Merge branch 'for-linus' of git://git.infradead.org/users/vkoul/slave-dma

Pull slave-dmaengine updates from Vinod Koul:

 - new drivers for:
        - Ingenic JZ4780 controller
        - APM X-Gene controller
        - Freescale RaidEngine device
        - Renesas USB Controller

  - remove device_alloc_chan_resources dummy handlers

  - sh driver cleanups for peri peri and related emmc and asoc patches
    as well

  - fixes and enhancements spread over the drivers

* 'for-linus' of git://git.infradead.org/users/vkoul/slave-dma: (59 commits)
  dmaengine: dw: don't prompt for DW_DMAC_CORE
  dmaengine: shdmac: avoid unused variable warnings
  dmaengine: fix platform_no_drv_owner.cocci warnings
  dmaengine: pch_dma: fix memory leak on failure path in pch_dma_probe()
  dmaengine: at_xdmac: unlock spin lock before return
  dmaengine: xgene: devm_ioremap() returns NULL on error
  dmaengine: xgene: buffer overflow in xgene_dma_init_channels()
  dmaengine: usb-dmac: Fix dereferencing freed memory 'desc'
  dmaengine: sa11x0: report slave capabilities to upper layers
  dmaengine: vdma: Fix compilation warnings
  dmaengine: fsl_raid: statify fsl_re_chan_probe
  dmaengine: Driver support for FSL RaidEngine device.
  dmaengine: xgene_dma_init_ring_mngr() can be static
  Documentation: dma: Add documentation for the APM X-Gene SoC DMA device DTS binding
  arm64: dts: Add APM X-Gene SoC DMA device and DMA clock DTS nodes
  dmaengine: Add support for APM X-Gene SoC DMA engine driver
  dmaengine: usb-dmac: Add Renesas USB DMA Controller (USB-DMAC) driver
  dmaengine: renesas,usb-dmac: Add device tree bindings documentation
  dmaengine: edma: fixed wrongly initialized data parameter to the edma callback
  dmaengine: ste_dma40: fix implicit conversion
  ...

49 files changed, 5446 insertions, 584 deletions

diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index 91eced044321..fd7ac13f2574 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -112,6 +112,17 @@ config FSL_DMA
 	  EloPlus is on mpc85xx and mpc86xx and Pxxx parts, and the Elo3 is on
 	  some Txxx and Bxxx parts.
 
+config FSL_RAID
+        tristate "Freescale RAID engine Support"
+        depends on FSL_SOC && !ASYNC_TX_ENABLE_CHANNEL_SWITCH
+        select DMA_ENGINE
+        select DMA_ENGINE_RAID
+        ---help---
+          Enable support for Freescale RAID Engine. RAID Engine is
+          available on some QorIQ SoCs (like P5020/P5040). It has
+          the capability to offload memcpy, xor and pq computation
+	  for raid5/6.
+
 source "drivers/dma/hsu/Kconfig"
 
 config MPC512X_DMA
@@ -347,6 +358,16 @@ config DMA_JZ4740
 	select DMA_ENGINE
 	select DMA_VIRTUAL_CHANNELS
 
+config DMA_JZ4780
+	tristate "JZ4780 DMA support"
+	depends on MACH_JZ4780
+	select DMA_ENGINE
+	select DMA_VIRTUAL_CHANNELS
+	help
+	  This selects support for the DMA controller in Ingenic JZ4780 SoCs.
+	  If you have a board based on such a SoC and wish to use DMA for
+	  devices which can use the DMA controller, say Y or M here.
+
 config K3_DMA
 	tristate "Hisilicon K3 DMA support"
 	depends on ARCH_HI3xxx
@@ -414,6 +435,14 @@ config IMG_MDC_DMA
 	help
 	  Enable support for the IMG multi-threaded DMA controller (MDC).
 
+config XGENE_DMA
+	tristate "APM X-Gene DMA support"
+	select DMA_ENGINE
+	select DMA_ENGINE_RAID
+	select ASYNC_TX_ENABLE_CHANNEL_SWITCH
+	help
+	  Enable support for the APM X-Gene SoC DMA engine.
+
 config DMA_ENGINE
 	bool
 
diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile
index 7e8301cb489d..69f77d5ba53b 100644
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@@ -41,9 +41,11 @@ obj-$(CONFIG_DMA_OMAP) += omap-dma.o
 obj-$(CONFIG_DMA_BCM2835) += bcm2835-dma.o
 obj-$(CONFIG_MMP_PDMA) += mmp_pdma.o
 obj-$(CONFIG_DMA_JZ4740) += dma-jz4740.o
+obj-$(CONFIG_DMA_JZ4780) += dma-jz4780.o
 obj-$(CONFIG_TI_CPPI41) += cppi41.o
 obj-$(CONFIG_K3_DMA) += k3dma.o
 obj-$(CONFIG_MOXART_DMA) += moxart-dma.o
+obj-$(CONFIG_FSL_RAID) += fsl_raid.o
 obj-$(CONFIG_FSL_EDMA) += fsl-edma.o
 obj-$(CONFIG_QCOM_BAM_DMA) += qcom_bam_dma.o
 obj-y += xilinx/
@@ -51,3 +53,4 @@ obj-$(CONFIG_INTEL_MIC_X100_DMA) += mic_x100_dma.o
 obj-$(CONFIG_NBPFAXI_DMA) += nbpfaxi.o
 obj-$(CONFIG_DMA_SUN6I) += sun6i-dma.o
 obj-$(CONFIG_IMG_MDC_DMA) += img-mdc-dma.o
+obj-$(CONFIG_XGENE_DMA) += xgene-dma.o
diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c
index 83aa55d6fa5d..49d396ec06e5 100644
--- a/drivers/dma/amba-pl08x.c
+++ b/drivers/dma/amba-pl08x.c
@@ -15,10 +15,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA  02111-1307, USA.
- *
  * The full GNU General Public License is in this distribution in the file
  * called COPYING.
  *
@@ -1195,11 +1191,6 @@ static void pl08x_free_txd_list(struct pl08x_driver_data *pl08x,
 /*
  * The DMA ENGINE API
  */
-static int pl08x_alloc_chan_resources(struct dma_chan *chan)
-{
-	return 0;
-}
-
 static void pl08x_free_chan_resources(struct dma_chan *chan)
 {
 	/* Ensure all queued descriptors are freed */
@@ -2066,7 +2057,6 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 	/* Initialize memcpy engine */
 	dma_cap_set(DMA_MEMCPY, pl08x->memcpy.cap_mask);
 	pl08x->memcpy.dev = &adev->dev;
-	pl08x->memcpy.device_alloc_chan_resources = pl08x_alloc_chan_resources;
 	pl08x->memcpy.device_free_chan_resources = pl08x_free_chan_resources;
 	pl08x->memcpy.device_prep_dma_memcpy = pl08x_prep_dma_memcpy;
 	pl08x->memcpy.device_prep_dma_interrupt = pl08x_prep_dma_interrupt;
@@ -2085,7 +2075,6 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 	dma_cap_set(DMA_SLAVE, pl08x->slave.cap_mask);
 	dma_cap_set(DMA_CYCLIC, pl08x->slave.cap_mask);
 	pl08x->slave.dev = &adev->dev;
-	pl08x->slave.device_alloc_chan_resources = pl08x_alloc_chan_resources;
 	pl08x->slave.device_free_chan_resources = pl08x_free_chan_resources;
 	pl08x->slave.device_prep_dma_interrupt = pl08x_prep_dma_interrupt;
 	pl08x->slave.device_tx_status = pl08x_dma_tx_status;
diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
index 0b4fc6fb48ce..57b2141ddddc 100644
--- a/drivers/dma/at_hdmac.c
+++ b/drivers/dma/at_hdmac.c
@@ -65,6 +65,21 @@ static void atc_issue_pending(struct dma_chan *chan);
 
 /*----------------------------------------------------------------------*/
 
+static inline unsigned int atc_get_xfer_width(dma_addr_t src, dma_addr_t dst,
+						size_t len)
+{
+	unsigned int width;
+
+	if (!((src | dst  | len) & 3))
+		width = 2;
+	else if (!((src | dst | len) & 1))
+		width = 1;
+	else
+		width = 0;
+
+	return width;
+}
+
 static struct at_desc *atc_first_active(struct at_dma_chan *atchan)
 {
 	return list_first_entry(&atchan->active_list,
@@ -659,16 +674,10 @@ atc_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 	 * We can be a lot more clever here, but this should take care
 	 * of the most common optimization.
 	 */
-	if (!((src | dest  | len) & 3)) {
-		ctrla = ATC_SRC_WIDTH_WORD | ATC_DST_WIDTH_WORD;
-		src_width = dst_width = 2;
-	} else if (!((src | dest | len) & 1)) {
-		ctrla = ATC_SRC_WIDTH_HALFWORD | ATC_DST_WIDTH_HALFWORD;
-		src_width = dst_width = 1;
-	} else {
-		ctrla = ATC_SRC_WIDTH_BYTE | ATC_DST_WIDTH_BYTE;
-		src_width = dst_width = 0;
-	}
+	src_width = dst_width = atc_get_xfer_width(src, dest, len);
+
+	ctrla = ATC_SRC_WIDTH(src_width) |
+		ATC_DST_WIDTH(dst_width);
 
 	for (offset = 0; offset < len; offset += xfer_count << src_width) {
 		xfer_count = min_t(size_t, (len - offset) >> src_width,
@@ -862,6 +871,144 @@ err:
 }
 
 /**
+ * atc_prep_dma_sg - prepare memory to memory scather-gather operation
+ * @chan: the channel to prepare operation on
+ * @dst_sg: destination scatterlist
+ * @dst_nents: number of destination scatterlist entries
+ * @src_sg: source scatterlist
+ * @src_nents: number of source scatterlist entries
+ * @flags: tx descriptor status flags
+ */
+static struct dma_async_tx_descriptor *
+atc_prep_dma_sg(struct dma_chan *chan,
+		struct scatterlist *dst_sg, unsigned int dst_nents,
+		struct scatterlist *src_sg, unsigned int src_nents,
+		unsigned long flags)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	struct at_desc		*desc = NULL;
+	struct at_desc		*first = NULL;
+	struct at_desc		*prev = NULL;
+	unsigned int		src_width;
+	unsigned int		dst_width;
+	size_t			xfer_count;
+	u32			ctrla;
+	u32			ctrlb;
+	size_t			dst_len = 0, src_len = 0;
+	dma_addr_t		dst = 0, src = 0;
+	size_t			len = 0, total_len = 0;
+
+	if (unlikely(dst_nents == 0 || src_nents == 0))
+		return NULL;
+
+	if (unlikely(dst_sg == NULL || src_sg == NULL))
+		return NULL;
+
+	ctrlb =   ATC_DEFAULT_CTRLB | ATC_IEN
+		| ATC_SRC_ADDR_MODE_INCR
+		| ATC_DST_ADDR_MODE_INCR
+		| ATC_FC_MEM2MEM;
+
+	/*
+	 * loop until there is either no more source or no more destination
+	 * scatterlist entry
+	 */
+	while (true) {
+
+		/* prepare the next transfer */
+		if (dst_len == 0) {
+
+			/* no more destination scatterlist entries */
+			if (!dst_sg || !dst_nents)
+				break;
+
+			dst = sg_dma_address(dst_sg);
+			dst_len = sg_dma_len(dst_sg);
+
+			dst_sg = sg_next(dst_sg);
+			dst_nents--;
+		}
+
+		if (src_len == 0) {
+
+			/* no more source scatterlist entries */
+			if (!src_sg || !src_nents)
+				break;
+
+			src = sg_dma_address(src_sg);
+			src_len = sg_dma_len(src_sg);
+
+			src_sg = sg_next(src_sg);
+			src_nents--;
+		}
+
+		len = min_t(size_t, src_len, dst_len);
+		if (len == 0)
+			continue;
+
+		/* take care for the alignment */
+		src_width = dst_width = atc_get_xfer_width(src, dst, len);
+
+		ctrla = ATC_SRC_WIDTH(src_width) |
+			ATC_DST_WIDTH(dst_width);
+
+		/*
+		 * The number of transfers to set up refer to the source width
+		 * that depends on the alignment.
+		 */
+		xfer_count = len >> src_width;
+		if (xfer_count > ATC_BTSIZE_MAX) {
+			xfer_count = ATC_BTSIZE_MAX;
+			len = ATC_BTSIZE_MAX << src_width;
+		}
+
+		/* create the transfer */
+		desc = atc_desc_get(atchan);
+		if (!desc)
+			goto err_desc_get;
+
+		desc->lli.saddr = src;
+		desc->lli.daddr = dst;
+		desc->lli.ctrla = ctrla | xfer_count;
+		desc->lli.ctrlb = ctrlb;
+
+		desc->txd.cookie = 0;
+		desc->len = len;
+
+		/*
+		 * Although we only need the transfer width for the first and
+		 * the last descriptor, its easier to set it to all descriptors.
+		 */
+		desc->tx_width = src_width;
+
+		atc_desc_chain(&first, &prev, desc);
+
+		/* update the lengths and addresses for the next loop cycle */
+		dst_len -= len;
+		src_len -= len;
+		dst += len;
+		src += len;
+
+		total_len += len;
+	}
+
+	/* First descriptor of the chain embedds additional information */
+	first->txd.cookie = -EBUSY;
+	first->total_len = total_len;
+
+	/* set end-of-link to the last link descriptor of list*/
+	set_desc_eol(desc);
+
+	first->txd.flags = flags; /* client is in control of this ack */
+
+	return &first->txd;
+
+err_desc_get:
+	atc_desc_put(atchan, first);
+	return NULL;
+}
+
+/**
  * atc_dma_cyclic_check_values
  * Check for too big/unaligned periods and unaligned DMA buffer
  */
@@ -1461,8 +1608,10 @@ static int __init at_dma_probe(struct platform_device *pdev)
 
 	/* setup platform data for each SoC */
 	dma_cap_set(DMA_MEMCPY, at91sam9rl_config.cap_mask);
+	dma_cap_set(DMA_SG, at91sam9rl_config.cap_mask);
 	dma_cap_set(DMA_MEMCPY, at91sam9g45_config.cap_mask);
 	dma_cap_set(DMA_SLAVE, at91sam9g45_config.cap_mask);
+	dma_cap_set(DMA_SG, at91sam9g45_config.cap_mask);
 
 	/* get DMA parameters from controller type */
 	plat_dat = at_dma_get_driver_data(pdev);
@@ -1582,11 +1731,15 @@ static int __init at_dma_probe(struct platform_device *pdev)
 		atdma->dma_common.residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
 	}
 
+	if (dma_has_cap(DMA_SG, atdma->dma_common.cap_mask))
+		atdma->dma_common.device_prep_dma_sg = atc_prep_dma_sg;
+
 	dma_writel(atdma, EN, AT_DMA_ENABLE);
 
-	dev_info(&pdev->dev, "Atmel AHB DMA Controller ( %s%s), %d channels\n",
+	dev_info(&pdev->dev, "Atmel AHB DMA Controller ( %s%s%s), %d channels\n",
 	  dma_has_cap(DMA_MEMCPY, atdma->dma_common.cap_mask) ? "cpy " : "",
 	  dma_has_cap(DMA_SLAVE, atdma->dma_common.cap_mask)  ? "slave " : "",
+	  dma_has_cap(DMA_SG, atdma->dma_common.cap_mask)  ? "sg-cpy " : "",
 	  plat_dat->nr_channels);
 
 	dma_async_device_register(&atdma->dma_common);
diff --git a/drivers/dma/at_xdmac.c b/drivers/dma/at_xdmac.c
index d9891d3461f6..933e4b338459 100644
--- a/drivers/dma/at_xdmac.c
+++ b/drivers/dma/at_xdmac.c
@@ -1154,8 +1154,10 @@ static int at_xdmac_device_resume(struct dma_chan *chan)
 	dev_dbg(chan2dev(chan), "%s\n", __func__);
 
 	spin_lock_bh(&atchan->lock);
-	if (!at_xdmac_chan_is_paused(atchan))
+	if (!at_xdmac_chan_is_paused(atchan)) {
+		spin_unlock_bh(&atchan->lock);
 		return 0;
+	}
 
 	at_xdmac_write(atxdmac, AT_XDMAC_GRWR, atchan->mask);
 	clear_bit(AT_XDMAC_CHAN_IS_PAUSED, &atchan->status);
diff --git a/drivers/dma/bestcomm/bestcomm.c b/drivers/dma/bestcomm/bestcomm.c
index fa378d88f6c8..180fedb418cc 100644
--- a/drivers/dma/bestcomm/bestcomm.c
+++ b/drivers/dma/bestcomm/bestcomm.c
@@ -30,7 +30,7 @@
 #define DRIVER_NAME "bestcomm-core"
 
 /* MPC5200 device tree match tables */
-static struct of_device_id mpc52xx_sram_ids[] = {
+static const struct of_device_id mpc52xx_sram_ids[] = {
 	{ .compatible = "fsl,mpc5200-sram", },
 	{ .compatible = "mpc5200-sram", },
 	{}
@@ -481,7 +481,7 @@ static int mpc52xx_bcom_remove(struct platform_device *op)
 	return 0;
 }
 
-static struct of_device_id mpc52xx_bcom_of_match[] = {
+static const struct of_device_id mpc52xx_bcom_of_match[] = {
 	{ .compatible = "fsl,mpc5200-bestcomm", },
 	{ .compatible = "mpc5200-bestcomm", },
 	{},
diff --git a/drivers/dma/dma-jz4740.c b/drivers/dma/dma-jz4740.c
index 84884418fd30..7638b24ce8d0 100644
--- a/drivers/dma/dma-jz4740.c
+++ b/drivers/dma/dma-jz4740.c
@@ -7,10 +7,6 @@
  *  Free Software Foundation;  either version 2 of the License, or (at your
  *  option) any later version.
  *
- *  You should have received a copy of the GNU General Public License along
- *  with this program; if not, write to the Free Software Foundation, Inc.,
- *  675 Mass Ave, Cambridge, MA 02139, USA.
- *
  */
 
 #include <linux/dmaengine.h>
@@ -343,7 +339,7 @@ static void jz4740_dma_chan_irq(struct jz4740_dmaengine_chan *chan)
 {
 	spin_lock(&chan->vchan.lock);
 	if (chan->desc) {
-		if (chan->desc && chan->desc->cyclic) {
+		if (chan->desc->cyclic) {
 			vchan_cyclic_callback(&chan->desc->vdesc);
 		} else {
 			if (chan->next_sg == chan->desc->num_sgs) {
@@ -496,11 +492,6 @@ static enum dma_status jz4740_dma_tx_status(struct dma_chan *c,
 	return status;
 }
 
-static int jz4740_dma_alloc_chan_resources(struct dma_chan *c)
-{
-	return 0;
-}
-
 static void jz4740_dma_free_chan_resources(struct dma_chan *c)
 {
 	vchan_free_chan_resources(to_virt_chan(c));
@@ -543,7 +534,6 @@ static int jz4740_dma_probe(struct platform_device *pdev)
 
