Adding upstream version 4.19.249.upstream/4.19.249

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

9 files changed, 4976 insertions, 0 deletions

diff --git a/drivers/soc/ti/Kconfig b/drivers/soc/ti/Kconfig
new file mode 100644
index 000000000..be4570baa
--- /dev/null
+++ b/drivers/soc/ti/Kconfig
@@ -0,0 +1,76 @@
+# 64-bit ARM SoCs from TI
+if ARM64
+
+if ARCH_K3
+
+config ARCH_K3_AM6_SOC
+	bool "K3 AM6 SoC"
+	help
+	  Enable support for TI's AM6 SoC Family support
+
+endif
+
+endif
+
+#
+# TI SOC drivers
+#
+menuconfig SOC_TI
+	bool "TI SOC drivers support"
+
+if SOC_TI
+
+config KEYSTONE_NAVIGATOR_QMSS
+	tristate "Keystone Queue Manager Sub System"
+	depends on ARCH_KEYSTONE
+	help
+	  Say y here to support the Keystone multicore Navigator Queue
+	  Manager support. The Queue Manager is a hardware module that
+	  is responsible for accelerating management of the packet queues.
+	  Packets are queued/de-queued by writing/reading descriptor address
+	  to a particular memory mapped location in the Queue Manager module.
+
+	  If unsure, say N.
+
+config KEYSTONE_NAVIGATOR_DMA
+	tristate "TI Keystone Navigator Packet DMA support"
+	depends on ARCH_KEYSTONE
+	help
+	  Say y tp enable support for the Keystone Navigator Packet DMA on
+	  on Keystone family of devices. It sets up the dma channels for the
+	  Queue Manager Sub System.
+
+	  If unsure, say N.
+
+config AMX3_PM
+	tristate "AMx3 Power Management"
+	depends on SOC_AM33XX || SOC_AM43XX
+	depends on WKUP_M3_IPC && TI_EMIF_SRAM && SRAM
+	help
+	  Enable power management on AM335x and AM437x. Required for suspend to mem
+	  and standby states on both AM335x and AM437x platforms and for deeper cpuidle
+	  c-states on AM335x.
+
+config WKUP_M3_IPC
+	tristate "TI AMx3 Wkup-M3 IPC Driver"
+	depends on WKUP_M3_RPROC
+	depends on OMAP2PLUS_MBOX
+	help
+	  TI AM33XX and AM43XX have a Cortex M3, the Wakeup M3, to handle
+	  low power transitions. This IPC driver provides the necessary API
+	  to communicate and use the Wakeup M3 for PM features like suspend
+	  resume and boots it using wkup_m3_rproc driver.
+
+config TI_SCI_PM_DOMAINS
+	tristate "TI SCI PM Domains Driver"
+	depends on TI_SCI_PROTOCOL
+	depends on PM_GENERIC_DOMAINS
+	help
+	  Generic power domain implementation for TI device implementing
+	  the TI SCI protocol.
+
+	  To compile this as a module, choose M here. The module will be
+	  called ti_sci_pm_domains. Note this is needed early in boot before
+	  rootfs may be available.
+
+endif # SOC_TI
diff --git a/drivers/soc/ti/Makefile b/drivers/soc/ti/Makefile
new file mode 100644
index 000000000..a22edc0b2
--- /dev/null
+++ b/drivers/soc/ti/Makefile
@@ -0,0 +1,10 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# TI Keystone SOC drivers
+#
+obj-$(CONFIG_KEYSTONE_NAVIGATOR_QMSS)	+= knav_qmss.o
+knav_qmss-y := knav_qmss_queue.o knav_qmss_acc.o
+obj-$(CONFIG_KEYSTONE_NAVIGATOR_DMA)	+= knav_dma.o
+obj-$(CONFIG_AMX3_PM)			+= pm33xx.o
+obj-$(CONFIG_WKUP_M3_IPC)		+= wkup_m3_ipc.o
+obj-$(CONFIG_TI_SCI_PM_DOMAINS)		+= ti_sci_pm_domains.o
diff --git a/drivers/soc/ti/knav_dma.c b/drivers/soc/ti/knav_dma.c
new file mode 100644
index 000000000..eb2e87229
--- /dev/null
+++ b/drivers/soc/ti/knav_dma.c
@@ -0,0 +1,831 @@
+/*
+ * Copyright (C) 2014 Texas Instruments Incorporated
+ * Authors:	Santosh Shilimkar <santosh.shilimkar@ti.com>
+ *		Sandeep Nair <sandeep_n@ti.com>
+ *		Cyril Chemparathy <cyril@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation version 2.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/io.h>
+#include <linux/sched.h>
+#include <linux/module.h>
+#include <linux/dma-direction.h>
+#include <linux/interrupt.h>
+#include <linux/pm_runtime.h>
+#include <linux/of_dma.h>
+#include <linux/of_address.h>
+#include <linux/platform_device.h>
+#include <linux/soc/ti/knav_dma.h>
+#include <linux/debugfs.h>
+#include <linux/seq_file.h>
+
+#define REG_MASK		0xffffffff
+
+#define DMA_LOOPBACK		BIT(31)
+#define DMA_ENABLE		BIT(31)
+#define DMA_TEARDOWN		BIT(30)
+
+#define DMA_TX_FILT_PSWORDS	BIT(29)
+#define DMA_TX_FILT_EINFO	BIT(30)
+#define DMA_TX_PRIO_SHIFT	0
+#define DMA_RX_PRIO_SHIFT	16
+#define DMA_PRIO_MASK		GENMASK(3, 0)
+#define DMA_PRIO_DEFAULT	0
+#define DMA_RX_TIMEOUT_DEFAULT	17500 /* cycles */
+#define DMA_RX_TIMEOUT_MASK	GENMASK(16, 0)
+#define DMA_RX_TIMEOUT_SHIFT	0
+
+#define CHAN_HAS_EPIB		BIT(30)
+#define CHAN_HAS_PSINFO		BIT(29)
+#define CHAN_ERR_RETRY		BIT(28)
+#define CHAN_PSINFO_AT_SOP	BIT(25)
+#define CHAN_SOP_OFF_SHIFT	16
+#define CHAN_SOP_OFF_MASK	GENMASK(9, 0)
+#define DESC_TYPE_SHIFT		26
+#define DESC_TYPE_MASK		GENMASK(2, 0)
+
+/*
+ * QMGR & QNUM together make up 14 bits with QMGR as the 2 MSb's in the logical
+ * navigator cloud mapping scheme.
+ * using the 14bit physical queue numbers directly maps into this scheme.
+ */
+#define CHAN_QNUM_MASK		GENMASK(14, 0)
+#define DMA_MAX_QMS		4
+#define DMA_TIMEOUT		1	/* msecs */
+#define DMA_INVALID_ID		0xffff
+
+struct reg_global {
+	u32	revision;
+	u32	perf_control;
+	u32	emulation_control;
+	u32	priority_control;
+	u32	qm_base_address[DMA_MAX_QMS];
+};
+
+struct reg_chan {
+	u32	control;
+	u32	mode;
+	u32	__rsvd[6];
+};
+
+struct reg_tx_sched {
+	u32	prio;
+};
+
+struct reg_rx_flow {
+	u32	control;
+	u32	tags;
+	u32	tag_sel;
+	u32	fdq_sel[2];
+	u32	thresh[3];
+};
+
+struct knav_dma_pool_device {
+	struct device			*dev;
+	struct list_head		list;
+};
+
+struct knav_dma_device {
+	bool				loopback, enable_all;
+	unsigned			tx_priority, rx_priority, rx_timeout;
+	unsigned			logical_queue_managers;
+	unsigned			qm_base_address[DMA_MAX_QMS];
+	struct reg_global __iomem	*reg_global;
+	struct reg_chan __iomem		*reg_tx_chan;
+	struct reg_rx_flow __iomem	*reg_rx_flow;
+	struct reg_chan __iomem		*reg_rx_chan;
+	struct reg_tx_sched __iomem	*reg_tx_sched;
+	unsigned			max_rx_chan, max_tx_chan;
+	unsigned			max_rx_flow;
+	char				name[32];
+	atomic_t			ref_count;
+	struct list_head		list;
+	struct list_head		chan_list;
+	spinlock_t			lock;
+};
+
+struct knav_dma_chan {
+	enum dma_transfer_direction	direction;
+	struct knav_dma_device		*dma;
+	atomic_t			ref_count;
+
+	/* registers */
+	struct reg_chan __iomem		*reg_chan;
+	struct reg_tx_sched __iomem	*reg_tx_sched;
+	struct reg_rx_flow __iomem	*reg_rx_flow;
+
+	/* configuration stuff */
+	unsigned			channel, flow;
+	struct knav_dma_cfg		cfg;
+	struct list_head		list;
+	spinlock_t			lock;
+};
+
+#define chan_number(ch)	((ch->direction == DMA_MEM_TO_DEV) ? \
+			ch->channel : ch->flow)
+
+static struct knav_dma_pool_device *kdev;
+
+static bool device_ready;
+bool knav_dma_device_ready(void)
+{
+	return device_ready;
+}
+EXPORT_SYMBOL_GPL(knav_dma_device_ready);
+
+static bool check_config(struct knav_dma_chan *chan, struct knav_dma_cfg *cfg)
+{
+	if (!memcmp(&chan->cfg, cfg, sizeof(*cfg)))
+		return true;
+	else
+		return false;
+}
+
+static int chan_start(struct knav_dma_chan *chan,
+			struct knav_dma_cfg *cfg)
+{
+	u32 v = 0;
+
+	spin_lock(&chan->lock);
+	if ((chan->direction == DMA_MEM_TO_DEV) && chan->reg_chan) {
+		if (cfg->u.tx.filt_pswords)
+			v |= DMA_TX_FILT_PSWORDS;
+		if (cfg->u.tx.filt_einfo)
+			v |= DMA_TX_FILT_EINFO;
+		writel_relaxed(v, &chan->reg_chan->mode);
+		writel_relaxed(DMA_ENABLE, &chan->reg_chan->control);
+	}
+
+	if (chan->reg_tx_sched)
+		writel_relaxed(cfg->u.tx.priority, &chan->reg_tx_sched->prio);
+
+	if (chan->reg_rx_flow) {
+		v = 0;
+
+		if (cfg->u.rx.einfo_present)
+			v |= CHAN_HAS_EPIB;
+		if (cfg->u.rx.psinfo_present)
+			v |= CHAN_HAS_PSINFO;
+		if (cfg->u.rx.err_mode == DMA_RETRY)
+			v |= CHAN_ERR_RETRY;
+		v |= (cfg->u.rx.desc_type & DESC_TYPE_MASK) << DESC_TYPE_SHIFT;
+		if (cfg->u.rx.psinfo_at_sop)
+			v |= CHAN_PSINFO_AT_SOP;
+		v |= (cfg->u.rx.sop_offset & CHAN_SOP_OFF_MASK)
+			<< CHAN_SOP_OFF_SHIFT;
+		v |= cfg->u.rx.dst_q & CHAN_QNUM_MASK;
+
+		writel_relaxed(v, &chan->reg_rx_flow->control);
+		writel_relaxed(0, &chan->reg_rx_flow->tags);
+		writel_relaxed(0, &chan->reg_rx_flow->tag_sel);
+
+		v =  cfg->u.rx.fdq[0] << 16;
+		v |=  cfg->u.rx.fdq[1] & CHAN_QNUM_MASK;
+		writel_relaxed(v, &chan->reg_rx_flow->fdq_sel[0]);
+
+		v =  cfg->u.rx.fdq[2] << 16;
+		v |=  cfg->u.rx.fdq[3] & CHAN_QNUM_MASK;
+		writel_relaxed(v, &chan->reg_rx_flow->fdq_sel[1]);
+
+		writel_relaxed(0, &chan->reg_rx_flow->thresh[0]);
+		writel_relaxed(0, &chan->reg_rx_flow->thresh[1]);
+		writel_relaxed(0, &chan->reg_rx_flow->thresh[2]);
+	}
+
+	/* Keep a copy of the cfg */
+	memcpy(&chan->cfg, cfg, sizeof(*cfg));
+	spin_unlock(&chan->lock);
+
+	return 0;
+}
+
+static int chan_teardown(struct knav_dma_chan *chan)
+{
+	unsigned long end, value;
+
+	if (!chan->reg_chan)
+		return 0;
+
+	/* indicate teardown */
+	writel_relaxed(DMA_TEARDOWN, &chan->reg_chan->control);
+
+	/* wait for the dma to shut itself down */
+	end = jiffies + msecs_to_jiffies(DMA_TIMEOUT);
+	do {
+		value = readl_relaxed(&chan->reg_chan->control);
+		if ((value & DMA_ENABLE) == 0)
+			break;
+	} while (time_after(end, jiffies));
+
+	if (readl_relaxed(&chan->reg_chan->control) & DMA_ENABLE) {
+		dev_err(kdev->dev, "timeout waiting for teardown\n");
+		return -ETIMEDOUT;
+	}
+
+	return 0;
+}
+
+static void chan_stop(struct knav_dma_chan *chan)
+{
+	spin_lock(&chan->lock);
+	if (chan->reg_rx_flow) {
+		/* first detach fdqs, starve out the flow */
+		writel_relaxed(0, &chan->reg_rx_flow->fdq_sel[0]);
+		writel_relaxed(0, &chan->reg_rx_flow->fdq_sel[1]);
+		writel_relaxed(0, &chan->reg_rx_flow->thresh[0]);
+		writel_relaxed(0, &chan->reg_rx_flow->thresh[1]);
+		writel_relaxed(0, &chan->reg_rx_flow->thresh[2]);
+	}
+
+	/* teardown the dma channel */
+	chan_teardown(chan);
+
+	/* then disconnect the completion side */
+	if (chan->reg_rx_flow) {
+		writel_relaxed(0, &chan->reg_rx_flow->control);
+		writel_relaxed(0, &chan->reg_rx_flow->tags);
+		writel_relaxed(0, &chan->reg_rx_flow->tag_sel);
+	}
+
+	memset(&chan->cfg, 0, sizeof(struct knav_dma_cfg));
+	spin_unlock(&chan->lock);
+
+	dev_dbg(kdev->dev, "channel stopped\n");
+}
+
+static void dma_hw_enable_all(struct knav_dma_device *dma)
+{
+	int i;
+
+	for (i = 0; i < dma->max_tx_chan; i++) {
+		writel_relaxed(0, &dma->reg_tx_chan[i].mode);
+		writel_relaxed(DMA_ENABLE, &dma->reg_tx_chan[i].control);
+	}
+}
+
+
+static void knav_dma_hw_init(struct knav_dma_device *dma)
+{
+	unsigned v;
+	int i;
+
+	spin_lock(&dma->lock);
+	v  = dma->loopback ? DMA_LOOPBACK : 0;
+	writel_relaxed(v, &dma->reg_global->emulation_control);
+
+	v = readl_relaxed(&dma->reg_global->perf_control);
+	v |= ((dma->rx_timeout & DMA_RX_TIMEOUT_MASK) << DMA_RX_TIMEOUT_SHIFT);
+	writel_relaxed(v, &dma->reg_global->perf_control);
+
+	v = ((dma->tx_priority << DMA_TX_PRIO_SHIFT) |
+	     (dma->rx_priority << DMA_RX_PRIO_SHIFT));
+
+	writel_relaxed(v, &dma->reg_global->priority_control);
+
+	/* Always enable all Rx channels. Rx paths are managed using flows */
+	for (i = 0; i < dma->max_rx_chan; i++)
+		writel_relaxed(DMA_ENABLE, &dma->reg_rx_chan[i].control);
+
+	for (i = 0; i < dma->logical_queue_managers; i++)
+		writel_relaxed(dma->qm_base_address[i],
+			       &dma->reg_global->qm_base_address[i]);
+	spin_unlock(&dma->lock);
+}
+
+static void knav_dma_hw_destroy(struct knav_dma_device *dma)
+{
+	int i;
+	unsigned v;
+
+	spin_lock(&dma->lock);
+	v = ~DMA_ENABLE & REG_MASK;
+
+	for (i = 0; i < dma->max_rx_chan; i++)
+		writel_relaxed(v, &dma->reg_rx_chan[i].control);
+
+	for (i = 0; i < dma->max_tx_chan; i++)
+		writel_relaxed(v, &dma->reg_tx_chan[i].control);
+	spin_unlock(&dma->lock);
+}
+
+static void dma_debug_show_channels(struct seq_file *s,
+					struct knav_dma_chan *chan)
+{
+	int i;
+
+	seq_printf(s, "\t%s %d:\t",
+		((chan->direction == DMA_MEM_TO_DEV) ? "tx chan" : "rx flow"),
+		chan_number(chan));
+
+	if (chan->direction == DMA_MEM_TO_DEV) {
+		seq_printf(s, "einfo - %d, pswords - %d, priority - %d\n",
+			chan->cfg.u.tx.filt_einfo,
+			chan->cfg.u.tx.filt_pswords,
+			chan->cfg.u.tx.priority);
+	} else {
+		seq_printf(s, "einfo - %d, psinfo - %d, desc_type - %d\n",
+			chan->cfg.u.rx.einfo_present,
+			chan->cfg.u.rx.psinfo_present,
+			chan->cfg.u.rx.desc_type);
+		seq_printf(s, "\t\t\tdst_q: [%d], thresh: %d fdq: ",
+			chan->cfg.u.rx.dst_q,
+			chan->cfg.u.rx.thresh);
+		for (i = 0; i < KNAV_DMA_FDQ_PER_CHAN; i++)
+			seq_printf(s, "[%d]", chan->cfg.u.rx.fdq[i]);
+		seq_printf(s, "\n");
+	}
+}
+
+static void dma_debug_show_devices(struct seq_file *s,
+					struct knav_dma_device *dma)
+{
+	struct knav_dma_chan *chan;
+
+	list_for_each_entry(chan, &dma->chan_list, list) {
+		if (atomic_read(&chan->ref_count))
+			dma_debug_show_channels(s, chan);
+	}
+}
+
+static int dma_debug_show(struct seq_file *s, void *v)
+{
+	struct knav_dma_device *dma;
+
+	list_for_each_entry(dma, &kdev->list, list) {
+		if (atomic_read(&dma->ref_count)) {
+			seq_printf(s, "%s : max_tx_chan: (%d), max_rx_flows: (%d)\n",
+			dma->name, dma->max_tx_chan, dma->max_rx_flow);
+			dma_debug_show_devices(s, dma);
+		}
+	}
+
+	return 0;
+}
+
+static int knav_dma_debug_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, dma_debug_show, NULL);
+}
+
+static const struct file_operations knav_dma_debug_ops = {
+	.open		= knav_dma_debug_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+static int of_channel_match_helper(struct device_node *np, const char *name,
+					const char **dma_instance)
+{
+	struct of_phandle_args args;
+	struct device_node *dma_node;
+	int index;
+
+	dma_node = of_parse_phandle(np, "ti,navigator-dmas", 0);
+	if (!dma_node)
+		return -ENODEV;
+
+	*dma_instance = dma_node->name;
+	index = of_property_match_string(np, "ti,navigator-dma-names", name);
+	if (index < 0) {
+		dev_err(kdev->dev, "No 'ti,navigator-dma-names' property\n");
+		return -ENODEV;
+	}
+
+	if (of_parse_phandle_with_fixed_args(np, "ti,navigator-dmas",
+					1, index, &args)) {
+		dev_err(kdev->dev, "Missing the phandle args name %s\n", name);
+		return -ENODEV;
+	}
+
+	if (args.args[0] < 0) {
+		dev_err(kdev->dev, "Missing args for %s\n", name);
+		return -ENODEV;
+	}
+
+	return args.args[0];
+}
+
+/**
+ * knav_dma_open_channel() - try to setup an exclusive slave channel
+ * @dev:	pointer to client device structure
+ * @name:	slave channel name
+ * @config:	dma configuration parameters
+ *
+ * Returns pointer to appropriate DMA channel on success or error.
