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

* 'next' of git://git.infradead.org/users/vkoul/slave-dma: (53 commits)
  ARM: mach-shmobile: specify CHCLR registers on SH7372
  dma: shdma: fix runtime PM: clear channel buffers on reset
  dma/imx-sdma: save irq flags when use spin_lock in sdma_tx_submit
  dmaengine/ste_dma40: clear LNK on channel startup
  dmaengine: intel_mid_dma: remove legacy pm interface
  ASoC: mxs: correct 'direction' of device_prep_dma_cyclic
  dmaengine: intel_mid_dma: error path fix
  dmaengine: intel_mid_dma: locking and freeing fixes
  mtd: gpmi-nand: move to dma_transfer_direction
  mtd: fix compile error for gpmi-nand
  mmc: mxs-mmc: fix the dma_transfer_direction migration
  dmaengine: add DMA_TRANS_NONE to dma_transfer_direction
  dma: mxs-dma: Don't use CLKGATE bits in CTRL0 to disable DMA channels
  dma: mxs-dma: make mxs_dma_prep_slave_sg() multi user safe
  dma: mxs-dma: Always leave mxs_dma_init() with the clock disabled.
  dma: mxs-dma: fix a typo in comment
  DMA: PL330: Remove pm_runtime_xxx calls from pl330 probe/remove
  video i.MX IPU: Fix display connections
  i.MX IPU DMA: Fix wrong burstsize settings
  dmaengine/ste_dma40: allow fixed physical channel
  ...

Fix up conflicts in drivers/dma/{Kconfig,mxs-dma.c,pl330.c}

The conflicts looked pretty trivial, but I'll ask people to verify them.

73 files changed, 1872 insertions, 530 deletions

diff --git a/Documentation/devicetree/bindings/dma/atmel-dma.txt b/Documentation/devicetree/bindings/dma/atmel-dma.txt
new file mode 100644
index 00000000000..3c046ee6e8b
--- /dev/null
+++ b/Documentation/devicetree/bindings/dma/atmel-dma.txt
@@ -0,0 +1,14 @@
+* Atmel Direct Memory Access Controller (DMA)
+
+Required properties:
+- compatible: Should be "atmel,<chip>-dma"
+- reg: Should contain DMA registers location and length
+- interrupts: Should contain DMA interrupt
+
+Examples:
+
+dma@ffffec00 {
+	compatible = "atmel,at91sam9g45-dma";
+	reg = <0xffffec00 0x200>;
+	interrupts = <21>;
+};
diff --git a/Documentation/dmaengine.txt b/Documentation/dmaengine.txt
index 94b7e0f96b3..bbe6cb3d185 100644
--- a/Documentation/dmaengine.txt
+++ b/Documentation/dmaengine.txt
@@ -75,6 +75,10 @@ The slave DMA usage consists of following steps:
    slave_sg	- DMA a list of scatter gather buffers from/to a peripheral
    dma_cyclic	- Perform a cyclic DMA operation from/to a peripheral till the
 		  operation is explicitly stopped.
+   interleaved_dma - This is common to Slave as well as M2M clients. For slave
+		 address of devices' fifo could be already known to the driver.
+		 Various types of operations could be expressed by setting
+		 appropriate values to the 'dma_interleaved_template' members.
 
    A non-NULL return of this transfer API represents a "descriptor" for
    the given transaction.
@@ -89,6 +93,10 @@ The slave DMA usage consists of following steps:
 		struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
 		size_t period_len, enum dma_data_direction direction);
 
+	struct dma_async_tx_descriptor *(*device_prep_interleaved_dma)(
+		struct dma_chan *chan, struct dma_interleaved_template *xt,
+		unsigned long flags);
+
    The peripheral driver is expected to have mapped the scatterlist for
    the DMA operation prior to calling device_prep_slave_sg, and must
    keep the scatterlist mapped until the DMA operation has completed.
diff --git a/MAINTAINERS b/MAINTAINERS
index ece8935025e..341dee3b02c 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -745,6 +745,7 @@ M:	Barry Song <baohua.song@csr.com>
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
 S:	Maintained
 F:	arch/arm/mach-prima2/
+F:	drivers/dma/sirf-dma*
 
 ARM/EBSA110 MACHINE SUPPORT
 M:	Russell King <linux@arm.linux.org.uk>
diff --git a/arch/arm/mach-ep93xx/include/mach/dma.h b/arch/arm/mach-ep93xx/include/mach/dma.h
index 46d4d876e6f..e82c642fa53 100644
--- a/arch/arm/mach-ep93xx/include/mach/dma.h
+++ b/arch/arm/mach-ep93xx/include/mach/dma.h
@@ -37,7 +37,7 @@
  */
 struct ep93xx_dma_data {
 	int				port;
-	enum dma_data_direction		direction;
+	enum dma_transfer_direction	direction;
 	const char			*name;
 };
 
@@ -80,14 +80,14 @@ static inline bool ep93xx_dma_chan_is_m2p(struct dma_chan *chan)
  * channel supports given DMA direction. Only M2P channels have such
  * limitation, for M2M channels the direction is configurable.
  */
-static inline enum dma_data_direction
+static inline enum dma_transfer_direction
 ep93xx_dma_chan_direction(struct dma_chan *chan)
 {
 	if (!ep93xx_dma_chan_is_m2p(chan))
 		return DMA_NONE;
 
 	/* even channels are for TX, odd for RX */
-	return (chan->chan_id % 2 == 0) ? DMA_TO_DEVICE : DMA_FROM_DEVICE;
+	return (chan->chan_id % 2 == 0) ? DMA_MEM_TO_DEV : DMA_DEV_TO_MEM;
 }
 
 #endif /* __ASM_ARCH_DMA_H */
diff --git a/arch/arm/mach-shmobile/setup-sh7372.c b/arch/arm/mach-shmobile/setup-sh7372.c
index 1ea89be63e2..6fcf304d3cd 100644
--- a/arch/arm/mach-shmobile/setup-sh7372.c
+++ b/arch/arm/mach-shmobile/setup-sh7372.c
@@ -445,31 +445,39 @@ static const struct sh_dmae_slave_config sh7372_dmae_slaves[] = {
 	},
 };
 
+#define SH7372_CHCLR 0x220
+
 static const struct sh_dmae_channel sh7372_dmae_channels[] = {
 	{
 		.offset = 0,
 		.dmars = 0,
 		.dmars_bit = 0,
+		.chclr_offset = SH7372_CHCLR + 0,
 	}, {
 		.offset = 0x10,
 		.dmars = 0,
 		.dmars_bit = 8,
+		.chclr_offset = SH7372_CHCLR + 0x10,
 	}, {
 		.offset = 0x20,
 		.dmars = 4,
 		.dmars_bit = 0,
+		.chclr_offset = SH7372_CHCLR + 0x20,
 	}, {
 		.offset = 0x30,
 		.dmars = 4,
 		.dmars_bit = 8,
+		.chclr_offset = SH7372_CHCLR + 0x30,
 	}, {
 		.offset = 0x50,
 		.dmars = 8,
 		.dmars_bit = 0,
+		.chclr_offset = SH7372_CHCLR + 0x50,
 	}, {
 		.offset = 0x60,
 		.dmars = 8,
 		.dmars_bit = 8,
+		.chclr_offset = SH7372_CHCLR + 0x60,
 	}
 };
 
@@ -487,6 +495,7 @@ static struct sh_dmae_pdata dma_platform_data = {
 	.ts_shift	= ts_shift,
 	.ts_shift_num	= ARRAY_SIZE(ts_shift),
 	.dmaor_init	= DMAOR_DME,
+	.chclr_present	= 1,
 };
 
 /* Resource order important! */
@@ -494,7 +503,7 @@ static struct resource sh7372_dmae0_resources[] = {
 	{
 		/* Channel registers and DMAOR */
 		.start	= 0xfe008020,
-		.end	= 0xfe00808f,
+		.end	= 0xfe00828f,
 		.flags	= IORESOURCE_MEM,
 	},
 	{
@@ -522,7 +531,7 @@ static struct resource sh7372_dmae1_resources[] = {
 	{
 		/* Channel registers and DMAOR */
 		.start	= 0xfe018020,
-		.end	= 0xfe01808f,
+		.end	= 0xfe01828f,
 		.flags	= IORESOURCE_MEM,
 	},
 	{
@@ -550,7 +559,7 @@ static struct resource sh7372_dmae2_resources[] = {
 	{
 		/* Channel registers and DMAOR */
 		.start	= 0xfe028020,
-		.end	= 0xfe02808f,
+		.end	= 0xfe02828f,
 		.flags	= IORESOURCE_MEM,
 	},
 	{
diff --git a/arch/arm/plat-mxc/include/mach/mx3fb.h b/arch/arm/plat-mxc/include/mach/mx3fb.h
index ac24c5c4bc8..fdbe6000154 100644
--- a/arch/arm/plat-mxc/include/mach/mx3fb.h
+++ b/arch/arm/plat-mxc/include/mach/mx3fb.h
@@ -22,6 +22,20 @@
 #define FB_SYNC_SWAP_RGB	0x04000000
 #define FB_SYNC_CLK_SEL_EN	0x02000000
 
+/*
+ * Specify the way your display is connected. The IPU can arbitrarily
+ * map the internal colors to the external data lines. We only support
+ * the following mappings at the moment.
+ */
+enum disp_data_mapping {
+	/* blue -> d[0..5], green -> d[6..11], red -> d[12..17] */
+	IPU_DISP_DATA_MAPPING_RGB666,
+	/* blue -> d[0..4], green -> d[5..10], red -> d[11..15] */
+	IPU_DISP_DATA_MAPPING_RGB565,
+	/* blue -> d[0..7], green -> d[8..15], red -> d[16..23] */
+	IPU_DISP_DATA_MAPPING_RGB888,
+};
+
 /**
  * struct mx3fb_platform_data - mx3fb platform data
  *
@@ -33,6 +47,7 @@ struct mx3fb_platform_data {
 	const char			*name;
 	const struct fb_videomode	*mode;
 	int				num_modes;
+	enum disp_data_mapping		disp_data_fmt;
 };
 
 #endif
diff --git a/arch/arm/plat-nomadik/include/plat/ste_dma40.h b/arch/arm/plat-nomadik/include/plat/ste_dma40.h
index 685c78716d9..fd0ee84c45d 100644
--- a/arch/arm/plat-nomadik/include/plat/ste_dma40.h
+++ b/arch/arm/plat-nomadik/include/plat/ste_dma40.h
@@ -113,7 +113,8 @@ struct stedma40_half_channel_info {
  * @dst_dev_type: Dst device type
  * @src_info: Parameters for dst half channel
  * @dst_info: Parameters for dst half channel
- *
+ * @use_fixed_channel: if true, use physical channel specified by phy_channel
+ * @phy_channel: physical channel to use, only if use_fixed_channel is true
  *
  * This structure has to be filled by the client drivers.
  * It is recommended to do all dma configurations for clients in the machine.
@@ -129,6 +130,9 @@ struct stedma40_chan_cfg {
 	int					 dst_dev_type;
 	struct stedma40_half_channel_info	 src_info;
 	struct stedma40_half_channel_info	 dst_info;
+
+	bool					 use_fixed_channel;
+	int					 phy_channel;
 };
 
 /**
@@ -153,6 +157,7 @@ struct stedma40_platform_data {
 	struct stedma40_chan_cfg	*memcpy_conf_phy;
 	struct stedma40_chan_cfg	*memcpy_conf_log;
 	int				 disabled_channels[STEDMA40_MAX_PHYS];
+	bool				 use_esram_lcla;
 };
 
 #ifdef CONFIG_STE_DMA40
@@ -187,7 +192,7 @@ static inline struct
 dma_async_tx_descriptor *stedma40_slave_mem(struct dma_chan *chan,
 					    dma_addr_t addr,
 					    unsigned int size,
-					    enum dma_data_direction direction,
+					    enum dma_transfer_direction direction,
 					    unsigned long flags)
 {
 	struct scatterlist sg;
@@ -209,7 +214,7 @@ static inline struct
 dma_async_tx_descriptor *stedma40_slave_mem(struct dma_chan *chan,
 					    dma_addr_t addr,
 					    unsigned int size,
-					    enum dma_data_direction direction,
+					    enum dma_transfer_direction direction,
 					    unsigned long flags)
 {
 	return NULL;
diff --git a/arch/arm/plat-samsung/dma-ops.c b/arch/arm/plat-samsung/dma-ops.c
index 2cded872f22..0747c77a2fd 100644
--- a/arch/arm/plat-samsung/dma-ops.c
+++ b/arch/arm/plat-samsung/dma-ops.c
@@ -37,14 +37,14 @@ static unsigned samsung_dmadev_request(enum dma_ch dma_ch,
 				(void *)dma_ch;
 	chan = dma_request_channel(mask, pl330_filter, filter_param);
 
-	if (info->direction == DMA_FROM_DEVICE) {
+	if (info->direction == DMA_DEV_TO_MEM) {
 		memset(&slave_config, 0, sizeof(struct dma_slave_config));
 		slave_config.direction = info->direction;
 		slave_config.src_addr = info->fifo;
 		slave_config.src_addr_width = info->width;
 		slave_config.src_maxburst = 1;
 		dmaengine_slave_config(chan, &slave_config);
-	} else if (info->direction == DMA_TO_DEVICE) {
+	} else if (info->direction == DMA_MEM_TO_DEV) {
 		memset(&slave_config, 0, sizeof(struct dma_slave_config));
 		slave_config.direction = info->direction;
 		slave_config.dst_addr = info->fifo;
diff --git a/arch/arm/plat-samsung/include/plat/dma-ops.h b/arch/arm/plat-samsung/include/plat/dma-ops.h
index 22eafc310bd..70b6325edb9 100644
--- a/arch/arm/plat-samsung/include/plat/dma-ops.h
+++ b/arch/arm/plat-samsung/include/plat/dma-ops.h
@@ -17,7 +17,7 @@
 
 struct samsung_dma_prep_info {
 	enum dma_transaction_type cap;
-	enum dma_data_direction direction;
+	enum dma_transfer_direction direction;
 	dma_addr_t buf;
 	unsigned long period;
 	unsigned long len;
@@ -27,7 +27,7 @@ struct samsung_dma_prep_info {
 
 struct samsung_dma_info {
 	enum dma_transaction_type cap;
-	enum dma_data_direction direction;
+	enum dma_transfer_direction direction;
 	enum dma_slave_buswidth width;
 	dma_addr_t fifo;
 	struct s3c2410_dma_client *client;
diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index 5a99bb3f255..f1a274994bb 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -124,7 +124,7 @@ config MV_XOR
 
 config MX3_IPU
 	bool "MX3x Image Processing Unit support"
-	depends on SOC_IMX31 || SOC_IMX35
+	depends on ARCH_MXC
 	select DMA_ENGINE
 	default y
 	help
@@ -187,6 +187,13 @@ config TIMB_DMA
 	help
 	  Enable support for the Timberdale FPGA DMA engine.
 
+config SIRF_DMA
+	tristate "CSR SiRFprimaII DMA support"
+	depends on ARCH_PRIMA2
+	select DMA_ENGINE
+	help
+	  Enable support for the CSR SiRFprimaII DMA engine.
+
 config ARCH_HAS_ASYNC_TX_FIND_CHANNEL
 	bool
 
@@ -201,26 +208,26 @@ config PL330_DMA
 	  platform_data for a dma-pl330 device.
 
 config PCH_DMA
-	tristate "Intel EG20T PCH / OKI Semi IOH(ML7213/ML7223) DMA support"
+	tristate "Intel EG20T PCH / LAPIS Semicon IOH(ML7213/ML7223/ML7831) DMA"
 	depends on PCI && X86
 	select DMA_ENGINE
 	help
 	  Enable support for Intel EG20T PCH DMA engine.
 
-	  This driver also can be used for OKI SEMICONDUCTOR IOH(Input/
-	  Output Hub), ML7213 and ML7223.
-	  ML7213 IOH is for IVI(In-Vehicle Infotainment) use and ML7223 IOH is
-	  for MP(Media Phone) use.
-	  ML7213/ML7223 is companion chip for Intel Atom E6xx series.
-	  ML7213/ML7223 is completely compatible for Intel EG20T PCH.
+	  This driver also can be used for LAPIS Semiconductor IOH(Input/
+	  Output Hub), ML7213, ML7223 and ML7831.
+	  ML7213 IOH is for IVI(In-Vehicle Infotainment) use, ML7223 IOH is
+	  for MP(Media Phone) use and ML7831 IOH is for general purpose use.
+	  ML7213/ML7223/ML7831 is companion chip for Intel Atom E6xx series.
+	  ML7213/ML7223/ML7831 is completely compatible for Intel EG20T PCH.
 
 config IMX_SDMA
 	tristate "i.MX SDMA support"
-	depends on ARCH_MX25 || SOC_IMX31 || SOC_IMX35 || ARCH_MX5
+	depends on ARCH_MXC
 	select DMA_ENGINE
 	help
 	  Support the i.MX SDMA engine. This engine is integrated into
-	  Freescale i.MX25/31/35/51 chips.
+	  Freescale i.MX25/31/35/51/53 chips.
 
 config IMX_DMA
 	tristate "i.MX DMA support"
diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile
index 30cf3b1f0c5..009a222e828 100644
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@@ -21,6 +21,7 @@ obj-$(CONFIG_IMX_SDMA) += imx-sdma.o
 obj-$(CONFIG_IMX_DMA) += imx-dma.o
 obj-$(CONFIG_MXS_DMA) += mxs-dma.o
 obj-$(CONFIG_TIMB_DMA) += timb_dma.o
+obj-$(CONFIG_SIRF_DMA) += sirf-dma.o
 obj-$(CONFIG_STE_DMA40) += ste_dma40.o ste_dma40_ll.o
 obj-$(CONFIG_PL330_DMA) += pl330.o
 obj-$(CONFIG_PCH_DMA) += pch_dma.o
diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c
index 0698695e8bf..8a281584458 100644
--- a/drivers/dma/amba-pl08x.c
+++ b/drivers/dma/amba-pl08x.c
@@ -854,8 +854,10 @@ static int prep_phy_channel(struct pl08x_dma_chan *plchan,
 	int ret;
 
 	/* Check if we already have a channel */
-	if (plchan->phychan)
-		return 0;
+	if (plchan->phychan) {
+		ch = plchan->phychan;
+		goto got_channel;
+	}
 
 	ch = pl08x_get_phy_channel(pl08x, plchan);
 	if (!ch) {
@@ -880,21 +882,22 @@ static int prep_phy_channel(struct pl08x_dma_chan *plchan,
 			return -EBUSY;
 		}
 		ch->signal = ret;
-
-		/* Assign the flow control signal to this channel */
-		if (txd->direction == DMA_TO_DEVICE)
-			txd->ccfg |= ch->signal << PL080_CONFIG_DST_SEL_SHIFT;
-		else if (txd->direction == DMA_FROM_DEVICE)
-			txd->ccfg |= ch->signal << PL080_CONFIG_SRC_SEL_SHIFT;
 	}
 
+	plchan->phychan = ch;
 	dev_dbg(&pl08x->adev->dev, "allocated physical channel %d and signal %d for xfer on %s\n",
 		 ch->id,
 		 ch->signal,
 		 plchan->name);
 
+got_channel:
+	/* Assign the flow control signal to this channel */
+	if (txd->direction == DMA_MEM_TO_DEV)
+		txd->ccfg |= ch->signal << PL080_CONFIG_DST_SEL_SHIFT;
+	else if (txd->direction == DMA_DEV_TO_MEM)
+		txd->ccfg |= ch->signal << PL080_CONFIG_SRC_SEL_SHIFT;
+
 	plchan->phychan_hold++;
-	plchan->phychan = ch;
 
 	return 0;
 }
@@ -1102,10 +1105,10 @@ static int dma_set_runtime_config(struct dma_chan *chan,
 
 	/* Transfer direction */
 	plchan->runtime_direction = config->direction;
-	if (config->direction == DMA_TO_DEVICE) {
+	if (config->direction == DMA_MEM_TO_DEV) {
 		addr_width = config->dst_addr_width;
 		maxburst = config->dst_maxburst;
-	} else if (config->direction == DMA_FROM_DEVICE) {
+	} else if (config->direction == DMA_DEV_TO_MEM) {
 		addr_width = config->src_addr_width;
 		maxburst = config->src_maxburst;
 	} else {
@@ -1136,7 +1139,7 @@ static int dma_set_runtime_config(struct dma_chan *chan,
 	cctl |= burst << PL080_CONTROL_SB_SIZE_SHIFT;
 	cctl |= burst << PL080_CONTROL_DB_SIZE_SHIFT;
 
-	if (plchan->runtime_direction == DMA_FROM_DEVICE) {
+	if (plchan->runtime_direction == DMA_DEV_TO_MEM) {
 		plchan->src_addr = config->src_addr;
 		plchan->src_cctl = pl08x_cctl(cctl) | PL080_CONTROL_DST_INCR |
 			pl08x_select_bus(plchan->cd->periph_buses,
@@ -1152,7 +1155,7 @@ static int dma_set_runtime_config(struct dma_chan *chan,
 		"configured channel %s (%s) for %s, data width %d, "
 		"maxburst %d words, LE, CCTL=0x%08x\n",
 		dma_chan_name(chan), plchan->name,
-		(config->direction == DMA_FROM_DEVICE) ? "RX" : "TX",
+		(config->direction == DMA_DEV_TO_MEM) ? "RX" : "TX",
 		addr_width,
 		maxburst,
 		cctl);
@@ -1322,7 +1325,7 @@ static struct dma_async_tx_descriptor *pl08x_prep_dma_memcpy(
 
 static struct dma_async_tx_descriptor *pl08x_prep_slave_sg(
 		struct dma_chan *chan, struct scatterlist *sgl,
-		unsigned int sg_len, enum dma_data_direction direction,
+		unsigned int sg_len, enum dma_transfer_direction direction,
 		unsigned long flags)
 {
 	struct pl08x_dma_chan *plchan = to_pl08x_chan(chan);
@@ -1354,10 +1357,10 @@ static struct dma_async_tx_descriptor *pl08x_prep_slave_sg(
 	 */
 	txd->direction = direction;
 
-	if (direction == DMA_TO_DEVICE) {
+	if (direction == DMA_MEM_TO_DEV) {
 		txd->cctl = plchan->dst_cctl;
 		slave_addr = plchan->dst_addr;
-	} else if (direction == DMA_FROM_DEVICE) {
+	} else if (direction == DMA_DEV_TO_MEM) {
 		txd->cctl = plchan->src_cctl;
 		slave_addr = plchan->src_addr;
 	} else {
@@ -1368,10 +1371,10 @@ static struct dma_async_tx_descriptor *pl08x_prep_slave_sg(
 	}
 
 	if (plchan->cd->device_fc)
-		tmp = (direction == DMA_TO_DEVICE) ? PL080_FLOW_MEM2PER_PER :
+		tmp = (direction == DMA_MEM_TO_DEV) ? PL080_FLOW_MEM2PER_PER :
 			PL080_FLOW_PER2MEM_PER;
 	else
-		tmp = (direction == DMA_TO_DEVICE) ? PL080_FLOW_MEM2PER :
+		tmp = (direction == DMA_MEM_TO_DEV) ? PL080_FLOW_MEM2PER :
 			PL080_FLOW_PER2MEM;
 
 	txd->ccfg |= tmp << PL080_CONFIG_FLOW_CONTROL_SHIFT;
@@ -1387,7 +1390,7 @@ static struct dma_async_tx_descriptor *pl08x_prep_slave_sg(
 		list_add_tail(&dsg->node, &txd->dsg_list);
 
 		dsg->len = sg_dma_len(sg);
-		if (direction == DMA_TO_DEVICE) {
+		if (direction == DMA_MEM_TO_DEV) {
 			dsg->src_addr = sg_phys(sg);
 			dsg->dst_addr = slave_addr;
 		} else {
diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
index fcfa0a8b5c5..97f87b29b9f 100644
--- a/drivers/dma/at_hdmac.c
+++ b/drivers/dma/at_hdmac.c
@@ -23,6 +23,8 @@
 #include <linux/module.h>
 #include <linux/platform_device.h>
 #include <linux/slab.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
 
 #include "at_hdmac_regs.h"
 
@@ -660,7 +662,7 @@ err_desc_get:
  */
 static struct dma_async_tx_descriptor *
 atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
-		unsigned int sg_len, enum dma_data_direction direction,
+		unsigned int sg_len, enum dma_transfer_direction direction,
 		unsigned long flags)
 {
 	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
@@ -678,7 +680,7 @@ atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 
 	dev_vdbg(chan2dev(chan), "prep_slave_sg (%d): %s f0x%lx\n",
 			sg_len,
-			direction == DMA_TO_DEVICE ? "TO DEVICE" : "FROM DEVICE",
+			direction == DMA_MEM_TO_DEV ? "TO DEVICE" : "FROM DEVICE",
 			flags);
 
 	if (unlikely(!atslave || !sg_len)) {
@@ -692,7 +694,7 @@ atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 	ctrlb = ATC_IEN;
 
 	switch (direction) {
-	case DMA_TO_DEVICE:
+	case DMA_MEM_TO_DEV:
 		ctrla |=  ATC_DST_WIDTH(reg_width);
 		ctrlb |=  ATC_DST_ADDR_MODE_FIXED
 			| ATC_SRC_ADDR_MODE_INCR
@@ -725,7 +727,7 @@ atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 			total_len += len;
 		}
 		break;
-	case DMA_FROM_DEVICE:
+	case DMA_DEV_TO_MEM:
 		ctrla |=  ATC_SRC_WIDTH(reg_width);
 		ctrlb |=  ATC_DST_ADDR_MODE_INCR
 			| ATC_SRC_ADDR_MODE_FIXED
@@ -787,7 +789,7 @@ err_desc_get:
  */
 static int
 atc_dma_cyclic_check_values(unsigned int reg_width, dma_addr_t buf_addr,
-		size_t period_len, enum dma_data_direction direction)
+		size_t period_len, enum dma_transfer_direction direction)
 {
 	if (period_len > (ATC_BTSIZE_MAX << reg_width))
 		goto err_out;
@@ -795,7 +797,7 @@ atc_dma_cyclic_check_values(unsigned int reg_width, dma_addr_t buf_addr,
 		goto err_out;
 	if (unlikely(buf_addr & ((1 << reg_width) - 1)))
 		goto err_out;
-	if (unlikely(!(direction & (DMA_TO_DEVICE | DMA_FROM_DEVICE))))
+	if (unlikely(!(direction & (DMA_DEV_TO_MEM | DMA_MEM_TO_DEV))))
 		goto err_out;
 
 	return 0;
@@ -810,7 +812,7 @@ err_out:
 static int
 atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc,
 		unsigned int period_index, dma_addr_t buf_addr,
-		size_t period_len, enum dma_data_direction direction)
+		size_t period_len, enum dma_transfer_direction direction)
 {
 	u32		ctrla;
 	unsigned int	reg_width = atslave->reg_width;
@@ -822,7 +824,7 @@ atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc,
 		| period_len >> reg_width;
 
 	switch (direction) {
-	case DMA_TO_DEVICE:
+	case DMA_MEM_TO_DEV:
 		desc->lli.saddr = buf_addr + (period_len * period_index);
 		desc->lli.daddr = atslave->tx_reg;
 		desc->lli.ctrla = ctrla;
@@ -833,7 +835,7 @@ atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc,
 				| ATC_DIF(AT_DMA_PER_IF);
 		break;
 
-	case DMA_FROM_DEVICE:
+	case DMA_DEV_TO_MEM:
 		desc->lli.saddr = atslave->rx_reg;
 		desc->lli.daddr = buf_addr + (period_len * period_index);
 		desc->lli.ctrla = ctrla;
@@ -861,7 +863,7 @@ atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc,
  */
 static struct dma_async_tx_descriptor *
 atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
-		size_t period_len, enum dma_data_direction direction)
+		size_t period_len, enum dma_transfer_direction direction)
 {
 	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
 	struct at_dma_slave	*atslave = chan->private;
@@ -872,7 +874,7 @@ atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
 	unsigned int		i;
 
 	dev_vdbg(chan2dev(chan), "prep_dma_cyclic: %s buf@0x%08x - %d (%d/%d)\n",
-			direction == DMA_TO_DEVICE ? "TO DEVICE" : "FROM DEVICE",
+			direction == DMA_MEM_TO_DEV ? "TO DEVICE" : "FROM DEVICE",
 			buf_addr,
 			periods, buf_len, period_len);
 
@@ -1175,6 +1177,56 @@ static void atc_free_chan_resources(struct dma_chan *chan)
 
 /*--  Module Management  -----------------------------------------------*/
 
+/* cap_mask is a multi-u32 bitfield, fill it with proper C code. */
+static struct at_dma_platform_data at91sam9rl_config = {
+	.nr_channels = 2,
+};
+static struct at_dma_platform_data at91sam9g45_config = {
+	.nr_channels = 8,
+};
+
+#if defined(CONFIG_OF)
+static const struct of_device_id atmel_dma_dt_ids[] = {
+	{
+		.compatible = "atmel,at91sam9rl-dma",
+		.data = &at91sam9rl_config,
+	}, {
+		.compatible = "atmel,at91sam9g45-dma",
+		.data = &at91sam9g45_config,
+	}, {
+		/* sentinel */
+	}
+};
+
+MODULE_DEVICE_TABLE(of, atmel_dma_dt_ids);
+#endif
+
+static const struct platform_device_id atdma_devtypes[] = {
+	{
+		.name = "at91sam9rl_dma",
+		.driver_data = (unsigned long) &at91sam9rl_config,
+	}, {
+		.name = "at91sam9g45_dma",
+		.driver_data = (unsigned long) &at91sam9g45_config,
+	}, {
+		/* sentinel */
+	}
+};
+
+static inline struct at_dma_platform_data * __init at_dma_get_driver_data(
+						struct platform_device *pdev)
+{
+	if (pdev->dev.of_node) {
+		const struct of_device_id *match;
+		match = of_match_node(atmel_dma_dt_ids, pdev->dev.of_node);
+		if (match == NULL)
+			return NULL;
+		return match->data;
+	}
+	return (struct at_dma_platform_data *)
+			platform_get_device_id(pdev)->driver_data;
+}
+
 /**
  * at_dma_off - disable DMA controller
  * @atdma: the Atmel HDAMC device
@@ -1193,18 +1245,23 @@ static void at_dma_off(struct at_dma *atdma)
 
 static int __init at_dma_probe(struct platform_device *pdev)
 {
-	struct at_dma_platform_data *pdata;
 	struct resource		*io;
 	struct at_dma		*atdma;
 	size_t			size;
 	int			irq;
 	int			err;
 	int			i;
+	struct at_dma_platform_data *plat_dat;
 