 	dma_cap_set(DMA_SLAVE, dd->cap_mask);
 	dma_cap_set(DMA_CYCLIC, dd->cap_mask);
-	dd->device_alloc_chan_resources = jz4740_dma_alloc_chan_resources;
 	dd->device_free_chan_resources = jz4740_dma_free_chan_resources;
 	dd->device_tx_status = jz4740_dma_tx_status;
 	dd->device_issue_pending = jz4740_dma_issue_pending;
diff --git a/drivers/dma/dma-jz4780.c b/drivers/dma/dma-jz4780.c
new file mode 100644
index 000000000000..26d2f0e09ea3
--- /dev/null
+++ b/drivers/dma/dma-jz4780.c
@@ -0,0 +1,877 @@
+/*
+ * Ingenic JZ4780 DMA controller
+ *
+ * Copyright (c) 2015 Imagination Technologies
+ * Author: Alex Smith <alex@alex-smith.me.uk>
+ *
+ * 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/clk.h>
+#include <linux/dmapool.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_dma.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+
+#include "dmaengine.h"
+#include "virt-dma.h"
+
+#define JZ_DMA_NR_CHANNELS	32
+
+/* Global registers. */
+#define JZ_DMA_REG_DMAC		0x1000
+#define JZ_DMA_REG_DIRQP	0x1004
+#define JZ_DMA_REG_DDR		0x1008
+#define JZ_DMA_REG_DDRS		0x100c
+#define JZ_DMA_REG_DMACP	0x101c
+#define JZ_DMA_REG_DSIRQP	0x1020
+#define JZ_DMA_REG_DSIRQM	0x1024
+#define JZ_DMA_REG_DCIRQP	0x1028
+#define JZ_DMA_REG_DCIRQM	0x102c
+
+/* Per-channel registers. */
+#define JZ_DMA_REG_CHAN(n)	(n * 0x20)
+#define JZ_DMA_REG_DSA(n)	(0x00 + JZ_DMA_REG_CHAN(n))
+#define JZ_DMA_REG_DTA(n)	(0x04 + JZ_DMA_REG_CHAN(n))
+#define JZ_DMA_REG_DTC(n)	(0x08 + JZ_DMA_REG_CHAN(n))
+#define JZ_DMA_REG_DRT(n)	(0x0c + JZ_DMA_REG_CHAN(n))
+#define JZ_DMA_REG_DCS(n)	(0x10 + JZ_DMA_REG_CHAN(n))
+#define JZ_DMA_REG_DCM(n)	(0x14 + JZ_DMA_REG_CHAN(n))
+#define JZ_DMA_REG_DDA(n)	(0x18 + JZ_DMA_REG_CHAN(n))
+#define JZ_DMA_REG_DSD(n)	(0x1c + JZ_DMA_REG_CHAN(n))
+
+#define JZ_DMA_DMAC_DMAE	BIT(0)
+#define JZ_DMA_DMAC_AR		BIT(2)
+#define JZ_DMA_DMAC_HLT		BIT(3)
+#define JZ_DMA_DMAC_FMSC	BIT(31)
+
+#define JZ_DMA_DRT_AUTO		0x8
+
+#define JZ_DMA_DCS_CTE		BIT(0)
+#define JZ_DMA_DCS_HLT		BIT(2)
+#define JZ_DMA_DCS_TT		BIT(3)
+#define JZ_DMA_DCS_AR		BIT(4)
+#define JZ_DMA_DCS_DES8		BIT(30)
+
+#define JZ_DMA_DCM_LINK		BIT(0)
+#define JZ_DMA_DCM_TIE		BIT(1)
+#define JZ_DMA_DCM_STDE		BIT(2)
+#define JZ_DMA_DCM_TSZ_SHIFT	8
+#define JZ_DMA_DCM_TSZ_MASK	(0x7 << JZ_DMA_DCM_TSZ_SHIFT)
+#define JZ_DMA_DCM_DP_SHIFT	12
+#define JZ_DMA_DCM_SP_SHIFT	14
+#define JZ_DMA_DCM_DAI		BIT(22)
+#define JZ_DMA_DCM_SAI		BIT(23)
+
+#define JZ_DMA_SIZE_4_BYTE	0x0
+#define JZ_DMA_SIZE_1_BYTE	0x1
+#define JZ_DMA_SIZE_2_BYTE	0x2
+#define JZ_DMA_SIZE_16_BYTE	0x3
+#define JZ_DMA_SIZE_32_BYTE	0x4
+#define JZ_DMA_SIZE_64_BYTE	0x5
+#define JZ_DMA_SIZE_128_BYTE	0x6
+
+#define JZ_DMA_WIDTH_32_BIT	0x0
+#define JZ_DMA_WIDTH_8_BIT	0x1
+#define JZ_DMA_WIDTH_16_BIT	0x2
+
+#define JZ_DMA_BUSWIDTHS	(BIT(DMA_SLAVE_BUSWIDTH_1_BYTE)	 | \
+				 BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) | \
+				 BIT(DMA_SLAVE_BUSWIDTH_4_BYTES))
+
+/**
+ * struct jz4780_dma_hwdesc - descriptor structure read by the DMA controller.
+ * @dcm: value for the DCM (channel command) register
+ * @dsa: source address
+ * @dta: target address
+ * @dtc: transfer count (number of blocks of the transfer size specified in DCM
+ * to transfer) in the low 24 bits, offset of the next descriptor from the
+ * descriptor base address in the upper 8 bits.
+ * @sd: target/source stride difference (in stride transfer mode).
+ * @drt: request type
+ */
+struct jz4780_dma_hwdesc {
+	uint32_t dcm;
+	uint32_t dsa;
+	uint32_t dta;
+	uint32_t dtc;
+	uint32_t sd;
+	uint32_t drt;
+	uint32_t reserved[2];
+};
+
+/* Size of allocations for hardware descriptor blocks. */
+#define JZ_DMA_DESC_BLOCK_SIZE	PAGE_SIZE
+#define JZ_DMA_MAX_DESC		\
+	(JZ_DMA_DESC_BLOCK_SIZE / sizeof(struct jz4780_dma_hwdesc))
+
+struct jz4780_dma_desc {
+	struct virt_dma_desc vdesc;
+
+	struct jz4780_dma_hwdesc *desc;
+	dma_addr_t desc_phys;
+	unsigned int count;
+	enum dma_transaction_type type;
+	uint32_t status;
+};
+
+struct jz4780_dma_chan {
+	struct virt_dma_chan vchan;
+	unsigned int id;
+	struct dma_pool *desc_pool;
+
+	uint32_t transfer_type;
+	uint32_t transfer_shift;
+	struct dma_slave_config	config;
+
+	struct jz4780_dma_desc *desc;
+	unsigned int curr_hwdesc;
+};
+
+struct jz4780_dma_dev {
+	struct dma_device dma_device;
+	void __iomem *base;
+	struct clk *clk;
+	unsigned int irq;
+
+	uint32_t chan_reserved;
+	struct jz4780_dma_chan chan[JZ_DMA_NR_CHANNELS];
+};
+
+struct jz4780_dma_data {
+	uint32_t transfer_type;
+	int channel;
+};
+
+static inline struct jz4780_dma_chan *to_jz4780_dma_chan(struct dma_chan *chan)
+{
+	return container_of(chan, struct jz4780_dma_chan, vchan.chan);
+}
+
+static inline struct jz4780_dma_desc *to_jz4780_dma_desc(
+	struct virt_dma_desc *vdesc)
+{
+	return container_of(vdesc, struct jz4780_dma_desc, vdesc);
+}
+
+static inline struct jz4780_dma_dev *jz4780_dma_chan_parent(
+	struct jz4780_dma_chan *jzchan)
+{
+	return container_of(jzchan->vchan.chan.device, struct jz4780_dma_dev,
+			    dma_device);
+}
+
+static inline uint32_t jz4780_dma_readl(struct jz4780_dma_dev *jzdma,
+	unsigned int reg)
+{
+	return readl(jzdma->base + reg);
+}
+
+static inline void jz4780_dma_writel(struct jz4780_dma_dev *jzdma,
+	unsigned int reg, uint32_t val)
+{
+	writel(val, jzdma->base + reg);
+}
+
+static struct jz4780_dma_desc *jz4780_dma_desc_alloc(
+	struct jz4780_dma_chan *jzchan, unsigned int count,
+	enum dma_transaction_type type)
+{
+	struct jz4780_dma_desc *desc;
+
+	if (count > JZ_DMA_MAX_DESC)
+		return NULL;
+
+	desc = kzalloc(sizeof(*desc), GFP_NOWAIT);
+	if (!desc)
+		return NULL;
+
+	desc->desc = dma_pool_alloc(jzchan->desc_pool, GFP_NOWAIT,
+				    &desc->desc_phys);
+	if (!desc->desc) {
+		kfree(desc);
+		return NULL;
+	}
+
+	desc->count = count;
+	desc->type = type;
+	return desc;
+}
+
+static void jz4780_dma_desc_free(struct virt_dma_desc *vdesc)
+{
+	struct jz4780_dma_desc *desc = to_jz4780_dma_desc(vdesc);
+	struct jz4780_dma_chan *jzchan = to_jz4780_dma_chan(vdesc->tx.chan);
+
+	dma_pool_free(jzchan->desc_pool, desc->desc, desc->desc_phys);
+	kfree(desc);
+}
+
+static uint32_t jz4780_dma_transfer_size(unsigned long val, int *ord)
+{
+	*ord = ffs(val) - 1;
+
+	switch (*ord) {
+	case 0:
+		return JZ_DMA_SIZE_1_BYTE;
+	case 1:
+		return JZ_DMA_SIZE_2_BYTE;
+	case 2:
+		return JZ_DMA_SIZE_4_BYTE;
+	case 4:
+		return JZ_DMA_SIZE_16_BYTE;
+	case 5:
+		return JZ_DMA_SIZE_32_BYTE;
+	case 6:
+		return JZ_DMA_SIZE_64_BYTE;
+	case 7:
+		return JZ_DMA_SIZE_128_BYTE;
+	default:
+		return -EINVAL;
+	}
+}
+
+static uint32_t jz4780_dma_setup_hwdesc(struct jz4780_dma_chan *jzchan,
+	struct jz4780_dma_hwdesc *desc, dma_addr_t addr, size_t len,
+	enum dma_transfer_direction direction)
+{
+	struct dma_slave_config *config = &jzchan->config;
+	uint32_t width, maxburst, tsz;
+	int ord;
+
+	if (direction == DMA_MEM_TO_DEV) {
+		desc->dcm = JZ_DMA_DCM_SAI;
+		desc->dsa = addr;
+		desc->dta = config->dst_addr;
+		desc->drt = jzchan->transfer_type;
+
+		width = config->dst_addr_width;
+		maxburst = config->dst_maxburst;
+	} else {
+		desc->dcm = JZ_DMA_DCM_DAI;
+		desc->dsa = config->src_addr;
+		desc->dta = addr;
+		desc->drt = jzchan->transfer_type;
+
+		width = config->src_addr_width;
+		maxburst = config->src_maxburst;
+	}
+
+	/*
+	 * This calculates the maximum transfer size that can be used with the
+	 * given address, length, width and maximum burst size. The address
+	 * must be aligned to the transfer size, the total length must be
+	 * divisible by the transfer size, and we must not use more than the
+	 * maximum burst specified by the user.
+	 */
+	tsz = jz4780_dma_transfer_size(addr | len | (width * maxburst), &ord);
+	jzchan->transfer_shift = ord;
+
+	switch (width) {
+	case DMA_SLAVE_BUSWIDTH_1_BYTE:
+	case DMA_SLAVE_BUSWIDTH_2_BYTES:
+		break;
+	case DMA_SLAVE_BUSWIDTH_4_BYTES:
+		width = JZ_DMA_WIDTH_32_BIT;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	desc->dcm |= tsz << JZ_DMA_DCM_TSZ_SHIFT;
+	desc->dcm |= width << JZ_DMA_DCM_SP_SHIFT;
+	desc->dcm |= width << JZ_DMA_DCM_DP_SHIFT;
+
+	desc->dtc = len >> ord;
+}
+
+static struct dma_async_tx_descriptor *jz4780_dma_prep_slave_sg(
+	struct dma_chan *chan, struct scatterlist *sgl, unsigned int sg_len,
+	enum dma_transfer_direction direction, unsigned long flags)
+{
+	struct jz4780_dma_chan *jzchan = to_jz4780_dma_chan(chan);
+	struct jz4780_dma_desc *desc;
+	unsigned int i;
+	int err;
+
+	desc = jz4780_dma_desc_alloc(jzchan, sg_len, DMA_SLAVE);
+	if (!desc)
+		return NULL;
+
+	for (i = 0; i < sg_len; i++) {
+		err = jz4780_dma_setup_hwdesc(jzchan, &desc->desc[i],
+					sg_dma_address(&sgl[i]),
+					sg_dma_len(&sgl[i]),
+					direction);
+		if (err < 0)
+			return ERR_PTR(err);
+
+
+		desc->desc[i].dcm |= JZ_DMA_DCM_TIE;
+
+		if (i != (sg_len - 1)) {
+			/* Automatically proceeed to the next descriptor. */
+			desc->desc[i].dcm |= JZ_DMA_DCM_LINK;
+
+			/*
+			 * The upper 8 bits of the DTC field in the descriptor
+			 * must be set to (offset from descriptor base of next
+			 * descriptor >> 4).
+			 */
+			desc->desc[i].dtc |=
+				(((i + 1) * sizeof(*desc->desc)) >> 4) << 24;
+		}
+	}
+
+	return vchan_tx_prep(&jzchan->vchan, &desc->vdesc, flags);
+}
+
+static struct dma_async_tx_descriptor *jz4780_dma_prep_dma_cyclic(
+	struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
+	size_t period_len, enum dma_transfer_direction direction,
+	unsigned long flags)
+{
+	struct jz4780_dma_chan *jzchan = to_jz4780_dma_chan(chan);
+	struct jz4780_dma_desc *desc;
+	unsigned int periods, i;
+	int err;
+
+	if (buf_len % period_len)
+		return NULL;
+
+	periods = buf_len / period_len;
+
+	desc = jz4780_dma_desc_alloc(jzchan, periods, DMA_CYCLIC);
+	if (!desc)
+		return NULL;
+
+	for (i = 0; i < periods; i++) {
+		err = jz4780_dma_setup_hwdesc(jzchan, &desc->desc[i], buf_addr,
+					period_len, direction);
+		if (err < 0)
+			return ERR_PTR(err);
+
+		buf_addr += period_len;
+
+		/*
+		 * Set the link bit to indicate that the controller should
+		 * automatically proceed to the next descriptor. In
+		 * jz4780_dma_begin(), this will be cleared if we need to issue
+		 * an interrupt after each period.
+		 */
+		desc->desc[i].dcm |= JZ_DMA_DCM_TIE | JZ_DMA_DCM_LINK;
+
+		/*
+		 * The upper 8 bits of the DTC field in the descriptor must be
+		 * set to (offset from descriptor base of next descriptor >> 4).
+		 * If this is the last descriptor, link it back to the first,
+		 * i.e. leave offset set to 0, otherwise point to the next one.
+		 */
+		if (i != (periods - 1)) {
+			desc->desc[i].dtc |=
+				(((i + 1) * sizeof(*desc->desc)) >> 4) << 24;
+		}
+	}
+
+	return vchan_tx_prep(&jzchan->vchan, &desc->vdesc, flags);
+}
+
+struct dma_async_tx_descriptor *jz4780_dma_prep_dma_memcpy(
+	struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
+	size_t len, unsigned long flags)
+{
+	struct jz4780_dma_chan *jzchan = to_jz4780_dma_chan(chan);
+	struct jz4780_dma_desc *desc;
+	uint32_t tsz;
+	int ord;
+
+	desc = jz4780_dma_desc_alloc(jzchan, 1, DMA_MEMCPY);
+	if (!desc)
+		return NULL;
+
+	tsz = jz4780_dma_transfer_size(dest | src | len, &ord);
+	if (tsz < 0)
+		return ERR_PTR(tsz);
+
+	desc->desc[0].dsa = src;
+	desc->desc[0].dta = dest;
+	desc->desc[0].drt = JZ_DMA_DRT_AUTO;
+	desc->desc[0].dcm = JZ_DMA_DCM_TIE | JZ_DMA_DCM_SAI | JZ_DMA_DCM_DAI |
+			    tsz << JZ_DMA_DCM_TSZ_SHIFT |
+			    JZ_DMA_WIDTH_32_BIT << JZ_DMA_DCM_SP_SHIFT |
+			    JZ_DMA_WIDTH_32_BIT << JZ_DMA_DCM_DP_SHIFT;
+	desc->desc[0].dtc = len >> ord;
+
+	return vchan_tx_prep(&jzchan->vchan, &desc->vdesc, flags);
+}
+
+static void jz4780_dma_begin(struct jz4780_dma_chan *jzchan)
+{
+	struct jz4780_dma_dev *jzdma = jz4780_dma_chan_parent(jzchan);
+	struct virt_dma_desc *vdesc;
+	unsigned int i;
+	dma_addr_t desc_phys;
+
+	if (!jzchan->desc) {
+		vdesc = vchan_next_desc(&jzchan->vchan);
+		if (!vdesc)
+			return;
+
+		list_del(&vdesc->node);
+
+		jzchan->desc = to_jz4780_dma_desc(vdesc);
+		jzchan->curr_hwdesc = 0;
+
+		if (jzchan->desc->type == DMA_CYCLIC && vdesc->tx.callback) {
+			/*
+			 * The DMA controller doesn't support triggering an
+			 * interrupt after processing each descriptor, only
+			 * after processing an entire terminated list of
+			 * descriptors. For a cyclic DMA setup the list of
+			 * descriptors is not terminated so we can never get an
+			 * interrupt.
+			 *
+			 * If the user requested a callback for a cyclic DMA
+			 * setup then we workaround this hardware limitation
+			 * here by degrading to a set of unlinked descriptors
+			 * which we will submit in sequence in response to the
+			 * completion of processing the previous descriptor.
+			 */
+			for (i = 0; i < jzchan->desc->count; i++)
+				jzchan->desc->desc[i].dcm &= ~JZ_DMA_DCM_LINK;
+		}
+	} else {
+		/*
+		 * There is an existing transfer, therefore this must be one
+		 * for which we unlinked the descriptors above. Advance to the
+		 * next one in the list.
+		 */
+		jzchan->curr_hwdesc =
+			(jzchan->curr_hwdesc + 1) % jzchan->desc->count;
+	}
+
+	/* Use 8-word descriptors. */
+	jz4780_dma_writel(jzdma, JZ_DMA_REG_DCS(jzchan->id), JZ_DMA_DCS_DES8);
+
+	/* Write descriptor address and initiate descriptor fetch. */
+	desc_phys = jzchan->desc->desc_phys +
+		    (jzchan->curr_hwdesc * sizeof(*jzchan->desc->desc));
+	jz4780_dma_writel(jzdma, JZ_DMA_REG_DDA(jzchan->id), desc_phys);
+	jz4780_dma_writel(jzdma, JZ_DMA_REG_DDRS, BIT(jzchan->id));
+
+	/* Enable the channel. */
+	jz4780_dma_writel(jzdma, JZ_DMA_REG_DCS(jzchan->id),
+			  JZ_DMA_DCS_DES8 | JZ_DMA_DCS_CTE);
+}
+
+static void jz4780_dma_issue_pending(struct dma_chan *chan)
+{
+	struct jz4780_dma_chan *jzchan = to_jz4780_dma_chan(chan);
+	unsigned long flags;
+
+	spin_lock_irqsave(&jzchan->vchan.lock, flags);
+
+	if (vchan_issue_pending(&jzchan->vchan) && !jzchan->desc)
+		jz4780_dma_begin(jzchan);
+
+	spin_unlock_irqrestore(&jzchan->vchan.lock, flags);
+}
+
+static int jz4780_dma_terminate_all(struct jz4780_dma_chan *jzchan)
+{
+	struct jz4780_dma_dev *jzdma = jz4780_dma_chan_parent(jzchan);
+	unsigned long flags;
+	LIST_HEAD(head);
+
+	spin_lock_irqsave(&jzchan->vchan.lock, flags);
+
+	/* Clear the DMA status and stop the transfer. */
+	jz4780_dma_writel(jzdma, JZ_DMA_REG_DCS(jzchan->id), 0);
+	if (jzchan->desc) {
+		jz4780_dma_desc_free(&jzchan->desc->vdesc);
+		jzchan->desc = NULL;
+	}
+
+	vchan_get_all_descriptors(&jzchan->vchan, &head);
+
+	spin_unlock_irqrestore(&jzchan->vchan.lock, flags);
+
+	vchan_dma_desc_free_list(&jzchan->vchan, &head);
+	return 0;
+}
+
+static int jz4780_dma_slave_config(struct jz4780_dma_chan *jzchan,
+	const struct dma_slave_config *config)
+{
+	if ((config->src_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES)
+	   || (config->dst_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES))
+		return -EINVAL;
+
+	/* Copy the reset of the slave configuration, it is used later. */
+	memcpy(&jzchan->config, config, sizeof(jzchan->config));
+
+	return 0;
+}
+
+static size_t jz4780_dma_desc_residue(struct jz4780_dma_chan *jzchan,
+	struct jz4780_dma_desc *desc, unsigned int next_sg)
+{
+	struct jz4780_dma_dev *jzdma = jz4780_dma_chan_parent(jzchan);
+	unsigned int residue, count;
+	unsigned int i;
+
+	residue = 0;
+
+	for (i = next_sg; i < desc->count; i++)
+		residue += desc->desc[i].dtc << jzchan->transfer_shift;
+
+	if (next_sg != 0) {
+		count = jz4780_dma_readl(jzdma,
+					 JZ_DMA_REG_DTC(jzchan->id));
+		residue += count << jzchan->transfer_shift;
+	}
+
+	return residue;
+}
+
+static enum dma_status jz4780_dma_tx_status(struct dma_chan *chan,
+	dma_cookie_t cookie, struct dma_tx_state *txstate)
+{
+	struct jz4780_dma_chan *jzchan = to_jz4780_dma_chan(chan);
+	struct virt_dma_desc *vdesc;
+	enum dma_status status;
+	unsigned long flags;
+
+	status = dma_cookie_status(chan, cookie, txstate);
+	if ((status == DMA_COMPLETE) || (txstate == NULL))
+		return status;
+
+	spin_lock_irqsave(&jzchan->vchan.lock, flags);
+
+	vdesc = vchan_find_desc(&jzchan->vchan, cookie);
+	if (vdesc) {
+		/* On the issued list, so hasn't been processed yet */
+		txstate->residue = jz4780_dma_desc_residue(jzchan,
+					to_jz4780_dma_desc(vdesc), 0);
+	} else if (cookie == jzchan->desc->vdesc.tx.cookie) {
+		txstate->residue = jz4780_dma_desc_residue(jzchan, jzchan->desc,
+			  (jzchan->curr_hwdesc + 1) % jzchan->desc->count);
+	} else
+		txstate->residue = 0;
+
+	if (vdesc && jzchan->desc && vdesc == &jzchan->desc->vdesc
+		&& jzchan->desc->status & (JZ_DMA_DCS_AR | JZ_DMA_DCS_HLT))
+			status = DMA_ERROR;
+
+	spin_unlock_irqrestore(&jzchan->vchan.lock, flags);
+	return status;
+}
+
+static void jz4780_dma_chan_irq(struct jz4780_dma_dev *jzdma,
+	struct jz4780_dma_chan *jzchan)
+{
+	uint32_t dcs;
+
+	spin_lock(&jzchan->vchan.lock);
+
+	dcs = jz4780_dma_readl(jzdma, JZ_DMA_REG_DCS(jzchan->id));
+	jz4780_dma_writel(jzdma, JZ_DMA_REG_DCS(jzchan->id), 0);
+
+	if (dcs & JZ_DMA_DCS_AR) {
+		dev_warn(&jzchan->vchan.chan.dev->device,
+			 "address error (DCS=0x%x)\n", dcs);
+	}
+
+	if (dcs & JZ_DMA_DCS_HLT) {
+		dev_warn(&jzchan->vchan.chan.dev->device,
+			 "channel halt (DCS=0x%x)\n", dcs);
+	}
+
+	if (jzchan->desc) {
+		jzchan->desc->status = dcs;
+
+		if ((dcs & (JZ_DMA_DCS_AR | JZ_DMA_DCS_HLT)) == 0) {
+			if (jzchan->desc->type == DMA_CYCLIC) {
+				vchan_cyclic_callback(&jzchan->desc->vdesc);
+			} else {
+				vchan_cookie_complete(&jzchan->desc->vdesc);
+				jzchan->desc = NULL;
+			}
+
+			jz4780_dma_begin(jzchan);
+		}
+	} else {
+		dev_err(&jzchan->vchan.chan.dev->device,
+			"channel IRQ with no active transfer\n");
+	}
+
+	spin_unlock(&jzchan->vchan.lock);
+}
+
+static irqreturn_t jz4780_dma_irq_handler(int irq, void *data)
+{
+	struct jz4780_dma_dev *jzdma = data;
+	uint32_t pending, dmac;
+	int i;
+
+	pending = jz4780_dma_readl(jzdma, JZ_DMA_REG_DIRQP);
+
+	for (i = 0; i < JZ_DMA_NR_CHANNELS; i++) {
+		if (!(pending & (1<<i)))
+			continue;
+
+		jz4780_dma_chan_irq(jzdma, &jzdma->chan[i]);
+	}
+
+	/* Clear halt and address error status of all channels. */
+	dmac = jz4780_dma_readl(jzdma, JZ_DMA_REG_DMAC);
+	dmac &= ~(JZ_DMA_DMAC_HLT | JZ_DMA_DMAC_AR);
+	jz4780_dma_writel(jzdma, JZ_DMA_REG_DMAC, dmac);
+
+	/* Clear interrupt pending status. */
+	jz4780_dma_writel(jzdma, JZ_DMA_REG_DIRQP, 0);
+
+	return IRQ_HANDLED;
+}
+
+static int jz4780_dma_alloc_chan_resources(struct dma_chan *chan)
+{
+	struct jz4780_dma_chan *jzchan = to_jz4780_dma_chan(chan);
+
+	jzchan->desc_pool = dma_pool_create(dev_name(&chan->dev->device),
+					    chan->device->dev,
+					    JZ_DMA_DESC_BLOCK_SIZE,
+					    PAGE_SIZE, 0);
+	if (!jzchan->desc_pool) {
+		dev_err(&chan->dev->device,
+			"failed to allocate descriptor pool\n");
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+static void jz4780_dma_free_chan_resources(struct dma_chan *chan)
+{
+	struct jz4780_dma_chan *jzchan = to_jz4780_dma_chan(chan);
+
+	vchan_free_chan_resources(&jzchan->vchan);
+	dma_pool_destroy(jzchan->desc_pool);
+	jzchan->desc_pool = NULL;
+}
+
+static bool jz4780_dma_filter_fn(struct dma_chan *chan, void *param)
+{
+	struct jz4780_dma_chan *jzchan = to_jz4780_dma_chan(chan);
+	struct jz4780_dma_dev *jzdma = jz4780_dma_chan_parent(jzchan);
+	struct jz4780_dma_data *data = param;
+
+	if (data->channel > -1) {
+		if (data->channel != jzchan->id)
+			return false;
+	} else if (jzdma->chan_reserved & BIT(jzchan->id)) {
+		return false;
+	}
+
+	jzchan->transfer_type = data->transfer_type;
+
+	return true;
+}
+
+static struct dma_chan *jz4780_of_dma_xlate(struct of_phandle_args *dma_spec,
+	struct of_dma *ofdma)
+{
+	struct jz4780_dma_dev *jzdma = ofdma->of_dma_data;
+	dma_cap_mask_t mask = jzdma->dma_device.cap_mask;
+	struct jz4780_dma_data data;
+
+	if (dma_spec->args_count != 2)
+		return NULL;
+
+	data.transfer_type = dma_spec->args[0];
+	data.channel = dma_spec->args[1];
+
+	if (data.channel > -1) {
+		if (data.channel >= JZ_DMA_NR_CHANNELS) {
+			dev_err(jzdma->dma_device.dev,
+				"device requested non-existent channel %u\n",
+				data.channel);
+			return NULL;
+		}
+
+		/* Can only select a channel marked as reserved. */
+		if (!(jzdma->chan_reserved & BIT(data.channel))) {
+			dev_err(jzdma->dma_device.dev,
+				"device requested unreserved channel %u\n",
+				data.channel);
+			return NULL;
+		}
+	}
+
+	return dma_request_channel(mask, jz4780_dma_filter_fn, &data);
+}
+
+static int jz4780_dma_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct jz4780_dma_dev *jzdma;
+	struct jz4780_dma_chan *jzchan;
+	struct dma_device *dd;
+	struct resource *res;
+	int i, ret;
+
+	jzdma = devm_kzalloc(dev, sizeof(*jzdma), GFP_KERNEL);
+	if (!jzdma)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, jzdma);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res) {
+		dev_err(dev, "failed to get I/O memory\n");
+		return -EINVAL;
+	}
+
+	jzdma->base = devm_ioremap_resource(dev, res);
+	if (IS_ERR(jzdma->base))
+		return PTR_ERR(jzdma->base);
+
+	jzdma->irq = platform_get_irq(pdev, 0);
+	if (jzdma->irq < 0) {
+		dev_err(dev, "failed to get IRQ: %d\n", ret);
+		return jzdma->irq;
+	}
+
+	ret = devm_request_irq(dev, jzdma->irq, jz4780_dma_irq_handler, 0,
+			       dev_name(dev), jzdma);
+	if (ret) {
+		dev_err(dev, "failed to request IRQ %u!\n", jzdma->irq);
+		return -EINVAL;
+	}
+
+	jzdma->clk = devm_clk_get(dev, NULL);
+	if (IS_ERR(jzdma->clk)) {
+		dev_err(dev, "failed to get clock\n");
+		return PTR_ERR(jzdma->clk);
+	}
+
+	clk_prepare_enable(jzdma->clk);
+
+	/* Property is optional, if it doesn't exist the value will remain 0. */
+	of_property_read_u32_index(dev->of_node, "ingenic,reserved-channels",
+				   0, &jzdma->chan_reserved);
+
+	dd = &jzdma->dma_device;
+
+	dma_cap_set(DMA_MEMCPY, dd->cap_mask);
+	dma_cap_set(DMA_SLAVE, dd->cap_mask);
+	dma_cap_set(DMA_CYCLIC, dd->cap_mask);
+
+	dd->dev = dev;
+	dd->copy_align = 2; /* 2^2 = 4 byte alignment */
+	dd->device_alloc_chan_resources = jz4780_dma_alloc_chan_resources;
+	dd->device_free_chan_resources = jz4780_dma_free_chan_resources;
+	dd->device_prep_slave_sg = jz4780_dma_prep_slave_sg;
+	dd->device_prep_dma_cyclic = jz4780_dma_prep_dma_cyclic;
+	dd->device_prep_dma_memcpy = jz4780_dma_prep_dma_memcpy;
+	dd->device_config = jz4780_dma_slave_config;
+	dd->device_terminate_all = jz4780_dma_terminate_all;
+	dd->device_tx_status = jz4780_dma_tx_status;
+	dd->device_issue_pending = jz4780_dma_issue_pending;
+	dd->src_addr_widths = JZ_DMA_BUSWIDTHS;
+	dd->dst_addr_widths = JZ_DMA_BUSWIDTHS;
+	dd->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	dd->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+
+
+	/*
+	 * Enable DMA controller, mark all channels as not programmable.
+	 * Also set the FMSC bit - it increases MSC performance, so it makes
+	 * little sense not to enable it.
+	 */
+	jz4780_dma_writel(jzdma, JZ_DMA_REG_DMAC,
+			  JZ_DMA_DMAC_DMAE | JZ_DMA_DMAC_FMSC);
+	jz4780_dma_writel(jzdma, JZ_DMA_REG_DMACP, 0);
+
+	INIT_LIST_HEAD(&dd->channels);
+
+	for (i = 0; i < JZ_DMA_NR_CHANNELS; i++) {
+		jzchan = &jzdma->chan[i];
+		jzchan->id = i;
+
+		vchan_init(&jzchan->vchan, dd);
+		jzchan->vchan.desc_free = jz4780_dma_desc_free;
+	}
+
+	ret = dma_async_device_register(dd);
+	if (ret) {
+		dev_err(dev, "failed to register device\n");
+		goto err_disable_clk;
+	}
+
+	/* Register with OF DMA helpers. */
+	ret = of_dma_controller_register(dev->of_node, jz4780_of_dma_xlate,
+					 jzdma);
+	if (ret) {
+		dev_err(dev, "failed to register OF DMA controller\n");
+		goto err_unregister_dev;
+	}
+
+	dev_info(dev, "JZ4780 DMA controller initialised\n");
+	return 0;
+
+err_unregister_dev:
+	dma_async_device_unregister(dd);
+
+err_disable_clk:
+	clk_disable_unprepare(jzdma->clk);
+	return ret;
+}
+
+static int jz4780_dma_remove(struct platform_device *pdev)
+{
+	struct jz4780_dma_dev *jzdma = platform_get_drvdata(pdev);
+
+	of_dma_controller_free(pdev->dev.of_node);
+	devm_free_irq(&pdev->dev, jzdma->irq, jzdma);
+	dma_async_device_unregister(&jzdma->dma_device);
+	return 0;
+}
+
+static const struct of_device_id jz4780_dma_dt_match[] = {
+	{ .compatible = "ingenic,jz4780-dma", .data = NULL },
+	{},
+};
+MODULE_DEVICE_TABLE(of, jz4780_dma_dt_match);
+
+static struct platform_driver jz4780_dma_driver = {
+	.probe		= jz4780_dma_probe,
+	.remove		= jz4780_dma_remove,
+	.driver	= {
+		.name	= "jz4780-dma",
+		.of_match_table = of_match_ptr(jz4780_dma_dt_match),
+	},
+};
+
+static int __init jz4780_dma_init(void)
+{
+	return platform_driver_register(&jz4780_dma_driver);
+}
+subsys_initcall(jz4780_dma_init);
+
+static void __exit jz4780_dma_exit(void)
+{
+	platform_driver_unregister(&jz4780_dma_driver);
+}
+module_exit(jz4780_dma_exit);
+
+MODULE_AUTHOR("Alex Smith <alex@alex-smith.me.uk>");
+MODULE_DESCRIPTION("Ingenic JZ4780 DMA controller driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index ac336a961dea..0e035a8cf401 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -11,10 +11,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA  02111-1307, USA.
- *
  * The full GNU General Public License is included in this distribution in the
  * file called COPYING.
  */
@@ -355,20 +351,6 @@ struct dma_chan *dma_find_channel(enum dma_transaction_type tx_type)
 }
 EXPORT_SYMBOL(dma_find_channel);
 
-/*
- * net_dma_find_channel - find a channel for net_dma
- * net_dma has alignment requirements
- */
-struct dma_chan *net_dma_find_channel(void)
-{
-	struct dma_chan *chan = dma_find_channel(DMA_MEMCPY);
-	if (chan && !is_dma_copy_aligned(chan->device, 1, 1, 1))
-		return NULL;
-
-	return chan;
-}
-EXPORT_SYMBOL(net_dma_find_channel);
-
 /**
  * dma_issue_pending_all - flush all pending operations across all channels
  */
diff --git a/drivers/dma/dw/Kconfig b/drivers/dma/dw/Kconfig
index dcfe964cc8dc..36e02f0f645e 100644
--- a/drivers/dma/dw/Kconfig
+++ b/drivers/dma/dw/Kconfig
@@ -3,7 +3,7 @@
 #
 
 config DW_DMAC_CORE
-	tristate "Synopsys DesignWare AHB DMA support"
+	tristate
 	select DMA_ENGINE
 
 config DW_DMAC
diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index a8ad05291b27..1022c2e1a2b0 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -230,7 +230,8 @@ static void dwc_dostart(struct dw_dma_chan *dwc, struct dw_desc *first)
 	/* ASSERT:  channel is idle */
 	if (dma_readl(dw, CH_EN) & dwc->mask) {
 		dev_err(chan2dev(&dwc->chan),
-			"BUG: Attempted to start non-idle channel\n");
+			"%s: BUG: Attempted to start non-idle channel\n",
+			__func__);
 		dwc_dump_chan_regs(dwc);
 
 		/* The tasklet will hopefully advance the queue... */
@@ -814,11 +815,8 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 
 slave_sg_todev_fill_desc:
 			desc = dwc_desc_get(dwc);
-			if (!desc) {
-				dev_err(chan2dev(chan),
-					"not enough descriptors available\n");
+			if (!desc)
 				goto err_desc_get;
-			}
 
 			desc->lli.sar = mem;
 			desc->lli.dar = reg;
@@ -874,11 +872,8 @@ slave_sg_todev_fill_desc:
 
 slave_sg_fromdev_fill_desc:
 			desc = dwc_desc_get(dwc);
-			if (!desc) {
-				dev_err(chan2dev(chan),
-						"not enough descriptors available\n");
+			if (!desc)
 				goto err_desc_get;
-			}
 
 			desc->lli.sar = reg;
 			desc->lli.dar = mem;
@@ -922,6 +917,8 @@ slave_sg_fromdev_fill_desc:
 	return &first->txd;
 
 err_desc_get:
+	dev_err(chan2dev(chan),
+		"not enough descriptors available. Direction %d\n", direction);
 	dwc_desc_put(dwc, first);
 	return NULL;
 }
@@ -1261,7 +1258,8 @@ int dw_dma_cyclic_start(struct dma_chan *chan)
 	/* Assert channel is idle */
 	if (dma_readl(dw, CH_EN) & dwc->mask) {
 		dev_err(chan2dev(&dwc->chan),
-			"BUG: Attempted to start non-idle channel\n");
+			"%s: BUG: Attempted to start non-idle channel\n",
+			__func__);
 		dwc_dump_chan_regs(dwc);
 		spin_unlock_irqrestore(&dwc->lock, flags);
 		return -EBUSY;
diff --git a/drivers/dma/edma.c b/drivers/dma/edma.c
index 53dbd3b3384c..bf09db7ca9ee 100644
--- a/drivers/dma/edma.c
+++ b/drivers/dma/edma.c
@@ -812,7 +812,7 @@ static int edma_alloc_chan_resources(struct dma_chan *chan)
 	LIST_HEAD(descs);
 
 	a_ch_num = edma_alloc_channel(echan->ch_num, edma_callback,
-					chan, EVENTQ_DEFAULT);
+					echan, EVENTQ_DEFAULT);
 
