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

Pull dmaengine updates from Vinod Koul:
 "This update brings:

   - the big cleanup up by Maxime for device control and slave
     capabilities.  This makes the API much cleaner.

   - new IMG MDC driver by Andrew

   - new Renesas R-Car Gen2 DMA Controller driver by Laurent along with
     bunch of fixes on rcar drivers

   - odd fixes and updates spread over driver"

* 'for-linus' of git://git.infradead.org/users/vkoul/slave-dma: (130 commits)
  dmaengine: pl330: add DMA_PAUSE feature
  dmaengine: pl330: improve pl330_tx_status() function
  dmaengine: rcar-dmac: Disable channel 0 when using IOMMU
  dmaengine: rcar-dmac: Work around descriptor mode IOMMU errata
  dmaengine: rcar-dmac: Allocate hardware descriptors with DMAC device
  dmaengine: rcar-dmac: Fix oops due to unintialized list in error ISR
  dmaengine: rcar-dmac: Fix spinlock issues in interrupt
  dmaenegine: edma: fix sparse warnings
  dmaengine: rcar-dmac: Fix uninitialized variable usage
  dmaengine: shdmac: extend PM methods
  dmaengine: shdmac: use SET_RUNTIME_PM_OPS()
  dmaengine: pl330: fix bug that cause start the same descs in cyclic
  dmaengine: at_xdmac: allow muliple dwidths when doing slave transfers
  dmaengine: at_xdmac: simplify channel configuration stuff
  dmaengine: at_xdmac: introduce save_cc field
  dmaengine: at_xdmac: wait for in-progress transaction to complete after pausing a channel
  ioat: fail self-test if wait_for_completion times out
  dmaengine: dw: define DW_DMA_MAX_NR_MASTERS
  dmaengine: dw: amend description of dma_dev field
  dmatest: move src_off, dst_off, len inside loop
  ...
diff --git a/Documentation/devicetree/bindings/dma/img-mdc-dma.txt b/Documentation/devicetree/bindings/dma/img-mdc-dma.txt
new file mode 100644
index 0000000..28c1341
--- /dev/null
+++ b/Documentation/devicetree/bindings/dma/img-mdc-dma.txt
@@ -0,0 +1,57 @@
+* IMG Multi-threaded DMA Controller (MDC)
+
+Required properties:
+- compatible: Must be "img,pistachio-mdc-dma".
+- reg: Must contain the base address and length of the MDC registers.
+- interrupts: Must contain all the per-channel DMA interrupts.
+- clocks: Must contain an entry for each entry in clock-names.
+  See ../clock/clock-bindings.txt for details.
+- clock-names: Must include the following entries:
+  - sys: MDC system interface clock.
+- img,cr-periph: Must contain a phandle to the peripheral control syscon
+  node which contains the DMA request to channel mapping registers.
+- img,max-burst-multiplier: Must be the maximum supported burst size multiplier.
+  The maximum burst size is this value multiplied by the hardware-reported bus
+  width.
+- #dma-cells: Must be 3:
+  - The first cell is the peripheral's DMA request line.
+  - The second cell is a bitmap specifying to which channels the DMA request
+    line may be mapped (i.e. bit N set indicates channel N is usable).
+  - The third cell is the thread ID to be used by the channel.
+
+Optional properties:
+- dma-channels: Number of supported DMA channels, up to 32.  If not specified
+  the number reported by the hardware is used.
+
+Example:
+
+mdc: dma-controller@18143000 {
+	compatible = "img,pistachio-mdc-dma";
+	reg = <0x18143000 0x1000>;
+	interrupts = <GIC_SHARED 27 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SHARED 28 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SHARED 29 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SHARED 30 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SHARED 31 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SHARED 32 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SHARED 33 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SHARED 34 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SHARED 35 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SHARED 36 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SHARED 37 IRQ_TYPE_LEVEL_HIGH>,
+		     <GIC_SHARED 38 IRQ_TYPE_LEVEL_HIGH>;
+	clocks = <&system_clk>;
+	clock-names = "sys";
+
+	img,max-burst-multiplier = <16>;
+	img,cr-periph = <&cr_periph>;
+
+	#dma-cells = <3>;
+};
+
+spi@18100f00 {
+	...
+	dmas = <&mdc 9 0xffffffff 0>, <&mdc 10 0xffffffff 0>;
+	dma-names = "tx", "rx";
+	...
+};
diff --git a/Documentation/devicetree/bindings/dma/renesas,rcar-dmac.txt b/Documentation/devicetree/bindings/dma/renesas,rcar-dmac.txt
index f7e21b1..09daeef 100644
--- a/Documentation/devicetree/bindings/dma/renesas,rcar-dmac.txt
+++ b/Documentation/devicetree/bindings/dma/renesas,rcar-dmac.txt
@@ -5,9 +5,6 @@
 can be dedicated to specific clients or shared between a large number of
 clients.
 
-DMA clients are connected to the DMAC ports referenced by an 8-bit identifier
-called MID/RID.
-
 Each DMA client is connected to one dedicated port of the DMAC, identified by
 an 8-bit port number called the MID/RID. A DMA controller can thus serve up to
 256 clients in total. When the number of hardware channels is lower than the
diff --git a/Documentation/devicetree/bindings/dma/snps-dma.txt b/Documentation/devicetree/bindings/dma/snps-dma.txt
index d58675e..c261598 100644
--- a/Documentation/devicetree/bindings/dma/snps-dma.txt
+++ b/Documentation/devicetree/bindings/dma/snps-dma.txt
@@ -38,7 +38,7 @@
 		chan_allocation_order = <1>;
 		chan_priority = <1>;
 		block_size = <0xfff>;
-		data_width = <3 3 0 0>;
+		data_width = <3 3>;
 	};
 
 DMA clients connected to the Designware DMA controller must use the format
diff --git a/Documentation/dmaengine/provider.txt b/Documentation/dmaengine/provider.txt
index 766658c..05d2280 100644
--- a/Documentation/dmaengine/provider.txt
+++ b/Documentation/dmaengine/provider.txt
@@ -113,6 +113,31 @@
   * channels:	should be initialized as a list using the
 		INIT_LIST_HEAD macro for example
 
+  * src_addr_widths:
+    - should contain a bitmask of the supported source transfer width
+
+  * dst_addr_widths:
+    - should contain a bitmask of the supported destination transfer
+      width
+
+  * directions:
+    - should contain a bitmask of the supported slave directions
+      (i.e. excluding mem2mem transfers)
+
+  * residue_granularity:
+    - Granularity of the transfer residue reported to dma_set_residue.
+    - This can be either:
+      + Descriptor
+        -> Your device doesn't support any kind of residue
+           reporting. The framework will only know that a particular
+           transaction descriptor is done.
+      + Segment
+        -> Your device is able to report which chunks have been
+           transferred
+      + Burst
+        -> Your device is able to report which burst have been
+           transferred
+
   * dev: 	should hold the pointer to the struct device associated
 		to your current driver instance.
 
@@ -274,48 +299,36 @@
        account the current period.
      - This function can be called in an interrupt context.
 
-   * device_control
-     - Used by client drivers to control and configure the channel it
-       has a handle on.
-     - Called with a command and an argument
-       + The command is one of the values listed by the enum
-         dma_ctrl_cmd. The valid commands are:
-         + DMA_PAUSE
-           + Pauses a transfer on the channel
-           + This command should operate synchronously on the channel,
-             pausing right away the work of the given channel
-         + DMA_RESUME
-           + Restarts a transfer on the channel
-           + This command should operate synchronously on the channel,
-             resuming right away the work of the given channel
-         + DMA_TERMINATE_ALL
-           + Aborts all the pending and ongoing transfers on the
-             channel
-           + This command should operate synchronously on the channel,
-             terminating right away all the channels
-         + DMA_SLAVE_CONFIG
-           + Reconfigures the channel with passed configuration
-           + This command should NOT perform synchronously, or on any
-             currently queued transfers, but only on subsequent ones
-           + In this case, the function will receive a
-             dma_slave_config structure pointer as an argument, that
-             will detail which configuration to use.
-           + Even though that structure contains a direction field,
-             this field is deprecated in favor of the direction
-             argument given to the prep_* functions
-         + FSLDMA_EXTERNAL_START
-           + TODO: Why does that even exist?
-       + The argument is an opaque unsigned long. This actually is a
-         pointer to a struct dma_slave_config that should be used only
-         in the DMA_SLAVE_CONFIG.
+   * device_config
+     - Reconfigures the channel with the configuration given as
+       argument
+     - This command should NOT perform synchronously, or on any
+       currently queued transfers, but only on subsequent ones
+     - In this case, the function will receive a dma_slave_config
+       structure pointer as an argument, that will detail which
+       configuration to use.
+     - Even though that structure contains a direction field, this
+       field is deprecated in favor of the direction argument given to
+       the prep_* functions
+     - This call is mandatory for slave operations only. This should NOT be
+       set or expected to be set for memcpy operations.
+       If a driver support both, it should use this call for slave
+       operations only and not for memcpy ones.
 
-  * device_slave_caps
-    - Called through the framework by client drivers in order to have
-      an idea of what are the properties of the channel allocated to
-      them.
-    - Such properties are the buswidth, available directions, etc.
-    - Required for every generic layer doing DMA transfers, such as
-      ASoC.
+   * device_pause
+     - Pauses a transfer on the channel
+     - This command should operate synchronously on the channel,
+       pausing right away the work of the given channel
+
+   * device_resume
+     - Resumes a transfer on the channel
+     - This command should operate synchronously on the channel,
+       pausing right away the work of the given channel
+
+   * device_terminate_all
+     - Aborts all the pending and ongoing transfers on the channel
+     - This command should operate synchronously on the channel,
+       terminating right away all the channels
 
 Misc notes (stuff that should be documented, but don't really know
 where to put them)
diff --git a/MAINTAINERS b/MAINTAINERS
index 7ac95f8..4f4915c 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -8503,6 +8503,7 @@
 M:	Viresh Kumar <viresh.linux@gmail.com>
 M:	Andy Shevchenko <andriy.shevchenko@linux.intel.com>
 S:	Maintained
+F:	include/linux/dma/dw.h
 F:	include/linux/platform_data/dma-dw.h
 F:	drivers/dma/dw/
 
diff --git a/arch/arc/boot/dts/abilis_tb10x.dtsi b/arch/arc/boot/dts/abilis_tb10x.dtsi
index a098d7c..cfb5052 100644
--- a/arch/arc/boot/dts/abilis_tb10x.dtsi
+++ b/arch/arc/boot/dts/abilis_tb10x.dtsi
@@ -112,7 +112,7 @@
 			chan_allocation_order = <0>;
 			chan_priority = <1>;
 			block_size = <0x7ff>;
-			data_width = <2 0 0 0>;
+			data_width = <2>;
 			clocks = <&ahb_clk>;
 			clock-names = "hclk";
 		};
diff --git a/arch/arm/boot/dts/spear13xx.dtsi b/arch/arm/boot/dts/spear13xx.dtsi
index a6eb543..40accc8 100644
--- a/arch/arm/boot/dts/spear13xx.dtsi
+++ b/arch/arm/boot/dts/spear13xx.dtsi
@@ -117,7 +117,7 @@
 			chan_priority = <1>;
 			block_size = <0xfff>;
 			dma-masters = <2>;
-			data_width = <3 3 0 0>;
+			data_width = <3 3>;
 		};
 
 		dma@eb000000 {
@@ -133,7 +133,7 @@
 			chan_allocation_order = <1>;
 			chan_priority = <1>;
 			block_size = <0xfff>;
-			data_width = <3 3 0 0>;
+			data_width = <3 3>;
 		};
 
 		fsmc: flash@b0000000 {
diff --git a/arch/avr32/mach-at32ap/at32ap700x.c b/arch/avr32/mach-at32ap/at32ap700x.c
index cc92cdb..1d8b147 100644
--- a/arch/avr32/mach-at32ap/at32ap700x.c
+++ b/arch/avr32/mach-at32ap/at32ap700x.c
@@ -607,7 +607,7 @@
 	.nr_channels	= 3,
 	.block_size	= 4095U,
 	.nr_masters	= 2,
-	.data_width	= { 2, 2, 0, 0 },
+	.data_width	= { 2, 2 },
 };
 
 static struct resource dw_dmac0_resource[] = {
diff --git a/drivers/crypto/ux500/cryp/cryp_core.c b/drivers/crypto/ux500/cryp/cryp_core.c
index d594ae96..fded0a5 100644
--- a/drivers/crypto/ux500/cryp/cryp_core.c
+++ b/drivers/crypto/ux500/cryp/cryp_core.c
@@ -606,12 +606,12 @@
 	dev_dbg(ctx->device->dev, "[%s]: ", __func__);
 
 	chan = ctx->device->dma.chan_mem2cryp;
-	dmaengine_device_control(chan, DMA_TERMINATE_ALL, 0);
+	dmaengine_terminate_all(chan);
 	dma_unmap_sg(chan->device->dev, ctx->device->dma.sg_src,
 		     ctx->device->dma.sg_src_len, DMA_TO_DEVICE);
 
 	chan = ctx->device->dma.chan_cryp2mem;
-	dmaengine_device_control(chan, DMA_TERMINATE_ALL, 0);
+	dmaengine_terminate_all(chan);
 	dma_unmap_sg(chan->device->dev, ctx->device->dma.sg_dst,
 		     ctx->device->dma.sg_dst_len, DMA_FROM_DEVICE);
 }
diff --git a/drivers/crypto/ux500/hash/hash_core.c b/drivers/crypto/ux500/hash/hash_core.c
index 70a2087..187a8fd 100644
--- a/drivers/crypto/ux500/hash/hash_core.c
+++ b/drivers/crypto/ux500/hash/hash_core.c
@@ -202,7 +202,7 @@
 	struct dma_chan *chan;
 
 	chan = ctx->device->dma.chan_mem2hash;
-	dmaengine_device_control(chan, DMA_TERMINATE_ALL, 0);
+	dmaengine_terminate_all(chan);
 	dma_unmap_sg(chan->device->dev, ctx->device->dma.sg,
 		     ctx->device->dma.sg_len, DMA_TO_DEVICE);
 }
diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index faf30a4..a874b6e 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -416,6 +416,15 @@
 	help
 	  Support for "Type-AXI" NBPF DMA IPs from Renesas
 
+config IMG_MDC_DMA
+	tristate "IMG MDC support"
+	depends on MIPS || COMPILE_TEST
+	depends on MFD_SYSCON
+	select DMA_ENGINE
+	select DMA_VIRTUAL_CHANNELS
+	help
+	  Enable support for the IMG multi-threaded DMA controller (MDC).
+
 config DMA_ENGINE
 	bool
 
diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile
index 2022b54..f915f61 100644
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@@ -19,7 +19,7 @@
 obj-$(CONFIG_AT_XDMAC) += at_xdmac.o
 obj-$(CONFIG_MX3_IPU) += ipu/
 obj-$(CONFIG_TXX9_DMAC) += txx9dmac.o
-obj-$(CONFIG_SH_DMAE_BASE) += sh/
+obj-$(CONFIG_RENESAS_DMA) += sh/
 obj-$(CONFIG_COH901318) += coh901318.o coh901318_lli.o
 obj-$(CONFIG_AMCC_PPC440SPE_ADMA) += ppc4xx/
 obj-$(CONFIG_IMX_SDMA) += imx-sdma.o
@@ -50,3 +50,4 @@
 obj-$(CONFIG_INTEL_MIC_X100_DMA) += mic_x100_dma.o
 obj-$(CONFIG_NBPFAXI_DMA) += nbpfaxi.o
 obj-$(CONFIG_DMA_SUN6I) += sun6i-dma.o
+obj-$(CONFIG_IMG_MDC_DMA) += img-mdc-dma.o
diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c
index 1364d00..4a5fd24 100644
--- a/drivers/dma/amba-pl08x.c
+++ b/drivers/dma/amba-pl08x.c
@@ -1386,32 +1386,6 @@
 	return pl08x_cctl(cctl);
 }
 
-static int dma_set_runtime_config(struct dma_chan *chan,
-				  struct dma_slave_config *config)
-{
-	struct pl08x_dma_chan *plchan = to_pl08x_chan(chan);
-	struct pl08x_driver_data *pl08x = plchan->host;
-
-	if (!plchan->slave)
-		return -EINVAL;
-
-	/* Reject definitely invalid configurations */
-	if (config->src_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES ||
-	    config->dst_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES)
-		return -EINVAL;
-
-	if (config->device_fc && pl08x->vd->pl080s) {
-		dev_err(&pl08x->adev->dev,
-			"%s: PL080S does not support peripheral flow control\n",
-			__func__);
-		return -EINVAL;
-	}
-
-	plchan->cfg = *config;
-
-	return 0;
-}
-
 /*
  * Slave transactions callback to the slave device to allow
  * synchronization of slave DMA signals with the DMAC enable
@@ -1693,20 +1667,71 @@
 	return vchan_tx_prep(&plchan->vc, &txd->vd, flags);
 }
 
-static int pl08x_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-			 unsigned long arg)
+static int pl08x_config(struct dma_chan *chan,
+			struct dma_slave_config *config)
+{
+	struct pl08x_dma_chan *plchan = to_pl08x_chan(chan);
+	struct pl08x_driver_data *pl08x = plchan->host;
+
+	if (!plchan->slave)
+		return -EINVAL;
+
+	/* Reject definitely invalid configurations */
+	if (config->src_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES ||
+	    config->dst_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES)
+		return -EINVAL;
+
+	if (config->device_fc && pl08x->vd->pl080s) {
+		dev_err(&pl08x->adev->dev,
+			"%s: PL080S does not support peripheral flow control\n",
+			__func__);
+		return -EINVAL;
+	}
+
+	plchan->cfg = *config;
+
+	return 0;
+}
+
+static int pl08x_terminate_all(struct dma_chan *chan)
 {
 	struct pl08x_dma_chan *plchan = to_pl08x_chan(chan);
 	struct pl08x_driver_data *pl08x = plchan->host;
 	unsigned long flags;
-	int ret = 0;
 
-	/* Controls applicable to inactive channels */
-	if (cmd == DMA_SLAVE_CONFIG) {
-		return dma_set_runtime_config(chan,
-					      (struct dma_slave_config *)arg);
+	spin_lock_irqsave(&plchan->vc.lock, flags);
+	if (!plchan->phychan && !plchan->at) {
+		spin_unlock_irqrestore(&plchan->vc.lock, flags);
+		return 0;
 	}
 
+	plchan->state = PL08X_CHAN_IDLE;
+
+	if (plchan->phychan) {
+		/*
+		 * Mark physical channel as free and free any slave
+		 * signal
+		 */
+		pl08x_phy_free(plchan);
+	}
+	/* Dequeue jobs and free LLIs */
+	if (plchan->at) {
+		pl08x_desc_free(&plchan->at->vd);
+		plchan->at = NULL;
+	}
+	/* Dequeue jobs not yet fired as well */
+	pl08x_free_txd_list(pl08x, plchan);
+
+	spin_unlock_irqrestore(&plchan->vc.lock, flags);
+
+	return 0;
+}
+
+static int pl08x_pause(struct dma_chan *chan)
+{
+	struct pl08x_dma_chan *plchan = to_pl08x_chan(chan);
+	unsigned long flags;
+
 	/*
 	 * Anything succeeds on channels with no physical allocation and
 	 * no queued transfers.
@@ -1717,42 +1742,35 @@
 		return 0;
 	}
 
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		plchan->state = PL08X_CHAN_IDLE;
-
-		if (plchan->phychan) {
-			/*
-			 * Mark physical channel as free and free any slave
-			 * signal
-			 */
-			pl08x_phy_free(plchan);
-		}
-		/* Dequeue jobs and free LLIs */
-		if (plchan->at) {
-			pl08x_desc_free(&plchan->at->vd);
-			plchan->at = NULL;
-		}
-		/* Dequeue jobs not yet fired as well */
-		pl08x_free_txd_list(pl08x, plchan);
-		break;
-	case DMA_PAUSE:
-		pl08x_pause_phy_chan(plchan->phychan);
-		plchan->state = PL08X_CHAN_PAUSED;
-		break;
-	case DMA_RESUME:
-		pl08x_resume_phy_chan(plchan->phychan);
-		plchan->state = PL08X_CHAN_RUNNING;
-		break;
-	default:
-		/* Unknown command */
-		ret = -ENXIO;
-		break;
-	}
+	pl08x_pause_phy_chan(plchan->phychan);
+	plchan->state = PL08X_CHAN_PAUSED;
 
 	spin_unlock_irqrestore(&plchan->vc.lock, flags);
 
-	return ret;
+	return 0;
+}
+
+static int pl08x_resume(struct dma_chan *chan)
+{
+	struct pl08x_dma_chan *plchan = to_pl08x_chan(chan);
+	unsigned long flags;
+
+	/*
+	 * Anything succeeds on channels with no physical allocation and
+	 * no queued transfers.
+	 */
+	spin_lock_irqsave(&plchan->vc.lock, flags);
+	if (!plchan->phychan && !plchan->at) {
+		spin_unlock_irqrestore(&plchan->vc.lock, flags);
+		return 0;
+	}
+
+	pl08x_resume_phy_chan(plchan->phychan);
+	plchan->state = PL08X_CHAN_RUNNING;
+
+	spin_unlock_irqrestore(&plchan->vc.lock, flags);
+
+	return 0;
 }
 
 bool pl08x_filter_id(struct dma_chan *chan, void *chan_id)
@@ -2048,7 +2066,10 @@
 	pl08x->memcpy.device_prep_dma_interrupt = pl08x_prep_dma_interrupt;
 	pl08x->memcpy.device_tx_status = pl08x_dma_tx_status;
 	pl08x->memcpy.device_issue_pending = pl08x_issue_pending;
-	pl08x->memcpy.device_control = pl08x_control;
+	pl08x->memcpy.device_config = pl08x_config;
+	pl08x->memcpy.device_pause = pl08x_pause;
+	pl08x->memcpy.device_resume = pl08x_resume;
+	pl08x->memcpy.device_terminate_all = pl08x_terminate_all;
 
 	/* Initialize slave engine */
 	dma_cap_set(DMA_SLAVE, pl08x->slave.cap_mask);
@@ -2061,7 +2082,10 @@
 	pl08x->slave.device_issue_pending = pl08x_issue_pending;
 	pl08x->slave.device_prep_slave_sg = pl08x_prep_slave_sg;
 	pl08x->slave.device_prep_dma_cyclic = pl08x_prep_dma_cyclic;
-	pl08x->slave.device_control = pl08x_control;
+	pl08x->slave.device_config = pl08x_config;
+	pl08x->slave.device_pause = pl08x_pause;
+	pl08x->slave.device_resume = pl08x_resume;
+	pl08x->slave.device_terminate_all = pl08x_terminate_all;
 
 	/* Get the platform data */
 	pl08x->pd = dev_get_platdata(&adev->dev);
diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
index ca9dd26..1e1a4c5 100644
--- a/drivers/dma/at_hdmac.c
+++ b/drivers/dma/at_hdmac.c
@@ -42,6 +42,11 @@
 #define	ATC_DEFAULT_CFG		(ATC_FIFOCFG_HALFFIFO)
 #define	ATC_DEFAULT_CTRLB	(ATC_SIF(AT_DMA_MEM_IF) \
 				|ATC_DIF(AT_DMA_MEM_IF))
+#define ATC_DMA_BUSWIDTHS\
+	(BIT(DMA_SLAVE_BUSWIDTH_UNDEFINED) |\
+	BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |\
+	BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) |\
+	BIT(DMA_SLAVE_BUSWIDTH_4_BYTES))
 
 /*
  * Initial number of descriptors to allocate for each channel. This could
@@ -972,11 +977,13 @@
 	return NULL;
 }
 
-static int set_runtime_config(struct dma_chan *chan,
-			      struct dma_slave_config *sconfig)
+static int atc_config(struct dma_chan *chan,
+		      struct dma_slave_config *sconfig)
 {
 	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
 
+	dev_vdbg(chan2dev(chan), "%s\n", __func__);
+
 	/* Check if it is chan is configured for slave transfers */
 	if (!chan->private)
 		return -EINVAL;
@@ -989,9 +996,7 @@
 	return 0;
 }
 
-
-static int atc_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-		       unsigned long arg)
+static int atc_pause(struct dma_chan *chan)
 {
 	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
 	struct at_dma		*atdma = to_at_dma(chan->device);
@@ -1000,60 +1005,82 @@
 
 	LIST_HEAD(list);
 
-	dev_vdbg(chan2dev(chan), "atc_control (%d)\n", cmd);
+	dev_vdbg(chan2dev(chan), "%s\n", __func__);
 
-	if (cmd == DMA_PAUSE) {
-		spin_lock_irqsave(&atchan->lock, flags);
+	spin_lock_irqsave(&atchan->lock, flags);
 
-		dma_writel(atdma, CHER, AT_DMA_SUSP(chan_id));
-		set_bit(ATC_IS_PAUSED, &atchan->status);
+	dma_writel(atdma, CHER, AT_DMA_SUSP(chan_id));
+	set_bit(ATC_IS_PAUSED, &atchan->status);
 
-		spin_unlock_irqrestore(&atchan->lock, flags);
-	} else if (cmd == DMA_RESUME) {
-		if (!atc_chan_is_paused(atchan))
-			return 0;
+	spin_unlock_irqrestore(&atchan->lock, flags);
 
-		spin_lock_irqsave(&atchan->lock, flags);
+	return 0;
+}
 
-		dma_writel(atdma, CHDR, AT_DMA_RES(chan_id));
-		clear_bit(ATC_IS_PAUSED, &atchan->status);
+static int atc_resume(struct dma_chan *chan)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	struct at_dma		*atdma = to_at_dma(chan->device);
+	int			chan_id = atchan->chan_common.chan_id;
+	unsigned long		flags;
 
-		spin_unlock_irqrestore(&atchan->lock, flags);
-	} else if (cmd == DMA_TERMINATE_ALL) {
-		struct at_desc	*desc, *_desc;
-		/*
-		 * This is only called when something went wrong elsewhere, so
-		 * we don't really care about the data. Just disable the
-		 * channel. We still have to poll the channel enable bit due
-		 * to AHB/HSB limitations.
-		 */
-		spin_lock_irqsave(&atchan->lock, flags);
+	LIST_HEAD(list);
 
-		/* disabling channel: must also remove suspend state */
-		dma_writel(atdma, CHDR, AT_DMA_RES(chan_id) | atchan->mask);
+	dev_vdbg(chan2dev(chan), "%s\n", __func__);
 
-		/* confirm that this channel is disabled */
-		while (dma_readl(atdma, CHSR) & atchan->mask)
-			cpu_relax();
+	if (!atc_chan_is_paused(atchan))
+		return 0;
 
-		/* active_list entries will end up before queued entries */
-		list_splice_init(&atchan->queue, &list);
-		list_splice_init(&atchan->active_list, &list);
+	spin_lock_irqsave(&atchan->lock, flags);
 
-		/* Flush all pending and queued descriptors */
-		list_for_each_entry_safe(desc, _desc, &list, desc_node)
-			atc_chain_complete(atchan, desc);
+	dma_writel(atdma, CHDR, AT_DMA_RES(chan_id));
+	clear_bit(ATC_IS_PAUSED, &atchan->status);
 
-		clear_bit(ATC_IS_PAUSED, &atchan->status);
-		/* if channel dedicated to cyclic operations, free it */
-		clear_bit(ATC_IS_CYCLIC, &atchan->status);
+	spin_unlock_irqrestore(&atchan->lock, flags);
 
-		spin_unlock_irqrestore(&atchan->lock, flags);
-	} else if (cmd == DMA_SLAVE_CONFIG) {
-		return set_runtime_config(chan, (struct dma_slave_config *)arg);
-	} else {
-		return -ENXIO;
-	}
+	return 0;
+}
+
+static int atc_terminate_all(struct dma_chan *chan)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	struct at_dma		*atdma = to_at_dma(chan->device);
+	int			chan_id = atchan->chan_common.chan_id;
+	struct at_desc		*desc, *_desc;
+	unsigned long		flags;
+
+	LIST_HEAD(list);
+
+	dev_vdbg(chan2dev(chan), "%s\n", __func__);
+
+	/*
+	 * This is only called when something went wrong elsewhere, so
+	 * we don't really care about the data. Just disable the
+	 * channel. We still have to poll the channel enable bit due
+	 * to AHB/HSB limitations.
+	 */
+	spin_lock_irqsave(&atchan->lock, flags);
+
+	/* disabling channel: must also remove suspend state */
+	dma_writel(atdma, CHDR, AT_DMA_RES(chan_id) | atchan->mask);
+
+	/* confirm that this channel is disabled */
+	while (dma_readl(atdma, CHSR) & atchan->mask)
+		cpu_relax();
+
+	/* active_list entries will end up before queued entries */
+	list_splice_init(&atchan->queue, &list);
+	list_splice_init(&atchan->active_list, &list);
+
+	/* Flush all pending and queued descriptors */
+	list_for_each_entry_safe(desc, _desc, &list, desc_node)
+		atc_chain_complete(atchan, desc);
+
+	clear_bit(ATC_IS_PAUSED, &atchan->status);
+	/* if channel dedicated to cyclic operations, free it */
+	clear_bit(ATC_IS_CYCLIC, &atchan->status);
+
+	spin_unlock_irqrestore(&atchan->lock, flags);
 
 	return 0;
 }
@@ -1505,7 +1532,14 @@
 		/* controller can do slave DMA: can trigger cyclic transfers */
 		dma_cap_set(DMA_CYCLIC, atdma->dma_common.cap_mask);
 		atdma->dma_common.device_prep_dma_cyclic = atc_prep_dma_cyclic;
-		atdma->dma_common.device_control = atc_control;
+		atdma->dma_common.device_config = atc_config;
+		atdma->dma_common.device_pause = atc_pause;
+		atdma->dma_common.device_resume = atc_resume;
+		atdma->dma_common.device_terminate_all = atc_terminate_all;
+		atdma->dma_common.src_addr_widths = ATC_DMA_BUSWIDTHS;
+		atdma->dma_common.dst_addr_widths = ATC_DMA_BUSWIDTHS;
+		atdma->dma_common.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+		atdma->dma_common.residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
 	}
 
 	dma_writel(atdma, EN, AT_DMA_ENABLE);
@@ -1622,7 +1656,7 @@
 	if (!atc_chan_is_paused(atchan)) {
 		dev_warn(chan2dev(chan),
 		"cyclic channel not paused, should be done by channel user\n");
-		atc_control(chan, DMA_PAUSE, 0);
+		atc_pause(chan);
 	}
 
 	/* now preserve additional data for cyclic operations */
diff --git a/drivers/dma/at_hdmac_regs.h b/drivers/dma/at_hdmac_regs.h
index 2787aba..d6bba6c 100644
--- a/drivers/dma/at_hdmac_regs.h
+++ b/drivers/dma/at_hdmac_regs.h
@@ -232,7 +232,8 @@
  * @save_dscr: for cyclic operations, preserve next descriptor address in
  *             the cyclic list on suspend/resume cycle
  * @remain_desc: to save remain desc length
- * @dma_sconfig: configuration for slave transfers, passed via DMA_SLAVE_CONFIG
+ * @dma_sconfig: configuration for slave transfers, passed via
+ * .device_config
  * @lock: serializes enqueue/dequeue operations to descriptors lists
  * @active_list: list of descriptors dmaengine is being running on
  * @queue: list of descriptors ready to be submitted to engine
diff --git a/drivers/dma/at_xdmac.c b/drivers/dma/at_xdmac.c
index b60d77a2..09e2825 100644
--- a/drivers/dma/at_xdmac.c
+++ b/drivers/dma/at_xdmac.c
@@ -25,6 +25,7 @@
 #include <linux/dmapool.h>
 #include <linux/interrupt.h>
 #include <linux/irq.h>
+#include <linux/kernel.h>
 #include <linux/list.h>
 #include <linux/module.h>
 #include <linux/of_dma.h>
@@ -174,6 +175,13 @@
 
 #define AT_XDMAC_MAX_CHAN	0x20
 
+#define AT_XDMAC_DMA_BUSWIDTHS\
+	(BIT(DMA_SLAVE_BUSWIDTH_UNDEFINED) |\
+	BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |\
+	BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) |\
+	BIT(DMA_SLAVE_BUSWIDTH_4_BYTES) |\
+	BIT(DMA_SLAVE_BUSWIDTH_8_BYTES))
+
 enum atc_status {
 	AT_XDMAC_CHAN_IS_CYCLIC = 0,
 	AT_XDMAC_CHAN_IS_PAUSED,
@@ -184,15 +192,15 @@
 	struct dma_chan			chan;
 	void __iomem			*ch_regs;
 	u32				mask;		/* Channel Mask */
-	u32				cfg[3];		/* Channel Configuration Register */
-	#define	AT_XDMAC_CUR_CFG	0		/* Current channel conf */
-	#define	AT_XDMAC_DEV_TO_MEM_CFG	1		/* Predifined dev to mem channel conf */
-	#define	AT_XDMAC_MEM_TO_DEV_CFG	2		/* Predifined mem to dev channel conf */
+	u32				cfg[2];		/* Channel Configuration Register */
+	#define	AT_XDMAC_DEV_TO_MEM_CFG	0		/* Predifined dev to mem channel conf */
+	#define	AT_XDMAC_MEM_TO_DEV_CFG	1		/* Predifined mem to dev channel conf */
 	u8				perid;		/* Peripheral ID */
 	u8				perif;		/* Peripheral Interface */
 	u8				memif;		/* Memory Interface */
 	u32				per_src_addr;
 	u32				per_dst_addr;
+	u32				save_cc;
 	u32				save_cim;
 	u32				save_cnda;
 	u32				save_cndc;
@@ -344,20 +352,13 @@
 	at_xdmac_chan_write(atchan, AT_XDMAC_CNDA, reg);
 
 	/*
-	 * When doing memory to memory transfer we need to use the next
+	 * When doing non cyclic transfer we need to use the next
 	 * descriptor view 2 since some fields of the configuration register
 	 * depend on transfer size and src/dest addresses.
 	 */
-	if (is_slave_direction(first->direction)) {
+	if (at_xdmac_chan_is_cyclic(atchan)) {
 		reg = AT_XDMAC_CNDC_NDVIEW_NDV1;
-		if (first->direction == DMA_MEM_TO_DEV)
-			atchan->cfg[AT_XDMAC_CUR_CFG] =
-				atchan->cfg[AT_XDMAC_MEM_TO_DEV_CFG];
-		else
-			atchan->cfg[AT_XDMAC_CUR_CFG] =
-				atchan->cfg[AT_XDMAC_DEV_TO_MEM_CFG];
-		at_xdmac_chan_write(atchan, AT_XDMAC_CC,
-				    atchan->cfg[AT_XDMAC_CUR_CFG]);
+		at_xdmac_chan_write(atchan, AT_XDMAC_CC, first->lld.mbr_cfg);
 	} else {
 		/*
 		 * No need to write AT_XDMAC_CC reg, it will be done when the
@@ -561,7 +562,6 @@
 	struct at_xdmac_desc	*first = NULL, *prev = NULL;
 	struct scatterlist	*sg;
 	int			i;
-	u32			cfg;
 	unsigned int		xfer_size = 0;
 
 	if (!sgl)
@@ -583,7 +583,7 @@
 	/* Prepare descriptors. */
 	for_each_sg(sgl, sg, sg_len, i) {
 		struct at_xdmac_desc	*desc = NULL;
-		u32			len, mem;
+		u32			len, mem, dwidth, fixed_dwidth;
 
 		len = sg_dma_len(sg);
 		mem = sg_dma_address(sg);
@@ -608,17 +608,21 @@
 		if (direction == DMA_DEV_TO_MEM) {
 			desc->lld.mbr_sa = atchan->per_src_addr;
 			desc->lld.mbr_da = mem;
-			cfg = atchan->cfg[AT_XDMAC_DEV_TO_MEM_CFG];
+			desc->lld.mbr_cfg = atchan->cfg[AT_XDMAC_DEV_TO_MEM_CFG];
 		} else {
 			desc->lld.mbr_sa = mem;
 			desc->lld.mbr_da = atchan->per_dst_addr;
-			cfg = atchan->cfg[AT_XDMAC_MEM_TO_DEV_CFG];
+			desc->lld.mbr_cfg = atchan->cfg[AT_XDMAC_MEM_TO_DEV_CFG];
 		}
-		desc->lld.mbr_ubc = AT_XDMAC_MBR_UBC_NDV1		/* next descriptor view */
-			| AT_XDMAC_MBR_UBC_NDEN				/* next descriptor dst parameter update */
-			| AT_XDMAC_MBR_UBC_NSEN				/* next descriptor src parameter update */
-			| (i == sg_len - 1 ? 0 : AT_XDMAC_MBR_UBC_NDE)	/* descriptor fetch */
-			| len / (1 << at_xdmac_get_dwidth(cfg));	/* microblock length */
+		dwidth = at_xdmac_get_dwidth(desc->lld.mbr_cfg);
+		fixed_dwidth = IS_ALIGNED(len, 1 << dwidth)
+			       ? at_xdmac_get_dwidth(desc->lld.mbr_cfg)
+			       : AT_XDMAC_CC_DWIDTH_BYTE;
+		desc->lld.mbr_ubc = AT_XDMAC_MBR_UBC_NDV2			/* next descriptor view */
+			| AT_XDMAC_MBR_UBC_NDEN					/* next descriptor dst parameter update */
+			| AT_XDMAC_MBR_UBC_NSEN					/* next descriptor src parameter update */
+			| (i == sg_len - 1 ? 0 : AT_XDMAC_MBR_UBC_NDE)		/* descriptor fetch */
+			| (len >> fixed_dwidth);				/* microblock length */
 		dev_dbg(chan2dev(chan),
 			 "%s: lld: mbr_sa=%pad, mbr_da=%pad, mbr_ubc=0x%08x\n",
 			 __func__, &desc->lld.mbr_sa, &desc->lld.mbr_da, desc->lld.mbr_ubc);
@@ -882,7 +886,7 @@
 	enum dma_status		ret;
 	int			residue;
 	u32			cur_nda, mask, value;
-	u8			dwidth = at_xdmac_get_dwidth(atchan->cfg[AT_XDMAC_CUR_CFG]);
+	u8			dwidth = 0;
 
 	ret = dma_cookie_status(chan, cookie, txstate);
 	if (ret == DMA_COMPLETE)
@@ -912,7 +916,7 @@
 	 */
 	mask = AT_XDMAC_CC_TYPE | AT_XDMAC_CC_DSYNC;
 	value = AT_XDMAC_CC_TYPE_PER_TRAN | AT_XDMAC_CC_DSYNC_PER2MEM;
-	if ((atchan->cfg[AT_XDMAC_CUR_CFG] & mask) == value) {
+	if ((desc->lld.mbr_cfg & mask) == value) {
 		at_xdmac_write(atxdmac, AT_XDMAC_GSWF, atchan->mask);
 		while (!(at_xdmac_chan_read(atchan, AT_XDMAC_CIS) & AT_XDMAC_CIS_FIS))
 			cpu_relax();
@@ -926,6 +930,7 @@
 	 */
 	descs_list = &desc->descs_list;
 	list_for_each_entry_safe(desc, _desc, descs_list, desc_node) {
+		dwidth = at_xdmac_get_dwidth(desc->lld.mbr_cfg);
 		residue -= (desc->lld.mbr_ubc & 0xffffff) << dwidth;
 		if ((desc->lld.mbr_nda & 0xfffffffc) == cur_nda)
 			break;
@@ -1107,58 +1112,80 @@
 	return;
 }
 
-static int at_xdmac_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-			    unsigned long arg)
+static int at_xdmac_device_config(struct dma_chan *chan,
+				  struct dma_slave_config *config)
+{
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+	int ret;
+
+	dev_dbg(chan2dev(chan), "%s\n", __func__);
+
+	spin_lock_bh(&atchan->lock);
+	ret = at_xdmac_set_slave_config(chan, config);
+	spin_unlock_bh(&atchan->lock);
+
+	return ret;
+}
+
+static int at_xdmac_device_pause(struct dma_chan *chan)
+{
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+	struct at_xdmac		*atxdmac = to_at_xdmac(atchan->chan.device);
+
+	dev_dbg(chan2dev(chan), "%s\n", __func__);
+
+	if (test_and_set_bit(AT_XDMAC_CHAN_IS_PAUSED, &atchan->status))
+		return 0;
+
+	spin_lock_bh(&atchan->lock);
+	at_xdmac_write(atxdmac, AT_XDMAC_GRWS, atchan->mask);
+	while (at_xdmac_chan_read(atchan, AT_XDMAC_CC)
+	       & (AT_XDMAC_CC_WRIP | AT_XDMAC_CC_RDIP))
+		cpu_relax();
+	spin_unlock_bh(&atchan->lock);
+
+	return 0;
+}
+
+static int at_xdmac_device_resume(struct dma_chan *chan)
+{
+	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
+	struct at_xdmac		*atxdmac = to_at_xdmac(atchan->chan.device);
+
+	dev_dbg(chan2dev(chan), "%s\n", __func__);
+
+	spin_lock_bh(&atchan->lock);
+	if (!at_xdmac_chan_is_paused(atchan))
+		return 0;
+
+	at_xdmac_write(atxdmac, AT_XDMAC_GRWR, atchan->mask);
+	clear_bit(AT_XDMAC_CHAN_IS_PAUSED, &atchan->status);
+	spin_unlock_bh(&atchan->lock);
+
+	return 0;
+}
+
+static int at_xdmac_device_terminate_all(struct dma_chan *chan)
 {
 	struct at_xdmac_desc	*desc, *_desc;
 	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
 	struct at_xdmac		*atxdmac = to_at_xdmac(atchan->chan.device);
-	int			ret = 0;
 
-	dev_dbg(chan2dev(chan), "%s: cmd=%d\n", __func__, cmd);
+	dev_dbg(chan2dev(chan), "%s\n", __func__);
 
 	spin_lock_bh(&atchan->lock);
+	at_xdmac_write(atxdmac, AT_XDMAC_GD, atchan->mask);
+	while (at_xdmac_read(atxdmac, AT_XDMAC_GS) & atchan->mask)
+		cpu_relax();
 
-	switch (cmd) {
-	case DMA_PAUSE:
-		at_xdmac_write(atxdmac, AT_XDMAC_GRWS, atchan->mask);
-		set_bit(AT_XDMAC_CHAN_IS_PAUSED, &atchan->status);
-		break;
+	/* Cancel all pending transfers. */
+	list_for_each_entry_safe(desc, _desc, &atchan->xfers_list, xfer_node)
+		at_xdmac_remove_xfer(atchan, desc);
 
-	case DMA_RESUME:
-		if (!at_xdmac_chan_is_paused(atchan))
-			break;
-
-		at_xdmac_write(atxdmac, AT_XDMAC_GRWR, atchan->mask);
-		clear_bit(AT_XDMAC_CHAN_IS_PAUSED, &atchan->status);
-		break;
-
-	case DMA_TERMINATE_ALL:
-		at_xdmac_write(atxdmac, AT_XDMAC_GD, atchan->mask);
-		while (at_xdmac_read(atxdmac, AT_XDMAC_GS) & atchan->mask)
-			cpu_relax();
-
-		/* Cancel all pending transfers. */
-		list_for_each_entry_safe(desc, _desc, &atchan->xfers_list, xfer_node)
-			at_xdmac_remove_xfer(atchan, desc);
-
-		clear_bit(AT_XDMAC_CHAN_IS_CYCLIC, &atchan->status);
-		break;
-
-	case DMA_SLAVE_CONFIG:
-		ret = at_xdmac_set_slave_config(chan,
-				(struct dma_slave_config *)arg);
-		break;
-
-	default:
-		dev_err(chan2dev(chan),
-			"unmanaged or unknown dma control cmd: %d\n", cmd);
-		ret = -ENXIO;
-	}
-
+	clear_bit(AT_XDMAC_CHAN_IS_CYCLIC, &atchan->status);
 	spin_unlock_bh(&atchan->lock);
 
-	return ret;
+	return 0;
 }
 
 static int at_xdmac_alloc_chan_resources(struct dma_chan *chan)
@@ -1217,27 +1244,6 @@
 	return;
 }
 
-#define AT_XDMAC_DMA_BUSWIDTHS\
-	(BIT(DMA_SLAVE_BUSWIDTH_UNDEFINED) |\
-	BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |\
-	BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) |\
-	BIT(DMA_SLAVE_BUSWIDTH_4_BYTES) |\
-	BIT(DMA_SLAVE_BUSWIDTH_8_BYTES))
-
-static int at_xdmac_device_slave_caps(struct dma_chan *dchan,
-				      struct dma_slave_caps *caps)
-{
-
-	caps->src_addr_widths = AT_XDMAC_DMA_BUSWIDTHS;
-	caps->dstn_addr_widths = AT_XDMAC_DMA_BUSWIDTHS;
-	caps->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
-	caps->cmd_pause = true;
-	caps->cmd_terminate = true;
-	caps->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
-
-	return 0;
-}
-
 #ifdef CONFIG_PM
 static int atmel_xdmac_prepare(struct device *dev)
 {
@@ -1268,9 +1274,10 @@
 	list_for_each_entry_safe(chan, _chan, &atxdmac->dma.channels, device_node) {
 		struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
 
