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

Pull slave-dmaengine update from Vinod Koul:
 "This includes the cookie cleanup by Russell, the addition of context
  parameter for dmaengine APIs, more arm dmaengine driver cleanup by
  moving code to dmaengine, this time for imx by Javier and pl330 by
  Boojin along with the usual driver fixes."

Fix up some fairly trivial conflicts with various other cleanups.

* 'next' of git://git.infradead.org/users/vkoul/slave-dma: (67 commits)
  dmaengine: imx: fix the build failure on x86_64
  dmaengine: i.MX: Fix merge of cookie branch.
  dmaengine: i.MX: Add support for interleaved transfers.
  dmaengine: imx-dma: use 'dev_dbg' and 'dev_warn' for messages.
  dmaengine: imx-dma: remove 'imx_dmav1_baseaddr' and 'dma_clk'.
  dmaengine: imx-dma: remove unused arg of imxdma_sg_next.
  dmaengine: imx-dma: remove internal structure.
  dmaengine: imx-dma: remove 'resbytes' field of 'internal' structure.
  dmaengine: imx-dma: remove 'in_use' field of 'internal' structure.
  dmaengine: imx-dma: remove sg member from internal structure.
  dmaengine: imx-dma: remove 'imxdma_setup_sg_hw' function.
  dmaengine: imx-dma: remove 'imxdma_config_channel_hw' function.
  dmaengine: imx-dma: remove 'imxdma_setup_mem2mem_hw' function.
  dmaengine: imx-dma: remove dma_mode member of internal structure.
  dmaengine: imx-dma: remove data member from internal structure.
  dmaengine: imx-dma: merge old dma-v1.c with imx-dma.c
  dmaengine: at_hdmac: add slave config operation
  dmaengine: add context parameter to prep_slave_sg and prep_dma_cyclic
  dmaengine/dma_slave: introduce inline wrappers
  dma: imx-sdma: Treat firmware messages as warnings instead of erros
  ...

37 files changed, 3547 insertions, 930 deletions

diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index 4a6c46dea8a0..cf9da362d64f 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -201,7 +201,6 @@ config PL330_DMA
 	tristate "DMA API Driver for PL330"
 	select DMA_ENGINE
 	depends on ARM_AMBA
-	select PL330
 	help
 	  Select if your platform has one or more PL330 DMACs.
 	  You need to provide platform specific settings via
@@ -231,7 +230,7 @@ config IMX_SDMA
 
 config IMX_DMA
 	tristate "i.MX DMA support"
-	depends on IMX_HAVE_DMA_V1
+	depends on ARCH_MXC
 	select DMA_ENGINE
 	help
 	  Support the i.MX DMA engine. This engine is integrated into
diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c
index 8a281584458b..c301a8ec31aa 100644
--- a/drivers/dma/amba-pl08x.c
+++ b/drivers/dma/amba-pl08x.c
@@ -85,6 +85,8 @@
 #include <linux/slab.h>
 #include <asm/hardware/pl080.h>
 
+#include "dmaengine.h"
+
 #define DRIVER_NAME	"pl08xdmac"
 
 static struct amba_driver pl08x_amba_driver;
@@ -649,7 +651,7 @@ static int pl08x_fill_llis_for_desc(struct pl08x_driver_data *pl08x,
 			}
 
 			if ((bd.srcbus.addr % bd.srcbus.buswidth) ||
-					(bd.srcbus.addr % bd.srcbus.buswidth)) {
+					(bd.dstbus.addr % bd.dstbus.buswidth)) {
 				dev_err(&pl08x->adev->dev,
 					"%s src & dst address must be aligned to src"
 					" & dst width if peripheral is flow controller",
@@ -919,13 +921,10 @@ static dma_cookie_t pl08x_tx_submit(struct dma_async_tx_descriptor *tx)
 	struct pl08x_dma_chan *plchan = to_pl08x_chan(tx->chan);
 	struct pl08x_txd *txd = to_pl08x_txd(tx);
 	unsigned long flags;
+	dma_cookie_t cookie;
 
 	spin_lock_irqsave(&plchan->lock, flags);
-
-	plchan->chan.cookie += 1;
-	if (plchan->chan.cookie < 0)
-		plchan->chan.cookie = 1;
-	tx->cookie = plchan->chan.cookie;
+	cookie = dma_cookie_assign(tx);
 
