summaryrefslogtreecommitdiffstats
path: root/drivers/infiniband/ulp/opa_vnic
diff options
context:
space:
mode:
authorDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-07 18:49:45 +0000
committerDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-07 18:49:45 +0000
commit2c3c1048746a4622d8c89a29670120dc8fab93c4 (patch)
tree848558de17fb3008cdf4d861b01ac7781903ce39 /drivers/infiniband/ulp/opa_vnic
parentInitial commit. (diff)
downloadlinux-2c3c1048746a4622d8c89a29670120dc8fab93c4.tar.xz
linux-2c3c1048746a4622d8c89a29670120dc8fab93c4.zip
Adding upstream version 6.1.76.upstream/6.1.76
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'drivers/infiniband/ulp/opa_vnic')
-rw-r--r--drivers/infiniband/ulp/opa_vnic/Kconfig9
-rw-r--r--drivers/infiniband/ulp/opa_vnic/Makefile9
-rw-r--r--drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c513
-rw-r--r--drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h524
-rw-r--r--drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c185
-rw-r--r--drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h329
-rw-r--r--drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c400
-rw-r--r--drivers/infiniband/ulp/opa_vnic/opa_vnic_vema.c1056
-rw-r--r--drivers/infiniband/ulp/opa_vnic/opa_vnic_vema_iface.c390
9 files changed, 3415 insertions, 0 deletions
diff --git a/drivers/infiniband/ulp/opa_vnic/Kconfig b/drivers/infiniband/ulp/opa_vnic/Kconfig
new file mode 100644
index 000000000..4d43d055f
--- /dev/null
+++ b/drivers/infiniband/ulp/opa_vnic/Kconfig
@@ -0,0 +1,9 @@
+# SPDX-License-Identifier: GPL-2.0-only
+config INFINIBAND_OPA_VNIC
+ tristate "Cornelis OPX VNIC support"
+ depends on X86_64 && INFINIBAND
+ help
+ This is Omni-Path Express (OPX) Virtual Network Interface Controller (VNIC)
+ driver for Ethernet over Omni-Path feature. It implements the HW
+ independent VNIC functionality. It interfaces with Linux stack for
+ data path and IB MAD for the control path.
diff --git a/drivers/infiniband/ulp/opa_vnic/Makefile b/drivers/infiniband/ulp/opa_vnic/Makefile
new file mode 100644
index 000000000..196183817
--- /dev/null
+++ b/drivers/infiniband/ulp/opa_vnic/Makefile
@@ -0,0 +1,9 @@
+# SPDX-License-Identifier: GPL-2.0-only
+# Makefile - Cornelis Omni-Path Express Virtual Network Controller driver
+# Copyright(c) 2017, Intel Corporation.
+# Copyright(c) 2021, Cornelis Networks.
+#
+obj-$(CONFIG_INFINIBAND_OPA_VNIC) += opa_vnic.o
+
+opa_vnic-y := opa_vnic_netdev.o opa_vnic_encap.o opa_vnic_ethtool.o \
+ opa_vnic_vema.o opa_vnic_vema_iface.o
diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c
new file mode 100644
index 000000000..31cd36141
--- /dev/null
+++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c
@@ -0,0 +1,513 @@
+/*
+ * Copyright(c) 2017 Intel Corporation.
+ *
+ * This file is provided under a dual BSD/GPLv2 license. When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * 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 GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * 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.
+ * - Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+/*
+ * This file contains OPA VNIC encapsulation/decapsulation function.
+ */
+
+#include <linux/if_ether.h>
+#include <linux/if_vlan.h>
+
+#include "opa_vnic_internal.h"
+
+/* OPA 16B Header fields */
+#define OPA_16B_LID_MASK 0xFFFFFull
+#define OPA_16B_SLID_HIGH_SHFT 8
+#define OPA_16B_SLID_MASK 0xF00ull
+#define OPA_16B_DLID_MASK 0xF000ull
+#define OPA_16B_DLID_HIGH_SHFT 12
+#define OPA_16B_LEN_SHFT 20
+#define OPA_16B_SC_SHFT 20
+#define OPA_16B_RC_SHFT 25
+#define OPA_16B_PKEY_SHFT 16
+
+#define OPA_VNIC_L4_HDR_SHFT 16
+
+/* L2+L4 hdr len is 20 bytes (5 quad words) */
+#define OPA_VNIC_HDR_QW_LEN 5
+
+static inline void opa_vnic_make_header(u8 *hdr, u32 slid, u32 dlid, u16 len,
+ u16 pkey, u16 entropy, u8 sc, u8 rc,
+ u8 l4_type, u16 l4_hdr)
+{
+ /* h[1]: LT=1, 16B L2=10 */
+ u32 h[OPA_VNIC_HDR_QW_LEN] = {0, 0xc0000000, 0, 0, 0};
+
+ h[2] = l4_type;
+ h[3] = entropy;
+ h[4] = l4_hdr << OPA_VNIC_L4_HDR_SHFT;
+
+ /* Extract and set 4 upper bits and 20 lower bits of the lids */
+ h[0] |= (slid & OPA_16B_LID_MASK);
+ h[2] |= ((slid >> (20 - OPA_16B_SLID_HIGH_SHFT)) & OPA_16B_SLID_MASK);
+
+ h[1] |= (dlid & OPA_16B_LID_MASK);
+ h[2] |= ((dlid >> (20 - OPA_16B_DLID_HIGH_SHFT)) & OPA_16B_DLID_MASK);
+
+ h[0] |= (len << OPA_16B_LEN_SHFT);
+ h[1] |= (rc << OPA_16B_RC_SHFT);
+ h[1] |= (sc << OPA_16B_SC_SHFT);
+ h[2] |= ((u32)pkey << OPA_16B_PKEY_SHFT);
+
+ memcpy(hdr, h, OPA_VNIC_HDR_LEN);
+}
+
+/*
+ * Using a simple hash table for mac table implementation with the last octet
+ * of mac address as a key.
+ */
+static void opa_vnic_free_mac_tbl(struct hlist_head *mactbl)
+{
+ struct opa_vnic_mac_tbl_node *node;
+ struct hlist_node *tmp;
+ int bkt;
+
+ if (!mactbl)
+ return;
+
+ vnic_hash_for_each_safe(mactbl, bkt, tmp, node, hlist) {
+ hash_del(&node->hlist);
+ kfree(node);
+ }
+ kfree(mactbl);
+}
+
+static struct hlist_head *opa_vnic_alloc_mac_tbl(void)
+{
+ u32 size = sizeof(struct hlist_head) * OPA_VNIC_MAC_TBL_SIZE;
+ struct hlist_head *mactbl;
+
+ mactbl = kzalloc(size, GFP_KERNEL);
+ if (!mactbl)
+ return ERR_PTR(-ENOMEM);
+
+ vnic_hash_init(mactbl);
+ return mactbl;
+}
+
+/* opa_vnic_release_mac_tbl - empty and free the mac table */
+void opa_vnic_release_mac_tbl(struct opa_vnic_adapter *adapter)
+{
+ struct hlist_head *mactbl;
+
+ mutex_lock(&adapter->mactbl_lock);
+ mactbl = rcu_access_pointer(adapter->mactbl);
+ rcu_assign_pointer(adapter->mactbl, NULL);
+ synchronize_rcu();
+ opa_vnic_free_mac_tbl(mactbl);
+ adapter->info.vport.mac_tbl_digest = 0;
+ mutex_unlock(&adapter->mactbl_lock);
+}
+
+/*
+ * opa_vnic_query_mac_tbl - query the mac table for a section
+ *
+ * This function implements query of specific function of the mac table.
+ * The function also expects the requested range to be valid.
+ */
+void opa_vnic_query_mac_tbl(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_mactable *tbl)
+{
+ struct opa_vnic_mac_tbl_node *node;
+ struct hlist_head *mactbl;
+ int bkt;
+ u16 loffset, lnum_entries;
+
+ rcu_read_lock();
+ mactbl = rcu_dereference(adapter->mactbl);
+ if (!mactbl)
+ goto get_mac_done;
+
+ loffset = be16_to_cpu(tbl->offset);
+ lnum_entries = be16_to_cpu(tbl->num_entries);
+
+ vnic_hash_for_each(mactbl, bkt, node, hlist) {
+ struct __opa_vnic_mactable_entry *nentry = &node->entry;
+ struct opa_veswport_mactable_entry *entry;
+
+ if ((node->index < loffset) ||
+ (node->index >= (loffset + lnum_entries)))
+ continue;
+
+ /* populate entry in the tbl corresponding to the index */
+ entry = &tbl->tbl_entries[node->index - loffset];
+ memcpy(entry->mac_addr, nentry->mac_addr,
+ ARRAY_SIZE(entry->mac_addr));
+ memcpy(entry->mac_addr_mask, nentry->mac_addr_mask,
+ ARRAY_SIZE(entry->mac_addr_mask));
+ entry->dlid_sd = cpu_to_be32(nentry->dlid_sd);
+ }
+ tbl->mac_tbl_digest = cpu_to_be32(adapter->info.vport.mac_tbl_digest);
+get_mac_done:
+ rcu_read_unlock();
+}
+
+/*
+ * opa_vnic_update_mac_tbl - update mac table section
+ *
+ * This function updates the specified section of the mac table.
+ * The procedure includes following steps.
+ * - Allocate a new mac (hash) table.
+ * - Add the specified entries to the new table.
+ * (except the ones that are requested to be deleted).
+ * - Add all the other entries from the old mac table.
+ * - If there is a failure, free the new table and return.
+ * - Switch to the new table.
+ * - Free the old table and return.
+ *
+ * The function also expects the requested range to be valid.
+ */
+int opa_vnic_update_mac_tbl(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_mactable *tbl)
+{
+ struct opa_vnic_mac_tbl_node *node, *new_node;
+ struct hlist_head *new_mactbl, *old_mactbl;
+ int i, bkt, rc = 0;
+ u8 key;
+ u16 loffset, lnum_entries;
+
+ mutex_lock(&adapter->mactbl_lock);
+ /* allocate new mac table */
+ new_mactbl = opa_vnic_alloc_mac_tbl();
+ if (IS_ERR(new_mactbl)) {
+ mutex_unlock(&adapter->mactbl_lock);
+ return PTR_ERR(new_mactbl);
+ }
+
+ loffset = be16_to_cpu(tbl->offset);
+ lnum_entries = be16_to_cpu(tbl->num_entries);
+
+ /* add updated entries to the new mac table */
+ for (i = 0; i < lnum_entries; i++) {
+ struct __opa_vnic_mactable_entry *nentry;
+ struct opa_veswport_mactable_entry *entry =
+ &tbl->tbl_entries[i];
+ u8 *mac_addr = entry->mac_addr;
+ u8 empty_mac[ETH_ALEN] = { 0 };
+
+ v_dbg("new mac entry %4d: %02x:%02x:%02x:%02x:%02x:%02x %x\n",
+ loffset + i, mac_addr[0], mac_addr[1], mac_addr[2],
+ mac_addr[3], mac_addr[4], mac_addr[5],
+ entry->dlid_sd);
+
+ /* if the entry is being removed, do not add it */
+ if (!memcmp(mac_addr, empty_mac, ARRAY_SIZE(empty_mac)))
+ continue;
+
+ node = kzalloc(sizeof(*node), GFP_KERNEL);
+ if (!node) {
+ rc = -ENOMEM;
+ goto updt_done;
+ }
+
+ node->index = loffset + i;
+ nentry = &node->entry;
+ memcpy(nentry->mac_addr, entry->mac_addr,
+ ARRAY_SIZE(nentry->mac_addr));
+ memcpy(nentry->mac_addr_mask, entry->mac_addr_mask,
+ ARRAY_SIZE(nentry->mac_addr_mask));
+ nentry->dlid_sd = be32_to_cpu(entry->dlid_sd);
+ key = node->entry.mac_addr[OPA_VNIC_MAC_HASH_IDX];
+ vnic_hash_add(new_mactbl, &node->hlist, key);
+ }
+
+ /* add other entries from current mac table to new mac table */
+ old_mactbl = rcu_access_pointer(adapter->mactbl);
+ if (!old_mactbl)
+ goto switch_tbl;
+
+ vnic_hash_for_each(old_mactbl, bkt, node, hlist) {
+ if ((node->index >= loffset) &&
+ (node->index < (loffset + lnum_entries)))
+ continue;
+
+ new_node = kzalloc(sizeof(*new_node), GFP_KERNEL);
+ if (!new_node) {
+ rc = -ENOMEM;
+ goto updt_done;
+ }
+
+ new_node->index = node->index;
+ memcpy(&new_node->entry, &node->entry, sizeof(node->entry));
+ key = new_node->entry.mac_addr[OPA_VNIC_MAC_HASH_IDX];
+ vnic_hash_add(new_mactbl, &new_node->hlist, key);
+ }
+
+switch_tbl:
+ /* switch to new table */
+ rcu_assign_pointer(adapter->mactbl, new_mactbl);
+ synchronize_rcu();
+
+ adapter->info.vport.mac_tbl_digest = be32_to_cpu(tbl->mac_tbl_digest);
+updt_done:
+ /* upon failure, free the new table; otherwise, free the old table */
+ if (rc)
+ opa_vnic_free_mac_tbl(new_mactbl);
+ else
+ opa_vnic_free_mac_tbl(old_mactbl);
+
+ mutex_unlock(&adapter->mactbl_lock);
+ return rc;
+}
+
+/* opa_vnic_chk_mac_tbl - check mac table for dlid */
+static uint32_t opa_vnic_chk_mac_tbl(struct opa_vnic_adapter *adapter,
+ struct ethhdr *mac_hdr)
+{
+ struct opa_vnic_mac_tbl_node *node;
+ struct hlist_head *mactbl;
+ u32 dlid = 0;
+ u8 key;
+
+ rcu_read_lock();
+ mactbl = rcu_dereference(adapter->mactbl);
+ if (unlikely(!mactbl))
+ goto chk_done;
+
+ key = mac_hdr->h_dest[OPA_VNIC_MAC_HASH_IDX];
+ vnic_hash_for_each_possible(mactbl, node, hlist, key) {
+ struct __opa_vnic_mactable_entry *entry = &node->entry;
+
+ /* if related to source mac, skip */
+ if (unlikely(OPA_VNIC_DLID_SD_IS_SRC_MAC(entry->dlid_sd)))
+ continue;
+
+ if (!memcmp(node->entry.mac_addr, mac_hdr->h_dest,
+ ARRAY_SIZE(node->entry.mac_addr))) {
+ /* mac address found */
+ dlid = OPA_VNIC_DLID_SD_GET_DLID(node->entry.dlid_sd);
+ break;
+ }
+ }
+
+chk_done:
+ rcu_read_unlock();
+ return dlid;
+}
+
+/* opa_vnic_get_dlid - find and return the DLID */
+static uint32_t opa_vnic_get_dlid(struct opa_vnic_adapter *adapter,
+ struct sk_buff *skb, u8 def_port)
+{
+ struct __opa_veswport_info *info = &adapter->info;
+ struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb);
+ u32 dlid;
+
+ dlid = opa_vnic_chk_mac_tbl(adapter, mac_hdr);
+ if (dlid)
+ return dlid;
+
+ if (is_multicast_ether_addr(mac_hdr->h_dest)) {
+ dlid = info->vesw.u_mcast_dlid;
+ } else {
+ if (is_local_ether_addr(mac_hdr->h_dest)) {
+ dlid = ((uint32_t)mac_hdr->h_dest[5] << 16) |