+		atchan->save_cc = at_xdmac_chan_read(atchan, AT_XDMAC_CC);
 		if (at_xdmac_chan_is_cyclic(atchan)) {
 			if (!at_xdmac_chan_is_paused(atchan))
-				at_xdmac_control(chan, DMA_PAUSE, 0);
+				at_xdmac_device_pause(chan);
 			atchan->save_cim = at_xdmac_chan_read(atchan, AT_XDMAC_CIM);
 			atchan->save_cnda = at_xdmac_chan_read(atchan, AT_XDMAC_CNDA);
 			atchan->save_cndc = at_xdmac_chan_read(atchan, AT_XDMAC_CNDC);
@@ -1290,7 +1297,6 @@
 	struct at_xdmac_chan	*atchan;
 	struct dma_chan		*chan, *_chan;
 	int			i;
-	u32			cfg;
 
 	clk_prepare_enable(atxdmac->clk);
 
@@ -1305,8 +1311,7 @@
 	at_xdmac_write(atxdmac, AT_XDMAC_GE, atxdmac->save_gs);
 	list_for_each_entry_safe(chan, _chan, &atxdmac->dma.channels, device_node) {
 		atchan = to_at_xdmac_chan(chan);
-		cfg = atchan->cfg[AT_XDMAC_CUR_CFG];
-		at_xdmac_chan_write(atchan, AT_XDMAC_CC, cfg);
+		at_xdmac_chan_write(atchan, AT_XDMAC_CC, atchan->save_cc);
 		if (at_xdmac_chan_is_cyclic(atchan)) {
 			at_xdmac_chan_write(atchan, AT_XDMAC_CNDA, atchan->save_cnda);
 			at_xdmac_chan_write(atchan, AT_XDMAC_CNDC, atchan->save_cndc);
@@ -1407,8 +1412,14 @@
 	atxdmac->dma.device_prep_dma_cyclic		= at_xdmac_prep_dma_cyclic;
 	atxdmac->dma.device_prep_dma_memcpy		= at_xdmac_prep_dma_memcpy;
 	atxdmac->dma.device_prep_slave_sg		= at_xdmac_prep_slave_sg;
-	atxdmac->dma.device_control			= at_xdmac_control;
-	atxdmac->dma.device_slave_caps			= at_xdmac_device_slave_caps;
+	atxdmac->dma.device_config			= at_xdmac_device_config;
+	atxdmac->dma.device_pause			= at_xdmac_device_pause;
+	atxdmac->dma.device_resume			= at_xdmac_device_resume;
+	atxdmac->dma.device_terminate_all		= at_xdmac_device_terminate_all;
+	atxdmac->dma.src_addr_widths = AT_XDMAC_DMA_BUSWIDTHS;
+	atxdmac->dma.dst_addr_widths = AT_XDMAC_DMA_BUSWIDTHS;
+	atxdmac->dma.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	atxdmac->dma.residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
 
 	/* Disable all chans and interrupts. */
 	at_xdmac_off(atxdmac);
@@ -1507,7 +1518,6 @@
 	.remove		= at_xdmac_remove,
 	.driver = {
 		.name		= "at_xdmac",
-		.owner		= THIS_MODULE,
 		.of_match_table	= of_match_ptr(atmel_xdmac_dt_ids),
 		.pm		= &atmel_xdmac_dev_pm_ops,
 	}
diff --git a/drivers/dma/bcm2835-dma.c b/drivers/dma/bcm2835-dma.c
index 918b7b3..0723096 100644
--- a/drivers/dma/bcm2835-dma.c
+++ b/drivers/dma/bcm2835-dma.c
@@ -436,9 +436,11 @@
 	return vchan_tx_prep(&c->vc, &d->vd, flags);
 }
 
-static int bcm2835_dma_slave_config(struct bcm2835_chan *c,
-		struct dma_slave_config *cfg)
+static int bcm2835_dma_slave_config(struct dma_chan *chan,
+				    struct dma_slave_config *cfg)
 {
+	struct bcm2835_chan *c = to_bcm2835_dma_chan(chan);
+
 	if ((cfg->direction == DMA_DEV_TO_MEM &&
 	     cfg->src_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES) ||
 	    (cfg->direction == DMA_MEM_TO_DEV &&
@@ -452,8 +454,9 @@
 	return 0;
 }
 
-static int bcm2835_dma_terminate_all(struct bcm2835_chan *c)
+static int bcm2835_dma_terminate_all(struct dma_chan *chan)
 {
+	struct bcm2835_chan *c = to_bcm2835_dma_chan(chan);
 	struct bcm2835_dmadev *d = to_bcm2835_dma_dev(c->vc.chan.device);
 	unsigned long flags;
 	int timeout = 10000;
@@ -495,24 +498,6 @@
 	return 0;
 }
 
-static int bcm2835_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-	unsigned long arg)
-{
-	struct bcm2835_chan *c = to_bcm2835_dma_chan(chan);
-
-	switch (cmd) {
-	case DMA_SLAVE_CONFIG:
-		return bcm2835_dma_slave_config(c,
-				(struct dma_slave_config *)arg);
-
-	case DMA_TERMINATE_ALL:
-		return bcm2835_dma_terminate_all(c);
-
-	default:
-		return -ENXIO;
-	}
-}
-
 static int bcm2835_dma_chan_init(struct bcm2835_dmadev *d, int chan_id, int irq)
 {
 	struct bcm2835_chan *c;
@@ -565,18 +550,6 @@
 	return chan;
 }
 
-static int bcm2835_dma_device_slave_caps(struct dma_chan *dchan,
-	struct dma_slave_caps *caps)
-{
-	caps->src_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_4_BYTES);
-	caps->dstn_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_4_BYTES);
-	caps->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
-	caps->cmd_pause = false;
-	caps->cmd_terminate = true;
-
-	return 0;
-}
-
 static int bcm2835_dma_probe(struct platform_device *pdev)
 {
 	struct bcm2835_dmadev *od;
@@ -615,9 +588,12 @@
 	od->ddev.device_free_chan_resources = bcm2835_dma_free_chan_resources;
 	od->ddev.device_tx_status = bcm2835_dma_tx_status;
 	od->ddev.device_issue_pending = bcm2835_dma_issue_pending;
-	od->ddev.device_slave_caps = bcm2835_dma_device_slave_caps;
 	od->ddev.device_prep_dma_cyclic = bcm2835_dma_prep_dma_cyclic;
-	od->ddev.device_control = bcm2835_dma_control;
+	od->ddev.device_config = bcm2835_dma_slave_config;
+	od->ddev.device_terminate_all = bcm2835_dma_terminate_all;
+	od->ddev.src_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_4_BYTES);
+	od->ddev.dst_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_4_BYTES);
+	od->ddev.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
 	od->ddev.dev = &pdev->dev;
 	INIT_LIST_HEAD(&od->ddev.channels);
 	spin_lock_init(&od->lock);
diff --git a/drivers/dma/coh901318.c b/drivers/dma/coh901318.c
index e88588d..fd22dd3 100644
--- a/drivers/dma/coh901318.c
+++ b/drivers/dma/coh901318.c
@@ -1690,7 +1690,7 @@
  * Pauses a transfer without losing data. Enables power save.
  * Use this function in conjunction with coh901318_resume.
  */
-static void coh901318_pause(struct dma_chan *chan)
+static int coh901318_pause(struct dma_chan *chan)
 {
 	u32 val;
 	unsigned long flags;
@@ -1730,12 +1730,13 @@
 	enable_powersave(cohc);
 
 	spin_unlock_irqrestore(&cohc->lock, flags);
+	return 0;
 }
 
 /* Resumes a transfer that has been stopped via 300_dma_stop(..).
    Power save is handled.
 */
-static void coh901318_resume(struct dma_chan *chan)
+static int coh901318_resume(struct dma_chan *chan)
 {
 	u32 val;
 	unsigned long flags;
@@ -1760,6 +1761,7 @@
 	}
 
 	spin_unlock_irqrestore(&cohc->lock, flags);
+	return 0;
 }
 
 bool coh901318_filter_id(struct dma_chan *chan, void *chan_id)
@@ -2114,6 +2116,57 @@
 	return IRQ_HANDLED;
 }
 
+static int coh901318_terminate_all(struct dma_chan *chan)
+{
+	unsigned long flags;
+	struct coh901318_chan *cohc = to_coh901318_chan(chan);
+	struct coh901318_desc *cohd;
+	void __iomem *virtbase = cohc->base->virtbase;
+
+	/* The remainder of this function terminates the transfer */
+	coh901318_pause(chan);
+	spin_lock_irqsave(&cohc->lock, flags);
+
+	/* Clear any pending BE or TC interrupt */
+	if (cohc->id < 32) {
+		writel(1 << cohc->id, virtbase + COH901318_BE_INT_CLEAR1);
+		writel(1 << cohc->id, virtbase + COH901318_TC_INT_CLEAR1);
+	} else {
+		writel(1 << (cohc->id - 32), virtbase +
+		       COH901318_BE_INT_CLEAR2);
+		writel(1 << (cohc->id - 32), virtbase +
+		       COH901318_TC_INT_CLEAR2);
+	}
+
+	enable_powersave(cohc);
+
+	while ((cohd = coh901318_first_active_get(cohc))) {
+		/* release the lli allocation*/
+		coh901318_lli_free(&cohc->base->pool, &cohd->lli);
+
+		/* return desc to free-list */
+		coh901318_desc_remove(cohd);
+		coh901318_desc_free(cohc, cohd);
+	}
+
+	while ((cohd = coh901318_first_queued(cohc))) {
+		/* release the lli allocation*/
+		coh901318_lli_free(&cohc->base->pool, &cohd->lli);
+
+		/* return desc to free-list */
+		coh901318_desc_remove(cohd);
+		coh901318_desc_free(cohc, cohd);
+	}
+
+
+	cohc->nbr_active_done = 0;
+	cohc->busy = 0;
+
+	spin_unlock_irqrestore(&cohc->lock, flags);
+
+	return 0;
+}
+
 static int coh901318_alloc_chan_resources(struct dma_chan *chan)
 {
 	struct coh901318_chan	*cohc = to_coh901318_chan(chan);
@@ -2156,7 +2209,7 @@
 
 	spin_unlock_irqrestore(&cohc->lock, flags);
 
-	dmaengine_terminate_all(chan);
+	coh901318_terminate_all(chan);
 }
 
 
@@ -2461,8 +2514,8 @@
 	},
 };
 
-static void coh901318_dma_set_runtimeconfig(struct dma_chan *chan,
-			struct dma_slave_config *config)
+static int coh901318_dma_set_runtimeconfig(struct dma_chan *chan,
+					   struct dma_slave_config *config)
 {
 	struct coh901318_chan *cohc = to_coh901318_chan(chan);
 	dma_addr_t addr;
@@ -2482,7 +2535,7 @@
 		maxburst = config->dst_maxburst;
 	} else {
 		dev_err(COHC_2_DEV(cohc), "illegal channel mode\n");
-		return;
+		return -EINVAL;
 	}
 
 	dev_dbg(COHC_2_DEV(cohc), "configure channel for %d byte transfers\n",
@@ -2528,7 +2581,7 @@
 	default:
 		dev_err(COHC_2_DEV(cohc),
 			"bad runtimeconfig: alien address width\n");
-		return;
+		return -EINVAL;
 	}
 
 	ctrl |= burst_sizes[i].reg;
@@ -2538,84 +2591,12 @@
 
 	cohc->addr = addr;
 	cohc->ctrl = ctrl;
-}
-
-static int
-coh901318_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-		  unsigned long arg)
-{
-	unsigned long flags;
-	struct coh901318_chan *cohc = to_coh901318_chan(chan);
-	struct coh901318_desc *cohd;
-	void __iomem *virtbase = cohc->base->virtbase;
-
-	if (cmd == DMA_SLAVE_CONFIG) {
-		struct dma_slave_config *config =
-			(struct dma_slave_config *) arg;
-
-		coh901318_dma_set_runtimeconfig(chan, config);
-		return 0;
-	  }
-
-	if (cmd == DMA_PAUSE) {
-		coh901318_pause(chan);
-		return 0;
-	}
-
-	if (cmd == DMA_RESUME) {
-		coh901318_resume(chan);
-		return 0;
-	}
-
-	if (cmd != DMA_TERMINATE_ALL)
-		return -ENXIO;
-
-	/* The remainder of this function terminates the transfer */
-	coh901318_pause(chan);
-	spin_lock_irqsave(&cohc->lock, flags);
-
-	/* Clear any pending BE or TC interrupt */
-	if (cohc->id < 32) {
-		writel(1 << cohc->id, virtbase + COH901318_BE_INT_CLEAR1);
-		writel(1 << cohc->id, virtbase + COH901318_TC_INT_CLEAR1);
-	} else {
-		writel(1 << (cohc->id - 32), virtbase +
-		       COH901318_BE_INT_CLEAR2);
-		writel(1 << (cohc->id - 32), virtbase +
-		       COH901318_TC_INT_CLEAR2);
-	}
-
-	enable_powersave(cohc);
-
-	while ((cohd = coh901318_first_active_get(cohc))) {
-		/* release the lli allocation*/
-		coh901318_lli_free(&cohc->base->pool, &cohd->lli);
-
-		/* return desc to free-list */
-		coh901318_desc_remove(cohd);
-		coh901318_desc_free(cohc, cohd);
-	}
-
-	while ((cohd = coh901318_first_queued(cohc))) {
-		/* release the lli allocation*/
-		coh901318_lli_free(&cohc->base->pool, &cohd->lli);
-
-		/* return desc to free-list */
-		coh901318_desc_remove(cohd);
-		coh901318_desc_free(cohc, cohd);
-	}
-
-
-	cohc->nbr_active_done = 0;
-	cohc->busy = 0;
-
-	spin_unlock_irqrestore(&cohc->lock, flags);
 
 	return 0;
 }
 
-void coh901318_base_init(struct dma_device *dma, const int *pick_chans,
-			 struct coh901318_base *base)
+static void coh901318_base_init(struct dma_device *dma, const int *pick_chans,
+				struct coh901318_base *base)
 {
 	int chans_i;
 	int i = 0;
@@ -2717,7 +2698,10 @@
 	base->dma_slave.device_prep_slave_sg = coh901318_prep_slave_sg;
 	base->dma_slave.device_tx_status = coh901318_tx_status;
 	base->dma_slave.device_issue_pending = coh901318_issue_pending;
-	base->dma_slave.device_control = coh901318_control;
+	base->dma_slave.device_config = coh901318_dma_set_runtimeconfig;
+	base->dma_slave.device_pause = coh901318_pause;
+	base->dma_slave.device_resume = coh901318_resume;
+	base->dma_slave.device_terminate_all = coh901318_terminate_all;
 	base->dma_slave.dev = &pdev->dev;
 
 	err = dma_async_device_register(&base->dma_slave);
@@ -2737,7 +2721,10 @@
 	base->dma_memcpy.device_prep_dma_memcpy = coh901318_prep_memcpy;
 	base->dma_memcpy.device_tx_status = coh901318_tx_status;
 	base->dma_memcpy.device_issue_pending = coh901318_issue_pending;
-	base->dma_memcpy.device_control = coh901318_control;
+	base->dma_memcpy.device_config = coh901318_dma_set_runtimeconfig;
+	base->dma_memcpy.device_pause = coh901318_pause;
+	base->dma_memcpy.device_resume = coh901318_resume;
+	base->dma_memcpy.device_terminate_all = coh901318_terminate_all;
 	base->dma_memcpy.dev = &pdev->dev;
 	/*
 	 * This controller can only access address at even 32bit boundaries,
diff --git a/drivers/dma/cppi41.c b/drivers/dma/cppi41.c
index b743adf..512cb8e 100644
--- a/drivers/dma/cppi41.c
+++ b/drivers/dma/cppi41.c
@@ -525,12 +525,6 @@
 	return &c->txd;
 }
 
-static int cpp41_cfg_chan(struct cppi41_channel *c,
-		struct dma_slave_config *cfg)
-{
-	return 0;
-}
-
 static void cppi41_compute_td_desc(struct cppi41_desc *d)
 {
 	d->pd0 = DESC_TYPE_TEARD << DESC_TYPE;
@@ -647,28 +641,6 @@
 	return 0;
 }
 
-static int cppi41_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-	unsigned long arg)
-{
-	struct cppi41_channel *c = to_cpp41_chan(chan);
-	int ret;
-
-	switch (cmd) {
-	case DMA_SLAVE_CONFIG:
-		ret = cpp41_cfg_chan(c, (struct dma_slave_config *) arg);
-		break;
-
-	case DMA_TERMINATE_ALL:
-		ret = cppi41_stop_chan(chan);
-		break;
-
-	default:
-		ret = -ENXIO;
-		break;
-	}
-	return ret;
-}
-
 static void cleanup_chans(struct cppi41_dd *cdd)
 {
 	while (!list_empty(&cdd->ddev.channels)) {
@@ -953,7 +925,7 @@
 	cdd->ddev.device_tx_status = cppi41_dma_tx_status;
 	cdd->ddev.device_issue_pending = cppi41_dma_issue_pending;
 	cdd->ddev.device_prep_slave_sg = cppi41_dma_prep_slave_sg;
-	cdd->ddev.device_control = cppi41_dma_control;
+	cdd->ddev.device_terminate_all = cppi41_stop_chan;
 	cdd->ddev.dev = dev;
 	INIT_LIST_HEAD(&cdd->ddev.channels);
 	cpp41_dma_info.dma_cap = cdd->ddev.cap_mask;
diff --git a/drivers/dma/dma-jz4740.c b/drivers/dma/dma-jz4740.c
index bdeafee..4527a3e 100644
--- a/drivers/dma/dma-jz4740.c
+++ b/drivers/dma/dma-jz4740.c
@@ -210,7 +210,7 @@
 }
 
 static int jz4740_dma_slave_config(struct dma_chan *c,
-	const struct dma_slave_config *config)
+				   struct dma_slave_config *config)
 {
 	struct jz4740_dmaengine_chan *chan = to_jz4740_dma_chan(c);
 	struct jz4740_dma_dev *dmadev = jz4740_dma_chan_get_dev(chan);
@@ -290,21 +290,6 @@
 	return 0;
 }
 
-static int jz4740_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-	unsigned long arg)
-{
-	struct dma_slave_config *config = (struct dma_slave_config *)arg;
-
-	switch (cmd) {
-	case DMA_SLAVE_CONFIG:
-		return jz4740_dma_slave_config(chan, config);
-	case DMA_TERMINATE_ALL:
-		return jz4740_dma_terminate_all(chan);
-	default:
-		return -ENOSYS;
-	}
-}
-
 static int jz4740_dma_start_transfer(struct jz4740_dmaengine_chan *chan)
 {
 	struct jz4740_dma_dev *dmadev = jz4740_dma_chan_get_dev(chan);
@@ -561,7 +546,8 @@
 	dd->device_issue_pending = jz4740_dma_issue_pending;
 	dd->device_prep_slave_sg = jz4740_dma_prep_slave_sg;
 	dd->device_prep_dma_cyclic = jz4740_dma_prep_dma_cyclic;
-	dd->device_control = jz4740_dma_control;
+	dd->device_config = jz4740_dma_slave_config;
+	dd->device_terminate_all = jz4740_dma_terminate_all;
 	dd->dev = &pdev->dev;
 	INIT_LIST_HEAD(&dd->channels);
 
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index e057935..f15712f 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -222,31 +222,35 @@
  */
 static int dma_chan_get(struct dma_chan *chan)
 {
-	int err = -ENODEV;
 	struct module *owner = dma_chan_to_owner(chan);
+	int ret;
 
+	/* The channel is already in use, update client count */
 	if (chan->client_count) {
 		__module_get(owner);
-		err = 0;
-	} else if (try_module_get(owner))
-		err = 0;
-
-	if (err == 0)
-		chan->client_count++;
-
-	/* allocate upon first client reference */
-	if (chan->client_count == 1 && err == 0) {
-		int desc_cnt = chan->device->device_alloc_chan_resources(chan);
-
-		if (desc_cnt < 0) {
-			err = desc_cnt;
-			chan->client_count = 0;
-			module_put(owner);
-		} else if (!dma_has_cap(DMA_PRIVATE, chan->device->cap_mask))
-			balance_ref_count(chan);
+		goto out;
 	}
 
-	return err;
+	if (!try_module_get(owner))
+		return -ENODEV;
+
+	/* allocate upon first client reference */
+	if (chan->device->device_alloc_chan_resources) {
+		ret = chan->device->device_alloc_chan_resources(chan);
+		if (ret < 0)
+			goto err_out;
+	}
+
+	if (!dma_has_cap(DMA_PRIVATE, chan->device->cap_mask))
+		balance_ref_count(chan);
+
+out:
+	chan->client_count++;
+	return 0;
+
+err_out:
+	module_put(owner);
+	return ret;
 }
 
 /**
@@ -257,11 +261,15 @@
  */
 static void dma_chan_put(struct dma_chan *chan)
 {
+	/* This channel is not in use, bail out */
 	if (!chan->client_count)
-		return; /* this channel failed alloc_chan_resources */
+		return;
+
 	chan->client_count--;
 	module_put(dma_chan_to_owner(chan));
-	if (chan->client_count == 0)
+
+	/* This channel is not in use anymore, free it */
+	if (!chan->client_count && chan->device->device_free_chan_resources)
 		chan->device->device_free_chan_resources(chan);
 }
 
@@ -471,6 +479,39 @@
 		}
 }
 
+int dma_get_slave_caps(struct dma_chan *chan, struct dma_slave_caps *caps)
+{
+	struct dma_device *device;
+
+	if (!chan || !caps)
+		return -EINVAL;
+
+	device = chan->device;
+
+	/* check if the channel supports slave transactions */
+	if (!test_bit(DMA_SLAVE, device->cap_mask.bits))
+		return -ENXIO;
+
+	/*
+	 * Check whether it reports it uses the generic slave
+	 * capabilities, if not, that means it doesn't support any
+	 * kind of slave capabilities reporting.
+	 */
+	if (!device->directions)
+		return -ENXIO;
+
+	caps->src_addr_widths = device->src_addr_widths;
+	caps->dst_addr_widths = device->dst_addr_widths;
+	caps->directions = device->directions;
+	caps->residue_granularity = device->residue_granularity;
+
+	caps->cmd_pause = !!device->device_pause;
+	caps->cmd_terminate = !!device->device_terminate_all;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(dma_get_slave_caps);
+
 static struct dma_chan *private_candidate(const dma_cap_mask_t *mask,
 					  struct dma_device *dev,
 					  dma_filter_fn fn, void *fn_param)
@@ -811,17 +852,16 @@
 		!device->device_prep_dma_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);
 	BUG_ON(!device->device_tx_status);
 	BUG_ON(!device->device_issue_pending);
 	BUG_ON(!device->dev);
 
+	WARN(dma_has_cap(DMA_SLAVE, device->cap_mask) && !device->directions,
+	     "this driver doesn't support generic slave capabilities reporting\n");
+
 	/* note: this only matters in the
 	 * CONFIG_ASYNC_TX_ENABLE_CHANNEL_SWITCH=n case
 	 */
diff --git a/drivers/dma/dmatest.c b/drivers/dma/dmatest.c
index a8d7809..220ee49 100644
--- a/drivers/dma/dmatest.c
+++ b/drivers/dma/dmatest.c
@@ -349,14 +349,14 @@
 		       unsigned long data)
 {
 	pr_debug("%s: result #%u: '%s' with src_off=0x%x dst_off=0x%x len=0x%x (%lu)\n",
-		   current->comm, n, err, src_off, dst_off, len, data);
+		 current->comm, n, err, src_off, dst_off, len, data);
 }
 
-#define verbose_result(err, n, src_off, dst_off, len, data) ({ \
-	if (verbose) \
-		result(err, n, src_off, dst_off, len, data); \
-	else \
-		dbg_result(err, n, src_off, dst_off, len, data); \
+#define verbose_result(err, n, src_off, dst_off, len, data) ({	\
+	if (verbose)						\
+		result(err, n, src_off, dst_off, len, data);	\
+	else							\
+		dbg_result(err, n, src_off, dst_off, len, data);\
 })
 
 static unsigned long long dmatest_persec(s64 runtime, unsigned int val)
@@ -405,7 +405,6 @@
 	struct dmatest_params	*params;
 	struct dma_chan		*chan;
 	struct dma_device	*dev;
-	unsigned int		src_off, dst_off, len;
 	unsigned int		error_count;
 	unsigned int		failed_tests = 0;
 	unsigned int		total_tests = 0;
@@ -484,6 +483,7 @@
 		struct dmaengine_unmap_data *um;
 		dma_addr_t srcs[src_cnt];
 		dma_addr_t *dsts;
+		unsigned int src_off, dst_off, len;
 		u8 align = 0;
 
 		total_tests++;
@@ -502,15 +502,21 @@
 			break;
 		}
 
-		if (params->noverify) {
+		if (params->noverify)
 			len = params->buf_size;
+		else
+			len = dmatest_random() % params->buf_size + 1;
+
+		len = (len >> align) << align;
+		if (!len)
+			len = 1 << align;
+
+		total_len += len;
+
+		if (params->noverify) {
 			src_off = 0;
 			dst_off = 0;
 		} else {
-			len = dmatest_random() % params->buf_size + 1;
-			len = (len >> align) << align;
-			if (!len)
-				len = 1 << align;
 			src_off = dmatest_random() % (params->buf_size - len + 1);
 			dst_off = dmatest_random() % (params->buf_size - len + 1);
 
@@ -523,11 +529,6 @@
 					  params->buf_size);
 		}
 
-		len = (len >> align) << align;
-		if (!len)
-			len = 1 << align;
-		total_len += len;
-
 		um = dmaengine_get_unmap_data(dev->dev, src_cnt+dst_cnt,
 					      GFP_KERNEL);
 		if (!um) {
diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index 5c062548..455b7a4 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -61,6 +61,13 @@
  */
 #define NR_DESCS_PER_CHANNEL	64
 
+/* The set of bus widths supported by the DMA controller */
+#define DW_DMA_BUSWIDTHS			  \
+	BIT(DMA_SLAVE_BUSWIDTH_UNDEFINED)	| \
+	BIT(DMA_SLAVE_BUSWIDTH_1_BYTE)		| \
+	BIT(DMA_SLAVE_BUSWIDTH_2_BYTES)		| \
+	BIT(DMA_SLAVE_BUSWIDTH_4_BYTES)
+
 /*----------------------------------------------------------------------*/
 
 static struct device *chan2dev(struct dma_chan *chan)
@@ -955,8 +962,7 @@
 		*maxburst = 0;
 }
 
-static int
-set_runtime_config(struct dma_chan *chan, struct dma_slave_config *sconfig)
+static int dwc_config(struct dma_chan *chan, struct dma_slave_config *sconfig)
 {
 	struct dw_dma_chan *dwc = to_dw_dma_chan(chan);
 
@@ -973,16 +979,25 @@
 	return 0;
 }
 
-static inline void dwc_chan_pause(struct dw_dma_chan *dwc)
+static int dwc_pause(struct dma_chan *chan)
 {
-	u32 cfglo = channel_readl(dwc, CFG_LO);
-	unsigned int count = 20;	/* timeout iterations */
+	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
+	unsigned long		flags;
+	unsigned int		count = 20;	/* timeout iterations */
+	u32			cfglo;
 
+	spin_lock_irqsave(&dwc->lock, flags);
+
+	cfglo = channel_readl(dwc, CFG_LO);
 	channel_writel(dwc, CFG_LO, cfglo | DWC_CFGL_CH_SUSP);
 	while (!(channel_readl(dwc, CFG_LO) & DWC_CFGL_FIFO_EMPTY) && count--)
 		udelay(2);
 
 	dwc->paused = true;
+
+	spin_unlock_irqrestore(&dwc->lock, flags);
+
+	return 0;
 }
 
 static inline void dwc_chan_resume(struct dw_dma_chan *dwc)
@@ -994,8 +1009,24 @@
 	dwc->paused = false;
 }
 
-static int dwc_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-		       unsigned long arg)
+static int dwc_resume(struct dma_chan *chan)
+{
+	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
+	unsigned long		flags;
+
+	if (!dwc->paused)
+		return 0;
+
+	spin_lock_irqsave(&dwc->lock, flags);
+
+	dwc_chan_resume(dwc);
+
+	spin_unlock_irqrestore(&dwc->lock, flags);
+
+	return 0;
+}
+
+static int dwc_terminate_all(struct dma_chan *chan)
 {
 	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
 	struct dw_dma		*dw = to_dw_dma(chan->device);
@@ -1003,44 +1034,23 @@
 	unsigned long		flags;
 	LIST_HEAD(list);
 
-	if (cmd == DMA_PAUSE) {
-		spin_lock_irqsave(&dwc->lock, flags);
+	spin_lock_irqsave(&dwc->lock, flags);
 
-		dwc_chan_pause(dwc);
+	clear_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags);
 
-		spin_unlock_irqrestore(&dwc->lock, flags);
-	} else if (cmd == DMA_RESUME) {
-		if (!dwc->paused)
-			return 0;
+	dwc_chan_disable(dw, dwc);
 
-		spin_lock_irqsave(&dwc->lock, flags);
+	dwc_chan_resume(dwc);
 
-		dwc_chan_resume(dwc);
+	/* active_list entries will end up before queued entries */
+	list_splice_init(&dwc->queue, &list);
+	list_splice_init(&dwc->active_list, &list);
 
-		spin_unlock_irqrestore(&dwc->lock, flags);
-	} else if (cmd == DMA_TERMINATE_ALL) {
-		spin_lock_irqsave(&dwc->lock, flags);
+	spin_unlock_irqrestore(&dwc->lock, flags);
 
-		clear_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags);
-
-		dwc_chan_disable(dw, dwc);
-
-		dwc_chan_resume(dwc);
-
-		/* active_list entries will end up before queued entries */
-		list_splice_init(&dwc->queue, &list);
-		list_splice_init(&dwc->active_list, &list);
-
-		spin_unlock_irqrestore(&dwc->lock, flags);
-
-		/* Flush all pending and queued descriptors */
-		list_for_each_entry_safe(desc, _desc, &list, desc_node)
-			dwc_descriptor_complete(dwc, desc, false);
-	} else if (cmd == DMA_SLAVE_CONFIG) {
-		return set_runtime_config(chan, (struct dma_slave_config *)arg);
-	} else {
-		return -ENXIO;
-	}
+	/* Flush all pending and queued descriptors */
+	list_for_each_entry_safe(desc, _desc, &list, desc_node)
+		dwc_descriptor_complete(dwc, desc, false);
 
 	return 0;
 }
@@ -1551,7 +1561,8 @@
 		}
 	} else {
 		dw->nr_masters = pdata->nr_masters;
-		memcpy(dw->data_width, pdata->data_width, 4);
+		for (i = 0; i < dw->nr_masters; i++)
+			dw->data_width[i] = pdata->data_width[i];
 	}
 
 	/* Calculate all channel mask before DMA setup */
@@ -1656,13 +1667,23 @@
 	dw->dma.device_free_chan_resources = dwc_free_chan_resources;
 
 	dw->dma.device_prep_dma_memcpy = dwc_prep_dma_memcpy;
-
 	dw->dma.device_prep_slave_sg = dwc_prep_slave_sg;
-	dw->dma.device_control = dwc_control;
+
+	dw->dma.device_config = dwc_config;
+	dw->dma.device_pause = dwc_pause;
+	dw->dma.device_resume = dwc_resume;
+	dw->dma.device_terminate_all = dwc_terminate_all;
 
 	dw->dma.device_tx_status = dwc_tx_status;
 	dw->dma.device_issue_pending = dwc_issue_pending;
 
+	/* DMA capabilities */
+	dw->dma.src_addr_widths = DW_DMA_BUSWIDTHS;
+	dw->dma.dst_addr_widths = DW_DMA_BUSWIDTHS;
+	dw->dma.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV) |
+			     BIT(DMA_MEM_TO_MEM);
+	dw->dma.residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+
 	err = dma_async_device_register(&dw->dma);
 	if (err)
 		goto err_dma_register;
diff --git a/drivers/dma/dw/platform.c b/drivers/dma/dw/platform.c
index 32ea1ac..6565a36 100644
--- a/drivers/dma/dw/platform.c
+++ b/drivers/dma/dw/platform.c
@@ -100,7 +100,7 @@
 {
 	struct device_node *np = pdev->dev.of_node;
 	struct dw_dma_platform_data *pdata;
-	u32 tmp, arr[4];
+	u32 tmp, arr[DW_DMA_MAX_NR_MASTERS];
 
 	if (!np) {
 		dev_err(&pdev->dev, "Missing DT data\n");
@@ -127,7 +127,7 @@
 		pdata->block_size = tmp;
 
 	if (!of_property_read_u32(np, "dma-masters", &tmp)) {
-		if (tmp > 4)
+		if (tmp > DW_DMA_MAX_NR_MASTERS)
 			return NULL;
 
 		pdata->nr_masters = tmp;
diff --git a/drivers/dma/dw/regs.h b/drivers/dma/dw/regs.h
index 848e232..241ff2b 100644
--- a/drivers/dma/dw/regs.h
+++ b/drivers/dma/dw/regs.h
@@ -252,7 +252,7 @@
 	u8			src_master;
 	u8			dst_master;
 
-	/* configuration passed via DMA_SLAVE_CONFIG */
+	/* configuration passed via .device_config */
 	struct dma_slave_config dma_sconfig;
 };
 
@@ -285,7 +285,7 @@
 
 	/* hardware configuration */
 	unsigned char		nr_masters;
-	unsigned char		data_width[4];
+	unsigned char		data_width[DW_DMA_MAX_NR_MASTERS];
 };
 
 static inline struct dw_dma_regs __iomem *__dw_regs(struct dw_dma *dw)
diff --git a/drivers/dma/edma.c b/drivers/dma/edma.c
index b969206..276157f 100644
--- a/drivers/dma/edma.c
+++ b/drivers/dma/edma.c
@@ -15,6 +15,7 @@
 
 #include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
+#include <linux/edma.h>
 #include <linux/err.h>
 #include <linux/init.h>
 #include <linux/interrupt.h>
@@ -244,8 +245,9 @@
 	}
 }
 
-static int edma_terminate_all(struct edma_chan *echan)
+static int edma_terminate_all(struct dma_chan *chan)
 {
+	struct edma_chan *echan = to_edma_chan(chan);
 	unsigned long flags;
 	LIST_HEAD(head);
 
@@ -273,9 +275,11 @@
 	return 0;
 }
 
-static int edma_slave_config(struct edma_chan *echan,
+static int edma_slave_config(struct dma_chan *chan,
 	struct dma_slave_config *cfg)
 {
+	struct edma_chan *echan = to_edma_chan(chan);
+
 	if (cfg->src_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES ||
 	    cfg->dst_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES)
 		return -EINVAL;
@@ -285,8 +289,10 @@
 	return 0;
 }
 
-static int edma_dma_pause(struct edma_chan *echan)
+static int edma_dma_pause(struct dma_chan *chan)
 {
+	struct edma_chan *echan = to_edma_chan(chan);
+
 	/* Pause/Resume only allowed with cyclic mode */
 	if (!echan->edesc || !echan->edesc->cyclic)
 		return -EINVAL;
@@ -295,8 +301,10 @@
 	return 0;
 }
 
-static int edma_dma_resume(struct edma_chan *echan)
+static int edma_dma_resume(struct dma_chan *chan)
 {
+	struct edma_chan *echan = to_edma_chan(chan);
+
 	/* Pause/Resume only allowed with cyclic mode */
 	if (!echan->edesc->cyclic)
 		return -EINVAL;
@@ -305,36 +313,6 @@
 	return 0;
 }
 
-static int edma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-			unsigned long arg)
-{
-	int ret = 0;
-	struct dma_slave_config *config;
-	struct edma_chan *echan = to_edma_chan(chan);
-
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		edma_terminate_all(echan);
-		break;
-	case DMA_SLAVE_CONFIG:
-		config = (struct dma_slave_config *)arg;
-		ret = edma_slave_config(echan, config);
-		break;
-	case DMA_PAUSE:
-		ret = edma_dma_pause(echan);
-		break;
-
-	case DMA_RESUME:
-		ret = edma_dma_resume(echan);
-		break;
-
-	default:
-		ret = -ENOSYS;
-	}
-
-	return ret;
-}
-
 /*
  * A PaRAM set configuration abstraction used by other modes
  * @chan: Channel who's PaRAM set we're configuring
@@ -557,7 +535,7 @@
 	return vchan_tx_prep(&echan->vchan, &edesc->vdesc, tx_flags);
 }
 
-struct dma_async_tx_descriptor *edma_prep_dma_memcpy(
+static struct dma_async_tx_descriptor *edma_prep_dma_memcpy(
 	struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 	size_t len, unsigned long tx_flags)
 {
@@ -994,19 +972,6 @@
 				 BIT(DMA_SLAVE_BUSWIDTH_3_BYTES) | \
 				 BIT(DMA_SLAVE_BUSWIDTH_4_BYTES))
 
-static int edma_dma_device_slave_caps(struct dma_chan *dchan,
-				      struct dma_slave_caps *caps)
-{
-	caps->src_addr_widths = EDMA_DMA_BUSWIDTHS;
-	caps->dstn_addr_widths = EDMA_DMA_BUSWIDTHS;
-	caps->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
-	caps->cmd_pause = true;
-	caps->cmd_terminate = true;
-	caps->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
-
-	return 0;
-}
-
 static void edma_dma_init(struct edma_cc *ecc, struct dma_device *dma,
 			  struct device *dev)
 {
@@ -1017,8 +982,16 @@
 	dma->device_free_chan_resources = edma_free_chan_resources;
 	dma->device_issue_pending = edma_issue_pending;
 	dma->device_tx_status = edma_tx_status;
-	dma->device_control = edma_control;
-	dma->device_slave_caps = edma_dma_device_slave_caps;
+	dma->device_config = edma_slave_config;
+	dma->device_pause = edma_dma_pause;
+	dma->device_resume = edma_dma_resume;
+	dma->device_terminate_all = edma_terminate_all;
+
+	dma->src_addr_widths = EDMA_DMA_BUSWIDTHS;
+	dma->dst_addr_widths = EDMA_DMA_BUSWIDTHS;
+	dma->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	dma->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+
 	dma->dev = dev;
 
 	/*
diff --git a/drivers/dma/ep93xx_dma.c b/drivers/dma/ep93xx_dma.c
index 7650470..24e5290 100644
--- a/drivers/dma/ep93xx_dma.c
+++ b/drivers/dma/ep93xx_dma.c
@@ -144,7 +144,7 @@
  * @queue: pending descriptors which are handled next
  * @free_list: list of free descriptors which can be used
  * @runtime_addr: physical address currently used as dest/src (M2M only). This
- *                is set via %DMA_SLAVE_CONFIG before slave operation is
+ *                is set via .device_config before slave operation is
  *                prepared
  * @runtime_ctrl: M2M runtime values for the control register.
  *
@@ -1164,13 +1164,14 @@
 
 /**
  * ep93xx_dma_terminate_all - terminate all transactions
- * @edmac: channel
+ * @chan: channel
  *
  * Stops all DMA transactions. All descriptors are put back to the
  * @edmac->free_list and callbacks are _not_ called.
  */
-static int ep93xx_dma_terminate_all(struct ep93xx_dma_chan *edmac)
+static int ep93xx_dma_terminate_all(struct dma_chan *chan)
 {
+	struct ep93xx_dma_chan *edmac = to_ep93xx_dma_chan(chan);
 	struct ep93xx_dma_desc *desc, *_d;
 	unsigned long flags;
 	LIST_HEAD(list);
@@ -1194,9 +1195,10 @@
 	return 0;
 }
 
-static int ep93xx_dma_slave_config(struct ep93xx_dma_chan *edmac,
+static int ep93xx_dma_slave_config(struct dma_chan *chan,
 				   struct dma_slave_config *config)
 {
+	struct ep93xx_dma_chan *edmac = to_ep93xx_dma_chan(chan);
 	enum dma_slave_buswidth width;
 	unsigned long flags;
 	u32 addr, ctrl;
@@ -1242,36 +1244,6 @@
 }
 
 /**
- * ep93xx_dma_control - manipulate all pending operations on a channel
- * @chan: channel
- * @cmd: control command to perform
- * @arg: optional argument
- *
- * Controls the channel. Function returns %0 in case of success or negative
- * error in case of failure.
- */
-static int ep93xx_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-			      unsigned long arg)
-{
-	struct ep93xx_dma_chan *edmac = to_ep93xx_dma_chan(chan);
-	struct dma_slave_config *config;
-
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		return ep93xx_dma_terminate_all(edmac);
-
-	case DMA_SLAVE_CONFIG:
-		config = (struct dma_slave_config *)arg;
-		return ep93xx_dma_slave_config(edmac, config);
-
-	default:
-		break;
-	}
-
-	return -ENOSYS;
-}
-
-/**
  * ep93xx_dma_tx_status - check if a transaction is completed
  * @chan: channel
  * @cookie: transaction specific cookie
@@ -1352,7 +1324,8 @@
 	dma_dev->device_free_chan_resources = ep93xx_dma_free_chan_resources;
 	dma_dev->device_prep_slave_sg = ep93xx_dma_prep_slave_sg;
 	dma_dev->device_prep_dma_cyclic = ep93xx_dma_prep_dma_cyclic;
-	dma_dev->device_control = ep93xx_dma_control;
+	dma_dev->device_config = ep93xx_dma_slave_config;
+	dma_dev->device_terminate_all = ep93xx_dma_terminate_all;
 	dma_dev->device_issue_pending = ep93xx_dma_issue_pending;
 	dma_dev->device_tx_status = ep93xx_dma_tx_status;
 
diff --git a/drivers/dma/fsl-edma.c b/drivers/dma/fsl-edma.c
index e9ebb89..09e2842 100644
--- a/drivers/dma/fsl-edma.c
+++ b/drivers/dma/fsl-edma.c
@@ -289,62 +289,69 @@
 	kfree(fsl_desc);
 }
 
-static int fsl_edma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-		unsigned long arg)
+static int fsl_edma_terminate_all(struct dma_chan *chan)
 {
 	struct fsl_edma_chan *fsl_chan = to_fsl_edma_chan(chan);
-	struct dma_slave_config *cfg = (void *)arg;
 	unsigned long flags;
 	LIST_HEAD(head);
 
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		spin_lock_irqsave(&fsl_chan->vchan.lock, flags);
+	spin_lock_irqsave(&fsl_chan->vchan.lock, flags);
+	fsl_edma_disable_request(fsl_chan);
+	fsl_chan->edesc = NULL;
+	vchan_get_all_descriptors(&fsl_chan->vchan, &head);
+	spin_unlock_irqrestore(&fsl_chan->vchan.lock, flags);
+	vchan_dma_desc_free_list(&fsl_chan->vchan, &head);
+	return 0;
+}
+
+static int fsl_edma_pause(struct dma_chan *chan)
+{
+	struct fsl_edma_chan *fsl_chan = to_fsl_edma_chan(chan);
+	unsigned long flags;
+
+	spin_lock_irqsave(&fsl_chan->vchan.lock, flags);
+	if (fsl_chan->edesc) {
 		fsl_edma_disable_request(fsl_chan);
-		fsl_chan->edesc = NULL;
-		vchan_get_all_descriptors(&fsl_chan->vchan, &head);
-		spin_unlock_irqrestore(&fsl_chan->vchan.lock, flags);
-		vchan_dma_desc_free_list(&fsl_chan->vchan, &head);
-		return 0;
-
-	case DMA_SLAVE_CONFIG:
-		fsl_chan->fsc.dir = cfg->direction;
-		if (cfg->direction == DMA_DEV_TO_MEM) {
-			fsl_chan->fsc.dev_addr = cfg->src_addr;
-			fsl_chan->fsc.addr_width = cfg->src_addr_width;
-			fsl_chan->fsc.burst = cfg->src_maxburst;
-			fsl_chan->fsc.attr = fsl_edma_get_tcd_attr(cfg->src_addr_width);
-		} else if (cfg->direction == DMA_MEM_TO_DEV) {
-			fsl_chan->fsc.dev_addr = cfg->dst_addr;
-			fsl_chan->fsc.addr_width = cfg->dst_addr_width;
-			fsl_chan->fsc.burst = cfg->dst_maxburst;
-			fsl_chan->fsc.attr = fsl_edma_get_tcd_attr(cfg->dst_addr_width);
-		} else {
-			return -EINVAL;
-		}
-		return 0;
-
-	case DMA_PAUSE:
-		spin_lock_irqsave(&fsl_chan->vchan.lock, flags);
-		if (fsl_chan->edesc) {
-			fsl_edma_disable_request(fsl_chan);
-			fsl_chan->status = DMA_PAUSED;
-		}
-		spin_unlock_irqrestore(&fsl_chan->vchan.lock, flags);
-		return 0;
-
-	case DMA_RESUME:
-		spin_lock_irqsave(&fsl_chan->vchan.lock, flags);
-		if (fsl_chan->edesc) {
-			fsl_edma_enable_request(fsl_chan);
-			fsl_chan->status = DMA_IN_PROGRESS;
-		}
-		spin_unlock_irqrestore(&fsl_chan->vchan.lock, flags);
-		return 0;
-
-	default:
-		return -ENXIO;
+		fsl_chan->status = DMA_PAUSED;
 	}
+	spin_unlock_irqrestore(&fsl_chan->vchan.lock, flags);
+	return 0;
+}
+
+static int fsl_edma_resume(struct dma_chan *chan)
+{
+	struct fsl_edma_chan *fsl_chan = to_fsl_edma_chan(chan);
+	unsigned long flags;
+
+	spin_lock_irqsave(&fsl_chan->vchan.lock, flags);
+	if (fsl_chan->edesc) {
+		fsl_edma_enable_request(fsl_chan);
+		fsl_chan->status = DMA_IN_PROGRESS;
+	}
+	spin_unlock_irqrestore(&fsl_chan->vchan.lock, flags);
+	return 0;
+}
+
+static int fsl_edma_slave_config(struct dma_chan *chan,
+				 struct dma_slave_config *cfg)
+{
+	struct fsl_edma_chan *fsl_chan = to_fsl_edma_chan(chan);
+
+	fsl_chan->fsc.dir = cfg->direction;
+	if (cfg->direction == DMA_DEV_TO_MEM) {
+		fsl_chan->fsc.dev_addr = cfg->src_addr;
+		fsl_chan->fsc.addr_width = cfg->src_addr_width;
+		fsl_chan->fsc.burst = cfg->src_maxburst;
+		fsl_chan->fsc.attr = fsl_edma_get_tcd_attr(cfg->src_addr_width);
+	} else if (cfg->direction == DMA_MEM_TO_DEV) {
+		fsl_chan->fsc.dev_addr = cfg->dst_addr;
+		fsl_chan->fsc.addr_width = cfg->dst_addr_width;
+		fsl_chan->fsc.burst = cfg->dst_maxburst;
+		fsl_chan->fsc.attr = fsl_edma_get_tcd_attr(cfg->dst_addr_width);
+	} else {
+			return -EINVAL;
+	}
+	return 0;
 }
 
 static size_t fsl_edma_desc_residue(struct fsl_edma_chan *fsl_chan,
@@ -780,18 +787,6 @@
 	fsl_chan->tcd_pool = NULL;
 }
 
