Merge branch 'origin' into devel-stable

Conflicts:
	arch/arm/mach-mx2/devices.c
	arch/arm/mach-mx2/devices.h
	sound/soc/pxa/pxa-ssp.c

19 files changed, 2291 insertions, 1057 deletions

diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index e02d74b1e89..c27f80e5d53 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -13,6 +13,22 @@ menuconfig DMADEVICES
 	  DMA Device drivers supported by the configured arch, it may
 	  be empty in some cases.
 
+config DMADEVICES_DEBUG
+        bool "DMA Engine debugging"
+        depends on DMADEVICES != n
+        help
+          This is an option for use by developers; most people should
+          say N here.  This enables DMA engine core and driver debugging.
+
+config DMADEVICES_VDEBUG
+        bool "DMA Engine verbose debugging"
+        depends on DMADEVICES_DEBUG != n
+        help
+          This is an option for use by developers; most people should
+          say N here.  This enables deeper (more verbose) debugging of
+          the DMA engine core and drivers.
+
+
 if DMADEVICES
 
 comment "DMA Devices"
@@ -69,6 +85,13 @@ config FSL_DMA
 	  The Elo is the DMA controller on some 82xx and 83xx parts, and the
 	  Elo Plus is the DMA controller on 85xx and 86xx parts.
 
+config MPC512X_DMA
+	tristate "Freescale MPC512x built-in DMA engine support"
+	depends on PPC_MPC512x
+	select DMA_ENGINE
+	---help---
+	  Enable support for the Freescale MPC512x built-in DMA engine.
+
 config MV_XOR
 	bool "Marvell XOR engine support"
 	depends on PLAT_ORION
diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile
index 807053d4823..22bba3d5e2b 100644
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@@ -1,9 +1,17 @@
+ifeq ($(CONFIG_DMADEVICES_DEBUG),y)
+	EXTRA_CFLAGS	+= -DDEBUG
+endif
+ifeq ($(CONFIG_DMADEVICES_VDEBUG),y)
+	EXTRA_CFLAGS	+= -DVERBOSE_DEBUG
+endif
+
 obj-$(CONFIG_DMA_ENGINE) += dmaengine.o
 obj-$(CONFIG_NET_DMA) += iovlock.o
 obj-$(CONFIG_DMATEST) += dmatest.o
 obj-$(CONFIG_INTEL_IOATDMA) += ioat/
 obj-$(CONFIG_INTEL_IOP_ADMA) += iop-adma.o
 obj-$(CONFIG_FSL_DMA) += fsldma.o
+obj-$(CONFIG_MPC512X_DMA) += mpc512x_dma.o
 obj-$(CONFIG_MV_XOR) += mv_xor.o
 obj-$(CONFIG_DW_DMAC) += dw_dmac.o
 obj-$(CONFIG_AT_HDMAC) += at_hdmac.o
diff --git a/drivers/dma/coh901318.c b/drivers/dma/coh901318.c
index 64a937262a4..1656fdcdb6c 100644
--- a/drivers/dma/coh901318.c
+++ b/drivers/dma/coh901318.c
@@ -39,7 +39,6 @@ struct coh901318_desc {
 	unsigned int sg_len;
 	struct coh901318_lli *data;
 	enum dma_data_direction dir;
-	int pending_irqs;
 	unsigned long flags;
 };
 
@@ -72,7 +71,6 @@ struct coh901318_chan {
 
 	unsigned long nbr_active_done;
 	unsigned long busy;
-	int pending_irqs;
 
 	struct coh901318_base *base;
 };
@@ -80,18 +78,16 @@ struct coh901318_chan {
 static void coh901318_list_print(struct coh901318_chan *cohc,
 				 struct coh901318_lli *lli)
 {
-	struct coh901318_lli *l;
-	dma_addr_t addr =  virt_to_phys(lli);
+	struct coh901318_lli *l = lli;
 	int i = 0;
 
-	while (addr) {
-		l = phys_to_virt(addr);
+	while (l) {
 		dev_vdbg(COHC_2_DEV(cohc), "i %d, lli %p, ctrl 0x%x, src 0x%x"
-			 ", dst 0x%x, link 0x%x link_virt 0x%p\n",
+			 ", dst 0x%x, link 0x%x virt_link_addr 0x%p\n",
 			 i, l, l->control, l->src_addr, l->dst_addr,
-			 l->link_addr, phys_to_virt(l->link_addr));
+			 l->link_addr, l->virt_link_addr);
 		i++;
-		addr = l->link_addr;
+		l = l->virt_link_addr;
 	}
 }
 
@@ -125,7 +121,7 @@ static int coh901318_debugfs_read(struct file *file, char __user *buf,
 		goto err_kmalloc;
 	tmp = dev_buf;
 