+ */
+void *knav_dma_open_channel(struct device *dev, const char *name,
+					struct knav_dma_cfg *config)
+{
+	struct knav_dma_chan *chan;
+	struct knav_dma_device *dma;
+	bool found = false;
+	int chan_num = -1;
+	const char *instance;
+
+	if (!kdev) {
+		pr_err("keystone-navigator-dma driver not registered\n");
+		return (void *)-EINVAL;
+	}
+
+	chan_num = of_channel_match_helper(dev->of_node, name, &instance);
+	if (chan_num < 0) {
+		dev_err(kdev->dev, "No DMA instace with name %s\n", name);
+		return (void *)-EINVAL;
+	}
+
+	dev_dbg(kdev->dev, "initializing %s channel %d from DMA %s\n",
+		  config->direction == DMA_MEM_TO_DEV ? "transmit" :
+		  config->direction == DMA_DEV_TO_MEM ? "receive"  :
+		  "unknown", chan_num, instance);
+
+	if (config->direction != DMA_MEM_TO_DEV &&
+	    config->direction != DMA_DEV_TO_MEM) {
+		dev_err(kdev->dev, "bad direction\n");
+		return (void *)-EINVAL;
+	}
+
+	/* Look for correct dma instance */
+	list_for_each_entry(dma, &kdev->list, list) {
+		if (!strcmp(dma->name, instance)) {
+			found = true;
+			break;
+		}
+	}
+	if (!found) {
+		dev_err(kdev->dev, "No DMA instace with name %s\n", instance);
+		return (void *)-EINVAL;
+	}
+
+	/* Look for correct dma channel from dma instance */
+	found = false;
+	list_for_each_entry(chan, &dma->chan_list, list) {
+		if (config->direction == DMA_MEM_TO_DEV) {
+			if (chan->channel == chan_num) {
+				found = true;
+				break;
+			}
+		} else {
+			if (chan->flow == chan_num) {
+				found = true;
+				break;
+			}
+		}
+	}
+	if (!found) {
+		dev_err(kdev->dev, "channel %d is not in DMA %s\n",
+				chan_num, instance);
+		return (void *)-EINVAL;
+	}
+
+	if (atomic_read(&chan->ref_count) >= 1) {
+		if (!check_config(chan, config)) {
+			dev_err(kdev->dev, "channel %d config miss-match\n",
+				chan_num);
+			return (void *)-EINVAL;
+		}
+	}
+
+	if (atomic_inc_return(&chan->dma->ref_count) <= 1)
+		knav_dma_hw_init(chan->dma);
+
+	if (atomic_inc_return(&chan->ref_count) <= 1)
+		chan_start(chan, config);
+
+	dev_dbg(kdev->dev, "channel %d opened from DMA %s\n",
+				chan_num, instance);
+
+	return chan;
+}
+EXPORT_SYMBOL_GPL(knav_dma_open_channel);
+
+/**
+ * knav_dma_close_channel()	- Destroy a dma channel
+ *
+ * channel:	dma channel handle
+ *
+ */
+void knav_dma_close_channel(void *channel)
+{
+	struct knav_dma_chan *chan = channel;
+
+	if (!kdev) {
+		pr_err("keystone-navigator-dma driver not registered\n");
+		return;
+	}
+
+	if (atomic_dec_return(&chan->ref_count) <= 0)
+		chan_stop(chan);
+
+	if (atomic_dec_return(&chan->dma->ref_count) <= 0)
+		knav_dma_hw_destroy(chan->dma);
+
+	dev_dbg(kdev->dev, "channel %d or flow %d closed from DMA %s\n",
+			chan->channel, chan->flow, chan->dma->name);
+}
+EXPORT_SYMBOL_GPL(knav_dma_close_channel);
+
+static void __iomem *pktdma_get_regs(struct knav_dma_device *dma,
+				struct device_node *node,
+				unsigned index, resource_size_t *_size)
+{
+	struct device *dev = kdev->dev;
+	struct resource res;
+	void __iomem *regs;
+	int ret;
+
+	ret = of_address_to_resource(node, index, &res);
+	if (ret) {
+		dev_err(dev, "Can't translate of node(%s) address for index(%d)\n",
+			node->name, index);
+		return ERR_PTR(ret);
+	}
+
+	regs = devm_ioremap_resource(kdev->dev, &res);
+	if (IS_ERR(regs))
+		dev_err(dev, "Failed to map register base for index(%d) node(%s)\n",
+			index, node->name);
+	if (_size)
+		*_size = resource_size(&res);
+
+	return regs;
+}
+
+static int pktdma_init_rx_chan(struct knav_dma_chan *chan, u32 flow)
+{
+	struct knav_dma_device *dma = chan->dma;
+
+	chan->flow = flow;
+	chan->reg_rx_flow = dma->reg_rx_flow + flow;
+	chan->channel = DMA_INVALID_ID;
+	dev_dbg(kdev->dev, "rx flow(%d) (%p)\n", chan->flow, chan->reg_rx_flow);
+
+	return 0;
+}
+
+static int pktdma_init_tx_chan(struct knav_dma_chan *chan, u32 channel)
+{
+	struct knav_dma_device *dma = chan->dma;
+
+	chan->channel = channel;
+	chan->reg_chan = dma->reg_tx_chan + channel;
+	chan->reg_tx_sched = dma->reg_tx_sched + channel;
+	chan->flow = DMA_INVALID_ID;
+	dev_dbg(kdev->dev, "tx channel(%d) (%p)\n", chan->channel, chan->reg_chan);
+
+	return 0;
+}
+
+static int pktdma_init_chan(struct knav_dma_device *dma,
+				enum dma_transfer_direction dir,
+				unsigned chan_num)
+{
+	struct device *dev = kdev->dev;
+	struct knav_dma_chan *chan;
+	int ret = -EINVAL;
+
+	chan = devm_kzalloc(dev, sizeof(*chan), GFP_KERNEL);
+	if (!chan)
+		return -ENOMEM;
+
+	INIT_LIST_HEAD(&chan->list);
+	chan->dma	= dma;
+	chan->direction	= DMA_NONE;
+	atomic_set(&chan->ref_count, 0);
+	spin_lock_init(&chan->lock);
+
+	if (dir == DMA_MEM_TO_DEV) {
+		chan->direction = dir;
+		ret = pktdma_init_tx_chan(chan, chan_num);
+	} else if (dir == DMA_DEV_TO_MEM) {
+		chan->direction = dir;
+		ret = pktdma_init_rx_chan(chan, chan_num);
+	} else {
+		dev_err(dev, "channel(%d) direction unknown\n", chan_num);
+	}
+
+	list_add_tail(&chan->list, &dma->chan_list);
+
+	return ret;
+}
+
+static int dma_init(struct device_node *cloud, struct device_node *dma_node)
+{
+	unsigned max_tx_chan, max_rx_chan, max_rx_flow, max_tx_sched;
+	struct device_node *node = dma_node;
+	struct knav_dma_device *dma;
+	int ret, len, num_chan = 0;
+	resource_size_t size;
+	u32 timeout;
+	u32 i;
+
+	dma = devm_kzalloc(kdev->dev, sizeof(*dma), GFP_KERNEL);
+	if (!dma) {
+		dev_err(kdev->dev, "could not allocate driver mem\n");
+		return -ENOMEM;
+	}
+	INIT_LIST_HEAD(&dma->list);
+	INIT_LIST_HEAD(&dma->chan_list);
+
+	if (!of_find_property(cloud, "ti,navigator-cloud-address", &len)) {
+		dev_err(kdev->dev, "unspecified navigator cloud addresses\n");
+		return -ENODEV;
+	}
+
+	dma->logical_queue_managers = len / sizeof(u32);
+	if (dma->logical_queue_managers > DMA_MAX_QMS) {
+		dev_warn(kdev->dev, "too many queue mgrs(>%d) rest ignored\n",
+			 dma->logical_queue_managers);
+		dma->logical_queue_managers = DMA_MAX_QMS;
+	}
+
+	ret = of_property_read_u32_array(cloud, "ti,navigator-cloud-address",
+					dma->qm_base_address,
+					dma->logical_queue_managers);
+	if (ret) {
+		dev_err(kdev->dev, "invalid navigator cloud addresses\n");
+		return -ENODEV;
+	}
+
+	dma->reg_global	 = pktdma_get_regs(dma, node, 0, &size);
+	if (!dma->reg_global)
+		return -ENODEV;
+	if (size < sizeof(struct reg_global)) {
+		dev_err(kdev->dev, "bad size %pa for global regs\n", &size);
+		return -ENODEV;
+	}
+
+	dma->reg_tx_chan = pktdma_get_regs(dma, node, 1, &size);
+	if (!dma->reg_tx_chan)
+		return -ENODEV;
+
+	max_tx_chan = size / sizeof(struct reg_chan);
+	dma->reg_rx_chan = pktdma_get_regs(dma, node, 2, &size);
+	if (!dma->reg_rx_chan)
+		return -ENODEV;
+
+	max_rx_chan = size / sizeof(struct reg_chan);
+	dma->reg_tx_sched = pktdma_get_regs(dma, node, 3, &size);
+	if (!dma->reg_tx_sched)
+		return -ENODEV;
+
+	max_tx_sched = size / sizeof(struct reg_tx_sched);
+	dma->reg_rx_flow = pktdma_get_regs(dma, node, 4, &size);
+	if (!dma->reg_rx_flow)
+		return -ENODEV;
+
+	max_rx_flow = size / sizeof(struct reg_rx_flow);
+	dma->rx_priority = DMA_PRIO_DEFAULT;
+	dma->tx_priority = DMA_PRIO_DEFAULT;
+
+	dma->enable_all	= (of_get_property(node, "ti,enable-all", NULL) != NULL);
+	dma->loopback	= (of_get_property(node, "ti,loop-back",  NULL) != NULL);
+
+	ret = of_property_read_u32(node, "ti,rx-retry-timeout", &timeout);
+	if (ret < 0) {
+		dev_dbg(kdev->dev, "unspecified rx timeout using value %d\n",
+			DMA_RX_TIMEOUT_DEFAULT);
+		timeout = DMA_RX_TIMEOUT_DEFAULT;
+	}
+
+	dma->rx_timeout = timeout;
+	dma->max_rx_chan = max_rx_chan;
+	dma->max_rx_flow = max_rx_flow;
+	dma->max_tx_chan = min(max_tx_chan, max_tx_sched);
+	atomic_set(&dma->ref_count, 0);
+	strcpy(dma->name, node->name);
+	spin_lock_init(&dma->lock);
+
+	for (i = 0; i < dma->max_tx_chan; i++) {
+		if (pktdma_init_chan(dma, DMA_MEM_TO_DEV, i) >= 0)
+			num_chan++;
+	}
+
+	for (i = 0; i < dma->max_rx_flow; i++) {
+		if (pktdma_init_chan(dma, DMA_DEV_TO_MEM, i) >= 0)
+			num_chan++;
+	}
+
+	list_add_tail(&dma->list, &kdev->list);
+
+	/*
+	 * For DSP software usecases or userpace transport software, setup all
+	 * the DMA hardware resources.
+	 */
+	if (dma->enable_all) {
+		atomic_inc(&dma->ref_count);
+		knav_dma_hw_init(dma);
+		dma_hw_enable_all(dma);
+	}
+
+	dev_info(kdev->dev, "DMA %s registered %d logical channels, flows %d, tx chans: %d, rx chans: %d%s\n",
+		dma->name, num_chan, dma->max_rx_flow,
+		dma->max_tx_chan, dma->max_rx_chan,
+		dma->loopback ? ", loopback" : "");
+
+	return 0;
+}
+
+static int knav_dma_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct device_node *node = pdev->dev.of_node;
+	struct device_node *child;
+	int ret = 0;
+
+	if (!node) {
+		dev_err(&pdev->dev, "could not find device info\n");
+		return -EINVAL;
+	}
+
+	kdev = devm_kzalloc(dev,
+			sizeof(struct knav_dma_pool_device), GFP_KERNEL);
+	if (!kdev) {
+		dev_err(dev, "could not allocate driver mem\n");
+		return -ENOMEM;
+	}
+
+	kdev->dev = dev;
+	INIT_LIST_HEAD(&kdev->list);
+
+	pm_runtime_enable(kdev->dev);
+	ret = pm_runtime_get_sync(kdev->dev);
+	if (ret < 0) {
+		pm_runtime_put_noidle(kdev->dev);
+		dev_err(kdev->dev, "unable to enable pktdma, err %d\n", ret);
+		goto err_pm_disable;
+	}
+
+	/* Initialise all packet dmas */
+	for_each_child_of_node(node, child) {
+		ret = dma_init(node, child);
+		if (ret) {
+			dev_err(&pdev->dev, "init failed with %d\n", ret);
+			break;
+		}
+	}
+
+	if (list_empty(&kdev->list)) {
+		dev_err(dev, "no valid dma instance\n");
+		ret = -ENODEV;
+		goto err_put_sync;
+	}
+
+	debugfs_create_file("knav_dma", S_IFREG | S_IRUGO, NULL, NULL,
+			    &knav_dma_debug_ops);
+
+	device_ready = true;
+	return ret;
+
+err_put_sync:
+	pm_runtime_put_sync(kdev->dev);
+err_pm_disable:
+	pm_runtime_disable(kdev->dev);
+
+	return ret;
+}
+
+static int knav_dma_remove(struct platform_device *pdev)
+{
+	struct knav_dma_device *dma;
+
+	list_for_each_entry(dma, &kdev->list, list) {
+		if (atomic_dec_return(&dma->ref_count) == 0)
+			knav_dma_hw_destroy(dma);
+	}
+
+	pm_runtime_put_sync(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+
+	return 0;
+}
+
+static struct of_device_id of_match[] = {
+	{ .compatible = "ti,keystone-navigator-dma", },
+	{},
+};
+
+MODULE_DEVICE_TABLE(of, of_match);
+
+static struct platform_driver knav_dma_driver = {
+	.probe	= knav_dma_probe,
+	.remove	= knav_dma_remove,
+	.driver = {
+		.name		= "keystone-navigator-dma",
+		.of_match_table	= of_match,
+	},
+};
+module_platform_driver(knav_dma_driver);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("TI Keystone Navigator Packet DMA driver");
+MODULE_AUTHOR("Sandeep Nair <sandeep_n@ti.com>");
+MODULE_AUTHOR("Santosh Shilimkar <santosh.shilimkar@ti.com>");
diff --git a/drivers/soc/ti/knav_qmss.h b/drivers/soc/ti/knav_qmss.h
new file mode 100644
index 000000000..bd040c29c
--- /dev/null
+++ b/drivers/soc/ti/knav_qmss.h
@@ -0,0 +1,395 @@
+/*
+ * Keystone Navigator QMSS driver internal header
+ *
+ * Copyright (C) 2014 Texas Instruments Incorporated - http://www.ti.com
+ * Author:	Sandeep Nair <sandeep_n@ti.com>
+ *		Cyril Chemparathy <cyril@ti.com>
+ *		Santosh Shilimkar <santosh.shilimkar@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * version 2 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.
+ */
+
+#ifndef __KNAV_QMSS_H__
+#define __KNAV_QMSS_H__
+
+#include <linux/percpu.h>
+
+#define THRESH_GTE	BIT(7)
+#define THRESH_LT	0
+
+#define PDSP_CTRL_PC_MASK	0xffff0000
+#define PDSP_CTRL_SOFT_RESET	BIT(0)
+#define PDSP_CTRL_ENABLE	BIT(1)
+#define PDSP_CTRL_RUNNING	BIT(15)
+
+#define ACC_MAX_CHANNEL		48
+#define ACC_DEFAULT_PERIOD	25 /* usecs */
+
+#define ACC_CHANNEL_INT_BASE		2
+
+#define ACC_LIST_ENTRY_TYPE		1
+#define ACC_LIST_ENTRY_WORDS		(1 << ACC_LIST_ENTRY_TYPE)
+#define ACC_LIST_ENTRY_QUEUE_IDX	0
+#define ACC_LIST_ENTRY_DESC_IDX	(ACC_LIST_ENTRY_WORDS - 1)
+
+#define ACC_CMD_DISABLE_CHANNEL	0x80
+#define ACC_CMD_ENABLE_CHANNEL	0x81
+#define ACC_CFG_MULTI_QUEUE		BIT(21)
+
+#define ACC_INTD_OFFSET_EOI		(0x0010)
+#define ACC_INTD_OFFSET_COUNT(ch)	(0x0300 + 4 * (ch))
+#define ACC_INTD_OFFSET_STATUS(ch)	(0x0200 + 4 * ((ch) / 32))
+
+#define RANGE_MAX_IRQS			64
+
+#define ACC_DESCS_MAX		SZ_1K
+#define ACC_DESCS_MASK		(ACC_DESCS_MAX - 1)
+#define DESC_SIZE_MASK		0xful
+#define DESC_PTR_MASK		(~DESC_SIZE_MASK)
+
+#define KNAV_NAME_SIZE			32
+
+enum knav_acc_result {
+	ACC_RET_IDLE,
+	ACC_RET_SUCCESS,
+	ACC_RET_INVALID_COMMAND,
+	ACC_RET_INVALID_CHANNEL,
+	ACC_RET_INACTIVE_CHANNEL,
+	ACC_RET_ACTIVE_CHANNEL,
+	ACC_RET_INVALID_QUEUE,
+	ACC_RET_INVALID_RET,
+};
+
+struct knav_reg_config {
+	u32		revision;
+	u32		__pad1;
+	u32		divert;
+	u32		link_ram_base0;
+	u32		link_ram_size0;
+	u32		link_ram_base1;
+	u32		__pad2[2];
+	u32		starvation[0];
+};
+
+struct knav_reg_region {
+	u32		base;
+	u32		start_index;
+	u32		size_count;
+	u32		__pad;
+};
+
+struct knav_reg_pdsp_regs {
+	u32		control;
+	u32		status;
+	u32		cycle_count;
+	u32		stall_count;
+};
+
+struct knav_reg_acc_command {
+	u32		command;
+	u32		queue_mask;
+	u32		list_dma;
+	u32		queue_num;
+	u32		timer_config;
+};
+
+struct knav_link_ram_block {
+	dma_addr_t	 dma;
+	void		*virt;
+	size_t		 size;
+};
+
+struct knav_acc_info {
+	u32			 pdsp_id;
+	u32			 start_channel;
+	u32			 list_entries;
+	u32			 pacing_mode;
+	u32			 timer_count;
+	int			 mem_size;
+	int			 list_size;
+	struct knav_pdsp_info	*pdsp;
+};
+
+struct knav_acc_channel {
+	u32			channel;
+	u32			list_index;
+	u32			open_mask;
+	u32			*list_cpu[2];
+	dma_addr_t		list_dma[2];
+	char			name[KNAV_NAME_SIZE];
+	atomic_t		retrigger_count;
+};
+
+struct knav_pdsp_info {
+	const char					*name;
+	struct knav_reg_pdsp_regs  __iomem		*regs;
+	union {
+		void __iomem				*command;
+		struct knav_reg_acc_command __iomem	*acc_command;
+		u32 __iomem				*qos_command;
+	};
+	void __iomem					*intd;
+	u32 __iomem					*iram;
+	u32						id;
+	struct list_head				list;
+	bool						loaded;
+	bool						started;
+};
+
+struct knav_qmgr_info {
+	unsigned			start_queue;
+	unsigned			num_queues;
+	struct knav_reg_config __iomem	*reg_config;
+	struct knav_reg_region __iomem	*reg_region;
+	struct knav_reg_queue __iomem	*reg_push, *reg_pop, *reg_peek;
+	void __iomem			*reg_status;
+	struct list_head		list;
+};
+
+#define KNAV_NUM_LINKRAM	2
+
+/**
+ * struct knav_queue_stats:	queue statistics
+ * pushes:			number of push operations
+ * pops:			number of pop operations
+ * push_errors:			number of push errors
+ * pop_errors:			number of pop errors
+ * notifies:			notifier counts
+ */
+struct knav_queue_stats {
+	unsigned int pushes;
+	unsigned int pops;
+	unsigned int push_errors;
+	unsigned int pop_errors;
+	unsigned int notifies;
+};
+
+/**
+ * struct knav_reg_queue:	queue registers
+ * @entry_count:		valid entries in the queue
+ * @byte_count:			total byte count in thhe queue
+ * @packet_size:		packet size for the queue
+ * @ptr_size_thresh:		packet pointer size threshold
+ */
+struct knav_reg_queue {
+	u32		entry_count;
+	u32		byte_count;
+	u32		packet_size;
+	u32		ptr_size_thresh;
+};
+
+/**
+ * struct knav_region:		qmss region info
+ * @dma_start, dma_end:		start and end dma address
+ * @virt_start, virt_end:	start and end virtual address
+ * @desc_size:			descriptor size
+ * @used_desc:			consumed descriptors
+ * @id:				region number
+ * @num_desc:			total descriptors
+ * @link_index:			index of the first descriptor
+ * @name:			region name
+ * @list:			instance in the device's region list
+ * @pools:			list of descriptor pools in the region
+ */
+struct knav_region {
+	dma_addr_t		dma_start, dma_end;
+	void			*virt_start, *virt_end;
+	unsigned		desc_size;
+	unsigned		used_desc;
+	unsigned		id;
+	unsigned		num_desc;
+	unsigned		link_index;
+	const char		*name;
+	struct list_head	list;
+	struct list_head	pools;
+};
+
+/**
+ * struct knav_pool:		qmss pools
+ * @dev:			device pointer
+ * @region:			qmss region info
+ * @queue:			queue registers
+ * @kdev:			qmss device pointer
+ * @region_offset:		offset from the base
+ * @num_desc:			total descriptors
+ * @desc_size:			descriptor size
+ * @region_id:			region number
+ * @name:			pool name
+ * @list:			list head
+ * @region_inst:		instance in the region's pool list
+ */
+struct knav_pool {
+	struct device			*dev;
+	struct knav_region		*region;
+	struct knav_queue		*queue;
+	struct knav_device		*kdev;
+	int				region_offset;
+	int				num_desc;
+	int				desc_size;
+	int				region_id;
+	const char			*name;
+	struct list_head		list;
+	struct list_head		region_inst;
+};
+
+/**
+ * struct knav_queue_inst:		qmss queue instace properties
+ * @descs:				descriptor pointer
+ * @desc_head, desc_tail, desc_count:	descriptor counters
+ * @acc:				accumulator channel pointer
+ * @kdev:				qmss device pointer
+ * @range:				range info
+ * @qmgr:				queue manager info
+ * @id:					queue instace id
+ * @irq_num:				irq line number
+ * @notify_needed:			notifier needed based on queue type
+ * @num_notifiers:			total notifiers
+ * @handles:				list head
+ * @name:				queue instance name
+ * @irq_name:				irq line name
+ */
+struct knav_queue_inst {
+	u32				*descs;
+	atomic_t			desc_head, desc_tail, desc_count;
+	struct knav_acc_channel	*acc;
+	struct knav_device		*kdev;
+	struct knav_range_info		*range;
+	struct knav_qmgr_info		*qmgr;
+	u32				id;
+	int				irq_num;
+	int				notify_needed;
+	atomic_t			num_notifiers;
+	struct list_head		handles;
+	const char			*name;
+	const char			*irq_name;
+};
+
+/**
+ * struct knav_queue:			qmss queue properties
+ * @reg_push, reg_pop, reg_peek:	push, pop queue registers
+ * @inst:				qmss queue instace properties
+ * @notifier_fn:			notifier function
+ * @notifier_fn_arg:			notifier function argument
+ * @notifier_enabled:			notier enabled for a give queue
+ * @rcu:				rcu head
+ * @flags:				queue flags
+ * @list:				list head
+ */
+struct knav_queue {
+	struct knav_reg_queue __iomem	*reg_push, *reg_pop, *reg_peek;
+	struct knav_queue_inst		*inst;
+	struct knav_queue_stats __percpu	*stats;
+	knav_queue_notify_fn		notifier_fn;
+	void				*notifier_fn_arg;
+	atomic_t			notifier_enabled;
+	struct rcu_head			rcu;
+	unsigned			flags;
+	struct list_head		list;
+};
+
+enum qmss_version {
+	QMSS,
+	QMSS_66AK2G,
+};
+
+struct knav_device {
+	struct device				*dev;
+	unsigned				base_id;
+	unsigned				num_queues;
+	unsigned				num_queues_in_use;
+	unsigned				inst_shift;
+	struct knav_link_ram_block		link_rams[KNAV_NUM_LINKRAM];
+	void					*instances;
+	struct list_head			regions;
+	struct list_head			queue_ranges;
+	struct list_head			pools;
+	struct list_head			pdsps;
+	struct list_head			qmgrs;
+	enum qmss_version			version;
+};
+
+struct knav_range_ops {
+	int	(*init_range)(struct knav_range_info *range);
+	int	(*free_range)(struct knav_range_info *range);
+	int	(*init_queue)(struct knav_range_info *range,
+			      struct knav_queue_inst *inst);
+	int	(*open_queue)(struct knav_range_info *range,
+			      struct knav_queue_inst *inst, unsigned flags);
+	int	(*close_queue)(struct knav_range_info *range,
+			       struct knav_queue_inst *inst);
+	int	(*set_notify)(struct knav_range_info *range,
+			      struct knav_queue_inst *inst, bool enabled);
+};
+
+struct knav_irq_info {
+	int		irq;
+	struct cpumask	*cpu_mask;
+};
+
+struct knav_range_info {
+	const char			*name;
+	struct knav_device		*kdev;
+	unsigned			queue_base;
+	unsigned			num_queues;
+	void				*queue_base_inst;
+	unsigned			flags;
+	struct list_head		list;
+	struct knav_range_ops		*ops;
+	struct knav_acc_info		acc_info;
+	struct knav_acc_channel	*acc;
+	unsigned			num_irqs;
+	struct knav_irq_info		irqs[RANGE_MAX_IRQS];
+};
+
+#define RANGE_RESERVED		BIT(0)
+#define RANGE_HAS_IRQ		BIT(1)
+#define RANGE_HAS_ACCUMULATOR	BIT(2)
+#define RANGE_MULTI_QUEUE	BIT(3)
+
+#define for_each_region(kdev, region)				\
+	list_for_each_entry(region, &kdev->regions, list)
+
+#define first_region(kdev)					\
+	list_first_entry_or_null(&kdev->regions, \
+				 struct knav_region, list)
+
+#define for_each_queue_range(kdev, range)			\
+	list_for_each_entry(range, &kdev->queue_ranges, list)
+
+#define first_queue_range(kdev)					\
+	list_first_entry_or_null(&kdev->queue_ranges, \
+				 struct knav_range_info, list)
+
+#define for_each_pool(kdev, pool)				\
+	list_for_each_entry(pool, &kdev->pools, list)
+
+#define for_each_pdsp(kdev, pdsp)				\
+	list_for_each_entry(pdsp, &kdev->pdsps, list)
+
+#define for_each_qmgr(kdev, qmgr)				\
+	list_for_each_entry(qmgr, &kdev->qmgrs, list)
+
+static inline struct knav_pdsp_info *
+knav_find_pdsp(struct knav_device *kdev, unsigned pdsp_id)
+{
+	struct knav_pdsp_info *pdsp;
+
+	for_each_pdsp(kdev, pdsp)
+		if (pdsp_id == pdsp->id)
+			return pdsp;
+	return NULL;
+}
+
+extern int knav_init_acc_range(struct knav_device *kdev,
+					struct device_node *node,
+					struct knav_range_info *range);
+extern void knav_queue_notify(struct knav_queue_inst *inst);
+
+#endif /* __KNAV_QMSS_H__ */
diff --git a/drivers/soc/ti/knav_qmss_acc.c b/drivers/soc/ti/knav_qmss_acc.c
new file mode 100644
index 000000000..2f7fb2dcc
--- /dev/null
+++ b/drivers/soc/ti/knav_qmss_acc.c
@@ -0,0 +1,592 @@
+/*
+ * Keystone accumulator queue manager
+ *
+ * Copyright (C) 2014 Texas Instruments Incorporated - http://www.ti.com
+ * Author:	Sandeep Nair <sandeep_n@ti.com>
+ *		Cyril Chemparathy <cyril@ti.com>
+ *		Santosh Shilimkar <santosh.shilimkar@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * version 2 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.