-static int fsl_dma_device_slave_caps(struct dma_chan *dchan,
-		struct dma_slave_caps *caps)
-{
-	caps->src_addr_widths = FSL_EDMA_BUSWIDTHS;
-	caps->dstn_addr_widths = FSL_EDMA_BUSWIDTHS;
-	caps->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
-	caps->cmd_pause = true;
-	caps->cmd_terminate = true;
-
-	return 0;
-}
-
 static int
 fsl_edma_irq_init(struct platform_device *pdev, struct fsl_edma_engine *fsl_edma)
 {
@@ -917,9 +912,15 @@
 	fsl_edma->dma_dev.device_tx_status = fsl_edma_tx_status;
 	fsl_edma->dma_dev.device_prep_slave_sg = fsl_edma_prep_slave_sg;
 	fsl_edma->dma_dev.device_prep_dma_cyclic = fsl_edma_prep_dma_cyclic;
-	fsl_edma->dma_dev.device_control = fsl_edma_control;
+	fsl_edma->dma_dev.device_config = fsl_edma_slave_config;
+	fsl_edma->dma_dev.device_pause = fsl_edma_pause;
+	fsl_edma->dma_dev.device_resume = fsl_edma_resume;
+	fsl_edma->dma_dev.device_terminate_all = fsl_edma_terminate_all;
 	fsl_edma->dma_dev.device_issue_pending = fsl_edma_issue_pending;
-	fsl_edma->dma_dev.device_slave_caps = fsl_dma_device_slave_caps;
+
+	fsl_edma->dma_dev.src_addr_widths = FSL_EDMA_BUSWIDTHS;
+	fsl_edma->dma_dev.dst_addr_widths = FSL_EDMA_BUSWIDTHS;
+	fsl_edma->dma_dev.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
 
 	platform_set_drvdata(pdev, fsl_edma);
 
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index 38821cd..300f821 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -941,37 +941,33 @@
 	return NULL;
 }
 
-/**
- * fsl_dma_prep_slave_sg - prepare descriptors for a DMA_SLAVE transaction
- * @chan: DMA channel
- * @sgl: scatterlist to transfer to/from
- * @sg_len: number of entries in @scatterlist
- * @direction: DMA direction
- * @flags: DMAEngine flags
- * @context: transaction context (ignored)
- *
- * Prepare a set of descriptors for a DMA_SLAVE transaction. Following the
- * DMA_SLAVE API, this gets the device-specific information from the
- * chan->private variable.
- */
-static struct dma_async_tx_descriptor *fsl_dma_prep_slave_sg(
-	struct dma_chan *dchan, struct scatterlist *sgl, unsigned int sg_len,
-	enum dma_transfer_direction direction, unsigned long flags,
-	void *context)
+static int fsl_dma_device_terminate_all(struct dma_chan *dchan)
 {
-	/*
-	 * This operation is not supported on the Freescale DMA controller
-	 *
-	 * However, we need to provide the function pointer to allow the
-	 * device_control() method to work.
-	 */
-	return NULL;
+	struct fsldma_chan *chan;
+
+	if (!dchan)
+		return -EINVAL;
+
+	chan = to_fsl_chan(dchan);
+
+	spin_lock_bh(&chan->desc_lock);
+
+	/* Halt the DMA engine */
+	dma_halt(chan);
+
+	/* Remove and free all of the descriptors in the LD queue */
+	fsldma_free_desc_list(chan, &chan->ld_pending);
+	fsldma_free_desc_list(chan, &chan->ld_running);
+	fsldma_free_desc_list(chan, &chan->ld_completed);
+	chan->idle = true;
+
+	spin_unlock_bh(&chan->desc_lock);
+	return 0;
 }
 
-static int fsl_dma_device_control(struct dma_chan *dchan,
-				  enum dma_ctrl_cmd cmd, unsigned long arg)
+static int fsl_dma_device_config(struct dma_chan *dchan,
+				 struct dma_slave_config *config)
 {
-	struct dma_slave_config *config;
 	struct fsldma_chan *chan;
 	int size;
 
@@ -980,45 +976,21 @@
 
 	chan = to_fsl_chan(dchan);
 
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		spin_lock_bh(&chan->desc_lock);
-
-		/* Halt the DMA engine */
-		dma_halt(chan);
-
-		/* Remove and free all of the descriptors in the LD queue */
-		fsldma_free_desc_list(chan, &chan->ld_pending);
-		fsldma_free_desc_list(chan, &chan->ld_running);
-		fsldma_free_desc_list(chan, &chan->ld_completed);
-		chan->idle = true;
-
-		spin_unlock_bh(&chan->desc_lock);
-		return 0;
-
-	case DMA_SLAVE_CONFIG:
-		config = (struct dma_slave_config *)arg;
-
-		/* make sure the channel supports setting burst size */
-		if (!chan->set_request_count)
-			return -ENXIO;
-
-		/* we set the controller burst size depending on direction */
-		if (config->direction == DMA_MEM_TO_DEV)
-			size = config->dst_addr_width * config->dst_maxburst;
-		else
-			size = config->src_addr_width * config->src_maxburst;
-
-		chan->set_request_count(chan, size);
-		return 0;
-
-	default:
+	/* make sure the channel supports setting burst size */
+	if (!chan->set_request_count)
 		return -ENXIO;
-	}
 
+	/* we set the controller burst size depending on direction */
+	if (config->direction == DMA_MEM_TO_DEV)
+		size = config->dst_addr_width * config->dst_maxburst;
+	else
+		size = config->src_addr_width * config->src_maxburst;
+
+	chan->set_request_count(chan, size);
 	return 0;
 }
 
+
 /**
  * fsl_dma_memcpy_issue_pending - Issue the DMA start command
  * @chan : Freescale DMA channel
@@ -1395,10 +1367,15 @@
 	fdev->common.device_prep_dma_sg = fsl_dma_prep_sg;
 	fdev->common.device_tx_status = fsl_tx_status;
 	fdev->common.device_issue_pending = fsl_dma_memcpy_issue_pending;
-	fdev->common.device_prep_slave_sg = fsl_dma_prep_slave_sg;
-	fdev->common.device_control = fsl_dma_device_control;
+	fdev->common.device_config = fsl_dma_device_config;
+	fdev->common.device_terminate_all = fsl_dma_device_terminate_all;
 	fdev->common.dev = &op->dev;
 
+	fdev->common.src_addr_widths = FSL_DMA_BUSWIDTHS;
+	fdev->common.dst_addr_widths = FSL_DMA_BUSWIDTHS;
+	fdev->common.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	fdev->common.residue_granularity = DMA_RESIDUE_GRANULARITY_DESCRIPTOR;
+
 	dma_set_mask(&(op->dev), DMA_BIT_MASK(36));
 
 	platform_set_drvdata(op, fdev);
diff --git a/drivers/dma/fsldma.h b/drivers/dma/fsldma.h
index 239c20c..31bffcc 100644
--- a/drivers/dma/fsldma.h
+++ b/drivers/dma/fsldma.h
@@ -83,6 +83,10 @@
 #define FSL_DMA_DGSR_EOSI	0x02
 #define FSL_DMA_DGSR_EOLSI	0x01
 
+#define FSL_DMA_BUSWIDTHS	(BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) | \
+				BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) | \
+				BIT(DMA_SLAVE_BUSWIDTH_4_BYTES) | \
+				BIT(DMA_SLAVE_BUSWIDTH_8_BYTES))
 typedef u64 __bitwise v64;
 typedef u32 __bitwise v32;
 
diff --git a/drivers/dma/img-mdc-dma.c b/drivers/dma/img-mdc-dma.c
new file mode 100644
index 0000000..ed045a9
--- /dev/null
+++ b/drivers/dma/img-mdc-dma.c
@@ -0,0 +1,1011 @@
+/*
+ * IMG Multi-threaded DMA Controller (MDC)
+ *
+ * Copyright (C) 2009,2012,2013 Imagination Technologies Ltd.
+ * Copyright (C) 2014 Google, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/dmapool.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/kernel.h>
+#include <linux/mfd/syscon.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/of_dma.h>
+#include <linux/platform_device.h>
+#include <linux/regmap.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+
+#include "dmaengine.h"
+#include "virt-dma.h"
+
+#define MDC_MAX_DMA_CHANNELS			32
+
+#define MDC_GENERAL_CONFIG			0x000
+#define MDC_GENERAL_CONFIG_LIST_IEN		BIT(31)
+#define MDC_GENERAL_CONFIG_IEN			BIT(29)
+#define MDC_GENERAL_CONFIG_LEVEL_INT		BIT(28)
+#define MDC_GENERAL_CONFIG_INC_W		BIT(12)
+#define MDC_GENERAL_CONFIG_INC_R		BIT(8)
+#define MDC_GENERAL_CONFIG_PHYSICAL_W		BIT(7)
+#define MDC_GENERAL_CONFIG_WIDTH_W_SHIFT	4
+#define MDC_GENERAL_CONFIG_WIDTH_W_MASK		0x7
+#define MDC_GENERAL_CONFIG_PHYSICAL_R		BIT(3)
+#define MDC_GENERAL_CONFIG_WIDTH_R_SHIFT	0
+#define MDC_GENERAL_CONFIG_WIDTH_R_MASK		0x7
+
+#define MDC_READ_PORT_CONFIG			0x004
+#define MDC_READ_PORT_CONFIG_STHREAD_SHIFT	28
+#define MDC_READ_PORT_CONFIG_STHREAD_MASK	0xf
+#define MDC_READ_PORT_CONFIG_RTHREAD_SHIFT	24
+#define MDC_READ_PORT_CONFIG_RTHREAD_MASK	0xf
+#define MDC_READ_PORT_CONFIG_WTHREAD_SHIFT	16
+#define MDC_READ_PORT_CONFIG_WTHREAD_MASK	0xf
+#define MDC_READ_PORT_CONFIG_BURST_SIZE_SHIFT	4
+#define MDC_READ_PORT_CONFIG_BURST_SIZE_MASK	0xff
+#define MDC_READ_PORT_CONFIG_DREQ_ENABLE	BIT(1)
+
+#define MDC_READ_ADDRESS			0x008
+
+#define MDC_WRITE_ADDRESS			0x00c
+
+#define MDC_TRANSFER_SIZE			0x010
+#define MDC_TRANSFER_SIZE_MASK			0xffffff
+
+#define MDC_LIST_NODE_ADDRESS			0x014
+
+#define MDC_CMDS_PROCESSED			0x018
+#define MDC_CMDS_PROCESSED_CMDS_PROCESSED_SHIFT	16
+#define MDC_CMDS_PROCESSED_CMDS_PROCESSED_MASK	0x3f
+#define MDC_CMDS_PROCESSED_INT_ACTIVE		BIT(8)
+#define MDC_CMDS_PROCESSED_CMDS_DONE_SHIFT	0
+#define MDC_CMDS_PROCESSED_CMDS_DONE_MASK	0x3f
+
+#define MDC_CONTROL_AND_STATUS			0x01c
+#define MDC_CONTROL_AND_STATUS_CANCEL		BIT(20)
+#define MDC_CONTROL_AND_STATUS_LIST_EN		BIT(4)
+#define MDC_CONTROL_AND_STATUS_EN		BIT(0)
+
+#define MDC_ACTIVE_TRANSFER_SIZE		0x030
+
+#define MDC_GLOBAL_CONFIG_A				0x900
+#define MDC_GLOBAL_CONFIG_A_THREAD_ID_WIDTH_SHIFT	16
+#define MDC_GLOBAL_CONFIG_A_THREAD_ID_WIDTH_MASK	0xff
+#define MDC_GLOBAL_CONFIG_A_DMA_CONTEXTS_SHIFT		8
+#define MDC_GLOBAL_CONFIG_A_DMA_CONTEXTS_MASK		0xff
+#define MDC_GLOBAL_CONFIG_A_SYS_DAT_WIDTH_SHIFT		0
+#define MDC_GLOBAL_CONFIG_A_SYS_DAT_WIDTH_MASK		0xff
+
+struct mdc_hw_list_desc {
+	u32 gen_conf;
+	u32 readport_conf;
+	u32 read_addr;
+	u32 write_addr;
+	u32 xfer_size;
+	u32 node_addr;
+	u32 cmds_done;
+	u32 ctrl_status;
+	/*
+	 * Not part of the list descriptor, but instead used by the CPU to
+	 * traverse the list.
+	 */
+	struct mdc_hw_list_desc *next_desc;
+};
+
+struct mdc_tx_desc {
+	struct mdc_chan *chan;
+	struct virt_dma_desc vd;
+	dma_addr_t list_phys;
+	struct mdc_hw_list_desc *list;
+	bool cyclic;
+	bool cmd_loaded;
+	unsigned int list_len;
+	unsigned int list_period_len;
+	size_t list_xfer_size;
+	unsigned int list_cmds_done;
+};
+
+struct mdc_chan {
+	struct mdc_dma *mdma;
+	struct virt_dma_chan vc;
+	struct dma_slave_config config;
+	struct mdc_tx_desc *desc;
+	int irq;
+	unsigned int periph;
+	unsigned int thread;
+	unsigned int chan_nr;
+};
+
+struct mdc_dma_soc_data {
+	void (*enable_chan)(struct mdc_chan *mchan);
+	void (*disable_chan)(struct mdc_chan *mchan);
+};
+
+struct mdc_dma {
+	struct dma_device dma_dev;
+	void __iomem *regs;
+	struct clk *clk;
+	struct dma_pool *desc_pool;
+	struct regmap *periph_regs;
+	spinlock_t lock;
+	unsigned int nr_threads;
+	unsigned int nr_channels;
+	unsigned int bus_width;
+	unsigned int max_burst_mult;
+	unsigned int max_xfer_size;
+	const struct mdc_dma_soc_data *soc;
+	struct mdc_chan channels[MDC_MAX_DMA_CHANNELS];
+};
+
+static inline u32 mdc_readl(struct mdc_dma *mdma, u32 reg)
+{
+	return readl(mdma->regs + reg);
+}
+
+static inline void mdc_writel(struct mdc_dma *mdma, u32 val, u32 reg)
+{
+	writel(val, mdma->regs + reg);
+}
+
+static inline u32 mdc_chan_readl(struct mdc_chan *mchan, u32 reg)
+{
+	return mdc_readl(mchan->mdma, mchan->chan_nr * 0x040 + reg);
+}
+
+static inline void mdc_chan_writel(struct mdc_chan *mchan, u32 val, u32 reg)
+{
+	mdc_writel(mchan->mdma, val, mchan->chan_nr * 0x040 + reg);
+}
+
+static inline struct mdc_chan *to_mdc_chan(struct dma_chan *c)
+{
+	return container_of(to_virt_chan(c), struct mdc_chan, vc);
+}
+
+static inline struct mdc_tx_desc *to_mdc_desc(struct dma_async_tx_descriptor *t)
+{
+	struct virt_dma_desc *vdesc = container_of(t, struct virt_dma_desc, tx);
+
+	return container_of(vdesc, struct mdc_tx_desc, vd);
+}
+
+static inline struct device *mdma2dev(struct mdc_dma *mdma)
+{
+	return mdma->dma_dev.dev;
+}
+
+static inline unsigned int to_mdc_width(unsigned int bytes)
+{
+	return ffs(bytes) - 1;
+}
+
+static inline void mdc_set_read_width(struct mdc_hw_list_desc *ldesc,
+				      unsigned int bytes)
+{
+	ldesc->gen_conf |= to_mdc_width(bytes) <<
+		MDC_GENERAL_CONFIG_WIDTH_R_SHIFT;
+}
+
+static inline void mdc_set_write_width(struct mdc_hw_list_desc *ldesc,
+				       unsigned int bytes)
+{
+	ldesc->gen_conf |= to_mdc_width(bytes) <<
+		MDC_GENERAL_CONFIG_WIDTH_W_SHIFT;
+}
+
+static void mdc_list_desc_config(struct mdc_chan *mchan,
+				 struct mdc_hw_list_desc *ldesc,
+				 enum dma_transfer_direction dir,
+				 dma_addr_t src, dma_addr_t dst, size_t len)
+{
+	struct mdc_dma *mdma = mchan->mdma;
+	unsigned int max_burst, burst_size;
+
+	ldesc->gen_conf = MDC_GENERAL_CONFIG_IEN | MDC_GENERAL_CONFIG_LIST_IEN |
+		MDC_GENERAL_CONFIG_LEVEL_INT | MDC_GENERAL_CONFIG_PHYSICAL_W |
+		MDC_GENERAL_CONFIG_PHYSICAL_R;
+	ldesc->readport_conf =
+		(mchan->thread << MDC_READ_PORT_CONFIG_STHREAD_SHIFT) |
+		(mchan->thread << MDC_READ_PORT_CONFIG_RTHREAD_SHIFT) |
+		(mchan->thread << MDC_READ_PORT_CONFIG_WTHREAD_SHIFT);
+	ldesc->read_addr = src;
+	ldesc->write_addr = dst;
+	ldesc->xfer_size = len - 1;
+	ldesc->node_addr = 0;
+	ldesc->cmds_done = 0;
+	ldesc->ctrl_status = MDC_CONTROL_AND_STATUS_LIST_EN |
+		MDC_CONTROL_AND_STATUS_EN;
+	ldesc->next_desc = NULL;
+
+	if (IS_ALIGNED(dst, mdma->bus_width) &&
+	    IS_ALIGNED(src, mdma->bus_width))
+		max_burst = mdma->bus_width * mdma->max_burst_mult;
+	else
+		max_burst = mdma->bus_width * (mdma->max_burst_mult - 1);
+
+	if (dir == DMA_MEM_TO_DEV) {
+		ldesc->gen_conf |= MDC_GENERAL_CONFIG_INC_R;
+		ldesc->readport_conf |= MDC_READ_PORT_CONFIG_DREQ_ENABLE;
+		mdc_set_read_width(ldesc, mdma->bus_width);
+		mdc_set_write_width(ldesc, mchan->config.dst_addr_width);
+		burst_size = min(max_burst, mchan->config.dst_maxburst *
+				 mchan->config.dst_addr_width);
+	} else if (dir == DMA_DEV_TO_MEM) {
+		ldesc->gen_conf |= MDC_GENERAL_CONFIG_INC_W;
+		ldesc->readport_conf |= MDC_READ_PORT_CONFIG_DREQ_ENABLE;
+		mdc_set_read_width(ldesc, mchan->config.src_addr_width);
+		mdc_set_write_width(ldesc, mdma->bus_width);
+		burst_size = min(max_burst, mchan->config.src_maxburst *
+				 mchan->config.src_addr_width);
+	} else {
+		ldesc->gen_conf |= MDC_GENERAL_CONFIG_INC_R |
+			MDC_GENERAL_CONFIG_INC_W;
+		mdc_set_read_width(ldesc, mdma->bus_width);
+		mdc_set_write_width(ldesc, mdma->bus_width);
+		burst_size = max_burst;
+	}
+	ldesc->readport_conf |= (burst_size - 1) <<
+		MDC_READ_PORT_CONFIG_BURST_SIZE_SHIFT;
+}
+
+static void mdc_list_desc_free(struct mdc_tx_desc *mdesc)
+{
+	struct mdc_dma *mdma = mdesc->chan->mdma;
+	struct mdc_hw_list_desc *curr, *next;
+	dma_addr_t curr_phys, next_phys;
+
+	curr = mdesc->list;
+	curr_phys = mdesc->list_phys;
+	while (curr) {
+		next = curr->next_desc;
+		next_phys = curr->node_addr;
+		dma_pool_free(mdma->desc_pool, curr, curr_phys);
+		curr = next;
+		curr_phys = next_phys;
+	}
+}
+
+static void mdc_desc_free(struct virt_dma_desc *vd)
+{
+	struct mdc_tx_desc *mdesc = to_mdc_desc(&vd->tx);
+
+	mdc_list_desc_free(mdesc);
+	kfree(mdesc);
+}
+
+static struct dma_async_tx_descriptor *mdc_prep_dma_memcpy(
+	struct dma_chan *chan, dma_addr_t dest, dma_addr_t src, size_t len,
+	unsigned long flags)
+{
+	struct mdc_chan *mchan = to_mdc_chan(chan);
+	struct mdc_dma *mdma = mchan->mdma;
+	struct mdc_tx_desc *mdesc;
+	struct mdc_hw_list_desc *curr, *prev = NULL;
+	dma_addr_t curr_phys, prev_phys;
+
+	if (!len)
+		return NULL;
+
+	mdesc = kzalloc(sizeof(*mdesc), GFP_NOWAIT);
+	if (!mdesc)
+		return NULL;
+	mdesc->chan = mchan;
+	mdesc->list_xfer_size = len;
+
+	while (len > 0) {
+		size_t xfer_size;
+
+		curr = dma_pool_alloc(mdma->desc_pool, GFP_NOWAIT, &curr_phys);
+		if (!curr)
+			goto free_desc;
+
+		if (prev) {
+			prev->node_addr = curr_phys;
+			prev->next_desc = curr;
+		} else {
+			mdesc->list_phys = curr_phys;
+			mdesc->list = curr;
+		}
+
+		xfer_size = min_t(size_t, mdma->max_xfer_size, len);
+
+		mdc_list_desc_config(mchan, curr, DMA_MEM_TO_MEM, src, dest,
+				     xfer_size);
+
+		prev = curr;
+		prev_phys = curr_phys;
+
+		mdesc->list_len++;
+		src += xfer_size;
+		dest += xfer_size;
+		len -= xfer_size;
+	}
+
+	return vchan_tx_prep(&mchan->vc, &mdesc->vd, flags);
+
+free_desc:
+	mdc_desc_free(&mdesc->vd);
+
+	return NULL;
+}
+
+static int mdc_check_slave_width(struct mdc_chan *mchan,
+				 enum dma_transfer_direction dir)
+{
+	enum dma_slave_buswidth width;
+
+	if (dir == DMA_MEM_TO_DEV)
+		width = mchan->config.dst_addr_width;
+	else
+		width = mchan->config.src_addr_width;
+
+	switch (width) {
+	case DMA_SLAVE_BUSWIDTH_1_BYTE:
+	case DMA_SLAVE_BUSWIDTH_2_BYTES:
+	case DMA_SLAVE_BUSWIDTH_4_BYTES:
+	case DMA_SLAVE_BUSWIDTH_8_BYTES:
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (width > mchan->mdma->bus_width)
+		return -EINVAL;
+
+	return 0;
+}
+
+static struct dma_async_tx_descriptor *mdc_prep_dma_cyclic(
+	struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
+	size_t period_len, enum dma_transfer_direction dir,
+	unsigned long flags)
+{
+	struct mdc_chan *mchan = to_mdc_chan(chan);
+	struct mdc_dma *mdma = mchan->mdma;
+	struct mdc_tx_desc *mdesc;
+	struct mdc_hw_list_desc *curr, *prev = NULL;
+	dma_addr_t curr_phys, prev_phys;
+
+	if (!buf_len && !period_len)
+		return NULL;
+
+	if (!is_slave_direction(dir))
+		return NULL;
+
+	if (mdc_check_slave_width(mchan, dir) < 0)
+		return NULL;
+
+	mdesc = kzalloc(sizeof(*mdesc), GFP_NOWAIT);
+	if (!mdesc)
+		return NULL;
+	mdesc->chan = mchan;
+	mdesc->cyclic = true;
+	mdesc->list_xfer_size = buf_len;
+	mdesc->list_period_len = DIV_ROUND_UP(period_len,
+					      mdma->max_xfer_size);
+
+	while (buf_len > 0) {
+		size_t remainder = min(period_len, buf_len);
+
+		while (remainder > 0) {
+			size_t xfer_size;
+
+			curr = dma_pool_alloc(mdma->desc_pool, GFP_NOWAIT,
+					      &curr_phys);
+			if (!curr)
+				goto free_desc;
+
+			if (!prev) {
+				mdesc->list_phys = curr_phys;
+				mdesc->list = curr;
+			} else {
+				prev->node_addr = curr_phys;
+				prev->next_desc = curr;
+			}
+
+			xfer_size = min_t(size_t, mdma->max_xfer_size,
+					  remainder);
+
+			if (dir == DMA_MEM_TO_DEV) {
+				mdc_list_desc_config(mchan, curr, dir,
+						     buf_addr,
+						     mchan->config.dst_addr,
+						     xfer_size);
+			} else {
+				mdc_list_desc_config(mchan, curr, dir,
+						     mchan->config.src_addr,
+						     buf_addr,
+						     xfer_size);
+			}
+
+			prev = curr;
+			prev_phys = curr_phys;
+
+			mdesc->list_len++;
+			buf_addr += xfer_size;
+			buf_len -= xfer_size;
+			remainder -= xfer_size;
+		}
+	}
+	prev->node_addr = mdesc->list_phys;
+
+	return vchan_tx_prep(&mchan->vc, &mdesc->vd, flags);
+
+free_desc:
+	mdc_desc_free(&mdesc->vd);
+
+	return NULL;
+}
+
+static struct dma_async_tx_descriptor *mdc_prep_slave_sg(
+	struct dma_chan *chan, struct scatterlist *sgl,
+	unsigned int sg_len, enum dma_transfer_direction dir,
+	unsigned long flags, void *context)
+{
+	struct mdc_chan *mchan = to_mdc_chan(chan);
+	struct mdc_dma *mdma = mchan->mdma;
+	struct mdc_tx_desc *mdesc;
+	struct scatterlist *sg;
+	struct mdc_hw_list_desc *curr, *prev = NULL;
+	dma_addr_t curr_phys, prev_phys;
+	unsigned int i;
+
+	if (!sgl)
+		return NULL;
+
+	if (!is_slave_direction(dir))
+		return NULL;
+
+	if (mdc_check_slave_width(mchan, dir) < 0)
+		return NULL;
+
+	mdesc = kzalloc(sizeof(*mdesc), GFP_NOWAIT);
+	if (!mdesc)
+		return NULL;
+	mdesc->chan = mchan;
+
+	for_each_sg(sgl, sg, sg_len, i) {
+		dma_addr_t buf = sg_dma_address(sg);
+		size_t buf_len = sg_dma_len(sg);
+
+		while (buf_len > 0) {
+			size_t xfer_size;
+
+			curr = dma_pool_alloc(mdma->desc_pool, GFP_NOWAIT,
+					      &curr_phys);
+			if (!curr)
+				goto free_desc;
+
+			if (!prev) {
+				mdesc->list_phys = curr_phys;
+				mdesc->list = curr;
+			} else {
+				prev->node_addr = curr_phys;
+				prev->next_desc = curr;
+			}
+
+			xfer_size = min_t(size_t, mdma->max_xfer_size,
+					  buf_len);
+
+			if (dir == DMA_MEM_TO_DEV) {
+				mdc_list_desc_config(mchan, curr, dir, buf,
+						     mchan->config.dst_addr,
+						     xfer_size);
+			} else {
+				mdc_list_desc_config(mchan, curr, dir,
+						     mchan->config.src_addr,
+						     buf, xfer_size);
+			}
+
+			prev = curr;
+			prev_phys = curr_phys;
+
+			mdesc->list_len++;
+			mdesc->list_xfer_size += xfer_size;
+			buf += xfer_size;
+			buf_len -= xfer_size;
+		}
+	}
+
+	return vchan_tx_prep(&mchan->vc, &mdesc->vd, flags);
+
+free_desc:
+	mdc_desc_free(&mdesc->vd);
+
+	return NULL;
+}
+
+static void mdc_issue_desc(struct mdc_chan *mchan)
+{
+	struct mdc_dma *mdma = mchan->mdma;
+	struct virt_dma_desc *vd;
+	struct mdc_tx_desc *mdesc;
+	u32 val;
+
+	vd = vchan_next_desc(&mchan->vc);
+	if (!vd)
+		return;
+
+	list_del(&vd->node);
+
+	mdesc = to_mdc_desc(&vd->tx);
+	mchan->desc = mdesc;
+
+	dev_dbg(mdma2dev(mdma), "Issuing descriptor on channel %d\n",
+		mchan->chan_nr);
+
+	mdma->soc->enable_chan(mchan);
+
+	val = mdc_chan_readl(mchan, MDC_GENERAL_CONFIG);
+	val |= MDC_GENERAL_CONFIG_LIST_IEN | MDC_GENERAL_CONFIG_IEN |
+		MDC_GENERAL_CONFIG_LEVEL_INT | MDC_GENERAL_CONFIG_PHYSICAL_W |
+		MDC_GENERAL_CONFIG_PHYSICAL_R;
+	mdc_chan_writel(mchan, val, MDC_GENERAL_CONFIG);
+	val = (mchan->thread << MDC_READ_PORT_CONFIG_STHREAD_SHIFT) |
+		(mchan->thread << MDC_READ_PORT_CONFIG_RTHREAD_SHIFT) |
+		(mchan->thread << MDC_READ_PORT_CONFIG_WTHREAD_SHIFT);
+	mdc_chan_writel(mchan, val, MDC_READ_PORT_CONFIG);
+	mdc_chan_writel(mchan, mdesc->list_phys, MDC_LIST_NODE_ADDRESS);
+	val = mdc_chan_readl(mchan, MDC_CONTROL_AND_STATUS);
+	val |= MDC_CONTROL_AND_STATUS_LIST_EN;
+	mdc_chan_writel(mchan, val, MDC_CONTROL_AND_STATUS);
+}
+
+static void mdc_issue_pending(struct dma_chan *chan)
+{
+	struct mdc_chan *mchan = to_mdc_chan(chan);
+	unsigned long flags;
+
+	spin_lock_irqsave(&mchan->vc.lock, flags);
+	if (vchan_issue_pending(&mchan->vc) && !mchan->desc)
+		mdc_issue_desc(mchan);
+	spin_unlock_irqrestore(&mchan->vc.lock, flags);
+}
+
+static enum dma_status mdc_tx_status(struct dma_chan *chan,
+	dma_cookie_t cookie, struct dma_tx_state *txstate)
+{
+	struct mdc_chan *mchan = to_mdc_chan(chan);
+	struct mdc_tx_desc *mdesc;
+	struct virt_dma_desc *vd;
+	unsigned long flags;
+	size_t bytes = 0;
+	int ret;
+
+	ret = dma_cookie_status(chan, cookie, txstate);
+	if (ret == DMA_COMPLETE)
+		return ret;
+
+	if (!txstate)
+		return ret;
+
+	spin_lock_irqsave(&mchan->vc.lock, flags);
+	vd = vchan_find_desc(&mchan->vc, cookie);
+	if (vd) {
+		mdesc = to_mdc_desc(&vd->tx);
+		bytes = mdesc->list_xfer_size;
+	} else if (mchan->desc && mchan->desc->vd.tx.cookie == cookie) {
+		struct mdc_hw_list_desc *ldesc;
+		u32 val1, val2, done, processed, residue;
+		int i, cmds;
+
+		mdesc = mchan->desc;
+
+		/*
+		 * Determine the number of commands that haven't been
+		 * processed (handled by the IRQ handler) yet.
+		 */
+		do {
+			val1 = mdc_chan_readl(mchan, MDC_CMDS_PROCESSED) &
+				~MDC_CMDS_PROCESSED_INT_ACTIVE;
+			residue = mdc_chan_readl(mchan,
+						 MDC_ACTIVE_TRANSFER_SIZE);
+			val2 = mdc_chan_readl(mchan, MDC_CMDS_PROCESSED) &
+				~MDC_CMDS_PROCESSED_INT_ACTIVE;
+		} while (val1 != val2);
+
+		done = (val1 >> MDC_CMDS_PROCESSED_CMDS_DONE_SHIFT) &
+			MDC_CMDS_PROCESSED_CMDS_DONE_MASK;
+		processed = (val1 >> MDC_CMDS_PROCESSED_CMDS_PROCESSED_SHIFT) &
+			MDC_CMDS_PROCESSED_CMDS_PROCESSED_MASK;
+		cmds = (done - processed) %
+			(MDC_CMDS_PROCESSED_CMDS_DONE_MASK + 1);
+
+		/*
+		 * If the command loaded event hasn't been processed yet, then
+		 * the difference above includes an extra command.
+		 */
+		if (!mdesc->cmd_loaded)
+			cmds--;
+		else
+			cmds += mdesc->list_cmds_done;
+
+		bytes = mdesc->list_xfer_size;
+		ldesc = mdesc->list;
+		for (i = 0; i < cmds; i++) {
+			bytes -= ldesc->xfer_size + 1;
+			ldesc = ldesc->next_desc;
+		}
+		if (ldesc) {
+			if (residue != MDC_TRANSFER_SIZE_MASK)
+				bytes -= ldesc->xfer_size - residue;
+			else
+				bytes -= ldesc->xfer_size + 1;
+		}
+	}
+	spin_unlock_irqrestore(&mchan->vc.lock, flags);
+
+	dma_set_residue(txstate, bytes);
+
+	return ret;
+}
+
+static int mdc_terminate_all(struct dma_chan *chan)
+{
+	struct mdc_chan *mchan = to_mdc_chan(chan);
+	struct mdc_tx_desc *mdesc;
+	unsigned long flags;
+	LIST_HEAD(head);
+
+	spin_lock_irqsave(&mchan->vc.lock, flags);
+
+	mdc_chan_writel(mchan, MDC_CONTROL_AND_STATUS_CANCEL,
+			MDC_CONTROL_AND_STATUS);
+
+	mdesc = mchan->desc;
+	mchan->desc = NULL;
+	vchan_get_all_descriptors(&mchan->vc, &head);
+
+	spin_unlock_irqrestore(&mchan->vc.lock, flags);
+
+	if (mdesc)
+		mdc_desc_free(&mdesc->vd);
+	vchan_dma_desc_free_list(&mchan->vc, &head);
+
+	return 0;
+}
+
+static int mdc_slave_config(struct dma_chan *chan,
+			    struct dma_slave_config *config)
+{
+	struct mdc_chan *mchan = to_mdc_chan(chan);
+	unsigned long flags;
+
+	spin_lock_irqsave(&mchan->vc.lock, flags);
+	mchan->config = *config;
+	spin_unlock_irqrestore(&mchan->vc.lock, flags);
+
+	return 0;
+}
+
+static int mdc_alloc_chan_resources(struct dma_chan *chan)
+{
+	return 0;
+}
+
+static void mdc_free_chan_resources(struct dma_chan *chan)
+{
+	struct mdc_chan *mchan = to_mdc_chan(chan);
+	struct mdc_dma *mdma = mchan->mdma;
+
+	mdc_terminate_all(chan);
+
+	mdma->soc->disable_chan(mchan);
+}
+
+static irqreturn_t mdc_chan_irq(int irq, void *dev_id)
+{
+	struct mdc_chan *mchan = (struct mdc_chan *)dev_id;
+	struct mdc_tx_desc *mdesc;
+	u32 val, processed, done1, done2;
+	unsigned int i;
+
+	spin_lock(&mchan->vc.lock);
+
+	val = mdc_chan_readl(mchan, MDC_CMDS_PROCESSED);
+	processed = (val >> MDC_CMDS_PROCESSED_CMDS_PROCESSED_SHIFT) &
+		MDC_CMDS_PROCESSED_CMDS_PROCESSED_MASK;
+	/*
+	 * CMDS_DONE may have incremented between reading CMDS_PROCESSED
+	 * and clearing INT_ACTIVE.  Re-read CMDS_PROCESSED to ensure we
+	 * didn't miss a command completion.
+	 */
+	do {
+		val = mdc_chan_readl(mchan, MDC_CMDS_PROCESSED);
+		done1 = (val >> MDC_CMDS_PROCESSED_CMDS_DONE_SHIFT) &
+			MDC_CMDS_PROCESSED_CMDS_DONE_MASK;
+		val &= ~((MDC_CMDS_PROCESSED_CMDS_PROCESSED_MASK <<
+			  MDC_CMDS_PROCESSED_CMDS_PROCESSED_SHIFT) |
+			 MDC_CMDS_PROCESSED_INT_ACTIVE);
+		val |= done1 << MDC_CMDS_PROCESSED_CMDS_PROCESSED_SHIFT;
+		mdc_chan_writel(mchan, val, MDC_CMDS_PROCESSED);
+		val = mdc_chan_readl(mchan, MDC_CMDS_PROCESSED);
+		done2 = (val >> MDC_CMDS_PROCESSED_CMDS_DONE_SHIFT) &
+			MDC_CMDS_PROCESSED_CMDS_DONE_MASK;
+	} while (done1 != done2);
+
+	dev_dbg(mdma2dev(mchan->mdma), "IRQ on channel %d\n", mchan->chan_nr);
+
+	mdesc = mchan->desc;
+	if (!mdesc) {
+		dev_warn(mdma2dev(mchan->mdma),
+			 "IRQ with no active descriptor on channel %d\n",
+			 mchan->chan_nr);
+		goto out;
+	}
+
+	for (i = processed; i != done1;
+	     i = (i + 1) % (MDC_CMDS_PROCESSED_CMDS_PROCESSED_MASK + 1)) {
+		/*
+		 * The first interrupt in a transfer indicates that the
+		 * command list has been loaded, not that a command has
+		 * been completed.
+		 */
+		if (!mdesc->cmd_loaded) {
+			mdesc->cmd_loaded = true;
+			continue;
+		}
+
+		mdesc->list_cmds_done++;
+		if (mdesc->cyclic) {
+			mdesc->list_cmds_done %= mdesc->list_len;
+			if (mdesc->list_cmds_done % mdesc->list_period_len == 0)
+				vchan_cyclic_callback(&mdesc->vd);
+		} else if (mdesc->list_cmds_done == mdesc->list_len) {
+			mchan->desc = NULL;
+			vchan_cookie_complete(&mdesc->vd);
+			mdc_issue_desc(mchan);
+			break;
+		}
+	}
+out:
+	spin_unlock(&mchan->vc.lock);
+
+	return IRQ_HANDLED;
+}
+
+static struct dma_chan *mdc_of_xlate(struct of_phandle_args *dma_spec,
+				     struct of_dma *ofdma)
+{
+	struct mdc_dma *mdma = ofdma->of_dma_data;
+	struct dma_chan *chan;
+
+	if (dma_spec->args_count != 3)
+		return NULL;
+
+	list_for_each_entry(chan, &mdma->dma_dev.channels, device_node) {
+		struct mdc_chan *mchan = to_mdc_chan(chan);
+
+		if (!(dma_spec->args[1] & BIT(mchan->chan_nr)))
+			continue;
+		if (dma_get_slave_channel(chan)) {
+			mchan->periph = dma_spec->args[0];
+			mchan->thread = dma_spec->args[2];
+			return chan;
+		}
+	}
+
+	return NULL;
+}
+
+#define PISTACHIO_CR_PERIPH_DMA_ROUTE(ch)	(0x120 + 0x4 * ((ch) / 4))
+#define PISTACHIO_CR_PERIPH_DMA_ROUTE_SHIFT(ch) (8 * ((ch) % 4))
+#define PISTACHIO_CR_PERIPH_DMA_ROUTE_MASK	0x3f
+
+static void pistachio_mdc_enable_chan(struct mdc_chan *mchan)
+{
+	struct mdc_dma *mdma = mchan->mdma;
+
+	regmap_update_bits(mdma->periph_regs,
+			   PISTACHIO_CR_PERIPH_DMA_ROUTE(mchan->chan_nr),
+			   PISTACHIO_CR_PERIPH_DMA_ROUTE_MASK <<
+			   PISTACHIO_CR_PERIPH_DMA_ROUTE_SHIFT(mchan->chan_nr),
+			   mchan->periph <<
+			   PISTACHIO_CR_PERIPH_DMA_ROUTE_SHIFT(mchan->chan_nr));
+}
+
+static void pistachio_mdc_disable_chan(struct mdc_chan *mchan)
+{
+	struct mdc_dma *mdma = mchan->mdma;
+
+	regmap_update_bits(mdma->periph_regs,
+			   PISTACHIO_CR_PERIPH_DMA_ROUTE(mchan->chan_nr),
+			   PISTACHIO_CR_PERIPH_DMA_ROUTE_MASK <<
+			   PISTACHIO_CR_PERIPH_DMA_ROUTE_SHIFT(mchan->chan_nr),
+			   0);
+}
+
+static const struct mdc_dma_soc_data pistachio_mdc_data = {
+	.enable_chan = pistachio_mdc_enable_chan,
+	.disable_chan = pistachio_mdc_disable_chan,
+};
+
+static const struct of_device_id mdc_dma_of_match[] = {
+	{ .compatible = "img,pistachio-mdc-dma", .data = &pistachio_mdc_data, },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, mdc_dma_of_match);
+
+static int mdc_dma_probe(struct platform_device *pdev)
+{
+	struct mdc_dma *mdma;
+	struct resource *res;
+	const struct of_device_id *match;
+	unsigned int i;
+	u32 val;
+	int ret;
+
+	mdma = devm_kzalloc(&pdev->dev, sizeof(*mdma), GFP_KERNEL);
+	if (!mdma)
+		return -ENOMEM;
+	platform_set_drvdata(pdev, mdma);
+
+	match = of_match_device(mdc_dma_of_match, &pdev->dev);
+	mdma->soc = match->data;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	mdma->regs = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(mdma->regs))
+		return PTR_ERR(mdma->regs);
+
+	mdma->periph_regs = syscon_regmap_lookup_by_phandle(pdev->dev.of_node,
+							    "img,cr-periph");
+	if (IS_ERR(mdma->periph_regs))
+		return PTR_ERR(mdma->periph_regs);
+
+	mdma->clk = devm_clk_get(&pdev->dev, "sys");
+	if (IS_ERR(mdma->clk))
+		return PTR_ERR(mdma->clk);
+
+	ret = clk_prepare_enable(mdma->clk);
+	if (ret)
+		return ret;
+
+	dma_cap_zero(mdma->dma_dev.cap_mask);
+	dma_cap_set(DMA_SLAVE, mdma->dma_dev.cap_mask);
+	dma_cap_set(DMA_PRIVATE, mdma->dma_dev.cap_mask);
+	dma_cap_set(DMA_CYCLIC, mdma->dma_dev.cap_mask);
+	dma_cap_set(DMA_MEMCPY, mdma->dma_dev.cap_mask);
+
+	val = mdc_readl(mdma, MDC_GLOBAL_CONFIG_A);
+	mdma->nr_channels = (val >> MDC_GLOBAL_CONFIG_A_DMA_CONTEXTS_SHIFT) &
+		MDC_GLOBAL_CONFIG_A_DMA_CONTEXTS_MASK;
+	mdma->nr_threads =
+		1 << ((val >> MDC_GLOBAL_CONFIG_A_THREAD_ID_WIDTH_SHIFT) &
+		      MDC_GLOBAL_CONFIG_A_THREAD_ID_WIDTH_MASK);
+	mdma->bus_width =
+		(1 << ((val >> MDC_GLOBAL_CONFIG_A_SYS_DAT_WIDTH_SHIFT) &
+		       MDC_GLOBAL_CONFIG_A_SYS_DAT_WIDTH_MASK)) / 8;
+	/*
+	 * Although transfer sizes of up to MDC_TRANSFER_SIZE_MASK + 1 bytes
+	 * are supported, this makes it possible for the value reported in
+	 * MDC_ACTIVE_TRANSFER_SIZE to be ambiguous - an active transfer size
+	 * of MDC_TRANSFER_SIZE_MASK may indicate either that 0 bytes or
+	 * MDC_TRANSFER_SIZE_MASK + 1 bytes are remaining.  To eliminate this
+	 * ambiguity, restrict transfer sizes to one bus-width less than the
+	 * actual maximum.
+	 */
+	mdma->max_xfer_size = MDC_TRANSFER_SIZE_MASK + 1 - mdma->bus_width;
+
+	of_property_read_u32(pdev->dev.of_node, "dma-channels",
+			     &mdma->nr_channels);
+	ret = of_property_read_u32(pdev->dev.of_node,
+				   "img,max-burst-multiplier",
+				   &mdma->max_burst_mult);
+	if (ret)
+		goto disable_clk;
+
+	mdma->dma_dev.dev = &pdev->dev;
+	mdma->dma_dev.device_prep_slave_sg = mdc_prep_slave_sg;
+	mdma->dma_dev.device_prep_dma_cyclic = mdc_prep_dma_cyclic;
+	mdma->dma_dev.device_prep_dma_memcpy = mdc_prep_dma_memcpy;
+	mdma->dma_dev.device_alloc_chan_resources = mdc_alloc_chan_resources;
+	mdma->dma_dev.device_free_chan_resources = mdc_free_chan_resources;
+	mdma->dma_dev.device_tx_status = mdc_tx_status;
+	mdma->dma_dev.device_issue_pending = mdc_issue_pending;
+	mdma->dma_dev.device_terminate_all = mdc_terminate_all;
+	mdma->dma_dev.device_config = mdc_slave_config;
+
+	mdma->dma_dev.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	mdma->dma_dev.residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+	for (i = 1; i <= mdma->bus_width; i <<= 1) {
+		mdma->dma_dev.src_addr_widths |= BIT(i);
+		mdma->dma_dev.dst_addr_widths |= BIT(i);
+	}
+
+	INIT_LIST_HEAD(&mdma->dma_dev.channels);
+	for (i = 0; i < mdma->nr_channels; i++) {
+		struct mdc_chan *mchan = &mdma->channels[i];
+
+		mchan->mdma = mdma;
+		mchan->chan_nr = i;
+		mchan->irq = platform_get_irq(pdev, i);
+		if (mchan->irq < 0) {
+			ret = mchan->irq;
+			goto disable_clk;
+		}
+		ret = devm_request_irq(&pdev->dev, mchan->irq, mdc_chan_irq,
+				       IRQ_TYPE_LEVEL_HIGH,
+				       dev_name(&pdev->dev), mchan);
+		if (ret < 0)
+			goto disable_clk;
+
+		mchan->vc.desc_free = mdc_desc_free;
+		vchan_init(&mchan->vc, &mdma->dma_dev);
+	}
+
+	mdma->desc_pool = dmam_pool_create(dev_name(&pdev->dev), &pdev->dev,
+					   sizeof(struct mdc_hw_list_desc),
+					   4, 0);
+	if (!mdma->desc_pool) {
+		ret = -ENOMEM;
+		goto disable_clk;
+	}
+
+	ret = dma_async_device_register(&mdma->dma_dev);
+	if (ret)
+		goto disable_clk;
+
+	ret = of_dma_controller_register(pdev->dev.of_node, mdc_of_xlate, mdma);
+	if (ret)
+		goto unregister;
+
+	dev_info(&pdev->dev, "MDC with %u channels and %u threads\n",
+		 mdma->nr_channels, mdma->nr_threads);
+
+	return 0;
+
+unregister:
+	dma_async_device_unregister(&mdma->dma_dev);
+disable_clk:
+	clk_disable_unprepare(mdma->clk);
+	return ret;
+}
+
+static int mdc_dma_remove(struct platform_device *pdev)
+{
+	struct mdc_dma *mdma = platform_get_drvdata(pdev);
+	struct mdc_chan *mchan, *next;
+
+	of_dma_controller_free(pdev->dev.of_node);
+	dma_async_device_unregister(&mdma->dma_dev);
+
+	list_for_each_entry_safe(mchan, next, &mdma->dma_dev.channels,
+				 vc.chan.device_node) {
+		list_del(&mchan->vc.chan.device_node);
+
+		synchronize_irq(mchan->irq);
+		devm_free_irq(&pdev->dev, mchan->irq, mchan);
+
+		tasklet_kill(&mchan->vc.task);
+	}
+
+	clk_disable_unprepare(mdma->clk);
+
+	return 0;
+}
+
+static struct platform_driver mdc_dma_driver = {
+	.driver = {
+		.name = "img-mdc-dma",
+		.of_match_table = of_match_ptr(mdc_dma_of_match),
+	},
+	.probe = mdc_dma_probe,
+	.remove = mdc_dma_remove,
+};
+module_platform_driver(mdc_dma_driver);
+
+MODULE_DESCRIPTION("IMG Multi-threaded DMA Controller (MDC) driver");
+MODULE_AUTHOR("Andrew Bresticker <abrestic@chromium.org>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/dma/imx-dma.c b/drivers/dma/imx-dma.c
index 10bbc0a..eed4059 100644
--- a/drivers/dma/imx-dma.c
+++ b/drivers/dma/imx-dma.c
@@ -230,11 +230,6 @@
 	return imxdma->devtype == IMX1_DMA;
 }
 
