR_F    RXFIFO_PRTY_ERR_V(1U)
+
+#define XGMAC_PORT_CFG2_A 0x1018
+
+#define PATEN_S    18
+#define PATEN_V(x) ((x) << PATEN_S)
+#define PATEN_F    PATEN_V(1U)
+
+#define MAGICEN_S    17
+#define MAGICEN_V(x) ((x) << MAGICEN_S)
+#define MAGICEN_F    MAGICEN_V(1U)
+
+#define XGMAC_PORT_MAGIC_MACID_LO 0x1024
+#define XGMAC_PORT_MAGIC_MACID_HI 0x1028
+
+#define XGMAC_PORT_EPIO_DATA0_A 0x10c0
+#define XGMAC_PORT_EPIO_DATA1_A 0x10c4
+#define XGMAC_PORT_EPIO_DATA2_A 0x10c8
+#define XGMAC_PORT_EPIO_DATA3_A 0x10cc
+#define XGMAC_PORT_EPIO_OP_A 0x10d0
+
+#define EPIOWR_S    8
+#define EPIOWR_V(x) ((x) << EPIOWR_S)
+#define EPIOWR_F    EPIOWR_V(1U)
+
+#define ADDRESS_S    0
+#define ADDRESS_V(x) ((x) << ADDRESS_S)
+
+#define MAC_PORT_INT_CAUSE_A 0x8dc
+#define XGMAC_PORT_INT_CAUSE_A 0x10dc
+
+#define TP_TX_MOD_QUEUE_REQ_MAP_A 0x7e28
+
+#define TP_TX_MOD_QUEUE_WEIGHT0_A 0x7e30
+#define TP_TX_MOD_CHANNEL_WEIGHT_A 0x7e34
+
+#define TX_MOD_QUEUE_REQ_MAP_S    0
+#define TX_MOD_QUEUE_REQ_MAP_V(x) ((x) << TX_MOD_QUEUE_REQ_MAP_S)
+
+#define TX_MODQ_WEIGHT3_S    24
+#define TX_MODQ_WEIGHT3_V(x) ((x) << TX_MODQ_WEIGHT3_S)
+
+#define TX_MODQ_WEIGHT2_S    16
+#define TX_MODQ_WEIGHT2_V(x) ((x) << TX_MODQ_WEIGHT2_S)
+
+#define TX_MODQ_WEIGHT1_S    8
+#define TX_MODQ_WEIGHT1_V(x) ((x) << TX_MODQ_WEIGHT1_S)
+
+#define TX_MODQ_WEIGHT0_S    0
+#define TX_MODQ_WEIGHT0_V(x) ((x) << TX_MODQ_WEIGHT0_S)
+
+#define TP_TX_SCHED_HDR_A 0x23
+#define TP_TX_SCHED_FIFO_A 0x24
+#define TP_TX_SCHED_PCMD_A 0x25
+
+#define NUM_MPS_CLS_SRAM_L_INSTANCES 336
+#define NUM_MPS_T5_CLS_SRAM_L_INSTANCES 512
+
+#define T5_PORT0_BASE 0x30000
+#define T5_PORT_STRIDE 0x4000
+#define T5_PORT_BASE(idx) (T5_PORT0_BASE + (idx) * T5_PORT_STRIDE)
+#define T5_PORT_REG(idx, reg) (T5_PORT_BASE(idx) + (reg))
+
+#define MC_0_BASE_ADDR 0x40000
+#define MC_1_BASE_ADDR 0x48000
+#define MC_STRIDE (MC_1_BASE_ADDR - MC_0_BASE_ADDR)
+#define MC_REG(reg, idx) (reg + MC_STRIDE * idx)
+
+#define MC_P_BIST_CMD_A			0x41400
+#define MC_P_BIST_CMD_ADDR_A		0x41404
+#define MC_P_BIST_CMD_LEN_A		0x41408
+#define MC_P_BIST_DATA_PATTERN_A	0x4140c
+#define MC_P_BIST_STATUS_RDATA_A	0x41488
+
+#define EDC_T50_BASE_ADDR		0x50000
+
+#define EDC_H_BIST_CMD_A		0x50004
+#define EDC_H_BIST_CMD_ADDR_A		0x50008
+#define EDC_H_BIST_CMD_LEN_A		0x5000c
+#define EDC_H_BIST_DATA_PATTERN_A	0x50010
+#define EDC_H_BIST_STATUS_RDATA_A	0x50028
+
+#define EDC_H_ECC_ERR_ADDR_A		0x50084
+#define EDC_T51_BASE_ADDR		0x50800
+
+#define EDC_T5_STRIDE (EDC_T51_BASE_ADDR - EDC_T50_BASE_ADDR)
+#define EDC_T5_REG(reg, idx) (reg + EDC_T5_STRIDE * idx)
+
+#define PL_VF_REV_A 0x4
+#define PL_VF_WHOAMI_A 0x0
+#define PL_VF_REVISION_A 0x8
+
+/* registers for module CIM */
+#define CIM_HOST_ACC_CTRL_A	0x7b50
+#define CIM_HOST_ACC_DATA_A	0x7b54
+#define UP_UP_DBG_LA_CFG_A	0x140
+#define UP_UP_DBG_LA_DATA_A	0x144
+
+#define HOSTBUSY_S	17
+#define HOSTBUSY_V(x)	((x) << HOSTBUSY_S)
+#define HOSTBUSY_F	HOSTBUSY_V(1U)
+
+#define HOSTWRITE_S	16
+#define HOSTWRITE_V(x)	((x) << HOSTWRITE_S)
+#define HOSTWRITE_F	HOSTWRITE_V(1U)
+
+#define CIM_IBQ_DBG_CFG_A 0x7b60
+
+#define IBQDBGADDR_S    16
+#define IBQDBGADDR_M    0xfffU
+#define IBQDBGADDR_V(x) ((x) << IBQDBGADDR_S)
+#define IBQDBGADDR_G(x) (((x) >> IBQDBGADDR_S) & IBQDBGADDR_M)
+
+#define IBQDBGBUSY_S    1
+#define IBQDBGBUSY_V(x) ((x) << IBQDBGBUSY_S)
+#define IBQDBGBUSY_F    IBQDBGBUSY_V(1U)
+
+#define IBQDBGEN_S    0
+#define IBQDBGEN_V(x) ((x) << IBQDBGEN_S)
+#define IBQDBGEN_F    IBQDBGEN_V(1U)
+
+#define CIM_OBQ_DBG_CFG_A 0x7b64
+
+#define OBQDBGADDR_S    16
+#define OBQDBGADDR_M    0xfffU
+#define OBQDBGADDR_V(x) ((x) << OBQDBGADDR_S)
+#define OBQDBGADDR_G(x) (((x) >> OBQDBGADDR_S) & OBQDBGADDR_M)
+
+#define OBQDBGBUSY_S    1
+#define OBQDBGBUSY_V(x) ((x) << OBQDBGBUSY_S)
+#define OBQDBGBUSY_F    OBQDBGBUSY_V(1U)
+
+#define OBQDBGEN_S    0
+#define OBQDBGEN_V(x) ((x) << OBQDBGEN_S)
+#define OBQDBGEN_F    OBQDBGEN_V(1U)
+
+#define CIM_IBQ_DBG_DATA_A 0x7b68
+#define CIM_OBQ_DBG_DATA_A 0x7b6c
+#define CIM_DEBUGCFG_A 0x7b70
+#define CIM_DEBUGSTS_A 0x7b74
+
+#define POLADBGRDPTR_S		23
+#define POLADBGRDPTR_M		0x1ffU
+#define POLADBGRDPTR_V(x)	((x) << POLADBGRDPTR_S)
+
+#define POLADBGWRPTR_S		16
+#define POLADBGWRPTR_M		0x1ffU
+#define POLADBGWRPTR_G(x)	(((x) >> POLADBGWRPTR_S) & POLADBGWRPTR_M)
+
+#define PILADBGRDPTR_S		14
+#define PILADBGRDPTR_M		0x1ffU
+#define PILADBGRDPTR_V(x)	((x) << PILADBGRDPTR_S)
+
+#define PILADBGWRPTR_S		0
+#define PILADBGWRPTR_M		0x1ffU
+#define PILADBGWRPTR_G(x)	(((x) >> PILADBGWRPTR_S) & PILADBGWRPTR_M)
+
+#define LADBGEN_S	12
+#define LADBGEN_V(x)	((x) << LADBGEN_S)
+#define LADBGEN_F	LADBGEN_V(1U)
+
+#define CIM_PO_LA_DEBUGDATA_A 0x7b78
+#define CIM_PI_LA_DEBUGDATA_A 0x7b7c
+#define CIM_PO_LA_MADEBUGDATA_A	0x7b80
+#define CIM_PI_LA_MADEBUGDATA_A	0x7b84
+
+#define UPDBGLARDEN_S		1
+#define UPDBGLARDEN_V(x)	((x) << UPDBGLARDEN_S)
+#define UPDBGLARDEN_F		UPDBGLARDEN_V(1U)
+
+#define UPDBGLAEN_S	0
+#define UPDBGLAEN_V(x)	((x) << UPDBGLAEN_S)
+#define UPDBGLAEN_F	UPDBGLAEN_V(1U)
+
+#define UPDBGLARDPTR_S		2
+#define UPDBGLARDPTR_M		0xfffU
+#define UPDBGLARDPTR_V(x)	((x) << UPDBGLARDPTR_S)
+
+#define UPDBGLAWRPTR_S    16
+#define UPDBGLAWRPTR_M    0xfffU
+#define UPDBGLAWRPTR_G(x) (((x) >> UPDBGLAWRPTR_S) & UPDBGLAWRPTR_M)
+
+#define UPDBGLACAPTPCONLY_S	30
+#define UPDBGLACAPTPCONLY_V(x)	((x) << UPDBGLACAPTPCONLY_S)
+#define UPDBGLACAPTPCONLY_F	UPDBGLACAPTPCONLY_V(1U)
+
+#define CIM_QUEUE_CONFIG_REF_A 0x7b48
+#define CIM_QUEUE_CONFIG_CTRL_A 0x7b4c
+
+#define CIMQSIZE_S    24
+#define CIMQSIZE_M    0x3fU
+#define CIMQSIZE_G(x) (((x) >> CIMQSIZE_S) & CIMQSIZE_M)
+
+#define CIMQBASE_S    16
+#define CIMQBASE_M    0x3fU
+#define CIMQBASE_G(x) (((x) >> CIMQBASE_S) & CIMQBASE_M)
+
+#define QUEFULLTHRSH_S    0
+#define QUEFULLTHRSH_M    0x1ffU
+#define QUEFULLTHRSH_G(x) (((x) >> QUEFULLTHRSH_S) & QUEFULLTHRSH_M)
+
+#define UP_IBQ_0_RDADDR_A 0x10
+#define UP_IBQ_0_SHADOW_RDADDR_A 0x280
+#define UP_OBQ_0_REALADDR_A 0x104
+#define UP_OBQ_0_SHADOW_REALADDR_A 0x394
+
+#define IBQRDADDR_S    0
+#define IBQRDADDR_M    0x1fffU
+#define IBQRDADDR_G(x) (((x) >> IBQRDADDR_S) & IBQRDADDR_M)
+
+#define IBQWRADDR_S    0
+#define IBQWRADDR_M    0x1fffU
+#define IBQWRADDR_G(x) (((x) >> IBQWRADDR_S) & IBQWRADDR_M)
+
+#define QUERDADDR_S    0
+#define QUERDADDR_M    0x7fffU
+#define QUERDADDR_G(x) (((x) >> QUERDADDR_S) & QUERDADDR_M)
+
+#define QUEREMFLITS_S    0
+#define QUEREMFLITS_M    0x7ffU
+#define QUEREMFLITS_G(x) (((x) >> QUEREMFLITS_S) & QUEREMFLITS_M)
+
+#define QUEEOPCNT_S    16
+#define QUEEOPCNT_M    0xfffU
+#define QUEEOPCNT_G(x) (((x) >> QUEEOPCNT_S) & QUEEOPCNT_M)
+
+#define QUESOPCNT_S    0
+#define QUESOPCNT_M    0xfffU
+#define QUESOPCNT_G(x) (((x) >> QUESOPCNT_S) & QUESOPCNT_M)
+
+#define OBQSELECT_S    4
+#define OBQSELECT_V(x) ((x) << OBQSELECT_S)
+#define OBQSELECT_F    OBQSELECT_V(1U)
+
+#define IBQSELECT_S    3
+#define IBQSELECT_V(x) ((x) << IBQSELECT_S)
+#define IBQSELECT_F    IBQSELECT_V(1U)
+
+#define QUENUMSELECT_S    0
+#define QUENUMSELECT_V(x) ((x) << QUENUMSELECT_S)
+
+#endif /* __T4_REGS_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_tcb.h b/drivers/net/ethernet/chelsio/cxgb4/t4_tcb.h
new file mode 100644
index 000000000..22a022012
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_tcb.h
@@ -0,0 +1,144 @@
+/*
+ * This file is part of the Chelsio T4/T5/T6 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2017 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __T4_TCB_H
+#define __T4_TCB_H
+
+#define TCB_L2T_IX_W		0
+#define TCB_L2T_IX_S		12
+#define TCB_L2T_IX_M		0xfffULL
+#define TCB_L2T_IX_V(x)		((x) << TCB_L2T_IX_S)
+
+#define TCB_T_FLAGS_W           1
+#define TCB_T_FLAGS_S           0
+#define TCB_T_FLAGS_M           0xffffffffffffffffULL
+#define TCB_T_FLAGS_V(x)        ((__u64)(x) << TCB_T_FLAGS_S)
+
+#define TCB_FIELD_COOKIE_TFLAG	1
+
+#define TCB_SMAC_SEL_W		0
+#define TCB_SMAC_SEL_S		24
+#define TCB_SMAC_SEL_M		0xffULL
+#define TCB_SMAC_SEL_V(x)	((x) << TCB_SMAC_SEL_S)
+
+#define TCB_T_FLAGS_W		1
+#define TCB_T_FLAGS_S		0
+#define TCB_T_FLAGS_M		0xffffffffffffffffULL
+#define TCB_T_FLAGS_V(x)	((__u64)(x) << TCB_T_FLAGS_S)
+
+#define TF_DROP_S		22
+#define TF_DIRECT_STEER_S	23
+#define TF_LPBK_S		59
+
+#define TF_CCTRL_ECE_S		60
+#define TF_CCTRL_CWR_S		61
+#define TF_CCTRL_RFR_S		62
+
+#define TCB_RSS_INFO_W		3
+#define TCB_RSS_INFO_S		0
+#define TCB_RSS_INFO_M		0x3ffULL
+#define TCB_RSS_INFO_V(x)	((x) << TCB_RSS_INFO_S)
+
+#define TCB_T_STATE_W		3
+#define TCB_T_STATE_S		16
+#define TCB_T_STATE_M		0xfULL
+#define TCB_T_STATE_V(x)	((x) << TCB_T_STATE_S)
+
+#define TCB_TIMESTAMP_W		5
+#define TCB_TIMESTAMP_S		0
+#define TCB_TIMESTAMP_M		0xffffffffULL
+#define TCB_TIMESTAMP_V(x)	((x) << TCB_TIMESTAMP_S)
+
+#define TCB_RTT_TS_RECENT_AGE_W		6
+#define TCB_RTT_TS_RECENT_AGE_S		0
+#define TCB_RTT_TS_RECENT_AGE_M		0xffffffffULL
+#define TCB_RTT_TS_RECENT_AGE_V(x)	((x) << TCB_RTT_TS_RECENT_AGE_S)
+
+#define TCB_T_RTSEQ_RECENT_W    7
+#define TCB_T_RTSEQ_RECENT_S    0
+#define TCB_T_RTSEQ_RECENT_M    0xffffffffULL
+#define TCB_T_RTSEQ_RECENT_V(x) ((x) << TCB_T_RTSEQ_RECENT_S)
+
+#define TCB_TX_MAX_W		9
+#define TCB_TX_MAX_S		0
+#define TCB_TX_MAX_M		0xffffffffULL
+#define TCB_TX_MAX_V(x)		((x) << TCB_TX_MAX_S)
+
+#define TCB_SND_UNA_RAW_W	10
+#define TCB_SND_UNA_RAW_S	0
+#define TCB_SND_UNA_RAW_M	0xfffffffULL
+#define TCB_SND_UNA_RAW_V(x)	((x) << TCB_SND_UNA_RAW_S)
+
+#define TCB_SND_NXT_RAW_W	10
+#define TCB_SND_NXT_RAW_S	28
+#define TCB_SND_NXT_RAW_M	0xfffffffULL
+#define TCB_SND_NXT_RAW_V(x)	((x) << TCB_SND_NXT_RAW_S)
+
+#define TCB_SND_MAX_RAW_W	11
+#define TCB_SND_MAX_RAW_S	24
+#define TCB_SND_MAX_RAW_M	0xfffffffULL
+#define TCB_SND_MAX_RAW_V(x)	((x) << TCB_SND_MAX_RAW_S)
+
+#define TCB_RCV_NXT_W		16
+#define TCB_RCV_NXT_S		10
+#define TCB_RCV_NXT_M		0xffffffffULL
+#define TCB_RCV_NXT_V(x)	((x) << TCB_RCV_NXT_S)
+
+#define TCB_RCV_WND_W		17
+#define TCB_RCV_WND_S		10
+#define TCB_RCV_WND_M		0xffffffULL
+#define TCB_RCV_WND_V(x)	((x) << TCB_RCV_WND_S)
+
+#define TCB_RX_FRAG2_PTR_RAW_W	27
+#define TCB_RX_FRAG3_LEN_RAW_W	29
+#define TCB_RX_FRAG3_START_IDX_OFFSET_RAW_W	30
+#define TCB_PDU_HDR_LEN_W	31
+
+#define TCB_RQ_START_W		30
+#define TCB_RQ_START_S		0
+#define TCB_RQ_START_M		0x3ffffffULL
+#define TCB_RQ_START_V(x)	((x) << TCB_RQ_START_S)
+
+#define TF_RX_PDU_OUT_S		49
+#define TF_RX_PDU_OUT_V(x)	((__u64)(x) << TF_RX_PDU_OUT_S)
+
+#define TF_CORE_BYPASS_S	63
+#define TF_CORE_BYPASS_V(x)	((__u64)(x) << TF_CORE_BYPASS_S)
+#define TF_CORE_BYPASS_F	TF_CORE_BYPASS_V(1)
+
+#define TF_NON_OFFLOAD_S	1
+#define TF_NON_OFFLOAD_V(x)	((x) << TF_NON_OFFLOAD_S)
+#define TF_NON_OFFLOAD_F	TF_NON_OFFLOAD_V(1)
+
+#endif /* __T4_TCB_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_values.h b/drivers/net/ethernet/chelsio/cxgb4/t4_values.h
new file mode 100644
index 000000000..eb1aa8214
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_values.h
@@ -0,0 +1,162 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __T4_VALUES_H__
+#define __T4_VALUES_H__
+
+/* This file contains definitions for various T4 register value hardware
+ * constants.  The types of values encoded here are predominantly those for
+ * register fields which control "modal" behavior.  For the most part, we do
+ * not include definitions for register fields which are simple numeric
+ * metrics, etc.
+ */
+
+/* SGE register field values.
+ */
+
+/* CONTROL1 register */
+#define RXPKTCPLMODE_SPLIT_X		1
+
+#define INGPCIEBOUNDARY_SHIFT_X		5
+#define INGPCIEBOUNDARY_32B_X		0
+
+#define INGPADBOUNDARY_SHIFT_X		5
+
+#define T6_INGPADBOUNDARY_SHIFT_X	3
+#define T6_INGPADBOUNDARY_8B_X		0
+#define T6_INGPADBOUNDARY_32B_X		2
+
+#define INGPADBOUNDARY_32B_X		0
+
+/* CONTROL2 register */
+#define INGPACKBOUNDARY_SHIFT_X		5
+#define INGPACKBOUNDARY_16B_X		0
+#define INGPACKBOUNDARY_64B_X		1
+
+/* GTS register */
+#define SGE_TIMERREGS			6
+#define TIMERREG_COUNTER0_X		0
+
+#define FETCHBURSTMIN_64B_X		2
+#define FETCHBURSTMIN_128B_X		3
+
+/* T6 and later use a single-bit encoding for FetchBurstMin */
+#define FETCHBURSTMIN_64B_T6_X		0
+#define FETCHBURSTMIN_128B_T6_X		1
+
+#define FETCHBURSTMAX_256B_X		2
+#define FETCHBURSTMAX_512B_X		3
+
+#define HOSTFCMODE_INGRESS_QUEUE_X	1
+#define HOSTFCMODE_STATUS_PAGE_X	2
+
+#define CIDXFLUSHTHRESH_32_X		5
+#define CIDXFLUSHTHRESH_128_X		7
+
+#define UPDATEDELIVERY_INTERRUPT_X	1
+
+#define RSPD_TYPE_FLBUF_X		0
+#define RSPD_TYPE_CPL_X			1
+#define RSPD_TYPE_INTR_X		2
+
+/* Congestion Manager Definitions.
+ */
+#define CONMCTXT_CNGTPMODE_S		19
+#define CONMCTXT_CNGTPMODE_V(x)		((x) << CONMCTXT_CNGTPMODE_S)
+#define CONMCTXT_CNGCHMAP_S		0
+#define CONMCTXT_CNGCHMAP_V(x)		((x) << CONMCTXT_CNGCHMAP_S)
+#define CONMCTXT_CNGTPMODE_CHANNEL_X	2
+#define CONMCTXT_CNGTPMODE_QUEUE_X	1
+
+/* T5 and later support a new BAR2-based doorbell mechanism for Egress Queues.
+ * The User Doorbells are each 128 bytes in length with a Simple Doorbell at
+ * offsets 8x and a Write Combining single 64-byte Egress Queue Unit
+ * (IDXSIZE_UNIT_X) Gather Buffer interface at offset 64.  For Ingress Queues,
+ * we have a Going To Sleep register at offsets 8x+4.
+ *
+ * As noted above, we have many instances of the Simple Doorbell and Going To
+ * Sleep registers at offsets 8x and 8x+4, respectively.  We want to use a
+ * non-64-byte aligned offset for the Simple Doorbell in order to attempt to
+ * avoid buffering of the writes to the Simple Doorbell and we want to use a
+ * non-contiguous offset for the Going To Sleep writes in order to avoid
+ * possible combining between them.
+ */
+#define SGE_UDB_SIZE		128
+#define SGE_UDB_KDOORBELL	8
+#define SGE_UDB_GTS		20
+#define SGE_UDB_WCDOORBELL	64
+
+/* CIM register field values.
+ */
+#define X_MBOWNER_FW			1
+#define X_MBOWNER_PL			2
+
+/* PCI-E definitions */
+#define WINDOW_SHIFT_X		10
+#define PCIEOFST_SHIFT_X	10
+
+/* TP_VLAN_PRI_MAP controls which subset of fields will be present in the
+ * Compressed Filter Tuple for LE filters.  Each bit set in TP_VLAN_PRI_MAP
+ * selects for a particular field being present.  These fields, when present
+ * in the Compressed Filter Tuple, have the following widths in bits.
+ */
+#define FT_FCOE_W                       1
+#define FT_PORT_W                       3
+#define FT_VNIC_ID_W                    17
+#define FT_VLAN_W                       17
+#define FT_TOS_W                        8
+#define FT_PROTOCOL_W                   8
+#define FT_ETHERTYPE_W                  16
+#define FT_MACMATCH_W                   9
+#define FT_MPSHITTYPE_W                 3
+#define FT_FRAGMENTATION_W              1
+
+/* Some of the Compressed Filter Tuple fields have internal structure.  These
+ * bit shifts/masks describe those structures.  All shifts are relative to the
+ * base position of the fields within the Compressed Filter Tuple
+ */
+#define FT_VLAN_VLD_S                   16
+#define FT_VLAN_VLD_V(x)                ((x) << FT_VLAN_VLD_S)
+#define FT_VLAN_VLD_F                   FT_VLAN_VLD_V(1U)
+
+#define FT_VNID_ID_VF_S                 0
+#define FT_VNID_ID_VF_V(x)              ((x) << FT_VNID_ID_VF_S)
+
+#define FT_VNID_ID_PF_S                 7
+#define FT_VNID_ID_PF_V(x)              ((x) << FT_VNID_ID_PF_S)
+
+#define FT_VNID_ID_VLD_S                16
+#define FT_VNID_ID_VLD_V(x)             ((x) << FT_VNID_ID_VLD_S)
+
+#endif /* __T4_VALUES_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h b/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h
new file mode 100644
index 000000000..2419459a0
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h
@@ -0,0 +1,4187 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2009-2016 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef _T4FW_INTERFACE_H_
+#define _T4FW_INTERFACE_H_
+
+enum fw_retval {
+	FW_SUCCESS		= 0,	/* completed successfully */
+	FW_EPERM		= 1,	/* operation not permitted */
+	FW_ENOENT		= 2,	/* no such file or directory */
+	FW_EIO			= 5,	/* input/output error; hw bad */
+	FW_ENOEXEC		= 8,	/* exec format error; inv microcode */
+	FW_EAGAIN		= 11,	/* try again */
+	FW_ENOMEM		= 12,	/* out of memory */
+	FW_EFAULT		= 14,	/* bad address; fw bad */
+	FW_EBUSY		= 16,	/* resource busy */
+	FW_EEXIST		= 17,	/* file exists */
+	FW_ENODEV		= 19,	/* no such device */
+	FW_EINVAL		= 22,	/* invalid argument */
+	FW_ENOSPC		= 28,	/* no space left on device */
+	FW_ENOSYS		= 38,	/* functionality not implemented */
+	FW_ENODATA		= 61,	/* no data available */
+	FW_EPROTO		= 71,	/* protocol error */
+	FW_EADDRINUSE		= 98,	/* address already in use */
+	FW_EADDRNOTAVAIL	= 99,	/* cannot assigned requested address */
+	FW_ENETDOWN		= 100,	/* network is down */
+	FW_ENETUNREACH		= 101,	/* network is unreachable */
+	FW_ENOBUFS		= 105,	/* no buffer space available */
+	FW_ETIMEDOUT		= 110,	/* timeout */
+	FW_EINPROGRESS		= 115,	/* fw internal */
+	FW_SCSI_ABORT_REQUESTED	= 128,	/* */
+	FW_SCSI_ABORT_TIMEDOUT	= 129,	/* */
+	FW_SCSI_ABORTED		= 130,	/* */
+	FW_SCSI_CLOSE_REQUESTED	= 131,	/* */
+	FW_ERR_LINK_DOWN	= 132,	/* */
+	FW_RDEV_NOT_READY	= 133,	/* */
+	FW_ERR_RDEV_LOST	= 134,	/* */
+	FW_ERR_RDEV_LOGO	= 135,	/* */
+	FW_FCOE_NO_XCHG		= 136,	/* */
+	FW_SCSI_RSP_ERR		= 137,	/* */
+	FW_ERR_RDEV_IMPL_LOGO	= 138,	/* */
+	FW_SCSI_UNDER_FLOW_ERR  = 139,	/* */
+	FW_SCSI_OVER_FLOW_ERR   = 140,	/* */
+	FW_SCSI_DDP_ERR		= 141,	/* DDP error*/
+	FW_SCSI_TASK_ERR	= 142,	/* No SCSI tasks available */
+};
+
+#define FW_T4VF_SGE_BASE_ADDR      0x0000
+#define FW_T4VF_MPS_BASE_ADDR      0x0100
+#define FW_T4VF_PL_BASE_ADDR       0x0200
+#define FW_T4VF_MBDATA_BASE_ADDR   0x0240
+#define FW_T4VF_CIM_BASE_ADDR      0x0300
+
+enum fw_wr_opcodes {
+	FW_FILTER_WR                   = 0x02,
+	FW_ULPTX_WR                    = 0x04,
+	FW_TP_WR                       = 0x05,
+	FW_ETH_TX_PKT_WR               = 0x08,
+	FW_ETH_TX_EO_WR                = 0x1c,
+	FW_OFLD_CONNECTION_WR          = 0x2f,
+	FW_FLOWC_WR                    = 0x0a,
+	FW_OFLD_TX_DATA_WR             = 0x0b,
+	FW_CMD_WR                      = 0x10,
+	FW_ETH_TX_PKT_VM_WR            = 0x11,
+	FW_RI_RES_WR                   = 0x0c,
+	FW_RI_INIT_WR                  = 0x0d,
+	FW_RI_RDMA_WRITE_WR            = 0x14,
+	FW_RI_SEND_WR                  = 0x15,
+	FW_RI_RDMA_READ_WR             = 0x16,
+	FW_RI_RECV_WR                  = 0x17,
+	FW_RI_BIND_MW_WR               = 0x18,
+	FW_RI_FR_NSMR_WR               = 0x19,
+	FW_RI_FR_NSMR_TPTE_WR	       = 0x20,
+	FW_RI_RDMA_WRITE_CMPL_WR       = 0x21,
+	FW_RI_INV_LSTAG_WR             = 0x1a,
+	FW_ISCSI_TX_DATA_WR	       = 0x45,
+	FW_PTP_TX_PKT_WR               = 0x46,
+	FW_TLSTX_DATA_WR	       = 0x68,
+	FW_CRYPTO_LOOKASIDE_WR         = 0X6d,
+	FW_LASTC2E_WR                  = 0x70,
+	FW_FILTER2_WR		       = 0x77
+};
+
+struct fw_wr_hdr {
+	__be32 hi;
+	__be32 lo;
+};
+
+/* work request opcode (hi) */
+#define FW_WR_OP_S	24
+#define FW_WR_OP_M      0xff
+#define FW_WR_OP_V(x)   ((x) << FW_WR_OP_S)
+#define FW_WR_OP_G(x)   (((x) >> FW_WR_OP_S) & FW_WR_OP_M)
+
+/* atomic flag (hi) - firmware encapsulates CPLs in CPL_BARRIER */
+#define FW_WR_ATOMIC_S		23
+#define FW_WR_ATOMIC_V(x)	((x) << FW_WR_ATOMIC_S)
+
+/* flush flag (hi) - firmware flushes flushable work request buffered
+ * in the flow context.