 	/* Put this onto the pending list */
 	list_add_tail(&txd->node, &plchan->pend_list);
@@ -945,7 +944,7 @@ static dma_cookie_t pl08x_tx_submit(struct dma_async_tx_descriptor *tx)
 
 	spin_unlock_irqrestore(&plchan->lock, flags);
 
-	return tx->cookie;
+	return cookie;
 }
 
 static struct dma_async_tx_descriptor *pl08x_prep_dma_interrupt(
@@ -965,31 +964,17 @@ static enum dma_status pl08x_dma_tx_status(struct dma_chan *chan,
 		dma_cookie_t cookie, struct dma_tx_state *txstate)
 {
 	struct pl08x_dma_chan *plchan = to_pl08x_chan(chan);
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
 	enum dma_status ret;
-	u32 bytesleft = 0;
 
-	last_used = plchan->chan.cookie;
-	last_complete = plchan->lc;
-
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
-	if (ret == DMA_SUCCESS) {
-		dma_set_tx_state(txstate, last_complete, last_used, 0);
+	ret = dma_cookie_status(chan, cookie, txstate);
+	if (ret == DMA_SUCCESS)
 		return ret;
-	}
 
 	/*
 	 * This cookie not complete yet
+	 * Get number of bytes left in the active transactions and queue
 	 */
-	last_used = plchan->chan.cookie;
-	last_complete = plchan->lc;
-
-	/* Get number of bytes left in the active transactions and queue */
-	bytesleft = pl08x_getbytes_chan(plchan);
-
-	dma_set_tx_state(txstate, last_complete, last_used,
-			 bytesleft);
+	dma_set_residue(txstate, pl08x_getbytes_chan(plchan));
 
 	if (plchan->state == PL08X_CHAN_PAUSED)
 		return DMA_PAUSED;
@@ -1139,6 +1124,8 @@ static int dma_set_runtime_config(struct dma_chan *chan,
 	cctl |= burst << PL080_CONTROL_SB_SIZE_SHIFT;
 	cctl |= burst << PL080_CONTROL_DB_SIZE_SHIFT;
 
+	plchan->device_fc = config->device_fc;
+
 	if (plchan->runtime_direction == DMA_DEV_TO_MEM) {
 		plchan->src_addr = config->src_addr;
 		plchan->src_cctl = pl08x_cctl(cctl) | PL080_CONTROL_DST_INCR |
@@ -1326,7 +1313,7 @@ static struct dma_async_tx_descriptor *pl08x_prep_dma_memcpy(
 static struct dma_async_tx_descriptor *pl08x_prep_slave_sg(
 		struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned int sg_len, enum dma_transfer_direction direction,
-		unsigned long flags)
+		unsigned long flags, void *context)
 {
 	struct pl08x_dma_chan *plchan = to_pl08x_chan(chan);
 	struct pl08x_driver_data *pl08x = plchan->host;
@@ -1370,7 +1357,7 @@ static struct dma_async_tx_descriptor *pl08x_prep_slave_sg(
 		return NULL;
 	}
 
-	if (plchan->cd->device_fc)
+	if (plchan->device_fc)
 		tmp = (direction == DMA_MEM_TO_DEV) ? PL080_FLOW_MEM2PER_PER :
 			PL080_FLOW_PER2MEM_PER;
 	else
@@ -1541,7 +1528,7 @@ static void pl08x_tasklet(unsigned long data)
 
 	if (txd) {
 		/* Update last completed */
-		plchan->lc = txd->tx.cookie;
+		dma_cookie_complete(&txd->tx);
 	}
 