-static inline int is_imx21_dma(struct imxdma_engine *imxdma)
-{
-	return imxdma->devtype == IMX21_DMA;
-}
-
 static inline int is_imx27_dma(struct imxdma_engine *imxdma)
 {
 	return imxdma->devtype == IMX27_DMA;
@@ -669,69 +664,67 @@
 
 }
 
-static int imxdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-		unsigned long arg)
+static int imxdma_terminate_all(struct dma_chan *chan)
 {
 	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
-	struct dma_slave_config *dmaengine_cfg = (void *)arg;
 	struct imxdma_engine *imxdma = imxdmac->imxdma;
 	unsigned long flags;
+
+	imxdma_disable_hw(imxdmac);
+
+	spin_lock_irqsave(&imxdma->lock, flags);
+	list_splice_tail_init(&imxdmac->ld_active, &imxdmac->ld_free);
+	list_splice_tail_init(&imxdmac->ld_queue, &imxdmac->ld_free);
+	spin_unlock_irqrestore(&imxdma->lock, flags);
+	return 0;
+}
+
+static int imxdma_config(struct dma_chan *chan,
+			 struct dma_slave_config *dmaengine_cfg)
+{
+	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
 	unsigned int mode = 0;
 
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		imxdma_disable_hw(imxdmac);
-
-		spin_lock_irqsave(&imxdma->lock, flags);
-		list_splice_tail_init(&imxdmac->ld_active, &imxdmac->ld_free);
-		list_splice_tail_init(&imxdmac->ld_queue, &imxdmac->ld_free);
-		spin_unlock_irqrestore(&imxdma->lock, flags);
-		return 0;
-	case DMA_SLAVE_CONFIG:
-		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;
-		} else {
-			imxdmac->per_address = dmaengine_cfg->dst_addr;
-			imxdmac->watermark_level = dmaengine_cfg->dst_maxburst;
-			imxdmac->word_size = dmaengine_cfg->dst_addr_width;
-		}
-
-		switch (imxdmac->word_size) {
-		case DMA_SLAVE_BUSWIDTH_1_BYTE:
-			mode = IMX_DMA_MEMSIZE_8;
-			break;
-		case DMA_SLAVE_BUSWIDTH_2_BYTES:
-			mode = IMX_DMA_MEMSIZE_16;
-			break;
-		default:
-		case DMA_SLAVE_BUSWIDTH_4_BYTES:
-			mode = IMX_DMA_MEMSIZE_32;
-			break;
-		}
-
-		imxdmac->hw_chaining = 0;
-
-		imxdmac->ccr_from_device = (mode | IMX_DMA_TYPE_FIFO) |
-			((IMX_DMA_MEMSIZE_32 | IMX_DMA_TYPE_LINEAR) << 2) |
-			CCR_REN;
-		imxdmac->ccr_to_device =
-			(IMX_DMA_MEMSIZE_32 | IMX_DMA_TYPE_LINEAR) |
-			((mode | IMX_DMA_TYPE_FIFO) << 2) | CCR_REN;
-		imx_dmav1_writel(imxdma, imxdmac->dma_request,
-				 DMA_RSSR(imxdmac->channel));
-
-		/* Set burst length */
-		imx_dmav1_writel(imxdma, imxdmac->watermark_level *
-				imxdmac->word_size, DMA_BLR(imxdmac->channel));
-
-		return 0;
-	default:
-		return -ENOSYS;
+	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;
+	} else {
+		imxdmac->per_address = dmaengine_cfg->dst_addr;
+		imxdmac->watermark_level = dmaengine_cfg->dst_maxburst;
+		imxdmac->word_size = dmaengine_cfg->dst_addr_width;
 	}
 
-	return -EINVAL;
+	switch (imxdmac->word_size) {
+	case DMA_SLAVE_BUSWIDTH_1_BYTE:
+		mode = IMX_DMA_MEMSIZE_8;
+		break;
+	case DMA_SLAVE_BUSWIDTH_2_BYTES:
+		mode = IMX_DMA_MEMSIZE_16;
+		break;
+	default:
+	case DMA_SLAVE_BUSWIDTH_4_BYTES:
+		mode = IMX_DMA_MEMSIZE_32;
+		break;
+	}
+
+	imxdmac->hw_chaining = 0;
+
+	imxdmac->ccr_from_device = (mode | IMX_DMA_TYPE_FIFO) |
+		((IMX_DMA_MEMSIZE_32 | IMX_DMA_TYPE_LINEAR) << 2) |
+		CCR_REN;
+	imxdmac->ccr_to_device =
+		(IMX_DMA_MEMSIZE_32 | IMX_DMA_TYPE_LINEAR) |
+		((mode | IMX_DMA_TYPE_FIFO) << 2) | CCR_REN;
+	imx_dmav1_writel(imxdma, imxdmac->dma_request,
+			 DMA_RSSR(imxdmac->channel));
+
+	/* Set burst length */
+	imx_dmav1_writel(imxdma, imxdmac->watermark_level *
+			 imxdmac->word_size, DMA_BLR(imxdmac->channel));
+
+	return 0;
 }
 
 static enum dma_status imxdma_tx_status(struct dma_chan *chan,
@@ -1184,7 +1177,8 @@
 	imxdma->dma_device.device_prep_dma_cyclic = imxdma_prep_dma_cyclic;
 	imxdma->dma_device.device_prep_dma_memcpy = imxdma_prep_dma_memcpy;
 	imxdma->dma_device.device_prep_interleaved_dma = imxdma_prep_dma_interleaved;
-	imxdma->dma_device.device_control = imxdma_control;
+	imxdma->dma_device.device_config = imxdma_config;
+	imxdma->dma_device.device_terminate_all = imxdma_terminate_all;
 	imxdma->dma_device.device_issue_pending = imxdma_issue_pending;
 
 	platform_set_drvdata(pdev, imxdma);
diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c
index d0df198..18c0a13 100644
--- a/drivers/dma/imx-sdma.c
+++ b/drivers/dma/imx-sdma.c
@@ -830,20 +830,29 @@
 	return ret;
 }
 
-static void sdma_disable_channel(struct sdma_channel *sdmac)
+static struct sdma_channel *to_sdma_chan(struct dma_chan *chan)
 {
+	return container_of(chan, struct sdma_channel, chan);
+}
+
+static int sdma_disable_channel(struct dma_chan *chan)
+{
+	struct sdma_channel *sdmac = to_sdma_chan(chan);
 	struct sdma_engine *sdma = sdmac->sdma;
 	int channel = sdmac->channel;
 
 	writel_relaxed(BIT(channel), sdma->regs + SDMA_H_STATSTOP);
 	sdmac->status = DMA_ERROR;
+
+	return 0;
 }
 
-static int sdma_config_channel(struct sdma_channel *sdmac)
+static int sdma_config_channel(struct dma_chan *chan)
 {
+	struct sdma_channel *sdmac = to_sdma_chan(chan);
 	int ret;
 
-	sdma_disable_channel(sdmac);
+	sdma_disable_channel(chan);
 
 	sdmac->event_mask[0] = 0;
 	sdmac->event_mask[1] = 0;
@@ -935,11 +944,6 @@
 	return ret;
 }
 
-static struct sdma_channel *to_sdma_chan(struct dma_chan *chan)
-{
-	return container_of(chan, struct sdma_channel, chan);
-}
-
 static dma_cookie_t sdma_tx_submit(struct dma_async_tx_descriptor *tx)
 {
 	unsigned long flags;
@@ -1004,7 +1008,7 @@
 	struct sdma_channel *sdmac = to_sdma_chan(chan);
 	struct sdma_engine *sdma = sdmac->sdma;
 
-	sdma_disable_channel(sdmac);
+	sdma_disable_channel(chan);
 
 	if (sdmac->event_id0)
 		sdma_event_disable(sdmac, sdmac->event_id0);
@@ -1203,35 +1207,24 @@
 	return NULL;
 }
 
-static int sdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-		unsigned long arg)
+static int sdma_config(struct dma_chan *chan,
+		       struct dma_slave_config *dmaengine_cfg)
 {
 	struct sdma_channel *sdmac = to_sdma_chan(chan);
-	struct dma_slave_config *dmaengine_cfg = (void *)arg;
 
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		sdma_disable_channel(sdmac);
-		return 0;
-	case DMA_SLAVE_CONFIG:
-		if (dmaengine_cfg->direction == DMA_DEV_TO_MEM) {
-			sdmac->per_address = dmaengine_cfg->src_addr;
-			sdmac->watermark_level = dmaengine_cfg->src_maxburst *
-						dmaengine_cfg->src_addr_width;
-			sdmac->word_size = dmaengine_cfg->src_addr_width;
-		} else {
-			sdmac->per_address = dmaengine_cfg->dst_addr;
-			sdmac->watermark_level = dmaengine_cfg->dst_maxburst *
-						dmaengine_cfg->dst_addr_width;
-			sdmac->word_size = dmaengine_cfg->dst_addr_width;
-		}
-		sdmac->direction = dmaengine_cfg->direction;
-		return sdma_config_channel(sdmac);
-	default:
-		return -ENOSYS;
+	if (dmaengine_cfg->direction == DMA_DEV_TO_MEM) {
+		sdmac->per_address = dmaengine_cfg->src_addr;
+		sdmac->watermark_level = dmaengine_cfg->src_maxburst *
+			dmaengine_cfg->src_addr_width;
+		sdmac->word_size = dmaengine_cfg->src_addr_width;
+	} else {
+		sdmac->per_address = dmaengine_cfg->dst_addr;
+		sdmac->watermark_level = dmaengine_cfg->dst_maxburst *
+			dmaengine_cfg->dst_addr_width;
+		sdmac->word_size = dmaengine_cfg->dst_addr_width;
 	}
-
-	return -EINVAL;
+	sdmac->direction = dmaengine_cfg->direction;
+	return sdma_config_channel(chan);
 }
 
 static enum dma_status sdma_tx_status(struct dma_chan *chan,
@@ -1303,15 +1296,15 @@
 	if (header->ram_code_start + header->ram_code_size > fw->size)
 		goto err_firmware;
 	switch (header->version_major) {
-		case 1:
-			sdma->script_number = SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V1;
-			break;
-		case 2:
-			sdma->script_number = SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V2;
-			break;
-		default:
-			dev_err(sdma->dev, "unknown firmware version\n");
-			goto err_firmware;
+	case 1:
+		sdma->script_number = SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V1;
+		break;
+	case 2:
+		sdma->script_number = SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V2;
+		break;
+	default:
+		dev_err(sdma->dev, "unknown firmware version\n");
+		goto err_firmware;
 	}
 
 	addr = (void *)header + header->script_addrs_start;
@@ -1479,7 +1472,7 @@
 	if (ret)
 		return ret;
 
-	sdma = kzalloc(sizeof(*sdma), GFP_KERNEL);
+	sdma = devm_kzalloc(&pdev->dev, sizeof(*sdma), GFP_KERNEL);
 	if (!sdma)
 		return -ENOMEM;
 
@@ -1488,48 +1481,34 @@
 	sdma->dev = &pdev->dev;
 	sdma->drvdata = drvdata;
 
-	iores = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	irq = platform_get_irq(pdev, 0);
-	if (!iores || irq < 0) {
-		ret = -EINVAL;
-		goto err_irq;
-	}
+	if (irq < 0)
+		return irq;
 
-	if (!request_mem_region(iores->start, resource_size(iores), pdev->name)) {
-		ret = -EBUSY;
-		goto err_request_region;
-	}
+	iores = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	sdma->regs = devm_ioremap_resource(&pdev->dev, iores);
+	if (IS_ERR(sdma->regs))
+		return PTR_ERR(sdma->regs);
 
 	sdma->clk_ipg = devm_clk_get(&pdev->dev, "ipg");
-	if (IS_ERR(sdma->clk_ipg)) {
-		ret = PTR_ERR(sdma->clk_ipg);
-		goto err_clk;
-	}
+	if (IS_ERR(sdma->clk_ipg))
+		return PTR_ERR(sdma->clk_ipg);
 
 	sdma->clk_ahb = devm_clk_get(&pdev->dev, "ahb");
-	if (IS_ERR(sdma->clk_ahb)) {
-		ret = PTR_ERR(sdma->clk_ahb);
-		goto err_clk;
-	}
+	if (IS_ERR(sdma->clk_ahb))
+		return PTR_ERR(sdma->clk_ahb);
 
 	clk_prepare(sdma->clk_ipg);
 	clk_prepare(sdma->clk_ahb);
 
-	sdma->regs = ioremap(iores->start, resource_size(iores));
-	if (!sdma->regs) {
-		ret = -ENOMEM;
-		goto err_ioremap;
-	}
-
-	ret = request_irq(irq, sdma_int_handler, 0, "sdma", sdma);
+	ret = devm_request_irq(&pdev->dev, irq, sdma_int_handler, 0, "sdma",
+			       sdma);
 	if (ret)
-		goto err_request_irq;
+		return ret;
 
 	sdma->script_addrs = kzalloc(sizeof(*sdma->script_addrs), GFP_KERNEL);
-	if (!sdma->script_addrs) {
-		ret = -ENOMEM;
-		goto err_alloc;
-	}
+	if (!sdma->script_addrs)
+		return -ENOMEM;
 
 	/* initially no scripts available */
 	saddr_arr = (s32 *)sdma->script_addrs;
@@ -1600,7 +1579,12 @@
 	sdma->dma_device.device_tx_status = sdma_tx_status;
 	sdma->dma_device.device_prep_slave_sg = sdma_prep_slave_sg;
 	sdma->dma_device.device_prep_dma_cyclic = sdma_prep_dma_cyclic;
-	sdma->dma_device.device_control = sdma_control;
+	sdma->dma_device.device_config = sdma_config;
+	sdma->dma_device.device_terminate_all = sdma_disable_channel;
+	sdma->dma_device.src_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_4_BYTES);
+	sdma->dma_device.dst_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_4_BYTES);
+	sdma->dma_device.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	sdma->dma_device.residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
 	sdma->dma_device.device_issue_pending = sdma_issue_pending;
 	sdma->dma_device.dev->dma_parms = &sdma->dma_parms;
 	dma_set_max_seg_size(sdma->dma_device.dev, 65535);
@@ -1629,38 +1613,22 @@
 	dma_async_device_unregister(&sdma->dma_device);
 err_init:
 	kfree(sdma->script_addrs);
-err_alloc:
-	free_irq(irq, sdma);
-err_request_irq:
-	iounmap(sdma->regs);
-err_ioremap:
-err_clk:
-	release_mem_region(iores->start, resource_size(iores));
-err_request_region:
-err_irq:
-	kfree(sdma);
 	return ret;
 }
 
 static int sdma_remove(struct platform_device *pdev)
 {
 	struct sdma_engine *sdma = platform_get_drvdata(pdev);
-	struct resource *iores = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	int irq = platform_get_irq(pdev, 0);
 	int i;
 
 	dma_async_device_unregister(&sdma->dma_device);
 	kfree(sdma->script_addrs);
-	free_irq(irq, sdma);
-	iounmap(sdma->regs);
-	release_mem_region(iores->start, resource_size(iores));
 	/* Kill the tasklet */
 	for (i = 0; i < MAX_DMA_CHANNELS; i++) {
 		struct sdma_channel *sdmac = &sdma->channel[i];
 
 		tasklet_kill(&sdmac->tasklet);
 	}
-	kfree(sdma);
 
 	platform_set_drvdata(pdev, NULL);
 	dev_info(&pdev->dev, "Removed...\n");
diff --git a/drivers/dma/intel_mid_dma.c b/drivers/dma/intel_mid_dma.c
index 1aab813..5aaead9 100644
--- a/drivers/dma/intel_mid_dma.c
+++ b/drivers/dma/intel_mid_dma.c
@@ -492,10 +492,10 @@
 	return ret;
 }
 
-static int dma_slave_control(struct dma_chan *chan, unsigned long arg)
+static int intel_mid_dma_config(struct dma_chan *chan,
+				struct dma_slave_config *slave)
 {
 	struct intel_mid_dma_chan	*midc = to_intel_mid_dma_chan(chan);
-	struct dma_slave_config  *slave = (struct dma_slave_config *)arg;
 	struct intel_mid_dma_slave *mid_slave;
 
 	BUG_ON(!midc);
@@ -509,28 +509,14 @@
 	midc->mid_slave = mid_slave;
 	return 0;
 }
-/**
- * intel_mid_dma_device_control -	DMA device control
- * @chan: chan for DMA control
- * @cmd: control cmd
- * @arg: cmd arg value
- *
- * Perform DMA control command
- */
-static int intel_mid_dma_device_control(struct dma_chan *chan,
-			enum dma_ctrl_cmd cmd, unsigned long arg)
+
+static int intel_mid_dma_terminate_all(struct dma_chan *chan)
 {
 	struct intel_mid_dma_chan	*midc = to_intel_mid_dma_chan(chan);
 	struct middma_device	*mid = to_middma_device(chan->device);
 	struct intel_mid_dma_desc	*desc, *_desc;
 	union intel_mid_dma_cfg_lo cfg_lo;
 
-	if (cmd == DMA_SLAVE_CONFIG)
-		return dma_slave_control(chan, arg);
-
-	if (cmd != DMA_TERMINATE_ALL)
-		return -ENXIO;
-
 	spin_lock_bh(&midc->lock);
 	if (midc->busy == false) {
 		spin_unlock_bh(&midc->lock);
@@ -1148,7 +1134,8 @@
 	dma->common.device_prep_dma_memcpy = intel_mid_dma_prep_memcpy;
 	dma->common.device_issue_pending = intel_mid_dma_issue_pending;
 	dma->common.device_prep_slave_sg = intel_mid_dma_prep_slave_sg;
-	dma->common.device_control = intel_mid_dma_device_control;
+	dma->common.device_config = intel_mid_dma_config;
+	dma->common.device_terminate_all = intel_mid_dma_terminate_all;
 
 	/*enable dma cntrl*/
 	iowrite32(REG_BIT0, dma->dma_base + DMA_CFG);
diff --git a/drivers/dma/ioat/dma_v3.c b/drivers/dma/ioat/dma_v3.c
index 32eae38..77a6dcf 100644
--- a/drivers/dma/ioat/dma_v3.c
+++ b/drivers/dma/ioat/dma_v3.c
@@ -214,6 +214,11 @@
 	case PCI_DEVICE_ID_INTEL_IOAT_BWD1:
 	case PCI_DEVICE_ID_INTEL_IOAT_BWD2:
 	case PCI_DEVICE_ID_INTEL_IOAT_BWD3:
+	/* even though not Atom, BDX-DE has same DMA silicon */
+	case PCI_DEVICE_ID_INTEL_IOAT_BDXDE0:
+	case PCI_DEVICE_ID_INTEL_IOAT_BDXDE1:
+	case PCI_DEVICE_ID_INTEL_IOAT_BDXDE2:
+	case PCI_DEVICE_ID_INTEL_IOAT_BDXDE3:
 		return true;
 	default:
 		return false;
@@ -489,6 +494,7 @@
 	struct ioat_chan_common *chan = &ioat->base;
 	struct pci_dev *pdev = to_pdev(chan);
 	struct ioat_dma_descriptor *hw;
+	struct dma_async_tx_descriptor *tx;
 	u64 phys_complete;
 	struct ioat_ring_ent *desc;
 	u32 err_handled = 0;
@@ -534,6 +540,16 @@
 		dev_err(to_dev(chan), "%s: fatal error (%x:%x)\n",
 			__func__, chanerr, err_handled);
 		BUG();
+	} else { /* cleanup the faulty descriptor */
+		tx = &desc->txd;
+		if (tx->cookie) {
+			dma_cookie_complete(tx);
+			dma_descriptor_unmap(tx);
+			if (tx->callback) {
+				tx->callback(tx->callback_param);
+				tx->callback = NULL;
+			}
+		}
 	}
 
 	writel(chanerr, chan->reg_base + IOAT_CHANERR_OFFSET);
@@ -1300,7 +1316,8 @@
 
 	tmo = wait_for_completion_timeout(&cmp, msecs_to_jiffies(3000));
 
-	if (dma->device_tx_status(dma_chan, cookie, NULL) != DMA_COMPLETE) {
+	if (tmo == 0 ||
+	    dma->device_tx_status(dma_chan, cookie, NULL) != DMA_COMPLETE) {
 		dev_err(dev, "Self-test xor timed out\n");
 		err = -ENODEV;
 		goto dma_unmap;
@@ -1366,7 +1383,8 @@
 
 	tmo = wait_for_completion_timeout(&cmp, msecs_to_jiffies(3000));
 
-	if (dma->device_tx_status(dma_chan, cookie, NULL) != DMA_COMPLETE) {
+	if (tmo == 0 ||
+	    dma->device_tx_status(dma_chan, cookie, NULL) != DMA_COMPLETE) {
 		dev_err(dev, "Self-test validate timed out\n");
 		err = -ENODEV;
 		goto dma_unmap;
@@ -1418,7 +1436,8 @@
 
 	tmo = wait_for_completion_timeout(&cmp, msecs_to_jiffies(3000));
 
-	if (dma->device_tx_status(dma_chan, cookie, NULL) != DMA_COMPLETE) {
+	if (tmo == 0 ||
+	    dma->device_tx_status(dma_chan, cookie, NULL) != DMA_COMPLETE) {
 		dev_err(dev, "Self-test 2nd validate timed out\n");
 		err = -ENODEV;
 		goto dma_unmap;
diff --git a/drivers/dma/ioat/hw.h b/drivers/dma/ioat/hw.h
index 62f83e9..02177ec 100644
--- a/drivers/dma/ioat/hw.h
+++ b/drivers/dma/ioat/hw.h
@@ -57,6 +57,11 @@
 #define PCI_DEVICE_ID_INTEL_IOAT_BWD2	0x0C52
 #define PCI_DEVICE_ID_INTEL_IOAT_BWD3	0x0C53
 
+#define PCI_DEVICE_ID_INTEL_IOAT_BDXDE0	0x6f50
+#define PCI_DEVICE_ID_INTEL_IOAT_BDXDE1	0x6f51
+#define PCI_DEVICE_ID_INTEL_IOAT_BDXDE2	0x6f52
+#define PCI_DEVICE_ID_INTEL_IOAT_BDXDE3	0x6f53
+
 #define IOAT_VER_1_2            0x12    /* Version 1.2 */
 #define IOAT_VER_2_0            0x20    /* Version 2.0 */
 #define IOAT_VER_3_0            0x30    /* Version 3.0 */
diff --git a/drivers/dma/ioat/pci.c b/drivers/dma/ioat/pci.c
index 1d051cd..5501eb0 100644
--- a/drivers/dma/ioat/pci.c
+++ b/drivers/dma/ioat/pci.c
@@ -111,6 +111,11 @@
 	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_IOAT_BWD2) },
 	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_IOAT_BWD3) },
 
+	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_IOAT_BDXDE0) },
+	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_IOAT_BDXDE1) },
+	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_IOAT_BDXDE2) },
+	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_IOAT_BDXDE3) },
+
 	{ 0, }
 };
 MODULE_DEVICE_TABLE(pci, ioat_pci_tbl);
diff --git a/drivers/dma/ipu/ipu_idmac.c b/drivers/dma/ipu/ipu_idmac.c
index c2b017a..b54f62d 100644
--- a/drivers/dma/ipu/ipu_idmac.c
+++ b/drivers/dma/ipu/ipu_idmac.c
@@ -1398,76 +1398,81 @@
 	 */
 }
 
-static int __idmac_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-			   unsigned long arg)
+static int idmac_pause(struct dma_chan *chan)
 {
 	struct idmac_channel *ichan = to_idmac_chan(chan);
 	struct idmac *idmac = to_idmac(chan->device);
 	struct ipu *ipu = to_ipu(idmac);
 	struct list_head *list, *tmp;
 	unsigned long flags;
-	int i;
 
-	switch (cmd) {
-	case DMA_PAUSE:
-		spin_lock_irqsave(&ipu->lock, flags);
-		ipu_ic_disable_task(ipu, chan->chan_id);
+	mutex_lock(&ichan->chan_mutex);
 
-		/* Return all descriptors into "prepared" state */
-		list_for_each_safe(list, tmp, &ichan->queue)
-			list_del_init(list);
+	spin_lock_irqsave(&ipu->lock, flags);
+	ipu_ic_disable_task(ipu, chan->chan_id);
 
-		ichan->sg[0] = NULL;
-		ichan->sg[1] = NULL;
+	/* Return all descriptors into "prepared" state */
+	list_for_each_safe(list, tmp, &ichan->queue)
+		list_del_init(list);
 
-		spin_unlock_irqrestore(&ipu->lock, flags);
+	ichan->sg[0] = NULL;
+	ichan->sg[1] = NULL;
 
-		ichan->status = IPU_CHANNEL_INITIALIZED;
-		break;
-	case DMA_TERMINATE_ALL:
-		ipu_disable_channel(idmac, ichan,
-				    ichan->status >= IPU_CHANNEL_ENABLED);
+	spin_unlock_irqrestore(&ipu->lock, flags);
 
-		tasklet_disable(&ipu->tasklet);
+	ichan->status = IPU_CHANNEL_INITIALIZED;
 
-		/* ichan->queue is modified in ISR, have to spinlock */
-		spin_lock_irqsave(&ichan->lock, flags);
-		list_splice_init(&ichan->queue, &ichan->free_list);
-
-		if (ichan->desc)
-			for (i = 0; i < ichan->n_tx_desc; i++) {
-				struct idmac_tx_desc *desc = ichan->desc + i;
-				if (list_empty(&desc->list))
-					/* Descriptor was prepared, but not submitted */
-					list_add(&desc->list, &ichan->free_list);
-
-				async_tx_clear_ack(&desc->txd);
-			}
-
-		ichan->sg[0] = NULL;
-		ichan->sg[1] = NULL;
-		spin_unlock_irqrestore(&ichan->lock, flags);
-
-		tasklet_enable(&ipu->tasklet);
-
-		ichan->status = IPU_CHANNEL_INITIALIZED;
-		break;
-	default:
-		return -ENOSYS;
-	}
+	mutex_unlock(&ichan->chan_mutex);
 
 	return 0;
 }
 
-static int idmac_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-			 unsigned long arg)
+static int __idmac_terminate_all(struct dma_chan *chan)
+{
+	struct idmac_channel *ichan = to_idmac_chan(chan);
+	struct idmac *idmac = to_idmac(chan->device);
+	struct ipu *ipu = to_ipu(idmac);
+	unsigned long flags;
+	int i;
+
+	ipu_disable_channel(idmac, ichan,
+			    ichan->status >= IPU_CHANNEL_ENABLED);
+
+	tasklet_disable(&ipu->tasklet);
+
+	/* ichan->queue is modified in ISR, have to spinlock */
+	spin_lock_irqsave(&ichan->lock, flags);
+	list_splice_init(&ichan->queue, &ichan->free_list);
+
+	if (ichan->desc)
+		for (i = 0; i < ichan->n_tx_desc; i++) {
+			struct idmac_tx_desc *desc = ichan->desc + i;
+			if (list_empty(&desc->list))
+				/* Descriptor was prepared, but not submitted */
+				list_add(&desc->list, &ichan->free_list);
+
+			async_tx_clear_ack(&desc->txd);
+		}
+
+	ichan->sg[0] = NULL;
+	ichan->sg[1] = NULL;
+	spin_unlock_irqrestore(&ichan->lock, flags);
+
+	tasklet_enable(&ipu->tasklet);
+
+	ichan->status = IPU_CHANNEL_INITIALIZED;
+
+	return 0;
+}
+
+static int idmac_terminate_all(struct dma_chan *chan)
 {
 	struct idmac_channel *ichan = to_idmac_chan(chan);
 	int ret;
 
 	mutex_lock(&ichan->chan_mutex);
 
-	ret = __idmac_control(chan, cmd, arg);
+	ret = __idmac_terminate_all(chan);
 
 	mutex_unlock(&ichan->chan_mutex);
 
@@ -1568,7 +1573,7 @@
 
 	mutex_lock(&ichan->chan_mutex);
 
-	__idmac_control(chan, DMA_TERMINATE_ALL, 0);
+	__idmac_terminate_all(chan);
 
 	if (ichan->status > IPU_CHANNEL_FREE) {
 #ifdef DEBUG
@@ -1622,7 +1627,8 @@
 
 	/* Compulsory for DMA_SLAVE fields */
 	dma->device_prep_slave_sg		= idmac_prep_slave_sg;
-	dma->device_control			= idmac_control;
+	dma->device_pause			= idmac_pause;
+	dma->device_terminate_all		= idmac_terminate_all;
 
 	INIT_LIST_HEAD(&dma->channels);
 	for (i = 0; i < IPU_CHANNELS_NUM; i++) {
@@ -1655,7 +1661,7 @@
 	for (i = 0; i < IPU_CHANNELS_NUM; i++) {
 		struct idmac_channel *ichan = ipu->channel + i;
 
-		idmac_control(&ichan->dma_chan, DMA_TERMINATE_ALL, 0);
+		idmac_terminate_all(&ichan->dma_chan);
 	}
 
 	dma_async_device_unregister(&idmac->dma);
diff --git a/drivers/dma/k3dma.c b/drivers/dma/k3dma.c
index a1de14a..6f7f435 100644
--- a/drivers/dma/k3dma.c
+++ b/drivers/dma/k3dma.c
@@ -441,7 +441,7 @@
 	num = 0;
 
 	if (!c->ccfg) {
-		/* default is memtomem, without calling device_control */
+		/* default is memtomem, without calling device_config */
 		c->ccfg = CX_CFG_SRCINCR | CX_CFG_DSTINCR | CX_CFG_EN;
 		c->ccfg |= (0xf << 20) | (0xf << 24);	/* burst = 16 */
 		c->ccfg |= (0x3 << 12) | (0x3 << 16);	/* width = 64 bit */
@@ -523,112 +523,126 @@
 	return vchan_tx_prep(&c->vc, &ds->vd, flags);
 }
 
-static int k3_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-	unsigned long arg)
+static int k3_dma_config(struct dma_chan *chan,
+			 struct dma_slave_config *cfg)
+{
+	struct k3_dma_chan *c = to_k3_chan(chan);
+	u32 maxburst = 0, val = 0;
+	enum dma_slave_buswidth width = DMA_SLAVE_BUSWIDTH_UNDEFINED;
+
+	if (cfg == NULL)
+		return -EINVAL;
+	c->dir = cfg->direction;
+	if (c->dir == DMA_DEV_TO_MEM) {
+		c->ccfg = CX_CFG_DSTINCR;
+		c->dev_addr = cfg->src_addr;
+		maxburst = cfg->src_maxburst;
+		width = cfg->src_addr_width;
+	} else if (c->dir == DMA_MEM_TO_DEV) {
+		c->ccfg = CX_CFG_SRCINCR;
+		c->dev_addr = cfg->dst_addr;
+		maxburst = cfg->dst_maxburst;
+		width = cfg->dst_addr_width;
+	}
+	switch (width) {
+	case DMA_SLAVE_BUSWIDTH_1_BYTE:
+	case DMA_SLAVE_BUSWIDTH_2_BYTES:
+	case DMA_SLAVE_BUSWIDTH_4_BYTES:
+	case DMA_SLAVE_BUSWIDTH_8_BYTES:
+		val =  __ffs(width);
+		break;
+	default:
+		val = 3;
+		break;
+	}
+	c->ccfg |= (val << 12) | (val << 16);
+
+	if ((maxburst == 0) || (maxburst > 16))
+		val = 16;
+	else
+		val = maxburst - 1;
+	c->ccfg |= (val << 20) | (val << 24);
+	c->ccfg |= CX_CFG_MEM2PER | CX_CFG_EN;
+
+	/* specific request line */
+	c->ccfg |= c->vc.chan.chan_id << 4;
+
+	return 0;
+}
+
+static int k3_dma_terminate_all(struct dma_chan *chan)
 {
 	struct k3_dma_chan *c = to_k3_chan(chan);
 	struct k3_dma_dev *d = to_k3_dma(chan->device);
-	struct dma_slave_config *cfg = (void *)arg;
 	struct k3_dma_phy *p = c->phy;
 	unsigned long flags;
-	u32 maxburst = 0, val = 0;
-	enum dma_slave_buswidth width = DMA_SLAVE_BUSWIDTH_UNDEFINED;
 	LIST_HEAD(head);
 
-	switch (cmd) {
-	case DMA_SLAVE_CONFIG:
-		if (cfg == NULL)
-			return -EINVAL;
-		c->dir = cfg->direction;
-		if (c->dir == DMA_DEV_TO_MEM) {
-			c->ccfg = CX_CFG_DSTINCR;
-			c->dev_addr = cfg->src_addr;
-			maxburst = cfg->src_maxburst;
-			width = cfg->src_addr_width;
-		} else if (c->dir == DMA_MEM_TO_DEV) {
-			c->ccfg = CX_CFG_SRCINCR;
-			c->dev_addr = cfg->dst_addr;
-			maxburst = cfg->dst_maxburst;
-			width = cfg->dst_addr_width;
-		}
-		switch (width) {
-		case DMA_SLAVE_BUSWIDTH_1_BYTE:
-		case DMA_SLAVE_BUSWIDTH_2_BYTES:
-		case DMA_SLAVE_BUSWIDTH_4_BYTES:
-		case DMA_SLAVE_BUSWIDTH_8_BYTES:
-			val =  __ffs(width);
-			break;
-		default:
-			val = 3;
-			break;
-		}
-		c->ccfg |= (val << 12) | (val << 16);
+	dev_dbg(d->slave.dev, "vchan %p: terminate all\n", &c->vc);
 
-		if ((maxburst == 0) || (maxburst > 16))
-			val = 16;
-		else
-			val = maxburst - 1;
-		c->ccfg |= (val << 20) | (val << 24);
-		c->ccfg |= CX_CFG_MEM2PER | CX_CFG_EN;
+	/* Prevent this channel being scheduled */
+	spin_lock(&d->lock);
+	list_del_init(&c->node);
+	spin_unlock(&d->lock);
 
-		/* specific request line */
-		c->ccfg |= c->vc.chan.chan_id << 4;
-		break;
-
-	case DMA_TERMINATE_ALL:
-		dev_dbg(d->slave.dev, "vchan %p: terminate all\n", &c->vc);
-
-		/* Prevent this channel being scheduled */
-		spin_lock(&d->lock);
-		list_del_init(&c->node);
-		spin_unlock(&d->lock);
-
-		/* Clear the tx descriptor lists */
-		spin_lock_irqsave(&c->vc.lock, flags);
-		vchan_get_all_descriptors(&c->vc, &head);
-		if (p) {
-			/* vchan is assigned to a pchan - stop the channel */
-			k3_dma_terminate_chan(p, d);
-			c->phy = NULL;
-			p->vchan = NULL;
-			p->ds_run = p->ds_done = NULL;
-		}
-		spin_unlock_irqrestore(&c->vc.lock, flags);
-		vchan_dma_desc_free_list(&c->vc, &head);
-		break;
-
-	case DMA_PAUSE:
-		dev_dbg(d->slave.dev, "vchan %p: pause\n", &c->vc);
-		if (c->status == DMA_IN_PROGRESS) {
-			c->status = DMA_PAUSED;
-			if (p) {
-				k3_dma_pause_dma(p, false);
-			} else {
-				spin_lock(&d->lock);
-				list_del_init(&c->node);
-				spin_unlock(&d->lock);
-			}
-		}
-		break;
-
-	case DMA_RESUME:
-		dev_dbg(d->slave.dev, "vchan %p: resume\n", &c->vc);
-		spin_lock_irqsave(&c->vc.lock, flags);
-		if (c->status == DMA_PAUSED) {
-			c->status = DMA_IN_PROGRESS;
-			if (p) {
-				k3_dma_pause_dma(p, true);
-			} else if (!list_empty(&c->vc.desc_issued)) {
-				spin_lock(&d->lock);
-				list_add_tail(&c->node, &d->chan_pending);
-				spin_unlock(&d->lock);
-			}
-		}
-		spin_unlock_irqrestore(&c->vc.lock, flags);
-		break;
-	default:
-		return -ENXIO;
+	/* Clear the tx descriptor lists */
+	spin_lock_irqsave(&c->vc.lock, flags);
+	vchan_get_all_descriptors(&c->vc, &head);
+	if (p) {
+		/* vchan is assigned to a pchan - stop the channel */
+		k3_dma_terminate_chan(p, d);
+		c->phy = NULL;
+		p->vchan = NULL;
+		p->ds_run = p->ds_done = NULL;
 	}
+	spin_unlock_irqrestore(&c->vc.lock, flags);
+	vchan_dma_desc_free_list(&c->vc, &head);
+
+	return 0;
+}
+
+static int k3_dma_transfer_pause(struct dma_chan *chan)
+{
+	struct k3_dma_chan *c = to_k3_chan(chan);
+	struct k3_dma_dev *d = to_k3_dma(chan->device);
+	struct k3_dma_phy *p = c->phy;
+
+	dev_dbg(d->slave.dev, "vchan %p: pause\n", &c->vc);
+	if (c->status == DMA_IN_PROGRESS) {
+		c->status = DMA_PAUSED;
+		if (p) {
+			k3_dma_pause_dma(p, false);
+		} else {
+			spin_lock(&d->lock);
+			list_del_init(&c->node);
+			spin_unlock(&d->lock);
+		}
+	}
+
+	return 0;
+}
+
+static int k3_dma_transfer_resume(struct dma_chan *chan)
+{
+	struct k3_dma_chan *c = to_k3_chan(chan);
+	struct k3_dma_dev *d = to_k3_dma(chan->device);
+	struct k3_dma_phy *p = c->phy;
+	unsigned long flags;
+
+	dev_dbg(d->slave.dev, "vchan %p: resume\n", &c->vc);
+	spin_lock_irqsave(&c->vc.lock, flags);
+	if (c->status == DMA_PAUSED) {
+		c->status = DMA_IN_PROGRESS;
+		if (p) {
+			k3_dma_pause_dma(p, true);
+		} else if (!list_empty(&c->vc.desc_issued)) {
+			spin_lock(&d->lock);
+			list_add_tail(&c->node, &d->chan_pending);
+			spin_unlock(&d->lock);
+		}
+	}
+	spin_unlock_irqrestore(&c->vc.lock, flags);
+
 	return 0;
 }
 
@@ -720,7 +734,10 @@
 	d->slave.device_prep_dma_memcpy = k3_dma_prep_memcpy;
 	d->slave.device_prep_slave_sg = k3_dma_prep_slave_sg;
 	d->slave.device_issue_pending = k3_dma_issue_pending;
-	d->slave.device_control = k3_dma_control;
+	d->slave.device_config = k3_dma_config;
+	d->slave.device_pause = k3_dma_transfer_pause;
+	d->slave.device_resume = k3_dma_transfer_resume;
+	d->slave.device_terminate_all = k3_dma_terminate_all;
 	d->slave.copy_align = DMA_ALIGN;
 
 	/* init virtual channel */
@@ -787,7 +804,7 @@
 }
 
 #ifdef CONFIG_PM_SLEEP
-static int k3_dma_suspend(struct device *dev)
+static int k3_dma_suspend_dev(struct device *dev)
 {
 	struct k3_dma_dev *d = dev_get_drvdata(dev);
 	u32 stat = 0;
@@ -803,7 +820,7 @@
 	return 0;
 }
 
-static int k3_dma_resume(struct device *dev)
+static int k3_dma_resume_dev(struct device *dev)
 {
 	struct k3_dma_dev *d = dev_get_drvdata(dev);
 	int ret = 0;
@@ -818,7 +835,7 @@
 }
 #endif
 
-static SIMPLE_DEV_PM_OPS(k3_dma_pmops, k3_dma_suspend, k3_dma_resume);
+static SIMPLE_DEV_PM_OPS(k3_dma_pmops, k3_dma_suspend_dev, k3_dma_resume_dev);
 
 static struct platform_driver k3_pdma_driver = {
 	.driver		= {
diff --git a/drivers/dma/mmp_pdma.c b/drivers/dma/mmp_pdma.c
index 8b8952f..8926f27 100644
--- a/drivers/dma/mmp_pdma.c
+++ b/drivers/dma/mmp_pdma.c
@@ -683,68 +683,70 @@
 	return NULL;
 }
 