+ */
+#define FW_WR_FLUSH_S     22
+#define FW_WR_FLUSH_V(x)  ((x) << FW_WR_FLUSH_S)
+
+/* completion flag (hi) - firmware generates a cpl_fw6_ack */
+#define FW_WR_COMPL_S     21
+#define FW_WR_COMPL_V(x)  ((x) << FW_WR_COMPL_S)
+#define FW_WR_COMPL_F     FW_WR_COMPL_V(1U)
+
+/* work request immediate data length (hi) */
+#define FW_WR_IMMDLEN_S 0
+#define FW_WR_IMMDLEN_M 0xff
+#define FW_WR_IMMDLEN_V(x)      ((x) << FW_WR_IMMDLEN_S)
+
+/* egress queue status update to associated ingress queue entry (lo) */
+#define FW_WR_EQUIQ_S           31
+#define FW_WR_EQUIQ_V(x)        ((x) << FW_WR_EQUIQ_S)
+#define FW_WR_EQUIQ_F           FW_WR_EQUIQ_V(1U)
+
+/* egress queue status update to egress queue status entry (lo) */
+#define FW_WR_EQUEQ_S           30
+#define FW_WR_EQUEQ_V(x)        ((x) << FW_WR_EQUEQ_S)
+#define FW_WR_EQUEQ_F           FW_WR_EQUEQ_V(1U)
+
+/* flow context identifier (lo) */
+#define FW_WR_FLOWID_S          8
+#define FW_WR_FLOWID_V(x)       ((x) << FW_WR_FLOWID_S)
+
+/* length in units of 16-bytes (lo) */
+#define FW_WR_LEN16_S           0
+#define FW_WR_LEN16_V(x)        ((x) << FW_WR_LEN16_S)
+
+#define HW_TPL_FR_MT_PR_IV_P_FC         0X32B
+#define HW_TPL_FR_MT_PR_OV_P_FC         0X327
+
+/* filter wr reply code in cookie in CPL_SET_TCB_RPL */
+enum fw_filter_wr_cookie {
+	FW_FILTER_WR_SUCCESS,
+	FW_FILTER_WR_FLT_ADDED,
+	FW_FILTER_WR_FLT_DELETED,
+	FW_FILTER_WR_SMT_TBL_FULL,
+	FW_FILTER_WR_EINVAL,
+};
+
+struct fw_filter_wr {
+	__be32 op_pkd;
+	__be32 len16_pkd;
+	__be64 r3;
+	__be32 tid_to_iq;
+	__be32 del_filter_to_l2tix;
+	__be16 ethtype;
+	__be16 ethtypem;
+	__u8   frag_to_ovlan_vldm;
+	__u8   smac_sel;
+	__be16 rx_chan_rx_rpl_iq;
+	__be32 maci_to_matchtypem;
+	__u8   ptcl;
+	__u8   ptclm;
+	__u8   ttyp;
+	__u8   ttypm;
+	__be16 ivlan;
+	__be16 ivlanm;
+	__be16 ovlan;
+	__be16 ovlanm;
+	__u8   lip[16];
+	__u8   lipm[16];
+	__u8   fip[16];
+	__u8   fipm[16];
+	__be16 lp;
+	__be16 lpm;
+	__be16 fp;
+	__be16 fpm;
+	__be16 r7;
+	__u8   sma[6];
+};
+
+struct fw_filter2_wr {
+	__be32 op_pkd;
+	__be32 len16_pkd;
+	__be64 r3;
+	__be32 tid_to_iq;
+	__be32 del_filter_to_l2tix;
+	__be16 ethtype;
+	__be16 ethtypem;
+	__u8   frag_to_ovlan_vldm;
+	__u8   smac_sel;
+	__be16 rx_chan_rx_rpl_iq;
+	__be32 maci_to_matchtypem;
+	__u8   ptcl;
+	__u8   ptclm;
+	__u8   ttyp;
+	__u8   ttypm;
+	__be16 ivlan;
+	__be16 ivlanm;
+	__be16 ovlan;
+	__be16 ovlanm;
+	__u8   lip[16];
+	__u8   lipm[16];
+	__u8   fip[16];
+	__u8   fipm[16];
+	__be16 lp;
+	__be16 lpm;
+	__be16 fp;
+	__be16 fpm;
+	__be16 r7;
+	__u8   sma[6];
+	__be16 r8;
+	__u8   filter_type_swapmac;
+	__u8   natmode_to_ulp_type;
+	__be16 newlport;
+	__be16 newfport;
+	__u8   newlip[16];
+	__u8   newfip[16];
+	__be32 natseqcheck;
+	__be32 r9;
+	__be64 r10;
+	__be64 r11;
+	__be64 r12;
+	__be64 r13;
+};
+
+#define FW_FILTER_WR_TID_S      12
+#define FW_FILTER_WR_TID_M      0xfffff
+#define FW_FILTER_WR_TID_V(x)   ((x) << FW_FILTER_WR_TID_S)
+#define FW_FILTER_WR_TID_G(x)   \
+	(((x) >> FW_FILTER_WR_TID_S) & FW_FILTER_WR_TID_M)
+
+#define FW_FILTER_WR_RQTYPE_S           11
+#define FW_FILTER_WR_RQTYPE_M           0x1
+#define FW_FILTER_WR_RQTYPE_V(x)        ((x) << FW_FILTER_WR_RQTYPE_S)
+#define FW_FILTER_WR_RQTYPE_G(x)        \
+	(((x) >> FW_FILTER_WR_RQTYPE_S) & FW_FILTER_WR_RQTYPE_M)
+#define FW_FILTER_WR_RQTYPE_F   FW_FILTER_WR_RQTYPE_V(1U)
+
+#define FW_FILTER_WR_NOREPLY_S          10
+#define FW_FILTER_WR_NOREPLY_M          0x1
+#define FW_FILTER_WR_NOREPLY_V(x)       ((x) << FW_FILTER_WR_NOREPLY_S)
+#define FW_FILTER_WR_NOREPLY_G(x)       \
+	(((x) >> FW_FILTER_WR_NOREPLY_S) & FW_FILTER_WR_NOREPLY_M)
+#define FW_FILTER_WR_NOREPLY_F  FW_FILTER_WR_NOREPLY_V(1U)
+
+#define FW_FILTER_WR_IQ_S       0
+#define FW_FILTER_WR_IQ_M       0x3ff
+#define FW_FILTER_WR_IQ_V(x)    ((x) << FW_FILTER_WR_IQ_S)
+#define FW_FILTER_WR_IQ_G(x)    \
+	(((x) >> FW_FILTER_WR_IQ_S) & FW_FILTER_WR_IQ_M)
+
+#define FW_FILTER_WR_DEL_FILTER_S       31
+#define FW_FILTER_WR_DEL_FILTER_M       0x1
+#define FW_FILTER_WR_DEL_FILTER_V(x)    ((x) << FW_FILTER_WR_DEL_FILTER_S)
+#define FW_FILTER_WR_DEL_FILTER_G(x)    \
+	(((x) >> FW_FILTER_WR_DEL_FILTER_S) & FW_FILTER_WR_DEL_FILTER_M)
+#define FW_FILTER_WR_DEL_FILTER_F       FW_FILTER_WR_DEL_FILTER_V(1U)
+
+#define FW_FILTER_WR_RPTTID_S           25
+#define FW_FILTER_WR_RPTTID_M           0x1
+#define FW_FILTER_WR_RPTTID_V(x)        ((x) << FW_FILTER_WR_RPTTID_S)
+#define FW_FILTER_WR_RPTTID_G(x)        \
+	(((x) >> FW_FILTER_WR_RPTTID_S) & FW_FILTER_WR_RPTTID_M)
+#define FW_FILTER_WR_RPTTID_F   FW_FILTER_WR_RPTTID_V(1U)
+
+#define FW_FILTER_WR_DROP_S     24
+#define FW_FILTER_WR_DROP_M     0x1
+#define FW_FILTER_WR_DROP_V(x)  ((x) << FW_FILTER_WR_DROP_S)
+#define FW_FILTER_WR_DROP_G(x)  \
+	(((x) >> FW_FILTER_WR_DROP_S) & FW_FILTER_WR_DROP_M)
+#define FW_FILTER_WR_DROP_F     FW_FILTER_WR_DROP_V(1U)
+
+#define FW_FILTER_WR_DIRSTEER_S         23
+#define FW_FILTER_WR_DIRSTEER_M         0x1
+#define FW_FILTER_WR_DIRSTEER_V(x)      ((x) << FW_FILTER_WR_DIRSTEER_S)
+#define FW_FILTER_WR_DIRSTEER_G(x)      \
+	(((x) >> FW_FILTER_WR_DIRSTEER_S) & FW_FILTER_WR_DIRSTEER_M)
+#define FW_FILTER_WR_DIRSTEER_F FW_FILTER_WR_DIRSTEER_V(1U)
+
+#define FW_FILTER_WR_MASKHASH_S         22
+#define FW_FILTER_WR_MASKHASH_M         0x1
+#define FW_FILTER_WR_MASKHASH_V(x)      ((x) << FW_FILTER_WR_MASKHASH_S)
+#define FW_FILTER_WR_MASKHASH_G(x)      \
+	(((x) >> FW_FILTER_WR_MASKHASH_S) & FW_FILTER_WR_MASKHASH_M)
+#define FW_FILTER_WR_MASKHASH_F FW_FILTER_WR_MASKHASH_V(1U)
+
+#define FW_FILTER_WR_DIRSTEERHASH_S     21
+#define FW_FILTER_WR_DIRSTEERHASH_M     0x1
+#define FW_FILTER_WR_DIRSTEERHASH_V(x)  ((x) << FW_FILTER_WR_DIRSTEERHASH_S)
+#define FW_FILTER_WR_DIRSTEERHASH_G(x)  \
+	(((x) >> FW_FILTER_WR_DIRSTEERHASH_S) & FW_FILTER_WR_DIRSTEERHASH_M)
+#define FW_FILTER_WR_DIRSTEERHASH_F     FW_FILTER_WR_DIRSTEERHASH_V(1U)
+
+#define FW_FILTER_WR_LPBK_S     20
+#define FW_FILTER_WR_LPBK_M     0x1
+#define FW_FILTER_WR_LPBK_V(x)  ((x) << FW_FILTER_WR_LPBK_S)
+#define FW_FILTER_WR_LPBK_G(x)  \
+	(((x) >> FW_FILTER_WR_LPBK_S) & FW_FILTER_WR_LPBK_M)
+#define FW_FILTER_WR_LPBK_F     FW_FILTER_WR_LPBK_V(1U)
+
+#define FW_FILTER_WR_DMAC_S     19
+#define FW_FILTER_WR_DMAC_M     0x1
+#define FW_FILTER_WR_DMAC_V(x)  ((x) << FW_FILTER_WR_DMAC_S)
+#define FW_FILTER_WR_DMAC_G(x)  \
+	(((x) >> FW_FILTER_WR_DMAC_S) & FW_FILTER_WR_DMAC_M)
+#define FW_FILTER_WR_DMAC_F     FW_FILTER_WR_DMAC_V(1U)
+
+#define FW_FILTER_WR_SMAC_S     18
+#define FW_FILTER_WR_SMAC_M     0x1
+#define FW_FILTER_WR_SMAC_V(x)  ((x) << FW_FILTER_WR_SMAC_S)
+#define FW_FILTER_WR_SMAC_G(x)  \
+	(((x) >> FW_FILTER_WR_SMAC_S) & FW_FILTER_WR_SMAC_M)
+#define FW_FILTER_WR_SMAC_F     FW_FILTER_WR_SMAC_V(1U)
+
+#define FW_FILTER_WR_INSVLAN_S          17
+#define FW_FILTER_WR_INSVLAN_M          0x1
+#define FW_FILTER_WR_INSVLAN_V(x)       ((x) << FW_FILTER_WR_INSVLAN_S)
+#define FW_FILTER_WR_INSVLAN_G(x)       \
+	(((x) >> FW_FILTER_WR_INSVLAN_S) & FW_FILTER_WR_INSVLAN_M)
+#define FW_FILTER_WR_INSVLAN_F  FW_FILTER_WR_INSVLAN_V(1U)
+
+#define FW_FILTER_WR_RMVLAN_S           16
+#define FW_FILTER_WR_RMVLAN_M           0x1
+#define FW_FILTER_WR_RMVLAN_V(x)        ((x) << FW_FILTER_WR_RMVLAN_S)
+#define FW_FILTER_WR_RMVLAN_G(x)        \
+	(((x) >> FW_FILTER_WR_RMVLAN_S) & FW_FILTER_WR_RMVLAN_M)
+#define FW_FILTER_WR_RMVLAN_F   FW_FILTER_WR_RMVLAN_V(1U)
+
+#define FW_FILTER_WR_HITCNTS_S          15
+#define FW_FILTER_WR_HITCNTS_M          0x1
+#define FW_FILTER_WR_HITCNTS_V(x)       ((x) << FW_FILTER_WR_HITCNTS_S)
+#define FW_FILTER_WR_HITCNTS_G(x)       \
+	(((x) >> FW_FILTER_WR_HITCNTS_S) & FW_FILTER_WR_HITCNTS_M)
+#define FW_FILTER_WR_HITCNTS_F  FW_FILTER_WR_HITCNTS_V(1U)
+
+#define FW_FILTER_WR_TXCHAN_S           13
+#define FW_FILTER_WR_TXCHAN_M           0x3
+#define FW_FILTER_WR_TXCHAN_V(x)        ((x) << FW_FILTER_WR_TXCHAN_S)
+#define FW_FILTER_WR_TXCHAN_G(x)        \
+	(((x) >> FW_FILTER_WR_TXCHAN_S) & FW_FILTER_WR_TXCHAN_M)
+
+#define FW_FILTER_WR_PRIO_S     12
+#define FW_FILTER_WR_PRIO_M     0x1
+#define FW_FILTER_WR_PRIO_V(x)  ((x) << FW_FILTER_WR_PRIO_S)
+#define FW_FILTER_WR_PRIO_G(x)  \
+	(((x) >> FW_FILTER_WR_PRIO_S) & FW_FILTER_WR_PRIO_M)
+#define FW_FILTER_WR_PRIO_F     FW_FILTER_WR_PRIO_V(1U)
+
+#define FW_FILTER_WR_L2TIX_S    0
+#define FW_FILTER_WR_L2TIX_M    0xfff
+#define FW_FILTER_WR_L2TIX_V(x) ((x) << FW_FILTER_WR_L2TIX_S)
+#define FW_FILTER_WR_L2TIX_G(x) \
+	(((x) >> FW_FILTER_WR_L2TIX_S) & FW_FILTER_WR_L2TIX_M)
+
+#define FW_FILTER_WR_FRAG_S     7
+#define FW_FILTER_WR_FRAG_M     0x1
+#define FW_FILTER_WR_FRAG_V(x)  ((x) << FW_FILTER_WR_FRAG_S)
+#define FW_FILTER_WR_FRAG_G(x)  \
+	(((x) >> FW_FILTER_WR_FRAG_S) & FW_FILTER_WR_FRAG_M)
+#define FW_FILTER_WR_FRAG_F     FW_FILTER_WR_FRAG_V(1U)
+
+#define FW_FILTER_WR_FRAGM_S    6
+#define FW_FILTER_WR_FRAGM_M    0x1
+#define FW_FILTER_WR_FRAGM_V(x) ((x) << FW_FILTER_WR_FRAGM_S)
+#define FW_FILTER_WR_FRAGM_G(x) \
+	(((x) >> FW_FILTER_WR_FRAGM_S) & FW_FILTER_WR_FRAGM_M)
+#define FW_FILTER_WR_FRAGM_F    FW_FILTER_WR_FRAGM_V(1U)
+
+#define FW_FILTER_WR_IVLAN_VLD_S        5
+#define FW_FILTER_WR_IVLAN_VLD_M        0x1
+#define FW_FILTER_WR_IVLAN_VLD_V(x)     ((x) << FW_FILTER_WR_IVLAN_VLD_S)
+#define FW_FILTER_WR_IVLAN_VLD_G(x)     \
+	(((x) >> FW_FILTER_WR_IVLAN_VLD_S) & FW_FILTER_WR_IVLAN_VLD_M)
+#define FW_FILTER_WR_IVLAN_VLD_F        FW_FILTER_WR_IVLAN_VLD_V(1U)
+
+#define FW_FILTER_WR_OVLAN_VLD_S        4
+#define FW_FILTER_WR_OVLAN_VLD_M        0x1
+#define FW_FILTER_WR_OVLAN_VLD_V(x)     ((x) << FW_FILTER_WR_OVLAN_VLD_S)
+#define FW_FILTER_WR_OVLAN_VLD_G(x)     \
+	(((x) >> FW_FILTER_WR_OVLAN_VLD_S) & FW_FILTER_WR_OVLAN_VLD_M)
+#define FW_FILTER_WR_OVLAN_VLD_F        FW_FILTER_WR_OVLAN_VLD_V(1U)
+
+#define FW_FILTER_WR_IVLAN_VLDM_S       3
+#define FW_FILTER_WR_IVLAN_VLDM_M       0x1
+#define FW_FILTER_WR_IVLAN_VLDM_V(x)    ((x) << FW_FILTER_WR_IVLAN_VLDM_S)
+#define FW_FILTER_WR_IVLAN_VLDM_G(x)    \
+	(((x) >> FW_FILTER_WR_IVLAN_VLDM_S) & FW_FILTER_WR_IVLAN_VLDM_M)
+#define FW_FILTER_WR_IVLAN_VLDM_F       FW_FILTER_WR_IVLAN_VLDM_V(1U)
+
+#define FW_FILTER_WR_OVLAN_VLDM_S       2
+#define FW_FILTER_WR_OVLAN_VLDM_M       0x1
+#define FW_FILTER_WR_OVLAN_VLDM_V(x)    ((x) << FW_FILTER_WR_OVLAN_VLDM_S)
+#define FW_FILTER_WR_OVLAN_VLDM_G(x)    \
+	(((x) >> FW_FILTER_WR_OVLAN_VLDM_S) & FW_FILTER_WR_OVLAN_VLDM_M)
+#define FW_FILTER_WR_OVLAN_VLDM_F       FW_FILTER_WR_OVLAN_VLDM_V(1U)
+
+#define FW_FILTER_WR_RX_CHAN_S          15
+#define FW_FILTER_WR_RX_CHAN_M          0x1
+#define FW_FILTER_WR_RX_CHAN_V(x)       ((x) << FW_FILTER_WR_RX_CHAN_S)
+#define FW_FILTER_WR_RX_CHAN_G(x)       \
+	(((x) >> FW_FILTER_WR_RX_CHAN_S) & FW_FILTER_WR_RX_CHAN_M)
+#define FW_FILTER_WR_RX_CHAN_F  FW_FILTER_WR_RX_CHAN_V(1U)
+
+#define FW_FILTER_WR_RX_RPL_IQ_S        0
+#define FW_FILTER_WR_RX_RPL_IQ_M        0x3ff
+#define FW_FILTER_WR_RX_RPL_IQ_V(x)     ((x) << FW_FILTER_WR_RX_RPL_IQ_S)
+#define FW_FILTER_WR_RX_RPL_IQ_G(x)     \
+	(((x) >> FW_FILTER_WR_RX_RPL_IQ_S) & FW_FILTER_WR_RX_RPL_IQ_M)
+
+#define FW_FILTER2_WR_FILTER_TYPE_S	1
+#define FW_FILTER2_WR_FILTER_TYPE_M	0x1
+#define FW_FILTER2_WR_FILTER_TYPE_V(x)	((x) << FW_FILTER2_WR_FILTER_TYPE_S)
+#define FW_FILTER2_WR_FILTER_TYPE_G(x)  \
+	(((x) >> FW_FILTER2_WR_FILTER_TYPE_S) & FW_FILTER2_WR_FILTER_TYPE_M)
+#define FW_FILTER2_WR_FILTER_TYPE_F	FW_FILTER2_WR_FILTER_TYPE_V(1U)
+
+#define FW_FILTER2_WR_NATMODE_S		5
+#define FW_FILTER2_WR_NATMODE_M		0x7
+#define FW_FILTER2_WR_NATMODE_V(x)	((x) << FW_FILTER2_WR_NATMODE_S)
+#define FW_FILTER2_WR_NATMODE_G(x)      \
+	(((x) >> FW_FILTER2_WR_NATMODE_S) & FW_FILTER2_WR_NATMODE_M)
+
+#define FW_FILTER2_WR_NATFLAGCHECK_S	4
+#define FW_FILTER2_WR_NATFLAGCHECK_M	0x1
+#define FW_FILTER2_WR_NATFLAGCHECK_V(x)	((x) << FW_FILTER2_WR_NATFLAGCHECK_S)
+#define FW_FILTER2_WR_NATFLAGCHECK_G(x) \
+	(((x) >> FW_FILTER2_WR_NATFLAGCHECK_S) & FW_FILTER2_WR_NATFLAGCHECK_M)
+#define FW_FILTER2_WR_NATFLAGCHECK_F	FW_FILTER2_WR_NATFLAGCHECK_V(1U)
+
+#define FW_FILTER2_WR_ULP_TYPE_S	0
+#define FW_FILTER2_WR_ULP_TYPE_M	0xf
+#define FW_FILTER2_WR_ULP_TYPE_V(x)	((x) << FW_FILTER2_WR_ULP_TYPE_S)
+#define FW_FILTER2_WR_ULP_TYPE_G(x)     \
+	(((x) >> FW_FILTER2_WR_ULP_TYPE_S) & FW_FILTER2_WR_ULP_TYPE_M)
+
+#define FW_FILTER_WR_MACI_S     23
+#define FW_FILTER_WR_MACI_M     0x1ff
+#define FW_FILTER_WR_MACI_V(x)  ((x) << FW_FILTER_WR_MACI_S)
+#define FW_FILTER_WR_MACI_G(x)  \
+	(((x) >> FW_FILTER_WR_MACI_S) & FW_FILTER_WR_MACI_M)
+
+#define FW_FILTER_WR_MACIM_S    14
+#define FW_FILTER_WR_MACIM_M    0x1ff
+#define FW_FILTER_WR_MACIM_V(x) ((x) << FW_FILTER_WR_MACIM_S)
+#define FW_FILTER_WR_MACIM_G(x) \
+	(((x) >> FW_FILTER_WR_MACIM_S) & FW_FILTER_WR_MACIM_M)
+
+#define FW_FILTER_WR_FCOE_S     13
+#define FW_FILTER_WR_FCOE_M     0x1
+#define FW_FILTER_WR_FCOE_V(x)  ((x) << FW_FILTER_WR_FCOE_S)
+#define FW_FILTER_WR_FCOE_G(x)  \
+	(((x) >> FW_FILTER_WR_FCOE_S) & FW_FILTER_WR_FCOE_M)
+#define FW_FILTER_WR_FCOE_F     FW_FILTER_WR_FCOE_V(1U)
+
+#define FW_FILTER_WR_FCOEM_S    12
+#define FW_FILTER_WR_FCOEM_M    0x1
+#define FW_FILTER_WR_FCOEM_V(x) ((x) << FW_FILTER_WR_FCOEM_S)
+#define FW_FILTER_WR_FCOEM_G(x) \
+	(((x) >> FW_FILTER_WR_FCOEM_S) & FW_FILTER_WR_FCOEM_M)
+#define FW_FILTER_WR_FCOEM_F    FW_FILTER_WR_FCOEM_V(1U)
+
+#define FW_FILTER_WR_PORT_S     9
+#define FW_FILTER_WR_PORT_M     0x7
+#define FW_FILTER_WR_PORT_V(x)  ((x) << FW_FILTER_WR_PORT_S)
+#define FW_FILTER_WR_PORT_G(x)  \
+	(((x) >> FW_FILTER_WR_PORT_S) & FW_FILTER_WR_PORT_M)
+
+#define FW_FILTER_WR_PORTM_S    6
+#define FW_FILTER_WR_PORTM_M    0x7
+#define FW_FILTER_WR_PORTM_V(x) ((x) << FW_FILTER_WR_PORTM_S)
+#define FW_FILTER_WR_PORTM_G(x) \
+	(((x) >> FW_FILTER_WR_PORTM_S) & FW_FILTER_WR_PORTM_M)
+
+#define FW_FILTER_WR_MATCHTYPE_S        3
+#define FW_FILTER_WR_MATCHTYPE_M        0x7
+#define FW_FILTER_WR_MATCHTYPE_V(x)     ((x) << FW_FILTER_WR_MATCHTYPE_S)
+#define FW_FILTER_WR_MATCHTYPE_G(x)     \
+	(((x) >> FW_FILTER_WR_MATCHTYPE_S) & FW_FILTER_WR_MATCHTYPE_M)
+
+#define FW_FILTER_WR_MATCHTYPEM_S       0
+#define FW_FILTER_WR_MATCHTYPEM_M       0x7
+#define FW_FILTER_WR_MATCHTYPEM_V(x)    ((x) << FW_FILTER_WR_MATCHTYPEM_S)
+#define FW_FILTER_WR_MATCHTYPEM_G(x)    \
+	(((x) >> FW_FILTER_WR_MATCHTYPEM_S) & FW_FILTER_WR_MATCHTYPEM_M)
+
+struct fw_ulptx_wr {
+	__be32 op_to_compl;
+	__be32 flowid_len16;
+	u64 cookie;
+};
+
+#define FW_ULPTX_WR_DATA_S      28
+#define FW_ULPTX_WR_DATA_M      0x1
+#define FW_ULPTX_WR_DATA_V(x)   ((x) << FW_ULPTX_WR_DATA_S)
+#define FW_ULPTX_WR_DATA_G(x)   \
+	(((x) >> FW_ULPTX_WR_DATA_S) & FW_ULPTX_WR_DATA_M)
+#define FW_ULPTX_WR_DATA_F      FW_ULPTX_WR_DATA_V(1U)
+
+struct fw_tp_wr {
+	__be32 op_to_immdlen;
+	__be32 flowid_len16;
+	u64 cookie;
+};
+
+struct fw_eth_tx_pkt_wr {
+	__be32 op_immdlen;
+	__be32 equiq_to_len16;
+	__be64 r3;
+};
+
+enum fw_eth_tx_eo_type {
+	FW_ETH_TX_EO_TYPE_UDPSEG = 0,
+	FW_ETH_TX_EO_TYPE_TCPSEG,
+};
+
+struct fw_eth_tx_eo_wr {
+	__be32 op_immdlen;
+	__be32 equiq_to_len16;
+	__be64 r3;
+	union fw_eth_tx_eo {
+		struct fw_eth_tx_eo_udpseg {
+			__u8   type;
+			__u8   ethlen;
+			__be16 iplen;
+			__u8   udplen;
+			__u8   rtplen;
+			__be16 r4;
+			__be16 mss;
+			__be16 schedpktsize;
+			__be32 plen;
+		} udpseg;
+		struct fw_eth_tx_eo_tcpseg {
+			__u8   type;
+			__u8   ethlen;
+			__be16 iplen;
+			__u8   tcplen;
+			__u8   tsclk_tsoff;
+			__be16 r4;
+			__be16 mss;
+			__be16 r5;
+			__be32 plen;
+		} tcpseg;
+	} u;
+};
+
+#define FW_ETH_TX_EO_WR_IMMDLEN_S	0
+#define FW_ETH_TX_EO_WR_IMMDLEN_M	0x1ff
+#define FW_ETH_TX_EO_WR_IMMDLEN_V(x)	((x) << FW_ETH_TX_EO_WR_IMMDLEN_S)
+#define FW_ETH_TX_EO_WR_IMMDLEN_G(x)	\
+	(((x) >> FW_ETH_TX_EO_WR_IMMDLEN_S) & FW_ETH_TX_EO_WR_IMMDLEN_M)
+
+struct fw_ofld_connection_wr {
+	__be32 op_compl;
+	__be32 len16_pkd;
+	__u64  cookie;
+	__be64 r2;
+	__be64 r3;
+	struct fw_ofld_connection_le {
+		__be32 version_cpl;
+		__be32 filter;
+		__be32 r1;
+		__be16 lport;
+		__be16 pport;
+		union fw_ofld_connection_leip {
+			struct fw_ofld_connection_le_ipv4 {
+				__be32 pip;
+				__be32 lip;
+				__be64 r0;
+				__be64 r1;
+				__be64 r2;
+			} ipv4;
+			struct fw_ofld_connection_le_ipv6 {
+				__be64 pip_hi;
+				__be64 pip_lo;
+				__be64 lip_hi;
+				__be64 lip_lo;
+			} ipv6;
+		} u;
+	} le;
+	struct fw_ofld_connection_tcb {
+		__be32 t_state_to_astid;
+		__be16 cplrxdataack_cplpassacceptrpl;
+		__be16 rcv_adv;
+		__be32 rcv_nxt;
+		__be32 tx_max;
+		__be64 opt0;
+		__be32 opt2;
+		__be32 r1;
+		__be64 r2;
+		__be64 r3;
+	} tcb;
+};
+
+#define FW_OFLD_CONNECTION_WR_VERSION_S                31
+#define FW_OFLD_CONNECTION_WR_VERSION_M                0x1
+#define FW_OFLD_CONNECTION_WR_VERSION_V(x)     \
+	((x) << FW_OFLD_CONNECTION_WR_VERSION_S)
+#define FW_OFLD_CONNECTION_WR_VERSION_G(x)     \
+	(((x) >> FW_OFLD_CONNECTION_WR_VERSION_S) & \
+	FW_OFLD_CONNECTION_WR_VERSION_M)
+#define FW_OFLD_CONNECTION_WR_VERSION_F        \
+	FW_OFLD_CONNECTION_WR_VERSION_V(1U)
+
+#define FW_OFLD_CONNECTION_WR_CPL_S    30
+#define FW_OFLD_CONNECTION_WR_CPL_M    0x1
+#define FW_OFLD_CONNECTION_WR_CPL_V(x) ((x) << FW_OFLD_CONNECTION_WR_CPL_S)
+#define FW_OFLD_CONNECTION_WR_CPL_G(x) \
+	(((x) >> FW_OFLD_CONNECTION_WR_CPL_S) & FW_OFLD_CONNECTION_WR_CPL_M)
+#define FW_OFLD_CONNECTION_WR_CPL_F    FW_OFLD_CONNECTION_WR_CPL_V(1U)
+
+#define FW_OFLD_CONNECTION_WR_T_STATE_S                28
+#define FW_OFLD_CONNECTION_WR_T_STATE_M                0xf
+#define FW_OFLD_CONNECTION_WR_T_STATE_V(x)     \
+	((x) << FW_OFLD_CONNECTION_WR_T_STATE_S)
+#define FW_OFLD_CONNECTION_WR_T_STATE_G(x)     \
+	(((x) >> FW_OFLD_CONNECTION_WR_T_STATE_S) & \
+	FW_OFLD_CONNECTION_WR_T_STATE_M)
+
+#define FW_OFLD_CONNECTION_WR_RCV_SCALE_S      24
+#define FW_OFLD_CONNECTION_WR_RCV_SCALE_M      0xf
+#define FW_OFLD_CONNECTION_WR_RCV_SCALE_V(x)   \
+	((x) << FW_OFLD_CONNECTION_WR_RCV_SCALE_S)
+#define FW_OFLD_CONNECTION_WR_RCV_SCALE_G(x)   \
+	(((x) >> FW_OFLD_CONNECTION_WR_RCV_SCALE_S) & \
+	FW_OFLD_CONNECTION_WR_RCV_SCALE_M)
+
+#define FW_OFLD_CONNECTION_WR_ASTID_S          0
+#define FW_OFLD_CONNECTION_WR_ASTID_M          0xffffff
+#define FW_OFLD_CONNECTION_WR_ASTID_V(x)       \
+	((x) << FW_OFLD_CONNECTION_WR_ASTID_S)
+#define FW_OFLD_CONNECTION_WR_ASTID_G(x)       \
+	(((x) >> FW_OFLD_CONNECTION_WR_ASTID_S) & FW_OFLD_CONNECTION_WR_ASTID_M)
+
+#define FW_OFLD_CONNECTION_WR_CPLRXDATAACK_S   15
+#define FW_OFLD_CONNECTION_WR_CPLRXDATAACK_M   0x1
+#define FW_OFLD_CONNECTION_WR_CPLRXDATAACK_V(x)        \
+	((x) << FW_OFLD_CONNECTION_WR_CPLRXDATAACK_S)
+#define FW_OFLD_CONNECTION_WR_CPLRXDATAACK_G(x)        \
+	(((x) >> FW_OFLD_CONNECTION_WR_CPLRXDATAACK_S) & \
+	FW_OFLD_CONNECTION_WR_CPLRXDATAACK_M)
+#define FW_OFLD_CONNECTION_WR_CPLRXDATAACK_F   \
+	FW_OFLD_CONNECTION_WR_CPLRXDATAACK_V(1U)
+
+#define FW_OFLD_CONNECTION_WR_CPLPASSACCEPTRPL_S       14
+#define FW_OFLD_CONNECTION_WR_CPLPASSACCEPTRPL_M       0x1
+#define FW_OFLD_CONNECTION_WR_CPLPASSACCEPTRPL_V(x)    \
+	((x) << FW_OFLD_CONNECTION_WR_CPLPASSACCEPTRPL_S)
+#define FW_OFLD_CONNECTION_WR_CPLPASSACCEPTRPL_G(x)    \
+	(((x) >> FW_OFLD_CONNECTION_WR_CPLPASSACCEPTRPL_S) & \
+	FW_OFLD_CONNECTION_WR_CPLPASSACCEPTRPL_M)
+#define FW_OFLD_CONNECTION_WR_CPLPASSACCEPTRPL_F       \
+	FW_OFLD_CONNECTION_WR_CPLPASSACCEPTRPL_V(1U)
+
+enum fw_flowc_mnem_tcpstate {
+	FW_FLOWC_MNEM_TCPSTATE_CLOSED   = 0, /* illegal */
+	FW_FLOWC_MNEM_TCPSTATE_LISTEN   = 1, /* illegal */
+	FW_FLOWC_MNEM_TCPSTATE_SYNSENT  = 2, /* illegal */
+	FW_FLOWC_MNEM_TCPSTATE_SYNRECEIVED = 3, /* illegal */
+	FW_FLOWC_MNEM_TCPSTATE_ESTABLISHED = 4, /* default */
+	FW_FLOWC_MNEM_TCPSTATE_CLOSEWAIT = 5, /* got peer close already */
+	FW_FLOWC_MNEM_TCPSTATE_FINWAIT1 = 6, /* haven't gotten ACK for FIN and
+					      * will resend FIN - equiv ESTAB
+					      */
+	FW_FLOWC_MNEM_TCPSTATE_CLOSING  = 7, /* haven't gotten ACK for FIN and
+					      * will resend FIN but have
+					      * received FIN
+					      */
+	FW_FLOWC_MNEM_TCPSTATE_LASTACK  = 8, /* haven't gotten ACK for FIN and
+					      * will resend FIN but have
+					      * received FIN
+					      */
+	FW_FLOWC_MNEM_TCPSTATE_FINWAIT2 = 9, /* sent FIN and got FIN + ACK,
+					      * waiting for FIN
+					      */
+	FW_FLOWC_MNEM_TCPSTATE_TIMEWAIT = 10, /* not expected */
+};
+
+enum fw_flowc_mnem_eostate {
+	FW_FLOWC_MNEM_EOSTATE_ESTABLISHED = 1, /* default */
+	/* graceful close, after sending outstanding payload */
+	FW_FLOWC_MNEM_EOSTATE_CLOSING = 2,
+};
+
+enum fw_flowc_mnem {
+	FW_FLOWC_MNEM_PFNVFN,		/* PFN [15:8] VFN [7:0] */
+	FW_FLOWC_MNEM_CH,
+	FW_FLOWC_MNEM_PORT,
+	FW_FLOWC_MNEM_IQID,
+	FW_FLOWC_MNEM_SNDNXT,
+	FW_FLOWC_MNEM_RCVNXT,
+	FW_FLOWC_MNEM_SNDBUF,
+	FW_FLOWC_MNEM_MSS,
+	FW_FLOWC_MNEM_TXDATAPLEN_MAX,
+	FW_FLOWC_MNEM_TCPSTATE,
+	FW_FLOWC_MNEM_EOSTATE,
+	FW_FLOWC_MNEM_SCHEDCLASS,
+	FW_FLOWC_MNEM_DCBPRIO,
+	FW_FLOWC_MNEM_SND_SCALE,
+	FW_FLOWC_MNEM_RCV_SCALE,
+	FW_FLOWC_MNEM_ULD_MODE,
+	FW_FLOWC_MNEM_MAX,
+};
+
+struct fw_flowc_mnemval {
+	u8 mnemonic;
+	u8 r4[3];
+	__be32 val;
+};
+
+struct fw_flowc_wr {
+	__be32 op_to_nparams;
+	__be32 flowid_len16;
+	struct fw_flowc_mnemval mnemval[];
+};
+
+#define FW_FLOWC_WR_NPARAMS_S           0
+#define FW_FLOWC_WR_NPARAMS_V(x)        ((x) << FW_FLOWC_WR_NPARAMS_S)
+
+struct fw_ofld_tx_data_wr {
+	__be32 op_to_immdlen;
+	__be32 flowid_len16;
+	__be32 plen;
+	__be32 tunnel_to_proxy;
+};
+
+#define FW_OFLD_TX_DATA_WR_ALIGNPLD_S   30
+#define FW_OFLD_TX_DATA_WR_ALIGNPLD_V(x) ((x) << FW_OFLD_TX_DATA_WR_ALIGNPLD_S)
+#define FW_OFLD_TX_DATA_WR_ALIGNPLD_F   FW_OFLD_TX_DATA_WR_ALIGNPLD_V(1U)
+
+#define FW_OFLD_TX_DATA_WR_SHOVE_S      29
+#define FW_OFLD_TX_DATA_WR_SHOVE_V(x)   ((x) << FW_OFLD_TX_DATA_WR_SHOVE_S)
+#define FW_OFLD_TX_DATA_WR_SHOVE_F      FW_OFLD_TX_DATA_WR_SHOVE_V(1U)
+
+#define FW_OFLD_TX_DATA_WR_TUNNEL_S     19
+#define FW_OFLD_TX_DATA_WR_TUNNEL_V(x)  ((x) << FW_OFLD_TX_DATA_WR_TUNNEL_S)
+
+#define FW_OFLD_TX_DATA_WR_SAVE_S       18
+#define FW_OFLD_TX_DATA_WR_SAVE_V(x)    ((x) << FW_OFLD_TX_DATA_WR_SAVE_S)
+
+#define FW_OFLD_TX_DATA_WR_FLUSH_S      17
+#define FW_OFLD_TX_DATA_WR_FLUSH_V(x)   ((x) << FW_OFLD_TX_DATA_WR_FLUSH_S)
+#define FW_OFLD_TX_DATA_WR_FLUSH_F      FW_OFLD_TX_DATA_WR_FLUSH_V(1U)
+
+#define FW_OFLD_TX_DATA_WR_URGENT_S     16
+#define FW_OFLD_TX_DATA_WR_URGENT_V(x)  ((x) << FW_OFLD_TX_DATA_WR_URGENT_S)
+
+#define FW_OFLD_TX_DATA_WR_MORE_S       15
+#define FW_OFLD_TX_DATA_WR_MORE_V(x)    ((x) << FW_OFLD_TX_DATA_WR_MORE_S)
+
+#define FW_OFLD_TX_DATA_WR_ULPMODE_S    10
+#define FW_OFLD_TX_DATA_WR_ULPMODE_V(x) ((x) << FW_OFLD_TX_DATA_WR_ULPMODE_S)
+
+#define FW_OFLD_TX_DATA_WR_ULPSUBMODE_S         6
+#define FW_OFLD_TX_DATA_WR_ULPSUBMODE_V(x)      \
+	((x) << FW_OFLD_TX_DATA_WR_ULPSUBMODE_S)
+
+struct fw_cmd_wr {
+	__be32 op_dma;
+	__be32 len16_pkd;
+	__be64 cookie_daddr;
+};
+
+#define FW_CMD_WR_DMA_S         17
+#define FW_CMD_WR_DMA_V(x)      ((x) << FW_CMD_WR_DMA_S)
+
+struct fw_eth_tx_pkt_vm_wr {
+	__be32 op_immdlen;
+	__be32 equiq_to_len16;
+	__be32 r3[2];
+	struct_group(firmware,
+		u8 ethmacdst[ETH_ALEN];
+		u8 ethmacsrc[ETH_ALEN];
+		__be16 ethtype;
+		__be16 vlantci;
+	);
+};
+
+#define FW_CMD_MAX_TIMEOUT 10000
+
+/*
+ * If a host driver does a HELLO and discovers that there's already a MASTER
+ * selected, we may have to wait for that MASTER to finish issuing RESET,
+ * configuration and INITIALIZE commands.  Also, there's a possibility that
+ * our own HELLO may get lost if it happens right as the MASTER is issuign a
+ * RESET command, so we need to be willing to make a few retries of our HELLO.