 	/* If a new descriptor is queued, set it up plchan->at is NULL here */
@@ -1722,8 +1709,7 @@ static int pl08x_dma_init_virtual_channels(struct pl08x_driver_data *pl08x,
 			 chan->name);
 
 		chan->chan.device = dmadev;
-		chan->chan.cookie = 0;
-		chan->lc = 0;
+		dma_cookie_init(&chan->chan);
 
 		spin_lock_init(&chan->lock);
 		INIT_LIST_HEAD(&chan->pend_list);
diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
index f4aed5fc2cb6..7aa58d204892 100644
--- a/drivers/dma/at_hdmac.c
+++ b/drivers/dma/at_hdmac.c
@@ -27,6 +27,7 @@
 #include <linux/of_device.h>
 
 #include "at_hdmac_regs.h"
+#include "dmaengine.h"
 
 /*
  * Glossary
@@ -192,27 +193,6 @@ static void atc_desc_chain(struct at_desc **first, struct at_desc **prev,
 }
 
 /**
- * atc_assign_cookie - compute and assign new cookie
- * @atchan: channel we work on
- * @desc: descriptor to assign cookie for
- *
- * Called with atchan->lock held and bh disabled
- */
-static dma_cookie_t
-atc_assign_cookie(struct at_dma_chan *atchan, struct at_desc *desc)
-{
-	dma_cookie_t cookie = atchan->chan_common.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	atchan->chan_common.cookie = cookie;
-	desc->txd.cookie = cookie;
-
-	return cookie;
-}
-
-/**
  * atc_dostart - starts the DMA engine for real
  * @atchan: the channel we want to start
  * @first: first descriptor in the list we want to begin with
@@ -269,7 +249,7 @@ atc_chain_complete(struct at_dma_chan *atchan, struct at_desc *desc)
 	dev_vdbg(chan2dev(&atchan->chan_common),
 		"descriptor %u complete\n", txd->cookie);
 
-	atchan->completed_cookie = txd->cookie;
+	dma_cookie_complete(txd);
 
 	/* move children to free_list */
 	list_splice_init(&desc->tx_list, &atchan->free_list);
@@ -547,7 +527,7 @@ static dma_cookie_t atc_tx_submit(struct dma_async_tx_descriptor *tx)
 	unsigned long		flags;
 
 	spin_lock_irqsave(&atchan->lock, flags);
-	cookie = atc_assign_cookie(atchan, desc);
+	cookie = dma_cookie_assign(tx);
 
 	if (list_empty(&atchan->active_list)) {
 		dev_vdbg(chan2dev(tx->chan), "tx_submit: started %u\n",
@@ -659,14 +639,16 @@ err_desc_get:
  * @sg_len: number of entries in @scatterlist
  * @direction: DMA direction
  * @flags: tx descriptor status flags
+ * @context: transaction context (ignored)
  */
 static struct dma_async_tx_descriptor *
 atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned int sg_len, enum dma_transfer_direction direction,
-		unsigned long flags)
+		unsigned long flags, void *context)
 {
 	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
 	struct at_dma_slave	*atslave = chan->private;
+	struct dma_slave_config	*sconfig = &atchan->dma_sconfig;
 	struct at_desc		*first = NULL;
 	struct at_desc		*prev = NULL;
 	u32			ctrla;
@@ -688,19 +670,18 @@ atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		return NULL;
 	}
 
-	reg_width = atslave->reg_width;
-
 	ctrla = ATC_DEFAULT_CTRLA | atslave->ctrla;
 	ctrlb = ATC_IEN;
 