+ */
+
+#include <linux/dma-mapping.h>
+#include <linux/io.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/soc/ti/knav_qmss.h>
+
+#include "knav_qmss.h"
+
+#define knav_range_offset_to_inst(kdev, range, q)	\
+	(range->queue_base_inst + (q << kdev->inst_shift))
+
+static void __knav_acc_notify(struct knav_range_info *range,
+				struct knav_acc_channel *acc)
+{
+	struct knav_device *kdev = range->kdev;
+	struct knav_queue_inst *inst;
+	int range_base, queue;
+
+	range_base = kdev->base_id + range->queue_base;
+
+	if (range->flags & RANGE_MULTI_QUEUE) {
+		for (queue = 0; queue < range->num_queues; queue++) {
+			inst = knav_range_offset_to_inst(kdev, range,
+								queue);
+			if (inst->notify_needed) {
+				inst->notify_needed = 0;
+				dev_dbg(kdev->dev, "acc-irq: notifying %d\n",
+					range_base + queue);
+				knav_queue_notify(inst);
+			}
+		}
+	} else {
+		queue = acc->channel - range->acc_info.start_channel;
+		inst = knav_range_offset_to_inst(kdev, range, queue);
+		dev_dbg(kdev->dev, "acc-irq: notifying %d\n",
+			range_base + queue);
+		knav_queue_notify(inst);
+	}
+}
+
+static int knav_acc_set_notify(struct knav_range_info *range,
+				struct knav_queue_inst *kq,
+				bool enabled)
+{
+	struct knav_pdsp_info *pdsp = range->acc_info.pdsp;
+	struct knav_device *kdev = range->kdev;
+	u32 mask, offset;
+
+	/*
+	 * when enabling, we need to re-trigger an interrupt if we
+	 * have descriptors pending
+	 */
+	if (!enabled || atomic_read(&kq->desc_count) <= 0)
+		return 0;
+
+	kq->notify_needed = 1;
+	atomic_inc(&kq->acc->retrigger_count);
+	mask = BIT(kq->acc->channel % 32);
+	offset = ACC_INTD_OFFSET_STATUS(kq->acc->channel);
+	dev_dbg(kdev->dev, "setup-notify: re-triggering irq for %s\n",
+		kq->acc->name);
+	writel_relaxed(mask, pdsp->intd + offset);
+	return 0;
+}
+
+static irqreturn_t knav_acc_int_handler(int irq, void *_instdata)
+{
+	struct knav_acc_channel *acc;
+	struct knav_queue_inst *kq = NULL;
+	struct knav_range_info *range;
+	struct knav_pdsp_info *pdsp;
+	struct knav_acc_info *info;
+	struct knav_device *kdev;
+
+	u32 *list, *list_cpu, val, idx, notifies;
+	int range_base, channel, queue = 0;
+	dma_addr_t list_dma;
+
+	range = _instdata;
+	info  = &range->acc_info;
+	kdev  = range->kdev;
+	pdsp  = range->acc_info.pdsp;
+	acc   = range->acc;
+
+	range_base = kdev->base_id + range->queue_base;
+	if ((range->flags & RANGE_MULTI_QUEUE) == 0) {
+		for (queue = 0; queue < range->num_irqs; queue++)
+			if (range->irqs[queue].irq == irq)
+				break;
+		kq = knav_range_offset_to_inst(kdev, range, queue);
+		acc += queue;
+	}
+
+	channel = acc->channel;
+	list_dma = acc->list_dma[acc->list_index];
+	list_cpu = acc->list_cpu[acc->list_index];
+	dev_dbg(kdev->dev, "acc-irq: channel %d, list %d, virt %p, dma %pad\n",
+		channel, acc->list_index, list_cpu, &list_dma);
+	if (atomic_read(&acc->retrigger_count)) {
+		atomic_dec(&acc->retrigger_count);
+		__knav_acc_notify(range, acc);
+		writel_relaxed(1, pdsp->intd + ACC_INTD_OFFSET_COUNT(channel));
+		/* ack the interrupt */
+		writel_relaxed(ACC_CHANNEL_INT_BASE + channel,
+			       pdsp->intd + ACC_INTD_OFFSET_EOI);
+
+		return IRQ_HANDLED;
+	}
+
+	notifies = readl_relaxed(pdsp->intd + ACC_INTD_OFFSET_COUNT(channel));
+	WARN_ON(!notifies);
+	dma_sync_single_for_cpu(kdev->dev, list_dma, info->list_size,
+				DMA_FROM_DEVICE);
+
+	for (list = list_cpu; list < list_cpu + (info->list_size / sizeof(u32));
+	     list += ACC_LIST_ENTRY_WORDS) {
+		if (ACC_LIST_ENTRY_WORDS == 1) {
+			dev_dbg(kdev->dev,
+				"acc-irq: list %d, entry @%p, %08x\n",
+				acc->list_index, list, list[0]);
+		} else if (ACC_LIST_ENTRY_WORDS == 2) {
+			dev_dbg(kdev->dev,
+				"acc-irq: list %d, entry @%p, %08x %08x\n",
+				acc->list_index, list, list[0], list[1]);
+		} else if (ACC_LIST_ENTRY_WORDS == 4) {
+			dev_dbg(kdev->dev,
+				"acc-irq: list %d, entry @%p, %08x %08x %08x %08x\n",
+				acc->list_index, list, list[0], list[1],
+				list[2], list[3]);
+		}
+
+		val = list[ACC_LIST_ENTRY_DESC_IDX];
+		if (!val)
+			break;
+
+		if (range->flags & RANGE_MULTI_QUEUE) {
+			queue = list[ACC_LIST_ENTRY_QUEUE_IDX] >> 16;
+			if (queue < range_base ||
+			    queue >= range_base + range->num_queues) {
+				dev_err(kdev->dev,
+					"bad queue %d, expecting %d-%d\n",
+					queue, range_base,
+					range_base + range->num_queues);
+				break;
+			}
+			queue -= range_base;
+			kq = knav_range_offset_to_inst(kdev, range,
+								queue);
+		}
+
+		if (atomic_inc_return(&kq->desc_count) >= ACC_DESCS_MAX) {
+			atomic_dec(&kq->desc_count);
+			dev_err(kdev->dev,
+				"acc-irq: queue %d full, entry dropped\n",
+				queue + range_base);
+			continue;
+		}
+
+		idx = atomic_inc_return(&kq->desc_tail) & ACC_DESCS_MASK;
+		kq->descs[idx] = val;
+		kq->notify_needed = 1;
+		dev_dbg(kdev->dev, "acc-irq: enqueue %08x at %d, queue %d\n",
+			val, idx, queue + range_base);
+	}
+
+	__knav_acc_notify(range, acc);
+	memset(list_cpu, 0, info->list_size);
+	dma_sync_single_for_device(kdev->dev, list_dma, info->list_size,
+				   DMA_TO_DEVICE);
+
+	/* flip to the other list */
+	acc->list_index ^= 1;
+
+	/* reset the interrupt counter */
+	writel_relaxed(1, pdsp->intd + ACC_INTD_OFFSET_COUNT(channel));
+
+	/* ack the interrupt */
+	writel_relaxed(ACC_CHANNEL_INT_BASE + channel,
+		       pdsp->intd + ACC_INTD_OFFSET_EOI);
+
+	return IRQ_HANDLED;
+}
+
+static int knav_range_setup_acc_irq(struct knav_range_info *range,
+				int queue, bool enabled)
+{
+	struct knav_device *kdev = range->kdev;
+	struct knav_acc_channel *acc;
+	struct cpumask *cpu_mask;
+	int ret = 0, irq;
+	u32 old, new;
+
+	if (range->flags & RANGE_MULTI_QUEUE) {
+		acc = range->acc;
+		irq = range->irqs[0].irq;
+		cpu_mask = range->irqs[0].cpu_mask;
+	} else {
+		acc = range->acc + queue;
+		irq = range->irqs[queue].irq;
+		cpu_mask = range->irqs[queue].cpu_mask;
+	}
+
+	old = acc->open_mask;
+	if (enabled)
+		new = old | BIT(queue);
+	else
+		new = old & ~BIT(queue);
+	acc->open_mask = new;
+
+	dev_dbg(kdev->dev,
+		"setup-acc-irq: open mask old %08x, new %08x, channel %s\n",
+		old, new, acc->name);
+
+	if (likely(new == old))
+		return 0;
+
+	if (new && !old) {
+		dev_dbg(kdev->dev,
+			"setup-acc-irq: requesting %s for channel %s\n",
+			acc->name, acc->name);
+		ret = request_irq(irq, knav_acc_int_handler, 0, acc->name,
+				  range);
+		if (!ret && cpu_mask) {
+			ret = irq_set_affinity_hint(irq, cpu_mask);
+			if (ret) {
+				dev_warn(range->kdev->dev,
+					 "Failed to set IRQ affinity\n");
+				return ret;
+			}
+		}
+	}
+
+	if (old && !new) {
+		dev_dbg(kdev->dev, "setup-acc-irq: freeing %s for channel %s\n",
+			acc->name, acc->name);
+		ret = irq_set_affinity_hint(irq, NULL);
+		if (ret)
+			dev_warn(range->kdev->dev,
+				 "Failed to set IRQ affinity\n");
+		free_irq(irq, range);
+	}
+
+	return ret;
+}
+
+static const char *knav_acc_result_str(enum knav_acc_result result)
+{
+	static const char * const result_str[] = {
+		[ACC_RET_IDLE]			= "idle",
+		[ACC_RET_SUCCESS]		= "success",
+		[ACC_RET_INVALID_COMMAND]	= "invalid command",
+		[ACC_RET_INVALID_CHANNEL]	= "invalid channel",
+		[ACC_RET_INACTIVE_CHANNEL]	= "inactive channel",
+		[ACC_RET_ACTIVE_CHANNEL]	= "active channel",
+		[ACC_RET_INVALID_QUEUE]		= "invalid queue",
+		[ACC_RET_INVALID_RET]		= "invalid return code",
+	};
+
+	if (result >= ARRAY_SIZE(result_str))
+		return result_str[ACC_RET_INVALID_RET];
+	else
+		return result_str[result];
+}
+
+static enum knav_acc_result
+knav_acc_write(struct knav_device *kdev, struct knav_pdsp_info *pdsp,
+		struct knav_reg_acc_command *cmd)
+{
+	u32 result;
+
+	dev_dbg(kdev->dev, "acc command %08x %08x %08x %08x %08x\n",
+		cmd->command, cmd->queue_mask, cmd->list_dma,
+		cmd->queue_num, cmd->timer_config);
+
+	writel_relaxed(cmd->timer_config, &pdsp->acc_command->timer_config);
+	writel_relaxed(cmd->queue_num, &pdsp->acc_command->queue_num);
+	writel_relaxed(cmd->list_dma, &pdsp->acc_command->list_dma);
+	writel_relaxed(cmd->queue_mask, &pdsp->acc_command->queue_mask);
+	writel_relaxed(cmd->command, &pdsp->acc_command->command);
+
+	/* wait for the command to clear */
+	do {
+		result = readl_relaxed(&pdsp->acc_command->command);
+	} while ((result >> 8) & 0xff);
+
+	return (result >> 24) & 0xff;
+}
+
+static void knav_acc_setup_cmd(struct knav_device *kdev,
+				struct knav_range_info *range,
+				struct knav_reg_acc_command *cmd,
+				int queue)
+{
+	struct knav_acc_info *info = &range->acc_info;
+	struct knav_acc_channel *acc;
+	int queue_base;
+	u32 queue_mask;
+
+	if (range->flags & RANGE_MULTI_QUEUE) {
+		acc = range->acc;
+		queue_base = range->queue_base;
+		queue_mask = BIT(range->num_queues) - 1;
+	} else {
+		acc = range->acc + queue;
+		queue_base = range->queue_base + queue;
+		queue_mask = 0;
+	}
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->command    = acc->channel;
+	cmd->queue_mask = queue_mask;
+	cmd->list_dma   = (u32)acc->list_dma[0];
+	cmd->queue_num  = info->list_entries << 16;
+	cmd->queue_num |= queue_base;
+
+	cmd->timer_config = ACC_LIST_ENTRY_TYPE << 18;
+	if (range->flags & RANGE_MULTI_QUEUE)
+		cmd->timer_config |= ACC_CFG_MULTI_QUEUE;
+	cmd->timer_config |= info->pacing_mode << 16;
+	cmd->timer_config |= info->timer_count;
+}
+
+static void knav_acc_stop(struct knav_device *kdev,
+				struct knav_range_info *range,
+				int queue)
+{
+	struct knav_reg_acc_command cmd;
+	struct knav_acc_channel *acc;
+	enum knav_acc_result result;
+
+	acc = range->acc + queue;
+
+	knav_acc_setup_cmd(kdev, range, &cmd, queue);
+	cmd.command |= ACC_CMD_DISABLE_CHANNEL << 8;
+	result = knav_acc_write(kdev, range->acc_info.pdsp, &cmd);
+
+	dev_dbg(kdev->dev, "stopped acc channel %s, result %s\n",
+		acc->name, knav_acc_result_str(result));
+}
+
+static enum knav_acc_result knav_acc_start(struct knav_device *kdev,
+						struct knav_range_info *range,
+						int queue)
+{
+	struct knav_reg_acc_command cmd;
+	struct knav_acc_channel *acc;
+	enum knav_acc_result result;
+
+	acc = range->acc + queue;
+
+	knav_acc_setup_cmd(kdev, range, &cmd, queue);
+	cmd.command |= ACC_CMD_ENABLE_CHANNEL << 8;
+	result = knav_acc_write(kdev, range->acc_info.pdsp, &cmd);
+
+	dev_dbg(kdev->dev, "started acc channel %s, result %s\n",
+		acc->name, knav_acc_result_str(result));
+
+	return result;
+}
+
+static int knav_acc_init_range(struct knav_range_info *range)
+{
+	struct knav_device *kdev = range->kdev;
+	struct knav_acc_channel *acc;
+	enum knav_acc_result result;
+	int queue;
+
+	for (queue = 0; queue < range->num_queues; queue++) {
+		acc = range->acc + queue;
+
+		knav_acc_stop(kdev, range, queue);
+		acc->list_index = 0;
+		result = knav_acc_start(kdev, range, queue);
+
+		if (result != ACC_RET_SUCCESS)
+			return -EIO;
+
+		if (range->flags & RANGE_MULTI_QUEUE)
+			return 0;
+	}
+	return 0;
+}
+
+static int knav_acc_init_queue(struct knav_range_info *range,
+				struct knav_queue_inst *kq)
+{
+	unsigned id = kq->id - range->queue_base;
+
+	kq->descs = devm_kcalloc(range->kdev->dev,
+				 ACC_DESCS_MAX, sizeof(u32), GFP_KERNEL);
+	if (!kq->descs)
+		return -ENOMEM;
+
+	kq->acc = range->acc;
+	if ((range->flags & RANGE_MULTI_QUEUE) == 0)
+		kq->acc += id;
+	return 0;
+}
+
+static int knav_acc_open_queue(struct knav_range_info *range,
+				struct knav_queue_inst *inst, unsigned flags)
+{
+	unsigned id = inst->id - range->queue_base;
+
+	return knav_range_setup_acc_irq(range, id, true);
+}
+
+static int knav_acc_close_queue(struct knav_range_info *range,
+					struct knav_queue_inst *inst)
+{
+	unsigned id = inst->id - range->queue_base;
+
+	return knav_range_setup_acc_irq(range, id, false);
+}
+
+static int knav_acc_free_range(struct knav_range_info *range)
+{
+	struct knav_device *kdev = range->kdev;
+	struct knav_acc_channel *acc;
+	struct knav_acc_info *info;
+	int channel, channels;
+
+	info = &range->acc_info;
+
+	if (range->flags & RANGE_MULTI_QUEUE)
+		channels = 1;
+	else
+		channels = range->num_queues;
+
+	for (channel = 0; channel < channels; channel++) {
+		acc = range->acc + channel;
+		if (!acc->list_cpu[0])
+			continue;
+		dma_unmap_single(kdev->dev, acc->list_dma[0],
+				 info->mem_size, DMA_BIDIRECTIONAL);
+		free_pages_exact(acc->list_cpu[0], info->mem_size);
+	}
+	devm_kfree(range->kdev->dev, range->acc);
+	return 0;
+}
+
+struct knav_range_ops knav_acc_range_ops = {
+	.set_notify	= knav_acc_set_notify,
+	.init_queue	= knav_acc_init_queue,
+	.open_queue	= knav_acc_open_queue,
+	.close_queue	= knav_acc_close_queue,
+	.init_range	= knav_acc_init_range,
+	.free_range	= knav_acc_free_range,
+};
+
+/**
+ * knav_init_acc_range: Initialise accumulator ranges
+ *
+ * @kdev:		qmss device
+ * @node:		device node
+ * @range:		qmms range information
+ *
+ * Return 0 on success or error
+ */
+int knav_init_acc_range(struct knav_device *kdev,
+			struct device_node *node,
+			struct knav_range_info *range)
+{
+	struct knav_acc_channel *acc;
+	struct knav_pdsp_info *pdsp;
+	struct knav_acc_info *info;
+	int ret, channel, channels;
+	int list_size, mem_size;
+	dma_addr_t list_dma;
+	void *list_mem;
+	u32 config[5];
+
+	range->flags |= RANGE_HAS_ACCUMULATOR;
+	info = &range->acc_info;
+
+	ret = of_property_read_u32_array(node, "accumulator", config, 5);
+	if (ret)
+		return ret;
+
+	info->pdsp_id		= config[0];
+	info->start_channel	= config[1];
+	info->list_entries	= config[2];
+	info->pacing_mode	= config[3];
+	info->timer_count	= config[4] / ACC_DEFAULT_PERIOD;
+
+	if (info->start_channel > ACC_MAX_CHANNEL) {
+		dev_err(kdev->dev, "channel %d invalid for range %s\n",
+			info->start_channel, range->name);
+		return -EINVAL;
+	}
+
+	if (info->pacing_mode > 3) {
+		dev_err(kdev->dev, "pacing mode %d invalid for range %s\n",
+			info->pacing_mode, range->name);
+		return -EINVAL;
+	}
+
+	pdsp = knav_find_pdsp(kdev, info->pdsp_id);
+	if (!pdsp) {
+		dev_err(kdev->dev, "pdsp id %d not found for range %s\n",
+			info->pdsp_id, range->name);
+		return -EINVAL;
+	}
+
+	if (!pdsp->started) {
+		dev_err(kdev->dev, "pdsp id %d not started for range %s\n",
+			info->pdsp_id, range->name);
+		return -ENODEV;
+	}
+
+	info->pdsp = pdsp;
+	channels = range->num_queues;
+	if (of_get_property(node, "multi-queue", NULL)) {
+		range->flags |= RANGE_MULTI_QUEUE;
+		channels = 1;
+		if (range->queue_base & (32 - 1)) {
+			dev_err(kdev->dev,
+				"misaligned multi-queue accumulator range %s\n",
+				range->name);
+			return -EINVAL;
+		}
+		if (range->num_queues > 32) {
+			dev_err(kdev->dev,
+				"too many queues in accumulator range %s\n",
+				range->name);
+			return -EINVAL;
+		}
+	}
+
+	/* figure out list size */
+	list_size  = info->list_entries;
+	list_size *= ACC_LIST_ENTRY_WORDS * sizeof(u32);
+	info->list_size = list_size;
+	mem_size   = PAGE_ALIGN(list_size * 2);
+	info->mem_size  = mem_size;
+	range->acc = devm_kcalloc(kdev->dev, channels, sizeof(*range->acc),
+				  GFP_KERNEL);
+	if (!range->acc)
+		return -ENOMEM;
+
+	for (channel = 0; channel < channels; channel++) {
+		acc = range->acc + channel;
+		acc->channel = info->start_channel + channel;
+
+		/* allocate memory for the two lists */
+		list_mem = alloc_pages_exact(mem_size, GFP_KERNEL | GFP_DMA);
+		if (!list_mem)
+			return -ENOMEM;
+
+		list_dma = dma_map_single(kdev->dev, list_mem, mem_size,
+					  DMA_BIDIRECTIONAL);
+		if (dma_mapping_error(kdev->dev, list_dma)) {
+			free_pages_exact(list_mem, mem_size);
+			return -ENOMEM;
+		}
+
+		memset(list_mem, 0, mem_size);
+		dma_sync_single_for_device(kdev->dev, list_dma, mem_size,
+					   DMA_TO_DEVICE);
+		scnprintf(acc->name, sizeof(acc->name), "hwqueue-acc-%d",
+			  acc->channel);
+		acc->list_cpu[0] = list_mem;
+		acc->list_cpu[1] = list_mem + list_size;
+		acc->list_dma[0] = list_dma;
+		acc->list_dma[1] = list_dma + list_size;
+		dev_dbg(kdev->dev, "%s: channel %d, dma %pad, virt %8p\n",
+			acc->name, acc->channel, &list_dma, list_mem);
+	}
+
+	range->ops = &knav_acc_range_ops;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(knav_init_acc_range);
diff --git a/drivers/soc/ti/knav_qmss_queue.c b/drivers/soc/ti/knav_qmss_queue.c
new file mode 100644
index 000000000..9f5ce52e6
--- /dev/null
+++ b/drivers/soc/ti/knav_qmss_queue.c
@@ -0,0 +1,1922 @@
+/*
+ * Keystone Queue Manager subsystem driver
+ *
+ * Copyright (C) 2014 Texas Instruments Incorporated - http://www.ti.com
+ * Authors:	Sandeep Nair <sandeep_n@ti.com>
+ *		Cyril Chemparathy <cyril@ti.com>
+ *		Santosh Shilimkar <santosh.shilimkar@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * version 2 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.