+ */
+#define FW_CMD_HELLO_TIMEOUT	(3 * FW_CMD_MAX_TIMEOUT)
+#define FW_CMD_HELLO_RETRIES	3
+
+
+enum fw_cmd_opcodes {
+	FW_LDST_CMD                    = 0x01,
+	FW_RESET_CMD                   = 0x03,
+	FW_HELLO_CMD                   = 0x04,
+	FW_BYE_CMD                     = 0x05,
+	FW_INITIALIZE_CMD              = 0x06,
+	FW_CAPS_CONFIG_CMD             = 0x07,
+	FW_PARAMS_CMD                  = 0x08,
+	FW_PFVF_CMD                    = 0x09,
+	FW_IQ_CMD                      = 0x10,
+	FW_EQ_MNGT_CMD                 = 0x11,
+	FW_EQ_ETH_CMD                  = 0x12,
+	FW_EQ_CTRL_CMD                 = 0x13,
+	FW_EQ_OFLD_CMD                 = 0x21,
+	FW_VI_CMD                      = 0x14,
+	FW_VI_MAC_CMD                  = 0x15,
+	FW_VI_RXMODE_CMD               = 0x16,
+	FW_VI_ENABLE_CMD               = 0x17,
+	FW_ACL_MAC_CMD                 = 0x18,
+	FW_ACL_VLAN_CMD                = 0x19,
+	FW_VI_STATS_CMD                = 0x1a,
+	FW_PORT_CMD                    = 0x1b,
+	FW_PORT_STATS_CMD              = 0x1c,
+	FW_PORT_LB_STATS_CMD           = 0x1d,
+	FW_PORT_TRACE_CMD              = 0x1e,
+	FW_PORT_TRACE_MMAP_CMD         = 0x1f,
+	FW_RSS_IND_TBL_CMD             = 0x20,
+	FW_RSS_GLB_CONFIG_CMD          = 0x22,
+	FW_RSS_VI_CONFIG_CMD           = 0x23,
+	FW_SCHED_CMD                   = 0x24,
+	FW_DEVLOG_CMD                  = 0x25,
+	FW_CLIP_CMD                    = 0x28,
+	FW_PTP_CMD                     = 0x3e,
+	FW_HMA_CMD                     = 0x3f,
+	FW_LASTC2E_CMD                 = 0x40,
+	FW_ERROR_CMD                   = 0x80,
+	FW_DEBUG_CMD                   = 0x81,
+};
+
+enum fw_cmd_cap {
+	FW_CMD_CAP_PF                  = 0x01,
+	FW_CMD_CAP_DMAQ                = 0x02,
+	FW_CMD_CAP_PORT                = 0x04,
+	FW_CMD_CAP_PORTPROMISC         = 0x08,
+	FW_CMD_CAP_PORTSTATS           = 0x10,
+	FW_CMD_CAP_VF                  = 0x80,
+};
+
+/*
+ * Generic command header flit0
+ */
+struct fw_cmd_hdr {
+	__be32 hi;
+	__be32 lo;
+};
+
+#define FW_CMD_OP_S             24
+#define FW_CMD_OP_M             0xff
+#define FW_CMD_OP_V(x)          ((x) << FW_CMD_OP_S)
+#define FW_CMD_OP_G(x)          (((x) >> FW_CMD_OP_S) & FW_CMD_OP_M)
+
+#define FW_CMD_REQUEST_S        23
+#define FW_CMD_REQUEST_V(x)     ((x) << FW_CMD_REQUEST_S)
+#define FW_CMD_REQUEST_F        FW_CMD_REQUEST_V(1U)
+
+#define FW_CMD_READ_S           22
+#define FW_CMD_READ_V(x)        ((x) << FW_CMD_READ_S)
+#define FW_CMD_READ_F           FW_CMD_READ_V(1U)
+
+#define FW_CMD_WRITE_S          21
+#define FW_CMD_WRITE_V(x)       ((x) << FW_CMD_WRITE_S)
+#define FW_CMD_WRITE_F          FW_CMD_WRITE_V(1U)
+
+#define FW_CMD_EXEC_S           20
+#define FW_CMD_EXEC_V(x)        ((x) << FW_CMD_EXEC_S)
+#define FW_CMD_EXEC_F           FW_CMD_EXEC_V(1U)
+
+#define FW_CMD_RAMASK_S         20
+#define FW_CMD_RAMASK_V(x)      ((x) << FW_CMD_RAMASK_S)
+
+#define FW_CMD_RETVAL_S         8
+#define FW_CMD_RETVAL_M         0xff
+#define FW_CMD_RETVAL_V(x)      ((x) << FW_CMD_RETVAL_S)
+#define FW_CMD_RETVAL_G(x)      (((x) >> FW_CMD_RETVAL_S) & FW_CMD_RETVAL_M)
+
+#define FW_CMD_LEN16_S          0
+#define FW_CMD_LEN16_V(x)       ((x) << FW_CMD_LEN16_S)
+
+#define FW_LEN16(fw_struct)	FW_CMD_LEN16_V(sizeof(fw_struct) / 16)
+
+enum fw_ldst_addrspc {
+	FW_LDST_ADDRSPC_FIRMWARE  = 0x0001,
+	FW_LDST_ADDRSPC_SGE_EGRC  = 0x0008,
+	FW_LDST_ADDRSPC_SGE_INGC  = 0x0009,
+	FW_LDST_ADDRSPC_SGE_FLMC  = 0x000a,
+	FW_LDST_ADDRSPC_SGE_CONMC = 0x000b,
+	FW_LDST_ADDRSPC_TP_PIO    = 0x0010,
+	FW_LDST_ADDRSPC_TP_TM_PIO = 0x0011,
+	FW_LDST_ADDRSPC_TP_MIB    = 0x0012,
+	FW_LDST_ADDRSPC_MDIO      = 0x0018,
+	FW_LDST_ADDRSPC_MPS       = 0x0020,
+	FW_LDST_ADDRSPC_FUNC      = 0x0028,
+	FW_LDST_ADDRSPC_FUNC_PCIE = 0x0029,
+	FW_LDST_ADDRSPC_I2C       = 0x0038,
+};
+
+enum fw_ldst_mps_fid {
+	FW_LDST_MPS_ATRB,
+	FW_LDST_MPS_RPLC
+};
+
+enum fw_ldst_func_access_ctl {
+	FW_LDST_FUNC_ACC_CTL_VIID,
+	FW_LDST_FUNC_ACC_CTL_FID
+};
+
+enum fw_ldst_func_mod_index {
+	FW_LDST_FUNC_MPS
+};
+
+struct fw_ldst_cmd {
+	__be32 op_to_addrspace;
+	__be32 cycles_to_len16;
+	union fw_ldst {
+		struct fw_ldst_addrval {
+			__be32 addr;
+			__be32 val;
+		} addrval;
+		struct fw_ldst_idctxt {
+			__be32 physid;
+			__be32 msg_ctxtflush;
+			__be32 ctxt_data7;
+			__be32 ctxt_data6;
+			__be32 ctxt_data5;
+			__be32 ctxt_data4;
+			__be32 ctxt_data3;
+			__be32 ctxt_data2;
+			__be32 ctxt_data1;
+			__be32 ctxt_data0;
+		} idctxt;
+		struct fw_ldst_mdio {
+			__be16 paddr_mmd;
+			__be16 raddr;
+			__be16 vctl;
+			__be16 rval;
+		} mdio;
+		struct fw_ldst_cim_rq {
+			u8 req_first64[8];
+			u8 req_second64[8];
+			u8 resp_first64[8];
+			u8 resp_second64[8];
+			__be32 r3[2];
+		} cim_rq;
+		union fw_ldst_mps {
+			struct fw_ldst_mps_rplc {
+				__be16 fid_idx;
+				__be16 rplcpf_pkd;
+				__be32 rplc255_224;
+				__be32 rplc223_192;
+				__be32 rplc191_160;
+				__be32 rplc159_128;
+				__be32 rplc127_96;
+				__be32 rplc95_64;
+				__be32 rplc63_32;
+				__be32 rplc31_0;
+			} rplc;
+			struct fw_ldst_mps_atrb {
+				__be16 fid_mpsid;
+				__be16 r2[3];
+				__be32 r3[2];
+				__be32 r4;
+				__be32 atrb;
+				__be16 vlan[16];
+			} atrb;
+		} mps;
+		struct fw_ldst_func {
+			u8 access_ctl;
+			u8 mod_index;
+			__be16 ctl_id;
+			__be32 offset;
+			__be64 data0;
+			__be64 data1;
+		} func;
+		struct fw_ldst_pcie {
+			u8 ctrl_to_fn;
+			u8 bnum;
+			u8 r;
+			u8 ext_r;
+			u8 select_naccess;
+			u8 pcie_fn;
+			__be16 nset_pkd;
+			__be32 data[12];
+		} pcie;
+		struct fw_ldst_i2c_deprecated {
+			u8 pid_pkd;
+			u8 base;
+			u8 boffset;
+			u8 data;
+			__be32 r9;
+		} i2c_deprecated;
+		struct fw_ldst_i2c {
+			u8 pid;
+			u8 did;
+			u8 boffset;
+			u8 blen;
+			__be32 r9;
+			__u8   data[48];
+		} i2c;
+		struct fw_ldst_le {
+			__be32 index;
+			__be32 r9;
+			u8 val[33];
+			u8 r11[7];
+		} le;
+	} u;
+};
+
+#define FW_LDST_CMD_ADDRSPACE_S		0
+#define FW_LDST_CMD_ADDRSPACE_V(x)	((x) << FW_LDST_CMD_ADDRSPACE_S)
+
+#define FW_LDST_CMD_MSG_S       31
+#define FW_LDST_CMD_MSG_V(x)	((x) << FW_LDST_CMD_MSG_S)
+
+#define FW_LDST_CMD_CTXTFLUSH_S		30
+#define FW_LDST_CMD_CTXTFLUSH_V(x)	((x) << FW_LDST_CMD_CTXTFLUSH_S)
+#define FW_LDST_CMD_CTXTFLUSH_F		FW_LDST_CMD_CTXTFLUSH_V(1U)
+
+#define FW_LDST_CMD_PADDR_S     8
+#define FW_LDST_CMD_PADDR_V(x)	((x) << FW_LDST_CMD_PADDR_S)
+
+#define FW_LDST_CMD_MMD_S       0
+#define FW_LDST_CMD_MMD_V(x)	((x) << FW_LDST_CMD_MMD_S)
+
+#define FW_LDST_CMD_FID_S       15
+#define FW_LDST_CMD_FID_V(x)	((x) << FW_LDST_CMD_FID_S)
+
+#define FW_LDST_CMD_IDX_S	0
+#define FW_LDST_CMD_IDX_V(x)	((x) << FW_LDST_CMD_IDX_S)
+
+#define FW_LDST_CMD_RPLCPF_S    0
+#define FW_LDST_CMD_RPLCPF_V(x)	((x) << FW_LDST_CMD_RPLCPF_S)
+
+#define FW_LDST_CMD_LC_S        4
+#define FW_LDST_CMD_LC_V(x)     ((x) << FW_LDST_CMD_LC_S)
+#define FW_LDST_CMD_LC_F	FW_LDST_CMD_LC_V(1U)
+
+#define FW_LDST_CMD_FN_S        0
+#define FW_LDST_CMD_FN_V(x)	((x) << FW_LDST_CMD_FN_S)
+
+#define FW_LDST_CMD_NACCESS_S           0
+#define FW_LDST_CMD_NACCESS_V(x)	((x) << FW_LDST_CMD_NACCESS_S)
+
+struct fw_reset_cmd {
+	__be32 op_to_write;
+	__be32 retval_len16;
+	__be32 val;
+	__be32 halt_pkd;
+};
+
+#define FW_RESET_CMD_HALT_S	31
+#define FW_RESET_CMD_HALT_M     0x1
+#define FW_RESET_CMD_HALT_V(x)	((x) << FW_RESET_CMD_HALT_S)
+#define FW_RESET_CMD_HALT_G(x)  \
+	(((x) >> FW_RESET_CMD_HALT_S) & FW_RESET_CMD_HALT_M)
+#define FW_RESET_CMD_HALT_F	FW_RESET_CMD_HALT_V(1U)
+
+enum fw_hellow_cmd {
+	fw_hello_cmd_stage_os		= 0x0
+};
+
+struct fw_hello_cmd {
+	__be32 op_to_write;
+	__be32 retval_len16;
+	__be32 err_to_clearinit;
+	__be32 fwrev;
+};
+
+#define FW_HELLO_CMD_ERR_S      31
+#define FW_HELLO_CMD_ERR_V(x)   ((x) << FW_HELLO_CMD_ERR_S)
+#define FW_HELLO_CMD_ERR_F	FW_HELLO_CMD_ERR_V(1U)
+
+#define FW_HELLO_CMD_INIT_S     30
+#define FW_HELLO_CMD_INIT_V(x)  ((x) << FW_HELLO_CMD_INIT_S)
+#define FW_HELLO_CMD_INIT_F	FW_HELLO_CMD_INIT_V(1U)
+
+#define FW_HELLO_CMD_MASTERDIS_S	29
+#define FW_HELLO_CMD_MASTERDIS_V(x)	((x) << FW_HELLO_CMD_MASTERDIS_S)
+
+#define FW_HELLO_CMD_MASTERFORCE_S      28
+#define FW_HELLO_CMD_MASTERFORCE_V(x)	((x) << FW_HELLO_CMD_MASTERFORCE_S)
+
+#define FW_HELLO_CMD_MBMASTER_S		24
+#define FW_HELLO_CMD_MBMASTER_M		0xfU
+#define FW_HELLO_CMD_MBMASTER_V(x)	((x) << FW_HELLO_CMD_MBMASTER_S)
+#define FW_HELLO_CMD_MBMASTER_G(x)	\
+	(((x) >> FW_HELLO_CMD_MBMASTER_S) & FW_HELLO_CMD_MBMASTER_M)
+
+#define FW_HELLO_CMD_MBASYNCNOTINT_S    23
+#define FW_HELLO_CMD_MBASYNCNOTINT_V(x)	((x) << FW_HELLO_CMD_MBASYNCNOTINT_S)
+
+#define FW_HELLO_CMD_MBASYNCNOT_S       20
+#define FW_HELLO_CMD_MBASYNCNOT_V(x)	((x) << FW_HELLO_CMD_MBASYNCNOT_S)
+
+#define FW_HELLO_CMD_STAGE_S		17
+#define FW_HELLO_CMD_STAGE_V(x)		((x) << FW_HELLO_CMD_STAGE_S)
+
+#define FW_HELLO_CMD_CLEARINIT_S        16
+#define FW_HELLO_CMD_CLEARINIT_V(x)     ((x) << FW_HELLO_CMD_CLEARINIT_S)
+#define FW_HELLO_CMD_CLEARINIT_F	FW_HELLO_CMD_CLEARINIT_V(1U)
+
+struct fw_bye_cmd {
+	__be32 op_to_write;
+	__be32 retval_len16;
+	__be64 r3;
+};
+
+struct fw_initialize_cmd {
+	__be32 op_to_write;
+	__be32 retval_len16;
+	__be64 r3;
+};
+
+enum fw_caps_config_hm {
+	FW_CAPS_CONFIG_HM_PCIE		= 0x00000001,
+	FW_CAPS_CONFIG_HM_PL		= 0x00000002,
+	FW_CAPS_CONFIG_HM_SGE		= 0x00000004,
+	FW_CAPS_CONFIG_HM_CIM		= 0x00000008,
+	FW_CAPS_CONFIG_HM_ULPTX		= 0x00000010,
+	FW_CAPS_CONFIG_HM_TP		= 0x00000020,
+	FW_CAPS_CONFIG_HM_ULPRX		= 0x00000040,
+	FW_CAPS_CONFIG_HM_PMRX		= 0x00000080,
+	FW_CAPS_CONFIG_HM_PMTX		= 0x00000100,
+	FW_CAPS_CONFIG_HM_MC		= 0x00000200,
+	FW_CAPS_CONFIG_HM_LE		= 0x00000400,
+	FW_CAPS_CONFIG_HM_MPS		= 0x00000800,
+	FW_CAPS_CONFIG_HM_XGMAC		= 0x00001000,
+	FW_CAPS_CONFIG_HM_CPLSWITCH	= 0x00002000,
+	FW_CAPS_CONFIG_HM_T4DBG		= 0x00004000,
+	FW_CAPS_CONFIG_HM_MI		= 0x00008000,
+	FW_CAPS_CONFIG_HM_I2CM		= 0x00010000,
+	FW_CAPS_CONFIG_HM_NCSI		= 0x00020000,
+	FW_CAPS_CONFIG_HM_SMB		= 0x00040000,
+	FW_CAPS_CONFIG_HM_MA		= 0x00080000,
+	FW_CAPS_CONFIG_HM_EDRAM		= 0x00100000,
+	FW_CAPS_CONFIG_HM_PMU		= 0x00200000,
+	FW_CAPS_CONFIG_HM_UART		= 0x00400000,
+	FW_CAPS_CONFIG_HM_SF		= 0x00800000,
+};
+
+enum fw_caps_config_nbm {
+	FW_CAPS_CONFIG_NBM_IPMI		= 0x00000001,
+	FW_CAPS_CONFIG_NBM_NCSI		= 0x00000002,
+};
+
+enum fw_caps_config_link {
+	FW_CAPS_CONFIG_LINK_PPP		= 0x00000001,
+	FW_CAPS_CONFIG_LINK_QFC		= 0x00000002,
+	FW_CAPS_CONFIG_LINK_DCBX	= 0x00000004,
+};
+
+enum fw_caps_config_switch {
+	FW_CAPS_CONFIG_SWITCH_INGRESS	= 0x00000001,
+	FW_CAPS_CONFIG_SWITCH_EGRESS	= 0x00000002,
+};
+
+enum fw_caps_config_nic {
+	FW_CAPS_CONFIG_NIC		= 0x00000001,
+	FW_CAPS_CONFIG_NIC_VM		= 0x00000002,
+	FW_CAPS_CONFIG_NIC_HASHFILTER	= 0x00000020,
+	FW_CAPS_CONFIG_NIC_ETHOFLD	= 0x00000040,
+};
+
+enum fw_caps_config_ofld {
+	FW_CAPS_CONFIG_OFLD		= 0x00000001,
+};
+
+enum fw_caps_config_rdma {
+	FW_CAPS_CONFIG_RDMA_RDDP	= 0x00000001,
+	FW_CAPS_CONFIG_RDMA_RDMAC	= 0x00000002,
+};
+
+enum fw_caps_config_iscsi {
+	FW_CAPS_CONFIG_ISCSI_INITIATOR_PDU = 0x00000001,
+	FW_CAPS_CONFIG_ISCSI_TARGET_PDU = 0x00000002,
+	FW_CAPS_CONFIG_ISCSI_INITIATOR_CNXOFLD = 0x00000004,
+	FW_CAPS_CONFIG_ISCSI_TARGET_CNXOFLD = 0x00000008,
+};
+
+enum fw_caps_config_crypto {
+	FW_CAPS_CONFIG_CRYPTO_LOOKASIDE = 0x00000001,
+	FW_CAPS_CONFIG_TLS_INLINE = 0x00000002,
+	FW_CAPS_CONFIG_IPSEC_INLINE = 0x00000004,
+	FW_CAPS_CONFIG_TLS_HW = 0x00000008,
+};
+
+enum fw_caps_config_fcoe {
+	FW_CAPS_CONFIG_FCOE_INITIATOR	= 0x00000001,
+	FW_CAPS_CONFIG_FCOE_TARGET	= 0x00000002,
+	FW_CAPS_CONFIG_FCOE_CTRL_OFLD	= 0x00000004,
+};
+
+enum fw_memtype_cf {
+	FW_MEMTYPE_CF_EDC0		= 0x0,
+	FW_MEMTYPE_CF_EDC1		= 0x1,
+	FW_MEMTYPE_CF_EXTMEM		= 0x2,
+	FW_MEMTYPE_CF_FLASH		= 0x4,
+	FW_MEMTYPE_CF_INTERNAL		= 0x5,
+	FW_MEMTYPE_CF_EXTMEM1           = 0x6,
+	FW_MEMTYPE_CF_HMA		= 0x7,
+};
+
+struct fw_caps_config_cmd {
+	__be32 op_to_write;
+	__be32 cfvalid_to_len16;
+	__be32 r2;
+	__be32 hwmbitmap;
+	__be16 nbmcaps;
+	__be16 linkcaps;
+	__be16 switchcaps;
+	__be16 r3;
+	__be16 niccaps;
+	__be16 ofldcaps;
+	__be16 rdmacaps;
+	__be16 cryptocaps;
+	__be16 iscsicaps;
+	__be16 fcoecaps;
+	__be32 cfcsum;
+	__be32 finiver;
+	__be32 finicsum;
+};
+
+#define FW_CAPS_CONFIG_CMD_CFVALID_S    27
+#define FW_CAPS_CONFIG_CMD_CFVALID_V(x) ((x) << FW_CAPS_CONFIG_CMD_CFVALID_S)
+#define FW_CAPS_CONFIG_CMD_CFVALID_F    FW_CAPS_CONFIG_CMD_CFVALID_V(1U)
+
+#define FW_CAPS_CONFIG_CMD_MEMTYPE_CF_S		24
+#define FW_CAPS_CONFIG_CMD_MEMTYPE_CF_V(x)	\
+	((x) << FW_CAPS_CONFIG_CMD_MEMTYPE_CF_S)
+
+#define FW_CAPS_CONFIG_CMD_MEMADDR64K_CF_S      16
+#define FW_CAPS_CONFIG_CMD_MEMADDR64K_CF_V(x)	\
+	((x) << FW_CAPS_CONFIG_CMD_MEMADDR64K_CF_S)
+
+/*
+ * params command mnemonics
+ */
+enum fw_params_mnem {
+	FW_PARAMS_MNEM_DEV		= 1,	/* device params */
+	FW_PARAMS_MNEM_PFVF		= 2,	/* function params */
+	FW_PARAMS_MNEM_REG		= 3,	/* limited register access */
+	FW_PARAMS_MNEM_DMAQ		= 4,	/* dma queue params */
+	FW_PARAMS_MNEM_CHNET            = 5,    /* chnet params */
+	FW_PARAMS_MNEM_LAST
+};
+
+/*
+ * device parameters
+ */
+
+#define FW_PARAMS_PARAM_FILTER_MODE_S 16
+#define FW_PARAMS_PARAM_FILTER_MODE_M 0xffff
+#define FW_PARAMS_PARAM_FILTER_MODE_V(x)          \
+	((x) << FW_PARAMS_PARAM_FILTER_MODE_S)
+#define FW_PARAMS_PARAM_FILTER_MODE_G(x)          \
+	(((x) >> FW_PARAMS_PARAM_FILTER_MODE_S) & \
+	FW_PARAMS_PARAM_FILTER_MODE_M)
+
+#define FW_PARAMS_PARAM_FILTER_MASK_S 0
+#define FW_PARAMS_PARAM_FILTER_MASK_M 0xffff
+#define FW_PARAMS_PARAM_FILTER_MASK_V(x)          \
+	((x) << FW_PARAMS_PARAM_FILTER_MASK_S)
+#define FW_PARAMS_PARAM_FILTER_MASK_G(x)          \
+	(((x) >> FW_PARAMS_PARAM_FILTER_MASK_S) & \
+	FW_PARAMS_PARAM_FILTER_MASK_M)
+
+enum fw_params_param_dev {
+	FW_PARAMS_PARAM_DEV_CCLK	= 0x00, /* chip core clock in khz */
+	FW_PARAMS_PARAM_DEV_PORTVEC	= 0x01, /* the port vector */
+	FW_PARAMS_PARAM_DEV_NTID	= 0x02, /* reads the number of TIDs
+						 * allocated by the device's
+						 * Lookup Engine
+						 */
+	FW_PARAMS_PARAM_DEV_FLOWC_BUFFIFO_SZ = 0x03,
+	FW_PARAMS_PARAM_DEV_INTVER_NIC	= 0x04,
+	FW_PARAMS_PARAM_DEV_INTVER_VNIC = 0x05,
+	FW_PARAMS_PARAM_DEV_INTVER_OFLD = 0x06,
+	FW_PARAMS_PARAM_DEV_INTVER_RI	= 0x07,
+	FW_PARAMS_PARAM_DEV_INTVER_ISCSIPDU = 0x08,
+	FW_PARAMS_PARAM_DEV_INTVER_ISCSI = 0x09,
+	FW_PARAMS_PARAM_DEV_INTVER_FCOE = 0x0A,
+	FW_PARAMS_PARAM_DEV_FWREV = 0x0B,
+	FW_PARAMS_PARAM_DEV_TPREV = 0x0C,
+	FW_PARAMS_PARAM_DEV_CF = 0x0D,
+	FW_PARAMS_PARAM_DEV_PHYFW = 0x0F,
+	FW_PARAMS_PARAM_DEV_DIAG = 0x11,
+	FW_PARAMS_PARAM_DEV_MAXORDIRD_QP = 0x13, /* max supported QP IRD/ORD */
+	FW_PARAMS_PARAM_DEV_MAXIRD_ADAPTER = 0x14, /* max supported adap IRD */
+	FW_PARAMS_PARAM_DEV_ULPTX_MEMWRITE_DSGL = 0x17,
+	FW_PARAMS_PARAM_DEV_FWCACHE = 0x18,
+	FW_PARAMS_PARAM_DEV_SCFGREV = 0x1A,
+	FW_PARAMS_PARAM_DEV_VPDREV = 0x1B,
+	FW_PARAMS_PARAM_DEV_RI_FR_NSMR_TPTE_WR	= 0x1C,
+	FW_PARAMS_PARAM_DEV_FILTER2_WR  = 0x1D,
+	FW_PARAMS_PARAM_DEV_MPSBGMAP	= 0x1E,
+	FW_PARAMS_PARAM_DEV_TPCHMAP     = 0x1F,
+	FW_PARAMS_PARAM_DEV_HMA_SIZE	= 0x20,
+	FW_PARAMS_PARAM_DEV_RDMA_WRITE_WITH_IMM = 0x21,
+	FW_PARAMS_PARAM_DEV_PPOD_EDRAM  = 0x23,
+	FW_PARAMS_PARAM_DEV_RI_WRITE_CMPL_WR    = 0x24,
+	FW_PARAMS_PARAM_DEV_HPFILTER_REGION_SUPPORT = 0x26,
+	FW_PARAMS_PARAM_DEV_OPAQUE_VIID_SMT_EXTN = 0x27,
+	FW_PARAMS_PARAM_DEV_HASHFILTER_WITH_OFLD = 0x28,
+	FW_PARAMS_PARAM_DEV_DBQ_TIMER	= 0x29,
+	FW_PARAMS_PARAM_DEV_DBQ_TIMERTICK = 0x2A,
+	FW_PARAMS_PARAM_DEV_NUM_TM_CLASS = 0x2B,
+	FW_PARAMS_PARAM_DEV_FILTER = 0x2E,
+	FW_PARAMS_PARAM_DEV_KTLS_HW = 0x31,
+};
+
+/*
+ * physical and virtual function parameters
+ */
+enum fw_params_param_pfvf {
+	FW_PARAMS_PARAM_PFVF_RWXCAPS	= 0x00,
+	FW_PARAMS_PARAM_PFVF_ROUTE_START = 0x01,
+	FW_PARAMS_PARAM_PFVF_ROUTE_END = 0x02,
+	FW_PARAMS_PARAM_PFVF_CLIP_START = 0x03,
+	FW_PARAMS_PARAM_PFVF_CLIP_END = 0x04,
+	FW_PARAMS_PARAM_PFVF_FILTER_START = 0x05,
+	FW_PARAMS_PARAM_PFVF_FILTER_END = 0x06,
+	FW_PARAMS_PARAM_PFVF_SERVER_START = 0x07,
+	FW_PARAMS_PARAM_PFVF_SERVER_END = 0x08,
+	FW_PARAMS_PARAM_PFVF_TDDP_START = 0x09,
+	FW_PARAMS_PARAM_PFVF_TDDP_END = 0x0A,
+	FW_PARAMS_PARAM_PFVF_ISCSI_START = 0x0B,
+	FW_PARAMS_PARAM_PFVF_ISCSI_END = 0x0C,
+	FW_PARAMS_PARAM_PFVF_STAG_START = 0x0D,
+	FW_PARAMS_PARAM_PFVF_STAG_END = 0x0E,
+	FW_PARAMS_PARAM_PFVF_RQ_START = 0x1F,
+	FW_PARAMS_PARAM_PFVF_RQ_END	= 0x10,
+	FW_PARAMS_PARAM_PFVF_PBL_START = 0x11,
+	FW_PARAMS_PARAM_PFVF_PBL_END	= 0x12,
+	FW_PARAMS_PARAM_PFVF_L2T_START = 0x13,
+	FW_PARAMS_PARAM_PFVF_L2T_END = 0x14,
+	FW_PARAMS_PARAM_PFVF_SQRQ_START = 0x15,
+	FW_PARAMS_PARAM_PFVF_SQRQ_END	= 0x16,
+	FW_PARAMS_PARAM_PFVF_CQ_START	= 0x17,
+	FW_PARAMS_PARAM_PFVF_CQ_END	= 0x18,
+	FW_PARAMS_PARAM_PFVF_SRQ_START  = 0x19,
+	FW_PARAMS_PARAM_PFVF_SRQ_END    = 0x1A,
+	FW_PARAMS_PARAM_PFVF_SCHEDCLASS_ETH = 0x20,
+	FW_PARAMS_PARAM_PFVF_VIID       = 0x24,
+	FW_PARAMS_PARAM_PFVF_CPMASK     = 0x25,
+	FW_PARAMS_PARAM_PFVF_OCQ_START  = 0x26,
+	FW_PARAMS_PARAM_PFVF_OCQ_END    = 0x27,
+	FW_PARAMS_PARAM_PFVF_CONM_MAP   = 0x28,
+	FW_PARAMS_PARAM_PFVF_IQFLINT_START = 0x29,
+	FW_PARAMS_PARAM_PFVF_IQFLINT_END = 0x2A,
+	FW_PARAMS_PARAM_PFVF_EQ_START	= 0x2B,
+	FW_PARAMS_PARAM_PFVF_EQ_END	= 0x2C,
+	FW_PARAMS_PARAM_PFVF_ACTIVE_FILTER_START = 0x2D,
+	FW_PARAMS_PARAM_PFVF_ACTIVE_FILTER_END = 0x2E,
+	FW_PARAMS_PARAM_PFVF_ETHOFLD_START = 0x2F,
+	FW_PARAMS_PARAM_PFVF_ETHOFLD_END = 0x30,
+	FW_PARAMS_PARAM_PFVF_CPLFW4MSG_ENCAP = 0x31,
+	FW_PARAMS_PARAM_PFVF_HPFILTER_START = 0x32,
+	FW_PARAMS_PARAM_PFVF_HPFILTER_END = 0x33,
+	FW_PARAMS_PARAM_PFVF_TLS_START = 0x34,
+	FW_PARAMS_PARAM_PFVF_TLS_END = 0x35,
+	FW_PARAMS_PARAM_PFVF_RAWF_START = 0x36,
+	FW_PARAMS_PARAM_PFVF_RAWF_END = 0x37,
+	FW_PARAMS_PARAM_PFVF_NCRYPTO_LOOKASIDE = 0x39,
+	FW_PARAMS_PARAM_PFVF_PORT_CAPS32 = 0x3A,
+	FW_PARAMS_PARAM_PFVF_PPOD_EDRAM_START = 0x3B,
+	FW_PARAMS_PARAM_PFVF_PPOD_EDRAM_END = 0x3C,
+	FW_PARAMS_PARAM_PFVF_LINK_STATE = 0x40,
+};
+
+/* Virtual link state as seen by the specified VF */
+enum vf_link_states {
+	FW_VF_LINK_STATE_AUTO		= 0x00,
+	FW_VF_LINK_STATE_ENABLE		= 0x01,
+	FW_VF_LINK_STATE_DISABLE	= 0x02,
+};
+
+/*
+ * dma queue parameters
+ */
+enum fw_params_param_dmaq {
+	FW_PARAMS_PARAM_DMAQ_IQ_DCAEN_DCACPU = 0x00,
+	FW_PARAMS_PARAM_DMAQ_IQ_INTCNTTHRESH = 0x01,
+	FW_PARAMS_PARAM_DMAQ_EQ_CMPLIQID_MNGT = 0x10,
+	FW_PARAMS_PARAM_DMAQ_EQ_CMPLIQID_CTRL = 0x11,
+	FW_PARAMS_PARAM_DMAQ_EQ_SCHEDCLASS_ETH = 0x12,
+	FW_PARAMS_PARAM_DMAQ_EQ_DCBPRIO_ETH = 0x13,
+	FW_PARAMS_PARAM_DMAQ_EQ_TIMERIX	= 0x15,
+	FW_PARAMS_PARAM_DMAQ_CONM_CTXT = 0x20,