 	switch (direction) {
 	case DMA_MEM_TO_DEV:
+		reg_width = convert_buswidth(sconfig->dst_addr_width);
 		ctrla |=  ATC_DST_WIDTH(reg_width);
 		ctrlb |=  ATC_DST_ADDR_MODE_FIXED
 			| ATC_SRC_ADDR_MODE_INCR
 			| ATC_FC_MEM2PER
 			| ATC_SIF(AT_DMA_MEM_IF) | ATC_DIF(AT_DMA_PER_IF);
-		reg = atslave->tx_reg;
+		reg = sconfig->dst_addr;
 		for_each_sg(sgl, sg, sg_len, i) {
 			struct at_desc	*desc;
 			u32		len;
@@ -728,13 +709,14 @@ atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		}
 		break;
 	case DMA_DEV_TO_MEM:
+		reg_width = convert_buswidth(sconfig->src_addr_width);
 		ctrla |=  ATC_SRC_WIDTH(reg_width);
 		ctrlb |=  ATC_DST_ADDR_MODE_INCR
 			| ATC_SRC_ADDR_MODE_FIXED
 			| ATC_FC_PER2MEM
 			| ATC_SIF(AT_DMA_PER_IF) | ATC_DIF(AT_DMA_MEM_IF);
 
-		reg = atslave->rx_reg;
+		reg = sconfig->src_addr;
 		for_each_sg(sgl, sg, sg_len, i) {
 			struct at_desc	*desc;
 			u32		len;
@@ -810,12 +792,15 @@ err_out:
  * atc_dma_cyclic_fill_desc - Fill one period decriptor
  */
 static int
-atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc,
+atc_dma_cyclic_fill_desc(struct dma_chan *chan, struct at_desc *desc,
 		unsigned int period_index, dma_addr_t buf_addr,
-		size_t period_len, enum dma_transfer_direction direction)
+		unsigned int reg_width, size_t period_len,
+		enum dma_transfer_direction direction)
 {
-	u32		ctrla;
-	unsigned int	reg_width = atslave->reg_width;
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	struct at_dma_slave	*atslave = chan->private;
+	struct dma_slave_config	*sconfig = &atchan->dma_sconfig;
+	u32			ctrla;
 
 	/* prepare common CRTLA value */
 	ctrla =   ATC_DEFAULT_CTRLA | atslave->ctrla
@@ -826,7 +811,7 @@ atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc,
 	switch (direction) {
 	case DMA_MEM_TO_DEV:
 		desc->lli.saddr = buf_addr + (period_len * period_index);
-		desc->lli.daddr = atslave->tx_reg;
+		desc->lli.daddr = sconfig->dst_addr;
 		desc->lli.ctrla = ctrla;
 		desc->lli.ctrlb = ATC_DST_ADDR_MODE_FIXED
 				| ATC_SRC_ADDR_MODE_INCR
@@ -836,7 +821,7 @@ atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc,
 		break;
 
 	case DMA_DEV_TO_MEM:
-		desc->lli.saddr = atslave->rx_reg;
+		desc->lli.saddr = sconfig->src_addr;
 		desc->lli.daddr = buf_addr + (period_len * period_index);
 		desc->lli.ctrla = ctrla;
 		desc->lli.ctrlb = ATC_DST_ADDR_MODE_INCR
@@ -860,16 +845,20 @@ atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc,
  * @buf_len: total number of bytes for the entire buffer
  * @period_len: number of bytes for each period
  * @direction: transfer direction, to or from device
+ * @context: transfer context (ignored)
  */
 static struct dma_async_tx_descriptor *
 atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
-		size_t period_len, enum dma_transfer_direction direction)
+		size_t period_len, enum dma_transfer_direction direction,
+		void *context)
 {
 	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
 	struct at_dma_slave	*atslave = chan->private;
+	struct dma_slave_config	*sconfig = &atchan->dma_sconfig;
 	struct at_desc		*first = NULL;
 	struct at_desc		*prev = NULL;
 	unsigned long		was_cyclic;
+	unsigned int		reg_width;
 	unsigned int		periods = buf_len / period_len;
 	unsigned int		i;
 
@@ -889,8 +878,13 @@ atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
 		return NULL;
 	}
 
+	if (sconfig->direction == DMA_MEM_TO_DEV)
+		reg_width = convert_buswidth(sconfig->dst_addr_width);
+	else
+		reg_width = convert_buswidth(sconfig->src_addr_width);
+
 	/* Check for too big/unaligned periods and unaligned DMA buffer */
-	if (atc_dma_cyclic_check_values(atslave->reg_width, buf_addr,
+	if (atc_dma_cyclic_check_values(reg_width, buf_addr,
 					period_len, direction))
 		goto err_out;
 