-static int mmp_pdma_control(struct dma_chan *dchan, enum dma_ctrl_cmd cmd,
-			    unsigned long arg)
+static int mmp_pdma_config(struct dma_chan *dchan,
+			   struct dma_slave_config *cfg)
 {
 	struct mmp_pdma_chan *chan = to_mmp_pdma_chan(dchan);
-	struct dma_slave_config *cfg = (void *)arg;
-	unsigned long flags;
 	u32 maxburst = 0, addr = 0;
 	enum dma_slave_buswidth width = DMA_SLAVE_BUSWIDTH_UNDEFINED;
 
 	if (!dchan)
 		return -EINVAL;
 
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		disable_chan(chan->phy);
-		mmp_pdma_free_phy(chan);
-		spin_lock_irqsave(&chan->desc_lock, flags);
-		mmp_pdma_free_desc_list(chan, &chan->chain_pending);
-		mmp_pdma_free_desc_list(chan, &chan->chain_running);
-		spin_unlock_irqrestore(&chan->desc_lock, flags);
-		chan->idle = true;
-		break;
-	case DMA_SLAVE_CONFIG:
-		if (cfg->direction == DMA_DEV_TO_MEM) {
-			chan->dcmd = DCMD_INCTRGADDR | DCMD_FLOWSRC;
-			maxburst = cfg->src_maxburst;
-			width = cfg->src_addr_width;
-			addr = cfg->src_addr;
-		} else if (cfg->direction == DMA_MEM_TO_DEV) {
-			chan->dcmd = DCMD_INCSRCADDR | DCMD_FLOWTRG;
-			maxburst = cfg->dst_maxburst;
-			width = cfg->dst_addr_width;
-			addr = cfg->dst_addr;
-		}
-
-		if (width == DMA_SLAVE_BUSWIDTH_1_BYTE)
-			chan->dcmd |= DCMD_WIDTH1;
-		else if (width == DMA_SLAVE_BUSWIDTH_2_BYTES)
-			chan->dcmd |= DCMD_WIDTH2;
-		else if (width == DMA_SLAVE_BUSWIDTH_4_BYTES)
-			chan->dcmd |= DCMD_WIDTH4;
-
-		if (maxburst == 8)
-			chan->dcmd |= DCMD_BURST8;
-		else if (maxburst == 16)
-			chan->dcmd |= DCMD_BURST16;
-		else if (maxburst == 32)
-			chan->dcmd |= DCMD_BURST32;
-
-		chan->dir = cfg->direction;
-		chan->dev_addr = addr;
-		/* FIXME: drivers should be ported over to use the filter
-		 * function. Once that's done, the following two lines can
-		 * be removed.
-		 */
-		if (cfg->slave_id)
-			chan->drcmr = cfg->slave_id;
-		break;
-	default:
-		return -ENOSYS;
+	if (cfg->direction == DMA_DEV_TO_MEM) {
+		chan->dcmd = DCMD_INCTRGADDR | DCMD_FLOWSRC;
+		maxburst = cfg->src_maxburst;
+		width = cfg->src_addr_width;
+		addr = cfg->src_addr;
+	} else if (cfg->direction == DMA_MEM_TO_DEV) {
+		chan->dcmd = DCMD_INCSRCADDR | DCMD_FLOWTRG;
+		maxburst = cfg->dst_maxburst;
+		width = cfg->dst_addr_width;
+		addr = cfg->dst_addr;
 	}
 
+	if (width == DMA_SLAVE_BUSWIDTH_1_BYTE)
+		chan->dcmd |= DCMD_WIDTH1;
+	else if (width == DMA_SLAVE_BUSWIDTH_2_BYTES)
+		chan->dcmd |= DCMD_WIDTH2;
+	else if (width == DMA_SLAVE_BUSWIDTH_4_BYTES)
+		chan->dcmd |= DCMD_WIDTH4;
+
+	if (maxburst == 8)
+		chan->dcmd |= DCMD_BURST8;
+	else if (maxburst == 16)
+		chan->dcmd |= DCMD_BURST16;
+	else if (maxburst == 32)
+		chan->dcmd |= DCMD_BURST32;
+
+	chan->dir = cfg->direction;
+	chan->dev_addr = addr;
+	/* FIXME: drivers should be ported over to use the filter
+	 * function. Once that's done, the following two lines can
+	 * be removed.
+	 */
+	if (cfg->slave_id)
+		chan->drcmr = cfg->slave_id;
+
+	return 0;
+}
+
+static int mmp_pdma_terminate_all(struct dma_chan *dchan)
+{
+	struct mmp_pdma_chan *chan = to_mmp_pdma_chan(dchan);
+	unsigned long flags;
+
+	if (!dchan)
+		return -EINVAL;
+
+	disable_chan(chan->phy);
+	mmp_pdma_free_phy(chan);
+	spin_lock_irqsave(&chan->desc_lock, flags);
+	mmp_pdma_free_desc_list(chan, &chan->chain_pending);
+	mmp_pdma_free_desc_list(chan, &chan->chain_running);
+	spin_unlock_irqrestore(&chan->desc_lock, flags);
+	chan->idle = true;
+
 	return 0;
 }
 
@@ -1061,7 +1063,8 @@
 	pdev->device.device_prep_slave_sg = mmp_pdma_prep_slave_sg;
 	pdev->device.device_prep_dma_cyclic = mmp_pdma_prep_dma_cyclic;
 	pdev->device.device_issue_pending = mmp_pdma_issue_pending;
-	pdev->device.device_control = mmp_pdma_control;
+	pdev->device.device_config = mmp_pdma_config;
+	pdev->device.device_terminate_all = mmp_pdma_terminate_all;
 	pdev->device.copy_align = PDMA_ALIGNMENT;
 
 	if (pdev->dev->coherent_dma_mask)
diff --git a/drivers/dma/mmp_tdma.c b/drivers/dma/mmp_tdma.c
index bfb4695..70c2fa9 100644
--- a/drivers/dma/mmp_tdma.c
+++ b/drivers/dma/mmp_tdma.c
@@ -19,7 +19,6 @@
 #include <linux/dmaengine.h>
 #include <linux/platform_device.h>
 #include <linux/device.h>
-#include <mach/regs-icu.h>
 #include <linux/platform_data/dma-mmp_tdma.h>
 #include <linux/of_device.h>
 #include <linux/of_dma.h>
@@ -164,33 +163,46 @@
 	tdmac->status = DMA_IN_PROGRESS;
 }
 
-static void mmp_tdma_disable_chan(struct mmp_tdma_chan *tdmac)
+static int mmp_tdma_disable_chan(struct dma_chan *chan)
 {
+	struct mmp_tdma_chan *tdmac = to_mmp_tdma_chan(chan);
+
 	writel(readl(tdmac->reg_base + TDCR) & ~TDCR_CHANEN,
 					tdmac->reg_base + TDCR);
 
 	tdmac->status = DMA_COMPLETE;
+
+	return 0;
 }
 
-static void mmp_tdma_resume_chan(struct mmp_tdma_chan *tdmac)
+static int mmp_tdma_resume_chan(struct dma_chan *chan)
 {
+	struct mmp_tdma_chan *tdmac = to_mmp_tdma_chan(chan);
+
 	writel(readl(tdmac->reg_base + TDCR) | TDCR_CHANEN,
 					tdmac->reg_base + TDCR);
 	tdmac->status = DMA_IN_PROGRESS;
+
+	return 0;
 }
 
-static void mmp_tdma_pause_chan(struct mmp_tdma_chan *tdmac)
+static int mmp_tdma_pause_chan(struct dma_chan *chan)
 {
+	struct mmp_tdma_chan *tdmac = to_mmp_tdma_chan(chan);
+
 	writel(readl(tdmac->reg_base + TDCR) & ~TDCR_CHANEN,
 					tdmac->reg_base + TDCR);
 	tdmac->status = DMA_PAUSED;
+
+	return 0;
 }
 
-static int mmp_tdma_config_chan(struct mmp_tdma_chan *tdmac)
+static int mmp_tdma_config_chan(struct dma_chan *chan)
 {
+	struct mmp_tdma_chan *tdmac = to_mmp_tdma_chan(chan);
 	unsigned int tdcr = 0;
 
-	mmp_tdma_disable_chan(tdmac);
+	mmp_tdma_disable_chan(chan);
 
 	if (tdmac->dir == DMA_MEM_TO_DEV)
 		tdcr = TDCR_DSTDIR_ADDR_HOLD | TDCR_SRCDIR_ADDR_INC;
@@ -452,42 +464,34 @@
 	return NULL;
 }
 
-static int mmp_tdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-		unsigned long arg)
+static int mmp_tdma_terminate_all(struct dma_chan *chan)
 {
 	struct mmp_tdma_chan *tdmac = to_mmp_tdma_chan(chan);
-	struct dma_slave_config *dmaengine_cfg = (void *)arg;
-	int ret = 0;
 
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		mmp_tdma_disable_chan(tdmac);
-		/* disable interrupt */
-		mmp_tdma_enable_irq(tdmac, false);
-		break;
-	case DMA_PAUSE:
-		mmp_tdma_pause_chan(tdmac);
-		break;
-	case DMA_RESUME:
-		mmp_tdma_resume_chan(tdmac);
-		break;
-	case DMA_SLAVE_CONFIG:
-		if (dmaengine_cfg->direction == DMA_DEV_TO_MEM) {
-			tdmac->dev_addr = dmaengine_cfg->src_addr;
-			tdmac->burst_sz = dmaengine_cfg->src_maxburst;
-			tdmac->buswidth = dmaengine_cfg->src_addr_width;
-		} else {
-			tdmac->dev_addr = dmaengine_cfg->dst_addr;
-			tdmac->burst_sz = dmaengine_cfg->dst_maxburst;
-			tdmac->buswidth = dmaengine_cfg->dst_addr_width;
-		}
-		tdmac->dir = dmaengine_cfg->direction;
-		return mmp_tdma_config_chan(tdmac);
-	default:
-		ret = -ENOSYS;
+	mmp_tdma_disable_chan(chan);
+	/* disable interrupt */
+	mmp_tdma_enable_irq(tdmac, false);
+
+	return 0;
+}
+
+static int mmp_tdma_config(struct dma_chan *chan,
+			   struct dma_slave_config *dmaengine_cfg)
+{
+	struct mmp_tdma_chan *tdmac = to_mmp_tdma_chan(chan);
+
+	if (dmaengine_cfg->direction == DMA_DEV_TO_MEM) {
+		tdmac->dev_addr = dmaengine_cfg->src_addr;
+		tdmac->burst_sz = dmaengine_cfg->src_maxburst;
+		tdmac->buswidth = dmaengine_cfg->src_addr_width;
+	} else {
+		tdmac->dev_addr = dmaengine_cfg->dst_addr;
+		tdmac->burst_sz = dmaengine_cfg->dst_maxburst;
+		tdmac->buswidth = dmaengine_cfg->dst_addr_width;
 	}
+	tdmac->dir = dmaengine_cfg->direction;
 
-	return ret;
+	return mmp_tdma_config_chan(chan);
 }
 
 static enum dma_status mmp_tdma_tx_status(struct dma_chan *chan,
@@ -668,7 +672,10 @@
 	tdev->device.device_prep_dma_cyclic = mmp_tdma_prep_dma_cyclic;
 	tdev->device.device_tx_status = mmp_tdma_tx_status;
 	tdev->device.device_issue_pending = mmp_tdma_issue_pending;
-	tdev->device.device_control = mmp_tdma_control;
+	tdev->device.device_config = mmp_tdma_config;
+	tdev->device.device_pause = mmp_tdma_pause_chan;
+	tdev->device.device_resume = mmp_tdma_resume_chan;
+	tdev->device.device_terminate_all = mmp_tdma_terminate_all;
 	tdev->device.copy_align = TDMA_ALIGNMENT;
 
 	dma_set_mask(&pdev->dev, DMA_BIT_MASK(64));
diff --git a/drivers/dma/moxart-dma.c b/drivers/dma/moxart-dma.c
index 53032ba..15cab7d 100644
--- a/drivers/dma/moxart-dma.c
+++ b/drivers/dma/moxart-dma.c
@@ -263,28 +263,6 @@
 	return 0;
 }
 
-static int moxart_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-			  unsigned long arg)
-{
-	int ret = 0;
-
-	switch (cmd) {
-	case DMA_PAUSE:
-	case DMA_RESUME:
-		return -EINVAL;
-	case DMA_TERMINATE_ALL:
-		moxart_terminate_all(chan);
-		break;
-	case DMA_SLAVE_CONFIG:
-		ret = moxart_slave_config(chan, (struct dma_slave_config *)arg);
-		break;
-	default:
-		ret = -ENOSYS;
-	}
-
-	return ret;
-}
-
 static struct dma_async_tx_descriptor *moxart_prep_slave_sg(
 	struct dma_chan *chan, struct scatterlist *sgl,
 	unsigned int sg_len, enum dma_transfer_direction dir,
@@ -531,7 +509,8 @@
 	dma->device_free_chan_resources		= moxart_free_chan_resources;
 	dma->device_issue_pending		= moxart_issue_pending;
 	dma->device_tx_status			= moxart_tx_status;
-	dma->device_control			= moxart_control;
+	dma->device_config			= moxart_slave_config;
+	dma->device_terminate_all		= moxart_terminate_all;
 	dma->dev				= dev;
 
 	INIT_LIST_HEAD(&dma->channels);
diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c
index 01bec40..57d2457 100644
--- a/drivers/dma/mpc512x_dma.c
+++ b/drivers/dma/mpc512x_dma.c
@@ -800,79 +800,69 @@
 	return NULL;
 }
 
-static int mpc_dma_device_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-							unsigned long arg)
+static int mpc_dma_device_config(struct dma_chan *chan,
+				 struct dma_slave_config *cfg)
 {
-	struct mpc_dma_chan *mchan;
-	struct mpc_dma *mdma;
-	struct dma_slave_config *cfg;
+	struct mpc_dma_chan *mchan = dma_chan_to_mpc_dma_chan(chan);
 	unsigned long flags;
 
-	mchan = dma_chan_to_mpc_dma_chan(chan);
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		/* Disable channel requests */
-		mdma = dma_chan_to_mpc_dma(chan);
+	/*
+	 * Software constraints:
+	 *  - only transfers between a peripheral device and
+	 *     memory are supported;
+	 *  - only peripheral devices with 4-byte FIFO access register
+	 *     are supported;
+	 *  - minimal transfer chunk is 4 bytes and consequently
+	 *     source and destination addresses must be 4-byte aligned
+	 *     and transfer size must be aligned on (4 * maxburst)
+	 *     boundary;
+	 *  - during the transfer RAM address is being incremented by
+	 *     the size of minimal transfer chunk;
+	 *  - peripheral port's address is constant during the transfer.
+	 */
 
-		spin_lock_irqsave(&mchan->lock, flags);
-
-		out_8(&mdma->regs->dmacerq, chan->chan_id);
-		list_splice_tail_init(&mchan->prepared, &mchan->free);
-		list_splice_tail_init(&mchan->queued, &mchan->free);
-		list_splice_tail_init(&mchan->active, &mchan->free);
-
-		spin_unlock_irqrestore(&mchan->lock, flags);
-
-		return 0;
-
-	case DMA_SLAVE_CONFIG:
-		/*
-		 * Software constraints:
-		 *  - only transfers between a peripheral device and
-		 *     memory are supported;
-		 *  - only peripheral devices with 4-byte FIFO access register
-		 *     are supported;
-		 *  - minimal transfer chunk is 4 bytes and consequently
-		 *     source and destination addresses must be 4-byte aligned
-		 *     and transfer size must be aligned on (4 * maxburst)
-		 *     boundary;
-		 *  - during the transfer RAM address is being incremented by
-		 *     the size of minimal transfer chunk;
-		 *  - peripheral port's address is constant during the transfer.
-		 */
-
-		cfg = (void *)arg;
-
-		if (cfg->src_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES ||
-		    cfg->dst_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES ||
-		    !IS_ALIGNED(cfg->src_addr, 4) ||
-		    !IS_ALIGNED(cfg->dst_addr, 4)) {
-			return -EINVAL;
-		}
-
-		spin_lock_irqsave(&mchan->lock, flags);
-
-		mchan->src_per_paddr = cfg->src_addr;
-		mchan->src_tcd_nunits = cfg->src_maxburst;
-		mchan->dst_per_paddr = cfg->dst_addr;
-		mchan->dst_tcd_nunits = cfg->dst_maxburst;
-
-		/* Apply defaults */
-		if (mchan->src_tcd_nunits == 0)
-			mchan->src_tcd_nunits = 1;
-		if (mchan->dst_tcd_nunits == 0)
-			mchan->dst_tcd_nunits = 1;
-
-		spin_unlock_irqrestore(&mchan->lock, flags);
-
-		return 0;
-
-	default:
-		/* Unknown command */
-		break;
+	if (cfg->src_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES ||
+	    cfg->dst_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES ||
+	    !IS_ALIGNED(cfg->src_addr, 4) ||
+	    !IS_ALIGNED(cfg->dst_addr, 4)) {
+		return -EINVAL;
 	}
 
-	return -ENXIO;
+	spin_lock_irqsave(&mchan->lock, flags);
+
+	mchan->src_per_paddr = cfg->src_addr;
+	mchan->src_tcd_nunits = cfg->src_maxburst;
+	mchan->dst_per_paddr = cfg->dst_addr;
+	mchan->dst_tcd_nunits = cfg->dst_maxburst;
+
+	/* Apply defaults */
+	if (mchan->src_tcd_nunits == 0)
+		mchan->src_tcd_nunits = 1;
+	if (mchan->dst_tcd_nunits == 0)
+		mchan->dst_tcd_nunits = 1;
+
+	spin_unlock_irqrestore(&mchan->lock, flags);
+
+	return 0;
+}
+
+static int mpc_dma_device_terminate_all(struct dma_chan *chan)
+{
+	struct mpc_dma_chan *mchan = dma_chan_to_mpc_dma_chan(chan);
+	struct mpc_dma *mdma = dma_chan_to_mpc_dma(chan);
+	unsigned long flags;
+
+	/* Disable channel requests */
+	spin_lock_irqsave(&mchan->lock, flags);
+
+	out_8(&mdma->regs->dmacerq, chan->chan_id);
+	list_splice_tail_init(&mchan->prepared, &mchan->free);
+	list_splice_tail_init(&mchan->queued, &mchan->free);
+	list_splice_tail_init(&mchan->active, &mchan->free);
+
+	spin_unlock_irqrestore(&mchan->lock, flags);
+
+	return 0;
 }
 
 static int mpc_dma_probe(struct platform_device *op)
@@ -963,7 +953,8 @@
 	dma->device_tx_status = mpc_dma_tx_status;
 	dma->device_prep_dma_memcpy = mpc_dma_prep_memcpy;
 	dma->device_prep_slave_sg = mpc_dma_prep_slave_sg;
-	dma->device_control = mpc_dma_device_control;
+	dma->device_config = mpc_dma_device_config;
+	dma->device_terminate_all = mpc_dma_device_terminate_all;
 
 	INIT_LIST_HEAD(&dma->channels);
 	dma_cap_set(DMA_MEMCPY, dma->cap_mask);
diff --git a/drivers/dma/mv_xor.c b/drivers/dma/mv_xor.c
index d7ac558..b03e813 100644
--- a/drivers/dma/mv_xor.c
+++ b/drivers/dma/mv_xor.c
@@ -928,14 +928,6 @@
 	return err;
 }
 
-/* This driver does not implement any of the optional DMA operations. */
-static int
-mv_xor_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-	       unsigned long arg)
-{
-	return -ENOSYS;
-}
-
 static int mv_xor_channel_remove(struct mv_xor_chan *mv_chan)
 {
 	struct dma_chan *chan, *_chan;
@@ -1008,7 +1000,6 @@
 	dma_dev->device_free_chan_resources = mv_xor_free_chan_resources;
 	dma_dev->device_tx_status = mv_xor_status;
 	dma_dev->device_issue_pending = mv_xor_issue_pending;
-	dma_dev->device_control = mv_xor_control;
 	dma_dev->dev = &pdev->dev;
 
 	/* set prep routines based on capability */
diff --git a/drivers/dma/mxs-dma.c b/drivers/dma/mxs-dma.c
index 5ea6120..829ec68 100644
--- a/drivers/dma/mxs-dma.c
+++ b/drivers/dma/mxs-dma.c
@@ -202,8 +202,9 @@
 	return container_of(chan, struct mxs_dma_chan, chan);
 }
 
-static void mxs_dma_reset_chan(struct mxs_dma_chan *mxs_chan)
+static void mxs_dma_reset_chan(struct dma_chan *chan)
 {
+	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
 	int chan_id = mxs_chan->chan.chan_id;
 
@@ -250,8 +251,9 @@
 	mxs_chan->status = DMA_COMPLETE;
 }
 
-static void mxs_dma_enable_chan(struct mxs_dma_chan *mxs_chan)
+static void mxs_dma_enable_chan(struct dma_chan *chan)
 {
+	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
 	int chan_id = mxs_chan->chan.chan_id;
 
@@ -272,13 +274,16 @@
 	mxs_chan->reset = false;
 }
 
-static void mxs_dma_disable_chan(struct mxs_dma_chan *mxs_chan)
+static void mxs_dma_disable_chan(struct dma_chan *chan)
 {
+	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
+
 	mxs_chan->status = DMA_COMPLETE;
 }
 
-static void mxs_dma_pause_chan(struct mxs_dma_chan *mxs_chan)
+static int mxs_dma_pause_chan(struct dma_chan *chan)
 {
+	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
 	int chan_id = mxs_chan->chan.chan_id;
 
@@ -291,10 +296,12 @@
 			mxs_dma->base + HW_APBHX_CHANNEL_CTRL + STMP_OFFSET_REG_SET);
 
 	mxs_chan->status = DMA_PAUSED;
+	return 0;
 }
 
-static void mxs_dma_resume_chan(struct mxs_dma_chan *mxs_chan)
+static int mxs_dma_resume_chan(struct dma_chan *chan)
 {
+	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
 	int chan_id = mxs_chan->chan.chan_id;
 
@@ -307,6 +314,7 @@
 			mxs_dma->base + HW_APBHX_CHANNEL_CTRL + STMP_OFFSET_REG_CLR);
 
 	mxs_chan->status = DMA_IN_PROGRESS;
+	return 0;
 }
 
 static dma_cookie_t mxs_dma_tx_submit(struct dma_async_tx_descriptor *tx)
@@ -383,7 +391,7 @@
 			"%s: error in channel %d\n", __func__,
 			chan);
 		mxs_chan->status = DMA_ERROR;
-		mxs_dma_reset_chan(mxs_chan);
+		mxs_dma_reset_chan(&mxs_chan->chan);
 	} else if (mxs_chan->status != DMA_COMPLETE) {
 		if (mxs_chan->flags & MXS_DMA_SG_LOOP) {
 			mxs_chan->status = DMA_IN_PROGRESS;
@@ -432,7 +440,7 @@
 	if (ret)
 		goto err_clk;
 
-	mxs_dma_reset_chan(mxs_chan);
+	mxs_dma_reset_chan(chan);
 
 	dma_async_tx_descriptor_init(&mxs_chan->desc, chan);
 	mxs_chan->desc.tx_submit = mxs_dma_tx_submit;
@@ -456,7 +464,7 @@
 	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
 
-	mxs_dma_disable_chan(mxs_chan);
+	mxs_dma_disable_chan(chan);
 
 	free_irq(mxs_chan->chan_irq, mxs_dma);
 
@@ -651,28 +659,12 @@
 	return NULL;
 }
 
-static int mxs_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-		unsigned long arg)
+static int mxs_dma_terminate_all(struct dma_chan *chan)
 {
-	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
-	int ret = 0;
+	mxs_dma_reset_chan(chan);
+	mxs_dma_disable_chan(chan);
 
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		mxs_dma_reset_chan(mxs_chan);
-		mxs_dma_disable_chan(mxs_chan);
-		break;
-	case DMA_PAUSE:
-		mxs_dma_pause_chan(mxs_chan);
-		break;
-	case DMA_RESUME:
-		mxs_dma_resume_chan(mxs_chan);
-		break;
-	default:
-		ret = -ENOSYS;
-	}
-
-	return ret;
+	return 0;
 }
 
 static enum dma_status mxs_dma_tx_status(struct dma_chan *chan,
@@ -701,13 +693,6 @@
 	return mxs_chan->status;
 }
 
-static void mxs_dma_issue_pending(struct dma_chan *chan)
-{
-	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
-
-	mxs_dma_enable_chan(mxs_chan);
-}
-
 static int __init mxs_dma_init(struct mxs_dma_engine *mxs_dma)
 {
 	int ret;
@@ -860,8 +845,14 @@
 	mxs_dma->dma_device.device_tx_status = mxs_dma_tx_status;
 	mxs_dma->dma_device.device_prep_slave_sg = mxs_dma_prep_slave_sg;
 	mxs_dma->dma_device.device_prep_dma_cyclic = mxs_dma_prep_dma_cyclic;
-	mxs_dma->dma_device.device_control = mxs_dma_control;
-	mxs_dma->dma_device.device_issue_pending = mxs_dma_issue_pending;
+	mxs_dma->dma_device.device_pause = mxs_dma_pause_chan;
+	mxs_dma->dma_device.device_resume = mxs_dma_resume_chan;
+	mxs_dma->dma_device.device_terminate_all = mxs_dma_terminate_all;
+	mxs_dma->dma_device.src_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_4_BYTES);
+	mxs_dma->dma_device.dst_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_4_BYTES);
+	mxs_dma->dma_device.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	mxs_dma->dma_device.residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+	mxs_dma->dma_device.device_issue_pending = mxs_dma_enable_chan;
 
 	ret = dma_async_device_register(&mxs_dma->dma_device);
 	if (ret) {
diff --git a/drivers/dma/nbpfaxi.c b/drivers/dma/nbpfaxi.c
index d7d61e1..88b77c9 100644
--- a/drivers/dma/nbpfaxi.c
+++ b/drivers/dma/nbpfaxi.c
@@ -504,7 +504,7 @@
 	 * pauses DMA and reads out data received via DMA as well as those left
 	 * in the Rx FIFO. For this to work with the RAM side using burst
 	 * transfers we enable the SBE bit and terminate the transfer in our
-	 * DMA_PAUSE handler.
+	 * .device_pause handler.
 	 */
 	mem_xfer = nbpf_xfer_ds(chan->nbpf, size);
 
@@ -565,13 +565,6 @@
 	nbpf_write(nbpf, NBPF_CTRL, NBPF_CTRL_LVINT);
 }
 
-static void nbpf_pause(struct nbpf_channel *chan)
-{
-	nbpf_chan_write(chan, NBPF_CHAN_CTRL, NBPF_CHAN_CTRL_SETSUS);
-	/* See comment in nbpf_prep_one() */
-	nbpf_chan_write(chan, NBPF_CHAN_CTRL, NBPF_CHAN_CTRL_CLREN);
-}
-
 /*		Generic part			*/
 
 /* DMA ENGINE functions */
@@ -837,54 +830,58 @@
 	}
 }
 
-static int nbpf_control(struct dma_chan *dchan, enum dma_ctrl_cmd cmd,
-			unsigned long arg)
+static int nbpf_pause(struct dma_chan *dchan)
 {
 	struct nbpf_channel *chan = nbpf_to_chan(dchan);
-	struct dma_slave_config *config;
 
-	dev_dbg(dchan->device->dev, "Entry %s(%d)\n", __func__, cmd);
+	dev_dbg(dchan->device->dev, "Entry %s\n", __func__);
 
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		dev_dbg(dchan->device->dev, "Terminating\n");
-		nbpf_chan_halt(chan);
-		nbpf_chan_idle(chan);
-		break;
+	chan->paused = true;
+	nbpf_chan_write(chan, NBPF_CHAN_CTRL, NBPF_CHAN_CTRL_SETSUS);
+	/* See comment in nbpf_prep_one() */
+	nbpf_chan_write(chan, NBPF_CHAN_CTRL, NBPF_CHAN_CTRL_CLREN);
 
-	case DMA_SLAVE_CONFIG:
-		if (!arg)
-			return -EINVAL;
-		config = (struct dma_slave_config *)arg;
+	return 0;
+}
 
-		/*
-		 * We could check config->slave_id to match chan->terminal here,
-		 * but with DT they would be coming from the same source, so
-		 * such a check would be superflous
-		 */
+static int nbpf_terminate_all(struct dma_chan *dchan)
+{
+	struct nbpf_channel *chan = nbpf_to_chan(dchan);
 
-		chan->slave_dst_addr = config->dst_addr;
-		chan->slave_dst_width = nbpf_xfer_size(chan->nbpf,
-						       config->dst_addr_width, 1);
-		chan->slave_dst_burst = nbpf_xfer_size(chan->nbpf,
-						       config->dst_addr_width,
-						       config->dst_maxburst);
-		chan->slave_src_addr = config->src_addr;
-		chan->slave_src_width = nbpf_xfer_size(chan->nbpf,
-						       config->src_addr_width, 1);
-		chan->slave_src_burst = nbpf_xfer_size(chan->nbpf,
-						       config->src_addr_width,
-						       config->src_maxburst);
-		break;
+	dev_dbg(dchan->device->dev, "Entry %s\n", __func__);
+	dev_dbg(dchan->device->dev, "Terminating\n");
 
-	case DMA_PAUSE:
-		chan->paused = true;
-		nbpf_pause(chan);
-		break;
+	nbpf_chan_halt(chan);
+	nbpf_chan_idle(chan);
 
-	default:
-		return -ENXIO;
-	}
+	return 0;
+}
+
+static int nbpf_config(struct dma_chan *dchan,
+		       struct dma_slave_config *config)
+{
+	struct nbpf_channel *chan = nbpf_to_chan(dchan);
+
+	dev_dbg(dchan->device->dev, "Entry %s\n", __func__);
+
+	/*
+	 * We could check config->slave_id to match chan->terminal here,
+	 * but with DT they would be coming from the same source, so
+	 * such a check would be superflous
+	 */
+
+	chan->slave_dst_addr = config->dst_addr;
+	chan->slave_dst_width = nbpf_xfer_size(chan->nbpf,
+					       config->dst_addr_width, 1);
+	chan->slave_dst_burst = nbpf_xfer_size(chan->nbpf,
+					       config->dst_addr_width,
+					       config->dst_maxburst);
+	chan->slave_src_addr = config->src_addr;
+	chan->slave_src_width = nbpf_xfer_size(chan->nbpf,
+					       config->src_addr_width, 1);
+	chan->slave_src_burst = nbpf_xfer_size(chan->nbpf,
+					       config->src_addr_width,
+					       config->src_maxburst);
 
 	return 0;
 }
@@ -1072,18 +1069,6 @@
 	}
 }
 
-static int nbpf_slave_caps(struct dma_chan *dchan,
-			   struct dma_slave_caps *caps)
-{
-	caps->src_addr_widths = NBPF_DMA_BUSWIDTHS;
-	caps->dstn_addr_widths = NBPF_DMA_BUSWIDTHS;
-	caps->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
-	caps->cmd_pause = false;
-	caps->cmd_terminate = true;
-
-	return 0;
-}
-
 static struct dma_chan *nbpf_of_xlate(struct of_phandle_args *dma_spec,
 				      struct of_dma *ofdma)
 {
@@ -1414,7 +1399,6 @@
 	dma_dev->device_prep_dma_memcpy = nbpf_prep_memcpy;
 	dma_dev->device_tx_status = nbpf_tx_status;
 	dma_dev->device_issue_pending = nbpf_issue_pending;
-	dma_dev->device_slave_caps = nbpf_slave_caps;
 
 	/*
 	 * If we drop support for unaligned MEMCPY buffer addresses and / or
@@ -1426,7 +1410,13 @@
 
 	/* Compulsory for DMA_SLAVE fields */
 	dma_dev->device_prep_slave_sg = nbpf_prep_slave_sg;
-	dma_dev->device_control = nbpf_control;
+	dma_dev->device_config = nbpf_config;
+	dma_dev->device_pause = nbpf_pause;
+	dma_dev->device_terminate_all = nbpf_terminate_all;
+
+	dma_dev->src_addr_widths = NBPF_DMA_BUSWIDTHS;
+	dma_dev->dst_addr_widths = NBPF_DMA_BUSWIDTHS;
+	dma_dev->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
 
 	platform_set_drvdata(pdev, nbpf);
 
diff --git a/drivers/dma/of-dma.c b/drivers/dma/of-dma.c
index d5fbeaa..ca31f1b 100644
--- a/drivers/dma/of-dma.c
+++ b/drivers/dma/of-dma.c
@@ -159,6 +159,10 @@
 		return ERR_PTR(-ENODEV);
 	}
 
+	/* Silently fail if there is not even the "dmas" property */
+	if (!of_find_property(np, "dmas", NULL))
+		return ERR_PTR(-ENODEV);
+
 	count = of_property_count_strings(np, "dma-names");
 	if (count < 0) {
 		pr_err("%s: dma-names property of node '%s' missing or empty\n",
diff --git a/drivers/dma/omap-dma.c b/drivers/dma/omap-dma.c
index c0016a6..7dd6dd1 100644
--- a/drivers/dma/omap-dma.c
+++ b/drivers/dma/omap-dma.c
@@ -948,8 +948,10 @@
 	return vchan_tx_prep(&c->vc, &d->vd, flags);
 }
 
-static int omap_dma_slave_config(struct omap_chan *c, struct dma_slave_config *cfg)
+static int omap_dma_slave_config(struct dma_chan *chan, struct dma_slave_config *cfg)
 {
+	struct omap_chan *c = to_omap_dma_chan(chan);
+
 	if (cfg->src_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES ||
 	    cfg->dst_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES)
 		return -EINVAL;
@@ -959,8 +961,9 @@
 	return 0;
 }
 
-static int omap_dma_terminate_all(struct omap_chan *c)
+static int omap_dma_terminate_all(struct dma_chan *chan)
 {
+	struct omap_chan *c = to_omap_dma_chan(chan);
 	struct omap_dmadev *d = to_omap_dma_dev(c->vc.chan.device);
 	unsigned long flags;
 	LIST_HEAD(head);
@@ -996,8 +999,10 @@
 	return 0;
 }
 
-static int omap_dma_pause(struct omap_chan *c)
+static int omap_dma_pause(struct dma_chan *chan)
 {
+	struct omap_chan *c = to_omap_dma_chan(chan);
+
 	/* Pause/Resume only allowed with cyclic mode */
 	if (!c->cyclic)
 		return -EINVAL;
@@ -1010,8 +1015,10 @@
 	return 0;
 }
 
-static int omap_dma_resume(struct omap_chan *c)
+static int omap_dma_resume(struct dma_chan *chan)
 {
+	struct omap_chan *c = to_omap_dma_chan(chan);
+
 	/* Pause/Resume only allowed with cyclic mode */
 	if (!c->cyclic)
 		return -EINVAL;
@@ -1029,37 +1036,6 @@
 	return 0;
 }
 
-static int omap_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-	unsigned long arg)
-{
-	struct omap_chan *c = to_omap_dma_chan(chan);
-	int ret;
-
-	switch (cmd) {
-	case DMA_SLAVE_CONFIG:
-		ret = omap_dma_slave_config(c, (struct dma_slave_config *)arg);
-		break;
-
-	case DMA_TERMINATE_ALL:
-		ret = omap_dma_terminate_all(c);
-		break;
-
-	case DMA_PAUSE:
-		ret = omap_dma_pause(c);
-		break;
-
-	case DMA_RESUME:
-		ret = omap_dma_resume(c);
-		break;
-
-	default:
-		ret = -ENXIO;
-		break;
-	}
-
-	return ret;
-}
-
 static int omap_dma_chan_init(struct omap_dmadev *od, int dma_sig)
 {
 	struct omap_chan *c;
@@ -1094,19 +1070,6 @@
 				 BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) | \
 				 BIT(DMA_SLAVE_BUSWIDTH_4_BYTES))
 
-static int omap_dma_device_slave_caps(struct dma_chan *dchan,
-				      struct dma_slave_caps *caps)
-{
-	caps->src_addr_widths = OMAP_DMA_BUSWIDTHS;
-	caps->dstn_addr_widths = OMAP_DMA_BUSWIDTHS;
-	caps->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
-	caps->cmd_pause = true;
-	caps->cmd_terminate = true;
-	caps->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
-
-	return 0;
-}
-
 static int omap_dma_probe(struct platform_device *pdev)
 {
 	struct omap_dmadev *od;
@@ -1136,8 +1099,14 @@
 	od->ddev.device_issue_pending = omap_dma_issue_pending;
 	od->ddev.device_prep_slave_sg = omap_dma_prep_slave_sg;
 	od->ddev.device_prep_dma_cyclic = omap_dma_prep_dma_cyclic;
-	od->ddev.device_control = omap_dma_control;
-	od->ddev.device_slave_caps = omap_dma_device_slave_caps;
+	od->ddev.device_config = omap_dma_slave_config;
+	od->ddev.device_pause = omap_dma_pause;
+	od->ddev.device_resume = omap_dma_resume;
+	od->ddev.device_terminate_all = omap_dma_terminate_all;
+	od->ddev.src_addr_widths = OMAP_DMA_BUSWIDTHS;
+	od->ddev.dst_addr_widths = OMAP_DMA_BUSWIDTHS;
+	od->ddev.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	od->ddev.residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
 	od->ddev.dev = &pdev->dev;
 	INIT_LIST_HEAD(&od->ddev.channels);
 	INIT_LIST_HEAD(&od->pending);
diff --git a/drivers/dma/pch_dma.c b/drivers/dma/pch_dma.c
index 6e0e47d..35c143c 100644
--- a/drivers/dma/pch_dma.c
+++ b/drivers/dma/pch_dma.c
@@ -665,16 +665,12 @@
 	return NULL;
 }
 
-static int pd_device_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-			     unsigned long arg)
+static int pd_device_terminate_all(struct dma_chan *chan)
 {
 	struct pch_dma_chan *pd_chan = to_pd_chan(chan);
 	struct pch_dma_desc *desc, *_d;
 	LIST_HEAD(list);
 
-	if (cmd != DMA_TERMINATE_ALL)
-		return -ENXIO;
-
 	spin_lock_irq(&pd_chan->lock);
 
 	pdc_set_mode(&pd_chan->chan, DMA_CTL0_DISABLE);
@@ -932,7 +928,7 @@
 	pd->dma.device_tx_status = pd_tx_status;
 	pd->dma.device_issue_pending = pd_issue_pending;
 	pd->dma.device_prep_slave_sg = pd_prep_slave_sg;
-	pd->dma.device_control = pd_device_control;
+	pd->dma.device_terminate_all = pd_device_terminate_all;
 
 	err = dma_async_device_register(&pd->dma);
 	if (err) {
diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c
index bdf40b5..0e1f5677 100644
--- a/drivers/dma/pl330.c
+++ b/drivers/dma/pl330.c
@@ -504,6 +504,9 @@
 
 	enum desc_status status;
 
+	int bytes_requested;
+	bool last;
+
 	/* The channel which currently holds this desc */
 	struct dma_pl330_chan *pchan;
 
@@ -1048,6 +1051,10 @@
 	if (!req)
 		return true;
 
+	/* Return if req is running */
+	if (idx == thrd->req_running)
+		return true;
+
 	desc = req->desc;
 
 	ns = desc->rqcfg.nonsecure ? 1 : 0;
@@ -1587,6 +1594,8 @@
 			descdone = thrd->req[active].desc;
 			thrd->req[active].desc = NULL;
 
+			thrd->req_running = -1;
+
 			/* Get going again ASAP */
 			_start(thrd);
 
@@ -2086,78 +2095,90 @@
 	return 1;
 }
 
-static int pl330_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, unsigned long arg)
+static int pl330_config(struct dma_chan *chan,
+			struct dma_slave_config *slave_config)
+{
+	struct dma_pl330_chan *pch = to_pchan(chan);
+
+	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_DEV_TO_MEM) {
+		if (slave_config->src_addr)
+			pch->fifo_addr = slave_config->src_addr;
+		if (slave_config->src_addr_width)
+			pch->burst_sz = __ffs(slave_config->src_addr_width);
+		if (slave_config->src_maxburst)
+			pch->burst_len = slave_config->src_maxburst;
+	}
+
+	return 0;
+}
+
+static int pl330_terminate_all(struct dma_chan *chan)
 {
 	struct dma_pl330_chan *pch = to_pchan(chan);
 	struct dma_pl330_desc *desc;
 	unsigned long flags;
 	struct pl330_dmac *pl330 = pch->dmac;
-	struct dma_slave_config *slave_config;
 	LIST_HEAD(list);
 
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		pm_runtime_get_sync(pl330->ddma.dev);
-		spin_lock_irqsave(&pch->lock, flags);
+	spin_lock_irqsave(&pch->lock, flags);
+	spin_lock(&pl330->lock);
+	_stop(pch->thread);
+	spin_unlock(&pl330->lock);
 
-		spin_lock(&pl330->lock);
-		_stop(pch->thread);
-		spin_unlock(&pl330->lock);
+	pch->thread->req[0].desc = NULL;
+	pch->thread->req[1].desc = NULL;
+	pch->thread->req_running = -1;
 
-		pch->thread->req[0].desc = NULL;
-		pch->thread->req[1].desc = NULL;
-		pch->thread->req_running = -1;
-
-		/* Mark all desc done */
-		list_for_each_entry(desc, &pch->submitted_list, node) {
-			desc->status = FREE;
-			dma_cookie_complete(&desc->txd);
-		}
-
-		list_for_each_entry(desc, &pch->work_list , node) {
-			desc->status = FREE;
-			dma_cookie_complete(&desc->txd);
-		}
-
-		list_for_each_entry(desc, &pch->completed_list , node) {
-			desc->status = FREE;
-			dma_cookie_complete(&desc->txd);
-		}
-
-		if (!list_empty(&pch->work_list))
-			pm_runtime_put(pl330->ddma.dev);
-
-		list_splice_tail_init(&pch->submitted_list, &pl330->desc_pool);
-		list_splice_tail_init(&pch->work_list, &pl330->desc_pool);
-		list_splice_tail_init(&pch->completed_list, &pl330->desc_pool);
-		spin_unlock_irqrestore(&pch->lock, flags);
-		pm_runtime_mark_last_busy(pl330->ddma.dev);
-		pm_runtime_put_autosuspend(pl330->ddma.dev);
-		break;
-	case DMA_SLAVE_CONFIG:
-		slave_config = (struct dma_slave_config *)arg;
-
-		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_DEV_TO_MEM) {
-			if (slave_config->src_addr)
-				pch->fifo_addr = slave_config->src_addr;
-			if (slave_config->src_addr_width)
-				pch->burst_sz = __ffs(slave_config->src_addr_width);
-			if (slave_config->src_maxburst)
-				pch->burst_len = slave_config->src_maxburst;
-		}
-		break;
-	default:
-		dev_err(pch->dmac->ddma.dev, "Not supported command.\n");
-		return -ENXIO;
+	/* Mark all desc done */
+	list_for_each_entry(desc, &pch->submitted_list, node) {
+		desc->status = FREE;
+		dma_cookie_complete(&desc->txd);
 	}
 
+	list_for_each_entry(desc, &pch->work_list , node) {
+		desc->status = FREE;
+		dma_cookie_complete(&desc->txd);
+	}
+
+	list_splice_tail_init(&pch->submitted_list, &pl330->desc_pool);
+	list_splice_tail_init(&pch->work_list, &pl330->desc_pool);
+	list_splice_tail_init(&pch->completed_list, &pl330->desc_pool);
+	spin_unlock_irqrestore(&pch->lock, flags);
+
+	return 0;
+}
+
+/*
+ * We don't support DMA_RESUME command because of hardware
+ * limitations, so after pausing the channel we cannot restore
+ * it to active state. We have to terminate channel and setup
+ * DMA transfer again. This pause feature was implemented to
+ * allow safely read residue before channel termination.
+ */
+int pl330_pause(struct dma_chan *chan)
+{
+	struct dma_pl330_chan *pch = to_pchan(chan);
+	struct pl330_dmac *pl330 = pch->dmac;
+	unsigned long flags;
+
+	pm_runtime_get_sync(pl330->ddma.dev);
+	spin_lock_irqsave(&pch->lock, flags);
+
+	spin_lock(&pl330->lock);
+	_stop(pch->thread);
+	spin_unlock(&pl330->lock);
+
+	spin_unlock_irqrestore(&pch->lock, flags);
+	pm_runtime_mark_last_busy(pl330->ddma.dev);
+	pm_runtime_put_autosuspend(pl330->ddma.dev);
+
 	return 0;
 }
 
