Merge branch 'topic/xilinx' into for-linus

4 files changed, 1596 insertions, 59 deletions

diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index f6c46d0..739f797 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -533,19 +533,31 @@ config XGENE_DMA
 	help
 	  Enable support for the APM X-Gene SoC DMA engine.
 
-config XILINX_VDMA
-	tristate "Xilinx AXI VDMA Engine"
+config XILINX_DMA
+	tristate "Xilinx AXI DMAS Engine"
 	depends on (ARCH_ZYNQ || MICROBLAZE || ARM64)
 	select DMA_ENGINE
 	help
 	  Enable support for Xilinx AXI VDMA Soft IP.
 
-	  This engine provides high-bandwidth direct memory access
+	  AXI VDMA engine provides high-bandwidth direct memory access
 	  between memory and AXI4-Stream video type target
 	  peripherals including peripherals which support AXI4-
 	  Stream Video Protocol.  It has two stream interfaces/
 	  channels, Memory Mapped to Stream (MM2S) and Stream to
 	  Memory Mapped (S2MM) for the data transfers.
+	  AXI CDMA engine provides high-bandwidth direct memory access
+	  between a memory-mapped source address and a memory-mapped
+	  destination address.
+	  AXI DMA engine provides high-bandwidth one dimensional direct
+	  memory access between memory and AXI4-Stream target peripherals.
+
+config XILINX_ZYNQMP_DMA
+	tristate "Xilinx ZynqMP DMA Engine"
+	depends on (ARCH_ZYNQ || MICROBLAZE || ARM64)
+	select DMA_ENGINE
+	help
+	  Enable support for Xilinx ZynqMP DMA controller.
 
 config ZX_DMA
 	tristate "ZTE ZX296702 DMA support"
diff --git a/drivers/dma/xilinx/Makefile b/drivers/dma/xilinx/Makefile
index 3c4e9f2..9e91f8f 100644
--- a/drivers/dma/xilinx/Makefile
+++ b/drivers/dma/xilinx/Makefile
@@ -1 +1,2 @@
-obj-$(CONFIG_XILINX_VDMA) += xilinx_vdma.o
+obj-$(CONFIG_XILINX_DMA) += xilinx_dma.o
+obj-$(CONFIG_XILINX_ZYNQMP_DMA) += zynqmp_dma.o
diff --git a/drivers/dma/xilinx/xilinx_vdma.c b/drivers/dma/xilinx/xilinx_dma.c
index df91185..4e223d0 100644
--- a/drivers/dma/xilinx/xilinx_vdma.c
+++ b/drivers/dma/xilinx/xilinx_dma.c
@@ -45,6 +45,7 @@
 #include <linux/of_irq.h>
 #include <linux/slab.h>
 #include <linux/clk.h>
+#include <linux/io-64-nonatomic-lo-hi.h>
 
 #include "../dmaengine.h"
 
@@ -113,7 +114,7 @@
 #define XILINX_VDMA_REG_START_ADDRESS_64(n)	(0x000c + 8 * (n))
 
 /* HW specific definitions */
-#define XILINX_DMA_MAX_CHANS_PER_DEVICE	0x2
+#define XILINX_DMA_MAX_CHANS_PER_DEVICE	0x20
 
 #define XILINX_DMA_DMAXR_ALL_IRQ_MASK	\
 		(XILINX_DMA_DMASR_FRM_CNT_IRQ | \
@@ -157,12 +158,25 @@
 /* AXI DMA Specific Masks/Bit fields */
 #define XILINX_DMA_MAX_TRANS_LEN	GENMASK(22, 0)
 #define XILINX_DMA_CR_COALESCE_MAX	GENMASK(23, 16)
+#define XILINX_DMA_CR_CYCLIC_BD_EN_MASK	BIT(4)
 #define XILINX_DMA_CR_COALESCE_SHIFT	16
 #define XILINX_DMA_BD_SOP		BIT(27)
 #define XILINX_DMA_BD_EOP		BIT(26)
 #define XILINX_DMA_COALESCE_MAX		255
 #define XILINX_DMA_NUM_APP_WORDS	5
 
+/* Multi-Channel DMA Descriptor offsets*/
+#define XILINX_DMA_MCRX_CDESC(x)	(0x40 + (x-1) * 0x20)
+#define XILINX_DMA_MCRX_TDESC(x)	(0x48 + (x-1) * 0x20)
+
+/* Multi-Channel DMA Masks/Shifts */
+#define XILINX_DMA_BD_HSIZE_MASK	GENMASK(15, 0)
+#define XILINX_DMA_BD_STRIDE_MASK	GENMASK(15, 0)
+#define XILINX_DMA_BD_VSIZE_MASK	GENMASK(31, 19)
+#define XILINX_DMA_BD_TDEST_MASK	GENMASK(4, 0)
+#define XILINX_DMA_BD_STRIDE_SHIFT	0
+#define XILINX_DMA_BD_VSIZE_SHIFT	19
+
 /* AXI CDMA Specific Registers/Offsets */
 #define XILINX_CDMA_REG_SRCADDR		0x18
 #define XILINX_CDMA_REG_DSTADDR		0x20
@@ -194,22 +208,22 @@ struct xilinx_vdma_desc_hw {
 /**
  * struct xilinx_axidma_desc_hw - Hardware Descriptor for AXI DMA
  * @next_desc: Next Descriptor Pointer @0x00
- * @pad1: Reserved @0x04
+ * @next_desc_msb: MSB of Next Descriptor Pointer @0x04
  * @buf_addr: Buffer address @0x08
- * @pad2: Reserved @0x0C
- * @pad3: Reserved @0x10
- * @pad4: Reserved @0x14
+ * @buf_addr_msb: MSB of Buffer address @0x0C
+ * @pad1: Reserved @0x10
+ * @pad2: Reserved @0x14
  * @control: Control field @0x18
  * @status: Status field @0x1C
  * @app: APP Fields @0x20 - 0x30
  */
 struct xilinx_axidma_desc_hw {
 	u32 next_desc;
-	u32 pad1;
+	u32 next_desc_msb;
 	u32 buf_addr;
-	u32 pad2;
-	u32 pad3;
-	u32 pad4;
+	u32 buf_addr_msb;
+	u32 mcdma_control;
+	u32 vsize_stride;
 	u32 control;
 	u32 status;
 	u32 app[XILINX_DMA_NUM_APP_WORDS];
@@ -218,21 +232,21 @@ struct xilinx_axidma_desc_hw {
 /**
  * struct xilinx_cdma_desc_hw - Hardware Descriptor
  * @next_desc: Next Descriptor Pointer @0x00
- * @pad1: Reserved @0x04
+ * @next_descmsb: Next Descriptor Pointer MSB @0x04
  * @src_addr: Source address @0x08
- * @pad2: Reserved @0x0C
+ * @src_addrmsb: Source address MSB @0x0C
  * @dest_addr: Destination address @0x10
- * @pad3: Reserved @0x14
+ * @dest_addrmsb: Destination address MSB @0x14
  * @control: Control field @0x18
  * @status: Status field @0x1C
  */
 struct xilinx_cdma_desc_hw {
 	u32 next_desc;
-	u32 pad1;
+	u32 next_desc_msb;
 	u32 src_addr;
-	u32 pad2;
+	u32 src_addr_msb;
 	u32 dest_addr;
-	u32 pad3;
+	u32 dest_addr_msb;
 	u32 control;
 	u32 status;
 } __aligned(64);
@@ -278,11 +292,13 @@ struct xilinx_cdma_tx_segment {
  * @async_tx: Async transaction descriptor
  * @segments: TX segments list
  * @node: Node in the channel descriptors list
+ * @cyclic: Check for cyclic transfers.
  */
 struct xilinx_dma_tx_descriptor {
 	struct dma_async_tx_descriptor async_tx;
 	struct list_head segments;
 	struct list_head node;
+	bool cyclic;
 };
 
 /**
@@ -302,6 +318,7 @@ struct xilinx_dma_tx_descriptor {
  * @direction: Transfer direction
  * @num_frms: Number of frames
  * @has_sg: Support scatter transfers
+ * @cyclic: Check for cyclic transfers.
  * @genlock: Support genlock mode
  * @err: Channel has errors
  * @tasklet: Cleanup work after irq
@@ -312,6 +329,7 @@ struct xilinx_dma_tx_descriptor {
  * @desc_submitcount: Descriptor h/w submitted count
  * @residue: Residue for AXI DMA
  * @seg_v: Statically allocated segments base
+ * @cyclic_seg_v: Statically allocated segment base for cyclic transfers
  * @start_transfer: Differentiate b/w DMA IP's transfer
  */
 struct xilinx_dma_chan {
@@ -330,6 +348,7 @@ struct xilinx_dma_chan {
 	enum dma_transfer_direction direction;
 	int num_frms;
 	bool has_sg;
+	bool cyclic;
 	bool genlock;
 	bool err;
 	struct tasklet_struct tasklet;
@@ -340,7 +359,9 @@ struct xilinx_dma_chan {
 	u32 desc_submitcount;
 	u32 residue;
 	struct xilinx_axidma_tx_segment *seg_v;
+	struct xilinx_axidma_tx_segment *cyclic_seg_v;
 	void (*start_transfer)(struct xilinx_dma_chan *chan);
+	u16 tdest;
 };
 
 struct xilinx_dma_config {
@@ -357,6 +378,7 @@ struct xilinx_dma_config {
  * @common: DMA device structure
  * @chan: Driver specific DMA channel
  * @has_sg: Specifies whether Scatter-Gather is present or not
+ * @mcdma: Specifies whether Multi-Channel is present or not
  * @flush_on_fsync: Flush on frame sync
  * @ext_addr: Indicates 64 bit addressing is supported by dma device
  * @pdev: Platform device structure pointer
@@ -366,6 +388,8 @@ struct xilinx_dma_config {
  * @txs_clk: DMA mm2s stream clock
  * @rx_clk: DMA s2mm clock
  * @rxs_clk: DMA s2mm stream clock
+ * @nr_channels: Number of channels DMA device supports
+ * @chan_id: DMA channel identifier
  */
 struct xilinx_dma_device {
 	void __iomem *regs;
@@ -373,6 +397,7 @@ struct xilinx_dma_device {
 	struct dma_device common;
 	struct xilinx_dma_chan *chan[XILINX_DMA_MAX_CHANS_PER_DEVICE];
 	bool has_sg;
+	bool mcdma;
 	u32 flush_on_fsync;
 	bool ext_addr;
 	struct platform_device  *pdev;
@@ -382,6 +407,8 @@ struct xilinx_dma_device {
 	struct clk *txs_clk;
 	struct clk *rx_clk;
 	struct clk *rxs_clk;
+	u32 nr_channels;
+	u32 chan_id;
 };
 
 /* Macros */
@@ -454,6 +481,34 @@ static inline void vdma_desc_write_64(struct xilinx_dma_chan *chan, u32 reg,
 	writel(value_msb, chan->xdev->regs + chan->desc_offset + reg + 4);
 }
 