@@ -902,8 +896,8 @@ atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
 		if (!desc)
 			goto err_desc_get;
 
-		if (atc_dma_cyclic_fill_desc(atslave, desc, i, buf_addr,
-						period_len, direction))
+		if (atc_dma_cyclic_fill_desc(chan, desc, i, buf_addr,
+					     reg_width, period_len, direction))
 			goto err_desc_get;
 
 		atc_desc_chain(&first, &prev, desc);
@@ -926,6 +920,23 @@ err_out:
 	return NULL;
 }
 
+static int set_runtime_config(struct dma_chan *chan,
+			      struct dma_slave_config *sconfig)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+
+	/* Check if it is chan is configured for slave transfers */
+	if (!chan->private)
+		return -EINVAL;
+
+	memcpy(&atchan->dma_sconfig, sconfig, sizeof(*sconfig));
+
+	convert_burst(&atchan->dma_sconfig.src_maxburst);
+	convert_burst(&atchan->dma_sconfig.dst_maxburst);
+
+	return 0;
+}
+
 
 static int atc_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 		       unsigned long arg)
@@ -986,6 +997,8 @@ static int atc_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 		clear_bit(ATC_IS_CYCLIC, &atchan->status);
 
 		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;
 	}
@@ -1016,26 +1029,20 @@ atc_tx_status(struct dma_chan *chan,
 
 	spin_lock_irqsave(&atchan->lock, flags);
 
-	last_complete = atchan->completed_cookie;
-	last_used = chan->cookie;
-
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
+	ret = dma_cookie_status(chan, cookie, txstate);
 	if (ret != DMA_SUCCESS) {
 		atc_cleanup_descriptors(atchan);
 
-		last_complete = atchan->completed_cookie;
-		last_used = chan->cookie;
-
-		ret = dma_async_is_complete(cookie, last_complete, last_used);
+		ret = dma_cookie_status(chan, cookie, txstate);
 	}
 
+	last_complete = chan->completed_cookie;
+	last_used = chan->cookie;
+
 	spin_unlock_irqrestore(&atchan->lock, flags);
 
 	if (ret != DMA_SUCCESS)
-		dma_set_tx_state(txstate, last_complete, last_used,
-			atc_first_active(atchan)->len);
-	else
-		dma_set_tx_state(txstate, last_complete, last_used, 0);
+		dma_set_residue(txstate, atc_first_active(atchan)->len);
 
 	if (atc_chan_is_paused(atchan))
 		ret = DMA_PAUSED;
@@ -1129,7 +1136,7 @@ static int atc_alloc_chan_resources(struct dma_chan *chan)
 	spin_lock_irqsave(&atchan->lock, flags);
 	atchan->descs_allocated = i;
 	list_splice(&tmp_list, &atchan->free_list);
-	atchan->completed_cookie = chan->cookie = 1;
+	dma_cookie_init(chan);
 	spin_unlock_irqrestore(&atchan->lock, flags);
 
 	/* channel parameters */
@@ -1329,7 +1336,7 @@ static int __init at_dma_probe(struct platform_device *pdev)
 		struct at_dma_chan	*atchan = &atdma->chan[i];
 
 		atchan->chan_common.device = &atdma->dma_common;
-		atchan->chan_common.cookie = atchan->completed_cookie = 1;
+		dma_cookie_init(&atchan->chan_common);
 		list_add_tail(&atchan->chan_common.device_node,
 				&atdma->dma_common.channels);
 