+ ((uint32_t)mac_hdr->h_dest[4] << 8) |
+ mac_hdr->h_dest[3];
+ if (unlikely(!dlid))
+ v_warn("Null dlid in MAC address\n");
+ } else if (def_port != OPA_VNIC_INVALID_PORT) {
+ if (def_port < OPA_VESW_MAX_NUM_DEF_PORT)
+ dlid = info->vesw.u_ucast_dlid[def_port];
+ }
+ }
+
+ return dlid;
+}
+
+/* opa_vnic_get_sc - return the service class */
+static u8 opa_vnic_get_sc(struct __opa_veswport_info *info,
+ struct sk_buff *skb)
+{
+ struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb);
+ u16 vlan_tci;
+ u8 sc;
+
+ if (!__vlan_get_tag(skb, &vlan_tci)) {
+ u8 pcp = OPA_VNIC_VLAN_PCP(vlan_tci);
+
+ if (is_multicast_ether_addr(mac_hdr->h_dest))
+ sc = info->vport.pcp_to_sc_mc[pcp];
+ else
+ sc = info->vport.pcp_to_sc_uc[pcp];
+ } else {
+ if (is_multicast_ether_addr(mac_hdr->h_dest))
+ sc = info->vport.non_vlan_sc_mc;
+ else
+ sc = info->vport.non_vlan_sc_uc;
+ }
+
+ return sc;
+}
+
+u8 opa_vnic_get_vl(struct opa_vnic_adapter *adapter, struct sk_buff *skb)
+{
+ struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb);
+ struct __opa_veswport_info *info = &adapter->info;
+ u8 vl;
+
+ if (skb_vlan_tag_present(skb)) {
+ u8 pcp = skb_vlan_tag_get(skb) >> VLAN_PRIO_SHIFT;
+
+ if (is_multicast_ether_addr(mac_hdr->h_dest))
+ vl = info->vport.pcp_to_vl_mc[pcp];
+ else
+ vl = info->vport.pcp_to_vl_uc[pcp];
+ } else {
+ if (is_multicast_ether_addr(mac_hdr->h_dest))
+ vl = info->vport.non_vlan_vl_mc;
+ else
+ vl = info->vport.non_vlan_vl_uc;
+ }
+
+ return vl;
+}
+
+/* opa_vnic_get_rc - return the routing control */
+static u8 opa_vnic_get_rc(struct __opa_veswport_info *info,
+ struct sk_buff *skb)
+{
+ u8 proto, rout_ctrl;
+
+ switch (vlan_get_protocol(skb)) {
+ case htons(ETH_P_IPV6):
+ proto = ipv6_hdr(skb)->nexthdr;
+ if (proto == IPPROTO_TCP)
+ rout_ctrl = OPA_VNIC_ENCAP_RC_EXT(info->vesw.rc,
+ IPV6_TCP);
+ else if (proto == IPPROTO_UDP)
+ rout_ctrl = OPA_VNIC_ENCAP_RC_EXT(info->vesw.rc,
+ IPV6_UDP);
+ else
+ rout_ctrl = OPA_VNIC_ENCAP_RC_EXT(info->vesw.rc, IPV6);
+ break;
+ case htons(ETH_P_IP):
+ proto = ip_hdr(skb)->protocol;
+ if (proto == IPPROTO_TCP)
+ rout_ctrl = OPA_VNIC_ENCAP_RC_EXT(info->vesw.rc,
+ IPV4_TCP);
+ else if (proto == IPPROTO_UDP)
+ rout_ctrl = OPA_VNIC_ENCAP_RC_EXT(info->vesw.rc,
+ IPV4_UDP);
+ else
+ rout_ctrl = OPA_VNIC_ENCAP_RC_EXT(info->vesw.rc, IPV4);
+ break;
+ default:
+ rout_ctrl = OPA_VNIC_ENCAP_RC_EXT(info->vesw.rc, DEFAULT);
+ }
+
+ return rout_ctrl;
+}
+
+/* opa_vnic_calc_entropy - calculate the packet entropy */
+u8 opa_vnic_calc_entropy(struct sk_buff *skb)
+{
+ u32 hash = skb_get_hash(skb);
+
+ /* store XOR of all bytes in lower 8 bits */
+ hash ^= hash >> 8;
+ hash ^= hash >> 16;
+
+ /* return lower 8 bits as entropy */
+ return (u8)(hash & 0xFF);
+}
+
+/* opa_vnic_get_def_port - get default port based on entropy */
+static inline u8 opa_vnic_get_def_port(struct opa_vnic_adapter *adapter,
+ u8 entropy)
+{
+ u8 flow_id;
+
+ /* Add the upper and lower 4-bits of entropy to get the flow id */
+ flow_id = ((entropy & 0xf) + (entropy >> 4));
+ return adapter->flow_tbl[flow_id & (OPA_VNIC_FLOW_TBL_SIZE - 1)];
+}
+
+/* Calculate packet length including OPA header, crc and padding */
+static inline int opa_vnic_wire_length(struct sk_buff *skb)
+{
+ u32 pad_len;
+
+ /* padding for 8 bytes size alignment */
+ pad_len = -(skb->len + OPA_VNIC_ICRC_TAIL_LEN) & 0x7;
+ pad_len += OPA_VNIC_ICRC_TAIL_LEN;
+
+ return (skb->len + pad_len) >> 3;
+}
+
+/* opa_vnic_encap_skb - encapsulate skb packet with OPA header and meta data */
+void opa_vnic_encap_skb(struct opa_vnic_adapter *adapter, struct sk_buff *skb)
+{
+ struct __opa_veswport_info *info = &adapter->info;
+ struct opa_vnic_skb_mdata *mdata;
+ u8 def_port, sc, rc, entropy, *hdr;
+ u16 len, l4_hdr;
+ u32 dlid;
+
+ hdr = skb_push(skb, OPA_VNIC_HDR_LEN);
+
+ entropy = opa_vnic_calc_entropy(skb);
+ def_port = opa_vnic_get_def_port(adapter, entropy);
+ len = opa_vnic_wire_length(skb);
+ dlid = opa_vnic_get_dlid(adapter, skb, def_port);
+ sc = opa_vnic_get_sc(info, skb);
+ rc = opa_vnic_get_rc(info, skb);
+ l4_hdr = info->vesw.vesw_id;
+
+ mdata = skb_push(skb, sizeof(*mdata));
+ mdata->vl = opa_vnic_get_vl(adapter, skb);
+ mdata->entropy = entropy;
+ mdata->flags = 0;
+ if (unlikely(!dlid)) {
+ mdata->flags = OPA_VNIC_SKB_MDATA_ENCAP_ERR;
+ return;
+ }
+
+ opa_vnic_make_header(hdr, info->vport.encap_slid, dlid, len,
+ info->vesw.pkey, entropy, sc, rc,
+ OPA_VNIC_L4_ETHR, l4_hdr);
+}
diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h
new file mode 100644
index 000000000..012fc27c5
--- /dev/null
+++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h
@@ -0,0 +1,524 @@
+#ifndef _OPA_VNIC_ENCAP_H
+#define _OPA_VNIC_ENCAP_H
+/*
+ * Copyright(c) 2017 Intel Corporation.
+ *
+ * This file is provided under a dual BSD/GPLv2 license. When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * 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 GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * 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.
+ * - Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+/*
+ * This file contains all OPA VNIC declaration required for encapsulation
+ * and decapsulation of Ethernet packets
+ */
+
+#include <linux/types.h>
+#include <rdma/ib_mad.h>
+
+/* EMA class version */
+#define OPA_EMA_CLASS_VERSION 0x80
+
+/*
+ * Define the Intel vendor management class for OPA
+ * ETHERNET MANAGEMENT
+ */
+#define OPA_MGMT_CLASS_INTEL_EMA 0x34
+
+/* EM attribute IDs */
+#define OPA_EM_ATTR_CLASS_PORT_INFO 0x0001
+#define OPA_EM_ATTR_VESWPORT_INFO 0x0011
+#define OPA_EM_ATTR_VESWPORT_MAC_ENTRIES 0x0012
+#define OPA_EM_ATTR_IFACE_UCAST_MACS 0x0013
+#define OPA_EM_ATTR_IFACE_MCAST_MACS 0x0014
+#define OPA_EM_ATTR_DELETE_VESW 0x0015
+#define OPA_EM_ATTR_VESWPORT_SUMMARY_COUNTERS 0x0020
+#define OPA_EM_ATTR_VESWPORT_ERROR_COUNTERS 0x0022
+
+/* VNIC configured and operational state values */
+#define OPA_VNIC_STATE_DROP_ALL 0x1
+#define OPA_VNIC_STATE_FORWARDING 0x3
+
+#define OPA_VESW_MAX_NUM_DEF_PORT 16
+#define OPA_VNIC_MAX_NUM_PCP 8
+
+#define OPA_VNIC_EMA_DATA (OPA_MGMT_MAD_SIZE - IB_MGMT_VENDOR_HDR)
+
+/* Defines for vendor specific notice(trap) attributes */
+#define OPA_INTEL_EMA_NOTICE_TYPE_INFO 0x04
+
+/* INTEL OUI */
+#define INTEL_OUI_1 0x00
+#define INTEL_OUI_2 0x06
+#define INTEL_OUI_3 0x6a
+
+/* Trap opcodes sent from VNIC */
+#define OPA_VESWPORT_TRAP_IFACE_UCAST_MAC_CHANGE 0x1
+#define OPA_VESWPORT_TRAP_IFACE_MCAST_MAC_CHANGE 0x2
+#define OPA_VESWPORT_TRAP_ETH_LINK_STATUS_CHANGE 0x3
+
+#define OPA_VNIC_DLID_SD_IS_SRC_MAC(dlid_sd) (!!((dlid_sd) & 0x20))
+#define OPA_VNIC_DLID_SD_GET_DLID(dlid_sd) ((dlid_sd) >> 8)
+
+/* VNIC Ethernet link status */
+#define OPA_VNIC_ETH_LINK_UP 1
+#define OPA_VNIC_ETH_LINK_DOWN 2
+
+/* routing control */
+#define OPA_VNIC_ENCAP_RC_DEFAULT 0
+#define OPA_VNIC_ENCAP_RC_IPV4 4
+#define OPA_VNIC_ENCAP_RC_IPV4_UDP 8
+#define OPA_VNIC_ENCAP_RC_IPV4_TCP 12
+#define OPA_VNIC_ENCAP_RC_IPV6 16
+#define OPA_VNIC_ENCAP_RC_IPV6_TCP 20
+#define OPA_VNIC_ENCAP_RC_IPV6_UDP 24
+
+#define OPA_VNIC_ENCAP_RC_EXT(w, b) (((w) >> OPA_VNIC_ENCAP_RC_ ## b) & 0x7)
+
+/**
+ * struct opa_vesw_info - OPA vnic switch information
+ * @fabric_id: 10-bit fabric id
+ * @vesw_id: 12-bit virtual ethernet switch id
+ * @rsvd0: reserved bytes
+ * @def_port_mask: bitmask of default ports
+ * @rsvd1: reserved bytes
+ * @pkey: partition key
+ * @rsvd2: reserved bytes
+ * @u_mcast_dlid: unknown multicast dlid
+ * @u_ucast_dlid: array of unknown unicast dlids
+ * @rsvd3: reserved bytes
+ * @rc: routing control
+ * @eth_mtu: Ethernet MTU
+ * @rsvd4: reserved bytes
+ */
+struct opa_vesw_info {
+ __be16 fabric_id;
+ __be16 vesw_id;
+
+ u8 rsvd0[6];
+ __be16 def_port_mask;
+
+ u8 rsvd1[2];
+ __be16 pkey;
+
+ u8 rsvd2[4];
+ __be32 u_mcast_dlid;
+ __be32 u_ucast_dlid[OPA_VESW_MAX_NUM_DEF_PORT];
+
+ __be32 rc;
+
+ u8 rsvd3[56];
+ __be16 eth_mtu;
+ u8 rsvd4[2];
+} __packed;
+
+/**
+ * struct opa_per_veswport_info - OPA vnic per port information
+ * @port_num: port number
+ * @eth_link_status: current ethernet link state
+ * @rsvd0: reserved bytes
+ * @base_mac_addr: base mac address
+ * @config_state: configured port state
+ * @oper_state: operational port state
+ * @max_mac_tbl_ent: max number of mac table entries
+ * @max_smac_ent: max smac entries in mac table
+ * @mac_tbl_digest: mac table digest
+ * @rsvd1: reserved bytes
+ * @encap_slid: base slid for the port
+ * @pcp_to_sc_uc: sc by pcp index for unicast ethernet packets
+ * @pcp_to_vl_uc: vl by pcp index for unicast ethernet packets
+ * @pcp_to_sc_mc: sc by pcp index for multicast ethernet packets
+ * @pcp_to_vl_mc: vl by pcp index for multicast ethernet packets
+ * @non_vlan_sc_uc: sc for non-vlan unicast ethernet packets
+ * @non_vlan_vl_uc: vl for non-vlan unicast ethernet packets
+ * @non_vlan_sc_mc: sc for non-vlan multicast ethernet packets
+ * @non_vlan_vl_mc: vl for non-vlan multicast ethernet packets
+ * @rsvd2: reserved bytes
+ * @uc_macs_gen_count: generation count for unicast macs list
+ * @mc_macs_gen_count: generation count for multicast macs list
+ * @rsvd3: reserved bytes
+ */
+struct opa_per_veswport_info {
+ __be32 port_num;
+
+ u8 eth_link_status;
+ u8 rsvd0[3];
+
+ u8 base_mac_addr[ETH_ALEN];
+ u8 config_state;
+ u8 oper_state;
+
+ __be16 max_mac_tbl_ent;
+ __be16 max_smac_ent;
+ __be32 mac_tbl_digest;
+ u8 rsvd1[4];
+
+ __be32 encap_slid;
+
+ u8 pcp_to_sc_uc[OPA_VNIC_MAX_NUM_PCP];
+ u8 pcp_to_vl_uc[OPA_VNIC_MAX_NUM_PCP];
+ u8 pcp_to_sc_mc[OPA_VNIC_MAX_NUM_PCP];
+ u8 pcp_to_vl_mc[OPA_VNIC_MAX_NUM_PCP];
+
+ u8 non_vlan_sc_uc;
+ u8 non_vlan_vl_uc;
+ u8 non_vlan_sc_mc;
+ u8 non_vlan_vl_mc;
+
+ u8 rsvd2[48];
+
+ __be16 uc_macs_gen_count;
+ __be16 mc_macs_gen_count;
+
+ u8 rsvd3[8];
+} __packed;
+
+/**
+ * struct opa_veswport_info - OPA vnic port information
+ * @vesw: OPA vnic switch information
+ * @vport: OPA vnic per port information
+ *
+ * On host, each of the virtual ethernet ports belongs
+ * to a different virtual ethernet switches.
+ */
+struct opa_veswport_info {
+ struct opa_vesw_info vesw;
+ struct opa_per_veswport_info vport;
+};
+
+/**
+ * struct opa_veswport_mactable_entry - single entry in the forwarding table
+ * @mac_addr: MAC address
+ * @mac_addr_mask: MAC address bit mask
+ * @dlid_sd: Matching DLID and side data
+ *
+ * On the host each virtual ethernet port will have
+ * a forwarding table. These tables are used to
+ * map a MAC to a LID and other data. For more
+ * details see struct opa_veswport_mactable_entries.
+ * This is the structure of a single mactable entry
+ */
+struct opa_veswport_mactable_entry {
+ u8 mac_addr[ETH_ALEN];
+ u8 mac_addr_mask[ETH_ALEN];
+ __be32 dlid_sd;
+} __packed;
+
+/**
+ * struct opa_veswport_mactable - Forwarding table array
+ * @offset: mac table starting offset
+ * @num_entries: Number of entries to get or set
+ * @mac_tbl_digest: mac table digest
+ * @tbl_entries: Array of table entries
+ *
+ * The EM sends down this structure in a MAD indicating
+ * the starting offset in the forwarding table that this
+ * entry is to be loaded into and the number of entries
+ * that that this MAD instance contains
+ * The mac_tbl_digest has been added to this MAD structure. It will be set by
+ * the EM and it will be used by the EM to check if there are any
+ * discrepancies with this value and the value
+ * maintained by the EM in the case of VNIC port being deleted or unloaded
+ * A new instantiation of a VNIC will always have a value of zero.
+ * This value is stored as part of the vnic adapter structure and will be
+ * accessed by the GET and SET routines for both the mactable entries and the
+ * veswport info.