 	if (a_ch_num < 0) {
 		ret = -ENODEV;
diff --git a/drivers/dma/fsl_raid.c b/drivers/dma/fsl_raid.c
new file mode 100644
index 000000000000..4d9470f16552
--- /dev/null
+++ b/drivers/dma/fsl_raid.c
@@ -0,0 +1,904 @@
+/*
+ * drivers/dma/fsl_raid.c
+ *
+ * Freescale RAID Engine device driver
+ *
+ * Author:
+ *	Harninder Rai <harninder.rai@freescale.com>
+ *	Naveen Burmi <naveenburmi@freescale.com>
+ *
+ * Rewrite:
+ *	Xuelin Shi <xuelin.shi@freescale.com>
+ *
+ * Copyright (c) 2010-2014 Freescale Semiconductor, Inc.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *     * Redistributions of source code must retain the above copyright
+ *       notice, this list of conditions and the following disclaimer.
+ *     * Redistributions in binary form must reproduce the above copyright
+ *       notice, this list of conditions and the following disclaimer in the
+ *       documentation and/or other materials provided with the distribution.
+ *     * Neither the name of Freescale Semiconductor nor the
+ *       names of its contributors may be used to endorse or promote products
+ *       derived from this software without specific prior written permission.
+ *
+ * ALTERNATIVELY, this software may be distributed under the terms of the
+ * GNU General Public License ("GPL") as published by the Free Software
+ * Foundation, either version 2 of that License or (at your option) any
+ * later version.
+ *
+ * THIS SOFTWARE IS PROVIDED BY Freescale Semiconductor ``AS IS'' AND ANY
+ * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL Freescale Semiconductor BE LIABLE FOR ANY
+ * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
+ * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * Theory of operation:
+ *
+ * General capabilities:
+ *	RAID Engine (RE) block is capable of offloading XOR, memcpy and P/Q
+ *	calculations required in RAID5 and RAID6 operations. RE driver
+ *	registers with Linux's ASYNC layer as dma driver. RE hardware
+ *	maintains strict ordering of the requests through chained
+ *	command queueing.
+ *
+ * Data flow:
+ *	Software RAID layer of Linux (MD layer) maintains RAID partitions,
+ *	strips, stripes etc. It sends requests to the underlying ASYNC layer
+ *	which further passes it to RE driver. ASYNC layer decides which request
+ *	goes to which job ring of RE hardware. For every request processed by
+ *	RAID Engine, driver gets an interrupt unless coalescing is set. The
+ *	per job ring interrupt handler checks the status register for errors,
+ *	clears the interrupt and leave the post interrupt processing to the irq
+ *	thread.
+ */
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/of_irq.h>
+#include <linux/of_address.h>
+#include <linux/of_platform.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmapool.h>
+#include <linux/dmaengine.h>
+#include <linux/io.h>
+#include <linux/spinlock.h>
+#include <linux/slab.h>
+
+#include "dmaengine.h"
+#include "fsl_raid.h"
+
+#define FSL_RE_MAX_XOR_SRCS	16
+#define FSL_RE_MAX_PQ_SRCS	16
+#define FSL_RE_MIN_DESCS	256
+#define FSL_RE_MAX_DESCS	(4 * FSL_RE_MIN_DESCS)
+#define FSL_RE_FRAME_FORMAT	0x1
+#define FSL_RE_MAX_DATA_LEN	(1024*1024)
+
+#define to_fsl_re_dma_desc(tx) container_of(tx, struct fsl_re_desc, async_tx)
+
+/* Add descriptors into per chan software queue - submit_q */
+static dma_cookie_t fsl_re_tx_submit(struct dma_async_tx_descriptor *tx)
+{
+	struct fsl_re_desc *desc;
+	struct fsl_re_chan *re_chan;
+	dma_cookie_t cookie;
+	unsigned long flags;
+
+	desc = to_fsl_re_dma_desc(tx);
+	re_chan = container_of(tx->chan, struct fsl_re_chan, chan);
+
+	spin_lock_irqsave(&re_chan->desc_lock, flags);
+	cookie = dma_cookie_assign(tx);
+	list_add_tail(&desc->node, &re_chan->submit_q);
+	spin_unlock_irqrestore(&re_chan->desc_lock, flags);
+
+	return cookie;
+}
+
+/* Copy descriptor from per chan software queue into hardware job ring */
+static void fsl_re_issue_pending(struct dma_chan *chan)
+{
+	struct fsl_re_chan *re_chan;
+	int avail;
+	struct fsl_re_desc *desc, *_desc;
+	unsigned long flags;
+
+	re_chan = container_of(chan, struct fsl_re_chan, chan);
+
+	spin_lock_irqsave(&re_chan->desc_lock, flags);
+	avail = FSL_RE_SLOT_AVAIL(
+		in_be32(&re_chan->jrregs->inbring_slot_avail));
+
+	list_for_each_entry_safe(desc, _desc, &re_chan->submit_q, node) {
+		if (!avail)
+			break;
+
+		list_move_tail(&desc->node, &re_chan->active_q);
+
+		memcpy(&re_chan->inb_ring_virt_addr[re_chan->inb_count],
+		       &desc->hwdesc, sizeof(struct fsl_re_hw_desc));
+
+		re_chan->inb_count = (re_chan->inb_count + 1) &
+						FSL_RE_RING_SIZE_MASK;
+		out_be32(&re_chan->jrregs->inbring_add_job, FSL_RE_ADD_JOB(1));
+		avail--;
+	}
+	spin_unlock_irqrestore(&re_chan->desc_lock, flags);
+}
+
+static void fsl_re_desc_done(struct fsl_re_desc *desc)
+{
+	dma_async_tx_callback callback;
+	void *callback_param;
+
+	dma_cookie_complete(&desc->async_tx);
+
+	callback = desc->async_tx.callback;
+	callback_param = desc->async_tx.callback_param;
+	if (callback)
+		callback(callback_param);
+
+	dma_descriptor_unmap(&desc->async_tx);
+}
+
+static void fsl_re_cleanup_descs(struct fsl_re_chan *re_chan)
+{
+	struct fsl_re_desc *desc, *_desc;
+	unsigned long flags;
+
+	spin_lock_irqsave(&re_chan->desc_lock, flags);
+	list_for_each_entry_safe(desc, _desc, &re_chan->ack_q, node) {
+		if (async_tx_test_ack(&desc->async_tx))
+			list_move_tail(&desc->node, &re_chan->free_q);
+	}
+	spin_unlock_irqrestore(&re_chan->desc_lock, flags);
+
+	fsl_re_issue_pending(&re_chan->chan);
+}
+
+static void fsl_re_dequeue(unsigned long data)
+{
+	struct fsl_re_chan *re_chan;
+	struct fsl_re_desc *desc, *_desc;
+	struct fsl_re_hw_desc *hwdesc;
+	unsigned long flags;
+	unsigned int count, oub_count;
+	int found;
+
+	re_chan = dev_get_drvdata((struct device *)data);
+
+	fsl_re_cleanup_descs(re_chan);
+
+	spin_lock_irqsave(&re_chan->desc_lock, flags);
+	count =	FSL_RE_SLOT_FULL(in_be32(&re_chan->jrregs->oubring_slot_full));
+	while (count--) {
+		found = 0;
+		hwdesc = &re_chan->oub_ring_virt_addr[re_chan->oub_count];
+		list_for_each_entry_safe(desc, _desc, &re_chan->active_q,
+					 node) {
+			/* compare the hw dma addr to find the completed */
+			if (desc->hwdesc.lbea32 == hwdesc->lbea32 &&
+			    desc->hwdesc.addr_low == hwdesc->addr_low) {
+				found = 1;
+				break;
+			}
+		}
+
+		if (found) {
+			fsl_re_desc_done(desc);
+			list_move_tail(&desc->node, &re_chan->ack_q);
+		} else {
+			dev_err(re_chan->dev,
+				"found hwdesc not in sw queue, discard it\n");
+		}
+
+		oub_count = (re_chan->oub_count + 1) & FSL_RE_RING_SIZE_MASK;
+		re_chan->oub_count = oub_count;
+
+		out_be32(&re_chan->jrregs->oubring_job_rmvd,
+			 FSL_RE_RMVD_JOB(1));
+	}
+	spin_unlock_irqrestore(&re_chan->desc_lock, flags);
+}
+
+/* Per Job Ring interrupt handler */
+static irqreturn_t fsl_re_isr(int irq, void *data)
+{
+	struct fsl_re_chan *re_chan;
+	u32 irqstate, status;
+
+	re_chan = dev_get_drvdata((struct device *)data);
+
+	irqstate = in_be32(&re_chan->jrregs->jr_interrupt_status);
+	if (!irqstate)
+		return IRQ_NONE;
+
+	/*
+	 * There's no way in upper layer (read MD layer) to recover from
+	 * error conditions except restart everything. In long term we
+	 * need to do something more than just crashing
+	 */
+	if (irqstate & FSL_RE_ERROR) {
+		status = in_be32(&re_chan->jrregs->jr_status);
+		dev_err(re_chan->dev, "chan error irqstate: %x, status: %x\n",
+			irqstate, status);
+	}
+
+	/* Clear interrupt */
+	out_be32(&re_chan->jrregs->jr_interrupt_status, FSL_RE_CLR_INTR);
+
+	tasklet_schedule(&re_chan->irqtask);
+
+	return IRQ_HANDLED;
+}
+
+static enum dma_status fsl_re_tx_status(struct dma_chan *chan,
+					dma_cookie_t cookie,
+					struct dma_tx_state *txstate)
+{
+	return dma_cookie_status(chan, cookie, txstate);
+}
+
+static void fill_cfd_frame(struct fsl_re_cmpnd_frame *cf, u8 index,
+			   size_t length, dma_addr_t addr, bool final)
+{
+	u32 efrl = length & FSL_RE_CF_LENGTH_MASK;
+
+	efrl |= final << FSL_RE_CF_FINAL_SHIFT;
+	cf[index].efrl32 = efrl;
+	cf[index].addr_high = upper_32_bits(addr);
+	cf[index].addr_low = lower_32_bits(addr);
+}
+
+static struct fsl_re_desc *fsl_re_init_desc(struct fsl_re_chan *re_chan,
+					    struct fsl_re_desc *desc,
+					    void *cf, dma_addr_t paddr)
+{
+	desc->re_chan = re_chan;
+	desc->async_tx.tx_submit = fsl_re_tx_submit;
+	dma_async_tx_descriptor_init(&desc->async_tx, &re_chan->chan);
+	INIT_LIST_HEAD(&desc->node);
+
+	desc->hwdesc.fmt32 = FSL_RE_FRAME_FORMAT << FSL_RE_HWDESC_FMT_SHIFT;
+	desc->hwdesc.lbea32 = upper_32_bits(paddr);
+	desc->hwdesc.addr_low = lower_32_bits(paddr);
+	desc->cf_addr = cf;
+	desc->cf_paddr = paddr;
+
+	desc->cdb_addr = (void *)(cf + FSL_RE_CF_DESC_SIZE);
+	desc->cdb_paddr = paddr + FSL_RE_CF_DESC_SIZE;
+
+	return desc;
+}
+
+static struct fsl_re_desc *fsl_re_chan_alloc_desc(struct fsl_re_chan *re_chan,
+						  unsigned long flags)
+{
+	struct fsl_re_desc *desc = NULL;
+	void *cf;
+	dma_addr_t paddr;
+	unsigned long lock_flag;
+
+	fsl_re_cleanup_descs(re_chan);
+
+	spin_lock_irqsave(&re_chan->desc_lock, lock_flag);
+	if (!list_empty(&re_chan->free_q)) {
+		/* take one desc from free_q */
+		desc = list_first_entry(&re_chan->free_q,
+					struct fsl_re_desc, node);
+		list_del(&desc->node);
+
+		desc->async_tx.flags = flags;
+	}
+	spin_unlock_irqrestore(&re_chan->desc_lock, lock_flag);
+
+	if (!desc) {
+		desc = kzalloc(sizeof(*desc), GFP_NOWAIT);
+		if (!desc)
+			return NULL;
+
+		cf = dma_pool_alloc(re_chan->re_dev->cf_desc_pool, GFP_NOWAIT,
+				    &paddr);
+		if (!cf) {
+			kfree(desc);
+			return NULL;
+		}
+
+		desc = fsl_re_init_desc(re_chan, desc, cf, paddr);
+		desc->async_tx.flags = flags;
+
+		spin_lock_irqsave(&re_chan->desc_lock, lock_flag);
+		re_chan->alloc_count++;
+		spin_unlock_irqrestore(&re_chan->desc_lock, lock_flag);
+	}
+
+	return desc;
+}
+
+static struct dma_async_tx_descriptor *fsl_re_prep_dma_genq(
+		struct dma_chan *chan, dma_addr_t dest, dma_addr_t *src,
+		unsigned int src_cnt, const unsigned char *scf, size_t len,
+		unsigned long flags)
+{
+	struct fsl_re_chan *re_chan;
+	struct fsl_re_desc *desc;
+	struct fsl_re_xor_cdb *xor;
+	struct fsl_re_cmpnd_frame *cf;
+	u32 cdb;
+	unsigned int i, j;
+	unsigned int save_src_cnt = src_cnt;
+	int cont_q = 0;
+
+	re_chan = container_of(chan, struct fsl_re_chan, chan);
+	if (len > FSL_RE_MAX_DATA_LEN) {
+		dev_err(re_chan->dev, "genq tx length %lu, max length %d\n",
+			len, FSL_RE_MAX_DATA_LEN);
+		return NULL;
+	}
+
+	desc = fsl_re_chan_alloc_desc(re_chan, flags);
+	if (desc <= 0)
+		return NULL;
+
+	if (scf && (flags & DMA_PREP_CONTINUE)) {
+		cont_q = 1;
+		src_cnt += 1;
+	}
+
+	/* Filling xor CDB */
+	cdb = FSL_RE_XOR_OPCODE << FSL_RE_CDB_OPCODE_SHIFT;
+	cdb |= (src_cnt - 1) << FSL_RE_CDB_NRCS_SHIFT;
+	cdb |= FSL_RE_BLOCK_SIZE << FSL_RE_CDB_BLKSIZE_SHIFT;
+	cdb |= FSL_RE_INTR_ON_ERROR << FSL_RE_CDB_ERROR_SHIFT;
+	cdb |= FSL_RE_DATA_DEP << FSL_RE_CDB_DEPEND_SHIFT;
+	xor = desc->cdb_addr;
+	xor->cdb32 = cdb;
+
+	if (scf) {
+		/* compute q = src0*coef0^src1*coef1^..., * is GF(8) mult */
+		for (i = 0; i < save_src_cnt; i++)
+			xor->gfm[i] = scf[i];
+		if (cont_q)
+			xor->gfm[i++] = 1;
+	} else {
+		/* compute P, that is XOR all srcs */
+		for (i = 0; i < src_cnt; i++)
+			xor->gfm[i] = 1;
+	}
+
+	/* Filling frame 0 of compound frame descriptor with CDB */
+	cf = desc->cf_addr;
+	fill_cfd_frame(cf, 0, sizeof(*xor), desc->cdb_paddr, 0);
+
+	/* Fill CFD's 1st frame with dest buffer */
+	fill_cfd_frame(cf, 1, len, dest, 0);
+
+	/* Fill CFD's rest of the frames with source buffers */
+	for (i = 2, j = 0; j < save_src_cnt; i++, j++)
+		fill_cfd_frame(cf, i, len, src[j], 0);
+
+	if (cont_q)
+		fill_cfd_frame(cf, i++, len, dest, 0);
+
+	/* Setting the final bit in the last source buffer frame in CFD */
+	cf[i - 1].efrl32 |= 1 << FSL_RE_CF_FINAL_SHIFT;
+
+	return &desc->async_tx;
+}
+
+/*
+ * Prep function for P parity calculation.In RAID Engine terminology,
+ * XOR calculation is called GenQ calculation done through GenQ command
+ */
+static struct dma_async_tx_descriptor *fsl_re_prep_dma_xor(
+		struct dma_chan *chan, dma_addr_t dest, dma_addr_t *src,
+		unsigned int src_cnt, size_t len, unsigned long flags)
+{
+	/* NULL let genq take all coef as 1 */
+	return fsl_re_prep_dma_genq(chan, dest, src, src_cnt, NULL, len, flags);
+}
+
+/*
+ * Prep function for P/Q parity calculation.In RAID Engine terminology,
+ * P/Q calculation is called GenQQ done through GenQQ command
+ */
+static struct dma_async_tx_descriptor *fsl_re_prep_dma_pq(
+		struct dma_chan *chan, dma_addr_t *dest, dma_addr_t *src,
+		unsigned int src_cnt, const unsigned char *scf, size_t len,
+		unsigned long flags)
+{
+	struct fsl_re_chan *re_chan;
+	struct fsl_re_desc *desc;
+	struct fsl_re_pq_cdb *pq;
+	struct fsl_re_cmpnd_frame *cf;
+	u32 cdb;
+	u8 *p;
+	int gfmq_len, i, j;
+	unsigned int save_src_cnt = src_cnt;
+
+	re_chan = container_of(chan, struct fsl_re_chan, chan);
+	if (len > FSL_RE_MAX_DATA_LEN) {
+		dev_err(re_chan->dev, "pq tx length is %lu, max length is %d\n",
+			len, FSL_RE_MAX_DATA_LEN);
+		return NULL;
+	}
+
+	/*
+	 * RE requires at least 2 sources, if given only one source, we pass the
+	 * second source same as the first one.
+	 * With only one source, generating P is meaningless, only generate Q.
+	 */
+	if (src_cnt == 1) {
+		struct dma_async_tx_descriptor *tx;
+		dma_addr_t dma_src[2];
+		unsigned char coef[2];
+
+		dma_src[0] = *src;
+		coef[0] = *scf;
+		dma_src[1] = *src;
+		coef[1] = 0;
+		tx = fsl_re_prep_dma_genq(chan, dest[1], dma_src, 2, coef, len,
+					  flags);
+		if (tx)
+			desc = to_fsl_re_dma_desc(tx);
+
+		return tx;
+	}
+
+	/*
+	 * During RAID6 array creation, Linux's MD layer gets P and Q
+	 * calculated separately in two steps. But our RAID Engine has
+	 * the capability to calculate both P and Q with a single command
+	 * Hence to merge well with MD layer, we need to provide a hook
+	 * here and call re_jq_prep_dma_genq() function
+	 */
+
+	if (flags & DMA_PREP_PQ_DISABLE_P)
+		return fsl_re_prep_dma_genq(chan, dest[1], src, src_cnt,
+				scf, len, flags);
+
+	if (flags & DMA_PREP_CONTINUE)
+		src_cnt += 3;
+
+	desc = fsl_re_chan_alloc_desc(re_chan, flags);
+	if (desc <= 0)
+		return NULL;
+
+	/* Filling GenQQ CDB */
+	cdb = FSL_RE_PQ_OPCODE << FSL_RE_CDB_OPCODE_SHIFT;
+	cdb |= (src_cnt - 1) << FSL_RE_CDB_NRCS_SHIFT;
+	cdb |= FSL_RE_BLOCK_SIZE << FSL_RE_CDB_BLKSIZE_SHIFT;
+	cdb |= FSL_RE_BUFFER_OUTPUT << FSL_RE_CDB_BUFFER_SHIFT;
+	cdb |= FSL_RE_DATA_DEP << FSL_RE_CDB_DEPEND_SHIFT;
+
+	pq = desc->cdb_addr;
+	pq->cdb32 = cdb;
+
+	p = pq->gfm_q1;
+	/* Init gfm_q1[] */
+	for (i = 0; i < src_cnt; i++)
+		p[i] = 1;
+
+	/* Align gfm[] to 32bit */
+	gfmq_len = ALIGN(src_cnt, 4);
+
+	/* Init gfm_q2[] */
+	p += gfmq_len;
+	for (i = 0; i < src_cnt; i++)
+		p[i] = scf[i];
+
+	/* Filling frame 0 of compound frame descriptor with CDB */
+	cf = desc->cf_addr;
+	fill_cfd_frame(cf, 0, sizeof(struct fsl_re_pq_cdb), desc->cdb_paddr, 0);
+
+	/* Fill CFD's 1st & 2nd frame with dest buffers */
+	for (i = 1, j = 0; i < 3; i++, j++)
+		fill_cfd_frame(cf, i, len, dest[j], 0);
+
+	/* Fill CFD's rest of the frames with source buffers */
+	for (i = 3, j = 0; j < save_src_cnt; i++, j++)
+		fill_cfd_frame(cf, i, len, src[j], 0);
+
+	/* PQ computation continuation */
+	if (flags & DMA_PREP_CONTINUE) {
+		if (src_cnt - save_src_cnt == 3) {
+			p[save_src_cnt] = 0;
+			p[save_src_cnt + 1] = 0;
+			p[save_src_cnt + 2] = 1;
+			fill_cfd_frame(cf, i++, len, dest[0], 0);
+			fill_cfd_frame(cf, i++, len, dest[1], 0);
+			fill_cfd_frame(cf, i++, len, dest[1], 0);
+		} else {
+			dev_err(re_chan->dev, "PQ tx continuation error!\n");
+			return NULL;
+		}
+	}
+
+	/* Setting the final bit in the last source buffer frame in CFD */
+	cf[i - 1].efrl32 |= 1 << FSL_RE_CF_FINAL_SHIFT;
+
+	return &desc->async_tx;
+}
+
+/*
+ * Prep function for memcpy. In RAID Engine, memcpy is done through MOVE
+ * command. Logic of this function will need to be modified once multipage
+ * support is added in Linux's MD/ASYNC Layer
+ */
+static struct dma_async_tx_descriptor *fsl_re_prep_dma_memcpy(
+		struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
+		size_t len, unsigned long flags)
+{
+	struct fsl_re_chan *re_chan;
+	struct fsl_re_desc *desc;
+	size_t length;
+	struct fsl_re_cmpnd_frame *cf;
+	struct fsl_re_move_cdb *move;
+	u32 cdb;
+
+	re_chan = container_of(chan, struct fsl_re_chan, chan);
+
+	if (len > FSL_RE_MAX_DATA_LEN) {
+		dev_err(re_chan->dev, "cp tx length is %lu, max length is %d\n",
+			len, FSL_RE_MAX_DATA_LEN);
+		return NULL;
+	}
+
+	desc = fsl_re_chan_alloc_desc(re_chan, flags);
+	if (desc <= 0)
+		return NULL;
+
+	/* Filling move CDB */
+	cdb = FSL_RE_MOVE_OPCODE << FSL_RE_CDB_OPCODE_SHIFT;
+	cdb |= FSL_RE_BLOCK_SIZE << FSL_RE_CDB_BLKSIZE_SHIFT;
+	cdb |= FSL_RE_INTR_ON_ERROR << FSL_RE_CDB_ERROR_SHIFT;
+	cdb |= FSL_RE_DATA_DEP << FSL_RE_CDB_DEPEND_SHIFT;
+
+	move = desc->cdb_addr;
+	move->cdb32 = cdb;
+
+	/* Filling frame 0 of CFD with move CDB */
+	cf = desc->cf_addr;
+	fill_cfd_frame(cf, 0, sizeof(*move), desc->cdb_paddr, 0);
+
+	length = min_t(size_t, len, FSL_RE_MAX_DATA_LEN);
+
+	/* Fill CFD's 1st frame with dest buffer */
+	fill_cfd_frame(cf, 1, length, dest, 0);
+
+	/* Fill CFD's 2nd frame with src buffer */
+	fill_cfd_frame(cf, 2, length, src, 1);
+
+	return &desc->async_tx;
+}
+
+static int fsl_re_alloc_chan_resources(struct dma_chan *chan)
+{
+	struct fsl_re_chan *re_chan;
+	struct fsl_re_desc *desc;
+	void *cf;
+	dma_addr_t paddr;
+	int i;
+
+	re_chan = container_of(chan, struct fsl_re_chan, chan);
+	for (i = 0; i < FSL_RE_MIN_DESCS; i++) {
+		desc = kzalloc(sizeof(*desc), GFP_KERNEL);
+		if (!desc)
+			break;
+
+		cf = dma_pool_alloc(re_chan->re_dev->cf_desc_pool, GFP_KERNEL,
+				    &paddr);
+		if (!cf) {
+			kfree(desc);
+			break;
+		}
+
+		INIT_LIST_HEAD(&desc->node);
+		fsl_re_init_desc(re_chan, desc, cf, paddr);
+
+		list_add_tail(&desc->node, &re_chan->free_q);
+		re_chan->alloc_count++;
+	}
+	return re_chan->alloc_count;
+}
+
+static void fsl_re_free_chan_resources(struct dma_chan *chan)
+{
+	struct fsl_re_chan *re_chan;
+	struct fsl_re_desc *desc;
+
+	re_chan = container_of(chan, struct fsl_re_chan, chan);
+	while (re_chan->alloc_count--) {
+		desc = list_first_entry(&re_chan->free_q,
+					struct fsl_re_desc,
+					node);
+
+		list_del(&desc->node);
+		dma_pool_free(re_chan->re_dev->cf_desc_pool, desc->cf_addr,
+			      desc->cf_paddr);
+		kfree(desc);
+	}
+
+	if (!list_empty(&re_chan->free_q))
+		dev_err(re_chan->dev, "chan resource cannot be cleaned!\n");
+}
+
+static int fsl_re_chan_probe(struct platform_device *ofdev,
+		      struct device_node *np, u8 q, u32 off)
+{
+	struct device *dev, *chandev;
+	struct fsl_re_drv_private *re_priv;
+	struct fsl_re_chan *chan;
+	struct dma_device *dma_dev;
+	u32 ptr;
+	u32 status;
+	int ret = 0, rc;
+	struct platform_device *chan_ofdev;
+
+	dev = &ofdev->dev;
+	re_priv = dev_get_drvdata(dev);
+	dma_dev = &re_priv->dma_dev;
+
+	chan = devm_kzalloc(dev, sizeof(*chan), GFP_KERNEL);
+	if (!chan)
+		return -ENOMEM;
+
+	/* create platform device for chan node */
+	chan_ofdev = of_platform_device_create(np, NULL, dev);
+	if (!chan_ofdev) {
+		dev_err(dev, "Not able to create ofdev for jr %d\n", q);
+		ret = -EINVAL;
+		goto err_free;
+	}
+
+	/* read reg property from dts */
+	rc = of_property_read_u32(np, "reg", &ptr);
+	if (rc) {
+		dev_err(dev, "Reg property not found in jr %d\n", q);
+		ret = -ENODEV;
+		goto err_free;
+	}
+
+	chan->jrregs = (struct fsl_re_chan_cfg *)((u8 *)re_priv->re_regs +
+			off + ptr);
+
+	/* read irq property from dts */
+	chan->irq = irq_of_parse_and_map(np, 0);
+	if (chan->irq == NO_IRQ) {
+		dev_err(dev, "No IRQ defined for JR %d\n", q);
+		ret = -ENODEV;
+		goto err_free;
+	}
+
+	snprintf(chan->name, sizeof(chan->name), "re_jr%02d", q);
+
+	chandev = &chan_ofdev->dev;
+	tasklet_init(&chan->irqtask, fsl_re_dequeue, (unsigned long)chandev);
+
+	ret = request_irq(chan->irq, fsl_re_isr, 0, chan->name, chandev);
+	if (ret) {
+		dev_err(dev, "Unable to register interrupt for JR %d\n", q);
+		ret = -EINVAL;
+		goto err_free;
+	}
+
+	re_priv->re_jrs[q] = chan;
+	chan->chan.device = dma_dev;
+	chan->chan.private = chan;
+	chan->dev = chandev;
+	chan->re_dev = re_priv;
+
+	spin_lock_init(&chan->desc_lock);
+	INIT_LIST_HEAD(&chan->ack_q);
+	INIT_LIST_HEAD(&chan->active_q);
+	INIT_LIST_HEAD(&chan->submit_q);
+	INIT_LIST_HEAD(&chan->free_q);
+
+	chan->inb_ring_virt_addr = dma_pool_alloc(chan->re_dev->hw_desc_pool,
+		GFP_KERNEL, &chan->inb_phys_addr);
+	if (!chan->inb_ring_virt_addr) {
+		dev_err(dev, "No dma memory for inb_ring_virt_addr\n");
+		ret = -ENOMEM;
+		goto err_free;
+	}
+
+	chan->oub_ring_virt_addr = dma_pool_alloc(chan->re_dev->hw_desc_pool,
+		GFP_KERNEL, &chan->oub_phys_addr);
+	if (!chan->oub_ring_virt_addr) {
+		dev_err(dev, "No dma memory for oub_ring_virt_addr\n");
+		ret = -ENOMEM;
+		goto err_free_1;
+	}
+
+	/* Program the Inbound/Outbound ring base addresses and size */
+	out_be32(&chan->jrregs->inbring_base_h,
+		 chan->inb_phys_addr & FSL_RE_ADDR_BIT_MASK);
+	out_be32(&chan->jrregs->oubring_base_h,
+		 chan->oub_phys_addr & FSL_RE_ADDR_BIT_MASK);
+	out_be32(&chan->jrregs->inbring_base_l,
+		 chan->inb_phys_addr >> FSL_RE_ADDR_BIT_SHIFT);
+	out_be32(&chan->jrregs->oubring_base_l,
+		 chan->oub_phys_addr >> FSL_RE_ADDR_BIT_SHIFT);
+	out_be32(&chan->jrregs->inbring_size,
+		 FSL_RE_RING_SIZE << FSL_RE_RING_SIZE_SHIFT);
+	out_be32(&chan->jrregs->oubring_size,
+		 FSL_RE_RING_SIZE << FSL_RE_RING_SIZE_SHIFT);
+
+	/* Read LIODN value from u-boot */
+	status = in_be32(&chan->jrregs->jr_config_1) & FSL_RE_REG_LIODN_MASK;
+
+	/* Program the CFG reg */
+	out_be32(&chan->jrregs->jr_config_1,
+		 FSL_RE_CFG1_CBSI | FSL_RE_CFG1_CBS0 | status);
+
+	dev_set_drvdata(chandev, chan);
+
+	/* Enable RE/CHAN */
+	out_be32(&chan->jrregs->jr_command, FSL_RE_ENABLE);
+
+	return 0;
+
+err_free_1:
+	dma_pool_free(chan->re_dev->hw_desc_pool, chan->inb_ring_virt_addr,
+		      chan->inb_phys_addr);
+err_free:
+	return ret;
+}
+
+/* Probe function for RAID Engine */
+static int fsl_re_probe(struct platform_device *ofdev)
+{
+	struct fsl_re_drv_private *re_priv;
+	struct device_node *np;
+	struct device_node *child;
+	u32 off;
+	u8 ridx = 0;
+	struct dma_device *dma_dev;
+	struct resource *res;
+	int rc;
+	struct device *dev = &ofdev->dev;
+
+	re_priv = devm_kzalloc(dev, sizeof(*re_priv), GFP_KERNEL);
+	if (!re_priv)
+		return -ENOMEM;
+
+	res = platform_get_resource(ofdev, IORESOURCE_MEM, 0);
+	if (!res)
+		return -ENODEV;
+
+	/* IOMAP the entire RAID Engine region */
+	re_priv->re_regs = devm_ioremap(dev, res->start, resource_size(res));
+	if (!re_priv->re_regs)
+		return -EBUSY;
+
+	/* Program the RE mode */
+	out_be32(&re_priv->re_regs->global_config, FSL_RE_NON_DPAA_MODE);
+
+	/* Program Galois Field polynomial */
+	out_be32(&re_priv->re_regs->galois_field_config, FSL_RE_GFM_POLY);
+
+	dev_info(dev, "version %x, mode %x, gfp %x\n",
+		 in_be32(&re_priv->re_regs->re_version_id),
+		 in_be32(&re_priv->re_regs->global_config),
+		 in_be32(&re_priv->re_regs->galois_field_config));
+
+	dma_dev = &re_priv->dma_dev;
+	dma_dev->dev = dev;
+	INIT_LIST_HEAD(&dma_dev->channels);
+	dma_set_mask(dev, DMA_BIT_MASK(40));
+
+	dma_dev->device_alloc_chan_resources = fsl_re_alloc_chan_resources;
+	dma_dev->device_tx_status = fsl_re_tx_status;
+	dma_dev->device_issue_pending = fsl_re_issue_pending;
+
+	dma_dev->max_xor = FSL_RE_MAX_XOR_SRCS;
+	dma_dev->device_prep_dma_xor = fsl_re_prep_dma_xor;
+	dma_cap_set(DMA_XOR, dma_dev->cap_mask);
+
+	dma_dev->max_pq = FSL_RE_MAX_PQ_SRCS;
+	dma_dev->device_prep_dma_pq = fsl_re_prep_dma_pq;
+	dma_cap_set(DMA_PQ, dma_dev->cap_mask);
+
+	dma_dev->device_prep_dma_memcpy = fsl_re_prep_dma_memcpy;
+	dma_cap_set(DMA_MEMCPY, dma_dev->cap_mask);
+
+	dma_dev->device_free_chan_resources = fsl_re_free_chan_resources;
+
+	re_priv->total_chans = 0;
+
+	re_priv->cf_desc_pool = dmam_pool_create("fsl_re_cf_desc_pool", dev,
+					FSL_RE_CF_CDB_SIZE,
+					FSL_RE_CF_CDB_ALIGN, 0);
+
+	if (!re_priv->cf_desc_pool) {
+		dev_err(dev, "No memory for fsl re_cf desc pool\n");
+		return -ENOMEM;
+	}
+
+	re_priv->hw_desc_pool = dmam_pool_create("fsl_re_hw_desc_pool", dev,
+			sizeof(struct fsl_re_hw_desc) * FSL_RE_RING_SIZE,
+			FSL_RE_FRAME_ALIGN, 0);
+	if (!re_priv->hw_desc_pool) {
+		dev_err(dev, "No memory for fsl re_hw desc pool\n");
+		return -ENOMEM;
+	}
+
+	dev_set_drvdata(dev, re_priv);
+
+	/* Parse Device tree to find out the total number of JQs present */
+	for_each_compatible_node(np, NULL, "fsl,raideng-v1.0-job-queue") {
+		rc = of_property_read_u32(np, "reg", &off);
+		if (rc) {
+			dev_err(dev, "Reg property not found in JQ node\n");
+			return -ENODEV;
+		}
+		/* Find out the Job Rings present under each JQ */
+		for_each_child_of_node(np, child) {
+			rc = of_device_is_compatible(child,
+					     "fsl,raideng-v1.0-job-ring");
+			if (rc) {
+				fsl_re_chan_probe(ofdev, child, ridx++, off);
+				re_priv->total_chans++;
+			}
+		}
+	}
+
+	dma_async_device_register(dma_dev);
+
+	return 0;
+}
+
+static void fsl_re_remove_chan(struct fsl_re_chan *chan)
+{
+	dma_pool_free(chan->re_dev->hw_desc_pool, chan->inb_ring_virt_addr,
+		      chan->inb_phys_addr);
+
+	dma_pool_free(chan->re_dev->hw_desc_pool, chan->oub_ring_virt_addr,
+		      chan->oub_phys_addr);
+}
+
+static int fsl_re_remove(struct platform_device *ofdev)
+{
+	struct fsl_re_drv_private *re_priv;
+	struct device *dev;
+	int i;
+
+	dev = &ofdev->dev;
+	re_priv = dev_get_drvdata(dev);
+
+	/* Cleanup chan related memory areas */
+	for (i = 0; i < re_priv->total_chans; i++)
+		fsl_re_remove_chan(re_priv->re_jrs[i]);
+
+	/* Unregister the driver */
+	dma_async_device_unregister(&re_priv->dma_dev);
+
+	return 0;
+}
+
+static struct of_device_id fsl_re_ids[] = {
+	{ .compatible = "fsl,raideng-v1.0", },
+	{}
+};
+
+static struct platform_driver fsl_re_driver = {
+	.driver = {
+		.name = "fsl-raideng",
+		.owner = THIS_MODULE,
+		.of_match_table = fsl_re_ids,
+	},
+	.probe = fsl_re_probe,
+	.remove = fsl_re_remove,
+};
+
+module_platform_driver(fsl_re_driver);
+
+MODULE_AUTHOR("Harninder Rai <harninder.rai@freescale.com>");
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("Freescale RAID Engine Device Driver");
diff --git a/drivers/dma/fsl_raid.h b/drivers/dma/fsl_raid.h
new file mode 100644
index 000000000000..69d743c04973
--- /dev/null
+++ b/drivers/dma/fsl_raid.h
@@ -0,0 +1,306 @@
+/*
+ * drivers/dma/fsl_raid.h
+ *
+ * Freescale RAID Engine device driver
+ *
+ * Author:
+ *	Harninder Rai <harninder.rai@freescale.com>
+ *	Naveen Burmi <naveenburmi@freescale.com>
+ *
+ * Rewrite:
+ *	Xuelin Shi <xuelin.shi@freescale.com>
+
+ * Copyright (c) 2010-2012 Freescale Semiconductor, Inc.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *     * Redistributions of source code must retain the above copyright
+ *       notice, this list of conditions and the following disclaimer.
+ *     * Redistributions in binary form must reproduce the above copyright
+ *       notice, this list of conditions and the following disclaimer in the
+ *       documentation and/or other materials provided with the distribution.
+ *     * Neither the name of Freescale Semiconductor nor the
+ *       names of its contributors may be used to endorse or promote products
+ *       derived from this software without specific prior written permission.
+ *
+ * ALTERNATIVELY, this software may be distributed under the terms of the
+ * GNU General Public License ("GPL") as published by the Free Software
+ * Foundation, either version 2 of that License or (at your option) any
+ * later version.
+ *
+ * THIS SOFTWARE IS PROVIDED BY Freescale Semiconductor ``AS IS'' AND ANY
+ * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL Freescale Semiconductor BE LIABLE FOR ANY
+ * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
+ * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+#define FSL_RE_MAX_CHANS		4
+#define FSL_RE_DPAA_MODE		BIT(30)
+#define FSL_RE_NON_DPAA_MODE		BIT(31)
+#define FSL_RE_GFM_POLY			0x1d000000
+#define FSL_RE_ADD_JOB(x)		((x) << 16)
+#define FSL_RE_RMVD_JOB(x)		((x) << 16)
+#define FSL_RE_CFG1_CBSI		0x08000000
+#define FSL_RE_CFG1_CBS0		0x00080000
+#define FSL_RE_SLOT_FULL_SHIFT		8
+#define FSL_RE_SLOT_FULL(x)		((x) >> FSL_RE_SLOT_FULL_SHIFT)
+#define FSL_RE_SLOT_AVAIL_SHIFT		8
+#define FSL_RE_SLOT_AVAIL(x)		((x) >> FSL_RE_SLOT_AVAIL_SHIFT)
+#define FSL_RE_PQ_OPCODE		0x1B
+#define FSL_RE_XOR_OPCODE		0x1A
+#define FSL_RE_MOVE_OPCODE		0x8
+#define FSL_RE_FRAME_ALIGN		16
+#define FSL_RE_BLOCK_SIZE		0x3 /* 4096 bytes */
+#define FSL_RE_CACHEABLE_IO		0x0
+#define FSL_RE_BUFFER_OUTPUT		0x0
+#define FSL_RE_INTR_ON_ERROR		0x1
+#define FSL_RE_DATA_DEP			0x1
+#define FSL_RE_ENABLE_DPI		0x0
+#define FSL_RE_RING_SIZE		0x400
+#define FSL_RE_RING_SIZE_MASK		(FSL_RE_RING_SIZE - 1)
+#define FSL_RE_RING_SIZE_SHIFT		8
+#define FSL_RE_ADDR_BIT_SHIFT		4
+#define FSL_RE_ADDR_BIT_MASK		(BIT(FSL_RE_ADDR_BIT_SHIFT) - 1)
+#define FSL_RE_ERROR			0x40000000
+#define FSL_RE_INTR			0x80000000
+#define FSL_RE_CLR_INTR			0x80000000
+#define FSL_RE_PAUSE			0x80000000
+#define FSL_RE_ENABLE			0x80000000
+#define FSL_RE_REG_LIODN_MASK		0x00000FFF
+
+#define FSL_RE_CDB_OPCODE_MASK		0xF8000000
+#define FSL_RE_CDB_OPCODE_SHIFT		27
+#define FSL_RE_CDB_EXCLEN_MASK		0x03000000
+#define FSL_RE_CDB_EXCLEN_SHIFT		24
+#define FSL_RE_CDB_EXCLQ1_MASK		0x00F00000
+#define FSL_RE_CDB_EXCLQ1_SHIFT		20
+#define FSL_RE_CDB_EXCLQ2_MASK		0x000F0000
+#define FSL_RE_CDB_EXCLQ2_SHIFT		16
+#define FSL_RE_CDB_BLKSIZE_MASK		0x0000C000
+#define FSL_RE_CDB_BLKSIZE_SHIFT	14
+#define FSL_RE_CDB_CACHE_MASK		0x00003000
+#define FSL_RE_CDB_CACHE_SHIFT		12
+#define FSL_RE_CDB_BUFFER_MASK		0x00000800
+#define FSL_RE_CDB_BUFFER_SHIFT		11
+#define FSL_RE_CDB_ERROR_MASK		0x00000400
+#define FSL_RE_CDB_ERROR_SHIFT		10
+#define FSL_RE_CDB_NRCS_MASK		0x0000003C
+#define FSL_RE_CDB_NRCS_SHIFT		6
+#define FSL_RE_CDB_DEPEND_MASK		0x00000008
+#define FSL_RE_CDB_DEPEND_SHIFT		3
+#define FSL_RE_CDB_DPI_MASK		0x00000004
+#define FSL_RE_CDB_DPI_SHIFT		2
+
+/*
+ * the largest cf block is 19*sizeof(struct cmpnd_frame), which is 304 bytes.
+ * here 19 = 1(cdb)+2(dest)+16(src), align to 64bytes, that is 320 bytes.
+ * the largest cdb block: struct pq_cdb which is 180 bytes, adding to cf block
+ * 320+180=500, align to 64bytes, that is 512 bytes.
+ */
+#define FSL_RE_CF_DESC_SIZE		320
+#define FSL_RE_CF_CDB_SIZE		512
+#define FSL_RE_CF_CDB_ALIGN		64
+
+struct fsl_re_ctrl {
+	/* General Configuration Registers */
+	__be32 global_config;	/* Global Configuration Register */
+	u8     rsvd1[4];
+	__be32 galois_field_config; /* Galois Field Configuration Register */
+	u8     rsvd2[4];
+	__be32 jq_wrr_config;   /* WRR Configuration register */
+	u8     rsvd3[4];
+	__be32 crc_config;	/* CRC Configuration register */
+	u8     rsvd4[228];
+	__be32 system_reset;	/* System Reset Register */
+	u8     rsvd5[252];
+	__be32 global_status;	/* Global Status Register */
+	u8     rsvd6[832];
+	__be32 re_liodn_base;	/* LIODN Base Register */
+	u8     rsvd7[1712];
+	__be32 re_version_id;	/* Version ID register of RE */
+	__be32 re_version_id_2; /* Version ID 2 register of RE */
+	u8     rsvd8[512];
+	__be32 host_config;	/* Host I/F Configuration Register */
+};
+
+struct fsl_re_chan_cfg {
+	/* Registers for JR interface */
+	__be32 jr_config_0;	/* Job Queue Configuration 0 Register */
+	__be32 jr_config_1;	/* Job Queue Configuration 1 Register */
+	__be32 jr_interrupt_status; /* Job Queue Interrupt Status Register */
+	u8     rsvd1[4];
+	__be32 jr_command;	/* Job Queue Command Register */
+	u8     rsvd2[4];
+	__be32 jr_status;	/* Job Queue Status Register */
+	u8     rsvd3[228];
+
+	/* Input Ring */
+	__be32 inbring_base_h;	/* Inbound Ring Base Address Register - High */
+	__be32 inbring_base_l;	/* Inbound Ring Base Address Register - Low */
+	__be32 inbring_size;	/* Inbound Ring Size Register */
+	u8     rsvd4[4];
+	__be32 inbring_slot_avail; /* Inbound Ring Slot Available Register */
+	u8     rsvd5[4];
+	__be32 inbring_add_job;	/* Inbound Ring Add Job Register */
+	u8     rsvd6[4];
+	__be32 inbring_cnsmr_indx; /* Inbound Ring Consumer Index Register */
+	u8     rsvd7[220];
+
+	/* Output Ring */
+	__be32 oubring_base_h;	/* Outbound Ring Base Address Register - High */
+	__be32 oubring_base_l;	/* Outbound Ring Base Address Register - Low */
+	__be32 oubring_size;	/* Outbound Ring Size Register */
+	u8     rsvd8[4];
+	__be32 oubring_job_rmvd; /* Outbound Ring Job Removed Register */
+	u8     rsvd9[4];
+	__be32 oubring_slot_full; /* Outbound Ring Slot Full Register */
+	u8     rsvd10[4];
+	__be32 oubring_prdcr_indx; /* Outbound Ring Producer Index */
+};
+
+/*
+ * Command Descriptor Block (CDB) for unicast move command.
+ * In RAID Engine terms, memcpy is done through move command
+ */
+struct fsl_re_move_cdb {
+	__be32 cdb32;
+};
+
+/* Data protection/integrity related fields */
+#define FSL_RE_DPI_APPS_MASK		0xC0000000
+#define FSL_RE_DPI_APPS_SHIFT		30
+#define FSL_RE_DPI_REF_MASK		0x30000000
+#define FSL_RE_DPI_REF_SHIFT		28
+#define FSL_RE_DPI_GUARD_MASK		0x0C000000
+#define FSL_RE_DPI_GUARD_SHIFT		26
+#define FSL_RE_DPI_ATTR_MASK		0x03000000
+#define FSL_RE_DPI_ATTR_SHIFT		24
+#define FSL_RE_DPI_META_MASK		0x0000FFFF
+
+struct fsl_re_dpi {
+	__be32 dpi32;
+	__be32 ref;
+};
+
+/*
+ * CDB for GenQ command. In RAID Engine terminology, XOR is
+ * done through this command
+ */
+struct fsl_re_xor_cdb {
+	__be32 cdb32;
+	u8 gfm[16];
+	struct fsl_re_dpi dpi_dest_spec;
+	struct fsl_re_dpi dpi_src_spec[16];
+};
+
+/* CDB for no-op command */
+struct fsl_re_noop_cdb {
+	__be32 cdb32;
+};
+
+/*
+ * CDB for GenQQ command. In RAID Engine terminology, P/Q is
+ * done through this command
+ */
+struct fsl_re_pq_cdb {
+	__be32 cdb32;
+	u8 gfm_q1[16];
+	u8 gfm_q2[16];
+	struct fsl_re_dpi dpi_dest_spec[2];
+	struct fsl_re_dpi dpi_src_spec[16];
+};
+
+/* Compound frame */
+#define FSL_RE_CF_ADDR_HIGH_MASK	0x000000FF
+#define FSL_RE_CF_EXT_MASK		0x80000000
+#define FSL_RE_CF_EXT_SHIFT		31
+#define FSL_RE_CF_FINAL_MASK		0x40000000
+#define FSL_RE_CF_FINAL_SHIFT		30
+#define FSL_RE_CF_LENGTH_MASK		0x000FFFFF
+#define FSL_RE_CF_BPID_MASK		0x00FF0000
+#define FSL_RE_CF_BPID_SHIFT		16
+#define FSL_RE_CF_OFFSET_MASK		0x00001FFF
+
+struct fsl_re_cmpnd_frame {
+	__be32 addr_high;
+	__be32 addr_low;
+	__be32 efrl32;
+	__be32 rbro32;
+};
+
+/* Frame descriptor */
+#define FSL_RE_HWDESC_LIODN_MASK	0x3F000000
+#define FSL_RE_HWDESC_LIODN_SHIFT	24
+#define FSL_RE_HWDESC_BPID_MASK		0x00FF0000
+#define FSL_RE_HWDESC_BPID_SHIFT	16
+#define FSL_RE_HWDESC_ELIODN_MASK	0x0000F000
+#define FSL_RE_HWDESC_ELIODN_SHIFT	12
+#define FSL_RE_HWDESC_FMT_SHIFT		29
+#define FSL_RE_HWDESC_FMT_MASK		(0x3 << FSL_RE_HWDESC_FMT_SHIFT)
+
+struct fsl_re_hw_desc {
+	__be32 lbea32;
+	__be32 addr_low;
+	__be32 fmt32;
+	__be32 status;
+};
+
+/* Raid Engine device private data */
+struct fsl_re_drv_private {
+	u8 total_chans;
+	struct dma_device dma_dev;
+	struct fsl_re_ctrl *re_regs;
+	struct fsl_re_chan *re_jrs[FSL_RE_MAX_CHANS];
+	struct dma_pool *cf_desc_pool;
+	struct dma_pool *hw_desc_pool;
+};
+
+/* Per job ring data structure */
+struct fsl_re_chan {
+	char name[16];
+	spinlock_t desc_lock; /* queue lock */
+	struct list_head ack_q;  /* wait to acked queue */
+	struct list_head active_q; /* already issued on hw, not completed */
+	struct list_head submit_q;
+	struct list_head free_q; /* alloc available queue */
+	struct device *dev;
+	struct fsl_re_drv_private *re_dev;
+	struct dma_chan chan;
+	struct fsl_re_chan_cfg *jrregs;
+	int irq;
+	struct tasklet_struct irqtask;
+	u32 alloc_count;
+
+	/* hw descriptor ring for inbound queue*/
+	dma_addr_t inb_phys_addr;
+	struct fsl_re_hw_desc *inb_ring_virt_addr;
+	u32 inb_count;
+
+	/* hw descriptor ring for outbound queue */
+	dma_addr_t oub_phys_addr;
+	struct fsl_re_hw_desc *oub_ring_virt_addr;
+	u32 oub_count;
+};
+
+/* Async transaction descriptor */
+struct fsl_re_desc {
+	struct dma_async_tx_descriptor async_tx;
+	struct list_head node;
+	struct fsl_re_hw_desc hwdesc;
+	struct fsl_re_chan *re_chan;
+
+	/* hwdesc will point to cf_addr */
+	void *cf_addr;
+	dma_addr_t cf_paddr;
+
+	void *cdb_addr;
+	dma_addr_t cdb_paddr;
+	int status;
+};
diff --git a/drivers/dma/img-mdc-dma.c b/drivers/dma/img-mdc-dma.c
index ed045a9ad634..9ca56830cc63 100644
--- a/drivers/dma/img-mdc-dma.c
+++ b/drivers/dma/img-mdc-dma.c
@@ -689,11 +689,6 @@ static int mdc_slave_config(struct dma_chan *chan,
 	return 0;
 }
 