@@ -2182,11 +2203,74 @@
 	pm_runtime_put_autosuspend(pch->dmac->ddma.dev);
 }
 
+int pl330_get_current_xferred_count(struct dma_pl330_chan *pch,
+		struct dma_pl330_desc *desc)
+{
+	struct pl330_thread *thrd = pch->thread;
+	struct pl330_dmac *pl330 = pch->dmac;
+	void __iomem *regs = thrd->dmac->base;
+	u32 val, addr;
+
+	pm_runtime_get_sync(pl330->ddma.dev);
+	val = addr = 0;
+	if (desc->rqcfg.src_inc) {
+		val = readl(regs + SA(thrd->id));
+		addr = desc->px.src_addr;
+	} else {
+		val = readl(regs + DA(thrd->id));
+		addr = desc->px.dst_addr;
+	}
+	pm_runtime_mark_last_busy(pch->dmac->ddma.dev);
+	pm_runtime_put_autosuspend(pl330->ddma.dev);
+	return val - addr;
+}
+
 static enum dma_status
 pl330_tx_status(struct dma_chan *chan, dma_cookie_t cookie,
 		 struct dma_tx_state *txstate)
 {
-	return dma_cookie_status(chan, cookie, txstate);
+	enum dma_status ret;
+	unsigned long flags;
+	struct dma_pl330_desc *desc, *running = NULL;
+	struct dma_pl330_chan *pch = to_pchan(chan);
+	unsigned int transferred, residual = 0;
+
+	ret = dma_cookie_status(chan, cookie, txstate);
+
+	if (!txstate)
+		return ret;
+
+	if (ret == DMA_COMPLETE)
+		goto out;
+
+	spin_lock_irqsave(&pch->lock, flags);
+
+	if (pch->thread->req_running != -1)
+		running = pch->thread->req[pch->thread->req_running].desc;
+
+	/* Check in pending list */
+	list_for_each_entry(desc, &pch->work_list, node) {
+		if (desc->status == DONE)
+			transferred = desc->bytes_requested;
+		else if (running && desc == running)
+			transferred =
+				pl330_get_current_xferred_count(pch, desc);
+		else
+			transferred = 0;
+		residual += desc->bytes_requested - transferred;
+		if (desc->txd.cookie == cookie) {
+			ret = desc->status;
+			break;
+		}
+		if (desc->last)
+			residual = 0;
+	}
+	spin_unlock_irqrestore(&pch->lock, flags);
+
+out:
+	dma_set_residue(txstate, residual);
+
+	return ret;
 }
 
 static void pl330_issue_pending(struct dma_chan *chan)
@@ -2231,12 +2315,14 @@
 			desc->txd.callback = last->txd.callback;
 			desc->txd.callback_param = last->txd.callback_param;
 		}
+		last->last = false;
 
 		dma_cookie_assign(&desc->txd);
 
 		list_move_tail(&desc->node, &pch->submitted_list);
 	}
 
+	last->last = true;
 	cookie = dma_cookie_assign(&last->txd);
 	list_add_tail(&last->node, &pch->submitted_list);
 	spin_unlock_irqrestore(&pch->lock, flags);
@@ -2459,6 +2545,7 @@
 		desc->rqtype = direction;
 		desc->rqcfg.brst_size = pch->burst_sz;
 		desc->rqcfg.brst_len = 1;
+		desc->bytes_requested = period_len;
 		fill_px(&desc->px, dst, src, period_len);
 
 		if (!first)
@@ -2601,6 +2688,7 @@
 		desc->rqcfg.brst_size = pch->burst_sz;
 		desc->rqcfg.brst_len = 1;
 		desc->rqtype = direction;
+		desc->bytes_requested = sg_dma_len(sg);
 	}
 
 	/* Return the last desc in the chain */
@@ -2623,19 +2711,6 @@
 	BIT(DMA_SLAVE_BUSWIDTH_4_BYTES) | \
 	BIT(DMA_SLAVE_BUSWIDTH_8_BYTES)
 
-static int pl330_dma_device_slave_caps(struct dma_chan *dchan,
-	struct dma_slave_caps *caps)
-{
-	caps->src_addr_widths = PL330_DMA_BUSWIDTHS;
-	caps->dstn_addr_widths = PL330_DMA_BUSWIDTHS;
-	caps->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
-	caps->cmd_pause = false;
-	caps->cmd_terminate = true;
-	caps->residue_granularity = DMA_RESIDUE_GRANULARITY_DESCRIPTOR;
-
-	return 0;
-}
-
 /*
  * Runtime PM callbacks are provided by amba/bus.c driver.
  *
@@ -2793,9 +2868,14 @@
 	pd->device_prep_dma_cyclic = pl330_prep_dma_cyclic;
 	pd->device_tx_status = pl330_tx_status;
 	pd->device_prep_slave_sg = pl330_prep_slave_sg;
-	pd->device_control = pl330_control;
+	pd->device_config = pl330_config;
+	pd->device_pause = pl330_pause;
+	pd->device_terminate_all = pl330_terminate_all;
 	pd->device_issue_pending = pl330_issue_pending;
-	pd->device_slave_caps = pl330_dma_device_slave_caps;
+	pd->src_addr_widths = PL330_DMA_BUSWIDTHS;
+	pd->dst_addr_widths = PL330_DMA_BUSWIDTHS;
+	pd->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	pd->residue_granularity = DMA_RESIDUE_GRANULARITY_SEGMENT;
 
 	ret = dma_async_device_register(pd);
 	if (ret) {
@@ -2847,7 +2927,7 @@
 
 		/* Flush the channel */
 		if (pch->thread) {
-			pl330_control(&pch->chan, DMA_TERMINATE_ALL, 0);
+			pl330_terminate_all(&pch->chan);
 			pl330_free_chan_resources(&pch->chan);
 		}
 	}
@@ -2878,7 +2958,7 @@
 
 		/* Flush the channel */
 		if (pch->thread) {
-			pl330_control(&pch->chan, DMA_TERMINATE_ALL, 0);
+			pl330_terminate_all(&pch->chan);
 			pl330_free_chan_resources(&pch->chan);
 		}
 	}
diff --git a/drivers/dma/qcom_bam_dma.c b/drivers/dma/qcom_bam_dma.c
index 3122a99..d7a33b3 100644
--- a/drivers/dma/qcom_bam_dma.c
+++ b/drivers/dma/qcom_bam_dma.c
@@ -530,11 +530,18 @@
  * Sets slave configuration for channel
  *
  */
-static void bam_slave_config(struct bam_chan *bchan,
-		struct dma_slave_config *cfg)
+static int bam_slave_config(struct dma_chan *chan,
+			    struct dma_slave_config *cfg)
 {
+	struct bam_chan *bchan = to_bam_chan(chan);
+	unsigned long flag;
+
+	spin_lock_irqsave(&bchan->vc.lock, flag);
 	memcpy(&bchan->slave, cfg, sizeof(*cfg));
 	bchan->reconfigure = 1;
+	spin_unlock_irqrestore(&bchan->vc.lock, flag);
+
+	return 0;
 }
 
 /**
@@ -627,8 +634,9 @@
  * No callbacks are done
  *
  */
-static void bam_dma_terminate_all(struct bam_chan *bchan)
+static int bam_dma_terminate_all(struct dma_chan *chan)
 {
+	struct bam_chan *bchan = to_bam_chan(chan);
 	unsigned long flag;
 	LIST_HEAD(head);
 
@@ -643,56 +651,46 @@
 	spin_unlock_irqrestore(&bchan->vc.lock, flag);
 
 	vchan_dma_desc_free_list(&bchan->vc, &head);
+
+	return 0;
 }
 
 /**
- * bam_control - DMA device control
+ * bam_pause - Pause DMA channel
  * @chan: dma channel
- * @cmd: control cmd
- * @arg: cmd argument
- *
- * Perform DMA control command
  *
  */
-static int bam_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-	unsigned long arg)
+static int bam_pause(struct dma_chan *chan)
 {
 	struct bam_chan *bchan = to_bam_chan(chan);
 	struct bam_device *bdev = bchan->bdev;
-	int ret = 0;
 	unsigned long flag;
 
-	switch (cmd) {
-	case DMA_PAUSE:
-		spin_lock_irqsave(&bchan->vc.lock, flag);
-		writel_relaxed(1, bam_addr(bdev, bchan->id, BAM_P_HALT));
-		bchan->paused = 1;
-		spin_unlock_irqrestore(&bchan->vc.lock, flag);
-		break;
+	spin_lock_irqsave(&bchan->vc.lock, flag);
+	writel_relaxed(1, bam_addr(bdev, bchan->id, BAM_P_HALT));
+	bchan->paused = 1;
+	spin_unlock_irqrestore(&bchan->vc.lock, flag);
 
-	case DMA_RESUME:
-		spin_lock_irqsave(&bchan->vc.lock, flag);
-		writel_relaxed(0, bam_addr(bdev, bchan->id, BAM_P_HALT));
-		bchan->paused = 0;
-		spin_unlock_irqrestore(&bchan->vc.lock, flag);
-		break;
+	return 0;
+}
 
-	case DMA_TERMINATE_ALL:
-		bam_dma_terminate_all(bchan);
-		break;
+/**
+ * bam_resume - Resume DMA channel operations
+ * @chan: dma channel
+ *
+ */
+static int bam_resume(struct dma_chan *chan)
+{
+	struct bam_chan *bchan = to_bam_chan(chan);
+	struct bam_device *bdev = bchan->bdev;
+	unsigned long flag;
 
-	case DMA_SLAVE_CONFIG:
-		spin_lock_irqsave(&bchan->vc.lock, flag);
-		bam_slave_config(bchan, (struct dma_slave_config *)arg);
-		spin_unlock_irqrestore(&bchan->vc.lock, flag);
-		break;
+	spin_lock_irqsave(&bchan->vc.lock, flag);
+	writel_relaxed(0, bam_addr(bdev, bchan->id, BAM_P_HALT));
+	bchan->paused = 0;
+	spin_unlock_irqrestore(&bchan->vc.lock, flag);
 
-	default:
-		ret = -ENXIO;
-		break;
-	}
-
-	return ret;
+	return 0;
 }
 
 /**
@@ -1148,7 +1146,10 @@
 	bdev->common.device_alloc_chan_resources = bam_alloc_chan;
 	bdev->common.device_free_chan_resources = bam_free_chan;
 	bdev->common.device_prep_slave_sg = bam_prep_slave_sg;
-	bdev->common.device_control = bam_control;
+	bdev->common.device_config = bam_slave_config;
+	bdev->common.device_pause = bam_pause;
+	bdev->common.device_resume = bam_resume;
+	bdev->common.device_terminate_all = bam_dma_terminate_all;
 	bdev->common.device_issue_pending = bam_issue_pending;
 	bdev->common.device_tx_status = bam_tx_status;
 	bdev->common.dev = bdev->dev;
@@ -1187,7 +1188,7 @@
 	devm_free_irq(bdev->dev, bdev->irq, bdev);
 
 	for (i = 0; i < bdev->num_channels; i++) {
-		bam_dma_terminate_all(&bdev->channels[i]);
+		bam_dma_terminate_all(&bdev->channels[i].vc.chan);
 		tasklet_kill(&bdev->channels[i].vc.task);
 
 		dma_free_writecombine(bdev->dev, BAM_DESC_FIFO_SIZE,
diff --git a/drivers/dma/s3c24xx-dma.c b/drivers/dma/s3c24xx-dma.c
index 6941a77..2f91da3 100644
--- a/drivers/dma/s3c24xx-dma.c
+++ b/drivers/dma/s3c24xx-dma.c
@@ -384,20 +384,30 @@
 	return tc * txd->width;
 }
 
-static int s3c24xx_dma_set_runtime_config(struct s3c24xx_dma_chan *s3cchan,
+static int s3c24xx_dma_set_runtime_config(struct dma_chan *chan,
 				  struct dma_slave_config *config)
 {
-	if (!s3cchan->slave)
-		return -EINVAL;
+	struct s3c24xx_dma_chan *s3cchan = to_s3c24xx_dma_chan(chan);
+	unsigned long flags;
+	int ret = 0;
 
 	/* Reject definitely invalid configurations */
 	if (config->src_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES ||
 	    config->dst_addr_width == DMA_SLAVE_BUSWIDTH_8_BYTES)
 		return -EINVAL;
 
+	spin_lock_irqsave(&s3cchan->vc.lock, flags);
+
+	if (!s3cchan->slave) {
+		ret = -EINVAL;
+		goto out;
+	}
+
 	s3cchan->cfg = *config;
 
-	return 0;
+out:
+	spin_unlock_irqrestore(&s3cchan->vc.lock, flags);
+	return ret;
 }
 
 /*
@@ -703,8 +713,7 @@
  * The DMA ENGINE API
  */
 
-static int s3c24xx_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-			 unsigned long arg)
+static int s3c24xx_dma_terminate_all(struct dma_chan *chan)
 {
 	struct s3c24xx_dma_chan *s3cchan = to_s3c24xx_dma_chan(chan);
 	struct s3c24xx_dma_engine *s3cdma = s3cchan->host;
@@ -713,40 +722,28 @@
 
 	spin_lock_irqsave(&s3cchan->vc.lock, flags);
 
-	switch (cmd) {
-	case DMA_SLAVE_CONFIG:
-		ret = s3c24xx_dma_set_runtime_config(s3cchan,
-					      (struct dma_slave_config *)arg);
-		break;
-	case DMA_TERMINATE_ALL:
-		if (!s3cchan->phy && !s3cchan->at) {
-			dev_err(&s3cdma->pdev->dev, "trying to terminate already stopped channel %d\n",
-				s3cchan->id);
-			ret = -EINVAL;
-			break;
-		}
-
-		s3cchan->state = S3C24XX_DMA_CHAN_IDLE;
-
-		 /* Mark physical channel as free */
-		if (s3cchan->phy)
-			s3c24xx_dma_phy_free(s3cchan);
-
-		/* Dequeue current job */
-		if (s3cchan->at) {
-			s3c24xx_dma_desc_free(&s3cchan->at->vd);
-			s3cchan->at = NULL;
-		}
-
-		/* Dequeue jobs not yet fired as well */
-		s3c24xx_dma_free_txd_list(s3cdma, s3cchan);
-		break;
-	default:
-		/* Unknown command */
-		ret = -ENXIO;
-		break;
+	if (!s3cchan->phy && !s3cchan->at) {
+		dev_err(&s3cdma->pdev->dev, "trying to terminate already stopped channel %d\n",
+			s3cchan->id);
+		ret = -EINVAL;
+		goto unlock;
 	}
 
+	s3cchan->state = S3C24XX_DMA_CHAN_IDLE;
+
+	/* Mark physical channel as free */
+	if (s3cchan->phy)
+		s3c24xx_dma_phy_free(s3cchan);
+
+	/* Dequeue current job */
+	if (s3cchan->at) {
+		s3c24xx_dma_desc_free(&s3cchan->at->vd);
+		s3cchan->at = NULL;
+	}
+
+	/* Dequeue jobs not yet fired as well */
+	s3c24xx_dma_free_txd_list(s3cdma, s3cchan);
+unlock:
 	spin_unlock_irqrestore(&s3cchan->vc.lock, flags);
 
 	return ret;
@@ -1300,7 +1297,8 @@
 	s3cdma->memcpy.device_prep_dma_memcpy = s3c24xx_dma_prep_memcpy;
 	s3cdma->memcpy.device_tx_status = s3c24xx_dma_tx_status;
 	s3cdma->memcpy.device_issue_pending = s3c24xx_dma_issue_pending;
-	s3cdma->memcpy.device_control = s3c24xx_dma_control;
+	s3cdma->memcpy.device_config = s3c24xx_dma_set_runtime_config;
+	s3cdma->memcpy.device_terminate_all = s3c24xx_dma_terminate_all;
 
 	/* Initialize slave engine for SoC internal dedicated peripherals */
 	dma_cap_set(DMA_SLAVE, s3cdma->slave.cap_mask);
@@ -1315,7 +1313,8 @@
 	s3cdma->slave.device_issue_pending = s3c24xx_dma_issue_pending;
 	s3cdma->slave.device_prep_slave_sg = s3c24xx_dma_prep_slave_sg;
 	s3cdma->slave.device_prep_dma_cyclic = s3c24xx_dma_prep_dma_cyclic;
-	s3cdma->slave.device_control = s3c24xx_dma_control;
+	s3cdma->slave.device_config = s3c24xx_dma_set_runtime_config;
+	s3cdma->slave.device_terminate_all = s3c24xx_dma_terminate_all;
 
 	/* Register as many memcpy channels as there are physical channels */
 	ret = s3c24xx_dma_init_virtual_channels(s3cdma, &s3cdma->memcpy,
diff --git a/drivers/dma/sa11x0-dma.c b/drivers/dma/sa11x0-dma.c
index 96bb62c..5adf540 100644
--- a/drivers/dma/sa11x0-dma.c
+++ b/drivers/dma/sa11x0-dma.c
@@ -669,8 +669,10 @@
 	return vchan_tx_prep(&c->vc, &txd->vd, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 }
 
-static int sa11x0_dma_slave_config(struct sa11x0_dma_chan *c, struct dma_slave_config *cfg)
+static int sa11x0_dma_device_config(struct dma_chan *chan,
+				    struct dma_slave_config *cfg)
 {
+	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
 	u32 ddar = c->ddar & ((0xf << 4) | DDAR_RW);
 	dma_addr_t addr;
 	enum dma_slave_buswidth width;
@@ -704,99 +706,101 @@
 	return 0;
 }
 
-static int sa11x0_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-	unsigned long arg)
+static int sa11x0_dma_device_pause(struct dma_chan *chan)
 {
 	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
 	struct sa11x0_dma_dev *d = to_sa11x0_dma(chan->device);
 	struct sa11x0_dma_phy *p;
 	LIST_HEAD(head);
 	unsigned long flags;
-	int ret;
 
-	switch (cmd) {
-	case DMA_SLAVE_CONFIG:
-		return sa11x0_dma_slave_config(c, (struct dma_slave_config *)arg);
-
-	case DMA_TERMINATE_ALL:
-		dev_dbg(d->slave.dev, "vchan %p: terminate all\n", &c->vc);
-		/* Clear the tx descriptor lists */
-		spin_lock_irqsave(&c->vc.lock, flags);
-		vchan_get_all_descriptors(&c->vc, &head);
+	dev_dbg(d->slave.dev, "vchan %p: pause\n", &c->vc);
+	spin_lock_irqsave(&c->vc.lock, flags);
+	if (c->status == DMA_IN_PROGRESS) {
+		c->status = DMA_PAUSED;
 
 		p = c->phy;
 		if (p) {
-			dev_dbg(d->slave.dev, "pchan %u: terminating\n", p->num);
-			/* vchan is assigned to a pchan - stop the channel */
-			writel(DCSR_RUN | DCSR_IE |
-				DCSR_STRTA | DCSR_DONEA |
-				DCSR_STRTB | DCSR_DONEB,
-				p->base + DMA_DCSR_C);
-
-			if (p->txd_load) {
-				if (p->txd_load != p->txd_done)
-					list_add_tail(&p->txd_load->vd.node, &head);
-				p->txd_load = NULL;
-			}
-			if (p->txd_done) {
-				list_add_tail(&p->txd_done->vd.node, &head);
-				p->txd_done = NULL;
-			}
-			c->phy = NULL;
+			writel(DCSR_RUN | DCSR_IE, p->base + DMA_DCSR_C);
+		} else {
 			spin_lock(&d->lock);
-			p->vchan = NULL;
+			list_del_init(&c->node);
 			spin_unlock(&d->lock);
-			tasklet_schedule(&d->task);
 		}
-		spin_unlock_irqrestore(&c->vc.lock, flags);
-		vchan_dma_desc_free_list(&c->vc, &head);
-		ret = 0;
-		break;
-
-	case DMA_PAUSE:
-		dev_dbg(d->slave.dev, "vchan %p: pause\n", &c->vc);
-		spin_lock_irqsave(&c->vc.lock, flags);
-		if (c->status == DMA_IN_PROGRESS) {
-			c->status = DMA_PAUSED;
-
-			p = c->phy;
-			if (p) {
-				writel(DCSR_RUN | DCSR_IE, p->base + DMA_DCSR_C);
-			} else {
-				spin_lock(&d->lock);
-				list_del_init(&c->node);
-				spin_unlock(&d->lock);
-			}
-		}
-		spin_unlock_irqrestore(&c->vc.lock, flags);
-		ret = 0;
-		break;
-
-	case DMA_RESUME:
-		dev_dbg(d->slave.dev, "vchan %p: resume\n", &c->vc);
-		spin_lock_irqsave(&c->vc.lock, flags);
-		if (c->status == DMA_PAUSED) {
-			c->status = DMA_IN_PROGRESS;
-
-			p = c->phy;
-			if (p) {
-				writel(DCSR_RUN | DCSR_IE, p->base + DMA_DCSR_S);
-			} else if (!list_empty(&c->vc.desc_issued)) {
-				spin_lock(&d->lock);
-				list_add_tail(&c->node, &d->chan_pending);
-				spin_unlock(&d->lock);
-			}
-		}
-		spin_unlock_irqrestore(&c->vc.lock, flags);
-		ret = 0;
-		break;
-
-	default:
-		ret = -ENXIO;
-		break;
 	}
+	spin_unlock_irqrestore(&c->vc.lock, flags);
 
-	return ret;
+	return 0;
+}
+
+static int sa11x0_dma_device_resume(struct dma_chan *chan)
+{
+	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
+	struct sa11x0_dma_dev *d = to_sa11x0_dma(chan->device);
+	struct sa11x0_dma_phy *p;
+	LIST_HEAD(head);
+	unsigned long flags;
+
+	dev_dbg(d->slave.dev, "vchan %p: resume\n", &c->vc);
+	spin_lock_irqsave(&c->vc.lock, flags);
+	if (c->status == DMA_PAUSED) {
+		c->status = DMA_IN_PROGRESS;
+
+		p = c->phy;
+		if (p) {
+			writel(DCSR_RUN | DCSR_IE, p->base + DMA_DCSR_S);
+		} else if (!list_empty(&c->vc.desc_issued)) {
+			spin_lock(&d->lock);
+			list_add_tail(&c->node, &d->chan_pending);
+			spin_unlock(&d->lock);
+		}
+	}
+	spin_unlock_irqrestore(&c->vc.lock, flags);
+
+	return 0;
+}
+
+static int sa11x0_dma_device_terminate_all(struct dma_chan *chan)
+{
+	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
+	struct sa11x0_dma_dev *d = to_sa11x0_dma(chan->device);
+	struct sa11x0_dma_phy *p;
+	LIST_HEAD(head);
+	unsigned long flags;
+
+	dev_dbg(d->slave.dev, "vchan %p: terminate all\n", &c->vc);
+	/* Clear the tx descriptor lists */
+	spin_lock_irqsave(&c->vc.lock, flags);
+	vchan_get_all_descriptors(&c->vc, &head);
+
+	p = c->phy;
+	if (p) {
+		dev_dbg(d->slave.dev, "pchan %u: terminating\n", p->num);
+		/* vchan is assigned to a pchan - stop the channel */
+		writel(DCSR_RUN | DCSR_IE |
+		       DCSR_STRTA | DCSR_DONEA |
+		       DCSR_STRTB | DCSR_DONEB,
+		       p->base + DMA_DCSR_C);
+
+		if (p->txd_load) {
+			if (p->txd_load != p->txd_done)
+				list_add_tail(&p->txd_load->vd.node, &head);
+			p->txd_load = NULL;
+		}
+		if (p->txd_done) {
+			list_add_tail(&p->txd_done->vd.node, &head);
+			p->txd_done = NULL;
+		}
+		c->phy = NULL;
+		spin_lock(&d->lock);
+		p->vchan = NULL;
+		spin_unlock(&d->lock);
+		tasklet_schedule(&d->task);
+	}
+	spin_unlock_irqrestore(&c->vc.lock, flags);
+	vchan_dma_desc_free_list(&c->vc, &head);
+
+	return 0;
 }
 
 struct sa11x0_dma_channel_desc {
@@ -833,7 +837,10 @@
 	dmadev->dev = dev;
 	dmadev->device_alloc_chan_resources = sa11x0_dma_alloc_chan_resources;
 	dmadev->device_free_chan_resources = sa11x0_dma_free_chan_resources;
-	dmadev->device_control = sa11x0_dma_control;
+	dmadev->device_config = sa11x0_dma_device_config;
+	dmadev->device_pause = sa11x0_dma_device_pause;
+	dmadev->device_resume = sa11x0_dma_device_resume;
+	dmadev->device_terminate_all = sa11x0_dma_device_terminate_all;
 	dmadev->device_tx_status = sa11x0_dma_tx_status;
 	dmadev->device_issue_pending = sa11x0_dma_issue_pending;
 
diff --git a/drivers/dma/sh/Kconfig b/drivers/dma/sh/Kconfig
index 0349125..8190ad2 100644
--- a/drivers/dma/sh/Kconfig
+++ b/drivers/dma/sh/Kconfig
@@ -2,6 +2,10 @@
 # DMA engine configuration for sh
 #
 
+config RENESAS_DMA
+	bool
+	select DMA_ENGINE
+
 #
 # DMA Engine Helpers
 #
@@ -12,7 +16,7 @@
 	depends on !SUPERH || SH_DMA
 	depends on !SH_DMA_API
 	default y
-	select DMA_ENGINE
+	select RENESAS_DMA
 	help
 	  Enable support for the Renesas SuperH DMA controllers.
 