+};
+
+enum fw_params_param_dev_ktls_hw {
+	FW_PARAMS_PARAM_DEV_KTLS_HW_DISABLE      = 0x00,
+	FW_PARAMS_PARAM_DEV_KTLS_HW_ENABLE       = 0x01,
+	FW_PARAMS_PARAM_DEV_KTLS_HW_USER_ENABLE  = 0x01,
+};
+
+enum fw_params_param_dev_phyfw {
+	FW_PARAMS_PARAM_DEV_PHYFW_DOWNLOAD = 0x00,
+	FW_PARAMS_PARAM_DEV_PHYFW_VERSION = 0x01,
+};
+
+enum fw_params_param_dev_diag {
+	FW_PARAM_DEV_DIAG_TMP		= 0x00,
+	FW_PARAM_DEV_DIAG_VDD		= 0x01,
+	FW_PARAM_DEV_DIAG_MAXTMPTHRESH	= 0x02,
+};
+
+enum fw_params_param_dev_filter {
+	FW_PARAM_DEV_FILTER_VNIC_MODE   = 0x00,
+	FW_PARAM_DEV_FILTER_MODE_MASK   = 0x01,
+};
+
+enum fw_params_param_dev_fwcache {
+	FW_PARAM_DEV_FWCACHE_FLUSH      = 0x00,
+	FW_PARAM_DEV_FWCACHE_FLUSHINV   = 0x01,
+};
+
+#define FW_PARAMS_MNEM_S	24
+#define FW_PARAMS_MNEM_V(x)	((x) << FW_PARAMS_MNEM_S)
+
+#define FW_PARAMS_PARAM_X_S     16
+#define FW_PARAMS_PARAM_X_V(x)	((x) << FW_PARAMS_PARAM_X_S)
+
+#define FW_PARAMS_PARAM_Y_S	8
+#define FW_PARAMS_PARAM_Y_M	0xffU
+#define FW_PARAMS_PARAM_Y_V(x)	((x) << FW_PARAMS_PARAM_Y_S)
+#define FW_PARAMS_PARAM_Y_G(x)	(((x) >> FW_PARAMS_PARAM_Y_S) &\
+		FW_PARAMS_PARAM_Y_M)
+
+#define FW_PARAMS_PARAM_Z_S	0
+#define FW_PARAMS_PARAM_Z_M	0xffu
+#define FW_PARAMS_PARAM_Z_V(x)	((x) << FW_PARAMS_PARAM_Z_S)
+#define FW_PARAMS_PARAM_Z_G(x)	(((x) >> FW_PARAMS_PARAM_Z_S) &\
+		FW_PARAMS_PARAM_Z_M)
+
+#define FW_PARAMS_PARAM_XYZ_S		0
+#define FW_PARAMS_PARAM_XYZ_V(x)	((x) << FW_PARAMS_PARAM_XYZ_S)
+
+#define FW_PARAMS_PARAM_YZ_S		0
+#define FW_PARAMS_PARAM_YZ_V(x)		((x) << FW_PARAMS_PARAM_YZ_S)
+
+struct fw_params_cmd {
+	__be32 op_to_vfn;
+	__be32 retval_len16;
+	struct fw_params_param {
+		__be32 mnem;
+		__be32 val;
+	} param[7];
+};
+
+#define FW_PARAMS_CMD_PFN_S     8
+#define FW_PARAMS_CMD_PFN_V(x)	((x) << FW_PARAMS_CMD_PFN_S)
+
+#define FW_PARAMS_CMD_VFN_S     0
+#define FW_PARAMS_CMD_VFN_V(x)	((x) << FW_PARAMS_CMD_VFN_S)
+
+struct fw_pfvf_cmd {
+	__be32 op_to_vfn;
+	__be32 retval_len16;
+	__be32 niqflint_niq;
+	__be32 type_to_neq;
+	__be32 tc_to_nexactf;
+	__be32 r_caps_to_nethctrl;
+	__be16 nricq;
+	__be16 nriqp;
+	__be32 r4;
+};
+
+#define FW_PFVF_CMD_PFN_S	8
+#define FW_PFVF_CMD_PFN_V(x)	((x) << FW_PFVF_CMD_PFN_S)
+
+#define FW_PFVF_CMD_VFN_S       0
+#define FW_PFVF_CMD_VFN_V(x)	((x) << FW_PFVF_CMD_VFN_S)
+
+#define FW_PFVF_CMD_NIQFLINT_S          20
+#define FW_PFVF_CMD_NIQFLINT_M          0xfff
+#define FW_PFVF_CMD_NIQFLINT_V(x)	((x) << FW_PFVF_CMD_NIQFLINT_S)
+#define FW_PFVF_CMD_NIQFLINT_G(x)	\
+	(((x) >> FW_PFVF_CMD_NIQFLINT_S) & FW_PFVF_CMD_NIQFLINT_M)
+
+#define FW_PFVF_CMD_NIQ_S       0
+#define FW_PFVF_CMD_NIQ_M       0xfffff
+#define FW_PFVF_CMD_NIQ_V(x)	((x) << FW_PFVF_CMD_NIQ_S)
+#define FW_PFVF_CMD_NIQ_G(x)	\
+	(((x) >> FW_PFVF_CMD_NIQ_S) & FW_PFVF_CMD_NIQ_M)
+
+#define FW_PFVF_CMD_TYPE_S      31
+#define FW_PFVF_CMD_TYPE_M      0x1
+#define FW_PFVF_CMD_TYPE_V(x)   ((x) << FW_PFVF_CMD_TYPE_S)
+#define FW_PFVF_CMD_TYPE_G(x)	\
+	(((x) >> FW_PFVF_CMD_TYPE_S) & FW_PFVF_CMD_TYPE_M)
+#define FW_PFVF_CMD_TYPE_F      FW_PFVF_CMD_TYPE_V(1U)
+
+#define FW_PFVF_CMD_CMASK_S     24
+#define FW_PFVF_CMD_CMASK_M	0xf
+#define FW_PFVF_CMD_CMASK_V(x)	((x) << FW_PFVF_CMD_CMASK_S)
+#define FW_PFVF_CMD_CMASK_G(x)	\
+	(((x) >> FW_PFVF_CMD_CMASK_S) & FW_PFVF_CMD_CMASK_M)
+
+#define FW_PFVF_CMD_PMASK_S     20
+#define FW_PFVF_CMD_PMASK_M	0xf
+#define FW_PFVF_CMD_PMASK_V(x)	((x) << FW_PFVF_CMD_PMASK_S)
+#define FW_PFVF_CMD_PMASK_G(x) \
+	(((x) >> FW_PFVF_CMD_PMASK_S) & FW_PFVF_CMD_PMASK_M)
+
+#define FW_PFVF_CMD_NEQ_S       0
+#define FW_PFVF_CMD_NEQ_M       0xfffff
+#define FW_PFVF_CMD_NEQ_V(x)	((x) << FW_PFVF_CMD_NEQ_S)
+#define FW_PFVF_CMD_NEQ_G(x)	\
+	(((x) >> FW_PFVF_CMD_NEQ_S) & FW_PFVF_CMD_NEQ_M)
+
+#define FW_PFVF_CMD_TC_S        24
+#define FW_PFVF_CMD_TC_M        0xff
+#define FW_PFVF_CMD_TC_V(x)	((x) << FW_PFVF_CMD_TC_S)
+#define FW_PFVF_CMD_TC_G(x)	(((x) >> FW_PFVF_CMD_TC_S) & FW_PFVF_CMD_TC_M)
+
+#define FW_PFVF_CMD_NVI_S       16
+#define FW_PFVF_CMD_NVI_M       0xff
+#define FW_PFVF_CMD_NVI_V(x)	((x) << FW_PFVF_CMD_NVI_S)
+#define FW_PFVF_CMD_NVI_G(x)	(((x) >> FW_PFVF_CMD_NVI_S) & FW_PFVF_CMD_NVI_M)
+
+#define FW_PFVF_CMD_NEXACTF_S           0
+#define FW_PFVF_CMD_NEXACTF_M           0xffff
+#define FW_PFVF_CMD_NEXACTF_V(x)	((x) << FW_PFVF_CMD_NEXACTF_S)
+#define FW_PFVF_CMD_NEXACTF_G(x)	\
+	(((x) >> FW_PFVF_CMD_NEXACTF_S) & FW_PFVF_CMD_NEXACTF_M)
+
+#define FW_PFVF_CMD_R_CAPS_S    24
+#define FW_PFVF_CMD_R_CAPS_M    0xff
+#define FW_PFVF_CMD_R_CAPS_V(x) ((x) << FW_PFVF_CMD_R_CAPS_S)
+#define FW_PFVF_CMD_R_CAPS_G(x) \
+	(((x) >> FW_PFVF_CMD_R_CAPS_S) & FW_PFVF_CMD_R_CAPS_M)
+
+#define FW_PFVF_CMD_WX_CAPS_S           16
+#define FW_PFVF_CMD_WX_CAPS_M           0xff
+#define FW_PFVF_CMD_WX_CAPS_V(x)	((x) << FW_PFVF_CMD_WX_CAPS_S)
+#define FW_PFVF_CMD_WX_CAPS_G(x)	\
+	(((x) >> FW_PFVF_CMD_WX_CAPS_S) & FW_PFVF_CMD_WX_CAPS_M)
+
+#define FW_PFVF_CMD_NETHCTRL_S          0
+#define FW_PFVF_CMD_NETHCTRL_M          0xffff
+#define FW_PFVF_CMD_NETHCTRL_V(x)	((x) << FW_PFVF_CMD_NETHCTRL_S)
+#define FW_PFVF_CMD_NETHCTRL_G(x)	\
+	(((x) >> FW_PFVF_CMD_NETHCTRL_S) & FW_PFVF_CMD_NETHCTRL_M)
+
+enum fw_iq_type {
+	FW_IQ_TYPE_FL_INT_CAP,
+	FW_IQ_TYPE_NO_FL_INT_CAP
+};
+
+enum fw_iq_iqtype {
+	FW_IQ_IQTYPE_OTHER,
+	FW_IQ_IQTYPE_NIC,
+	FW_IQ_IQTYPE_OFLD,
+};
+
+struct fw_iq_cmd {
+	__be32 op_to_vfn;
+	__be32 alloc_to_len16;
+	__be16 physiqid;
+	__be16 iqid;
+	__be16 fl0id;
+	__be16 fl1id;
+	__be32 type_to_iqandstindex;
+	__be16 iqdroprss_to_iqesize;
+	__be16 iqsize;
+	__be64 iqaddr;
+	__be32 iqns_to_fl0congen;
+	__be16 fl0dcaen_to_fl0cidxfthresh;
+	__be16 fl0size;
+	__be64 fl0addr;
+	__be32 fl1cngchmap_to_fl1congen;
+	__be16 fl1dcaen_to_fl1cidxfthresh;
+	__be16 fl1size;
+	__be64 fl1addr;
+};
+
+#define FW_IQ_CMD_PFN_S		8
+#define FW_IQ_CMD_PFN_V(x)	((x) << FW_IQ_CMD_PFN_S)
+
+#define FW_IQ_CMD_VFN_S		0
+#define FW_IQ_CMD_VFN_V(x)	((x) << FW_IQ_CMD_VFN_S)
+
+#define FW_IQ_CMD_ALLOC_S	31
+#define FW_IQ_CMD_ALLOC_V(x)	((x) << FW_IQ_CMD_ALLOC_S)
+#define FW_IQ_CMD_ALLOC_F	FW_IQ_CMD_ALLOC_V(1U)
+
+#define FW_IQ_CMD_FREE_S	30
+#define FW_IQ_CMD_FREE_V(x)	((x) << FW_IQ_CMD_FREE_S)
+#define FW_IQ_CMD_FREE_F	FW_IQ_CMD_FREE_V(1U)
+
+#define FW_IQ_CMD_MODIFY_S	29
+#define FW_IQ_CMD_MODIFY_V(x)	((x) << FW_IQ_CMD_MODIFY_S)
+#define FW_IQ_CMD_MODIFY_F	FW_IQ_CMD_MODIFY_V(1U)
+
+#define FW_IQ_CMD_IQSTART_S	28
+#define FW_IQ_CMD_IQSTART_V(x)	((x) << FW_IQ_CMD_IQSTART_S)
+#define FW_IQ_CMD_IQSTART_F	FW_IQ_CMD_IQSTART_V(1U)
+
+#define FW_IQ_CMD_IQSTOP_S	27
+#define FW_IQ_CMD_IQSTOP_V(x)	((x) << FW_IQ_CMD_IQSTOP_S)
+#define FW_IQ_CMD_IQSTOP_F	FW_IQ_CMD_IQSTOP_V(1U)
+
+#define FW_IQ_CMD_TYPE_S	29
+#define FW_IQ_CMD_TYPE_V(x)	((x) << FW_IQ_CMD_TYPE_S)
+
+#define FW_IQ_CMD_IQASYNCH_S	28
+#define FW_IQ_CMD_IQASYNCH_V(x)	((x) << FW_IQ_CMD_IQASYNCH_S)
+
+#define FW_IQ_CMD_VIID_S	16
+#define FW_IQ_CMD_VIID_V(x)	((x) << FW_IQ_CMD_VIID_S)
+
+#define FW_IQ_CMD_IQANDST_S	15
+#define FW_IQ_CMD_IQANDST_V(x)	((x) << FW_IQ_CMD_IQANDST_S)
+
+#define FW_IQ_CMD_IQANUS_S	14
+#define FW_IQ_CMD_IQANUS_V(x)	((x) << FW_IQ_CMD_IQANUS_S)
+
+#define FW_IQ_CMD_IQANUD_S	12
+#define FW_IQ_CMD_IQANUD_V(x)	((x) << FW_IQ_CMD_IQANUD_S)
+
+#define FW_IQ_CMD_IQANDSTINDEX_S	0
+#define FW_IQ_CMD_IQANDSTINDEX_V(x)	((x) << FW_IQ_CMD_IQANDSTINDEX_S)
+
+#define FW_IQ_CMD_IQDROPRSS_S		15
+#define FW_IQ_CMD_IQDROPRSS_V(x)	((x) << FW_IQ_CMD_IQDROPRSS_S)
+#define FW_IQ_CMD_IQDROPRSS_F	FW_IQ_CMD_IQDROPRSS_V(1U)
+
+#define FW_IQ_CMD_IQGTSMODE_S		14
+#define FW_IQ_CMD_IQGTSMODE_V(x)	((x) << FW_IQ_CMD_IQGTSMODE_S)
+#define FW_IQ_CMD_IQGTSMODE_F		FW_IQ_CMD_IQGTSMODE_V(1U)
+
+#define FW_IQ_CMD_IQPCIECH_S	12
+#define FW_IQ_CMD_IQPCIECH_V(x)	((x) << FW_IQ_CMD_IQPCIECH_S)
+
+#define FW_IQ_CMD_IQDCAEN_S	11
+#define FW_IQ_CMD_IQDCAEN_V(x)	((x) << FW_IQ_CMD_IQDCAEN_S)
+
+#define FW_IQ_CMD_IQDCACPU_S	6
+#define FW_IQ_CMD_IQDCACPU_V(x)	((x) << FW_IQ_CMD_IQDCACPU_S)
+
+#define FW_IQ_CMD_IQINTCNTTHRESH_S	4
+#define FW_IQ_CMD_IQINTCNTTHRESH_V(x)	((x) << FW_IQ_CMD_IQINTCNTTHRESH_S)
+
+#define FW_IQ_CMD_IQO_S		3
+#define FW_IQ_CMD_IQO_V(x)	((x) << FW_IQ_CMD_IQO_S)
+#define FW_IQ_CMD_IQO_F		FW_IQ_CMD_IQO_V(1U)
+
+#define FW_IQ_CMD_IQCPRIO_S	2
+#define FW_IQ_CMD_IQCPRIO_V(x)	((x) << FW_IQ_CMD_IQCPRIO_S)
+
+#define FW_IQ_CMD_IQESIZE_S	0
+#define FW_IQ_CMD_IQESIZE_V(x)	((x) << FW_IQ_CMD_IQESIZE_S)
+
+#define FW_IQ_CMD_IQNS_S	31
+#define FW_IQ_CMD_IQNS_V(x)	((x) << FW_IQ_CMD_IQNS_S)
+
+#define FW_IQ_CMD_IQRO_S	30
+#define FW_IQ_CMD_IQRO_V(x)	((x) << FW_IQ_CMD_IQRO_S)
+
+#define FW_IQ_CMD_IQFLINTIQHSEN_S	28
+#define FW_IQ_CMD_IQFLINTIQHSEN_V(x)	((x) << FW_IQ_CMD_IQFLINTIQHSEN_S)
+
+#define FW_IQ_CMD_IQFLINTCONGEN_S	27
+#define FW_IQ_CMD_IQFLINTCONGEN_V(x)	((x) << FW_IQ_CMD_IQFLINTCONGEN_S)
+#define FW_IQ_CMD_IQFLINTCONGEN_F	FW_IQ_CMD_IQFLINTCONGEN_V(1U)
+
+#define FW_IQ_CMD_IQFLINTISCSIC_S	26
+#define FW_IQ_CMD_IQFLINTISCSIC_V(x)	((x) << FW_IQ_CMD_IQFLINTISCSIC_S)
+
+#define FW_IQ_CMD_IQTYPE_S		24
+#define FW_IQ_CMD_IQTYPE_M		0x3
+#define FW_IQ_CMD_IQTYPE_V(x)		((x) << FW_IQ_CMD_IQTYPE_S)
+#define FW_IQ_CMD_IQTYPE_G(x)		\
+	(((x) >> FW_IQ_CMD_IQTYPE_S) & FW_IQ_CMD_IQTYPE_M)
+
+#define FW_IQ_CMD_FL0CNGCHMAP_S		20
+#define FW_IQ_CMD_FL0CNGCHMAP_V(x)	((x) << FW_IQ_CMD_FL0CNGCHMAP_S)
+
+#define FW_IQ_CMD_FL0CACHELOCK_S	15
+#define FW_IQ_CMD_FL0CACHELOCK_V(x)	((x) << FW_IQ_CMD_FL0CACHELOCK_S)
+
+#define FW_IQ_CMD_FL0DBP_S	14
+#define FW_IQ_CMD_FL0DBP_V(x)	((x) << FW_IQ_CMD_FL0DBP_S)
+
+#define FW_IQ_CMD_FL0DATANS_S		13
+#define FW_IQ_CMD_FL0DATANS_V(x)	((x) << FW_IQ_CMD_FL0DATANS_S)
+
+#define FW_IQ_CMD_FL0DATARO_S		12
+#define FW_IQ_CMD_FL0DATARO_V(x)	((x) << FW_IQ_CMD_FL0DATARO_S)
+#define FW_IQ_CMD_FL0DATARO_F		FW_IQ_CMD_FL0DATARO_V(1U)
+
+#define FW_IQ_CMD_FL0CONGCIF_S		11
+#define FW_IQ_CMD_FL0CONGCIF_V(x)	((x) << FW_IQ_CMD_FL0CONGCIF_S)
+#define FW_IQ_CMD_FL0CONGCIF_F		FW_IQ_CMD_FL0CONGCIF_V(1U)
+
+#define FW_IQ_CMD_FL0ONCHIP_S		10
+#define FW_IQ_CMD_FL0ONCHIP_V(x)	((x) << FW_IQ_CMD_FL0ONCHIP_S)
+
+#define FW_IQ_CMD_FL0STATUSPGNS_S	9
+#define FW_IQ_CMD_FL0STATUSPGNS_V(x)	((x) << FW_IQ_CMD_FL0STATUSPGNS_S)
+
+#define FW_IQ_CMD_FL0STATUSPGRO_S	8
+#define FW_IQ_CMD_FL0STATUSPGRO_V(x)	((x) << FW_IQ_CMD_FL0STATUSPGRO_S)
+
+#define FW_IQ_CMD_FL0FETCHNS_S		7
+#define FW_IQ_CMD_FL0FETCHNS_V(x)	((x) << FW_IQ_CMD_FL0FETCHNS_S)
+
+#define FW_IQ_CMD_FL0FETCHRO_S		6
+#define FW_IQ_CMD_FL0FETCHRO_V(x)	((x) << FW_IQ_CMD_FL0FETCHRO_S)
+#define FW_IQ_CMD_FL0FETCHRO_F		FW_IQ_CMD_FL0FETCHRO_V(1U)
+
+#define FW_IQ_CMD_FL0HOSTFCMODE_S	4
+#define FW_IQ_CMD_FL0HOSTFCMODE_V(x)	((x) << FW_IQ_CMD_FL0HOSTFCMODE_S)
+
+#define FW_IQ_CMD_FL0CPRIO_S	3
+#define FW_IQ_CMD_FL0CPRIO_V(x)	((x) << FW_IQ_CMD_FL0CPRIO_S)
+
+#define FW_IQ_CMD_FL0PADEN_S	2
+#define FW_IQ_CMD_FL0PADEN_V(x)	((x) << FW_IQ_CMD_FL0PADEN_S)
+#define FW_IQ_CMD_FL0PADEN_F	FW_IQ_CMD_FL0PADEN_V(1U)
+
+#define FW_IQ_CMD_FL0PACKEN_S		1
+#define FW_IQ_CMD_FL0PACKEN_V(x)	((x) << FW_IQ_CMD_FL0PACKEN_S)
+#define FW_IQ_CMD_FL0PACKEN_F		FW_IQ_CMD_FL0PACKEN_V(1U)
+
+#define FW_IQ_CMD_FL0CONGEN_S		0
+#define FW_IQ_CMD_FL0CONGEN_V(x)	((x) << FW_IQ_CMD_FL0CONGEN_S)
+#define FW_IQ_CMD_FL0CONGEN_F		FW_IQ_CMD_FL0CONGEN_V(1U)
+
+#define FW_IQ_CMD_FL0DCAEN_S	15
+#define FW_IQ_CMD_FL0DCAEN_V(x)	((x) << FW_IQ_CMD_FL0DCAEN_S)
+
+#define FW_IQ_CMD_FL0DCACPU_S		10
+#define FW_IQ_CMD_FL0DCACPU_V(x)	((x) << FW_IQ_CMD_FL0DCACPU_S)
+
+#define FW_IQ_CMD_FL0FBMIN_S	7
+#define FW_IQ_CMD_FL0FBMIN_V(x)	((x) << FW_IQ_CMD_FL0FBMIN_S)
+
+#define FW_IQ_CMD_FL0FBMAX_S	4
+#define FW_IQ_CMD_FL0FBMAX_V(x)	((x) << FW_IQ_CMD_FL0FBMAX_S)
+
+#define FW_IQ_CMD_FL0CIDXFTHRESHO_S	3
+#define FW_IQ_CMD_FL0CIDXFTHRESHO_V(x)	((x) << FW_IQ_CMD_FL0CIDXFTHRESHO_S)
+#define FW_IQ_CMD_FL0CIDXFTHRESHO_F	FW_IQ_CMD_FL0CIDXFTHRESHO_V(1U)
+
+#define FW_IQ_CMD_FL0CIDXFTHRESH_S	0
+#define FW_IQ_CMD_FL0CIDXFTHRESH_V(x)	((x) << FW_IQ_CMD_FL0CIDXFTHRESH_S)
+
+#define FW_IQ_CMD_FL1CNGCHMAP_S		20
+#define FW_IQ_CMD_FL1CNGCHMAP_V(x)	((x) << FW_IQ_CMD_FL1CNGCHMAP_S)
+
+#define FW_IQ_CMD_FL1CACHELOCK_S	15
+#define FW_IQ_CMD_FL1CACHELOCK_V(x)	((x) << FW_IQ_CMD_FL1CACHELOCK_S)
+
+#define FW_IQ_CMD_FL1DBP_S	14
+#define FW_IQ_CMD_FL1DBP_V(x)	((x) << FW_IQ_CMD_FL1DBP_S)
+
+#define FW_IQ_CMD_FL1DATANS_S		13
+#define FW_IQ_CMD_FL1DATANS_V(x)	((x) << FW_IQ_CMD_FL1DATANS_S)
+
+#define FW_IQ_CMD_FL1DATARO_S		12
+#define FW_IQ_CMD_FL1DATARO_V(x)	((x) << FW_IQ_CMD_FL1DATARO_S)
+
+#define FW_IQ_CMD_FL1CONGCIF_S		11
+#define FW_IQ_CMD_FL1CONGCIF_V(x)	((x) << FW_IQ_CMD_FL1CONGCIF_S)
+
+#define FW_IQ_CMD_FL1ONCHIP_S		10
+#define FW_IQ_CMD_FL1ONCHIP_V(x)	((x) << FW_IQ_CMD_FL1ONCHIP_S)
+
+#define FW_IQ_CMD_FL1STATUSPGNS_S	9
+#define FW_IQ_CMD_FL1STATUSPGNS_V(x)	((x) << FW_IQ_CMD_FL1STATUSPGNS_S)
+
+#define FW_IQ_CMD_FL1STATUSPGRO_S	8
+#define FW_IQ_CMD_FL1STATUSPGRO_V(x)	((x) << FW_IQ_CMD_FL1STATUSPGRO_S)
+
+#define FW_IQ_CMD_FL1FETCHNS_S		7
+#define FW_IQ_CMD_FL1FETCHNS_V(x)	((x) << FW_IQ_CMD_FL1FETCHNS_S)
+
+#define FW_IQ_CMD_FL1FETCHRO_S		6
+#define FW_IQ_CMD_FL1FETCHRO_V(x)	((x) << FW_IQ_CMD_FL1FETCHRO_S)
+
+#define FW_IQ_CMD_FL1HOSTFCMODE_S	4
+#define FW_IQ_CMD_FL1HOSTFCMODE_V(x)	((x) << FW_IQ_CMD_FL1HOSTFCMODE_S)
+
+#define FW_IQ_CMD_FL1CPRIO_S	3
+#define FW_IQ_CMD_FL1CPRIO_V(x)	((x) << FW_IQ_CMD_FL1CPRIO_S)
+
+#define FW_IQ_CMD_FL1PADEN_S	2
+#define FW_IQ_CMD_FL1PADEN_V(x)	((x) << FW_IQ_CMD_FL1PADEN_S)
+#define FW_IQ_CMD_FL1PADEN_F	FW_IQ_CMD_FL1PADEN_V(1U)
+
+#define FW_IQ_CMD_FL1PACKEN_S		1
+#define FW_IQ_CMD_FL1PACKEN_V(x)	((x) << FW_IQ_CMD_FL1PACKEN_S)
+#define FW_IQ_CMD_FL1PACKEN_F	FW_IQ_CMD_FL1PACKEN_V(1U)
+
+#define FW_IQ_CMD_FL1CONGEN_S		0
+#define FW_IQ_CMD_FL1CONGEN_V(x)	((x) << FW_IQ_CMD_FL1CONGEN_S)
+#define FW_IQ_CMD_FL1CONGEN_F	FW_IQ_CMD_FL1CONGEN_V(1U)
+
+#define FW_IQ_CMD_FL1DCAEN_S	15
+#define FW_IQ_CMD_FL1DCAEN_V(x)	((x) << FW_IQ_CMD_FL1DCAEN_S)
+
+#define FW_IQ_CMD_FL1DCACPU_S		10
+#define FW_IQ_CMD_FL1DCACPU_V(x)	((x) << FW_IQ_CMD_FL1DCACPU_S)
+
+#define FW_IQ_CMD_FL1FBMIN_S	7
+#define FW_IQ_CMD_FL1FBMIN_V(x)	((x) << FW_IQ_CMD_FL1FBMIN_S)
+
+#define FW_IQ_CMD_FL1FBMAX_S	4
+#define FW_IQ_CMD_FL1FBMAX_V(x)	((x) << FW_IQ_CMD_FL1FBMAX_S)
+
+#define FW_IQ_CMD_FL1CIDXFTHRESHO_S	3
+#define FW_IQ_CMD_FL1CIDXFTHRESHO_V(x)	((x) << FW_IQ_CMD_FL1CIDXFTHRESHO_S)
+#define FW_IQ_CMD_FL1CIDXFTHRESHO_F	FW_IQ_CMD_FL1CIDXFTHRESHO_V(1U)
+
+#define FW_IQ_CMD_FL1CIDXFTHRESH_S	0
+#define FW_IQ_CMD_FL1CIDXFTHRESH_V(x)	((x) << FW_IQ_CMD_FL1CIDXFTHRESH_S)
+
+struct fw_eq_eth_cmd {
+	__be32 op_to_vfn;
+	__be32 alloc_to_len16;
+	__be32 eqid_pkd;
+	__be32 physeqid_pkd;
+	__be32 fetchszm_to_iqid;
+	__be32 dcaen_to_eqsize;
+	__be64 eqaddr;
+	__be32 autoequiqe_to_viid;
+	__be32 timeren_timerix;
+	__be64 r9;
+};
+
+#define FW_EQ_ETH_CMD_PFN_S	8
+#define FW_EQ_ETH_CMD_PFN_V(x)	((x) << FW_EQ_ETH_CMD_PFN_S)
+
+#define FW_EQ_ETH_CMD_VFN_S	0
+#define FW_EQ_ETH_CMD_VFN_V(x)	((x) << FW_EQ_ETH_CMD_VFN_S)
+
+#define FW_EQ_ETH_CMD_ALLOC_S		31
+#define FW_EQ_ETH_CMD_ALLOC_V(x)	((x) << FW_EQ_ETH_CMD_ALLOC_S)
+#define FW_EQ_ETH_CMD_ALLOC_F	FW_EQ_ETH_CMD_ALLOC_V(1U)
+
+#define FW_EQ_ETH_CMD_FREE_S	30
+#define FW_EQ_ETH_CMD_FREE_V(x)	((x) << FW_EQ_ETH_CMD_FREE_S)
+#define FW_EQ_ETH_CMD_FREE_F	FW_EQ_ETH_CMD_FREE_V(1U)
+
+#define FW_EQ_ETH_CMD_MODIFY_S		29
+#define FW_EQ_ETH_CMD_MODIFY_V(x)	((x) << FW_EQ_ETH_CMD_MODIFY_S)
+#define FW_EQ_ETH_CMD_MODIFY_F	FW_EQ_ETH_CMD_MODIFY_V(1U)
+
+#define FW_EQ_ETH_CMD_EQSTART_S		28
+#define FW_EQ_ETH_CMD_EQSTART_V(x)	((x) << FW_EQ_ETH_CMD_EQSTART_S)
+#define FW_EQ_ETH_CMD_EQSTART_F	FW_EQ_ETH_CMD_EQSTART_V(1U)
+
+#define FW_EQ_ETH_CMD_EQSTOP_S		27
+#define FW_EQ_ETH_CMD_EQSTOP_V(x)	((x) << FW_EQ_ETH_CMD_EQSTOP_S)
+#define FW_EQ_ETH_CMD_EQSTOP_F	FW_EQ_ETH_CMD_EQSTOP_V(1U)
+
+#define FW_EQ_ETH_CMD_EQID_S	0
+#define FW_EQ_ETH_CMD_EQID_M	0xfffff
+#define FW_EQ_ETH_CMD_EQID_V(x)	((x) << FW_EQ_ETH_CMD_EQID_S)
+#define FW_EQ_ETH_CMD_EQID_G(x)	\
+	(((x) >> FW_EQ_ETH_CMD_EQID_S) & FW_EQ_ETH_CMD_EQID_M)
+
+#define FW_EQ_ETH_CMD_PHYSEQID_S	0
+#define FW_EQ_ETH_CMD_PHYSEQID_M	0xfffff
+#define FW_EQ_ETH_CMD_PHYSEQID_V(x)	((x) << FW_EQ_ETH_CMD_PHYSEQID_S)
+#define FW_EQ_ETH_CMD_PHYSEQID_G(x)	\
+	(((x) >> FW_EQ_ETH_CMD_PHYSEQID_S) & FW_EQ_ETH_CMD_PHYSEQID_M)
+
+#define FW_EQ_ETH_CMD_FETCHSZM_S	26
+#define FW_EQ_ETH_CMD_FETCHSZM_V(x)	((x) << FW_EQ_ETH_CMD_FETCHSZM_S)
+#define FW_EQ_ETH_CMD_FETCHSZM_F	FW_EQ_ETH_CMD_FETCHSZM_V(1U)
+
+#define FW_EQ_ETH_CMD_STATUSPGNS_S	25
+#define FW_EQ_ETH_CMD_STATUSPGNS_V(x)	((x) << FW_EQ_ETH_CMD_STATUSPGNS_S)
+
+#define FW_EQ_ETH_CMD_STATUSPGRO_S	24
+#define FW_EQ_ETH_CMD_STATUSPGRO_V(x)	((x) << FW_EQ_ETH_CMD_STATUSPGRO_S)
+
+#define FW_EQ_ETH_CMD_FETCHNS_S		23
+#define FW_EQ_ETH_CMD_FETCHNS_V(x)	((x) << FW_EQ_ETH_CMD_FETCHNS_S)
+
+#define FW_EQ_ETH_CMD_FETCHRO_S		22
+#define FW_EQ_ETH_CMD_FETCHRO_V(x)	((x) << FW_EQ_ETH_CMD_FETCHRO_S)
+#define FW_EQ_ETH_CMD_FETCHRO_F		FW_EQ_ETH_CMD_FETCHRO_V(1U)
+
+#define FW_EQ_ETH_CMD_HOSTFCMODE_S	20
+#define FW_EQ_ETH_CMD_HOSTFCMODE_V(x)	((x) << FW_EQ_ETH_CMD_HOSTFCMODE_S)
+
+#define FW_EQ_ETH_CMD_CPRIO_S		19
+#define FW_EQ_ETH_CMD_CPRIO_V(x)	((x) << FW_EQ_ETH_CMD_CPRIO_S)
+
+#define FW_EQ_ETH_CMD_ONCHIP_S		18
+#define FW_EQ_ETH_CMD_ONCHIP_V(x)	((x) << FW_EQ_ETH_CMD_ONCHIP_S)
+
+#define FW_EQ_ETH_CMD_PCIECHN_S		16
+#define FW_EQ_ETH_CMD_PCIECHN_V(x)	((x) << FW_EQ_ETH_CMD_PCIECHN_S)
+
+#define FW_EQ_ETH_CMD_IQID_S	0
+#define FW_EQ_ETH_CMD_IQID_V(x)	((x) << FW_EQ_ETH_CMD_IQID_S)
+
+#define FW_EQ_ETH_CMD_DCAEN_S		31
+#define FW_EQ_ETH_CMD_DCAEN_V(x)	((x) << FW_EQ_ETH_CMD_DCAEN_S)
+
+#define FW_EQ_ETH_CMD_DCACPU_S		26
+#define FW_EQ_ETH_CMD_DCACPU_V(x)	((x) << FW_EQ_ETH_CMD_DCACPU_S)
+
+#define FW_EQ_ETH_CMD_FBMIN_S		23
+#define FW_EQ_ETH_CMD_FBMIN_V(x)	((x) << FW_EQ_ETH_CMD_FBMIN_S)
+
+#define FW_EQ_ETH_CMD_FBMAX_S		20
+#define FW_EQ_ETH_CMD_FBMAX_V(x)	((x) << FW_EQ_ETH_CMD_FBMAX_S)
+
+#define FW_EQ_ETH_CMD_CIDXFTHRESHO_S	19
+#define FW_EQ_ETH_CMD_CIDXFTHRESHO_V(x)	((x) << FW_EQ_ETH_CMD_CIDXFTHRESHO_S)
+
+#define FW_EQ_ETH_CMD_CIDXFTHRESH_S	16
+#define FW_EQ_ETH_CMD_CIDXFTHRESH_V(x)	((x) << FW_EQ_ETH_CMD_CIDXFTHRESH_S)
+
+#define FW_EQ_ETH_CMD_EQSIZE_S		0
+#define FW_EQ_ETH_CMD_EQSIZE_V(x)	((x) << FW_EQ_ETH_CMD_EQSIZE_S)
+
+#define FW_EQ_ETH_CMD_AUTOEQUIQE_S	31
+#define FW_EQ_ETH_CMD_AUTOEQUIQE_V(x)	((x) << FW_EQ_ETH_CMD_AUTOEQUIQE_S)
+#define FW_EQ_ETH_CMD_AUTOEQUIQE_F	FW_EQ_ETH_CMD_AUTOEQUIQE_V(1U)
+
+#define FW_EQ_ETH_CMD_AUTOEQUEQE_S	30
+#define FW_EQ_ETH_CMD_AUTOEQUEQE_V(x)	((x) << FW_EQ_ETH_CMD_AUTOEQUEQE_S)