-static int mdc_alloc_chan_resources(struct dma_chan *chan)
-{
-	return 0;
-}
-
 static void mdc_free_chan_resources(struct dma_chan *chan)
 {
 	struct mdc_chan *mchan = to_mdc_chan(chan);
@@ -910,7 +905,6 @@ static int mdc_dma_probe(struct platform_device *pdev)
 	mdma->dma_dev.device_prep_slave_sg = mdc_prep_slave_sg;
 	mdma->dma_dev.device_prep_dma_cyclic = mdc_prep_dma_cyclic;
 	mdma->dma_dev.device_prep_dma_memcpy = mdc_prep_dma_memcpy;
-	mdma->dma_dev.device_alloc_chan_resources = mdc_alloc_chan_resources;
 	mdma->dma_dev.device_free_chan_resources = mdc_free_chan_resources;
 	mdma->dma_dev.device_tx_status = mdc_tx_status;
 	mdma->dma_dev.device_issue_pending = mdc_issue_pending;
diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c
index 66a0efb9651d..62bbd79338e0 100644
--- a/drivers/dma/imx-sdma.c
+++ b/drivers/dma/imx-sdma.c
@@ -1260,6 +1260,7 @@ static void sdma_issue_pending(struct dma_chan *chan)
 
 #define SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V1	34
 #define SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V2	38
+#define SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V3	41
 
 static void sdma_add_scripts(struct sdma_engine *sdma,
 		const struct sdma_script_start_addrs *addr)
@@ -1306,6 +1307,9 @@ static void sdma_load_firmware(const struct firmware *fw, void *context)
 	case 2:
 		sdma->script_number = SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V2;
 		break;
+	case 3:
+		sdma->script_number = SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V3;
+		break;
 	default:
 		dev_err(sdma->dev, "unknown firmware version\n");
 		goto err_firmware;
diff --git a/drivers/dma/ioat/dca.c b/drivers/dma/ioat/dca.c
index 3b55bb8d969a..ea1e107ae884 100644
--- a/drivers/dma/ioat/dca.c
+++ b/drivers/dma/ioat/dca.c
@@ -11,10 +11,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc.,
- * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
- *
  * The full GNU General Public License is included in this distribution in
  * the file called "COPYING".
  *
diff --git a/drivers/dma/ioat/dma.c b/drivers/dma/ioat/dma.c
index 940c1502a8b5..ee0aa9f4ccfa 100644
--- a/drivers/dma/ioat/dma.c
+++ b/drivers/dma/ioat/dma.c
@@ -11,10 +11,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc.,
- * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
- *
  * The full GNU General Public License is included in this distribution in
  * the file called "COPYING".
  *
diff --git a/drivers/dma/ioat/dma.h b/drivers/dma/ioat/dma.h
index d63f68b1aa35..30f5c7eede16 100644
--- a/drivers/dma/ioat/dma.h
+++ b/drivers/dma/ioat/dma.h
@@ -11,10 +11,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA  02111-1307, USA.
- *
  * The full GNU General Public License is included in this distribution in the
  * file called COPYING.
  */
diff --git a/drivers/dma/ioat/dma_v2.c b/drivers/dma/ioat/dma_v2.c
index 695483e6be32..69c7dfcad023 100644
--- a/drivers/dma/ioat/dma_v2.c
+++ b/drivers/dma/ioat/dma_v2.c
@@ -11,10 +11,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc.,
- * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
- *
  * The full GNU General Public License is included in this distribution in
  * the file called "COPYING".
  *
diff --git a/drivers/dma/ioat/dma_v2.h b/drivers/dma/ioat/dma_v2.h
index 470292767e68..bf24ebe874b0 100644
--- a/drivers/dma/ioat/dma_v2.h
+++ b/drivers/dma/ioat/dma_v2.h
@@ -11,10 +11,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA  02111-1307, USA.
- *
  * The full GNU General Public License is included in this distribution in the
  * file called COPYING.
  */
diff --git a/drivers/dma/ioat/dma_v3.c b/drivers/dma/ioat/dma_v3.c
index 194ec20c9408..64790a45ef5d 100644
--- a/drivers/dma/ioat/dma_v3.c
+++ b/drivers/dma/ioat/dma_v3.c
@@ -15,10 +15,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc.,
- * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
- *
  * The full GNU General Public License is included in this distribution in
  * the file called "COPYING".
  *
diff --git a/drivers/dma/ioat/hw.h b/drivers/dma/ioat/hw.h
index 02177ecf09f8..a3e731edce57 100644
--- a/drivers/dma/ioat/hw.h
+++ b/drivers/dma/ioat/hw.h
@@ -11,10 +11,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA  02111-1307, USA.
- *
  * The full GNU General Public License is included in this distribution in the
  * file called COPYING.
  */
diff --git a/drivers/dma/ioat/pci.c b/drivers/dma/ioat/pci.c
index 5501eb072d69..76f0dc688a19 100644
--- a/drivers/dma/ioat/pci.c
+++ b/drivers/dma/ioat/pci.c
@@ -11,10 +11,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc.,
- * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
- *
  * The full GNU General Public License is included in this distribution in
  * the file called "COPYING".
  *
diff --git a/drivers/dma/ioat/registers.h b/drivers/dma/ioat/registers.h
index 2f1cfa0f1f47..909352f74c89 100644
--- a/drivers/dma/ioat/registers.h
+++ b/drivers/dma/ioat/registers.h
@@ -11,10 +11,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA  02111-1307, USA.
- *
  * The full GNU General Public License is included in this distribution in the
  * file called COPYING.
  */
diff --git a/drivers/dma/iop-adma.c b/drivers/dma/iop-adma.c
index 263d9f6a207e..998826854fdd 100644
--- a/drivers/dma/iop-adma.c
+++ b/drivers/dma/iop-adma.c
@@ -11,10 +11,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc.,
- * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
- *
  */
 
 /*
diff --git a/drivers/dma/k3dma.c b/drivers/dma/k3dma.c
index 6f7f43529ccb..647e362f01fd 100644
--- a/drivers/dma/k3dma.c
+++ b/drivers/dma/k3dma.c
@@ -313,11 +313,6 @@ static void k3_dma_tasklet(unsigned long arg)
 	}
 }
 
-static int k3_dma_alloc_chan_resources(struct dma_chan *chan)
-{
-	return 0;
-}
-
 static void k3_dma_free_chan_resources(struct dma_chan *chan)
 {
 	struct k3_dma_chan *c = to_k3_chan(chan);
@@ -654,7 +649,7 @@ static void k3_dma_free_desc(struct virt_dma_desc *vd)
 	kfree(ds);
 }
 
-static struct of_device_id k3_pdma_dt_ids[] = {
+static const struct of_device_id k3_pdma_dt_ids[] = {
 	{ .compatible = "hisilicon,k3-dma-1.0", },
 	{}
 };
@@ -728,7 +723,6 @@ static int k3_dma_probe(struct platform_device *op)
 	dma_cap_set(DMA_SLAVE, d->slave.cap_mask);
 	dma_cap_set(DMA_MEMCPY, d->slave.cap_mask);
 	d->slave.dev = &op->dev;
-	d->slave.device_alloc_chan_resources = k3_dma_alloc_chan_resources;
 	d->slave.device_free_chan_resources = k3_dma_free_chan_resources;
 	d->slave.device_tx_status = k3_dma_tx_status;
 	d->slave.device_prep_dma_memcpy = k3_dma_prep_memcpy;
diff --git a/drivers/dma/mmp_pdma.c b/drivers/dma/mmp_pdma.c
index eb410044e1af..462a0229a743 100644
--- a/drivers/dma/mmp_pdma.c
+++ b/drivers/dma/mmp_pdma.c
@@ -973,7 +973,7 @@ static int mmp_pdma_chan_init(struct mmp_pdma_device *pdev, int idx, int irq)
 	return 0;
 }
 
-static struct of_device_id mmp_pdma_dt_ids[] = {
+static const struct of_device_id mmp_pdma_dt_ids[] = {
 	{ .compatible = "marvell,pdma-1.0", },
 	{}
 };
diff --git a/drivers/dma/mmp_tdma.c b/drivers/dma/mmp_tdma.c
index b6f4e1fc9c78..449e785def17 100644
--- a/drivers/dma/mmp_tdma.c
+++ b/drivers/dma/mmp_tdma.c
@@ -613,7 +613,7 @@ struct dma_chan *mmp_tdma_xlate(struct of_phandle_args *dma_spec,
 	return dma_request_channel(mask, mmp_tdma_filter_fn, &param);
 }
 
-static struct of_device_id mmp_tdma_dt_ids[] = {
+static const struct of_device_id mmp_tdma_dt_ids[] = {
 	{ .compatible = "marvell,adma-1.0", .data = (void *)MMP_AUD_TDMA},
 	{ .compatible = "marvell,pxa910-squ", .data = (void *)PXA910_SQU},
 	{}
diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c
index 57d2457545f3..e6281e7aa46e 100644
--- a/drivers/dma/mpc512x_dma.c
+++ b/drivers/dma/mpc512x_dma.c
@@ -21,10 +21,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA  02111-1307, USA.
- *
  * The full GNU General Public License is included in this distribution in the
  * file called COPYING.
  */
@@ -1072,7 +1068,7 @@ static int mpc_dma_remove(struct platform_device *op)
 	return 0;
 }
 
-static struct of_device_id mpc_dma_match[] = {
+static const struct of_device_id mpc_dma_match[] = {
 	{ .compatible = "fsl,mpc5121-dma", },
 	{ .compatible = "fsl,mpc8308-dma", },
 	{},
diff --git a/drivers/dma/mv_xor.c b/drivers/dma/mv_xor.c
index b03e8137b918..1c56001df676 100644
--- a/drivers/dma/mv_xor.c
+++ b/drivers/dma/mv_xor.c
@@ -10,10 +10,6 @@
  * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc.,
- * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
  */
 
 #include <linux/init.h>
@@ -1249,7 +1245,7 @@ static int mv_xor_remove(struct platform_device *pdev)
 }
 
 #ifdef CONFIG_OF
-static struct of_device_id mv_xor_dt_ids[] = {
+static const struct of_device_id mv_xor_dt_ids[] = {
        { .compatible = "marvell,orion-xor", },
        {},
 };
diff --git a/drivers/dma/mv_xor.h b/drivers/dma/mv_xor.h
index 78edc7e44569..91958dba39a2 100644
--- a/drivers/dma/mv_xor.h
+++ b/drivers/dma/mv_xor.h
@@ -9,10 +9,6 @@
  * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
  * for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software Foundation,
- * Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
  */
 
 #ifndef MV_XOR_H
diff --git a/drivers/dma/pch_dma.c b/drivers/dma/pch_dma.c
index 35c143cb88da..b859792dde95 100644
--- a/drivers/dma/pch_dma.c
+++ b/drivers/dma/pch_dma.c
@@ -949,6 +949,7 @@ err_free_res:
 err_disable_pdev:
 	pci_disable_device(pdev);
 err_free_mem:
+	kfree(pd);
 	return err;
 }
 
diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c
index 0e1f56772855..a7d9d3029b14 100644
--- a/drivers/dma/pl330.c
+++ b/drivers/dma/pl330.c
@@ -556,7 +556,7 @@ static inline u32 _emit_ADDH(unsigned dry_run, u8 buf[],
 
 	buf[0] = CMD_DMAADDH;
 	buf[0] |= (da << 1);
-	*((u16 *)&buf[1]) = val;
+	*((__le16 *)&buf[1]) = cpu_to_le16(val);
 
 	PL330_DBGCMD_DUMP(SZ_DMAADDH, "\tDMAADDH %s %u\n",
 		da == 1 ? "DA" : "SA", val);
@@ -710,7 +710,7 @@ static inline u32 _emit_MOV(unsigned dry_run, u8 buf[],
 
 	buf[0] = CMD_DMAMOV;
 	buf[1] = dst;
-	*((u32 *)&buf[2]) = val;
+	*((__le32 *)&buf[2]) = cpu_to_le32(val);
 
 	PL330_DBGCMD_DUMP(SZ_DMAMOV, "\tDMAMOV %s 0x%x\n",
 		dst == SAR ? "SAR" : (dst == DAR ? "DAR" : "CCR"), val);
@@ -888,7 +888,7 @@ static inline u32 _emit_GO(unsigned dry_run, u8 buf[],
 
 	buf[1] = chan & 0x7;
 
-	*((u32 *)&buf[2]) = addr;
+	*((__le32 *)&buf[2]) = cpu_to_le32(addr);
 
 	return SZ_DMAGO;
 }
@@ -928,7 +928,7 @@ static inline void _execute_DBGINSN(struct pl330_thread *thrd,
 	}
 	writel(val, regs + DBGINST0);
 
-	val = *((u32 *)&insn[2]);
+	val = le32_to_cpu(*((__le32 *)&insn[2]));
 	writel(val, regs + DBGINST1);
 
 	/* If timed out due to halted state-machine */
@@ -2162,7 +2162,7 @@ static int pl330_terminate_all(struct dma_chan *chan)
  * DMA transfer again. This pause feature was implemented to
  * allow safely read residue before channel termination.
  */
-int pl330_pause(struct dma_chan *chan)
+static int pl330_pause(struct dma_chan *chan)
 {
 	struct dma_pl330_chan *pch = to_pchan(chan);
 	struct pl330_dmac *pl330 = pch->dmac;
@@ -2203,8 +2203,8 @@ static void pl330_free_chan_resources(struct dma_chan *chan)
 	pm_runtime_put_autosuspend(pch->dmac->ddma.dev);
 }
 
-int pl330_get_current_xferred_count(struct dma_pl330_chan *pch,
-		struct dma_pl330_desc *desc)
+static int pl330_get_current_xferred_count(struct dma_pl330_chan *pch,
+					   struct dma_pl330_desc *desc)
 {
 	struct pl330_thread *thrd = pch->thread;
 	struct pl330_dmac *pl330 = pch->dmac;
@@ -2259,7 +2259,17 @@ pl330_tx_status(struct dma_chan *chan, dma_cookie_t cookie,
 			transferred = 0;
 		residual += desc->bytes_requested - transferred;
 		if (desc->txd.cookie == cookie) {
-			ret = desc->status;
+			switch (desc->status) {
+			case DONE:
+				ret = DMA_COMPLETE;
+				break;
+			case PREP:
+			case BUSY:
+				ret = DMA_IN_PROGRESS;
+				break;
+			default:
+				WARN_ON(1);
+			}
 			break;
 		}
 		if (desc->last)
diff --git a/drivers/dma/ppc4xx/adma.c b/drivers/dma/ppc4xx/adma.c
index fa764a39cd36..9217f893b0d1 100644
--- a/drivers/dma/ppc4xx/adma.c
+++ b/drivers/dma/ppc4xx/adma.c
@@ -16,10 +16,6 @@
  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  * more details.
  *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA  02111-1307, USA.
- *
  * The full GNU General Public License is included in this distribution in the
  * file called COPYING.
  */
diff --git a/drivers/dma/qcom_bam_dma.c b/drivers/dma/qcom_bam_dma.c
index 9c914d625906..5a250cdc8376 100644
--- a/drivers/dma/qcom_bam_dma.c
+++ b/drivers/dma/qcom_bam_dma.c
@@ -171,6 +171,35 @@ static const struct reg_offset_data bam_v1_4_reg_info[] = {
 	[BAM_P_FIFO_SIZES]	= { 0x1820, 0x00, 0x1000, 0x00 },
 };
 
+static const struct reg_offset_data bam_v1_7_reg_info[] = {
+	[BAM_CTRL]		= { 0x00000, 0x00, 0x00, 0x00 },
+	[BAM_REVISION]		= { 0x01000, 0x00, 0x00, 0x00 },
+	[BAM_NUM_PIPES]		= { 0x01008, 0x00, 0x00, 0x00 },
+	[BAM_DESC_CNT_TRSHLD]	= { 0x00008, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS]		= { 0x03010, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS_MSK]	= { 0x03014, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS_UNMASKED]	= { 0x03018, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_STTS]		= { 0x00014, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_CLR]		= { 0x00018, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_EN]		= { 0x0001C, 0x00, 0x00, 0x00 },
+	[BAM_CNFG_BITS]		= { 0x0007C, 0x00, 0x00, 0x00 },
+	[BAM_IRQ_SRCS_EE]	= { 0x03000, 0x00, 0x00, 0x1000 },
+	[BAM_IRQ_SRCS_MSK_EE]	= { 0x03004, 0x00, 0x00, 0x1000 },
+	[BAM_P_CTRL]		= { 0x13000, 0x1000, 0x00, 0x00 },
+	[BAM_P_RST]		= { 0x13004, 0x1000, 0x00, 0x00 },
+	[BAM_P_HALT]		= { 0x13008, 0x1000, 0x00, 0x00 },
+	[BAM_P_IRQ_STTS]	= { 0x13010, 0x1000, 0x00, 0x00 },
+	[BAM_P_IRQ_CLR]		= { 0x13014, 0x1000, 0x00, 0x00 },
+	[BAM_P_IRQ_EN]		= { 0x13018, 0x1000, 0x00, 0x00 },
+	[BAM_P_EVNT_DEST_ADDR]	= { 0x1382C, 0x00, 0x1000, 0x00 },
+	[BAM_P_EVNT_REG]	= { 0x13818, 0x00, 0x1000, 0x00 },
+	[BAM_P_SW_OFSTS]	= { 0x13800, 0x00, 0x1000, 0x00 },
+	[BAM_P_DATA_FIFO_ADDR]	= { 0x13824, 0x00, 0x1000, 0x00 },
+	[BAM_P_DESC_FIFO_ADDR]	= { 0x1381C, 0x00, 0x1000, 0x00 },
+	[BAM_P_EVNT_GEN_TRSHLD]	= { 0x13828, 0x00, 0x1000, 0x00 },
+	[BAM_P_FIFO_SIZES]	= { 0x13820, 0x00, 0x1000, 0x00 },
+};
+
 /* BAM CTRL */
 #define BAM_SW_RST			BIT(0)
 #define BAM_EN				BIT(1)
@@ -1051,6 +1080,7 @@ static void bam_channel_init(struct bam_device *bdev, struct bam_chan *bchan,
 static const struct of_device_id bam_of_match[] = {
 	{ .compatible = "qcom,bam-v1.3.0", .data = &bam_v1_3_reg_info },
 	{ .compatible = "qcom,bam-v1.4.0", .data = &bam_v1_4_reg_info },
+	{ .compatible = "qcom,bam-v1.7.0", .data = &bam_v1_7_reg_info },
 	{}
 };
 
@@ -1113,7 +1143,7 @@ static int bam_dma_probe(struct platform_device *pdev)
 
 	if (!bdev->channels) {
 		ret = -ENOMEM;
-		goto err_disable_clk;
+		goto err_tasklet_kill;
 	}
 
 	/* allocate and initialize channels */
@@ -1125,7 +1155,7 @@ static int bam_dma_probe(struct platform_device *pdev)
 	ret = devm_request_irq(bdev->dev, bdev->irq, bam_dma_irq,
 			IRQF_TRIGGER_HIGH, "bam_dma", bdev);
 	if (ret)
-		goto err_disable_clk;
+		goto err_bam_channel_exit;
 
 	/* set max dma segment size */
 	bdev->common.dev = bdev->dev;
@@ -1133,7 +1163,7 @@ static int bam_dma_probe(struct platform_device *pdev)
 	ret = dma_set_max_seg_size(bdev->common.dev, BAM_MAX_DATA_SIZE);
 	if (ret) {
 		dev_err(bdev->dev, "cannot set maximum segment size\n");
-		goto err_disable_clk;
+		goto err_bam_channel_exit;
 	}
 
 	platform_set_drvdata(pdev, bdev);
@@ -1161,7 +1191,7 @@ static int bam_dma_probe(struct platform_device *pdev)
 	ret = dma_async_device_register(&bdev->common);
 	if (ret) {
 		dev_err(bdev->dev, "failed to register dma async device\n");
-		goto err_disable_clk;
+		goto err_bam_channel_exit;
 	}
 
 	ret = of_dma_controller_register(pdev->dev.of_node, bam_dma_xlate,
@@ -1173,8 +1203,14 @@ static int bam_dma_probe(struct platform_device *pdev)
 
 err_unregister_dma:
 	dma_async_device_unregister(&bdev->common);
+err_bam_channel_exit:
+	for (i = 0; i < bdev->num_channels; i++)
+		tasklet_kill(&bdev->channels[i].vc.task);
+err_tasklet_kill:
+	tasklet_kill(&bdev->task);
 err_disable_clk:
 	clk_disable_unprepare(bdev->bamclk);
+
 	return ret;
 }
 
diff --git a/drivers/dma/s3c24xx-dma.c b/drivers/dma/s3c24xx-dma.c
index 2f91da3db836..01dcaf21b988 100644
--- a/drivers/dma/s3c24xx-dma.c
+++ b/drivers/dma/s3c24xx-dma.c
@@ -749,11 +749,6 @@ unlock:
 	return ret;
 }
 
-static int s3c24xx_dma_alloc_chan_resources(struct dma_chan *chan)
-{
-	return 0;
-}
-
 static void s3c24xx_dma_free_chan_resources(struct dma_chan *chan)
 {
 	/* Ensure all queued descriptors are freed */
@@ -1238,7 +1233,7 @@ static int s3c24xx_dma_probe(struct platform_device *pdev)
 	if (!s3cdma->phy_chans)
 		return -ENOMEM;
 
-	/* aquire irqs and clocks for all physical channels */
+	/* acquire irqs and clocks for all physical channels */
 	for (i = 0; i < pdata->num_phy_channels; i++) {
 		struct s3c24xx_dma_phy *phy = &s3cdma->phy_chans[i];
 		char clk_name[6];
@@ -1266,7 +1261,7 @@ static int s3c24xx_dma_probe(struct platform_device *pdev)
 			sprintf(clk_name, "dma.%d", i);
 			phy->clk = devm_clk_get(&pdev->dev, clk_name);
 			if (IS_ERR(phy->clk) && sdata->has_clocks) {
-				dev_err(&pdev->dev, "unable to aquire clock for channel %d, error %lu",
+				dev_err(&pdev->dev, "unable to acquire clock for channel %d, error %lu\n",
 					i, PTR_ERR(phy->clk));
 				continue;
 			}
@@ -1290,8 +1285,6 @@ static int s3c24xx_dma_probe(struct platform_device *pdev)
 	dma_cap_set(DMA_MEMCPY, s3cdma->memcpy.cap_mask);
 	dma_cap_set(DMA_PRIVATE, s3cdma->memcpy.cap_mask);
 	s3cdma->memcpy.dev = &pdev->dev;
-	s3cdma->memcpy.device_alloc_chan_resources =
-					s3c24xx_dma_alloc_chan_resources;
 	s3cdma->memcpy.device_free_chan_resources =
 					s3c24xx_dma_free_chan_resources;
 	s3cdma->memcpy.device_prep_dma_memcpy = s3c24xx_dma_prep_memcpy;
@@ -1305,8 +1298,6 @@ static int s3c24xx_dma_probe(struct platform_device *pdev)
 	dma_cap_set(DMA_CYCLIC, s3cdma->slave.cap_mask);
 	dma_cap_set(DMA_PRIVATE, s3cdma->slave.cap_mask);
 	s3cdma->slave.dev = &pdev->dev;
-	s3cdma->slave.device_alloc_chan_resources =
-					s3c24xx_dma_alloc_chan_resources;
 	s3cdma->slave.device_free_chan_resources =
 					s3c24xx_dma_free_chan_resources;
 	s3cdma->slave.device_tx_status = s3c24xx_dma_tx_status;
diff --git a/drivers/dma/sa11x0-dma.c b/drivers/dma/sa11x0-dma.c
index 5adf5407a8cb..43db255050d2 100644
--- a/drivers/dma/sa11x0-dma.c
+++ b/drivers/dma/sa11x0-dma.c
@@ -389,11 +389,6 @@ static void sa11x0_dma_tasklet(unsigned long arg)
 }
 
 
-static int sa11x0_dma_alloc_chan_resources(struct dma_chan *chan)
-{
-	return 0;
-}
-
 static void sa11x0_dma_free_chan_resources(struct dma_chan *chan)
 {
 	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
@@ -835,7 +830,6 @@ static int sa11x0_dma_init_dmadev(struct dma_device *dmadev,
 
 	INIT_LIST_HEAD(&dmadev->channels);
 	dmadev->dev = dev;
-	dmadev->device_alloc_chan_resources = sa11x0_dma_alloc_chan_resources;
 	dmadev->device_free_chan_resources = sa11x0_dma_free_chan_resources;
 	dmadev->device_config = sa11x0_dma_device_config;
 	dmadev->device_pause = sa11x0_dma_device_pause;
@@ -948,6 +942,12 @@ static int sa11x0_dma_probe(struct platform_device *pdev)
 	dma_cap_set(DMA_CYCLIC, d->slave.cap_mask);
 	d->slave.device_prep_slave_sg = sa11x0_dma_prep_slave_sg;
 	d->slave.device_prep_dma_cyclic = sa11x0_dma_prep_dma_cyclic;
+	d->slave.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	d->slave.residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+	d->slave.src_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |
+				   BIT(DMA_SLAVE_BUSWIDTH_2_BYTES);
+	d->slave.dst_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |
+				   BIT(DMA_SLAVE_BUSWIDTH_2_BYTES);
 	ret = sa11x0_dma_init_dmadev(&d->slave, &pdev->dev);
 	if (ret) {
 		dev_warn(d->slave.dev, "failed to register slave async device: %d\n",
diff --git a/drivers/dma/sh/Kconfig b/drivers/dma/sh/Kconfig
index 8190ad225a1b..0f371524a4d9 100644
--- a/drivers/dma/sh/Kconfig
+++ b/drivers/dma/sh/Kconfig
@@ -51,12 +51,6 @@ config RCAR_HPB_DMAE
 	help
 	  Enable support for the Renesas R-Car series DMA controllers.
 