+static inline void dma_writeq(struct xilinx_dma_chan *chan, u32 reg, u64 value)
+{
+	lo_hi_writeq(value, chan->xdev->regs + chan->ctrl_offset + reg);
+}
+
+static inline void xilinx_write(struct xilinx_dma_chan *chan, u32 reg,
+				dma_addr_t addr)
+{
+	if (chan->ext_addr)
+		dma_writeq(chan, reg, addr);
+	else
+		dma_ctrl_write(chan, reg, addr);
+}
+
+static inline void xilinx_axidma_buf(struct xilinx_dma_chan *chan,
+				     struct xilinx_axidma_desc_hw *hw,
+				     dma_addr_t buf_addr, size_t sg_used,
+				     size_t period_len)
+{
+	if (chan->ext_addr) {
+		hw->buf_addr = lower_32_bits(buf_addr + sg_used + period_len);
+		hw->buf_addr_msb = upper_32_bits(buf_addr + sg_used +
+						 period_len);
+	} else {
+		hw->buf_addr = buf_addr + sg_used + period_len;
+	}
+}
+
 /* -----------------------------------------------------------------------------
  * Descriptors and segments alloc and free
  */
@@ -491,11 +546,10 @@ xilinx_cdma_alloc_tx_segment(struct xilinx_dma_chan *chan)
 	struct xilinx_cdma_tx_segment *segment;
 	dma_addr_t phys;
 
-	segment = dma_pool_alloc(chan->desc_pool, GFP_ATOMIC, &phys);
+	segment = dma_pool_zalloc(chan->desc_pool, GFP_ATOMIC, &phys);
 	if (!segment)
 		return NULL;
 
-	memset(segment, 0, sizeof(*segment));
 	segment->phys = phys;
 
 	return segment;
@@ -513,11 +567,10 @@ xilinx_axidma_alloc_tx_segment(struct xilinx_dma_chan *chan)
 	struct xilinx_axidma_tx_segment *segment;
 	dma_addr_t phys;
 
-	segment = dma_pool_alloc(chan->desc_pool, GFP_ATOMIC, &phys);
+	segment = dma_pool_zalloc(chan->desc_pool, GFP_ATOMIC, &phys);
 	if (!segment)
 		return NULL;
 
-	memset(segment, 0, sizeof(*segment));
 	segment->phys = phys;
 
 	return segment;
@@ -660,13 +713,37 @@ static void xilinx_dma_free_chan_resources(struct dma_chan *dchan)
 	dev_dbg(chan->dev, "Free all channel resources.\n");
 
 	xilinx_dma_free_descriptors(chan);
-	if (chan->xdev->dma_config->dmatype == XDMA_TYPE_AXIDMA)
+	if (chan->xdev->dma_config->dmatype == XDMA_TYPE_AXIDMA) {
+		xilinx_dma_free_tx_segment(chan, chan->cyclic_seg_v);
 		xilinx_dma_free_tx_segment(chan, chan->seg_v);
+	}
 	dma_pool_destroy(chan->desc_pool);
 	chan->desc_pool = NULL;
 }
 
 /**
+ * xilinx_dma_chan_handle_cyclic - Cyclic dma callback
+ * @chan: Driver specific dma channel
+ * @desc: dma transaction descriptor
+ * @flags: flags for spin lock
+ */
+static void xilinx_dma_chan_handle_cyclic(struct xilinx_dma_chan *chan,
+					  struct xilinx_dma_tx_descriptor *desc,
+					  unsigned long *flags)
+{
+	dma_async_tx_callback callback;
+	void *callback_param;
+
+	callback = desc->async_tx.callback;
+	callback_param = desc->async_tx.callback_param;
+	if (callback) {
+		spin_unlock_irqrestore(&chan->lock, *flags);
+		callback(callback_param);
+		spin_lock_irqsave(&chan->lock, *flags);
+	}
+}
+
+/**
  * xilinx_dma_chan_desc_cleanup - Clean channel descriptors
  * @chan: Driver specific DMA channel
  */
@@ -681,6 +758,11 @@ static void xilinx_dma_chan_desc_cleanup(struct xilinx_dma_chan *chan)
 		dma_async_tx_callback callback;
 		void *callback_param;
 
+		if (desc->cyclic) {
+			xilinx_dma_chan_handle_cyclic(chan, desc, &flags);
+			break;
+		}
+
 		/* Remove from the list of running transactions */
 		list_del(&desc->node);
 
@@ -757,7 +839,7 @@ static int xilinx_dma_alloc_chan_resources(struct dma_chan *dchan)
 		return -ENOMEM;
 	}
 
-	if (chan->xdev->dma_config->dmatype == XDMA_TYPE_AXIDMA)
+	if (chan->xdev->dma_config->dmatype == XDMA_TYPE_AXIDMA) {
 		/*
 		 * For AXI DMA case after submitting a pending_list, keep
 		 * an extra segment allocated so that the "next descriptor"
@@ -768,6 +850,15 @@ static int xilinx_dma_alloc_chan_resources(struct dma_chan *dchan)
 		 */
 		chan->seg_v = xilinx_axidma_alloc_tx_segment(chan);
 
+		/*
+		 * For cyclic DMA mode we need to program the tail Descriptor
+		 * register with a value which is not a part of the BD chain
+		 * so allocating a desc segment during channel allocation for
+		 * programming tail descriptor.
+		 */
+		chan->cyclic_seg_v = xilinx_axidma_alloc_tx_segment(chan);
+	}
+
 	dma_cookie_init(dchan);
 
 	if (chan->xdev->dma_config->dmatype == XDMA_TYPE_AXIDMA) {
@@ -1065,12 +1156,12 @@ static void xilinx_cdma_start_transfer(struct xilinx_dma_chan *chan)
 	}
 
 	if (chan->has_sg) {
-		dma_ctrl_write(chan, XILINX_DMA_REG_CURDESC,
-			   head_desc->async_tx.phys);
+		xilinx_write(chan, XILINX_DMA_REG_CURDESC,
+			     head_desc->async_tx.phys);
 
 		/* Update tail ptr register which will start the transfer */
-		dma_ctrl_write(chan, XILINX_DMA_REG_TAILDESC,
-			       tail_segment->phys);
+		xilinx_write(chan, XILINX_DMA_REG_TAILDESC,
+			     tail_segment->phys);
 	} else {
 		/* In simple mode */
 		struct xilinx_cdma_tx_segment *segment;
@@ -1082,8 +1173,8 @@ static void xilinx_cdma_start_transfer(struct xilinx_dma_chan *chan)
 
 		hw = &segment->hw;
 
-		dma_ctrl_write(chan, XILINX_CDMA_REG_SRCADDR, hw->src_addr);
-		dma_ctrl_write(chan, XILINX_CDMA_REG_DSTADDR, hw->dest_addr);
+		xilinx_write(chan, XILINX_CDMA_REG_SRCADDR, hw->src_addr);
+		xilinx_write(chan, XILINX_CDMA_REG_DSTADDR, hw->dest_addr);
 
 		/* Start the transfer */
 		dma_ctrl_write(chan, XILINX_DMA_REG_BTT,
@@ -1124,18 +1215,20 @@ static void xilinx_dma_start_transfer(struct xilinx_dma_chan *chan)
 	tail_segment = list_last_entry(&tail_desc->segments,
 				       struct xilinx_axidma_tx_segment, node);
 
-	old_head = list_first_entry(&head_desc->segments,
-				struct xilinx_axidma_tx_segment, node);
-	new_head = chan->seg_v;
-	/* Copy Buffer Descriptor fields. */
-	new_head->hw = old_head->hw;
+	if (chan->has_sg && !chan->xdev->mcdma) {
+		old_head = list_first_entry(&head_desc->segments,
+					struct xilinx_axidma_tx_segment, node);
+		new_head = chan->seg_v;
+		/* Copy Buffer Descriptor fields. */
+		new_head->hw = old_head->hw;
 
-	/* Swap and save new reserve */
-	list_replace_init(&old_head->node, &new_head->node);
-	chan->seg_v = old_head;
+		/* Swap and save new reserve */
+		list_replace_init(&old_head->node, &new_head->node);
+		chan->seg_v = old_head;
 
-	tail_segment->hw.next_desc = chan->seg_v->phys;
-	head_desc->async_tx.phys = new_head->phys;
+		tail_segment->hw.next_desc = chan->seg_v->phys;
+		head_desc->async_tx.phys = new_head->phys;
+	}
 
 	reg = dma_ctrl_read(chan, XILINX_DMA_REG_DMACR);
 
@@ -1146,9 +1239,25 @@ static void xilinx_dma_start_transfer(struct xilinx_dma_chan *chan)
 		dma_ctrl_write(chan, XILINX_DMA_REG_DMACR, reg);
 	}
 
-	if (chan->has_sg)
-		dma_ctrl_write(chan, XILINX_DMA_REG_CURDESC,
-			       head_desc->async_tx.phys);
+	if (chan->has_sg && !chan->xdev->mcdma)
+		xilinx_write(chan, XILINX_DMA_REG_CURDESC,
+			     head_desc->async_tx.phys);
+
+	if (chan->has_sg && chan->xdev->mcdma) {
+		if (chan->direction == DMA_MEM_TO_DEV) {
+			dma_ctrl_write(chan, XILINX_DMA_REG_CURDESC,
+				       head_desc->async_tx.phys);
+		} else {
+			if (!chan->tdest) {
+				dma_ctrl_write(chan, XILINX_DMA_REG_CURDESC,
+				       head_desc->async_tx.phys);
+			} else {
+				dma_ctrl_write(chan,
+					XILINX_DMA_MCRX_CDESC(chan->tdest),
+				       head_desc->async_tx.phys);
+			}
+		}
+	}
 
 	xilinx_dma_start(chan);
 
@@ -1156,9 +1265,27 @@ static void xilinx_dma_start_transfer(struct xilinx_dma_chan *chan)
 		return;
 
 	/* Start the transfer */
-	if (chan->has_sg) {
-		dma_ctrl_write(chan, XILINX_DMA_REG_TAILDESC,
+	if (chan->has_sg && !chan->xdev->mcdma) {
+		if (chan->cyclic)
+			xilinx_write(chan, XILINX_DMA_REG_TAILDESC,
+				     chan->cyclic_seg_v->phys);
+		else
+			xilinx_write(chan, XILINX_DMA_REG_TAILDESC,
+				     tail_segment->phys);
+	} else if (chan->has_sg && chan->xdev->mcdma) {
+		if (chan->direction == DMA_MEM_TO_DEV) {
+			dma_ctrl_write(chan, XILINX_DMA_REG_TAILDESC,
 			       tail_segment->phys);
+		} else {
+			if (!chan->tdest) {
+				dma_ctrl_write(chan, XILINX_DMA_REG_TAILDESC,
+					       tail_segment->phys);
+			} else {
+				dma_ctrl_write(chan,
+					XILINX_DMA_MCRX_TDESC(chan->tdest),
+					tail_segment->phys);
+			}
+		}
 	} else {
 		struct xilinx_axidma_tx_segment *segment;
 		struct xilinx_axidma_desc_hw *hw;
@@ -1168,7 +1295,7 @@ static void xilinx_dma_start_transfer(struct xilinx_dma_chan *chan)
 					   node);
 		hw = &segment->hw;
 
