soc: ti: add Keystone Navigator DMA support

The Keystone Navigator DMA driver sets up the dma channels and flows for
the QMSS(Queue Manager SubSystem) who triggers the actual data movements
across clients using destination queues. Every client modules like
NETCP(Network Coprocessor), SRIO(Serial Rapid IO) and CRYPTO
Engines has its own instance of packet dma hardware. QMSS has also
an internal packet DMA module which is used as an infrastructure
DMA with zero copy.

Initially this driver was proposed as DMA engine driver but since the
hardware is not typical DMA engine and hence doesn't comply with typical
DMA engine driver needs, that approach was naked. Link to that
discussion -
	https://lkml.org/lkml/2014/3/18/340

As aligned, now we pair the Navigator DMA with its companion Navigator
QMSS subsystem driver.

Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Kumar Gala <galak@codeaurora.org>
Cc: Olof Johansson <olof@lixom.net>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Grant Likely <grant.likely@linaro.org>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Sandeep Nair <sandeep_n@ti.com>
Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>

3 files changed, 826 insertions, 0 deletions

diff --git a/drivers/soc/ti/Kconfig b/drivers/soc/ti/Kconfig
index f73896f762e8..7266b2165183 100644
--- a/drivers/soc/ti/Kconfig
+++ b/drivers/soc/ti/Kconfig
@@ -18,4 +18,14 @@ config KEYSTONE_NAVIGATOR_QMSS
 
 	  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.
+
 endif # SOC_TI
diff --git a/drivers/soc/ti/Makefile b/drivers/soc/ti/Makefile
index bf85cacd5b85..6bed611e1934 100644
--- a/drivers/soc/ti/Makefile
+++ b/drivers/soc/ti/Makefile
@@ -2,3 +2,4 @@
 # TI Keystone SOC drivers
 #
 obj-$(CONFIG_KEYSTONE_NAVIGATOR_QMSS)	+= knav_qmss_queue.o knav_qmss_acc.o
+obj-$(CONFIG_KEYSTONE_NAVIGATOR_DMA)	+= knav_dma.o
diff --git a/drivers/soc/ti/knav_dma.c b/drivers/soc/ti/knav_dma.c
new file mode 100644
index 000000000000..17264275f32b
--- /dev/null
+++ b/drivers/soc/ti/knav_dma.c
@@ -0,0 +1,815 @@
+/*
+ * 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 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' propery\n");
+		return -ENODEV;
+	}
+
+	if (of_parse_phandle_with_fixed_args(np, "ti,navigator-dmas",
+					1, index, &args)) {
+		dev_err(kdev->dev, "Missing the pahndle 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 NULL.
+ */
+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) {
+		dev_err(kdev->dev, "unable to enable pktdma, err %d\n", ret);
+		return ret;
+	}
+
+	/* 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");
+		return -ENODEV;
+	}
+
+	debugfs_create_file("knav_dma", S_IFREG | S_IRUGO, NULL, NULL,
+			    &knav_dma_debug_ops);
+
+	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",
+		.owner		= THIS_MODULE,
+		.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>");