+ */
+struct opa_veswport_mactable {
+ __be16 offset;
+ __be16 num_entries;
+ __be32 mac_tbl_digest;
+ struct opa_veswport_mactable_entry tbl_entries[];
+} __packed;
+
+/**
+ * struct opa_veswport_summary_counters - summary counters
+ * @vp_instance: vport instance on the OPA port
+ * @vesw_id: virtual ethernet switch id
+ * @veswport_num: virtual ethernet switch port number
+ * @tx_errors: transmit errors
+ * @rx_errors: receive errors
+ * @tx_packets: transmit packets
+ * @rx_packets: receive packets
+ * @tx_bytes: transmit bytes
+ * @rx_bytes: receive bytes
+ * @tx_unicast: unicast packets transmitted
+ * @tx_mcastbcast: multicast/broadcast packets transmitted
+ * @tx_untagged: non-vlan packets transmitted
+ * @tx_vlan: vlan packets transmitted
+ * @tx_64_size: transmit packet length is 64 bytes
+ * @tx_65_127: transmit packet length is >=65 and < 127 bytes
+ * @tx_128_255: transmit packet length is >=128 and < 255 bytes
+ * @tx_256_511: transmit packet length is >=256 and < 511 bytes
+ * @tx_512_1023: transmit packet length is >=512 and < 1023 bytes
+ * @tx_1024_1518: transmit packet length is >=1024 and < 1518 bytes
+ * @tx_1519_max: transmit packet length >= 1519 bytes
+ * @rx_unicast: unicast packets received
+ * @rx_mcastbcast: multicast/broadcast packets received
+ * @rx_untagged: non-vlan packets received
+ * @rx_vlan: vlan packets received
+ * @rx_64_size: received packet length is 64 bytes
+ * @rx_65_127: received packet length is >=65 and < 127 bytes
+ * @rx_128_255: received packet length is >=128 and < 255 bytes
+ * @rx_256_511: received packet length is >=256 and < 511 bytes
+ * @rx_512_1023: received packet length is >=512 and < 1023 bytes
+ * @rx_1024_1518: received packet length is >=1024 and < 1518 bytes
+ * @rx_1519_max: received packet length >= 1519 bytes
+ * @reserved: reserved bytes
+ *
+ * All the above are counters of corresponding conditions.
+ */
+struct opa_veswport_summary_counters {
+ __be16 vp_instance;
+ __be16 vesw_id;
+ __be32 veswport_num;
+
+ __be64 tx_errors;
+ __be64 rx_errors;
+ __be64 tx_packets;
+ __be64 rx_packets;
+ __be64 tx_bytes;
+ __be64 rx_bytes;
+
+ __be64 tx_unicast;
+ __be64 tx_mcastbcast;
+
+ __be64 tx_untagged;
+ __be64 tx_vlan;
+
+ __be64 tx_64_size;
+ __be64 tx_65_127;
+ __be64 tx_128_255;
+ __be64 tx_256_511;
+ __be64 tx_512_1023;
+ __be64 tx_1024_1518;
+ __be64 tx_1519_max;
+
+ __be64 rx_unicast;
+ __be64 rx_mcastbcast;
+
+ __be64 rx_untagged;
+ __be64 rx_vlan;
+
+ __be64 rx_64_size;
+ __be64 rx_65_127;
+ __be64 rx_128_255;
+ __be64 rx_256_511;
+ __be64 rx_512_1023;
+ __be64 rx_1024_1518;
+ __be64 rx_1519_max;
+
+ __be64 reserved[16];
+} __packed;
+
+/**
+ * struct opa_veswport_error_counters - error counters
+ * @vp_instance: vport instance on the OPA port
+ * @vesw_id: virtual ethernet switch id
+ * @veswport_num: virtual ethernet switch port number
+ * @tx_errors: transmit errors
+ * @rx_errors: receive errors
+ * @rsvd0: reserved bytes
+ * @tx_smac_filt: smac filter errors
+ * @rsvd1: reserved bytes
+ * @rsvd2: reserved bytes
+ * @rsvd3: reserved bytes
+ * @tx_dlid_zero: transmit packets with invalid dlid
+ * @rsvd4: reserved bytes
+ * @tx_logic: other transmit errors
+ * @rsvd5: reserved bytes
+ * @tx_drop_state: packet tansmission in non-forward port state
+ * @rx_bad_veswid: received packet with invalid vesw id
+ * @rsvd6: reserved bytes
+ * @rx_runt: received ethernet packet with length < 64 bytes
+ * @rx_oversize: received ethernet packet with length > MTU size
+ * @rsvd7: reserved bytes
+ * @rx_eth_down: received packets when interface is down
+ * @rx_drop_state: received packets in non-forwarding port state
+ * @rx_logic: other receive errors
+ * @rsvd8: reserved bytes
+ * @rsvd9: reserved bytes
+ *
+ * All the above are counters of corresponding error conditions.
+ */
+struct opa_veswport_error_counters {
+ __be16 vp_instance;
+ __be16 vesw_id;
+ __be32 veswport_num;
+
+ __be64 tx_errors;
+ __be64 rx_errors;
+
+ __be64 rsvd0;
+ __be64 tx_smac_filt;
+ __be64 rsvd1;
+ __be64 rsvd2;
+ __be64 rsvd3;
+ __be64 tx_dlid_zero;
+ __be64 rsvd4;
+ __be64 tx_logic;
+ __be64 rsvd5;
+ __be64 tx_drop_state;
+
+ __be64 rx_bad_veswid;
+ __be64 rsvd6;
+ __be64 rx_runt;
+ __be64 rx_oversize;
+ __be64 rsvd7;
+ __be64 rx_eth_down;
+ __be64 rx_drop_state;
+ __be64 rx_logic;
+ __be64 rsvd8;
+
+ __be64 rsvd9[16];
+} __packed;
+
+/**
+ * struct opa_veswport_trap - Trap message sent to EM by VNIC
+ * @fabric_id: 10 bit fabric id
+ * @veswid: 12 bit virtual ethernet switch id
+ * @veswportnum: logical port number on the Virtual switch
+ * @opaportnum: physical port num (redundant on host)
+ * @veswportindex: switch port index on opa port 0 based
+ * @opcode: operation
+ * @reserved: 32 bit for alignment
+ *
+ * The VNIC will send trap messages to the Ethernet manager to
+ * inform it about changes to the VNIC config, behaviour etc.
+ * This is the format of the trap payload.
+ */
+struct opa_veswport_trap {
+ __be16 fabric_id;
+ __be16 veswid;
+ __be32 veswportnum;
+ __be16 opaportnum;
+ u8 veswportindex;
+ u8 opcode;
+ __be32 reserved;
+} __packed;
+
+/**
+ * struct opa_vnic_iface_mac_entry - single entry in the mac list
+ * @mac_addr: MAC address
+ */
+struct opa_vnic_iface_mac_entry {
+ u8 mac_addr[ETH_ALEN];
+};
+
+/**
+ * struct opa_veswport_iface_macs - Msg to set globally administered MAC
+ * @start_idx: position of first entry (0 based)
+ * @num_macs_in_msg: number of MACs in this message
+ * @tot_macs_in_lst: The total number of MACs the agent has
+ * @gen_count: gen_count to indicate change
+ * @entry: The mac list entry
+ *
+ * Same attribute IDS and attribute modifiers as in locally administered
+ * addresses used to set globally administered addresses
+ */
+struct opa_veswport_iface_macs {
+ __be16 start_idx;
+ __be16 num_macs_in_msg;
+ __be16 tot_macs_in_lst;
+ __be16 gen_count;
+ struct opa_vnic_iface_mac_entry entry[];
+} __packed;
+
+/**
+ * struct opa_vnic_vema_mad - Generic VEMA MAD
+ * @mad_hdr: Generic MAD header
+ * @rmpp_hdr: RMPP header for vendor specific MADs
+ * @reserved: reserved bytes
+ * @oui: Unique org identifier
+ * @data: MAD data
+ */
+struct opa_vnic_vema_mad {
+ struct ib_mad_hdr mad_hdr;
+ struct ib_rmpp_hdr rmpp_hdr;
+ u8 reserved;
+ u8 oui[3];
+ u8 data[OPA_VNIC_EMA_DATA];
+};
+
+/**
+ * struct opa_vnic_notice_attr - Generic Notice MAD
+ * @gen_type: Generic/Specific bit and type of notice
+ * @oui_1: Vendor ID byte 1
+ * @oui_2: Vendor ID byte 2
+ * @oui_3: Vendor ID byte 3
+ * @trap_num: Trap number
+ * @toggle_count: Notice toggle bit and count value
+ * @issuer_lid: Trap issuer's lid
+ * @reserved: reserved bytes
+ * @issuer_gid: Issuer GID (only if Report method)
+ * @raw_data: Trap message body
+ */
+struct opa_vnic_notice_attr {
+ u8 gen_type;
+ u8 oui_1;
+ u8 oui_2;
+ u8 oui_3;
+ __be16 trap_num;
+ __be16 toggle_count;
+ __be32 issuer_lid;
+ __be32 reserved;
+ u8 issuer_gid[16];
+ u8 raw_data[64];
+} __packed;
+
+/**
+ * struct opa_vnic_vema_mad_trap - Generic VEMA MAD Trap
+ * @mad_hdr: Generic MAD header
+ * @rmpp_hdr: RMPP header for vendor specific MADs
+ * @reserved: reserved bytes
+ * @oui: Unique org identifier
+ * @notice: Notice structure
+ */
+struct opa_vnic_vema_mad_trap {
+ struct ib_mad_hdr mad_hdr;
+ struct ib_rmpp_hdr rmpp_hdr;
+ u8 reserved;
+ u8 oui[3];
+ struct opa_vnic_notice_attr notice;
+};
+
+#endif /* _OPA_VNIC_ENCAP_H */
diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c
new file mode 100644
index 000000000..29b3d8fce
--- /dev/null
+++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c
@@ -0,0 +1,185 @@
+/*
+ * Copyright(c) 2017 Intel Corporation.
+ *
+ * This file is provided under a dual BSD/GPLv2 license. When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * 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 GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * 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.
+ * - Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+/*
+ * This file contains OPA VNIC ethtool functions
+ */
+
+#include <linux/ethtool.h>
+
+#include "opa_vnic_internal.h"
+
+enum {NETDEV_STATS, VNIC_STATS};
+
+struct vnic_stats {
+ char stat_string[ETH_GSTRING_LEN];
+ struct {
+ int sizeof_stat;
+ int stat_offset;
+ };
+};
+
+#define VNIC_STAT(m) { sizeof_field(struct opa_vnic_stats, m), \
+ offsetof(struct opa_vnic_stats, m) }
+
+static struct vnic_stats vnic_gstrings_stats[] = {
+ /* NETDEV stats */
+ {"rx_packets", VNIC_STAT(netstats.rx_packets)},
+ {"tx_packets", VNIC_STAT(netstats.tx_packets)},
+ {"rx_bytes", VNIC_STAT(netstats.rx_bytes)},
+ {"tx_bytes", VNIC_STAT(netstats.tx_bytes)},
+ {"rx_errors", VNIC_STAT(netstats.rx_errors)},
+ {"tx_errors", VNIC_STAT(netstats.tx_errors)},
+ {"rx_dropped", VNIC_STAT(netstats.rx_dropped)},
+ {"tx_dropped", VNIC_STAT(netstats.tx_dropped)},
+
+ /* SUMMARY counters */
+ {"tx_unicast", VNIC_STAT(tx_grp.unicast)},
+ {"tx_mcastbcast", VNIC_STAT(tx_grp.mcastbcast)},
+ {"tx_untagged", VNIC_STAT(tx_grp.untagged)},
+ {"tx_vlan", VNIC_STAT(tx_grp.vlan)},
+
+ {"tx_64_size", VNIC_STAT(tx_grp.s_64)},
+ {"tx_65_127", VNIC_STAT(tx_grp.s_65_127)},
+ {"tx_128_255", VNIC_STAT(tx_grp.s_128_255)},
+ {"tx_256_511", VNIC_STAT(tx_grp.s_256_511)},
+ {"tx_512_1023", VNIC_STAT(tx_grp.s_512_1023)},
+ {"tx_1024_1518", VNIC_STAT(tx_grp.s_1024_1518)},
+ {"tx_1519_max", VNIC_STAT(tx_grp.s_1519_max)},
+
+ {"rx_unicast", VNIC_STAT(rx_grp.unicast)},
+ {"rx_mcastbcast", VNIC_STAT(rx_grp.mcastbcast)},
+ {"rx_untagged", VNIC_STAT(rx_grp.untagged)},
+ {"rx_vlan", VNIC_STAT(rx_grp.vlan)},
+
+ {"rx_64_size", VNIC_STAT(rx_grp.s_64)},
+ {"rx_65_127", VNIC_STAT(rx_grp.s_65_127)},
+ {"rx_128_255", VNIC_STAT(rx_grp.s_128_255)},
+ {"rx_256_511", VNIC_STAT(rx_grp.s_256_511)},
+ {"rx_512_1023", VNIC_STAT(rx_grp.s_512_1023)},
+ {"rx_1024_1518", VNIC_STAT(rx_grp.s_1024_1518)},
+ {"rx_1519_max", VNIC_STAT(rx_grp.s_1519_max)},
+
+ /* ERROR counters */
+ {"rx_fifo_errors", VNIC_STAT(netstats.rx_fifo_errors)},
+ {"rx_length_errors", VNIC_STAT(netstats.rx_length_errors)},
+
+ {"tx_fifo_errors", VNIC_STAT(netstats.tx_fifo_errors)},
+ {"tx_carrier_errors", VNIC_STAT(netstats.tx_carrier_errors)},
+
+ {"tx_dlid_zero", VNIC_STAT(tx_dlid_zero)},
+ {"tx_drop_state", VNIC_STAT(tx_drop_state)},
+ {"rx_drop_state", VNIC_STAT(rx_drop_state)},
+ {"rx_oversize", VNIC_STAT(rx_oversize)},
+ {"rx_runt", VNIC_STAT(rx_runt)},
+};
+
+#define VNIC_STATS_LEN ARRAY_SIZE(vnic_gstrings_stats)
+
+/* vnic_get_drvinfo - get driver info */
+static void vnic_get_drvinfo(struct net_device *netdev,
+ struct ethtool_drvinfo *drvinfo)
+{
+ strscpy(drvinfo->driver, opa_vnic_driver_name, sizeof(drvinfo->driver));
+ strscpy(drvinfo->bus_info, dev_name(netdev->dev.parent),
+ sizeof(drvinfo->bus_info));
+}
+
+/* vnic_get_sset_count - get string set count */
+static int vnic_get_sset_count(struct net_device *netdev, int sset)
+{
+ return (sset == ETH_SS_STATS) ? VNIC_STATS_LEN : -EOPNOTSUPP;
+}
+
+/* vnic_get_ethtool_stats - get statistics */
+static void vnic_get_ethtool_stats(struct net_device *netdev,
+ struct ethtool_stats *stats, u64 *data)
+{
+ struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev);
+ struct opa_vnic_stats vstats;
+ int i;
+
+ memset(&vstats, 0, sizeof(vstats));
+ spin_lock(&adapter->stats_lock);
+ adapter->rn_ops->ndo_get_stats64(netdev, &vstats.netstats);
+ spin_unlock(&adapter->stats_lock);
+ for (i = 0; i < VNIC_STATS_LEN; i++) {
+ char *p = (char *)&vstats + vnic_gstrings_stats[i].stat_offset;
+
+ data[i] = (vnic_gstrings_stats[i].sizeof_stat ==
+ sizeof(u64)) ? *(u64 *)p : *(u32 *)p;
+ }
+}
+
+/* vnic_get_strings - get strings */
+static void vnic_get_strings(struct net_device *netdev, u32 stringset, u8 *data)
+{
+ int i;
+
+ if (stringset != ETH_SS_STATS)
+ return;
+
+ for (i = 0; i < VNIC_STATS_LEN; i++)
+ memcpy(data + i * ETH_GSTRING_LEN,
+ vnic_gstrings_stats[i].stat_string,
+ ETH_GSTRING_LEN);
+}
+
+/* ethtool ops */
+static const struct ethtool_ops opa_vnic_ethtool_ops = {
+ .get_drvinfo = vnic_get_drvinfo,
+ .get_link = ethtool_op_get_link,
+ .get_strings = vnic_get_strings,
+ .get_sset_count = vnic_get_sset_count,
+ .get_ethtool_stats = vnic_get_ethtool_stats,
+};
+
+/* opa_vnic_set_ethtool_ops - set ethtool ops */
+void opa_vnic_set_ethtool_ops(struct net_device *netdev)
+{
+ netdev->ethtool_ops = &opa_vnic_ethtool_ops;
+}
diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h b/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h
new file mode 100644
index 000000000..dd942dd64
--- /dev/null
+++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h
@@ -0,0 +1,329 @@
+#ifndef _OPA_VNIC_INTERNAL_H
+#define _OPA_VNIC_INTERNAL_H
+/*
+ * Copyright(c) 2017 Intel Corporation.