+ */
+
+#include <linux/debugfs.h>
+#include <linux/dma-mapping.h>
+#include <linux/firmware.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/of_device.h>
+#include <linux/of_irq.h>
+#include <linux/pm_runtime.h>
+#include <linux/slab.h>
+#include <linux/soc/ti/knav_qmss.h>
+
+#include "knav_qmss.h"
+
+static struct knav_device *kdev;
+static DEFINE_MUTEX(knav_dev_lock);
+
+/* Queue manager register indices in DTS */
+#define KNAV_QUEUE_PEEK_REG_INDEX	0
+#define KNAV_QUEUE_STATUS_REG_INDEX	1
+#define KNAV_QUEUE_CONFIG_REG_INDEX	2
+#define KNAV_QUEUE_REGION_REG_INDEX	3
+#define KNAV_QUEUE_PUSH_REG_INDEX	4
+#define KNAV_QUEUE_POP_REG_INDEX	5
+
+/* Queue manager register indices in DTS for QMSS in K2G NAVSS.
+ * There are no status and vbusm push registers on this version
+ * of QMSS. Push registers are same as pop, So all indices above 1
+ * are to be re-defined
+ */
+#define KNAV_L_QUEUE_CONFIG_REG_INDEX	1
+#define KNAV_L_QUEUE_REGION_REG_INDEX	2
+#define KNAV_L_QUEUE_PUSH_REG_INDEX	3
+
+/* PDSP register indices in DTS */
+#define KNAV_QUEUE_PDSP_IRAM_REG_INDEX	0
+#define KNAV_QUEUE_PDSP_REGS_REG_INDEX	1
+#define KNAV_QUEUE_PDSP_INTD_REG_INDEX	2
+#define KNAV_QUEUE_PDSP_CMD_REG_INDEX	3
+
+#define knav_queue_idx_to_inst(kdev, idx)			\
+	(kdev->instances + (idx << kdev->inst_shift))
+
+#define for_each_handle_rcu(qh, inst)			\
+	list_for_each_entry_rcu(qh, &inst->handles, list)
+
+#define for_each_instance(idx, inst, kdev)		\
+	for (idx = 0, inst = kdev->instances;		\
+	     idx < (kdev)->num_queues_in_use;			\
+	     idx++, inst = knav_queue_idx_to_inst(kdev, idx))
+
+/* All firmware file names end up here. List the firmware file names below.
+ * Newest followed by older ones. Search is done from start of the array
+ * until a firmware file is found.
+ */
+const char *knav_acc_firmwares[] = {"ks2_qmss_pdsp_acc48.bin"};
+
+static bool device_ready;
+bool knav_qmss_device_ready(void)
+{
+	return device_ready;
+}
+EXPORT_SYMBOL_GPL(knav_qmss_device_ready);
+
+/**
+ * knav_queue_notify: qmss queue notfier call
+ *
+ * @inst:		qmss queue instance like accumulator
+ */
+void knav_queue_notify(struct knav_queue_inst *inst)
+{
+	struct knav_queue *qh;
+
+	if (!inst)
+		return;
+
+	rcu_read_lock();
+	for_each_handle_rcu(qh, inst) {
+		if (atomic_read(&qh->notifier_enabled) <= 0)
+			continue;
+		if (WARN_ON(!qh->notifier_fn))
+			continue;
+		this_cpu_inc(qh->stats->notifies);
+		qh->notifier_fn(qh->notifier_fn_arg);
+	}
+	rcu_read_unlock();
+}
+EXPORT_SYMBOL_GPL(knav_queue_notify);
+
+static irqreturn_t knav_queue_int_handler(int irq, void *_instdata)
+{
+	struct knav_queue_inst *inst = _instdata;
+
+	knav_queue_notify(inst);
+	return IRQ_HANDLED;
+}
+
+static int knav_queue_setup_irq(struct knav_range_info *range,
+			  struct knav_queue_inst *inst)
+{
+	unsigned queue = inst->id - range->queue_base;
+	int ret = 0, irq;
+
+	if (range->flags & RANGE_HAS_IRQ) {
+		irq = range->irqs[queue].irq;
+		ret = request_irq(irq, knav_queue_int_handler, 0,
+					inst->irq_name, inst);
+		if (ret)
+			return ret;
+		disable_irq(irq);
+		if (range->irqs[queue].cpu_mask) {
+			ret = irq_set_affinity_hint(irq, range->irqs[queue].cpu_mask);
+			if (ret) {
+				dev_warn(range->kdev->dev,
+					 "Failed to set IRQ affinity\n");
+				return ret;
+			}
+		}
+	}
+	return ret;
+}
+
+static void knav_queue_free_irq(struct knav_queue_inst *inst)
+{
+	struct knav_range_info *range = inst->range;
+	unsigned queue = inst->id - inst->range->queue_base;
+	int irq;
+
+	if (range->flags & RANGE_HAS_IRQ) {
+		irq = range->irqs[queue].irq;
+		irq_set_affinity_hint(irq, NULL);
+		free_irq(irq, inst);
+	}
+}
+
+static inline bool knav_queue_is_busy(struct knav_queue_inst *inst)
+{
+	return !list_empty(&inst->handles);
+}
+
+static inline bool knav_queue_is_reserved(struct knav_queue_inst *inst)
+{
+	return inst->range->flags & RANGE_RESERVED;
+}
+
+static inline bool knav_queue_is_shared(struct knav_queue_inst *inst)
+{
+	struct knav_queue *tmp;
+
+	rcu_read_lock();
+	for_each_handle_rcu(tmp, inst) {
+		if (tmp->flags & KNAV_QUEUE_SHARED) {
+			rcu_read_unlock();
+			return true;
+		}
+	}
+	rcu_read_unlock();
+	return false;
+}
+
+static inline bool knav_queue_match_type(struct knav_queue_inst *inst,
+						unsigned type)
+{
+	if ((type == KNAV_QUEUE_QPEND) &&
+	    (inst->range->flags & RANGE_HAS_IRQ)) {
+		return true;
+	} else if ((type == KNAV_QUEUE_ACC) &&
+		(inst->range->flags & RANGE_HAS_ACCUMULATOR)) {
+		return true;
+	} else if ((type == KNAV_QUEUE_GP) &&
+		!(inst->range->flags &
+			(RANGE_HAS_ACCUMULATOR | RANGE_HAS_IRQ))) {
+		return true;
+	}
+	return false;
+}
+
+static inline struct knav_queue_inst *
+knav_queue_match_id_to_inst(struct knav_device *kdev, unsigned id)
+{
+	struct knav_queue_inst *inst;
+	int idx;
+
+	for_each_instance(idx, inst, kdev) {
+		if (inst->id == id)
+			return inst;
+	}
+	return NULL;
+}
+
+static inline struct knav_queue_inst *knav_queue_find_by_id(int id)
+{
+	if (kdev->base_id <= id &&
+	    kdev->base_id + kdev->num_queues > id) {
+		id -= kdev->base_id;
+		return knav_queue_match_id_to_inst(kdev, id);
+	}
+	return NULL;
+}
+
+static struct knav_queue *__knav_queue_open(struct knav_queue_inst *inst,
+				      const char *name, unsigned flags)
+{
+	struct knav_queue *qh;
+	unsigned id;
+	int ret = 0;
+
+	qh = devm_kzalloc(inst->kdev->dev, sizeof(*qh), GFP_KERNEL);
+	if (!qh)
+		return ERR_PTR(-ENOMEM);
+
+	qh->stats = alloc_percpu(struct knav_queue_stats);
+	if (!qh->stats) {
+		ret = -ENOMEM;
+		goto err;
+	}
+
+	qh->flags = flags;
+	qh->inst = inst;
+	id = inst->id - inst->qmgr->start_queue;
+	qh->reg_push = &inst->qmgr->reg_push[id];
+	qh->reg_pop = &inst->qmgr->reg_pop[id];
+	qh->reg_peek = &inst->qmgr->reg_peek[id];
+
+	/* first opener? */
+	if (!knav_queue_is_busy(inst)) {
+		struct knav_range_info *range = inst->range;
+
+		inst->name = kstrndup(name, KNAV_NAME_SIZE - 1, GFP_KERNEL);
+		if (range->ops && range->ops->open_queue)
+			ret = range->ops->open_queue(range, inst, flags);
+
+		if (ret)
+			goto err;
+	}
+	list_add_tail_rcu(&qh->list, &inst->handles);
+	return qh;
+
+err:
+	if (qh->stats)
+		free_percpu(qh->stats);
+	devm_kfree(inst->kdev->dev, qh);
+	return ERR_PTR(ret);
+}
+
+static struct knav_queue *
+knav_queue_open_by_id(const char *name, unsigned id, unsigned flags)
+{
+	struct knav_queue_inst *inst;
+	struct knav_queue *qh;
+
+	mutex_lock(&knav_dev_lock);
+
+	qh = ERR_PTR(-ENODEV);
+	inst = knav_queue_find_by_id(id);
+	if (!inst)
+		goto unlock_ret;
+
+	qh = ERR_PTR(-EEXIST);
+	if (!(flags & KNAV_QUEUE_SHARED) && knav_queue_is_busy(inst))
+		goto unlock_ret;
+
+	qh = ERR_PTR(-EBUSY);
+	if ((flags & KNAV_QUEUE_SHARED) &&
+	    (knav_queue_is_busy(inst) && !knav_queue_is_shared(inst)))
+		goto unlock_ret;
+
+	qh = __knav_queue_open(inst, name, flags);
+
+unlock_ret:
+	mutex_unlock(&knav_dev_lock);
+
+	return qh;
+}
+
+static struct knav_queue *knav_queue_open_by_type(const char *name,
+						unsigned type, unsigned flags)
+{
+	struct knav_queue_inst *inst;
+	struct knav_queue *qh = ERR_PTR(-EINVAL);
+	int idx;
+
+	mutex_lock(&knav_dev_lock);
+
+	for_each_instance(idx, inst, kdev) {
+		if (knav_queue_is_reserved(inst))
+			continue;
+		if (!knav_queue_match_type(inst, type))
+			continue;
+		if (knav_queue_is_busy(inst))
+			continue;
+		qh = __knav_queue_open(inst, name, flags);
+		goto unlock_ret;
+	}
+
+unlock_ret:
+	mutex_unlock(&knav_dev_lock);
+	return qh;
+}
+
+static void knav_queue_set_notify(struct knav_queue_inst *inst, bool enabled)
+{
+	struct knav_range_info *range = inst->range;
+
+	if (range->ops && range->ops->set_notify)
+		range->ops->set_notify(range, inst, enabled);
+}
+
+static int knav_queue_enable_notifier(struct knav_queue *qh)
+{
+	struct knav_queue_inst *inst = qh->inst;
+	bool first;
+
+	if (WARN_ON(!qh->notifier_fn))
+		return -EINVAL;
+
+	/* Adjust the per handle notifier count */
+	first = (atomic_inc_return(&qh->notifier_enabled) == 1);
+	if (!first)
+		return 0; /* nothing to do */
+
+	/* Now adjust the per instance notifier count */
+	first = (atomic_inc_return(&inst->num_notifiers) == 1);
+	if (first)
+		knav_queue_set_notify(inst, true);
+
+	return 0;
+}
+
+static int knav_queue_disable_notifier(struct knav_queue *qh)
+{
+	struct knav_queue_inst *inst = qh->inst;
+	bool last;
+
+	last = (atomic_dec_return(&qh->notifier_enabled) == 0);
+	if (!last)
+		return 0; /* nothing to do */
+
+	last = (atomic_dec_return(&inst->num_notifiers) == 0);
+	if (last)
+		knav_queue_set_notify(inst, false);
+
+	return 0;
+}
+
+static int knav_queue_set_notifier(struct knav_queue *qh,
+				struct knav_queue_notify_config *cfg)
+{
+	knav_queue_notify_fn old_fn = qh->notifier_fn;
+
+	if (!cfg)
+		return -EINVAL;
+
+	if (!(qh->inst->range->flags & (RANGE_HAS_ACCUMULATOR | RANGE_HAS_IRQ)))
+		return -ENOTSUPP;
+
+	if (!cfg->fn && old_fn)
+		knav_queue_disable_notifier(qh);
+
+	qh->notifier_fn = cfg->fn;
+	qh->notifier_fn_arg = cfg->fn_arg;
+
+	if (cfg->fn && !old_fn)
+		knav_queue_enable_notifier(qh);
+
+	return 0;
+}
+
+static int knav_gp_set_notify(struct knav_range_info *range,
+			       struct knav_queue_inst *inst,
+			       bool enabled)
+{
+	unsigned queue;
+
+	if (range->flags & RANGE_HAS_IRQ) {
+		queue = inst->id - range->queue_base;
+		if (enabled)
+			enable_irq(range->irqs[queue].irq);
+		else
+			disable_irq_nosync(range->irqs[queue].irq);
+	}
+	return 0;
+}
+
+static int knav_gp_open_queue(struct knav_range_info *range,
+				struct knav_queue_inst *inst, unsigned flags)
+{
+	return knav_queue_setup_irq(range, inst);
+}
+
+static int knav_gp_close_queue(struct knav_range_info *range,
+				struct knav_queue_inst *inst)
+{
+	knav_queue_free_irq(inst);
+	return 0;
+}
+
+struct knav_range_ops knav_gp_range_ops = {
+	.set_notify	= knav_gp_set_notify,
+	.open_queue	= knav_gp_open_queue,
+	.close_queue	= knav_gp_close_queue,
+};
+
+
+static int knav_queue_get_count(void *qhandle)
+{
+	struct knav_queue *qh = qhandle;
+	struct knav_queue_inst *inst = qh->inst;
+
+	return readl_relaxed(&qh->reg_peek[0].entry_count) +
+		atomic_read(&inst->desc_count);
+}
+
+static void knav_queue_debug_show_instance(struct seq_file *s,
+					struct knav_queue_inst *inst)
+{
+	struct knav_device *kdev = inst->kdev;
+	struct knav_queue *qh;
+	int cpu = 0;
+	int pushes = 0;
+	int pops = 0;
+	int push_errors = 0;
+	int pop_errors = 0;
+	int notifies = 0;
+
+	if (!knav_queue_is_busy(inst))
+		return;
+
+	seq_printf(s, "\tqueue id %d (%s)\n",
+		   kdev->base_id + inst->id, inst->name);
+	for_each_handle_rcu(qh, inst) {
+		for_each_possible_cpu(cpu) {
+			pushes += per_cpu_ptr(qh->stats, cpu)->pushes;
+			pops += per_cpu_ptr(qh->stats, cpu)->pops;
+			push_errors += per_cpu_ptr(qh->stats, cpu)->push_errors;
+			pop_errors += per_cpu_ptr(qh->stats, cpu)->pop_errors;
+			notifies += per_cpu_ptr(qh->stats, cpu)->notifies;
+		}
+
+		seq_printf(s, "\t\thandle %p: pushes %8d, pops %8d, count %8d, notifies %8d, push errors %8d, pop errors %8d\n",
+				qh,
+				pushes,
+				pops,
+				knav_queue_get_count(qh),
+				notifies,
+				push_errors,
+				pop_errors);
+	}
+}
+
+static int knav_queue_debug_show(struct seq_file *s, void *v)
+{
+	struct knav_queue_inst *inst;
+	int idx;
+
+	mutex_lock(&knav_dev_lock);
+	seq_printf(s, "%s: %u-%u\n",
+		   dev_name(kdev->dev), kdev->base_id,
+		   kdev->base_id + kdev->num_queues - 1);
+	for_each_instance(idx, inst, kdev)
+		knav_queue_debug_show_instance(s, inst);
+	mutex_unlock(&knav_dev_lock);
+
+	return 0;
+}
+
+static int knav_queue_debug_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, knav_queue_debug_show, NULL);
+}
+
+static const struct file_operations knav_queue_debug_ops = {
+	.open		= knav_queue_debug_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+static inline int knav_queue_pdsp_wait(u32 * __iomem addr, unsigned timeout,
+					u32 flags)
+{
+	unsigned long end;
+	u32 val = 0;
+
+	end = jiffies + msecs_to_jiffies(timeout);
+	while (time_after(end, jiffies)) {
+		val = readl_relaxed(addr);
+		if (flags)
+			val &= flags;
+		if (!val)
+			break;
+		cpu_relax();
+	}
+	return val ? -ETIMEDOUT : 0;
+}
+
+
+static int knav_queue_flush(struct knav_queue *qh)
+{
+	struct knav_queue_inst *inst = qh->inst;
+	unsigned id = inst->id - inst->qmgr->start_queue;
+
+	atomic_set(&inst->desc_count, 0);
+	writel_relaxed(0, &inst->qmgr->reg_push[id].ptr_size_thresh);
+	return 0;
+}
+
+/**
+ * knav_queue_open()	- open a hardware queue
+ * @name		- name to give the queue handle
+ * @id			- desired queue number if any or specifes the type
+ *			  of queue
+ * @flags		- the following flags are applicable to queues:
+ *	KNAV_QUEUE_SHARED - allow the queue to be shared. Queues are
+ *			     exclusive by default.
+ *			     Subsequent attempts to open a shared queue should
+ *			     also have this flag.
+ *
+ * Returns a handle to the open hardware queue if successful. Use IS_ERR()
+ * to check the returned value for error codes.
+ */
+void *knav_queue_open(const char *name, unsigned id,
+					unsigned flags)
+{
+	struct knav_queue *qh = ERR_PTR(-EINVAL);
+
+	switch (id) {
+	case KNAV_QUEUE_QPEND:
+	case KNAV_QUEUE_ACC:
+	case KNAV_QUEUE_GP:
+		qh = knav_queue_open_by_type(name, id, flags);
+		break;
+
+	default:
+		qh = knav_queue_open_by_id(name, id, flags);
+		break;
+	}
+	return qh;
+}
+EXPORT_SYMBOL_GPL(knav_queue_open);
+
+/**
+ * knav_queue_close()	- close a hardware queue handle
+ * @qh			- handle to close
+ */
+void knav_queue_close(void *qhandle)
+{
+	struct knav_queue *qh = qhandle;
+	struct knav_queue_inst *inst = qh->inst;
+
+	while (atomic_read(&qh->notifier_enabled) > 0)
+		knav_queue_disable_notifier(qh);
+
+	mutex_lock(&knav_dev_lock);
+	list_del_rcu(&qh->list);
+	mutex_unlock(&knav_dev_lock);
+	synchronize_rcu();
+	if (!knav_queue_is_busy(inst)) {
+		struct knav_range_info *range = inst->range;
+
+		if (range->ops && range->ops->close_queue)
+			range->ops->close_queue(range, inst);
+	}
+	free_percpu(qh->stats);
+	devm_kfree(inst->kdev->dev, qh);
+}
+EXPORT_SYMBOL_GPL(knav_queue_close);
+
+/**
+ * knav_queue_device_control()	- Perform control operations on a queue
+ * @qh				- queue handle
+ * @cmd				- control commands
+ * @arg				- command argument
+ *
+ * Returns 0 on success, errno otherwise.