-	/* get DMA Controller parameters from platform */
-	pdata = pdev->dev.platform_data;
-	if (!pdata || pdata->nr_channels > AT_DMA_MAX_NR_CHANNELS)
-		return -EINVAL;
+	/* setup platform data for each SoC */
+	dma_cap_set(DMA_MEMCPY, at91sam9rl_config.cap_mask);
+	dma_cap_set(DMA_MEMCPY, at91sam9g45_config.cap_mask);
+	dma_cap_set(DMA_SLAVE, at91sam9g45_config.cap_mask);
+
+	/* get DMA parameters from controller type */
+	plat_dat = at_dma_get_driver_data(pdev);
+	if (!plat_dat)
+		return -ENODEV;
 
 	io = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	if (!io)
@@ -1215,14 +1272,14 @@ static int __init at_dma_probe(struct platform_device *pdev)
 		return irq;
 
 	size = sizeof(struct at_dma);
-	size += pdata->nr_channels * sizeof(struct at_dma_chan);
+	size += plat_dat->nr_channels * sizeof(struct at_dma_chan);
 	atdma = kzalloc(size, GFP_KERNEL);
 	if (!atdma)
 		return -ENOMEM;
 
-	/* discover transaction capabilites from the platform data */
-	atdma->dma_common.cap_mask = pdata->cap_mask;
-	atdma->all_chan_mask = (1 << pdata->nr_channels) - 1;
+	/* discover transaction capabilities */
+	atdma->dma_common.cap_mask = plat_dat->cap_mask;
+	atdma->all_chan_mask = (1 << plat_dat->nr_channels) - 1;
 
 	size = resource_size(io);
 	if (!request_mem_region(io->start, size, pdev->dev.driver->name)) {
@@ -1268,7 +1325,7 @@ static int __init at_dma_probe(struct platform_device *pdev)
 
 	/* initialize channels related values */
 	INIT_LIST_HEAD(&atdma->dma_common.channels);
-	for (i = 0; i < pdata->nr_channels; i++) {
+	for (i = 0; i < plat_dat->nr_channels; i++) {
 		struct at_dma_chan	*atchan = &atdma->chan[i];
 
 		atchan->chan_common.device = &atdma->dma_common;
@@ -1313,7 +1370,7 @@ static int __init at_dma_probe(struct platform_device *pdev)
 	dev_info(&pdev->dev, "Atmel AHB DMA Controller ( %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 " : "",
-	  pdata->nr_channels);
+	  plat_dat->nr_channels);
 
 	dma_async_device_register(&atdma->dma_common);
 
@@ -1495,9 +1552,11 @@ static const struct dev_pm_ops at_dma_dev_pm_ops = {
 static struct platform_driver at_dma_driver = {
 	.remove		= __exit_p(at_dma_remove),
 	.shutdown	= at_dma_shutdown,
+	.id_table	= atdma_devtypes,
 	.driver = {
 		.name	= "at_hdmac",
 		.pm	= &at_dma_dev_pm_ops,
+		.of_match_table	= of_match_ptr(atmel_dma_dt_ids),
 	},
 };
 
diff --git a/drivers/dma/at_hdmac_regs.h b/drivers/dma/at_hdmac_regs.h
index aa4c9aebab7..dcaedfc181c 100644
--- a/drivers/dma/at_hdmac_regs.h
+++ b/drivers/dma/at_hdmac_regs.h
@@ -251,6 +251,7 @@ static inline struct at_dma_chan *to_at_dma_chan(struct dma_chan *dchan)
 /**
  * struct at_dma - internal representation of an Atmel HDMA Controller
  * @chan_common: common dmaengine dma_device object members
+ * @atdma_devtype: identifier of DMA controller compatibility
  * @ch_regs: memory mapped register base
  * @clk: dma controller clock
  * @save_imr: interrupt mask register that is saved on suspend/resume cycle
diff --git a/drivers/dma/coh901318.c b/drivers/dma/coh901318.c
index 4234f416ef1..d65a718c0f9 100644
--- a/drivers/dma/coh901318.c
+++ b/drivers/dma/coh901318.c
@@ -39,7 +39,7 @@ struct coh901318_desc {
 	struct scatterlist *sg;
 	unsigned int sg_len;
 	struct coh901318_lli *lli;
-	enum dma_data_direction dir;
+	enum dma_transfer_direction dir;
 	unsigned long flags;
 	u32 head_config;
 	u32 head_ctrl;
@@ -1034,7 +1034,7 @@ coh901318_prep_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 
 static struct dma_async_tx_descriptor *
 coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
-			unsigned int sg_len, enum dma_data_direction direction,
+			unsigned int sg_len, enum dma_transfer_direction direction,
 			unsigned long flags)
 {
 	struct coh901318_chan *cohc = to_coh901318_chan(chan);
@@ -1077,7 +1077,7 @@ coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 	ctrl_last |= cohc->runtime_ctrl;
 	ctrl |= cohc->runtime_ctrl;
 
-	if (direction == DMA_TO_DEVICE) {
+	if (direction == DMA_MEM_TO_DEV) {
 		u32 tx_flags = COH901318_CX_CTRL_PRDD_SOURCE |
 			COH901318_CX_CTRL_SRC_ADDR_INC_ENABLE;
 
@@ -1085,7 +1085,7 @@ coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		ctrl_chained |= tx_flags;
 		ctrl_last |= tx_flags;
 		ctrl |= tx_flags;
-	} else if (direction == DMA_FROM_DEVICE) {
+	} else if (direction == DMA_DEV_TO_MEM) {
 		u32 rx_flags = COH901318_CX_CTRL_PRDD_DEST |
 			COH901318_CX_CTRL_DST_ADDR_INC_ENABLE;
 
@@ -1274,11 +1274,11 @@ static void coh901318_dma_set_runtimeconfig(struct dma_chan *chan,
 	int i = 0;
 
 	/* We only support mem to per or per to mem transfers */
-	if (config->direction == DMA_FROM_DEVICE) {
+	if (config->direction == DMA_DEV_TO_MEM) {
 		addr = config->src_addr;
 		addr_width = config->src_addr_width;
 		maxburst = config->src_maxburst;
-	} else if (config->direction == DMA_TO_DEVICE) {
+	} else if (config->direction == DMA_MEM_TO_DEV) {
 		addr = config->dst_addr;
 		addr_width = config->dst_addr_width;
 		maxburst = config->dst_maxburst;
diff --git a/drivers/dma/coh901318_lli.c b/drivers/dma/coh901318_lli.c
index 9f7e0e6a7ee..6c0e2d4c668 100644
--- a/drivers/dma/coh901318_lli.c
+++ b/drivers/dma/coh901318_lli.c
@@ -7,11 +7,10 @@
  * Author: Per Friden <per.friden@stericsson.com>
  */
 
-#include <linux/dma-mapping.h>
 #include <linux/spinlock.h>
-#include <linux/dmapool.h>
 #include <linux/memory.h>
 #include <linux/gfp.h>
+#include <linux/dmapool.h>
 #include <mach/coh901318.h>
 
 #include "coh901318_lli.h"
@@ -177,18 +176,18 @@ coh901318_lli_fill_single(struct coh901318_pool *pool,
 			  struct coh901318_lli *lli,
 			  dma_addr_t buf, unsigned int size,
 			  dma_addr_t dev_addr, u32 ctrl_chained, u32 ctrl_eom,
-			  enum dma_data_direction dir)
+			  enum dma_transfer_direction dir)
 {
 	int s = size;
 	dma_addr_t src;
 	dma_addr_t dst;
 
 
-	if (dir == DMA_TO_DEVICE) {
+	if (dir == DMA_MEM_TO_DEV) {
 		src = buf;
 		dst = dev_addr;
 
-	} else if (dir == DMA_FROM_DEVICE) {
+	} else if (dir == DMA_DEV_TO_MEM) {
 
 		src = dev_addr;
 		dst = buf;
@@ -215,9 +214,9 @@ coh901318_lli_fill_single(struct coh901318_pool *pool,
 
 		lli = coh901318_lli_next(lli);
 
-		if (dir == DMA_TO_DEVICE)
+		if (dir == DMA_MEM_TO_DEV)
 			src += block_size;
-		else if (dir == DMA_FROM_DEVICE)
+		else if (dir == DMA_DEV_TO_MEM)
 			dst += block_size;
 	}
 
@@ -234,7 +233,7 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool,
 		      struct scatterlist *sgl, unsigned int nents,
 		      dma_addr_t dev_addr, u32 ctrl_chained, u32 ctrl,
 		      u32 ctrl_last,
-		      enum dma_data_direction dir, u32 ctrl_irq_mask)
+		      enum dma_transfer_direction dir, u32 ctrl_irq_mask)
 {
 	int i;
 	struct scatterlist *sg;
@@ -249,9 +248,9 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool,
 
 	spin_lock(&pool->lock);
 
-	if (dir == DMA_TO_DEVICE)
+	if (dir == DMA_MEM_TO_DEV)
 		dst = dev_addr;
-	else if (dir == DMA_FROM_DEVICE)
+	else if (dir == DMA_DEV_TO_MEM)
 		src = dev_addr;
 	else
 		goto err;
@@ -269,7 +268,7 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool,
 			ctrl_sg = ctrl ? ctrl : ctrl_last;
 
 
-		if (dir == DMA_TO_DEVICE)
+		if (dir == DMA_MEM_TO_DEV)
 			/* increment source address */
 			src = sg_phys(sg);
 		else
@@ -293,7 +292,7 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool,
 			lli->src_addr = src;
 			lli->dst_addr = dst;
 
-			if (dir == DMA_FROM_DEVICE)
+			if (dir == DMA_DEV_TO_MEM)
 				dst += elem_size;
 			else
 				src += elem_size;
diff --git a/drivers/dma/coh901318_lli.h b/drivers/dma/coh901318_lli.h
index 7a5c80990e9..abff3714fdd 100644
--- a/drivers/dma/coh901318_lli.h
+++ b/drivers/dma/coh901318_lli.h
@@ -97,7 +97,7 @@ coh901318_lli_fill_single(struct coh901318_pool *pool,
 			  struct coh901318_lli *lli,
 			  dma_addr_t buf, unsigned int size,
 			  dma_addr_t dev_addr, u32 ctrl_chained, u32 ctrl_last,
-			  enum dma_data_direction dir);
+			  enum dma_transfer_direction dir);
 
 /**
  * coh901318_lli_fill_single() - Prepares the lli:s for dma scatter list transfer
@@ -119,6 +119,6 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool,
 		      struct scatterlist *sg, unsigned int nents,
 		      dma_addr_t dev_addr, u32 ctrl_chained,
 		      u32 ctrl, u32 ctrl_last,
-		      enum dma_data_direction dir, u32 ctrl_irq_mask);
+		      enum dma_transfer_direction dir, u32 ctrl_irq_mask);
 
 #endif /* COH901318_LLI_H */
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index b48967b499d..a6c6051ec85 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -693,12 +693,12 @@ int dma_async_device_register(struct dma_device *device)
 		!device->device_prep_dma_interrupt);
 	BUG_ON(dma_has_cap(DMA_SG, device->cap_mask) &&
 		!device->device_prep_dma_sg);
-	BUG_ON(dma_has_cap(DMA_SLAVE, device->cap_mask) &&
-		!device->device_prep_slave_sg);
 	BUG_ON(dma_has_cap(DMA_CYCLIC, device->cap_mask) &&
 		!device->device_prep_dma_cyclic);
 	BUG_ON(dma_has_cap(DMA_SLAVE, device->cap_mask) &&
 		!device->device_control);
+	BUG_ON(dma_has_cap(DMA_INTERLEAVE, device->cap_mask) &&
+		!device->device_prep_interleaved_dma);
 
 	BUG_ON(!device->device_alloc_chan_resources);
 	BUG_ON(!device->device_free_chan_resources);
diff --git a/drivers/dma/dw_dmac.c b/drivers/dma/dw_dmac.c
index 9bfd6d36071..9b592b02b5f 100644
--- a/drivers/dma/dw_dmac.c
+++ b/drivers/dma/dw_dmac.c
@@ -166,6 +166,38 @@ dwc_assign_cookie(struct dw_dma_chan *dwc, struct dw_desc *desc)
 	return cookie;
 }
 
+static void dwc_initialize(struct dw_dma_chan *dwc)
+{
+	struct dw_dma *dw = to_dw_dma(dwc->chan.device);
+	struct dw_dma_slave *dws = dwc->chan.private;
+	u32 cfghi = DWC_CFGH_FIFO_MODE;
+	u32 cfglo = DWC_CFGL_CH_PRIOR(dwc->priority);
+
+	if (dwc->initialized == true)
+		return;
+
+	if (dws) {
+		/*
+		 * We need controller-specific data to set up slave
+		 * transfers.
+		 */
+		BUG_ON(!dws->dma_dev || dws->dma_dev != dw->dma.dev);
+
+		cfghi = dws->cfg_hi;
+		cfglo |= dws->cfg_lo & ~DWC_CFGL_CH_PRIOR_MASK;
+	}
+
+	channel_writel(dwc, CFG_LO, cfglo);
+	channel_writel(dwc, CFG_HI, cfghi);
+
+	/* Enable interrupts */
+	channel_set_bit(dw, MASK.XFER, dwc->mask);
+	channel_set_bit(dw, MASK.BLOCK, dwc->mask);
+	channel_set_bit(dw, MASK.ERROR, dwc->mask);
+
+	dwc->initialized = true;
+}
+
 /*----------------------------------------------------------------------*/
 
 /* Called with dwc->lock held and bh disabled */
@@ -189,6 +221,8 @@ static void dwc_dostart(struct dw_dma_chan *dwc, struct dw_desc *first)
 		return;
 	}
 
+	dwc_initialize(dwc);
+
 	channel_writel(dwc, LLP, first->txd.phys);
 	channel_writel(dwc, CTL_LO,
 			DWC_CTLL_LLP_D_EN | DWC_CTLL_LLP_S_EN);
@@ -696,7 +730,7 @@ err_desc_get:
 
 static struct dma_async_tx_descriptor *
 dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
-		unsigned int sg_len, enum dma_data_direction direction,
+		unsigned int sg_len, enum dma_transfer_direction direction,
 		unsigned long flags)
 {
 	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
@@ -720,7 +754,7 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 	prev = first = NULL;
 
 	switch (direction) {
-	case DMA_TO_DEVICE:
+	case DMA_MEM_TO_DEV:
 		ctllo = (DWC_DEFAULT_CTLLO(chan->private)
 				| DWC_CTLL_DST_WIDTH(reg_width)
 				| DWC_CTLL_DST_FIX
@@ -777,7 +811,7 @@ slave_sg_todev_fill_desc:
 				goto slave_sg_todev_fill_desc;
 		}
 		break;
-	case DMA_FROM_DEVICE:
+	case DMA_DEV_TO_MEM:
 		ctllo = (DWC_DEFAULT_CTLLO(chan->private)
 				| DWC_CTLL_SRC_WIDTH(reg_width)
 				| DWC_CTLL_DST_INC
@@ -959,10 +993,7 @@ static int dwc_alloc_chan_resources(struct dma_chan *chan)
 	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
 	struct dw_dma		*dw = to_dw_dma(chan->device);
 	struct dw_desc		*desc;
-	struct dw_dma_slave	*dws;
 	int			i;
-	u32			cfghi;
-	u32			cfglo;
 	unsigned long		flags;
 
 	dev_vdbg(chan2dev(chan), "alloc_chan_resources\n");
@@ -975,26 +1006,6 @@ static int dwc_alloc_chan_resources(struct dma_chan *chan)
 
 	dwc->completed = chan->cookie = 1;
 
-	cfghi = DWC_CFGH_FIFO_MODE;
-	cfglo = 0;
-
-	dws = chan->private;
-	if (dws) {
-		/*
-		 * We need controller-specific data to set up slave
-		 * transfers.
-		 */
-		BUG_ON(!dws->dma_dev || dws->dma_dev != dw->dma.dev);
-
-		cfghi = dws->cfg_hi;
-		cfglo = dws->cfg_lo & ~DWC_CFGL_CH_PRIOR_MASK;
-	}
-
-	cfglo |= DWC_CFGL_CH_PRIOR(dwc->priority);
-
-	channel_writel(dwc, CFG_LO, cfglo);
-	channel_writel(dwc, CFG_HI, cfghi);
-
 	/*
 	 * NOTE: some controllers may have additional features that we
 	 * need to initialize here, like "scatter-gather" (which
@@ -1026,11 +1037,6 @@ static int dwc_alloc_chan_resources(struct dma_chan *chan)
 		i = ++dwc->descs_allocated;
 	}
 
-	/* Enable interrupts */
-	channel_set_bit(dw, MASK.XFER, dwc->mask);
-	channel_set_bit(dw, MASK.BLOCK, dwc->mask);
-	channel_set_bit(dw, MASK.ERROR, dwc->mask);
-
 	spin_unlock_irqrestore(&dwc->lock, flags);
 
 	dev_dbg(chan2dev(chan),
@@ -1058,6 +1064,7 @@ static void dwc_free_chan_resources(struct dma_chan *chan)
 	spin_lock_irqsave(&dwc->lock, flags);
 	list_splice_init(&dwc->free_list, &list);
 	dwc->descs_allocated = 0;
+	dwc->initialized = false;
 
 	/* Disable interrupts */
 	channel_clear_bit(dw, MASK.XFER, dwc->mask);
@@ -1165,7 +1172,7 @@ EXPORT_SYMBOL(dw_dma_cyclic_stop);
  */
 struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan,
 		dma_addr_t buf_addr, size_t buf_len, size_t period_len,
-		enum dma_data_direction direction)
+		enum dma_transfer_direction direction)
 {
 	struct dw_dma_chan		*dwc = to_dw_dma_chan(chan);
 	struct dw_cyclic_desc		*cdesc;
@@ -1206,7 +1213,7 @@ struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan,
 		goto out_err;
 	if (unlikely(buf_addr & ((1 << reg_width) - 1)))
 		goto out_err;
-	if (unlikely(!(direction & (DMA_TO_DEVICE | DMA_FROM_DEVICE))))
+	if (unlikely(!(direction & (DMA_MEM_TO_DEV | DMA_DEV_TO_MEM))))
 		goto out_err;
 
 	retval = ERR_PTR(-ENOMEM);
@@ -1228,7 +1235,7 @@ struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan,
 			goto out_err_desc_get;
 
 		switch (direction) {
-		case DMA_TO_DEVICE:
+		case DMA_MEM_TO_DEV:
 			desc->lli.dar = dws->tx_reg;
 			desc->lli.sar = buf_addr + (period_len * i);
 			desc->lli.ctllo = (DWC_DEFAULT_CTLLO(chan->private)
@@ -1239,7 +1246,7 @@ struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan,
 					| DWC_CTLL_FC(dws->fc)
 					| DWC_CTLL_INT_EN);
 			break;
-		case DMA_FROM_DEVICE:
+		case DMA_DEV_TO_MEM:
 			desc->lli.dar = buf_addr + (period_len * i);
 			desc->lli.sar = dws->rx_reg;
 			desc->lli.ctllo = (DWC_DEFAULT_CTLLO(chan->private)
@@ -1335,6 +1342,8 @@ EXPORT_SYMBOL(dw_dma_cyclic_free);
 
 static void dw_dma_off(struct dw_dma *dw)
 {
+	int i;
+
 	dma_writel(dw, CFG, 0);
 
 	channel_clear_bit(dw, MASK.XFER, dw->all_chan_mask);
@@ -1345,6 +1354,9 @@ static void dw_dma_off(struct dw_dma *dw)
 
 	while (dma_readl(dw, CFG) & DW_CFG_DMA_EN)
 		cpu_relax();
+
+	for (i = 0; i < dw->dma.chancnt; i++)
+		dw->chan[i].initialized = false;
 }
 
 static int __init dw_probe(struct platform_device *pdev)
@@ -1533,6 +1545,7 @@ static int dw_suspend_noirq(struct device *dev)
 
 	dw_dma_off(platform_get_drvdata(pdev));
 	clk_disable(dw->clk);
+
 	return 0;
 }
 
diff --git a/drivers/dma/dw_dmac_regs.h b/drivers/dma/dw_dmac_regs.h
index c3419518d70..5eef6946a36 100644
--- a/drivers/dma/dw_dmac_regs.h
+++ b/drivers/dma/dw_dmac_regs.h
@@ -140,6 +140,7 @@ struct dw_dma_chan {
 	u8			mask;
 	u8			priority;
 	bool			paused;
+	bool			initialized;
 
 	spinlock_t		lock;
 
diff --git a/drivers/dma/ep93xx_dma.c b/drivers/dma/ep93xx_dma.c
index b47e2b803fa..59e7a965772 100644
--- a/drivers/dma/ep93xx_dma.c
+++ b/drivers/dma/ep93xx_dma.c
@@ -246,6 +246,9 @@ static void ep93xx_dma_set_active(struct ep93xx_dma_chan *edmac,
 static struct ep93xx_dma_desc *
 ep93xx_dma_get_active(struct ep93xx_dma_chan *edmac)
 {
+	if (list_empty(&edmac->active))
+		return NULL;
+
 	return list_first_entry(&edmac->active, struct ep93xx_dma_desc, node);
 }
 
@@ -263,16 +266,22 @@ ep93xx_dma_get_active(struct ep93xx_dma_chan *edmac)
  */
 static bool ep93xx_dma_advance_active(struct ep93xx_dma_chan *edmac)
 {
+	struct ep93xx_dma_desc *desc;
+
 	list_rotate_left(&edmac->active);
 
 	if (test_bit(EP93XX_DMA_IS_CYCLIC, &edmac->flags))
 		return true;
 
+	desc = ep93xx_dma_get_active(edmac);
+	if (!desc)
+		return false;
+
 	/*
 	 * If txd.cookie is set it means that we are back in the first
 	 * descriptor in the chain and hence done with it.
 	 */
-	return !ep93xx_dma_get_active(edmac)->txd.cookie;
+	return !desc->txd.cookie;
 }
 
 /*
@@ -327,10 +336,16 @@ static void m2p_hw_shutdown(struct ep93xx_dma_chan *edmac)
 
 static void m2p_fill_desc(struct ep93xx_dma_chan *edmac)
 {
-	struct ep93xx_dma_desc *desc = ep93xx_dma_get_active(edmac);
+	struct ep93xx_dma_desc *desc;
 	u32 bus_addr;
 
-	if (ep93xx_dma_chan_direction(&edmac->chan) == DMA_TO_DEVICE)
+	desc = ep93xx_dma_get_active(edmac);
+	if (!desc) {
+		dev_warn(chan2dev(edmac), "M2P: empty descriptor list\n");
+		return;
+	}
+
+	if (ep93xx_dma_chan_direction(&edmac->chan) == DMA_MEM_TO_DEV)
 		bus_addr = desc->src_addr;
 	else
 		bus_addr = desc->dst_addr;
@@ -443,7 +458,7 @@ static int m2m_hw_setup(struct ep93xx_dma_chan *edmac)
 		control = (5 << M2M_CONTROL_PWSC_SHIFT);
 		control |= M2M_CONTROL_NO_HDSK;
 
-		if (data->direction == DMA_TO_DEVICE) {
+		if (data->direction == DMA_MEM_TO_DEV) {
 			control |= M2M_CONTROL_DAH;
 			control |= M2M_CONTROL_TM_TX;
 			control |= M2M_CONTROL_RSS_SSPTX;
@@ -459,11 +474,7 @@ static int m2m_hw_setup(struct ep93xx_dma_chan *edmac)
 		 * This IDE part is totally untested. Values below are taken
 		 * from the EP93xx Users's Guide and might not be correct.
 		 */
-		control |= M2M_CONTROL_NO_HDSK;
-		control |= M2M_CONTROL_RSS_IDE;
-		control |= M2M_CONTROL_PW_16;
-
-		if (data->direction == DMA_TO_DEVICE) {
+		if (data->direction == DMA_MEM_TO_DEV) {
 			/* Worst case from the UG */
 			control = (3 << M2M_CONTROL_PWSC_SHIFT);
 			control |= M2M_CONTROL_DAH;
@@ -473,6 +484,10 @@ static int m2m_hw_setup(struct ep93xx_dma_chan *edmac)
 			control |= M2M_CONTROL_SAH;
 			control |= M2M_CONTROL_TM_RX;
 		}
+
+		control |= M2M_CONTROL_NO_HDSK;
+		control |= M2M_CONTROL_RSS_IDE;
+		control |= M2M_CONTROL_PW_16;
 		break;
 
 	default:
@@ -491,7 +506,13 @@ static void m2m_hw_shutdown(struct ep93xx_dma_chan *edmac)
 
 static void m2m_fill_desc(struct ep93xx_dma_chan *edmac)
 {
-	struct ep93xx_dma_desc *desc = ep93xx_dma_get_active(edmac);
+	struct ep93xx_dma_desc *desc;
+
+	desc = ep93xx_dma_get_active(edmac);
+	if (!desc) {
+		dev_warn(chan2dev(edmac), "M2M: empty descriptor list\n");
+		return;
+	}
 
 	if (edmac->buffer == 0) {
 		writel(desc->src_addr, edmac->regs + M2M_SAR_BASE0);
@@ -669,24 +690,30 @@ static void ep93xx_dma_tasklet(unsigned long data)
 {
 	struct ep93xx_dma_chan *edmac = (struct ep93xx_dma_chan *)data;
 	struct ep93xx_dma_desc *desc, *d;
-	dma_async_tx_callback callback;
-	void *callback_param;
+	dma_async_tx_callback callback = NULL;
+	void *callback_param = NULL;
 	LIST_HEAD(list);
 
 	spin_lock_irq(&edmac->lock);
+	/*
+	 * If dma_terminate_all() was called before we get to run, the active
+	 * list has become empty. If that happens we aren't supposed to do
+	 * anything more than call ep93xx_dma_advance_work().
+	 */
 	desc = ep93xx_dma_get_active(edmac);
-	if (desc->complete) {
-		edmac->last_completed = desc->txd.cookie;
-		list_splice_init(&edmac->active, &list);
+	if (desc) {
+		if (desc->complete) {
+			edmac->last_completed = desc->txd.cookie;
+			list_splice_init(&edmac->active, &list);
+		}
+		callback = desc->txd.callback;
+		callback_param = desc->txd.callback_param;
 	}
 	spin_unlock_irq(&edmac->lock);
 
 	/* Pick up the next descriptor from the queue */
 	ep93xx_dma_advance_work(edmac);
 
-	callback = desc->txd.callback;
-	callback_param = desc->txd.callback_param;
-
 	/* Now we can release all the chained descriptors */
 	list_for_each_entry_safe(desc, d, &list, node) {
 		/*
@@ -706,13 +733,22 @@ static void ep93xx_dma_tasklet(unsigned long data)
 static irqreturn_t ep93xx_dma_interrupt(int irq, void *dev_id)
 {
 	struct ep93xx_dma_chan *edmac = dev_id;
+	struct ep93xx_dma_desc *desc;
 	irqreturn_t ret = IRQ_HANDLED;
 
 	spin_lock(&edmac->lock);
 
+	desc = ep93xx_dma_get_active(edmac);
+	if (!desc) {
+		dev_warn(chan2dev(edmac),
+			 "got interrupt while active list is empty\n");
+		spin_unlock(&edmac->lock);
+		return IRQ_NONE;
+	}
+
 	switch (edmac->edma->hw_interrupt(edmac)) {
 	case INTERRUPT_DONE:
-		ep93xx_dma_get_active(edmac)->complete = true;
+		desc->complete = true;
 		tasklet_schedule(&edmac->tasklet);
 		break;
 
@@ -803,8 +839,8 @@ static int ep93xx_dma_alloc_chan_resources(struct dma_chan *chan)
 			switch (data->port) {
 			case EP93XX_DMA_SSP:
 			case EP93XX_DMA_IDE:
-				if (data->direction != DMA_TO_DEVICE &&
-				    data->direction != DMA_FROM_DEVICE)
+				if (data->direction != DMA_MEM_TO_DEV &&
+				    data->direction != DMA_DEV_TO_MEM)
 					return -EINVAL;
 				break;
 			default:
@@ -952,7 +988,7 @@ fail:
  */
 static struct dma_async_tx_descriptor *
 ep93xx_dma_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
-			 unsigned int sg_len, enum dma_data_direction dir,
+			 unsigned int sg_len, enum dma_transfer_direction dir,
 			 unsigned long flags)
 {
 	struct ep93xx_dma_chan *edmac = to_ep93xx_dma_chan(chan);
@@ -988,7 +1024,7 @@ ep93xx_dma_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 			goto fail;
 		}
 
-		if (dir == DMA_TO_DEVICE) {
+		if (dir == DMA_MEM_TO_DEV) {
 			desc->src_addr = sg_dma_address(sg);
 			desc->dst_addr = edmac->runtime_addr;
 		} else {
@@ -1032,7 +1068,7 @@ fail:
 static struct dma_async_tx_descriptor *
 ep93xx_dma_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t dma_addr,
 			   size_t buf_len, size_t period_len,
-			   enum dma_data_direction dir)
+			   enum dma_transfer_direction dir)
 {
 	struct ep93xx_dma_chan *edmac = to_ep93xx_dma_chan(chan);
 	struct ep93xx_dma_desc *desc, *first;
@@ -1065,7 +1101,7 @@ ep93xx_dma_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t dma_addr,
 			goto fail;
 		}
 
-		if (dir == DMA_TO_DEVICE) {
+		if (dir == DMA_MEM_TO_DEV) {
 			desc->src_addr = dma_addr + offset;
 			desc->dst_addr = edmac->runtime_addr;
 		} else {
@@ -1133,12 +1169,12 @@ static int ep93xx_dma_slave_config(struct ep93xx_dma_chan *edmac,
 		return -EINVAL;
 
 	switch (config->direction) {
-	case DMA_FROM_DEVICE:
+	case DMA_DEV_TO_MEM:
 		width = config->src_addr_width;
 		addr = config->src_addr;
 		break;
 
-	case DMA_TO_DEVICE:
+	case DMA_MEM_TO_DEV:
 		width = config->dst_addr_width;
 		addr = config->dst_addr;
 		break;
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index 8a781540590..b98070c33ca 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -772,7 +772,7 @@ fail:
  */
 static struct dma_async_tx_descriptor *fsl_dma_prep_slave_sg(
 	struct dma_chan *dchan, struct scatterlist *sgl, unsigned int sg_len,
-	enum dma_data_direction direction, unsigned long flags)
+	enum dma_transfer_direction direction, unsigned long flags)
 {
 	/*
 	 * This operation is not supported on the Freescale DMA controller
@@ -819,7 +819,7 @@ static int fsl_dma_device_control(struct dma_chan *dchan,
 			return -ENXIO;
 