+ *
+ * This file is provided under a dual BSD/GPLv2 license. When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * 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 GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * 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.
+ * - Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+/*
+ * This file contains OPA VNIC driver internal declarations
+ */
+
+#include <linux/bitops.h>
+#include <linux/etherdevice.h>
+#include <linux/hashtable.h>
+#include <linux/sizes.h>
+#include <rdma/opa_vnic.h>
+
+#include "opa_vnic_encap.h"
+
+#define OPA_VNIC_VLAN_PCP(vlan_tci) \
+ (((vlan_tci) & VLAN_PRIO_MASK) >> VLAN_PRIO_SHIFT)
+
+/* Flow to default port redirection table size */
+#define OPA_VNIC_FLOW_TBL_SIZE 32
+
+/* Invalid port number */
+#define OPA_VNIC_INVALID_PORT 0xff
+
+struct opa_vnic_adapter;
+
+/*
+ * struct __opa_vesw_info - OPA vnic virtual switch info
+ *
+ * Same as opa_vesw_info without bitwise attribute.
+ */
+struct __opa_vesw_info {
+ u16 fabric_id;
+ u16 vesw_id;
+
+ u8 rsvd0[6];
+ u16 def_port_mask;
+
+ u8 rsvd1[2];
+ u16 pkey;
+
+ u8 rsvd2[4];
+ u32 u_mcast_dlid;
+ u32 u_ucast_dlid[OPA_VESW_MAX_NUM_DEF_PORT];
+
+ u32 rc;
+
+ u8 rsvd3[56];
+ u16 eth_mtu;
+ u8 rsvd4[2];
+} __packed;
+
+/*
+ * struct __opa_per_veswport_info - OPA vnic per port info
+ *
+ * Same as opa_per_veswport_info without bitwise attribute.
+ */
+struct __opa_per_veswport_info {
+ u32 port_num;
+
+ u8 eth_link_status;
+ u8 rsvd0[3];
+
+ u8 base_mac_addr[ETH_ALEN];
+ u8 config_state;
+ u8 oper_state;
+
+ u16 max_mac_tbl_ent;
+ u16 max_smac_ent;
+ u32 mac_tbl_digest;
+ u8 rsvd1[4];
+
+ u32 encap_slid;
+
+ u8 pcp_to_sc_uc[OPA_VNIC_MAX_NUM_PCP];
+ u8 pcp_to_vl_uc[OPA_VNIC_MAX_NUM_PCP];
+ u8 pcp_to_sc_mc[OPA_VNIC_MAX_NUM_PCP];
+ u8 pcp_to_vl_mc[OPA_VNIC_MAX_NUM_PCP];
+
+ u8 non_vlan_sc_uc;
+ u8 non_vlan_vl_uc;
+ u8 non_vlan_sc_mc;
+ u8 non_vlan_vl_mc;
+
+ u8 rsvd2[48];
+
+ u16 uc_macs_gen_count;
+ u16 mc_macs_gen_count;
+
+ u8 rsvd3[8];
+} __packed;
+
+/*
+ * struct __opa_veswport_info - OPA vnic port info
+ *
+ * Same as opa_veswport_info without bitwise attribute.
+ */
+struct __opa_veswport_info {
+ struct __opa_vesw_info vesw;
+ struct __opa_per_veswport_info vport;
+};
+
+/*
+ * struct __opa_veswport_trap - OPA vnic trap info
+ *
+ * Same as opa_veswport_trap without bitwise attribute.
+ */
+struct __opa_veswport_trap {
+ u16 fabric_id;
+ u16 veswid;
+ u32 veswportnum;
+ u16 opaportnum;
+ u8 veswportindex;
+ u8 opcode;
+ u32 reserved;
+} __packed;
+
+/**
+ * struct opa_vnic_ctrl_port - OPA virtual NIC control port
+ * @ibdev: pointer to ib device
+ * @ops: opa vnic control operations
+ * @num_ports: number of opa ports
+ */
+struct opa_vnic_ctrl_port {
+ struct ib_device *ibdev;
+ struct opa_vnic_ctrl_ops *ops;
+ u8 num_ports;
+};
+
+/**
+ * struct opa_vnic_adapter - OPA VNIC netdev private data structure
+ * @netdev: pointer to associated netdev
+ * @ibdev: ib device
+ * @cport: pointer to opa vnic control port
+ * @rn_ops: rdma netdev's net_device_ops
+ * @port_num: OPA port number
+ * @vport_num: vesw port number
+ * @lock: adapter lock
+ * @info: virtual ethernet switch port information
+ * @vema_mac_addr: mac address configured by vema
+ * @umac_hash: unicast maclist hash
+ * @mmac_hash: multicast maclist hash
+ * @mactbl: hash table of MAC entries
+ * @mactbl_lock: mac table lock
+ * @stats_lock: statistics lock
+ * @flow_tbl: flow to default port redirection table
+ * @trap_timeout: trap timeout
+ * @trap_count: no. of traps allowed within timeout period
+ */
+struct opa_vnic_adapter {
+ struct net_device *netdev;
+ struct ib_device *ibdev;
+ struct opa_vnic_ctrl_port *cport;
+ const struct net_device_ops *rn_ops;
+
+ u8 port_num;
+ u8 vport_num;
+
+ /* Lock used around concurrent updates to netdev */
+ struct mutex lock;
+
+ struct __opa_veswport_info info;
+ u8 vema_mac_addr[ETH_ALEN];
+ u32 umac_hash;
+ u32 mmac_hash;
+ struct hlist_head __rcu *mactbl;
+
+ /* Lock used to protect updates to mac table */
+ struct mutex mactbl_lock;
+
+ /* Lock used to protect access to vnic counters */
+ spinlock_t stats_lock;
+
+ u8 flow_tbl[OPA_VNIC_FLOW_TBL_SIZE];
+
+ unsigned long trap_timeout;
+ u8 trap_count;
+};
+
+/* Same as opa_veswport_mactable_entry, but without bitwise attribute */
+struct __opa_vnic_mactable_entry {
+ u8 mac_addr[ETH_ALEN];
+ u8 mac_addr_mask[ETH_ALEN];
+ u32 dlid_sd;
+} __packed;
+
+/**
+ * struct opa_vnic_mac_tbl_node - OPA VNIC mac table node
+ * @hlist: hash list handle
+ * @index: index of entry in the mac table
+ * @entry: entry in the table
+ */
+struct opa_vnic_mac_tbl_node {
+ struct hlist_node hlist;
+ u16 index;
+ struct __opa_vnic_mactable_entry entry;
+};
+
+#define v_dbg(format, arg...) \
+ netdev_dbg(adapter->netdev, format, ## arg)
+#define v_err(format, arg...) \
+ netdev_err(adapter->netdev, format, ## arg)
+#define v_info(format, arg...) \
+ netdev_info(adapter->netdev, format, ## arg)
+#define v_warn(format, arg...) \
+ netdev_warn(adapter->netdev, format, ## arg)
+
+#define c_err(format, arg...) \
+ dev_err(&cport->ibdev->dev, format, ## arg)
+#define c_info(format, arg...) \
+ dev_info(&cport->ibdev->dev, format, ## arg)
+#define c_dbg(format, arg...) \
+ dev_dbg(&cport->ibdev->dev, format, ## arg)
+
+/* The maximum allowed entries in the mac table */
+#define OPA_VNIC_MAC_TBL_MAX_ENTRIES 2048
+/* Limit of smac entries in mac table */
+#define OPA_VNIC_MAX_SMAC_LIMIT 256
+
+/* The last octet of the MAC address is used as the key to the hash table */
+#define OPA_VNIC_MAC_HASH_IDX 5
+
+/* The VNIC MAC hash table is of size 2^8 */
+#define OPA_VNIC_MAC_TBL_HASH_BITS 8
+#define OPA_VNIC_MAC_TBL_SIZE BIT(OPA_VNIC_MAC_TBL_HASH_BITS)
+
+/* VNIC HASH MACROS */
+#define vnic_hash_init(hashtable) __hash_init(hashtable, OPA_VNIC_MAC_TBL_SIZE)
+
+#define vnic_hash_add(hashtable, node, key) \
+ hlist_add_head(node, \
+ &hashtable[hash_min(key, ilog2(OPA_VNIC_MAC_TBL_SIZE))])
+
+#define vnic_hash_for_each_safe(name, bkt, tmp, obj, member) \
+ for ((bkt) = 0, obj = NULL; \
+ !obj && (bkt) < OPA_VNIC_MAC_TBL_SIZE; (bkt)++) \
+ hlist_for_each_entry_safe(obj, tmp, &name[bkt], member)
+
+#define vnic_hash_for_each_possible(name, obj, member, key) \
+ hlist_for_each_entry(obj, \
+ &name[hash_min(key, ilog2(OPA_VNIC_MAC_TBL_SIZE))], member)
+
+#define vnic_hash_for_each(name, bkt, obj, member) \
+ for ((bkt) = 0, obj = NULL; \
+ !obj && (bkt) < OPA_VNIC_MAC_TBL_SIZE; (bkt)++) \
+ hlist_for_each_entry(obj, &name[bkt], member)
+
+extern char opa_vnic_driver_name[];
+
+struct opa_vnic_adapter *opa_vnic_add_netdev(struct ib_device *ibdev,
+ u8 port_num, u8 vport_num);
+void opa_vnic_rem_netdev(struct opa_vnic_adapter *adapter);
+void opa_vnic_encap_skb(struct opa_vnic_adapter *adapter, struct sk_buff *skb);
+u8 opa_vnic_get_vl(struct opa_vnic_adapter *adapter, struct sk_buff *skb);
+u8 opa_vnic_calc_entropy(struct sk_buff *skb);
+void opa_vnic_process_vema_config(struct opa_vnic_adapter *adapter);
+void opa_vnic_release_mac_tbl(struct opa_vnic_adapter *adapter);
+void opa_vnic_query_mac_tbl(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_mactable *tbl);
+int opa_vnic_update_mac_tbl(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_mactable *tbl);
+void opa_vnic_query_ucast_macs(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_iface_macs *macs);
+void opa_vnic_query_mcast_macs(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_iface_macs *macs);
+void opa_vnic_get_summary_counters(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_summary_counters *cntrs);
+void opa_vnic_get_error_counters(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_error_counters *cntrs);
+void opa_vnic_get_vesw_info(struct opa_vnic_adapter *adapter,
+ struct opa_vesw_info *info);
+void opa_vnic_set_vesw_info(struct opa_vnic_adapter *adapter,
+ struct opa_vesw_info *info);
+void opa_vnic_get_per_veswport_info(struct opa_vnic_adapter *adapter,
+ struct opa_per_veswport_info *info);
+void opa_vnic_set_per_veswport_info(struct opa_vnic_adapter *adapter,
+ struct opa_per_veswport_info *info);
+void opa_vnic_vema_report_event(struct opa_vnic_adapter *adapter, u8 event);
+void opa_vnic_set_ethtool_ops(struct net_device *netdev);
+void opa_vnic_vema_send_trap(struct opa_vnic_adapter *adapter,
+ struct __opa_veswport_trap *data, u32 lid);
+
+#endif /* _OPA_VNIC_INTERNAL_H */
diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c
new file mode 100644
index 000000000..071f35711
--- /dev/null
+++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c
@@ -0,0 +1,400 @@
+/*
+ * Copyright(c) 2017 Intel Corporation.
+ *
+ * This file is provided under a dual BSD/GPLv2 license. When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * 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 GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * 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.
+ * - Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+/*
+ * This file contains OPA Virtual Network Interface Controller (VNIC) driver
+ * netdev functionality.
+ */
+
+#include <linux/if_vlan.h>
+#include <linux/crc32.h>
+
+#include "opa_vnic_internal.h"
+
+#define OPA_TX_TIMEOUT_MS 1000
+
+#define OPA_VNIC_SKB_HEADROOM \
+ ALIGN((OPA_VNIC_HDR_LEN + OPA_VNIC_SKB_MDATA_LEN), 8)
+
+/* This function is overloaded for opa_vnic specific implementation */
+static void opa_vnic_get_stats64(struct net_device *netdev,
+ struct rtnl_link_stats64 *stats)
+{
+ struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev);
+ struct opa_vnic_stats vstats;
+
+ memset(&vstats, 0, sizeof(vstats));
+ spin_lock(&adapter->stats_lock);
+ adapter->rn_ops->ndo_get_stats64(netdev, &vstats.netstats);
+ spin_unlock(&adapter->stats_lock);
+ memcpy(stats, &vstats.netstats, sizeof(*stats));
+}
+
+/* opa_netdev_start_xmit - transmit function */
+static netdev_tx_t opa_netdev_start_xmit(struct sk_buff *skb,
+ struct net_device *netdev)
+{
+ struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev);
+
+ v_dbg("xmit: queue %d skb len %d\n", skb->queue_mapping, skb->len);
+ /* pad to ensure mininum ethernet packet length */
+ if (unlikely(skb->len < ETH_ZLEN)) {
+ if (skb_padto(skb, ETH_ZLEN))
+ return NETDEV_TX_OK;
+
+ skb_put(skb, ETH_ZLEN - skb->len);
+ }
+
+ opa_vnic_encap_skb(adapter, skb);
+ return adapter->rn_ops->ndo_start_xmit(skb, netdev);
+}
+
+static u16 opa_vnic_select_queue(struct net_device *netdev, struct sk_buff *skb,
+ struct net_device *sb_dev)
+{
+ struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev);
+ struct opa_vnic_skb_mdata *mdata;
+ int rc;
+
+ /* pass entropy and vl as metadata in skb */
+ mdata = skb_push(skb, sizeof(*mdata));
+ mdata->entropy = opa_vnic_calc_entropy(skb);
+ mdata->vl = opa_vnic_get_vl(adapter, skb);
+ rc = adapter->rn_ops->ndo_select_queue(netdev, skb, sb_dev);
+ skb_pull(skb, sizeof(*mdata));
+ return rc;
+}
+
+static void opa_vnic_update_state(struct opa_vnic_adapter *adapter, bool up)
+{
+ struct __opa_veswport_info *info = &adapter->info;
+
+ mutex_lock(&adapter->lock);
+ /* Operational state can only be DROP_ALL or FORWARDING */
+ if ((info->vport.config_state == OPA_VNIC_STATE_FORWARDING) && up) {
+ info->vport.oper_state = OPA_VNIC_STATE_FORWARDING;
+ info->vport.eth_link_status = OPA_VNIC_ETH_LINK_UP;
+ } else {
+ info->vport.oper_state = OPA_VNIC_STATE_DROP_ALL;
+ info->vport.eth_link_status = OPA_VNIC_ETH_LINK_DOWN;
+ }
+
+ if (info->vport.config_state == OPA_VNIC_STATE_FORWARDING)
+ netif_dormant_off(adapter->netdev);
+ else
+ netif_dormant_on(adapter->netdev);
+ mutex_unlock(&adapter->lock);
+}
+
+/* opa_vnic_process_vema_config - process vema configuration updates */
+void opa_vnic_process_vema_config(struct opa_vnic_adapter *adapter)
+{
+ struct __opa_veswport_info *info = &adapter->info;
+ struct rdma_netdev *rn = netdev_priv(adapter->netdev);
+ u8 port_num[OPA_VESW_MAX_NUM_DEF_PORT] = { 0 };
+ struct net_device *netdev = adapter->netdev;
+ u8 i, port_count = 0;
+ u16 port_mask;
+
+ /* If the base_mac_addr is changed, update the interface mac address */
+ if (memcmp(info->vport.base_mac_addr, adapter->vema_mac_addr,
+ ARRAY_SIZE(info->vport.base_mac_addr))) {
+ struct sockaddr saddr;
+
+ memcpy(saddr.sa_data, info->vport.base_mac_addr,
+ ARRAY_SIZE(info->vport.base_mac_addr));
+ mutex_lock(&adapter->lock);
+ eth_commit_mac_addr_change(netdev, &saddr);
+ memcpy(adapter->vema_mac_addr,
+ info->vport.base_mac_addr, ETH_ALEN);
+ mutex_unlock(&adapter->lock);
+ }
+
+ rn->set_id(netdev, info->vesw.vesw_id);
+
+ /* Handle MTU limit change */
+ rtnl_lock();
+ netdev->max_mtu = max_t(unsigned int, info->vesw.eth_mtu,
+ netdev->min_mtu);
+ if (netdev->mtu > netdev->max_mtu)
+ dev_set_mtu(netdev, netdev->max_mtu);
+ rtnl_unlock();
+
+ /* Update flow to default port redirection table */
+ port_mask = info->vesw.def_port_mask;
+ for (i = 0; i < OPA_VESW_MAX_NUM_DEF_PORT; i++) {
+ if (port_mask & 1)
+ port_num[port_count++] = i;
+ port_mask >>= 1;
+ }
+
+ /*
+ * Build the flow table. Flow table is required when destination LID
+ * is not available. Up to OPA_VNIC_FLOW_TBL_SIZE flows supported.