+ */
+int knav_queue_device_control(void *qhandle, enum knav_queue_ctrl_cmd cmd,
+				unsigned long arg)
+{
+	struct knav_queue *qh = qhandle;
+	struct knav_queue_notify_config *cfg;
+	int ret;
+
+	switch ((int)cmd) {
+	case KNAV_QUEUE_GET_ID:
+		ret = qh->inst->kdev->base_id + qh->inst->id;
+		break;
+
+	case KNAV_QUEUE_FLUSH:
+		ret = knav_queue_flush(qh);
+		break;
+
+	case KNAV_QUEUE_SET_NOTIFIER:
+		cfg = (void *)arg;
+		ret = knav_queue_set_notifier(qh, cfg);
+		break;
+
+	case KNAV_QUEUE_ENABLE_NOTIFY:
+		ret = knav_queue_enable_notifier(qh);
+		break;
+
+	case KNAV_QUEUE_DISABLE_NOTIFY:
+		ret = knav_queue_disable_notifier(qh);
+		break;
+
+	case KNAV_QUEUE_GET_COUNT:
+		ret = knav_queue_get_count(qh);
+		break;
+
+	default:
+		ret = -ENOTSUPP;
+		break;
+	}
+	return ret;
+}
+EXPORT_SYMBOL_GPL(knav_queue_device_control);
+
+
+
+/**
+ * knav_queue_push()	- push data (or descriptor) to the tail of a queue
+ * @qh			- hardware queue handle
+ * @data		- data to push
+ * @size		- size of data to push
+ * @flags		- can be used to pass additional information
+ *
+ * Returns 0 on success, errno otherwise.
+ */
+int knav_queue_push(void *qhandle, dma_addr_t dma,
+					unsigned size, unsigned flags)
+{
+	struct knav_queue *qh = qhandle;
+	u32 val;
+
+	val = (u32)dma | ((size / 16) - 1);
+	writel_relaxed(val, &qh->reg_push[0].ptr_size_thresh);
+
+	this_cpu_inc(qh->stats->pushes);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(knav_queue_push);
+
+/**
+ * knav_queue_pop()	- pop data (or descriptor) from the head of a queue
+ * @qh			- hardware queue handle
+ * @size		- (optional) size of the data pop'ed.
+ *
+ * Returns a DMA address on success, 0 on failure.
+ */
+dma_addr_t knav_queue_pop(void *qhandle, unsigned *size)
+{
+	struct knav_queue *qh = qhandle;
+	struct knav_queue_inst *inst = qh->inst;
+	dma_addr_t dma;
+	u32 val, idx;
+
+	/* are we accumulated? */
+	if (inst->descs) {
+		if (unlikely(atomic_dec_return(&inst->desc_count) < 0)) {
+			atomic_inc(&inst->desc_count);
+			return 0;
+		}
+		idx  = atomic_inc_return(&inst->desc_head);
+		idx &= ACC_DESCS_MASK;
+		val = inst->descs[idx];
+	} else {
+		val = readl_relaxed(&qh->reg_pop[0].ptr_size_thresh);
+		if (unlikely(!val))
+			return 0;
+	}
+
+	dma = val & DESC_PTR_MASK;
+	if (size)
+		*size = ((val & DESC_SIZE_MASK) + 1) * 16;
+
+	this_cpu_inc(qh->stats->pops);
+	return dma;
+}
+EXPORT_SYMBOL_GPL(knav_queue_pop);
+
+/* carve out descriptors and push into queue */
+static void kdesc_fill_pool(struct knav_pool *pool)
+{
+	struct knav_region *region;
+	int i;
+
+	region = pool->region;
+	pool->desc_size = region->desc_size;
+	for (i = 0; i < pool->num_desc; i++) {
+		int index = pool->region_offset + i;
+		dma_addr_t dma_addr;
+		unsigned dma_size;
+		dma_addr = region->dma_start + (region->desc_size * index);
+		dma_size = ALIGN(pool->desc_size, SMP_CACHE_BYTES);
+		dma_sync_single_for_device(pool->dev, dma_addr, dma_size,
+					   DMA_TO_DEVICE);
+		knav_queue_push(pool->queue, dma_addr, dma_size, 0);
+	}
+}
+
+/* pop out descriptors and close the queue */
+static void kdesc_empty_pool(struct knav_pool *pool)
+{
+	dma_addr_t dma;
+	unsigned size;
+	void *desc;
+	int i;
+
+	if (!pool->queue)
+		return;
+
+	for (i = 0;; i++) {
+		dma = knav_queue_pop(pool->queue, &size);
+		if (!dma)
+			break;
+		desc = knav_pool_desc_dma_to_virt(pool, dma);
+		if (!desc) {
+			dev_dbg(pool->kdev->dev,
+				"couldn't unmap desc, continuing\n");
+			continue;
+		}
+	}
+	WARN_ON(i != pool->num_desc);
+	knav_queue_close(pool->queue);
+}
+
+
+/* Get the DMA address of a descriptor */
+dma_addr_t knav_pool_desc_virt_to_dma(void *ph, void *virt)
+{
+	struct knav_pool *pool = ph;
+	return pool->region->dma_start + (virt - pool->region->virt_start);
+}
+EXPORT_SYMBOL_GPL(knav_pool_desc_virt_to_dma);
+
+void *knav_pool_desc_dma_to_virt(void *ph, dma_addr_t dma)
+{
+	struct knav_pool *pool = ph;
+	return pool->region->virt_start + (dma - pool->region->dma_start);
+}
+EXPORT_SYMBOL_GPL(knav_pool_desc_dma_to_virt);
+
+/**
+ * knav_pool_create()	- Create a pool of descriptors
+ * @name		- name to give the pool handle
+ * @num_desc		- numbers of descriptors in the pool
+ * @region_id		- QMSS region id from which the descriptors are to be
+ *			  allocated.
+ *
+ * Returns a pool handle on success.
+ * Use IS_ERR_OR_NULL() to identify error values on return.
+ */
+void *knav_pool_create(const char *name,
+					int num_desc, int region_id)
+{
+	struct knav_region *reg_itr, *region = NULL;
+	struct knav_pool *pool, *pi;
+	struct list_head *node;
+	unsigned last_offset;
+	bool slot_found;
+	int ret;
+
+	if (!kdev)
+		return ERR_PTR(-EPROBE_DEFER);
+
+	if (!kdev->dev)
+		return ERR_PTR(-ENODEV);
+
+	pool = devm_kzalloc(kdev->dev, sizeof(*pool), GFP_KERNEL);
+	if (!pool) {
+		dev_err(kdev->dev, "out of memory allocating pool\n");
+		return ERR_PTR(-ENOMEM);
+	}
+
+	for_each_region(kdev, reg_itr) {
+		if (reg_itr->id != region_id)
+			continue;
+		region = reg_itr;
+		break;
+	}
+
+	if (!region) {
+		dev_err(kdev->dev, "region-id(%d) not found\n", region_id);
+		ret = -EINVAL;
+		goto err;
+	}
+
+	pool->queue = knav_queue_open(name, KNAV_QUEUE_GP, 0);
+	if (IS_ERR_OR_NULL(pool->queue)) {
+		dev_err(kdev->dev,
+			"failed to open queue for pool(%s), error %ld\n",
+			name, PTR_ERR(pool->queue));
+		ret = PTR_ERR(pool->queue);
+		goto err;
+	}
+
+	pool->name = kstrndup(name, KNAV_NAME_SIZE - 1, GFP_KERNEL);
+	pool->kdev = kdev;
+	pool->dev = kdev->dev;
+
+	mutex_lock(&knav_dev_lock);
+
+	if (num_desc > (region->num_desc - region->used_desc)) {
+		dev_err(kdev->dev, "out of descs in region(%d) for pool(%s)\n",
+			region_id, name);
+		ret = -ENOMEM;
+		goto err_unlock;
+	}
+
+	/* Region maintains a sorted (by region offset) list of pools
+	 * use the first free slot which is large enough to accomodate
+	 * the request
+	 */
+	last_offset = 0;
+	slot_found = false;
+	node = &region->pools;
+	list_for_each_entry(pi, &region->pools, region_inst) {
+		if ((pi->region_offset - last_offset) >= num_desc) {
+			slot_found = true;
+			break;
+		}
+		last_offset = pi->region_offset + pi->num_desc;
+	}
+	node = &pi->region_inst;
+
+	if (slot_found) {
+		pool->region = region;
+		pool->num_desc = num_desc;
+		pool->region_offset = last_offset;
+		region->used_desc += num_desc;
+		list_add_tail(&pool->list, &kdev->pools);
+		list_add_tail(&pool->region_inst, node);
+	} else {
+		dev_err(kdev->dev, "pool(%s) create failed: fragmented desc pool in region(%d)\n",
+			name, region_id);
+		ret = -ENOMEM;
+		goto err_unlock;
+	}
+
+	mutex_unlock(&knav_dev_lock);
+	kdesc_fill_pool(pool);
+	return pool;
+
+err_unlock:
+	mutex_unlock(&knav_dev_lock);
+err:
+	kfree(pool->name);
+	devm_kfree(kdev->dev, pool);
+	return ERR_PTR(ret);
+}
+EXPORT_SYMBOL_GPL(knav_pool_create);
+
+/**
+ * knav_pool_destroy()	- Free a pool of descriptors
+ * @pool		- pool handle
+ */
+void knav_pool_destroy(void *ph)
+{
+	struct knav_pool *pool = ph;
+
+	if (!pool)
+		return;
+
+	if (!pool->region)
+		return;
+
+	kdesc_empty_pool(pool);
+	mutex_lock(&knav_dev_lock);
+
+	pool->region->used_desc -= pool->num_desc;
+	list_del(&pool->region_inst);
+	list_del(&pool->list);
+
+	mutex_unlock(&knav_dev_lock);
+	kfree(pool->name);
+	devm_kfree(kdev->dev, pool);
+}
+EXPORT_SYMBOL_GPL(knav_pool_destroy);
+
+
+/**
+ * knav_pool_desc_get()	- Get a descriptor from the pool
+ * @pool			- pool handle
+ *
+ * Returns descriptor from the pool.
+ */
+void *knav_pool_desc_get(void *ph)
+{
+	struct knav_pool *pool = ph;
+	dma_addr_t dma;
+	unsigned size;
+	void *data;
+
+	dma = knav_queue_pop(pool->queue, &size);
+	if (unlikely(!dma))
+		return ERR_PTR(-ENOMEM);
+	data = knav_pool_desc_dma_to_virt(pool, dma);
+	return data;
+}
+EXPORT_SYMBOL_GPL(knav_pool_desc_get);
+
+/**
+ * knav_pool_desc_put()	- return a descriptor to the pool
+ * @pool			- pool handle
+ */
+void knav_pool_desc_put(void *ph, void *desc)
+{
+	struct knav_pool *pool = ph;
+	dma_addr_t dma;
+	dma = knav_pool_desc_virt_to_dma(pool, desc);
+	knav_queue_push(pool->queue, dma, pool->region->desc_size, 0);
+}
+EXPORT_SYMBOL_GPL(knav_pool_desc_put);
+
+/**
+ * knav_pool_desc_map()	- Map descriptor for DMA transfer
+ * @pool			- pool handle
+ * @desc			- address of descriptor to map
+ * @size			- size of descriptor to map
+ * @dma				- DMA address return pointer
+ * @dma_sz			- adjusted return pointer
+ *
+ * Returns 0 on success, errno otherwise.
+ */
+int knav_pool_desc_map(void *ph, void *desc, unsigned size,
+					dma_addr_t *dma, unsigned *dma_sz)
+{
+	struct knav_pool *pool = ph;
+	*dma = knav_pool_desc_virt_to_dma(pool, desc);
+	size = min(size, pool->region->desc_size);
+	size = ALIGN(size, SMP_CACHE_BYTES);
+	*dma_sz = size;
+	dma_sync_single_for_device(pool->dev, *dma, size, DMA_TO_DEVICE);
+
+	/* Ensure the descriptor reaches to the memory */
+	__iowmb();
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(knav_pool_desc_map);
+
+/**
+ * knav_pool_desc_unmap()	- Unmap descriptor after DMA transfer
+ * @pool			- pool handle
+ * @dma				- DMA address of descriptor to unmap
+ * @dma_sz			- size of descriptor to unmap
+ *
+ * Returns descriptor address on success, Use IS_ERR_OR_NULL() to identify
+ * error values on return.
+ */
+void *knav_pool_desc_unmap(void *ph, dma_addr_t dma, unsigned dma_sz)
+{
+	struct knav_pool *pool = ph;
+	unsigned desc_sz;
+	void *desc;
+
+	desc_sz = min(dma_sz, pool->region->desc_size);
+	desc = knav_pool_desc_dma_to_virt(pool, dma);
+	dma_sync_single_for_cpu(pool->dev, dma, desc_sz, DMA_FROM_DEVICE);
+	prefetch(desc);
+	return desc;
+}
+EXPORT_SYMBOL_GPL(knav_pool_desc_unmap);
+
+/**
+ * knav_pool_count()	- Get the number of descriptors in pool.
+ * @pool		- pool handle
+ * Returns number of elements in the pool.
+ */
+int knav_pool_count(void *ph)
+{
+	struct knav_pool *pool = ph;
+	return knav_queue_get_count(pool->queue);
+}
+EXPORT_SYMBOL_GPL(knav_pool_count);
+
+static void knav_queue_setup_region(struct knav_device *kdev,
+					struct knav_region *region)
+{
+	unsigned hw_num_desc, hw_desc_size, size;
+	struct knav_reg_region __iomem  *regs;
+	struct knav_qmgr_info *qmgr;
+	struct knav_pool *pool;
+	int id = region->id;
+	struct page *page;
+
+	/* unused region? */
+	if (!region->num_desc) {
+		dev_warn(kdev->dev, "unused region %s\n", region->name);
+		return;
+	}
+
+	/* get hardware descriptor value */
+	hw_num_desc = ilog2(region->num_desc - 1) + 1;
+
+	/* did we force fit ourselves into nothingness? */
+	if (region->num_desc < 32) {
+		region->num_desc = 0;
+		dev_warn(kdev->dev, "too few descriptors in region %s\n",
+			 region->name);
+		return;
+	}
+
+	size = region->num_desc * region->desc_size;
+	region->virt_start = alloc_pages_exact(size, GFP_KERNEL | GFP_DMA |
+						GFP_DMA32);
+	if (!region->virt_start) {
+		region->num_desc = 0;
+		dev_err(kdev->dev, "memory alloc failed for region %s\n",
+			region->name);
+		return;
+	}
+	region->virt_end = region->virt_start + size;
+	page = virt_to_page(region->virt_start);
+
+	region->dma_start = dma_map_page(kdev->dev, page, 0, size,
+					 DMA_BIDIRECTIONAL);
+	if (dma_mapping_error(kdev->dev, region->dma_start)) {
+		dev_err(kdev->dev, "dma map failed for region %s\n",
+			region->name);
+		goto fail;
+	}
+	region->dma_end = region->dma_start + size;
+
+	pool = devm_kzalloc(kdev->dev, sizeof(*pool), GFP_KERNEL);
+	if (!pool) {
+		dev_err(kdev->dev, "out of memory allocating dummy pool\n");
+		goto fail;
+	}
+	pool->num_desc = 0;
+	pool->region_offset = region->num_desc;
+	list_add(&pool->region_inst, &region->pools);
+
+	dev_dbg(kdev->dev,
+		"region %s (%d): size:%d, link:%d@%d, dma:%pad-%pad, virt:%p-%p\n",
+		region->name, id, region->desc_size, region->num_desc,
+		region->link_index, &region->dma_start, &region->dma_end,
+		region->virt_start, region->virt_end);
+
+	hw_desc_size = (region->desc_size / 16) - 1;
+	hw_num_desc -= 5;
+
+	for_each_qmgr(kdev, qmgr) {
+		regs = qmgr->reg_region + id;
+		writel_relaxed((u32)region->dma_start, &regs->base);
+		writel_relaxed(region->link_index, &regs->start_index);
+		writel_relaxed(hw_desc_size << 16 | hw_num_desc,
+			       &regs->size_count);
+	}
+	return;
+
+fail:
+	if (region->dma_start)
+		dma_unmap_page(kdev->dev, region->dma_start, size,
+				DMA_BIDIRECTIONAL);
+	if (region->virt_start)
+		free_pages_exact(region->virt_start, size);
+	region->num_desc = 0;
+	return;
+}
+
+static const char *knav_queue_find_name(struct device_node *node)
+{
+	const char *name;
+
+	if (of_property_read_string(node, "label", &name) < 0)
+		name = node->name;
+	if (!name)
+		name = "unknown";
+	return name;
+}
+
+static int knav_queue_setup_regions(struct knav_device *kdev,
+					struct device_node *regions)
+{
+	struct device *dev = kdev->dev;
+	struct knav_region *region;
+	struct device_node *child;
+	u32 temp[2];
+	int ret;
+
+	for_each_child_of_node(regions, child) {
+		region = devm_kzalloc(dev, sizeof(*region), GFP_KERNEL);
+		if (!region) {
+			dev_err(dev, "out of memory allocating region\n");
+			return -ENOMEM;
+		}
+
+		region->name = knav_queue_find_name(child);
+		of_property_read_u32(child, "id", &region->id);
+		ret = of_property_read_u32_array(child, "region-spec", temp, 2);
+		if (!ret) {
+			region->num_desc  = temp[0];
+			region->desc_size = temp[1];
+		} else {
+			dev_err(dev, "invalid region info %s\n", region->name);
+			devm_kfree(dev, region);
+			continue;
+		}
+
+		if (!of_get_property(child, "link-index", NULL)) {
+			dev_err(dev, "No link info for %s\n", region->name);
+			devm_kfree(dev, region);
+			continue;
+		}
+		ret = of_property_read_u32(child, "link-index",
+					   &region->link_index);
+		if (ret) {
+			dev_err(dev, "link index not found for %s\n",
+				region->name);
+			devm_kfree(dev, region);
+			continue;
+		}
+
+		INIT_LIST_HEAD(&region->pools);
+		list_add_tail(&region->list, &kdev->regions);
+	}
+	if (list_empty(&kdev->regions)) {
+		dev_err(dev, "no valid region information found\n");
+		return -ENODEV;
+	}
+
+	/* Next, we run through the regions and set things up */
+	for_each_region(kdev, region)
+		knav_queue_setup_region(kdev, region);
+
+	return 0;
+}
+
+static int knav_get_link_ram(struct knav_device *kdev,
+				       const char *name,
+				       struct knav_link_ram_block *block)
+{
+	struct platform_device *pdev = to_platform_device(kdev->dev);
+	struct device_node *node = pdev->dev.of_node;
+	u32 temp[2];
+
+	/*
+	 * Note: link ram resources are specified in "entry" sized units. In
+	 * reality, although entries are ~40bits in hardware, we treat them as
+	 * 64-bit entities here.
+	 *
+	 * For example, to specify the internal link ram for Keystone-I class
+	 * devices, we would set the linkram0 resource to 0x80000-0x83fff.
+	 *
+	 * This gets a bit weird when other link rams are used.  For example,
+	 * if the range specified is 0x0c000000-0x0c003fff (i.e., 16K entries
+	 * in MSMC SRAM), the actual memory used is 0x0c000000-0x0c020000,
+	 * which accounts for 64-bits per entry, for 16K entries.