diff --git a/drivers/dma/at_hdmac_regs.h b/drivers/dma/at_hdmac_regs.h
index a8d3277d60b5..897a8bcaec90 100644
--- a/drivers/dma/at_hdmac_regs.h
+++ b/drivers/dma/at_hdmac_regs.h
@@ -207,8 +207,8 @@ enum atc_status {
  * @save_cfg: configuration register that is saved on suspend/resume cycle
  * @save_dscr: for cyclic operations, preserve next descriptor address in
  *             the cyclic list on suspend/resume cycle
+ * @dma_sconfig: configuration for slave transfers, passed via DMA_SLAVE_CONFIG
  * @lock: serializes enqueue/dequeue operations to descriptors lists
- * @completed_cookie: identifier for the most recently completed operation
  * @active_list: list of descriptors dmaengine is being running on
  * @queue: list of descriptors ready to be submitted to engine
  * @free_list: list of descriptors usable by the channel
@@ -223,11 +223,11 @@ struct at_dma_chan {
 	struct tasklet_struct	tasklet;
 	u32			save_cfg;
 	u32			save_dscr;
+	struct dma_slave_config dma_sconfig;
 
 	spinlock_t		lock;
 
 	/* these other elements are all protected by lock */
-	dma_cookie_t		completed_cookie;
 	struct list_head	active_list;
 	struct list_head	queue;
 	struct list_head	free_list;
@@ -245,6 +245,36 @@ static inline struct at_dma_chan *to_at_dma_chan(struct dma_chan *dchan)
 	return container_of(dchan, struct at_dma_chan, chan_common);
 }
 
+/*
+ * Fix sconfig's burst size according to at_hdmac. We need to convert them as:
+ * 1 -> 0, 4 -> 1, 8 -> 2, 16 -> 3, 32 -> 4, 64 -> 5, 128 -> 6, 256 -> 7.
+ *
+ * This can be done by finding most significant bit set.
+ */
+static inline void convert_burst(u32 *maxburst)
+{
+	if (*maxburst > 1)
+		*maxburst = fls(*maxburst) - 2;
+	else
+		*maxburst = 0;
+}
+
+/*
+ * Fix sconfig's bus width according to at_hdmac.
+ * 1 byte -> 0, 2 bytes -> 1, 4 bytes -> 2.
+ */
+static inline u8 convert_buswidth(enum dma_slave_buswidth addr_width)
+{
+	switch (addr_width) {
+	case DMA_SLAVE_BUSWIDTH_2_BYTES:
+		return 1;
+	case DMA_SLAVE_BUSWIDTH_4_BYTES:
+		return 2;
+	default:
+		/* For 1 byte width or fallback */
+		return 0;
+	}
+}
 
 /*--  Controller  ------------------------------------------------------*/
 
diff --git a/drivers/dma/coh901318.c b/drivers/dma/coh901318.c
index d65a718c0f9b..dc89455f5550 100644
--- a/drivers/dma/coh901318.c
+++ b/drivers/dma/coh901318.c
@@ -24,6 +24,7 @@
 #include <mach/coh901318.h>
 
 #include "coh901318_lli.h"
+#include "dmaengine.h"
 
 #define COHC_2_DEV(cohc) (&cohc->chan.dev->device)
 
@@ -59,7 +60,6 @@ struct coh901318_base {
 struct coh901318_chan {
 	spinlock_t lock;
 	int allocated;
-	int completed;
 	int id;
 	int stopped;
 
@@ -318,20 +318,6 @@ static int coh901318_prep_linked_list(struct coh901318_chan *cohc,
 
 	return 0;
 }
-static dma_cookie_t
-coh901318_assign_cookie(struct coh901318_chan *cohc,
-			struct coh901318_desc *cohd)
-{
-	dma_cookie_t cookie = cohc->chan.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	cohc->chan.cookie = cookie;
-	cohd->desc.cookie = cookie;
-
-	return cookie;
-}
 
 static struct coh901318_desc *
 coh901318_desc_get(struct coh901318_chan *cohc)
@@ -705,7 +691,7 @@ static void dma_tasklet(unsigned long data)
 	callback_param = cohd_fin->desc.callback_param;
 
 	/* sign this job as completed on the channel */
-	cohc->completed = cohd_fin->desc.cookie;
+	dma_cookie_complete(&cohd_fin->desc);
 