+ * Each flow need a default port number to get its dlid from the
+ * u_ucast_dlid array.
+ */
+ for (i = 0; i < OPA_VNIC_FLOW_TBL_SIZE; i++)
+ adapter->flow_tbl[i] = port_count ? port_num[i % port_count] :
+ OPA_VNIC_INVALID_PORT;
+
+ /* update state */
+ opa_vnic_update_state(adapter, !!(netdev->flags & IFF_UP));
+}
+
+/*
+ * Set the power on default values in adapter's vema interface structure.
+ */
+static inline void opa_vnic_set_pod_values(struct opa_vnic_adapter *adapter)
+{
+ adapter->info.vport.max_mac_tbl_ent = OPA_VNIC_MAC_TBL_MAX_ENTRIES;
+ adapter->info.vport.max_smac_ent = OPA_VNIC_MAX_SMAC_LIMIT;
+ adapter->info.vport.config_state = OPA_VNIC_STATE_DROP_ALL;
+ adapter->info.vport.eth_link_status = OPA_VNIC_ETH_LINK_DOWN;
+ adapter->info.vesw.eth_mtu = ETH_DATA_LEN;
+}
+
+/* opa_vnic_set_mac_addr - change mac address */
+static int opa_vnic_set_mac_addr(struct net_device *netdev, void *addr)
+{
+ struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev);
+ struct sockaddr *sa = addr;
+ int rc;
+
+ if (!memcmp(netdev->dev_addr, sa->sa_data, ETH_ALEN))
+ return 0;
+
+ mutex_lock(&adapter->lock);
+ rc = eth_mac_addr(netdev, addr);
+ mutex_unlock(&adapter->lock);
+ if (rc)
+ return rc;
+
+ adapter->info.vport.uc_macs_gen_count++;
+ opa_vnic_vema_report_event(adapter,
+ OPA_VESWPORT_TRAP_IFACE_UCAST_MAC_CHANGE);
+ return 0;
+}
+
+/*
+ * opa_vnic_mac_send_event - post event on possible mac list exchange
+ * Send trap when digest from uc/mc mac list differs from previous run.
+ * Digest is evaluated similar to how cksum does.
+ */
+static void opa_vnic_mac_send_event(struct net_device *netdev, u8 event)
+{
+ struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev);
+ struct netdev_hw_addr *ha;
+ struct netdev_hw_addr_list *hw_list;
+ u32 *ref_crc;
+ u32 l, crc = 0;
+
+ switch (event) {
+ case OPA_VESWPORT_TRAP_IFACE_UCAST_MAC_CHANGE:
+ hw_list = &netdev->uc;
+ adapter->info.vport.uc_macs_gen_count++;
+ ref_crc = &adapter->umac_hash;
+ break;
+ case OPA_VESWPORT_TRAP_IFACE_MCAST_MAC_CHANGE:
+ hw_list = &netdev->mc;
+ adapter->info.vport.mc_macs_gen_count++;
+ ref_crc = &adapter->mmac_hash;
+ break;
+ default:
+ return;
+ }
+ netdev_hw_addr_list_for_each(ha, hw_list) {
+ crc = crc32_le(crc, ha->addr, ETH_ALEN);
+ }
+ l = netdev_hw_addr_list_count(hw_list) * ETH_ALEN;
+ crc = ~crc32_le(crc, (void *)&l, sizeof(l));
+
+ if (crc != *ref_crc) {
+ *ref_crc = crc;
+ opa_vnic_vema_report_event(adapter, event);
+ }
+}
+
+/* opa_vnic_set_rx_mode - handle uc/mc mac list change */
+static void opa_vnic_set_rx_mode(struct net_device *netdev)
+{
+ opa_vnic_mac_send_event(netdev,
+ OPA_VESWPORT_TRAP_IFACE_UCAST_MAC_CHANGE);
+
+ opa_vnic_mac_send_event(netdev,
+ OPA_VESWPORT_TRAP_IFACE_MCAST_MAC_CHANGE);
+}
+
+/* opa_netdev_open - activate network interface */
+static int opa_netdev_open(struct net_device *netdev)
+{
+ struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev);
+ int rc;
+
+ rc = adapter->rn_ops->ndo_open(adapter->netdev);
+ if (rc) {
+ v_dbg("open failed %d\n", rc);
+ return rc;
+ }
+
+ /* Update status and send trap */
+ opa_vnic_update_state(adapter, true);
+ opa_vnic_vema_report_event(adapter,
+ OPA_VESWPORT_TRAP_ETH_LINK_STATUS_CHANGE);
+ return 0;
+}
+
+/* opa_netdev_close - disable network interface */
+static int opa_netdev_close(struct net_device *netdev)
+{
+ struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev);
+ int rc;
+
+ rc = adapter->rn_ops->ndo_stop(adapter->netdev);
+ if (rc) {
+ v_dbg("close failed %d\n", rc);
+ return rc;
+ }
+
+ /* Update status and send trap */
+ opa_vnic_update_state(adapter, false);
+ opa_vnic_vema_report_event(adapter,
+ OPA_VESWPORT_TRAP_ETH_LINK_STATUS_CHANGE);
+ return 0;
+}
+
+/* netdev ops */
+static const struct net_device_ops opa_netdev_ops = {
+ .ndo_open = opa_netdev_open,
+ .ndo_stop = opa_netdev_close,
+ .ndo_start_xmit = opa_netdev_start_xmit,
+ .ndo_get_stats64 = opa_vnic_get_stats64,
+ .ndo_set_rx_mode = opa_vnic_set_rx_mode,
+ .ndo_select_queue = opa_vnic_select_queue,
+ .ndo_set_mac_address = opa_vnic_set_mac_addr,
+};
+
+/* opa_vnic_add_netdev - create vnic netdev interface */
+struct opa_vnic_adapter *opa_vnic_add_netdev(struct ib_device *ibdev,
+ u8 port_num, u8 vport_num)
+{
+ struct opa_vnic_adapter *adapter;
+ struct net_device *netdev;
+ struct rdma_netdev *rn;
+ int rc;
+
+ netdev = ibdev->ops.alloc_rdma_netdev(ibdev, port_num,
+ RDMA_NETDEV_OPA_VNIC,
+ "veth%d", NET_NAME_UNKNOWN,
+ ether_setup);
+ if (!netdev)
+ return ERR_PTR(-ENOMEM);
+ else if (IS_ERR(netdev))
+ return ERR_CAST(netdev);
+
+ rn = netdev_priv(netdev);
+ adapter = kzalloc(sizeof(*adapter), GFP_KERNEL);
+ if (!adapter) {
+ rc = -ENOMEM;
+ goto adapter_err;
+ }
+
+ rn->clnt_priv = adapter;
+ rn->hca = ibdev;
+ rn->port_num = port_num;
+ adapter->netdev = netdev;
+ adapter->ibdev = ibdev;
+ adapter->port_num = port_num;
+ adapter->vport_num = vport_num;
+ adapter->rn_ops = netdev->netdev_ops;
+
+ netdev->netdev_ops = &opa_netdev_ops;
+ netdev->priv_flags |= IFF_LIVE_ADDR_CHANGE;
+ netdev->hard_header_len += OPA_VNIC_SKB_HEADROOM;
+ mutex_init(&adapter->lock);
+ mutex_init(&adapter->mactbl_lock);
+ spin_lock_init(&adapter->stats_lock);
+
+ SET_NETDEV_DEV(netdev, ibdev->dev.parent);
+
+ opa_vnic_set_ethtool_ops(netdev);
+
+ opa_vnic_set_pod_values(adapter);
+
+ rc = register_netdev(netdev);
+ if (rc)
+ goto netdev_err;
+
+ netif_carrier_off(netdev);
+ netif_dormant_on(netdev);
+ v_info("initialized\n");
+
+ return adapter;
+netdev_err:
+ mutex_destroy(&adapter->lock);
+ mutex_destroy(&adapter->mactbl_lock);
+ kfree(adapter);
+adapter_err:
+ rn->free_rdma_netdev(netdev);
+
+ return ERR_PTR(rc);
+}
+
+/* opa_vnic_rem_netdev - remove vnic netdev interface */
+void opa_vnic_rem_netdev(struct opa_vnic_adapter *adapter)
+{
+ struct net_device *netdev = adapter->netdev;
+ struct rdma_netdev *rn = netdev_priv(netdev);
+
+ v_info("removing\n");
+ unregister_netdev(netdev);
+ opa_vnic_release_mac_tbl(adapter);
+ mutex_destroy(&adapter->lock);
+ mutex_destroy(&adapter->mactbl_lock);
+ kfree(adapter);
+ rn->free_rdma_netdev(netdev);
+}
diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema.c
new file mode 100644
index 000000000..21c6cea8b
--- /dev/null
+++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema.c
@@ -0,0 +1,1056 @@
+/*
+ * Copyright(c) 2017 Intel Corporation.
+ * Copyright(c) 2021 Cornelis Networks.
+ *
+ * This file is provided under a dual BSD/GPLv2 license. When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * 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 GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * 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.
+ * - Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+/*
+ * This file contains OPX Virtual Network Interface Controller (VNIC)
+ * Ethernet Management Agent (EMA) driver
+ */
+
+#include <linux/module.h>
+#include <linux/xarray.h>
+#include <rdma/ib_addr.h>
+#include <rdma/ib_verbs.h>
+#include <rdma/opa_smi.h>
+#include <rdma/opa_port_info.h>
+
+#include "opa_vnic_internal.h"
+
+char opa_vnic_driver_name[] = "opa_vnic";
+
+/*
+ * The trap service level is kept in bits 3 to 7 in the trap_sl_rsvd
+ * field in the class port info MAD.
+ */
+#define GET_TRAP_SL_FROM_CLASS_PORT_INFO(x) (((x) >> 3) & 0x1f)
+
+/* Cap trap bursts to a reasonable limit good for normal cases */
+#define OPA_VNIC_TRAP_BURST_LIMIT 4
+
+/*
+ * VNIC trap limit timeout.
+ * Inverse of cap2_mask response time out (1.0737 secs) = 0.9
+ * secs approx IB spec 13.4.6.2.1 PortInfoSubnetTimeout and
+ * 13.4.9 Traps.
+ */
+#define OPA_VNIC_TRAP_TIMEOUT ((4096 * (1UL << 18)) / 1000)
+
+#define OPA_VNIC_UNSUP_ATTR \
+ cpu_to_be16(IB_MGMT_MAD_STATUS_UNSUPPORTED_METHOD_ATTRIB)
+
+#define OPA_VNIC_INVAL_ATTR \
+ cpu_to_be16(IB_MGMT_MAD_STATUS_INVALID_ATTRIB_VALUE)
+
+#define OPA_VNIC_CLASS_CAP_TRAP 0x1
+
+/* Maximum number of VNIC ports supported */
+#define OPA_VNIC_MAX_NUM_VPORT 255
+
+/**
+ * struct opa_vnic_vema_port -- VNIC VEMA port details
+ * @cport: pointer to port
+ * @mad_agent: pointer to mad agent for port
+ * @class_port_info: Class port info information.
+ * @tid: Transaction id
+ * @port_num: OPA port number
+ * @vports: vnic ports
+ * @event_handler: ib event handler
+ * @lock: adapter interface lock
+ */
+struct opa_vnic_vema_port {
+ struct opa_vnic_ctrl_port *cport;
+ struct ib_mad_agent *mad_agent;
+ struct opa_class_port_info class_port_info;
+ u64 tid;
+ u8 port_num;
+ struct xarray vports;
+ struct ib_event_handler event_handler;
+
+ /* Lock to query/update network adapter */
+ struct mutex lock;
+};
+
+static int opa_vnic_vema_add_one(struct ib_device *device);
+static void opa_vnic_vema_rem_one(struct ib_device *device,
+ void *client_data);
+
+static struct ib_client opa_vnic_client = {
+ .name = opa_vnic_driver_name,
+ .add = opa_vnic_vema_add_one,
+ .remove = opa_vnic_vema_rem_one,
+};
+
+/**
+ * vema_get_vport_num -- Get the vnic from the mad
+ * @recvd_mad: Received mad
+ *
+ * Return: returns value of the vnic port number
+ */
+static inline u8 vema_get_vport_num(struct opa_vnic_vema_mad *recvd_mad)
+{
+ return be32_to_cpu(recvd_mad->mad_hdr.attr_mod) & 0xff;
+}
+
+/**
+ * vema_get_vport_adapter -- Get vnic port adapter from recvd mad
+ * @recvd_mad: received mad
+ * @port: ptr to port struct on which MAD was recvd
+ *
+ * Return: vnic adapter
+ */
+static inline struct opa_vnic_adapter *
+vema_get_vport_adapter(struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_port *port)
+{
+ u8 vport_num = vema_get_vport_num(recvd_mad);
+
+ return xa_load(&port->vports, vport_num);
+}
+
+/**
+ * vema_mac_tbl_req_ok -- Check if mac request has correct values
+ * @mac_tbl: mac table
+ *
+ * This function checks for the validity of the offset and number of
+ * entries required.
+ *
+ * Return: true if offset and num_entries are valid
+ */
+static inline bool vema_mac_tbl_req_ok(struct opa_veswport_mactable *mac_tbl)
+{
+ u16 offset, num_entries;
+ u16 req_entries = ((OPA_VNIC_EMA_DATA - sizeof(*mac_tbl)) /
+ sizeof(mac_tbl->tbl_entries[0]));
+
+ offset = be16_to_cpu(mac_tbl->offset);
+ num_entries = be16_to_cpu(mac_tbl->num_entries);
+
+ return ((num_entries <= req_entries) &&
+ (offset + num_entries <= OPA_VNIC_MAC_TBL_MAX_ENTRIES));
+}
+
+/*
+ * Return the power on default values in the port info structure
+ * in big endian format as required by MAD.