 		/* we set the controller burst size depending on direction */
-		if (config->direction == DMA_TO_DEVICE)
+		if (config->direction == DMA_MEM_TO_DEV)
 			size = config->dst_addr_width * config->dst_maxburst;
 		else
 			size = config->src_addr_width * config->src_maxburst;
diff --git a/drivers/dma/imx-dma.c b/drivers/dma/imx-dma.c
index 4be55f9bb6c..e4383ee2c9a 100644
--- a/drivers/dma/imx-dma.c
+++ b/drivers/dma/imx-dma.c
@@ -107,7 +107,7 @@ static int imxdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 		imx_dma_disable(imxdmac->imxdma_channel);
 		return 0;
 	case DMA_SLAVE_CONFIG:
-		if (dmaengine_cfg->direction == DMA_FROM_DEVICE) {
+		if (dmaengine_cfg->direction == DMA_DEV_TO_MEM) {
 			imxdmac->per_address = dmaengine_cfg->src_addr;
 			imxdmac->watermark_level = dmaengine_cfg->src_maxburst;
 			imxdmac->word_size = dmaengine_cfg->src_addr_width;
@@ -224,7 +224,7 @@ static void imxdma_free_chan_resources(struct dma_chan *chan)
 
 static struct dma_async_tx_descriptor *imxdma_prep_slave_sg(
 		struct dma_chan *chan, struct scatterlist *sgl,
-		unsigned int sg_len, enum dma_data_direction direction,
+		unsigned int sg_len, enum dma_transfer_direction direction,
 		unsigned long flags)
 {
 	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
@@ -241,7 +241,7 @@ static struct dma_async_tx_descriptor *imxdma_prep_slave_sg(
 		dma_length += sg->length;
 	}
 
-	if (direction == DMA_FROM_DEVICE)
+	if (direction == DMA_DEV_TO_MEM)
 		dmamode = DMA_MODE_READ;
 	else
 		dmamode = DMA_MODE_WRITE;
@@ -271,7 +271,7 @@ static struct dma_async_tx_descriptor *imxdma_prep_slave_sg(
 
 static struct dma_async_tx_descriptor *imxdma_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len,
-		size_t period_len, enum dma_data_direction direction)
+		size_t period_len, enum dma_transfer_direction direction)
 {
 	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
 	struct imxdma_engine *imxdma = imxdmac->imxdma;
@@ -317,7 +317,7 @@ static struct dma_async_tx_descriptor *imxdma_prep_dma_cyclic(
 	imxdmac->sg_list[periods].page_link =
 		((unsigned long)imxdmac->sg_list | 0x01) & ~0x02;
 
-	if (direction == DMA_FROM_DEVICE)
+	if (direction == DMA_DEV_TO_MEM)
 		dmamode = DMA_MODE_READ;
 	else
 		dmamode = DMA_MODE_WRITE;
diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c
index f993955a640..a8af379680c 100644
--- a/drivers/dma/imx-sdma.c
+++ b/drivers/dma/imx-sdma.c
@@ -247,7 +247,7 @@ struct sdma_engine;
 struct sdma_channel {
 	struct sdma_engine		*sdma;
 	unsigned int			channel;
-	enum dma_data_direction		direction;
+	enum dma_transfer_direction		direction;
 	enum sdma_peripheral_type	peripheral_type;
 	unsigned int			event_id0;
 	unsigned int			event_id1;
@@ -268,6 +268,8 @@ struct sdma_channel {
 	struct dma_async_tx_descriptor	desc;
 	dma_cookie_t			last_completed;
 	enum dma_status			status;
+	unsigned int			chn_count;
+	unsigned int			chn_real_count;
 };
 
 #define IMX_DMA_SG_LOOP		(1 << 0)
@@ -503,6 +505,7 @@ static void mxc_sdma_handle_channel_normal(struct sdma_channel *sdmac)
 	struct sdma_buffer_descriptor *bd;
 	int i, error = 0;
 
+	sdmac->chn_real_count = 0;
 	/*
 	 * non loop mode. Iterate over all descriptors, collect
 	 * errors and call callback function
@@ -512,6 +515,7 @@ static void mxc_sdma_handle_channel_normal(struct sdma_channel *sdmac)
 
 		 if (bd->mode.status & (BD_DONE | BD_RROR))
 			error = -EIO;
+		 sdmac->chn_real_count += bd->mode.count;
 	}
 
 	if (error)
@@ -519,9 +523,9 @@ static void mxc_sdma_handle_channel_normal(struct sdma_channel *sdmac)
 	else
 		sdmac->status = DMA_SUCCESS;
 
+	sdmac->last_completed = sdmac->desc.cookie;
 	if (sdmac->desc.callback)
 		sdmac->desc.callback(sdmac->desc.callback_param);
-	sdmac->last_completed = sdmac->desc.cookie;
 }
 
 static void mxc_sdma_handle_channel(struct sdma_channel *sdmac)
@@ -650,7 +654,7 @@ static int sdma_load_context(struct sdma_channel *sdmac)
 	struct sdma_buffer_descriptor *bd0 = sdma->channel[0].bd;
 	int ret;
 
-	if (sdmac->direction == DMA_FROM_DEVICE) {
+	if (sdmac->direction == DMA_DEV_TO_MEM) {
 		load_address = sdmac->pc_from_device;
 	} else {
 		load_address = sdmac->pc_to_device;
@@ -832,17 +836,18 @@ static struct sdma_channel *to_sdma_chan(struct dma_chan *chan)
 
 static dma_cookie_t sdma_tx_submit(struct dma_async_tx_descriptor *tx)
 {
+	unsigned long flags;
 	struct sdma_channel *sdmac = to_sdma_chan(tx->chan);
 	struct sdma_engine *sdma = sdmac->sdma;
 	dma_cookie_t cookie;
 
-	spin_lock_irq(&sdmac->lock);
+	spin_lock_irqsave(&sdmac->lock, flags);
 
 	cookie = sdma_assign_cookie(sdmac);
 
 	sdma_enable_channel(sdma, sdmac->channel);
 
-	spin_unlock_irq(&sdmac->lock);
+	spin_unlock_irqrestore(&sdmac->lock, flags);
 
 	return cookie;
 }
@@ -911,7 +916,7 @@ static void sdma_free_chan_resources(struct dma_chan *chan)
 
 static struct dma_async_tx_descriptor *sdma_prep_slave_sg(
 		struct dma_chan *chan, struct scatterlist *sgl,
-		unsigned int sg_len, enum dma_data_direction direction,
+		unsigned int sg_len, enum dma_transfer_direction direction,
 		unsigned long flags)
 {
 	struct sdma_channel *sdmac = to_sdma_chan(chan);
@@ -941,6 +946,7 @@ static struct dma_async_tx_descriptor *sdma_prep_slave_sg(
 		goto err_out;
 	}
 
+	sdmac->chn_count = 0;
 	for_each_sg(sgl, sg, sg_len, i) {
 		struct sdma_buffer_descriptor *bd = &sdmac->bd[i];
 		int param;
@@ -957,6 +963,7 @@ static struct dma_async_tx_descriptor *sdma_prep_slave_sg(
 		}
 
 		bd->mode.count = count;
+		sdmac->chn_count += count;
 
 		if (sdmac->word_size > DMA_SLAVE_BUSWIDTH_4_BYTES) {
 			ret =  -EINVAL;
@@ -1008,7 +1015,7 @@ err_out:
 
 static struct dma_async_tx_descriptor *sdma_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len,
-		size_t period_len, enum dma_data_direction direction)
+		size_t period_len, enum dma_transfer_direction direction)
 {
 	struct sdma_channel *sdmac = to_sdma_chan(chan);
 	struct sdma_engine *sdma = sdmac->sdma;
@@ -1093,7 +1100,7 @@ static int sdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 		sdma_disable_channel(sdmac);
 		return 0;
 	case DMA_SLAVE_CONFIG:
-		if (dmaengine_cfg->direction == DMA_FROM_DEVICE) {
+		if (dmaengine_cfg->direction == DMA_DEV_TO_MEM) {
 			sdmac->per_address = dmaengine_cfg->src_addr;
 			sdmac->watermark_level = dmaengine_cfg->src_maxburst;
 			sdmac->word_size = dmaengine_cfg->src_addr_width;
@@ -1102,6 +1109,7 @@ static int sdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 			sdmac->watermark_level = dmaengine_cfg->dst_maxburst;
 			sdmac->word_size = dmaengine_cfg->dst_addr_width;
 		}
+		sdmac->direction = dmaengine_cfg->direction;
 		return sdma_config_channel(sdmac);
 	default:
 		return -ENOSYS;
@@ -1119,7 +1127,8 @@ static enum dma_status sdma_tx_status(struct dma_chan *chan,
 
 	last_used = chan->cookie;
 
-	dma_set_tx_state(txstate, sdmac->last_completed, last_used, 0);
+	dma_set_tx_state(txstate, sdmac->last_completed, last_used,
+			sdmac->chn_count - sdmac->chn_real_count);
 
 	return sdmac->status;
 }
diff --git a/drivers/dma/intel_mid_dma.c b/drivers/dma/intel_mid_dma.c
index 19a0c64d45d..74f70aadf9e 100644
--- a/drivers/dma/intel_mid_dma.c
+++ b/drivers/dma/intel_mid_dma.c
@@ -280,7 +280,8 @@ static void midc_dostart(struct intel_mid_dma_chan *midc,
  * callbacks but must be called with the lock held.
  */
 static void midc_descriptor_complete(struct intel_mid_dma_chan *midc,
-	       struct intel_mid_dma_desc *desc)
+		struct intel_mid_dma_desc *desc)
+		__releases(&midc->lock) __acquires(&midc->lock)
 {
 	struct dma_async_tx_descriptor	*txd = &desc->txd;
 	dma_async_tx_callback callback_txd = NULL;
@@ -311,6 +312,7 @@ static void midc_descriptor_complete(struct intel_mid_dma_chan *midc,
 			pci_pool_free(desc->lli_pool, desc->lli,
 						desc->lli_phys);
 			pci_pool_destroy(desc->lli_pool);
+			desc->lli = NULL;
 		}
 		list_move(&desc->desc_node, &midc->free_list);
 		midc->busy = false;
@@ -395,10 +397,10 @@ static int midc_lli_fill_sg(struct intel_mid_dma_chan *midc,
 							midc->dma->block_size);
 		/*Populate SAR and DAR values*/
 		sg_phy_addr = sg_phys(sg);
-		if (desc->dirn ==  DMA_TO_DEVICE) {
+		if (desc->dirn ==  DMA_MEM_TO_DEV) {
 			lli_bloc_desc->sar  = sg_phy_addr;
 			lli_bloc_desc->dar  = mids->dma_slave.dst_addr;
-		} else if (desc->dirn ==  DMA_FROM_DEVICE) {
+		} else if (desc->dirn ==  DMA_DEV_TO_MEM) {
 			lli_bloc_desc->sar  = mids->dma_slave.src_addr;
 			lli_bloc_desc->dar  = sg_phy_addr;
 		}
@@ -490,7 +492,9 @@ static enum dma_status intel_mid_dma_tx_status(struct dma_chan *chan,
 
 	ret = dma_async_is_complete(cookie, last_complete, last_used);
 	if (ret != DMA_SUCCESS) {
+		spin_lock_bh(&midc->lock);
 		midc_scan_descriptors(to_middma_device(chan->device), midc);
+		spin_unlock_bh(&midc->lock);
 
 		last_complete = midc->completed;
 		last_used = chan->cookie;
@@ -566,6 +570,7 @@ static int intel_mid_dma_device_control(struct dma_chan *chan,
 			pci_pool_free(desc->lli_pool, desc->lli,
 						desc->lli_phys);
 			pci_pool_destroy(desc->lli_pool);
+			desc->lli = NULL;
 		}
 		list_move(&desc->desc_node, &midc->free_list);
 	}
@@ -632,13 +637,13 @@ static struct dma_async_tx_descriptor *intel_mid_dma_prep_memcpy(
 		if (midc->dma->pimr_mask) {
 			cfg_hi.cfgx.protctl = 0x0; /*default value*/
 			cfg_hi.cfgx.fifo_mode = 1;
-			if (mids->dma_slave.direction == DMA_TO_DEVICE) {
+			if (mids->dma_slave.direction == DMA_MEM_TO_DEV) {
 				cfg_hi.cfgx.src_per = 0;
 				if (mids->device_instance == 0)
 					cfg_hi.cfgx.dst_per = 3;
 				if (mids->device_instance == 1)
 					cfg_hi.cfgx.dst_per = 1;
-			} else if (mids->dma_slave.direction == DMA_FROM_DEVICE) {
+			} else if (mids->dma_slave.direction == DMA_DEV_TO_MEM) {
 				if (mids->device_instance == 0)
 					cfg_hi.cfgx.src_per = 2;
 				if (mids->device_instance == 1)
@@ -682,11 +687,11 @@ static struct dma_async_tx_descriptor *intel_mid_dma_prep_memcpy(
 		ctl_lo.ctlx.sinc = 0;
 		ctl_lo.ctlx.dinc = 0;
 	} else {
-		if (mids->dma_slave.direction == DMA_TO_DEVICE) {
+		if (mids->dma_slave.direction == DMA_MEM_TO_DEV) {
 			ctl_lo.ctlx.sinc = 0;
 			ctl_lo.ctlx.dinc = 2;
 			ctl_lo.ctlx.tt_fc = 1;
-		} else if (mids->dma_slave.direction == DMA_FROM_DEVICE) {
+		} else if (mids->dma_slave.direction == DMA_DEV_TO_MEM) {
 			ctl_lo.ctlx.sinc = 2;
 			ctl_lo.ctlx.dinc = 0;
 			ctl_lo.ctlx.tt_fc = 2;
@@ -732,7 +737,7 @@ err_desc_get:
  */
 static struct dma_async_tx_descriptor *intel_mid_dma_prep_slave_sg(
 			struct dma_chan *chan, struct scatterlist *sgl,
-			unsigned int sg_len, enum dma_data_direction direction,
+			unsigned int sg_len, enum dma_transfer_direction direction,
 			unsigned long flags)
 {
 	struct intel_mid_dma_chan *midc = NULL;
@@ -868,7 +873,7 @@ static int intel_mid_dma_alloc_chan_resources(struct dma_chan *chan)
 	pm_runtime_get_sync(&mid->pdev->dev);
 
 	if (mid->state == SUSPENDED) {
-		if (dma_resume(mid->pdev)) {
+		if (dma_resume(&mid->pdev->dev)) {
 			pr_err("ERR_MDMA: resume failed");
 			return -EFAULT;
 		}
@@ -1099,7 +1104,8 @@ static int mid_setup_dma(struct pci_dev *pdev)
 					LNW_PERIPHRAL_MASK_SIZE);
 		if (dma->mask_reg == NULL) {
 			pr_err("ERR_MDMA:Can't map periphral intr space !!\n");
-			return -ENOMEM;
+			err = -ENOMEM;
+			goto err_ioremap;
 		}
 	} else
 		dma->mask_reg = NULL;
@@ -1196,6 +1202,9 @@ static int mid_setup_dma(struct pci_dev *pdev)
 err_engine:
 	free_irq(pdev->irq, dma);
 err_irq:
+	if (dma->mask_reg)
+		iounmap(dma->mask_reg);
+err_ioremap:
 	pci_pool_destroy(dma->dma_pool);
 err_dma_pool:
 	pr_err("ERR_MDMA:setup_dma failed: %d\n", err);
@@ -1337,8 +1346,9 @@ static void __devexit intel_mid_dma_remove(struct pci_dev *pdev)
 *
 * This function is called by OS when a power event occurs
 */
-int dma_suspend(struct pci_dev *pci, pm_message_t state)
+static int dma_suspend(struct device *dev)
 {
+	struct pci_dev *pci = to_pci_dev(dev);
 	int i;
 	struct middma_device *device = pci_get_drvdata(pci);
 	pr_debug("MDMA: dma_suspend called\n");
@@ -1362,8 +1372,9 @@ int dma_suspend(struct pci_dev *pci, pm_message_t state)
 *
 * This function is called by OS when a power event occurs
 */
-int dma_resume(struct pci_dev *pci)
+int dma_resume(struct device *dev)
 {
+	struct pci_dev *pci = to_pci_dev(dev);
 	int ret;
 	struct middma_device *device = pci_get_drvdata(pci);
 
@@ -1429,6 +1440,8 @@ static const struct dev_pm_ops intel_mid_dma_pm = {
 	.runtime_suspend = dma_runtime_suspend,
 	.runtime_resume = dma_runtime_resume,
 	.runtime_idle = dma_runtime_idle,
+	.suspend = dma_suspend,
+	.resume = dma_resume,
 };
 
 static struct pci_driver intel_mid_dma_pci_driver = {
@@ -1437,8 +1450,6 @@ static struct pci_driver intel_mid_dma_pci_driver = {
 	.probe		=	intel_mid_dma_probe,
 	.remove		=	__devexit_p(intel_mid_dma_remove),
 #ifdef CONFIG_PM
-	.suspend = dma_suspend,
-	.resume = dma_resume,
 	.driver = {
 		.pm = &intel_mid_dma_pm,
 	},
diff --git a/drivers/dma/intel_mid_dma_regs.h b/drivers/dma/intel_mid_dma_regs.h
index aea5ee88ce0..c83d35b97bd 100644
--- a/drivers/dma/intel_mid_dma_regs.h
+++ b/drivers/dma/intel_mid_dma_regs.h
@@ -262,7 +262,7 @@ struct intel_mid_dma_desc {
 	unsigned int			lli_length;
 	unsigned int			current_lli;
 	dma_addr_t			next;
-	enum dma_data_direction		dirn;
+	enum dma_transfer_direction		dirn;
 	enum dma_status			status;
 	enum dma_slave_buswidth		width; /*width of DMA txn*/
 	enum intel_mid_dma_mode		cfg_mode; /*mode configuration*/
@@ -296,6 +296,6 @@ static inline struct intel_mid_dma_slave *to_intel_mid_dma_slave
 }
 
 
-int dma_resume(struct pci_dev *pci);
+int dma_resume(struct device *dev);
 
 #endif /*__INTEL_MID_DMAC_REGS_H__*/
diff --git a/drivers/dma/iop-adma.c b/drivers/dma/iop-adma.c
index e03f811a83d..04be90b645b 100644
--- a/drivers/dma/iop-adma.c
+++ b/drivers/dma/iop-adma.c
@@ -1735,8 +1735,6 @@ static void iop_chan_start_null_xor(struct iop_adma_chan *iop_chan)
 	spin_unlock_bh(&iop_chan->lock);
 }
 
-MODULE_ALIAS("platform:iop-adma");
-
 static struct platform_driver iop_adma_driver = {
 	.probe		= iop_adma_probe,
 	.remove		= __devexit_p(iop_adma_remove),
@@ -1746,19 +1744,9 @@ static struct platform_driver iop_adma_driver = {
 	},
 };
 
-static int __init iop_adma_init (void)
-{
-	return platform_driver_register(&iop_adma_driver);
-}
-
-static void __exit iop_adma_exit (void)
-{
-	platform_driver_unregister(&iop_adma_driver);
-	return;
-}
-module_exit(iop_adma_exit);
-module_init(iop_adma_init);
+module_platform_driver(iop_adma_driver);
 
 MODULE_AUTHOR("Intel Corporation");
 MODULE_DESCRIPTION("IOP ADMA Engine Driver");
 MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:iop-adma");
diff --git a/drivers/dma/ipu/ipu_idmac.c b/drivers/dma/ipu/ipu_idmac.c
index 0e5ef33f90a..6212b16e8cf 100644
--- a/drivers/dma/ipu/ipu_idmac.c
+++ b/drivers/dma/ipu/ipu_idmac.c
@@ -312,7 +312,7 @@ static void ipu_ch_param_set_size(union chan_param_mem *params,
 	case IPU_PIX_FMT_RGB565:
 		params->ip.bpp	= 2;
 		params->ip.pfs	= 4;
-		params->ip.npb	= 7;
+		params->ip.npb	= 15;
 		params->ip.sat	= 2;		/* SAT = 32-bit access */
 		params->ip.ofs0	= 0;		/* Red bit offset */
 		params->ip.ofs1	= 5;		/* Green bit offset */
@@ -422,12 +422,6 @@ static void ipu_ch_param_set_size(union chan_param_mem *params,
 	params->pp.nsb = 1;
 }
 
-static void ipu_ch_param_set_burst_size(union chan_param_mem *params,
-					uint16_t burst_pixels)
-{
-	params->pp.npb = burst_pixels - 1;
-}
-
 static void ipu_ch_param_set_buffer(union chan_param_mem *params,
 				    dma_addr_t buf0, dma_addr_t buf1)
 {
@@ -690,23 +684,6 @@ static int ipu_init_channel_buffer(struct idmac_channel *ichan,
 	ipu_ch_param_set_size(&params, pixel_fmt, width, height, stride_bytes);
 	ipu_ch_param_set_buffer(&params, phyaddr_0, phyaddr_1);
 	ipu_ch_param_set_rotation(&params, rot_mode);
-	/* Some channels (rotation) have restriction on burst length */
-	switch (channel) {
-	case IDMAC_IC_7:	/* Hangs with burst 8, 16, other values
-				   invalid - Table 44-30 */
-/*
-		ipu_ch_param_set_burst_size(&params, 8);
- */
-		break;
-	case IDMAC_SDC_0:
-	case IDMAC_SDC_1:
-		/* In original code only IPU_PIX_FMT_RGB565 was setting burst */
-		ipu_ch_param_set_burst_size(&params, 16);
-		break;
-	case IDMAC_IC_0:
-	default:
-		break;
-	}
 
 	spin_lock_irqsave(&ipu->lock, flags);
 
@@ -1364,7 +1341,7 @@ static void ipu_gc_tasklet(unsigned long arg)
 /* Allocate and initialise a transfer descriptor. */
 static struct dma_async_tx_descriptor *idmac_prep_slave_sg(struct dma_chan *chan,
 		struct scatterlist *sgl, unsigned int sg_len,
-		enum dma_data_direction direction, unsigned long tx_flags)
+		enum dma_transfer_direction direction, unsigned long tx_flags)
 {
 	struct idmac_channel *ichan = to_idmac_chan(chan);
 	struct idmac_tx_desc *desc = NULL;
@@ -1376,7 +1353,7 @@ static struct dma_async_tx_descriptor *idmac_prep_slave_sg(struct dma_chan *chan
 	    chan->chan_id != IDMAC_IC_7)
 		return NULL;
 
-	if (direction != DMA_FROM_DEVICE && direction != DMA_TO_DEVICE) {
+	if (direction != DMA_DEV_TO_MEM && direction != DMA_MEM_TO_DEV) {
 		dev_err(chan->device->dev, "Invalid DMA direction %d!\n", direction);
 		return NULL;
 	}
diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c
index 8ba4edc6185..4d6d4cf6694 100644
--- a/drivers/dma/mpc512x_dma.c
+++ b/drivers/dma/mpc512x_dma.c
@@ -835,17 +835,7 @@ static struct platform_driver mpc_dma_driver = {
 	},
 };
 
-static int __init mpc_dma_init(void)
-{
-	return platform_driver_register(&mpc_dma_driver);
-}
-module_init(mpc_dma_init);
-
-static void __exit mpc_dma_exit(void)
-{
-	platform_driver_unregister(&mpc_dma_driver);
-}
-module_exit(mpc_dma_exit);
+module_platform_driver(mpc_dma_driver);
 
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Piotr Ziecik <kosmo@semihalf.com>");
diff --git a/drivers/dma/mxs-dma.c b/drivers/dma/mxs-dma.c
index fc903c0ed23..b06cd4ca626 100644
--- a/drivers/dma/mxs-dma.c
+++ b/drivers/dma/mxs-dma.c
@@ -44,7 +44,6 @@
 #define HW_APBHX_CTRL0				0x000
 #define BM_APBH_CTRL0_APB_BURST8_EN		(1 << 29)
 #define BM_APBH_CTRL0_APB_BURST_EN		(1 << 28)
-#define BP_APBH_CTRL0_CLKGATE_CHANNEL		8
 #define BP_APBH_CTRL0_RESET_CHANNEL		16
 #define HW_APBHX_CTRL1				0x010
 #define HW_APBHX_CTRL2				0x020
@@ -111,6 +110,7 @@ struct mxs_dma_chan {
 	int				chan_irq;
 	struct mxs_dma_ccw		*ccw;
 	dma_addr_t			ccw_phys;
+	int				desc_count;
 	dma_cookie_t			last_completed;
 	enum dma_status			status;
 	unsigned int			flags;
@@ -130,23 +130,6 @@ struct mxs_dma_engine {
 	struct mxs_dma_chan		mxs_chans[MXS_DMA_CHANNELS];
 };
 
-static inline void mxs_dma_clkgate(struct mxs_dma_chan *mxs_chan, int enable)
-{
-	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
-	int chan_id = mxs_chan->chan.chan_id;
-	int set_clr = enable ? MXS_CLR_ADDR : MXS_SET_ADDR;
-
-	/* enable apbh channel clock */
-	if (dma_is_apbh()) {
-		if (apbh_is_old())
-			writel(1 << (chan_id + BP_APBH_CTRL0_CLKGATE_CHANNEL),
-				mxs_dma->base + HW_APBHX_CTRL0 + set_clr);
-		else
-			writel(1 << chan_id,
-				mxs_dma->base + HW_APBHX_CTRL0 + set_clr);
-	}
-}
-
 static void mxs_dma_reset_chan(struct mxs_dma_chan *mxs_chan)
 {
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
@@ -165,9 +148,6 @@ static void mxs_dma_enable_chan(struct mxs_dma_chan *mxs_chan)
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
 	int chan_id = mxs_chan->chan.chan_id;
 
-	/* clkgate needs to be enabled before writing other registers */
-	mxs_dma_clkgate(mxs_chan, 1);
-
 	/* set cmd_addr up */
 	writel(mxs_chan->ccw_phys,
 		mxs_dma->base + HW_APBHX_CHn_NXTCMDAR(chan_id));
@@ -178,9 +158,6 @@ static void mxs_dma_enable_chan(struct mxs_dma_chan *mxs_chan)
 
 static void mxs_dma_disable_chan(struct mxs_dma_chan *mxs_chan)
 {
-	/* disable apbh channel clock */
-	mxs_dma_clkgate(mxs_chan, 0);
-
 	mxs_chan->status = DMA_SUCCESS;
 }
 
@@ -268,7 +245,7 @@ static irqreturn_t mxs_dma_int_handler(int irq, void *dev_id)
 	/*
 	 * When both completion and error of termination bits set at the
 	 * same time, we do not take it as an error.  IOW, it only becomes
-	 * an error we need to handler here in case of ether it's (1) an bus
+	 * an error we need to handle here in case of either it's (1) a bus
 	 * error or (2) a termination error with no completion.
 	 */
 	stat2 = ((stat2 >> MXS_DMA_CHANNELS) & stat2) | /* (1) */
@@ -338,10 +315,7 @@ static int mxs_dma_alloc_chan_resources(struct dma_chan *chan)
 	if (ret)
 		goto err_clk;
 
-	/* clkgate needs to be enabled for reset to finish */
-	mxs_dma_clkgate(mxs_chan, 1);
 	mxs_dma_reset_chan(mxs_chan);
-	mxs_dma_clkgate(mxs_chan, 0);
 
 	dma_async_tx_descriptor_init(&mxs_chan->desc, chan);
 	mxs_chan->desc.tx_submit = mxs_dma_tx_submit;
@@ -377,7 +351,7 @@ static void mxs_dma_free_chan_resources(struct dma_chan *chan)
 
 static struct dma_async_tx_descriptor *mxs_dma_prep_slave_sg(
 		struct dma_chan *chan, struct scatterlist *sgl,
-		unsigned int sg_len, enum dma_data_direction direction,
+		unsigned int sg_len, enum dma_transfer_direction direction,
 		unsigned long append)
 {
 	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
@@ -386,7 +360,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_slave_sg(
 	struct scatterlist *sg;
 	int i, j;
 	u32 *pio;
-	static int idx;
+	int idx = append ? mxs_chan->desc_count : 0;
 
 	if (mxs_chan->status == DMA_IN_PROGRESS && !append)
 		return NULL;
@@ -417,7 +391,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_slave_sg(
 		idx = 0;
 	}
 
-	if (direction == DMA_NONE) {
+	if (direction == DMA_TRANS_NONE) {
 		ccw = &mxs_chan->ccw[idx++];
 		pio = (u32 *) sgl;
 
@@ -450,7 +424,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_slave_sg(
 			ccw->bits |= CCW_CHAIN;
 			ccw->bits |= CCW_HALT_ON_TERM;
 			ccw->bits |= CCW_TERM_FLUSH;
-			ccw->bits |= BF_CCW(direction == DMA_FROM_DEVICE ?
+			ccw->bits |= BF_CCW(direction == DMA_DEV_TO_MEM ?
 					MXS_DMA_CMD_WRITE : MXS_DMA_CMD_READ,
 					COMMAND);
 
@@ -462,6 +436,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_slave_sg(
 			}
 		}
 	}
+	mxs_chan->desc_count = idx;
 
 	return &mxs_chan->desc;
 
@@ -472,7 +447,7 @@ err_out:
 
 static struct dma_async_tx_descriptor *mxs_dma_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len,
-		size_t period_len, enum dma_data_direction direction)
+		size_t period_len, enum dma_transfer_direction direction)
 {
 	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
@@ -515,7 +490,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_dma_cyclic(
 		ccw->bits |= CCW_IRQ;
 		ccw->bits |= CCW_HALT_ON_TERM;
 		ccw->bits |= CCW_TERM_FLUSH;
-		ccw->bits |= BF_CCW(direction == DMA_FROM_DEVICE ?
+		ccw->bits |= BF_CCW(direction == DMA_DEV_TO_MEM ?
 				MXS_DMA_CMD_WRITE : MXS_DMA_CMD_READ, COMMAND);
 
 		dma_addr += period_len;
@@ -523,6 +498,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_dma_cyclic(
 
 		i++;
 	}
+	mxs_chan->desc_count = i;
 
 	return &mxs_chan->desc;
 
@@ -539,8 +515,8 @@ static int mxs_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 
 	switch (cmd) {
 	case DMA_TERMINATE_ALL:
-		mxs_dma_disable_chan(mxs_chan);
 		mxs_dma_reset_chan(mxs_chan);
+		mxs_dma_disable_chan(mxs_chan);
 		break;
 	case DMA_PAUSE:
 		mxs_dma_pause_chan(mxs_chan);
@@ -580,7 +556,7 @@ static int __init mxs_dma_init(struct mxs_dma_engine *mxs_dma)
 
 	ret = clk_prepare_enable(mxs_dma->clk);
 	if (ret)
-		goto err_out;
+		return ret;
 
 	ret = mxs_reset_block(mxs_dma->base);
 	if (ret)
@@ -604,11 +580,8 @@ static int __init mxs_dma_init(struct mxs_dma_engine *mxs_dma)
 	writel(MXS_DMA_CHANNELS_MASK << MXS_DMA_CHANNELS,
 		mxs_dma->base + HW_APBHX_CTRL1 + MXS_SET_ADDR);
 