-		dma_ctrl_write(chan, XILINX_DMA_REG_SRCDSTADDR, hw->buf_addr);
+		xilinx_write(chan, XILINX_DMA_REG_SRCDSTADDR, hw->buf_addr);
 
 		/* Start the transfer */
 		dma_ctrl_write(chan, XILINX_DMA_REG_BTT,
@@ -1209,7 +1336,8 @@ static void xilinx_dma_complete_descriptor(struct xilinx_dma_chan *chan)
 
 	list_for_each_entry_safe(desc, next, &chan->active_list, node) {
 		list_del(&desc->node);
-		dma_cookie_complete(&desc->async_tx);
+		if (!desc->cyclic)
+			dma_cookie_complete(&desc->async_tx);
 		list_add_tail(&desc->node, &chan->done_list);
 	}
 }
@@ -1397,6 +1525,11 @@ static dma_cookie_t xilinx_dma_tx_submit(struct dma_async_tx_descriptor *tx)
 	unsigned long flags;
 	int err;
 
+	if (chan->cyclic) {
+		xilinx_dma_free_tx_descriptor(chan, desc);
+		return -EBUSY;
+	}
+
 	if (chan->err) {
 		/*
 		 * If reset fails, need to hard reset the system.
@@ -1414,6 +1547,9 @@ static dma_cookie_t xilinx_dma_tx_submit(struct dma_async_tx_descriptor *tx)
 	/* Put this transaction onto the tail of the pending queue */
 	append_desc_queue(chan, desc);
 
+	if (desc->cyclic)
+		chan->cyclic = true;
+
 	spin_unlock_irqrestore(&chan->lock, flags);
 
 	return cookie;
@@ -1541,6 +1677,10 @@ xilinx_cdma_prep_memcpy(struct dma_chan *dchan, dma_addr_t dma_dst,
 	hw->control = len;
 	hw->src_addr = dma_src;
 	hw->dest_addr = dma_dst;
+	if (chan->ext_addr) {
+		hw->src_addr_msb = upper_32_bits(dma_src);
+		hw->dest_addr_msb = upper_32_bits(dma_dst);
+	}
 
 	/* Fill the previous next descriptor with current */
 	prev = list_last_entry(&desc->segments,
@@ -1623,7 +1763,8 @@ static struct dma_async_tx_descriptor *xilinx_dma_prep_slave_sg(
 			hw = &segment->hw;
 
 			/* Fill in the descriptor */
-			hw->buf_addr = sg_dma_address(sg) + sg_used;
+			xilinx_axidma_buf(chan, hw, sg_dma_address(sg),
+					  sg_used, 0);
 
 			hw->control = copy;
 
@@ -1669,12 +1810,204 @@ error:
 }
 
 /**
+ * xilinx_dma_prep_dma_cyclic - prepare descriptors for a DMA_SLAVE transaction
+ * @chan: DMA channel
+ * @sgl: scatterlist to transfer to/from
+ * @sg_len: number of entries in @scatterlist
+ * @direction: DMA direction
+ * @flags: transfer ack flags
+ */
+static struct dma_async_tx_descriptor *xilinx_dma_prep_dma_cyclic(
+	struct dma_chan *dchan, dma_addr_t buf_addr, size_t buf_len,
+	size_t period_len, enum dma_transfer_direction direction,
+	unsigned long flags)
+{
+	struct xilinx_dma_chan *chan = to_xilinx_chan(dchan);
+	struct xilinx_dma_tx_descriptor *desc;
+	struct xilinx_axidma_tx_segment *segment, *head_segment, *prev = NULL;
+	size_t copy, sg_used;
+	unsigned int num_periods;
+	int i;
+	u32 reg;
+
+	if (!period_len)
+		return NULL;
+
+	num_periods = buf_len / period_len;
+
+	if (!num_periods)
+		return NULL;
+
+	if (!is_slave_direction(direction))
+		return NULL;
+
+	/* Allocate a transaction descriptor. */
+	desc = xilinx_dma_alloc_tx_descriptor(chan);
+	if (!desc)
+		return NULL;
+
+	chan->direction = direction;
+	dma_async_tx_descriptor_init(&desc->async_tx, &chan->common);
+	desc->async_tx.tx_submit = xilinx_dma_tx_submit;
+
+	for (i = 0; i < num_periods; ++i) {
+		sg_used = 0;
+
+		while (sg_used < period_len) {
+			struct xilinx_axidma_desc_hw *hw;
+
+			/* Get a free segment */
+			segment = xilinx_axidma_alloc_tx_segment(chan);
+			if (!segment)
+				goto error;
+
+			/*
+			 * Calculate the maximum number of bytes to transfer,
+			 * making sure it is less than the hw limit
+			 */
+			copy = min_t(size_t, period_len - sg_used,
+				     XILINX_DMA_MAX_TRANS_LEN);
+			hw = &segment->hw;
+			xilinx_axidma_buf(chan, hw, buf_addr, sg_used,
+					  period_len * i);
+			hw->control = copy;
+
+			if (prev)
+				prev->hw.next_desc = segment->phys;
+
+			prev = segment;
+			sg_used += copy;
+
+			/*
+			 * Insert the segment into the descriptor segments
+			 * list.
+			 */
+			list_add_tail(&segment->node, &desc->segments);
+		}
+	}
+
+	head_segment = list_first_entry(&desc->segments,
+				   struct xilinx_axidma_tx_segment, node);
+	desc->async_tx.phys = head_segment->phys;
+
+	desc->cyclic = true;
+	reg = dma_ctrl_read(chan, XILINX_DMA_REG_DMACR);
+	reg |= XILINX_DMA_CR_CYCLIC_BD_EN_MASK;
+	dma_ctrl_write(chan, XILINX_DMA_REG_DMACR, reg);
+
+	segment = list_last_entry(&desc->segments,
+				  struct xilinx_axidma_tx_segment,
+				  node);
+	segment->hw.next_desc = (u32) head_segment->phys;
+
+	/* For the last DMA_MEM_TO_DEV transfer, set EOP */
+	if (direction == DMA_MEM_TO_DEV) {
+		head_segment->hw.control |= XILINX_DMA_BD_SOP;
+		segment->hw.control |= XILINX_DMA_BD_EOP;
+	}
+
+	return &desc->async_tx;
+
+error:
+	xilinx_dma_free_tx_descriptor(chan, desc);
+	return NULL;
+}
+
+/**
+ * xilinx_dma_prep_interleaved - prepare a descriptor for a
+ *	DMA_SLAVE transaction
+ * @dchan: DMA channel
+ * @xt: Interleaved template pointer
+ * @flags: transfer ack flags
+ *
+ * Return: Async transaction descriptor on success and NULL on failure
+ */
+static struct dma_async_tx_descriptor *
+xilinx_dma_prep_interleaved(struct dma_chan *dchan,
+				 struct dma_interleaved_template *xt,
+				 unsigned long flags)
+{
+	struct xilinx_dma_chan *chan = to_xilinx_chan(dchan);
+	struct xilinx_dma_tx_descriptor *desc;
+	struct xilinx_axidma_tx_segment *segment;
+	struct xilinx_axidma_desc_hw *hw;
+
+	if (!is_slave_direction(xt->dir))
+		return NULL;
+
+	if (!xt->numf || !xt->sgl[0].size)
+		return NULL;
+
+	if (xt->frame_size != 1)
+		return NULL;
+
+	/* Allocate a transaction descriptor. */
+	desc = xilinx_dma_alloc_tx_descriptor(chan);
+	if (!desc)
+		return NULL;
+
+	chan->direction = xt->dir;
+	dma_async_tx_descriptor_init(&desc->async_tx, &chan->common);
+	desc->async_tx.tx_submit = xilinx_dma_tx_submit;
+
+	/* Get a free segment */
+	segment = xilinx_axidma_alloc_tx_segment(chan);
+	if (!segment)
+		goto error;
+
+	hw = &segment->hw;
+
+	/* Fill in the descriptor */
+	if (xt->dir != DMA_MEM_TO_DEV)
+		hw->buf_addr = xt->dst_start;
+	else
+		hw->buf_addr = xt->src_start;
+
+	hw->mcdma_control = chan->tdest & XILINX_DMA_BD_TDEST_MASK;
+	hw->vsize_stride = (xt->numf << XILINX_DMA_BD_VSIZE_SHIFT) &
+			    XILINX_DMA_BD_VSIZE_MASK;
+	hw->vsize_stride |= (xt->sgl[0].icg + xt->sgl[0].size) &
+			    XILINX_DMA_BD_STRIDE_MASK;
+	hw->control = xt->sgl[0].size & XILINX_DMA_BD_HSIZE_MASK;
+
+	/*
+	 * Insert the segment into the descriptor segments
+	 * list.
+	 */
+	list_add_tail(&segment->node, &desc->segments);
+
+
+	segment = list_first_entry(&desc->segments,
+				   struct xilinx_axidma_tx_segment, node);
+	desc->async_tx.phys = segment->phys;
+
+	/* For the last DMA_MEM_TO_DEV transfer, set EOP */
+	if (xt->dir == DMA_MEM_TO_DEV) {
+		segment->hw.control |= XILINX_DMA_BD_SOP;
+		segment = list_last_entry(&desc->segments,
+					  struct xilinx_axidma_tx_segment,
+					  node);
+		segment->hw.control |= XILINX_DMA_BD_EOP;
+	}
+
+	return &desc->async_tx;
+
+error:
+	xilinx_dma_free_tx_descriptor(chan, desc);
+	return NULL;
+}
+
+/**
  * xilinx_dma_terminate_all - Halt the channel and free descriptors
  * @chan: Driver specific DMA Channel pointer
  */
 static int xilinx_dma_terminate_all(struct dma_chan *dchan)
 {
 	struct xilinx_dma_chan *chan = to_xilinx_chan(dchan);
+	u32 reg;
+
+	if (chan->cyclic)
+		xilinx_dma_chan_reset(chan);
 
 	/* Halt the DMA engine */
 	xilinx_dma_halt(chan);
@@ -1682,6 +2015,13 @@ static int xilinx_dma_terminate_all(struct dma_chan *dchan)
 	/* Remove and free all of the descriptors in the lists */
 	xilinx_dma_free_descriptors(chan);
 
+	if (chan->cyclic) {
+		reg = dma_ctrl_read(chan, XILINX_DMA_REG_DMACR);
+		reg &= ~XILINX_DMA_CR_CYCLIC_BD_EN_MASK;
+		dma_ctrl_write(chan, XILINX_DMA_REG_DMACR, reg);
+		chan->cyclic = false;
+	}
+
 	return 0;
 }
 