-config RCAR_AUDMAC_PP
-	tristate "Renesas R-Car Audio DMAC Peripheral Peripheral support"
-	depends on SH_DMAE_BASE
-	help
-	  Enable support for the Renesas R-Car Audio DMAC Peripheral Peripheral controllers.
-
 config RCAR_DMAC
 	tristate "Renesas R-Car Gen2 DMA Controller"
 	depends on ARCH_SHMOBILE || COMPILE_TEST
@@ -64,3 +58,12 @@ config RCAR_DMAC
 	help
 	  This driver supports the general purpose DMA controller found in the
 	  Renesas R-Car second generation SoCs.
+
+config RENESAS_USB_DMAC
+	tristate "Renesas USB-DMA Controller"
+	depends on ARCH_SHMOBILE || COMPILE_TEST
+	select RENESAS_DMA
+	select DMA_VIRTUAL_CHANNELS
+	help
+	  This driver supports the USB-DMA controller found in the Renesas
+	  SoCs.
diff --git a/drivers/dma/sh/Makefile b/drivers/dma/sh/Makefile
index 2852f9db61a4..b8a598066ce2 100644
--- a/drivers/dma/sh/Makefile
+++ b/drivers/dma/sh/Makefile
@@ -15,5 +15,5 @@ obj-$(CONFIG_SH_DMAE) += shdma.o
 
 obj-$(CONFIG_SUDMAC) += sudmac.o
 obj-$(CONFIG_RCAR_HPB_DMAE) += rcar-hpbdma.o
-obj-$(CONFIG_RCAR_AUDMAC_PP) += rcar-audmapp.o
 obj-$(CONFIG_RCAR_DMAC) += rcar-dmac.o
+obj-$(CONFIG_RENESAS_USB_DMAC) += usb-dmac.o
diff --git a/drivers/dma/sh/rcar-audmapp.c b/drivers/dma/sh/rcar-audmapp.c
deleted file mode 100644
index d95bbdd721f4..000000000000
--- a/drivers/dma/sh/rcar-audmapp.c
+++ /dev/null
@@ -1,376 +0,0 @@
-/*
- * This is for Renesas R-Car Audio-DMAC-peri-peri.
- *
- * Copyright (C) 2014 Renesas Electronics Corporation
- * Copyright (C) 2014 Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
- *
- * based on the drivers/dma/sh/shdma.c
- *
- * Copyright (C) 2011-2012 Guennadi Liakhovetski <g.liakhovetski@gmx.de>
- * Copyright (C) 2009 Nobuhiro Iwamatsu <iwamatsu.nobuhiro@renesas.com>
- * Copyright (C) 2009 Renesas Solutions, Inc. All rights reserved.
- * Copyright (C) 2007 Freescale Semiconductor, Inc. All rights reserved.
- *
- * This 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/delay.h>
-#include <linux/init.h>
-#include <linux/module.h>
-#include <linux/slab.h>
-#include <linux/dmaengine.h>
-#include <linux/of_dma.h>
-#include <linux/platform_data/dma-rcar-audmapp.h>
-#include <linux/platform_device.h>
-#include <linux/shdma-base.h>
-
-/*
- * DMA register
- */
-#define PDMASAR		0x00
-#define PDMADAR		0x04
-#define PDMACHCR	0x0c
-
-/* PDMACHCR */
-#define PDMACHCR_DE		(1 << 0)
-
-#define AUDMAPP_MAX_CHANNELS	29
-
-/* Default MEMCPY transfer size = 2^2 = 4 bytes */
-#define LOG2_DEFAULT_XFER_SIZE	2
-#define AUDMAPP_SLAVE_NUMBER	256
-#define AUDMAPP_LEN_MAX		(16 * 1024 * 1024)
-
-struct audmapp_chan {
-	struct shdma_chan shdma_chan;
-	void __iomem *base;
-	dma_addr_t slave_addr;
-	u32 chcr;
-};
-
-struct audmapp_device {
-	struct shdma_dev shdma_dev;
-	struct audmapp_pdata *pdata;
-	struct device *dev;
-	void __iomem *chan_reg;
-};
-
-struct audmapp_desc {
-	struct shdma_desc shdma_desc;
-	dma_addr_t src;
-	dma_addr_t dst;
-};
-
-#define to_shdma_chan(c) container_of(c, struct shdma_chan, dma_chan)
-
-#define to_chan(chan) container_of(chan, struct audmapp_chan, shdma_chan)
-#define to_desc(sdesc) container_of(sdesc, struct audmapp_desc, shdma_desc)
-#define to_dev(chan) container_of(chan->shdma_chan.dma_chan.device,	\
-				  struct audmapp_device, shdma_dev.dma_dev)
-
-static void audmapp_write(struct audmapp_chan *auchan, u32 data, u32 reg)
-{
-	struct audmapp_device *audev = to_dev(auchan);
-	struct device *dev = audev->dev;
-
-	dev_dbg(dev, "w %p : %08x\n", auchan->base + reg, data);
-
-	iowrite32(data, auchan->base + reg);
-}
-
-static u32 audmapp_read(struct audmapp_chan *auchan, u32 reg)
-{
-	return ioread32(auchan->base + reg);
-}
-
-static void audmapp_halt(struct shdma_chan *schan)
-{
-	struct audmapp_chan *auchan = to_chan(schan);
-	int i;
-
-	audmapp_write(auchan, 0, PDMACHCR);
-
-	for (i = 0; i < 1024; i++) {
-		if (0 == audmapp_read(auchan, PDMACHCR))
-			return;
-		udelay(1);
-	}
-}
-
-static void audmapp_start_xfer(struct shdma_chan *schan,
-			       struct shdma_desc *sdesc)
-{
-	struct audmapp_chan *auchan = to_chan(schan);
-	struct audmapp_device *audev = to_dev(auchan);
-	struct audmapp_desc *desc = to_desc(sdesc);
-	struct device *dev = audev->dev;
-	u32 chcr = auchan->chcr | PDMACHCR_DE;
-
-	dev_dbg(dev, "src/dst/chcr = %pad/%pad/%08x\n",
-		&desc->src, &desc->dst, chcr);
-
-	audmapp_write(auchan, desc->src,	PDMASAR);
-	audmapp_write(auchan, desc->dst,	PDMADAR);
-	audmapp_write(auchan, chcr,	PDMACHCR);
-}
-
-static int audmapp_get_config(struct audmapp_chan *auchan, int slave_id,
-			      u32 *chcr, dma_addr_t *dst)
-{
-	struct audmapp_device *audev = to_dev(auchan);
-	struct audmapp_pdata *pdata = audev->pdata;
-	struct audmapp_slave_config *cfg;
-	int i;
-
-	*chcr	= 0;
-	*dst	= 0;
-
-	if (!pdata) { /* DT */
-		*chcr = ((u32)slave_id) << 16;
-		auchan->shdma_chan.slave_id = (slave_id) >> 8;
-		return 0;
-	}
-
-	/* non-DT */
-
-	if (slave_id >= AUDMAPP_SLAVE_NUMBER)
-		return -ENXIO;
-
-	for (i = 0, cfg = pdata->slave; i < pdata->slave_num; i++, cfg++)
-		if (cfg->slave_id == slave_id) {
-			*chcr	= cfg->chcr;
-			*dst	= cfg->dst;
-			return 0;
-		}
-
-	return -ENXIO;
-}
-
-static int audmapp_set_slave(struct shdma_chan *schan, int slave_id,
-			     dma_addr_t slave_addr, bool try)
-{
-	struct audmapp_chan *auchan = to_chan(schan);
-	u32 chcr;
-	dma_addr_t dst;
-	int ret;
-
-	ret = audmapp_get_config(auchan, slave_id, &chcr, &dst);
-	if (ret < 0)
-		return ret;
-
-	if (try)
-		return 0;
-
-	auchan->chcr		= chcr;
-	auchan->slave_addr	= slave_addr ? : dst;
-
-	return 0;
-}
-
-static int audmapp_desc_setup(struct shdma_chan *schan,
-			      struct shdma_desc *sdesc,
-			      dma_addr_t src, dma_addr_t dst, size_t *len)
-{
-	struct audmapp_desc *desc = to_desc(sdesc);
-
-	if (*len > (size_t)AUDMAPP_LEN_MAX)
-		*len = (size_t)AUDMAPP_LEN_MAX;
-
-	desc->src = src;
-	desc->dst = dst;
-
-	return 0;
-}
-
-static void audmapp_setup_xfer(struct shdma_chan *schan,
-			       int slave_id)
-{
-}
-
-static dma_addr_t audmapp_slave_addr(struct shdma_chan *schan)
-{
-	struct audmapp_chan *auchan = to_chan(schan);
-
-	return auchan->slave_addr;
-}
-
-static bool audmapp_channel_busy(struct shdma_chan *schan)
-{
-	struct audmapp_chan *auchan = to_chan(schan);
-	u32 chcr = audmapp_read(auchan, PDMACHCR);
-
-	return chcr & ~PDMACHCR_DE;
-}
-
-static bool audmapp_desc_completed(struct shdma_chan *schan,
-				   struct shdma_desc *sdesc)
-{
-	return true;
-}
-
-static struct shdma_desc *audmapp_embedded_desc(void *buf, int i)
-{
-	return &((struct audmapp_desc *)buf)[i].shdma_desc;
-}
-
-static const struct shdma_ops audmapp_shdma_ops = {
-	.halt_channel	= audmapp_halt,
-	.desc_setup	= audmapp_desc_setup,
-	.set_slave	= audmapp_set_slave,
-	.start_xfer	= audmapp_start_xfer,
-	.embedded_desc	= audmapp_embedded_desc,
-	.setup_xfer	= audmapp_setup_xfer,
-	.slave_addr	= audmapp_slave_addr,
-	.channel_busy	= audmapp_channel_busy,
-	.desc_completed	= audmapp_desc_completed,
-};
-
-static int audmapp_chan_probe(struct platform_device *pdev,
-			      struct audmapp_device *audev, int id)
-{
-	struct shdma_dev *sdev = &audev->shdma_dev;
-	struct audmapp_chan *auchan;
-	struct shdma_chan *schan;
-	struct device *dev = audev->dev;
-
-	auchan = devm_kzalloc(dev, sizeof(*auchan), GFP_KERNEL);
-	if (!auchan)
-		return -ENOMEM;
-
-	schan = &auchan->shdma_chan;
-	schan->max_xfer_len = AUDMAPP_LEN_MAX;
-
-	shdma_chan_probe(sdev, schan, id);
-
-	auchan->base = audev->chan_reg + 0x20 + (0x10 * id);
-	dev_dbg(dev, "%02d : %p / %p", id, auchan->base, audev->chan_reg);
-
-	return 0;
-}
-
-static void audmapp_chan_remove(struct audmapp_device *audev)
-{
-	struct shdma_chan *schan;
-	int i;
-
-	shdma_for_each_chan(schan, &audev->shdma_dev, i) {
-		BUG_ON(!schan);
-		shdma_chan_remove(schan);
-	}
-}
-
-static struct dma_chan *audmapp_of_xlate(struct of_phandle_args *dma_spec,
-					 struct of_dma *ofdma)
-{
-	dma_cap_mask_t mask;
-	struct dma_chan *chan;
-	u32 chcr = dma_spec->args[0];
-
-	if (dma_spec->args_count != 1)
-		return NULL;
-
-	dma_cap_zero(mask);
-	dma_cap_set(DMA_SLAVE, mask);
-
-	chan = dma_request_channel(mask, shdma_chan_filter, NULL);
-	if (chan)
-		to_shdma_chan(chan)->hw_req = chcr;
-
-	return chan;
-}
-
-static int audmapp_probe(struct platform_device *pdev)
-{
-	struct audmapp_pdata *pdata = pdev->dev.platform_data;
-	struct device_node *np = pdev->dev.of_node;
-	struct audmapp_device *audev;
-	struct shdma_dev *sdev;
-	struct dma_device *dma_dev;
-	struct resource *res;
-	int err, i;
-
-	if (np)
-		of_dma_controller_register(np, audmapp_of_xlate, pdev);
-	else if (!pdata)
-		return -ENODEV;
-
-	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-
-	audev = devm_kzalloc(&pdev->dev, sizeof(*audev), GFP_KERNEL);
-	if (!audev)
-		return -ENOMEM;
-
-	audev->dev	= &pdev->dev;
-	audev->pdata	= pdata;
-	audev->chan_reg	= devm_ioremap_resource(&pdev->dev, res);
-	if (IS_ERR(audev->chan_reg))
-		return PTR_ERR(audev->chan_reg);
-
-	sdev		= &audev->shdma_dev;
-	sdev->ops	= &audmapp_shdma_ops;
-	sdev->desc_size	= sizeof(struct audmapp_desc);
-
-	dma_dev			= &sdev->dma_dev;
-	dma_dev->copy_align	= LOG2_DEFAULT_XFER_SIZE;
-	dma_cap_set(DMA_SLAVE, dma_dev->cap_mask);
-
-	err = shdma_init(&pdev->dev, sdev, AUDMAPP_MAX_CHANNELS);
-	if (err < 0)
-		return err;
-
-	platform_set_drvdata(pdev, audev);
-
-	/* Create DMA Channel */
-	for (i = 0; i < AUDMAPP_MAX_CHANNELS; i++) {
-		err = audmapp_chan_probe(pdev, audev, i);
-		if (err)
-			goto chan_probe_err;
-	}
-
-	err = dma_async_device_register(dma_dev);
-	if (err < 0)
-		goto chan_probe_err;
-
-	return err;
-
-chan_probe_err:
-	audmapp_chan_remove(audev);
-	shdma_cleanup(sdev);
-
-	return err;
-}
-
-static int audmapp_remove(struct platform_device *pdev)
-{
-	struct audmapp_device *audev = platform_get_drvdata(pdev);
-	struct dma_device *dma_dev = &audev->shdma_dev.dma_dev;
-
-	dma_async_device_unregister(dma_dev);
-
-	audmapp_chan_remove(audev);
-	shdma_cleanup(&audev->shdma_dev);
-
-	return 0;
-}
-
-static const struct of_device_id audmapp_of_match[] = {
-	{ .compatible = "renesas,rcar-audmapp", },
-	{},
-};
-
-static struct platform_driver audmapp_driver = {
-	.probe		= audmapp_probe,
-	.remove		= audmapp_remove,
-	.driver		= {
-		.name	= "rcar-audmapp-engine",
-		.of_match_table = audmapp_of_match,
-	},
-};
-module_platform_driver(audmapp_driver);
-
-MODULE_AUTHOR("Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>");
-MODULE_DESCRIPTION("Renesas R-Car Audio DMAC peri-peri driver");
-MODULE_LICENSE("GPL");
diff --git a/drivers/dma/sh/shdma-base.c b/drivers/dma/sh/shdma-base.c
index 8ee383d339a5..10fcabad80f3 100644
--- a/drivers/dma/sh/shdma-base.c
+++ b/drivers/dma/sh/shdma-base.c
@@ -171,8 +171,7 @@ static struct shdma_desc *shdma_get_desc(struct shdma_chan *schan)
 	return NULL;
 }
 
-static int shdma_setup_slave(struct shdma_chan *schan, int slave_id,
-			     dma_addr_t slave_addr)
+static int shdma_setup_slave(struct shdma_chan *schan, dma_addr_t slave_addr)
 {
 	struct shdma_dev *sdev = to_shdma_dev(schan->dma_chan.device);
 	const struct shdma_ops *ops = sdev->ops;
@@ -183,25 +182,23 @@ static int shdma_setup_slave(struct shdma_chan *schan, int slave_id,
 		ret = ops->set_slave(schan, match, slave_addr, true);
 		if (ret < 0)
 			return ret;
-
-		slave_id = schan->slave_id;
 	} else {
-		match = slave_id;
+		match = schan->real_slave_id;
 	}
 
-	if (slave_id < 0 || slave_id >= slave_num)
+	if (schan->real_slave_id < 0 || schan->real_slave_id >= slave_num)
 		return -EINVAL;
 
-	if (test_and_set_bit(slave_id, shdma_slave_used))
+	if (test_and_set_bit(schan->real_slave_id, shdma_slave_used))
 		return -EBUSY;
 
 	ret = ops->set_slave(schan, match, slave_addr, false);
 	if (ret < 0) {
-		clear_bit(slave_id, shdma_slave_used);
+		clear_bit(schan->real_slave_id, shdma_slave_used);
 		return ret;
 	}
 
-	schan->slave_id = slave_id;
+	schan->slave_id = schan->real_slave_id;
 
 	return 0;
 }
@@ -221,10 +218,12 @@ static int shdma_alloc_chan_resources(struct dma_chan *chan)
 	 */
 	if (slave) {
 		/* Legacy mode: .private is set in filter */
-		ret = shdma_setup_slave(schan, slave->slave_id, 0);
+		schan->real_slave_id = slave->slave_id;
+		ret = shdma_setup_slave(schan, 0);
 		if (ret < 0)
 			goto esetslave;
 	} else {
+		/* Normal mode: real_slave_id was set by filter */
 		schan->slave_id = -EINVAL;
 	}
 
@@ -258,11 +257,14 @@ esetslave:
 
 /*
  * This is the standard shdma filter function to be used as a replacement to the
- * "old" method, using the .private pointer. If for some reason you allocate a
- * channel without slave data, use something like ERR_PTR(-EINVAL) as a filter
+ * "old" method, using the .private pointer.
+ * You always have to pass a valid slave id as the argument, old drivers that
+ * pass ERR_PTR(-EINVAL) as a filter parameter and set it up in dma_slave_config
+ * need to be updated so we can remove the slave_id field from dma_slave_config.
  * parameter. If this filter is used, the slave driver, after calling
  * dma_request_channel(), will also have to call dmaengine_slave_config() with
- * .slave_id, .direction, and either .src_addr or .dst_addr set.
+ * .direction, and either .src_addr or .dst_addr set.
+ *
  * NOTE: this filter doesn't support multiple DMAC drivers with the DMA_SLAVE
  * capability! If this becomes a requirement, hardware glue drivers, using this
  * services would have to provide their own filters, which first would check
@@ -276,7 +278,7 @@ bool shdma_chan_filter(struct dma_chan *chan, void *arg)
 {
 	struct shdma_chan *schan;
 	struct shdma_dev *sdev;
-	int match = (long)arg;
+	int slave_id = (long)arg;
 	int ret;
 
 	/* Only support channels handled by this driver. */
@@ -284,19 +286,39 @@ bool shdma_chan_filter(struct dma_chan *chan, void *arg)
 	    shdma_alloc_chan_resources)
 		return false;
 
-	if (match < 0)
+	schan = to_shdma_chan(chan);
+	sdev = to_shdma_dev(chan->device);
+
+	/*
+	 * For DT, the schan->slave_id field is generated by the
+	 * set_slave function from the slave ID that is passed in
+	 * from xlate. For the non-DT case, the slave ID is
+	 * directly passed into the filter function by the driver
+	 */
+	if (schan->dev->of_node) {
+		ret = sdev->ops->set_slave(schan, slave_id, 0, true);
+		if (ret < 0)
+			return false;
+
+		schan->real_slave_id = schan->slave_id;
+		return true;
+	}
+
+	if (slave_id < 0) {
 		/* No slave requested - arbitrary channel */
+		dev_warn(sdev->dma_dev.dev, "invalid slave ID passed to dma_request_slave\n");
 		return true;
+	}
 
-	schan = to_shdma_chan(chan);
-	if (!schan->dev->of_node && match >= slave_num)
+	if (slave_id >= slave_num)
 		return false;
 
-	sdev = to_shdma_dev(schan->dma_chan.device);
-	ret = sdev->ops->set_slave(schan, match, 0, true);
+	ret = sdev->ops->set_slave(schan, slave_id, 0, true);
 	if (ret < 0)
 		return false;
 
+	schan->real_slave_id = slave_id;
+
 	return true;
 }
 EXPORT_SYMBOL(shdma_chan_filter);
@@ -452,6 +474,8 @@ static void shdma_free_chan_resources(struct dma_chan *chan)
 		chan->private = NULL;
 	}
 
+	schan->real_slave_id = 0;
+
 	spin_lock_irq(&schan->chan_lock);
 
 	list_splice_init(&schan->ld_free, &list);
@@ -764,11 +788,20 @@ static int shdma_config(struct dma_chan *chan,
 	 */
 	if (!config)
 		return -EINVAL;
+
+	/*
+	 * overriding the slave_id through dma_slave_config is deprecated,
+	 * but possibly some out-of-tree drivers still do it.
+	 */
+	if (WARN_ON_ONCE(config->slave_id &&
+			 config->slave_id != schan->real_slave_id))
+		schan->real_slave_id = config->slave_id;
+
 	/*
 	 * We could lock this, but you shouldn't be configuring the
 	 * channel, while using it...
 	 */
-	return shdma_setup_slave(schan, config->slave_id,
+	return shdma_setup_slave(schan,
 				 config->direction == DMA_DEV_TO_MEM ?
 				 config->src_addr : config->dst_addr);
 }
diff --git a/drivers/dma/sh/shdmac.c b/drivers/dma/sh/shdmac.c
index 9f1d4c7dbab8..11707df1a689 100644
--- a/drivers/dma/sh/shdmac.c
+++ b/drivers/dma/sh/shdmac.c
@@ -443,7 +443,7 @@ static bool sh_dmae_reset(struct sh_dmae_device *shdev)
 	return ret;
 }
 
-#if defined(CONFIG_CPU_SH4) || defined(CONFIG_ARM)
+#if defined(CONFIG_CPU_SH4) || defined(CONFIG_ARCH_SHMOBILE)
 static irqreturn_t sh_dmae_err(int irq, void *data)
 {
 	struct sh_dmae_device *shdev = data;
@@ -689,7 +689,7 @@ static int sh_dmae_probe(struct platform_device *pdev)
 	const struct sh_dmae_pdata *pdata;
 	unsigned long chan_flag[SH_DMAE_MAX_CHANNELS] = {};
 	int chan_irq[SH_DMAE_MAX_CHANNELS];
-#if defined(CONFIG_CPU_SH4) || defined(CONFIG_ARM)
+#if defined(CONFIG_CPU_SH4) || defined(CONFIG_ARCH_SHMOBILE)
 	unsigned long irqflags = 0;
 	int errirq;
 #endif
diff --git a/drivers/dma/sh/usb-dmac.c b/drivers/dma/sh/usb-dmac.c
new file mode 100644
index 000000000000..f705798ce3eb
--- /dev/null
+++ b/drivers/dma/sh/usb-dmac.c
@@ -0,0 +1,910 @@
+/*
+ * Renesas USB DMA Controller Driver
+ *
+ * Copyright (C) 2015 Renesas Electronics Corporation
+ *
+ * based on rcar-dmac.c
+ * Copyright (C) 2014 Renesas Electronics Inc.
+ * Author: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
+ *
+ * This is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/interrupt.h>
+#include <linux/list.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_dma.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+
+#include "../dmaengine.h"
+#include "../virt-dma.h"
+
+/*
+ * struct usb_dmac_sg - Descriptor for a hardware transfer
+ * @mem_addr: memory address
+ * @size: transfer size in bytes
+ */
+struct usb_dmac_sg {
+	dma_addr_t mem_addr;
+	u32 size;
+};
+
+/*
+ * struct usb_dmac_desc - USB DMA Transfer Descriptor
+ * @vd: base virtual channel DMA transaction descriptor
+ * @direction: direction of the DMA transfer
+ * @sg_allocated_len: length of allocated sg
+ * @sg_len: length of sg
+ * @sg_index: index of sg
+ * @residue: residue after the DMAC completed a transfer
+ * @node: node for desc_got and desc_freed
+ * @done_cookie: cookie after the DMAC completed a transfer
+ * @sg: information for the transfer
+ */
+struct usb_dmac_desc {
+	struct virt_dma_desc vd;
+	enum dma_transfer_direction direction;
+	unsigned int sg_allocated_len;
+	unsigned int sg_len;
+	unsigned int sg_index;
+	u32 residue;
+	struct list_head node;
+	dma_cookie_t done_cookie;
+	struct usb_dmac_sg sg[0];
+};
+
+#define to_usb_dmac_desc(vd)	container_of(vd, struct usb_dmac_desc, vd)
+
+/*
+ * struct usb_dmac_chan - USB DMA Controller Channel
+ * @vc: base virtual DMA channel object
+ * @iomem: channel I/O memory base
+ * @index: index of this channel in the controller
+ * @irq: irq number of this channel
+ * @desc: the current descriptor
+ * @descs_allocated: number of descriptors allocated
+ * @desc_got: got descriptors
+ * @desc_freed: freed descriptors after the DMAC completed a transfer
+ */
+struct usb_dmac_chan {
+	struct virt_dma_chan vc;
+	void __iomem *iomem;
+	unsigned int index;
+	int irq;
+	struct usb_dmac_desc *desc;
+	int descs_allocated;
+	struct list_head desc_got;
+	struct list_head desc_freed;
+};
+
+#define to_usb_dmac_chan(c) container_of(c, struct usb_dmac_chan, vc.chan)
+
+/*
+ * struct usb_dmac - USB DMA Controller
+ * @engine: base DMA engine object
+ * @dev: the hardware device
+ * @iomem: remapped I/O memory base
+ * @n_channels: number of available channels
+ * @channels: array of DMAC channels
+ */
+struct usb_dmac {
+	struct dma_device engine;
+	struct device *dev;
+	void __iomem *iomem;
+
+	unsigned int n_channels;
+	struct usb_dmac_chan *channels;
+};
+
+#define to_usb_dmac(d)		container_of(d, struct usb_dmac, engine)
+
+/* -----------------------------------------------------------------------------
+ * Registers
+ */
+
+#define USB_DMAC_CHAN_OFFSET(i)		(0x20 + 0x20 * (i))
+
+#define USB_DMASWR			0x0008
+#define USB_DMASWR_SWR			(1 << 0)
+#define USB_DMAOR			0x0060
+#define USB_DMAOR_AE			(1 << 2)
+#define USB_DMAOR_DME			(1 << 0)
+
+#define USB_DMASAR			0x0000
+#define USB_DMADAR			0x0004
+#define USB_DMATCR			0x0008
+#define USB_DMATCR_MASK			0x00ffffff
+#define USB_DMACHCR			0x0014
+#define USB_DMACHCR_FTE			(1 << 24)
+#define USB_DMACHCR_NULLE		(1 << 16)
+#define USB_DMACHCR_NULL		(1 << 12)
+#define USB_DMACHCR_TS_8B		((0 << 7) | (0 << 6))
+#define USB_DMACHCR_TS_16B		((0 << 7) | (1 << 6))
+#define USB_DMACHCR_TS_32B		((1 << 7) | (0 << 6))
+#define USB_DMACHCR_IE			(1 << 5)
+#define USB_DMACHCR_SP			(1 << 2)
+#define USB_DMACHCR_TE			(1 << 1)
+#define USB_DMACHCR_DE			(1 << 0)
+#define USB_DMATEND			0x0018
+
+/* Hardcode the xfer_shift to 5 (32bytes) */
+#define USB_DMAC_XFER_SHIFT	5
+#define USB_DMAC_XFER_SIZE	(1 << USB_DMAC_XFER_SHIFT)
+#define USB_DMAC_CHCR_TS	USB_DMACHCR_TS_32B
+#define USB_DMAC_SLAVE_BUSWIDTH	DMA_SLAVE_BUSWIDTH_32_BYTES
+
+/* for descriptors */
+#define USB_DMAC_INITIAL_NR_DESC	16
+#define USB_DMAC_INITIAL_NR_SG		8
+
+/* -----------------------------------------------------------------------------
+ * Device access
+ */
+
+static void usb_dmac_write(struct usb_dmac *dmac, u32 reg, u32 data)
+{
+	writel(data, dmac->iomem + reg);
+}
+
+static u32 usb_dmac_read(struct usb_dmac *dmac, u32 reg)
+{
+	return readl(dmac->iomem + reg);
+}
+
+static u32 usb_dmac_chan_read(struct usb_dmac_chan *chan, u32 reg)
+{
+	return readl(chan->iomem + reg);
+}
+
+static void usb_dmac_chan_write(struct usb_dmac_chan *chan, u32 reg, u32 data)
+{
+	writel(data, chan->iomem + reg);
+}
+
+/* -----------------------------------------------------------------------------
+ * Initialization and configuration
+ */
+
+static bool usb_dmac_chan_is_busy(struct usb_dmac_chan *chan)
+{
+	u32 chcr = usb_dmac_chan_read(chan, USB_DMACHCR);
+
+	return (chcr & (USB_DMACHCR_DE | USB_DMACHCR_TE)) == USB_DMACHCR_DE;
+}
+
+static u32 usb_dmac_calc_tend(u32 size)
+{
+	/*
+	 * Please refer to the Figure "Example of Final Transaction Valid
+	 * Data Transfer Enable (EDTEN) Setting" in the data sheet.
+	 */
+	return 0xffffffff << (32 - (size % USB_DMAC_XFER_SIZE ?	:
+						USB_DMAC_XFER_SIZE));
+}
+
+/* This function is already held by vc.lock */
+static void usb_dmac_chan_start_sg(struct usb_dmac_chan *chan,
+				   unsigned int index)
+{
+	struct usb_dmac_desc *desc = chan->desc;
+	struct usb_dmac_sg *sg = desc->sg + index;
+	dma_addr_t src_addr = 0, dst_addr = 0;
+
+	WARN_ON_ONCE(usb_dmac_chan_is_busy(chan));
+
+	if (desc->direction == DMA_DEV_TO_MEM)
+		dst_addr = sg->mem_addr;
+	else
+		src_addr = sg->mem_addr;
+
+	dev_dbg(chan->vc.chan.device->dev,
+		"chan%u: queue sg %p: %u@%pad -> %pad\n",
+		chan->index, sg, sg->size, &src_addr, &dst_addr);
+
+	usb_dmac_chan_write(chan, USB_DMASAR, src_addr & 0xffffffff);
+	usb_dmac_chan_write(chan, USB_DMADAR, dst_addr & 0xffffffff);
+	usb_dmac_chan_write(chan, USB_DMATCR,
+			    DIV_ROUND_UP(sg->size, USB_DMAC_XFER_SIZE));
+	usb_dmac_chan_write(chan, USB_DMATEND, usb_dmac_calc_tend(sg->size));
+
+	usb_dmac_chan_write(chan, USB_DMACHCR, USB_DMAC_CHCR_TS |
+			USB_DMACHCR_NULLE | USB_DMACHCR_IE | USB_DMACHCR_DE);
+}
+
+/* This function is already held by vc.lock */
+static void usb_dmac_chan_start_desc(struct usb_dmac_chan *chan)
+{
+	struct virt_dma_desc *vd;
+
+	vd = vchan_next_desc(&chan->vc);
+	if (!vd) {
+		chan->desc = NULL;
+		return;
+	}
+
+	/*
+	 * Remove this request from vc->desc_issued. Otherwise, this driver
+	 * will get the previous value from vchan_next_desc() after a transfer
+	 * was completed.
+	 */
+	list_del(&vd->node);
+
+	chan->desc = to_usb_dmac_desc(vd);
+	chan->desc->sg_index = 0;
+	usb_dmac_chan_start_sg(chan, 0);
+}
+
+static int usb_dmac_init(struct usb_dmac *dmac)
+{
+	u16 dmaor;
+
+	/* Clear all channels and enable the DMAC globally. */
+	usb_dmac_write(dmac, USB_DMAOR, USB_DMAOR_DME);
+
+	dmaor = usb_dmac_read(dmac, USB_DMAOR);
+	if ((dmaor & (USB_DMAOR_AE | USB_DMAOR_DME)) != USB_DMAOR_DME) {
+		dev_warn(dmac->dev, "DMAOR initialization failed.\n");
+		return -EIO;
+	}
+
+	return 0;
+}
+
+/* -----------------------------------------------------------------------------
+ * Descriptors allocation and free
+ */
+static int usb_dmac_desc_alloc(struct usb_dmac_chan *chan, unsigned int sg_len,
+			       gfp_t gfp)
+{
+	struct usb_dmac_desc *desc;
+	unsigned long flags;
+
+	desc = kzalloc(sizeof(*desc) + sg_len * sizeof(desc->sg[0]), gfp);
+	if (!desc)
+		return -ENOMEM;
+
+	desc->sg_allocated_len = sg_len;
+	INIT_LIST_HEAD(&desc->node);
+
+	spin_lock_irqsave(&chan->vc.lock, flags);
+	list_add_tail(&desc->node, &chan->desc_freed);
+	spin_unlock_irqrestore(&chan->vc.lock, flags);
+
+	return 0;
+}
+
+static void usb_dmac_desc_free(struct usb_dmac_chan *chan)
+{
+	struct usb_dmac_desc *desc, *_desc;
+	LIST_HEAD(list);
+
+	list_splice_init(&chan->desc_freed, &list);
+	list_splice_init(&chan->desc_got, &list);
+
+	list_for_each_entry_safe(desc, _desc, &list, node) {
+		list_del(&desc->node);
+		kfree(desc);
+	}
+	chan->descs_allocated = 0;
+}
+
+static struct usb_dmac_desc *usb_dmac_desc_get(struct usb_dmac_chan *chan,
+					       unsigned int sg_len, gfp_t gfp)
+{
+	struct usb_dmac_desc *desc = NULL;
+	unsigned long flags;
+
+	/* Get a freed descritpor */
+	spin_lock_irqsave(&chan->vc.lock, flags);
+	list_for_each_entry(desc, &chan->desc_freed, node) {
+		if (sg_len <= desc->sg_allocated_len) {
+			list_move_tail(&desc->node, &chan->desc_got);
+			spin_unlock_irqrestore(&chan->vc.lock, flags);
+			return desc;
+		}
+	}
+	spin_unlock_irqrestore(&chan->vc.lock, flags);
+
+	/* Allocate a new descriptor */
+	if (!usb_dmac_desc_alloc(chan, sg_len, gfp)) {
+		/* If allocated the desc, it was added to tail of the list */
+		spin_lock_irqsave(&chan->vc.lock, flags);
+		desc = list_last_entry(&chan->desc_freed, struct usb_dmac_desc,
+				       node);
+		list_move_tail(&desc->node, &chan->desc_got);
+		spin_unlock_irqrestore(&chan->vc.lock, flags);
+		return desc;
+	}
+
+	return NULL;
+}
+
+static void usb_dmac_desc_put(struct usb_dmac_chan *chan,
+			      struct usb_dmac_desc *desc)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&chan->vc.lock, flags);
+	list_move_tail(&desc->node, &chan->desc_freed);
+	spin_unlock_irqrestore(&chan->vc.lock, flags);
+}
+
+/* -----------------------------------------------------------------------------
+ * Stop and reset
+ */
+
+static void usb_dmac_soft_reset(struct usb_dmac_chan *uchan)
+{
+	struct dma_chan *chan = &uchan->vc.chan;
+	struct usb_dmac *dmac = to_usb_dmac(chan->device);
+	int i;
+
+	/* Don't issue soft reset if any one of channels is busy */
+	for (i = 0; i < dmac->n_channels; ++i) {
+		if (usb_dmac_chan_is_busy(uchan))
+			return;
+	}
+
+	usb_dmac_write(dmac, USB_DMAOR, 0);
+	usb_dmac_write(dmac, USB_DMASWR, USB_DMASWR_SWR);
+	udelay(100);
+	usb_dmac_write(dmac, USB_DMASWR, 0);
+	usb_dmac_write(dmac, USB_DMAOR, 1);
+}
+
+static void usb_dmac_chan_halt(struct usb_dmac_chan *chan)
+{
+	u32 chcr = usb_dmac_chan_read(chan, USB_DMACHCR);
+
+	chcr &= ~(USB_DMACHCR_IE | USB_DMACHCR_TE | USB_DMACHCR_DE);
+	usb_dmac_chan_write(chan, USB_DMACHCR, chcr);
+
+	usb_dmac_soft_reset(chan);
+}
+
+static void usb_dmac_stop(struct usb_dmac *dmac)
+{
+	usb_dmac_write(dmac, USB_DMAOR, 0);
+}
+
+/* -----------------------------------------------------------------------------
+ * DMA engine operations
+ */
+
+static int usb_dmac_alloc_chan_resources(struct dma_chan *chan)
+{
+	struct usb_dmac_chan *uchan = to_usb_dmac_chan(chan);
+	int ret;
+
+	while (uchan->descs_allocated < USB_DMAC_INITIAL_NR_DESC) {
+		ret = usb_dmac_desc_alloc(uchan, USB_DMAC_INITIAL_NR_SG,
+					  GFP_KERNEL);
+		if (ret < 0) {
+			usb_dmac_desc_free(uchan);
+			return ret;
+		}
+		uchan->descs_allocated++;
+	}
+
+	return pm_runtime_get_sync(chan->device->dev);
+}
+
+static void usb_dmac_free_chan_resources(struct dma_chan *chan)
+{
+	struct usb_dmac_chan *uchan = to_usb_dmac_chan(chan);
+	unsigned long flags;
+
+	/* Protect against ISR */
+	spin_lock_irqsave(&uchan->vc.lock, flags);
+	usb_dmac_chan_halt(uchan);
+	spin_unlock_irqrestore(&uchan->vc.lock, flags);
+
+	usb_dmac_desc_free(uchan);
+	vchan_free_chan_resources(&uchan->vc);
+
+	pm_runtime_put(chan->device->dev);
+}
+
+static struct dma_async_tx_descriptor *
+usb_dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
+		       unsigned int sg_len, enum dma_transfer_direction dir,
+		       unsigned long dma_flags, void *context)
+{
+	struct usb_dmac_chan *uchan = to_usb_dmac_chan(chan);
+	struct usb_dmac_desc *desc;
+	struct scatterlist *sg;
+	int i;
+
+	if (!sg_len) {
+		dev_warn(chan->device->dev,
+			 "%s: bad parameter: len=%d\n", __func__, sg_len);
+		return NULL;
+	}
+
+	desc = usb_dmac_desc_get(uchan, sg_len, GFP_NOWAIT);
+	if (!desc)
+		return NULL;
+
+	desc->direction = dir;
+	desc->sg_len = sg_len;
+	for_each_sg(sgl, sg, sg_len, i) {
+		desc->sg[i].mem_addr = sg_dma_address(sg);
+		desc->sg[i].size = sg_dma_len(sg);
+	}
+
+	return vchan_tx_prep(&uchan->vc, &desc->vd, dma_flags);
+}
+
+static int usb_dmac_chan_terminate_all(struct dma_chan *chan)
+{
+	struct usb_dmac_chan *uchan = to_usb_dmac_chan(chan);
+	struct usb_dmac_desc *desc;
+	unsigned long flags;
+	LIST_HEAD(head);
+	LIST_HEAD(list);
+
+	spin_lock_irqsave(&uchan->vc.lock, flags);
+	usb_dmac_chan_halt(uchan);
+	vchan_get_all_descriptors(&uchan->vc, &head);
+	if (uchan->desc)
+		uchan->desc = NULL;
+	list_splice_init(&uchan->desc_got, &list);
+	list_for_each_entry(desc, &list, node)
+		list_move_tail(&desc->node, &uchan->desc_freed);
+	spin_unlock_irqrestore(&uchan->vc.lock, flags);
+	vchan_dma_desc_free_list(&uchan->vc, &head);
+
+	return 0;
+}
+
+static unsigned int usb_dmac_get_current_residue(struct usb_dmac_chan *chan,
+						 struct usb_dmac_desc *desc,
+						 int sg_index)
+{
+	struct usb_dmac_sg *sg = desc->sg + sg_index;
+	u32 mem_addr = sg->mem_addr & 0xffffffff;
+	unsigned int residue = sg->size;
+
+	/*
+	 * We cannot use USB_DMATCR to calculate residue because USB_DMATCR
+	 * has unsuited value to calculate.
+	 */
+	if (desc->direction == DMA_DEV_TO_MEM)
+		residue -= usb_dmac_chan_read(chan, USB_DMADAR) - mem_addr;
+	else
+		residue -= usb_dmac_chan_read(chan, USB_DMASAR) - mem_addr;
+
+	return residue;
+}
+
+static u32 usb_dmac_chan_get_residue_if_complete(struct usb_dmac_chan *chan,
+						 dma_cookie_t cookie)
+{
+	struct usb_dmac_desc *desc;
+	u32 residue = 0;
+
+	list_for_each_entry_reverse(desc, &chan->desc_freed, node) {
+		if (desc->done_cookie == cookie) {
+			residue = desc->residue;
+			break;
+		}
+	}
+
+	return residue;
+}
+
+static u32 usb_dmac_chan_get_residue(struct usb_dmac_chan *chan,
+				     dma_cookie_t cookie)
+{
+	u32 residue = 0;
+	struct virt_dma_desc *vd;
+	struct usb_dmac_desc *desc = chan->desc;
+	int i;
+
+	if (!desc) {
+		vd = vchan_find_desc(&chan->vc, cookie);
+		if (!vd)
+			return 0;
+		desc = to_usb_dmac_desc(vd);
+	}
+
+	/* Compute the size of all usb_dmac_sg still to be transferred */
+	for (i = desc->sg_index + 1; i < desc->sg_len; i++)
+		residue += desc->sg[i].size;
+
+	/* Add the residue for the current sg */
+	residue += usb_dmac_get_current_residue(chan, desc, desc->sg_index);
+
+	return residue;
+}
+
+static enum dma_status usb_dmac_tx_status(struct dma_chan *chan,
+					  dma_cookie_t cookie,
+					  struct dma_tx_state *txstate)
+{
+	struct usb_dmac_chan *uchan = to_usb_dmac_chan(chan);
+	enum dma_status status;
+	unsigned int residue = 0;
+	unsigned long flags;
+
+	status = dma_cookie_status(chan, cookie, txstate);
+	/* a client driver will get residue after DMA_COMPLETE */
+	if (!txstate)
+		return status;
+
+	spin_lock_irqsave(&uchan->vc.lock, flags);
+	if (status == DMA_COMPLETE)
+		residue = usb_dmac_chan_get_residue_if_complete(uchan, cookie);
+	else
+		residue = usb_dmac_chan_get_residue(uchan, cookie);
+	spin_unlock_irqrestore(&uchan->vc.lock, flags);
+
+	dma_set_residue(txstate, residue);
+
+	return status;
+}
+
+static void usb_dmac_issue_pending(struct dma_chan *chan)
+{
+	struct usb_dmac_chan *uchan = to_usb_dmac_chan(chan);
+	unsigned long flags;
+
+	spin_lock_irqsave(&uchan->vc.lock, flags);
+	if (vchan_issue_pending(&uchan->vc) && !uchan->desc)
+		usb_dmac_chan_start_desc(uchan);
+	spin_unlock_irqrestore(&uchan->vc.lock, flags);
+}
+
+static void usb_dmac_virt_desc_free(struct virt_dma_desc *vd)
+{
+	struct usb_dmac_desc *desc = to_usb_dmac_desc(vd);
+	struct usb_dmac_chan *chan = to_usb_dmac_chan(vd->tx.chan);
+
+	usb_dmac_desc_put(chan, desc);
+}
+
+/* -----------------------------------------------------------------------------
+ * IRQ handling
+ */
+
+static void usb_dmac_isr_transfer_end(struct usb_dmac_chan *chan)
+{
+	struct usb_dmac_desc *desc = chan->desc;
+
+	BUG_ON(!desc);
+
+	if (++desc->sg_index < desc->sg_len) {
+		usb_dmac_chan_start_sg(chan, desc->sg_index);
+	} else {
+		desc->residue = usb_dmac_get_current_residue(chan, desc,
+							desc->sg_index - 1);
+		desc->done_cookie = desc->vd.tx.cookie;
+		vchan_cookie_complete(&desc->vd);
+
+		/* Restart the next transfer if this driver has a next desc */
+		usb_dmac_chan_start_desc(chan);
+	}
+}
+
+static irqreturn_t usb_dmac_isr_channel(int irq, void *dev)
+{
+	struct usb_dmac_chan *chan = dev;
+	irqreturn_t ret = IRQ_NONE;
+	u32 mask = USB_DMACHCR_TE;
+	u32 check_bits = USB_DMACHCR_TE | USB_DMACHCR_SP;
+	u32 chcr;
+
+	spin_lock(&chan->vc.lock);
+
+	chcr = usb_dmac_chan_read(chan, USB_DMACHCR);
+	if (chcr & check_bits)
+		mask |= USB_DMACHCR_DE | check_bits;
+	if (chcr & USB_DMACHCR_NULL) {
+		/* An interruption of TE will happen after we set FTE */
+		mask |= USB_DMACHCR_NULL;
+		chcr |= USB_DMACHCR_FTE;
+		ret |= IRQ_HANDLED;
+	}
+	usb_dmac_chan_write(chan, USB_DMACHCR, chcr & ~mask);
+
+	if (chcr & check_bits) {
+		usb_dmac_isr_transfer_end(chan);
+		ret |= IRQ_HANDLED;
+	}
+
+	spin_unlock(&chan->vc.lock);
+
+	return ret;
+}
+
+/* -----------------------------------------------------------------------------
+ * OF xlate and channel filter
+ */
+
+static bool usb_dmac_chan_filter(struct dma_chan *chan, void *arg)
+{
+	struct usb_dmac_chan *uchan = to_usb_dmac_chan(chan);
+	struct of_phandle_args *dma_spec = arg;
+
+	if (dma_spec->np != chan->device->dev->of_node)
+		return false;
+
+	/* USB-DMAC should be used with fixed usb controller's FIFO */
+	if (uchan->index != dma_spec->args[0])
+		return false;
+
+	return true;
+}
+
+static struct dma_chan *usb_dmac_of_xlate(struct of_phandle_args *dma_spec,
+					  struct of_dma *ofdma)
+{
+	struct usb_dmac_chan *uchan;
+	struct dma_chan *chan;
+	dma_cap_mask_t mask;
+
+	if (dma_spec->args_count != 1)
+		return NULL;
+
+	/* Only slave DMA channels can be allocated via DT */
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	chan = dma_request_channel(mask, usb_dmac_chan_filter, dma_spec);
+	if (!chan)
+		return NULL;
+
+	uchan = to_usb_dmac_chan(chan);
+
+	return chan;
+}
+
+/* -----------------------------------------------------------------------------
+ * Power management
+ */
+
+static int usb_dmac_runtime_suspend(struct device *dev)
+{
+	struct usb_dmac *dmac = dev_get_drvdata(dev);
+	int i;
+
+	for (i = 0; i < dmac->n_channels; ++i)
+		usb_dmac_chan_halt(&dmac->channels[i]);
+
+	return 0;
+}
+
+static int usb_dmac_runtime_resume(struct device *dev)
+{
+	struct usb_dmac *dmac = dev_get_drvdata(dev);
+
+	return usb_dmac_init(dmac);
+}
+
+static const struct dev_pm_ops usb_dmac_pm = {
+	SET_RUNTIME_PM_OPS(usb_dmac_runtime_suspend, usb_dmac_runtime_resume,
+			   NULL)
+};
+
+/* -----------------------------------------------------------------------------
+ * Probe and remove
+ */
+
+static int usb_dmac_chan_probe(struct usb_dmac *dmac,
+			       struct usb_dmac_chan *uchan,
+			       unsigned int index)
+{
+	struct platform_device *pdev = to_platform_device(dmac->dev);
+	char pdev_irqname[5];
+	char *irqname;
+	int ret;
+
+	uchan->index = index;
+	uchan->iomem = dmac->iomem + USB_DMAC_CHAN_OFFSET(index);
+
+	/* Request the channel interrupt. */
+	sprintf(pdev_irqname, "ch%u", index);
+	uchan->irq = platform_get_irq_byname(pdev, pdev_irqname);
+	if (uchan->irq < 0) {
+		dev_err(dmac->dev, "no IRQ specified for channel %u\n", index);
+		return -ENODEV;
+	}
+
+	irqname = devm_kasprintf(dmac->dev, GFP_KERNEL, "%s:%u",
+				 dev_name(dmac->dev), index);
+	if (!irqname)
+		return -ENOMEM;
+
+	ret = devm_request_irq(dmac->dev, uchan->irq, usb_dmac_isr_channel,
+			       IRQF_SHARED, irqname, uchan);
+	if (ret) {
+		dev_err(dmac->dev, "failed to request IRQ %u (%d)\n",
+			uchan->irq, ret);
+		return ret;
+	}
+
+	uchan->vc.desc_free = usb_dmac_virt_desc_free;
+	vchan_init(&uchan->vc, &dmac->engine);
+	INIT_LIST_HEAD(&uchan->desc_freed);
+	INIT_LIST_HEAD(&uchan->desc_got);
+
+	return 0;
+}
+
+static int usb_dmac_parse_of(struct device *dev, struct usb_dmac *dmac)
+{
+	struct device_node *np = dev->of_node;
+	int ret;
+
+	ret = of_property_read_u32(np, "dma-channels", &dmac->n_channels);
+	if (ret < 0) {
+		dev_err(dev, "unable to read dma-channels property\n");
+		return ret;
+	}
+
+	if (dmac->n_channels <= 0 || dmac->n_channels >= 100) {
+		dev_err(dev, "invalid number of channels %u\n",
+			dmac->n_channels);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int usb_dmac_probe(struct platform_device *pdev)
+{
+	const enum dma_slave_buswidth widths = USB_DMAC_SLAVE_BUSWIDTH;
+	struct dma_device *engine;
+	struct usb_dmac *dmac;
+	struct resource *mem;
+	unsigned int i;
+	int ret;
+
+	dmac = devm_kzalloc(&pdev->dev, sizeof(*dmac), GFP_KERNEL);
+	if (!dmac)
+		return -ENOMEM;
+
+	dmac->dev = &pdev->dev;
+	platform_set_drvdata(pdev, dmac);
+
+	ret = usb_dmac_parse_of(&pdev->dev, dmac);
+	if (ret < 0)
+		return ret;
+
+	dmac->channels = devm_kcalloc(&pdev->dev, dmac->n_channels,
+				      sizeof(*dmac->channels), GFP_KERNEL);
+	if (!dmac->channels)
+		return -ENOMEM;
+
+	/* Request resources. */
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	dmac->iomem = devm_ioremap_resource(&pdev->dev, mem);
+	if (IS_ERR(dmac->iomem))
+		return PTR_ERR(dmac->iomem);
+
+	/* Enable runtime PM and initialize the device. */
+	pm_runtime_enable(&pdev->dev);
+	ret = pm_runtime_get_sync(&pdev->dev);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "runtime PM get sync failed (%d)\n", ret);
+		return ret;
+	}
+
+	ret = usb_dmac_init(dmac);
+	pm_runtime_put(&pdev->dev);
+
+	if (ret) {
+		dev_err(&pdev->dev, "failed to reset device\n");
+		goto error;
+	}
+
+	/* Initialize the channels. */
+	INIT_LIST_HEAD(&dmac->engine.channels);
+
+	for (i = 0; i < dmac->n_channels; ++i) {
+		ret = usb_dmac_chan_probe(dmac, &dmac->channels[i], i);
+		if (ret < 0)
+			goto error;
+	}
+
+	/* Register the DMAC as a DMA provider for DT. */
+	ret = of_dma_controller_register(pdev->dev.of_node, usb_dmac_of_xlate,
+					 NULL);
+	if (ret < 0)
+		goto error;
+
+	/*
+	 * Register the DMA engine device.
+	 *
+	 * Default transfer size of 32 bytes requires 32-byte alignment.
+	 */
+	engine = &dmac->engine;
+	dma_cap_set(DMA_SLAVE, engine->cap_mask);
+
+	engine->dev = &pdev->dev;
+
+	engine->src_addr_widths = widths;
+	engine->dst_addr_widths = widths;
+	engine->directions = BIT(DMA_MEM_TO_DEV) | BIT(DMA_DEV_TO_MEM);
+	engine->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+
+	engine->device_alloc_chan_resources = usb_dmac_alloc_chan_resources;
+	engine->device_free_chan_resources = usb_dmac_free_chan_resources;
+	engine->device_prep_slave_sg = usb_dmac_prep_slave_sg;
+	engine->device_terminate_all = usb_dmac_chan_terminate_all;
+	engine->device_tx_status = usb_dmac_tx_status;
+	engine->device_issue_pending = usb_dmac_issue_pending;
+
+	ret = dma_async_device_register(engine);
+	if (ret < 0)
+		goto error;
+
+	return 0;
+
+error:
+	of_dma_controller_free(pdev->dev.of_node);
+	pm_runtime_disable(&pdev->dev);
+	return ret;
+}
+
+static void usb_dmac_chan_remove(struct usb_dmac *dmac,
+				 struct usb_dmac_chan *uchan)
+{
+	usb_dmac_chan_halt(uchan);
+	devm_free_irq(dmac->dev, uchan->irq, uchan);
+}
+
+static int usb_dmac_remove(struct platform_device *pdev)
+{
+	struct usb_dmac *dmac = platform_get_drvdata(pdev);
+	int i;
+
+	for (i = 0; i < dmac->n_channels; ++i)
+		usb_dmac_chan_remove(dmac, &dmac->channels[i]);
+	of_dma_controller_free(pdev->dev.of_node);
+	dma_async_device_unregister(&dmac->engine);
+
+	pm_runtime_disable(&pdev->dev);
+
+	return 0;
+}
+
+static void usb_dmac_shutdown(struct platform_device *pdev)
+{
+	struct usb_dmac *dmac = platform_get_drvdata(pdev);
+
+	usb_dmac_stop(dmac);
+}
+
+static const struct of_device_id usb_dmac_of_ids[] = {
+	{ .compatible = "renesas,usb-dmac", },
+	{ /* Sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, usb_dmac_of_ids);
+
+static struct platform_driver usb_dmac_driver = {
+	.driver		= {
+		.pm	= &usb_dmac_pm,
+		.name	= "usb-dmac",
+		.of_match_table = usb_dmac_of_ids,
+	},
+	.probe		= usb_dmac_probe,
+	.remove		= usb_dmac_remove,
+	.shutdown	= usb_dmac_shutdown,
+};
+
+module_platform_driver(usb_dmac_driver);
+
+MODULE_DESCRIPTION("Renesas USB DMA Controller Driver");
+MODULE_AUTHOR("Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/dma/sirf-dma.c b/drivers/dma/sirf-dma.c
index d0086e9f2082..a1afda43b8ef 100644
--- a/drivers/dma/sirf-dma.c
+++ b/drivers/dma/sirf-dma.c
@@ -896,7 +896,7 @@ static const struct dev_pm_ops sirfsoc_dma_pm_ops = {
 	SET_SYSTEM_SLEEP_PM_OPS(sirfsoc_dma_pm_suspend, sirfsoc_dma_pm_resume)
 };
 