+	 */
+	if (!of_property_read_u32_array(node, name , temp, 2)) {
+		if (temp[0]) {
+			/*
+			 * queue_base specified => using internal or onchip
+			 * link ram WARNING - we do not "reserve" this block
+			 */
+			block->dma = (dma_addr_t)temp[0];
+			block->virt = NULL;
+			block->size = temp[1];
+		} else {
+			block->size = temp[1];
+			/* queue_base not specific => allocate requested size */
+			block->virt = dmam_alloc_coherent(kdev->dev,
+						  8 * block->size, &block->dma,
+						  GFP_KERNEL);
+			if (!block->virt) {
+				dev_err(kdev->dev, "failed to alloc linkram\n");
+				return -ENOMEM;
+			}
+		}
+	} else {
+		return -ENODEV;
+	}
+	return 0;
+}
+
+static int knav_queue_setup_link_ram(struct knav_device *kdev)
+{
+	struct knav_link_ram_block *block;
+	struct knav_qmgr_info *qmgr;
+
+	for_each_qmgr(kdev, qmgr) {
+		block = &kdev->link_rams[0];
+		dev_dbg(kdev->dev, "linkram0: dma:%pad, virt:%p, size:%x\n",
+			&block->dma, block->virt, block->size);
+		writel_relaxed((u32)block->dma, &qmgr->reg_config->link_ram_base0);
+		if (kdev->version == QMSS_66AK2G)
+			writel_relaxed(block->size,
+				       &qmgr->reg_config->link_ram_size0);
+		else
+			writel_relaxed(block->size - 1,
+				       &qmgr->reg_config->link_ram_size0);
+		block++;
+		if (!block->size)
+			continue;
+
+		dev_dbg(kdev->dev, "linkram1: dma:%pad, virt:%p, size:%x\n",
+			&block->dma, block->virt, block->size);
+		writel_relaxed(block->dma, &qmgr->reg_config->link_ram_base1);
+	}
+
+	return 0;
+}
+
+static int knav_setup_queue_range(struct knav_device *kdev,
+					struct device_node *node)
+{
+	struct device *dev = kdev->dev;
+	struct knav_range_info *range;
+	struct knav_qmgr_info *qmgr;
+	u32 temp[2], start, end, id, index;
+	int ret, i;
+
+	range = devm_kzalloc(dev, sizeof(*range), GFP_KERNEL);
+	if (!range) {
+		dev_err(dev, "out of memory allocating range\n");
+		return -ENOMEM;
+	}
+
+	range->kdev = kdev;
+	range->name = knav_queue_find_name(node);
+	ret = of_property_read_u32_array(node, "qrange", temp, 2);
+	if (!ret) {
+		range->queue_base = temp[0] - kdev->base_id;
+		range->num_queues = temp[1];
+	} else {
+		dev_err(dev, "invalid queue range %s\n", range->name);
+		devm_kfree(dev, range);
+		return -EINVAL;
+	}
+
+	for (i = 0; i < RANGE_MAX_IRQS; i++) {
+		struct of_phandle_args oirq;
+
+		if (of_irq_parse_one(node, i, &oirq))
+			break;
+
+		range->irqs[i].irq = irq_create_of_mapping(&oirq);
+		if (range->irqs[i].irq == IRQ_NONE)
+			break;
+
+		range->num_irqs++;
+
+		if (IS_ENABLED(CONFIG_SMP) && oirq.args_count == 3) {
+			unsigned long mask;
+			int bit;
+
+			range->irqs[i].cpu_mask = devm_kzalloc(dev,
+							       cpumask_size(), GFP_KERNEL);
+			if (!range->irqs[i].cpu_mask)
+				return -ENOMEM;
+
+			mask = (oirq.args[2] & 0x0000ff00) >> 8;
+			for_each_set_bit(bit, &mask, BITS_PER_LONG)
+				cpumask_set_cpu(bit, range->irqs[i].cpu_mask);
+		}
+	}
+
+	range->num_irqs = min(range->num_irqs, range->num_queues);
+	if (range->num_irqs)
+		range->flags |= RANGE_HAS_IRQ;
+
+	if (of_get_property(node, "qalloc-by-id", NULL))
+		range->flags |= RANGE_RESERVED;
+
+	if (of_get_property(node, "accumulator", NULL)) {
+		ret = knav_init_acc_range(kdev, node, range);
+		if (ret < 0) {
+			devm_kfree(dev, range);
+			return ret;
+		}
+	} else {
+		range->ops = &knav_gp_range_ops;
+	}
+
+	/* set threshold to 1, and flush out the queues */
+	for_each_qmgr(kdev, qmgr) {
+		start = max(qmgr->start_queue, range->queue_base);
+		end   = min(qmgr->start_queue + qmgr->num_queues,
+			    range->queue_base + range->num_queues);
+		for (id = start; id < end; id++) {
+			index = id - qmgr->start_queue;
+			writel_relaxed(THRESH_GTE | 1,
+				       &qmgr->reg_peek[index].ptr_size_thresh);
+			writel_relaxed(0,
+				       &qmgr->reg_push[index].ptr_size_thresh);
+		}
+	}
+
+	list_add_tail(&range->list, &kdev->queue_ranges);
+	dev_dbg(dev, "added range %s: %d-%d, %d irqs%s%s%s\n",
+		range->name, range->queue_base,
+		range->queue_base + range->num_queues - 1,
+		range->num_irqs,
+		(range->flags & RANGE_HAS_IRQ) ? ", has irq" : "",
+		(range->flags & RANGE_RESERVED) ? ", reserved" : "",
+		(range->flags & RANGE_HAS_ACCUMULATOR) ? ", acc" : "");
+	kdev->num_queues_in_use += range->num_queues;
+	return 0;
+}
+
+static int knav_setup_queue_pools(struct knav_device *kdev,
+				   struct device_node *queue_pools)
+{
+	struct device_node *type, *range;
+	int ret;
+
+	for_each_child_of_node(queue_pools, type) {
+		for_each_child_of_node(type, range) {
+			ret = knav_setup_queue_range(kdev, range);
+			/* return value ignored, we init the rest... */
+		}
+	}
+
+	/* ... and barf if they all failed! */
+	if (list_empty(&kdev->queue_ranges)) {
+		dev_err(kdev->dev, "no valid queue range found\n");
+		return -ENODEV;
+	}
+	return 0;
+}
+
+static void knav_free_queue_range(struct knav_device *kdev,
+				  struct knav_range_info *range)
+{
+	if (range->ops && range->ops->free_range)
+		range->ops->free_range(range);
+	list_del(&range->list);
+	devm_kfree(kdev->dev, range);
+}
+
+static void knav_free_queue_ranges(struct knav_device *kdev)
+{
+	struct knav_range_info *range;
+
+	for (;;) {
+		range = first_queue_range(kdev);
+		if (!range)
+			break;
+		knav_free_queue_range(kdev, range);
+	}
+}
+
+static void knav_queue_free_regions(struct knav_device *kdev)
+{
+	struct knav_region *region;
+	struct knav_pool *pool, *tmp;
+	unsigned size;
+
+	for (;;) {
+		region = first_region(kdev);
+		if (!region)
+			break;
+		list_for_each_entry_safe(pool, tmp, &region->pools, region_inst)
+			knav_pool_destroy(pool);
+
+		size = region->virt_end - region->virt_start;
+		if (size)
+			free_pages_exact(region->virt_start, size);
+		list_del(&region->list);
+		devm_kfree(kdev->dev, region);
+	}
+}
+
+static void __iomem *knav_queue_map_reg(struct knav_device *kdev,
+					struct device_node *node, int index)
+{
+	struct resource res;
+	void __iomem *regs;
+	int ret;
+
+	ret = of_address_to_resource(node, index, &res);
+	if (ret) {
+		dev_err(kdev->dev, "Can't translate of node(%s) address for index(%d)\n",
+			node->name, index);
+		return ERR_PTR(ret);
+	}
+
+	regs = devm_ioremap_resource(kdev->dev, &res);
+	if (IS_ERR(regs))
+		dev_err(kdev->dev, "Failed to map register base for index(%d) node(%s)\n",
+			index, node->name);
+	return regs;
+}
+
+static int knav_queue_init_qmgrs(struct knav_device *kdev,
+					struct device_node *qmgrs)
+{
+	struct device *dev = kdev->dev;
+	struct knav_qmgr_info *qmgr;
+	struct device_node *child;
+	u32 temp[2];
+	int ret;
+
+	for_each_child_of_node(qmgrs, child) {
+		qmgr = devm_kzalloc(dev, sizeof(*qmgr), GFP_KERNEL);
+		if (!qmgr) {
+			dev_err(dev, "out of memory allocating qmgr\n");
+			return -ENOMEM;
+		}
+
+		ret = of_property_read_u32_array(child, "managed-queues",
+						 temp, 2);
+		if (!ret) {
+			qmgr->start_queue = temp[0];
+			qmgr->num_queues = temp[1];
+		} else {
+			dev_err(dev, "invalid qmgr queue range\n");
+			devm_kfree(dev, qmgr);
+			continue;
+		}
+
+		dev_info(dev, "qmgr start queue %d, number of queues %d\n",
+			 qmgr->start_queue, qmgr->num_queues);
+
+		qmgr->reg_peek =
+			knav_queue_map_reg(kdev, child,
+					   KNAV_QUEUE_PEEK_REG_INDEX);
+
+		if (kdev->version == QMSS) {
+			qmgr->reg_status =
+				knav_queue_map_reg(kdev, child,
+						   KNAV_QUEUE_STATUS_REG_INDEX);
+		}
+
+		qmgr->reg_config =
+			knav_queue_map_reg(kdev, child,
+					   (kdev->version == QMSS_66AK2G) ?
+					   KNAV_L_QUEUE_CONFIG_REG_INDEX :
+					   KNAV_QUEUE_CONFIG_REG_INDEX);
+		qmgr->reg_region =
+			knav_queue_map_reg(kdev, child,
+					   (kdev->version == QMSS_66AK2G) ?
+					   KNAV_L_QUEUE_REGION_REG_INDEX :
+					   KNAV_QUEUE_REGION_REG_INDEX);
+
+		qmgr->reg_push =
+			knav_queue_map_reg(kdev, child,
+					   (kdev->version == QMSS_66AK2G) ?
+					    KNAV_L_QUEUE_PUSH_REG_INDEX :
+					    KNAV_QUEUE_PUSH_REG_INDEX);
+
+		if (kdev->version == QMSS) {
+			qmgr->reg_pop =
+				knav_queue_map_reg(kdev, child,
+						   KNAV_QUEUE_POP_REG_INDEX);
+		}
+
+		if (IS_ERR(qmgr->reg_peek) ||
+		    ((kdev->version == QMSS) &&
+		    (IS_ERR(qmgr->reg_status) || IS_ERR(qmgr->reg_pop))) ||
+		    IS_ERR(qmgr->reg_config) || IS_ERR(qmgr->reg_region) ||
+		    IS_ERR(qmgr->reg_push)) {
+			dev_err(dev, "failed to map qmgr regs\n");
+			if (kdev->version == QMSS) {
+				if (!IS_ERR(qmgr->reg_status))
+					devm_iounmap(dev, qmgr->reg_status);
+				if (!IS_ERR(qmgr->reg_pop))
+					devm_iounmap(dev, qmgr->reg_pop);
+			}
+			if (!IS_ERR(qmgr->reg_peek))
+				devm_iounmap(dev, qmgr->reg_peek);
+			if (!IS_ERR(qmgr->reg_config))
+				devm_iounmap(dev, qmgr->reg_config);
+			if (!IS_ERR(qmgr->reg_region))
+				devm_iounmap(dev, qmgr->reg_region);
+			if (!IS_ERR(qmgr->reg_push))
+				devm_iounmap(dev, qmgr->reg_push);
+			devm_kfree(dev, qmgr);
+			continue;
+		}
+
+		/* Use same push register for pop as well */
+		if (kdev->version == QMSS_66AK2G)
+			qmgr->reg_pop = qmgr->reg_push;
+
+		list_add_tail(&qmgr->list, &kdev->qmgrs);
+		dev_info(dev, "added qmgr start queue %d, num of queues %d, reg_peek %p, reg_status %p, reg_config %p, reg_region %p, reg_push %p, reg_pop %p\n",
+			 qmgr->start_queue, qmgr->num_queues,
+			 qmgr->reg_peek, qmgr->reg_status,
+			 qmgr->reg_config, qmgr->reg_region,
+			 qmgr->reg_push, qmgr->reg_pop);
+	}
+	return 0;
+}
+
+static int knav_queue_init_pdsps(struct knav_device *kdev,
+					struct device_node *pdsps)
+{
+	struct device *dev = kdev->dev;
+	struct knav_pdsp_info *pdsp;
+	struct device_node *child;
+
+	for_each_child_of_node(pdsps, child) {
+		pdsp = devm_kzalloc(dev, sizeof(*pdsp), GFP_KERNEL);
+		if (!pdsp) {
+			dev_err(dev, "out of memory allocating pdsp\n");
+			return -ENOMEM;
+		}
+		pdsp->name = knav_queue_find_name(child);
+		pdsp->iram =
+			knav_queue_map_reg(kdev, child,
+					   KNAV_QUEUE_PDSP_IRAM_REG_INDEX);
+		pdsp->regs =
+			knav_queue_map_reg(kdev, child,
+					   KNAV_QUEUE_PDSP_REGS_REG_INDEX);
+		pdsp->intd =
+			knav_queue_map_reg(kdev, child,
+					   KNAV_QUEUE_PDSP_INTD_REG_INDEX);
+		pdsp->command =
+			knav_queue_map_reg(kdev, child,
+					   KNAV_QUEUE_PDSP_CMD_REG_INDEX);
+
+		if (IS_ERR(pdsp->command) || IS_ERR(pdsp->iram) ||
+		    IS_ERR(pdsp->regs) || IS_ERR(pdsp->intd)) {
+			dev_err(dev, "failed to map pdsp %s regs\n",
+				pdsp->name);
+			if (!IS_ERR(pdsp->command))
+				devm_iounmap(dev, pdsp->command);
+			if (!IS_ERR(pdsp->iram))
+				devm_iounmap(dev, pdsp->iram);
+			if (!IS_ERR(pdsp->regs))
+				devm_iounmap(dev, pdsp->regs);
+			if (!IS_ERR(pdsp->intd))
+				devm_iounmap(dev, pdsp->intd);
+			devm_kfree(dev, pdsp);
+			continue;
+		}
+		of_property_read_u32(child, "id", &pdsp->id);
+		list_add_tail(&pdsp->list, &kdev->pdsps);
+		dev_dbg(dev, "added pdsp %s: command %p, iram %p, regs %p, intd %p\n",
+			pdsp->name, pdsp->command, pdsp->iram, pdsp->regs,
+			pdsp->intd);
+	}
+	return 0;
+}
+
+static int knav_queue_stop_pdsp(struct knav_device *kdev,
+			  struct knav_pdsp_info *pdsp)
+{
+	u32 val, timeout = 1000;
+	int ret;
+
+	val = readl_relaxed(&pdsp->regs->control) & ~PDSP_CTRL_ENABLE;
+	writel_relaxed(val, &pdsp->regs->control);
+	ret = knav_queue_pdsp_wait(&pdsp->regs->control, timeout,
+					PDSP_CTRL_RUNNING);
+	if (ret < 0) {
+		dev_err(kdev->dev, "timed out on pdsp %s stop\n", pdsp->name);
+		return ret;
+	}
+	pdsp->loaded = false;
+	pdsp->started = false;
+	return 0;
+}
+
+static int knav_queue_load_pdsp(struct knav_device *kdev,
+			  struct knav_pdsp_info *pdsp)
+{
+	int i, ret, fwlen;
+	const struct firmware *fw;
+	bool found = false;
+	u32 *fwdata;
+
+	for (i = 0; i < ARRAY_SIZE(knav_acc_firmwares); i++) {
+		if (knav_acc_firmwares[i]) {
+			ret = request_firmware_direct(&fw,
+						      knav_acc_firmwares[i],
+						      kdev->dev);
+			if (!ret) {
+				found = true;
+				break;
+			}
+		}
+	}
+
+	if (!found) {
+		dev_err(kdev->dev, "failed to get firmware for pdsp\n");
+		return -ENODEV;
+	}
+
+	dev_info(kdev->dev, "firmware file %s downloaded for PDSP\n",
+		 knav_acc_firmwares[i]);
+
+	writel_relaxed(pdsp->id + 1, pdsp->command + 0x18);
+	/* download the firmware */
+	fwdata = (u32 *)fw->data;
+	fwlen = (fw->size + sizeof(u32) - 1) / sizeof(u32);
+	for (i = 0; i < fwlen; i++)
+		writel_relaxed(be32_to_cpu(fwdata[i]), pdsp->iram + i);
+
+	release_firmware(fw);
+	return 0;
+}
+
+static int knav_queue_start_pdsp(struct knav_device *kdev,
+			   struct knav_pdsp_info *pdsp)
+{
+	u32 val, timeout = 1000;
+	int ret;
+
+	/* write a command for sync */
+	writel_relaxed(0xffffffff, pdsp->command);
+	while (readl_relaxed(pdsp->command) != 0xffffffff)
+		cpu_relax();
+
+	/* soft reset the PDSP */
+	val  = readl_relaxed(&pdsp->regs->control);
+	val &= ~(PDSP_CTRL_PC_MASK | PDSP_CTRL_SOFT_RESET);
+	writel_relaxed(val, &pdsp->regs->control);
+
+	/* enable pdsp */
+	val = readl_relaxed(&pdsp->regs->control) | PDSP_CTRL_ENABLE;
+	writel_relaxed(val, &pdsp->regs->control);
+
+	/* wait for command register to clear */
+	ret = knav_queue_pdsp_wait(pdsp->command, timeout, 0);
+	if (ret < 0) {
+		dev_err(kdev->dev,
+			"timed out on pdsp %s command register wait\n",
+			pdsp->name);
+		return ret;
+	}
+	return 0;
+}
+
+static void knav_queue_stop_pdsps(struct knav_device *kdev)
+{
+	struct knav_pdsp_info *pdsp;
+
+	/* disable all pdsps */
+	for_each_pdsp(kdev, pdsp)
+		knav_queue_stop_pdsp(kdev, pdsp);
+}
+
+static int knav_queue_start_pdsps(struct knav_device *kdev)
+{
+	struct knav_pdsp_info *pdsp;
+	int ret;
+
+	knav_queue_stop_pdsps(kdev);
+	/* now load them all. We return success even if pdsp
+	 * is not loaded as acc channels are optional on having
+	 * firmware availability in the system. We set the loaded
+	 * and stated flag and when initialize the acc range, check
+	 * it and init the range only if pdsp is started.
+	 */
+	for_each_pdsp(kdev, pdsp) {
+		ret = knav_queue_load_pdsp(kdev, pdsp);
+		if (!ret)
+			pdsp->loaded = true;
+	}
+
+	for_each_pdsp(kdev, pdsp) {
+		if (pdsp->loaded) {
+			ret = knav_queue_start_pdsp(kdev, pdsp);
+			if (!ret)
+				pdsp->started = true;
+		}
+	}
+	return 0;
+}
+
+static inline struct knav_qmgr_info *knav_find_qmgr(unsigned id)
+{
+	struct knav_qmgr_info *qmgr;
+
+	for_each_qmgr(kdev, qmgr) {
+		if ((id >= qmgr->start_queue) &&
+		    (id < qmgr->start_queue + qmgr->num_queues))
+			return qmgr;
+	}
+	return NULL;
+}
+
+static int knav_queue_init_queue(struct knav_device *kdev,
+					struct knav_range_info *range,
+					struct knav_queue_inst *inst,
+					unsigned id)
+{
+	char irq_name[KNAV_NAME_SIZE];
+	inst->qmgr = knav_find_qmgr(id);
+	if (!inst->qmgr)
+		return -1;
+
+	INIT_LIST_HEAD(&inst->handles);
+	inst->kdev = kdev;
+	inst->range = range;
+	inst->irq_num = -1;
+	inst->id = id;
+	scnprintf(irq_name, sizeof(irq_name), "hwqueue-%d", id);
+	inst->irq_name = kstrndup(irq_name, sizeof(irq_name), GFP_KERNEL);
+
+	if (range->ops && range->ops->init_queue)
+		return range->ops->init_queue(range, inst);
+	else
+		return 0;
+}
+
+static int knav_queue_init_queues(struct knav_device *kdev)
+{
+	struct knav_range_info *range;
+	int size, id, base_idx;
+	int idx = 0, ret = 0;
+
+	/* how much do we need for instance data? */
+	size = sizeof(struct knav_queue_inst);
+
+	/* round this up to a power of 2, keep the index to instance
+	 * arithmetic fast.