+#define FW_EQ_ETH_CMD_AUTOEQUEQE_F	FW_EQ_ETH_CMD_AUTOEQUEQE_V(1U)
+
+#define FW_EQ_ETH_CMD_VIID_S	16
+#define FW_EQ_ETH_CMD_VIID_V(x)	((x) << FW_EQ_ETH_CMD_VIID_S)
+
+#define FW_EQ_ETH_CMD_TIMEREN_S		3
+#define FW_EQ_ETH_CMD_TIMEREN_M		0x1
+#define FW_EQ_ETH_CMD_TIMEREN_V(x)	((x) << FW_EQ_ETH_CMD_TIMEREN_S)
+#define FW_EQ_ETH_CMD_TIMEREN_G(x)	\
+    (((x) >> FW_EQ_ETH_CMD_TIMEREN_S) & FW_EQ_ETH_CMD_TIMEREN_M)
+#define FW_EQ_ETH_CMD_TIMEREN_F	FW_EQ_ETH_CMD_TIMEREN_V(1U)
+
+#define FW_EQ_ETH_CMD_TIMERIX_S		0
+#define FW_EQ_ETH_CMD_TIMERIX_M		0x7
+#define FW_EQ_ETH_CMD_TIMERIX_V(x)	((x) << FW_EQ_ETH_CMD_TIMERIX_S)
+#define FW_EQ_ETH_CMD_TIMERIX_G(x)	\
+    (((x) >> FW_EQ_ETH_CMD_TIMERIX_S) & FW_EQ_ETH_CMD_TIMERIX_M)
+
+struct fw_eq_ctrl_cmd {
+	__be32 op_to_vfn;
+	__be32 alloc_to_len16;
+	__be32 cmpliqid_eqid;
+	__be32 physeqid_pkd;
+	__be32 fetchszm_to_iqid;
+	__be32 dcaen_to_eqsize;
+	__be64 eqaddr;
+};
+
+#define FW_EQ_CTRL_CMD_PFN_S	8
+#define FW_EQ_CTRL_CMD_PFN_V(x)	((x) << FW_EQ_CTRL_CMD_PFN_S)
+
+#define FW_EQ_CTRL_CMD_VFN_S	0
+#define FW_EQ_CTRL_CMD_VFN_V(x)	((x) << FW_EQ_CTRL_CMD_VFN_S)
+
+#define FW_EQ_CTRL_CMD_ALLOC_S		31
+#define FW_EQ_CTRL_CMD_ALLOC_V(x)	((x) << FW_EQ_CTRL_CMD_ALLOC_S)
+#define FW_EQ_CTRL_CMD_ALLOC_F		FW_EQ_CTRL_CMD_ALLOC_V(1U)
+
+#define FW_EQ_CTRL_CMD_FREE_S		30
+#define FW_EQ_CTRL_CMD_FREE_V(x)	((x) << FW_EQ_CTRL_CMD_FREE_S)
+#define FW_EQ_CTRL_CMD_FREE_F		FW_EQ_CTRL_CMD_FREE_V(1U)
+
+#define FW_EQ_CTRL_CMD_MODIFY_S		29
+#define FW_EQ_CTRL_CMD_MODIFY_V(x)	((x) << FW_EQ_CTRL_CMD_MODIFY_S)
+#define FW_EQ_CTRL_CMD_MODIFY_F		FW_EQ_CTRL_CMD_MODIFY_V(1U)
+
+#define FW_EQ_CTRL_CMD_EQSTART_S	28
+#define FW_EQ_CTRL_CMD_EQSTART_V(x)	((x) << FW_EQ_CTRL_CMD_EQSTART_S)
+#define FW_EQ_CTRL_CMD_EQSTART_F	FW_EQ_CTRL_CMD_EQSTART_V(1U)
+
+#define FW_EQ_CTRL_CMD_EQSTOP_S		27
+#define FW_EQ_CTRL_CMD_EQSTOP_V(x)	((x) << FW_EQ_CTRL_CMD_EQSTOP_S)
+#define FW_EQ_CTRL_CMD_EQSTOP_F		FW_EQ_CTRL_CMD_EQSTOP_V(1U)
+
+#define FW_EQ_CTRL_CMD_CMPLIQID_S	20
+#define FW_EQ_CTRL_CMD_CMPLIQID_V(x)	((x) << FW_EQ_CTRL_CMD_CMPLIQID_S)
+
+#define FW_EQ_CTRL_CMD_EQID_S		0
+#define FW_EQ_CTRL_CMD_EQID_M		0xfffff
+#define FW_EQ_CTRL_CMD_EQID_V(x)	((x) << FW_EQ_CTRL_CMD_EQID_S)
+#define FW_EQ_CTRL_CMD_EQID_G(x)	\
+	(((x) >> FW_EQ_CTRL_CMD_EQID_S) & FW_EQ_CTRL_CMD_EQID_M)
+
+#define FW_EQ_CTRL_CMD_PHYSEQID_S	0
+#define FW_EQ_CTRL_CMD_PHYSEQID_M	0xfffff
+#define FW_EQ_CTRL_CMD_PHYSEQID_G(x)	\
+	(((x) >> FW_EQ_CTRL_CMD_PHYSEQID_S) & FW_EQ_CTRL_CMD_PHYSEQID_M)
+
+#define FW_EQ_CTRL_CMD_FETCHSZM_S	26
+#define FW_EQ_CTRL_CMD_FETCHSZM_V(x)	((x) << FW_EQ_CTRL_CMD_FETCHSZM_S)
+#define FW_EQ_CTRL_CMD_FETCHSZM_F	FW_EQ_CTRL_CMD_FETCHSZM_V(1U)
+
+#define FW_EQ_CTRL_CMD_STATUSPGNS_S	25
+#define FW_EQ_CTRL_CMD_STATUSPGNS_V(x)	((x) << FW_EQ_CTRL_CMD_STATUSPGNS_S)
+#define FW_EQ_CTRL_CMD_STATUSPGNS_F	FW_EQ_CTRL_CMD_STATUSPGNS_V(1U)
+
+#define FW_EQ_CTRL_CMD_STATUSPGRO_S	24
+#define FW_EQ_CTRL_CMD_STATUSPGRO_V(x)	((x) << FW_EQ_CTRL_CMD_STATUSPGRO_S)
+#define FW_EQ_CTRL_CMD_STATUSPGRO_F	FW_EQ_CTRL_CMD_STATUSPGRO_V(1U)
+
+#define FW_EQ_CTRL_CMD_FETCHNS_S	23
+#define FW_EQ_CTRL_CMD_FETCHNS_V(x)	((x) << FW_EQ_CTRL_CMD_FETCHNS_S)
+#define FW_EQ_CTRL_CMD_FETCHNS_F	FW_EQ_CTRL_CMD_FETCHNS_V(1U)
+
+#define FW_EQ_CTRL_CMD_FETCHRO_S	22
+#define FW_EQ_CTRL_CMD_FETCHRO_V(x)	((x) << FW_EQ_CTRL_CMD_FETCHRO_S)
+#define FW_EQ_CTRL_CMD_FETCHRO_F	FW_EQ_CTRL_CMD_FETCHRO_V(1U)
+
+#define FW_EQ_CTRL_CMD_HOSTFCMODE_S	20
+#define FW_EQ_CTRL_CMD_HOSTFCMODE_V(x)	((x) << FW_EQ_CTRL_CMD_HOSTFCMODE_S)
+
+#define FW_EQ_CTRL_CMD_CPRIO_S		19
+#define FW_EQ_CTRL_CMD_CPRIO_V(x)	((x) << FW_EQ_CTRL_CMD_CPRIO_S)
+
+#define FW_EQ_CTRL_CMD_ONCHIP_S		18
+#define FW_EQ_CTRL_CMD_ONCHIP_V(x)	((x) << FW_EQ_CTRL_CMD_ONCHIP_S)
+
+#define FW_EQ_CTRL_CMD_PCIECHN_S	16
+#define FW_EQ_CTRL_CMD_PCIECHN_V(x)	((x) << FW_EQ_CTRL_CMD_PCIECHN_S)
+
+#define FW_EQ_CTRL_CMD_IQID_S		0
+#define FW_EQ_CTRL_CMD_IQID_V(x)	((x) << FW_EQ_CTRL_CMD_IQID_S)
+
+#define FW_EQ_CTRL_CMD_DCAEN_S		31
+#define FW_EQ_CTRL_CMD_DCAEN_V(x)	((x) << FW_EQ_CTRL_CMD_DCAEN_S)
+
+#define FW_EQ_CTRL_CMD_DCACPU_S		26
+#define FW_EQ_CTRL_CMD_DCACPU_V(x)	((x) << FW_EQ_CTRL_CMD_DCACPU_S)
+
+#define FW_EQ_CTRL_CMD_FBMIN_S		23
+#define FW_EQ_CTRL_CMD_FBMIN_V(x)	((x) << FW_EQ_CTRL_CMD_FBMIN_S)
+
+#define FW_EQ_CTRL_CMD_FBMAX_S		20
+#define FW_EQ_CTRL_CMD_FBMAX_V(x)	((x) << FW_EQ_CTRL_CMD_FBMAX_S)
+
+#define FW_EQ_CTRL_CMD_CIDXFTHRESHO_S		19
+#define FW_EQ_CTRL_CMD_CIDXFTHRESHO_V(x)	\
+	((x) << FW_EQ_CTRL_CMD_CIDXFTHRESHO_S)
+
+#define FW_EQ_CTRL_CMD_CIDXFTHRESH_S	16
+#define FW_EQ_CTRL_CMD_CIDXFTHRESH_V(x)	((x) << FW_EQ_CTRL_CMD_CIDXFTHRESH_S)
+
+#define FW_EQ_CTRL_CMD_EQSIZE_S		0
+#define FW_EQ_CTRL_CMD_EQSIZE_V(x)	((x) << FW_EQ_CTRL_CMD_EQSIZE_S)
+
+struct fw_eq_ofld_cmd {
+	__be32 op_to_vfn;
+	__be32 alloc_to_len16;
+	__be32 eqid_pkd;
+	__be32 physeqid_pkd;
+	__be32 fetchszm_to_iqid;
+	__be32 dcaen_to_eqsize;
+	__be64 eqaddr;
+};
+
+#define FW_EQ_OFLD_CMD_PFN_S	8
+#define FW_EQ_OFLD_CMD_PFN_V(x)	((x) << FW_EQ_OFLD_CMD_PFN_S)
+
+#define FW_EQ_OFLD_CMD_VFN_S	0
+#define FW_EQ_OFLD_CMD_VFN_V(x)	((x) << FW_EQ_OFLD_CMD_VFN_S)
+
+#define FW_EQ_OFLD_CMD_ALLOC_S		31
+#define FW_EQ_OFLD_CMD_ALLOC_V(x)	((x) << FW_EQ_OFLD_CMD_ALLOC_S)
+#define FW_EQ_OFLD_CMD_ALLOC_F		FW_EQ_OFLD_CMD_ALLOC_V(1U)
+
+#define FW_EQ_OFLD_CMD_FREE_S		30
+#define FW_EQ_OFLD_CMD_FREE_V(x)	((x) << FW_EQ_OFLD_CMD_FREE_S)
+#define FW_EQ_OFLD_CMD_FREE_F		FW_EQ_OFLD_CMD_FREE_V(1U)
+
+#define FW_EQ_OFLD_CMD_MODIFY_S		29
+#define FW_EQ_OFLD_CMD_MODIFY_V(x)	((x) << FW_EQ_OFLD_CMD_MODIFY_S)
+#define FW_EQ_OFLD_CMD_MODIFY_F		FW_EQ_OFLD_CMD_MODIFY_V(1U)
+
+#define FW_EQ_OFLD_CMD_EQSTART_S	28
+#define FW_EQ_OFLD_CMD_EQSTART_V(x)	((x) << FW_EQ_OFLD_CMD_EQSTART_S)
+#define FW_EQ_OFLD_CMD_EQSTART_F	FW_EQ_OFLD_CMD_EQSTART_V(1U)
+
+#define FW_EQ_OFLD_CMD_EQSTOP_S		27
+#define FW_EQ_OFLD_CMD_EQSTOP_V(x)	((x) << FW_EQ_OFLD_CMD_EQSTOP_S)
+#define FW_EQ_OFLD_CMD_EQSTOP_F		FW_EQ_OFLD_CMD_EQSTOP_V(1U)
+
+#define FW_EQ_OFLD_CMD_EQID_S		0
+#define FW_EQ_OFLD_CMD_EQID_M		0xfffff
+#define FW_EQ_OFLD_CMD_EQID_V(x)	((x) << FW_EQ_OFLD_CMD_EQID_S)
+#define FW_EQ_OFLD_CMD_EQID_G(x)	\
+	(((x) >> FW_EQ_OFLD_CMD_EQID_S) & FW_EQ_OFLD_CMD_EQID_M)
+
+#define FW_EQ_OFLD_CMD_PHYSEQID_S	0
+#define FW_EQ_OFLD_CMD_PHYSEQID_M	0xfffff
+#define FW_EQ_OFLD_CMD_PHYSEQID_G(x)	\
+	(((x) >> FW_EQ_OFLD_CMD_PHYSEQID_S) & FW_EQ_OFLD_CMD_PHYSEQID_M)
+
+#define FW_EQ_OFLD_CMD_FETCHSZM_S	26
+#define FW_EQ_OFLD_CMD_FETCHSZM_V(x)	((x) << FW_EQ_OFLD_CMD_FETCHSZM_S)
+
+#define FW_EQ_OFLD_CMD_STATUSPGNS_S	25
+#define FW_EQ_OFLD_CMD_STATUSPGNS_V(x)	((x) << FW_EQ_OFLD_CMD_STATUSPGNS_S)
+
+#define FW_EQ_OFLD_CMD_STATUSPGRO_S	24
+#define FW_EQ_OFLD_CMD_STATUSPGRO_V(x)	((x) << FW_EQ_OFLD_CMD_STATUSPGRO_S)
+
+#define FW_EQ_OFLD_CMD_FETCHNS_S	23
+#define FW_EQ_OFLD_CMD_FETCHNS_V(x)	((x) << FW_EQ_OFLD_CMD_FETCHNS_S)
+
+#define FW_EQ_OFLD_CMD_FETCHRO_S	22
+#define FW_EQ_OFLD_CMD_FETCHRO_V(x)	((x) << FW_EQ_OFLD_CMD_FETCHRO_S)
+#define FW_EQ_OFLD_CMD_FETCHRO_F	FW_EQ_OFLD_CMD_FETCHRO_V(1U)
+
+#define FW_EQ_OFLD_CMD_HOSTFCMODE_S	20
+#define FW_EQ_OFLD_CMD_HOSTFCMODE_V(x)	((x) << FW_EQ_OFLD_CMD_HOSTFCMODE_S)
+
+#define FW_EQ_OFLD_CMD_CPRIO_S		19
+#define FW_EQ_OFLD_CMD_CPRIO_V(x)	((x) << FW_EQ_OFLD_CMD_CPRIO_S)
+
+#define FW_EQ_OFLD_CMD_ONCHIP_S		18
+#define FW_EQ_OFLD_CMD_ONCHIP_V(x)	((x) << FW_EQ_OFLD_CMD_ONCHIP_S)
+
+#define FW_EQ_OFLD_CMD_PCIECHN_S	16
+#define FW_EQ_OFLD_CMD_PCIECHN_V(x)	((x) << FW_EQ_OFLD_CMD_PCIECHN_S)
+
+#define FW_EQ_OFLD_CMD_IQID_S		0
+#define FW_EQ_OFLD_CMD_IQID_V(x)	((x) << FW_EQ_OFLD_CMD_IQID_S)
+
+#define FW_EQ_OFLD_CMD_DCAEN_S		31
+#define FW_EQ_OFLD_CMD_DCAEN_V(x)	((x) << FW_EQ_OFLD_CMD_DCAEN_S)
+
+#define FW_EQ_OFLD_CMD_DCACPU_S		26
+#define FW_EQ_OFLD_CMD_DCACPU_V(x)	((x) << FW_EQ_OFLD_CMD_DCACPU_S)
+
+#define FW_EQ_OFLD_CMD_FBMIN_S		23
+#define FW_EQ_OFLD_CMD_FBMIN_V(x)	((x) << FW_EQ_OFLD_CMD_FBMIN_S)
+
+#define FW_EQ_OFLD_CMD_FBMAX_S		20
+#define FW_EQ_OFLD_CMD_FBMAX_V(x)	((x) << FW_EQ_OFLD_CMD_FBMAX_S)
+
+#define FW_EQ_OFLD_CMD_CIDXFTHRESHO_S		19
+#define FW_EQ_OFLD_CMD_CIDXFTHRESHO_V(x)	\
+	((x) << FW_EQ_OFLD_CMD_CIDXFTHRESHO_S)
+
+#define FW_EQ_OFLD_CMD_CIDXFTHRESH_S	16
+#define FW_EQ_OFLD_CMD_CIDXFTHRESH_V(x)	((x) << FW_EQ_OFLD_CMD_CIDXFTHRESH_S)
+
+#define FW_EQ_OFLD_CMD_EQSIZE_S		0
+#define FW_EQ_OFLD_CMD_EQSIZE_V(x)	((x) << FW_EQ_OFLD_CMD_EQSIZE_S)
+
+/*
+ * Macros for VIID parsing:
+ * VIID - [10:8] PFN, [7] VI Valid, [6:0] VI number
+ */
+
+#define FW_VIID_PFN_S           8
+#define FW_VIID_PFN_M           0x7
+#define FW_VIID_PFN_G(x)        (((x) >> FW_VIID_PFN_S) & FW_VIID_PFN_M)
+
+#define FW_VIID_VIVLD_S		7
+#define FW_VIID_VIVLD_M		0x1
+#define FW_VIID_VIVLD_G(x)	(((x) >> FW_VIID_VIVLD_S) & FW_VIID_VIVLD_M)
+
+#define FW_VIID_VIN_S		0
+#define FW_VIID_VIN_M		0x7F
+#define FW_VIID_VIN_G(x)	(((x) >> FW_VIID_VIN_S) & FW_VIID_VIN_M)
+
+struct fw_vi_cmd {
+	__be32 op_to_vfn;
+	__be32 alloc_to_len16;
+	__be16 type_viid;
+	u8 mac[6];
+	u8 portid_pkd;
+	u8 nmac;
+	u8 nmac0[6];
+	__be16 rsssize_pkd;
+	u8 nmac1[6];
+	__be16 idsiiq_pkd;
+	u8 nmac2[6];
+	__be16 idseiq_pkd;
+	u8 nmac3[6];
+	__be64 r9;
+	__be64 r10;
+};
+
+#define FW_VI_CMD_PFN_S		8
+#define FW_VI_CMD_PFN_V(x)	((x) << FW_VI_CMD_PFN_S)
+
+#define FW_VI_CMD_VFN_S		0
+#define FW_VI_CMD_VFN_V(x)	((x) << FW_VI_CMD_VFN_S)
+
+#define FW_VI_CMD_ALLOC_S	31
+#define FW_VI_CMD_ALLOC_V(x)	((x) << FW_VI_CMD_ALLOC_S)
+#define FW_VI_CMD_ALLOC_F	FW_VI_CMD_ALLOC_V(1U)
+
+#define FW_VI_CMD_FREE_S	30
+#define FW_VI_CMD_FREE_V(x)	((x) << FW_VI_CMD_FREE_S)
+#define FW_VI_CMD_FREE_F	FW_VI_CMD_FREE_V(1U)
+
+#define FW_VI_CMD_VFVLD_S	24
+#define FW_VI_CMD_VFVLD_M	0x1
+#define FW_VI_CMD_VFVLD_V(x)	((x) << FW_VI_CMD_VFVLD_S)
+#define FW_VI_CMD_VFVLD_G(x)	\
+	(((x) >> FW_VI_CMD_VFVLD_S) & FW_VI_CMD_VFVLD_M)
+#define FW_VI_CMD_VFVLD_F	FW_VI_CMD_VFVLD_V(1U)
+
+#define FW_VI_CMD_VIN_S		16
+#define FW_VI_CMD_VIN_M		0xff
+#define FW_VI_CMD_VIN_V(x)	((x) << FW_VI_CMD_VIN_S)
+#define FW_VI_CMD_VIN_G(x)	\
+	(((x) >> FW_VI_CMD_VIN_S) & FW_VI_CMD_VIN_M)
+
+#define FW_VI_CMD_VIID_S	0
+#define FW_VI_CMD_VIID_M	0xfff
+#define FW_VI_CMD_VIID_V(x)	((x) << FW_VI_CMD_VIID_S)
+#define FW_VI_CMD_VIID_G(x)	(((x) >> FW_VI_CMD_VIID_S) & FW_VI_CMD_VIID_M)
+
+#define FW_VI_CMD_PORTID_S	4
+#define FW_VI_CMD_PORTID_M	0xf
+#define FW_VI_CMD_PORTID_V(x)	((x) << FW_VI_CMD_PORTID_S)
+#define FW_VI_CMD_PORTID_G(x)	\
+	(((x) >> FW_VI_CMD_PORTID_S) & FW_VI_CMD_PORTID_M)
+
+#define FW_VI_CMD_RSSSIZE_S	0
+#define FW_VI_CMD_RSSSIZE_M	0x7ff
+#define FW_VI_CMD_RSSSIZE_G(x)	\
+	(((x) >> FW_VI_CMD_RSSSIZE_S) & FW_VI_CMD_RSSSIZE_M)
+
+/* Special VI_MAC command index ids */
+#define FW_VI_MAC_ADD_MAC		0x3FF
+#define FW_VI_MAC_ADD_PERSIST_MAC	0x3FE
+#define FW_VI_MAC_MAC_BASED_FREE	0x3FD
+#define FW_VI_MAC_ID_BASED_FREE		0x3FC
+#define FW_CLS_TCAM_NUM_ENTRIES		336
+
+enum fw_vi_mac_smac {
+	FW_VI_MAC_MPS_TCAM_ENTRY,
+	FW_VI_MAC_MPS_TCAM_ONLY,
+	FW_VI_MAC_SMT_ONLY,
+	FW_VI_MAC_SMT_AND_MPSTCAM
+};
+
+enum fw_vi_mac_result {
+	FW_VI_MAC_R_SUCCESS,
+	FW_VI_MAC_R_F_NONEXISTENT_NOMEM,
+	FW_VI_MAC_R_SMAC_FAIL,
+	FW_VI_MAC_R_F_ACL_CHECK
+};
+
+enum fw_vi_mac_entry_types {
+	FW_VI_MAC_TYPE_EXACTMAC,
+	FW_VI_MAC_TYPE_HASHVEC,
+	FW_VI_MAC_TYPE_RAW,
+	FW_VI_MAC_TYPE_EXACTMAC_VNI,
+};
+
+struct fw_vi_mac_cmd {
+	__be32 op_to_viid;
+	__be32 freemacs_to_len16;
+	union fw_vi_mac {
+		struct fw_vi_mac_exact {
+			__be16 valid_to_idx;
+			u8 macaddr[6];
+		} exact[7];
+		struct fw_vi_mac_hash {
+			__be64 hashvec;
+		} hash;
+		struct fw_vi_mac_raw {
+			__be32 raw_idx_pkd;
+			__be32 data0_pkd;
+			__be32 data1[2];
+			__be64 data0m_pkd;
+			__be32 data1m[2];
+		} raw;
+		struct fw_vi_mac_vni {
+			__be16 valid_to_idx;
+			__u8 macaddr[6];
+			__be16 r7;
+			__u8 macaddr_mask[6];
+			__be32 lookup_type_to_vni;
+			__be32 vni_mask_pkd;
+		} exact_vni[2];
+	} u;
+};
+
+#define FW_VI_MAC_CMD_SMTID_S		12
+#define FW_VI_MAC_CMD_SMTID_M		0xff
+#define FW_VI_MAC_CMD_SMTID_V(x)	((x) << FW_VI_MAC_CMD_SMTID_S)
+#define FW_VI_MAC_CMD_SMTID_G(x)	\
+	(((x) >> FW_VI_MAC_CMD_SMTID_S) & FW_VI_MAC_CMD_SMTID_M)
+
+#define FW_VI_MAC_CMD_VIID_S	0
+#define FW_VI_MAC_CMD_VIID_V(x)	((x) << FW_VI_MAC_CMD_VIID_S)
+
+#define FW_VI_MAC_CMD_FREEMACS_S	31
+#define FW_VI_MAC_CMD_FREEMACS_V(x)	((x) << FW_VI_MAC_CMD_FREEMACS_S)
+
+#define FW_VI_MAC_CMD_ENTRY_TYPE_S      23
+#define FW_VI_MAC_CMD_ENTRY_TYPE_M      0x7
+#define FW_VI_MAC_CMD_ENTRY_TYPE_V(x)   ((x) << FW_VI_MAC_CMD_ENTRY_TYPE_S)
+#define FW_VI_MAC_CMD_ENTRY_TYPE_G(x)	\
+	(((x) >> FW_VI_MAC_CMD_ENTRY_TYPE_S) & FW_VI_MAC_CMD_ENTRY_TYPE_M)
+
+#define FW_VI_MAC_CMD_HASHVECEN_S	23
+#define FW_VI_MAC_CMD_HASHVECEN_V(x)	((x) << FW_VI_MAC_CMD_HASHVECEN_S)
+#define FW_VI_MAC_CMD_HASHVECEN_F	FW_VI_MAC_CMD_HASHVECEN_V(1U)
+
+#define FW_VI_MAC_CMD_HASHUNIEN_S	22
+#define FW_VI_MAC_CMD_HASHUNIEN_V(x)	((x) << FW_VI_MAC_CMD_HASHUNIEN_S)
+
+#define FW_VI_MAC_CMD_VALID_S		15
+#define FW_VI_MAC_CMD_VALID_V(x)	((x) << FW_VI_MAC_CMD_VALID_S)
+#define FW_VI_MAC_CMD_VALID_F	FW_VI_MAC_CMD_VALID_V(1U)
+
+#define FW_VI_MAC_CMD_PRIO_S	12
+#define FW_VI_MAC_CMD_PRIO_V(x)	((x) << FW_VI_MAC_CMD_PRIO_S)
+
+#define FW_VI_MAC_CMD_SMAC_RESULT_S	10
+#define FW_VI_MAC_CMD_SMAC_RESULT_M	0x3
+#define FW_VI_MAC_CMD_SMAC_RESULT_V(x)	((x) << FW_VI_MAC_CMD_SMAC_RESULT_S)
+#define FW_VI_MAC_CMD_SMAC_RESULT_G(x)	\
+	(((x) >> FW_VI_MAC_CMD_SMAC_RESULT_S) & FW_VI_MAC_CMD_SMAC_RESULT_M)
+
+#define FW_VI_MAC_CMD_IDX_S	0
+#define FW_VI_MAC_CMD_IDX_M	0x3ff
+#define FW_VI_MAC_CMD_IDX_V(x)	((x) << FW_VI_MAC_CMD_IDX_S)
+#define FW_VI_MAC_CMD_IDX_G(x)	\
+	(((x) >> FW_VI_MAC_CMD_IDX_S) & FW_VI_MAC_CMD_IDX_M)
+
+#define FW_VI_MAC_CMD_RAW_IDX_S         16
+#define FW_VI_MAC_CMD_RAW_IDX_M         0xffff
+#define FW_VI_MAC_CMD_RAW_IDX_V(x)      ((x) << FW_VI_MAC_CMD_RAW_IDX_S)
+#define FW_VI_MAC_CMD_RAW_IDX_G(x)      \
+	(((x) >> FW_VI_MAC_CMD_RAW_IDX_S) & FW_VI_MAC_CMD_RAW_IDX_M)
+
+#define FW_VI_MAC_CMD_LOOKUP_TYPE_S	31
+#define FW_VI_MAC_CMD_LOOKUP_TYPE_M	0x1
+#define FW_VI_MAC_CMD_LOOKUP_TYPE_V(x)	((x) << FW_VI_MAC_CMD_LOOKUP_TYPE_S)
+#define FW_VI_MAC_CMD_LOOKUP_TYPE_G(x)	\
+	(((x) >> FW_VI_MAC_CMD_LOOKUP_TYPE_S) & FW_VI_MAC_CMD_LOOKUP_TYPE_M)
+#define FW_VI_MAC_CMD_LOOKUP_TYPE_F	FW_VI_MAC_CMD_LOOKUP_TYPE_V(1U)
+
+#define FW_VI_MAC_CMD_DIP_HIT_S		30
+#define FW_VI_MAC_CMD_DIP_HIT_M		0x1
+#define FW_VI_MAC_CMD_DIP_HIT_V(x)	((x) << FW_VI_MAC_CMD_DIP_HIT_S)
+#define FW_VI_MAC_CMD_DIP_HIT_G(x)	\
+	(((x) >> FW_VI_MAC_CMD_DIP_HIT_S) & FW_VI_MAC_CMD_DIP_HIT_M)
+#define FW_VI_MAC_CMD_DIP_HIT_F		FW_VI_MAC_CMD_DIP_HIT_V(1U)
+
+#define FW_VI_MAC_CMD_VNI_S		0
+#define FW_VI_MAC_CMD_VNI_M		0xffffff
+#define FW_VI_MAC_CMD_VNI_V(x)		((x) << FW_VI_MAC_CMD_VNI_S)
+#define FW_VI_MAC_CMD_VNI_G(x)		\
+	(((x) >> FW_VI_MAC_CMD_VNI_S) & FW_VI_MAC_CMD_VNI_M)
+
+#define FW_VI_MAC_CMD_VNI_MASK_S	0
+#define FW_VI_MAC_CMD_VNI_MASK_M	0xffffff
+#define FW_VI_MAC_CMD_VNI_MASK_V(x)	((x) << FW_VI_MAC_CMD_VNI_MASK_S)
+#define FW_VI_MAC_CMD_VNI_MASK_G(x)	\
+	(((x) >> FW_VI_MAC_CMD_VNI_MASK_S) & FW_VI_MAC_CMD_VNI_MASK_M)
+
+#define FW_RXMODE_MTU_NO_CHG	65535
+
+struct fw_vi_rxmode_cmd {
+	__be32 op_to_viid;
+	__be32 retval_len16;
+	__be32 mtu_to_vlanexen;
+	__be32 r4_lo;
+};
+
+#define FW_VI_RXMODE_CMD_VIID_S		0
+#define FW_VI_RXMODE_CMD_VIID_V(x)	((x) << FW_VI_RXMODE_CMD_VIID_S)
+
+#define FW_VI_RXMODE_CMD_MTU_S		16
+#define FW_VI_RXMODE_CMD_MTU_M		0xffff
+#define FW_VI_RXMODE_CMD_MTU_V(x)	((x) << FW_VI_RXMODE_CMD_MTU_S)
+
+#define FW_VI_RXMODE_CMD_PROMISCEN_S	14
+#define FW_VI_RXMODE_CMD_PROMISCEN_M	0x3
+#define FW_VI_RXMODE_CMD_PROMISCEN_V(x)	((x) << FW_VI_RXMODE_CMD_PROMISCEN_S)
+
+#define FW_VI_RXMODE_CMD_ALLMULTIEN_S		12
+#define FW_VI_RXMODE_CMD_ALLMULTIEN_M		0x3
+#define FW_VI_RXMODE_CMD_ALLMULTIEN_V(x)	\
+	((x) << FW_VI_RXMODE_CMD_ALLMULTIEN_S)
+
+#define FW_VI_RXMODE_CMD_BROADCASTEN_S		10
+#define FW_VI_RXMODE_CMD_BROADCASTEN_M		0x3
+#define FW_VI_RXMODE_CMD_BROADCASTEN_V(x)	\
+	((x) << FW_VI_RXMODE_CMD_BROADCASTEN_S)
+
+#define FW_VI_RXMODE_CMD_VLANEXEN_S	8
+#define FW_VI_RXMODE_CMD_VLANEXEN_M	0x3
+#define FW_VI_RXMODE_CMD_VLANEXEN_V(x)	((x) << FW_VI_RXMODE_CMD_VLANEXEN_S)
+
+struct fw_vi_enable_cmd {
+	__be32 op_to_viid;
+	__be32 ien_to_len16;
+	__be16 blinkdur;
+	__be16 r3;
+	__be32 r4;
+};
+
+#define FW_VI_ENABLE_CMD_VIID_S         0
+#define FW_VI_ENABLE_CMD_VIID_V(x)      ((x) << FW_VI_ENABLE_CMD_VIID_S)
+
+#define FW_VI_ENABLE_CMD_IEN_S		31
+#define FW_VI_ENABLE_CMD_IEN_V(x)	((x) << FW_VI_ENABLE_CMD_IEN_S)
+
+#define FW_VI_ENABLE_CMD_EEN_S		30
+#define FW_VI_ENABLE_CMD_EEN_V(x)	((x) << FW_VI_ENABLE_CMD_EEN_S)
+
+#define FW_VI_ENABLE_CMD_LED_S		29
+#define FW_VI_ENABLE_CMD_LED_V(x)	((x) << FW_VI_ENABLE_CMD_LED_S)
+#define FW_VI_ENABLE_CMD_LED_F	FW_VI_ENABLE_CMD_LED_V(1U)
+
+#define FW_VI_ENABLE_CMD_DCB_INFO_S	28
+#define FW_VI_ENABLE_CMD_DCB_INFO_V(x)	((x) << FW_VI_ENABLE_CMD_DCB_INFO_S)
+
+/* VI VF stats offset definitions */
+#define VI_VF_NUM_STATS	16
+enum fw_vi_stats_vf_index {
+	FW_VI_VF_STAT_TX_BCAST_BYTES_IX,
+	FW_VI_VF_STAT_TX_BCAST_FRAMES_IX,
+	FW_VI_VF_STAT_TX_MCAST_BYTES_IX,
+	FW_VI_VF_STAT_TX_MCAST_FRAMES_IX,
+	FW_VI_VF_STAT_TX_UCAST_BYTES_IX,
+	FW_VI_VF_STAT_TX_UCAST_FRAMES_IX,
+	FW_VI_VF_STAT_TX_DROP_FRAMES_IX,
+	FW_VI_VF_STAT_TX_OFLD_BYTES_IX,
+	FW_VI_VF_STAT_TX_OFLD_FRAMES_IX,
+	FW_VI_VF_STAT_RX_BCAST_BYTES_IX,
+	FW_VI_VF_STAT_RX_BCAST_FRAMES_IX,
+	FW_VI_VF_STAT_RX_MCAST_BYTES_IX,
+	FW_VI_VF_STAT_RX_MCAST_FRAMES_IX,
+	FW_VI_VF_STAT_RX_UCAST_BYTES_IX,
+	FW_VI_VF_STAT_RX_UCAST_FRAMES_IX,
+	FW_VI_VF_STAT_RX_ERR_FRAMES_IX
+};
+
+/* VI PF stats offset definitions */
+#define VI_PF_NUM_STATS	17
+enum fw_vi_stats_pf_index {
+	FW_VI_PF_STAT_TX_BCAST_BYTES_IX,
+	FW_VI_PF_STAT_TX_BCAST_FRAMES_IX,
+	FW_VI_PF_STAT_TX_MCAST_BYTES_IX,
+	FW_VI_PF_STAT_TX_MCAST_FRAMES_IX,
+	FW_VI_PF_STAT_TX_UCAST_BYTES_IX,
+	FW_VI_PF_STAT_TX_UCAST_FRAMES_IX,
+	FW_VI_PF_STAT_TX_OFLD_BYTES_IX,
+	FW_VI_PF_STAT_TX_OFLD_FRAMES_IX,
+	FW_VI_PF_STAT_RX_BYTES_IX,
+	FW_VI_PF_STAT_RX_FRAMES_IX,
+	FW_VI_PF_STAT_RX_BCAST_BYTES_IX,
+	FW_VI_PF_STAT_RX_BCAST_FRAMES_IX,
+	FW_VI_PF_STAT_RX_MCAST_BYTES_IX,
+	FW_VI_PF_STAT_RX_MCAST_FRAMES_IX,
+	FW_VI_PF_STAT_RX_UCAST_BYTES_IX,
+	FW_VI_PF_STAT_RX_UCAST_FRAMES_IX,
+	FW_VI_PF_STAT_RX_ERR_FRAMES_IX
+};
+
+struct fw_vi_stats_cmd {
+	__be32 op_to_viid;
+	__be32 retval_len16;
+	union fw_vi_stats {
+		struct fw_vi_stats_ctl {
+			__be16 nstats_ix;
+			__be16 r6;
+			__be32 r7;
+			__be64 stat0;
+			__be64 stat1;
+			__be64 stat2;
+			__be64 stat3;