-	clk_disable_unprepare(mxs_dma->clk);
-
-	return 0;
-
 err_out:
+	clk_disable_unprepare(mxs_dma->clk);
 	return ret;
 }
 
diff --git a/drivers/dma/pch_dma.c b/drivers/dma/pch_dma.c
index a6d0e3dbed0..823f58179f9 100644
--- a/drivers/dma/pch_dma.c
+++ b/drivers/dma/pch_dma.c
@@ -1,7 +1,7 @@
 /*
  * Topcliff PCH DMA controller driver
  * Copyright (c) 2010 Intel Corporation
- * Copyright (C) 2011 OKI SEMICONDUCTOR CO., LTD.
+ * Copyright (C) 2011 LAPIS Semiconductor Co., Ltd.
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 as
@@ -99,7 +99,7 @@ struct pch_dma_desc {
 struct pch_dma_chan {
 	struct dma_chan		chan;
 	void __iomem *membase;
-	enum dma_data_direction	dir;
+	enum dma_transfer_direction dir;
 	struct tasklet_struct	tasklet;
 	unsigned long		err_status;
 
@@ -224,7 +224,7 @@ static void pdc_set_dir(struct dma_chan *chan)
 		mask_ctl = DMA_MASK_CTL0_MODE & ~(DMA_CTL0_MODE_MASK_BITS <<
 				       (DMA_CTL0_BITS_PER_CH * chan->chan_id));
 		val &= mask_mode;
-		if (pd_chan->dir == DMA_TO_DEVICE)
+		if (pd_chan->dir == DMA_MEM_TO_DEV)
 			val |= 0x1 << (DMA_CTL0_BITS_PER_CH * chan->chan_id +
 				       DMA_CTL0_DIR_SHIFT_BITS);
 		else
@@ -242,7 +242,7 @@ static void pdc_set_dir(struct dma_chan *chan)
 		mask_ctl = DMA_MASK_CTL2_MODE & ~(DMA_CTL0_MODE_MASK_BITS <<
 						 (DMA_CTL0_BITS_PER_CH * ch));
 		val &= mask_mode;
-		if (pd_chan->dir == DMA_TO_DEVICE)
+		if (pd_chan->dir == DMA_MEM_TO_DEV)
 			val |= 0x1 << (DMA_CTL0_BITS_PER_CH * ch +
 				       DMA_CTL0_DIR_SHIFT_BITS);
 		else
@@ -607,7 +607,7 @@ static void pd_issue_pending(struct dma_chan *chan)
 
 static struct dma_async_tx_descriptor *pd_prep_slave_sg(struct dma_chan *chan,
 			struct scatterlist *sgl, unsigned int sg_len,
-			enum dma_data_direction direction, unsigned long flags)
+			enum dma_transfer_direction direction, unsigned long flags)
 {
 	struct pch_dma_chan *pd_chan = to_pd_chan(chan);
 	struct pch_dma_slave *pd_slave = chan->private;
@@ -623,9 +623,9 @@ static struct dma_async_tx_descriptor *pd_prep_slave_sg(struct dma_chan *chan,
 		return NULL;
 	}
 
-	if (direction == DMA_FROM_DEVICE)
+	if (direction == DMA_DEV_TO_MEM)
 		reg = pd_slave->rx_reg;
-	else if (direction == DMA_TO_DEVICE)
+	else if (direction == DMA_MEM_TO_DEV)
 		reg = pd_slave->tx_reg;
 	else
 		return NULL;
@@ -1018,6 +1018,8 @@ static void __devexit pch_dma_remove(struct pci_dev *pdev)
 #define PCI_DEVICE_ID_ML7223_DMA2_4CH	0x800E
 #define PCI_DEVICE_ID_ML7223_DMA3_4CH	0x8017
 #define PCI_DEVICE_ID_ML7223_DMA4_4CH	0x803B
+#define PCI_DEVICE_ID_ML7831_DMA1_8CH	0x8810
+#define PCI_DEVICE_ID_ML7831_DMA2_4CH	0x8815
 
 DEFINE_PCI_DEVICE_TABLE(pch_dma_id_table) = {
 	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_EG20T_PCH_DMA_8CH), 8 },
@@ -1030,6 +1032,8 @@ DEFINE_PCI_DEVICE_TABLE(pch_dma_id_table) = {
 	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA2_4CH), 4}, /* Video SPI */
 	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA3_4CH), 4}, /* Security */
 	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA4_4CH), 4}, /* FPGA */
+	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7831_DMA1_8CH), 8}, /* UART */
+	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7831_DMA2_4CH), 4}, /* SPI */
 	{ 0, },
 };
 
@@ -1057,7 +1061,7 @@ static void __exit pch_dma_exit(void)
 module_init(pch_dma_init);
 module_exit(pch_dma_exit);
 
-MODULE_DESCRIPTION("Intel EG20T PCH / OKI SEMICONDUCTOR ML7213 IOH "
+MODULE_DESCRIPTION("Intel EG20T PCH / LAPIS Semicon ML7213/ML7223/ML7831 IOH "
 		   "DMA controller driver");
 MODULE_AUTHOR("Yong Wang <yong.y.wang@intel.com>");
 MODULE_LICENSE("GPL v2");
diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c
index 09adcfcd953..b8ec03ee8e2 100644
--- a/drivers/dma/pl330.c
+++ b/drivers/dma/pl330.c
@@ -350,14 +350,14 @@ static int pl330_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, unsigned
 	case DMA_SLAVE_CONFIG:
 		slave_config = (struct dma_slave_config *)arg;
 
-		if (slave_config->direction == DMA_TO_DEVICE) {
+		if (slave_config->direction == DMA_MEM_TO_DEV) {
 			if (slave_config->dst_addr)
 				pch->fifo_addr = slave_config->dst_addr;
 			if (slave_config->dst_addr_width)
 				pch->burst_sz = __ffs(slave_config->dst_addr_width);
 			if (slave_config->dst_maxburst)
 				pch->burst_len = slave_config->dst_maxburst;
-		} else if (slave_config->direction == DMA_FROM_DEVICE) {
+		} else if (slave_config->direction == DMA_DEV_TO_MEM) {
 			if (slave_config->src_addr)
 				pch->fifo_addr = slave_config->src_addr;
 			if (slave_config->src_addr_width)
@@ -621,7 +621,7 @@ static inline int get_burst_len(struct dma_pl330_desc *desc, size_t len)
 
 static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t len,
-		size_t period_len, enum dma_data_direction direction)
+		size_t period_len, enum dma_transfer_direction direction)
 {
 	struct dma_pl330_desc *desc;
 	struct dma_pl330_chan *pch = to_pchan(chan);
@@ -636,14 +636,14 @@ static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic(
 	}
 
 	switch (direction) {
-	case DMA_TO_DEVICE:
+	case DMA_MEM_TO_DEV:
 		desc->rqcfg.src_inc = 1;
 		desc->rqcfg.dst_inc = 0;
 		desc->req.rqtype = MEMTODEV;
 		src = dma_addr;
 		dst = pch->fifo_addr;
 		break;
-	case DMA_FROM_DEVICE:
+	case DMA_DEV_TO_MEM:
 		desc->rqcfg.src_inc = 0;
 		desc->rqcfg.dst_inc = 1;
 		desc->req.rqtype = DEVTOMEM;
@@ -710,7 +710,7 @@ pl330_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dst,
 
 static struct dma_async_tx_descriptor *
 pl330_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
-		unsigned int sg_len, enum dma_data_direction direction,
+		unsigned int sg_len, enum dma_transfer_direction direction,
 		unsigned long flg)
 {
 	struct dma_pl330_desc *first, *desc = NULL;
@@ -759,7 +759,7 @@ pl330_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		else
 			list_add_tail(&desc->node, &first->node);
 
-		if (direction == DMA_TO_DEVICE) {
+		if (direction == DMA_MEM_TO_DEV) {
 			desc->rqcfg.src_inc = 1;
 			desc->rqcfg.dst_inc = 0;
 			desc->req.rqtype = MEMTODEV;
@@ -834,17 +834,7 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 
 	amba_set_drvdata(adev, pdmac);
 
-#ifdef CONFIG_PM_RUNTIME
-	/* to use the runtime PM helper functions */
-	pm_runtime_enable(&adev->dev);
-
-	/* enable the power domain */
-	if (pm_runtime_get_sync(&adev->dev)) {
-		dev_err(&adev->dev, "failed to get runtime pm\n");
-		ret = -ENODEV;
-		goto probe_err1;
-	}
-#else
+#ifndef CONFIG_PM_RUNTIME
 	/* enable dma clk */
 	clk_enable(pdmac->clk);
 #endif
@@ -977,10 +967,7 @@ static int __devexit pl330_remove(struct amba_device *adev)
 	res = &adev->res;
 	release_mem_region(res->start, resource_size(res));
 
-#ifdef CONFIG_PM_RUNTIME
-	pm_runtime_put(&adev->dev);
-	pm_runtime_disable(&adev->dev);
-#else
+#ifndef CONFIG_PM_RUNTIME
 	clk_disable(pdmac->clk);
 #endif
 
diff --git a/drivers/dma/shdma.c b/drivers/dma/shdma.c
index 81809c2b46a..54043cd831c 100644
--- a/drivers/dma/shdma.c
+++ b/drivers/dma/shdma.c
@@ -23,7 +23,6 @@
 #include <linux/interrupt.h>
 #include <linux/dmaengine.h>
 #include <linux/delay.h>
-#include <linux/dma-mapping.h>
 #include <linux/platform_device.h>
 #include <linux/pm_runtime.h>
 #include <linux/sh_dma.h>
@@ -57,6 +56,15 @@ static LIST_HEAD(sh_dmae_devices);
 static unsigned long sh_dmae_slave_used[BITS_TO_LONGS(SH_DMA_SLAVE_NUMBER)];
 
 static void sh_dmae_chan_ld_cleanup(struct sh_dmae_chan *sh_chan, bool all);
+static void sh_chan_xfer_ld_queue(struct sh_dmae_chan *sh_chan);
+
+static void chclr_write(struct sh_dmae_chan *sh_dc, u32 data)
+{
+	struct sh_dmae_device *shdev = to_sh_dev(sh_dc);
+
+	__raw_writel(data, shdev->chan_reg +
+		     shdev->pdata->channel[sh_dc->id].chclr_offset);
+}
 
 static void sh_dmae_writel(struct sh_dmae_chan *sh_dc, u32 data, u32 reg)
 {
@@ -129,6 +137,15 @@ static int sh_dmae_rst(struct sh_dmae_device *shdev)
 
 	dmaor = dmaor_read(shdev) & ~(DMAOR_NMIF | DMAOR_AE | DMAOR_DME);
 
+	if (shdev->pdata->chclr_present) {
+		int i;
+		for (i = 0; i < shdev->pdata->channel_num; i++) {
+			struct sh_dmae_chan *sh_chan = shdev->chan[i];
+			if (sh_chan)
+				chclr_write(sh_chan, 0);
+		}
+	}
+
 	dmaor_write(shdev, dmaor | shdev->pdata->dmaor_init);
 
 	dmaor = dmaor_read(shdev);
@@ -139,6 +156,10 @@ static int sh_dmae_rst(struct sh_dmae_device *shdev)
 		dev_warn(shdev->common.dev, "Can't initialize DMAOR.\n");
 		return -EIO;
 	}
+	if (shdev->pdata->dmaor_init & ~dmaor)
+		dev_warn(shdev->common.dev,
+			 "DMAOR=0x%x hasn't latched the initial value 0x%x.\n",
+			 dmaor, shdev->pdata->dmaor_init);
 	return 0;
 }
 
@@ -259,8 +280,6 @@ static int dmae_set_dmars(struct sh_dmae_chan *sh_chan, u16 val)
 	return 0;
 }
 
-static void sh_chan_xfer_ld_queue(struct sh_dmae_chan *sh_chan);
-
 static dma_cookie_t sh_dmae_tx_submit(struct dma_async_tx_descriptor *tx)
 {
 	struct sh_desc *desc = tx_to_sh_desc(tx), *chunk, *last = desc, *c;
@@ -340,6 +359,8 @@ static dma_cookie_t sh_dmae_tx_submit(struct dma_async_tx_descriptor *tx)
 				sh_chan_xfer_ld_queue(sh_chan);
 			sh_chan->pm_state = DMAE_PM_ESTABLISHED;
 		}
+	} else {
+		sh_chan->pm_state = DMAE_PM_PENDING;
 	}
 
 	spin_unlock_irq(&sh_chan->desc_lock);
@@ -479,19 +500,19 @@ static void sh_dmae_free_chan_resources(struct dma_chan *chan)
  * @sh_chan:	DMA channel
  * @flags:	DMA transfer flags
  * @dest:	destination DMA address, incremented when direction equals
- *		DMA_FROM_DEVICE or DMA_BIDIRECTIONAL
+ *		DMA_DEV_TO_MEM
  * @src:	source DMA address, incremented when direction equals
- *		DMA_TO_DEVICE or DMA_BIDIRECTIONAL
+ *		DMA_MEM_TO_DEV
  * @len:	DMA transfer length
  * @first:	if NULL, set to the current descriptor and cookie set to -EBUSY
  * @direction:	needed for slave DMA to decide which address to keep constant,
- *		equals DMA_BIDIRECTIONAL for MEMCPY
+ *		equals DMA_MEM_TO_MEM for MEMCPY
  * Returns 0 or an error
  * Locks: called with desc_lock held
  */
 static struct sh_desc *sh_dmae_add_desc(struct sh_dmae_chan *sh_chan,
 	unsigned long flags, dma_addr_t *dest, dma_addr_t *src, size_t *len,
-	struct sh_desc **first, enum dma_data_direction direction)
+	struct sh_desc **first, enum dma_transfer_direction direction)
 {
 	struct sh_desc *new;
 	size_t copy_size;
@@ -531,9 +552,9 @@ static struct sh_desc *sh_dmae_add_desc(struct sh_dmae_chan *sh_chan,
 	new->direction = direction;
 
 	*len -= copy_size;
-	if (direction == DMA_BIDIRECTIONAL || direction == DMA_TO_DEVICE)
+	if (direction == DMA_MEM_TO_MEM || direction == DMA_MEM_TO_DEV)
 		*src += copy_size;
-	if (direction == DMA_BIDIRECTIONAL || direction == DMA_FROM_DEVICE)
+	if (direction == DMA_MEM_TO_MEM || direction == DMA_DEV_TO_MEM)
 		*dest += copy_size;
 
 	return new;
@@ -546,12 +567,12 @@ static struct sh_desc *sh_dmae_add_desc(struct sh_dmae_chan *sh_chan,
  * converted to scatter-gather to guarantee consistent locking and a correct
  * list manipulation. For slave DMA direction carries the usual meaning, and,
  * logically, the SG list is RAM and the addr variable contains slave address,
- * e.g., the FIFO I/O register. For MEMCPY direction equals DMA_BIDIRECTIONAL
+ * e.g., the FIFO I/O register. For MEMCPY direction equals DMA_MEM_TO_MEM
  * and the SG list contains only one element and points at the source buffer.
  */
 static struct dma_async_tx_descriptor *sh_dmae_prep_sg(struct sh_dmae_chan *sh_chan,
 	struct scatterlist *sgl, unsigned int sg_len, dma_addr_t *addr,
-	enum dma_data_direction direction, unsigned long flags)
+	enum dma_transfer_direction direction, unsigned long flags)
 {
 	struct scatterlist *sg;
 	struct sh_desc *first = NULL, *new = NULL /* compiler... */;
@@ -592,7 +613,7 @@ static struct dma_async_tx_descriptor *sh_dmae_prep_sg(struct sh_dmae_chan *sh_c
 			dev_dbg(sh_chan->dev, "Add SG #%d@%p[%d], dma %llx\n",
 				i, sg, len, (unsigned long long)sg_addr);
 
-			if (direction == DMA_FROM_DEVICE)
+			if (direction == DMA_DEV_TO_MEM)
 				new = sh_dmae_add_desc(sh_chan, flags,
 						&sg_addr, addr, &len, &first,
 						direction);
@@ -646,13 +667,13 @@ static struct dma_async_tx_descriptor *sh_dmae_prep_memcpy(
 	sg_dma_address(&sg) = dma_src;
 	sg_dma_len(&sg) = len;
 
-	return sh_dmae_prep_sg(sh_chan, &sg, 1, &dma_dest, DMA_BIDIRECTIONAL,
+	return sh_dmae_prep_sg(sh_chan, &sg, 1, &dma_dest, DMA_MEM_TO_MEM,
 			       flags);
 }
 
 static struct dma_async_tx_descriptor *sh_dmae_prep_slave_sg(
 	struct dma_chan *chan, struct scatterlist *sgl, unsigned int sg_len,
-	enum dma_data_direction direction, unsigned long flags)
+	enum dma_transfer_direction direction, unsigned long flags)
 {
 	struct sh_dmae_slave *param;
 	struct sh_dmae_chan *sh_chan;
@@ -996,7 +1017,7 @@ static void dmae_do_tasklet(unsigned long data)
 	spin_lock_irq(&sh_chan->desc_lock);
 	list_for_each_entry(desc, &sh_chan->ld_queue, node) {
 		if (desc->mark == DESC_SUBMITTED &&
-		    ((desc->direction == DMA_FROM_DEVICE &&
+		    ((desc->direction == DMA_DEV_TO_MEM &&
 		      (desc->hw.dar + desc->hw.tcr) == dar_buf) ||
 		     (desc->hw.sar + desc->hw.tcr) == sar_buf)) {
 			dev_dbg(sh_chan->dev, "done #%d@%p dst %u\n",
@@ -1225,6 +1246,8 @@ static int __init sh_dmae_probe(struct platform_device *pdev)
 
 	platform_set_drvdata(pdev, shdev);
 
+	shdev->common.dev = &pdev->dev;
+
 	pm_runtime_enable(&pdev->dev);
 	pm_runtime_get_sync(&pdev->dev);
 
@@ -1254,7 +1277,6 @@ static int __init sh_dmae_probe(struct platform_device *pdev)
 	shdev->common.device_prep_slave_sg = sh_dmae_prep_slave_sg;
 	shdev->common.device_control = sh_dmae_control;
 
-	shdev->common.dev = &pdev->dev;
 	/* Default transfer size of 32 bytes requires 32-byte alignment */
 	shdev->common.copy_align = LOG2_DEFAULT_XFER_SIZE;
 
@@ -1435,22 +1457,17 @@ static int sh_dmae_runtime_resume(struct device *dev)
 #ifdef CONFIG_PM
 static int sh_dmae_suspend(struct device *dev)
 {
-	struct sh_dmae_device *shdev = dev_get_drvdata(dev);
-	int i;
-
-	for (i = 0; i < shdev->pdata->channel_num; i++) {
-		struct sh_dmae_chan *sh_chan = shdev->chan[i];
-		if (sh_chan->descs_allocated)
-			sh_chan->pm_error = pm_runtime_put_sync(dev);
-	}
-
 	return 0;
 }
 
 static int sh_dmae_resume(struct device *dev)
 {
 	struct sh_dmae_device *shdev = dev_get_drvdata(dev);
-	int i;
+	int i, ret;
+
+	ret = sh_dmae_rst(shdev);
+	if (ret < 0)
+		dev_err(dev, "Failed to reset!\n");
 
 	for (i = 0; i < shdev->pdata->channel_num; i++) {
 		struct sh_dmae_chan *sh_chan = shdev->chan[i];
@@ -1459,9 +1476,6 @@ static int sh_dmae_resume(struct device *dev)
 		if (!sh_chan->descs_allocated)
 			continue;
 