@@ -1972,7 +2312,7 @@ static void xdma_disable_allclks(struct xilinx_dma_device *xdev)
  * Return: '0' on success and failure value on error
  */
 static int xilinx_dma_chan_probe(struct xilinx_dma_device *xdev,
-				  struct device_node *node)
+				  struct device_node *node, int chan_id)
 {
 	struct xilinx_dma_chan *chan;
 	bool has_dre = false;
@@ -2014,9 +2354,12 @@ static int xilinx_dma_chan_probe(struct xilinx_dma_device *xdev,
 	if (!has_dre)
 		xdev->common.copy_align = fls(width - 1);
 
-	if (of_device_is_compatible(node, "xlnx,axi-vdma-mm2s-channel")) {
+	if (of_device_is_compatible(node, "xlnx,axi-vdma-mm2s-channel") ||
+	    of_device_is_compatible(node, "xlnx,axi-dma-mm2s-channel") ||
+	    of_device_is_compatible(node, "xlnx,axi-cdma-channel")) {
 		chan->direction = DMA_MEM_TO_DEV;
-		chan->id = 0;
+		chan->id = chan_id;
+		chan->tdest = chan_id;
 
 		chan->ctrl_offset = XILINX_DMA_MM2S_CTRL_OFFSET;
 		if (xdev->dma_config->dmatype == XDMA_TYPE_VDMA) {
@@ -2027,9 +2370,12 @@ static int xilinx_dma_chan_probe(struct xilinx_dma_device *xdev,
 				chan->flush_on_fsync = true;
 		}
 	} else if (of_device_is_compatible(node,
-					    "xlnx,axi-vdma-s2mm-channel")) {
+					   "xlnx,axi-vdma-s2mm-channel") ||
+		   of_device_is_compatible(node,
+					   "xlnx,axi-dma-s2mm-channel")) {
 		chan->direction = DMA_DEV_TO_MEM;
-		chan->id = 1;
+		chan->id = chan_id;
+		chan->tdest = chan_id - xdev->nr_channels;
 
 		chan->ctrl_offset = XILINX_DMA_S2MM_CTRL_OFFSET;
 		if (xdev->dma_config->dmatype == XDMA_TYPE_VDMA) {
@@ -2084,6 +2430,32 @@ static int xilinx_dma_chan_probe(struct xilinx_dma_device *xdev,
 }
 
 /**
+ * xilinx_dma_child_probe - Per child node probe
+ * It get number of dma-channels per child node from
+ * device-tree and initializes all the channels.
+ *
+ * @xdev: Driver specific device structure
+ * @node: Device node
+ *
+ * Return: 0 always.
+ */
+static int xilinx_dma_child_probe(struct xilinx_dma_device *xdev,
+				    struct device_node *node) {
+	int ret, i, nr_channels = 1;
+
+	ret = of_property_read_u32(node, "dma-channels", &nr_channels);
+	if ((ret < 0) && xdev->mcdma)
+		dev_warn(xdev->dev, "missing dma-channels property\n");
+
+	for (i = 0; i < nr_channels; i++)
+		xilinx_dma_chan_probe(xdev, node, xdev->chan_id++);
+
+	xdev->nr_channels += nr_channels;
+
+	return 0;
+}
+
+/**
  * of_dma_xilinx_xlate - Translation function
  * @dma_spec: Pointer to DMA specifier as found in the device tree
  * @ofdma: Pointer to DMA controller data
@@ -2096,7 +2468,7 @@ static struct dma_chan *of_dma_xilinx_xlate(struct of_phandle_args *dma_spec,
 	struct xilinx_dma_device *xdev = ofdma->of_dma_data;
 	int chan_id = dma_spec->args[0];
 
-	if (chan_id >= XILINX_DMA_MAX_CHANS_PER_DEVICE || !xdev->chan[chan_id])
+	if (chan_id >= xdev->nr_channels || !xdev->chan[chan_id])
 		return NULL;
 
 	return dma_get_slave_channel(&xdev->chan[chan_id]->common);
@@ -2172,6 +2544,8 @@ static int xilinx_dma_probe(struct platform_device *pdev)
 
 	/* Retrieve the DMA engine properties from the device tree */
 	xdev->has_sg = of_property_read_bool(node, "xlnx,include-sg");
+	if (xdev->dma_config->dmatype == XDMA_TYPE_AXIDMA)
+		xdev->mcdma = of_property_read_bool(node, "xlnx,mcdma");
 
 	if (xdev->dma_config->dmatype == XDMA_TYPE_VDMA) {
 		err = of_property_read_u32(node, "xlnx,num-fstores",
@@ -2218,7 +2592,12 @@ static int xilinx_dma_probe(struct platform_device *pdev)
 	xdev->common.device_tx_status = xilinx_dma_tx_status;
 	xdev->common.device_issue_pending = xilinx_dma_issue_pending;
 	if (xdev->dma_config->dmatype == XDMA_TYPE_AXIDMA) {
+		dma_cap_set(DMA_CYCLIC, xdev->common.cap_mask);
 		xdev->common.device_prep_slave_sg = xilinx_dma_prep_slave_sg;
+		xdev->common.device_prep_dma_cyclic =
+					  xilinx_dma_prep_dma_cyclic;
+		xdev->common.device_prep_interleaved_dma =
+					xilinx_dma_prep_interleaved;
 		/* Residue calculation is supported by only AXI DMA */
 		xdev->common.residue_granularity =
 					  DMA_RESIDUE_GRANULARITY_SEGMENT;
@@ -2234,13 +2613,13 @@ static int xilinx_dma_probe(struct platform_device *pdev)
 
 	/* Initialize the channels */
 	for_each_child_of_node(node, child) {
-		err = xilinx_dma_chan_probe(xdev, child);
+		err = xilinx_dma_child_probe(xdev, child);
 		if (err < 0)
 			goto disable_clks;
 	}
 
 	if (xdev->dma_config->dmatype == XDMA_TYPE_VDMA) {
-		for (i = 0; i < XILINX_DMA_MAX_CHANS_PER_DEVICE; i++)
+		for (i = 0; i < xdev->nr_channels; i++)
 			if (xdev->chan[i])
 				xdev->chan[i]->num_frms = num_frames;
 	}
@@ -2263,7 +2642,7 @@ static int xilinx_dma_probe(struct platform_device *pdev)
 disable_clks:
 	xdma_disable_allclks(xdev);
 error:
-	for (i = 0; i < XILINX_DMA_MAX_CHANS_PER_DEVICE; i++)
+	for (i = 0; i < xdev->nr_channels; i++)
 		if (xdev->chan[i])
 			xilinx_dma_chan_remove(xdev->chan[i]);
 
@@ -2285,7 +2664,7 @@ static int xilinx_dma_remove(struct platform_device *pdev)
 
 	dma_async_device_unregister(&xdev->common);
 
-	for (i = 0; i < XILINX_DMA_MAX_CHANS_PER_DEVICE; i++)
+	for (i = 0; i < xdev->nr_channels; i++)
 		if (xdev->chan[i])
 			xilinx_dma_chan_remove(xdev->chan[i]);
 