+ */
+static inline void vema_get_pod_values(struct opa_veswport_info *port_info)
+{
+ memset(port_info, 0, sizeof(*port_info));
+ port_info->vport.max_mac_tbl_ent =
+ cpu_to_be16(OPA_VNIC_MAC_TBL_MAX_ENTRIES);
+ port_info->vport.max_smac_ent =
+ cpu_to_be16(OPA_VNIC_MAX_SMAC_LIMIT);
+ port_info->vport.oper_state = OPA_VNIC_STATE_DROP_ALL;
+ port_info->vport.config_state = OPA_VNIC_STATE_DROP_ALL;
+ port_info->vesw.eth_mtu = cpu_to_be16(ETH_DATA_LEN);
+}
+
+/**
+ * vema_add_vport -- Add a new vnic port
+ * @port: ptr to opa_vnic_vema_port struct
+ * @vport_num: vnic port number (to be added)
+ *
+ * Return a pointer to the vnic adapter structure
+ */
+static struct opa_vnic_adapter *vema_add_vport(struct opa_vnic_vema_port *port,
+ u8 vport_num)
+{
+ struct opa_vnic_ctrl_port *cport = port->cport;
+ struct opa_vnic_adapter *adapter;
+
+ adapter = opa_vnic_add_netdev(cport->ibdev, port->port_num, vport_num);
+ if (!IS_ERR(adapter)) {
+ int rc;
+
+ adapter->cport = cport;
+ rc = xa_insert(&port->vports, vport_num, adapter, GFP_KERNEL);
+ if (rc < 0) {
+ opa_vnic_rem_netdev(adapter);
+ adapter = ERR_PTR(rc);
+ }
+ }
+
+ return adapter;
+}
+
+/**
+ * vema_get_class_port_info -- Get class info for port
+ * @port: Port on whic MAD was received
+ * @recvd_mad: pointer to the received mad
+ * @rsp_mad: pointer to respose mad
+ *
+ * This function copies the latest class port info value set for the
+ * port and stores it for generating traps
+ */
+static void vema_get_class_port_info(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad)
+{
+ struct opa_class_port_info *port_info;
+
+ port_info = (struct opa_class_port_info *)rsp_mad->data;
+ memcpy(port_info, &port->class_port_info, sizeof(*port_info));
+ port_info->base_version = OPA_MGMT_BASE_VERSION;
+ port_info->class_version = OPA_EMA_CLASS_VERSION;
+
+ /*
+ * Set capability mask bit indicating agent generates traps,
+ * and set the maximum number of VNIC ports supported.
+ */
+ port_info->cap_mask = cpu_to_be16((OPA_VNIC_CLASS_CAP_TRAP |
+ (OPA_VNIC_MAX_NUM_VPORT << 8)));
+
+ /*
+ * Since a get routine is always sent by the EM first we
+ * set the expected response time to
+ * 4.096 usec * 2^18 == 1.0737 sec here.
+ */
+ port_info->cap_mask2_resp_time = cpu_to_be32(18);
+}
+
+/**
+ * vema_set_class_port_info -- Get class info for port
+ * @port: Port on whic MAD was received
+ * @recvd_mad: pointer to the received mad
+ * @rsp_mad: pointer to respose mad
+ *
+ * This function updates the port class info for the specific vnic
+ * and sets up the response mad data
+ */
+static void vema_set_class_port_info(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad)
+{
+ memcpy(&port->class_port_info, recvd_mad->data,
+ sizeof(port->class_port_info));
+
+ vema_get_class_port_info(port, recvd_mad, rsp_mad);
+}
+
+/**
+ * vema_get_veswport_info -- Get veswport info
+ * @port: source port on which MAD was received
+ * @recvd_mad: pointer to the received mad
+ * @rsp_mad: pointer to respose mad
+ */
+static void vema_get_veswport_info(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad)
+{
+ struct opa_veswport_info *port_info =
+ (struct opa_veswport_info *)rsp_mad->data;
+ struct opa_vnic_adapter *adapter;
+
+ adapter = vema_get_vport_adapter(recvd_mad, port);
+ if (adapter) {
+ memset(port_info, 0, sizeof(*port_info));
+ opa_vnic_get_vesw_info(adapter, &port_info->vesw);
+ opa_vnic_get_per_veswport_info(adapter,
+ &port_info->vport);
+ } else {
+ vema_get_pod_values(port_info);
+ }
+}
+
+/**
+ * vema_set_veswport_info -- Set veswport info
+ * @port: source port on which MAD was received
+ * @recvd_mad: pointer to the received mad
+ * @rsp_mad: pointer to respose mad
+ *
+ * This function gets the port class infor for vnic
+ */
+static void vema_set_veswport_info(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad)
+{
+ struct opa_vnic_ctrl_port *cport = port->cport;
+ struct opa_veswport_info *port_info;
+ struct opa_vnic_adapter *adapter;
+ u8 vport_num;
+
+ vport_num = vema_get_vport_num(recvd_mad);
+
+ adapter = vema_get_vport_adapter(recvd_mad, port);
+ if (!adapter) {
+ adapter = vema_add_vport(port, vport_num);
+ if (IS_ERR(adapter)) {
+ c_err("failed to add vport %d: %ld\n",
+ vport_num, PTR_ERR(adapter));
+ goto err_exit;
+ }
+ }
+
+ port_info = (struct opa_veswport_info *)recvd_mad->data;
+ opa_vnic_set_vesw_info(adapter, &port_info->vesw);
+ opa_vnic_set_per_veswport_info(adapter, &port_info->vport);
+
+ /* Process the new config settings */
+ opa_vnic_process_vema_config(adapter);
+
+ vema_get_veswport_info(port, recvd_mad, rsp_mad);
+ return;
+
+err_exit:
+ rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR;
+}
+
+/**
+ * vema_get_mac_entries -- Get MAC entries in VNIC MAC table
+ * @port: source port on which MAD was received
+ * @recvd_mad: pointer to the received mad
+ * @rsp_mad: pointer to respose mad
+ *
+ * This function gets the MAC entries that are programmed into
+ * the VNIC MAC forwarding table. It checks for the validity of
+ * the index into the MAC table and the number of entries that
+ * are to be retrieved.
+ */
+static void vema_get_mac_entries(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad)
+{
+ struct opa_veswport_mactable *mac_tbl_in, *mac_tbl_out;
+ struct opa_vnic_adapter *adapter;
+
+ adapter = vema_get_vport_adapter(recvd_mad, port);
+ if (!adapter) {
+ rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR;
+ return;
+ }
+
+ mac_tbl_in = (struct opa_veswport_mactable *)recvd_mad->data;
+ mac_tbl_out = (struct opa_veswport_mactable *)rsp_mad->data;
+
+ if (vema_mac_tbl_req_ok(mac_tbl_in)) {
+ mac_tbl_out->offset = mac_tbl_in->offset;
+ mac_tbl_out->num_entries = mac_tbl_in->num_entries;
+ opa_vnic_query_mac_tbl(adapter, mac_tbl_out);
+ } else {
+ rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR;
+ }
+}
+
+/**
+ * vema_set_mac_entries -- Set MAC entries in VNIC MAC table
+ * @port: source port on which MAD was received
+ * @recvd_mad: pointer to the received mad
+ * @rsp_mad: pointer to respose mad
+ *
+ * This function sets the MAC entries in the VNIC forwarding table
+ * It checks for the validity of the index and the number of forwarding
+ * table entries to be programmed.
+ */
+static void vema_set_mac_entries(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad)
+{
+ struct opa_veswport_mactable *mac_tbl;
+ struct opa_vnic_adapter *adapter;
+
+ adapter = vema_get_vport_adapter(recvd_mad, port);
+ if (!adapter) {
+ rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR;
+ return;
+ }
+
+ mac_tbl = (struct opa_veswport_mactable *)recvd_mad->data;
+ if (vema_mac_tbl_req_ok(mac_tbl)) {
+ if (opa_vnic_update_mac_tbl(adapter, mac_tbl))
+ rsp_mad->mad_hdr.status = OPA_VNIC_UNSUP_ATTR;
+ } else {
+ rsp_mad->mad_hdr.status = OPA_VNIC_UNSUP_ATTR;
+ }
+ vema_get_mac_entries(port, recvd_mad, rsp_mad);
+}
+
+/**
+ * vema_set_delete_vesw -- Reset VESW info to POD values
+ * @port: source port on which MAD was received
+ * @recvd_mad: pointer to the received mad
+ * @rsp_mad: pointer to respose mad
+ *
+ * This function clears all the fields of veswport info for the requested vesw
+ * and sets them back to the power-on default values. It does not delete the
+ * vesw.
+ */
+static void vema_set_delete_vesw(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad)
+{
+ struct opa_veswport_info *port_info =
+ (struct opa_veswport_info *)rsp_mad->data;
+ struct opa_vnic_adapter *adapter;
+
+ adapter = vema_get_vport_adapter(recvd_mad, port);
+ if (!adapter) {
+ rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR;
+ return;
+ }
+
+ vema_get_pod_values(port_info);
+ opa_vnic_set_vesw_info(adapter, &port_info->vesw);
+ opa_vnic_set_per_veswport_info(adapter, &port_info->vport);
+
+ /* Process the new config settings */
+ opa_vnic_process_vema_config(adapter);
+
+ opa_vnic_release_mac_tbl(adapter);
+
+ vema_get_veswport_info(port, recvd_mad, rsp_mad);
+}
+
+/**
+ * vema_get_mac_list -- Get the unicast/multicast macs.
+ * @port: source port on which MAD was received
+ * @recvd_mad: Received mad contains fields to set vnic parameters
+ * @rsp_mad: Response mad to be built
+ * @attr_id: Attribute ID indicating multicast or unicast mac list
+ */
+static void vema_get_mac_list(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad,
+ u16 attr_id)
+{
+ struct opa_veswport_iface_macs *macs_in, *macs_out;
+ int max_entries = (OPA_VNIC_EMA_DATA - sizeof(*macs_out)) / ETH_ALEN;
+ struct opa_vnic_adapter *adapter;
+
+ adapter = vema_get_vport_adapter(recvd_mad, port);
+ if (!adapter) {
+ rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR;
+ return;
+ }
+
+ macs_in = (struct opa_veswport_iface_macs *)recvd_mad->data;
+ macs_out = (struct opa_veswport_iface_macs *)rsp_mad->data;
+
+ macs_out->start_idx = macs_in->start_idx;
+ if (macs_in->num_macs_in_msg)
+ macs_out->num_macs_in_msg = macs_in->num_macs_in_msg;
+ else
+ macs_out->num_macs_in_msg = cpu_to_be16(max_entries);
+
+ if (attr_id == OPA_EM_ATTR_IFACE_MCAST_MACS)
+ opa_vnic_query_mcast_macs(adapter, macs_out);
+ else
+ opa_vnic_query_ucast_macs(adapter, macs_out);
+}
+
+/**
+ * vema_get_summary_counters -- Gets summary counters.
+ * @port: source port on which MAD was received
+ * @recvd_mad: Received mad contains fields to set vnic parameters
+ * @rsp_mad: Response mad to be built
+ */
+static void vema_get_summary_counters(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad)
+{
+ struct opa_veswport_summary_counters *cntrs;
+ struct opa_vnic_adapter *adapter;
+
+ adapter = vema_get_vport_adapter(recvd_mad, port);
+ if (adapter) {
+ cntrs = (struct opa_veswport_summary_counters *)rsp_mad->data;
+ opa_vnic_get_summary_counters(adapter, cntrs);
+ } else {
+ rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR;
+ }
+}
+
+/**
+ * vema_get_error_counters -- Gets summary counters.
+ * @port: source port on which MAD was received
+ * @recvd_mad: Received mad contains fields to set vnic parameters
+ * @rsp_mad: Response mad to be built
+ */
+static void vema_get_error_counters(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad)
+{
+ struct opa_veswport_error_counters *cntrs;
+ struct opa_vnic_adapter *adapter;
+
+ adapter = vema_get_vport_adapter(recvd_mad, port);
+ if (adapter) {
+ cntrs = (struct opa_veswport_error_counters *)rsp_mad->data;
+ opa_vnic_get_error_counters(adapter, cntrs);
+ } else {
+ rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR;
+ }
+}
+
+/**
+ * vema_get -- Process received get MAD
+ * @port: source port on which MAD was received
+ * @recvd_mad: Received mad
+ * @rsp_mad: Response mad to be built
+ */
+static void vema_get(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad)
+{
+ u16 attr_id = be16_to_cpu(recvd_mad->mad_hdr.attr_id);
+
+ switch (attr_id) {
+ case OPA_EM_ATTR_CLASS_PORT_INFO:
+ vema_get_class_port_info(port, recvd_mad, rsp_mad);
+ break;
+ case OPA_EM_ATTR_VESWPORT_INFO:
+ vema_get_veswport_info(port, recvd_mad, rsp_mad);
+ break;
+ case OPA_EM_ATTR_VESWPORT_MAC_ENTRIES:
+ vema_get_mac_entries(port, recvd_mad, rsp_mad);
+ break;
+ case OPA_EM_ATTR_IFACE_UCAST_MACS:
+ case OPA_EM_ATTR_IFACE_MCAST_MACS:
+ vema_get_mac_list(port, recvd_mad, rsp_mad, attr_id);
+ break;
+ case OPA_EM_ATTR_VESWPORT_SUMMARY_COUNTERS:
+ vema_get_summary_counters(port, recvd_mad, rsp_mad);
+ break;
+ case OPA_EM_ATTR_VESWPORT_ERROR_COUNTERS:
+ vema_get_error_counters(port, recvd_mad, rsp_mad);
+ break;
+ default:
+ rsp_mad->mad_hdr.status = OPA_VNIC_UNSUP_ATTR;
+ break;
+ }
+}
+
+/**
+ * vema_set -- Process received set MAD
+ * @port: source port on which MAD was received
+ * @recvd_mad: Received mad contains fields to set vnic parameters
+ * @rsp_mad: Response mad to be built
+ */
+static void vema_set(struct opa_vnic_vema_port *port,
+ struct opa_vnic_vema_mad *recvd_mad,
+ struct opa_vnic_vema_mad *rsp_mad)
+{
+ u16 attr_id = be16_to_cpu(recvd_mad->mad_hdr.attr_id);
+
+ switch (attr_id) {
+ case OPA_EM_ATTR_CLASS_PORT_INFO:
+ vema_set_class_port_info(port, recvd_mad, rsp_mad);
+ break;
+ case OPA_EM_ATTR_VESWPORT_INFO:
+ vema_set_veswport_info(port, recvd_mad, rsp_mad);
+ break;
+ case OPA_EM_ATTR_VESWPORT_MAC_ENTRIES:
+ vema_set_mac_entries(port, recvd_mad, rsp_mad);
+ break;
+ case OPA_EM_ATTR_DELETE_VESW:
+ vema_set_delete_vesw(port, recvd_mad, rsp_mad);
+ break;
+ default:
+ rsp_mad->mad_hdr.status = OPA_VNIC_UNSUP_ATTR;
+ break;
+ }
+}
+
+/**
+ * vema_send -- Send handler for VEMA MAD agent
+ * @mad_agent: pointer to the mad agent
+ * @mad_wc: pointer to mad send work completion information
+ *
+ * Free all the data structures associated with the sent MAD
+ */
+static void vema_send(struct ib_mad_agent *mad_agent,
+ struct ib_mad_send_wc *mad_wc)
+{
+ rdma_destroy_ah(mad_wc->send_buf->ah, RDMA_DESTROY_AH_SLEEPABLE);
+ ib_free_send_mad(mad_wc->send_buf);
+}
+
+/**
+ * vema_recv -- Recv handler for VEMA MAD agent
+ * @mad_agent: pointer to the mad agent
+ * @send_buf: Send buffer if found, else NULL
+ * @mad_wc: pointer to mad send work completion information
+ *
+ * Handle only set and get methods and respond to other methods
+ * as unsupported. Allocate response buffer and address handle
+ * for the response MAD.