-		if (!sh_chan->pm_error)
-			pm_runtime_get_sync(dev);
-
 		if (param) {
 			const struct sh_dmae_slave_config *cfg = param->config;
 			dmae_set_dmars(sh_chan, cfg->mid_rid);
diff --git a/drivers/dma/sirf-dma.c b/drivers/dma/sirf-dma.c
new file mode 100644
index 00000000000..2333810d168
--- /dev/null
+++ b/drivers/dma/sirf-dma.c
@@ -0,0 +1,707 @@
+/*
+ * DMA controller driver for CSR SiRFprimaII
+ *
+ * Copyright (c) 2011 Cambridge Silicon Radio Limited, a CSR plc group company.
+ *
+ * Licensed under GPLv2 or later.
+ */
+
+#include <linux/module.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/slab.h>
+#include <linux/of_irq.h>
+#include <linux/of_address.h>
+#include <linux/of_device.h>
+#include <linux/of_platform.h>
+#include <linux/sirfsoc_dma.h>
+
+#define SIRFSOC_DMA_DESCRIPTORS                 16
+#define SIRFSOC_DMA_CHANNELS                    16
+
+#define SIRFSOC_DMA_CH_ADDR                     0x00
+#define SIRFSOC_DMA_CH_XLEN                     0x04
+#define SIRFSOC_DMA_CH_YLEN                     0x08
+#define SIRFSOC_DMA_CH_CTRL                     0x0C
+
+#define SIRFSOC_DMA_WIDTH_0                     0x100
+#define SIRFSOC_DMA_CH_VALID                    0x140
+#define SIRFSOC_DMA_CH_INT                      0x144
+#define SIRFSOC_DMA_INT_EN                      0x148
+#define SIRFSOC_DMA_CH_LOOP_CTRL                0x150
+
+#define SIRFSOC_DMA_MODE_CTRL_BIT               4
+#define SIRFSOC_DMA_DIR_CTRL_BIT                5
+
+/* xlen and dma_width register is in 4 bytes boundary */
+#define SIRFSOC_DMA_WORD_LEN			4
+
+struct sirfsoc_dma_desc {
+	struct dma_async_tx_descriptor	desc;
+	struct list_head		node;
+
+	/* SiRFprimaII 2D-DMA parameters */
+
+	int             xlen;           /* DMA xlen */
+	int             ylen;           /* DMA ylen */
+	int             width;          /* DMA width */
+	int             dir;
+	bool            cyclic;         /* is loop DMA? */
+	u32             addr;		/* DMA buffer address */
+};
+
+struct sirfsoc_dma_chan {
+	struct dma_chan			chan;
+	struct list_head		free;
+	struct list_head		prepared;
+	struct list_head		queued;
+	struct list_head		active;
+	struct list_head		completed;
+	dma_cookie_t			completed_cookie;
+	unsigned long			happened_cyclic;
+	unsigned long			completed_cyclic;
+
+	/* Lock for this structure */
+	spinlock_t			lock;
+
+	int				mode;
+};
+
+struct sirfsoc_dma {
+	struct dma_device		dma;
+	struct tasklet_struct		tasklet;
+	struct sirfsoc_dma_chan		channels[SIRFSOC_DMA_CHANNELS];
+	void __iomem			*base;
+	int				irq;
+};
+
+#define DRV_NAME	"sirfsoc_dma"
+
+/* Convert struct dma_chan to struct sirfsoc_dma_chan */
+static inline
+struct sirfsoc_dma_chan *dma_chan_to_sirfsoc_dma_chan(struct dma_chan *c)
+{
+	return container_of(c, struct sirfsoc_dma_chan, chan);
+}
+
+/* Convert struct dma_chan to struct sirfsoc_dma */
+static inline struct sirfsoc_dma *dma_chan_to_sirfsoc_dma(struct dma_chan *c)
+{
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(c);
+	return container_of(schan, struct sirfsoc_dma, channels[c->chan_id]);
+}
+
+/* Execute all queued DMA descriptors */
+static void sirfsoc_dma_execute(struct sirfsoc_dma_chan *schan)
+{
+	struct sirfsoc_dma *sdma = dma_chan_to_sirfsoc_dma(&schan->chan);
+	int cid = schan->chan.chan_id;
+	struct sirfsoc_dma_desc *sdesc = NULL;
+
+	/*
+	 * lock has been held by functions calling this, so we don't hold
+	 * lock again
+	 */
+
+	sdesc = list_first_entry(&schan->queued, struct sirfsoc_dma_desc,
+		node);
+	/* Move the first queued descriptor to active list */
+	list_move_tail(&schan->queued, &schan->active);
+
+	/* Start the DMA transfer */
+	writel_relaxed(sdesc->width, sdma->base + SIRFSOC_DMA_WIDTH_0 +
+		cid * 4);
+	writel_relaxed(cid | (schan->mode << SIRFSOC_DMA_MODE_CTRL_BIT) |
+		(sdesc->dir << SIRFSOC_DMA_DIR_CTRL_BIT),
+		sdma->base + cid * 0x10 + SIRFSOC_DMA_CH_CTRL);
+	writel_relaxed(sdesc->xlen, sdma->base + cid * 0x10 +
+		SIRFSOC_DMA_CH_XLEN);
+	writel_relaxed(sdesc->ylen, sdma->base + cid * 0x10 +
+		SIRFSOC_DMA_CH_YLEN);
+	writel_relaxed(readl_relaxed(sdma->base + SIRFSOC_DMA_INT_EN) |
+		(1 << cid), sdma->base + SIRFSOC_DMA_INT_EN);
+
+	/*
+	 * writel has an implict memory write barrier to make sure data is
+	 * flushed into memory before starting DMA
+	 */
+	writel(sdesc->addr >> 2, sdma->base + cid * 0x10 + SIRFSOC_DMA_CH_ADDR);
+
+	if (sdesc->cyclic) {
+		writel((1 << cid) | 1 << (cid + 16) |
+			readl_relaxed(sdma->base + SIRFSOC_DMA_CH_LOOP_CTRL),
+			sdma->base + SIRFSOC_DMA_CH_LOOP_CTRL);
+		schan->happened_cyclic = schan->completed_cyclic = 0;
+	}
+}
+
+/* Interrupt handler */
+static irqreturn_t sirfsoc_dma_irq(int irq, void *data)
+{
+	struct sirfsoc_dma *sdma = data;
+	struct sirfsoc_dma_chan *schan;
+	struct sirfsoc_dma_desc *sdesc = NULL;
+	u32 is;
+	int ch;
+
+	is = readl(sdma->base + SIRFSOC_DMA_CH_INT);
+	while ((ch = fls(is) - 1) >= 0) {
+		is &= ~(1 << ch);
+		writel_relaxed(1 << ch, sdma->base + SIRFSOC_DMA_CH_INT);
+		schan = &sdma->channels[ch];
+
+		spin_lock(&schan->lock);
+
+		sdesc = list_first_entry(&schan->active, struct sirfsoc_dma_desc,
+			node);
+		if (!sdesc->cyclic) {
+			/* Execute queued descriptors */
+			list_splice_tail_init(&schan->active, &schan->completed);
+			if (!list_empty(&schan->queued))
+				sirfsoc_dma_execute(schan);
+		} else
+			schan->happened_cyclic++;
+
+		spin_unlock(&schan->lock);
+	}
+
+	/* Schedule tasklet */
+	tasklet_schedule(&sdma->tasklet);
+
+	return IRQ_HANDLED;
+}
+
+/* process completed descriptors */
+static void sirfsoc_dma_process_completed(struct sirfsoc_dma *sdma)
+{
+	dma_cookie_t last_cookie = 0;
+	struct sirfsoc_dma_chan *schan;
+	struct sirfsoc_dma_desc *sdesc;
+	struct dma_async_tx_descriptor *desc;
+	unsigned long flags;
+	unsigned long happened_cyclic;
+	LIST_HEAD(list);
+	int i;
+
+	for (i = 0; i < sdma->dma.chancnt; i++) {
+		schan = &sdma->channels[i];
+
+		/* Get all completed descriptors */
+		spin_lock_irqsave(&schan->lock, flags);
+		if (!list_empty(&schan->completed)) {
+			list_splice_tail_init(&schan->completed, &list);
+			spin_unlock_irqrestore(&schan->lock, flags);
+
+			/* Execute callbacks and run dependencies */
+			list_for_each_entry(sdesc, &list, node) {
+				desc = &sdesc->desc;
+
+				if (desc->callback)
+					desc->callback(desc->callback_param);
+
+				last_cookie = desc->cookie;
+				dma_run_dependencies(desc);
+			}
+
+			/* Free descriptors */
+			spin_lock_irqsave(&schan->lock, flags);
+			list_splice_tail_init(&list, &schan->free);
+			schan->completed_cookie = last_cookie;
+			spin_unlock_irqrestore(&schan->lock, flags);
+		} else {
+			/* for cyclic channel, desc is always in active list */
+			sdesc = list_first_entry(&schan->active, struct sirfsoc_dma_desc,
+				node);
+
+			if (!sdesc || (sdesc && !sdesc->cyclic)) {
+				/* without active cyclic DMA */
+				spin_unlock_irqrestore(&schan->lock, flags);
+				continue;
+			}
+
+			/* cyclic DMA */
+			happened_cyclic = schan->happened_cyclic;
+			spin_unlock_irqrestore(&schan->lock, flags);
+
+			desc = &sdesc->desc;
+			while (happened_cyclic != schan->completed_cyclic) {
+				if (desc->callback)
+					desc->callback(desc->callback_param);
+				schan->completed_cyclic++;
+			}
+		}
+	}
+}
+
+/* DMA Tasklet */
+static void sirfsoc_dma_tasklet(unsigned long data)
+{
+	struct sirfsoc_dma *sdma = (void *)data;
+
+	sirfsoc_dma_process_completed(sdma);
+}
+
+/* Submit descriptor to hardware */
+static dma_cookie_t sirfsoc_dma_tx_submit(struct dma_async_tx_descriptor *txd)
+{
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(txd->chan);
+	struct sirfsoc_dma_desc *sdesc;
+	unsigned long flags;
+	dma_cookie_t cookie;
+
+	sdesc = container_of(txd, struct sirfsoc_dma_desc, desc);
+
+	spin_lock_irqsave(&schan->lock, flags);
+
+	/* Move descriptor to queue */
+	list_move_tail(&sdesc->node, &schan->queued);
+
+	/* Update cookie */
+	cookie = schan->chan.cookie + 1;
+	if (cookie <= 0)
+		cookie = 1;
+
+	schan->chan.cookie = cookie;
+	sdesc->desc.cookie = cookie;
+
+	spin_unlock_irqrestore(&schan->lock, flags);
+
+	return cookie;
+}
+
+static int sirfsoc_dma_slave_config(struct sirfsoc_dma_chan *schan,
+	struct dma_slave_config *config)
+{
+	unsigned long flags;
+
+	if ((config->src_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES) ||
+		(config->dst_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES))
+		return -EINVAL;
+
+	spin_lock_irqsave(&schan->lock, flags);
+	schan->mode = (config->src_maxburst == 4 ? 1 : 0);
+	spin_unlock_irqrestore(&schan->lock, flags);
+
+	return 0;
+}
+
+static int sirfsoc_dma_terminate_all(struct sirfsoc_dma_chan *schan)
+{
+	struct sirfsoc_dma *sdma = dma_chan_to_sirfsoc_dma(&schan->chan);
+	int cid = schan->chan.chan_id;
+	unsigned long flags;
+
+	writel_relaxed(readl_relaxed(sdma->base + SIRFSOC_DMA_INT_EN) &
+		~(1 << cid), sdma->base + SIRFSOC_DMA_INT_EN);
+	writel_relaxed(1 << cid, sdma->base + SIRFSOC_DMA_CH_VALID);
+
+	writel_relaxed(readl_relaxed(sdma->base + SIRFSOC_DMA_CH_LOOP_CTRL)
+		& ~((1 << cid) | 1 << (cid + 16)),
+			sdma->base + SIRFSOC_DMA_CH_LOOP_CTRL);
+
+	spin_lock_irqsave(&schan->lock, flags);
+	list_splice_tail_init(&schan->active, &schan->free);
+	list_splice_tail_init(&schan->queued, &schan->free);
+	spin_unlock_irqrestore(&schan->lock, flags);
+
+	return 0;
+}
+
+static int sirfsoc_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
+	unsigned long arg)
+{
+	struct dma_slave_config *config;
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
+
+	switch (cmd) {
+	case DMA_TERMINATE_ALL:
+		return sirfsoc_dma_terminate_all(schan);
+	case DMA_SLAVE_CONFIG:
+		config = (struct dma_slave_config *)arg;
+		return sirfsoc_dma_slave_config(schan, config);
+
+	default:
+		break;
+	}
+
+	return -ENOSYS;
+}
+
+/* Alloc channel resources */
+static int sirfsoc_dma_alloc_chan_resources(struct dma_chan *chan)
+{
+	struct sirfsoc_dma *sdma = dma_chan_to_sirfsoc_dma(chan);
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
+	struct sirfsoc_dma_desc *sdesc;
+	unsigned long flags;
+	LIST_HEAD(descs);
+	int i;
+
+	/* Alloc descriptors for this channel */
+	for (i = 0; i < SIRFSOC_DMA_DESCRIPTORS; i++) {
+		sdesc = kzalloc(sizeof(*sdesc), GFP_KERNEL);
+		if (!sdesc) {
+			dev_notice(sdma->dma.dev, "Memory allocation error. "
+				"Allocated only %u descriptors\n", i);
+			break;
+		}
+
+		dma_async_tx_descriptor_init(&sdesc->desc, chan);
+		sdesc->desc.flags = DMA_CTRL_ACK;
+		sdesc->desc.tx_submit = sirfsoc_dma_tx_submit;
+
+		list_add_tail(&sdesc->node, &descs);
+	}
+
+	/* Return error only if no descriptors were allocated */
+	if (i == 0)
+		return -ENOMEM;
+
+	spin_lock_irqsave(&schan->lock, flags);
+
+	list_splice_tail_init(&descs, &schan->free);
+	spin_unlock_irqrestore(&schan->lock, flags);
+
+	return i;
+}
+
+/* Free channel resources */
+static void sirfsoc_dma_free_chan_resources(struct dma_chan *chan)
+{
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
+	struct sirfsoc_dma_desc *sdesc, *tmp;
+	unsigned long flags;
+	LIST_HEAD(descs);
+
+	spin_lock_irqsave(&schan->lock, flags);
+
+	/* Channel must be idle */
+	BUG_ON(!list_empty(&schan->prepared));
+	BUG_ON(!list_empty(&schan->queued));
+	BUG_ON(!list_empty(&schan->active));
+	BUG_ON(!list_empty(&schan->completed));
+
+	/* Move data */
+	list_splice_tail_init(&schan->free, &descs);
+
+	spin_unlock_irqrestore(&schan->lock, flags);
+
+	/* Free descriptors */
+	list_for_each_entry_safe(sdesc, tmp, &descs, node)
+		kfree(sdesc);
+}
+
+/* Send pending descriptor to hardware */
+static void sirfsoc_dma_issue_pending(struct dma_chan *chan)
+{
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
+	unsigned long flags;
+
+	spin_lock_irqsave(&schan->lock, flags);
+
+	if (list_empty(&schan->active) && !list_empty(&schan->queued))
+		sirfsoc_dma_execute(schan);
+
+	spin_unlock_irqrestore(&schan->lock, flags);
+}
+
+/* Check request completion status */
+static enum dma_status
+sirfsoc_dma_tx_status(struct dma_chan *chan, dma_cookie_t cookie,
+	struct dma_tx_state *txstate)
+{
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
+	unsigned long flags;
+	dma_cookie_t last_used;
+	dma_cookie_t last_complete;
+
+	spin_lock_irqsave(&schan->lock, flags);
+	last_used = schan->chan.cookie;
+	last_complete = schan->completed_cookie;
+	spin_unlock_irqrestore(&schan->lock, flags);
+
+	dma_set_tx_state(txstate, last_complete, last_used, 0);
+	return dma_async_is_complete(cookie, last_complete, last_used);
+}
+
+static struct dma_async_tx_descriptor *sirfsoc_dma_prep_interleaved(
+	struct dma_chan *chan, struct dma_interleaved_template *xt,
+	unsigned long flags)
+{
+	struct sirfsoc_dma *sdma = dma_chan_to_sirfsoc_dma(chan);
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
+	struct sirfsoc_dma_desc *sdesc = NULL;
+	unsigned long iflags;
+	int ret;
+
+	if ((xt->dir != DMA_MEM_TO_DEV) || (xt->dir != DMA_DEV_TO_MEM)) {
+		ret = -EINVAL;
+		goto err_dir;
+	}
+
+	/* Get free descriptor */
+	spin_lock_irqsave(&schan->lock, iflags);
+	if (!list_empty(&schan->free)) {
+		sdesc = list_first_entry(&schan->free, struct sirfsoc_dma_desc,
+			node);
+		list_del(&sdesc->node);
+	}
+	spin_unlock_irqrestore(&schan->lock, iflags);
+
+	if (!sdesc) {
+		/* try to free completed descriptors */
+		sirfsoc_dma_process_completed(sdma);
+		ret = 0;
+		goto no_desc;
+	}
+
+	/* Place descriptor in prepared list */
+	spin_lock_irqsave(&schan->lock, iflags);
+
+	/*
+	 * Number of chunks in a frame can only be 1 for prima2
+	 * and ylen (number of frame - 1) must be at least 0
+	 */
+	if ((xt->frame_size == 1) && (xt->numf > 0)) {
+		sdesc->cyclic = 0;
+		sdesc->xlen = xt->sgl[0].size / SIRFSOC_DMA_WORD_LEN;
+		sdesc->width = (xt->sgl[0].size + xt->sgl[0].icg) /
+				SIRFSOC_DMA_WORD_LEN;
+		sdesc->ylen = xt->numf - 1;
+		if (xt->dir == DMA_MEM_TO_DEV) {
+			sdesc->addr = xt->src_start;
+			sdesc->dir = 1;
+		} else {
+			sdesc->addr = xt->dst_start;
+			sdesc->dir = 0;
+		}
+
+		list_add_tail(&sdesc->node, &schan->prepared);
+	} else {
+		pr_err("sirfsoc DMA Invalid xfer\n");
+		ret = -EINVAL;
+		goto err_xfer;
+	}
+	spin_unlock_irqrestore(&schan->lock, iflags);
+
+	return &sdesc->desc;
+err_xfer:
+	spin_unlock_irqrestore(&schan->lock, iflags);
+no_desc:
+err_dir:
+	return ERR_PTR(ret);
+}
+
+static struct dma_async_tx_descriptor *
+sirfsoc_dma_prep_cyclic(struct dma_chan *chan, dma_addr_t addr,
+	size_t buf_len, size_t period_len,
+	enum dma_transfer_direction direction)
+{
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
+	struct sirfsoc_dma_desc *sdesc = NULL;
+	unsigned long iflags;
+
+	/*
+	 * we only support cycle transfer with 2 period
+	 * If the X-length is set to 0, it would be the loop mode.
+	 * The DMA address keeps increasing until reaching the end of a loop
+	 * area whose size is defined by (DMA_WIDTH x (Y_LENGTH + 1)). Then
+	 * the DMA address goes back to the beginning of this area.
+	 * In loop mode, the DMA data region is divided into two parts, BUFA
+	 * and BUFB. DMA controller generates interrupts twice in each loop:
+	 * when the DMA address reaches the end of BUFA or the end of the
+	 * BUFB
+	 */
+	if (buf_len !=  2 * period_len)
+		return ERR_PTR(-EINVAL);
+
+	/* Get free descriptor */
+	spin_lock_irqsave(&schan->lock, iflags);
+	if (!list_empty(&schan->free)) {
+		sdesc = list_first_entry(&schan->free, struct sirfsoc_dma_desc,
+			node);
+		list_del(&sdesc->node);
+	}
+	spin_unlock_irqrestore(&schan->lock, iflags);
+
+	if (!sdesc)
+		return 0;
+
+	/* Place descriptor in prepared list */
+	spin_lock_irqsave(&schan->lock, iflags);
+	sdesc->addr = addr;
+	sdesc->cyclic = 1;
+	sdesc->xlen = 0;
+	sdesc->ylen = buf_len / SIRFSOC_DMA_WORD_LEN - 1;
+	sdesc->width = 1;
+	list_add_tail(&sdesc->node, &schan->prepared);
+	spin_unlock_irqrestore(&schan->lock, iflags);
+
+	return &sdesc->desc;
+}
+
+/*
+ * The DMA controller consists of 16 independent DMA channels.
+ * Each channel is allocated to a different function
+ */
+bool sirfsoc_dma_filter_id(struct dma_chan *chan, void *chan_id)
+{
+	unsigned int ch_nr = (unsigned int) chan_id;
+
+	if (ch_nr == chan->chan_id +
+		chan->device->dev_id * SIRFSOC_DMA_CHANNELS)
+		return true;
+
+	return false;
+}
+EXPORT_SYMBOL(sirfsoc_dma_filter_id);
+
+static int __devinit sirfsoc_dma_probe(struct platform_device *op)
+{
+	struct device_node *dn = op->dev.of_node;
+	struct device *dev = &op->dev;
+	struct dma_device *dma;
+	struct sirfsoc_dma *sdma;
+	struct sirfsoc_dma_chan *schan;
+	struct resource res;
+	ulong regs_start, regs_size;
+	u32 id;
+	int ret, i;
+
+	sdma = devm_kzalloc(dev, sizeof(*sdma), GFP_KERNEL);
+	if (!sdma) {
+		dev_err(dev, "Memory exhausted!\n");
+		return -ENOMEM;
+	}
+
+	if (of_property_read_u32(dn, "cell-index", &id)) {
+		dev_err(dev, "Fail to get DMAC index\n");
+		ret = -ENODEV;
+		goto free_mem;
+	}
+
+	sdma->irq = irq_of_parse_and_map(dn, 0);
+	if (sdma->irq == NO_IRQ) {
+		dev_err(dev, "Error mapping IRQ!\n");
+		ret = -EINVAL;
+		goto free_mem;
+	}
+
+	ret = of_address_to_resource(dn, 0, &res);
+	if (ret) {
+		dev_err(dev, "Error parsing memory region!\n");
+		goto free_mem;
+	}
+
+	regs_start = res.start;
+	regs_size = resource_size(&res);
+
+	sdma->base = devm_ioremap(dev, regs_start, regs_size);
+	if (!sdma->base) {
+		dev_err(dev, "Error mapping memory region!\n");
+		ret = -ENOMEM;
+		goto irq_dispose;
+	}
+
+	ret = devm_request_irq(dev, sdma->irq, &sirfsoc_dma_irq, 0, DRV_NAME,
+		sdma);
+	if (ret) {
+		dev_err(dev, "Error requesting IRQ!\n");
+		ret = -EINVAL;
+		goto unmap_mem;
+	}
+
+	dma = &sdma->dma;
+	dma->dev = dev;
+	dma->chancnt = SIRFSOC_DMA_CHANNELS;
+
+	dma->device_alloc_chan_resources = sirfsoc_dma_alloc_chan_resources;
+	dma->device_free_chan_resources = sirfsoc_dma_free_chan_resources;
+	dma->device_issue_pending = sirfsoc_dma_issue_pending;
+	dma->device_control = sirfsoc_dma_control;
+	dma->device_tx_status = sirfsoc_dma_tx_status;
+	dma->device_prep_interleaved_dma = sirfsoc_dma_prep_interleaved;
+	dma->device_prep_dma_cyclic = sirfsoc_dma_prep_cyclic;
+
+	INIT_LIST_HEAD(&dma->channels);
+	dma_cap_set(DMA_SLAVE, dma->cap_mask);
+	dma_cap_set(DMA_CYCLIC, dma->cap_mask);
+	dma_cap_set(DMA_INTERLEAVE, dma->cap_mask);
+	dma_cap_set(DMA_PRIVATE, dma->cap_mask);
+
+	for (i = 0; i < dma->chancnt; i++) {
+		schan = &sdma->channels[i];
+
+		schan->chan.device = dma;
+		schan->chan.cookie = 1;
+		schan->completed_cookie = schan->chan.cookie;
+
+		INIT_LIST_HEAD(&schan->free);
+		INIT_LIST_HEAD(&schan->prepared);
+		INIT_LIST_HEAD(&schan->queued);
+		INIT_LIST_HEAD(&schan->active);
+		INIT_LIST_HEAD(&schan->completed);
+
+		spin_lock_init(&schan->lock);
+		list_add_tail(&schan->chan.device_node, &dma->channels);
+	}
+
+	tasklet_init(&sdma->tasklet, sirfsoc_dma_tasklet, (unsigned long)sdma);
+
+	/* Register DMA engine */
+	dev_set_drvdata(dev, sdma);
+	ret = dma_async_device_register(dma);
+	if (ret)
+		goto free_irq;
+
+	dev_info(dev, "initialized SIRFSOC DMAC driver\n");
+
+	return 0;
+
+free_irq:
+	devm_free_irq(dev, sdma->irq, sdma);
+irq_dispose:
+	irq_dispose_mapping(sdma->irq);
+unmap_mem:
+	iounmap(sdma->base);
+free_mem:
+	devm_kfree(dev, sdma);
+	return ret;
+}
+
+static int __devexit sirfsoc_dma_remove(struct platform_device *op)
+{
+	struct device *dev = &op->dev;
+	struct sirfsoc_dma *sdma = dev_get_drvdata(dev);
+
+	dma_async_device_unregister(&sdma->dma);
+	devm_free_irq(dev, sdma->irq, sdma);
+	irq_dispose_mapping(sdma->irq);
+	iounmap(sdma->base);
+	devm_kfree(dev, sdma);
+	return 0;
+}
+
+static struct of_device_id sirfsoc_dma_match[] = {
+	{ .compatible = "sirf,prima2-dmac", },
+	{},
+};
+
+static struct platform_driver sirfsoc_dma_driver = {
+	.probe		= sirfsoc_dma_probe,
+	.remove		= __devexit_p(sirfsoc_dma_remove),
+	.driver = {
+		.name = DRV_NAME,
+		.owner = THIS_MODULE,
+		.of_match_table	= sirfsoc_dma_match,
+	},
+};
+
+module_platform_driver(sirfsoc_dma_driver);
+
+MODULE_AUTHOR("Rongjun Ying <rongjun.ying@csr.com>, "
+	"Barry Song <baohua.song@csr.com>");
+MODULE_DESCRIPTION("SIRFSOC DMA control driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c
index 13259cad0ce..cc5ecbc067a 100644
--- a/drivers/dma/ste_dma40.c
+++ b/drivers/dma/ste_dma40.c
@@ -14,6 +14,8 @@
 #include <linux/platform_device.h>
 #include <linux/clk.h>
 #include <linux/delay.h>
+#include <linux/pm.h>
+#include <linux/pm_runtime.h>
 #include <linux/err.h>
 #include <linux/amba/bus.h>
 
@@ -32,6 +34,9 @@
 /* Maximum iterations taken before giving up suspending a channel */
 #define D40_SUSPEND_MAX_IT 500
 
+/* Milliseconds */
+#define DMA40_AUTOSUSPEND_DELAY	100
+
 /* Hardware requirement on LCLA alignment */
 #define LCLA_ALIGNMENT 0x40000
 
@@ -62,6 +67,55 @@ enum d40_command {
 	D40_DMA_SUSPENDED	= 3
 };
 
+/*
+ * These are the registers that has to be saved and later restored
+ * when the DMA hw is powered off.
+ * TODO: Add save/restore of D40_DREG_GCC on dma40 v3 or later, if that works.
+ */
+static u32 d40_backup_regs[] = {
+	D40_DREG_LCPA,
+	D40_DREG_LCLA,
+	D40_DREG_PRMSE,
+	D40_DREG_PRMSO,
+	D40_DREG_PRMOE,
+	D40_DREG_PRMOO,
+};
+
+#define BACKUP_REGS_SZ ARRAY_SIZE(d40_backup_regs)
+
+/* TODO: Check if all these registers have to be saved/restored on dma40 v3 */
+static u32 d40_backup_regs_v3[] = {
+	D40_DREG_PSEG1,
+	D40_DREG_PSEG2,
+	D40_DREG_PSEG3,
+	D40_DREG_PSEG4,
+	D40_DREG_PCEG1,
+	D40_DREG_PCEG2,
+	D40_DREG_PCEG3,
+	D40_DREG_PCEG4,
+	D40_DREG_RSEG1,
+	D40_DREG_RSEG2,
+	D40_DREG_RSEG3,
+	D40_DREG_RSEG4,
+	D40_DREG_RCEG1,
+	D40_DREG_RCEG2,
+	D40_DREG_RCEG3,
+	D40_DREG_RCEG4,
+};
+
+#define BACKUP_REGS_SZ_V3 ARRAY_SIZE(d40_backup_regs_v3)
+
+static u32 d40_backup_regs_chan[] = {
+	D40_CHAN_REG_SSCFG,
+	D40_CHAN_REG_SSELT,
+	D40_CHAN_REG_SSPTR,
+	D40_CHAN_REG_SSLNK,
+	D40_CHAN_REG_SDCFG,
+	D40_CHAN_REG_SDELT,
+	D40_CHAN_REG_SDPTR,
+	D40_CHAN_REG_SDLNK,
+};
+
 /**
  * struct d40_lli_pool - Structure for keeping LLIs in memory
  *
@@ -96,7 +150,7 @@ struct d40_lli_pool {
  * during a transfer.
  * @node: List entry.
  * @is_in_client_list: true if the client owns this descriptor.
- * the previous one.
+ * @cyclic: true if this is a cyclic job
  *
  * This descriptor is used for both logical and physical transfers.
  */
@@ -143,6 +197,7 @@ struct d40_lcla_pool {
  * channels.
  *
  * @lock: A lock protection this entity.
+ * @reserved: True if used by secure world or otherwise.
  * @num: The physical channel number of this entity.
  * @allocated_src: Bit mapped to show which src event line's are mapped to
  * this physical channel. Can also be free or physically allocated.
@@ -152,6 +207,7 @@ struct d40_lcla_pool {
  */
 struct d40_phy_res {
 	spinlock_t lock;
+	bool	   reserved;
 	int	   num;
 	u32	   allocated_src;
 	u32	   allocated_dst;
@@ -185,7 +241,6 @@ struct d40_base;
  * @src_def_cfg: Default cfg register setting for src.
  * @dst_def_cfg: Default cfg register setting for dst.
  * @log_def: Default logical channel settings.
- * @lcla: Space for one dst src pair for logical channel transfers.
  * @lcpa: Pointer to dst and src lcpa settings.
  * @runtime_addr: runtime configured address.
  * @runtime_direction: runtime configured direction.
@@ -217,7 +272,7 @@ struct d40_chan {
 	struct d40_log_lli_full		*lcpa;
 	/* Runtime reconfiguration */
 	dma_addr_t			runtime_addr;
-	enum dma_data_direction		runtime_direction;
+	enum dma_transfer_direction	runtime_direction;
 };
 
 /**
@@ -241,6 +296,7 @@ struct d40_chan {
  * @dma_both: dma_device channels that can do both memcpy and slave transfers.
  * @dma_slave: dma_device channels that can do only do slave transfers.
  * @dma_memcpy: dma_device channels that can do only do memcpy transfers.
+ * @phy_chans: Room for all possible physical channels in system.
  * @log_chans: Room for all possible logical channels in system.
  * @lookup_log_chans: Used to map interrupt number to logical channel. Points
  * to log_chans entries.
@@ -248,12 +304,20 @@ struct d40_chan {
  * to phy_chans entries.
  * @plat_data: Pointer to provided platform_data which is the driver
  * configuration.
+ * @lcpa_regulator: Pointer to hold the regulator for the esram bank for lcla.
  * @phy_res: Vector containing all physical channels.
  * @lcla_pool: lcla pool settings and data.
  * @lcpa_base: The virtual mapped address of LCPA.
  * @phy_lcpa: The physical address of the LCPA.
  * @lcpa_size: The size of the LCPA area.
  * @desc_slab: cache for descriptors.
+ * @reg_val_backup: Here the values of some hardware registers are stored
+ * before the DMA is powered off. They are restored when the power is back on.
+ * @reg_val_backup_v3: Backup of registers that only exits on dma40 v3 and
+ * later.
+ * @reg_val_backup_chan: Backup data for standard channel parameter registers.
+ * @gcc_pwr_off_mask: Mask to maintain the channels that can be turned off.
+ * @initialized: true if the dma has been initialized
  */
 struct d40_base {
 	spinlock_t			 interrupt_lock;
@@ -275,6 +339,7 @@ struct d40_base {
 	struct d40_chan			**lookup_log_chans;
 	struct d40_chan			**lookup_phy_chans;
 	struct stedma40_platform_data	 *plat_data;
+	struct regulator		 *lcpa_regulator;
 	/* Physical half channels */
 	struct d40_phy_res		 *phy_res;
 	struct d40_lcla_pool		  lcla_pool;
@@ -282,6 +347,11 @@ struct d40_base {
 	dma_addr_t			  phy_lcpa;
 	resource_size_t			  lcpa_size;
 	struct kmem_cache		 *desc_slab;
+	u32				  reg_val_backup[BACKUP_REGS_SZ];
+	u32				  reg_val_backup_v3[BACKUP_REGS_SZ_V3];
+	u32				 *reg_val_backup_chan;
+	u16				  gcc_pwr_off_mask;
+	bool				  initialized;
 };
 
 /**
@@ -479,13 +549,14 @@ static struct d40_desc *d40_desc_get(struct d40_chan *d40c)
 		struct d40_desc *d;
 		struct d40_desc *_d;
 
-		list_for_each_entry_safe(d, _d, &d40c->client, node)
+		list_for_each_entry_safe(d, _d, &d40c->client, node) {
 			if (async_tx_test_ack(&d->txd)) {
 				d40_desc_remove(d);
 				desc = d;
 				memset(desc, 0, sizeof(*desc));
 				break;
 			}
+		}
 	}
 
 	if (!desc)
@@ -536,6 +607,7 @@ static void d40_log_lli_to_lcxa(struct d40_chan *chan, struct d40_desc *desc)
 	bool cyclic = desc->cyclic;
 	int curr_lcla = -EINVAL;
 	int first_lcla = 0;
+	bool use_esram_lcla = chan->base->plat_data->use_esram_lcla;
 	bool linkback;
 
 	/*
@@ -608,11 +680,16 @@ static void d40_log_lli_to_lcxa(struct d40_chan *chan, struct d40_desc *desc)
 				       &lli->src[lli_current],
 				       next_lcla, flags);
 
-		dma_sync_single_range_for_device(chan->base->dev,
-					pool->dma_addr, lcla_offset,
-					2 * sizeof(struct d40_log_lli),
-					DMA_TO_DEVICE);
-
+		/*
+		 * Cache maintenance is not needed if lcla is
+		 * mapped in esram
+		 */
+		if (!use_esram_lcla) {
+			dma_sync_single_range_for_device(chan->base->dev,
+						pool->dma_addr, lcla_offset,
+						2 * sizeof(struct d40_log_lli),
+						DMA_TO_DEVICE);
+		}
 		curr_lcla = next_lcla;
 
 		if (curr_lcla == -EINVAL || curr_lcla == first_lcla) {
@@ -740,7 +817,61 @@ static int d40_sg_2_dmalen(struct scatterlist *sgl, int sg_len,
 	return len;
 }
 
-/* Support functions for logical channels */
+
+#ifdef CONFIG_PM
+static void dma40_backup(void __iomem *baseaddr, u32 *backup,
+			 u32 *regaddr, int num, bool save)
+{
+	int i;
+
+	for (i = 0; i < num; i++) {
+		void __iomem *addr = baseaddr + regaddr[i];
+
+		if (save)
+			backup[i] = readl_relaxed(addr);
+		else
+			writel_relaxed(backup[i], addr);
+	}
+}
+
+static void d40_save_restore_registers(struct d40_base *base, bool save)
+{
+	int i;
+
+	/* Save/Restore channel specific registers */
+	for (i = 0; i < base->num_phy_chans; i++) {
+		void __iomem *addr;
+		int idx;
+
+		if (base->phy_res[i].reserved)
+			continue;
+
+		addr = base->virtbase + D40_DREG_PCBASE + i * D40_DREG_PCDELTA;
+		idx = i * ARRAY_SIZE(d40_backup_regs_chan);
+
+		dma40_backup(addr, &base->reg_val_backup_chan[idx],
+			     d40_backup_regs_chan,
+			     ARRAY_SIZE(d40_backup_regs_chan),
+			     save);
+	}
+
+	/* Save/Restore global registers */
+	dma40_backup(base->virtbase, base->reg_val_backup,
+		     d40_backup_regs, ARRAY_SIZE(d40_backup_regs),
+		     save);
+
+	/* Save/Restore registers only existing on dma40 v3 and later */
+	if (base->rev >= 3)
+		dma40_backup(base->virtbase, base->reg_val_backup_v3,
+			     d40_backup_regs_v3,
+			     ARRAY_SIZE(d40_backup_regs_v3),
+			     save);
+}
+#else
+static void d40_save_restore_registers(struct d40_base *base, bool save)
+{
+}
+#endif
 
 static int d40_channel_execute_command(struct d40_chan *d40c,
 				       enum d40_command command)
@@ -973,6 +1104,10 @@ static void d40_config_write(struct d40_chan *d40c)
 		/* Set LIDX for lcla */
 		writel(lidx, chanbase + D40_CHAN_REG_SSELT);
 		writel(lidx, chanbase + D40_CHAN_REG_SDELT);
+
+		/* Clear LNK which will be used by d40_chan_has_events() */
+		writel(0, chanbase + D40_CHAN_REG_SSLNK);
+		writel(0, chanbase + D40_CHAN_REG_SDLNK);
 	}
 }
 
@@ -1013,6 +1148,7 @@ static int d40_pause(struct d40_chan *d40c)
 	if (!d40c->busy)
 		return 0;
 
+	pm_runtime_get_sync(d40c->base->dev);
 	spin_lock_irqsave(&d40c->lock, flags);
 
 	res = d40_channel_execute_command(d40c, D40_DMA_SUSPEND_REQ);
@@ -1025,7 +1161,8 @@ static int d40_pause(struct d40_chan *d40c)
 								  D40_DMA_RUN);
 		}
 	}
-
+	pm_runtime_mark_last_busy(d40c->base->dev);
+	pm_runtime_put_autosuspend(d40c->base->dev);
 	spin_unlock_irqrestore(&d40c->lock, flags);
 	return res;
 }
@@ -1039,7 +1176,7 @@ static int d40_resume(struct d40_chan *d40c)
 		return 0;
 
 	spin_lock_irqsave(&d40c->lock, flags);
-
+	pm_runtime_get_sync(d40c->base->dev);
 	if (d40c->base->rev == 0)
 		if (chan_is_logical(d40c)) {
 			res = d40_channel_execute_command(d40c,
@@ -1057,6 +1194,8 @@ static int d40_resume(struct d40_chan *d40c)
 	}
 
 no_suspend:
+	pm_runtime_mark_last_busy(d40c->base->dev);
+	pm_runtime_put_autosuspend(d40c->base->dev);
 	spin_unlock_irqrestore(&d40c->lock, flags);
 	return res;
 }
@@ -1129,7 +1268,10 @@ static struct d40_desc *d40_queue_start(struct d40_chan *d40c)
 	d40d = d40_first_queued(d40c);
 
 	if (d40d != NULL) {
-		d40c->busy = true;
+		if (!d40c->busy)
+			d40c->busy = true;
+
+		pm_runtime_get_sync(d40c->base->dev);
 
 		/* Remove from queue */
 		d40_desc_remove(d40d);
@@ -1190,6 +1332,8 @@ static void dma_tc_handle(struct d40_chan *d40c)
 
 		if (d40_queue_start(d40c) == NULL)
 			d40c->busy = false;
+		pm_runtime_mark_last_busy(d40c->base->dev);
+		pm_runtime_put_autosuspend(d40c->base->dev);
 	}
 
 	d40c->pending_tx++;
@@ -1405,11 +1549,16 @@ static int d40_validate_conf(struct d40_chan *d40c,
 	return res;
 }
 
-static bool d40_alloc_mask_set(struct d40_phy_res *phy, bool is_src,
-			       int log_event_line, bool is_log)
+static bool d40_alloc_mask_set(struct d40_phy_res *phy,
+			       bool is_src, int log_event_line, bool is_log,
+			       bool *first_user)
 {
 	unsigned long flags;
 	spin_lock_irqsave(&phy->lock, flags);
+
+	*first_user = ((phy->allocated_src | phy->allocated_dst)
+			== D40_ALLOC_FREE);
+
 	if (!is_log) {
 		/* Physical interrupts are masked per physical full channel */
 		if (phy->allocated_src == D40_ALLOC_FREE &&
@@ -1490,7 +1639,7 @@ out:
 	return is_free;
 }
 
-static int d40_allocate_channel(struct d40_chan *d40c)
+static int d40_allocate_channel(struct d40_chan *d40c, bool *first_phy_user)
 {
 	int dev_type;
 	int event_group;
@@ -1526,7 +1675,8 @@ static int d40_allocate_channel(struct d40_chan *d40c)
 			for (i = 0; i < d40c->base->num_phy_chans; i++) {
 
 				if (d40_alloc_mask_set(&phys[i], is_src,
-						       0, is_log))
+						       0, is_log,
+						       first_phy_user))
 					goto found_phy;
 			}
 		} else
@@ -1536,7 +1686,8 @@ static int d40_allocate_channel(struct d40_chan *d40c)
 					if (d40_alloc_mask_set(&phys[i],
 							       is_src,
 							       0,
-							       is_log))
+							       is_log,
+							       first_phy_user))
 						goto found_phy;
 				}
 			}
@@ -1552,6 +1703,25 @@ found_phy:
 	/* Find logical channel */
 	for (j = 0; j < d40c->base->num_phy_chans; j += 8) {
 		int phy_num = j + event_group * 2;
+
+		if (d40c->dma_cfg.use_fixed_channel) {
+			i = d40c->dma_cfg.phy_channel;
+
+			if ((i != phy_num) && (i != phy_num + 1)) {
+				dev_err(chan2dev(d40c),
+					"invalid fixed phy channel %d\n", i);
+				return -EINVAL;
+			}
+
+			if (d40_alloc_mask_set(&phys[i], is_src, event_line,
+					       is_log, first_phy_user))
+				goto found_log;
+
+			dev_err(chan2dev(d40c),
+				"could not allocate fixed phy channel %d\n", i);
+			return -EINVAL;
+		}
+
 		/*
 		 * Spread logical channels across all available physical rather
 		 * than pack every logical channel at the first available phy
@@ -1560,13 +1730,15 @@ found_phy:
 		if (is_src) {
 			for (i = phy_num; i < phy_num + 2; i++) {
 				if (d40_alloc_mask_set(&phys[i], is_src,
-						       event_line, is_log))
+						       event_line, is_log,
+						       first_phy_user))
 					goto found_log;
 			}
 		} else {
 			for (i = phy_num + 1; i >= phy_num; i--) {
 				if (d40_alloc_mask_set(&phys[i], is_src,
-						       event_line, is_log))
+						       event_line, is_log,
+						       first_phy_user))
 					goto found_log;
 			}
 		}
@@ -1643,10 +1815,11 @@ static int d40_free_dma(struct d40_chan *d40c)
 		return -EINVAL;
 	}
 