+	 * */
+	kdev->inst_shift = order_base_2(size);
+	size = (1 << kdev->inst_shift) * kdev->num_queues_in_use;
+	kdev->instances = devm_kzalloc(kdev->dev, size, GFP_KERNEL);
+	if (!kdev->instances)
+		return -ENOMEM;
+
+	for_each_queue_range(kdev, range) {
+		if (range->ops && range->ops->init_range)
+			range->ops->init_range(range);
+		base_idx = idx;
+		for (id = range->queue_base;
+		     id < range->queue_base + range->num_queues; id++, idx++) {
+			ret = knav_queue_init_queue(kdev, range,
+					knav_queue_idx_to_inst(kdev, idx), id);
+			if (ret < 0)
+				return ret;
+		}
+		range->queue_base_inst =
+			knav_queue_idx_to_inst(kdev, base_idx);
+	}
+	return 0;
+}
+
+/* Match table for of_platform binding */
+static const struct of_device_id keystone_qmss_of_match[] = {
+	{
+		.compatible = "ti,keystone-navigator-qmss",
+	},
+	{
+		.compatible = "ti,66ak2g-navss-qm",
+		.data	= (void *)QMSS_66AK2G,
+	},
+	{},
+};
+MODULE_DEVICE_TABLE(of, keystone_qmss_of_match);
+
+static int knav_queue_probe(struct platform_device *pdev)
+{
+	struct device_node *node = pdev->dev.of_node;
+	struct device_node *qmgrs, *queue_pools, *regions, *pdsps;
+	const struct of_device_id *match;
+	struct device *dev = &pdev->dev;
+	u32 temp[2];
+	int ret;
+
+	if (!node) {
+		dev_err(dev, "device tree info unavailable\n");
+		return -ENODEV;
+	}
+
+	kdev = devm_kzalloc(dev, sizeof(struct knav_device), GFP_KERNEL);
+	if (!kdev) {
+		dev_err(dev, "memory allocation failed\n");
+		return -ENOMEM;
+	}
+
+	match = of_match_device(of_match_ptr(keystone_qmss_of_match), dev);
+	if (match && match->data)
+		kdev->version = QMSS_66AK2G;
+
+	platform_set_drvdata(pdev, kdev);
+	kdev->dev = dev;
+	INIT_LIST_HEAD(&kdev->queue_ranges);
+	INIT_LIST_HEAD(&kdev->qmgrs);
+	INIT_LIST_HEAD(&kdev->pools);
+	INIT_LIST_HEAD(&kdev->regions);
+	INIT_LIST_HEAD(&kdev->pdsps);
+
+	pm_runtime_enable(&pdev->dev);
+	ret = pm_runtime_get_sync(&pdev->dev);
+	if (ret < 0) {
+		pm_runtime_put_noidle(&pdev->dev);
+		dev_err(dev, "Failed to enable QMSS\n");
+		return ret;
+	}
+
+	if (of_property_read_u32_array(node, "queue-range", temp, 2)) {
+		dev_err(dev, "queue-range not specified\n");
+		ret = -ENODEV;
+		goto err;
+	}
+	kdev->base_id    = temp[0];
+	kdev->num_queues = temp[1];
+
+	/* Initialize queue managers using device tree configuration */
+	qmgrs =  of_get_child_by_name(node, "qmgrs");
+	if (!qmgrs) {
+		dev_err(dev, "queue manager info not specified\n");
+		ret = -ENODEV;
+		goto err;
+	}
+	ret = knav_queue_init_qmgrs(kdev, qmgrs);
+	of_node_put(qmgrs);
+	if (ret)
+		goto err;
+
+	/* get pdsp configuration values from device tree */
+	pdsps =  of_get_child_by_name(node, "pdsps");
+	if (pdsps) {
+		ret = knav_queue_init_pdsps(kdev, pdsps);
+		if (ret)
+			goto err;
+
+		ret = knav_queue_start_pdsps(kdev);
+		if (ret)
+			goto err;
+	}
+	of_node_put(pdsps);
+
+	/* get usable queue range values from device tree */
+	queue_pools = of_get_child_by_name(node, "queue-pools");
+	if (!queue_pools) {
+		dev_err(dev, "queue-pools not specified\n");
+		ret = -ENODEV;
+		goto err;
+	}
+	ret = knav_setup_queue_pools(kdev, queue_pools);
+	of_node_put(queue_pools);
+	if (ret)
+		goto err;
+
+	ret = knav_get_link_ram(kdev, "linkram0", &kdev->link_rams[0]);
+	if (ret) {
+		dev_err(kdev->dev, "could not setup linking ram\n");
+		goto err;
+	}
+
+	ret = knav_get_link_ram(kdev, "linkram1", &kdev->link_rams[1]);
+	if (ret) {
+		/*
+		 * nothing really, we have one linking ram already, so we just
+		 * live within our means
+		 */
+	}
+
+	ret = knav_queue_setup_link_ram(kdev);
+	if (ret)
+		goto err;
+
+	regions = of_get_child_by_name(node, "descriptor-regions");
+	if (!regions) {
+		dev_err(dev, "descriptor-regions not specified\n");
+		ret = -ENODEV;
+		goto err;
+	}
+	ret = knav_queue_setup_regions(kdev, regions);
+	of_node_put(regions);
+	if (ret)
+		goto err;
+
+	ret = knav_queue_init_queues(kdev);
+	if (ret < 0) {
+		dev_err(dev, "hwqueue initialization failed\n");
+		goto err;
+	}
+
+	debugfs_create_file("qmss", S_IFREG | S_IRUGO, NULL, NULL,
+			    &knav_queue_debug_ops);
+	device_ready = true;
+	return 0;
+
+err:
+	knav_queue_stop_pdsps(kdev);
+	knav_queue_free_regions(kdev);
+	knav_free_queue_ranges(kdev);
+	pm_runtime_put_sync(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+	return ret;
+}
+
+static int knav_queue_remove(struct platform_device *pdev)
+{
+	/* TODO: Free resources */
+	pm_runtime_put_sync(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+	return 0;
+}
+
+static struct platform_driver keystone_qmss_driver = {
+	.probe		= knav_queue_probe,
+	.remove		= knav_queue_remove,
+	.driver		= {
+		.name	= "keystone-navigator-qmss",
+		.of_match_table = keystone_qmss_of_match,
+	},
+};
+module_platform_driver(keystone_qmss_driver);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("TI QMSS driver for Keystone SOCs");
+MODULE_AUTHOR("Sandeep Nair <sandeep_n@ti.com>");
+MODULE_AUTHOR("Santosh Shilimkar <santosh.shilimkar@ti.com>");
diff --git a/drivers/soc/ti/pm33xx.c b/drivers/soc/ti/pm33xx.c
new file mode 100644
index 000000000..d0dab3236
--- /dev/null
+++ b/drivers/soc/ti/pm33xx.c
@@ -0,0 +1,363 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * AM33XX Power Management Routines
+ *
+ * Copyright (C) 2012-2018 Texas Instruments Incorporated - http://www.ti.com/
+ *	Vaibhav Bedia, Dave Gerlach
+ */
+
+#include <linux/cpu.h>
+#include <linux/err.h>
+#include <linux/genalloc.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_data/pm33xx.h>
+#include <linux/platform_device.h>
+#include <linux/sizes.h>
+#include <linux/sram.h>
+#include <linux/suspend.h>
+#include <linux/ti-emif-sram.h>
+#include <linux/wkup_m3_ipc.h>
+
+#include <asm/proc-fns.h>
+#include <asm/suspend.h>
+#include <asm/system_misc.h>
+
+#define AMX3_PM_SRAM_SYMBOL_OFFSET(sym) ((unsigned long)(sym) - \
+					 (unsigned long)pm_sram->do_wfi)
+
+static int (*am33xx_do_wfi_sram)(unsigned long unused);
+static phys_addr_t am33xx_do_wfi_sram_phys;
+
+static struct gen_pool *sram_pool, *sram_pool_data;
+static unsigned long ocmcram_location, ocmcram_location_data;
+
+static struct am33xx_pm_platform_data *pm_ops;
+static struct am33xx_pm_sram_addr *pm_sram;
+
+static struct device *pm33xx_dev;
+static struct wkup_m3_ipc *m3_ipc;
+
+static unsigned long suspend_wfi_flags;
+
+static u32 sram_suspend_address(unsigned long addr)
+{
+	return ((unsigned long)am33xx_do_wfi_sram +
+		AMX3_PM_SRAM_SYMBOL_OFFSET(addr));
+}
+
+#ifdef CONFIG_SUSPEND
+static int am33xx_pm_suspend(suspend_state_t suspend_state)
+{
+	int i, ret = 0;
+
+	ret = pm_ops->soc_suspend((unsigned long)suspend_state,
+				  am33xx_do_wfi_sram, suspend_wfi_flags);
+
+	if (ret) {
+		dev_err(pm33xx_dev, "PM: Kernel suspend failure\n");
+	} else {
+		i = m3_ipc->ops->request_pm_status(m3_ipc);
+
+		switch (i) {
+		case 0:
+			dev_info(pm33xx_dev,
+				 "PM: Successfully put all powerdomains to target state\n");
+			break;
+		case 1:
+			dev_err(pm33xx_dev,
+				"PM: Could not transition all powerdomains to target state\n");
+			ret = -1;
+			break;
+		default:
+			dev_err(pm33xx_dev,
+				"PM: CM3 returned unknown result = %d\n", i);
+			ret = -1;
+		}
+	}
+
+	return ret;
+}
+
+static int am33xx_pm_enter(suspend_state_t suspend_state)
+{
+	int ret = 0;
+
+	switch (suspend_state) {
+	case PM_SUSPEND_MEM:
+	case PM_SUSPEND_STANDBY:
+		ret = am33xx_pm_suspend(suspend_state);
+		break;
+	default:
+		ret = -EINVAL;
+	}
+
+	return ret;
+}
+
+static int am33xx_pm_begin(suspend_state_t state)
+{
+	int ret = -EINVAL;
+
+	switch (state) {
+	case PM_SUSPEND_MEM:
+		ret = m3_ipc->ops->prepare_low_power(m3_ipc, WKUP_M3_DEEPSLEEP);
+		break;
+	case PM_SUSPEND_STANDBY:
+		ret = m3_ipc->ops->prepare_low_power(m3_ipc, WKUP_M3_STANDBY);
+		break;
+	}
+
+	return ret;
+}
+
+static void am33xx_pm_end(void)
+{
+	m3_ipc->ops->finish_low_power(m3_ipc);
+}
+
+static int am33xx_pm_valid(suspend_state_t state)
+{
+	switch (state) {
+	case PM_SUSPEND_STANDBY:
+	case PM_SUSPEND_MEM:
+		return 1;
+	default:
+		return 0;
+	}
+}
+
+static const struct platform_suspend_ops am33xx_pm_ops = {
+	.begin		= am33xx_pm_begin,
+	.end		= am33xx_pm_end,
+	.enter		= am33xx_pm_enter,
+	.valid		= am33xx_pm_valid,
+};
+#endif /* CONFIG_SUSPEND */
+
+static void am33xx_pm_set_ipc_ops(void)
+{
+	u32 resume_address;
+	int temp;
+
+	temp = ti_emif_get_mem_type();
+	if (temp < 0) {
+		dev_err(pm33xx_dev, "PM: Cannot determine memory type, no PM available\n");
+		return;
+	}
+	m3_ipc->ops->set_mem_type(m3_ipc, temp);
+
+	/* Physical resume address to be used by ROM code */
+	resume_address = am33xx_do_wfi_sram_phys +
+			 *pm_sram->resume_offset + 0x4;
+
+	m3_ipc->ops->set_resume_address(m3_ipc, (void *)resume_address);
+}
+
+static void am33xx_pm_free_sram(void)
+{
+	gen_pool_free(sram_pool, ocmcram_location, *pm_sram->do_wfi_sz);
+	gen_pool_free(sram_pool_data, ocmcram_location_data,
+		      sizeof(struct am33xx_pm_ro_sram_data));
+}
+
+/*
+ * Push the minimal suspend-resume code to SRAM
+ */
+static int am33xx_pm_alloc_sram(void)
+{
+	struct device_node *np;
+	int ret = 0;
+
+	np = of_find_compatible_node(NULL, NULL, "ti,omap3-mpu");
+	if (!np) {
+		np = of_find_compatible_node(NULL, NULL, "ti,omap4-mpu");
+		if (!np) {
+			dev_err(pm33xx_dev, "PM: %s: Unable to find device node for mpu\n",
+				__func__);
+			return -ENODEV;
+		}
+	}
+
+	sram_pool = of_gen_pool_get(np, "pm-sram", 0);
+	if (!sram_pool) {
+		dev_err(pm33xx_dev, "PM: %s: Unable to get sram pool for ocmcram\n",
+			__func__);
+		ret = -ENODEV;
+		goto mpu_put_node;
+	}
+
+	sram_pool_data = of_gen_pool_get(np, "pm-sram", 1);
+	if (!sram_pool_data) {
+		dev_err(pm33xx_dev, "PM: %s: Unable to get sram data pool for ocmcram\n",
+			__func__);
+		ret = -ENODEV;
+		goto mpu_put_node;
+	}
+
+	ocmcram_location = gen_pool_alloc(sram_pool, *pm_sram->do_wfi_sz);
+	if (!ocmcram_location) {
+		dev_err(pm33xx_dev, "PM: %s: Unable to allocate memory from ocmcram\n",
+			__func__);
+		ret = -ENOMEM;
+		goto mpu_put_node;
+	}
+
+	ocmcram_location_data = gen_pool_alloc(sram_pool_data,
+					       sizeof(struct emif_regs_amx3));
+	if (!ocmcram_location_data) {
+		dev_err(pm33xx_dev, "PM: Unable to allocate memory from ocmcram\n");
+		gen_pool_free(sram_pool, ocmcram_location, *pm_sram->do_wfi_sz);
+		ret = -ENOMEM;
+	}
+
+mpu_put_node:
+	of_node_put(np);
+	return ret;
+}
+
+static int am33xx_push_sram_idle(void)
+{
+	struct am33xx_pm_ro_sram_data ro_sram_data;
+	int ret;
+	u32 table_addr, ro_data_addr;
+	void *copy_addr;
+
+	ro_sram_data.amx3_pm_sram_data_virt = ocmcram_location_data;
+	ro_sram_data.amx3_pm_sram_data_phys =
+		gen_pool_virt_to_phys(sram_pool_data, ocmcram_location_data);
+	ro_sram_data.rtc_base_virt = pm_ops->get_rtc_base_addr();
+
+	/* Save physical address to calculate resume offset during pm init */
+	am33xx_do_wfi_sram_phys = gen_pool_virt_to_phys(sram_pool,
+							ocmcram_location);
+
+	am33xx_do_wfi_sram = sram_exec_copy(sram_pool, (void *)ocmcram_location,
+					    pm_sram->do_wfi,
+					    *pm_sram->do_wfi_sz);
+	if (!am33xx_do_wfi_sram) {
+		dev_err(pm33xx_dev,
+			"PM: %s: am33xx_do_wfi copy to sram failed\n",
+			__func__);
+		return -ENODEV;
+	}
+
+	table_addr =
+		sram_suspend_address((unsigned long)pm_sram->emif_sram_table);
+	ret = ti_emif_copy_pm_function_table(sram_pool, (void *)table_addr);
+	if (ret) {
+		dev_dbg(pm33xx_dev,
+			"PM: %s: EMIF function copy failed\n", __func__);
+		return -EPROBE_DEFER;
+	}
+
+	ro_data_addr =
+		sram_suspend_address((unsigned long)pm_sram->ro_sram_data);
+	copy_addr = sram_exec_copy(sram_pool, (void *)ro_data_addr,
+				   &ro_sram_data,
+				   sizeof(ro_sram_data));
+	if (!copy_addr) {
+		dev_err(pm33xx_dev,
+			"PM: %s: ro_sram_data copy to sram failed\n",
+			__func__);
+		return -ENODEV;
+	}
+
+	return 0;
+}
+
+static int am33xx_pm_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	int ret;
+
+	if (!of_machine_is_compatible("ti,am33xx") &&
+	    !of_machine_is_compatible("ti,am43"))
+		return -ENODEV;
+
+	pm_ops = dev->platform_data;
+	if (!pm_ops) {
+		dev_err(dev, "PM: Cannot get core PM ops!\n");
+		return -ENODEV;
+	}
+
+	pm_sram = pm_ops->get_sram_addrs();
+	if (!pm_sram) {
+		dev_err(dev, "PM: Cannot get PM asm function addresses!!\n");
+		return -ENODEV;
+	}
+
+	pm33xx_dev = dev;
+
+	ret = am33xx_pm_alloc_sram();
+	if (ret)
+		return ret;
+
+	ret = am33xx_push_sram_idle();
+	if (ret)
+		goto err_free_sram;
+
+	m3_ipc = wkup_m3_ipc_get();
+	if (!m3_ipc) {
+		dev_dbg(dev, "PM: Cannot get wkup_m3_ipc handle\n");
+		ret = -EPROBE_DEFER;
+		goto err_free_sram;
+	}
+
+	am33xx_pm_set_ipc_ops();
+
+#ifdef CONFIG_SUSPEND
+	suspend_set_ops(&am33xx_pm_ops);
+#endif /* CONFIG_SUSPEND */
+
+	/*
+	 * For a system suspend we must flush the caches, we want
+	 * the DDR in self-refresh, we want to save the context
+	 * of the EMIF, and we want the wkup_m3 to handle low-power
+	 * transition.
+	 */
+	suspend_wfi_flags |= WFI_FLAG_FLUSH_CACHE;
+	suspend_wfi_flags |= WFI_FLAG_SELF_REFRESH;
+	suspend_wfi_flags |= WFI_FLAG_SAVE_EMIF;
+	suspend_wfi_flags |= WFI_FLAG_WAKE_M3;
+
+	ret = pm_ops->init();
+	if (ret) {
+		dev_err(dev, "Unable to call core pm init!\n");
+		ret = -ENODEV;
+		goto err_put_wkup_m3_ipc;
+	}
+
+	return 0;
+
+err_put_wkup_m3_ipc:
+	wkup_m3_ipc_put(m3_ipc);
+err_free_sram:
+	am33xx_pm_free_sram();
+	pm33xx_dev = NULL;
+	return ret;
+}
+
+static int am33xx_pm_remove(struct platform_device *pdev)
+{
+	suspend_set_ops(NULL);
+	wkup_m3_ipc_put(m3_ipc);
+	am33xx_pm_free_sram();
+	return 0;
+}
+
+static struct platform_driver am33xx_pm_driver = {
+	.driver = {
+		.name   = "pm33xx",
+	},
+	.probe = am33xx_pm_probe,
+	.remove = am33xx_pm_remove,
+};
+module_platform_driver(am33xx_pm_driver);
+
+MODULE_ALIAS("platform:pm33xx");
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("am33xx power management driver");
diff --git a/drivers/soc/ti/ti_sci_pm_domains.c b/drivers/soc/ti/ti_sci_pm_domains.c
new file mode 100644
index 000000000..de31b9389
--- /dev/null
+++ b/drivers/soc/ti/ti_sci_pm_domains.c
@@ -0,0 +1,204 @@
+/*
+ * TI SCI Generic Power Domain Driver
+ *
+ * Copyright (C) 2015-2017 Texas Instruments Incorporated - http://www.ti.com/
+ *	J Keerthy <j-keerthy@ti.com>
+ *	Dave Gerlach <d-gerlach@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * version 2 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.
+ */
+
+#include <linux/err.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/pm_domain.h>
+#include <linux/slab.h>
+#include <linux/soc/ti/ti_sci_protocol.h>
+
+/**
+ * struct ti_sci_genpd_dev_data: holds data needed for every device attached
+ *				 to this genpd
+ * @idx: index of the device that identifies it with the system
+ *	 control processor.
+ */
+struct ti_sci_genpd_dev_data {
+	int idx;
+};
+
+/**
+ * struct ti_sci_pm_domain: TI specific data needed for power domain
+ * @ti_sci: handle to TI SCI protocol driver that provides ops to
+ *	    communicate with system control processor.
+ * @dev: pointer to dev for the driver for devm allocs
+ * @pd: generic_pm_domain for use with the genpd framework
+ */
+struct ti_sci_pm_domain {
+	const struct ti_sci_handle *ti_sci;
+	struct device *dev;
+	struct generic_pm_domain pd;
+};
+
+#define genpd_to_ti_sci_pd(gpd) container_of(gpd, struct ti_sci_pm_domain, pd)
+
+/**
+ * ti_sci_dev_id(): get prepopulated ti_sci id from struct dev
+ * @dev: pointer to device associated with this genpd
+ *
+ * Returns device_id stored from ti,sci_id property
+ */
+static int ti_sci_dev_id(struct device *dev)
+{
+	struct generic_pm_domain_data *genpd_data = dev_gpd_data(dev);
+	struct ti_sci_genpd_dev_data *sci_dev_data = genpd_data->data;
+
+	return sci_dev_data->idx;
+}
+
+/**
+ * ti_sci_dev_to_sci_handle(): get pointer to ti_sci_handle
+ * @dev: pointer to device associated with this genpd
+ *
+ * Returns ti_sci_handle to be used to communicate with system
+ *	   control processor.
+ */
+static const struct ti_sci_handle *ti_sci_dev_to_sci_handle(struct device *dev)
+{
+	struct generic_pm_domain *pd = pd_to_genpd(dev->pm_domain);
+	struct ti_sci_pm_domain *ti_sci_genpd = genpd_to_ti_sci_pd(pd);
+
+	return ti_sci_genpd->ti_sci;
+}
+
+/**
+ * ti_sci_dev_start(): genpd device start hook called to turn device on
+ * @dev: pointer to device associated with this genpd to be powered on
+ */
+static int ti_sci_dev_start(struct device *dev)
+{
+	const struct ti_sci_handle *ti_sci = ti_sci_dev_to_sci_handle(dev);
+	int idx = ti_sci_dev_id(dev);
+
+	return ti_sci->ops.dev_ops.get_device(ti_sci, idx);
+}
+
+/**
+ * ti_sci_dev_stop(): genpd device stop hook called to turn device off
+ * @dev: pointer to device associated with this genpd to be powered off
+ */
+static int ti_sci_dev_stop(struct device *dev)
+{
+	const struct ti_sci_handle *ti_sci = ti_sci_dev_to_sci_handle(dev);
+	int idx = ti_sci_dev_id(dev);
+
+	return ti_sci->ops.dev_ops.put_device(ti_sci, idx);
+}
+
+static int ti_sci_pd_attach_dev(struct generic_pm_domain *domain,
+				struct device *dev)
+{
+	struct device_node *np = dev->of_node;
+	struct of_phandle_args pd_args;
+	struct ti_sci_pm_domain *ti_sci_genpd = genpd_to_ti_sci_pd(domain);
+	const struct ti_sci_handle *ti_sci = ti_sci_genpd->ti_sci;
+	struct ti_sci_genpd_dev_data *sci_dev_data;
+	struct generic_pm_domain_data *genpd_data;
+	int idx, ret = 0;
+
+	ret = of_parse_phandle_with_args(np, "power-domains",
+					 "#power-domain-cells", 0, &pd_args);
+	if (ret < 0)
+		return ret;
+
+	if (pd_args.args_count != 1)
+		return -EINVAL;
+
+	idx = pd_args.args[0];
+
+	/*
+	 * Check the validity of the requested idx, if the index is not valid
+	 * the PMMC will return a NAK here and we will not allocate it.
+	 */
+	ret = ti_sci->ops.dev_ops.is_valid(ti_sci, idx);
+	if (ret)
+		return -EINVAL;
+
+	sci_dev_data = kzalloc(sizeof(*sci_dev_data), GFP_KERNEL);
+	if (!sci_dev_data)
+		return -ENOMEM;
+
+	sci_dev_data->idx = idx;
+
+	genpd_data = dev_gpd_data(dev);
+	genpd_data->data = sci_dev_data;
+
+	return 0;
+}
+
+static void ti_sci_pd_detach_dev(struct generic_pm_domain *domain,
+				 struct device *dev)
+{
+	struct generic_pm_domain_data *genpd_data = dev_gpd_data(dev);
+	struct ti_sci_genpd_dev_data *sci_dev_data = genpd_data->data;
+
+	kfree(sci_dev_data);
+	genpd_data->data = NULL;
+}
+
+static const struct of_device_id ti_sci_pm_domain_matches[] = {
+	{ .compatible = "ti,sci-pm-domain", },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, ti_sci_pm_domain_matches);
+
+static int ti_sci_pm_domain_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct device_node *np = dev->of_node;
+	struct ti_sci_pm_domain *ti_sci_pd;
+	int ret;
+
+	ti_sci_pd = devm_kzalloc(dev, sizeof(*ti_sci_pd), GFP_KERNEL);
+	if (!ti_sci_pd)
+		return -ENOMEM;
+
+	ti_sci_pd->ti_sci = devm_ti_sci_get_handle(dev);
+	if (IS_ERR(ti_sci_pd->ti_sci))
+		return PTR_ERR(ti_sci_pd->ti_sci);
+
+	ti_sci_pd->dev = dev;
+
+	ti_sci_pd->pd.name = "ti_sci_pd";
+
+	ti_sci_pd->pd.attach_dev = ti_sci_pd_attach_dev;
+	ti_sci_pd->pd.detach_dev = ti_sci_pd_detach_dev;
+
+	ti_sci_pd->pd.dev_ops.start = ti_sci_dev_start;
+	ti_sci_pd->pd.dev_ops.stop = ti_sci_dev_stop;
+
+	pm_genpd_init(&ti_sci_pd->pd, NULL, true);
+
+	ret = of_genpd_add_provider_simple(np, &ti_sci_pd->pd);
+
+	return ret;
+}
+
+static struct platform_driver ti_sci_pm_domains_driver = {
+	.probe = ti_sci_pm_domain_probe,
+	.driver = {
+		.name = "ti_sci_pm_domains",
+		.of_match_table = ti_sci_pm_domain_matches,
+	},
+};
+module_platform_driver(ti_sci_pm_domains_driver);
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("TI System Control Interface (SCI) Power Domain driver");
+MODULE_AUTHOR("Dave Gerlach");
diff --git a/drivers/soc/ti/wkup_m3_ipc.c b/drivers/soc/ti/wkup_m3_ipc.c
new file mode 100644
index 000000000..8358b9750
--- /dev/null
+++ b/drivers/soc/ti/wkup_m3_ipc.c
@@ -0,0 +1,583 @@
+/*
+ * AMx3 Wkup M3 IPC driver
+ *
+ * Copyright (C) 2015 Texas Instruments, Inc.