-static struct of_device_id sirfsoc_dma_match[] = {
+static const struct of_device_id sirfsoc_dma_match[] = {
 	{ .compatible = "sirf,prima2-dmac", },
 	{ .compatible = "sirf,marco-dmac", },
 	{},
diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c
index 1332b1d4d541..3c10f034d4b9 100644
--- a/drivers/dma/ste_dma40.c
+++ b/drivers/dma/ste_dma40.c
@@ -2514,7 +2514,8 @@ static struct dma_async_tx_descriptor *d40_prep_memcpy(struct dma_chan *chan,
 	sg_dma_len(&dst_sg) = size;
 	sg_dma_len(&src_sg) = size;
 
-	return d40_prep_sg(chan, &src_sg, &dst_sg, 1, DMA_NONE, dma_flags);
+	return d40_prep_sg(chan, &src_sg, &dst_sg, 1,
+			   DMA_MEM_TO_MEM, dma_flags);
 }
 
 static struct dma_async_tx_descriptor *
@@ -2526,7 +2527,8 @@ d40_prep_memcpy_sg(struct dma_chan *chan,
 	if (dst_nents != src_nents)
 		return NULL;
 
-	return d40_prep_sg(chan, src_sg, dst_sg, src_nents, DMA_NONE, dma_flags);
+	return d40_prep_sg(chan, src_sg, dst_sg, src_nents,
+			   DMA_MEM_TO_MEM, dma_flags);
 }
 
 static struct dma_async_tx_descriptor *
diff --git a/drivers/dma/sun6i-dma.c b/drivers/dma/sun6i-dma.c
index 7ebcf9bec698..11e536586812 100644
--- a/drivers/dma/sun6i-dma.c
+++ b/drivers/dma/sun6i-dma.c
@@ -796,11 +796,6 @@ static void sun6i_dma_issue_pending(struct dma_chan *chan)
 	spin_unlock_irqrestore(&vchan->vc.lock, flags);
 }
 
-static int sun6i_dma_alloc_chan_resources(struct dma_chan *chan)
-{
-	return 0;
-}
-
 static void sun6i_dma_free_chan_resources(struct dma_chan *chan)
 {
 	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(chan->device);
@@ -896,7 +891,7 @@ static struct sun6i_dma_config sun8i_a23_dma_cfg = {
 	.nr_max_vchans   = 37,
 };
 
-static struct of_device_id sun6i_dma_match[] = {
+static const struct of_device_id sun6i_dma_match[] = {
 	{ .compatible = "allwinner,sun6i-a31-dma", .data = &sun6i_a31_dma_cfg },
 	{ .compatible = "allwinner,sun8i-a23-dma", .data = &sun8i_a23_dma_cfg },
 	{ /* sentinel */ }
@@ -957,7 +952,6 @@ static int sun6i_dma_probe(struct platform_device *pdev)
 	dma_cap_set(DMA_SLAVE, sdc->slave.cap_mask);
 