+	pm_runtime_get_sync(d40c->base->dev);
 	res = d40_channel_execute_command(d40c, D40_DMA_SUSPEND_REQ);
 	if (res) {
 		chan_err(d40c, "suspend failed\n");
-		return res;
+		goto out;
 	}
 
 	if (chan_is_logical(d40c)) {
@@ -1664,13 +1837,11 @@ static int d40_free_dma(struct d40_chan *d40c)
 			if (d40_chan_has_events(d40c)) {
 				res = d40_channel_execute_command(d40c,
 								  D40_DMA_RUN);
-				if (res) {
+				if (res)
 					chan_err(d40c,
 						"Executing RUN command\n");
-					return res;
-				}
 			}
-			return 0;
+			goto out;
 		}
 	} else {
 		(void) d40_alloc_mask_free(phy, is_src, 0);
@@ -1680,13 +1851,23 @@ static int d40_free_dma(struct d40_chan *d40c)
 	res = d40_channel_execute_command(d40c, D40_DMA_STOP);
 	if (res) {
 		chan_err(d40c, "Failed to stop channel\n");
-		return res;
+		goto out;
 	}
+
+	if (d40c->busy) {
+		pm_runtime_mark_last_busy(d40c->base->dev);
+		pm_runtime_put_autosuspend(d40c->base->dev);
+	}
+
+	d40c->busy = false;
 	d40c->phy_chan = NULL;
 	d40c->configured = false;
 	d40c->base->lookup_phy_chans[phy->num] = NULL;
+out:
 
-	return 0;
+	pm_runtime_mark_last_busy(d40c->base->dev);
+	pm_runtime_put_autosuspend(d40c->base->dev);
+	return res;
 }
 
 static bool d40_is_paused(struct d40_chan *d40c)
@@ -1855,7 +2036,7 @@ err:
 }
 
 static dma_addr_t
-d40_get_dev_addr(struct d40_chan *chan, enum dma_data_direction direction)
+d40_get_dev_addr(struct d40_chan *chan, enum dma_transfer_direction direction)
 {
 	struct stedma40_platform_data *plat = chan->base->plat_data;
 	struct stedma40_chan_cfg *cfg = &chan->dma_cfg;
@@ -1864,9 +2045,9 @@ d40_get_dev_addr(struct d40_chan *chan, enum dma_data_direction direction)
 	if (chan->runtime_addr)
 		return chan->runtime_addr;
 
-	if (direction == DMA_FROM_DEVICE)
+	if (direction == DMA_DEV_TO_MEM)
 		addr = plat->dev_rx[cfg->src_dev_type];
-	else if (direction == DMA_TO_DEVICE)
+	else if (direction == DMA_MEM_TO_DEV)
 		addr = plat->dev_tx[cfg->dst_dev_type];
 
 	return addr;
@@ -1875,7 +2056,7 @@ d40_get_dev_addr(struct d40_chan *chan, enum dma_data_direction direction)
 static struct dma_async_tx_descriptor *
 d40_prep_sg(struct dma_chan *dchan, struct scatterlist *sg_src,
 	    struct scatterlist *sg_dst, unsigned int sg_len,
-	    enum dma_data_direction direction, unsigned long dma_flags)
+	    enum dma_transfer_direction direction, unsigned long dma_flags)
 {
 	struct d40_chan *chan = container_of(dchan, struct d40_chan, chan);
 	dma_addr_t src_dev_addr = 0;
@@ -1902,9 +2083,9 @@ d40_prep_sg(struct dma_chan *dchan, struct scatterlist *sg_src,
 	if (direction != DMA_NONE) {
 		dma_addr_t dev_addr = d40_get_dev_addr(chan, direction);
 
-		if (direction == DMA_FROM_DEVICE)
+		if (direction == DMA_DEV_TO_MEM)
 			src_dev_addr = dev_addr;
-		else if (direction == DMA_TO_DEVICE)
+		else if (direction == DMA_MEM_TO_DEV)
 			dst_dev_addr = dev_addr;
 	}
 
@@ -2011,14 +2192,15 @@ static int d40_alloc_chan_resources(struct dma_chan *chan)
 			goto fail;
 		}
 	}
-	is_free_phy = (d40c->phy_chan == NULL);
 
-	err = d40_allocate_channel(d40c);
+	err = d40_allocate_channel(d40c, &is_free_phy);
 	if (err) {
 		chan_err(d40c, "Failed to allocate channel\n");
+		d40c->configured = false;
 		goto fail;
 	}
 
+	pm_runtime_get_sync(d40c->base->dev);
 	/* Fill in basic CFG register values */
 	d40_phy_cfg(&d40c->dma_cfg, &d40c->src_def_cfg,
 		    &d40c->dst_def_cfg, chan_is_logical(d40c));
@@ -2038,6 +2220,12 @@ static int d40_alloc_chan_resources(struct dma_chan *chan)
 			  D40_LCPA_CHAN_SIZE + D40_LCPA_CHAN_DST_DELTA;
 	}
 
+	dev_dbg(chan2dev(d40c), "allocated %s channel (phy %d%s)\n",
+		 chan_is_logical(d40c) ? "logical" : "physical",
+		 d40c->phy_chan->num,
+		 d40c->dma_cfg.use_fixed_channel ? ", fixed" : "");
+
+
 	/*
 	 * Only write channel configuration to the DMA if the physical
 	 * resource is free. In case of multiple logical channels
@@ -2046,6 +2234,8 @@ static int d40_alloc_chan_resources(struct dma_chan *chan)
 	if (is_free_phy)
 		d40_config_write(d40c);
 fail:
+	pm_runtime_mark_last_busy(d40c->base->dev);
+	pm_runtime_put_autosuspend(d40c->base->dev);
 	spin_unlock_irqrestore(&d40c->lock, flags);
 	return err;
 }
@@ -2108,10 +2298,10 @@ d40_prep_memcpy_sg(struct dma_chan *chan,
 static struct dma_async_tx_descriptor *d40_prep_slave_sg(struct dma_chan *chan,
 							 struct scatterlist *sgl,
 							 unsigned int sg_len,
-							 enum dma_data_direction direction,
+							 enum dma_transfer_direction direction,
 							 unsigned long dma_flags)
 {
-	if (direction != DMA_FROM_DEVICE && direction != DMA_TO_DEVICE)
+	if (direction != DMA_DEV_TO_MEM && direction != DMA_MEM_TO_DEV)
 		return NULL;
 
 	return d40_prep_sg(chan, sgl, sgl, sg_len, direction, dma_flags);
@@ -2120,7 +2310,7 @@ static struct dma_async_tx_descriptor *d40_prep_slave_sg(struct dma_chan *chan,
 static struct dma_async_tx_descriptor *
 dma40_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t dma_addr,
 		     size_t buf_len, size_t period_len,
-		     enum dma_data_direction direction)
+		     enum dma_transfer_direction direction)
 {
 	unsigned int periods = buf_len / period_len;
 	struct dma_async_tx_descriptor *txd;
@@ -2269,7 +2459,7 @@ static int d40_set_runtime_config(struct dma_chan *chan,
 	dst_addr_width = config->dst_addr_width;
 	dst_maxburst = config->dst_maxburst;
 
-	if (config->direction == DMA_FROM_DEVICE) {
+	if (config->direction == DMA_DEV_TO_MEM) {
 		dma_addr_t dev_addr_rx =
 			d40c->base->plat_data->dev_rx[cfg->src_dev_type];
 
@@ -2292,7 +2482,7 @@ static int d40_set_runtime_config(struct dma_chan *chan,
 		if (dst_maxburst == 0)
 			dst_maxburst = src_maxburst;
 
-	} else if (config->direction == DMA_TO_DEVICE) {
+	} else if (config->direction == DMA_MEM_TO_DEV) {
 		dma_addr_t dev_addr_tx =
 			d40c->base->plat_data->dev_tx[cfg->dst_dev_type];
 
@@ -2357,7 +2547,7 @@ static int d40_set_runtime_config(struct dma_chan *chan,
 		"configured channel %s for %s, data width %d/%d, "
 		"maxburst %d/%d elements, LE, no flow control\n",
 		dma_chan_name(chan),
-		(config->direction == DMA_FROM_DEVICE) ? "RX" : "TX",
+		(config->direction == DMA_DEV_TO_MEM) ? "RX" : "TX",
 		src_addr_width, dst_addr_width,
 		src_maxburst, dst_maxburst);
 
@@ -2519,6 +2709,72 @@ failure1:
 	return err;
 }
 
+/* Suspend resume functionality */
+#ifdef CONFIG_PM
+static int dma40_pm_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct d40_base *base = platform_get_drvdata(pdev);
+	int ret = 0;
+	if (!pm_runtime_suspended(dev))
+		return -EBUSY;
+
+	if (base->lcpa_regulator)
+		ret = regulator_disable(base->lcpa_regulator);
+	return ret;
+}
+
+static int dma40_runtime_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct d40_base *base = platform_get_drvdata(pdev);
+
+	d40_save_restore_registers(base, true);
+
+	/* Don't disable/enable clocks for v1 due to HW bugs */
+	if (base->rev != 1)
+		writel_relaxed(base->gcc_pwr_off_mask,
+			       base->virtbase + D40_DREG_GCC);
+
+	return 0;
+}
+
+static int dma40_runtime_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct d40_base *base = platform_get_drvdata(pdev);
+
+	if (base->initialized)
+		d40_save_restore_registers(base, false);
+
+	writel_relaxed(D40_DREG_GCC_ENABLE_ALL,
+		       base->virtbase + D40_DREG_GCC);
+	return 0;
+}
+
+static int dma40_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct d40_base *base = platform_get_drvdata(pdev);
+	int ret = 0;
+
+	if (base->lcpa_regulator)
+		ret = regulator_enable(base->lcpa_regulator);
+
+	return ret;
+}
+
+static const struct dev_pm_ops dma40_pm_ops = {
+	.suspend		= dma40_pm_suspend,
+	.runtime_suspend	= dma40_runtime_suspend,
+	.runtime_resume		= dma40_runtime_resume,
+	.resume			= dma40_resume,
+};
+#define DMA40_PM_OPS	(&dma40_pm_ops)
+#else
+#define DMA40_PM_OPS	NULL
+#endif
+
 /* Initialization functions. */
 
 static int __init d40_phy_res_init(struct d40_base *base)
@@ -2527,6 +2783,7 @@ static int __init d40_phy_res_init(struct d40_base *base)
 	int num_phy_chans_avail = 0;
 	u32 val[2];
 	int odd_even_bit = -2;
+	int gcc = D40_DREG_GCC_ENA;
 
 	val[0] = readl(base->virtbase + D40_DREG_PRSME);
 	val[1] = readl(base->virtbase + D40_DREG_PRSMO);
@@ -2538,9 +2795,17 @@ static int __init d40_phy_res_init(struct d40_base *base)
 			/* Mark security only channels as occupied */
 			base->phy_res[i].allocated_src = D40_ALLOC_PHY;
 			base->phy_res[i].allocated_dst = D40_ALLOC_PHY;
+			base->phy_res[i].reserved = true;
+			gcc |= D40_DREG_GCC_EVTGRP_ENA(D40_PHYS_TO_GROUP(i),
+						       D40_DREG_GCC_SRC);
+			gcc |= D40_DREG_GCC_EVTGRP_ENA(D40_PHYS_TO_GROUP(i),
+						       D40_DREG_GCC_DST);
+
+
 		} else {
 			base->phy_res[i].allocated_src = D40_ALLOC_FREE;
 			base->phy_res[i].allocated_dst = D40_ALLOC_FREE;
+			base->phy_res[i].reserved = false;
 			num_phy_chans_avail++;
 		}
 		spin_lock_init(&base->phy_res[i].lock);
@@ -2552,6 +2817,11 @@ static int __init d40_phy_res_init(struct d40_base *base)
 
 		base->phy_res[chan].allocated_src = D40_ALLOC_PHY;
 		base->phy_res[chan].allocated_dst = D40_ALLOC_PHY;
+		base->phy_res[chan].reserved = true;
+		gcc |= D40_DREG_GCC_EVTGRP_ENA(D40_PHYS_TO_GROUP(chan),
+					       D40_DREG_GCC_SRC);
+		gcc |= D40_DREG_GCC_EVTGRP_ENA(D40_PHYS_TO_GROUP(chan),
+					       D40_DREG_GCC_DST);
 		num_phy_chans_avail--;
 	}
 
@@ -2572,6 +2842,15 @@ static int __init d40_phy_res_init(struct d40_base *base)
 		val[0] = val[0] >> 2;
 	}
 
+	/*
+	 * To keep things simple, Enable all clocks initially.
+	 * The clocks will get managed later post channel allocation.
+	 * The clocks for the event lines on which reserved channels exists
+	 * are not managed here.
+	 */
+	writel(D40_DREG_GCC_ENABLE_ALL, base->virtbase + D40_DREG_GCC);
+	base->gcc_pwr_off_mask = gcc;
+
 	return num_phy_chans_avail;
 }
 
@@ -2699,10 +2978,15 @@ static struct d40_base * __init d40_hw_detect_init(struct platform_device *pdev)
 			goto failure;
 	}
 
-	base->lcla_pool.alloc_map = kzalloc(num_phy_chans *
-					    sizeof(struct d40_desc *) *
-					    D40_LCLA_LINK_PER_EVENT_GRP,
+	base->reg_val_backup_chan = kmalloc(base->num_phy_chans *
+					    sizeof(d40_backup_regs_chan),
 					    GFP_KERNEL);
+	if (!base->reg_val_backup_chan)
+		goto failure;
+
+	base->lcla_pool.alloc_map =
+		kzalloc(num_phy_chans * sizeof(struct d40_desc *)
+			* D40_LCLA_LINK_PER_EVENT_GRP, GFP_KERNEL);
 	if (!base->lcla_pool.alloc_map)
 		goto failure;
 
@@ -2741,9 +3025,9 @@ failure:
 static void __init d40_hw_init(struct d40_base *base)
 {
 
-	static const struct d40_reg_val dma_init_reg[] = {
+	static struct d40_reg_val dma_init_reg[] = {
 		/* Clock every part of the DMA block from start */
-		{ .reg = D40_DREG_GCC,    .val = 0x0000ff01},
+		{ .reg = D40_DREG_GCC,    .val = D40_DREG_GCC_ENABLE_ALL},
 
 		/* Interrupts on all logical channels */
 		{ .reg = D40_DREG_LCMIS0, .val = 0xFFFFFFFF},
@@ -2943,11 +3227,31 @@ static int __init d40_probe(struct platform_device *pdev)
 		d40_err(&pdev->dev, "Failed to ioremap LCPA region\n");
 		goto failure;
 	}
+	/* If lcla has to be located in ESRAM we don't need to allocate */
+	if (base->plat_data->use_esram_lcla) {
+		res = platform_get_resource_byname(pdev, IORESOURCE_MEM,
+							"lcla_esram");
+		if (!res) {
+			ret = -ENOENT;
+			d40_err(&pdev->dev,
+				"No \"lcla_esram\" memory resource\n");
+			goto failure;
+		}
+		base->lcla_pool.base = ioremap(res->start,
+						resource_size(res));
+		if (!base->lcla_pool.base) {
+			ret = -ENOMEM;
+			d40_err(&pdev->dev, "Failed to ioremap LCLA region\n");
+			goto failure;
+		}
+		writel(res->start, base->virtbase + D40_DREG_LCLA);
 
-	ret = d40_lcla_allocate(base);
-	if (ret) {
-		d40_err(&pdev->dev, "Failed to allocate LCLA area\n");
-		goto failure;
+	} else {
+		ret = d40_lcla_allocate(base);
+		if (ret) {
+			d40_err(&pdev->dev, "Failed to allocate LCLA area\n");
+			goto failure;
+		}
 	}
 
 	spin_lock_init(&base->lcla_pool.lock);
@@ -2960,6 +3264,32 @@ static int __init d40_probe(struct platform_device *pdev)
 		goto failure;
 	}
 
+	pm_runtime_irq_safe(base->dev);
+	pm_runtime_set_autosuspend_delay(base->dev, DMA40_AUTOSUSPEND_DELAY);
+	pm_runtime_use_autosuspend(base->dev);
+	pm_runtime_enable(base->dev);
+	pm_runtime_resume(base->dev);
+
+	if (base->plat_data->use_esram_lcla) {
+
+		base->lcpa_regulator = regulator_get(base->dev, "lcla_esram");
+		if (IS_ERR(base->lcpa_regulator)) {
+			d40_err(&pdev->dev, "Failed to get lcpa_regulator\n");
+			base->lcpa_regulator = NULL;
+			goto failure;
+		}
+
+		ret = regulator_enable(base->lcpa_regulator);
+		if (ret) {
+			d40_err(&pdev->dev,
+				"Failed to enable lcpa_regulator\n");
+			regulator_put(base->lcpa_regulator);
+			base->lcpa_regulator = NULL;
+			goto failure;
+		}
+	}
+
+	base->initialized = true;
 	err = d40_dmaengine_init(base, num_reserved_chans);
 	if (err)
 		goto failure;
@@ -2976,6 +3306,11 @@ failure:
 		if (base->virtbase)
 			iounmap(base->virtbase);
 
+		if (base->lcla_pool.base && base->plat_data->use_esram_lcla) {
+			iounmap(base->lcla_pool.base);
+			base->lcla_pool.base = NULL;
+		}
+
 		if (base->lcla_pool.dma_addr)
 			dma_unmap_single(base->dev, base->lcla_pool.dma_addr,
 					 SZ_1K * base->num_phy_chans,
@@ -2998,6 +3333,11 @@ failure:
 			clk_put(base->clk);
 		}
 
+		if (base->lcpa_regulator) {
+			regulator_disable(base->lcpa_regulator);
+			regulator_put(base->lcpa_regulator);
+		}
+
 		kfree(base->lcla_pool.alloc_map);
 		kfree(base->lookup_log_chans);
 		kfree(base->lookup_phy_chans);
@@ -3013,6 +3353,7 @@ static struct platform_driver d40_driver = {
 	.driver = {
 		.owner = THIS_MODULE,
 		.name  = D40_NAME,
+		.pm = DMA40_PM_OPS,
 	},
 };
 
diff --git a/drivers/dma/ste_dma40_ll.h b/drivers/dma/ste_dma40_ll.h
index b44c455158d..8d3d490968a 100644
--- a/drivers/dma/ste_dma40_ll.h
+++ b/drivers/dma/ste_dma40_ll.h
@@ -16,6 +16,8 @@
 
 #define D40_TYPE_TO_GROUP(type) (type / 16)
 #define D40_TYPE_TO_EVENT(type) (type % 16)
+#define D40_GROUP_SIZE 8
+#define D40_PHYS_TO_GROUP(phys) ((phys & (D40_GROUP_SIZE - 1)) / 2)
 
 /* Most bits of the CFG register are the same in log as in phy mode */
 #define D40_SREG_CFG_MST_POS		15
@@ -123,6 +125,15 @@
 
 /* DMA Register Offsets */
 #define D40_DREG_GCC		0x000
+#define D40_DREG_GCC_ENA	0x1
+/* This assumes that there are only 4 event groups */
+#define D40_DREG_GCC_ENABLE_ALL	0xff01
+#define D40_DREG_GCC_EVTGRP_POS 8
+#define D40_DREG_GCC_SRC 0
+#define D40_DREG_GCC_DST 1
+#define D40_DREG_GCC_EVTGRP_ENA(x, y) \
+	(1 << (D40_DREG_GCC_EVTGRP_POS + 2 * x + y))
+
 #define D40_DREG_PRTYP		0x004
 #define D40_DREG_PRSME		0x008
 #define D40_DREG_PRSMO		0x00C
diff --git a/drivers/dma/timb_dma.c b/drivers/dma/timb_dma.c
index a4a398f2ef6..a6f9c1684a0 100644
--- a/drivers/dma/timb_dma.c
+++ b/drivers/dma/timb_dma.c
@@ -90,7 +90,7 @@ struct timb_dma_chan {
 	struct list_head	queue;
 	struct list_head	free_list;
 	unsigned int		bytes_per_line;
-	enum dma_data_direction	direction;
+	enum dma_transfer_direction	direction;
 	unsigned int		descs; /* Descriptors to allocate */
 	unsigned int		desc_elems; /* number of elems per descriptor */
 };
@@ -166,10 +166,10 @@ static void __td_unmap_desc(struct timb_dma_chan *td_chan, const u8 *dma_desc,
 
 	if (single)
 		dma_unmap_single(chan2dev(&td_chan->chan), addr, len,
-			td_chan->direction);
+			DMA_TO_DEVICE);
 	else
 		dma_unmap_page(chan2dev(&td_chan->chan), addr, len,
-			td_chan->direction);
+			DMA_TO_DEVICE);
 }
 
 static void __td_unmap_descs(struct timb_dma_desc *td_desc, bool single)
@@ -235,7 +235,7 @@ static void __td_start_dma(struct timb_dma_chan *td_chan)
 		"td_chan: %p, chan: %d, membase: %p\n",
 		td_chan, td_chan->chan.chan_id, td_chan->membase);
 
-	if (td_chan->direction == DMA_FROM_DEVICE) {
+	if (td_chan->direction == DMA_DEV_TO_MEM) {
 
 		/* descriptor address */
 		iowrite32(0, td_chan->membase + TIMBDMA_OFFS_RX_DHAR);
@@ -278,7 +278,7 @@ static void __td_finish(struct timb_dma_chan *td_chan)
 		txd->cookie);
 
 	/* make sure to stop the transfer */
-	if (td_chan->direction == DMA_FROM_DEVICE)
+	if (td_chan->direction == DMA_DEV_TO_MEM)
 		iowrite32(0, td_chan->membase + TIMBDMA_OFFS_RX_ER);
 /* Currently no support for stopping DMA transfers
 	else
@@ -558,7 +558,7 @@ static void td_issue_pending(struct dma_chan *chan)
 
 static struct dma_async_tx_descriptor *td_prep_slave_sg(struct dma_chan *chan,
 	struct scatterlist *sgl, unsigned int sg_len,
-	enum dma_data_direction direction, unsigned long flags)
+	enum dma_transfer_direction direction, unsigned long flags)
 {
 	struct timb_dma_chan *td_chan =
 		container_of(chan, struct timb_dma_chan, chan);
@@ -606,7 +606,7 @@ static struct dma_async_tx_descriptor *td_prep_slave_sg(struct dma_chan *chan,
 	}
 
 	dma_sync_single_for_device(chan2dmadev(chan), td_desc->txd.phys,
-		td_desc->desc_list_len, DMA_TO_DEVICE);
+		td_desc->desc_list_len, DMA_MEM_TO_DEV);
 
 	return &td_desc->txd;
 }
@@ -775,8 +775,8 @@ static int __devinit td_probe(struct platform_device *pdev)
 		td_chan->descs = pchan->descriptors;
 		td_chan->desc_elems = pchan->descriptor_elements;
 		td_chan->bytes_per_line = pchan->bytes_per_line;
-		td_chan->direction = pchan->rx ? DMA_FROM_DEVICE :
-			DMA_TO_DEVICE;
+		td_chan->direction = pchan->rx ? DMA_DEV_TO_MEM :
+			DMA_MEM_TO_DEV;
 
 		td_chan->membase = td->membase +
 			(i / 2) * TIMBDMA_INSTANCE_OFFSET +
@@ -841,17 +841,7 @@ static struct platform_driver td_driver = {
 	.remove	= __exit_p(td_remove),
 };
 
-static int __init td_init(void)
-{
-	return platform_driver_register(&td_driver);
-}
-module_init(td_init);
-
-static void __exit td_exit(void)
-{
-	platform_driver_unregister(&td_driver);
-}
-module_exit(td_exit);
+module_platform_driver(td_driver);
 
 MODULE_LICENSE("GPL v2");
 MODULE_DESCRIPTION("Timberdale DMA controller driver");
diff --git a/drivers/dma/txx9dmac.c b/drivers/dma/txx9dmac.c
index cbd83e362b5..6122c364cf1 100644
--- a/drivers/dma/txx9dmac.c
+++ b/drivers/dma/txx9dmac.c
@@ -845,7 +845,7 @@ txx9dmac_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 
 static struct dma_async_tx_descriptor *
 txx9dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
-		unsigned int sg_len, enum dma_data_direction direction,
+		unsigned int sg_len, enum dma_transfer_direction direction,
 		unsigned long flags)
 {
 	struct txx9dmac_chan *dc = to_txx9dmac_chan(chan);
@@ -860,9 +860,9 @@ txx9dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 
 	BUG_ON(!ds || !ds->reg_width);
 	if (ds->tx_reg)
-		BUG_ON(direction != DMA_TO_DEVICE);
+		BUG_ON(direction != DMA_MEM_TO_DEV);
 	else
-		BUG_ON(direction != DMA_FROM_DEVICE);
+		BUG_ON(direction != DMA_DEV_TO_MEM);
 	if (unlikely(!sg_len))
 		return NULL;
 
@@ -882,7 +882,7 @@ txx9dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		mem = sg_dma_address(sg);
 
 		if (__is_dmac64(ddev)) {
-			if (direction == DMA_TO_DEVICE) {
+			if (direction == DMA_MEM_TO_DEV) {
 				desc->hwdesc.SAR = mem;
 				desc->hwdesc.DAR = ds->tx_reg;
 			} else {
@@ -891,7 +891,7 @@ txx9dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 			}
 			desc->hwdesc.CNTR = sg_dma_len(sg);
 		} else {
-			if (direction == DMA_TO_DEVICE) {
+			if (direction == DMA_MEM_TO_DEV) {
 				desc->hwdesc32.SAR = mem;
 				desc->hwdesc32.DAR = ds->tx_reg;
 			} else {
@@ -900,7 +900,7 @@ txx9dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 			}
 			desc->hwdesc32.CNTR = sg_dma_len(sg);
 		}
-		if (direction == DMA_TO_DEVICE) {
+		if (direction == DMA_MEM_TO_DEV) {
 			sai = ds->reg_width;
 			dai = 0;
 		} else {
diff --git a/drivers/media/video/mx3_camera.c b/drivers/media/video/mx3_camera.c
index 0cb461dd396..74522773e93 100644
--- a/drivers/media/video/mx3_camera.c
+++ b/drivers/media/video/mx3_camera.c
@@ -287,7 +287,7 @@ static void mx3_videobuf_queue(struct vb2_buffer *vb)
 		sg_dma_len(sg)		= new_size;
 
 		txd = ichan->dma_chan.device->device_prep_slave_sg(
-			&ichan->dma_chan, sg, 1, DMA_FROM_DEVICE,
+			&ichan->dma_chan, sg, 1, DMA_DEV_TO_MEM,
 			DMA_PREP_INTERRUPT);
 		if (!txd)
 			goto error;
diff --git a/drivers/media/video/timblogiw.c b/drivers/media/video/timblogiw.c
index 0a2d75f0406..4ed1c7c28ae 100644
--- a/drivers/media/video/timblogiw.c
+++ b/drivers/media/video/timblogiw.c
@@ -565,7 +565,7 @@ static void buffer_queue(struct videobuf_queue *vq, struct videobuf_buffer *vb)
 	spin_unlock_irq(&fh->queue_lock);
 
 	desc = fh->chan->device->device_prep_slave_sg(fh->chan,
-		buf->sg, sg_elems, DMA_FROM_DEVICE,
+		buf->sg, sg_elems, DMA_DEV_TO_MEM,
 		DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_SRC_UNMAP);
 	if (!desc) {
 		spin_lock_irq(&fh->queue_lock);
diff --git a/drivers/misc/carma/carma-fpga-program.c b/drivers/misc/carma/carma-fpga-program.c
index eb5cd28bc6d..a2d25e4857e 100644
--- a/drivers/misc/carma/carma-fpga-program.c
+++ b/drivers/misc/carma/carma-fpga-program.c
@@ -513,7 +513,7 @@ static noinline int fpga_program_dma(struct fpga_dev *priv)
 	 * transaction, and then put it under external control
 	 */
 	memset(&config, 0, sizeof(config));
-	config.direction = DMA_TO_DEVICE;
+	config.direction = DMA_MEM_TO_DEV;
 	config.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
 	config.dst_maxburst = fpga_fifo_size(priv->regs) / 2 / 4;
 	ret = chan->device->device_control(chan, DMA_SLAVE_CONFIG,
diff --git a/drivers/mmc/host/atmel-mci.c b/drivers/mmc/host/atmel-mci.c
index a7ee5027146..fcfe1eb5acc 100644
--- a/drivers/mmc/host/atmel-mci.c
+++ b/drivers/mmc/host/atmel-mci.c
@@ -823,6 +823,7 @@ atmci_prepare_data_dma(struct atmel_mci *host, struct mmc_data *data)
 	struct scatterlist		*sg;
 	unsigned int			i;
 	enum dma_data_direction		direction;
+	enum dma_transfer_direction	slave_dirn;
 	unsigned int			sglen;
 	u32 iflags;
 
@@ -860,16 +861,19 @@ atmci_prepare_data_dma(struct atmel_mci *host, struct mmc_data *data)
 	if (host->caps.has_dma)
 		atmci_writel(host, ATMCI_DMA, ATMCI_DMA_CHKSIZE(3) | ATMCI_DMAEN);
 
-	if (data->flags & MMC_DATA_READ)
+	if (data->flags & MMC_DATA_READ) {
 		direction = DMA_FROM_DEVICE;
-	else
+		slave_dirn = DMA_DEV_TO_MEM;
+	} else {
 		direction = DMA_TO_DEVICE;
+		slave_dirn = DMA_MEM_TO_DEV;
+	}
 
 	sglen = dma_map_sg(chan->device->dev, data->sg,
 			data->sg_len, direction);
 
 	desc = chan->device->device_prep_slave_sg(chan,
-			data->sg, sglen, direction,
+			data->sg, sglen, slave_dirn,
 			DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc)
 		goto unmap_exit;
diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c
index ece03b491c7..0d955ffaf44 100644
--- a/drivers/mmc/host/mmci.c
+++ b/drivers/mmc/host/mmci.c
@@ -374,6 +374,7 @@ static int mmci_dma_prep_data(struct mmci_host *host, struct mmc_data *data,
 	struct dma_chan *chan;
 	struct dma_device *device;
 	struct dma_async_tx_descriptor *desc;
+	enum dma_data_direction buffer_dirn;
 	int nr_sg;
 