 	/* release the lli allocation and remove the descriptor */
 	coh901318_lli_free(&cohc->base->pool, &cohd_fin->lli);
@@ -929,7 +915,7 @@ static int coh901318_alloc_chan_resources(struct dma_chan *chan)
 	coh901318_config(cohc, NULL);
 
 	cohc->allocated = 1;
-	cohc->completed = chan->cookie = 1;
+	dma_cookie_init(chan);
 
 	spin_unlock_irqrestore(&cohc->lock, flags);
 
@@ -966,16 +952,16 @@ coh901318_tx_submit(struct dma_async_tx_descriptor *tx)
 						   desc);
 	struct coh901318_chan *cohc = to_coh901318_chan(tx->chan);
 	unsigned long flags;
+	dma_cookie_t cookie;
 
 	spin_lock_irqsave(&cohc->lock, flags);
-
-	tx->cookie = coh901318_assign_cookie(cohc, cohd);
+	cookie = dma_cookie_assign(tx);
 
 	coh901318_desc_queue(cohc, cohd);
 
 	spin_unlock_irqrestore(&cohc->lock, flags);
 
-	return tx->cookie;
+	return cookie;
 }
 
 static struct dma_async_tx_descriptor *
@@ -1035,7 +1021,7 @@ coh901318_prep_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 static struct dma_async_tx_descriptor *
 coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 			unsigned int sg_len, enum dma_transfer_direction direction,
-			unsigned long flags)
+			unsigned long flags, void *context)
 {
 	struct coh901318_chan *cohc = to_coh901318_chan(chan);
 	struct coh901318_lli *lli;
@@ -1165,17 +1151,12 @@ coh901318_tx_status(struct dma_chan *chan, dma_cookie_t cookie,
 		 struct dma_tx_state *txstate)
 {
 	struct coh901318_chan *cohc = to_coh901318_chan(chan);
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
-	int ret;
-
-	last_complete = cohc->completed;
-	last_used = chan->cookie;
+	enum dma_status ret;
 
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
+	ret = dma_cookie_status(chan, cookie, txstate);
+	/* FIXME: should be conditional on ret != DMA_SUCCESS? */
+	dma_set_residue(txstate, coh901318_get_bytes_left(chan));
 
-	dma_set_tx_state(txstate, last_complete, last_used,
-			 coh901318_get_bytes_left(chan));
 	if (ret == DMA_IN_PROGRESS && cohc->stopped)
 		ret = DMA_PAUSED;
 
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index a6c6051ec858..767bcc31b365 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -510,8 +510,8 @@ struct dma_chan *__dma_request_channel(dma_cap_mask_t *mask, dma_filter_fn fn, v
 					 dma_chan_name(chan));
 				list_del_rcu(&device->global_node);
 			} else if (err)
-				pr_debug("dmaengine: failed to get %s: (%d)\n",
-					 dma_chan_name(chan), err);
+				pr_debug("%s: failed to get %s: (%d)\n",
+					__func__, dma_chan_name(chan), err);
 			else
 				break;
 			if (--device->privatecnt == 0)
@@ -564,8 +564,8 @@ void dmaengine_get(void)
 				list_del_rcu(&device->global_node);
 				break;
 			} else if (err)
-				pr_err("dmaengine: failed to get %s: (%d)\n",
-				       dma_chan_name(chan), err);
+				pr_err("%s: failed to get %s: (%d)\n",
+					__func__, dma_chan_name(chan), err);
 		}
 	}
 