 	INIT_LIST_HEAD(&sdc->slave.channels);
-	sdc->slave.device_alloc_chan_resources	= sun6i_dma_alloc_chan_resources;
 	sdc->slave.device_free_chan_resources	= sun6i_dma_free_chan_resources;
 	sdc->slave.device_tx_status		= sun6i_dma_tx_status;
 	sdc->slave.device_issue_pending		= sun6i_dma_issue_pending;
diff --git a/drivers/dma/xgene-dma.c b/drivers/dma/xgene-dma.c
new file mode 100755
index 000000000000..f52e37502254
--- /dev/null
+++ b/drivers/dma/xgene-dma.c
@@ -0,0 +1,2089 @@
+/*
+ * Applied Micro X-Gene SoC DMA engine Driver
+ *
+ * Copyright (c) 2015, Applied Micro Circuits Corporation
+ * Authors: Rameshwar Prasad Sahu <rsahu@apm.com>
+ *	    Loc Ho <lho@apm.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;  either version 2 of the  License, or (at your
+ * option) any later version.
+ *
+ * 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.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ * NOTE: PM support is currently not available.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/dmapool.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+
+#include "dmaengine.h"
+
+/* X-Gene DMA ring csr registers and bit definations */
+#define XGENE_DMA_RING_CONFIG			0x04
+#define XGENE_DMA_RING_ENABLE			BIT(31)
+#define XGENE_DMA_RING_ID			0x08
+#define XGENE_DMA_RING_ID_SETUP(v)		((v) | BIT(31))
+#define XGENE_DMA_RING_ID_BUF			0x0C
+#define XGENE_DMA_RING_ID_BUF_SETUP(v)		(((v) << 9) | BIT(21))
+#define XGENE_DMA_RING_THRESLD0_SET1		0x30
+#define XGENE_DMA_RING_THRESLD0_SET1_VAL	0X64
+#define XGENE_DMA_RING_THRESLD1_SET1		0x34
+#define XGENE_DMA_RING_THRESLD1_SET1_VAL	0xC8
+#define XGENE_DMA_RING_HYSTERESIS		0x68
+#define XGENE_DMA_RING_HYSTERESIS_VAL		0xFFFFFFFF
+#define XGENE_DMA_RING_STATE			0x6C
+#define XGENE_DMA_RING_STATE_WR_BASE		0x70
+#define XGENE_DMA_RING_NE_INT_MODE		0x017C
+#define XGENE_DMA_RING_NE_INT_MODE_SET(m, v)	\
+	((m) = ((m) & ~BIT(31 - (v))) | BIT(31 - (v)))
+#define XGENE_DMA_RING_NE_INT_MODE_RESET(m, v)	\
+	((m) &= (~BIT(31 - (v))))
+#define XGENE_DMA_RING_CLKEN			0xC208
+#define XGENE_DMA_RING_SRST			0xC200
+#define XGENE_DMA_RING_MEM_RAM_SHUTDOWN		0xD070
+#define XGENE_DMA_RING_BLK_MEM_RDY		0xD074
+#define XGENE_DMA_RING_BLK_MEM_RDY_VAL		0xFFFFFFFF
+#define XGENE_DMA_RING_DESC_CNT(v)		(((v) & 0x0001FFFE) >> 1)
+#define XGENE_DMA_RING_ID_GET(owner, num)	(((owner) << 6) | (num))
+#define XGENE_DMA_RING_DST_ID(v)		((1 << 10) | (v))
+#define XGENE_DMA_RING_CMD_OFFSET		0x2C
+#define XGENE_DMA_RING_CMD_BASE_OFFSET(v)	((v) << 6)
+#define XGENE_DMA_RING_COHERENT_SET(m)		\
+	(((u32 *)(m))[2] |= BIT(4))
+#define XGENE_DMA_RING_ADDRL_SET(m, v)		\
+	(((u32 *)(m))[2] |= (((v) >> 8) << 5))
+#define XGENE_DMA_RING_ADDRH_SET(m, v)		\
+	(((u32 *)(m))[3] |= ((v) >> 35))
+#define XGENE_DMA_RING_ACCEPTLERR_SET(m)	\
+	(((u32 *)(m))[3] |= BIT(19))
+#define XGENE_DMA_RING_SIZE_SET(m, v)		\
+	(((u32 *)(m))[3] |= ((v) << 23))
+#define XGENE_DMA_RING_RECOMBBUF_SET(m)		\
+	(((u32 *)(m))[3] |= BIT(27))
+#define XGENE_DMA_RING_RECOMTIMEOUTL_SET(m)	\
+	(((u32 *)(m))[3] |= (0x7 << 28))
+#define XGENE_DMA_RING_RECOMTIMEOUTH_SET(m)	\
+	(((u32 *)(m))[4] |= 0x3)
+#define XGENE_DMA_RING_SELTHRSH_SET(m)		\
+	(((u32 *)(m))[4] |= BIT(3))
+#define XGENE_DMA_RING_TYPE_SET(m, v)		\
+	(((u32 *)(m))[4] |= ((v) << 19))
+
+/* X-Gene DMA device csr registers and bit definitions */
+#define XGENE_DMA_IPBRR				0x0
+#define XGENE_DMA_DEV_ID_RD(v)			((v) & 0x00000FFF)
+#define XGENE_DMA_BUS_ID_RD(v)			(((v) >> 12) & 3)
+#define XGENE_DMA_REV_NO_RD(v)			(((v) >> 14) & 3)
+#define XGENE_DMA_GCR				0x10
+#define XGENE_DMA_CH_SETUP(v)			\
+	((v) = ((v) & ~0x000FFFFF) | 0x000AAFFF)
+#define XGENE_DMA_ENABLE(v)			((v) |= BIT(31))
+#define XGENE_DMA_DISABLE(v)			((v) &= ~BIT(31))
+#define XGENE_DMA_RAID6_CONT			0x14
+#define XGENE_DMA_RAID6_MULTI_CTRL(v)		((v) << 24)
+#define XGENE_DMA_INT				0x70
+#define XGENE_DMA_INT_MASK			0x74
+#define XGENE_DMA_INT_ALL_MASK			0xFFFFFFFF
+#define XGENE_DMA_INT_ALL_UNMASK		0x0
+#define XGENE_DMA_INT_MASK_SHIFT		0x14
+#define XGENE_DMA_RING_INT0_MASK		0x90A0
+#define XGENE_DMA_RING_INT1_MASK		0x90A8
+#define XGENE_DMA_RING_INT2_MASK		0x90B0
+#define XGENE_DMA_RING_INT3_MASK		0x90B8
+#define XGENE_DMA_RING_INT4_MASK		0x90C0
+#define XGENE_DMA_CFG_RING_WQ_ASSOC		0x90E0
+#define XGENE_DMA_ASSOC_RING_MNGR1		0xFFFFFFFF
+#define XGENE_DMA_MEM_RAM_SHUTDOWN		0xD070
+#define XGENE_DMA_BLK_MEM_RDY			0xD074
+#define XGENE_DMA_BLK_MEM_RDY_VAL		0xFFFFFFFF
+
+/* X-Gene SoC EFUSE csr register and bit defination */
+#define XGENE_SOC_JTAG1_SHADOW			0x18
+#define XGENE_DMA_PQ_DISABLE_MASK		BIT(13)
+
+/* X-Gene DMA Descriptor format */
+#define XGENE_DMA_DESC_NV_BIT			BIT_ULL(50)
+#define XGENE_DMA_DESC_IN_BIT			BIT_ULL(55)
+#define XGENE_DMA_DESC_C_BIT			BIT_ULL(63)
+#define XGENE_DMA_DESC_DR_BIT			BIT_ULL(61)
+#define XGENE_DMA_DESC_ELERR_POS		46
+#define XGENE_DMA_DESC_RTYPE_POS		56
+#define XGENE_DMA_DESC_LERR_POS			60
+#define XGENE_DMA_DESC_FLYBY_POS		4
+#define XGENE_DMA_DESC_BUFLEN_POS		48
+#define XGENE_DMA_DESC_HOENQ_NUM_POS		48
+
+#define XGENE_DMA_DESC_NV_SET(m)		\
+	(((u64 *)(m))[0] |= XGENE_DMA_DESC_NV_BIT)
+#define XGENE_DMA_DESC_IN_SET(m)		\
+	(((u64 *)(m))[0] |= XGENE_DMA_DESC_IN_BIT)
+#define XGENE_DMA_DESC_RTYPE_SET(m, v)		\
+	(((u64 *)(m))[0] |= ((u64)(v) << XGENE_DMA_DESC_RTYPE_POS))
+#define XGENE_DMA_DESC_BUFADDR_SET(m, v)	\
+	(((u64 *)(m))[0] |= (v))
+#define XGENE_DMA_DESC_BUFLEN_SET(m, v)		\
+	(((u64 *)(m))[0] |= ((u64)(v) << XGENE_DMA_DESC_BUFLEN_POS))
+#define XGENE_DMA_DESC_C_SET(m)			\
+	(((u64 *)(m))[1] |= XGENE_DMA_DESC_C_BIT)
+#define XGENE_DMA_DESC_FLYBY_SET(m, v)		\
+	(((u64 *)(m))[2] |= ((v) << XGENE_DMA_DESC_FLYBY_POS))
+#define XGENE_DMA_DESC_MULTI_SET(m, v, i)	\
+	(((u64 *)(m))[2] |= ((u64)(v) << (((i) + 1) * 8)))
+#define XGENE_DMA_DESC_DR_SET(m)		\
+	(((u64 *)(m))[2] |= XGENE_DMA_DESC_DR_BIT)
+#define XGENE_DMA_DESC_DST_ADDR_SET(m, v)	\
+	(((u64 *)(m))[3] |= (v))
+#define XGENE_DMA_DESC_H0ENQ_NUM_SET(m, v)	\
+	(((u64 *)(m))[3] |= ((u64)(v) << XGENE_DMA_DESC_HOENQ_NUM_POS))
+#define XGENE_DMA_DESC_ELERR_RD(m)		\
+	(((m) >> XGENE_DMA_DESC_ELERR_POS) & 0x3)
+#define XGENE_DMA_DESC_LERR_RD(m)		\
+	(((m) >> XGENE_DMA_DESC_LERR_POS) & 0x7)
+#define XGENE_DMA_DESC_STATUS(elerr, lerr)	\
+	(((elerr) << 4) | (lerr))
+
+/* X-Gene DMA descriptor empty s/w signature */
+#define XGENE_DMA_DESC_EMPTY_INDEX		0
+#define XGENE_DMA_DESC_EMPTY_SIGNATURE		~0ULL
+#define XGENE_DMA_DESC_SET_EMPTY(m)		\
+	(((u64 *)(m))[XGENE_DMA_DESC_EMPTY_INDEX] =	\
+	 XGENE_DMA_DESC_EMPTY_SIGNATURE)
+#define XGENE_DMA_DESC_IS_EMPTY(m)		\
+	(((u64 *)(m))[XGENE_DMA_DESC_EMPTY_INDEX] ==	\
+	 XGENE_DMA_DESC_EMPTY_SIGNATURE)
+
+/* X-Gene DMA configurable parameters defines */
+#define XGENE_DMA_RING_NUM		512
+#define XGENE_DMA_BUFNUM		0x0
+#define XGENE_DMA_CPU_BUFNUM		0x18
+#define XGENE_DMA_RING_OWNER_DMA	0x03
+#define XGENE_DMA_RING_OWNER_CPU	0x0F
+#define XGENE_DMA_RING_TYPE_REGULAR	0x01
+#define XGENE_DMA_RING_WQ_DESC_SIZE	32	/* 32 Bytes */
+#define XGENE_DMA_RING_NUM_CONFIG	5
+#define XGENE_DMA_MAX_CHANNEL		4
+#define XGENE_DMA_XOR_CHANNEL		0
+#define XGENE_DMA_PQ_CHANNEL		1
+#define XGENE_DMA_MAX_BYTE_CNT		0x4000	/* 16 KB */
+#define XGENE_DMA_MAX_64B_DESC_BYTE_CNT	0x14000	/* 80 KB */
+#define XGENE_DMA_XOR_ALIGNMENT		6	/* 64 Bytes */
+#define XGENE_DMA_MAX_XOR_SRC		5
+#define XGENE_DMA_16K_BUFFER_LEN_CODE	0x0
+#define XGENE_DMA_INVALID_LEN_CODE	0x7800
+
+/* X-Gene DMA descriptor error codes */
+#define ERR_DESC_AXI			0x01
+#define ERR_BAD_DESC			0x02
+#define ERR_READ_DATA_AXI		0x03
+#define ERR_WRITE_DATA_AXI		0x04
+#define ERR_FBP_TIMEOUT			0x05
+#define ERR_ECC				0x06
+#define ERR_DIFF_SIZE			0x08
+#define ERR_SCT_GAT_LEN			0x09
+#define ERR_CRC_ERR			0x11
+#define ERR_CHKSUM			0x12
+#define ERR_DIF				0x13
+
+/* X-Gene DMA error interrupt codes */
+#define ERR_DIF_SIZE_INT		0x0
+#define ERR_GS_ERR_INT			0x1
+#define ERR_FPB_TIMEO_INT		0x2
+#define ERR_WFIFO_OVF_INT		0x3
+#define ERR_RFIFO_OVF_INT		0x4
+#define ERR_WR_TIMEO_INT		0x5
+#define ERR_RD_TIMEO_INT		0x6
+#define ERR_WR_ERR_INT			0x7
+#define ERR_RD_ERR_INT			0x8
+#define ERR_BAD_DESC_INT		0x9
+#define ERR_DESC_DST_INT		0xA
+#define ERR_DESC_SRC_INT		0xB
+
+/* X-Gene DMA flyby operation code */
+#define FLYBY_2SRC_XOR			0x8
+#define FLYBY_3SRC_XOR			0x9
+#define FLYBY_4SRC_XOR			0xA
+#define FLYBY_5SRC_XOR			0xB
+
+/* X-Gene DMA SW descriptor flags */
+#define XGENE_DMA_FLAG_64B_DESC		BIT(0)
+
+/* Define to dump X-Gene DMA descriptor */
+#define XGENE_DMA_DESC_DUMP(desc, m)	\
+	print_hex_dump(KERN_ERR, (m),	\
+			DUMP_PREFIX_ADDRESS, 16, 8, (desc), 32, 0)
+
+#define to_dma_desc_sw(tx)		\
+	container_of(tx, struct xgene_dma_desc_sw, tx)
+#define to_dma_chan(dchan)		\
+	container_of(dchan, struct xgene_dma_chan, dma_chan)
+
+#define chan_dbg(chan, fmt, arg...)	\
+	dev_dbg(chan->dev, "%s: " fmt, chan->name, ##arg)
+#define chan_err(chan, fmt, arg...)	\
+	dev_err(chan->dev, "%s: " fmt, chan->name, ##arg)
+
+struct xgene_dma_desc_hw {
+	u64 m0;
+	u64 m1;
+	u64 m2;
+	u64 m3;
+};
+
+enum xgene_dma_ring_cfgsize {
+	XGENE_DMA_RING_CFG_SIZE_512B,
+	XGENE_DMA_RING_CFG_SIZE_2KB,
+	XGENE_DMA_RING_CFG_SIZE_16KB,
+	XGENE_DMA_RING_CFG_SIZE_64KB,
+	XGENE_DMA_RING_CFG_SIZE_512KB,
+	XGENE_DMA_RING_CFG_SIZE_INVALID
+};
+
+struct xgene_dma_ring {
+	struct xgene_dma *pdma;
+	u8 buf_num;
+	u16 id;
+	u16 num;
+	u16 head;
+	u16 owner;
+	u16 slots;
+	u16 dst_ring_num;
+	u32 size;
+	void __iomem *cmd;
+	void __iomem *cmd_base;
+	dma_addr_t desc_paddr;
+	u32 state[XGENE_DMA_RING_NUM_CONFIG];
+	enum xgene_dma_ring_cfgsize cfgsize;
+	union {
+		void *desc_vaddr;
+		struct xgene_dma_desc_hw *desc_hw;
+	};
+};
+
+struct xgene_dma_desc_sw {
+	struct xgene_dma_desc_hw desc1;
+	struct xgene_dma_desc_hw desc2;
+	u32 flags;
+	struct list_head node;
+	struct list_head tx_list;
+	struct dma_async_tx_descriptor tx;
+};
+
+/**
+ * struct xgene_dma_chan - internal representation of an X-Gene DMA channel
+ * @dma_chan: dmaengine channel object member
+ * @pdma: X-Gene DMA device structure reference
+ * @dev: struct device reference for dma mapping api
+ * @id: raw id of this channel
+ * @rx_irq: channel IRQ
+ * @name: name of X-Gene DMA channel
+ * @lock: serializes enqueue/dequeue operations to the descriptor pool
+ * @pending: number of transaction request pushed to DMA controller for
+ *	execution, but still waiting for completion,
+ * @max_outstanding: max number of outstanding request we can push to channel
+ * @ld_pending: descriptors which are queued to run, but have not yet been
+ *	submitted to the hardware for execution
+ * @ld_running: descriptors which are currently being executing by the hardware
+ * @ld_completed: descriptors which have finished execution by the hardware.
+ *	These descriptors have already had their cleanup actions run. They
+ *	are waiting for the ACK bit to be set by the async tx API.
+ * @desc_pool: descriptor pool for DMA operations
+ * @tasklet: bottom half where all completed descriptors cleans
+ * @tx_ring: transmit ring descriptor that we use to prepare actual
+ *	descriptors for further executions
+ * @rx_ring: receive ring descriptor that we use to get completed DMA
+ *	descriptors during cleanup time
+ */
+struct xgene_dma_chan {
+	struct dma_chan dma_chan;
+	struct xgene_dma *pdma;
+	struct device *dev;
+	int id;
+	int rx_irq;
+	char name[10];
+	spinlock_t lock;
+	int pending;
+	int max_outstanding;
+	struct list_head ld_pending;
+	struct list_head ld_running;
+	struct list_head ld_completed;
+	struct dma_pool *desc_pool;
+	struct tasklet_struct tasklet;
+	struct xgene_dma_ring tx_ring;
+	struct xgene_dma_ring rx_ring;
+};
+
+/**
+ * struct xgene_dma - internal representation of an X-Gene DMA device
+ * @err_irq: DMA error irq number
+ * @ring_num: start id number for DMA ring
+ * @csr_dma: base for DMA register access
+ * @csr_ring: base for DMA ring register access
+ * @csr_ring_cmd: base for DMA ring command register access
+ * @csr_efuse: base for efuse register access
+ * @dma_dev: embedded struct dma_device
+ * @chan: reference to X-Gene DMA channels
+ */
+struct xgene_dma {
+	struct device *dev;
+	struct clk *clk;
+	int err_irq;
+	int ring_num;
+	void __iomem *csr_dma;
+	void __iomem *csr_ring;
+	void __iomem *csr_ring_cmd;
+	void __iomem *csr_efuse;
+	struct dma_device dma_dev[XGENE_DMA_MAX_CHANNEL];
+	struct xgene_dma_chan chan[XGENE_DMA_MAX_CHANNEL];
+};
+
+static const char * const xgene_dma_desc_err[] = {
+	[ERR_DESC_AXI] = "AXI error when reading src/dst link list",
+	[ERR_BAD_DESC] = "ERR or El_ERR fields not set to zero in desc",
+	[ERR_READ_DATA_AXI] = "AXI error when reading data",
+	[ERR_WRITE_DATA_AXI] = "AXI error when writing data",
+	[ERR_FBP_TIMEOUT] = "Timeout on bufpool fetch",
+	[ERR_ECC] = "ECC double bit error",
+	[ERR_DIFF_SIZE] = "Bufpool too small to hold all the DIF result",
+	[ERR_SCT_GAT_LEN] = "Gather and scatter data length not same",
+	[ERR_CRC_ERR] = "CRC error",
+	[ERR_CHKSUM] = "Checksum error",
+	[ERR_DIF] = "DIF error",
+};
+
+static const char * const xgene_dma_err[] = {
+	[ERR_DIF_SIZE_INT] = "DIF size error",
+	[ERR_GS_ERR_INT] = "Gather scatter not same size error",
+	[ERR_FPB_TIMEO_INT] = "Free pool time out error",
+	[ERR_WFIFO_OVF_INT] = "Write FIFO over flow error",
+	[ERR_RFIFO_OVF_INT] = "Read FIFO over flow error",
+	[ERR_WR_TIMEO_INT] = "Write time out error",
+	[ERR_RD_TIMEO_INT] = "Read time out error",
+	[ERR_WR_ERR_INT] = "HBF bus write error",
+	[ERR_RD_ERR_INT] = "HBF bus read error",
+	[ERR_BAD_DESC_INT] = "Ring descriptor HE0 not set error",
+	[ERR_DESC_DST_INT] = "HFB reading dst link address error",
+	[ERR_DESC_SRC_INT] = "HFB reading src link address error",
+};
+
+static bool is_pq_enabled(struct xgene_dma *pdma)
+{
+	u32 val;
+
+	val = ioread32(pdma->csr_efuse + XGENE_SOC_JTAG1_SHADOW);
+	return !(val & XGENE_DMA_PQ_DISABLE_MASK);
+}
+
+static void xgene_dma_cpu_to_le64(u64 *desc, int count)
+{
+	int i;
+
+	for (i = 0; i < count; i++)
+		desc[i] = cpu_to_le64(desc[i]);
+}
+
+static u16 xgene_dma_encode_len(u32 len)
+{
+	return (len < XGENE_DMA_MAX_BYTE_CNT) ?
+		len : XGENE_DMA_16K_BUFFER_LEN_CODE;
+}
+
+static u8 xgene_dma_encode_xor_flyby(u32 src_cnt)
+{
+	static u8 flyby_type[] = {
+		FLYBY_2SRC_XOR, /* Dummy */
+		FLYBY_2SRC_XOR, /* Dummy */
+		FLYBY_2SRC_XOR,
+		FLYBY_3SRC_XOR,
+		FLYBY_4SRC_XOR,
+		FLYBY_5SRC_XOR
+	};
+
+	return flyby_type[src_cnt];
+}
+
+static u32 xgene_dma_ring_desc_cnt(struct xgene_dma_ring *ring)
+{
+	u32 __iomem *cmd_base = ring->cmd_base;
+	u32 ring_state = ioread32(&cmd_base[1]);
+
+	return XGENE_DMA_RING_DESC_CNT(ring_state);
+}
+
+static void xgene_dma_set_src_buffer(void *ext8, size_t *len,
+				     dma_addr_t *paddr)
+{
+	size_t nbytes = (*len < XGENE_DMA_MAX_BYTE_CNT) ?
+			*len : XGENE_DMA_MAX_BYTE_CNT;
+
+	XGENE_DMA_DESC_BUFADDR_SET(ext8, *paddr);
+	XGENE_DMA_DESC_BUFLEN_SET(ext8, xgene_dma_encode_len(nbytes));
+	*len -= nbytes;
+	*paddr += nbytes;
+}
+
+static void xgene_dma_invalidate_buffer(void *ext8)
+{
+	XGENE_DMA_DESC_BUFLEN_SET(ext8, XGENE_DMA_INVALID_LEN_CODE);
+}
+
+static void *xgene_dma_lookup_ext8(u64 *desc, int idx)
+{
+	return (idx % 2) ? (desc + idx - 1) : (desc + idx + 1);
+}
+
+static void xgene_dma_init_desc(void *desc, u16 dst_ring_num)
+{
+	XGENE_DMA_DESC_C_SET(desc); /* Coherent IO */
+	XGENE_DMA_DESC_IN_SET(desc);
+	XGENE_DMA_DESC_H0ENQ_NUM_SET(desc, dst_ring_num);
+	XGENE_DMA_DESC_RTYPE_SET(desc, XGENE_DMA_RING_OWNER_DMA);
+}
+
+static void xgene_dma_prep_cpy_desc(struct xgene_dma_chan *chan,
+				    struct xgene_dma_desc_sw *desc_sw,
+				    dma_addr_t dst, dma_addr_t src,
+				    size_t len)
+{
+	void *desc1, *desc2;
+	int i;
+
+	/* Get 1st descriptor */
+	desc1 = &desc_sw->desc1;
+	xgene_dma_init_desc(desc1, chan->tx_ring.dst_ring_num);
+
+	/* Set destination address */
+	XGENE_DMA_DESC_DR_SET(desc1);
+	XGENE_DMA_DESC_DST_ADDR_SET(desc1, dst);
+
+	/* Set 1st source address */
+	xgene_dma_set_src_buffer(desc1 + 8, &len, &src);
+
+	if (len <= 0) {
+		desc2 = NULL;
+		goto skip_additional_src;
+	}
+
+	/*
+	 * We need to split this source buffer,
+	 * and need to use 2nd descriptor
+	 */
+	desc2 = &desc_sw->desc2;
+	XGENE_DMA_DESC_NV_SET(desc1);
+
+	/* Set 2nd to 5th source address */
+	for (i = 0; i < 4 && len; i++)
+		xgene_dma_set_src_buffer(xgene_dma_lookup_ext8(desc2, i),
+					 &len, &src);
+
+	/* Invalidate unused source address field */
+	for (; i < 4; i++)
+		xgene_dma_invalidate_buffer(xgene_dma_lookup_ext8(desc2, i));
+
+	/* Updated flag that we have prepared 64B descriptor */
+	desc_sw->flags |= XGENE_DMA_FLAG_64B_DESC;
+
+skip_additional_src:
+	/* Hardware stores descriptor in little endian format */
+	xgene_dma_cpu_to_le64(desc1, 4);
+	if (desc2)
+		xgene_dma_cpu_to_le64(desc2, 4);
+}
+
+static void xgene_dma_prep_xor_desc(struct xgene_dma_chan *chan,
+				    struct xgene_dma_desc_sw *desc_sw,
+				    dma_addr_t *dst, dma_addr_t *src,
+				    u32 src_cnt, size_t *nbytes,
+				    const u8 *scf)
+{
+	void *desc1, *desc2;
+	size_t len = *nbytes;
+	int i;
+
+	desc1 = &desc_sw->desc1;
+	desc2 = &desc_sw->desc2;
+
+	/* Initialize DMA descriptor */
+	xgene_dma_init_desc(desc1, chan->tx_ring.dst_ring_num);
+
+	/* Set destination address */
+	XGENE_DMA_DESC_DR_SET(desc1);
+	XGENE_DMA_DESC_DST_ADDR_SET(desc1, *dst);
+
+	/* We have multiple source addresses, so need to set NV bit*/
+	XGENE_DMA_DESC_NV_SET(desc1);
+
+	/* Set flyby opcode */
+	XGENE_DMA_DESC_FLYBY_SET(desc1, xgene_dma_encode_xor_flyby(src_cnt));
+
+	/* Set 1st to 5th source addresses */
+	for (i = 0; i < src_cnt; i++) {
+		len = *nbytes;
+		xgene_dma_set_src_buffer((i == 0) ? (desc1 + 8) :
+					 xgene_dma_lookup_ext8(desc2, i - 1),
+					 &len, &src[i]);
+		XGENE_DMA_DESC_MULTI_SET(desc1, scf[i], i);
+	}
+
+	/* Hardware stores descriptor in little endian format */
+	xgene_dma_cpu_to_le64(desc1, 4);
+	xgene_dma_cpu_to_le64(desc2, 4);
+
+	/* Update meta data */
+	*nbytes = len;
+	*dst += XGENE_DMA_MAX_BYTE_CNT;
+
+	/* We need always 64B descriptor to perform xor or pq operations */
+	desc_sw->flags |= XGENE_DMA_FLAG_64B_DESC;
+}
+
+static dma_cookie_t xgene_dma_tx_submit(struct dma_async_tx_descriptor *tx)
+{
+	struct xgene_dma_desc_sw *desc;
+	struct xgene_dma_chan *chan;
+	dma_cookie_t cookie;
+
+	if (unlikely(!tx))
+		return -EINVAL;
+
+	chan = to_dma_chan(tx->chan);
+	desc = to_dma_desc_sw(tx);
+
+	spin_lock_bh(&chan->lock);
+
+	cookie = dma_cookie_assign(tx);
+
+	/* Add this transaction list onto the tail of the pending queue */
+	list_splice_tail_init(&desc->tx_list, &chan->ld_pending);
+
+	spin_unlock_bh(&chan->lock);
+
+	return cookie;
+}
+
+static void xgene_dma_clean_descriptor(struct xgene_dma_chan *chan,
+				       struct xgene_dma_desc_sw *desc)
+{
+	list_del(&desc->node);
+	chan_dbg(chan, "LD %p free\n", desc);
+	dma_pool_free(chan->desc_pool, desc, desc->tx.phys);
+}
+
+static struct xgene_dma_desc_sw *xgene_dma_alloc_descriptor(
+				 struct xgene_dma_chan *chan)
+{
+	struct xgene_dma_desc_sw *desc;
+	dma_addr_t phys;
+
+	desc = dma_pool_alloc(chan->desc_pool, GFP_NOWAIT, &phys);
+	if (!desc) {
+		chan_err(chan, "Failed to allocate LDs\n");
+		return NULL;
+	}
+
+	memset(desc, 0, sizeof(*desc));
+
+	INIT_LIST_HEAD(&desc->tx_list);
+	desc->tx.phys = phys;
+	desc->tx.tx_submit = xgene_dma_tx_submit;
+	dma_async_tx_descriptor_init(&desc->tx, &chan->dma_chan);
+
+	chan_dbg(chan, "LD %p allocated\n", desc);
+
+	return desc;
+}
+
+/**
+ * xgene_dma_clean_completed_descriptor - free all descriptors which
+ * has been completed and acked
+ * @chan: X-Gene DMA channel
+ *
+ * This function is used on all completed and acked descriptors.
+ */
+static void xgene_dma_clean_completed_descriptor(struct xgene_dma_chan *chan)
+{
+	struct xgene_dma_desc_sw *desc, *_desc;
+
+	/* Run the callback for each descriptor, in order */
+	list_for_each_entry_safe(desc, _desc, &chan->ld_completed, node) {
+		if (async_tx_test_ack(&desc->tx))
+			xgene_dma_clean_descriptor(chan, desc);
+	}
+}
+
+/**
+ * xgene_dma_run_tx_complete_actions - cleanup a single link descriptor
+ * @chan: X-Gene DMA channel
+ * @desc: descriptor to cleanup and free
+ *
+ * This function is used on a descriptor which has been executed by the DMA
+ * controller. It will run any callbacks, submit any dependencies.
+ */
+static void xgene_dma_run_tx_complete_actions(struct xgene_dma_chan *chan,
+					      struct xgene_dma_desc_sw *desc)
+{
+	struct dma_async_tx_descriptor *tx = &desc->tx;
+
+	/*
+	 * If this is not the last transaction in the group,
+	 * then no need to complete cookie and run any callback as
+	 * this is not the tx_descriptor which had been sent to caller
+	 * of this DMA request
+	 */
+
+	if (tx->cookie == 0)
+		return;
+
+	dma_cookie_complete(tx);
+
+	/* Run the link descriptor callback function */
+	if (tx->callback)
+		tx->callback(tx->callback_param);
+
+	dma_descriptor_unmap(tx);
+
+	/* Run any dependencies */
+	dma_run_dependencies(tx);
+}
+
+/**
+ * xgene_dma_clean_running_descriptor - move the completed descriptor from
+ * ld_running to ld_completed
+ * @chan: X-Gene DMA channel
+ * @desc: the descriptor which is completed
+ *
+ * Free the descriptor directly if acked by async_tx api,
+ * else move it to queue ld_completed.
+ */
+static void xgene_dma_clean_running_descriptor(struct xgene_dma_chan *chan,
+					       struct xgene_dma_desc_sw *desc)
+{
+	/* Remove from the list of running transactions */
+	list_del(&desc->node);
+
+	/*
+	 * the client is allowed to attach dependent operations
+	 * until 'ack' is set
+	 */
+	if (!async_tx_test_ack(&desc->tx)) {
+		/*
+		 * Move this descriptor to the list of descriptors which is
+		 * completed, but still awaiting the 'ack' bit to be set.
+		 */
+		list_add_tail(&desc->node, &chan->ld_completed);
+		return;
+	}
+
+	chan_dbg(chan, "LD %p free\n", desc);
+	dma_pool_free(chan->desc_pool, desc, desc->tx.phys);
+}
+
+static int xgene_chan_xfer_request(struct xgene_dma_ring *ring,
+				   struct xgene_dma_desc_sw *desc_sw)
+{
+	struct xgene_dma_desc_hw *desc_hw;
+
+	/* Check if can push more descriptor to hw for execution */
+	if (xgene_dma_ring_desc_cnt(ring) > (ring->slots - 2))
+		return -EBUSY;
+
+	/* Get hw descriptor from DMA tx ring */
+	desc_hw = &ring->desc_hw[ring->head];
+
+	/*
+	 * Increment the head count to point next
+	 * descriptor for next time
+	 */
+	if (++ring->head == ring->slots)
+		ring->head = 0;
+
+	/* Copy prepared sw descriptor data to hw descriptor */
+	memcpy(desc_hw, &desc_sw->desc1, sizeof(*desc_hw));
+
+	/*
+	 * Check if we have prepared 64B descriptor,
+	 * in this case we need one more hw descriptor
+	 */
+	if (desc_sw->flags & XGENE_DMA_FLAG_64B_DESC) {
+		desc_hw = &ring->desc_hw[ring->head];
+
+		if (++ring->head == ring->slots)
+			ring->head = 0;
+
+		memcpy(desc_hw, &desc_sw->desc2, sizeof(*desc_hw));
+	}
+
+	/* Notify the hw that we have descriptor ready for execution */
+	iowrite32((desc_sw->flags & XGENE_DMA_FLAG_64B_DESC) ?
+		  2 : 1, ring->cmd);
+
+	return 0;
+}
+
+/**
+ * xgene_chan_xfer_ld_pending - push any pending transactions to hw
+ * @chan : X-Gene DMA channel
+ *
+ * LOCKING: must hold chan->desc_lock
+ */
+static void xgene_chan_xfer_ld_pending(struct xgene_dma_chan *chan)
+{
+	struct xgene_dma_desc_sw *desc_sw, *_desc_sw;
+	int ret;
+
+	/*
+	 * If the list of pending descriptors is empty, then we
+	 * don't need to do any work at all
+	 */
+	if (list_empty(&chan->ld_pending)) {
+		chan_dbg(chan, "No pending LDs\n");
+		return;
+	}
+
+	/*
+	 * Move elements from the queue of pending transactions onto the list
+	 * of running transactions and push it to hw for further executions
+	 */
+	list_for_each_entry_safe(desc_sw, _desc_sw, &chan->ld_pending, node) {
+		/*
+		 * Check if have pushed max number of transactions to hw
+		 * as capable, so let's stop here and will push remaining
+		 * elements from pening ld queue after completing some
+		 * descriptors that we have already pushed
+		 */
+		if (chan->pending >= chan->max_outstanding)
+			return;
+
+		ret = xgene_chan_xfer_request(&chan->tx_ring, desc_sw);
+		if (ret)
+			return;
+
+		/*
+		 * Delete this element from ld pending queue and append it to
+		 * ld running queue
+		 */
+		list_move_tail(&desc_sw->node, &chan->ld_running);
+
+		/* Increment the pending transaction count */
+		chan->pending++;
+	}
+}
+
+/**
+ * xgene_dma_cleanup_descriptors - cleanup link descriptors which are completed
+ * and move them to ld_completed to free until flag 'ack' is set
+ * @chan: X-Gene DMA channel
+ *
+ * This function is used on descriptors which have been executed by the DMA
+ * controller. It will run any callbacks, submit any dependencies, then
+ * free these descriptors if flag 'ack' is set.
+ */
+static void xgene_dma_cleanup_descriptors(struct xgene_dma_chan *chan)
+{
+	struct xgene_dma_ring *ring = &chan->rx_ring;
+	struct xgene_dma_desc_sw *desc_sw, *_desc_sw;
+	struct xgene_dma_desc_hw *desc_hw;
+	u8 status;
+
+	/* Clean already completed and acked descriptors */
+	xgene_dma_clean_completed_descriptor(chan);
+
+	/* Run the callback for each descriptor, in order */
+	list_for_each_entry_safe(desc_sw, _desc_sw, &chan->ld_running, node) {
+		/* Get subsequent hw descriptor from DMA rx ring */
+		desc_hw = &ring->desc_hw[ring->head];
+
+		/* Check if this descriptor has been completed */
+		if (unlikely(XGENE_DMA_DESC_IS_EMPTY(desc_hw)))
+			break;
+
+		if (++ring->head == ring->slots)
+			ring->head = 0;
+
+		/* Check if we have any error with DMA transactions */
+		status = XGENE_DMA_DESC_STATUS(
+				XGENE_DMA_DESC_ELERR_RD(le64_to_cpu(
+							desc_hw->m0)),
+				XGENE_DMA_DESC_LERR_RD(le64_to_cpu(
+						       desc_hw->m0)));
+		if (status) {
+			/* Print the DMA error type */
+			chan_err(chan, "%s\n", xgene_dma_desc_err[status]);
+
+			/*
+			 * We have DMA transactions error here. Dump DMA Tx
+			 * and Rx descriptors for this request */
+			XGENE_DMA_DESC_DUMP(&desc_sw->desc1,
+					    "X-Gene DMA TX DESC1: ");
+
+			if (desc_sw->flags & XGENE_DMA_FLAG_64B_DESC)
+				XGENE_DMA_DESC_DUMP(&desc_sw->desc2,
+						    "X-Gene DMA TX DESC2: ");
+
+			XGENE_DMA_DESC_DUMP(desc_hw,
+					    "X-Gene DMA RX ERR DESC: ");
+		}
+
+		/* Notify the hw about this completed descriptor */
+		iowrite32(-1, ring->cmd);
+
+		/* Mark this hw descriptor as processed */
+		XGENE_DMA_DESC_SET_EMPTY(desc_hw);
+
+		xgene_dma_run_tx_complete_actions(chan, desc_sw);
+
+		xgene_dma_clean_running_descriptor(chan, desc_sw);
+
+		/*
+		 * Decrement the pending transaction count
+		 * as we have processed one
+		 */
+		chan->pending--;
+	}
+
+	/*
+	 * Start any pending transactions automatically
+	 * In the ideal case, we keep the DMA controller busy while we go
+	 * ahead and free the descriptors below.
+	 */
+	xgene_chan_xfer_ld_pending(chan);
+}
+
+static int xgene_dma_alloc_chan_resources(struct dma_chan *dchan)
+{
+	struct xgene_dma_chan *chan = to_dma_chan(dchan);
+
+	/* Has this channel already been allocated? */
+	if (chan->desc_pool)
+		return 1;
+
+	chan->desc_pool = dma_pool_create(chan->name, chan->dev,
+					  sizeof(struct xgene_dma_desc_sw),
+					  0, 0);
+	if (!chan->desc_pool) {
+		chan_err(chan, "Failed to allocate descriptor pool\n");
+		return -ENOMEM;
+	}
+
+	chan_dbg(chan, "Allocate descripto pool\n");
+
+	return 1;
+}
+
+/**
+ * xgene_dma_free_desc_list - Free all descriptors in a queue
+ * @chan: X-Gene DMA channel
+ * @list: the list to free
+ *
+ * LOCKING: must hold chan->desc_lock
+ */
+static void xgene_dma_free_desc_list(struct xgene_dma_chan *chan,
+				     struct list_head *list)
+{
+	struct xgene_dma_desc_sw *desc, *_desc;
+
+	list_for_each_entry_safe(desc, _desc, list, node)
+		xgene_dma_clean_descriptor(chan, desc);
+}
+
+static void xgene_dma_free_tx_desc_list(struct xgene_dma_chan *chan,
+					struct list_head *list)
+{
+	struct xgene_dma_desc_sw *desc, *_desc;
+
+	list_for_each_entry_safe(desc, _desc, list, node)
+		xgene_dma_clean_descriptor(chan, desc);
+}
+
+static void xgene_dma_free_chan_resources(struct dma_chan *dchan)
+{
+	struct xgene_dma_chan *chan = to_dma_chan(dchan);
+
+	chan_dbg(chan, "Free all resources\n");
+
+	if (!chan->desc_pool)
+		return;
+
+	spin_lock_bh(&chan->lock);
+
+	/* Process all running descriptor */
+	xgene_dma_cleanup_descriptors(chan);
+
+	/* Clean all link descriptor queues */
+	xgene_dma_free_desc_list(chan, &chan->ld_pending);
+	xgene_dma_free_desc_list(chan, &chan->ld_running);
+	xgene_dma_free_desc_list(chan, &chan->ld_completed);
+
+	spin_unlock_bh(&chan->lock);
+
+	/* Delete this channel DMA pool */
+	dma_pool_destroy(chan->desc_pool);
+	chan->desc_pool = NULL;
+}
+
+static struct dma_async_tx_descriptor *xgene_dma_prep_memcpy(
+	struct dma_chan *dchan, dma_addr_t dst, dma_addr_t src,
+	size_t len, unsigned long flags)
+{
+	struct xgene_dma_desc_sw *first = NULL, *new;
+	struct xgene_dma_chan *chan;
+	size_t copy;
+
+	if (unlikely(!dchan || !len))
+		return NULL;
+
+	chan = to_dma_chan(dchan);
+
+	do {
+		/* Allocate the link descriptor from DMA pool */
+		new = xgene_dma_alloc_descriptor(chan);
+		if (!new)
+			goto fail;
+
+		/* Create the largest transaction possible */
+		copy = min_t(size_t, len, XGENE_DMA_MAX_64B_DESC_BYTE_CNT);
+
+		/* Prepare DMA descriptor */
+		xgene_dma_prep_cpy_desc(chan, new, dst, src, copy);
+
+		if (!first)
+			first = new;
+
+		new->tx.cookie = 0;
+		async_tx_ack(&new->tx);
+
+		/* Update metadata */
+		len -= copy;
+		dst += copy;
+		src += copy;
+
+		/* Insert the link descriptor to the LD ring */