 	/* Check if next job is already prepared */
@@ -387,10 +388,12 @@ static int mmci_dma_prep_data(struct mmci_host *host, struct mmc_data *data,
 	}
 
 	if (data->flags & MMC_DATA_READ) {
-		conf.direction = DMA_FROM_DEVICE;
+		conf.direction = DMA_DEV_TO_MEM;
+		buffer_dirn = DMA_FROM_DEVICE;
 		chan = host->dma_rx_channel;
 	} else {
-		conf.direction = DMA_TO_DEVICE;
+		conf.direction = DMA_MEM_TO_DEV;
+		buffer_dirn = DMA_TO_DEVICE;
 		chan = host->dma_tx_channel;
 	}
 
@@ -403,7 +406,7 @@ static int mmci_dma_prep_data(struct mmci_host *host, struct mmc_data *data,
 		return -EINVAL;
 
 	device = chan->device;
-	nr_sg = dma_map_sg(device->dev, data->sg, data->sg_len, conf.direction);
+	nr_sg = dma_map_sg(device->dev, data->sg, data->sg_len, buffer_dirn);
 	if (nr_sg == 0)
 		return -EINVAL;
 
@@ -426,7 +429,7 @@ static int mmci_dma_prep_data(struct mmci_host *host, struct mmc_data *data,
  unmap_exit:
 	if (!next)
 		dmaengine_terminate_all(chan);
-	dma_unmap_sg(device->dev, data->sg, data->sg_len, conf.direction);
+	dma_unmap_sg(device->dev, data->sg, data->sg_len, buffer_dirn);
 	return -ENOMEM;
 }
 
diff --git a/drivers/mmc/host/mxcmmc.c b/drivers/mmc/host/mxcmmc.c
index 7088b40f957..4184b7946bb 100644
--- a/drivers/mmc/host/mxcmmc.c
+++ b/drivers/mmc/host/mxcmmc.c
@@ -218,6 +218,7 @@ static int mxcmci_setup_data(struct mxcmci_host *host, struct mmc_data *data)
 	unsigned int blksz = data->blksz;
 	unsigned int datasize = nob * blksz;
 	struct scatterlist *sg;
+	enum dma_transfer_direction slave_dirn;
 	int i, nents;
 
 	if (data->flags & MMC_DATA_STREAM)
@@ -240,10 +241,13 @@ static int mxcmci_setup_data(struct mxcmci_host *host, struct mmc_data *data)
 		}
 	}
 
-	if (data->flags & MMC_DATA_READ)
+	if (data->flags & MMC_DATA_READ) {
 		host->dma_dir = DMA_FROM_DEVICE;
-	else
+		slave_dirn = DMA_DEV_TO_MEM;
+	} else {
 		host->dma_dir = DMA_TO_DEVICE;
+		slave_dirn = DMA_MEM_TO_DEV;
+	}
 
 	nents = dma_map_sg(host->dma->device->dev, data->sg,
 				     data->sg_len,  host->dma_dir);
@@ -251,7 +255,7 @@ static int mxcmci_setup_data(struct mxcmci_host *host, struct mmc_data *data)
 		return -EINVAL;
 
 	host->desc = host->dma->device->device_prep_slave_sg(host->dma,
-		data->sg, data->sg_len, host->dma_dir,
+		data->sg, data->sg_len, slave_dirn,
 		DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 
 	if (!host->desc) {
diff --git a/drivers/mmc/host/mxs-mmc.c b/drivers/mmc/host/mxs-mmc.c
index 4e2e019dd5c..382c835d217 100644
--- a/drivers/mmc/host/mxs-mmc.c
+++ b/drivers/mmc/host/mxs-mmc.c
@@ -154,6 +154,7 @@ struct mxs_mmc_host {
 	struct dma_chan         	*dmach;
 	struct mxs_dma_data		dma_data;
 	unsigned int			dma_dir;
+	enum dma_transfer_direction	slave_dirn;
 	u32				ssp_pio_words[SSP_PIO_NUM];
 
 	unsigned int			version;
@@ -324,7 +325,7 @@ static struct dma_async_tx_descriptor *mxs_mmc_prep_dma(
 	}
 
 	desc = host->dmach->device->device_prep_slave_sg(host->dmach,
-				sgl, sg_len, host->dma_dir, append);
+				sgl, sg_len, host->slave_dirn, append);
 	if (desc) {
 		desc->callback = mxs_mmc_dma_irq_callback;
 		desc->callback_param = host;
@@ -356,6 +357,7 @@ static void mxs_mmc_bc(struct mxs_mmc_host *host)
 	host->ssp_pio_words[1] = cmd0;
 	host->ssp_pio_words[2] = cmd1;
 	host->dma_dir = DMA_NONE;
+	host->slave_dirn = DMA_TRANS_NONE;
 	desc = mxs_mmc_prep_dma(host, 0);
 	if (!desc)
 		goto out;
@@ -395,6 +397,7 @@ static void mxs_mmc_ac(struct mxs_mmc_host *host)
 	host->ssp_pio_words[1] = cmd0;
 	host->ssp_pio_words[2] = cmd1;
 	host->dma_dir = DMA_NONE;
+	host->slave_dirn = DMA_TRANS_NONE;
 	desc = mxs_mmc_prep_dma(host, 0);
 	if (!desc)
 		goto out;
@@ -433,6 +436,7 @@ static void mxs_mmc_adtc(struct mxs_mmc_host *host)
 	int i;
 
 	unsigned short dma_data_dir, timeout;
+	enum dma_transfer_direction slave_dirn;
 	unsigned int data_size = 0, log2_blksz;
 	unsigned int blocks = data->blocks;
 
@@ -448,9 +452,11 @@ static void mxs_mmc_adtc(struct mxs_mmc_host *host)
 
 	if (data->flags & MMC_DATA_WRITE) {
 		dma_data_dir = DMA_TO_DEVICE;
+		slave_dirn = DMA_MEM_TO_DEV;
 		read = 0;
 	} else {
 		dma_data_dir = DMA_FROM_DEVICE;
+		slave_dirn = DMA_DEV_TO_MEM;
 		read = BM_SSP_CTRL0_READ;
 	}
 
@@ -510,6 +516,7 @@ static void mxs_mmc_adtc(struct mxs_mmc_host *host)
 	host->ssp_pio_words[1] = cmd0;
 	host->ssp_pio_words[2] = cmd1;
 	host->dma_dir = DMA_NONE;
+	host->slave_dirn = DMA_TRANS_NONE;
 	desc = mxs_mmc_prep_dma(host, 0);
 	if (!desc)
 		goto out;
@@ -518,6 +525,7 @@ static void mxs_mmc_adtc(struct mxs_mmc_host *host)
 	WARN_ON(host->data != NULL);
 	host->data = data;
 	host->dma_dir = dma_data_dir;
+	host->slave_dirn = slave_dirn;
 	desc = mxs_mmc_prep_dma(host, 1);
 	if (!desc)
 		goto out;
diff --git a/drivers/mmc/host/sh_mmcif.c b/drivers/mmc/host/sh_mmcif.c
index 4a2c5b2355f..f5d8b53be33 100644
--- a/drivers/mmc/host/sh_mmcif.c
+++ b/drivers/mmc/host/sh_mmcif.c
@@ -286,7 +286,7 @@ static void sh_mmcif_start_dma_rx(struct sh_mmcif_host *host)
 	if (ret > 0) {
 		host->dma_active = true;
 		desc = chan->device->device_prep_slave_sg(chan, sg, ret,
-			DMA_FROM_DEVICE, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+			DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	}
 
 	if (desc) {
@@ -335,7 +335,7 @@ static void sh_mmcif_start_dma_tx(struct sh_mmcif_host *host)
 	if (ret > 0) {
 		host->dma_active = true;
 		desc = chan->device->device_prep_slave_sg(chan, sg, ret,
-			DMA_TO_DEVICE, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+			DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	}
 
 	if (desc) {
diff --git a/drivers/mmc/host/tmio_mmc_dma.c b/drivers/mmc/host/tmio_mmc_dma.c
index 86f259cdfcb..7a6e6cc8f8b 100644
--- a/drivers/mmc/host/tmio_mmc_dma.c
+++ b/drivers/mmc/host/tmio_mmc_dma.c
@@ -77,7 +77,7 @@ static void tmio_mmc_start_dma_rx(struct tmio_mmc_host *host)
 	ret = dma_map_sg(chan->device->dev, sg, host->sg_len, DMA_FROM_DEVICE);
 	if (ret > 0)
 		desc = chan->device->device_prep_slave_sg(chan, sg, ret,
-			DMA_FROM_DEVICE, DMA_CTRL_ACK);
+			DMA_DEV_TO_MEM, DMA_CTRL_ACK);
 
 	if (desc) {
 		cookie = dmaengine_submit(desc);
@@ -158,7 +158,7 @@ static void tmio_mmc_start_dma_tx(struct tmio_mmc_host *host)
 	ret = dma_map_sg(chan->device->dev, sg, host->sg_len, DMA_TO_DEVICE);
 	if (ret > 0)
 		desc = chan->device->device_prep_slave_sg(chan, sg, ret,
-			DMA_TO_DEVICE, DMA_CTRL_ACK);
+			DMA_MEM_TO_DEV, DMA_CTRL_ACK);
 
 	if (desc) {
 		cookie = dmaengine_submit(desc);
diff --git a/drivers/mtd/nand/gpmi-nand/gpmi-lib.c b/drivers/mtd/nand/gpmi-nand/gpmi-lib.c
index 2a56fc6f399..7f680420bfa 100644
--- a/drivers/mtd/nand/gpmi-nand/gpmi-lib.c
+++ b/drivers/mtd/nand/gpmi-nand/gpmi-lib.c
@@ -827,7 +827,7 @@ int gpmi_send_command(struct gpmi_nand_data *this)
 	pio[1] = pio[2] = 0;
 	desc = channel->device->device_prep_slave_sg(channel,
 					(struct scatterlist *)pio,
-					ARRAY_SIZE(pio), DMA_NONE, 0);
+					ARRAY_SIZE(pio), DMA_TRANS_NONE, 0);
 	if (!desc) {
 		pr_err("step 1 error\n");
 		return -1;
@@ -839,7 +839,7 @@ int gpmi_send_command(struct gpmi_nand_data *this)
 	sg_init_one(sgl, this->cmd_buffer, this->command_length);
 	dma_map_sg(this->dev, sgl, 1, DMA_TO_DEVICE);
 	desc = channel->device->device_prep_slave_sg(channel,
-					sgl, 1, DMA_TO_DEVICE, 1);
+					sgl, 1, DMA_MEM_TO_DEV, 1);
 	if (!desc) {
 		pr_err("step 2 error\n");
 		return -1;
@@ -872,7 +872,7 @@ int gpmi_send_data(struct gpmi_nand_data *this)
 	pio[1] = 0;
 	desc = channel->device->device_prep_slave_sg(channel,
 					(struct scatterlist *)pio,
-					ARRAY_SIZE(pio), DMA_NONE, 0);
+					ARRAY_SIZE(pio), DMA_TRANS_NONE, 0);
 	if (!desc) {
 		pr_err("step 1 error\n");
 		return -1;
@@ -881,7 +881,7 @@ int gpmi_send_data(struct gpmi_nand_data *this)
 	/* [2] send DMA request */
 	prepare_data_dma(this, DMA_TO_DEVICE);
 	desc = channel->device->device_prep_slave_sg(channel, &this->data_sgl,
-						1, DMA_TO_DEVICE, 1);
+						1, DMA_MEM_TO_DEV, 1);
 	if (!desc) {
 		pr_err("step 2 error\n");
 		return -1;
@@ -908,7 +908,7 @@ int gpmi_read_data(struct gpmi_nand_data *this)
 	pio[1] = 0;
 	desc = channel->device->device_prep_slave_sg(channel,
 					(struct scatterlist *)pio,
-					ARRAY_SIZE(pio), DMA_NONE, 0);
+					ARRAY_SIZE(pio), DMA_TRANS_NONE, 0);
 	if (!desc) {
 		pr_err("step 1 error\n");
 		return -1;
@@ -917,7 +917,7 @@ int gpmi_read_data(struct gpmi_nand_data *this)
 	/* [2] : send DMA request */
 	prepare_data_dma(this, DMA_FROM_DEVICE);
 	desc = channel->device->device_prep_slave_sg(channel, &this->data_sgl,
-						1, DMA_FROM_DEVICE, 1);
+						1, DMA_DEV_TO_MEM, 1);
 	if (!desc) {
 		pr_err("step 2 error\n");
 		return -1;
@@ -964,7 +964,7 @@ int gpmi_send_page(struct gpmi_nand_data *this,
 
 	desc = channel->device->device_prep_slave_sg(channel,
 					(struct scatterlist *)pio,
-					ARRAY_SIZE(pio), DMA_NONE, 0);
+					ARRAY_SIZE(pio), DMA_TRANS_NONE, 0);
 	if (!desc) {
 		pr_err("step 2 error\n");
 		return -1;
@@ -998,7 +998,8 @@ int gpmi_read_page(struct gpmi_nand_data *this,
 		| BF_GPMI_CTRL0_XFER_COUNT(0);
 	pio[1] = 0;
 	desc = channel->device->device_prep_slave_sg(channel,
-				(struct scatterlist *)pio, 2, DMA_NONE, 0);
+				(struct scatterlist *)pio, 2,
+				DMA_TRANS_NONE, 0);
 	if (!desc) {
 		pr_err("step 1 error\n");
 		return -1;
@@ -1027,7 +1028,7 @@ int gpmi_read_page(struct gpmi_nand_data *this,
 	pio[5] = auxiliary;
 	desc = channel->device->device_prep_slave_sg(channel,
 					(struct scatterlist *)pio,
-					ARRAY_SIZE(pio), DMA_NONE, 1);
+					ARRAY_SIZE(pio), DMA_TRANS_NONE, 1);
 	if (!desc) {
 		pr_err("step 2 error\n");
 		return -1;
@@ -1045,7 +1046,8 @@ int gpmi_read_page(struct gpmi_nand_data *this,
 		| BF_GPMI_CTRL0_XFER_COUNT(geo->page_size);
 	pio[1] = 0;
 	desc = channel->device->device_prep_slave_sg(channel,
-				(struct scatterlist *)pio, 2, DMA_NONE, 1);
+				(struct scatterlist *)pio, 2,
+				DMA_TRANS_NONE, 1);
 	if (!desc) {
 		pr_err("step 3 error\n");
 		return -1;
diff --git a/drivers/net/ethernet/micrel/ks8842.c b/drivers/net/ethernet/micrel/ks8842.c
index 75ec87a822b..0a85690a132 100644
--- a/drivers/net/ethernet/micrel/ks8842.c
+++ b/drivers/net/ethernet/micrel/ks8842.c
@@ -459,7 +459,7 @@ static int ks8842_tx_frame_dma(struct sk_buff *skb, struct net_device *netdev)
 		sg_dma_len(&ctl->sg) += 4 - sg_dma_len(&ctl->sg) % 4;
 
 	ctl->adesc = ctl->chan->device->device_prep_slave_sg(ctl->chan,
-		&ctl->sg, 1, DMA_TO_DEVICE,
+		&ctl->sg, 1, DMA_MEM_TO_DEV,
 		DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_SRC_UNMAP);
 	if (!ctl->adesc)
 		return NETDEV_TX_BUSY;
@@ -571,7 +571,7 @@ static int __ks8842_start_new_rx_dma(struct net_device *netdev)
 		sg_dma_len(sg) = DMA_BUFFER_SIZE;
 
 		ctl->adesc = ctl->chan->device->device_prep_slave_sg(ctl->chan,
-			sg, 1, DMA_FROM_DEVICE,
+			sg, 1, DMA_DEV_TO_MEM,
 			DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_SRC_UNMAP);
 
 		if (!ctl->adesc)
diff --git a/drivers/spi/spi-dw-mid.c b/drivers/spi/spi-dw-mid.c
index e743a45ee92..8418eb03665 100644
--- a/drivers/spi/spi-dw-mid.c
+++ b/drivers/spi/spi-dw-mid.c
@@ -131,7 +131,7 @@ static int mid_spi_dma_transfer(struct dw_spi *dws, int cs_change)
 	rxchan = dws->rxchan;
 
 	/* 2. Prepare the TX dma transfer */
-	txconf.direction = DMA_TO_DEVICE;
+	txconf.direction = DMA_MEM_TO_DEV;
 	txconf.dst_addr = dws->dma_addr;
 	txconf.dst_maxburst = LNW_DMA_MSIZE_16;
 	txconf.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
@@ -147,13 +147,13 @@ static int mid_spi_dma_transfer(struct dw_spi *dws, int cs_change)
 	txdesc = txchan->device->device_prep_slave_sg(txchan,
 				&dws->tx_sgl,
 				1,
-				DMA_TO_DEVICE,
+				DMA_MEM_TO_DEV,
 				DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_DEST_UNMAP);
 	txdesc->callback = dw_spi_dma_done;
 	txdesc->callback_param = dws;
 
 	/* 3. Prepare the RX dma transfer */
-	rxconf.direction = DMA_FROM_DEVICE;
+	rxconf.direction = DMA_DEV_TO_MEM;
 	rxconf.src_addr = dws->dma_addr;
 	rxconf.src_maxburst = LNW_DMA_MSIZE_16;
 	rxconf.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
@@ -169,7 +169,7 @@ static int mid_spi_dma_transfer(struct dw_spi *dws, int cs_change)
 	rxdesc = rxchan->device->device_prep_slave_sg(rxchan,
 				&dws->rx_sgl,
 				1,
-				DMA_FROM_DEVICE,
+				DMA_DEV_TO_MEM,
 				DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_DEST_UNMAP);
 	rxdesc->callback = dw_spi_dma_done;
 	rxdesc->callback_param = dws;
diff --git a/drivers/spi/spi-ep93xx.c b/drivers/spi/spi-ep93xx.c
index 0a282e5fcc9..d46e55c720b 100644
--- a/drivers/spi/spi-ep93xx.c
+++ b/drivers/spi/spi-ep93xx.c
@@ -551,6 +551,7 @@ ep93xx_spi_dma_prepare(struct ep93xx_spi *espi, enum dma_data_direction dir)
 	struct dma_async_tx_descriptor *txd;
 	enum dma_slave_buswidth buswidth;
 	struct dma_slave_config conf;
+	enum dma_transfer_direction slave_dirn;
 	struct scatterlist *sg;
 	struct sg_table *sgt;
 	struct dma_chan *chan;
@@ -573,6 +574,7 @@ ep93xx_spi_dma_prepare(struct ep93xx_spi *espi, enum dma_data_direction dir)
 
 		conf.src_addr = espi->sspdr_phys;
 		conf.src_addr_width = buswidth;
+		slave_dirn = DMA_DEV_TO_MEM;
 	} else {
 		chan = espi->dma_tx;
 		buf = t->tx_buf;
@@ -580,6 +582,7 @@ ep93xx_spi_dma_prepare(struct ep93xx_spi *espi, enum dma_data_direction dir)
 
 		conf.dst_addr = espi->sspdr_phys;
 		conf.dst_addr_width = buswidth;
+		slave_dirn = DMA_MEM_TO_DEV;
 	}
 
 	ret = dmaengine_slave_config(chan, &conf);
@@ -631,7 +634,7 @@ ep93xx_spi_dma_prepare(struct ep93xx_spi *espi, enum dma_data_direction dir)
 		return ERR_PTR(-ENOMEM);
 
 	txd = chan->device->device_prep_slave_sg(chan, sgt->sgl, nents,
-						 dir, DMA_CTRL_ACK);
+						 slave_dirn, DMA_CTRL_ACK);
 	if (!txd) {
 		dma_unmap_sg(chan->device->dev, sgt->sgl, sgt->nents, dir);
 		return ERR_PTR(-ENOMEM);
@@ -979,7 +982,7 @@ static int ep93xx_spi_setup_dma(struct ep93xx_spi *espi)
 	dma_cap_set(DMA_SLAVE, mask);
 
 	espi->dma_rx_data.port = EP93XX_DMA_SSP;
-	espi->dma_rx_data.direction = DMA_FROM_DEVICE;
+	espi->dma_rx_data.direction = DMA_DEV_TO_MEM;
 	espi->dma_rx_data.name = "ep93xx-spi-rx";
 
 	espi->dma_rx = dma_request_channel(mask, ep93xx_spi_dma_filter,
@@ -990,7 +993,7 @@ static int ep93xx_spi_setup_dma(struct ep93xx_spi *espi)
 	}
 
 	espi->dma_tx_data.port = EP93XX_DMA_SSP;
-	espi->dma_tx_data.direction = DMA_TO_DEVICE;
+	espi->dma_tx_data.direction = DMA_MEM_TO_DEV;
 	espi->dma_tx_data.name = "ep93xx-spi-tx";
 
 	espi->dma_tx = dma_request_channel(mask, ep93xx_spi_dma_filter,
diff --git a/drivers/spi/spi-pl022.c b/drivers/spi/spi-pl022.c
index f1f5efbc340..2f9cb43a239 100644
--- a/drivers/spi/spi-pl022.c
+++ b/drivers/spi/spi-pl022.c
@@ -900,11 +900,11 @@ static int configure_dma(struct pl022 *pl022)
 {
 	struct dma_slave_config rx_conf = {
 		.src_addr = SSP_DR(pl022->phybase),
-		.direction = DMA_FROM_DEVICE,
+		.direction = DMA_DEV_TO_MEM,
 	};
 	struct dma_slave_config tx_conf = {
 		.dst_addr = SSP_DR(pl022->phybase),
-		.direction = DMA_TO_DEVICE,
+		.direction = DMA_MEM_TO_DEV,
 	};
 	unsigned int pages;
 	int ret;
@@ -1041,7 +1041,7 @@ static int configure_dma(struct pl022 *pl022)
 	rxdesc = rxchan->device->device_prep_slave_sg(rxchan,
 				      pl022->sgt_rx.sgl,
 				      rx_sglen,
-				      DMA_FROM_DEVICE,
+				      DMA_DEV_TO_MEM,
 				      DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!rxdesc)
 		goto err_rxdesc;
@@ -1049,7 +1049,7 @@ static int configure_dma(struct pl022 *pl022)
 	txdesc = txchan->device->device_prep_slave_sg(txchan,
 				      pl022->sgt_tx.sgl,
 				      tx_sglen,
-				      DMA_TO_DEVICE,
+				      DMA_MEM_TO_DEV,
 				      DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!txdesc)
 		goto err_txdesc;
diff --git a/drivers/spi/spi-topcliff-pch.c b/drivers/spi/spi-topcliff-pch.c
index 7086583b910..2a6429d8c36 100644
--- a/drivers/spi/spi-topcliff-pch.c
+++ b/drivers/spi/spi-topcliff-pch.c
@@ -1079,7 +1079,7 @@ static void pch_spi_handle_dma(struct pch_spi_data *data, int *bpw)
 	}
 	sg = dma->sg_rx_p;
 	desc_rx = dma->chan_rx->device->device_prep_slave_sg(dma->chan_rx, sg,
-					num, DMA_FROM_DEVICE,
+					num, DMA_DEV_TO_MEM,
 					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc_rx) {
 		dev_err(&data->master->dev, "%s:device_prep_slave_sg Failed\n",
@@ -1124,7 +1124,7 @@ static void pch_spi_handle_dma(struct pch_spi_data *data, int *bpw)
 	}
 	sg = dma->sg_tx_p;
 	desc_tx = dma->chan_tx->device->device_prep_slave_sg(dma->chan_tx,
-					sg, num, DMA_TO_DEVICE,
+					sg, num, DMA_MEM_TO_DEV,
 					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc_tx) {
 		dev_err(&data->master->dev, "%s:device_prep_slave_sg Failed\n",
diff --git a/drivers/tty/serial/amba-pl011.c b/drivers/tty/serial/amba-pl011.c
index 6958594f2fc..9ae024025ff 100644
--- a/drivers/tty/serial/amba-pl011.c
+++ b/drivers/tty/serial/amba-pl011.c
@@ -268,7 +268,7 @@ static void pl011_dma_probe_initcall(struct uart_amba_port *uap)
 	struct dma_slave_config tx_conf = {
 		.dst_addr = uap->port.mapbase + UART01x_DR,
 		.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE,
-		.direction = DMA_TO_DEVICE,
+		.direction = DMA_MEM_TO_DEV,
 		.dst_maxburst = uap->fifosize >> 1,
 	};
 	struct dma_chan *chan;
@@ -301,7 +301,7 @@ static void pl011_dma_probe_initcall(struct uart_amba_port *uap)
 		struct dma_slave_config rx_conf = {
 			.src_addr = uap->port.mapbase + UART01x_DR,
 			.src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE,
-			.direction = DMA_FROM_DEVICE,
+			.direction = DMA_DEV_TO_MEM,
 			.src_maxburst = uap->fifosize >> 1,
 		};
 
@@ -480,7 +480,7 @@ static int pl011_dma_tx_refill(struct uart_amba_port *uap)
 		return -EBUSY;
 	}
 
-	desc = dma_dev->device_prep_slave_sg(chan, &dmatx->sg, 1, DMA_TO_DEVICE,
+	desc = dma_dev->device_prep_slave_sg(chan, &dmatx->sg, 1, DMA_MEM_TO_DEV,
 					     DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc) {
 		dma_unmap_sg(dma_dev->dev, &dmatx->sg, 1, DMA_TO_DEVICE);
@@ -676,7 +676,7 @@ static int pl011_dma_rx_trigger_dma(struct uart_amba_port *uap)
 		&uap->dmarx.sgbuf_b : &uap->dmarx.sgbuf_a;
 	dma_dev = rxchan->device;
 	desc = rxchan->device->device_prep_slave_sg(rxchan, &sgbuf->sg, 1,
-					DMA_FROM_DEVICE,
+					DMA_DEV_TO_MEM,
 					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	/*
 	 * If the DMA engine is busy and cannot prepare a
diff --git a/drivers/tty/serial/pch_uart.c b/drivers/tty/serial/pch_uart.c
index de0f613ed6f..17ae65762d1 100644
--- a/drivers/tty/serial/pch_uart.c
+++ b/drivers/tty/serial/pch_uart.c
@@ -764,7 +764,7 @@ static int dma_handle_rx(struct eg20t_port *priv)
 	sg_dma_address(sg) = priv->rx_buf_dma;
 
 	desc = priv->chan_rx->device->device_prep_slave_sg(priv->chan_rx,
-			sg, 1, DMA_FROM_DEVICE,
+			sg, 1, DMA_DEV_TO_MEM,
 			DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 
 	if (!desc)
@@ -923,7 +923,7 @@ static unsigned int dma_handle_tx(struct eg20t_port *priv)
 	}
 
 	desc = priv->chan_tx->device->device_prep_slave_sg(priv->chan_tx,
-					priv->sg_tx_p, nent, DMA_TO_DEVICE,
+					priv->sg_tx_p, nent, DMA_MEM_TO_DEV,
 					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc) {
 		dev_err(priv->port.dev, "%s:device_prep_slave_sg Failed\n",
diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c
index 9e62349b3d9..75085795528 100644
--- a/drivers/tty/serial/sh-sci.c
+++ b/drivers/tty/serial/sh-sci.c
@@ -1339,7 +1339,7 @@ static void sci_submit_rx(struct sci_port *s)
 		struct dma_async_tx_descriptor *desc;
 
 		desc = chan->device->device_prep_slave_sg(chan,
-			sg, 1, DMA_FROM_DEVICE, DMA_PREP_INTERRUPT);
+			sg, 1, DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT);
 
 		if (desc) {
 			s->desc_rx[i] = desc;
@@ -1454,7 +1454,7 @@ static void work_fn_tx(struct work_struct *work)
 	BUG_ON(!sg_dma_len(sg));
 
 	desc = chan->device->device_prep_slave_sg(chan,
-			sg, s->sg_len_tx, DMA_TO_DEVICE,
+			sg, s->sg_len_tx, DMA_MEM_TO_DEV,
 			DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc) {
 		/* switch to PIO */
diff --git a/drivers/usb/musb/ux500_dma.c b/drivers/usb/musb/ux500_dma.c
index a163632877a..97cb45916c4 100644
--- a/drivers/usb/musb/ux500_dma.c
+++ b/drivers/usb/musb/ux500_dma.c
@@ -84,7 +84,7 @@ static bool ux500_configure_channel(struct dma_channel *channel,
 	struct musb_hw_ep *hw_ep = ux500_channel->hw_ep;
 	struct dma_chan *dma_chan = ux500_channel->dma_chan;
 	struct dma_async_tx_descriptor *dma_desc;
-	enum dma_data_direction direction;
+	enum dma_transfer_direction direction;
 	struct scatterlist sg;
 	struct dma_slave_config slave_conf;
 	enum dma_slave_buswidth addr_width;
@@ -104,7 +104,7 @@ static bool ux500_configure_channel(struct dma_channel *channel,
 	sg_dma_address(&sg) = dma_addr;
 	sg_dma_len(&sg) = len;
 
-	direction = ux500_channel->is_tx ? DMA_TO_DEVICE : DMA_FROM_DEVICE;
+	direction = ux500_channel->is_tx ? DMA_MEM_TO_DEV : DMA_DEV_TO_MEM;
 	addr_width = (len & 0x3) ? DMA_SLAVE_BUSWIDTH_1_BYTE :
 					DMA_SLAVE_BUSWIDTH_4_BYTES;
 
diff --git a/drivers/usb/renesas_usbhs/fifo.c b/drivers/usb/renesas_usbhs/fifo.c
index b51fcd80d24..72339bd6fca 100644
--- a/drivers/usb/renesas_usbhs/fifo.c
+++ b/drivers/usb/renesas_usbhs/fifo.c
@@ -772,10 +772,10 @@ static void usbhsf_dma_prepare_tasklet(unsigned long data)
 	struct dma_async_tx_descriptor *desc;
 	struct dma_chan *chan = usbhsf_dma_chan_get(fifo, pkt);
 	struct device *dev = usbhs_priv_to_dev(priv);
-	enum dma_data_direction dir;
+	enum dma_transfer_direction dir;
 	dma_cookie_t cookie;
 
-	dir = usbhs_pipe_is_dir_in(pipe) ? DMA_FROM_DEVICE : DMA_TO_DEVICE;
+	dir = usbhs_pipe_is_dir_in(pipe) ? DMA_DEV_TO_MEM : DMA_MEM_TO_DEV;
 
 	sg_init_table(&sg, 1);
 	sg_set_page(&sg, virt_to_page(pkt->dma),
diff --git a/drivers/video/mx3fb.c b/drivers/video/mx3fb.c
index e3406ab3130..727a5149d81 100644
--- a/drivers/video/mx3fb.c
+++ b/drivers/video/mx3fb.c
@@ -245,6 +245,7 @@ struct mx3fb_data {
 
 	uint32_t		h_start_width;
 	uint32_t		v_start_width;
+	enum disp_data_mapping	disp_data_fmt;
 };
 
 struct dma_chan_request {
@@ -287,11 +288,14 @@ static void mx3fb_write_reg(struct mx3fb_data *mx3fb, u32 value, unsigned long r
 	__raw_writel(value, mx3fb->reg_base + reg);
 }
 