diff --git a/drivers/dma/dmaengine.h b/drivers/dma/dmaengine.h
new file mode 100644
index 000000000000..17f983a4e9ba
--- /dev/null
+++ b/drivers/dma/dmaengine.h
@@ -0,0 +1,89 @@
+/*
+ * The contents of this file are private to DMA engine drivers, and is not
+ * part of the API to be used by DMA engine users.
+ */
+#ifndef DMAENGINE_H
+#define DMAENGINE_H
+
+#include <linux/bug.h>
+#include <linux/dmaengine.h>
+
+/**
+ * dma_cookie_init - initialize the cookies for a DMA channel
+ * @chan: dma channel to initialize
+ */
+static inline void dma_cookie_init(struct dma_chan *chan)
+{
+	chan->cookie = DMA_MIN_COOKIE;
+	chan->completed_cookie = DMA_MIN_COOKIE;
+}
+
+/**
+ * dma_cookie_assign - assign a DMA engine cookie to the descriptor
+ * @tx: descriptor needing cookie
+ *
+ * Assign a unique non-zero per-channel cookie to the descriptor.
+ * Note: caller is expected to hold a lock to prevent concurrency.
+ */
+static inline dma_cookie_t dma_cookie_assign(struct dma_async_tx_descriptor *tx)
+{
+	struct dma_chan *chan = tx->chan;
+	dma_cookie_t cookie;
+
+	cookie = chan->cookie + 1;
+	if (cookie < DMA_MIN_COOKIE)
+		cookie = DMA_MIN_COOKIE;
+	tx->cookie = chan->cookie = cookie;
+
+	return cookie;
+}
+
+/**
+ * dma_cookie_complete - complete a descriptor
+ * @tx: descriptor to complete
+ *
+ * Mark this descriptor complete by updating the channels completed
+ * cookie marker.  Zero the descriptors cookie to prevent accidental
+ * repeated completions.
+ *
+ * Note: caller is expected to hold a lock to prevent concurrency.
+ */
+static inline void dma_cookie_complete(struct dma_async_tx_descriptor *tx)
+{
+	BUG_ON(tx->cookie < DMA_MIN_COOKIE);
+	tx->chan->completed_cookie = tx->cookie;
+	tx->cookie = 0;
+}
+
+/**
+ * dma_cookie_status - report cookie status
+ * @chan: dma channel
+ * @cookie: cookie we are interested in
+ * @state: dma_tx_state structure to return last/used cookies
+ *
+ * Report the status of the cookie, filling in the state structure if
+ * non-NULL.  No locking is required.
+ */
+static inline enum dma_status dma_cookie_status(struct dma_chan *chan,
+	dma_cookie_t cookie, struct dma_tx_state *state)
+{
+	dma_cookie_t used, complete;
+
+	used = chan->cookie;
+	complete = chan->completed_cookie;
+	barrier();
+	if (state) {
+		state->last = complete;
+		state->used = used;
+		state->residue = 0;
+	}
+	return dma_async_is_complete(cookie, complete, used);
+}
+
+static inline void dma_set_residue(struct dma_tx_state *state, u32 residue)
+{
+	if (state)
+		state->residue = residue;
+}
+
+#endif
diff --git a/drivers/dma/dw_dmac.c b/drivers/dma/dw_dmac.c
index 9b592b02b5f4..7439079f5eed 100644
--- a/drivers/dma/dw_dmac.c
+++ b/drivers/dma/dw_dmac.c
@@ -9,6 +9,7 @@
  * it under the terms of the GNU General Public License version 2 as
  * published by the Free Software Foundation.
  */
+#include <linux/bitops.h>
 #include <linux/clk.h>
 #include <linux/delay.h>
 #include <linux/dmaengine.h>
@@ -22,6 +23,7 @@
 #include <linux/slab.h>
 
 #include "dw_dmac_regs.h"
+#include "dmaengine.h"
 
 /*
  * This supports the Synopsys "DesignWare AHB Central DMA Controller",
@@ -33,19 +35,23 @@
  * which does not support descriptor writeback.
  */
 
-#define DWC_DEFAULT_CTLLO(private) ({				\
-		struct dw_dma_slave *__slave = (private);	\
-		int dms = __slave ? __slave->dst_master : 0;	\
-		int sms = __slave ? __slave->src_master : 1;	\
-		u8 smsize = __slave ? __slave->src_msize : DW_DMA_MSIZE_16; \
-		u8 dmsize = __slave ? __slave->dst_msize : DW_DM