@@ -52,3 +56,11 @@
 	depends on SH_DMAE_BASE
 	help
 	  Enable support for the Renesas R-Car Audio DMAC Peripheral Peripheral controllers.
+
+config RCAR_DMAC
+	tristate "Renesas R-Car Gen2 DMA Controller"
+	depends on ARCH_SHMOBILE || COMPILE_TEST
+	select RENESAS_DMA
+	help
+	  This driver supports the general purpose DMA controller found in the
+	  Renesas R-Car second generation SoCs.
diff --git a/drivers/dma/sh/Makefile b/drivers/dma/sh/Makefile
index 0a5cfdb..2852f9d 100644
--- a/drivers/dma/sh/Makefile
+++ b/drivers/dma/sh/Makefile
@@ -16,3 +16,4 @@
 obj-$(CONFIG_SUDMAC) += sudmac.o
 obj-$(CONFIG_RCAR_HPB_DMAE) += rcar-hpbdma.o
 obj-$(CONFIG_RCAR_AUDMAC_PP) += rcar-audmapp.o
+obj-$(CONFIG_RCAR_DMAC) += rcar-dmac.o
diff --git a/drivers/dma/sh/rcar-dmac.c b/drivers/dma/sh/rcar-dmac.c
new file mode 100644
index 0000000..a18d16c
--- /dev/null
+++ b/drivers/dma/sh/rcar-dmac.c
@@ -0,0 +1,1770 @@
+/*
+ * Renesas R-Car Gen2 DMA Controller Driver
+ *
+ * Copyright (C) 2014 Renesas Electronics Inc.
+ *
+ * Author: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
+ *
+ * This is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/interrupt.h>
+#include <linux/list.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/of.h>
+#include <linux/of_dma.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+
+#include "../dmaengine.h"
+
+/*
+ * struct rcar_dmac_xfer_chunk - Descriptor for a hardware transfer
+ * @node: entry in the parent's chunks list
+ * @src_addr: device source address
+ * @dst_addr: device destination address
+ * @size: transfer size in bytes
+ */
+struct rcar_dmac_xfer_chunk {
+	struct list_head node;
+
+	dma_addr_t src_addr;
+	dma_addr_t dst_addr;
+	u32 size;
+};
+
+/*
+ * struct rcar_dmac_hw_desc - Hardware descriptor for a transfer chunk
+ * @sar: value of the SAR register (source address)
+ * @dar: value of the DAR register (destination address)
+ * @tcr: value of the TCR register (transfer count)
+ */
+struct rcar_dmac_hw_desc {
+	u32 sar;
+	u32 dar;
+	u32 tcr;
+	u32 reserved;
+} __attribute__((__packed__));
+
+/*
+ * struct rcar_dmac_desc - R-Car Gen2 DMA Transfer Descriptor
+ * @async_tx: base DMA asynchronous transaction descriptor
+ * @direction: direction of the DMA transfer
+ * @xfer_shift: log2 of the transfer size
+ * @chcr: value of the channel configuration register for this transfer
+ * @node: entry in the channel's descriptors lists
+ * @chunks: list of transfer chunks for this transfer
+ * @running: the transfer chunk being currently processed
+ * @nchunks: number of transfer chunks for this transfer
+ * @hwdescs.use: whether the transfer descriptor uses hardware descriptors
+ * @hwdescs.mem: hardware descriptors memory for the transfer
+ * @hwdescs.dma: device address of the hardware descriptors memory
+ * @hwdescs.size: size of the hardware descriptors in bytes
+ * @size: transfer size in bytes
+ * @cyclic: when set indicates that the DMA transfer is cyclic
+ */
+struct rcar_dmac_desc {
+	struct dma_async_tx_descriptor async_tx;
+	enum dma_transfer_direction direction;
+	unsigned int xfer_shift;
+	u32 chcr;
+
+	struct list_head node;
+	struct list_head chunks;
+	struct rcar_dmac_xfer_chunk *running;
+	unsigned int nchunks;
+
+	struct {
+		bool use;
+		struct rcar_dmac_hw_desc *mem;
+		dma_addr_t dma;
+		size_t size;
+	} hwdescs;
+
+	unsigned int size;
+	bool cyclic;
+};
+
+#define to_rcar_dmac_desc(d)	container_of(d, struct rcar_dmac_desc, async_tx)
+
+/*
+ * struct rcar_dmac_desc_page - One page worth of descriptors
+ * @node: entry in the channel's pages list
+ * @descs: array of DMA descriptors
+ * @chunks: array of transfer chunk descriptors
+ */
+struct rcar_dmac_desc_page {
+	struct list_head node;
+
+	union {
+		struct rcar_dmac_desc descs[0];
+		struct rcar_dmac_xfer_chunk chunks[0];
+	};
+};
+
+#define RCAR_DMAC_DESCS_PER_PAGE					\
+	((PAGE_SIZE - offsetof(struct rcar_dmac_desc_page, descs)) /	\
+	sizeof(struct rcar_dmac_desc))
+#define RCAR_DMAC_XFER_CHUNKS_PER_PAGE					\
+	((PAGE_SIZE - offsetof(struct rcar_dmac_desc_page, chunks)) /	\
+	sizeof(struct rcar_dmac_xfer_chunk))
+
+/*
+ * struct rcar_dmac_chan - R-Car Gen2 DMA Controller Channel
+ * @chan: base DMA channel object
+ * @iomem: channel I/O memory base
+ * @index: index of this channel in the controller
+ * @src_xfer_size: size (in bytes) of hardware transfers on the source side
+ * @dst_xfer_size: size (in bytes) of hardware transfers on the destination side
+ * @src_slave_addr: slave source memory address
+ * @dst_slave_addr: slave destination memory address
+ * @mid_rid: hardware MID/RID for the DMA client using this channel
+ * @lock: protects the channel CHCR register and the desc members
+ * @desc.free: list of free descriptors
+ * @desc.pending: list of pending descriptors (submitted with tx_submit)
+ * @desc.active: list of active descriptors (activated with issue_pending)
+ * @desc.done: list of completed descriptors
+ * @desc.wait: list of descriptors waiting for an ack
+ * @desc.running: the descriptor being processed (a member of the active list)
+ * @desc.chunks_free: list of free transfer chunk descriptors
+ * @desc.pages: list of pages used by allocated descriptors
+ */
+struct rcar_dmac_chan {
+	struct dma_chan chan;
+	void __iomem *iomem;
+	unsigned int index;
+
+	unsigned int src_xfer_size;
+	unsigned int dst_xfer_size;
+	dma_addr_t src_slave_addr;
+	dma_addr_t dst_slave_addr;
+	int mid_rid;
+
+	spinlock_t lock;
+
+	struct {
+		struct list_head free;
+		struct list_head pending;
+		struct list_head active;
+		struct list_head done;
+		struct list_head wait;
+		struct rcar_dmac_desc *running;
+
+		struct list_head chunks_free;
+
+		struct list_head pages;
+	} desc;
+};
+
+#define to_rcar_dmac_chan(c)	container_of(c, struct rcar_dmac_chan, chan)
+
+/*
+ * struct rcar_dmac - R-Car Gen2 DMA Controller
+ * @engine: base DMA engine object
+ * @dev: the hardware device
+ * @iomem: remapped I/O memory base
+ * @n_channels: number of available channels
+ * @channels: array of DMAC channels
+ * @modules: bitmask of client modules in use
+ */
+struct rcar_dmac {
+	struct dma_device engine;
+	struct device *dev;
+	void __iomem *iomem;
+
+	unsigned int n_channels;
+	struct rcar_dmac_chan *channels;
+
+	unsigned long modules[256 / BITS_PER_LONG];
+};
+
+#define to_rcar_dmac(d)		container_of(d, struct rcar_dmac, engine)
+
+/* -----------------------------------------------------------------------------
+ * Registers
+ */
+
+#define RCAR_DMAC_CHAN_OFFSET(i)	(0x8000 + 0x80 * (i))
+
+#define RCAR_DMAISTA			0x0020
+#define RCAR_DMASEC			0x0030
+#define RCAR_DMAOR			0x0060
+#define RCAR_DMAOR_PRI_FIXED		(0 << 8)
+#define RCAR_DMAOR_PRI_ROUND_ROBIN	(3 << 8)
+#define RCAR_DMAOR_AE			(1 << 2)
+#define RCAR_DMAOR_DME			(1 << 0)
+#define RCAR_DMACHCLR			0x0080
+#define RCAR_DMADPSEC			0x00a0
+
+#define RCAR_DMASAR			0x0000
+#define RCAR_DMADAR			0x0004
+#define RCAR_DMATCR			0x0008
+#define RCAR_DMATCR_MASK		0x00ffffff
+#define RCAR_DMATSR			0x0028
+#define RCAR_DMACHCR			0x000c
+#define RCAR_DMACHCR_CAE		(1 << 31)
+#define RCAR_DMACHCR_CAIE		(1 << 30)
+#define RCAR_DMACHCR_DPM_DISABLED	(0 << 28)
+#define RCAR_DMACHCR_DPM_ENABLED	(1 << 28)
+#define RCAR_DMACHCR_DPM_REPEAT		(2 << 28)
+#define RCAR_DMACHCR_DPM_INFINITE	(3 << 28)
+#define RCAR_DMACHCR_RPT_SAR		(1 << 27)
+#define RCAR_DMACHCR_RPT_DAR		(1 << 26)
+#define RCAR_DMACHCR_RPT_TCR		(1 << 25)
+#define RCAR_DMACHCR_DPB		(1 << 22)
+#define RCAR_DMACHCR_DSE		(1 << 19)
+#define RCAR_DMACHCR_DSIE		(1 << 18)
+#define RCAR_DMACHCR_TS_1B		((0 << 20) | (0 << 3))
+#define RCAR_DMACHCR_TS_2B		((0 << 20) | (1 << 3))
+#define RCAR_DMACHCR_TS_4B		((0 << 20) | (2 << 3))
+#define RCAR_DMACHCR_TS_16B		((0 << 20) | (3 << 3))
+#define RCAR_DMACHCR_TS_32B		((1 << 20) | (0 << 3))
+#define RCAR_DMACHCR_TS_64B		((1 << 20) | (1 << 3))
+#define RCAR_DMACHCR_TS_8B		((1 << 20) | (3 << 3))
+#define RCAR_DMACHCR_DM_FIXED		(0 << 14)
+#define RCAR_DMACHCR_DM_INC		(1 << 14)
+#define RCAR_DMACHCR_DM_DEC		(2 << 14)
+#define RCAR_DMACHCR_SM_FIXED		(0 << 12)
+#define RCAR_DMACHCR_SM_INC		(1 << 12)
+#define RCAR_DMACHCR_SM_DEC		(2 << 12)
+#define RCAR_DMACHCR_RS_AUTO		(4 << 8)
+#define RCAR_DMACHCR_RS_DMARS		(8 << 8)
+#define RCAR_DMACHCR_IE			(1 << 2)
+#define RCAR_DMACHCR_TE			(1 << 1)
+#define RCAR_DMACHCR_DE			(1 << 0)
+#define RCAR_DMATCRB			0x0018
+#define RCAR_DMATSRB			0x0038
+#define RCAR_DMACHCRB			0x001c
+#define RCAR_DMACHCRB_DCNT(n)		((n) << 24)
+#define RCAR_DMACHCRB_DPTR_MASK		(0xff << 16)
+#define RCAR_DMACHCRB_DPTR_SHIFT	16
+#define RCAR_DMACHCRB_DRST		(1 << 15)
+#define RCAR_DMACHCRB_DTS		(1 << 8)
+#define RCAR_DMACHCRB_SLM_NORMAL	(0 << 4)
+#define RCAR_DMACHCRB_SLM_CLK(n)	((8 | (n)) << 4)
+#define RCAR_DMACHCRB_PRI(n)		((n) << 0)
+#define RCAR_DMARS			0x0040
+#define RCAR_DMABUFCR			0x0048
+#define RCAR_DMABUFCR_MBU(n)		((n) << 16)
+#define RCAR_DMABUFCR_ULB(n)		((n) << 0)
+#define RCAR_DMADPBASE			0x0050
+#define RCAR_DMADPBASE_MASK		0xfffffff0
+#define RCAR_DMADPBASE_SEL		(1 << 0)
+#define RCAR_DMADPCR			0x0054
+#define RCAR_DMADPCR_DIPT(n)		((n) << 24)
+#define RCAR_DMAFIXSAR			0x0010
+#define RCAR_DMAFIXDAR			0x0014
+#define RCAR_DMAFIXDPBASE		0x0060
+
+/* Hardcode the MEMCPY transfer size to 4 bytes. */
+#define RCAR_DMAC_MEMCPY_XFER_SIZE	4
+
+/* -----------------------------------------------------------------------------
+ * Device access
+ */
+
+static void rcar_dmac_write(struct rcar_dmac *dmac, u32 reg, u32 data)
+{
+	if (reg == RCAR_DMAOR)
+		writew(data, dmac->iomem + reg);
+	else
+		writel(data, dmac->iomem + reg);
+}
+
+static u32 rcar_dmac_read(struct rcar_dmac *dmac, u32 reg)
+{
+	if (reg == RCAR_DMAOR)
+		return readw(dmac->iomem + reg);
+	else
+		return readl(dmac->iomem + reg);
+}
+
+static u32 rcar_dmac_chan_read(struct rcar_dmac_chan *chan, u32 reg)
+{
+	if (reg == RCAR_DMARS)
+		return readw(chan->iomem + reg);
+	else
+		return readl(chan->iomem + reg);
+}
+
+static void rcar_dmac_chan_write(struct rcar_dmac_chan *chan, u32 reg, u32 data)
+{
+	if (reg == RCAR_DMARS)
+		writew(data, chan->iomem + reg);
+	else
+		writel(data, chan->iomem + reg);
+}
+
+/* -----------------------------------------------------------------------------
+ * Initialization and configuration
+ */
+
+static bool rcar_dmac_chan_is_busy(struct rcar_dmac_chan *chan)
+{
+	u32 chcr = rcar_dmac_chan_read(chan, RCAR_DMACHCR);
+
+	return (chcr & (RCAR_DMACHCR_DE | RCAR_DMACHCR_TE)) == RCAR_DMACHCR_DE;
+}
+
+static void rcar_dmac_chan_start_xfer(struct rcar_dmac_chan *chan)
+{
+	struct rcar_dmac_desc *desc = chan->desc.running;
+	u32 chcr = desc->chcr;
+
+	WARN_ON_ONCE(rcar_dmac_chan_is_busy(chan));
+
+	if (chan->mid_rid >= 0)
+		rcar_dmac_chan_write(chan, RCAR_DMARS, chan->mid_rid);
+
+	if (desc->hwdescs.use) {
+		struct rcar_dmac_xfer_chunk *chunk;
+
+		dev_dbg(chan->chan.device->dev,
+			"chan%u: queue desc %p: %u@%pad\n",
+			chan->index, desc, desc->nchunks, &desc->hwdescs.dma);
+
+#ifdef CONFIG_ARCH_DMA_ADDR_T_64BIT
+		rcar_dmac_chan_write(chan, RCAR_DMAFIXDPBASE,
+				     desc->hwdescs.dma >> 32);
+#endif
+		rcar_dmac_chan_write(chan, RCAR_DMADPBASE,
+				     (desc->hwdescs.dma & 0xfffffff0) |
+				     RCAR_DMADPBASE_SEL);
+		rcar_dmac_chan_write(chan, RCAR_DMACHCRB,
+				     RCAR_DMACHCRB_DCNT(desc->nchunks - 1) |
+				     RCAR_DMACHCRB_DRST);
+
+		/*
+		 * Errata: When descriptor memory is accessed through an IOMMU
+		 * the DMADAR register isn't initialized automatically from the
+		 * first descriptor at beginning of transfer by the DMAC like it
+		 * should. Initialize it manually with the destination address
+		 * of the first chunk.
+		 */
+		chunk = list_first_entry(&desc->chunks,
+					 struct rcar_dmac_xfer_chunk, node);
+		rcar_dmac_chan_write(chan, RCAR_DMADAR,
+				     chunk->dst_addr & 0xffffffff);
+
+		/*
+		 * Program the descriptor stage interrupt to occur after the end
+		 * of the first stage.
+		 */
+		rcar_dmac_chan_write(chan, RCAR_DMADPCR, RCAR_DMADPCR_DIPT(1));
+
+		chcr |= RCAR_DMACHCR_RPT_SAR | RCAR_DMACHCR_RPT_DAR
+		     |  RCAR_DMACHCR_RPT_TCR | RCAR_DMACHCR_DPB;
+
+		/*
+		 * If the descriptor isn't cyclic enable normal descriptor mode
+		 * and the transfer completion interrupt.
+		 */
+		if (!desc->cyclic)
+			chcr |= RCAR_DMACHCR_DPM_ENABLED | RCAR_DMACHCR_IE;
+		/*
+		 * If the descriptor is cyclic and has a callback enable the
+		 * descriptor stage interrupt in infinite repeat mode.
+		 */
+		else if (desc->async_tx.callback)
+			chcr |= RCAR_DMACHCR_DPM_INFINITE | RCAR_DMACHCR_DSIE;
+		/*
+		 * Otherwise just select infinite repeat mode without any
+		 * interrupt.
+		 */
+		else
+			chcr |= RCAR_DMACHCR_DPM_INFINITE;
+	} else {
+		struct rcar_dmac_xfer_chunk *chunk = desc->running;
+
+		dev_dbg(chan->chan.device->dev,
+			"chan%u: queue chunk %p: %u@%pad -> %pad\n",
+			chan->index, chunk, chunk->size, &chunk->src_addr,
+			&chunk->dst_addr);
+
+#ifdef CONFIG_ARCH_DMA_ADDR_T_64BIT
+		rcar_dmac_chan_write(chan, RCAR_DMAFIXSAR,
+				     chunk->src_addr >> 32);
+		rcar_dmac_chan_write(chan, RCAR_DMAFIXDAR,
+				     chunk->dst_addr >> 32);
+#endif
+		rcar_dmac_chan_write(chan, RCAR_DMASAR,
+				     chunk->src_addr & 0xffffffff);
+		rcar_dmac_chan_write(chan, RCAR_DMADAR,
+				     chunk->dst_addr & 0xffffffff);
+		rcar_dmac_chan_write(chan, RCAR_DMATCR,
+				     chunk->size >> desc->xfer_shift);
+
+		chcr |= RCAR_DMACHCR_DPM_DISABLED | RCAR_DMACHCR_IE;
+	}
+
+	rcar_dmac_chan_write(chan, RCAR_DMACHCR, chcr | RCAR_DMACHCR_DE);
+}
+
+static int rcar_dmac_init(struct rcar_dmac *dmac)
+{
+	u16 dmaor;
+
+	/* Clear all channels and enable the DMAC globally. */
+	rcar_dmac_write(dmac, RCAR_DMACHCLR, 0x7fff);
+	rcar_dmac_write(dmac, RCAR_DMAOR,
+			RCAR_DMAOR_PRI_FIXED | RCAR_DMAOR_DME);
+
+	dmaor = rcar_dmac_read(dmac, RCAR_DMAOR);
+	if ((dmaor & (RCAR_DMAOR_AE | RCAR_DMAOR_DME)) != RCAR_DMAOR_DME) {
+		dev_warn(dmac->dev, "DMAOR initialization failed.\n");
+		return -EIO;
+	}
+
+	return 0;
+}
+
+/* -----------------------------------------------------------------------------
+ * Descriptors submission
+ */
+
+static dma_cookie_t rcar_dmac_tx_submit(struct dma_async_tx_descriptor *tx)
+{
+	struct rcar_dmac_chan *chan = to_rcar_dmac_chan(tx->chan);
+	struct rcar_dmac_desc *desc = to_rcar_dmac_desc(tx);
+	unsigned long flags;
+	dma_cookie_t cookie;
+
+	spin_lock_irqsave(&chan->lock, flags);
+
+	cookie = dma_cookie_assign(tx);
+
+	dev_dbg(chan->chan.device->dev, "chan%u: submit #%d@%p\n",
+		chan->index, tx->cookie, desc);
+
+	list_add_tail(&desc->node, &chan->desc.pending);
+	desc->running = list_first_entry(&desc->chunks,
+					 struct rcar_dmac_xfer_chunk, node);
+
+	spin_unlock_irqrestore(&chan->lock, flags);
+
+	return cookie;
+}
+
+/* -----------------------------------------------------------------------------
+ * Descriptors allocation and free
+ */
+
+/*
+ * rcar_dmac_desc_alloc - Allocate a page worth of DMA descriptors
+ * @chan: the DMA channel
+ * @gfp: allocation flags
+ */
+static int rcar_dmac_desc_alloc(struct rcar_dmac_chan *chan, gfp_t gfp)
+{
+	struct rcar_dmac_desc_page *page;
+	LIST_HEAD(list);
+	unsigned int i;
+
+	page = (void *)get_zeroed_page(gfp);
+	if (!page)
+		return -ENOMEM;
+
+	for (i = 0; i < RCAR_DMAC_DESCS_PER_PAGE; ++i) {
+		struct rcar_dmac_desc *desc = &page->descs[i];
+
+		dma_async_tx_descriptor_init(&desc->async_tx, &chan->chan);
+		desc->async_tx.tx_submit = rcar_dmac_tx_submit;
+		INIT_LIST_HEAD(&desc->chunks);
+
+		list_add_tail(&desc->node, &list);
+	}
+
+	spin_lock_irq(&chan->lock);
+	list_splice_tail(&list, &chan->desc.free);
+	list_add_tail(&page->node, &chan->desc.pages);
+	spin_unlock_irq(&chan->lock);
+
+	return 0;
+}
+
+/*
+ * rcar_dmac_desc_put - Release a DMA transfer descriptor
+ * @chan: the DMA channel
+ * @desc: the descriptor
+ *
+ * Put the descriptor and its transfer chunk descriptors back in the channel's
+ * free descriptors lists. The descriptor's chunks list will be reinitialized to
+ * an empty list as a result.
+ *
+ * The descriptor must have been removed from the channel's lists before calling
+ * this function.
+ */
+static void rcar_dmac_desc_put(struct rcar_dmac_chan *chan,
+			       struct rcar_dmac_desc *desc)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&chan->lock, flags);
+	list_splice_tail_init(&desc->chunks, &chan->desc.chunks_free);
+	list_add_tail(&desc->node, &chan->desc.free);
+	spin_unlock_irqrestore(&chan->lock, flags);
+}
+
+static void rcar_dmac_desc_recycle_acked(struct rcar_dmac_chan *chan)
+{
+	struct rcar_dmac_desc *desc, *_desc;
+	LIST_HEAD(list);
+
+	/*
+	 * We have to temporarily move all descriptors from the wait list to a
+	 * local list as iterating over the wait list, even with
+	 * list_for_each_entry_safe, isn't safe if we release the channel lock
+	 * around the rcar_dmac_desc_put() call.
+	 */
+	spin_lock_irq(&chan->lock);
+	list_splice_init(&chan->desc.wait, &list);
+	spin_unlock_irq(&chan->lock);
+
+	list_for_each_entry_safe(desc, _desc, &list, node) {
+		if (async_tx_test_ack(&desc->async_tx)) {
+			list_del(&desc->node);
+			rcar_dmac_desc_put(chan, desc);
+		}
+	}
+
+	if (list_empty(&list))
+		return;
+
+	/* Put the remaining descriptors back in the wait list. */
+	spin_lock_irq(&chan->lock);
+	list_splice(&list, &chan->desc.wait);
+	spin_unlock_irq(&chan->lock);
+}
+
+/*
+ * rcar_dmac_desc_get - Allocate a descriptor for a DMA transfer
+ * @chan: the DMA channel
+ *
+ * Locking: This function must be called in a non-atomic context.
+ *
+ * Return: A pointer to the allocated descriptor or NULL if no descriptor can
+ * be allocated.
+ */
+static struct rcar_dmac_desc *rcar_dmac_desc_get(struct rcar_dmac_chan *chan)
+{
+	struct rcar_dmac_desc *desc;
+	int ret;
+
+	/* Recycle acked descriptors before attempting allocation. */
+	rcar_dmac_desc_recycle_acked(chan);
+
+	spin_lock_irq(&chan->lock);
+
+	while (list_empty(&chan->desc.free)) {
+		/*
+		 * No free descriptors, allocate a page worth of them and try
+		 * again, as someone else could race us to get the newly
+		 * allocated descriptors. If the allocation fails return an
+		 * error.
+		 */
+		spin_unlock_irq(&chan->lock);
+		ret = rcar_dmac_desc_alloc(chan, GFP_NOWAIT);
+		if (ret < 0)
+			return NULL;
+		spin_lock_irq(&chan->lock);
+	}
+
+	desc = list_first_entry(&chan->desc.free, struct rcar_dmac_desc, node);
+	list_del(&desc->node);
+
+	spin_unlock_irq(&chan->lock);
+
+	return desc;
+}
+
+/*
+ * rcar_dmac_xfer_chunk_alloc - Allocate a page worth of transfer chunks
+ * @chan: the DMA channel
+ * @gfp: allocation flags
+ */
+static int rcar_dmac_xfer_chunk_alloc(struct rcar_dmac_chan *chan, gfp_t gfp)
+{
+	struct rcar_dmac_desc_page *page;
+	LIST_HEAD(list);
+	unsigned int i;
+
+	page = (void *)get_zeroed_page(gfp);
+	if (!page)
+		return -ENOMEM;
+
+	for (i = 0; i < RCAR_DMAC_XFER_CHUNKS_PER_PAGE; ++i) {
+		struct rcar_dmac_xfer_chunk *chunk = &page->chunks[i];
+
+		list_add_tail(&chunk->node, &list);
+	}
+
+	spin_lock_irq(&chan->lock);
+	list_splice_tail(&list, &chan->desc.chunks_free);
+	list_add_tail(&page->node, &chan->desc.pages);
+	spin_unlock_irq(&chan->lock);
+
+	return 0;
+}
+
+/*
+ * rcar_dmac_xfer_chunk_get - Allocate a transfer chunk for a DMA transfer
+ * @chan: the DMA channel
+ *
+ * Locking: This function must be called in a non-atomic context.
+ *
+ * Return: A pointer to the allocated transfer chunk descriptor or NULL if no
+ * descriptor can be allocated.
+ */
+static struct rcar_dmac_xfer_chunk *
+rcar_dmac_xfer_chunk_get(struct rcar_dmac_chan *chan)
+{
+	struct rcar_dmac_xfer_chunk *chunk;
+	int ret;
+
+	spin_lock_irq(&chan->lock);
+
+	while (list_empty(&chan->desc.chunks_free)) {
+		/*
+		 * No free descriptors, allocate a page worth of them and try
+		 * again, as someone else could race us to get the newly
+		 * allocated descriptors. If the allocation fails return an
+		 * error.
+		 */
+		spin_unlock_irq(&chan->lock);
+		ret = rcar_dmac_xfer_chunk_alloc(chan, GFP_NOWAIT);
+		if (ret < 0)
+			return NULL;
+		spin_lock_irq(&chan->lock);
+	}
+
+	chunk = list_first_entry(&chan->desc.chunks_free,
+				 struct rcar_dmac_xfer_chunk, node);
+	list_del(&chunk->node);
+
+	spin_unlock_irq(&chan->lock);
+
+	return chunk;
+}
+
+static void rcar_dmac_realloc_hwdesc(struct rcar_dmac_chan *chan,
+				     struct rcar_dmac_desc *desc, size_t size)
+{
+	/*
+	 * dma_alloc_coherent() allocates memory in page size increments. To
+	 * avoid reallocating the hardware descriptors when the allocated size
+	 * wouldn't change align the requested size to a multiple of the page
+	 * size.
+	 */
+	size = PAGE_ALIGN(size);
+
+	if (desc->hwdescs.size == size)
+		return;
+
+	if (desc->hwdescs.mem) {
+		dma_free_coherent(chan->chan.device->dev, desc->hwdescs.size,
+				  desc->hwdescs.mem, desc->hwdescs.dma);
+		desc->hwdescs.mem = NULL;
+		desc->hwdescs.size = 0;
+	}
+
+	if (!size)
+		return;
+
+	desc->hwdescs.mem = dma_alloc_coherent(chan->chan.device->dev, size,
+					       &desc->hwdescs.dma, GFP_NOWAIT);
+	if (!desc->hwdescs.mem)
+		return;
+
+	desc->hwdescs.size = size;
+}
+
+static int rcar_dmac_fill_hwdesc(struct rcar_dmac_chan *chan,
+				 struct rcar_dmac_desc *desc)
+{
+	struct rcar_dmac_xfer_chunk *chunk;
+	struct rcar_dmac_hw_desc *hwdesc;
+
+	rcar_dmac_realloc_hwdesc(chan, desc, desc->nchunks * sizeof(*hwdesc));
+
+	hwdesc = desc->hwdescs.mem;
+	if (!hwdesc)
+		return -ENOMEM;
+
+	list_for_each_entry(chunk, &desc->chunks, node) {
+		hwdesc->sar = chunk->src_addr;
+		hwdesc->dar = chunk->dst_addr;
+		hwdesc->tcr = chunk->size >> desc->xfer_shift;
+		hwdesc++;
+	}
+
+	return 0;
+}
+
+/* -----------------------------------------------------------------------------
+ * Stop and reset
+ */
+
+static void rcar_dmac_chan_halt(struct rcar_dmac_chan *chan)
+{
+	u32 chcr = rcar_dmac_chan_read(chan, RCAR_DMACHCR);
+
+	chcr &= ~(RCAR_DMACHCR_DSE | RCAR_DMACHCR_DSIE | RCAR_DMACHCR_IE |
+		  RCAR_DMACHCR_TE | RCAR_DMACHCR_DE);
+	rcar_dmac_chan_write(chan, RCAR_DMACHCR, chcr);
+}
+
+static void rcar_dmac_chan_reinit(struct rcar_dmac_chan *chan)
+{
+	struct rcar_dmac_desc *desc, *_desc;
+	unsigned long flags;
+	LIST_HEAD(descs);
+
+	spin_lock_irqsave(&chan->lock, flags);
+
+	/* Move all non-free descriptors to the local lists. */
+	list_splice_init(&chan->desc.pending, &descs);
+	list_splice_init(&chan->desc.active, &descs);
+	list_splice_init(&chan->desc.done, &descs);
+	list_splice_init(&chan->desc.wait, &descs);
+
+	chan->desc.running = NULL;
+
+	spin_unlock_irqrestore(&chan->lock, flags);
+
+	list_for_each_entry_safe(desc, _desc, &descs, node) {
+		list_del(&desc->node);
+		rcar_dmac_desc_put(chan, desc);
+	}
+}
+
+static void rcar_dmac_stop(struct rcar_dmac *dmac)
+{
+	rcar_dmac_write(dmac, RCAR_DMAOR, 0);
+}
+
+static void rcar_dmac_abort(struct rcar_dmac *dmac)
+{
+	unsigned int i;
+
+	/* Stop all channels. */
+	for (i = 0; i < dmac->n_channels; ++i) {
+		struct rcar_dmac_chan *chan = &dmac->channels[i];
+
+		/* Stop and reinitialize the channel. */
+		spin_lock(&chan->lock);
+		rcar_dmac_chan_halt(chan);
+		spin_unlock(&chan->lock);
+
+		rcar_dmac_chan_reinit(chan);
+	}
+}
+
+/* -----------------------------------------------------------------------------
+ * Descriptors preparation
+ */
+
+static void rcar_dmac_chan_configure_desc(struct rcar_dmac_chan *chan,
+					  struct rcar_dmac_desc *desc)
+{
+	static const u32 chcr_ts[] = {
+		RCAR_DMACHCR_TS_1B, RCAR_DMACHCR_TS_2B,
+		RCAR_DMACHCR_TS_4B, RCAR_DMACHCR_TS_8B,
+		RCAR_DMACHCR_TS_16B, RCAR_DMACHCR_TS_32B,
+		RCAR_DMACHCR_TS_64B,
+	};
+
+	unsigned int xfer_size;
+	u32 chcr;
+
+	switch (desc->direction) {
+	case DMA_DEV_TO_MEM:
+		chcr = RCAR_DMACHCR_DM_INC | RCAR_DMACHCR_SM_FIXED
+		     | RCAR_DMACHCR_RS_DMARS;
+		xfer_size = chan->src_xfer_size;
+		break;
+
+	case DMA_MEM_TO_DEV:
+		chcr = RCAR_DMACHCR_DM_FIXED | RCAR_DMACHCR_SM_INC
+		     | RCAR_DMACHCR_RS_DMARS;
+		xfer_size = chan->dst_xfer_size;
+		break;
+
+	case DMA_MEM_TO_MEM:
+	default:
+		chcr = RCAR_DMACHCR_DM_INC | RCAR_DMACHCR_SM_INC
+		     | RCAR_DMACHCR_RS_AUTO;
+		xfer_size = RCAR_DMAC_MEMCPY_XFER_SIZE;
+		break;
+	}
+
+	desc->xfer_shift = ilog2(xfer_size);
+	desc->chcr = chcr | chcr_ts[desc->xfer_shift];
+}
+
+/*
+ * rcar_dmac_chan_prep_sg - prepare transfer descriptors from an SG list
+ *
+ * Common routine for public (MEMCPY) and slave DMA. The MEMCPY case is also
+ * 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_MEM_TO_MEM
+ * and the SG list contains only one element and points at the source buffer.
+ */
+static struct dma_async_tx_descriptor *
+rcar_dmac_chan_prep_sg(struct rcar_dmac_chan *chan, struct scatterlist *sgl,
+		       unsigned int sg_len, dma_addr_t dev_addr,
+		       enum dma_transfer_direction dir, unsigned long dma_flags,
+		       bool cyclic)
+{
+	struct rcar_dmac_xfer_chunk *chunk;
+	struct rcar_dmac_desc *desc;
+	struct scatterlist *sg;
+	unsigned int nchunks = 0;
+	unsigned int max_chunk_size;
+	unsigned int full_size = 0;
+	bool highmem = false;
+	unsigned int i;
+
+	desc = rcar_dmac_desc_get(chan);
+	if (!desc)
+		return NULL;
+
+	desc->async_tx.flags = dma_flags;
+	desc->async_tx.cookie = -EBUSY;
+
+	desc->cyclic = cyclic;
+	desc->direction = dir;
+
+	rcar_dmac_chan_configure_desc(chan, desc);
+
+	max_chunk_size = (RCAR_DMATCR_MASK + 1) << desc->xfer_shift;
+
+	/*
+	 * Allocate and fill the transfer chunk descriptors. We own the only
+	 * reference to the DMA descriptor, there's no need for locking.
+	 */
+	for_each_sg(sgl, sg, sg_len, i) {
+		dma_addr_t mem_addr = sg_dma_address(sg);
+		unsigned int len = sg_dma_len(sg);
+
+		full_size += len;
+
+		while (len) {
+			unsigned int size = min(len, max_chunk_size);
+
+#ifdef CONFIG_ARCH_DMA_ADDR_T_64BIT
+			/*
+			 * Prevent individual transfers from crossing 4GB
+			 * boundaries.
+			 */
+			if (dev_addr >> 32 != (dev_addr + size - 1) >> 32)
+				size = ALIGN(dev_addr, 1ULL << 32) - dev_addr;
+			if (mem_addr >> 32 != (mem_addr + size - 1) >> 32)
+				size = ALIGN(mem_addr, 1ULL << 32) - mem_addr;
+
+			/*
+			 * Check if either of the source or destination address
+			 * can't be expressed in 32 bits. If so we can't use
+			 * hardware descriptor lists.
+			 */
+			if (dev_addr >> 32 || mem_addr >> 32)
+				highmem = true;
+#endif
+
+			chunk = rcar_dmac_xfer_chunk_get(chan);
+			if (!chunk) {
+				rcar_dmac_desc_put(chan, desc);
+				return NULL;
+			}
+
+			if (dir == DMA_DEV_TO_MEM) {
+				chunk->src_addr = dev_addr;
+				chunk->dst_addr = mem_addr;
+			} else {
+				chunk->src_addr = mem_addr;
+				chunk->dst_addr = dev_addr;
+			}
+
+			chunk->size = size;
+
+			dev_dbg(chan->chan.device->dev,
+				"chan%u: chunk %p/%p sgl %u@%p, %u/%u %pad -> %pad\n",
+				chan->index, chunk, desc, i, sg, size, len,
+				&chunk->src_addr, &chunk->dst_addr);
+
+			mem_addr += size;
+			if (dir == DMA_MEM_TO_MEM)
+				dev_addr += size;
+
+			len -= size;
+
+			list_add_tail(&chunk->node, &desc->chunks);
+			nchunks++;
+		}
+	}
+
+	desc->nchunks = nchunks;
+	desc->size = full_size;
+
+	/*
+	 * Use hardware descriptor lists if possible when more than one chunk
+	 * needs to be transferred (otherwise they don't make much sense).
+	 *
+	 * The highmem check currently covers the whole transfer. As an
+	 * optimization we could use descriptor lists for consecutive lowmem
+	 * chunks and direct manual mode for highmem chunks. Whether the
+	 * performance improvement would be significant enough compared to the
+	 * additional complexity remains to be investigated.
+	 */
+	desc->hwdescs.use = !highmem && nchunks > 1;
+	if (desc->hwdescs.use) {
+		if (rcar_dmac_fill_hwdesc(chan, desc) < 0)
+			desc->hwdescs.use = false;
+	}
+
+	return &desc->async_tx;
+}
+
+/* -----------------------------------------------------------------------------
+ * DMA engine operations
+ */
+
+static int rcar_dmac_alloc_chan_resources(struct dma_chan *chan)
+{
+	struct rcar_dmac_chan *rchan = to_rcar_dmac_chan(chan);
+	int ret;
+
+	INIT_LIST_HEAD(&rchan->desc.chunks_free);
+	INIT_LIST_HEAD(&rchan->desc.pages);
+
+	/* Preallocate descriptors. */
+	ret = rcar_dmac_xfer_chunk_alloc(rchan, GFP_KERNEL);
+	if (ret < 0)
+		return -ENOMEM;
+
+	ret = rcar_dmac_desc_alloc(rchan, GFP_KERNEL);
+	if (ret < 0)
+		return -ENOMEM;
+
+	return pm_runtime_get_sync(chan->device->dev);
+}
+
+static void rcar_dmac_free_chan_resources(struct dma_chan *chan)
+{
+	struct rcar_dmac_chan *rchan = to_rcar_dmac_chan(chan);
+	struct rcar_dmac *dmac = to_rcar_dmac(chan->device);
+	struct rcar_dmac_desc_page *page, *_page;
+	struct rcar_dmac_desc *desc;
+	LIST_HEAD(list);
+
+	/* Protect against ISR */
+	spin_lock_irq(&rchan->lock);
+	rcar_dmac_chan_halt(rchan);
+	spin_unlock_irq(&rchan->lock);
+
+	/* Now no new interrupts will occur */
+
+	if (rchan->mid_rid >= 0) {
+		/* The caller is holding dma_list_mutex */
+		clear_bit(rchan->mid_rid, dmac->modules);
+		rchan->mid_rid = -EINVAL;
+	}
+
+	list_splice_init(&rchan->desc.free, &list);
+	list_splice_init(&rchan->desc.pending, &list);
+	list_splice_init(&rchan->desc.active, &list);
+	list_splice_init(&rchan->desc.done, &list);
+	list_splice_init(&rchan->desc.wait, &list);
+
+	list_for_each_entry(desc, &list, node)
+		rcar_dmac_realloc_hwdesc(rchan, desc, 0);
+
+	list_for_each_entry_safe(page, _page, &rchan->desc.pages, node) {
+		list_del(&page->node);
+		free_page((unsigned long)page);
+	}
+
+	pm_runtime_put(chan->device->dev);
+}
+
+static struct dma_async_tx_descriptor *
+rcar_dmac_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dma_dest,
+			  dma_addr_t dma_src, size_t len, unsigned long flags)
+{
+	struct rcar_dmac_chan *rchan = to_rcar_dmac_chan(chan);
+	struct scatterlist sgl;
+
+	if (!len)
+		return NULL;
+
+	sg_init_table(&sgl, 1);
+	sg_set_page(&sgl, pfn_to_page(PFN_DOWN(dma_src)), len,
+		    offset_in_page(dma_src));
+	sg_dma_address(&sgl) = dma_src;
+	sg_dma_len(&sgl) = len;
+
+	return rcar_dmac_chan_prep_sg(rchan, &sgl, 1, dma_dest,
+				      DMA_MEM_TO_MEM, flags, false);
+}
+
+static struct dma_async_tx_descriptor *
+rcar_dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
+			unsigned int sg_len, enum dma_transfer_direction dir,
+			unsigned long flags, void *context)
+{
+	struct rcar_dmac_chan *rchan = to_rcar_dmac_chan(chan);
+	dma_addr_t dev_addr;
+
+	/* Someone calling slave DMA on a generic channel? */
+	if (rchan->mid_rid < 0 || !sg_len) {
+		dev_warn(chan->device->dev,
+			 "%s: bad parameter: len=%d, id=%d\n",
+			 __func__, sg_len, rchan->mid_rid);
+		return NULL;
+	}
+
+	dev_addr = dir == DMA_DEV_TO_MEM
+		 ? rchan->src_slave_addr : rchan->dst_slave_addr;
+	return rcar_dmac_chan_prep_sg(rchan, sgl, sg_len, dev_addr,
+				      dir, flags, false);
+}
+
+#define RCAR_DMAC_MAX_SG_LEN	32
+
+static struct dma_async_tx_descriptor *
+rcar_dmac_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr,
+			  size_t buf_len, size_t period_len,
+			  enum dma_transfer_direction dir, unsigned long flags)
+{
+	struct rcar_dmac_chan *rchan = to_rcar_dmac_chan(chan);
+	struct dma_async_tx_descriptor *desc;
+	struct scatterlist *sgl;
+	dma_addr_t dev_addr;
+	unsigned int sg_len;
+	unsigned int i;
+
+	/* Someone calling slave DMA on a generic channel? */
+	if (rchan->mid_rid < 0 || buf_len < period_len) {
+		dev_warn(chan->device->dev,
+			"%s: bad parameter: buf_len=%zu, period_len=%zu, id=%d\n",
+			__func__, buf_len, period_len, rchan->mid_rid);
+		return NULL;
+	}
+
+	sg_len = buf_len / period_len;
+	if (sg_len > RCAR_DMAC_MAX_SG_LEN) {
+		dev_err(chan->device->dev,
+			"chan%u: sg length %d exceds limit %d",
+			rchan->index, sg_len, RCAR_DMAC_MAX_SG_LEN);
+		return NULL;
+	}
+
+	/*
+	 * Allocate the sg list dynamically as it would consume too much stack
+	 * space.
+	 */
+	sgl = kcalloc(sg_len, sizeof(*sgl), GFP_NOWAIT);
+	if (!sgl)
+		return NULL;
+
+	sg_init_table(sgl, sg_len);
+
+	for (i = 0; i < sg_len; ++i) {
+		dma_addr_t src = buf_addr + (period_len * i);
+
+		sg_set_page(&sgl[i], pfn_to_page(PFN_DOWN(src)), period_len,
+			    offset_in_page(src));
+		sg_dma_address(&sgl[i]) = src;
+		sg_dma_len(&sgl[i]) = period_len;
+	}
+
+	dev_addr = dir == DMA_DEV_TO_MEM
+		 ? rchan->src_slave_addr : rchan->dst_slave_addr;
+	desc = rcar_dmac_chan_prep_sg(rchan, sgl, sg_len, dev_addr,
+				      dir, flags, true);
+
+	kfree(sgl);
+	return desc;
+}
+
+static int rcar_dmac_device_config(struct dma_chan *chan,
+				   struct dma_slave_config *cfg)
+{
+	struct rcar_dmac_chan *rchan = to_rcar_dmac_chan(chan);
+
+	/*
+	 * We could lock this, but you shouldn't be configuring the
+	 * channel, while using it...
+	 */
+	rchan->src_slave_addr = cfg->src_addr;
+	rchan->dst_slave_addr = cfg->dst_addr;
+	rchan->src_xfer_size = cfg->src_addr_width;
+	rchan->dst_xfer_size = cfg->dst_addr_width;
+
+	return 0;
+}
+
+static int rcar_dmac_chan_terminate_all(struct dma_chan *chan)
+{
+	struct rcar_dmac_chan *rchan = to_rcar_dmac_chan(chan);
+	unsigned long flags;
+
+	spin_lock_irqsave(&rchan->lock, flags);
+	rcar_dmac_chan_halt(rchan);
+	spin_unlock_irqrestore(&rchan->lock, flags);
+
+	/*
+	 * FIXME: No new interrupt can occur now, but the IRQ thread might still
+	 * be running.
+	 */
+
+	rcar_dmac_chan_reinit(rchan);
+
+	return 0;
+}
+
+static unsigned int rcar_dmac_chan_get_residue(struct rcar_dmac_chan *chan,
+					       dma_cookie_t cookie)
+{
+	struct rcar_dmac_desc *desc = chan->desc.running;
+	struct rcar_dmac_xfer_chunk *running = NULL;
+	struct rcar_dmac_xfer_chunk *chunk;
+	unsigned int residue = 0;
+	unsigned int dptr = 0;
+
+	if (!desc)
+		return 0;
+
+	/*
+	 * If the cookie doesn't correspond to the currently running transfer
+	 * then the descriptor hasn't been processed yet, and the residue is
+	 * equal to the full descriptor size.
+	 */
+	if (cookie != desc->async_tx.cookie)
+		return desc->size;
+
+	/*
+	 * In descriptor mode the descriptor running pointer is not maintained
+	 * by the interrupt handler, find the running descriptor from the
+	 * descriptor pointer field in the CHCRB register. In non-descriptor
+	 * mode just use the running descriptor pointer.
+	 */
+	if (desc->hwdescs.use) {
+		dptr = (rcar_dmac_chan_read(chan, RCAR_DMACHCRB) &
+			RCAR_DMACHCRB_DPTR_MASK) >> RCAR_DMACHCRB_DPTR_SHIFT;
+		WARN_ON(dptr >= desc->nchunks);
+	} else {
+		running = desc->running;
+	}
+
+	/* Compute the size of all chunks still to be transferred. */
+	list_for_each_entry_reverse(chunk, &desc->chunks, node) {
+		if (chunk == running || ++dptr == desc->nchunks)
+			break;
+
+		residue += chunk->size;
+	}
+
+	/* Add the residue for the current chunk. */
+	residue += rcar_dmac_chan_read(chan, RCAR_DMATCR) << desc->xfer_shift;
+
+	return residue;
+}
+
+static enum dma_status rcar_dmac_tx_status(struct dma_chan *chan,
+					   dma_cookie_t cookie,
+					   struct dma_tx_state *txstate)
+{
+	struct rcar_dmac_chan *rchan = to_rcar_dmac_chan(chan);
+	enum dma_status status;
+	unsigned long flags;
+	unsigned int residue;
+
+	status = dma_cookie_status(chan, cookie, txstate);
+	if (status == DMA_COMPLETE || !txstate)
+		return status;
+
+	spin_lock_irqsave(&rchan->lock, flags);
+	residue = rcar_dmac_chan_get_residue(rchan, cookie);
+	spin_unlock_irqrestore(&rchan->lock, flags);
+
+	dma_set_residue(txstate, residue);
+
+	return status;
+}
+
+static void rcar_dmac_issue_pending(struct dma_chan *chan)
+{
+	struct rcar_dmac_chan *rchan = to_rcar_dmac_chan(chan);
+	unsigned long flags;
+
+	spin_lock_irqsave(&rchan->lock, flags);
+
+	if (list_empty(&rchan->desc.pending))
+		goto done;
+
+	/* Append the pending list to the active list. */
+	list_splice_tail_init(&rchan->desc.pending, &rchan->desc.active);
+
+	/*
+	 * If no transfer is running pick the first descriptor from the active
+	 * list and start the transfer.
+	 */
+	if (!rchan->desc.running) {
+		struct rcar_dmac_desc *desc;
+
+		desc = list_first_entry(&rchan->desc.active,
+					struct rcar_dmac_desc, node);
+		rchan->desc.running = desc;
+
+		rcar_dmac_chan_start_xfer(rchan);
+	}
+
+done:
+	spin_unlock_irqrestore(&rchan->lock, flags);
+}
+
+/* -----------------------------------------------------------------------------
+ * IRQ handling
+ */
+
+static irqreturn_t rcar_dmac_isr_desc_stage_end(struct rcar_dmac_chan *chan)
+{
+	struct rcar_dmac_desc *desc = chan->desc.running;
+	unsigned int stage;
+
+	if (WARN_ON(!desc || !desc->cyclic)) {
+		/*
+		 * This should never happen, there should always be a running
+		 * cyclic descriptor when a descriptor stage end interrupt is
+		 * triggered. Warn and return.
+		 */
+		return IRQ_NONE;
+	}
+
+	/* Program the interrupt pointer to the next stage. */
+	stage = (rcar_dmac_chan_read(chan, RCAR_DMACHCRB) &
+		 RCAR_DMACHCRB_DPTR_MASK) >> RCAR_DMACHCRB_DPTR_SHIFT;
+	rcar_dmac_chan_write(chan, RCAR_DMADPCR, RCAR_DMADPCR_DIPT(stage));
+
+	return IRQ_WAKE_THREAD;
+}
+
+static irqreturn_t rcar_dmac_isr_transfer_end(struct rcar_dmac_chan *chan)
+{
+	struct rcar_dmac_desc *desc = chan->desc.running;
+	irqreturn_t ret = IRQ_WAKE_THREAD;
+
+	if (WARN_ON_ONCE(!desc)) {
+		/*
+		 * This should never happen, there should always be a running
+		 * descriptor when a transfer end interrupt is triggered. Warn
+		 * and return.
+		 */
+		return IRQ_NONE;
+	}
+
+	/*
+	 * The transfer end interrupt isn't generated for each chunk when using
+	 * descriptor mode. Only update the running chunk pointer in
+	 * non-descriptor mode.
+	 */
+	if (!desc->hwdescs.use) {
+		/*
+		 * If we haven't completed the last transfer chunk simply move
+		 * to the next one. Only wake the IRQ thread if the transfer is
+		 * cyclic.
+		 */
+		if (!list_is_last(&desc->running->node, &desc->chunks)) {
+			desc->running = list_next_entry(desc->running, node);
+			if (!desc->cyclic)
+				ret = IRQ_HANDLED;
+			goto done;
+		}
+
+		/*
+		 * We've completed the last transfer chunk. If the transfer is
+		 * cyclic, move back to the first one.
+		 */
+		if (desc->cyclic) {
+			desc->running =
+				list_first_entry(&desc->chunks,
+						 struct rcar_dmac_xfer_chunk,
+						 node);
+			goto done;
+		}
+	}
+
+	/* The descriptor is complete, move it to the done list. */
+	list_move_tail(&desc->node, &chan->desc.done);
+
+	/* Queue the next descriptor, if any. */
+	if (!list_empty(&chan->desc.active))
+		chan->desc.running = list_first_entry(&chan->desc.active,
+						      struct rcar_dmac_desc,
+						      node);
+	else
+		chan->desc.running = NULL;
+
+done:
+	if (chan->desc.running)
+		rcar_dmac_chan_start_xfer(chan);
+
+	return ret;
+}
+
+static irqreturn_t rcar_dmac_isr_channel(int irq, void *dev)
+{
+	u32 mask = RCAR_DMACHCR_DSE | RCAR_DMACHCR_TE;
+	struct rcar_dmac_chan *chan = dev;
+	irqreturn_t ret = IRQ_NONE;
+	u32 chcr;
+
+	spin_lock(&chan->lock);
+
+	chcr = rcar_dmac_chan_read(chan, RCAR_DMACHCR);
+	if (chcr & RCAR_DMACHCR_TE)
+		mask |= RCAR_DMACHCR_DE;
+	rcar_dmac_chan_write(chan, RCAR_DMACHCR, chcr & ~mask);
+
+	if (chcr & RCAR_DMACHCR_DSE)
+		ret |= rcar_dmac_isr_desc_stage_end(chan);
+
+	if (chcr & RCAR_DMACHCR_TE)
+		ret |= rcar_dmac_isr_transfer_end(chan);
+
+	spin_unlock(&chan->lock);
+
+	return ret;
+}
+
+static irqreturn_t rcar_dmac_isr_channel_thread(int irq, void *dev)
+{
+	struct rcar_dmac_chan *chan = dev;
+	struct rcar_dmac_desc *desc;
+
+	spin_lock_irq(&chan->lock);
+
+	/* For cyclic transfers notify the user after every chunk. */
+	if (chan->desc.running && chan->desc.running->cyclic) {
+		dma_async_tx_callback callback;
+		void *callback_param;
+
+		desc = chan->desc.running;
+		callback = desc->async_tx.callback;
+		callback_param = desc->async_tx.callback_param;
+
+		if (callback) {
+			spin_unlock_irq(&chan->lock);
+			callback(callback_param);
+			spin_lock_irq(&chan->lock);
+		}
+	}
+
+	/*
+	 * Call the callback function for all descriptors on the done list and
+	 * move them to the ack wait list.
+	 */
+	while (!list_empty(&chan->desc.done)) {
+		desc = list_first_entry(&chan->desc.done, struct rcar_dmac_desc,
+					node);
+		dma_cookie_complete(&desc->async_tx);
+		list_del(&desc->node);
+
+		if (desc->async_tx.callback) {
+			spin_unlock_irq(&chan->lock);
+			/*
+			 * We own the only reference to this descriptor, we can
+			 * safely dereference it without holding the channel
+			 * lock.
+			 */
+			desc->async_tx.callback(desc->async_tx.callback_param);
+			spin_lock_irq(&chan->lock);
+		}
+
+		list_add_tail(&desc->node, &chan->desc.wait);
+	}
+
+	spin_unlock_irq(&chan->lock);
+
+	/* Recycle all acked descriptors. */
+	rcar_dmac_desc_recycle_acked(chan);
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t rcar_dmac_isr_error(int irq, void *data)
+{
+	struct rcar_dmac *dmac = data;
+
+	if (!(rcar_dmac_read(dmac, RCAR_DMAOR) & RCAR_DMAOR_AE))
+		return IRQ_NONE;
+
+	/*
+	 * An unrecoverable error occurred on an unknown channel. Halt the DMAC,
+	 * abort transfers on all channels, and reinitialize the DMAC.
+	 */
+	rcar_dmac_stop(dmac);
+	rcar_dmac_abort(dmac);
+	rcar_dmac_init(dmac);
+
+	return IRQ_HANDLED;
+}
+
+/* -----------------------------------------------------------------------------
+ * OF xlate and channel filter
+ */
+
+static bool rcar_dmac_chan_filter(struct dma_chan *chan, void *arg)
+{
+	struct rcar_dmac *dmac = to_rcar_dmac(chan->device);
+	struct of_phandle_args *dma_spec = arg;
+
+	/*
+	 * FIXME: Using a filter on OF platforms is a nonsense. The OF xlate
+	 * function knows from which device it wants to allocate a channel from,
+	 * and would be perfectly capable of selecting the channel it wants.
+	 * Forcing it to call dma_request_channel() and iterate through all
+	 * channels from all controllers is just pointless.
+	 */
+	if (chan->device->device_config != rcar_dmac_device_config ||
+	    dma_spec->np != chan->device->dev->of_node)
+		return false;
+
+	return !test_and_set_bit(dma_spec->args[0], dmac->modules);
+}
+
+static struct dma_chan *rcar_dmac_of_xlate(struct of_phandle_args *dma_spec,
+					   struct of_dma *ofdma)
+{
+	struct rcar_dmac_chan *rchan;
+	struct dma_chan *chan;
+	dma_cap_mask_t mask;
+
+	if (dma_spec->args_count != 1)
+		return NULL;
+
+	/* Only slave DMA channels can be allocated via DT */
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	chan = dma_request_channel(mask, rcar_dmac_chan_filter, dma_spec);
+	if (!chan)
+		return NULL;
+
+	rchan = to_rcar_dmac_chan(chan);
+	rchan->mid_rid = dma_spec->args[0];
+
+	return chan;
+}
+
+/* -----------------------------------------------------------------------------
+ * Power management
+ */
+
+#ifdef CONFIG_PM_SLEEP
+static int rcar_dmac_sleep_suspend(struct device *dev)
+{
+	/*
+	 * TODO: Wait for the current transfer to complete and stop the device.
+	 */
+	return 0;
+}
+
+static int rcar_dmac_sleep_resume(struct device *dev)
+{
+	/* TODO: Resume transfers, if any. */
+	return 0;
+}
+#endif
+
+#ifdef CONFIG_PM
+static int rcar_dmac_runtime_suspend(struct device *dev)
+{
+	return 0;
+}
+
+static int rcar_dmac_runtime_resume(struct device *dev)
+{
+	struct rcar_dmac *dmac = dev_get_drvdata(dev);
+
+	return rcar_dmac_init(dmac);
+}
+#endif
+
+static const struct dev_pm_ops rcar_dmac_pm = {
+	SET_SYSTEM_SLEEP_PM_OPS(rcar_dmac_sleep_suspend, rcar_dmac_sleep_resume)
+	SET_RUNTIME_PM_OPS(rcar_dmac_runtime_suspend, rcar_dmac_runtime_resume,
+			   NULL)
+};
+
+/* -----------------------------------------------------------------------------
+ * Probe and remove
+ */
+
+static int rcar_dmac_chan_probe(struct rcar_dmac *dmac,
+				struct rcar_dmac_chan *rchan,
+				unsigned int index)
+{
+	struct platform_device *pdev = to_platform_device(dmac->dev);
+	struct dma_chan *chan = &rchan->chan;
+	char pdev_irqname[5];
+	char *irqname;
+	int irq;
+	int ret;
+
+	rchan->index = index;
+	rchan->iomem = dmac->iomem + RCAR_DMAC_CHAN_OFFSET(index);
+	rchan->mid_rid = -EINVAL;
+
+	spin_lock_init(&rchan->lock);
+
+	INIT_LIST_HEAD(&rchan->desc.free);
+	INIT_LIST_HEAD(&rchan->desc.pending);
+	INIT_LIST_HEAD(&rchan->desc.active);
+	INIT_LIST_HEAD(&rchan->desc.done);
+	INIT_LIST_HEAD(&rchan->desc.wait);
+
+	/* Request the channel interrupt. */
+	sprintf(pdev_irqname, "ch%u", index);
+	irq = platform_get_irq_byname(pdev, pdev_irqname);
+	if (irq < 0) {
+		dev_err(dmac->dev, "no IRQ specified for channel %u\n", index);
+		return -ENODEV;
+	}
+
+	irqname = devm_kasprintf(dmac->dev, GFP_KERNEL, "%s:%u",
+				 dev_name(dmac->dev), index);
+	if (!irqname)
+		return -ENOMEM;
+
+	ret = devm_request_threaded_irq(dmac->dev, irq, rcar_dmac_isr_channel,
+					rcar_dmac_isr_channel_thread, 0,
+					irqname, rchan);
+	if (ret) {
+		dev_err(dmac->dev, "failed to request IRQ %u (%d)\n", irq, ret);
+		return ret;
+	}
+
+	/*
+	 * Initialize the DMA engine channel and add it to the DMA engine
+	 * channels list.
+	 */
+	chan->device = &dmac->engine;
+	dma_cookie_init(chan);
+
+	list_add_tail(&chan->device_node, &dmac->engine.channels);
+
+	return 0;
+}
+
+static int rcar_dmac_parse_of(struct device *dev, struct rcar_dmac *dmac)
+{
+	struct device_node *np = dev->of_node;
+	int ret;
+
+	ret = of_property_read_u32(np, "dma-channels", &dmac->n_channels);
+	if (ret < 0) {
+		dev_err(dev, "unable to read dma-channels property\n");
+		return ret;
+	}
+
+	if (dmac->n_channels <= 0 || dmac->n_channels >= 100) {
+		dev_err(dev, "invalid number of channels %u\n",
+			dmac->n_channels);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int rcar_dmac_probe(struct platform_device *pdev)
+{
+	const enum dma_slave_buswidth widths = DMA_SLAVE_BUSWIDTH_1_BYTE |
+		DMA_SLAVE_BUSWIDTH_2_BYTES | DMA_SLAVE_BUSWIDTH_4_BYTES |
+		DMA_SLAVE_BUSWIDTH_8_BYTES | DMA_SLAVE_BUSWIDTH_16_BYTES |
+		DMA_SLAVE_BUSWIDTH_32_BYTES | DMA_SLAVE_BUSWIDTH_64_BYTES;
+	unsigned int channels_offset = 0;
+	struct dma_device *engine;
+	struct rcar_dmac *dmac;
+	struct resource *mem;
+	unsigned int i;
+	char *irqname;
+	int irq;
+	int ret;
+
+	dmac = devm_kzalloc(&pdev->dev, sizeof(*dmac), GFP_KERNEL);
+	if (!dmac)
+		return -ENOMEM;
+
+	dmac->dev = &pdev->dev;
+	platform_set_drvdata(pdev, dmac);
+
+	ret = rcar_dmac_parse_of(&pdev->dev, dmac);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * A still unconfirmed hardware bug prevents the IPMMU microTLB 0 to be
+	 * flushed correctly, resulting in memory corruption. DMAC 0 channel 0
+	 * is connected to microTLB 0 on currently supported platforms, so we
+	 * can't use it with the IPMMU. As the IOMMU API operates at the device
+	 * level we can't disable it selectively, so ignore channel 0 for now if
+	 * the device is part of an IOMMU group.
+	 */
+	if (pdev->dev.iommu_group) {
+		dmac->n_channels--;
+		channels_offset = 1;
+	}
+
+	dmac->channels = devm_kcalloc(&pdev->dev, dmac->n_channels,
+				      sizeof(*dmac->channels), GFP_KERNEL);
+	if (!dmac->channels)
+		return -ENOMEM;
+
+	/* Request resources. */
+	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	dmac->iomem = devm_ioremap_resource(&pdev->dev, mem);
+	if (IS_ERR(dmac->iomem))
+		return PTR_ERR(dmac->iomem);
+
+	irq = platform_get_irq_byname(pdev, "error");
+	if (irq < 0) {
+		dev_err(&pdev->dev, "no error IRQ specified\n");
+		return -ENODEV;
+	}
+
+	irqname = devm_kasprintf(dmac->dev, GFP_KERNEL, "%s:error",
+				 dev_name(dmac->dev));
+	if (!irqname)
+		return -ENOMEM;
+
+	ret = devm_request_irq(&pdev->dev, irq, rcar_dmac_isr_error, 0,
+			       irqname, dmac);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to request IRQ %u (%d)\n",
+			irq, ret);
+		return ret;
+	}
+
+	/* Enable runtime PM and initialize the device. */
+	pm_runtime_enable(&pdev->dev);
+	ret = pm_runtime_get_sync(&pdev->dev);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "runtime PM get sync failed (%d)\n", ret);
+		return ret;
+	}
+
+	ret = rcar_dmac_init(dmac);
+	pm_runtime_put(&pdev->dev);
+
+	if (ret) {
+		dev_err(&pdev->dev, "failed to reset device\n");
+		goto error;
+	}
+
+	/* Initialize the channels. */
+	INIT_LIST_HEAD(&dmac->engine.channels);
+
+	for (i = 0; i < dmac->n_channels; ++i) {
+		ret = rcar_dmac_chan_probe(dmac, &dmac->channels[i],
+					   i + channels_offset);
+		if (ret < 0)
+			goto error;
+	}
+
+	/* Register the DMAC as a DMA provider for DT. */
+	ret = of_dma_controller_register(pdev->dev.of_node, rcar_dmac_of_xlate,
+					 NULL);
+	if (ret < 0)
+		goto error;
+
+	/*
+	 * Register the DMA engine device.
+	 *
+	 * Default transfer size of 32 bytes requires 32-byte alignment.
+	 */
+	engine = &dmac->engine;
+	dma_cap_set(DMA_MEMCPY, engine->cap_mask);
+	dma_cap_set(DMA_SLAVE, engine->cap_mask);
+
+	engine->dev = &pdev->dev;
+	engine->copy_align = ilog2(RCAR_DMAC_MEMCPY_XFER_SIZE);
+
+	engine->src_addr_widths = widths;
+	engine->dst_addr_widths = widths;
+	engine->directions = BIT(DMA_MEM_TO_DEV) | BIT(DMA_DEV_TO_MEM);
+	engine->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+
+	engine->device_alloc_chan_resources = rcar_dmac_alloc_chan_resources;
+	engine->device_free_chan_resources = rcar_dmac_free_chan_resources;
+	engine->device_prep_dma_memcpy = rcar_dmac_prep_dma_memcpy;
+	engine->device_prep_slave_sg = rcar_dmac_prep_slave_sg;
+	engine->device_prep_dma_cyclic = rcar_dmac_prep_dma_cyclic;
+	engine->device_config = rcar_dmac_device_config;
+	engine->device_terminate_all = rcar_dmac_chan_terminate_all;
+	engine->device_tx_status = rcar_dmac_tx_status;
+	engine->device_issue_pending = rcar_dmac_issue_pending;
+
+	ret = dma_async_device_register(engine);
+	if (ret < 0)
+		goto error;
+
+	return 0;
+
+error:
+	of_dma_controller_free(pdev->dev.of_node);
+	pm_runtime_disable(&pdev->dev);
+	return ret;
+}
+
+static int rcar_dmac_remove(struct platform_device *pdev)
+{
+	struct rcar_dmac *dmac = platform_get_drvdata(pdev);
+
+	of_dma_controller_free(pdev->dev.of_node);
+	dma_async_device_unregister(&dmac->engine);
+
+	pm_runtime_disable(&pdev->dev);
+
+	return 0;
+}
+
+static void rcar_dmac_shutdown(struct platform_device *pdev)
+{
+	struct rcar_dmac *dmac = platform_get_drvdata(pdev);
+
+	rcar_dmac_stop(dmac);
+}
+
+static const struct of_device_id rcar_dmac_of_ids[] = {
+	{ .compatible = "renesas,rcar-dmac", },
+	{ /* Sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, rcar_dmac_of_ids);
+
+static struct platform_driver rcar_dmac_driver = {
+	.driver		= {
+		.pm	= &rcar_dmac_pm,
+		.name	= "rcar-dmac",
+		.of_match_table = rcar_dmac_of_ids,
+	},
+	.probe		= rcar_dmac_probe,
+	.remove		= rcar_dmac_remove,
+	.shutdown	= rcar_dmac_shutdown,
+};
+
+module_platform_driver(rcar_dmac_driver);
+
+MODULE_DESCRIPTION("R-Car Gen2 DMA Controller Driver");
+MODULE_AUTHOR("Laurent Pinchart <laurent.pinchart@ideasonboard.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/dma/sh/rcar-hpbdma.c b/drivers/dma/sh/rcar-hpbdma.c
index 20a6f6f..749f26e 100644
--- a/drivers/dma/sh/rcar-hpbdma.c
+++ b/drivers/dma/sh/rcar-hpbdma.c
@@ -534,6 +534,8 @@
 
 static int hpb_dmae_probe(struct platform_device *pdev)
 {
+	const enum dma_slave_buswidth widths = DMA_SLAVE_BUSWIDTH_1_BYTE |
+		DMA_SLAVE_BUSWIDTH_2_BYTES | DMA_SLAVE_BUSWIDTH_4_BYTES;
 	struct hpb_dmae_pdata *pdata = pdev->dev.platform_data;
 	struct hpb_dmae_device *hpbdev;
 	struct dma_device *dma_dev;
@@ -595,6 +597,10 @@
 
 	dma_cap_set(DMA_MEMCPY, dma_dev->cap_mask);
 	dma_cap_set(DMA_SLAVE, dma_dev->cap_mask);
+	dma_dev->src_addr_widths = widths;
+	dma_dev->dst_addr_widths = widths;
+	dma_dev->directions = BIT(DMA_MEM_TO_DEV) | BIT(DMA_DEV_TO_MEM);
+	dma_dev->residue_granularity = DMA_RESIDUE_GRANULARITY_DESCRIPTOR;
 
 	hpbdev->shdma_dev.ops = &hpb_dmae_ops;
 	hpbdev->shdma_dev.desc_size = sizeof(struct hpb_desc);
diff --git a/drivers/dma/sh/shdma-base.c b/drivers/dma/sh/shdma-base.c
index 3a2adb1..8ee383d 100644
--- a/drivers/dma/sh/shdma-base.c
+++ b/drivers/dma/sh/shdma-base.c
@@ -729,57 +729,50 @@
 	return desc;
 }
 