diff --git a/drivers/dma/xilinx/zynqmp_dma.c b/drivers/dma/xilinx/zynqmp_dma.c
new file mode 100644
index 0000000..f777a5b
--- /dev/null
+++ b/drivers/dma/xilinx/zynqmp_dma.c
@@ -0,0 +1,1145 @@
+/*
+ * DMA driver for Xilinx ZynqMP DMA Engine
+ *
+ * Copyright (C) 2016 Xilinx, Inc. All rights reserved.
+ *
+ * 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, either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/bitops.h>
+#include <linux/dmapool.h>
+#include <linux/dma/xilinx_dma.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/of_dma.h>
+#include <linux/of_irq.h>
+#include <linux/of_platform.h>
+#include <linux/slab.h>
+#include <linux/clk.h>
+#include <linux/io-64-nonatomic-lo-hi.h>
+
+#include "../dmaengine.h"
+
+/* Register Offsets */
+#define ZYNQMP_DMA_ISR			0x100
+#define ZYNQMP_DMA_IMR			0x104
+#define ZYNQMP_DMA_IER			0x108
+#define ZYNQMP_DMA_IDS			0x10C
+#define ZYNQMP_DMA_CTRL0		0x110
+#define ZYNQMP_DMA_CTRL1		0x114
+#define ZYNQMP_DMA_DATA_ATTR		0x120
+#define ZYNQMP_DMA_DSCR_ATTR		0x124
+#define ZYNQMP_DMA_SRC_DSCR_WRD0	0x128
+#define ZYNQMP_DMA_SRC_DSCR_WRD1	0x12C
+#define ZYNQMP_DMA_SRC_DSCR_WRD2	0x130
+#define ZYNQMP_DMA_SRC_DSCR_WRD3	0x134
+#define ZYNQMP_DMA_DST_DSCR_WRD0	0x138
+#define ZYNQMP_DMA_DST_DSCR_WRD1	0x13C
+#define ZYNQMP_DMA_DST_DSCR_WRD2	0x140
+#define ZYNQMP_DMA_DST_DSCR_WRD3	0x144
+#define ZYNQMP_DMA_SRC_START_LSB	0x158
+#define ZYNQMP_DMA_SRC_START_MSB	0x15C
+#define ZYNQMP_DMA_DST_START_LSB	0x160
+#define ZYNQMP_DMA_DST_START_MSB	0x164
+#define ZYNQMP_DMA_RATE_CTRL		0x18C
+#define ZYNQMP_DMA_IRQ_SRC_ACCT		0x190
+#define ZYNQMP_DMA_IRQ_DST_ACCT		0x194
+#define ZYNQMP_DMA_CTRL2		0x200
+
+/* Interrupt registers bit field definitions */
+#define ZYNQMP_DMA_DONE			BIT(10)
+#define ZYNQMP_DMA_AXI_WR_DATA		BIT(9)
+#define ZYNQMP_DMA_AXI_RD_DATA		BIT(8)
+#define ZYNQMP_DMA_AXI_RD_DST_DSCR	BIT(7)
+#define ZYNQMP_DMA_AXI_RD_SRC_DSCR	BIT(6)
+#define ZYNQMP_DMA_IRQ_DST_ACCT_ERR	BIT(5)
+#define ZYNQMP_DMA_IRQ_SRC_ACCT_ERR	BIT(4)
+#define ZYNQMP_DMA_BYTE_CNT_OVRFL	BIT(3)
+#define ZYNQMP_DMA_DST_DSCR_DONE	BIT(2)
+#define ZYNQMP_DMA_INV_APB		BIT(0)
+
+/* Control 0 register bit field definitions */
+#define ZYNQMP_DMA_OVR_FETCH		BIT(7)
+#define ZYNQMP_DMA_POINT_TYPE_SG	BIT(6)
+#define ZYNQMP_DMA_RATE_CTRL_EN		BIT(3)
+
+/* Control 1 register bit field definitions */
+#define ZYNQMP_DMA_SRC_ISSUE		GENMASK(4, 0)
+
+/* Data Attribute register bit field definitions */
+#define ZYNQMP_DMA_ARBURST		GENMASK(27, 26)
+#define ZYNQMP_DMA_ARCACHE		GENMASK(25, 22)
+#define ZYNQMP_DMA_ARCACHE_OFST		22
+#define ZYNQMP_DMA_ARQOS		GENMASK(21, 18)
+#define ZYNQMP_DMA_ARQOS_OFST		18
+#define ZYNQMP_DMA_ARLEN		GENMASK(17, 14)
+#define ZYNQMP_DMA_ARLEN_OFST		14
+#define ZYNQMP_DMA_AWBURST		GENMASK(13, 12)
+#define ZYNQMP_DMA_AWCACHE		GENMASK(11, 8)
+#define ZYNQMP_DMA_AWCACHE_OFST		8
+#define ZYNQMP_DMA_AWQOS		GENMASK(7, 4)
+#define ZYNQMP_DMA_AWQOS_OFST		4
+#define ZYNQMP_DMA_AWLEN		GENMASK(3, 0)
+#define ZYNQMP_DMA_AWLEN_OFST		0
+
+/* Descriptor Attribute register bit field definitions */
+#define ZYNQMP_DMA_AXCOHRNT		BIT(8)
+#define ZYNQMP_DMA_AXCACHE		GENMASK(7, 4)
+#define ZYNQMP_DMA_AXCACHE_OFST		4
+#define ZYNQMP_DMA_AXQOS		GENMASK(3, 0)
+#define ZYNQMP_DMA_AXQOS_OFST		0
+
+/* Control register 2 bit field definitions */
+#define ZYNQMP_DMA_ENABLE		BIT(0)
+
+/* Buffer Descriptor definitions */
+#define ZYNQMP_DMA_DESC_CTRL_STOP	0x10
+#define ZYNQMP_DMA_DESC_CTRL_COMP_INT	0x4
+#define ZYNQMP_DMA_DESC_CTRL_SIZE_256	0x2
+#define ZYNQMP_DMA_DESC_CTRL_COHRNT	0x1
+
+/* Interrupt Mask specific definitions */
+#define ZYNQMP_DMA_INT_ERR	(ZYNQMP_DMA_AXI_RD_DATA | \
+				ZYNQMP_DMA_AXI_WR_DATA | \
+				ZYNQMP_DMA_AXI_RD_DST_DSCR | \
+				ZYNQMP_DMA_AXI_RD_SRC_DSCR | \
+				ZYNQMP_DMA_INV_APB)
+#define ZYNQMP_DMA_INT_OVRFL	(ZYNQMP_DMA_BYTE_CNT_OVRFL | \
+				ZYNQMP_DMA_IRQ_SRC_ACCT_ERR | \
+				ZYNQMP_DMA_IRQ_DST_ACCT_ERR)
+#define ZYNQMP_DMA_INT_DONE	(ZYNQMP_DMA_DONE | ZYNQMP_DMA_DST_DSCR_DONE)
+#define ZYNQMP_DMA_INT_EN_DEFAULT_MASK	(ZYNQMP_DMA_INT_DONE | \
+					ZYNQMP_DMA_INT_ERR | \
+					ZYNQMP_DMA_INT_OVRFL | \
+					ZYNQMP_DMA_DST_DSCR_DONE)
+
+/* Max number of descriptors per channel */
+#define ZYNQMP_DMA_NUM_DESCS	32
+
+/* Max transfer size per descriptor */
+#define ZYNQMP_DMA_MAX_TRANS_LEN	0x40000000
+
+/* Reset values for data attributes */
+#define ZYNQMP_DMA_AXCACHE_VAL		0xF
+#define ZYNQMP_DMA_ARLEN_RST_VAL	0xF
+#define ZYNQMP_DMA_AWLEN_RST_VAL	0xF
+
+#define ZYNQMP_DMA_SRC_ISSUE_RST_VAL	0x1F
+
+#define ZYNQMP_DMA_IDS_DEFAULT_MASK	0xFFF
+
+/* Bus width in bits */
+#define ZYNQMP_DMA_BUS_WIDTH_64		64
+#define ZYNQMP_DMA_BUS_WIDTH_128	128
+
+#define ZYNQMP_DMA_DESC_SIZE(chan)	(chan->desc_size)
+
+#define to_chan(chan)		container_of(chan, struct zynqmp_dma_chan, \
+					     common)
+#define tx_to_desc(tx)		container_of(tx, struct zynqmp_dma_desc_sw, \
+					     async_tx)
+
+/**
+ * struct zynqmp_dma_desc_ll - Hw linked list descriptor
+ * @addr: Buffer address
+ * @size: Size of the buffer
+ * @ctrl: Control word
+ * @nxtdscraddr: Next descriptor base address
+ * @rsvd: Reserved field and for Hw internal use.
+ */
+struct zynqmp_dma_desc_ll {
+	u64 addr;
+	u32 size;
+	u32 ctrl;
+	u64 nxtdscraddr;
+	u64 rsvd;
+}; __aligned(64)
+
+/**
+ * struct zynqmp_dma_desc_sw - Per Transaction structure
+ * @src: Source address for simple mode dma
+ * @dst: Destination address for simple mode dma
+ * @len: Transfer length for simple mode dma
+ * @node: Node in the channel descriptor list
+ * @tx_list: List head for the current transfer
+ * @async_tx: Async transaction descriptor
+ * @src_v: Virtual address of the src descriptor
+ * @src_p: Physical address of the src descriptor
+ * @dst_v: Virtual address of the dst descriptor
+ * @dst_p: Physical address of the dst descriptor
+ */
+struct zynqmp_dma_desc_sw {
+	u64 src;
+	u64 dst;
+	u32 len;
+	struct list_head node;
+	struct list_head tx_list;
+	struct dma_async_tx_descriptor async_tx;
+	struct zynqmp_dma_desc_ll *src_v;
+	dma_addr_t src_p;
+	struct zynqmp_dma_desc_ll *dst_v;
+	dma_addr_t dst_p;
+};
+
+/**
+ * struct zynqmp_dma_chan - Driver specific DMA channel structure
+ * @zdev: Driver specific device structure
+ * @regs: Control registers offset
+ * @lock: Descriptor operation lock
+ * @pending_list: Descriptors waiting
+ * @free_list: Descriptors free
+ * @active_list: Descriptors active
+ * @sw_desc_pool: SW descriptor pool
+ * @done_list: Complete descriptors
+ * @common: DMA common channel
+ * @desc_pool_v: Statically allocated descriptor base
+ * @desc_pool_p: Physical allocated descriptor base
+ * @desc_free_cnt: Descriptor available count
+ * @dev: The dma device
+ * @irq: Channel IRQ
+ * @is_dmacoherent: Tells whether dma operations are coherent or not
+ * @tasklet: Cleanup work after irq
+ * @idle : Channel status;
+ * @desc_size: Size of the low level descriptor
+ * @err: Channel has errors
+ * @bus_width: Bus width
+ * @src_burst_len: Source burst length
+ * @dst_burst_len: Dest burst length
+ * @clk_main: Pointer to main clock
+ * @clk_apb: Pointer to apb clock
+ */
+struct zynqmp_dma_chan {
+	struct zynqmp_dma_device *zdev;
+	void __iomem *regs;
+	spinlock_t lock;
+	struct list_head pending_list;
+	struct list_head free_list;
+	struct list_head active_list;
+	struct zynqmp_dma_desc_sw *sw_desc_pool;
+	struct list_head done_list;
+	struct dma_chan common;
+	void *desc_pool_v;
+	dma_addr_t desc_pool_p;
+	u32 desc_free_cnt;
+	struct device *dev;
+	int irq;
+	bool is_dmacoherent;
+	struct tasklet_struct tasklet;
+	bool idle;
+	u32 desc_size;
+	bool err;
+	u32 bus_width;
+	u32 src_burst_len;
+	u32 dst_burst_len;
+	struct clk *clk_main;
+	struct clk *clk_apb;
+};
+
+/**
+ * struct zynqmp_dma_device - DMA device structure
+ * @dev: Device Structure
+ * @common: DMA device structure
+ * @chan: Driver specific DMA channel
+ */
+struct zynqmp_dma_device {
+	struct device *dev;
+	struct dma_device common;
+	struct zynqmp_dma_chan *chan;
+};
+
+static inline void zynqmp_dma_writeq(struct zynqmp_dma_chan *chan, u32 reg,
+				     u64 value)
+{
+	lo_hi_writeq(value, chan->regs + reg);
+}
+
+/**
+ * zynqmp_dma_update_desc_to_ctrlr - Updates descriptor to the controller