+ */
+static void vema_recv(struct ib_mad_agent *mad_agent,
+ struct ib_mad_send_buf *send_buf,
+ struct ib_mad_recv_wc *mad_wc)
+{
+ struct opa_vnic_vema_port *port;
+ struct ib_ah *ah;
+ struct ib_mad_send_buf *rsp;
+ struct opa_vnic_vema_mad *vema_mad;
+
+ if (!mad_wc || !mad_wc->recv_buf.mad)
+ return;
+
+ port = mad_agent->context;
+ ah = ib_create_ah_from_wc(mad_agent->qp->pd, mad_wc->wc,
+ mad_wc->recv_buf.grh, mad_agent->port_num);
+ if (IS_ERR(ah))
+ goto free_recv_mad;
+
+ rsp = ib_create_send_mad(mad_agent, mad_wc->wc->src_qp,
+ mad_wc->wc->pkey_index, 0,
+ IB_MGMT_VENDOR_HDR, OPA_VNIC_EMA_DATA,
+ GFP_KERNEL, OPA_MGMT_BASE_VERSION);
+ if (IS_ERR(rsp))
+ goto err_rsp;
+
+ rsp->ah = ah;
+ vema_mad = rsp->mad;
+ memcpy(vema_mad, mad_wc->recv_buf.mad, IB_MGMT_VENDOR_HDR);
+ vema_mad->mad_hdr.method = IB_MGMT_METHOD_GET_RESP;
+ vema_mad->mad_hdr.status = 0;
+
+ /* Lock ensures network adapter is not removed */
+ mutex_lock(&port->lock);
+
+ switch (mad_wc->recv_buf.mad->mad_hdr.method) {
+ case IB_MGMT_METHOD_GET:
+ vema_get(port, (struct opa_vnic_vema_mad *)mad_wc->recv_buf.mad,
+ vema_mad);
+ break;
+ case IB_MGMT_METHOD_SET:
+ vema_set(port, (struct opa_vnic_vema_mad *)mad_wc->recv_buf.mad,
+ vema_mad);
+ break;
+ default:
+ vema_mad->mad_hdr.status = OPA_VNIC_UNSUP_ATTR;
+ break;
+ }
+ mutex_unlock(&port->lock);
+
+ if (!ib_post_send_mad(rsp, NULL)) {
+ /*
+ * with post send successful ah and send mad
+ * will be destroyed in send handler
+ */
+ goto free_recv_mad;
+ }
+
+ ib_free_send_mad(rsp);
+
+err_rsp:
+ rdma_destroy_ah(ah, RDMA_DESTROY_AH_SLEEPABLE);
+free_recv_mad:
+ ib_free_recv_mad(mad_wc);
+}
+
+/**
+ * vema_get_port -- Gets the opa_vnic_vema_port
+ * @cport: pointer to control dev
+ * @port_num: Port number
+ *
+ * This function loops through the ports and returns
+ * the opa_vnic_vema port structure that is associated
+ * with the OPA port number
+ *
+ * Return: ptr to requested opa_vnic_vema_port strucure
+ * if success, NULL if not
+ */
+static struct opa_vnic_vema_port *
+vema_get_port(struct opa_vnic_ctrl_port *cport, u8 port_num)
+{
+ struct opa_vnic_vema_port *port = (void *)cport + sizeof(*cport);
+
+ if (port_num > cport->num_ports)
+ return NULL;
+
+ return port + (port_num - 1);
+}
+
+/**
+ * opa_vnic_vema_send_trap -- This function sends a trap to the EM
+ * @adapter: pointer to vnic adapter
+ * @data: pointer to trap data filled by calling function
+ * @lid: issuers lid (encap_slid from vesw_port_info)
+ *
+ * This function is called from the VNIC driver to send a trap if there
+ * is somethng the EM should be notified about. These events currently
+ * are
+ * 1) UNICAST INTERFACE MACADDRESS changes
+ * 2) MULTICAST INTERFACE MACADDRESS changes
+ * 3) ETHERNET LINK STATUS changes
+ * While allocating the send mad the remote site qpn used is 1
+ * as this is the well known QP.
+ *
+ */
+void opa_vnic_vema_send_trap(struct opa_vnic_adapter *adapter,
+ struct __opa_veswport_trap *data, u32 lid)
+{
+ struct opa_vnic_ctrl_port *cport = adapter->cport;
+ struct ib_mad_send_buf *send_buf;
+ struct opa_vnic_vema_port *port;
+ struct ib_device *ibp;
+ struct opa_vnic_vema_mad_trap *trap_mad;
+ struct opa_class_port_info *class;
+ struct rdma_ah_attr ah_attr;
+ struct ib_ah *ah;
+ struct opa_veswport_trap *trap;
+ u32 trap_lid;
+ u16 pkey_idx;
+
+ if (!cport)
+ goto err_exit;
+ ibp = cport->ibdev;
+ port = vema_get_port(cport, data->opaportnum);
+ if (!port || !port->mad_agent)
+ goto err_exit;
+
+ if (time_before(jiffies, adapter->trap_timeout)) {
+ if (adapter->trap_count == OPA_VNIC_TRAP_BURST_LIMIT) {
+ v_warn("Trap rate exceeded\n");
+ goto err_exit;
+ } else {
+ adapter->trap_count++;
+ }
+ } else {
+ adapter->trap_count = 0;
+ }
+
+ class = &port->class_port_info;
+ /* Set up address handle */
+ memset(&ah_attr, 0, sizeof(ah_attr));
+ ah_attr.type = rdma_ah_find_type(ibp, port->port_num);
+ rdma_ah_set_sl(&ah_attr,
+ GET_TRAP_SL_FROM_CLASS_PORT_INFO(class->trap_sl_rsvd));
+ rdma_ah_set_port_num(&ah_attr, port->port_num);
+ trap_lid = be32_to_cpu(class->trap_lid);
+ /*
+ * check for trap lid validity, must not be zero
+ * The trap sink could change after we fashion the MAD but since traps
+ * are not guaranteed we won't use a lock as anyway the change will take
+ * place even with locking.
+ */
+ if (!trap_lid) {
+ c_err("%s: Invalid dlid\n", __func__);
+ goto err_exit;
+ }
+
+ rdma_ah_set_dlid(&ah_attr, trap_lid);
+ ah = rdma_create_ah(port->mad_agent->qp->pd, &ah_attr, 0);
+ if (IS_ERR(ah)) {
+ c_err("%s:Couldn't create new AH = %p\n", __func__, ah);
+ c_err("%s:dlid = %d, sl = %d, port = %d\n", __func__,
+ rdma_ah_get_dlid(&ah_attr), rdma_ah_get_sl(&ah_attr),
+ rdma_ah_get_port_num(&ah_attr));
+ goto err_exit;
+ }
+
+ if (ib_find_pkey(ibp, data->opaportnum, IB_DEFAULT_PKEY_FULL,
+ &pkey_idx) < 0) {
+ c_err("%s:full key not found, defaulting to partial\n",
+ __func__);
+ if (ib_find_pkey(ibp, data->opaportnum, IB_DEFAULT_PKEY_PARTIAL,
+ &pkey_idx) < 0)
+ pkey_idx = 1;
+ }
+
+ send_buf = ib_create_send_mad(port->mad_agent, 1, pkey_idx, 0,
+ IB_MGMT_VENDOR_HDR, IB_MGMT_MAD_DATA,
+ GFP_ATOMIC, OPA_MGMT_BASE_VERSION);
+ if (IS_ERR(send_buf)) {
+ c_err("%s:Couldn't allocate send buf\n", __func__);
+ goto err_sndbuf;
+ }
+
+ send_buf->ah = ah;
+
+ /* Set up common MAD hdr */
+ trap_mad = send_buf->mad;
+ trap_mad->mad_hdr.base_version = OPA_MGMT_BASE_VERSION;
+ trap_mad->mad_hdr.mgmt_class = OPA_MGMT_CLASS_INTEL_EMA;
+ trap_mad->mad_hdr.class_version = OPA_EMA_CLASS_VERSION;
+ trap_mad->mad_hdr.method = IB_MGMT_METHOD_TRAP;
+ port->tid++;
+ trap_mad->mad_hdr.tid = cpu_to_be64(port->tid);
+ trap_mad->mad_hdr.attr_id = IB_SMP_ATTR_NOTICE;
+
+ /* Set up vendor OUI */
+ trap_mad->oui[0] = INTEL_OUI_1;
+ trap_mad->oui[1] = INTEL_OUI_2;
+ trap_mad->oui[2] = INTEL_OUI_3;
+
+ /* Setup notice attribute portion */
+ trap_mad->notice.gen_type = OPA_INTEL_EMA_NOTICE_TYPE_INFO << 1;
+ trap_mad->notice.oui_1 = INTEL_OUI_1;
+ trap_mad->notice.oui_2 = INTEL_OUI_2;
+ trap_mad->notice.oui_3 = INTEL_OUI_3;
+ trap_mad->notice.issuer_lid = cpu_to_be32(lid);
+
+ /* copy the actual trap data */
+ trap = (struct opa_veswport_trap *)trap_mad->notice.raw_data;
+ trap->fabric_id = cpu_to_be16(data->fabric_id);
+ trap->veswid = cpu_to_be16(data->veswid);
+ trap->veswportnum = cpu_to_be32(data->veswportnum);
+ trap->opaportnum = cpu_to_be16(data->opaportnum);
+ trap->veswportindex = data->veswportindex;
+ trap->opcode = data->opcode;
+
+ /* If successful send set up rate limit timeout else bail */
+ if (ib_post_send_mad(send_buf, NULL)) {
+ ib_free_send_mad(send_buf);
+ } else {
+ if (adapter->trap_count)
+ return;
+ adapter->trap_timeout = jiffies +
+ usecs_to_jiffies(OPA_VNIC_TRAP_TIMEOUT);
+ return;
+ }
+
+err_sndbuf:
+ rdma_destroy_ah(ah, 0);
+err_exit:
+ v_err("Aborting trap\n");
+}
+
+static void opa_vnic_event(struct ib_event_handler *handler,
+ struct ib_event *record)
+{
+ struct opa_vnic_vema_port *port =
+ container_of(handler, struct opa_vnic_vema_port, event_handler);
+ struct opa_vnic_ctrl_port *cport = port->cport;
+ struct opa_vnic_adapter *adapter;
+ unsigned long index;
+
+ if (record->element.port_num != port->port_num)
+ return;
+
+ c_dbg("OPA_VNIC received event %d on device %s port %d\n",
+ record->event, dev_name(&record->device->dev),
+ record->element.port_num);
+
+ if (record->event != IB_EVENT_PORT_ERR &&
+ record->event != IB_EVENT_PORT_ACTIVE)
+ return;
+
+ xa_for_each(&port->vports, index, adapter) {
+ if (record->event == IB_EVENT_PORT_ACTIVE)
+ netif_carrier_on(adapter->netdev);
+ else
+ netif_carrier_off(adapter->netdev);
+ }
+}
+
+/**
+ * vema_unregister -- Unregisters agent
+ * @cport: pointer to control port
+ *
+ * This deletes the registration by VEMA for MADs
+ */
+static void vema_unregister(struct opa_vnic_ctrl_port *cport)
+{
+ struct opa_vnic_adapter *adapter;
+ unsigned long index;
+ int i;
+
+ for (i = 1; i <= cport->num_ports; i++) {
+ struct opa_vnic_vema_port *port = vema_get_port(cport, i);
+
+ if (!port->mad_agent)
+ continue;
+
+ /* Lock ensures no MAD is being processed */
+ mutex_lock(&port->lock);
+ xa_for_each(&port->vports, index, adapter)
+ opa_vnic_rem_netdev(adapter);
+ mutex_unlock(&port->lock);
+
+ ib_unregister_mad_agent(port->mad_agent);
+ port->mad_agent = NULL;
+ mutex_destroy(&port->lock);
+ xa_destroy(&port->vports);
+ ib_unregister_event_handler(&port->event_handler);
+ }
+}
+
+/**
+ * vema_register -- Registers agent
+ * @cport: pointer to control port
+ *
+ * This function registers the handlers for the VEMA MADs
+ *
+ * Return: returns 0 on success. non zero otherwise
+ */
+static int vema_register(struct opa_vnic_ctrl_port *cport)
+{
+ struct ib_mad_reg_req reg_req = {
+ .mgmt_class = OPA_MGMT_CLASS_INTEL_EMA,
+ .mgmt_class_version = OPA_MGMT_BASE_VERSION,
+ .oui = { INTEL_OUI_1, INTEL_OUI_2, INTEL_OUI_3 }
+ };
+ int i;
+
+ set_bit(IB_MGMT_METHOD_GET, reg_req.method_mask);
+ set_bit(IB_MGMT_METHOD_SET, reg_req.method_mask);
+
+ /* register ib event handler and mad agent for each port on dev */
+ for (i = 1; i <= cport->num_ports; i++) {
+ struct opa_vnic_vema_port *port = vema_get_port(cport, i);
+ int ret;
+
+ port->cport = cport;
+ port->port_num = i;
+
+ INIT_IB_EVENT_HANDLER(&port->event_handler,
+ cport->ibdev, opa_vnic_event);
+ ib_register_event_handler(&port->event_handler);
+
+ xa_init(&port->vports);
+ mutex_init(&port->lock);
+ port->mad_agent = ib_register_mad_agent(cport->ibdev, i,
+ IB_QPT_GSI, &reg_req,
+ IB_MGMT_RMPP_VERSION,
+ vema_send, vema_recv,
+ port, 0);
+ if (IS_ERR(port->mad_agent)) {
+ ret = PTR_ERR(port->mad_agent);
+ port->mad_agent = NULL;
+ mutex_destroy(&port->lock);
+ vema_unregister(cport);
+ return ret;
+ }
+ }
+
+ return 0;
+}
+
+/**
+ * opa_vnic_ctrl_config_dev -- This function sends a trap to the EM
+ * by way of ib_modify_port to indicate support for ethernet on the
+ * fabric.
+ * @cport: pointer to control port
+ * @en: enable or disable ethernet on fabric support
+ */
+static void opa_vnic_ctrl_config_dev(struct opa_vnic_ctrl_port *cport, bool en)
+{
+ struct ib_port_modify pm = { 0 };
+ int i;
+
+ if (en)
+ pm.set_port_cap_mask = OPA_CAP_MASK3_IsEthOnFabricSupported;
+ else
+ pm.clr_port_cap_mask = OPA_CAP_MASK3_IsEthOnFabricSupported;
+
+ for (i = 1; i <= cport->num_ports; i++)
+ ib_modify_port(cport->ibdev, i, IB_PORT_OPA_MASK_CHG, &pm);
+}
+
+/**
+ * opa_vnic_vema_add_one -- Handle new ib device
+ * @device: ib device pointer
+ *
+ * Allocate the vnic control port and initialize it.
+ */
+static int opa_vnic_vema_add_one(struct ib_device *device)
+{
+ struct opa_vnic_ctrl_port *cport;
+ int rc, size = sizeof(*cport);
+
+ if (!rdma_cap_opa_vnic(device))
+ return -EOPNOTSUPP;
+
+ size += device->phys_port_cnt * sizeof(struct opa_vnic_vema_port);
+ cport = kzalloc(size, GFP_KERNEL);
+ if (!cport)
+ return -ENOMEM;
+
+ cport->num_ports = device->phys_port_cnt;
+ cport->ibdev = device;
+
+ /* Initialize opa vnic management agent (vema) */
+ rc = vema_register(cport);
+ if (!rc)
+ c_info("VNIC client initialized\n");
+
+ ib_set_client_data(device, &opa_vnic_client, cport);
+ opa_vnic_ctrl_config_dev(cport, true);
+ return 0;
+}
+
+/**
+ * opa_vnic_vema_rem_one -- Handle ib device removal
+ * @device: ib device pointer
+ * @client_data: ib client data
+ *
+ * Uninitialize and free the vnic control port.
+ */
+static void opa_vnic_vema_rem_one(struct ib_device *device,
+ void *client_data)
+{
+ struct opa_vnic_ctrl_port *cport = client_data;
+
+ c_info("removing VNIC client\n");
+ opa_vnic_ctrl_config_dev(cport, false);
+ vema_unregister(cport);
+ kfree(cport);
+}
+
+static int __init opa_vnic_init(void)
+{
+ int rc;
+
+ rc = ib_register_client(&opa_vnic_client);
+ if (rc)
+ pr_err("VNIC driver register failed %d\n", rc);
+
+ return rc;
+}
+module_init(opa_vnic_init);
+
+static void opa_vnic_deinit(void)
+{
+ ib_unregister_client(&opa_vnic_client);
+}
+module_exit(opa_vnic_deinit);
+
+MODULE_LICENSE("Dual BSD/GPL");
+MODULE_AUTHOR("Cornelis Networks");
+MODULE_DESCRIPTION("Cornelis OPX Virtual Network driver");
diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema_iface.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema_iface.c
new file mode 100644
index 000000000..292c037aa
--- /dev/null
+++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema_iface.c
@@ -0,0 +1,390 @@
+/*
+ * Copyright(c) 2017 Intel Corporation.
+ *
+ * This file is provided under a dual BSD/GPLv2 license. When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * 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 GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * 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.
+ * - Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+/*
+ * This file contains OPA VNIC EMA Interface functions.
+ */
+
+#include "opa_vnic_internal.h"
+
+/**
+ * opa_vnic_vema_report_event - sent trap to report the specified event
+ * @adapter: vnic port adapter
+ * @event: event to be reported
+ *
+ * This function calls vema api to sent a trap for the given event.