+			__be64 stat4;
+			__be64 stat5;
+		} ctl;
+		struct fw_vi_stats_pf {
+			__be64 tx_bcast_bytes;
+			__be64 tx_bcast_frames;
+			__be64 tx_mcast_bytes;
+			__be64 tx_mcast_frames;
+			__be64 tx_ucast_bytes;
+			__be64 tx_ucast_frames;
+			__be64 tx_offload_bytes;
+			__be64 tx_offload_frames;
+			__be64 rx_pf_bytes;
+			__be64 rx_pf_frames;
+			__be64 rx_bcast_bytes;
+			__be64 rx_bcast_frames;
+			__be64 rx_mcast_bytes;
+			__be64 rx_mcast_frames;
+			__be64 rx_ucast_bytes;
+			__be64 rx_ucast_frames;
+			__be64 rx_err_frames;
+		} pf;
+		struct fw_vi_stats_vf {
+			__be64 tx_bcast_bytes;
+			__be64 tx_bcast_frames;
+			__be64 tx_mcast_bytes;
+			__be64 tx_mcast_frames;
+			__be64 tx_ucast_bytes;
+			__be64 tx_ucast_frames;
+			__be64 tx_drop_frames;
+			__be64 tx_offload_bytes;
+			__be64 tx_offload_frames;
+			__be64 rx_bcast_bytes;
+			__be64 rx_bcast_frames;
+			__be64 rx_mcast_bytes;
+			__be64 rx_mcast_frames;
+			__be64 rx_ucast_bytes;
+			__be64 rx_ucast_frames;
+			__be64 rx_err_frames;
+		} vf;
+	} u;
+};
+
+#define FW_VI_STATS_CMD_VIID_S		0
+#define FW_VI_STATS_CMD_VIID_V(x)	((x) << FW_VI_STATS_CMD_VIID_S)
+
+#define FW_VI_STATS_CMD_NSTATS_S	12
+#define FW_VI_STATS_CMD_NSTATS_V(x)	((x) << FW_VI_STATS_CMD_NSTATS_S)
+
+#define FW_VI_STATS_CMD_IX_S	0
+#define FW_VI_STATS_CMD_IX_V(x)	((x) << FW_VI_STATS_CMD_IX_S)
+
+struct fw_acl_mac_cmd {
+	__be32 op_to_vfn;
+	__be32 en_to_len16;
+	u8 nmac;
+	u8 r3[7];
+	__be16 r4;
+	u8 macaddr0[6];
+	__be16 r5;
+	u8 macaddr1[6];
+	__be16 r6;
+	u8 macaddr2[6];
+	__be16 r7;
+	u8 macaddr3[6];
+};
+
+#define FW_ACL_MAC_CMD_PFN_S	8
+#define FW_ACL_MAC_CMD_PFN_V(x)	((x) << FW_ACL_MAC_CMD_PFN_S)
+
+#define FW_ACL_MAC_CMD_VFN_S	0
+#define FW_ACL_MAC_CMD_VFN_V(x)	((x) << FW_ACL_MAC_CMD_VFN_S)
+
+#define FW_ACL_MAC_CMD_EN_S	31
+#define FW_ACL_MAC_CMD_EN_V(x)	((x) << FW_ACL_MAC_CMD_EN_S)
+
+struct fw_acl_vlan_cmd {
+	__be32 op_to_vfn;
+	__be32 en_to_len16;
+	u8 nvlan;
+	u8 dropnovlan_fm;
+	u8 r3_lo[6];
+	__be16 vlanid[16];
+};
+
+#define FW_ACL_VLAN_CMD_PFN_S		8
+#define FW_ACL_VLAN_CMD_PFN_V(x)	((x) << FW_ACL_VLAN_CMD_PFN_S)
+
+#define FW_ACL_VLAN_CMD_VFN_S		0
+#define FW_ACL_VLAN_CMD_VFN_V(x)	((x) << FW_ACL_VLAN_CMD_VFN_S)
+
+#define FW_ACL_VLAN_CMD_EN_S		31
+#define FW_ACL_VLAN_CMD_EN_M		0x1
+#define FW_ACL_VLAN_CMD_EN_V(x)		((x) << FW_ACL_VLAN_CMD_EN_S)
+#define FW_ACL_VLAN_CMD_EN_G(x)         \
+	(((x) >> S_FW_ACL_VLAN_CMD_EN_S) & FW_ACL_VLAN_CMD_EN_M)
+#define FW_ACL_VLAN_CMD_EN_F            FW_ACL_VLAN_CMD_EN_V(1U)
+
+#define FW_ACL_VLAN_CMD_DROPNOVLAN_S	7
+#define FW_ACL_VLAN_CMD_DROPNOVLAN_V(x)	((x) << FW_ACL_VLAN_CMD_DROPNOVLAN_S)
+#define FW_ACL_VLAN_CMD_DROPNOVLAN_F    FW_ACL_VLAN_CMD_DROPNOVLAN_V(1U)
+
+#define FW_ACL_VLAN_CMD_FM_S		6
+#define FW_ACL_VLAN_CMD_FM_M		0x1
+#define FW_ACL_VLAN_CMD_FM_V(x)         ((x) << FW_ACL_VLAN_CMD_FM_S)
+#define FW_ACL_VLAN_CMD_FM_G(x)         \
+	(((x) >> FW_ACL_VLAN_CMD_FM_S) & FW_ACL_VLAN_CMD_FM_M)
+#define FW_ACL_VLAN_CMD_FM_F            FW_ACL_VLAN_CMD_FM_V(1U)
+
+/* old 16-bit port capabilities bitmap (fw_port_cap16_t) */
+enum fw_port_cap {
+	FW_PORT_CAP_SPEED_100M		= 0x0001,
+	FW_PORT_CAP_SPEED_1G		= 0x0002,
+	FW_PORT_CAP_SPEED_25G		= 0x0004,
+	FW_PORT_CAP_SPEED_10G		= 0x0008,
+	FW_PORT_CAP_SPEED_40G		= 0x0010,
+	FW_PORT_CAP_SPEED_100G		= 0x0020,
+	FW_PORT_CAP_FC_RX		= 0x0040,
+	FW_PORT_CAP_FC_TX		= 0x0080,
+	FW_PORT_CAP_ANEG		= 0x0100,
+	FW_PORT_CAP_MDIAUTO		= 0x0200,
+	FW_PORT_CAP_MDISTRAIGHT		= 0x0400,
+	FW_PORT_CAP_FEC_RS		= 0x0800,
+	FW_PORT_CAP_FEC_BASER_RS	= 0x1000,
+	FW_PORT_CAP_FORCE_PAUSE		= 0x2000,
+	FW_PORT_CAP_802_3_PAUSE		= 0x4000,
+	FW_PORT_CAP_802_3_ASM_DIR	= 0x8000,
+};
+
+#define FW_PORT_CAP_SPEED_S     0
+#define FW_PORT_CAP_SPEED_M     0x3f
+#define FW_PORT_CAP_SPEED_V(x)  ((x) << FW_PORT_CAP_SPEED_S)
+#define FW_PORT_CAP_SPEED_G(x) \
+	(((x) >> FW_PORT_CAP_SPEED_S) & FW_PORT_CAP_SPEED_M)
+
+enum fw_port_mdi {
+	FW_PORT_CAP_MDI_UNCHANGED,
+	FW_PORT_CAP_MDI_AUTO,
+	FW_PORT_CAP_MDI_F_STRAIGHT,
+	FW_PORT_CAP_MDI_F_CROSSOVER
+};
+
+#define FW_PORT_CAP_MDI_S 9
+#define FW_PORT_CAP_MDI_V(x) ((x) << FW_PORT_CAP_MDI_S)
+
+/* new 32-bit port capabilities bitmap (fw_port_cap32_t) */
+#define	FW_PORT_CAP32_SPEED_100M	0x00000001UL
+#define	FW_PORT_CAP32_SPEED_1G		0x00000002UL
+#define	FW_PORT_CAP32_SPEED_10G		0x00000004UL
+#define	FW_PORT_CAP32_SPEED_25G		0x00000008UL
+#define	FW_PORT_CAP32_SPEED_40G		0x00000010UL
+#define	FW_PORT_CAP32_SPEED_50G		0x00000020UL
+#define	FW_PORT_CAP32_SPEED_100G	0x00000040UL
+#define	FW_PORT_CAP32_SPEED_200G	0x00000080UL
+#define	FW_PORT_CAP32_SPEED_400G	0x00000100UL
+#define	FW_PORT_CAP32_SPEED_RESERVED1	0x00000200UL
+#define	FW_PORT_CAP32_SPEED_RESERVED2	0x00000400UL
+#define	FW_PORT_CAP32_SPEED_RESERVED3	0x00000800UL
+#define	FW_PORT_CAP32_RESERVED1		0x0000f000UL
+#define	FW_PORT_CAP32_FC_RX		0x00010000UL
+#define	FW_PORT_CAP32_FC_TX		0x00020000UL
+#define	FW_PORT_CAP32_802_3_PAUSE	0x00040000UL
+#define	FW_PORT_CAP32_802_3_ASM_DIR	0x00080000UL
+#define	FW_PORT_CAP32_ANEG		0x00100000UL
+#define	FW_PORT_CAP32_MDIAUTO		0x00200000UL
+#define	FW_PORT_CAP32_MDISTRAIGHT	0x00400000UL
+#define	FW_PORT_CAP32_FEC_RS		0x00800000UL
+#define	FW_PORT_CAP32_FEC_BASER_RS	0x01000000UL
+#define	FW_PORT_CAP32_FEC_RESERVED1	0x02000000UL
+#define	FW_PORT_CAP32_FEC_RESERVED2	0x04000000UL
+#define	FW_PORT_CAP32_FEC_RESERVED3	0x08000000UL
+#define FW_PORT_CAP32_FORCE_PAUSE	0x10000000UL
+#define FW_PORT_CAP32_RESERVED2		0xe0000000UL
+
+#define FW_PORT_CAP32_SPEED_S	0
+#define FW_PORT_CAP32_SPEED_M	0xfff
+#define FW_PORT_CAP32_SPEED_V(x)	((x) << FW_PORT_CAP32_SPEED_S)
+#define FW_PORT_CAP32_SPEED_G(x) \
+	(((x) >> FW_PORT_CAP32_SPEED_S) & FW_PORT_CAP32_SPEED_M)
+
+#define FW_PORT_CAP32_FC_S	16
+#define FW_PORT_CAP32_FC_M	0x3
+#define FW_PORT_CAP32_FC_V(x)	((x) << FW_PORT_CAP32_FC_S)
+#define FW_PORT_CAP32_FC_G(x) \
+	(((x) >> FW_PORT_CAP32_FC_S) & FW_PORT_CAP32_FC_M)
+
+#define FW_PORT_CAP32_802_3_S	18
+#define FW_PORT_CAP32_802_3_M	0x3
+#define FW_PORT_CAP32_802_3_V(x)	((x) << FW_PORT_CAP32_802_3_S)
+#define FW_PORT_CAP32_802_3_G(x) \
+	(((x) >> FW_PORT_CAP32_802_3_S) & FW_PORT_CAP32_802_3_M)
+
+#define FW_PORT_CAP32_ANEG_S	20
+#define FW_PORT_CAP32_ANEG_M	0x1
+#define FW_PORT_CAP32_ANEG_V(x)	((x) << FW_PORT_CAP32_ANEG_S)
+#define FW_PORT_CAP32_ANEG_G(x) \
+	(((x) >> FW_PORT_CAP32_ANEG_S) & FW_PORT_CAP32_ANEG_M)
+
+enum fw_port_mdi32 {
+	FW_PORT_CAP32_MDI_UNCHANGED,
+	FW_PORT_CAP32_MDI_AUTO,
+	FW_PORT_CAP32_MDI_F_STRAIGHT,
+	FW_PORT_CAP32_MDI_F_CROSSOVER
+};
+
+#define FW_PORT_CAP32_MDI_S 21
+#define FW_PORT_CAP32_MDI_M 3
+#define FW_PORT_CAP32_MDI_V(x) ((x) << FW_PORT_CAP32_MDI_S)
+#define FW_PORT_CAP32_MDI_G(x) \
+	(((x) >> FW_PORT_CAP32_MDI_S) & FW_PORT_CAP32_MDI_M)
+
+#define FW_PORT_CAP32_FEC_S	23
+#define FW_PORT_CAP32_FEC_M	0x1f
+#define FW_PORT_CAP32_FEC_V(x)	((x) << FW_PORT_CAP32_FEC_S)
+#define FW_PORT_CAP32_FEC_G(x) \
+	(((x) >> FW_PORT_CAP32_FEC_S) & FW_PORT_CAP32_FEC_M)
+
+/* macros to isolate various 32-bit Port Capabilities sub-fields */
+#define CAP32_SPEED(__cap32) \
+	(FW_PORT_CAP32_SPEED_V(FW_PORT_CAP32_SPEED_M) & __cap32)
+
+#define CAP32_FEC(__cap32) \
+	(FW_PORT_CAP32_FEC_V(FW_PORT_CAP32_FEC_M) & __cap32)
+
+enum fw_port_action {
+	FW_PORT_ACTION_L1_CFG		= 0x0001,
+	FW_PORT_ACTION_L2_CFG		= 0x0002,
+	FW_PORT_ACTION_GET_PORT_INFO	= 0x0003,
+	FW_PORT_ACTION_L2_PPP_CFG	= 0x0004,
+	FW_PORT_ACTION_L2_DCB_CFG	= 0x0005,
+	FW_PORT_ACTION_DCB_READ_TRANS	= 0x0006,
+	FW_PORT_ACTION_DCB_READ_RECV	= 0x0007,
+	FW_PORT_ACTION_DCB_READ_DET	= 0x0008,
+	FW_PORT_ACTION_L1_CFG32		= 0x0009,
+	FW_PORT_ACTION_GET_PORT_INFO32	= 0x000a,
+	FW_PORT_ACTION_LOW_PWR_TO_NORMAL = 0x0010,
+	FW_PORT_ACTION_L1_LOW_PWR_EN	= 0x0011,
+	FW_PORT_ACTION_L2_WOL_MODE_EN	= 0x0012,
+	FW_PORT_ACTION_LPBK_TO_NORMAL	= 0x0020,
+	FW_PORT_ACTION_L1_LPBK		= 0x0021,
+	FW_PORT_ACTION_L1_PMA_LPBK	= 0x0022,
+	FW_PORT_ACTION_L1_PCS_LPBK	= 0x0023,
+	FW_PORT_ACTION_L1_PHYXS_CSIDE_LPBK = 0x0024,
+	FW_PORT_ACTION_L1_PHYXS_ESIDE_LPBK = 0x0025,
+	FW_PORT_ACTION_PHY_RESET	= 0x0040,
+	FW_PORT_ACTION_PMA_RESET	= 0x0041,
+	FW_PORT_ACTION_PCS_RESET	= 0x0042,
+	FW_PORT_ACTION_PHYXS_RESET	= 0x0043,
+	FW_PORT_ACTION_DTEXS_REEST	= 0x0044,
+	FW_PORT_ACTION_AN_RESET		= 0x0045
+};
+
+enum fw_port_l2cfg_ctlbf {
+	FW_PORT_L2_CTLBF_OVLAN0	= 0x01,
+	FW_PORT_L2_CTLBF_OVLAN1	= 0x02,
+	FW_PORT_L2_CTLBF_OVLAN2	= 0x04,
+	FW_PORT_L2_CTLBF_OVLAN3	= 0x08,
+	FW_PORT_L2_CTLBF_IVLAN	= 0x10,
+	FW_PORT_L2_CTLBF_TXIPG	= 0x20
+};
+
+enum fw_port_dcb_versions {
+	FW_PORT_DCB_VER_UNKNOWN,
+	FW_PORT_DCB_VER_CEE1D0,
+	FW_PORT_DCB_VER_CEE1D01,
+	FW_PORT_DCB_VER_IEEE,
+	FW_PORT_DCB_VER_AUTO = 7
+};
+
+enum fw_port_dcb_cfg {
+	FW_PORT_DCB_CFG_PG	= 0x01,
+	FW_PORT_DCB_CFG_PFC	= 0x02,
+	FW_PORT_DCB_CFG_APPL	= 0x04
+};
+
+enum fw_port_dcb_cfg_rc {
+	FW_PORT_DCB_CFG_SUCCESS	= 0x0,
+	FW_PORT_DCB_CFG_ERROR	= 0x1
+};
+
+enum fw_port_dcb_type {
+	FW_PORT_DCB_TYPE_PGID		= 0x00,
+	FW_PORT_DCB_TYPE_PGRATE		= 0x01,
+	FW_PORT_DCB_TYPE_PRIORATE	= 0x02,
+	FW_PORT_DCB_TYPE_PFC		= 0x03,
+	FW_PORT_DCB_TYPE_APP_ID		= 0x04,
+	FW_PORT_DCB_TYPE_CONTROL	= 0x05,
+};
+
+enum fw_port_dcb_feature_state {
+	FW_PORT_DCB_FEATURE_STATE_PENDING = 0x0,
+	FW_PORT_DCB_FEATURE_STATE_SUCCESS = 0x1,
+	FW_PORT_DCB_FEATURE_STATE_ERROR	= 0x2,
+	FW_PORT_DCB_FEATURE_STATE_TIMEOUT = 0x3,
+};
+
+struct fw_port_cmd {
+	__be32 op_to_portid;
+	__be32 action_to_len16;
+	union fw_port {
+		struct fw_port_l1cfg {
+			__be32 rcap;
+			__be32 r;
+		} l1cfg;
+		struct fw_port_l2cfg {
+			__u8   ctlbf;
+			__u8   ovlan3_to_ivlan0;
+			__be16 ivlantype;
+			__be16 txipg_force_pinfo;
+			__be16 mtu;
+			__be16 ovlan0mask;
+			__be16 ovlan0type;
+			__be16 ovlan1mask;
+			__be16 ovlan1type;
+			__be16 ovlan2mask;
+			__be16 ovlan2type;
+			__be16 ovlan3mask;
+			__be16 ovlan3type;
+		} l2cfg;
+		struct fw_port_info {
+			__be32 lstatus_to_modtype;
+			__be16 pcap;
+			__be16 acap;
+			__be16 mtu;
+			__u8   cbllen;
+			__u8   auxlinfo;
+			__u8   dcbxdis_pkd;
+			__u8   r8_lo;
+			__be16 lpacap;
+			__be64 r9;
+		} info;
+		struct fw_port_diags {
+			__u8   diagop;
+			__u8   r[3];
+			__be32 diagval;
+		} diags;
+		union fw_port_dcb {
+			struct fw_port_dcb_pgid {
+				__u8   type;
+				__u8   apply_pkd;
+				__u8   r10_lo[2];
+				__be32 pgid;
+				__be64 r11;
+			} pgid;
+			struct fw_port_dcb_pgrate {
+				__u8   type;
+				__u8   apply_pkd;
+				__u8   r10_lo[5];
+				__u8   num_tcs_supported;
+				__u8   pgrate[8];
+				__u8   tsa[8];
+			} pgrate;
+			struct fw_port_dcb_priorate {
+				__u8   type;
+				__u8   apply_pkd;
+				__u8   r10_lo[6];
+				__u8   strict_priorate[8];
+			} priorate;
+			struct fw_port_dcb_pfc {
+				__u8   type;
+				__u8   pfcen;
+				__u8   r10[5];
+				__u8   max_pfc_tcs;
+				__be64 r11;
+			} pfc;
+			struct fw_port_app_priority {
+				__u8   type;
+				__u8   r10[2];
+				__u8   idx;
+				__u8   user_prio_map;
+				__u8   sel_field;
+				__be16 protocolid;
+				__be64 r12;
+			} app_priority;
+			struct fw_port_dcb_control {
+				__u8   type;
+				__u8   all_syncd_pkd;
+				__be16 dcb_version_to_app_state;
+				__be32 r11;
+				__be64 r12;
+			} control;
+		} dcb;
+		struct fw_port_l1cfg32 {
+			__be32 rcap32;
+			__be32 r;
+		} l1cfg32;
+		struct fw_port_info32 {
+			__be32 lstatus32_to_cbllen32;
+			__be32 auxlinfo32_mtu32;
+			__be32 linkattr32;
+			__be32 pcaps32;
+			__be32 acaps32;
+			__be32 lpacaps32;
+		} info32;
+	} u;
+};
+
+#define FW_PORT_CMD_READ_S	22
+#define FW_PORT_CMD_READ_V(x)	((x) << FW_PORT_CMD_READ_S)
+#define FW_PORT_CMD_READ_F	FW_PORT_CMD_READ_V(1U)
+
+#define FW_PORT_CMD_PORTID_S	0
+#define FW_PORT_CMD_PORTID_M	0xf
+#define FW_PORT_CMD_PORTID_V(x)	((x) << FW_PORT_CMD_PORTID_S)
+#define FW_PORT_CMD_PORTID_G(x)	\
+	(((x) >> FW_PORT_CMD_PORTID_S) & FW_PORT_CMD_PORTID_M)
+
+#define FW_PORT_CMD_ACTION_S	16
+#define FW_PORT_CMD_ACTION_M	0xffff
+#define FW_PORT_CMD_ACTION_V(x)	((x) << FW_PORT_CMD_ACTION_S)
+#define FW_PORT_CMD_ACTION_G(x)	\
+	(((x) >> FW_PORT_CMD_ACTION_S) & FW_PORT_CMD_ACTION_M)
+
+#define FW_PORT_CMD_OVLAN3_S	7
+#define FW_PORT_CMD_OVLAN3_V(x)	((x) << FW_PORT_CMD_OVLAN3_S)
+
+#define FW_PORT_CMD_OVLAN2_S	6
+#define FW_PORT_CMD_OVLAN2_V(x)	((x) << FW_PORT_CMD_OVLAN2_S)
+
+#define FW_PORT_CMD_OVLAN1_S	5
+#define FW_PORT_CMD_OVLAN1_V(x)	((x) << FW_PORT_CMD_OVLAN1_S)
+
+#define FW_PORT_CMD_OVLAN0_S	4
+#define FW_PORT_CMD_OVLAN0_V(x)	((x) << FW_PORT_CMD_OVLAN0_S)
+
+#define FW_PORT_CMD_IVLAN0_S	3
+#define FW_PORT_CMD_IVLAN0_V(x)	((x) << FW_PORT_CMD_IVLAN0_S)
+
+#define FW_PORT_CMD_TXIPG_S	3
+#define FW_PORT_CMD_TXIPG_V(x)	((x) << FW_PORT_CMD_TXIPG_S)
+
+#define FW_PORT_CMD_LSTATUS_S           31
+#define FW_PORT_CMD_LSTATUS_M           0x1
+#define FW_PORT_CMD_LSTATUS_V(x)        ((x) << FW_PORT_CMD_LSTATUS_S)
+#define FW_PORT_CMD_LSTATUS_G(x)        \
+	(((x) >> FW_PORT_CMD_LSTATUS_S) & FW_PORT_CMD_LSTATUS_M)
+#define FW_PORT_CMD_LSTATUS_F   FW_PORT_CMD_LSTATUS_V(1U)
+
+#define FW_PORT_CMD_LSPEED_S	24
+#define FW_PORT_CMD_LSPEED_M	0x3f
+#define FW_PORT_CMD_LSPEED_V(x)	((x) << FW_PORT_CMD_LSPEED_S)
+#define FW_PORT_CMD_LSPEED_G(x)	\
+	(((x) >> FW_PORT_CMD_LSPEED_S) & FW_PORT_CMD_LSPEED_M)
+
+#define FW_PORT_CMD_TXPAUSE_S		23
+#define FW_PORT_CMD_TXPAUSE_V(x)	((x) << FW_PORT_CMD_TXPAUSE_S)
+#define FW_PORT_CMD_TXPAUSE_F	FW_PORT_CMD_TXPAUSE_V(1U)
+
+#define FW_PORT_CMD_RXPAUSE_S		22
+#define FW_PORT_CMD_RXPAUSE_V(x)	((x) << FW_PORT_CMD_RXPAUSE_S)
+#define FW_PORT_CMD_RXPAUSE_F	FW_PORT_CMD_RXPAUSE_V(1U)
+
+#define FW_PORT_CMD_MDIOCAP_S		21
+#define FW_PORT_CMD_MDIOCAP_V(x)	((x) << FW_PORT_CMD_MDIOCAP_S)
+#define FW_PORT_CMD_MDIOCAP_F	FW_PORT_CMD_MDIOCAP_V(1U)
+
+#define FW_PORT_CMD_MDIOADDR_S		16
+#define FW_PORT_CMD_MDIOADDR_M		0x1f
+#define FW_PORT_CMD_MDIOADDR_G(x)	\
+	(((x) >> FW_PORT_CMD_MDIOADDR_S) & FW_PORT_CMD_MDIOADDR_M)
+
+#define FW_PORT_CMD_LPTXPAUSE_S		15
+#define FW_PORT_CMD_LPTXPAUSE_V(x)	((x) << FW_PORT_CMD_LPTXPAUSE_S)
+#define FW_PORT_CMD_LPTXPAUSE_F	FW_PORT_CMD_LPTXPAUSE_V(1U)
+
+#define FW_PORT_CMD_LPRXPAUSE_S		14
+#define FW_PORT_CMD_LPRXPAUSE_V(x)	((x) << FW_PORT_CMD_LPRXPAUSE_S)
+#define FW_PORT_CMD_LPRXPAUSE_F	FW_PORT_CMD_LPRXPAUSE_V(1U)
+
+#define FW_PORT_CMD_PTYPE_S	8
+#define FW_PORT_CMD_PTYPE_M	0x1f
+#define FW_PORT_CMD_PTYPE_G(x)	\
+	(((x) >> FW_PORT_CMD_PTYPE_S) & FW_PORT_CMD_PTYPE_M)
+
+#define FW_PORT_CMD_LINKDNRC_S		5
+#define FW_PORT_CMD_LINKDNRC_M		0x7
+#define FW_PORT_CMD_LINKDNRC_G(x)	\
+	(((x) >> FW_PORT_CMD_LINKDNRC_S) & FW_PORT_CMD_LINKDNRC_M)
+
+#define FW_PORT_CMD_MODTYPE_S		0
+#define FW_PORT_CMD_MODTYPE_M		0x1f
+#define FW_PORT_CMD_MODTYPE_V(x)	((x) << FW_PORT_CMD_MODTYPE_S)
+#define FW_PORT_CMD_MODTYPE_G(x)	\
+	(((x) >> FW_PORT_CMD_MODTYPE_S) & FW_PORT_CMD_MODTYPE_M)
+
+#define FW_PORT_CMD_DCBXDIS_S		7
+#define FW_PORT_CMD_DCBXDIS_V(x)	((x) << FW_PORT_CMD_DCBXDIS_S)
+#define FW_PORT_CMD_DCBXDIS_F	FW_PORT_CMD_DCBXDIS_V(1U)
+
+#define FW_PORT_CMD_APPLY_S	7
+#define FW_PORT_CMD_APPLY_V(x)	((x) << FW_PORT_CMD_APPLY_S)
+#define FW_PORT_CMD_APPLY_F	FW_PORT_CMD_APPLY_V(1U)
+
+#define FW_PORT_CMD_ALL_SYNCD_S		7
+#define FW_PORT_CMD_ALL_SYNCD_V(x)	((x) << FW_PORT_CMD_ALL_SYNCD_S)
+#define FW_PORT_CMD_ALL_SYNCD_F	FW_PORT_CMD_ALL_SYNCD_V(1U)
+
+#define FW_PORT_CMD_DCB_VERSION_S	12
+#define FW_PORT_CMD_DCB_VERSION_M	0x7
+#define FW_PORT_CMD_DCB_VERSION_G(x)	\
+	(((x) >> FW_PORT_CMD_DCB_VERSION_S) & FW_PORT_CMD_DCB_VERSION_M)
+
+#define FW_PORT_CMD_LSTATUS32_S		31
+#define FW_PORT_CMD_LSTATUS32_M		0x1
+#define FW_PORT_CMD_LSTATUS32_V(x)	((x) << FW_PORT_CMD_LSTATUS32_S)
+#define FW_PORT_CMD_LSTATUS32_G(x)	\
+	(((x) >> FW_PORT_CMD_LSTATUS32_S) & FW_PORT_CMD_LSTATUS32_M)
+#define FW_PORT_CMD_LSTATUS32_F	FW_PORT_CMD_LSTATUS32_V(1U)
+
+#define FW_PORT_CMD_LINKDNRC32_S	28
+#define FW_PORT_CMD_LINKDNRC32_M	0x7
+#define FW_PORT_CMD_LINKDNRC32_V(x)	((x) << FW_PORT_CMD_LINKDNRC32_S)
+#define FW_PORT_CMD_LINKDNRC32_G(x)	\
+	(((x) >> FW_PORT_CMD_LINKDNRC32_S) & FW_PORT_CMD_LINKDNRC32_M)
+
+#define FW_PORT_CMD_DCBXDIS32_S		27
+#define FW_PORT_CMD_DCBXDIS32_M		0x1
+#define FW_PORT_CMD_DCBXDIS32_V(x)	((x) << FW_PORT_CMD_DCBXDIS32_S)
+#define FW_PORT_CMD_DCBXDIS32_G(x)	\
+	(((x) >> FW_PORT_CMD_DCBXDIS32_S) & FW_PORT_CMD_DCBXDIS32_M)
+#define FW_PORT_CMD_DCBXDIS32_F	FW_PORT_CMD_DCBXDIS32_V(1U)
+
+#define FW_PORT_CMD_MDIOCAP32_S		26
+#define FW_PORT_CMD_MDIOCAP32_M		0x1
+#define FW_PORT_CMD_MDIOCAP32_V(x)	((x) << FW_PORT_CMD_MDIOCAP32_S)
+#define FW_PORT_CMD_MDIOCAP32_G(x)	\
+	(((x) >> FW_PORT_CMD_MDIOCAP32_S) & FW_PORT_CMD_MDIOCAP32_M)
+#define FW_PORT_CMD_MDIOCAP32_F	FW_PORT_CMD_MDIOCAP32_V(1U)
+
+#define FW_PORT_CMD_MDIOADDR32_S	21
+#define FW_PORT_CMD_MDIOADDR32_M	0x1f
+#define FW_PORT_CMD_MDIOADDR32_V(x)	((x) << FW_PORT_CMD_MDIOADDR32_S)
+#define FW_PORT_CMD_MDIOADDR32_G(x)	\
+	(((x) >> FW_PORT_CMD_MDIOADDR32_S) & FW_PORT_CMD_MDIOADDR32_M)
+
+#define FW_PORT_CMD_PORTTYPE32_S	13
+#define FW_PORT_CMD_PORTTYPE32_M	0xff
+#define FW_PORT_CMD_PORTTYPE32_V(x)	((x) << FW_PORT_CMD_PORTTYPE32_S)
+#define FW_PORT_CMD_PORTTYPE32_G(x)	\
+	(((x) >> FW_PORT_CMD_PORTTYPE32_S) & FW_PORT_CMD_PORTTYPE32_M)
+
+#define FW_PORT_CMD_MODTYPE32_S		8
+#define FW_PORT_CMD_MODTYPE32_M		0x1f
+#define FW_PORT_CMD_MODTYPE32_V(x)	((x) << FW_PORT_CMD_MODTYPE32_S)
+#define FW_PORT_CMD_MODTYPE32_G(x)	\
+	(((x) >> FW_PORT_CMD_MODTYPE32_S) & FW_PORT_CMD_MODTYPE32_M)
+
+#define FW_PORT_CMD_CBLLEN32_S		0
+#define FW_PORT_CMD_CBLLEN32_M		0xff
+#define FW_PORT_CMD_CBLLEN32_V(x)	((x) << FW_PORT_CMD_CBLLEN32_S)
+#define FW_PORT_CMD_CBLLEN32_G(x)	\
+	(((x) >> FW_PORT_CMD_CBLLEN32_S) & FW_PORT_CMD_CBLLEN32_M)
+
+#define FW_PORT_CMD_AUXLINFO32_S	24
+#define FW_PORT_CMD_AUXLINFO32_M	0xff
+#define FW_PORT_CMD_AUXLINFO32_V(x)	((x) << FW_PORT_CMD_AUXLINFO32_S)
+#define FW_PORT_CMD_AUXLINFO32_G(x)	\
+	(((x) >> FW_PORT_CMD_AUXLINFO32_S) & FW_PORT_CMD_AUXLINFO32_M)
+
+#define FW_PORT_AUXLINFO32_KX4_S	2
+#define FW_PORT_AUXLINFO32_KX4_M	0x1
+#define FW_PORT_AUXLINFO32_KX4_V(x) \
+	((x) << FW_PORT_AUXLINFO32_KX4_S)
+#define FW_PORT_AUXLINFO32_KX4_G(x) \
+	(((x) >> FW_PORT_AUXLINFO32_KX4_S) & FW_PORT_AUXLINFO32_KX4_M)
+#define FW_PORT_AUXLINFO32_KX4_F	FW_PORT_AUXLINFO32_KX4_V(1U)
+
+#define FW_PORT_AUXLINFO32_KR_S	1
+#define FW_PORT_AUXLINFO32_KR_M	0x1
+#define FW_PORT_AUXLINFO32_KR_V(x) \
+	((x) << FW_PORT_AUXLINFO32_KR_S)
+#define FW_PORT_AUXLINFO32_KR_G(x) \
+	(((x) >> FW_PORT_AUXLINFO32_KR_S) & FW_PORT_AUXLINFO32_KR_M)
+#define FW_PORT_AUXLINFO32_KR_F	FW_PORT_AUXLINFO32_KR_V(1U)