+ * @chan: ZynqMP DMA DMA channel pointer
+ * @desc: Transaction descriptor pointer
+ */
+static void zynqmp_dma_update_desc_to_ctrlr(struct zynqmp_dma_chan *chan,
+				      struct zynqmp_dma_desc_sw *desc)
+{
+	dma_addr_t addr;
+
+	addr = desc->src_p;
+	zynqmp_dma_writeq(chan, ZYNQMP_DMA_SRC_START_LSB, addr);
+	addr = desc->dst_p;
+	zynqmp_dma_writeq(chan, ZYNQMP_DMA_DST_START_LSB, addr);
+}
+
+/**
+ * zynqmp_dma_desc_config_eod - Mark the descriptor as end descriptor
+ * @chan: ZynqMP DMA channel pointer
+ * @desc: Hw descriptor pointer
+ */
+static void zynqmp_dma_desc_config_eod(struct zynqmp_dma_chan *chan,
+				       void *desc)
+{
+	struct zynqmp_dma_desc_ll *hw = (struct zynqmp_dma_desc_ll *)desc;
+
+	hw->ctrl |= ZYNQMP_DMA_DESC_CTRL_STOP;
+	hw++;
+	hw->ctrl |= ZYNQMP_DMA_DESC_CTRL_COMP_INT | ZYNQMP_DMA_DESC_CTRL_STOP;
+}
+
+/**
+ * zynqmp_dma_config_sg_ll_desc - Configure the linked list descriptor
+ * @chan: ZynqMP DMA channel pointer
+ * @sdesc: Hw descriptor pointer
+ * @src: Source buffer address
+ * @dst: Destination buffer address
+ * @len: Transfer length
+ * @prev: Previous hw descriptor pointer
+ */
+static void zynqmp_dma_config_sg_ll_desc(struct zynqmp_dma_chan *chan,
+				   struct zynqmp_dma_desc_ll *sdesc,
+				   dma_addr_t src, dma_addr_t dst, size_t len,
+				   struct zynqmp_dma_desc_ll *prev)
+{
+	struct zynqmp_dma_desc_ll *ddesc = sdesc + 1;
+
+	sdesc->size = ddesc->size = len;
+	sdesc->addr = src;
+	ddesc->addr = dst;
+
+	sdesc->ctrl = ddesc->ctrl = ZYNQMP_DMA_DESC_CTRL_SIZE_256;
+	if (chan->is_dmacoherent) {
+		sdesc->ctrl |= ZYNQMP_DMA_DESC_CTRL_COHRNT;
+		ddesc->ctrl |= ZYNQMP_DMA_DESC_CTRL_COHRNT;
+	}
+
+	if (prev) {
+		dma_addr_t addr = chan->desc_pool_p +
+			    ((uintptr_t)sdesc - (uintptr_t)chan->desc_pool_v);
+		ddesc = prev + 1;
+		prev->nxtdscraddr = addr;
+		ddesc->nxtdscraddr = addr + ZYNQMP_DMA_DESC_SIZE(chan);
+	}
+}
+
+/**
+ * zynqmp_dma_init - Initialize the channel
+ * @chan: ZynqMP DMA channel pointer
+ */
+static void zynqmp_dma_init(struct zynqmp_dma_chan *chan)
+{
+	u32 val;
+
+	writel(ZYNQMP_DMA_IDS_DEFAULT_MASK, chan->regs + ZYNQMP_DMA_IDS);
+	val = readl(chan->regs + ZYNQMP_DMA_ISR);
+	writel(val, chan->regs + ZYNQMP_DMA_ISR);
+
+	if (chan->is_dmacoherent) {
+		val = ZYNQMP_DMA_AXCOHRNT;
+		val = (val & ~ZYNQMP_DMA_AXCACHE) |
+			(ZYNQMP_DMA_AXCACHE_VAL << ZYNQMP_DMA_AXCACHE_OFST);
+		writel(val, chan->regs + ZYNQMP_DMA_DSCR_ATTR);
+	}
+
+	val = readl(chan->regs + ZYNQMP_DMA_DATA_ATTR);
+	if (chan->is_dmacoherent) {
+		val = (val & ~ZYNQMP_DMA_ARCACHE) |
+			(ZYNQMP_DMA_AXCACHE_VAL << ZYNQMP_DMA_ARCACHE_OFST);
+		val = (val & ~ZYNQMP_DMA_AWCACHE) |
+			(ZYNQMP_DMA_AXCACHE_VAL << ZYNQMP_DMA_AWCACHE_OFST);
+	}
+	writel(val, chan->regs + ZYNQMP_DMA_DATA_ATTR);
+
+	/* Clearing the interrupt account rgisters */
+	val = readl(chan->regs + ZYNQMP_DMA_IRQ_SRC_ACCT);
+	val = readl(chan->regs + ZYNQMP_DMA_IRQ_DST_ACCT);
+
+	chan->idle = true;
+}
+
+/**
+ * zynqmp_dma_tx_submit - Submit DMA transaction
+ * @tx: Async transaction descriptor pointer
+ *
+ * Return: cookie value
+ */
+static dma_cookie_t zynqmp_dma_tx_submit(struct dma_async_tx_descriptor *tx)
+{
+	struct zynqmp_dma_chan *chan = to_chan(tx->chan);
+	struct zynqmp_dma_desc_sw *desc, *new;
+	dma_cookie_t cookie;
+
+	new = tx_to_desc(tx);
+	spin_lock_bh(&chan->lock);
+	cookie = dma_cookie_assign(tx);
+
+	if (!list_empty(&chan->pending_list)) {
+		desc = list_last_entry(&chan->pending_list,
+				     struct zynqmp_dma_desc_sw, node);
+		if (!list_empty(&desc->tx_list))
+			desc = list_last_entry(&desc->tx_list,
+					       struct zynqmp_dma_desc_sw, node);
+		desc->src_v->nxtdscraddr = new->src_p;
+		desc->src_v->ctrl &= ~ZYNQMP_DMA_DESC_CTRL_STOP;
+		desc->dst_v->nxtdscraddr = new->dst_p;
+		desc->dst_v->ctrl &= ~ZYNQMP_DMA_DESC_CTRL_STOP;
+	}
+
+	list_add_tail(&new->node, &chan->pending_list);
+	spin_unlock_bh(&chan->lock);
+
+	return cookie;
+}
+
+/**
+ * zynqmp_dma_get_descriptor - Get the sw descriptor from the pool
+ * @chan: ZynqMP DMA channel pointer
+ *
+ * Return: The sw descriptor
+ */
+static struct zynqmp_dma_desc_sw *
+zynqmp_dma_get_descriptor(struct zynqmp_dma_chan *chan)
+{
+	struct zynqmp_dma_desc_sw *desc;
+
+	spin_lock_bh(&chan->lock);
+	desc = list_first_entry(&chan->free_list,
+				struct zynqmp_dma_desc_sw, node);
+	list_del(&desc->node);
+	spin_unlock_bh(&chan->lock);
+
+	INIT_LIST_HEAD(&desc->tx_list);
+	/* Clear the src and dst descriptor memory */
+	memset((void *)desc->src_v, 0, ZYNQMP_DMA_DESC_SIZE(chan));
+	memset((void *)desc->dst_v, 0, ZYNQMP_DMA_DESC_SIZE(chan));
+
+	return desc;
+}
+
+/**
+ * zynqmp_dma_free_descriptor - Issue pending transactions
+ * @chan: ZynqMP DMA channel pointer
+ * @sdesc: Transaction descriptor pointer
+ */
+static void zynqmp_dma_free_descriptor(struct zynqmp_dma_chan *chan,
+				 struct zynqmp_dma_desc_sw *sdesc)
+{
+	struct zynqmp_dma_desc_sw *child, *next;
+
+	chan->desc_free_cnt++;
+	list_add_tail(&sdesc->node, &chan->free_list);
+	list_for_each_entry_safe(child, next, &sdesc->tx_list, node) {
+		chan->desc_free_cnt++;
+		list_move_tail(&child->node, &chan->free_list);
+	}
+}
+
+/**
+ * zynqmp_dma_free_desc_list - Free descriptors list
+ * @chan: ZynqMP DMA channel pointer
+ * @list: List to parse and delete the descriptor
+ */
+static void zynqmp_dma_free_desc_list(struct zynqmp_dma_chan *chan,
+				      struct list_head *list)
+{
+	struct zynqmp_dma_desc_sw *desc, *next;
+
+	list_for_each_entry_safe(desc, next, list, node)
+		zynqmp_dma_free_descriptor(chan, desc);
+}
+
+/**
+ * zynqmp_dma_alloc_chan_resources - Allocate channel resources
+ * @dchan: DMA channel
+ *
+ * Return: Number of descriptors on success and failure value on error
+ */
+static int zynqmp_dma_alloc_chan_resources(struct dma_chan *dchan)
+{
+	struct zynqmp_dma_chan *chan = to_chan(dchan);
+	struct zynqmp_dma_desc_sw *desc;
+	int i;
+
+	chan->sw_desc_pool = kzalloc(sizeof(*desc) * ZYNQMP_DMA_NUM_DESCS,
+				     GFP_KERNEL);
+	if (!chan->sw_desc_pool)
+		return -ENOMEM;
+
+	chan->idle = true;
+	chan->desc_free_cnt = ZYNQMP_DMA_NUM_DESCS;
+
+	INIT_LIST_HEAD(&chan->free_list);
+
+	for (i = 0; i < ZYNQMP_DMA_NUM_DESCS; i++) {
+		desc = chan->sw_desc_pool + i;
+		dma_async_tx_descriptor_init(&desc->async_tx, &chan->common);
+		desc->async_tx.tx_submit = zynqmp_dma_tx_submit;
+		list_add_tail(&desc->node, &chan->free_list);
+	}
+
+	chan->desc_pool_v = dma_zalloc_coherent(chan->dev,
+				(2 * chan->desc_size * ZYNQMP_DMA_NUM_DESCS),
+				&chan->desc_pool_p, GFP_KERNEL);
+	if (!chan->desc_pool_v)
+		return -ENOMEM;
+
+	for (i = 0; i < ZYNQMP_DMA_NUM_DESCS; i++) {
+		desc = chan->sw_desc_pool + i;
+		desc->src_v = (struct zynqmp_dma_desc_ll *) (chan->desc_pool_v +
+					(i * ZYNQMP_DMA_DESC_SIZE(chan) * 2));
+		desc->dst_v = (struct zynqmp_dma_desc_ll *) (desc->src_v + 1);
+		desc->src_p = chan->desc_pool_p +
+				(i * ZYNQMP_DMA_DESC_SIZE(chan) * 2);
+		desc->dst_p = desc->src_p + ZYNQMP_DMA_DESC_SIZE(chan);
+	}
+
+	return ZYNQMP_DMA_NUM_DESCS;
+}
+
+/**
+ * zynqmp_dma_start - Start DMA channel
+ * @chan: ZynqMP DMA channel pointer
+ */
+static void zynqmp_dma_start(struct zynqmp_dma_chan *chan)
+{
+	writel(ZYNQMP_DMA_INT_EN_DEFAULT_MASK, chan->regs + ZYNQMP_DMA_IER);
+	chan->idle = false;
+	writel(ZYNQMP_DMA_ENABLE, chan->regs + ZYNQMP_DMA_CTRL2);
+}
+
+/**
+ * zynqmp_dma_handle_ovfl_int - Process the overflow interrupt
+ * @chan: ZynqMP DMA channel pointer
+ * @status: Interrupt status value
+ */
+static void zynqmp_dma_handle_ovfl_int(struct zynqmp_dma_chan *chan, u32 status)
+{
+	u32 val;
+
+	if (status & ZYNQMP_DMA_IRQ_DST_ACCT_ERR)
+		val = readl(chan->regs + ZYNQMP_DMA_IRQ_DST_ACCT);
+	if (status & ZYNQMP_DMA_IRQ_SRC_ACCT_ERR)
+		val = readl(chan->regs + ZYNQMP_DMA_IRQ_SRC_ACCT);
+}
+
+static void zynqmp_dma_config(struct zynqmp_dma_chan *chan)
+{
+	u32 val;
+
+	val = readl(chan->regs + ZYNQMP_DMA_CTRL0);
+	val |= ZYNQMP_DMA_POINT_TYPE_SG;
+	writel(val, chan->regs + ZYNQMP_DMA_CTRL0);
+
+	val = readl(chan->regs + ZYNQMP_DMA_DATA_ATTR);
+	val = (val & ~ZYNQMP_DMA_ARLEN) |
+		(chan->src_burst_len << ZYNQMP_DMA_ARLEN_OFST);
+	val = (val & ~ZYNQMP_DMA_AWLEN) |
+		(chan->dst_burst_len << ZYNQMP_DMA_AWLEN_OFST);