+ */
+void opa_vnic_vema_report_event(struct opa_vnic_adapter *adapter, u8 event)
+{
+ struct __opa_veswport_info *info = &adapter->info;
+ struct __opa_veswport_trap trap_data;
+
+ trap_data.fabric_id = info->vesw.fabric_id;
+ trap_data.veswid = info->vesw.vesw_id;
+ trap_data.veswportnum = info->vport.port_num;
+ trap_data.opaportnum = adapter->port_num;
+ trap_data.veswportindex = adapter->vport_num;
+ trap_data.opcode = event;
+
+ opa_vnic_vema_send_trap(adapter, &trap_data, info->vport.encap_slid);
+}
+
+/**
+ * opa_vnic_get_summary_counters - get summary counters
+ * @adapter: vnic port adapter
+ * @cntrs: pointer to destination summary counters structure
+ *
+ * This function populates the summary counters that is maintained by the
+ * given adapter to destination address provided.
+ */
+void opa_vnic_get_summary_counters(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_summary_counters *cntrs)
+{
+ struct opa_vnic_stats vstats;
+ __be64 *dst;
+ u64 *src;
+
+ memset(&vstats, 0, sizeof(vstats));
+ spin_lock(&adapter->stats_lock);
+ adapter->rn_ops->ndo_get_stats64(adapter->netdev, &vstats.netstats);
+ spin_unlock(&adapter->stats_lock);
+
+ cntrs->vp_instance = cpu_to_be16(adapter->vport_num);
+ cntrs->vesw_id = cpu_to_be16(adapter->info.vesw.vesw_id);
+ cntrs->veswport_num = cpu_to_be32(adapter->port_num);
+
+ cntrs->tx_errors = cpu_to_be64(vstats.netstats.tx_errors);
+ cntrs->rx_errors = cpu_to_be64(vstats.netstats.rx_errors);
+ cntrs->tx_packets = cpu_to_be64(vstats.netstats.tx_packets);
+ cntrs->rx_packets = cpu_to_be64(vstats.netstats.rx_packets);
+ cntrs->tx_bytes = cpu_to_be64(vstats.netstats.tx_bytes);
+ cntrs->rx_bytes = cpu_to_be64(vstats.netstats.rx_bytes);
+
+ /*
+ * This loop depends on layout of
+ * opa_veswport_summary_counters opa_vnic_stats structures.
+ */
+ for (dst = &cntrs->tx_unicast, src = &vstats.tx_grp.unicast;
+ dst < &cntrs->reserved[0]; dst++, src++) {
+ *dst = cpu_to_be64(*src);
+ }
+}
+
+/**
+ * opa_vnic_get_error_counters - get error counters
+ * @adapter: vnic port adapter
+ * @cntrs: pointer to destination error counters structure
+ *
+ * This function populates the error counters that is maintained by the
+ * given adapter to destination address provided.
+ */
+void opa_vnic_get_error_counters(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_error_counters *cntrs)
+{
+ struct opa_vnic_stats vstats;
+
+ memset(&vstats, 0, sizeof(vstats));
+ spin_lock(&adapter->stats_lock);
+ adapter->rn_ops->ndo_get_stats64(adapter->netdev, &vstats.netstats);
+ spin_unlock(&adapter->stats_lock);
+
+ cntrs->vp_instance = cpu_to_be16(adapter->vport_num);
+ cntrs->vesw_id = cpu_to_be16(adapter->info.vesw.vesw_id);
+ cntrs->veswport_num = cpu_to_be32(adapter->port_num);
+
+ cntrs->tx_errors = cpu_to_be64(vstats.netstats.tx_errors);
+ cntrs->rx_errors = cpu_to_be64(vstats.netstats.rx_errors);
+ cntrs->tx_dlid_zero = cpu_to_be64(vstats.tx_dlid_zero);
+ cntrs->tx_drop_state = cpu_to_be64(vstats.tx_drop_state);
+ cntrs->tx_logic = cpu_to_be64(vstats.netstats.tx_fifo_errors +
+ vstats.netstats.tx_carrier_errors);
+
+ cntrs->rx_bad_veswid = cpu_to_be64(vstats.netstats.rx_nohandler);
+ cntrs->rx_runt = cpu_to_be64(vstats.rx_runt);
+ cntrs->rx_oversize = cpu_to_be64(vstats.rx_oversize);
+ cntrs->rx_drop_state = cpu_to_be64(vstats.rx_drop_state);
+ cntrs->rx_logic = cpu_to_be64(vstats.netstats.rx_fifo_errors);
+}
+
+/**
+ * opa_vnic_get_vesw_info -- Get the vesw information
+ * @adapter: vnic port adapter
+ * @info: pointer to destination vesw info structure
+ *
+ * This function copies the vesw info that is maintained by the
+ * given adapter to destination address provided.
+ */
+void opa_vnic_get_vesw_info(struct opa_vnic_adapter *adapter,
+ struct opa_vesw_info *info)
+{
+ struct __opa_vesw_info *src = &adapter->info.vesw;
+ int i;
+
+ info->fabric_id = cpu_to_be16(src->fabric_id);
+ info->vesw_id = cpu_to_be16(src->vesw_id);
+ memcpy(info->rsvd0, src->rsvd0, ARRAY_SIZE(src->rsvd0));
+ info->def_port_mask = cpu_to_be16(src->def_port_mask);
+ memcpy(info->rsvd1, src->rsvd1, ARRAY_SIZE(src->rsvd1));
+ info->pkey = cpu_to_be16(src->pkey);
+
+ memcpy(info->rsvd2, src->rsvd2, ARRAY_SIZE(src->rsvd2));
+ info->u_mcast_dlid = cpu_to_be32(src->u_mcast_dlid);
+ for (i = 0; i < OPA_VESW_MAX_NUM_DEF_PORT; i++)
+ info->u_ucast_dlid[i] = cpu_to_be32(src->u_ucast_dlid[i]);
+
+ info->rc = cpu_to_be32(src->rc);
+
+ memcpy(info->rsvd3, src->rsvd3, ARRAY_SIZE(src->rsvd3));
+ info->eth_mtu = cpu_to_be16(src->eth_mtu);
+ memcpy(info->rsvd4, src->rsvd4, ARRAY_SIZE(src->rsvd4));
+}
+
+/**
+ * opa_vnic_set_vesw_info -- Set the vesw information
+ * @adapter: vnic port adapter
+ * @info: pointer to vesw info structure
+ *
+ * This function updates the vesw info that is maintained by the
+ * given adapter with vesw info provided. Reserved fields are stored
+ * and returned back to EM as is.
+ */
+void opa_vnic_set_vesw_info(struct opa_vnic_adapter *adapter,
+ struct opa_vesw_info *info)
+{
+ struct __opa_vesw_info *dst = &adapter->info.vesw;
+ int i;
+
+ dst->fabric_id = be16_to_cpu(info->fabric_id);
+ dst->vesw_id = be16_to_cpu(info->vesw_id);
+ memcpy(dst->rsvd0, info->rsvd0, ARRAY_SIZE(info->rsvd0));
+ dst->def_port_mask = be16_to_cpu(info->def_port_mask);
+ memcpy(dst->rsvd1, info->rsvd1, ARRAY_SIZE(info->rsvd1));
+ dst->pkey = be16_to_cpu(info->pkey);
+
+ memcpy(dst->rsvd2, info->rsvd2, ARRAY_SIZE(info->rsvd2));
+ dst->u_mcast_dlid = be32_to_cpu(info->u_mcast_dlid);
+ for (i = 0; i < OPA_VESW_MAX_NUM_DEF_PORT; i++)
+ dst->u_ucast_dlid[i] = be32_to_cpu(info->u_ucast_dlid[i]);
+
+ dst->rc = be32_to_cpu(info->rc);
+
+ memcpy(dst->rsvd3, info->rsvd3, ARRAY_SIZE(info->rsvd3));
+ dst->eth_mtu = be16_to_cpu(info->eth_mtu);
+ memcpy(dst->rsvd4, info->rsvd4, ARRAY_SIZE(info->rsvd4));
+}
+
+/**
+ * opa_vnic_get_per_veswport_info -- Get the vesw per port information
+ * @adapter: vnic port adapter
+ * @info: pointer to destination vport info structure
+ *
+ * This function copies the vesw per port info that is maintained by the
+ * given adapter to destination address provided.
+ * Note that the read only fields are not copied.
+ */
+void opa_vnic_get_per_veswport_info(struct opa_vnic_adapter *adapter,
+ struct opa_per_veswport_info *info)
+{
+ struct __opa_per_veswport_info *src = &adapter->info.vport;
+
+ info->port_num = cpu_to_be32(src->port_num);
+ info->eth_link_status = src->eth_link_status;
+ memcpy(info->rsvd0, src->rsvd0, ARRAY_SIZE(src->rsvd0));
+
+ memcpy(info->base_mac_addr, src->base_mac_addr,
+ ARRAY_SIZE(info->base_mac_addr));
+ info->config_state = src->config_state;
+ info->oper_state = src->oper_state;
+ info->max_mac_tbl_ent = cpu_to_be16(src->max_mac_tbl_ent);
+ info->max_smac_ent = cpu_to_be16(src->max_smac_ent);
+ info->mac_tbl_digest = cpu_to_be32(src->mac_tbl_digest);
+ memcpy(info->rsvd1, src->rsvd1, ARRAY_SIZE(src->rsvd1));
+
+ info->encap_slid = cpu_to_be32(src->encap_slid);
+ memcpy(info->pcp_to_sc_uc, src->pcp_to_sc_uc,
+ ARRAY_SIZE(info->pcp_to_sc_uc));
+ memcpy(info->pcp_to_vl_uc, src->pcp_to_vl_uc,
+ ARRAY_SIZE(info->pcp_to_vl_uc));
+ memcpy(info->pcp_to_sc_mc, src->pcp_to_sc_mc,
+ ARRAY_SIZE(info->pcp_to_sc_mc));
+ memcpy(info->pcp_to_vl_mc, src->pcp_to_vl_mc,
+ ARRAY_SIZE(info->pcp_to_vl_mc));
+ info->non_vlan_sc_uc = src->non_vlan_sc_uc;
+ info->non_vlan_vl_uc = src->non_vlan_vl_uc;
+ info->non_vlan_sc_mc = src->non_vlan_sc_mc;
+ info->non_vlan_vl_mc = src->non_vlan_vl_mc;
+ memcpy(info->rsvd2, src->rsvd2, ARRAY_SIZE(src->rsvd2));
+
+ info->uc_macs_gen_count = cpu_to_be16(src->uc_macs_gen_count);
+ info->mc_macs_gen_count = cpu_to_be16(src->mc_macs_gen_count);
+ memcpy(info->rsvd3, src->rsvd3, ARRAY_SIZE(src->rsvd3));
+}
+
+/**
+ * opa_vnic_set_per_veswport_info -- Set vesw per port information
+ * @adapter: vnic port adapter
+ * @info: pointer to vport info structure
+ *
+ * This function updates the vesw per port info that is maintained by the
+ * given adapter with vesw per port info provided. Reserved fields are
+ * stored and returned back to EM as is.
+ */
+void opa_vnic_set_per_veswport_info(struct opa_vnic_adapter *adapter,
+ struct opa_per_veswport_info *info)
+{
+ struct __opa_per_veswport_info *dst = &adapter->info.vport;
+
+ dst->port_num = be32_to_cpu(info->port_num);
+ memcpy(dst->rsvd0, info->rsvd0, ARRAY_SIZE(info->rsvd0));
+
+ memcpy(dst->base_mac_addr, info->base_mac_addr,
+ ARRAY_SIZE(dst->base_mac_addr));
+ dst->config_state = info->config_state;
+ memcpy(dst->rsvd1, info->rsvd1, ARRAY_SIZE(info->rsvd1));
+
+ dst->encap_slid = be32_to_cpu(info->encap_slid);
+ memcpy(dst->pcp_to_sc_uc, info->pcp_to_sc_uc,
+ ARRAY_SIZE(dst->pcp_to_sc_uc));
+ memcpy(dst->pcp_to_vl_uc, info->pcp_to_vl_uc,
+ ARRAY_SIZE(dst->pcp_to_vl_uc));
+ memcpy(dst->pcp_to_sc_mc, info->pcp_to_sc_mc,
+ ARRAY_SIZE(dst->pcp_to_sc_mc));
+ memcpy(dst->pcp_to_vl_mc, info->pcp_to_vl_mc,
+ ARRAY_SIZE(dst->pcp_to_vl_mc));
+ dst->non_vlan_sc_uc = info->non_vlan_sc_uc;
+ dst->non_vlan_vl_uc = info->non_vlan_vl_uc;
+ dst->non_vlan_sc_mc = info->non_vlan_sc_mc;
+ dst->non_vlan_vl_mc = info->non_vlan_vl_mc;
+ memcpy(dst->rsvd2, info->rsvd2, ARRAY_SIZE(info->rsvd2));
+ memcpy(dst->rsvd3, info->rsvd3, ARRAY_SIZE(info->rsvd3));
+}
+
+/**
+ * opa_vnic_query_mcast_macs - query multicast mac list
+ * @adapter: vnic port adapter
+ * @macs: pointer mac list
+ *
+ * This function populates the provided mac list with the configured
+ * multicast addresses in the adapter.
+ */
+void opa_vnic_query_mcast_macs(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_iface_macs *macs)
+{
+ u16 start_idx, num_macs, idx = 0, count = 0;
+ struct netdev_hw_addr *ha;
+
+ start_idx = be16_to_cpu(macs->start_idx);
+ num_macs = be16_to_cpu(macs->num_macs_in_msg);
+ netdev_for_each_mc_addr(ha, adapter->netdev) {
+ struct opa_vnic_iface_mac_entry *entry = &macs->entry[count];
+
+ if (start_idx > idx++)
+ continue;
+ else if (num_macs == count)
+ break;
+ memcpy(entry, ha->addr, sizeof(*entry));
+ count++;
+ }
+
+ macs->tot_macs_in_lst = cpu_to_be16(netdev_mc_count(adapter->netdev));
+ macs->num_macs_in_msg = cpu_to_be16(count);
+ macs->gen_count = cpu_to_be16(adapter->info.vport.mc_macs_gen_count);
+}
+
+/**
+ * opa_vnic_query_ucast_macs - query unicast mac list
+ * @adapter: vnic port adapter
+ * @macs: pointer mac list
+ *
+ * This function populates the provided mac list with the configured
+ * unicast addresses in the adapter.
+ */
+void opa_vnic_query_ucast_macs(struct opa_vnic_adapter *adapter,
+ struct opa_veswport_iface_macs *macs)
+{
+ u16 start_idx, tot_macs, num_macs, idx = 0, count = 0, em_macs = 0;
+ struct netdev_hw_addr *ha;
+
+ start_idx = be16_to_cpu(macs->start_idx);
+ num_macs = be16_to_cpu(macs->num_macs_in_msg);
+ /* loop through dev_addrs list first */
+ for_each_dev_addr(adapter->netdev, ha) {
+ struct opa_vnic_iface_mac_entry *entry = &macs->entry[count];
+
+ /* Do not include EM specified MAC address */
+ if (!memcmp(adapter->info.vport.base_mac_addr, ha->addr,
+ ARRAY_SIZE(adapter->info.vport.base_mac_addr))) {
+ em_macs++;
+ continue;
+ }
+
+ if (start_idx > idx++)
+ continue;
+ else if (num_macs == count)
+ break;
+ memcpy(entry, ha->addr, sizeof(*entry));
+ count++;
+ }
+
+ /* loop through uc list */
+ netdev_for_each_uc_addr(ha, adapter->netdev) {
+ struct opa_vnic_iface_mac_entry *entry = &macs->entry[count];
+
+ if (start_idx > idx++)
+ continue;
+ else if (num_macs == count)
+ break;
+ memcpy(entry, ha->addr, sizeof(*entry));
+ count++;
+ }
+
+ tot_macs = netdev_hw_addr_list_count(&adapter->netdev->dev_addrs) +
+ netdev_uc_count(adapter->netdev) - em_macs;
+ macs->tot_macs_in_lst = cpu_to_be16(tot_macs);
+ macs->num_macs_in_msg = cpu_to_be16(count);
+ macs->gen_count = cpu_to_be16(adapter->info.vport.uc_macs_gen_count);
+}