+
+#define FW_PORT_CMD_MTU32_S	0
+#define FW_PORT_CMD_MTU32_M	0xffff
+#define FW_PORT_CMD_MTU32_V(x)	((x) << FW_PORT_CMD_MTU32_S)
+#define FW_PORT_CMD_MTU32_G(x)	\
+	(((x) >> FW_PORT_CMD_MTU32_S) & FW_PORT_CMD_MTU32_M)
+
+enum fw_port_type {
+	FW_PORT_TYPE_FIBER_XFI,
+	FW_PORT_TYPE_FIBER_XAUI,
+	FW_PORT_TYPE_BT_SGMII,
+	FW_PORT_TYPE_BT_XFI,
+	FW_PORT_TYPE_BT_XAUI,
+	FW_PORT_TYPE_KX4,
+	FW_PORT_TYPE_CX4,
+	FW_PORT_TYPE_KX,
+	FW_PORT_TYPE_KR,
+	FW_PORT_TYPE_SFP,
+	FW_PORT_TYPE_BP_AP,
+	FW_PORT_TYPE_BP4_AP,
+	FW_PORT_TYPE_QSFP_10G,
+	FW_PORT_TYPE_QSA,
+	FW_PORT_TYPE_QSFP,
+	FW_PORT_TYPE_BP40_BA,
+	FW_PORT_TYPE_KR4_100G,
+	FW_PORT_TYPE_CR4_QSFP,
+	FW_PORT_TYPE_CR_QSFP,
+	FW_PORT_TYPE_CR2_QSFP,
+	FW_PORT_TYPE_SFP28,
+	FW_PORT_TYPE_KR_SFP28,
+	FW_PORT_TYPE_KR_XLAUI,
+
+	FW_PORT_TYPE_NONE = FW_PORT_CMD_PTYPE_M
+};
+
+enum fw_port_module_type {
+	FW_PORT_MOD_TYPE_NA,
+	FW_PORT_MOD_TYPE_LR,
+	FW_PORT_MOD_TYPE_SR,
+	FW_PORT_MOD_TYPE_ER,
+	FW_PORT_MOD_TYPE_TWINAX_PASSIVE,
+	FW_PORT_MOD_TYPE_TWINAX_ACTIVE,
+	FW_PORT_MOD_TYPE_LRM,
+	FW_PORT_MOD_TYPE_ERROR		= FW_PORT_CMD_MODTYPE_M - 3,
+	FW_PORT_MOD_TYPE_UNKNOWN	= FW_PORT_CMD_MODTYPE_M - 2,
+	FW_PORT_MOD_TYPE_NOTSUPPORTED	= FW_PORT_CMD_MODTYPE_M - 1,
+
+	FW_PORT_MOD_TYPE_NONE = FW_PORT_CMD_MODTYPE_M
+};
+
+enum fw_port_mod_sub_type {
+	FW_PORT_MOD_SUB_TYPE_NA,
+	FW_PORT_MOD_SUB_TYPE_MV88E114X = 0x1,
+	FW_PORT_MOD_SUB_TYPE_TN8022 = 0x2,
+	FW_PORT_MOD_SUB_TYPE_AQ1202 = 0x3,
+	FW_PORT_MOD_SUB_TYPE_88x3120 = 0x4,
+	FW_PORT_MOD_SUB_TYPE_BCM84834 = 0x5,
+	FW_PORT_MOD_SUB_TYPE_BT_VSC8634 = 0x8,
+
+	/* The following will never been in the VPD.  They are TWINAX cable
+	 * lengths decoded from SFP+ module i2c PROMs.  These should
+	 * almost certainly go somewhere else ...
+	 */
+	FW_PORT_MOD_SUB_TYPE_TWINAX_1 = 0x9,
+	FW_PORT_MOD_SUB_TYPE_TWINAX_3 = 0xA,
+	FW_PORT_MOD_SUB_TYPE_TWINAX_5 = 0xB,
+	FW_PORT_MOD_SUB_TYPE_TWINAX_7 = 0xC,
+};
+
+enum fw_port_stats_tx_index {
+	FW_STAT_TX_PORT_BYTES_IX = 0,
+	FW_STAT_TX_PORT_FRAMES_IX,
+	FW_STAT_TX_PORT_BCAST_IX,
+	FW_STAT_TX_PORT_MCAST_IX,
+	FW_STAT_TX_PORT_UCAST_IX,
+	FW_STAT_TX_PORT_ERROR_IX,
+	FW_STAT_TX_PORT_64B_IX,
+	FW_STAT_TX_PORT_65B_127B_IX,
+	FW_STAT_TX_PORT_128B_255B_IX,
+	FW_STAT_TX_PORT_256B_511B_IX,
+	FW_STAT_TX_PORT_512B_1023B_IX,
+	FW_STAT_TX_PORT_1024B_1518B_IX,
+	FW_STAT_TX_PORT_1519B_MAX_IX,
+	FW_STAT_TX_PORT_DROP_IX,
+	FW_STAT_TX_PORT_PAUSE_IX,
+	FW_STAT_TX_PORT_PPP0_IX,
+	FW_STAT_TX_PORT_PPP1_IX,
+	FW_STAT_TX_PORT_PPP2_IX,
+	FW_STAT_TX_PORT_PPP3_IX,
+	FW_STAT_TX_PORT_PPP4_IX,
+	FW_STAT_TX_PORT_PPP5_IX,
+	FW_STAT_TX_PORT_PPP6_IX,
+	FW_STAT_TX_PORT_PPP7_IX,
+	FW_NUM_PORT_TX_STATS
+};
+
+enum fw_port_stat_rx_index {
+	FW_STAT_RX_PORT_BYTES_IX = 0,
+	FW_STAT_RX_PORT_FRAMES_IX,
+	FW_STAT_RX_PORT_BCAST_IX,
+	FW_STAT_RX_PORT_MCAST_IX,
+	FW_STAT_RX_PORT_UCAST_IX,
+	FW_STAT_RX_PORT_MTU_ERROR_IX,
+	FW_STAT_RX_PORT_MTU_CRC_ERROR_IX,
+	FW_STAT_RX_PORT_CRC_ERROR_IX,
+	FW_STAT_RX_PORT_LEN_ERROR_IX,
+	FW_STAT_RX_PORT_SYM_ERROR_IX,
+	FW_STAT_RX_PORT_64B_IX,
+	FW_STAT_RX_PORT_65B_127B_IX,
+	FW_STAT_RX_PORT_128B_255B_IX,
+	FW_STAT_RX_PORT_256B_511B_IX,
+	FW_STAT_RX_PORT_512B_1023B_IX,
+	FW_STAT_RX_PORT_1024B_1518B_IX,
+	FW_STAT_RX_PORT_1519B_MAX_IX,
+	FW_STAT_RX_PORT_PAUSE_IX,
+	FW_STAT_RX_PORT_PPP0_IX,
+	FW_STAT_RX_PORT_PPP1_IX,
+	FW_STAT_RX_PORT_PPP2_IX,
+	FW_STAT_RX_PORT_PPP3_IX,
+	FW_STAT_RX_PORT_PPP4_IX,
+	FW_STAT_RX_PORT_PPP5_IX,
+	FW_STAT_RX_PORT_PPP6_IX,
+	FW_STAT_RX_PORT_PPP7_IX,
+	FW_STAT_RX_PORT_LESS_64B_IX,
+	FW_STAT_RX_PORT_MAC_ERROR_IX,
+	FW_NUM_PORT_RX_STATS
+};
+
+/* port stats */
+#define FW_NUM_PORT_STATS (FW_NUM_PORT_TX_STATS + FW_NUM_PORT_RX_STATS)
+
+struct fw_port_stats_cmd {
+	__be32 op_to_portid;
+	__be32 retval_len16;
+	union fw_port_stats {
+		struct fw_port_stats_ctl {
+			u8 nstats_bg_bm;
+			u8 tx_ix;
+			__be16 r6;
+			__be32 r7;
+			__be64 stat0;
+			__be64 stat1;
+			__be64 stat2;
+			__be64 stat3;
+			__be64 stat4;
+			__be64 stat5;
+		} ctl;
+		struct fw_port_stats_all {
+			__be64 tx_bytes;
+			__be64 tx_frames;
+			__be64 tx_bcast;
+			__be64 tx_mcast;
+			__be64 tx_ucast;
+			__be64 tx_error;
+			__be64 tx_64b;
+			__be64 tx_65b_127b;
+			__be64 tx_128b_255b;
+			__be64 tx_256b_511b;
+			__be64 tx_512b_1023b;
+			__be64 tx_1024b_1518b;
+			__be64 tx_1519b_max;
+			__be64 tx_drop;
+			__be64 tx_pause;
+			__be64 tx_ppp0;
+			__be64 tx_ppp1;
+			__be64 tx_ppp2;
+			__be64 tx_ppp3;
+			__be64 tx_ppp4;
+			__be64 tx_ppp5;
+			__be64 tx_ppp6;
+			__be64 tx_ppp7;
+			__be64 rx_bytes;
+			__be64 rx_frames;
+			__be64 rx_bcast;
+			__be64 rx_mcast;
+			__be64 rx_ucast;
+			__be64 rx_mtu_error;
+			__be64 rx_mtu_crc_error;
+			__be64 rx_crc_error;
+			__be64 rx_len_error;
+			__be64 rx_sym_error;
+			__be64 rx_64b;
+			__be64 rx_65b_127b;
+			__be64 rx_128b_255b;
+			__be64 rx_256b_511b;
+			__be64 rx_512b_1023b;
+			__be64 rx_1024b_1518b;
+			__be64 rx_1519b_max;
+			__be64 rx_pause;
+			__be64 rx_ppp0;
+			__be64 rx_ppp1;
+			__be64 rx_ppp2;
+			__be64 rx_ppp3;
+			__be64 rx_ppp4;
+			__be64 rx_ppp5;
+			__be64 rx_ppp6;
+			__be64 rx_ppp7;
+			__be64 rx_less_64b;
+			__be64 rx_bg_drop;
+			__be64 rx_bg_trunc;
+		} all;
+	} u;
+};
+
+/* port loopback stats */
+#define FW_NUM_LB_STATS 16
+enum fw_port_lb_stats_index {
+	FW_STAT_LB_PORT_BYTES_IX,
+	FW_STAT_LB_PORT_FRAMES_IX,
+	FW_STAT_LB_PORT_BCAST_IX,
+	FW_STAT_LB_PORT_MCAST_IX,
+	FW_STAT_LB_PORT_UCAST_IX,
+	FW_STAT_LB_PORT_ERROR_IX,
+	FW_STAT_LB_PORT_64B_IX,
+	FW_STAT_LB_PORT_65B_127B_IX,
+	FW_STAT_LB_PORT_128B_255B_IX,
+	FW_STAT_LB_PORT_256B_511B_IX,
+	FW_STAT_LB_PORT_512B_1023B_IX,
+	FW_STAT_LB_PORT_1024B_1518B_IX,
+	FW_STAT_LB_PORT_1519B_MAX_IX,
+	FW_STAT_LB_PORT_DROP_FRAMES_IX
+};
+
+struct fw_port_lb_stats_cmd {
+	__be32 op_to_lbport;
+	__be32 retval_len16;
+	union fw_port_lb_stats {
+		struct fw_port_lb_stats_ctl {
+			u8 nstats_bg_bm;
+			u8 ix_pkd;
+			__be16 r6;
+			__be32 r7;
+			__be64 stat0;
+			__be64 stat1;
+			__be64 stat2;
+			__be64 stat3;
+			__be64 stat4;
+			__be64 stat5;
+		} ctl;
+		struct fw_port_lb_stats_all {
+			__be64 tx_bytes;
+			__be64 tx_frames;
+			__be64 tx_bcast;
+			__be64 tx_mcast;
+			__be64 tx_ucast;
+			__be64 tx_error;
+			__be64 tx_64b;
+			__be64 tx_65b_127b;
+			__be64 tx_128b_255b;
+			__be64 tx_256b_511b;
+			__be64 tx_512b_1023b;
+			__be64 tx_1024b_1518b;
+			__be64 tx_1519b_max;
+			__be64 rx_lb_drop;
+			__be64 rx_lb_trunc;
+		} all;
+	} u;
+};
+
+enum fw_ptp_subop {
+	/* none */
+	FW_PTP_SC_INIT_TIMER            = 0x00,
+	FW_PTP_SC_TX_TYPE               = 0x01,
+	/* init */
+	FW_PTP_SC_RXTIME_STAMP          = 0x08,
+	FW_PTP_SC_RDRX_TYPE             = 0x09,
+	/* ts */
+	FW_PTP_SC_ADJ_FREQ              = 0x10,
+	FW_PTP_SC_ADJ_TIME              = 0x11,
+	FW_PTP_SC_ADJ_FTIME             = 0x12,
+	FW_PTP_SC_WALL_CLOCK            = 0x13,
+	FW_PTP_SC_GET_TIME              = 0x14,
+	FW_PTP_SC_SET_TIME              = 0x15,
+};
+
+struct fw_ptp_cmd {
+	__be32 op_to_portid;
+	__be32 retval_len16;
+	union fw_ptp {
+		struct fw_ptp_sc {
+			__u8   sc;
+			__u8   r3[7];
+		} scmd;
+		struct fw_ptp_init {
+			__u8   sc;
+			__u8   txchan;
+			__be16 absid;
+			__be16 mode;
+			__be16 r3;
+		} init;
+		struct fw_ptp_ts {
+			__u8   sc;
+			__u8   sign;
+			__be16 r3;
+			__be32 ppb;
+			__be64 tm;
+		} ts;
+	} u;
+	__be64 r3;
+};
+
+#define FW_PTP_CMD_PORTID_S             0
+#define FW_PTP_CMD_PORTID_M             0xf
+#define FW_PTP_CMD_PORTID_V(x)          ((x) << FW_PTP_CMD_PORTID_S)
+#define FW_PTP_CMD_PORTID_G(x)          \
+	(((x) >> FW_PTP_CMD_PORTID_S) & FW_PTP_CMD_PORTID_M)
+
+struct fw_rss_ind_tbl_cmd {
+	__be32 op_to_viid;
+	__be32 retval_len16;
+	__be16 niqid;
+	__be16 startidx;
+	__be32 r3;
+	__be32 iq0_to_iq2;
+	__be32 iq3_to_iq5;
+	__be32 iq6_to_iq8;
+	__be32 iq9_to_iq11;
+	__be32 iq12_to_iq14;
+	__be32 iq15_to_iq17;
+	__be32 iq18_to_iq20;
+	__be32 iq21_to_iq23;
+	__be32 iq24_to_iq26;
+	__be32 iq27_to_iq29;
+	__be32 iq30_iq31;
+	__be32 r15_lo;
+};
+
+#define FW_RSS_IND_TBL_CMD_VIID_S	0
+#define FW_RSS_IND_TBL_CMD_VIID_V(x)	((x) << FW_RSS_IND_TBL_CMD_VIID_S)
+
+#define FW_RSS_IND_TBL_CMD_IQ0_S	20
+#define FW_RSS_IND_TBL_CMD_IQ0_V(x)	((x) << FW_RSS_IND_TBL_CMD_IQ0_S)
+
+#define FW_RSS_IND_TBL_CMD_IQ1_S	10
+#define FW_RSS_IND_TBL_CMD_IQ1_V(x)	((x) << FW_RSS_IND_TBL_CMD_IQ1_S)
+
+#define FW_RSS_IND_TBL_CMD_IQ2_S	0
+#define FW_RSS_IND_TBL_CMD_IQ2_V(x)	((x) << FW_RSS_IND_TBL_CMD_IQ2_S)
+
+struct fw_rss_glb_config_cmd {
+	__be32 op_to_write;
+	__be32 retval_len16;
+	union fw_rss_glb_config {
+		struct fw_rss_glb_config_manual {
+			__be32 mode_pkd;
+			__be32 r3;
+			__be64 r4;
+			__be64 r5;
+		} manual;
+		struct fw_rss_glb_config_basicvirtual {
+			__be32 mode_pkd;
+			__be32 synmapen_to_hashtoeplitz;
+			__be64 r8;
+			__be64 r9;
+		} basicvirtual;
+	} u;
+};
+
+#define FW_RSS_GLB_CONFIG_CMD_MODE_S	28
+#define FW_RSS_GLB_CONFIG_CMD_MODE_M	0xf
+#define FW_RSS_GLB_CONFIG_CMD_MODE_V(x)	((x) << FW_RSS_GLB_CONFIG_CMD_MODE_S)
+#define FW_RSS_GLB_CONFIG_CMD_MODE_G(x)	\
+	(((x) >> FW_RSS_GLB_CONFIG_CMD_MODE_S) & FW_RSS_GLB_CONFIG_CMD_MODE_M)
+
+#define FW_RSS_GLB_CONFIG_CMD_MODE_MANUAL	0
+#define FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL	1
+
+#define FW_RSS_GLB_CONFIG_CMD_SYNMAPEN_S	8
+#define FW_RSS_GLB_CONFIG_CMD_SYNMAPEN_V(x)	\
+	((x) << FW_RSS_GLB_CONFIG_CMD_SYNMAPEN_S)
+#define FW_RSS_GLB_CONFIG_CMD_SYNMAPEN_F	\
+	FW_RSS_GLB_CONFIG_CMD_SYNMAPEN_V(1U)
+
+#define FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV6_S		7
+#define FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV6_V(x)	\
+	((x) << FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV6_S)
+#define FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV6_F	\
+	FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV6_V(1U)
+
+#define FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV6_S		6
+#define FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV6_V(x)	\
+	((x) << FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV6_S)
+#define FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV6_F	\
+	FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV6_V(1U)
+
+#define FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV4_S		5
+#define FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV4_V(x)	\
+	((x) << FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV4_S)
+#define FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV4_F	\
+	FW_RSS_GLB_CONFIG_CMD_SYN4TUPENIPV4_V(1U)
+
+#define FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV4_S		4
+#define FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV4_V(x)	\
+	((x) << FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV4_S)
+#define FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV4_F	\
+	FW_RSS_GLB_CONFIG_CMD_SYN2TUPENIPV4_V(1U)
+
+#define FW_RSS_GLB_CONFIG_CMD_OFDMAPEN_S	3
+#define FW_RSS_GLB_CONFIG_CMD_OFDMAPEN_V(x)	\
+	((x) << FW_RSS_GLB_CONFIG_CMD_OFDMAPEN_S)
+#define FW_RSS_GLB_CONFIG_CMD_OFDMAPEN_F	\
+	FW_RSS_GLB_CONFIG_CMD_OFDMAPEN_V(1U)
+
+#define FW_RSS_GLB_CONFIG_CMD_TNLMAPEN_S	2
+#define FW_RSS_GLB_CONFIG_CMD_TNLMAPEN_V(x)	\
+	((x) << FW_RSS_GLB_CONFIG_CMD_TNLMAPEN_S)
+#define FW_RSS_GLB_CONFIG_CMD_TNLMAPEN_F	\
+	FW_RSS_GLB_CONFIG_CMD_TNLMAPEN_V(1U)
+
+#define FW_RSS_GLB_CONFIG_CMD_TNLALLLKP_S	1
+#define FW_RSS_GLB_CONFIG_CMD_TNLALLLKP_V(x)	\
+	((x) << FW_RSS_GLB_CONFIG_CMD_TNLALLLKP_S)
+#define FW_RSS_GLB_CONFIG_CMD_TNLALLLKP_F	\
+	FW_RSS_GLB_CONFIG_CMD_TNLALLLKP_V(1U)
+
+#define FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ_S	0
+#define FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ_V(x)	\
+	((x) << FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ_S)
+#define FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ_F	\
+	FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ_V(1U)
+
+struct fw_rss_vi_config_cmd {
+	__be32 op_to_viid;
+#define FW_RSS_VI_CONFIG_CMD_VIID(x) ((x) << 0)
+	__be32 retval_len16;
+	union fw_rss_vi_config {
+		struct fw_rss_vi_config_manual {
+			__be64 r3;
+			__be64 r4;
+			__be64 r5;
+		} manual;
+		struct fw_rss_vi_config_basicvirtual {
+			__be32 r6;
+			__be32 defaultq_to_udpen;
+			__be64 r9;
+			__be64 r10;
+		} basicvirtual;
+	} u;
+};
+
+#define FW_RSS_VI_CONFIG_CMD_VIID_S	0
+#define FW_RSS_VI_CONFIG_CMD_VIID_V(x)	((x) << FW_RSS_VI_CONFIG_CMD_VIID_S)
+
+#define FW_RSS_VI_CONFIG_CMD_DEFAULTQ_S		16
+#define FW_RSS_VI_CONFIG_CMD_DEFAULTQ_M		0x3ff
+#define FW_RSS_VI_CONFIG_CMD_DEFAULTQ_V(x)	\
+	((x) << FW_RSS_VI_CONFIG_CMD_DEFAULTQ_S)
+#define FW_RSS_VI_CONFIG_CMD_DEFAULTQ_G(x)	\
+	(((x) >> FW_RSS_VI_CONFIG_CMD_DEFAULTQ_S) & \
+	 FW_RSS_VI_CONFIG_CMD_DEFAULTQ_M)
+
+#define FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN_S	4
+#define FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN_V(x)	\
+	((x) << FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN_S)
+#define FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN_F	\
+	FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN_V(1U)
+
+#define FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN_S	3
+#define FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN_V(x)	\
+	((x) << FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN_S)
+#define FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN_F	\
+	FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN_V(1U)
+
+#define FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN_S	2
+#define FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN_V(x)	\
+	((x) << FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN_S)
+#define FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN_F	\
+	FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN_V(1U)
+
+#define FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN_S	1
+#define FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN_V(x)	\
+	((x) << FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN_S)
+#define FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN_F	\
+	FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN_V(1U)
+
+#define FW_RSS_VI_CONFIG_CMD_UDPEN_S	0
+#define FW_RSS_VI_CONFIG_CMD_UDPEN_V(x)	((x) << FW_RSS_VI_CONFIG_CMD_UDPEN_S)
+#define FW_RSS_VI_CONFIG_CMD_UDPEN_F	FW_RSS_VI_CONFIG_CMD_UDPEN_V(1U)
+
+enum fw_sched_sc {
+	FW_SCHED_SC_PARAMS		= 1,
+};
+
+struct fw_sched_cmd {
+	__be32 op_to_write;
+	__be32 retval_len16;
+	union fw_sched {
+		struct fw_sched_config {
+			__u8   sc;
+			__u8   type;
+			__u8   minmaxen;
+			__u8   r3[5];
+			__u8   nclasses[4];
+			__be32 r4;
+		} config;
+		struct fw_sched_params {
+			__u8   sc;
+			__u8   type;
+			__u8   level;
+			__u8   mode;
+			__u8   unit;
+			__u8   rate;
+			__u8   ch;
+			__u8   cl;
+			__be32 min;
+			__be32 max;
+			__be16 weight;
+			__be16 pktsize;
+			__be16 burstsize;
+			__be16 r4;
+		} params;
+	} u;
+};
+
+struct fw_clip_cmd {
+	__be32 op_to_write;
+	__be32 alloc_to_len16;
+	__be64 ip_hi;
+	__be64 ip_lo;
+	__be32 r4[2];
+};
+
+#define FW_CLIP_CMD_ALLOC_S     31
+#define FW_CLIP_CMD_ALLOC_V(x)  ((x) << FW_CLIP_CMD_ALLOC_S)
+#define FW_CLIP_CMD_ALLOC_F     FW_CLIP_CMD_ALLOC_V(1U)
+
+#define FW_CLIP_CMD_FREE_S      30
+#define FW_CLIP_CMD_FREE_V(x)   ((x) << FW_CLIP_CMD_FREE_S)
+#define FW_CLIP_CMD_FREE_F      FW_CLIP_CMD_FREE_V(1U)
+
+enum fw_error_type {
+	FW_ERROR_TYPE_EXCEPTION		= 0x0,
+	FW_ERROR_TYPE_HWMODULE		= 0x1,
+	FW_ERROR_TYPE_WR		= 0x2,
+	FW_ERROR_TYPE_ACL		= 0x3,
+};
+
+struct fw_error_cmd {
+	__be32 op_to_type;
+	__be32 len16_pkd;
+	union fw_error {
+		struct fw_error_exception {
+			__be32 info[6];
+		} exception;
+		struct fw_error_hwmodule {
+			__be32 regaddr;
+			__be32 regval;
+		} hwmodule;
+		struct fw_error_wr {
+			__be16 cidx;
+			__be16 pfn_vfn;
+			__be32 eqid;
+			u8 wrhdr[16];
+		} wr;
+		struct fw_error_acl {
+			__be16 cidx;
+			__be16 pfn_vfn;
+			__be32 eqid;
+			__be16 mv_pkd;
+			u8 val[6];
+			__be64 r4;
+		} acl;
+	} u;
+};
+
+struct fw_debug_cmd {
+	__be32 op_type;
+	__be32 len16_pkd;
+	union fw_debug {
+		struct fw_debug_assert {
+			__be32 fcid;
+			__be32 line;
+			__be32 x;
+			__be32 y;
+			u8 filename_0_7[8];
+			u8 filename_8_15[8];
+			__be64 r3;
+		} assert;
+		struct fw_debug_prt {
+			__be16 dprtstridx;
+			__be16 r3[3];
+			__be32 dprtstrparam0;
+			__be32 dprtstrparam1;
+			__be32 dprtstrparam2;
+			__be32 dprtstrparam3;
+		} prt;
+	} u;
+};
+
+#define FW_DEBUG_CMD_TYPE_S	0
+#define FW_DEBUG_CMD_TYPE_M	0xff
+#define FW_DEBUG_CMD_TYPE_G(x)	\
+	(((x) >> FW_DEBUG_CMD_TYPE_S) & FW_DEBUG_CMD_TYPE_M)
+
+struct fw_hma_cmd {
+	__be32 op_pkd;
+	__be32 retval_len16;
+	__be32 mode_to_pcie_params;
+	__be32 naddr_size;
+	__be32 addr_size_pkd;
+	__be32 r6;
+	__be64 phy_address[5];
+};
+
+#define FW_HMA_CMD_MODE_S	31
+#define FW_HMA_CMD_MODE_M	0x1
+#define FW_HMA_CMD_MODE_V(x)	((x) << FW_HMA_CMD_MODE_S)
+#define FW_HMA_CMD_MODE_G(x)	\
+	(((x) >> FW_HMA_CMD_MODE_S) & FW_HMA_CMD_MODE_M)
+#define FW_HMA_CMD_MODE_F	FW_HMA_CMD_MODE_V(1U)
+
+#define FW_HMA_CMD_SOC_S	30
+#define FW_HMA_CMD_SOC_M	0x1
+#define FW_HMA_CMD_SOC_V(x)	((x) << FW_HMA_CMD_SOC_S)
+#define FW_HMA_CMD_SOC_G(x)	(((x) >> FW_HMA_CMD_SOC_S) & FW_HMA_CMD_SOC_M)
+#define FW_HMA_CMD_SOC_F	FW_HMA_CMD_SOC_V(1U)
+
+#define FW_HMA_CMD_EOC_S	29
+#define FW_HMA_CMD_EOC_M	0x1
+#define FW_HMA_CMD_EOC_V(x)	((x) << FW_HMA_CMD_EOC_S)
+#define FW_HMA_CMD_EOC_G(x)	(((x) >> FW_HMA_CMD_EOC_S) & FW_HMA_CMD_EOC_M)
+#define FW_HMA_CMD_EOC_F	FW_HMA_CMD_EOC_V(1U)
+
+#define FW_HMA_CMD_PCIE_PARAMS_S	0
+#define FW_HMA_CMD_PCIE_PARAMS_M	0x7ffffff
+#define FW_HMA_CMD_PCIE_PARAMS_V(x)	((x) << FW_HMA_CMD_PCIE_PARAMS_S)
+#define FW_HMA_CMD_PCIE_PARAMS_G(x)	\
+	(((x) >> FW_HMA_CMD_PCIE_PARAMS_S) & FW_HMA_CMD_PCIE_PARAMS_M)
+
+#define FW_HMA_CMD_NADDR_S	12
+#define FW_HMA_CMD_NADDR_M	0x3f
+#define FW_HMA_CMD_NADDR_V(x)	((x) << FW_HMA_CMD_NADDR_S)
+#define FW_HMA_CMD_NADDR_G(x)	\
+	(((x) >> FW_HMA_CMD_NADDR_S) & FW_HMA_CMD_NADDR_M)
+
+#define FW_HMA_CMD_SIZE_S	0
+#define FW_HMA_CMD_SIZE_M	0xfff
+#define FW_HMA_CMD_SIZE_V(x)	((x) << FW_HMA_CMD_SIZE_S)
+#define FW_HMA_CMD_SIZE_G(x)	\
+	(((x) >> FW_HMA_CMD_SIZE_S) & FW_HMA_CMD_SIZE_M)
+
+#define FW_HMA_CMD_ADDR_SIZE_S		11
+#define FW_HMA_CMD_ADDR_SIZE_M		0x1fffff
+#define FW_HMA_CMD_ADDR_SIZE_V(x)	((x) << FW_HMA_CMD_ADDR_SIZE_S)
+#define FW_HMA_CMD_ADDR_SIZE_G(x)	\
+	(((x) >> FW_HMA_CMD_ADDR_SIZE_S) & FW_HMA_CMD_ADDR_SIZE_M)
+
+enum pcie_fw_eval {
+	PCIE_FW_EVAL_CRASH = 0,
+};
+
+#define PCIE_FW_ERR_S		31
+#define PCIE_FW_ERR_V(x)	((x) << PCIE_FW_ERR_S)