+	writel(val, chan->regs + ZYNQMP_DMA_DATA_ATTR);
+}
+
+/**
+ * zynqmp_dma_device_config - Zynqmp dma device configuration
+ * @dchan: DMA channel
+ * @config: DMA device config
+ */
+static int zynqmp_dma_device_config(struct dma_chan *dchan,
+				    struct dma_slave_config *config)
+{
+	struct zynqmp_dma_chan *chan = to_chan(dchan);
+
+	chan->src_burst_len = config->src_maxburst;
+	chan->dst_burst_len = config->dst_maxburst;
+
+	return 0;
+}
+
+/**
+ * zynqmp_dma_start_transfer - Initiate the new transfer
+ * @chan: ZynqMP DMA channel pointer
+ */
+static void zynqmp_dma_start_transfer(struct zynqmp_dma_chan *chan)
+{
+	struct zynqmp_dma_desc_sw *desc;
+
+	if (!chan->idle)
+		return;
+
+	zynqmp_dma_config(chan);
+
+	desc = list_first_entry_or_null(&chan->pending_list,
+					struct zynqmp_dma_desc_sw, node);
+	if (!desc)
+		return;
+
+	list_splice_tail_init(&chan->pending_list, &chan->active_list);
+	zynqmp_dma_update_desc_to_ctrlr(chan, desc);
+	zynqmp_dma_start(chan);
+}
+
+
+/**
+ * zynqmp_dma_chan_desc_cleanup - Cleanup the completed descriptors
+ * @chan: ZynqMP DMA channel
+ */
+static void zynqmp_dma_chan_desc_cleanup(struct zynqmp_dma_chan *chan)
+{
+	struct zynqmp_dma_desc_sw *desc, *next;
+
+	list_for_each_entry_safe(desc, next, &chan->done_list, node) {
+		dma_async_tx_callback callback;
+		void *callback_param;
+
+		list_del(&desc->node);
+
+		callback = desc->async_tx.callback;
+		callback_param = desc->async_tx.callback_param;
+		if (callback) {
+			spin_unlock(&chan->lock);
+			callback(callback_param);
+			spin_lock(&chan->lock);
+		}
+
+		/* Run any dependencies, then free the descriptor */
+		zynqmp_dma_free_descriptor(chan, desc);
+	}
+}
+
+/**
+ * zynqmp_dma_complete_descriptor - Mark the active descriptor as complete
+ * @chan: ZynqMP DMA channel pointer
+ */
+static void zynqmp_dma_complete_descriptor(struct zynqmp_dma_chan *chan)
+{
+	struct zynqmp_dma_desc_sw *desc;
+
+	desc = list_first_entry_or_null(&chan->active_list,
+					struct zynqmp_dma_desc_sw, node);
+	if (!desc)
+		return;
+	list_del(&desc->node);
+	dma_cookie_complete(&desc->async_tx);
+	list_add_tail(&desc->node, &chan->done_list);
+}
+
+/**
+ * zynqmp_dma_issue_pending - Issue pending transactions
+ * @dchan: DMA channel pointer
+ */
+static void zynqmp_dma_issue_pending(struct dma_chan *dchan)
+{
+	struct zynqmp_dma_chan *chan = to_chan(dchan);
+
+	spin_lock_bh(&chan->lock);
+	zynqmp_dma_start_transfer(chan);
+	spin_unlock_bh(&chan->lock);
+}
+
+/**
+ * zynqmp_dma_free_descriptors - Free channel descriptors
+ * @dchan: DMA channel pointer
+ */
+static void zynqmp_dma_free_descriptors(struct zynqmp_dma_chan *chan)
+{
+	zynqmp_dma_free_desc_list(chan, &chan->active_list);
+	zynqmp_dma_free_desc_list(chan, &chan->pending_list);
+	zynqmp_dma_free_desc_list(chan, &chan->done_list);
+}
+
+/**
+ * zynqmp_dma_free_chan_resources - Free channel resources
+ * @dchan: DMA channel pointer
+ */
+static void zynqmp_dma_free_chan_resources(struct dma_chan *dchan)
+{
+	struct zynqmp_dma_chan *chan = to_chan(dchan);
+
+	spin_lock_bh(&chan->lock);
+	zynqmp_dma_free_descriptors(chan);
+	spin_unlock_bh(&chan->lock);
+	dma_free_coherent(chan->dev,
+		(2 * ZYNQMP_DMA_DESC_SIZE(chan) * ZYNQMP_DMA_NUM_DESCS),
+		chan->desc_pool_v, chan->desc_pool_p);
+	kfree(chan->sw_desc_pool);
+}
+
+/**
+ * zynqmp_dma_reset - Reset the channel
+ * @chan: ZynqMP DMA channel pointer
+ */
+static void zynqmp_dma_reset(struct zynqmp_dma_chan *chan)
+{
+	writel(ZYNQMP_DMA_IDS_DEFAULT_MASK, chan->regs + ZYNQMP_DMA_IDS);
+
+	zynqmp_dma_complete_descriptor(chan);
+	zynqmp_dma_chan_desc_cleanup(chan);
+	zynqmp_dma_free_descriptors(chan);
+	zynqmp_dma_init(chan);
+}
+
+/**
+ * zynqmp_dma_irq_handler - ZynqMP DMA Interrupt handler
+ * @irq: IRQ number
+ * @data: Pointer to the ZynqMP DMA channel structure
+ *
+ * Return: IRQ_HANDLED/IRQ_NONE
+ */
+static irqreturn_t zynqmp_dma_irq_handler(int irq, void *data)
+{
+	struct zynqmp_dma_chan *chan = (struct zynqmp_dma_chan *)data;
+	u32 isr, imr, status;
+	irqreturn_t ret = IRQ_NONE;
+
+	isr = readl(chan->regs + ZYNQMP_DMA_ISR);
+	imr = readl(chan->regs + ZYNQMP_DMA_IMR);
+	status = isr & ~imr;
+
+	writel(isr, chan->regs + ZYNQMP_DMA_ISR);
+	if (status & ZYNQMP_DMA_INT_DONE) {
+		tasklet_schedule(&chan->tasklet);
+		ret = IRQ_HANDLED;
+	}
+
+	if (status & ZYNQMP_DMA_DONE)
+		chan->idle = true;
+
+	if (status & ZYNQMP_DMA_INT_ERR) {
+		chan->err = true;
+		tasklet_schedule(&chan->tasklet);
+		dev_err(chan->dev, "Channel %p has errors\n", chan);
+		ret = IRQ_HANDLED;
+	}
+
+	if (status & ZYNQMP_DMA_INT_OVRFL) {
+		zynqmp_dma_handle_ovfl_int(chan, status);
+		dev_info(chan->dev, "Channel %p overflow interrupt\n", chan);
+		ret = IRQ_HANDLED;
+	}
+
+	return ret;
+}
+
+/**
+ * zynqmp_dma_do_tasklet - Schedule completion tasklet
+ * @data: Pointer to the ZynqMP DMA channel structure
+ */
+static void zynqmp_dma_do_tasklet(unsigned long data)
+{
+	struct zynqmp_dma_chan *chan = (struct zynqmp_dma_chan *)data;
+	u32 count;
+
+	spin_lock(&chan->lock);
+
+	if (chan->err) {
+		zynqmp_dma_reset(chan);
+		chan->err = false;
+		goto unlock;
+	}
+
+	count = readl(chan->regs + ZYNQMP_DMA_IRQ_DST_ACCT);
+
+	while (count) {
+		zynqmp_dma_complete_descriptor(chan);
+		zynqmp_dma_chan_desc_cleanup(chan);
+		count--;
+	}
+
+	if (chan->idle)
+		zynqmp_dma_start_transfer(chan);
+
+unlock:
+	spin_unlock(&chan->lock);
+}
+
+/**
+ * zynqmp_dma_device_terminate_all - Aborts all transfers on a channel
+ * @dchan: DMA channel pointer
+ *
+ * Return: Always '0'
+ */
+static int zynqmp_dma_device_terminate_all(struct dma_chan *dchan)
+{
+	struct zynqmp_dma_chan *chan = to_chan(dchan);
+
+	spin_lock_bh(&chan->lock);
+	writel(ZYNQMP_DMA_IDS_DEFAULT_MASK, chan->regs + ZYNQMP_DMA_IDS);
+	zynqmp_dma_free_descriptors(chan);
+	spin_unlock_bh(&chan->lock);
+
+	return 0;
+}
+
+/**
+ * zynqmp_dma_prep_memcpy - prepare descriptors for memcpy transaction
+ * @dchan: DMA channel
+ * @dma_dst: Destination buffer address
+ * @dma_src: Source buffer address
+ * @len: Transfer length
+ * @flags: transfer ack flags
+ *
+ * Return: Async transaction descriptor on success and NULL on failure
+ */
+static struct dma_async_tx_descriptor *zynqmp_dma_prep_memcpy(
+				struct dma_chan *dchan, dma_addr_t dma_dst,
+				dma_addr_t dma_src, size_t len, ulong flags)
+{
+	struct zynqmp_dma_chan *chan;
+	struct zynqmp_dma_desc_sw *new, *first = NULL;
+	void *desc = NULL, *prev = NULL;
+	size_t copy;
+	u32 desc_cnt;
+
+	chan = to_chan(dchan);
+
+	if (len > ZYNQMP_DMA_MAX_TRANS_LEN)
+		return NULL;
+
+	desc_cnt = DIV_ROUND_UP(len, ZYNQMP_DMA_MAX_TRANS_LEN);
+
+	spin_lock_bh(&chan->lock);
+	if (desc_cnt > chan->desc_free_cnt) {
+		spin_unlock_bh(&chan->lock);
+		dev_dbg(chan->dev, "chan %p descs are not available\n", chan);
+		return NULL;
+	}
+	chan->desc_free_cnt = chan->desc_free_cnt - desc_cnt;
+	spin_unlock_bh(&chan->lock);
+
+	do {
+		/* Allocate and populate the descriptor */
+		new = zynqmp_dma_get_descriptor(chan);
+
+		copy = min_t(size_t, len, ZYNQMP_DMA_MAX_TRANS_LEN);
+		desc = (struct zynqmp_dma_desc_ll *)new->src_v;
+		zynqmp_dma_config_sg_ll_desc(chan, desc, dma_src,
+					     dma_dst, copy, prev);
+		prev = desc;
+		len -= copy;
+		dma_src += copy;
+		dma_dst += copy;
+		if (!first)
+			first = new;
+		else
+			list_add_tail(&new->node, &first->tx_list);
+	} while (len);
+
+	zynqmp_dma_desc_config_eod(chan, desc);
+	async_tx_ack(&first->async_tx);
+	first->async_tx.flags = flags;
+	return &first->async_tx;
+}
+
+/**
+ * zynqmp_dma_prep_slave_sg - prepare descriptors for a memory sg transaction
+ * @dchan: DMA channel
+ * @dst_sg: Destination scatter list
+ * @dst_sg_len: Number of entries in destination scatter list
+ * @src_sg: Source scatter list
+ * @src_sg_len: Number of entries in source scatter list
+ * @flags: transfer ack flags
+ *
+ * Return: Async transaction descriptor on success and NULL on failure
+ */
+static struct dma_async_tx_descriptor *zynqmp_dma_prep_sg(
+			struct dma_chan *dchan, struct scatterlist *dst_sg,
+			unsigned int dst_sg_len, struct scatterlist *src_sg,
+			unsigned int src_sg_len, unsigned long flags)
+{
+	struct zynqmp_dma_desc_sw *new, *first = NULL;
+	struct zynqmp_dma_chan *chan = to_chan(dchan);