-static int shdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-			  unsigned long arg)
+static int shdma_terminate_all(struct dma_chan *chan)
 {
 	struct shdma_chan *schan = to_shdma_chan(chan);
 	struct shdma_dev *sdev = to_shdma_dev(chan->device);
 	const struct shdma_ops *ops = sdev->ops;
-	struct dma_slave_config *config;
 	unsigned long flags;
-	int ret;
 
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		spin_lock_irqsave(&schan->chan_lock, flags);
-		ops->halt_channel(schan);
+	spin_lock_irqsave(&schan->chan_lock, flags);
+	ops->halt_channel(schan);
 
-		if (ops->get_partial && !list_empty(&schan->ld_queue)) {
-			/* Record partial transfer */
-			struct shdma_desc *desc = list_first_entry(&schan->ld_queue,
-						struct shdma_desc, node);
-			desc->partial = ops->get_partial(schan, desc);
-		}
-
-		spin_unlock_irqrestore(&schan->chan_lock, flags);
-
-		shdma_chan_ld_cleanup(schan, true);
-		break;
-	case DMA_SLAVE_CONFIG:
-		/*
-		 * So far only .slave_id is used, but the slave drivers are
-		 * encouraged to also set a transfer direction and an address.
-		 */
-		if (!arg)
-			return -EINVAL;
-		/*
-		 * We could lock this, but you shouldn't be configuring the
-		 * channel, while using it...
-		 */
-		config = (struct dma_slave_config *)arg;
-		ret = shdma_setup_slave(schan, config->slave_id,
-					config->direction == DMA_DEV_TO_MEM ?
-					config->src_addr : config->dst_addr);
-		if (ret < 0)
-			return ret;
-		break;
-	default:
-		return -ENXIO;
+	if (ops->get_partial && !list_empty(&schan->ld_queue)) {
+		/* Record partial transfer */
+		struct shdma_desc *desc = list_first_entry(&schan->ld_queue,
+							   struct shdma_desc, node);
+		desc->partial = ops->get_partial(schan, desc);
 	}
 
+	spin_unlock_irqrestore(&schan->chan_lock, flags);
+
+	shdma_chan_ld_cleanup(schan, true);
+
 	return 0;
 }
 
+static int shdma_config(struct dma_chan *chan,
+			struct dma_slave_config *config)
+{
+	struct shdma_chan *schan = to_shdma_chan(chan);
+
+	/*
+	 * So far only .slave_id is used, but the slave drivers are
+	 * encouraged to also set a transfer direction and an address.
+	 */
+	if (!config)
+		return -EINVAL;
+	/*
+	 * We could lock this, but you shouldn't be configuring the
+	 * channel, while using it...
+	 */
+	return shdma_setup_slave(schan, config->slave_id,
+				 config->direction == DMA_DEV_TO_MEM ?
+				 config->src_addr : config->dst_addr);
+}
+
 static void shdma_issue_pending(struct dma_chan *chan)
 {
 	struct shdma_chan *schan = to_shdma_chan(chan);
@@ -1002,7 +995,8 @@
 	/* Compulsory for DMA_SLAVE fields */
 	dma_dev->device_prep_slave_sg = shdma_prep_slave_sg;
 	dma_dev->device_prep_dma_cyclic = shdma_prep_dma_cyclic;
-	dma_dev->device_control = shdma_control;
+	dma_dev->device_config = shdma_config;
+	dma_dev->device_terminate_all = shdma_terminate_all;
 
 	dma_dev->dev = dev;
 
diff --git a/drivers/dma/sh/shdmac.c b/drivers/dma/sh/shdmac.c
index aec8a84..b2431aa 100644
--- a/drivers/dma/sh/shdmac.c
+++ b/drivers/dma/sh/shdmac.c
@@ -588,6 +588,7 @@
 	sh_dmae_ctl_stop(shdev);
 }
 
+#ifdef CONFIG_PM
 static int sh_dmae_runtime_suspend(struct device *dev)
 {
 	return 0;
@@ -599,8 +600,9 @@
 
 	return sh_dmae_rst(shdev);
 }
+#endif
 
-#ifdef CONFIG_PM
+#ifdef CONFIG_PM_SLEEP
 static int sh_dmae_suspend(struct device *dev)
 {
 	return 0;
@@ -632,16 +634,12 @@
 
 	return 0;
 }
-#else
-#define sh_dmae_suspend NULL
-#define sh_dmae_resume NULL
 #endif
 
 static const struct dev_pm_ops sh_dmae_pm = {
-	.suspend		= sh_dmae_suspend,
-	.resume			= sh_dmae_resume,
-	.runtime_suspend	= sh_dmae_runtime_suspend,
-	.runtime_resume		= sh_dmae_runtime_resume,
+	SET_SYSTEM_SLEEP_PM_OPS(sh_dmae_suspend, sh_dmae_resume)
+	SET_RUNTIME_PM_OPS(sh_dmae_runtime_suspend, sh_dmae_runtime_resume,
+			   NULL)
 };
 
 static dma_addr_t sh_dmae_slave_addr(struct shdma_chan *schan)
@@ -684,6 +682,10 @@
 
 static int sh_dmae_probe(struct platform_device *pdev)
 {
+	const enum dma_slave_buswidth widths =
+		DMA_SLAVE_BUSWIDTH_1_BYTE   | DMA_SLAVE_BUSWIDTH_2_BYTES |
+		DMA_SLAVE_BUSWIDTH_4_BYTES  | DMA_SLAVE_BUSWIDTH_8_BYTES |
+		DMA_SLAVE_BUSWIDTH_16_BYTES | DMA_SLAVE_BUSWIDTH_32_BYTES;
 	const struct sh_dmae_pdata *pdata;
 	unsigned long chan_flag[SH_DMAE_MAX_CHANNELS] = {};
 	int chan_irq[SH_DMAE_MAX_CHANNELS];
@@ -746,6 +748,11 @@
 			return PTR_ERR(shdev->dmars);
 	}
 
+	dma_dev->src_addr_widths = widths;
+	dma_dev->dst_addr_widths = widths;
+	dma_dev->directions = BIT(DMA_MEM_TO_DEV) | BIT(DMA_DEV_TO_MEM);
+	dma_dev->residue_granularity = DMA_RESIDUE_GRANULARITY_DESCRIPTOR;
+
 	if (!pdata->slave_only)
 		dma_cap_set(DMA_MEMCPY, dma_dev->cap_mask);
 	if (pdata->slave && pdata->slave_num)
diff --git a/drivers/dma/sirf-dma.c b/drivers/dma/sirf-dma.c
index 3492a5f..d0086e9 100644
--- a/drivers/dma/sirf-dma.c
+++ b/drivers/dma/sirf-dma.c
@@ -281,9 +281,10 @@
 	return cookie;
 }
 
-static int sirfsoc_dma_slave_config(struct sirfsoc_dma_chan *schan,
-	struct dma_slave_config *config)
+static int sirfsoc_dma_slave_config(struct dma_chan *chan,
+				    struct dma_slave_config *config)
 {
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
 	unsigned long flags;
 
 	if ((config->src_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES) ||
@@ -297,8 +298,9 @@
 	return 0;
 }
 
-static int sirfsoc_dma_terminate_all(struct sirfsoc_dma_chan *schan)
+static int sirfsoc_dma_terminate_all(struct dma_chan *chan)
 {
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
 	struct sirfsoc_dma *sdma = dma_chan_to_sirfsoc_dma(&schan->chan);
 	int cid = schan->chan.chan_id;
 	unsigned long flags;
@@ -327,8 +329,9 @@
 	return 0;
 }
 
-static int sirfsoc_dma_pause_chan(struct sirfsoc_dma_chan *schan)
+static int sirfsoc_dma_pause_chan(struct dma_chan *chan)
 {
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
 	struct sirfsoc_dma *sdma = dma_chan_to_sirfsoc_dma(&schan->chan);
 	int cid = schan->chan.chan_id;
 	unsigned long flags;
@@ -348,8 +351,9 @@
 	return 0;
 }
 
-static int sirfsoc_dma_resume_chan(struct sirfsoc_dma_chan *schan)
+static int sirfsoc_dma_resume_chan(struct dma_chan *chan)
 {
+	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
 	struct sirfsoc_dma *sdma = dma_chan_to_sirfsoc_dma(&schan->chan);
 	int cid = schan->chan.chan_id;
 	unsigned long flags;
@@ -369,30 +373,6 @@
 	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_PAUSE:
-		return sirfsoc_dma_pause_chan(schan);
-	case DMA_RESUME:
-		return sirfsoc_dma_resume_chan(schan);
-	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)
 {
@@ -648,18 +628,6 @@
 	BIT(DMA_SLAVE_BUSWIDTH_4_BYTES) | \
 	BIT(DMA_SLAVE_BUSWIDTH_8_BYTES))
 
-static int sirfsoc_dma_device_slave_caps(struct dma_chan *dchan,
-	struct dma_slave_caps *caps)
-{
-	caps->src_addr_widths = SIRFSOC_DMA_BUSWIDTHS;
-	caps->dstn_addr_widths = SIRFSOC_DMA_BUSWIDTHS;
-	caps->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
-	caps->cmd_pause = true;
-	caps->cmd_terminate = true;
-
-	return 0;
-}
-
 static struct dma_chan *of_dma_sirfsoc_xlate(struct of_phandle_args *dma_spec,
 	struct of_dma *ofdma)
 {
@@ -739,11 +707,16 @@
 	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_config = sirfsoc_dma_slave_config;
+	dma->device_pause = sirfsoc_dma_pause_chan;
+	dma->device_resume = sirfsoc_dma_resume_chan;
+	dma->device_terminate_all = sirfsoc_dma_terminate_all;
 	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;
-	dma->device_slave_caps = sirfsoc_dma_device_slave_caps;
+	dma->src_addr_widths = SIRFSOC_DMA_BUSWIDTHS;
+	dma->dst_addr_widths = SIRFSOC_DMA_BUSWIDTHS;
+	dma->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
 
 	INIT_LIST_HEAD(&dma->channels);
 	dma_cap_set(DMA_SLAVE, dma->cap_mask);
diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c
index 15d4946..68aca33 100644
--- a/drivers/dma/ste_dma40.c
+++ b/drivers/dma/ste_dma40.c
@@ -1429,11 +1429,17 @@
 	return is_link;
 }
 
-static int d40_pause(struct d40_chan *d40c)
+static int d40_pause(struct dma_chan *chan)
 {
+	struct d40_chan *d40c = container_of(chan, struct d40_chan, chan);
 	int res = 0;
 	unsigned long flags;
 
+	if (d40c->phy_chan == NULL) {
+		chan_err(d40c, "Channel is not allocated!\n");
+		return -EINVAL;
+	}
+
 	if (!d40c->busy)
 		return 0;
 
@@ -1448,11 +1454,17 @@
 	return res;
 }
 
-static int d40_resume(struct d40_chan *d40c)
+static int d40_resume(struct dma_chan *chan)
 {
+	struct d40_chan *d40c = container_of(chan, struct d40_chan, chan);
 	int res = 0;
 	unsigned long flags;
 
+	if (d40c->phy_chan == NULL) {
+		chan_err(d40c, "Channel is not allocated!\n");
+		return -EINVAL;
+	}
+
 	if (!d40c->busy)
 		return 0;
 
@@ -2604,12 +2616,17 @@
 	spin_unlock_irqrestore(&d40c->lock, flags);
 }
 
-static void d40_terminate_all(struct dma_chan *chan)
+static int d40_terminate_all(struct dma_chan *chan)
 {
 	unsigned long flags;
 	struct d40_chan *d40c = container_of(chan, struct d40_chan, chan);
 	int ret;
 
+	if (d40c->phy_chan == NULL) {
+		chan_err(d40c, "Channel is not allocated!\n");
+		return -EINVAL;
+	}
+
 	spin_lock_irqsave(&d40c->lock, flags);
 
 	pm_runtime_get_sync(d40c->base->dev);
@@ -2627,6 +2644,7 @@
 	d40c->busy = false;
 
 	spin_unlock_irqrestore(&d40c->lock, flags);
+	return 0;
 }
 
 static int
@@ -2673,6 +2691,11 @@
 	u32 src_maxburst, dst_maxburst;
 	int ret;
 
+	if (d40c->phy_chan == NULL) {
+		chan_err(d40c, "Channel is not allocated!\n");
+		return -EINVAL;
+	}
+
 	src_addr_width = config->src_addr_width;
 	src_maxburst = config->src_maxburst;
 	dst_addr_width = config->dst_addr_width;
@@ -2781,35 +2804,6 @@
 	return 0;
 }
 
-static int d40_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-		       unsigned long arg)
-{
-	struct d40_chan *d40c = container_of(chan, struct d40_chan, chan);
-
-	if (d40c->phy_chan == NULL) {
-		chan_err(d40c, "Channel is not allocated!\n");
-		return -EINVAL;
-	}
-
-	switch (cmd) {
-	case DMA_TERMINATE_ALL:
-		d40_terminate_all(chan);
-		return 0;
-	case DMA_PAUSE:
-		return d40_pause(d40c);
-	case DMA_RESUME:
-		return d40_resume(d40c);
-	case DMA_SLAVE_CONFIG:
-		return d40_set_runtime_config(chan,
-			(struct dma_slave_config *) arg);
-	default:
-		break;
-	}
-
-	/* Other commands are unimplemented */
-	return -ENXIO;
-}
-
 /* Initialization functions */
 
 static void __init d40_chan_init(struct d40_base *base, struct dma_device *dma,
@@ -2870,7 +2864,10 @@
 	dev->device_free_chan_resources = d40_free_chan_resources;
 	dev->device_issue_pending = d40_issue_pending;
 	dev->device_tx_status = d40_tx_status;
-	dev->device_control = d40_control;
+	dev->device_config = d40_set_runtime_config;
+	dev->device_pause = d40_pause;
+	dev->device_resume = d40_resume;
+	dev->device_terminate_all = d40_terminate_all;
 	dev->dev = base->dev;
 }
 
diff --git a/drivers/dma/sun6i-dma.c b/drivers/dma/sun6i-dma.c
index 159f173..7ebcf9b 100644
--- a/drivers/dma/sun6i-dma.c
+++ b/drivers/dma/sun6i-dma.c
@@ -355,38 +355,6 @@
 	kfree(txd);
 }
 
-static int sun6i_dma_terminate_all(struct sun6i_vchan *vchan)
-{
-	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(vchan->vc.chan.device);
-	struct sun6i_pchan *pchan = vchan->phy;
-	unsigned long flags;
-	LIST_HEAD(head);
-
-	spin_lock(&sdev->lock);
-	list_del_init(&vchan->node);
-	spin_unlock(&sdev->lock);
-
-	spin_lock_irqsave(&vchan->vc.lock, flags);
-
-	vchan_get_all_descriptors(&vchan->vc, &head);
-
-	if (pchan) {
-		writel(DMA_CHAN_ENABLE_STOP, pchan->base + DMA_CHAN_ENABLE);
-		writel(DMA_CHAN_PAUSE_RESUME, pchan->base + DMA_CHAN_PAUSE);
-
-		vchan->phy = NULL;
-		pchan->vchan = NULL;
-		pchan->desc = NULL;
-		pchan->done = NULL;
-	}
-
-	spin_unlock_irqrestore(&vchan->vc.lock, flags);
-
-	vchan_dma_desc_free_list(&vchan->vc, &head);
-
-	return 0;
-}
-
 static int sun6i_dma_start_desc(struct sun6i_vchan *vchan)
 {
 	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(vchan->vc.chan.device);
@@ -675,57 +643,92 @@
 	return NULL;
 }
 
-static int sun6i_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-		       unsigned long arg)
+static int sun6i_dma_config(struct dma_chan *chan,
+			    struct dma_slave_config *config)
+{
+	struct sun6i_vchan *vchan = to_sun6i_vchan(chan);
+
+	memcpy(&vchan->cfg, config, sizeof(*config));
+
+	return 0;
+}
+
+static int sun6i_dma_pause(struct dma_chan *chan)
+{
+	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(chan->device);
+	struct sun6i_vchan *vchan = to_sun6i_vchan(chan);
+	struct sun6i_pchan *pchan = vchan->phy;
+
+	dev_dbg(chan2dev(chan), "vchan %p: pause\n", &vchan->vc);
+
+	if (pchan) {
+		writel(DMA_CHAN_PAUSE_PAUSE,
+		       pchan->base + DMA_CHAN_PAUSE);
+	} else {
+		spin_lock(&sdev->lock);
+		list_del_init(&vchan->node);
+		spin_unlock(&sdev->lock);
+	}
+
+	return 0;
+}
+
+static int sun6i_dma_resume(struct dma_chan *chan)
 {
 	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(chan->device);
 	struct sun6i_vchan *vchan = to_sun6i_vchan(chan);
 	struct sun6i_pchan *pchan = vchan->phy;
 	unsigned long flags;
-	int ret = 0;
 
-	switch (cmd) {
-	case DMA_RESUME:
-		dev_dbg(chan2dev(chan), "vchan %p: resume\n", &vchan->vc);
+	dev_dbg(chan2dev(chan), "vchan %p: resume\n", &vchan->vc);
 
-		spin_lock_irqsave(&vchan->vc.lock, flags);
+	spin_lock_irqsave(&vchan->vc.lock, flags);
 
-		if (pchan) {
-			writel(DMA_CHAN_PAUSE_RESUME,
-			       pchan->base + DMA_CHAN_PAUSE);
-		} else if (!list_empty(&vchan->vc.desc_issued)) {
-			spin_lock(&sdev->lock);
-			list_add_tail(&vchan->node, &sdev->pending);
-			spin_unlock(&sdev->lock);
-		}
-
-		spin_unlock_irqrestore(&vchan->vc.lock, flags);
-		break;
-
-	case DMA_PAUSE:
-		dev_dbg(chan2dev(chan), "vchan %p: pause\n", &vchan->vc);
-
-		if (pchan) {
-			writel(DMA_CHAN_PAUSE_PAUSE,
-			       pchan->base + DMA_CHAN_PAUSE);
-		} else {
-			spin_lock(&sdev->lock);
-			list_del_init(&vchan->node);
-			spin_unlock(&sdev->lock);
-		}
-		break;
-
-	case DMA_TERMINATE_ALL:
-		ret = sun6i_dma_terminate_all(vchan);
-		break;
-	case DMA_SLAVE_CONFIG:
-		memcpy(&vchan->cfg, (void *)arg, sizeof(struct dma_slave_config));
-		break;
-	default:
-		ret = -ENXIO;
-		break;
+	if (pchan) {
+		writel(DMA_CHAN_PAUSE_RESUME,
+		       pchan->base + DMA_CHAN_PAUSE);
+	} else if (!list_empty(&vchan->vc.desc_issued)) {
+		spin_lock(&sdev->lock);
+		list_add_tail(&vchan->node, &sdev->pending);
+		spin_unlock(&sdev->lock);
 	}
-	return ret;
+
+	spin_unlock_irqrestore(&vchan->vc.lock, flags);
+
+	return 0;
+}
+
+static int sun6i_dma_terminate_all(struct dma_chan *chan)
+{
+	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(chan->device);
+	struct sun6i_vchan *vchan = to_sun6i_vchan(chan);
+	struct sun6i_pchan *pchan = vchan->phy;
+	unsigned long flags;
+	LIST_HEAD(head);
+
+	spin_lock(&sdev->lock);
+	list_del_init(&vchan->node);
+	spin_unlock(&sdev->lock);
+
+	spin_lock_irqsave(&vchan->vc.lock, flags);
+
+	vchan_get_all_descriptors(&vchan->vc, &head);
+
+	if (pchan) {
+		writel(DMA_CHAN_ENABLE_STOP, pchan->base + DMA_CHAN_ENABLE);
+		writel(DMA_CHAN_PAUSE_RESUME, pchan->base + DMA_CHAN_PAUSE);
+
+		vchan->phy = NULL;
+		pchan->vchan = NULL;
+		pchan->desc = NULL;
+		pchan->done = NULL;
+	}
+
+	spin_unlock_irqrestore(&vchan->vc.lock, flags);
+
+	vchan_dma_desc_free_list(&vchan->vc, &head);
+
+	return 0;
 }
 
 static enum dma_status sun6i_dma_tx_status(struct dma_chan *chan,
@@ -960,9 +963,20 @@
 	sdc->slave.device_issue_pending		= sun6i_dma_issue_pending;
 	sdc->slave.device_prep_slave_sg		= sun6i_dma_prep_slave_sg;
 	sdc->slave.device_prep_dma_memcpy	= sun6i_dma_prep_dma_memcpy;
-	sdc->slave.device_control		= sun6i_dma_control;
 	sdc->slave.copy_align			= 4;
-
+	sdc->slave.device_config		= sun6i_dma_config;
+	sdc->slave.device_pause			= sun6i_dma_pause;
+	sdc->slave.device_resume		= sun6i_dma_resume;
+	sdc->slave.device_terminate_all		= sun6i_dma_terminate_all;
+	sdc->slave.src_addr_widths		= BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |
+						  BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) |
+						  BIT(DMA_SLAVE_BUSWIDTH_4_BYTES);
+	sdc->slave.dst_addr_widths		= BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |
+						  BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) |
+						  BIT(DMA_SLAVE_BUSWIDTH_4_BYTES);
+	sdc->slave.directions			= BIT(DMA_DEV_TO_MEM) |
+						  BIT(DMA_MEM_TO_DEV);
+	sdc->slave.residue_granularity		= DMA_RESIDUE_GRANULARITY_BURST;
 	sdc->slave.dev = &pdev->dev;
 
 	sdc->pchans = devm_kcalloc(&pdev->dev, sdc->cfg->nr_max_channels,
diff --git a/drivers/dma/tegra20-apb-dma.c b/drivers/dma/tegra20-apb-dma.c
index d8450c3..eaf585e 100644
--- a/drivers/dma/tegra20-apb-dma.c
+++ b/drivers/dma/tegra20-apb-dma.c
@@ -723,7 +723,7 @@
 	return;
 }
 
-static void tegra_dma_terminate_all(struct dma_chan *dc)
+static int tegra_dma_terminate_all(struct dma_chan *dc)
 {
 	struct tegra_dma_channel *tdc = to_tegra_dma_chan(dc);
 	struct tegra_dma_sg_req *sgreq;
@@ -736,7 +736,7 @@
 	spin_lock_irqsave(&tdc->lock, flags);
 	if (list_empty(&tdc->pending_sg_req)) {
 		spin_unlock_irqrestore(&tdc->lock, flags);
-		return;
+		return 0;
 	}
 
 	if (!tdc->busy)
@@ -777,6 +777,7 @@
 		dma_desc->cb_count = 0;
 	}
 	spin_unlock_irqrestore(&tdc->lock, flags);
+	return 0;
 }
 
 static enum dma_status tegra_dma_tx_status(struct dma_chan *dc,
@@ -827,25 +828,6 @@
 	return ret;
 }
 
-static int tegra_dma_device_control(struct dma_chan *dc, enum dma_ctrl_cmd cmd,
-			unsigned long arg)
-{
-	switch (cmd) {
-	case DMA_SLAVE_CONFIG:
-		return tegra_dma_slave_config(dc,
-				(struct dma_slave_config *)arg);
-
-	case DMA_TERMINATE_ALL:
-		tegra_dma_terminate_all(dc);
-		return 0;
-
-	default:
-		break;
-	}
-
-	return -ENXIO;
-}
-
 static inline int get_bus_width(struct tegra_dma_channel *tdc,
 		enum dma_slave_buswidth slave_bw)
 {
@@ -1443,7 +1425,23 @@
 					tegra_dma_free_chan_resources;
 	tdma->dma_dev.device_prep_slave_sg = tegra_dma_prep_slave_sg;
 	tdma->dma_dev.device_prep_dma_cyclic = tegra_dma_prep_dma_cyclic;
-	tdma->dma_dev.device_control = tegra_dma_device_control;
+	tdma->dma_dev.src_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |
+		BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) |
+		BIT(DMA_SLAVE_BUSWIDTH_4_BYTES) |
+		BIT(DMA_SLAVE_BUSWIDTH_8_BYTES);
+	tdma->dma_dev.dst_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) |
+		BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) |
+		BIT(DMA_SLAVE_BUSWIDTH_4_BYTES) |
+		BIT(DMA_SLAVE_BUSWIDTH_8_BYTES);
+	tdma->dma_dev.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	/*
+	 * XXX The hardware appears to support
+	 * DMA_RESIDUE_GRANULARITY_BURST-level reporting, but it's
+	 * only used by this driver during tegra_dma_terminate_all()
+	 */
+	tdma->dma_dev.residue_granularity = DMA_RESIDUE_GRANULARITY_SEGMENT;
+	tdma->dma_dev.device_config = tegra_dma_slave_config;
+	tdma->dma_dev.device_terminate_all = tegra_dma_terminate_all;
 	tdma->dma_dev.device_tx_status = tegra_dma_tx_status;
 	tdma->dma_dev.device_issue_pending = tegra_dma_issue_pending;
 
diff --git a/drivers/dma/timb_dma.c b/drivers/dma/timb_dma.c
index 2407ccf..c4c3d93 100644
--- a/drivers/dma/timb_dma.c
+++ b/drivers/dma/timb_dma.c
@@ -561,8 +561,7 @@
 	return &td_desc->txd;
 }
 
-static int td_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-		      unsigned long arg)
+static int td_terminate_all(struct dma_chan *chan)
 {
 	struct timb_dma_chan *td_chan =
 		container_of(chan, struct timb_dma_chan, chan);
@@ -570,9 +569,6 @@
 
 	dev_dbg(chan2dev(chan), "%s: Entry\n", __func__);
 
-	if (cmd != DMA_TERMINATE_ALL)
-		return -ENXIO;
-
 	/* first the easy part, put the queue into the free list */
 	spin_lock_bh(&td_chan->lock);
 	list_for_each_entry_safe(td_desc, _td_desc, &td_chan->queue,
@@ -697,7 +693,7 @@
 	dma_cap_set(DMA_SLAVE, td->dma.cap_mask);
 	dma_cap_set(DMA_PRIVATE, td->dma.cap_mask);
 	td->dma.device_prep_slave_sg = td_prep_slave_sg;
-	td->dma.device_control = td_control;
+	td->dma.device_terminate_all = td_terminate_all;
 
 	td->dma.dev = &pdev->dev;
 
diff --git a/drivers/dma/txx9dmac.c b/drivers/dma/txx9dmac.c
index 0659ec9..8849318 100644
--- a/drivers/dma/txx9dmac.c
+++ b/drivers/dma/txx9dmac.c
@@ -901,17 +901,12 @@
 	return &first->txd;
 }
 
-static int txx9dmac_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
-			    unsigned long arg)
+static int txx9dmac_terminate_all(struct dma_chan *chan)
 {
 	struct txx9dmac_chan *dc = to_txx9dmac_chan(chan);
 	struct txx9dmac_desc *desc, *_desc;
 	LIST_HEAD(list);
 
-	/* Only supports DMA_TERMINATE_ALL */
-	if (cmd != DMA_TERMINATE_ALL)
-		return -EINVAL;
-
 	dev_vdbg(chan2dev(chan), "terminate_all\n");
 	spin_lock_bh(&dc->lock);
 
@@ -1109,7 +1104,7 @@
 	dc->dma.dev = &pdev->dev;
 	dc->dma.device_alloc_chan_resources = txx9dmac_alloc_chan_resources;
 	dc->dma.device_free_chan_resources = txx9dmac_free_chan_resources;
-	dc->dma.device_control = txx9dmac_control;
+	dc->dma.device_terminate_all = txx9dmac_terminate_all;
 	dc->dma.device_tx_status = txx9dmac_tx_status;
 	dc->dma.device_issue_pending = txx9dmac_issue_pending;
 	if (pdata && pdata->memcpy_chan == ch) {
diff --git a/drivers/dma/xilinx/xilinx_vdma.c b/drivers/dma/xilinx/xilinx_vdma.c
index 4a3a8f3..bdd2a5d 100644
--- a/drivers/dma/xilinx/xilinx_vdma.c
+++ b/drivers/dma/xilinx/xilinx_vdma.c
@@ -1001,13 +1001,17 @@
  * xilinx_vdma_terminate_all - Halt the channel and free descriptors
  * @chan: Driver specific VDMA Channel pointer
  */
-static void xilinx_vdma_terminate_all(struct xilinx_vdma_chan *chan)
+static int xilinx_vdma_terminate_all(struct dma_chan *dchan)
 {
+	struct xilinx_vdma_chan *chan = to_xilinx_chan(dchan);
+
 	/* Halt the DMA engine */
 	xilinx_vdma_halt(chan);
 
 	/* Remove and free all of the descriptors in the lists */
 	xilinx_vdma_free_descriptors(chan);
+
+	return 0;
 }
 
 /**
@@ -1075,27 +1079,6 @@
 }
 EXPORT_SYMBOL(xilinx_vdma_channel_set_config);
 
-/**
- * xilinx_vdma_device_control - Configure DMA channel of the device
- * @dchan: DMA Channel pointer
- * @cmd: DMA control command
- * @arg: Channel configuration
- *
- * Return: '0' on success and failure value on error
- */
-static int xilinx_vdma_device_control(struct dma_chan *dchan,
-				      enum dma_ctrl_cmd cmd, unsigned long arg)
-{
-	struct xilinx_vdma_chan *chan = to_xilinx_chan(dchan);
-
-	if (cmd != DMA_TERMINATE_ALL)
-		return -ENXIO;
-
-	xilinx_vdma_terminate_all(chan);
-
-	return 0;
-}
-
 /* -----------------------------------------------------------------------------
  * Probe and remove
  */
@@ -1300,7 +1283,7 @@
 				xilinx_vdma_free_chan_resources;
 	xdev->common.device_prep_interleaved_dma =
 				xilinx_vdma_dma_prep_interleaved;
-	xdev->common.device_control = xilinx_vdma_device_control;
+	xdev->common.device_terminate_all = xilinx_vdma_terminate_all;
 	xdev->common.device_tx_status = xilinx_vdma_tx_status;
 	xdev->common.device_issue_pending = xilinx_vdma_issue_pending;
 
diff --git a/drivers/rapidio/devices/tsi721_dma.c b/drivers/rapidio/devices/tsi721_dma.c
index f64c5de..4729594 100644
--- a/drivers/rapidio/devices/tsi721_dma.c
+++ b/drivers/rapidio/devices/tsi721_dma.c
@@ -815,8 +815,7 @@
 	return txd;
 }
 
-static int tsi721_device_control(struct dma_chan *dchan, enum dma_ctrl_cmd cmd,
-			     unsigned long arg)
+static int tsi721_terminate_all(struct dma_chan *dchan)
 {
 	struct tsi721_bdma_chan *bdma_chan = to_tsi721_chan(dchan);
 	struct tsi721_tx_desc *desc, *_d;
@@ -825,9 +824,6 @@
 
 	dev_dbg(dchan->device->dev, "%s: Entry\n", __func__);
 
-	if (cmd != DMA_TERMINATE_ALL)
-		return -ENOSYS;
-
 	spin_lock_bh(&bdma_chan->lock);
 
 	bdma_chan->active = false;
@@ -901,7 +897,7 @@
 	mport->dma.device_tx_status = tsi721_tx_status;
 	mport->dma.device_issue_pending = tsi721_issue_pending;
 	mport->dma.device_prep_slave_sg = tsi721_prep_rio_sg;
-	mport->dma.device_control = tsi721_device_control;
+	mport->dma.device_terminate_all = tsi721_terminate_all;
 
 	err = dma_async_device_register(&mport->dma);
 	if (err)
diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index 40cd75e..b6997a0 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -189,25 +189,6 @@
 };
 
 /**
- * enum dma_ctrl_cmd - DMA operations that can optionally be exercised
- * on a running channel.
- * @DMA_TERMINATE_ALL: terminate all ongoing transfers
- * @DMA_PAUSE: pause ongoing transfers
- * @DMA_RESUME: resume paused transfer
- * @DMA_SLAVE_CONFIG: this command is only implemented by DMA controllers
- * that need to runtime reconfigure the slave channels (as opposed to passing
- * configuration data in statically from the platform). An additional
- * argument of struct dma_slave_config must be passed in with this
- * command.
- */
-enum dma_ctrl_cmd {
-	DMA_TERMINATE_ALL,
-	DMA_PAUSE,
-	DMA_RESUME,
-	DMA_SLAVE_CONFIG,
-};
-
-/**
  * enum sum_check_bits - bit position of pq_check_flags
  */
 enum sum_check_bits {
@@ -298,6 +279,9 @@
 	DMA_SLAVE_BUSWIDTH_3_BYTES = 3,
 	DMA_SLAVE_BUSWIDTH_4_BYTES = 4,
 	DMA_SLAVE_BUSWIDTH_8_BYTES = 8,
+	DMA_SLAVE_BUSWIDTH_16_BYTES = 16,
+	DMA_SLAVE_BUSWIDTH_32_BYTES = 32,
+	DMA_SLAVE_BUSWIDTH_64_BYTES = 64,
 };
 
 /**
@@ -336,9 +320,8 @@
  * This struct is passed in as configuration data to a DMA engine
  * in order to set up a certain channel for DMA transport at runtime.
  * The DMA device/engine has to provide support for an additional
- * command in the channel config interface, DMA_SLAVE_CONFIG
- * and this struct will then be passed in as an argument to the
- * DMA engine device_control() function.
+ * callback in the dma_device structure, device_config and this struct
+ * will then be passed in as an argument to the function.
  *
  * The rationale for adding configuration information to this struct is as
  * follows: if it is likely that more than one DMA slave controllers in
@@ -387,7 +370,7 @@
 /* struct dma_slave_caps - expose capabilities of a slave channel only
  *
  * @src_addr_widths: bit mask of src addr widths the channel supports
- * @dstn_addr_widths: bit mask of dstn addr widths the channel supports
+ * @dst_addr_widths: bit mask of dstn addr widths the channel supports
  * @directions: bit mask of slave direction the channel supported
  * 	since the enum dma_transfer_direction is not defined as bits for each
  * 	type of direction, the dma controller should fill (1 << <TYPE>) and same
@@ -398,7 +381,7 @@
  */
 struct dma_slave_caps {
 	u32 src_addr_widths;
-	u32 dstn_addr_widths;
+	u32 dst_addr_widths;
 	u32 directions;
 	bool cmd_pause;
 	bool cmd_terminate;
@@ -594,6 +577,14 @@
  * @fill_align: alignment shift for memset operations
  * @dev_id: unique device ID
  * @dev: struct device reference for dma mapping api
+ * @src_addr_widths: bit mask of src addr widths the device supports
+ * @dst_addr_widths: bit mask of dst addr widths the device supports
+ * @directions: bit mask of slave direction the device supports since
+ * 	the enum dma_transfer_direction is not defined as bits for
+ * 	each type of direction, the dma controller should fill (1 <<
+ * 	<TYPE>) and same should be checked by controller as well
+ * @residue_granularity: granularity of the transfer residue reported
+ *	by tx_status
  * @device_alloc_chan_resources: allocate resources and return the
  *	number of allocated descriptors
  * @device_free_chan_resources: release DMA channel's resources
@@ -608,14 +599,19 @@
  *	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_config: Pushes a new configuration to a channel, return 0 or an error
+ *	code
+ * @device_pause: Pauses any transfer happening on a channel. Returns
+ *	0 or an error code
+ * @device_resume: Resumes any transfer on a channel previously
+ *	paused. Returns 0 or an error code
+ * @device_terminate_all: Aborts all transfers on a channel. Returns 0
+ *	or an error code
  * @device_tx_status: poll for transaction completion, the optional
  *	txstate parameter can be supplied with a pointer to get a
  *	struct with auxiliary transfer status information, otherwise the call
  *	will just return a simple status code
  * @device_issue_pending: push pending transactions to hardware
- * @device_slave_caps: return the slave channel capabilities
  */
 struct dma_device {
 
@@ -635,14 +631,19 @@
 	int dev_id;
 	struct device *dev;
 
+	u32 src_addr_widths;
+	u32 dst_addr_widths;
+	u32 directions;
+	enum dma_residue_granularity residue_granularity;
+
 	int (*device_alloc_chan_resources)(struct dma_chan *chan);
 	void (*device_free_chan_resources)(struct dma_chan *chan);
 
 	struct dma_async_tx_descriptor *(*device_prep_dma_memcpy)(
-		struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
+		struct dma_chan *chan, dma_addr_t dst, dma_addr_t src,
 		size_t len, unsigned long flags);
 	struct dma_async_tx_descriptor *(*device_prep_dma_xor)(
-		struct dma_chan *chan, dma_addr_t dest, dma_addr_t *src,
+		struct dma_chan *chan, dma_addr_t dst, dma_addr_t *src,
 		unsigned int src_cnt, size_t len, unsigned long flags);
 	struct dma_async_tx_descriptor *(*device_prep_dma_xor_val)(
 		struct dma_chan *chan, dma_addr_t *src,	unsigned int src_cnt,
@@ -674,31 +675,26 @@
 	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);
+
+	int (*device_config)(struct dma_chan *chan,
+			     struct dma_slave_config *config);
+	int (*device_pause)(struct dma_chan *chan);
+	int (*device_resume)(struct dma_chan *chan);
+	int (*device_terminate_all)(struct dma_chan *chan);
 
 	enum dma_status (*device_tx_status)(struct dma_chan *chan,
 					    dma_cookie_t cookie,
 					    struct dma_tx_state *txstate);
 	void (*device_issue_pending)(struct dma_chan *chan);
-	int (*device_slave_caps)(struct dma_chan *chan, struct dma_slave_caps *caps);
 };
 
-static inline int dmaengine_device_control(struct dma_chan *chan,
-					   enum dma_ctrl_cmd cmd,
-					   unsigned long arg)
-{
-	if (chan->device->device_control)
-		return chan->device->device_control(chan, cmd, arg);
-
-	return -ENOSYS;
-}
-
 static inline int dmaengine_slave_config(struct dma_chan *chan,
 					  struct dma_slave_config *config)
 {
-	return dmaengine_device_control(chan, DMA_SLAVE_CONFIG,
-			(unsigned long)config);
+	if (chan->device->device_config)
+		return chan->device->device_config(chan, config);
+
+	return -ENOSYS;
 }
 
 static inline bool is_slave_direction(enum dma_transfer_direction direction)
@@ -765,34 +761,28 @@
 			src_sg, src_nents, flags);
 }
 
-static inline int dma_get_slave_caps(struct dma_chan *chan, struct dma_slave_caps *caps)
-{
-	if (!chan || !caps)
-		return -EINVAL;
-
-	/* check if the channel supports slave transactions */
-	if (!test_bit(DMA_SLAVE, chan->device->cap_mask.bits))
-		return -ENXIO;
-
-	if (chan->device->device_slave_caps)
-		return chan->device->device_slave_caps(chan, caps);
-
-	return -ENXIO;
-}
-
 static inline int dmaengine_terminate_all(struct dma_chan *chan)
 {
-	return dmaengine_device_control(chan, DMA_TERMINATE_ALL, 0);
+	if (chan->device->device_terminate_all)
+		return chan->device->device_terminate_all(chan);
+
+	return -ENOSYS;
 }
 
 static inline int dmaengine_pause(struct dma_chan *chan)
 {
-	return dmaengine_device_control(chan, DMA_PAUSE, 0);
+	if (chan->device->device_pause)
+		return chan->device->device_pause(chan);
+
+	return -ENOSYS;
 }
 
 static inline int dmaengine_resume(struct dma_chan *chan)
 {
-	return dmaengine_device_control(chan, DMA_RESUME, 0);
+	if (chan->device->device_resume)
+		return chan->device->device_resume(chan);
+
+	return -ENOSYS;
 }
 
 static inline enum dma_status dmaengine_tx_status(struct dma_chan *chan,
@@ -1059,6 +1049,7 @@
 						  const char *name);
 struct dma_chan *dma_request_slave_channel(struct device *dev, const char *name);
 void dma_release_channel(struct dma_chan *chan);
+int dma_get_slave_caps(struct dma_chan *chan, struct dma_slave_caps *caps);
 #else
 static inline struct dma_chan *dma_find_channel(enum dma_transaction_type tx_type)
 {
@@ -1093,6 +1084,11 @@
 static inline void dma_release_channel(struct dma_chan *chan)
 {
 }
+static inline int dma_get_slave_caps(struct dma_chan *chan,
+				     struct dma_slave_caps *caps)
+{
+	return -ENXIO;
+}
 #endif
 
 /* --- DMA device --- */
diff --git a/include/linux/platform_data/dma-dw.h b/include/linux/platform_data/dma-dw.h
index d8155c0..87ac14c 100644
--- a/include/linux/platform_data/dma-dw.h
+++ b/include/linux/platform_data/dma-dw.h
@@ -13,10 +13,12 @@
 
 #include <linux/device.h>
 
+#define DW_DMA_MAX_NR_MASTERS	4
+
 /**
  * struct dw_dma_slave - Controller-specific information about a slave
  *
- * @dma_dev: required DMA master device. Depricated.
+ * @dma_dev:	required DMA master device
  * @src_id:	src request line
  * @dst_id:	dst request line
  * @src_master: src master for transfers on allocated channel.
@@ -53,7 +55,7 @@
 	unsigned char	chan_priority;
 	unsigned short	block_size;
 	unsigned char	nr_masters;
-	unsigned char	data_width[4];
+	unsigned char	data_width[DW_DMA_MAX_NR_MASTERS];
 };
 
 #endif /* _PLATFORM_DATA_DMA_DW_H */
diff --git a/include/linux/platform_data/dma-mmp_tdma.h b/include/linux/platform_data/dma-mmp_tdma.h
index 66574ea..0c72886 100644
--- a/include/linux/platform_data/dma-mmp_tdma.h
+++ b/include/linux/platform_data/dma-mmp_tdma.h
@@ -28,6 +28,13 @@
 	int granularity;
 };
 
+#ifdef CONFIG_ARM
 extern struct gen_pool *sram_get_gpool(char *pool_name);
+#else
+static inline struct gen_pool *sram_get_gpool(char *pool_name)
+{
+	return NULL;
+}
+#endif
 
 #endif /* __DMA_MMP_TDMA_H */
diff --git a/sound/soc/soc-generic-dmaengine-pcm.c b/sound/soc/soc-generic-dmaengine-pcm.c
index 4864392..c9917ca 100644
--- a/sound/soc/soc-generic-dmaengine-pcm.c
+++ b/sound/soc/soc-generic-dmaengine-pcm.c
@@ -151,7 +151,7 @@
 			hw.info |= SNDRV_PCM_INFO_BATCH;
 
 		if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK)
-			addr_widths = dma_caps.dstn_addr_widths;
+			addr_widths = dma_caps.dst_addr_widths;
 		else
 			addr_widths = dma_caps.src_addr_widths;
 	}