+		list_add_tail(&new->node, &first->tx_list);
+	} while (len);
+
+	new->tx.flags = flags; /* client is in control of this ack */
+	new->tx.cookie = -EBUSY;
+	list_splice(&first->tx_list, &new->tx_list);
+
+	return &new->tx;
+
+fail:
+	if (!first)
+		return NULL;
+
+	xgene_dma_free_tx_desc_list(chan, &first->tx_list);
+	return NULL;
+}
+
+static struct dma_async_tx_descriptor *xgene_dma_prep_sg(
+	struct dma_chan *dchan, struct scatterlist *dst_sg,
+	u32 dst_nents, struct scatterlist *src_sg,
+	u32 src_nents, unsigned long flags)
+{
+	struct xgene_dma_desc_sw *first = NULL, *new = NULL;
+	struct xgene_dma_chan *chan;
+	size_t dst_avail, src_avail;
+	dma_addr_t dst, src;
+	size_t len;
+
+	if (unlikely(!dchan))
+		return NULL;
+
+	if (unlikely(!dst_nents || !src_nents))
+		return NULL;
+
+	if (unlikely(!dst_sg || !src_sg))
+		return NULL;
+
+	chan = to_dma_chan(dchan);
+
+	/* Get prepared for the loop */
+	dst_avail = sg_dma_len(dst_sg);
+	src_avail = sg_dma_len(src_sg);
+	dst_nents--;
+	src_nents--;
+
+	/* Run until we are out of scatterlist entries */
+	while (true) {
+		/* Create the largest transaction possible */
+		len = min_t(size_t, src_avail, dst_avail);
+		len = min_t(size_t, len, XGENE_DMA_MAX_64B_DESC_BYTE_CNT);
+		if (len == 0)
+			goto fetch;
+
+		dst = sg_dma_address(dst_sg) + sg_dma_len(dst_sg) - dst_avail;
+		src = sg_dma_address(src_sg) + sg_dma_len(src_sg) - src_avail;
+
+		/* Allocate the link descriptor from DMA pool */
+		new = xgene_dma_alloc_descriptor(chan);
+		if (!new)
+			goto fail;
+
+		/* Prepare DMA descriptor */
+		xgene_dma_prep_cpy_desc(chan, new, dst, src, len);
+
+		if (!first)
+			first = new;
+
+		new->tx.cookie = 0;
+		async_tx_ack(&new->tx);
+
+		/* update metadata */
+		dst_avail -= len;
+		src_avail -= len;
+
+		/* Insert the link descriptor to the LD ring */
+		list_add_tail(&new->node, &first->tx_list);
+
+fetch:
+		/* fetch the next dst scatterlist entry */
+		if (dst_avail == 0) {
+			/* no more entries: we're done */
+			if (dst_nents == 0)
+				break;
+
+			/* fetch the next entry: if there are no more: done */
+			dst_sg = sg_next(dst_sg);
+			if (!dst_sg)
+				break;
+
+			dst_nents--;
+			dst_avail = sg_dma_len(dst_sg);
+		}
+
+		/* fetch the next src scatterlist entry */
+		if (src_avail == 0) {
+			/* no more entries: we're done */
+			if (src_nents == 0)
+				break;
+
+			/* fetch the next entry: if there are no more: done */
+			src_sg = sg_next(src_sg);
+			if (!src_sg)
+				break;
+
+			src_nents--;
+			src_avail = sg_dma_len(src_sg);
+		}
+	}
+
+	if (!new)
+		return NULL;
+
+	new->tx.flags = flags; /* client is in control of this ack */
+	new->tx.cookie = -EBUSY;
+	list_splice(&first->tx_list, &new->tx_list);
+
+	return &new->tx;
+fail:
+	if (!first)
+		return NULL;
+
+	xgene_dma_free_tx_desc_list(chan, &first->tx_list);
+	return NULL;
+}
+
+static struct dma_async_tx_descriptor *xgene_dma_prep_xor(
+	struct dma_chan *dchan, dma_addr_t dst,	dma_addr_t *src,
+	u32 src_cnt, size_t len, unsigned long flags)
+{
+	struct xgene_dma_desc_sw *first = NULL, *new;
+	struct xgene_dma_chan *chan;
+	static u8 multi[XGENE_DMA_MAX_XOR_SRC] = {
+				0x01, 0x01, 0x01, 0x01, 0x01};
+
+	if (unlikely(!dchan || !len))
+		return NULL;
+
+	chan = to_dma_chan(dchan);
+
+	do {
+		/* Allocate the link descriptor from DMA pool */
+		new = xgene_dma_alloc_descriptor(chan);
+		if (!new)
+			goto fail;
+
+		/* Prepare xor DMA descriptor */
+		xgene_dma_prep_xor_desc(chan, new, &dst, src,
+					src_cnt, &len, multi);
+
+		if (!first)
+			first = new;
+
+		new->tx.cookie = 0;
+		async_tx_ack(&new->tx);
+
+		/* Insert the link descriptor to the LD ring */
+		list_add_tail(&new->node, &first->tx_list);
+	} while (len);
+
+	new->tx.flags = flags; /* client is in control of this ack */
+	new->tx.cookie = -EBUSY;
+	list_splice(&first->tx_list, &new->tx_list);
+
+	return &new->tx;
+
+fail:
+	if (!first)
+		return NULL;
+
+	xgene_dma_free_tx_desc_list(chan, &first->tx_list);
+	return NULL;
+}
+
+static struct dma_async_tx_descriptor *xgene_dma_prep_pq(
+	struct dma_chan *dchan, dma_addr_t *dst, dma_addr_t *src,
+	u32 src_cnt, const u8 *scf, size_t len, unsigned long flags)
+{
+	struct xgene_dma_desc_sw *first = NULL, *new;
+	struct xgene_dma_chan *chan;
+	size_t _len = len;
+	dma_addr_t _src[XGENE_DMA_MAX_XOR_SRC];
+	static u8 multi[XGENE_DMA_MAX_XOR_SRC] = {0x01, 0x01, 0x01, 0x01, 0x01};
+
+	if (unlikely(!dchan || !len))
+		return NULL;
+
+	chan = to_dma_chan(dchan);
+
+	/*
+	 * Save source addresses on local variable, may be we have to
+	 * prepare two descriptor to generate P and Q if both enabled
+	 * in the flags by client
+	 */
+	memcpy(_src, src, sizeof(*src) * src_cnt);
+
+	if (flags & DMA_PREP_PQ_DISABLE_P)
+		len = 0;
+
+	if (flags & DMA_PREP_PQ_DISABLE_Q)
+		_len = 0;
+
+	do {
+		/* Allocate the link descriptor from DMA pool */
+		new = xgene_dma_alloc_descriptor(chan);
+		if (!new)
+			goto fail;
+
+		if (!first)
+			first = new;
+
+		new->tx.cookie = 0;
+		async_tx_ack(&new->tx);
+
+		/* Insert the link descriptor to the LD ring */
+		list_add_tail(&new->node, &first->tx_list);
+
+		/*
+		 * Prepare DMA descriptor to generate P,
+		 * if DMA_PREP_PQ_DISABLE_P flag is not set
+		 */
+		if (len) {
+			xgene_dma_prep_xor_desc(chan, new, &dst[0], src,
+						src_cnt, &len, multi);
+			continue;
+		}
+
+		/*
+		 * Prepare DMA descriptor to generate Q,
+		 * if DMA_PREP_PQ_DISABLE_Q flag is not set
+		 */
+		if (_len) {
+			xgene_dma_prep_xor_desc(chan, new, &dst[1], _src,
+						src_cnt, &_len, scf);
+		}
+	} while (len || _len);
+
+	new->tx.flags = flags; /* client is in control of this ack */
+	new->tx.cookie = -EBUSY;
+	list_splice(&first->tx_list, &new->tx_list);
+
+	return &new->tx;
+
+fail:
+	if (!first)
+		return NULL;
+
+	xgene_dma_free_tx_desc_list(chan, &first->tx_list);
+	return NULL;
+}
+
+static void xgene_dma_issue_pending(struct dma_chan *dchan)
+{
+	struct xgene_dma_chan *chan = to_dma_chan(dchan);
+
+	spin_lock_bh(&chan->lock);
+	xgene_chan_xfer_ld_pending(chan);
+	spin_unlock_bh(&chan->lock);
+}
+
+static enum dma_status xgene_dma_tx_status(struct dma_chan *dchan,
+					   dma_cookie_t cookie,
+					   struct dma_tx_state *txstate)
+{
+	return dma_cookie_status(dchan, cookie, txstate);
+}
+
+static void xgene_dma_tasklet_cb(unsigned long data)
+{
+	struct xgene_dma_chan *chan = (struct xgene_dma_chan *)data;
+
+	spin_lock_bh(&chan->lock);
+
+	/* Run all cleanup for descriptors which have been completed */
+	xgene_dma_cleanup_descriptors(chan);
+
+	/* Re-enable DMA channel IRQ */
+	enable_irq(chan->rx_irq);
+
+	spin_unlock_bh(&chan->lock);
+}
+
+static irqreturn_t xgene_dma_chan_ring_isr(int irq, void *id)
+{
+	struct xgene_dma_chan *chan = (struct xgene_dma_chan *)id;
+
+	BUG_ON(!chan);
+
+	/*
+	 * Disable DMA channel IRQ until we process completed
+	 * descriptors
+	 */
+	disable_irq_nosync(chan->rx_irq);
+
+	/*
+	 * Schedule the tasklet to handle all cleanup of the current
+	 * transaction. It will start a new transaction if there is
+	 * one pending.
+	 */
+	tasklet_schedule(&chan->tasklet);
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t xgene_dma_err_isr(int irq, void *id)
+{
+	struct xgene_dma *pdma = (struct xgene_dma *)id;
+	unsigned long int_mask;
+	u32 val, i;
+
+	val = ioread32(pdma->csr_dma + XGENE_DMA_INT);
+
+	/* Clear DMA interrupts */
+	iowrite32(val, pdma->csr_dma + XGENE_DMA_INT);
+
+	/* Print DMA error info */
+	int_mask = val >> XGENE_DMA_INT_MASK_SHIFT;
+	for_each_set_bit(i, &int_mask, ARRAY_SIZE(xgene_dma_err))
+		dev_err(pdma->dev,
+			"Interrupt status 0x%08X %s\n", val, xgene_dma_err[i]);
+
+	return IRQ_HANDLED;
+}
+
+static void xgene_dma_wr_ring_state(struct xgene_dma_ring *ring)
+{
+	int i;
+
+	iowrite32(ring->num, ring->pdma->csr_ring + XGENE_DMA_RING_STATE);
+
+	for (i = 0; i < XGENE_DMA_RING_NUM_CONFIG; i++)
+		iowrite32(ring->state[i], ring->pdma->csr_ring +
+			  XGENE_DMA_RING_STATE_WR_BASE + (i * 4));
+}
+
+static void xgene_dma_clr_ring_state(struct xgene_dma_ring *ring)
+{
+	memset(ring->state, 0, sizeof(u32) * XGENE_DMA_RING_NUM_CONFIG);
+	xgene_dma_wr_ring_state(ring);
+}
+
+static void xgene_dma_setup_ring(struct xgene_dma_ring *ring)
+{
+	void *ring_cfg = ring->state;
+	u64 addr = ring->desc_paddr;
+	void *desc;
+	u32 i, val;
+
+	ring->slots = ring->size / XGENE_DMA_RING_WQ_DESC_SIZE;
+
+	/* Clear DMA ring state */
+	xgene_dma_clr_ring_state(ring);
+
+	/* Set DMA ring type */
+	XGENE_DMA_RING_TYPE_SET(ring_cfg, XGENE_DMA_RING_TYPE_REGULAR);
+
+	if (ring->owner == XGENE_DMA_RING_OWNER_DMA) {
+		/* Set recombination buffer and timeout */
+		XGENE_DMA_RING_RECOMBBUF_SET(ring_cfg);
+		XGENE_DMA_RING_RECOMTIMEOUTL_SET(ring_cfg);
+		XGENE_DMA_RING_RECOMTIMEOUTH_SET(ring_cfg);
+	}
+
+	/* Initialize DMA ring state */
+	XGENE_DMA_RING_SELTHRSH_SET(ring_cfg);
+	XGENE_DMA_RING_ACCEPTLERR_SET(ring_cfg);
+	XGENE_DMA_RING_COHERENT_SET(ring_cfg);
+	XGENE_DMA_RING_ADDRL_SET(ring_cfg, addr);
+	XGENE_DMA_RING_ADDRH_SET(ring_cfg, addr);
+	XGENE_DMA_RING_SIZE_SET(ring_cfg, ring->cfgsize);
+
+	/* Write DMA ring configurations */
+	xgene_dma_wr_ring_state(ring);
+
+	/* Set DMA ring id */
+	iowrite32(XGENE_DMA_RING_ID_SETUP(ring->id),
+		  ring->pdma->csr_ring + XGENE_DMA_RING_ID);
+
+	/* Set DMA ring buffer */
+	iowrite32(XGENE_DMA_RING_ID_BUF_SETUP(ring->num),
+		  ring->pdma->csr_ring + XGENE_DMA_RING_ID_BUF);
+
+	if (ring->owner != XGENE_DMA_RING_OWNER_CPU)
+		return;
+
+	/* Set empty signature to DMA Rx ring descriptors */
+	for (i = 0; i < ring->slots; i++) {
+		desc = &ring->desc_hw[i];
+		XGENE_DMA_DESC_SET_EMPTY(desc);
+	}
+
+	/* Enable DMA Rx ring interrupt */
+	val = ioread32(ring->pdma->csr_ring + XGENE_DMA_RING_NE_INT_MODE);
+	XGENE_DMA_RING_NE_INT_MODE_SET(val, ring->buf_num);
+	iowrite32(val, ring->pdma->csr_ring + XGENE_DMA_RING_NE_INT_MODE);
+}
+
+static void xgene_dma_clear_ring(struct xgene_dma_ring *ring)
+{
+	u32 ring_id, val;
+
+	if (ring->owner == XGENE_DMA_RING_OWNER_CPU) {
+		/* Disable DMA Rx ring interrupt */
+		val = ioread32(ring->pdma->csr_ring +
+			       XGENE_DMA_RING_NE_INT_MODE);
+		XGENE_DMA_RING_NE_INT_MODE_RESET(val, ring->buf_num);
+		iowrite32(val, ring->pdma->csr_ring +
+			  XGENE_DMA_RING_NE_INT_MODE);
+	}
+
+	/* Clear DMA ring state */
+	ring_id = XGENE_DMA_RING_ID_SETUP(ring->id);
+	iowrite32(ring_id, ring->pdma->csr_ring + XGENE_DMA_RING_ID);
+
+	iowrite32(0, ring->pdma->csr_ring + XGENE_DMA_RING_ID_BUF);
+	xgene_dma_clr_ring_state(ring);
+}
+
+static void xgene_dma_set_ring_cmd(struct xgene_dma_ring *ring)
+{
+	ring->cmd_base = ring->pdma->csr_ring_cmd +
+				XGENE_DMA_RING_CMD_BASE_OFFSET((ring->num -
+							  XGENE_DMA_RING_NUM));
+
+	ring->cmd = ring->cmd_base + XGENE_DMA_RING_CMD_OFFSET;
+}
+
+static int xgene_dma_get_ring_size(struct xgene_dma_chan *chan,
+				   enum xgene_dma_ring_cfgsize cfgsize)
+{
+	int size;
+
+	switch (cfgsize) {
+	case XGENE_DMA_RING_CFG_SIZE_512B:
+		size = 0x200;
+		break;
+	case XGENE_DMA_RING_CFG_SIZE_2KB:
+		size = 0x800;
+		break;
+	case XGENE_DMA_RING_CFG_SIZE_16KB:
+		size = 0x4000;
+		break;
+	case XGENE_DMA_RING_CFG_SIZE_64KB:
+		size = 0x10000;
+		break;
+	case XGENE_DMA_RING_CFG_SIZE_512KB:
+		size = 0x80000;
+		break;
+	default:
+		chan_err(chan, "Unsupported cfg ring size %d\n", cfgsize);
+		return -EINVAL;
+	}
+
+	return size;
+}
+
+static void xgene_dma_delete_ring_one(struct xgene_dma_ring *ring)
+{
+	/* Clear DMA ring configurations */
+	xgene_dma_clear_ring(ring);
+
+	/* De-allocate DMA ring descriptor */
+	if (ring->desc_vaddr) {
+		dma_free_coherent(ring->pdma->dev, ring->size,
+				  ring->desc_vaddr, ring->desc_paddr);
+		ring->desc_vaddr = NULL;
+	}
+}
+
+static void xgene_dma_delete_chan_rings(struct xgene_dma_chan *chan)
+{
+	xgene_dma_delete_ring_one(&chan->rx_ring);
+	xgene_dma_delete_ring_one(&chan->tx_ring);
+}
+
+static int xgene_dma_create_ring_one(struct xgene_dma_chan *chan,
+				     struct xgene_dma_ring *ring,
+				     enum xgene_dma_ring_cfgsize cfgsize)
+{
+	/* Setup DMA ring descriptor variables */
+	ring->pdma = chan->pdma;
+	ring->cfgsize = cfgsize;
+	ring->num = chan->pdma->ring_num++;
+	ring->id = XGENE_DMA_RING_ID_GET(ring->owner, ring->buf_num);
+
+	ring->size = xgene_dma_get_ring_size(chan, cfgsize);
+	if (ring->size <= 0)
+		return ring->size;
+
+	/* Allocate memory for DMA ring descriptor */
+	ring->desc_vaddr = dma_zalloc_coherent(chan->dev, ring->size,
+					       &ring->desc_paddr, GFP_KERNEL);
+	if (!ring->desc_vaddr) {
+		chan_err(chan, "Failed to allocate ring desc\n");
+		return -ENOMEM;
+	}
+
+	/* Configure and enable DMA ring */
+	xgene_dma_set_ring_cmd(ring);
+	xgene_dma_setup_ring(ring);
+
+	return 0;
+}
+
+static int xgene_dma_create_chan_rings(struct xgene_dma_chan *chan)
+{
+	struct xgene_dma_ring *rx_ring = &chan->rx_ring;
+	struct xgene_dma_ring *tx_ring = &chan->tx_ring;
+	int ret;
+
+	/* Create DMA Rx ring descriptor */
+	rx_ring->owner = XGENE_DMA_RING_OWNER_CPU;
+	rx_ring->buf_num = XGENE_DMA_CPU_BUFNUM + chan->id;
+
+	ret = xgene_dma_create_ring_one(chan, rx_ring,
+					XGENE_DMA_RING_CFG_SIZE_64KB);
+	if (ret)
+		return ret;
+
+	chan_dbg(chan, "Rx ring id 0x%X num %d desc 0x%p\n",
+		 rx_ring->id, rx_ring->num, rx_ring->desc_vaddr);
+
+	/* Create DMA Tx ring descriptor */
+	tx_ring->owner = XGENE_DMA_RING_OWNER_DMA;
+	tx_ring->buf_num = XGENE_DMA_BUFNUM + chan->id;
+
+	ret = xgene_dma_create_ring_one(chan, tx_ring,
+					XGENE_DMA_RING_CFG_SIZE_64KB);
+	if (ret) {
+		xgene_dma_delete_ring_one(rx_ring);
+		return ret;
+	}
+
+	tx_ring->dst_ring_num = XGENE_DMA_RING_DST_ID(rx_ring->num);
+
+	chan_dbg(chan,
+		 "Tx ring id 0x%X num %d desc 0x%p\n",
+		 tx_ring->id, tx_ring->num, tx_ring->desc_vaddr);
+
+	/* Set the max outstanding request possible to this channel */
+	chan->max_outstanding = rx_ring->slots;
+
+	return ret;
+}
+
+static int xgene_dma_init_rings(struct xgene_dma *pdma)
+{
+	int ret, i, j;
+
+	for (i = 0; i < XGENE_DMA_MAX_CHANNEL; i++) {
+		ret = xgene_dma_create_chan_rings(&pdma->chan[i]);
+		if (ret) {
+			for (j = 0; j < i; j++)
+				xgene_dma_delete_chan_rings(&pdma->chan[j]);
+			return ret;
+		}
+	}
+
+	return ret;
+}
+
+static void xgene_dma_enable(struct xgene_dma *pdma)
+{
+	u32 val;
+
+	/* Configure and enable DMA engine */
+	val = ioread32(pdma->csr_dma + XGENE_DMA_GCR);
+	XGENE_DMA_CH_SETUP(val);
+	XGENE_DMA_ENABLE(val);
+	iowrite32(val, pdma->csr_dma + XGENE_DMA_GCR);
+}
+
+static void xgene_dma_disable(struct xgene_dma *pdma)
+{
+	u32 val;
+
+	val = ioread32(pdma->csr_dma + XGENE_DMA_GCR);
+	XGENE_DMA_DISABLE(val);
+	iowrite32(val, pdma->csr_dma + XGENE_DMA_GCR);
+}
+
+static void xgene_dma_mask_interrupts(struct xgene_dma *pdma)
+{
+	/*
+	 * Mask DMA ring overflow, underflow and
+	 * AXI write/read error interrupts
+	 */
+	iowrite32(XGENE_DMA_INT_ALL_MASK,
+		  pdma->csr_dma + XGENE_DMA_RING_INT0_MASK);
+	iowrite32(XGENE_DMA_INT_ALL_MASK,
+		  pdma->csr_dma + XGENE_DMA_RING_INT1_MASK);
+	iowrite32(XGENE_DMA_INT_ALL_MASK,
+		  pdma->csr_dma + XGENE_DMA_RING_INT2_MASK);
+	iowrite32(XGENE_DMA_INT_ALL_MASK,
+		  pdma->csr_dma + XGENE_DMA_RING_INT3_MASK);
+	iowrite32(XGENE_DMA_INT_ALL_MASK,
+		  pdma->csr_dma + XGENE_DMA_RING_INT4_MASK);
+
+	/* Mask DMA error interrupts */
+	iowrite32(XGENE_DMA_INT_ALL_MASK, pdma->csr_dma + XGENE_DMA_INT_MASK);
+}
+
+static void xgene_dma_unmask_interrupts(struct xgene_dma *pdma)
+{
+	/*
+	 * Unmask DMA ring overflow, underflow and
+	 * AXI write/read error interrupts
+	 */
+	iowrite32(XGENE_DMA_INT_ALL_UNMASK,
+		  pdma->csr_dma + XGENE_DMA_RING_INT0_MASK);
+	iowrite32(XGENE_DMA_INT_ALL_UNMASK,
+		  pdma->csr_dma + XGENE_DMA_RING_INT1_MASK);
+	iowrite32(XGENE_DMA_INT_ALL_UNMASK,
+		  pdma->csr_dma + XGENE_DMA_RING_INT2_MASK);
+	iowrite32(XGENE_DMA_INT_ALL_UNMASK,
+		  pdma->csr_dma + XGENE_DMA_RING_INT3_MASK);
+	iowrite32(XGENE_DMA_INT_ALL_UNMASK,
+		  pdma->csr_dma + XGENE_DMA_RING_INT4_MASK);
+
+	/* Unmask DMA error interrupts */
+	iowrite32(XGENE_DMA_INT_ALL_UNMASK,
+		  pdma->csr_dma + XGENE_DMA_INT_MASK);
+}
+
+static void xgene_dma_init_hw(struct xgene_dma *pdma)
+{
+	u32 val;
+
+	/* Associate DMA ring to corresponding ring HW */
+	iowrite32(XGENE_DMA_ASSOC_RING_MNGR1,
+		  pdma->csr_dma + XGENE_DMA_CFG_RING_WQ_ASSOC);
+
+	/* Configure RAID6 polynomial control setting */
+	if (is_pq_enabled(pdma))
+		iowrite32(XGENE_DMA_RAID6_MULTI_CTRL(0x1D),
+			  pdma->csr_dma + XGENE_DMA_RAID6_CONT);
+	else
+		dev_info(pdma->dev, "PQ is disabled in HW\n");
+
+	xgene_dma_enable(pdma);
+	xgene_dma_unmask_interrupts(pdma);
+
+	/* Get DMA id and version info */
+	val = ioread32(pdma->csr_dma + XGENE_DMA_IPBRR);
+
+	/* DMA device info */
+	dev_info(pdma->dev,
+		 "X-Gene DMA v%d.%02d.%02d driver registered %d channels",
+		 XGENE_DMA_REV_NO_RD(val), XGENE_DMA_BUS_ID_RD(val),
+		 XGENE_DMA_DEV_ID_RD(val), XGENE_DMA_MAX_CHANNEL);
+}
+
+static int xgene_dma_init_ring_mngr(struct xgene_dma *pdma)
+{
+	if (ioread32(pdma->csr_ring + XGENE_DMA_RING_CLKEN) &&
+	    (!ioread32(pdma->csr_ring + XGENE_DMA_RING_SRST)))
+		return 0;
+
+	iowrite32(0x3, pdma->csr_ring + XGENE_DMA_RING_CLKEN);
+	iowrite32(0x0, pdma->csr_ring + XGENE_DMA_RING_SRST);
+
+	/* Bring up memory */
+	iowrite32(0x0, pdma->csr_ring + XGENE_DMA_RING_MEM_RAM_SHUTDOWN);
+
+	/* Force a barrier */
+	ioread32(pdma->csr_ring + XGENE_DMA_RING_MEM_RAM_SHUTDOWN);
+
+	/* reset may take up to 1ms */
+	usleep_range(1000, 1100);
+
+	if (ioread32(pdma->csr_ring + XGENE_DMA_RING_BLK_MEM_RDY)
+		!= XGENE_DMA_RING_BLK_MEM_RDY_VAL) {
+		dev_err(pdma->dev,
+			"Failed to release ring mngr memory from shutdown\n");
+		return -ENODEV;
+	}
+
+	/* program threshold set 1 and all hysteresis */
+	iowrite32(XGENE_DMA_RING_THRESLD0_SET1_VAL,
+		  pdma->csr_ring + XGENE_DMA_RING_THRESLD0_SET1);
+	iowrite32(XGENE_DMA_RING_THRESLD1_SET1_VAL,
+		  pdma->csr_ring + XGENE_DMA_RING_THRESLD1_SET1);
+	iowrite32(XGENE_DMA_RING_HYSTERESIS_VAL,
+		  pdma->csr_ring + XGENE_DMA_RING_HYSTERESIS);
+
+	/* Enable QPcore and assign error queue */
+	iowrite32(XGENE_DMA_RING_ENABLE,
+		  pdma->csr_ring + XGENE_DMA_RING_CONFIG);
+
+	return 0;
+}
+
+static int xgene_dma_init_mem(struct xgene_dma *pdma)
+{
+	int ret;
+
+	ret = xgene_dma_init_ring_mngr(pdma);
+	if (ret)
+		return ret;
+
+	/* Bring up memory */
+	iowrite32(0x0, pdma->csr_dma + XGENE_DMA_MEM_RAM_SHUTDOWN);
+
+	/* Force a barrier */
+	ioread32(pdma->csr_dma + XGENE_DMA_MEM_RAM_SHUTDOWN);
+
+	/* reset may take up to 1ms */
+	usleep_range(1000, 1100);
+
+	if (ioread32(pdma->csr_dma + XGENE_DMA_BLK_MEM_RDY)
+		!= XGENE_DMA_BLK_MEM_RDY_VAL) {
+		dev_err(pdma->dev,
+			"Failed to release DMA memory from shutdown\n");
+		return -ENODEV;
+	}
+
+	return 0;
+}
+
+static int xgene_dma_request_irqs(struct xgene_dma *pdma)
+{
+	struct xgene_dma_chan *chan;
+	int ret, i, j;
+
+	/* Register DMA error irq */
+	ret = devm_request_irq(pdma->dev, pdma->err_irq, xgene_dma_err_isr,
+			       0, "dma_error", pdma);
+	if (ret) {
+		dev_err(pdma->dev,
+			"Failed to register error IRQ %d\n", pdma->err_irq);
+		return ret;
+	}
+
+	/* Register DMA channel rx irq */
+	for (i = 0; i < XGENE_DMA_MAX_CHANNEL; i++) {
+		chan = &pdma->chan[i];
+		ret = devm_request_irq(chan->dev, chan->rx_irq,
+				       xgene_dma_chan_ring_isr,
+				       0, chan->name, chan);
+		if (ret) {
+			chan_err(chan, "Failed to register Rx IRQ %d\n",
+				 chan->rx_irq);
+			devm_free_irq(pdma->dev, pdma->err_irq, pdma);
+
+			for (j = 0; j < i; j++) {
+				chan = &pdma->chan[i];
+				devm_free_irq(chan->dev, chan->rx_irq, chan);
+			}
+
+			return ret;
+		}
+	}
+
+	return 0;
+}
+
+static void xgene_dma_free_irqs(struct xgene_dma *pdma)
+{
+	struct xgene_dma_chan *chan;
+	int i;
+
+	/* Free DMA device error irq */
+	devm_free_irq(pdma->dev, pdma->err_irq, pdma);
+
+	for (i = 0; i < XGENE_DMA_MAX_CHANNEL; i++) {
+		chan = &pdma->chan[i];
+		devm_free_irq(chan->dev, chan->rx_irq, chan);
+	}
+}
+
+static void xgene_dma_set_caps(struct xgene_dma_chan *chan,
+			       struct dma_device *dma_dev)
+{
+	/* Initialize DMA device capability mask */
+	dma_cap_zero(dma_dev->cap_mask);
+
+	/* Set DMA device capability */
+	dma_cap_set(DMA_MEMCPY, dma_dev->cap_mask);
+	dma_cap_set(DMA_SG, dma_dev->cap_mask);
+
+	/* Basically here, the X-Gene SoC DMA engine channel 0 supports XOR
+	 * and channel 1 supports XOR, PQ both. First thing here is we have
+	 * mechanism in hw to enable/disable PQ/XOR supports on channel 1,
+	 * we can make sure this by reading SoC Efuse register.
+	 * Second thing, we have hw errata that if we run channel 0 and
+	 * channel 1 simultaneously with executing XOR and PQ request,
+	 * suddenly DMA engine hangs, So here we enable XOR on channel 0 only
+	 * if XOR and PQ supports on channel 1 is disabled.
+	 */
+	if ((chan->id == XGENE_DMA_PQ_CHANNEL) &&
+	    is_pq_enabled(chan->pdma)) {
+		dma_cap_set(DMA_PQ, dma_dev->cap_mask);
+		dma_cap_set(DMA_XOR, dma_dev->cap_mask);
+	} else if ((chan->id == XGENE_DMA_XOR_CHANNEL) &&
+		   !is_pq_enabled(chan->pdma)) {
+		dma_cap_set(DMA_XOR, dma_dev->cap_mask);
+	}
+
+	/* Set base and prep routines */
+	dma_dev->dev = chan->dev;
+	dma_dev->device_alloc_chan_resources = xgene_dma_alloc_chan_resources;
+	dma_dev->device_free_chan_resources = xgene_dma_free_chan_resources;
+	dma_dev->device_issue_pending = xgene_dma_issue_pending;
+	dma_dev->device_tx_status = xgene_dma_tx_status;
+	dma_dev->device_prep_dma_memcpy = xgene_dma_prep_memcpy;
+	dma_dev->device_prep_dma_sg = xgene_dma_prep_sg;
+
+	if (dma_has_cap(DMA_XOR, dma_dev->cap_mask)) {
+		dma_dev->device_prep_dma_xor = xgene_dma_prep_xor;
+		dma_dev->max_xor = XGENE_DMA_MAX_XOR_SRC;
+		dma_dev->xor_align = XGENE_DMA_XOR_ALIGNMENT;
+	}
+
+	if (dma_has_cap(DMA_PQ, dma_dev->cap_mask)) {
+		dma_dev->device_prep_dma_pq = xgene_dma_prep_pq;
+		dma_dev->max_pq = XGENE_DMA_MAX_XOR_SRC;
+		dma_dev->pq_align = XGENE_DMA_XOR_ALIGNMENT;
+	}
+}
+
+static int xgene_dma_async_register(struct xgene_dma *pdma, int id)
+{
+	struct xgene_dma_chan *chan = &pdma->chan[id];
+	struct dma_device *dma_dev = &pdma->dma_dev[id];
+	int ret;
+
+	chan->dma_chan.device = dma_dev;
+
+	spin_lock_init(&chan->lock);
+	INIT_LIST_HEAD(&chan->ld_pending);
+	INIT_LIST_HEAD(&chan->ld_running);
+	INIT_LIST_HEAD(&chan->ld_completed);
+	tasklet_init(&chan->tasklet, xgene_dma_tasklet_cb,
+		     (unsigned long)chan);
+
+	chan->pending = 0;
+	chan->desc_pool = NULL;
+	dma_cookie_init(&chan->dma_chan);
+
+	/* Setup dma device capabilities and prep routines */
+	xgene_dma_set_caps(chan, dma_dev);
+
+	/* Initialize DMA device list head */
+	INIT_LIST_HEAD(&dma_dev->channels);
+	list_add_tail(&chan->dma_chan.device_node, &dma_dev->channels);
+
+	/* Register with Linux async DMA framework*/
+	ret = dma_async_device_register(dma_dev);
+	if (ret) {
+		chan_err(chan, "Failed to register async device %d", ret);
+		tasklet_kill(&chan->tasklet);
+
+		return ret;
+	}
+
+	/* DMA capability info */
+	dev_info(pdma->dev,
+		 "%s: CAPABILITY ( %s%s%s%s)\n", dma_chan_name(&chan->dma_chan),
+		 dma_has_cap(DMA_MEMCPY, dma_dev->cap_mask) ? "MEMCPY " : "",
+		 dma_has_cap(DMA_SG, dma_dev->cap_mask) ? "SGCPY " : "",
+		 dma_has_cap(DMA_XOR, dma_dev->cap_mask) ? "XOR " : "",
+		 dma_has_cap(DMA_PQ, dma_dev->cap_mask) ? "PQ " : "");
+
+	return 0;
+}
+
+static int xgene_dma_init_async(struct xgene_dma *pdma)
+{
+	int ret, i, j;
+
+	for (i = 0; i < XGENE_DMA_MAX_CHANNEL ; i++) {
+		ret = xgene_dma_async_register(pdma, i);
+		if (ret) {
+			for (j = 0; j < i; j++) {
+				dma_async_device_unregister(&pdma->dma_dev[j]);
+				tasklet_kill(&pdma->chan[j].tasklet);
+			}
+
+			return ret;
+		}
+	}
+
+	return ret;
+}
+
+static void xgene_dma_async_unregister(struct xgene_dma *pdma)
+{
+	int i;
+
+	for (i = 0; i < XGENE_DMA_MAX_CHANNEL; i++)
+		dma_async_device_unregister(&pdma->dma_dev[i]);
+}
+
+static void xgene_dma_init_channels(struct xgene_dma *pdma)
+{
+	struct xgene_dma_chan *chan;
+	int i;
+
+	pdma->ring_num = XGENE_DMA_RING_NUM;
+
+	for (i = 0; i < XGENE_DMA_MAX_CHANNEL; i++) {
+		chan = &pdma->chan[i];
+		chan->dev = pdma->dev;
+		chan->pdma = pdma;
+		chan->id = i;
+		snprintf(chan->name, sizeof(chan->name), "dmachan%d", chan->id);
+	}
+}
+
+static int xgene_dma_get_resources(struct platform_device *pdev,
+				   struct xgene_dma *pdma)
+{
+	struct resource *res;
+	int irq, i;
+
+	/* Get DMA csr region */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res) {
+		dev_err(&pdev->dev, "Failed to get csr region\n");
+		return -ENXIO;
+	}
+
+	pdma->csr_dma = devm_ioremap(&pdev->dev, res->start,
+				     resource_size(res));
+	if (!pdma->csr_dma) {
+		dev_err(&pdev->dev, "Failed to ioremap csr region");
+		return -ENOMEM;
+	}
+
+	/* Get DMA ring csr region */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	if (!res) {
+		dev_err(&pdev->dev, "Failed to get ring csr region\n");
+		return -ENXIO;
+	}
+
+	pdma->csr_ring =  devm_ioremap(&pdev->dev, res->start,
+				       resource_size(res));
+	if (!pdma->csr_ring) {
+		dev_err(&pdev->dev, "Failed to ioremap ring csr region");
+		return -ENOMEM;
+	}
+
+	/* Get DMA ring cmd csr region */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 2);
+	if (!res) {
+		dev_err(&pdev->dev, "Failed to get ring cmd csr region\n");
+		return -ENXIO;
+	}
+
+	pdma->csr_ring_cmd = devm_ioremap(&pdev->dev, res->start,
+					  resource_size(res));
+	if (!pdma->csr_ring_cmd) {
+		dev_err(&pdev->dev, "Failed to ioremap ring cmd csr region");
+		return -ENOMEM;
+	}
+
+	/* Get efuse csr region */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 3);
+	if (!res) {
+		dev_err(&pdev->dev, "Failed to get efuse csr region\n");
+		return -ENXIO;
+	}
+
+	pdma->csr_efuse = devm_ioremap(&pdev->dev, res->start,
+				       resource_size(res));
+	if (!pdma->csr_efuse) {
+		dev_err(&pdev->dev, "Failed to ioremap efuse csr region");
+		return -ENOMEM;
+	}
+
+	/* Get DMA error interrupt */
+	irq = platform_get_irq(pdev, 0);
+	if (irq <= 0) {
+		dev_err(&pdev->dev, "Failed to get Error IRQ\n");
+		return -ENXIO;
+	}
+
+	pdma->err_irq = irq;
+
+	/* Get DMA Rx ring descriptor interrupts for all DMA channels */
+	for (i = 1; i <= XGENE_DMA_MAX_CHANNEL; i++) {
+		irq = platform_get_irq(pdev, i);
+		if (irq <= 0) {
+			dev_err(&pdev->dev, "Failed to get Rx IRQ\n");
+			return -ENXIO;
+		}
+
+		pdma->chan[i - 1].rx_irq = irq;
+	}
+
+	return 0;
+}
+
+static int xgene_dma_probe(struct platform_device *pdev)
+{
+	struct xgene_dma *pdma;
+	int ret, i;
+
+	pdma = devm_kzalloc(&pdev->dev, sizeof(*pdma), GFP_KERNEL);
+	if (!pdma)
+		return -ENOMEM;
+
+	pdma->dev = &pdev->dev;
+	platform_set_drvdata(pdev, pdma);
+
+	ret = xgene_dma_get_resources(pdev, pdma);
+	if (ret)
+		return ret;
+
+	pdma->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(pdma->clk)) {
+		dev_err(&pdev->dev, "Failed to get clk\n");
+		return PTR_ERR(pdma->clk);
+	}
+
+	/* Enable clk before accessing registers */
+	ret = clk_prepare_enable(pdma->clk);
+	if (ret) {
+		dev_err(&pdev->dev, "Failed to enable clk %d\n", ret);
+		return ret;
+	}
+
+	/* Remove DMA RAM out of shutdown */
+	ret = xgene_dma_init_mem(pdma);
+	if (ret)
+		goto err_clk_enable;
+
+	ret = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(42));
+	if (ret) {
+		dev_err(&pdev->dev, "No usable DMA configuration\n");
+		goto err_dma_mask;
+	}
+
+	/* Initialize DMA channels software state */
+	xgene_dma_init_channels(pdma);
+
+	/* Configue DMA rings */
+	ret = xgene_dma_init_rings(pdma);
+	if (ret)
+		goto err_clk_enable;
+
+	ret = xgene_dma_request_irqs(pdma);
+	if (ret)
+		goto err_request_irq;
+
+	/* Configure and enable DMA engine */
+	xgene_dma_init_hw(pdma);
+
+	/* Register DMA device with linux async framework */
+	ret = xgene_dma_init_async(pdma);
+	if (ret)
+		goto err_async_init;
+
+	return 0;
+
+err_async_init:
+	xgene_dma_free_irqs(pdma);
+
+err_request_irq:
+	for (i = 0; i < XGENE_DMA_MAX_CHANNEL; i++)
+		xgene_dma_delete_chan_rings(&pdma->chan[i]);
+
+err_dma_mask:
+err_clk_enable:
+	clk_disable_unprepare(pdma->clk);
+
+	return ret;
+}
+
+static int xgene_dma_remove(struct platform_device *pdev)
+{
+	struct xgene_dma *pdma = platform_get_drvdata(pdev);
+	struct xgene_dma_chan *chan;
+	int i;
+
+	xgene_dma_async_unregister(pdma);
+
+	/* Mask interrupts and disable DMA engine */
+	xgene_dma_mask_interrupts(pdma);
+	xgene_dma_disable(pdma);
+	xgene_dma_free_irqs(pdma);
+
+	for (i = 0; i < XGENE_DMA_MAX_CHANNEL; i++) {
+		chan = &pdma->chan[i];
+		tasklet_kill(&chan->tasklet);
+		xgene_dma_delete_chan_rings(chan);
+	}
+
+	clk_disable_unprepare(pdma->clk);
+
+	return 0;
+}
+
+static const struct of_device_id xgene_dma_of_match_ptr[] = {
+	{.compatible = "apm,xgene-storm-dma",},
+	{},
+};
+MODULE_DEVICE_TABLE(of, xgene_dma_of_match_ptr);
+
+static struct platform_driver xgene_dma_driver = {
+	.probe = xgene_dma_probe,
+	.remove = xgene_dma_remove,
+	.driver = {
+		.name = "X-Gene-DMA",
+		.of_match_table = xgene_dma_of_match_ptr,
+	},
+};
+
+module_platform_driver(xgene_dma_driver);
+
+MODULE_DESCRIPTION("APM X-Gene SoC DMA driver");
+MODULE_AUTHOR("Rameshwar Prasad Sahu <rsahu@apm.com>");
+MODULE_AUTHOR("Loc Ho <lho@apm.com>");
+MODULE_LICENSE("GPL");
+MODULE_VERSION("1.0");
diff --git a/drivers/dma/xilinx/xilinx_vdma.c b/drivers/dma/xilinx/xilinx_vdma.c
index bdd2a5dd7220..d8434d465885 100644
--- a/drivers/dma/xilinx/xilinx_vdma.c
+++ b/drivers/dma/xilinx/xilinx_vdma.c
@@ -22,9 +22,9 @@
  * (at your option) any later version.
  */
 
-#include <linux/amba/xilinx_dma.h>
 #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>