+	void *desc = NULL, *prev = NULL;
+	size_t len, dst_avail, src_avail;
+	dma_addr_t dma_dst, dma_src;
+	u32 desc_cnt = 0, i;
+	struct scatterlist *sg;
+
+	for_each_sg(src_sg, sg, src_sg_len, i)
+		desc_cnt += DIV_ROUND_UP(sg_dma_len(sg),
+					 ZYNQMP_DMA_MAX_TRANS_LEN);
+
+	spin_lock_bh(&chan->lock);
+	if (desc_cnt > chan->desc_free_cnt) {
+		spin_unlock_bh(&chan->lock);
+		dev_dbg(chan->dev, "chan %p descs are not available\n", chan);
+		return NULL;
+	}
+	chan->desc_free_cnt = chan->desc_free_cnt - desc_cnt;
+	spin_unlock_bh(&chan->lock);
+
+	dst_avail = sg_dma_len(dst_sg);
+	src_avail = sg_dma_len(src_sg);
+
+	/* Run until we are out of scatterlist entries */
+	while (true) {
+		/* Allocate and populate the descriptor */
+		new = zynqmp_dma_get_descriptor(chan);
+		desc = (struct zynqmp_dma_desc_ll *)new->src_v;
+		len = min_t(size_t, src_avail, dst_avail);
+		len = min_t(size_t, len, ZYNQMP_DMA_MAX_TRANS_LEN);
+		if (len == 0)
+			goto fetch;
+		dma_dst = sg_dma_address(dst_sg) + sg_dma_len(dst_sg) -
+			dst_avail;
+		dma_src = sg_dma_address(src_sg) + sg_dma_len(src_sg) -
+			src_avail;
+
+		zynqmp_dma_config_sg_ll_desc(chan, desc, dma_src, dma_dst,
+					     len, prev);
+		prev = desc;
+		dst_avail -= len;
+		src_avail -= len;
+
+		if (!first)
+			first = new;
+		else
+			list_add_tail(&new->node, &first->tx_list);
+fetch:
+		/* Fetch the next dst scatterlist entry */
+		if (dst_avail == 0) {
+			if (dst_sg_len == 0)
+				break;
+			dst_sg = sg_next(dst_sg);
+			if (dst_sg == NULL)
+				break;
+			dst_sg_len--;
+			dst_avail = sg_dma_len(dst_sg);
+		}
+		/* Fetch the next src scatterlist entry */
+		if (src_avail == 0) {
+			if (src_sg_len == 0)
+				break;
+			src_sg = sg_next(src_sg);
+			if (src_sg == NULL)
+				break;
+			src_sg_len--;
+			src_avail = sg_dma_len(src_sg);
+		}
+	}
+
+	zynqmp_dma_desc_config_eod(chan, desc);
+	first->async_tx.flags = flags;
+	return &first->async_tx;
+}
+
+/**
+ * zynqmp_dma_chan_remove - Channel remove function
+ * @chan: ZynqMP DMA channel pointer
+ */
+static void zynqmp_dma_chan_remove(struct zynqmp_dma_chan *chan)
+{
+	if (!chan)
+		return;
+
+	devm_free_irq(chan->zdev->dev, chan->irq, chan);
+	tasklet_kill(&chan->tasklet);
+	list_del(&chan->common.device_node);
+	clk_disable_unprepare(chan->clk_apb);
+	clk_disable_unprepare(chan->clk_main);
+}
+
+/**
+ * zynqmp_dma_chan_probe - Per Channel Probing
+ * @zdev: Driver specific device structure
+ * @pdev: Pointer to the platform_device structure
+ *
+ * Return: '0' on success and failure value on error
+ */
+static int zynqmp_dma_chan_probe(struct zynqmp_dma_device *zdev,
+			   struct platform_device *pdev)
+{
+	struct zynqmp_dma_chan *chan;
+	struct resource *res;
+	struct device_node *node = pdev->dev.of_node;
+	int err;
+
+	chan = devm_kzalloc(zdev->dev, sizeof(*chan), GFP_KERNEL);
+	if (!chan)
+		return -ENOMEM;
+	chan->dev = zdev->dev;
+	chan->zdev = zdev;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	chan->regs = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(chan->regs))
+		return PTR_ERR(chan->regs);
+
+	chan->bus_width = ZYNQMP_DMA_BUS_WIDTH_64;
+	chan->dst_burst_len = ZYNQMP_DMA_AWLEN_RST_VAL;
+	chan->src_burst_len = ZYNQMP_DMA_ARLEN_RST_VAL;
+	err = of_property_read_u32(node, "xlnx,bus-width", &chan->bus_width);
+	if ((err < 0) && ((chan->bus_width != ZYNQMP_DMA_BUS_WIDTH_64) ||
+			  (chan->bus_width != ZYNQMP_DMA_BUS_WIDTH_128))) {
+		dev_err(zdev->dev, "invalid bus-width value");
+		return err;
+	}
+
+	chan->is_dmacoherent =  of_property_read_bool(node, "dma-coherent");
+	zdev->chan = chan;
+	tasklet_init(&chan->tasklet, zynqmp_dma_do_tasklet, (ulong)chan);
+	spin_lock_init(&chan->lock);
+	INIT_LIST_HEAD(&chan->active_list);
+	INIT_LIST_HEAD(&chan->pending_list);
+	INIT_LIST_HEAD(&chan->done_list);
+	INIT_LIST_HEAD(&chan->free_list);
+
+	dma_cookie_init(&chan->common);
+	chan->common.device = &zdev->common;
+	list_add_tail(&chan->common.device_node, &zdev->common.channels);
+
+	zynqmp_dma_init(chan);
+	chan->irq = platform_get_irq(pdev, 0);
+	if (chan->irq < 0)
+		return -ENXIO;
+	err = devm_request_irq(&pdev->dev, chan->irq, zynqmp_dma_irq_handler, 0,
+			       "zynqmp-dma", chan);
+	if (err)
+		return err;
+	chan->clk_main = devm_clk_get(&pdev->dev, "clk_main");
+	if (IS_ERR(chan->clk_main)) {
+		dev_err(&pdev->dev, "main clock not found.\n");
+		return PTR_ERR(chan->clk_main);
+	}
+
+	chan->clk_apb = devm_clk_get(&pdev->dev, "clk_apb");
+	if (IS_ERR(chan->clk_apb)) {
+		dev_err(&pdev->dev, "apb clock not found.\n");
+		return PTR_ERR(chan->clk_apb);
+	}
+
+	err = clk_prepare_enable(chan->clk_main);
+	if (err) {
+		dev_err(&pdev->dev, "Unable to enable main clock.\n");
+		return err;
+	}
+
+	err = clk_prepare_enable(chan->clk_apb);
+	if (err) {
+		clk_disable_unprepare(chan->clk_main);
+		dev_err(&pdev->dev, "Unable to enable apb clock.\n");
+		return err;
+	}
+
+	chan->desc_size = sizeof(struct zynqmp_dma_desc_ll);
+	chan->idle = true;
+	return 0;
+}
+
+/**
+ * of_zynqmp_dma_xlate - Translation function
+ * @dma_spec: Pointer to DMA specifier as found in the device tree
+ * @ofdma: Pointer to DMA controller data
+ *
+ * Return: DMA channel pointer on success and NULL on error
+ */
+static struct dma_chan *of_zynqmp_dma_xlate(struct of_phandle_args *dma_spec,
+					    struct of_dma *ofdma)
+{
+	struct zynqmp_dma_device *zdev = ofdma->of_dma_data;
+
+	return dma_get_slave_channel(&zdev->chan->common);
+}
+
+/**
+ * zynqmp_dma_probe - Driver probe function
+ * @pdev: Pointer to the platform_device structure
+ *
+ * Return: '0' on success and failure value on error
+ */
+static int zynqmp_dma_probe(struct platform_device *pdev)
+{
+	struct zynqmp_dma_device *zdev;
+	struct dma_device *p;
+	int ret;
+
+	zdev = devm_kzalloc(&pdev->dev, sizeof(*zdev), GFP_KERNEL);
+	if (!zdev)
+		return -ENOMEM;
+
+	zdev->dev = &pdev->dev;
+	INIT_LIST_HEAD(&zdev->common.channels);
+
+	dma_set_mask(&pdev->dev, DMA_BIT_MASK(44));
+	dma_cap_set(DMA_SG, zdev->common.cap_mask);
+	dma_cap_set(DMA_MEMCPY, zdev->common.cap_mask);
+
+	p = &zdev->common;
+	p->device_prep_dma_sg = zynqmp_dma_prep_sg;
+	p->device_prep_dma_memcpy = zynqmp_dma_prep_memcpy;
+	p->device_terminate_all = zynqmp_dma_device_terminate_all;
+	p->device_issue_pending = zynqmp_dma_issue_pending;
+	p->device_alloc_chan_resources = zynqmp_dma_alloc_chan_resources;
+	p->device_free_chan_resources = zynqmp_dma_free_chan_resources;
+	p->device_tx_status = dma_cookie_status;
+	p->device_config = zynqmp_dma_device_config;
+	p->dev = &pdev->dev;
+
+	platform_set_drvdata(pdev, zdev);
+
+	ret = zynqmp_dma_chan_probe(zdev, pdev);
+	if (ret) {
+		dev_err(&pdev->dev, "Probing channel failed\n");
+		goto free_chan_resources;
+	}
+
+	p->dst_addr_widths = BIT(zdev->chan->bus_width / 8);
+	p->src_addr_widths = BIT(zdev->chan->bus_width / 8);
+
+	dma_async_device_register(&zdev->common);
+
+	ret = of_dma_controller_register(pdev->dev.of_node,
+					 of_zynqmp_dma_xlate, zdev);
+	if (ret) {
+		dev_err(&pdev->dev, "Unable to register DMA to DT\n");
+		dma_async_device_unregister(&zdev->common);
+		goto free_chan_resources;
+	}
+
+	dev_info(&pdev->dev, "ZynqMP DMA driver Probe success\n");
+
+	return 0;
+
+free_chan_resources:
+	zynqmp_dma_chan_remove(zdev->chan);
+	return ret;
+}
+
+/**
+ * zynqmp_dma_remove - Driver remove function
+ * @pdev: Pointer to the platform_device structure
+ *
+ * Return: Always '0'
+ */
+static int zynqmp_dma_remove(struct platform_device *pdev)
+{
+	struct zynqmp_dma_device *zdev = platform_get_drvdata(pdev);
+
+	of_dma_controller_free(pdev->dev.of_node);
+	dma_async_device_unregister(&zdev->common);
+
+	zynqmp_dma_chan_remove(zdev->chan);
+
+	return 0;
+}
+
+static const struct of_device_id zynqmp_dma_of_match[] = {
+	{ .compatible = "xlnx,zynqmp-dma-1.0", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, zynqmp_dma_of_match);
+
+static struct platform_driver zynqmp_dma_driver = {
+	.driver = {
+		.name = "xilinx-zynqmp-dma",
+		.of_match_table = zynqmp_dma_of_match,
+	},
+	.probe = zynqmp_dma_probe,
+	.remove = zynqmp_dma_remove,
+};
+
+module_platform_driver(zynqmp_dma_driver);
+
+MODULE_AUTHOR("Xilinx, Inc.");
+MODULE_DESCRIPTION("Xilinx ZynqMP DMA driver");