+#define PCIE_FW_ERR_F		PCIE_FW_ERR_V(1U)
+
+#define PCIE_FW_INIT_S		30
+#define PCIE_FW_INIT_V(x)	((x) << PCIE_FW_INIT_S)
+#define PCIE_FW_INIT_F		PCIE_FW_INIT_V(1U)
+
+#define PCIE_FW_HALT_S          29
+#define PCIE_FW_HALT_V(x)       ((x) << PCIE_FW_HALT_S)
+#define PCIE_FW_HALT_F          PCIE_FW_HALT_V(1U)
+
+#define PCIE_FW_EVAL_S		24
+#define PCIE_FW_EVAL_M		0x7
+#define PCIE_FW_EVAL_G(x)	(((x) >> PCIE_FW_EVAL_S) & PCIE_FW_EVAL_M)
+
+#define PCIE_FW_MASTER_VLD_S	15
+#define PCIE_FW_MASTER_VLD_V(x)	((x) << PCIE_FW_MASTER_VLD_S)
+#define PCIE_FW_MASTER_VLD_F	PCIE_FW_MASTER_VLD_V(1U)
+
+#define PCIE_FW_MASTER_S	12
+#define PCIE_FW_MASTER_M	0x7
+#define PCIE_FW_MASTER_V(x)	((x) << PCIE_FW_MASTER_S)
+#define PCIE_FW_MASTER_G(x)	(((x) >> PCIE_FW_MASTER_S) & PCIE_FW_MASTER_M)
+
+struct fw_hdr {
+	u8 ver;
+	u8 chip;			/* terminator chip type */
+	__be16	len512;			/* bin length in units of 512-bytes */
+	__be32	fw_ver;			/* firmware version */
+	__be32	tp_microcode_ver;
+	u8 intfver_nic;
+	u8 intfver_vnic;
+	u8 intfver_ofld;
+	u8 intfver_ri;
+	u8 intfver_iscsipdu;
+	u8 intfver_iscsi;
+	u8 intfver_fcoepdu;
+	u8 intfver_fcoe;
+	__u32   reserved2;
+	__u32   reserved3;
+	__u32   reserved4;
+	__be32  flags;
+	__be32  reserved6[23];
+};
+
+enum fw_hdr_chip {
+	FW_HDR_CHIP_T4,
+	FW_HDR_CHIP_T5,
+	FW_HDR_CHIP_T6
+};
+
+#define FW_HDR_FW_VER_MAJOR_S	24
+#define FW_HDR_FW_VER_MAJOR_M	0xff
+#define FW_HDR_FW_VER_MAJOR_V(x) \
+	((x) << FW_HDR_FW_VER_MAJOR_S)
+#define FW_HDR_FW_VER_MAJOR_G(x) \
+	(((x) >> FW_HDR_FW_VER_MAJOR_S) & FW_HDR_FW_VER_MAJOR_M)
+
+#define FW_HDR_FW_VER_MINOR_S	16
+#define FW_HDR_FW_VER_MINOR_M	0xff
+#define FW_HDR_FW_VER_MINOR_V(x) \
+	((x) << FW_HDR_FW_VER_MINOR_S)
+#define FW_HDR_FW_VER_MINOR_G(x) \
+	(((x) >> FW_HDR_FW_VER_MINOR_S) & FW_HDR_FW_VER_MINOR_M)
+
+#define FW_HDR_FW_VER_MICRO_S	8
+#define FW_HDR_FW_VER_MICRO_M	0xff
+#define FW_HDR_FW_VER_MICRO_V(x) \
+	((x) << FW_HDR_FW_VER_MICRO_S)
+#define FW_HDR_FW_VER_MICRO_G(x) \
+	(((x) >> FW_HDR_FW_VER_MICRO_S) & FW_HDR_FW_VER_MICRO_M)
+
+#define FW_HDR_FW_VER_BUILD_S	0
+#define FW_HDR_FW_VER_BUILD_M	0xff
+#define FW_HDR_FW_VER_BUILD_V(x) \
+	((x) << FW_HDR_FW_VER_BUILD_S)
+#define FW_HDR_FW_VER_BUILD_G(x) \
+	(((x) >> FW_HDR_FW_VER_BUILD_S) & FW_HDR_FW_VER_BUILD_M)
+
+enum fw_hdr_intfver {
+	FW_HDR_INTFVER_NIC      = 0x00,
+	FW_HDR_INTFVER_VNIC     = 0x00,
+	FW_HDR_INTFVER_OFLD     = 0x00,
+	FW_HDR_INTFVER_RI       = 0x00,
+	FW_HDR_INTFVER_ISCSIPDU = 0x00,
+	FW_HDR_INTFVER_ISCSI    = 0x00,
+	FW_HDR_INTFVER_FCOEPDU  = 0x00,
+	FW_HDR_INTFVER_FCOE     = 0x00,
+};
+
+enum fw_hdr_flags {
+	FW_HDR_FLAGS_RESET_HALT = 0x00000001,
+};
+
+/* length of the formatting string  */
+#define FW_DEVLOG_FMT_LEN	192
+
+/* maximum number of the formatting string parameters */
+#define FW_DEVLOG_FMT_PARAMS_NUM 8
+
+/* priority levels */
+enum fw_devlog_level {
+	FW_DEVLOG_LEVEL_EMERG	= 0x0,
+	FW_DEVLOG_LEVEL_CRIT	= 0x1,
+	FW_DEVLOG_LEVEL_ERR	= 0x2,
+	FW_DEVLOG_LEVEL_NOTICE	= 0x3,
+	FW_DEVLOG_LEVEL_INFO	= 0x4,
+	FW_DEVLOG_LEVEL_DEBUG	= 0x5,
+	FW_DEVLOG_LEVEL_MAX	= 0x5,
+};
+
+/* facilities that may send a log message */
+enum fw_devlog_facility {
+	FW_DEVLOG_FACILITY_CORE		= 0x00,
+	FW_DEVLOG_FACILITY_CF		= 0x01,
+	FW_DEVLOG_FACILITY_SCHED	= 0x02,
+	FW_DEVLOG_FACILITY_TIMER	= 0x04,
+	FW_DEVLOG_FACILITY_RES		= 0x06,
+	FW_DEVLOG_FACILITY_HW		= 0x08,
+	FW_DEVLOG_FACILITY_FLR		= 0x10,
+	FW_DEVLOG_FACILITY_DMAQ		= 0x12,
+	FW_DEVLOG_FACILITY_PHY		= 0x14,
+	FW_DEVLOG_FACILITY_MAC		= 0x16,
+	FW_DEVLOG_FACILITY_PORT		= 0x18,
+	FW_DEVLOG_FACILITY_VI		= 0x1A,
+	FW_DEVLOG_FACILITY_FILTER	= 0x1C,
+	FW_DEVLOG_FACILITY_ACL		= 0x1E,
+	FW_DEVLOG_FACILITY_TM		= 0x20,
+	FW_DEVLOG_FACILITY_QFC		= 0x22,
+	FW_DEVLOG_FACILITY_DCB		= 0x24,
+	FW_DEVLOG_FACILITY_ETH		= 0x26,
+	FW_DEVLOG_FACILITY_OFLD		= 0x28,
+	FW_DEVLOG_FACILITY_RI		= 0x2A,
+	FW_DEVLOG_FACILITY_ISCSI	= 0x2C,
+	FW_DEVLOG_FACILITY_FCOE		= 0x2E,
+	FW_DEVLOG_FACILITY_FOISCSI	= 0x30,
+	FW_DEVLOG_FACILITY_FOFCOE	= 0x32,
+	FW_DEVLOG_FACILITY_CHNET        = 0x34,
+	FW_DEVLOG_FACILITY_MAX          = 0x34,
+};
+
+/* log message format */
+struct fw_devlog_e {
+	__be64	timestamp;
+	__be32	seqno;
+	__be16	reserved1;
+	__u8	level;
+	__u8	facility;
+	__u8	fmt[FW_DEVLOG_FMT_LEN];
+	__be32	params[FW_DEVLOG_FMT_PARAMS_NUM];
+	__be32	reserved3[4];
+};
+
+struct fw_devlog_cmd {
+	__be32 op_to_write;
+	__be32 retval_len16;
+	__u8   level;
+	__u8   r2[7];
+	__be32 memtype_devlog_memaddr16_devlog;
+	__be32 memsize_devlog;
+	__be32 r3[2];
+};
+
+#define FW_DEVLOG_CMD_MEMTYPE_DEVLOG_S		28
+#define FW_DEVLOG_CMD_MEMTYPE_DEVLOG_M		0xf
+#define FW_DEVLOG_CMD_MEMTYPE_DEVLOG_G(x)	\
+	(((x) >> FW_DEVLOG_CMD_MEMTYPE_DEVLOG_S) & \
+	 FW_DEVLOG_CMD_MEMTYPE_DEVLOG_M)
+
+#define FW_DEVLOG_CMD_MEMADDR16_DEVLOG_S	0
+#define FW_DEVLOG_CMD_MEMADDR16_DEVLOG_M	0xfffffff
+#define FW_DEVLOG_CMD_MEMADDR16_DEVLOG_G(x)	\
+	(((x) >> FW_DEVLOG_CMD_MEMADDR16_DEVLOG_S) & \
+	 FW_DEVLOG_CMD_MEMADDR16_DEVLOG_M)
+
+/* P C I E   F W   P F 7   R E G I S T E R */
+
+/* PF7 stores the Firmware Device Log parameters which allows Host Drivers to
+ * access the "devlog" which needing to contact firmware.  The encoding is
+ * mostly the same as that returned by the DEVLOG command except for the size
+ * which is encoded as the number of entries in multiples-1 of 128 here rather
+ * than the memory size as is done in the DEVLOG command.  Thus, 0 means 128
+ * and 15 means 2048.  This of course in turn constrains the allowed values
+ * for the devlog size ...
+ */
+#define PCIE_FW_PF_DEVLOG		7
+
+#define PCIE_FW_PF_DEVLOG_NENTRIES128_S	28
+#define PCIE_FW_PF_DEVLOG_NENTRIES128_M	0xf
+#define PCIE_FW_PF_DEVLOG_NENTRIES128_V(x) \
+	((x) << PCIE_FW_PF_DEVLOG_NENTRIES128_S)
+#define PCIE_FW_PF_DEVLOG_NENTRIES128_G(x) \
+	(((x) >> PCIE_FW_PF_DEVLOG_NENTRIES128_S) & \
+	 PCIE_FW_PF_DEVLOG_NENTRIES128_M)
+
+#define PCIE_FW_PF_DEVLOG_ADDR16_S	4
+#define PCIE_FW_PF_DEVLOG_ADDR16_M	0xffffff
+#define PCIE_FW_PF_DEVLOG_ADDR16_V(x)	((x) << PCIE_FW_PF_DEVLOG_ADDR16_S)
+#define PCIE_FW_PF_DEVLOG_ADDR16_G(x) \
+	(((x) >> PCIE_FW_PF_DEVLOG_ADDR16_S) & PCIE_FW_PF_DEVLOG_ADDR16_M)
+
+#define PCIE_FW_PF_DEVLOG_MEMTYPE_S	0
+#define PCIE_FW_PF_DEVLOG_MEMTYPE_M	0xf
+#define PCIE_FW_PF_DEVLOG_MEMTYPE_V(x)	((x) << PCIE_FW_PF_DEVLOG_MEMTYPE_S)
+#define PCIE_FW_PF_DEVLOG_MEMTYPE_G(x) \
+	(((x) >> PCIE_FW_PF_DEVLOG_MEMTYPE_S) & PCIE_FW_PF_DEVLOG_MEMTYPE_M)
+
+#define MAX_IMM_OFLD_TX_DATA_WR_LEN (0xff + sizeof(struct fw_ofld_tx_data_wr))
+
+struct fw_crypto_lookaside_wr {
+	__be32 op_to_cctx_size;
+	__be32 len16_pkd;
+	__be32 session_id;
+	__be32 rx_chid_to_rx_q_id;
+	__be32 key_addr;
+	__be32 pld_size_hash_size;
+	__be64 cookie;
+};
+
+#define FW_CRYPTO_LOOKASIDE_WR_OPCODE_S 24
+#define FW_CRYPTO_LOOKASIDE_WR_OPCODE_M 0xff
+#define FW_CRYPTO_LOOKASIDE_WR_OPCODE_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_OPCODE_S)
+#define FW_CRYPTO_LOOKASIDE_WR_OPCODE_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_OPCODE_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_OPCODE_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_COMPL_S 23
+#define FW_CRYPTO_LOOKASIDE_WR_COMPL_M 0x1
+#define FW_CRYPTO_LOOKASIDE_WR_COMPL_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_COMPL_S)
+#define FW_CRYPTO_LOOKASIDE_WR_COMPL_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_COMPL_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_COMPL_M)
+#define FW_CRYPTO_LOOKASIDE_WR_COMPL_F FW_CRYPTO_LOOKASIDE_WR_COMPL_V(1U)
+
+#define FW_CRYPTO_LOOKASIDE_WR_IMM_LEN_S 15
+#define FW_CRYPTO_LOOKASIDE_WR_IMM_LEN_M 0xff
+#define FW_CRYPTO_LOOKASIDE_WR_IMM_LEN_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_IMM_LEN_S)
+#define FW_CRYPTO_LOOKASIDE_WR_IMM_LEN_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_IMM_LEN_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_IMM_LEN_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_CCTX_LOC_S 5
+#define FW_CRYPTO_LOOKASIDE_WR_CCTX_LOC_M 0x3
+#define FW_CRYPTO_LOOKASIDE_WR_CCTX_LOC_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_CCTX_LOC_S)
+#define FW_CRYPTO_LOOKASIDE_WR_CCTX_LOC_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_CCTX_LOC_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_CCTX_LOC_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_CCTX_SIZE_S 0
+#define FW_CRYPTO_LOOKASIDE_WR_CCTX_SIZE_M 0x1f
+#define FW_CRYPTO_LOOKASIDE_WR_CCTX_SIZE_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_CCTX_SIZE_S)
+#define FW_CRYPTO_LOOKASIDE_WR_CCTX_SIZE_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_CCTX_SIZE_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_CCTX_SIZE_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_LEN16_S 0
+#define FW_CRYPTO_LOOKASIDE_WR_LEN16_M 0xff
+#define FW_CRYPTO_LOOKASIDE_WR_LEN16_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_LEN16_S)
+#define FW_CRYPTO_LOOKASIDE_WR_LEN16_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_LEN16_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_LEN16_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_RX_CHID_S 29
+#define FW_CRYPTO_LOOKASIDE_WR_RX_CHID_M 0x3
+#define FW_CRYPTO_LOOKASIDE_WR_RX_CHID_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_RX_CHID_S)
+#define FW_CRYPTO_LOOKASIDE_WR_RX_CHID_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_RX_CHID_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_RX_CHID_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_LCB_S  27
+#define FW_CRYPTO_LOOKASIDE_WR_LCB_M  0x3
+#define FW_CRYPTO_LOOKASIDE_WR_LCB_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_LCB_S)
+#define FW_CRYPTO_LOOKASIDE_WR_LCB_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_LCB_S) & FW_CRYPTO_LOOKASIDE_WR_LCB_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_PHASH_S 25
+#define FW_CRYPTO_LOOKASIDE_WR_PHASH_M 0x3
+#define FW_CRYPTO_LOOKASIDE_WR_PHASH_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_PHASH_S)
+#define FW_CRYPTO_LOOKASIDE_WR_PHASH_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_PHASH_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_PHASH_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_IV_S   23
+#define FW_CRYPTO_LOOKASIDE_WR_IV_M   0x3
+#define FW_CRYPTO_LOOKASIDE_WR_IV_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_IV_S)
+#define FW_CRYPTO_LOOKASIDE_WR_IV_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_IV_S) & FW_CRYPTO_LOOKASIDE_WR_IV_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_FQIDX_S   15
+#define FW_CRYPTO_LOOKASIDE_WR_FQIDX_M   0xff
+#define FW_CRYPTO_LOOKASIDE_WR_FQIDX_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_FQIDX_S)
+#define FW_CRYPTO_LOOKASIDE_WR_FQIDX_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_FQIDX_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_FQIDX_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_TX_CH_S 10
+#define FW_CRYPTO_LOOKASIDE_WR_TX_CH_M 0x3
+#define FW_CRYPTO_LOOKASIDE_WR_TX_CH_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_TX_CH_S)
+#define FW_CRYPTO_LOOKASIDE_WR_TX_CH_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_TX_CH_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_TX_CH_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_RX_Q_ID_S 0
+#define FW_CRYPTO_LOOKASIDE_WR_RX_Q_ID_M 0x3ff
+#define FW_CRYPTO_LOOKASIDE_WR_RX_Q_ID_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_RX_Q_ID_S)
+#define FW_CRYPTO_LOOKASIDE_WR_RX_Q_ID_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_RX_Q_ID_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_RX_Q_ID_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_PLD_SIZE_S 24
+#define FW_CRYPTO_LOOKASIDE_WR_PLD_SIZE_M 0xff
+#define FW_CRYPTO_LOOKASIDE_WR_PLD_SIZE_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_PLD_SIZE_S)
+#define FW_CRYPTO_LOOKASIDE_WR_PLD_SIZE_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_PLD_SIZE_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_PLD_SIZE_M)
+
+#define FW_CRYPTO_LOOKASIDE_WR_HASH_SIZE_S 17
+#define FW_CRYPTO_LOOKASIDE_WR_HASH_SIZE_M 0x7f
+#define FW_CRYPTO_LOOKASIDE_WR_HASH_SIZE_V(x) \
+	((x) << FW_CRYPTO_LOOKASIDE_WR_HASH_SIZE_S)
+#define FW_CRYPTO_LOOKASIDE_WR_HASH_SIZE_G(x) \
+	(((x) >> FW_CRYPTO_LOOKASIDE_WR_HASH_SIZE_S) & \
+	 FW_CRYPTO_LOOKASIDE_WR_HASH_SIZE_M)
+
+struct fw_tlstx_data_wr {
+	__be32 op_to_immdlen;
+	__be32 flowid_len16;
+	__be32 plen;
+	__be32 lsodisable_to_flags;
+	__be32 r5;
+	__be32 ctxloc_to_exp;
+	__be16 mfs;
+	__be16 adjustedplen_pkd;
+	__be16 expinplenmax_pkd;
+	u8   pdusinplenmax_pkd;
+	u8   r10;
+};
+
+#define FW_TLSTX_DATA_WR_OPCODE_S       24
+#define FW_TLSTX_DATA_WR_OPCODE_M       0xff
+#define FW_TLSTX_DATA_WR_OPCODE_V(x)    ((x) << FW_TLSTX_DATA_WR_OPCODE_S)
+#define FW_TLSTX_DATA_WR_OPCODE_G(x)    \
+	(((x) >> FW_TLSTX_DATA_WR_OPCODE_S) & FW_TLSTX_DATA_WR_OPCODE_M)
+
+#define FW_TLSTX_DATA_WR_COMPL_S        21
+#define FW_TLSTX_DATA_WR_COMPL_M        0x1
+#define FW_TLSTX_DATA_WR_COMPL_V(x)     ((x) << FW_TLSTX_DATA_WR_COMPL_S)
+#define FW_TLSTX_DATA_WR_COMPL_G(x)     \
+	(((x) >> FW_TLSTX_DATA_WR_COMPL_S) & FW_TLSTX_DATA_WR_COMPL_M)
+#define FW_TLSTX_DATA_WR_COMPL_F        FW_TLSTX_DATA_WR_COMPL_V(1U)
+
+#define FW_TLSTX_DATA_WR_IMMDLEN_S      0
+#define FW_TLSTX_DATA_WR_IMMDLEN_M      0xff
+#define FW_TLSTX_DATA_WR_IMMDLEN_V(x)   ((x) << FW_TLSTX_DATA_WR_IMMDLEN_S)
+#define FW_TLSTX_DATA_WR_IMMDLEN_G(x)   \
+	(((x) >> FW_TLSTX_DATA_WR_IMMDLEN_S) & FW_TLSTX_DATA_WR_IMMDLEN_M)
+
+#define FW_TLSTX_DATA_WR_FLOWID_S       8
+#define FW_TLSTX_DATA_WR_FLOWID_M       0xfffff
+#define FW_TLSTX_DATA_WR_FLOWID_V(x)    ((x) << FW_TLSTX_DATA_WR_FLOWID_S)
+#define FW_TLSTX_DATA_WR_FLOWID_G(x)    \
+	(((x) >> FW_TLSTX_DATA_WR_FLOWID_S) & FW_TLSTX_DATA_WR_FLOWID_M)
+
+#define FW_TLSTX_DATA_WR_LEN16_S        0
+#define FW_TLSTX_DATA_WR_LEN16_M        0xff
+#define FW_TLSTX_DATA_WR_LEN16_V(x)     ((x) << FW_TLSTX_DATA_WR_LEN16_S)
+#define FW_TLSTX_DATA_WR_LEN16_G(x)     \
+	(((x) >> FW_TLSTX_DATA_WR_LEN16_S) & FW_TLSTX_DATA_WR_LEN16_M)
+
+#define FW_TLSTX_DATA_WR_LSODISABLE_S   31
+#define FW_TLSTX_DATA_WR_LSODISABLE_M   0x1
+#define FW_TLSTX_DATA_WR_LSODISABLE_V(x) \
+	((x) << FW_TLSTX_DATA_WR_LSODISABLE_S)
+#define FW_TLSTX_DATA_WR_LSODISABLE_G(x) \
+	(((x) >> FW_TLSTX_DATA_WR_LSODISABLE_S) & FW_TLSTX_DATA_WR_LSODISABLE_M)
+#define FW_TLSTX_DATA_WR_LSODISABLE_F   FW_TLSTX_DATA_WR_LSODISABLE_V(1U)
+
+#define FW_TLSTX_DATA_WR_ALIGNPLD_S     30
+#define FW_TLSTX_DATA_WR_ALIGNPLD_M     0x1
+#define FW_TLSTX_DATA_WR_ALIGNPLD_V(x)  ((x) << FW_TLSTX_DATA_WR_ALIGNPLD_S)
+#define FW_TLSTX_DATA_WR_ALIGNPLD_G(x)  \
+	(((x) >> FW_TLSTX_DATA_WR_ALIGNPLD_S) & FW_TLSTX_DATA_WR_ALIGNPLD_M)
+#define FW_TLSTX_DATA_WR_ALIGNPLD_F     FW_TLSTX_DATA_WR_ALIGNPLD_V(1U)
+
+#define FW_TLSTX_DATA_WR_ALIGNPLDSHOVE_S 29
+#define FW_TLSTX_DATA_WR_ALIGNPLDSHOVE_M 0x1
+#define FW_TLSTX_DATA_WR_ALIGNPLDSHOVE_V(x) \
+	((x) << FW_TLSTX_DATA_WR_ALIGNPLDSHOVE_S)
+#define FW_TLSTX_DATA_WR_ALIGNPLDSHOVE_G(x) \
+	(((x) >> FW_TLSTX_DATA_WR_ALIGNPLDSHOVE_S) & \
+	FW_TLSTX_DATA_WR_ALIGNPLDSHOVE_M)
+#define FW_TLSTX_DATA_WR_ALIGNPLDSHOVE_F FW_TLSTX_DATA_WR_ALIGNPLDSHOVE_V(1U)
+
+#define FW_TLSTX_DATA_WR_FLAGS_S        0
+#define FW_TLSTX_DATA_WR_FLAGS_M        0xfffffff
+#define FW_TLSTX_DATA_WR_FLAGS_V(x)     ((x) << FW_TLSTX_DATA_WR_FLAGS_S)
+#define FW_TLSTX_DATA_WR_FLAGS_G(x)     \
+	(((x) >> FW_TLSTX_DATA_WR_FLAGS_S) & FW_TLSTX_DATA_WR_FLAGS_M)
+
+#define FW_TLSTX_DATA_WR_CTXLOC_S       30
+#define FW_TLSTX_DATA_WR_CTXLOC_M       0x3
+#define FW_TLSTX_DATA_WR_CTXLOC_V(x)    ((x) << FW_TLSTX_DATA_WR_CTXLOC_S)
+#define FW_TLSTX_DATA_WR_CTXLOC_G(x)    \
+	(((x) >> FW_TLSTX_DATA_WR_CTXLOC_S) & FW_TLSTX_DATA_WR_CTXLOC_M)
+
+#define FW_TLSTX_DATA_WR_IVDSGL_S       29
+#define FW_TLSTX_DATA_WR_IVDSGL_M       0x1
+#define FW_TLSTX_DATA_WR_IVDSGL_V(x)    ((x) << FW_TLSTX_DATA_WR_IVDSGL_S)
+#define FW_TLSTX_DATA_WR_IVDSGL_G(x)    \
+	(((x) >> FW_TLSTX_DATA_WR_IVDSGL_S) & FW_TLSTX_DATA_WR_IVDSGL_M)
+#define FW_TLSTX_DATA_WR_IVDSGL_F       FW_TLSTX_DATA_WR_IVDSGL_V(1U)
+
+#define FW_TLSTX_DATA_WR_KEYSIZE_S      24
+#define FW_TLSTX_DATA_WR_KEYSIZE_M      0x1f
+#define FW_TLSTX_DATA_WR_KEYSIZE_V(x)   ((x) << FW_TLSTX_DATA_WR_KEYSIZE_S)
+#define FW_TLSTX_DATA_WR_KEYSIZE_G(x)   \
+	(((x) >> FW_TLSTX_DATA_WR_KEYSIZE_S) & FW_TLSTX_DATA_WR_KEYSIZE_M)
+
+#define FW_TLSTX_DATA_WR_NUMIVS_S       14
+#define FW_TLSTX_DATA_WR_NUMIVS_M       0xff
+#define FW_TLSTX_DATA_WR_NUMIVS_V(x)    ((x) << FW_TLSTX_DATA_WR_NUMIVS_S)
+#define FW_TLSTX_DATA_WR_NUMIVS_G(x)    \
+	(((x) >> FW_TLSTX_DATA_WR_NUMIVS_S) & FW_TLSTX_DATA_WR_NUMIVS_M)
+
+#define FW_TLSTX_DATA_WR_EXP_S          0
+#define FW_TLSTX_DATA_WR_EXP_M          0x3fff
+#define FW_TLSTX_DATA_WR_EXP_V(x)       ((x) << FW_TLSTX_DATA_WR_EXP_S)
+#define FW_TLSTX_DATA_WR_EXP_G(x)       \
+	(((x) >> FW_TLSTX_DATA_WR_EXP_S) & FW_TLSTX_DATA_WR_EXP_M)
+
+#define FW_TLSTX_DATA_WR_ADJUSTEDPLEN_S 1
+#define FW_TLSTX_DATA_WR_ADJUSTEDPLEN_V(x) \
+	((x) << FW_TLSTX_DATA_WR_ADJUSTEDPLEN_S)
+
+#define FW_TLSTX_DATA_WR_EXPINPLENMAX_S 4
+#define FW_TLSTX_DATA_WR_EXPINPLENMAX_V(x) \
+	((x) << FW_TLSTX_DATA_WR_EXPINPLENMAX_S)
+
+#define FW_TLSTX_DATA_WR_PDUSINPLENMAX_S 2
+#define FW_TLSTX_DATA_WR_PDUSINPLENMAX_V(x) \
+	((x) << FW_TLSTX_DATA_WR_PDUSINPLENMAX_S)
+
+#endif /* _T4FW_INTERFACE_H_ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4fw_version.h b/drivers/net/ethernet/chelsio/cxgb4/t4fw_version.h
new file mode 100644
index 000000000..a02b1dff4
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4fw_version.h
@@ -0,0 +1,64 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __T4FW_VERSION_H__
+#define __T4FW_VERSION_H__
+
+#define T4FW_VERSION_MAJOR 0x01
+#define T4FW_VERSION_MINOR 0x17
+#define T4FW_VERSION_MICRO 0x03
+#define T4FW_VERSION_BUILD 0x00
+
+#define T4FW_MIN_VERSION_MAJOR 0x01
+#define T4FW_MIN_VERSION_MINOR 0x04
+#define T4FW_MIN_VERSION_MICRO 0x00
+
+#define T5FW_VERSION_MAJOR 0x01
+#define T5FW_VERSION_MINOR 0x17
+#define T5FW_VERSION_MICRO 0x03
+#define T5FW_VERSION_BUILD 0x00
+
+#define T5FW_MIN_VERSION_MAJOR 0x00
+#define T5FW_MIN_VERSION_MINOR 0x00
+#define T5FW_MIN_VERSION_MICRO 0x00
+
+#define T6FW_VERSION_MAJOR 0x01
+#define T6FW_VERSION_MINOR 0x17
+#define T6FW_VERSION_MICRO 0x03
+#define T6FW_VERSION_BUILD 0x00
+
+#define T6FW_MIN_VERSION_MAJOR 0x00
+#define T6FW_MIN_VERSION_MINOR 0x00
+#define T6FW_MIN_VERSION_MICRO 0x00
+#endif