+ *
+ * Dave Gerlach <d-gerlach@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * version 2 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.
+ */
+
+#include <linux/err.h>
+#include <linux/kernel.h>
+#include <linux/kthread.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/omap-mailbox.h>
+#include <linux/platform_device.h>
+#include <linux/remoteproc.h>
+#include <linux/suspend.h>
+#include <linux/wkup_m3_ipc.h>
+
+#define AM33XX_CTRL_IPC_REG_COUNT	0x8
+#define AM33XX_CTRL_IPC_REG_OFFSET(m)	(0x4 + 4 * (m))
+
+/* AM33XX M3_TXEV_EOI register */
+#define AM33XX_CONTROL_M3_TXEV_EOI	0x00
+
+#define AM33XX_M3_TXEV_ACK		(0x1 << 0)
+#define AM33XX_M3_TXEV_ENABLE		(0x0 << 0)
+
+#define IPC_CMD_DS0			0x4
+#define IPC_CMD_STANDBY			0xc
+#define IPC_CMD_IDLE			0x10
+#define IPC_CMD_RESET			0xe
+#define DS_IPC_DEFAULT			0xffffffff
+#define M3_VERSION_UNKNOWN		0x0000ffff
+#define M3_BASELINE_VERSION		0x191
+#define M3_STATUS_RESP_MASK		(0xffff << 16)
+#define M3_FW_VERSION_MASK		0xffff
+#define M3_WAKE_SRC_MASK		0xff
+
+#define M3_STATE_UNKNOWN		0
+#define M3_STATE_RESET			1
+#define M3_STATE_INITED			2
+#define M3_STATE_MSG_FOR_LP		3
+#define M3_STATE_MSG_FOR_RESET		4
+
+static struct wkup_m3_ipc *m3_ipc_state;
+
+static const struct wkup_m3_wakeup_src wakeups[] = {
+	{.irq_nr = 35,	.src = "USB0_PHY"},
+	{.irq_nr = 36,	.src = "USB1_PHY"},
+	{.irq_nr = 40,	.src = "I2C0"},
+	{.irq_nr = 41,	.src = "RTC Timer"},
+	{.irq_nr = 42,	.src = "RTC Alarm"},
+	{.irq_nr = 43,	.src = "Timer0"},
+	{.irq_nr = 44,	.src = "Timer1"},
+	{.irq_nr = 45,	.src = "UART"},
+	{.irq_nr = 46,	.src = "GPIO0"},
+	{.irq_nr = 48,	.src = "MPU_WAKE"},
+	{.irq_nr = 49,	.src = "WDT0"},
+	{.irq_nr = 50,	.src = "WDT1"},
+	{.irq_nr = 51,	.src = "ADC_TSC"},
+	{.irq_nr = 0,	.src = "Unknown"},
+};
+
+static void am33xx_txev_eoi(struct wkup_m3_ipc *m3_ipc)
+{
+	writel(AM33XX_M3_TXEV_ACK,
+	       m3_ipc->ipc_mem_base + AM33XX_CONTROL_M3_TXEV_EOI);
+}
+
+static void am33xx_txev_enable(struct wkup_m3_ipc *m3_ipc)
+{
+	writel(AM33XX_M3_TXEV_ENABLE,
+	       m3_ipc->ipc_mem_base + AM33XX_CONTROL_M3_TXEV_EOI);
+}
+
+static void wkup_m3_ctrl_ipc_write(struct wkup_m3_ipc *m3_ipc,
+				   u32 val, int ipc_reg_num)
+{
+	if (WARN(ipc_reg_num < 0 || ipc_reg_num > AM33XX_CTRL_IPC_REG_COUNT,
+		 "ipc register operation out of range"))
+		return;
+
+	writel(val, m3_ipc->ipc_mem_base +
+	       AM33XX_CTRL_IPC_REG_OFFSET(ipc_reg_num));
+}
+
+static unsigned int wkup_m3_ctrl_ipc_read(struct wkup_m3_ipc *m3_ipc,
+					  int ipc_reg_num)
+{
+	if (WARN(ipc_reg_num < 0 || ipc_reg_num > AM33XX_CTRL_IPC_REG_COUNT,
+		 "ipc register operation out of range"))
+		return 0;
+
+	return readl(m3_ipc->ipc_mem_base +
+		     AM33XX_CTRL_IPC_REG_OFFSET(ipc_reg_num));
+}
+
+static int wkup_m3_fw_version_read(struct wkup_m3_ipc *m3_ipc)
+{
+	int val;
+
+	val = wkup_m3_ctrl_ipc_read(m3_ipc, 2);
+
+	return val & M3_FW_VERSION_MASK;
+}
+
+static irqreturn_t wkup_m3_txev_handler(int irq, void *ipc_data)
+{
+	struct wkup_m3_ipc *m3_ipc = ipc_data;
+	struct device *dev = m3_ipc->dev;
+	int ver = 0;
+
+	am33xx_txev_eoi(m3_ipc);
+
+	switch (m3_ipc->state) {
+	case M3_STATE_RESET:
+		ver = wkup_m3_fw_version_read(m3_ipc);
+
+		if (ver == M3_VERSION_UNKNOWN ||
+		    ver < M3_BASELINE_VERSION) {
+			dev_warn(dev, "CM3 Firmware Version %x not supported\n",
+				 ver);
+		} else {
+			dev_info(dev, "CM3 Firmware Version = 0x%x\n", ver);
+		}
+
+		m3_ipc->state = M3_STATE_INITED;
+		complete(&m3_ipc->sync_complete);
+		break;
+	case M3_STATE_MSG_FOR_RESET:
+		m3_ipc->state = M3_STATE_INITED;
+		complete(&m3_ipc->sync_complete);
+		break;
+	case M3_STATE_MSG_FOR_LP:
+		complete(&m3_ipc->sync_complete);
+		break;
+	case M3_STATE_UNKNOWN:
+		dev_warn(dev, "Unknown CM3 State\n");
+	}
+
+	am33xx_txev_enable(m3_ipc);
+
+	return IRQ_HANDLED;
+}
+
+static int wkup_m3_ping(struct wkup_m3_ipc *m3_ipc)
+{
+	struct device *dev = m3_ipc->dev;
+	mbox_msg_t dummy_msg = 0;
+	int ret;
+
+	if (!m3_ipc->mbox) {
+		dev_err(dev,
+			"No IPC channel to communicate with wkup_m3!\n");
+		return -EIO;
+	}
+
+	/*
+	 * Write a dummy message to the mailbox in order to trigger the RX
+	 * interrupt to alert the M3 that data is available in the IPC
+	 * registers. We must enable the IRQ here and disable it after in
+	 * the RX callback to avoid multiple interrupts being received
+	 * by the CM3.
+	 */
+	ret = mbox_send_message(m3_ipc->mbox, &dummy_msg);
+	if (ret < 0) {
+		dev_err(dev, "%s: mbox_send_message() failed: %d\n",
+			__func__, ret);
+		return ret;
+	}
+
+	ret = wait_for_completion_timeout(&m3_ipc->sync_complete,
+					  msecs_to_jiffies(500));
+	if (!ret) {
+		dev_err(dev, "MPU<->CM3 sync failure\n");
+		m3_ipc->state = M3_STATE_UNKNOWN;
+		return -EIO;
+	}
+
+	mbox_client_txdone(m3_ipc->mbox, 0);
+	return 0;
+}
+
+static int wkup_m3_ping_noirq(struct wkup_m3_ipc *m3_ipc)
+{
+	struct device *dev = m3_ipc->dev;
+	mbox_msg_t dummy_msg = 0;
+	int ret;
+
+	if (!m3_ipc->mbox) {
+		dev_err(dev,
+			"No IPC channel to communicate with wkup_m3!\n");
+		return -EIO;
+	}
+
+	ret = mbox_send_message(m3_ipc->mbox, &dummy_msg);
+	if (ret < 0) {
+		dev_err(dev, "%s: mbox_send_message() failed: %d\n",
+			__func__, ret);
+		return ret;
+	}
+
+	mbox_client_txdone(m3_ipc->mbox, 0);
+	return 0;
+}
+
+static int wkup_m3_is_available(struct wkup_m3_ipc *m3_ipc)
+{
+	return ((m3_ipc->state != M3_STATE_RESET) &&
+		(m3_ipc->state != M3_STATE_UNKNOWN));
+}
+
+/* Public functions */
+/**
+ * wkup_m3_set_mem_type - Pass wkup_m3 which type of memory is in use
+ * @mem_type: memory type value read directly from emif
+ *
+ * wkup_m3 must know what memory type is in use to properly suspend
+ * and resume.
+ */
+static void wkup_m3_set_mem_type(struct wkup_m3_ipc *m3_ipc, int mem_type)
+{
+	m3_ipc->mem_type = mem_type;
+}
+
+/**
+ * wkup_m3_set_resume_address - Pass wkup_m3 resume address
+ * @addr: Physical address from which resume code should execute
+ */
+static void wkup_m3_set_resume_address(struct wkup_m3_ipc *m3_ipc, void *addr)
+{
+	m3_ipc->resume_addr = (unsigned long)addr;
+}
+
+/**
+ * wkup_m3_request_pm_status - Retrieve wkup_m3 status code after suspend
+ *
+ * Returns code representing the status of a low power mode transition.
+ *	0 - Successful transition
+ *	1 - Failure to transition to low power state
+ */
+static int wkup_m3_request_pm_status(struct wkup_m3_ipc *m3_ipc)
+{
+	unsigned int i;
+	int val;
+
+	val = wkup_m3_ctrl_ipc_read(m3_ipc, 1);
+
+	i = M3_STATUS_RESP_MASK & val;
+	i >>= __ffs(M3_STATUS_RESP_MASK);
+
+	return i;
+}
+
+/**
+ * wkup_m3_prepare_low_power - Request preparation for transition to
+ *			       low power state
+ * @state: A kernel suspend state to enter, either MEM or STANDBY
+ *
+ * Returns 0 if preparation was successful, otherwise returns error code
+ */
+static int wkup_m3_prepare_low_power(struct wkup_m3_ipc *m3_ipc, int state)
+{
+	struct device *dev = m3_ipc->dev;
+	int m3_power_state;
+	int ret = 0;
+
+	if (!wkup_m3_is_available(m3_ipc))
+		return -ENODEV;
+
+	switch (state) {
+	case WKUP_M3_DEEPSLEEP:
+		m3_power_state = IPC_CMD_DS0;
+		break;
+	case WKUP_M3_STANDBY:
+		m3_power_state = IPC_CMD_STANDBY;
+		break;
+	case WKUP_M3_IDLE:
+		m3_power_state = IPC_CMD_IDLE;
+		break;
+	default:
+		return 1;
+	}
+
+	/* Program each required IPC register then write defaults to others */
+	wkup_m3_ctrl_ipc_write(m3_ipc, m3_ipc->resume_addr, 0);
+	wkup_m3_ctrl_ipc_write(m3_ipc, m3_power_state, 1);
+	wkup_m3_ctrl_ipc_write(m3_ipc, m3_ipc->mem_type, 4);
+
+	wkup_m3_ctrl_ipc_write(m3_ipc, DS_IPC_DEFAULT, 2);
+	wkup_m3_ctrl_ipc_write(m3_ipc, DS_IPC_DEFAULT, 3);
+	wkup_m3_ctrl_ipc_write(m3_ipc, DS_IPC_DEFAULT, 5);
+	wkup_m3_ctrl_ipc_write(m3_ipc, DS_IPC_DEFAULT, 6);
+	wkup_m3_ctrl_ipc_write(m3_ipc, DS_IPC_DEFAULT, 7);
+
+	m3_ipc->state = M3_STATE_MSG_FOR_LP;
+
+	if (state == WKUP_M3_IDLE)
+		ret = wkup_m3_ping_noirq(m3_ipc);
+	else
+		ret = wkup_m3_ping(m3_ipc);
+
+	if (ret) {
+		dev_err(dev, "Unable to ping CM3\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+/**
+ * wkup_m3_finish_low_power - Return m3 to reset state
+ *
+ * Returns 0 if reset was successful, otherwise returns error code
+ */
+static int wkup_m3_finish_low_power(struct wkup_m3_ipc *m3_ipc)
+{
+	struct device *dev = m3_ipc->dev;
+	int ret = 0;
+
+	if (!wkup_m3_is_available(m3_ipc))
+		return -ENODEV;
+
+	wkup_m3_ctrl_ipc_write(m3_ipc, IPC_CMD_RESET, 1);
+	wkup_m3_ctrl_ipc_write(m3_ipc, DS_IPC_DEFAULT, 2);
+
+	m3_ipc->state = M3_STATE_MSG_FOR_RESET;
+
+	ret = wkup_m3_ping(m3_ipc);
+	if (ret) {
+		dev_err(dev, "Unable to ping CM3\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+/**
+ * wkup_m3_request_wake_src - Get the wakeup source info passed from wkup_m3
+ * @m3_ipc: Pointer to wkup_m3_ipc context
+ */
+static const char *wkup_m3_request_wake_src(struct wkup_m3_ipc *m3_ipc)
+{
+	unsigned int wakeup_src_idx;
+	int j, val;
+
+	val = wkup_m3_ctrl_ipc_read(m3_ipc, 6);
+
+	wakeup_src_idx = val & M3_WAKE_SRC_MASK;
+
+	for (j = 0; j < ARRAY_SIZE(wakeups) - 1; j++) {
+		if (wakeups[j].irq_nr == wakeup_src_idx)
+			return wakeups[j].src;
+	}
+	return wakeups[j].src;
+}
+
+/**
+ * wkup_m3_set_rtc_only - Set the rtc_only flag
+ * @wkup_m3_wakeup: struct wkup_m3_wakeup_src * gets assigned the
+ *                  wakeup src value
+ */
+static void wkup_m3_set_rtc_only(struct wkup_m3_ipc *m3_ipc)
+{
+	if (m3_ipc_state)
+		m3_ipc_state->is_rtc_only = true;
+}
+
+static struct wkup_m3_ipc_ops ipc_ops = {
+	.set_mem_type = wkup_m3_set_mem_type,
+	.set_resume_address = wkup_m3_set_resume_address,
+	.prepare_low_power = wkup_m3_prepare_low_power,
+	.finish_low_power = wkup_m3_finish_low_power,
+	.request_pm_status = wkup_m3_request_pm_status,
+	.request_wake_src = wkup_m3_request_wake_src,
+	.set_rtc_only = wkup_m3_set_rtc_only,
+};
+
+/**
+ * wkup_m3_ipc_get - Return handle to wkup_m3_ipc
+ *
+ * Returns NULL if the wkup_m3 is not yet available, otherwise returns
+ * pointer to wkup_m3_ipc struct.
+ */
+struct wkup_m3_ipc *wkup_m3_ipc_get(void)
+{
+	if (m3_ipc_state)
+		get_device(m3_ipc_state->dev);
+	else
+		return NULL;
+
+	return m3_ipc_state;
+}
+EXPORT_SYMBOL_GPL(wkup_m3_ipc_get);
+
+/**
+ * wkup_m3_ipc_put - Free handle to wkup_m3_ipc returned from wkup_m3_ipc_get
+ * @m3_ipc: A pointer to wkup_m3_ipc struct returned by wkup_m3_ipc_get
+ */
+void wkup_m3_ipc_put(struct wkup_m3_ipc *m3_ipc)
+{
+	if (m3_ipc_state)
+		put_device(m3_ipc_state->dev);
+}
+EXPORT_SYMBOL_GPL(wkup_m3_ipc_put);
+
+static void wkup_m3_rproc_boot_thread(struct wkup_m3_ipc *m3_ipc)
+{
+	struct device *dev = m3_ipc->dev;
+	int ret;
+
+	init_completion(&m3_ipc->sync_complete);
+
+	ret = rproc_boot(m3_ipc->rproc);
+	if (ret)
+		dev_err(dev, "rproc_boot failed\n");
+	else
+		m3_ipc_state = m3_ipc;
+
+	do_exit(0);
+}
+
+static int wkup_m3_ipc_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	int irq, ret;
+	phandle rproc_phandle;
+	struct rproc *m3_rproc;
+	struct resource *res;
+	struct task_struct *task;
+	struct wkup_m3_ipc *m3_ipc;
+
+	m3_ipc = devm_kzalloc(dev, sizeof(*m3_ipc), GFP_KERNEL);
+	if (!m3_ipc)
+		return -ENOMEM;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	m3_ipc->ipc_mem_base = devm_ioremap_resource(dev, res);
+	if (IS_ERR(m3_ipc->ipc_mem_base)) {
+		dev_err(dev, "could not ioremap ipc_mem\n");
+		return PTR_ERR(m3_ipc->ipc_mem_base);
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "no irq resource\n");
+		return irq;
+	}
+
+	ret = devm_request_irq(dev, irq, wkup_m3_txev_handler,
+			       0, "wkup_m3_txev", m3_ipc);
+	if (ret) {
+		dev_err(dev, "request_irq failed\n");
+		return ret;
+	}
+
+	m3_ipc->mbox_client.dev = dev;
+	m3_ipc->mbox_client.tx_done = NULL;
+	m3_ipc->mbox_client.tx_prepare = NULL;
+	m3_ipc->mbox_client.rx_callback = NULL;
+	m3_ipc->mbox_client.tx_block = false;
+	m3_ipc->mbox_client.knows_txdone = false;
+
+	m3_ipc->mbox = mbox_request_channel(&m3_ipc->mbox_client, 0);
+
+	if (IS_ERR(m3_ipc->mbox)) {
+		dev_err(dev, "IPC Request for A8->M3 Channel failed! %ld\n",
+			PTR_ERR(m3_ipc->mbox));
+		return PTR_ERR(m3_ipc->mbox);
+	}
+
+	if (of_property_read_u32(dev->of_node, "ti,rproc", &rproc_phandle)) {
+		dev_err(&pdev->dev, "could not get rproc phandle\n");
+		ret = -ENODEV;
+		goto err_free_mbox;
+	}
+
+	m3_rproc = rproc_get_by_phandle(rproc_phandle);
+	if (!m3_rproc) {
+		dev_err(&pdev->dev, "could not get rproc handle\n");
+		ret = -EPROBE_DEFER;
+		goto err_free_mbox;
+	}
+
+	m3_ipc->rproc = m3_rproc;
+	m3_ipc->dev = dev;
+	m3_ipc->state = M3_STATE_RESET;
+
+	m3_ipc->ops = &ipc_ops;
+
+	/*
+	 * Wait for firmware loading completion in a thread so we
+	 * can boot the wkup_m3 as soon as it's ready without holding
+	 * up kernel boot
+	 */
+	task = kthread_run((void *)wkup_m3_rproc_boot_thread, m3_ipc,
+			   "wkup_m3_rproc_loader");
+
+	if (IS_ERR(task)) {
+		dev_err(dev, "can't create rproc_boot thread\n");
+		ret = PTR_ERR(task);
+		goto err_put_rproc;
+	}
+
+	return 0;
+
+err_put_rproc:
+	rproc_put(m3_rproc);
+err_free_mbox:
+	mbox_free_channel(m3_ipc->mbox);
+	return ret;
+}
+
+static int wkup_m3_ipc_remove(struct platform_device *pdev)
+{
+	mbox_free_channel(m3_ipc_state->mbox);
+
+	rproc_shutdown(m3_ipc_state->rproc);
+	rproc_put(m3_ipc_state->rproc);
+
+	m3_ipc_state = NULL;
+
+	return 0;
+}
+
+static int __maybe_unused wkup_m3_ipc_suspend(struct device *dev)
+{
+	/*
+	 * Nothing needs to be done on suspend even with rtc_only flag set
+	 */
+	return 0;
+}
+
+static int __maybe_unused wkup_m3_ipc_resume(struct device *dev)
+{
+	if (m3_ipc_state->is_rtc_only) {
+		rproc_shutdown(m3_ipc_state->rproc);
+		rproc_boot(m3_ipc_state->rproc);
+	}
+
+	m3_ipc_state->is_rtc_only = false;
+
+	return 0;
+}
+
+static const struct dev_pm_ops wkup_m3_ipc_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(wkup_m3_ipc_suspend, wkup_m3_ipc_resume)
+};
+
+static const struct of_device_id wkup_m3_ipc_of_match[] = {
+	{ .compatible = "ti,am3352-wkup-m3-ipc", },
+	{ .compatible = "ti,am4372-wkup-m3-ipc", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, wkup_m3_ipc_of_match);
+
+static struct platform_driver wkup_m3_ipc_driver = {
+	.probe = wkup_m3_ipc_probe,
+	.remove = wkup_m3_ipc_remove,
+	.driver = {
+		.name = "wkup_m3_ipc",
+		.of_match_table = wkup_m3_ipc_of_match,
+		.pm = &wkup_m3_ipc_pm_ops,
+	},
+};
+
+module_platform_driver(wkup_m3_ipc_driver);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("wkup m3 remote processor ipc driver");
+MODULE_AUTHOR("Dave Gerlach <d-gerlach@ti.com>");