-static const uint32_t di_mappings[] = {
-	0x1600AAAA, 0x00E05555, 0x00070000, 3,	/* RGB888 */
-	0x0005000F, 0x000B000F, 0x0011000F, 1,	/* RGB666 */
-	0x0011000F, 0x000B000F, 0x0005000F, 1,	/* BGR666 */
-	0x0004003F, 0x000A000F, 0x000F003F, 1	/* RGB565 */
+struct di_mapping {
+	uint32_t b0, b1, b2;
+};
+
+static const struct di_mapping di_mappings[] = {
+	[IPU_DISP_DATA_MAPPING_RGB666] = { 0x0005000f, 0x000b000f, 0x0011000f },
+	[IPU_DISP_DATA_MAPPING_RGB565] = { 0x0004003f, 0x000a000f, 0x000f003f },
+	[IPU_DISP_DATA_MAPPING_RGB888] = { 0x00070000, 0x000f0000, 0x00170000 },
 };
 
 static void sdc_fb_init(struct mx3fb_info *fbi)
@@ -334,7 +338,7 @@ static void sdc_enable_channel(struct mx3fb_info *mx3_fbi)
 	/* This enables the channel */
 	if (mx3_fbi->cookie < 0) {
 		mx3_fbi->txd = dma_chan->device->device_prep_slave_sg(dma_chan,
-		      &mx3_fbi->sg[0], 1, DMA_TO_DEVICE, DMA_PREP_INTERRUPT);
+		      &mx3_fbi->sg[0], 1, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT);
 		if (!mx3_fbi->txd) {
 			dev_err(mx3fb->dev, "Cannot allocate descriptor on %d\n",
 				dma_chan->chan_id);
@@ -425,7 +429,6 @@ static int sdc_set_window_pos(struct mx3fb_data *mx3fb, enum ipu_channel channel
  * @pixel_clk:		desired pixel clock frequency in Hz.
  * @width:		width of panel in pixels.
  * @height:		height of panel in pixels.
- * @pixel_fmt:		pixel format of buffer as FOURCC ASCII code.
  * @h_start_width:	number of pixel clocks between the HSYNC signal pulse
  *			and the start of valid data.
  * @h_sync_width:	width of the HSYNC signal in units of pixel clocks.
@@ -442,7 +445,6 @@ static int sdc_set_window_pos(struct mx3fb_data *mx3fb, enum ipu_channel channel
 static int sdc_init_panel(struct mx3fb_data *mx3fb, enum ipu_panel panel,
 			  uint32_t pixel_clk,
 			  uint16_t width, uint16_t height,
-			  enum pixel_fmt pixel_fmt,
 			  uint16_t h_start_width, uint16_t h_sync_width,
 			  uint16_t h_end_width, uint16_t v_start_width,
 			  uint16_t v_sync_width, uint16_t v_end_width,
@@ -453,6 +455,7 @@ static int sdc_init_panel(struct mx3fb_data *mx3fb, enum ipu_panel panel,
 	uint32_t old_conf;
 	uint32_t div;
 	struct clk *ipu_clk;
+	const struct di_mapping *map;
 
 	dev_dbg(mx3fb->dev, "panel size = %d x %d", width, height);
 
@@ -540,36 +543,10 @@ static int sdc_init_panel(struct mx3fb_data *mx3fb, enum ipu_panel panel,
 		sig.Vsync_pol << DI_D3_VSYNC_POL_SHIFT;
 	mx3fb_write_reg(mx3fb, old_conf, DI_DISP_SIG_POL);
 
-	switch (pixel_fmt) {
-	case IPU_PIX_FMT_RGB24:
-		mx3fb_write_reg(mx3fb, di_mappings[0], DI_DISP3_B0_MAP);
-		mx3fb_write_reg(mx3fb, di_mappings[1], DI_DISP3_B1_MAP);
-		mx3fb_write_reg(mx3fb, di_mappings[2], DI_DISP3_B2_MAP);
-		mx3fb_write_reg(mx3fb, mx3fb_read_reg(mx3fb, DI_DISP_ACC_CC) |
-			     ((di_mappings[3] - 1) << 12), DI_DISP_ACC_CC);
-		break;
-	case IPU_PIX_FMT_RGB666:
-		mx3fb_write_reg(mx3fb, di_mappings[4], DI_DISP3_B0_MAP);
-		mx3fb_write_reg(mx3fb, di_mappings[5], DI_DISP3_B1_MAP);
-		mx3fb_write_reg(mx3fb, di_mappings[6], DI_DISP3_B2_MAP);
-		mx3fb_write_reg(mx3fb, mx3fb_read_reg(mx3fb, DI_DISP_ACC_CC) |
-			     ((di_mappings[7] - 1) << 12), DI_DISP_ACC_CC);
-		break;
-	case IPU_PIX_FMT_BGR666:
-		mx3fb_write_reg(mx3fb, di_mappings[8], DI_DISP3_B0_MAP);
-		mx3fb_write_reg(mx3fb, di_mappings[9], DI_DISP3_B1_MAP);
-		mx3fb_write_reg(mx3fb, di_mappings[10], DI_DISP3_B2_MAP);
-		mx3fb_write_reg(mx3fb, mx3fb_read_reg(mx3fb, DI_DISP_ACC_CC) |
-			     ((di_mappings[11] - 1) << 12), DI_DISP_ACC_CC);
-		break;
-	default:
-		mx3fb_write_reg(mx3fb, di_mappings[12], DI_DISP3_B0_MAP);
-		mx3fb_write_reg(mx3fb, di_mappings[13], DI_DISP3_B1_MAP);
-		mx3fb_write_reg(mx3fb, di_mappings[14], DI_DISP3_B2_MAP);
-		mx3fb_write_reg(mx3fb, mx3fb_read_reg(mx3fb, DI_DISP_ACC_CC) |
-			     ((di_mappings[15] - 1) << 12), DI_DISP_ACC_CC);
-		break;
-	}
+	map = &di_mappings[mx3fb->disp_data_fmt];
+	mx3fb_write_reg(mx3fb, map->b0, DI_DISP3_B0_MAP);
+	mx3fb_write_reg(mx3fb, map->b1, DI_DISP3_B1_MAP);
+	mx3fb_write_reg(mx3fb, map->b2, DI_DISP3_B2_MAP);
 
 	spin_unlock_irqrestore(&mx3fb->lock, lock_flags);
 
@@ -780,8 +757,6 @@ static int __set_par(struct fb_info *fbi, bool lock)
 		if (sdc_init_panel(mx3fb, mode,
 				   (PICOS2KHZ(fbi->var.pixclock)) * 1000UL,
 				   fbi->var.xres, fbi->var.yres,
-				   (fbi->var.sync & FB_SYNC_SWAP_RGB) ?
-				   IPU_PIX_FMT_BGR666 : IPU_PIX_FMT_RGB666,
 				   fbi->var.left_margin,
 				   fbi->var.hsync_len,
 				   fbi->var.right_margin +
@@ -1117,7 +1092,7 @@ static int mx3fb_pan_display(struct fb_var_screeninfo *var,
 		async_tx_ack(mx3_fbi->txd);
 
 	txd = dma_chan->device->device_prep_slave_sg(dma_chan, sg +
-		mx3_fbi->cur_ipu_buf, 1, DMA_TO_DEVICE, DMA_PREP_INTERRUPT);
+		mx3_fbi->cur_ipu_buf, 1, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT);
 	if (!txd) {
 		dev_err(fbi->device,
 			"Error preparing a DMA transaction descriptor.\n");
@@ -1349,6 +1324,12 @@ static int init_fb_chan(struct mx3fb_data *mx3fb, struct idmac_channel *ichan)
 	const struct fb_videomode *mode;
 	int ret, num_modes;
 
+	if (mx3fb_pdata->disp_data_fmt >= ARRAY_SIZE(di_mappings)) {
+		dev_err(dev, "Illegal display data format %d\n",
+				mx3fb_pdata->disp_data_fmt);
+		return -EINVAL;
+	}
+
 	ichan->client = mx3fb;
 	irq = ichan->eof_irq;
 
@@ -1402,6 +1383,8 @@ static int init_fb_chan(struct mx3fb_data *mx3fb, struct idmac_channel *ichan)
 	mx3fbi->mx3fb		= mx3fb;
 	mx3fbi->blank		= FB_BLANK_NORMAL;
 
+	mx3fb->disp_data_fmt	= mx3fb_pdata->disp_data_fmt;
+
 	init_completion(&mx3fbi->flip_cmpl);
 	disable_irq(ichan->eof_irq);
 	dev_dbg(mx3fb->dev, "disabling irq %d\n", ichan->eof_irq);
diff --git a/include/linux/amba/pl08x.h b/include/linux/amba/pl08x.h
index 9eabffbc4e5..033f6aa670d 100644
--- a/include/linux/amba/pl08x.h
+++ b/include/linux/amba/pl08x.h
@@ -134,7 +134,7 @@ struct pl08x_txd {
 	struct dma_async_tx_descriptor tx;
 	struct list_head node;
 	struct list_head dsg_list;
-	enum dma_data_direction	direction;
+	enum dma_transfer_direction direction;
 	dma_addr_t llis_bus;
 	struct pl08x_lli *llis_va;
 	/* Default cctl value for LLIs */
@@ -197,7 +197,7 @@ struct pl08x_dma_chan {
 	dma_addr_t dst_addr;
 	u32 src_cctl;
 	u32 dst_cctl;
-	enum dma_data_direction	runtime_direction;
+	enum dma_transfer_direction runtime_direction;
 	dma_cookie_t lc;
 	struct list_head pend_list;
 	struct pl08x_txd *at;
diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index 75f53f874b2..679b349d9b6 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -23,7 +23,6 @@
 
 #include <linux/device.h>
 #include <linux/uio.h>
-#include <linux/dma-direction.h>
 #include <linux/scatterlist.h>
 #include <linux/bitmap.h>
 #include <asm/page.h>
@@ -72,11 +71,93 @@ enum dma_transaction_type {
 	DMA_ASYNC_TX,
 	DMA_SLAVE,
 	DMA_CYCLIC,
+	DMA_INTERLEAVE,
+/* last transaction type for creation of the capabilities mask */
+	DMA_TX_TYPE_END,
 };
 
-/* last transaction type for creation of the capabilities mask */
-#define DMA_TX_TYPE_END (DMA_CYCLIC + 1)
+/**
+ * enum dma_transfer_direction - dma transfer mode and direction indicator
+ * @DMA_MEM_TO_MEM: Async/Memcpy mode
+ * @DMA_MEM_TO_DEV: Slave mode & From Memory to Device
+ * @DMA_DEV_TO_MEM: Slave mode & From Device to Memory
+ * @DMA_DEV_TO_DEV: Slave mode & From Device to Device
+ */
+enum dma_transfer_direction {
+	DMA_MEM_TO_MEM,
+	DMA_MEM_TO_DEV,
+	DMA_DEV_TO_MEM,
+	DMA_DEV_TO_DEV,
+	DMA_TRANS_NONE,
+};
+
+/**
+ * Interleaved Transfer Request
+ * ----------------------------
+ * A chunk is collection of contiguous bytes to be transfered.
+ * The gap(in bytes) between two chunks is called inter-chunk-gap(ICG).
+ * ICGs may or maynot change between chunks.
+ * A FRAME is the smallest series of contiguous {chunk,icg} pairs,
+ *  that when repeated an integral number of times, specifies the transfer.
+ * A transfer template is specification of a Frame, the number of times
+ *  it is to be repeated and other per-transfer attributes.
+ *
+ * Practically, a client driver would have ready a template for each
+ *  type of transfer it is going to need during its lifetime and
+ *  set only 'src_start' and 'dst_start' before submitting the requests.
+ *
+ *
+ *  |      Frame-1        |       Frame-2       | ~ |       Frame-'numf'  |
+ *  |====....==.===...=...|====....==.===...=...| ~ |====....==.===...=...|
+ *
+ *    ==  Chunk size
+ *    ... ICG
+ */
+
+/**
+ * struct data_chunk - Element of scatter-gather list that makes a frame.
+ * @size: Number of bytes to read from source.
+ *	  size_dst := fn(op, size_src), so doesn't mean much for destination.
+ * @icg: Number of bytes to jump after last src/dst address of this
+ *	 chunk and before first src/dst address for next chunk.
+ *	 Ignored for dst(assumed 0), if dst_inc is true and dst_sgl is false.
+ *	 Ignored for src(assumed 0), if src_inc is true and src_sgl is false.
+ */
+struct data_chunk {
+	size_t size;
+	size_t icg;
+};
 
+/**
+ * struct dma_interleaved_template - Template to convey DMAC the transfer pattern
+ *	 and attributes.
+ * @src_start: Bus address of source for the first chunk.
+ * @dst_start: Bus address of destination for the first chunk.
+ * @dir: Specifies the type of Source and Destination.
+ * @src_inc: If the source address increments after reading from it.
+ * @dst_inc: If the destination address increments after writing to it.
+ * @src_sgl: If the 'icg' of sgl[] applies to Source (scattered read).
+ *		Otherwise, source is read contiguously (icg ignored).
+ *		Ignored if src_inc is false.
+ * @dst_sgl: If the 'icg' of sgl[] applies to Destination (scattered write).
+ *		Otherwise, destination is filled contiguously (icg ignored).
+ *		Ignored if dst_inc is false.
+ * @numf: Number of frames in this template.
+ * @frame_size: Number of chunks in a frame i.e, size of sgl[].
+ * @sgl: Array of {chunk,icg} pairs that make up a frame.
+ */
+struct dma_interleaved_template {
+	dma_addr_t src_start;
+	dma_addr_t dst_start;
+	enum dma_transfer_direction dir;
+	bool src_inc;
+	bool dst_inc;
+	bool src_sgl;
+	bool dst_sgl;
+	size_t numf;
+	size_t frame_size;
+	struct data_chunk sgl[0];
+};
 
 /**
  * enum dma_ctrl_flags - DMA flags to augment operation preparation,
@@ -269,7 +350,7 @@ enum dma_slave_buswidth {
  * struct, if applicable.
  */
 struct dma_slave_config {
-	enum dma_data_direction direction;
+	enum dma_transfer_direction direction;
 	dma_addr_t src_addr;
 	dma_addr_t dst_addr;
 	enum dma_slave_buswidth src_addr_width;
@@ -433,6 +514,7 @@ struct dma_tx_state {
  * @device_prep_dma_cyclic: prepare a cyclic dma operation suitable for audio.
  *	The function takes a buffer of size buf_len. The callback function will
  *	be called after period_len bytes have been transferred.
+ * @device_prep_interleaved_dma: Transfer expression in a generic way.
  * @device_control: manipulate all pending operations on a channel, returns
  *	zero or error code
  * @device_tx_status: poll for transaction completion, the optional
@@ -492,11 +574,14 @@ struct dma_device {
 
 	struct dma_async_tx_descriptor *(*device_prep_slave_sg)(
 		struct dma_chan *chan, struct scatterlist *sgl,
-		unsigned int sg_len, enum dma_data_direction direction,
+		unsigned int sg_len, enum dma_transfer_direction direction,
 		unsigned long flags);
 	struct dma_async_tx_descriptor *(*device_prep_dma_cyclic)(
 		struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
-		size_t period_len, enum dma_data_direction direction);
+		size_t period_len, enum dma_transfer_direction direction);
+	struct dma_async_tx_descriptor *(*device_prep_interleaved_dma)(
+		struct dma_chan *chan, struct dma_interleaved_template *xt,
+		unsigned long flags);
 	int (*device_control)(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 		unsigned long arg);
 
@@ -522,7 +607,7 @@ static inline int dmaengine_slave_config(struct dma_chan *chan,
 
 static inline struct dma_async_tx_descriptor *dmaengine_prep_slave_single(
 	struct dma_chan *chan, void *buf, size_t len,
-	enum dma_data_direction dir, unsigned long flags)
+	enum dma_transfer_direction dir, unsigned long flags)
 {
 	struct scatterlist sg;
 	sg_init_one(&sg, buf, len);
diff --git a/include/linux/dw_dmac.h b/include/linux/dw_dmac.h
index 4bfe0a2f7d5..f2c64f92c4a 100644
--- a/include/linux/dw_dmac.h
+++ b/include/linux/dw_dmac.h
@@ -127,7 +127,7 @@ struct dw_cyclic_desc {
 
 struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan,
 		dma_addr_t buf_addr, size_t buf_len, size_t period_len,
-		enum dma_data_direction direction);
+		enum dma_transfer_direction direction);
 void dw_dma_cyclic_free(struct dma_chan *chan);
 int dw_dma_cyclic_start(struct dma_chan *chan);
 void dw_dma_cyclic_stop(struct dma_chan *chan);
diff --git a/include/linux/mtd/gpmi-nand.h b/include/linux/mtd/gpmi-nand.h
new file mode 100644
index 00000000000..69b6dbf46b5
--- /dev/null
+++ b/include/linux/mtd/gpmi-nand.h
@@ -0,0 +1,68 @@
+/*
+ * Copyright (C) 2011 Freescale Semiconductor, Inc. All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * 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, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ */
+
+#ifndef __MACH_MXS_GPMI_NAND_H__
+#define __MACH_MXS_GPMI_NAND_H__
+
+/* The size of the resources is fixed. */
+#define GPMI_NAND_RES_SIZE	6
+
+/* Resource names for the GPMI NAND driver. */
+#define GPMI_NAND_GPMI_REGS_ADDR_RES_NAME  "GPMI NAND GPMI Registers"
+#define GPMI_NAND_GPMI_INTERRUPT_RES_NAME  "GPMI NAND GPMI Interrupt"
+#define GPMI_NAND_BCH_REGS_ADDR_RES_NAME   "GPMI NAND BCH Registers"
+#define GPMI_NAND_BCH_INTERRUPT_RES_NAME   "GPMI NAND BCH Interrupt"
+#define GPMI_NAND_DMA_CHANNELS_RES_NAME    "GPMI NAND DMA Channels"
+#define GPMI_NAND_DMA_INTERRUPT_RES_NAME   "GPMI NAND DMA Interrupt"
+
+/**
+ * struct gpmi_nand_platform_data - GPMI NAND driver platform data.
+ *
+ * This structure communicates platform-specific information to the GPMI NAND
+ * driver that can't be expressed as resources.
+ *
+ * @platform_init:           A pointer to a function the driver will call to
+ *                           initialize the platform (e.g., set up the pin mux).
+ * @min_prop_delay_in_ns:    Minimum propagation delay of GPMI signals to and
+ *                           from the NAND Flash device, in nanoseconds.
+ * @max_prop_delay_in_ns:    Maximum propagation delay of GPMI signals to and
+ *                           from the NAND Flash device, in nanoseconds.
+ * @max_chip_count:          The maximum number of chips for which the driver
+ *                           should configure the hardware. This value most
+ *                           likely reflects the number of pins that are
+ *                           connected to a NAND Flash device. If this is
+ *                           greater than the SoC hardware can support, the
+ *                           driver will print a message and fail to initialize.
+ * @partitions:              An optional pointer to an array of partition
+ *                           descriptions.
+ * @partition_count:         The number of elements in the partitions array.
+ */
+struct gpmi_nand_platform_data {
+	/* SoC hardware information. */
+	int		(*platform_init)(void);
+
+	/* NAND Flash information. */
+	unsigned int	min_prop_delay_in_ns;
+	unsigned int	max_prop_delay_in_ns;
+	unsigned int	max_chip_count;
+
+	/* Medium information. */
+	struct		mtd_partition *partitions;
+	unsigned	partition_count;
+};
+#endif
diff --git a/include/linux/sh_dma.h b/include/linux/sh_dma.h
index cb2dd118cc0..8cd7fe59cf1 100644
--- a/include/linux/sh_dma.h
+++ b/include/linux/sh_dma.h
@@ -30,7 +30,7 @@ struct sh_desc {
 	struct sh_dmae_regs hw;
 	struct list_head node;
 	struct dma_async_tx_descriptor async_tx;
-	enum dma_data_direction direction;
+	enum dma_transfer_direction direction;
 	dma_cookie_t cookie;
 	size_t partial;
 	int chunks;
@@ -48,6 +48,7 @@ struct sh_dmae_channel {
 	unsigned int	offset;
 	unsigned int	dmars;
 	unsigned int	dmars_bit;
+	unsigned int	chclr_offset;
 };
 
 struct sh_dmae_pdata {
@@ -68,6 +69,7 @@ struct sh_dmae_pdata {
 	unsigned int dmaor_is_32bit:1;
 	unsigned int needs_tend_set:1;
 	unsigned int no_dmars:1;
+	unsigned int chclr_present:1;
 };
 
 /* DMA register */
diff --git a/include/linux/sirfsoc_dma.h b/include/linux/sirfsoc_dma.h
new file mode 100644
index 00000000000..29d959333d8
--- /dev/null
+++ b/include/linux/sirfsoc_dma.h
@@ -0,0 +1,6 @@
+#ifndef _SIRFSOC_DMA_H_
+#define _SIRFSOC_DMA_H_
+
+bool sirfsoc_dma_filter_id(struct dma_chan *chan, void *chan_id);
+
+#endif
diff --git a/sound/atmel/abdac.c b/sound/atmel/abdac.c
index 6fd9391b3a6..4fa1dbd8ee8 100644
--- a/sound/atmel/abdac.c
+++ b/sound/atmel/abdac.c
@@ -133,7 +133,7 @@ static int atmel_abdac_prepare_dma(struct atmel_abdac *dac,
 	period_len = frames_to_bytes(runtime, runtime->period_size);
 
 	cdesc = dw_dma_cyclic_prep(chan, runtime->dma_addr, buffer_len,
-			period_len, DMA_TO_DEVICE);
+			period_len, DMA_MEM_TO_DEV);
 	if (IS_ERR(cdesc)) {
 		dev_dbg(&dac->pdev->dev, "could not prepare cyclic DMA\n");
 		return PTR_ERR(cdesc);
diff --git a/sound/atmel/ac97c.c b/sound/atmel/ac97c.c
index 73516f69ac7..61dade69835 100644
--- a/sound/atmel/ac97c.c
+++ b/sound/atmel/ac97c.c
@@ -102,7 +102,7 @@ static void atmel_ac97c_dma_capture_period_done(void *arg)
 
 static int atmel_ac97c_prepare_dma(struct atmel_ac97c *chip,
 		struct snd_pcm_substream *substream,
-		enum dma_data_direction direction)
+		enum dma_transfer_direction direction)
 {
 	struct dma_chan			*chan;
 	struct dw_cyclic_desc		*cdesc;
@@ -118,7 +118,7 @@ static int atmel_ac97c_prepare_dma(struct atmel_ac97c *chip,
 		return -EINVAL;
 	}
 
-	if (direction == DMA_TO_DEVICE)
+	if (direction == DMA_MEM_TO_DEV)
 		chan = chip->dma.tx_chan;
 	else
 		chan = chip->dma.rx_chan;
@@ -133,7 +133,7 @@ static int atmel_ac97c_prepare_dma(struct atmel_ac97c *chip,
 		return PTR_ERR(cdesc);
 	}
 
-	if (direction == DMA_TO_DEVICE) {
+	if (direction == DMA_MEM_TO_DEV) {
 		cdesc->period_callback = atmel_ac97c_dma_playback_period_done;
 		set_bit(DMA_TX_READY, &chip->flags);
 	} else {
@@ -393,7 +393,7 @@ static int atmel_ac97c_playback_prepare(struct snd_pcm_substream *substream)
 	if (cpu_is_at32ap7000()) {
 		if (!test_bit(DMA_TX_READY, &chip->flags))
 			retval = atmel_ac97c_prepare_dma(chip, substream,
-					DMA_TO_DEVICE);
+					DMA_MEM_TO_DEV);
 	} else {
 		/* Initialize and start the PDC */
 		writel(runtime->dma_addr, chip->regs + ATMEL_PDC_TPR);
@@ -484,7 +484,7 @@ static int atmel_ac97c_capture_prepare(struct snd_pcm_substream *substream)
 	if (cpu_is_at32ap7000()) {
 		if (!test_bit(DMA_RX_READY, &chip->flags))
 			retval = atmel_ac97c_prepare_dma(chip, substream,
-					DMA_FROM_DEVICE);
+					DMA_DEV_TO_MEM);
 	} else {
 		/* Initialize and start the PDC */
 		writel(runtime->dma_addr, chip->regs + ATMEL_PDC_RPR);
diff --git a/sound/soc/ep93xx/ep93xx-pcm.c b/sound/soc/ep93xx/ep93xx-pcm.c
index 3fc96130d1a..de839044987 100644
--- a/sound/soc/ep93xx/ep93xx-pcm.c
+++ b/sound/soc/ep93xx/ep93xx-pcm.c
@@ -113,9 +113,9 @@ static int ep93xx_pcm_open(struct snd_pcm_substream *substream)
 	rtd->dma_data.name = dma_params->name;
 
 	if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK)
-		rtd->dma_data.direction = DMA_TO_DEVICE;
+		rtd->dma_data.direction = DMA_MEM_TO_DEV;
 	else
-		rtd->dma_data.direction = DMA_FROM_DEVICE;
+		rtd->dma_data.direction = DMA_DEV_TO_MEM;
 
 	rtd->dma_chan = dma_request_channel(mask, ep93xx_pcm_dma_filter,
 					    &rtd->dma_data);
diff --git a/sound/soc/imx/imx-pcm-dma-mx2.c b/sound/soc/imx/imx-pcm-dma-mx2.c
index 1cf2fe889f6..aecdba9f65a 100644
--- a/sound/soc/imx/imx-pcm-dma-mx2.c
+++ b/sound/soc/imx/imx-pcm-dma-mx2.c
@@ -107,12 +107,12 @@ static int imx_ssi_dma_alloc(struct snd_pcm_substream *substream,
 	}
 
 	if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) {
-		slave_config.direction = DMA_TO_DEVICE;
+		slave_config.direction = DMA_MEM_TO_DEV;
 		slave_config.dst_addr = dma_params->dma_addr;
 		slave_config.dst_addr_width = buswidth;
 		slave_config.dst_maxburst = dma_params->burstsize;
 	} else {
-		slave_config.direction = DMA_FROM_DEVICE;
+		slave_config.direction = DMA_DEV_TO_MEM;
 		slave_config.src_addr = dma_params->dma_addr;
 		slave_config.src_addr_width = buswidth;
 		slave_config.src_maxburst = dma_params->burstsize;
@@ -159,7 +159,7 @@ static int snd_imx_pcm_hw_params(struct snd_pcm_substream *substream,
 			iprtd->period_bytes * iprtd->periods,
 			iprtd->period_bytes,
 			substream->stream == SNDRV_PCM_STREAM_PLAYBACK ?
-			DMA_TO_DEVICE : DMA_FROM_DEVICE);
+			DMA_MEM_TO_DEV : DMA_DEV_TO_MEM);
 	if (!iprtd->desc) {
 		dev_err(&chan->dev->device, "cannot prepare slave dma\n");
 		return -EINVAL;
diff --git a/sound/soc/mxs/mxs-pcm.c b/sound/soc/mxs/mxs-pcm.c
index 0e12f4e0a76..105f42a394d 100644
--- a/sound/soc/mxs/mxs-pcm.c
+++ b/sound/soc/mxs/mxs-pcm.c
@@ -136,7 +136,7 @@ static int snd_mxs_pcm_hw_params(struct snd_pcm_substream *substream,
 			iprtd->period_bytes * iprtd->periods,
 			iprtd->period_bytes,
 			substream->stream == SNDRV_PCM_STREAM_PLAYBACK ?
-			DMA_TO_DEVICE : DMA_FROM_DEVICE);
+			DMA_MEM_TO_DEV : DMA_DEV_TO_MEM);
 	if (!iprtd->desc) {
 		dev_err(&chan->dev->device, "cannot prepare slave dma\n");
 		return -EINVAL;
diff --git a/sound/soc/samsung/dma.c b/sound/soc/samsung/dma.c
index 427ae0d9817..e4ba17ce6b3 100644
--- a/sound/soc/samsung/dma.c
+++ b/sound/soc/samsung/dma.c
@@ -86,7 +86,7 @@ static void dma_enqueue(struct snd_pcm_substream *substream)
 	dma_info.cap = (samsung_dma_has_circular() ? DMA_CYCLIC : DMA_SLAVE);
 	dma_info.direction =
 		(substream->stream == SNDRV_PCM_STREAM_PLAYBACK
-		? DMA_TO_DEVICE : DMA_FROM_DEVICE);
+		? DMA_MEM_TO_DEV : DMA_DEV_TO_MEM);
 	dma_info.fp = audio_buffdone;
 	dma_info.fp_param = substream;
 	dma_info.period = prtd->dma_period;
@@ -171,7 +171,7 @@ static int dma_hw_params(struct snd_pcm_substream *substream,
 		dma_info.client = prtd->params->client;
 		dma_info.direction =
 			(substream->stream == SNDRV_PCM_STREAM_PLAYBACK
-			? DMA_TO_DEVICE : DMA_FROM_DEVICE);
+			? DMA_MEM_TO_DEV : DMA_DEV_TO_MEM);
 		dma_info.width = prtd->params->dma_size;
 		dma_info.fifo = prtd->params->dma_addr;
 		prtd->params->ch = prtd->params->ops->request(
diff --git a/sound/soc/sh/siu_pcm.c b/sound/soc/sh/siu_pcm.c
index f8f681690a7..0193e595d41 100644
--- a/sound/soc/sh/siu_pcm.c
+++ b/sound/soc/sh/siu_pcm.c
@@ -131,7 +131,7 @@ static int siu_pcm_wr_set(struct siu_port *port_info,
 	sg_dma_address(&sg) = buff;
 
 	desc = siu_stream->chan->device->device_prep_slave_sg(siu_stream->chan,
-		&sg, 1, DMA_TO_DEVICE, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		&sg, 1, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc) {
 		dev_err(dev, "Failed to allocate a dma descriptor\n");
 		return -ENOMEM;
@@ -181,7 +181,7 @@ static int siu_pcm_rd_set(struct siu_port *port_info,
 	sg_dma_address(&sg) = buff;
 
 	desc = siu_stream->chan->device->device_prep_slave_sg(siu_stream->chan,
-		&sg, 1, DMA_FROM_DEVICE, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		&sg, 1, DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc) {
 		dev_err(dev, "Failed to allocate dma descriptor\n");
 		return -ENOMEM;
diff --git a/sound/soc/txx9/txx9aclc.c b/sound/soc/txx9/txx9aclc.c
index 93931def0dc..21554611557 100644
--- a/sound/soc/txx9/txx9aclc.c
+++ b/sound/soc/txx9/txx9aclc.c
@@ -134,7 +134,7 @@ txx9aclc_dma_submit(struct txx9aclc_dmadata *dmadata, dma_addr_t buf_dma_addr)
 	sg_dma_address(&sg) = buf_dma_addr;
 	desc = chan->device->device_prep_slave_sg(chan, &sg, 1,
 		dmadata->substream->stream == SNDRV_PCM_STREAM_PLAYBACK ?
-		DMA_TO_DEVICE : DMA_FROM_DEVICE,
+		DMA_MEM_TO_DEV : DMA_DEV_TO_MEM,
 		DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc) {
 		dev_err(&chan->dev->device, "cannot prepare slave dma\n");