-	tmp += sprintf(tmp, "DMA -- enable dma channels\n");
+	tmp += sprintf(tmp, "DMA -- enabled dma channels\n");
 
 	for (i = 0; i < debugfs_dma_base->platform->max_channels; i++)
 		if (started_channels & (1 << i))
@@ -337,16 +333,22 @@ coh901318_desc_get(struct coh901318_chan *cohc)
 		 * TODO: alloc a pile of descs instead of just one,
 		 * avoid many small allocations.
 		 */
-		desc = kmalloc(sizeof(struct coh901318_desc), GFP_NOWAIT);
+		desc = kzalloc(sizeof(struct coh901318_desc), GFP_NOWAIT);
 		if (desc == NULL)
 			goto out;
 		INIT_LIST_HEAD(&desc->node);
+		dma_async_tx_descriptor_init(&desc->desc, &cohc->chan);
 	} else {
 		/* Reuse an old desc. */
 		desc = list_first_entry(&cohc->free,
 					struct coh901318_desc,
 					node);
 		list_del(&desc->node);
+		/* Initialize it a bit so it's not insane */
+		desc->sg = NULL;
+		desc->sg_len = 0;
+		desc->desc.callback = NULL;
+		desc->desc.callback_param = NULL;
 	}
 
  out:
@@ -364,10 +366,6 @@ static void
 coh901318_desc_submit(struct coh901318_chan *cohc, struct coh901318_desc *desc)
 {
 	list_add_tail(&desc->node, &cohc->active);
-
-	BUG_ON(cohc->pending_irqs != 0);
-
-	cohc->pending_irqs = desc->pending_irqs;
 }
 
 static struct coh901318_desc *
@@ -592,6 +590,10 @@ static struct coh901318_desc *coh901318_queue_start(struct coh901318_chan *cohc)
 	return cohd_que;
 }
 
+/*
+ * This tasklet is called from the interrupt handler to
+ * handle each descriptor (DMA job) that is sent to a channel.
+ */
 static void dma_tasklet(unsigned long data)
 {
 	struct coh901318_chan *cohc = (struct coh901318_chan *) data;
@@ -600,55 +602,58 @@ static void dma_tasklet(unsigned long data)
 	dma_async_tx_callback callback;
 	void *callback_param;
 
+	dev_vdbg(COHC_2_DEV(cohc), "[%s] chan_id %d"
+		 " nbr_active_done %ld\n", __func__,
+		 cohc->id, cohc->nbr_active_done);
+
 	spin_lock_irqsave(&cohc->lock, flags);
 
-	/* get first active entry from list */
+	/* get first active descriptor entry from list */
 	cohd_fin = coh901318_first_active_get(cohc);
 
-	BUG_ON(cohd_fin->pending_irqs == 0);
-
 	if (cohd_fin == NULL)
 		goto err;
 
-	cohd_fin->pending_irqs--;
-	cohc->completed = cohd_fin->desc.cookie;
+	/* locate callback to client */
+	callback = cohd_fin->desc.callback;
+	callback_param = cohd_fin->desc.callback_param;
 
-	if (cohc->nbr_active_done == 0)
-		return;
+	/* sign this job as completed on the channel */
+	cohc->completed = cohd_fin->desc.cookie;
 
-	if (!cohd_fin->pending_irqs) {
-		/* release the lli allocation*/
-		coh901318_lli_free(&cohc->base->pool, &cohd_fin->data);
-	}
+	/* release the lli allocation and remove the descriptor */
+	coh901318_lli_free(&cohc->base->pool, &cohd_fin->data);
 
-	dev_vdbg(COHC_2_DEV(cohc), "[%s] chan_id %d pending_irqs %d"
-		 " nbr_active_done %ld\n", __func__,
-		 cohc->id, cohc->pending_irqs, cohc->nbr_active_done);
+	/* return desc to free-list */
+	coh901318_desc_remove(cohd_fin);
+	coh901318_desc_free(cohc, cohd_fin);
 
-	/* callback to client */
-	callback = cohd_fin->desc.callback;
-	callback_param = cohd_fin->desc.callback_param;
-
-	if (!cohd_fin->pending_irqs) {
-		coh901318_desc_remove(cohd_fin);
+	spin_unlock_irqrestore(&cohc->lock, flags);
 
-		/* return desc to free-list */
-		coh901318_desc_free(cohc, cohd_fin);
-	}
+	/* Call the callback when we're done */
+	if (callback)
+		callback(callback_param);
 
-	if (cohc->nbr_active_done)
-		cohc->nbr_active_done--;
+	spin_lock_irqsave(&cohc->lock, flags);
 
+	/*
+	 * If another interrupt fired while the tasklet was scheduling,
+	 * we don't get called twice, so we have this number of active
+	 * counter that keep track of the number of IRQs expected to
+	 * be handled for this channel. If there happen to be more than
+	 * one IRQ to be ack:ed, we simply schedule this tasklet again.
+	 */
+	cohc->nbr_active_done--;
 	if (cohc->nbr_active_done) {
+		dev_dbg(COHC_2_DEV(cohc), "scheduling tasklet again, new IRQs "
+			"came in while we were scheduling this tasklet\n");
 		if (cohc_chan_conf(cohc)->priority_high)
 			tasklet_hi_schedule(&cohc->tasklet);
 		else
 			tasklet_schedule(&cohc->tasklet);
 	}
-	spin_unlock_irqrestore(&cohc->lock, flags);
 
-	if (callback)
-		callback(callback_param);
+	spin_unlock_irqrestore(&cohc->lock, flags);
 
 	return;
 
@@ -667,16 +672,17 @@ static void dma_tc_handle(struct coh901318_chan *cohc)
 	if (!cohc->allocated)
 		return;
 
-	BUG_ON(cohc->pending_irqs == 0);
+	spin_lock(&cohc->lock);
 
-	cohc->pending_irqs--;
 	cohc->nbr_active_done++;
 
-	if (cohc->pending_irqs == 0 && coh901318_queue_start(cohc) == NULL)
+	if (coh901318_queue_start(cohc) == NULL)
 		cohc->busy = 0;
 
 	BUG_ON(list_empty(&cohc->active));
 
+	spin_unlock(&cohc->lock);
+
 	if (cohc_chan_conf(cohc)->priority_high)
 		tasklet_hi_schedule(&cohc->tasklet);
 	else
@@ -870,6 +876,7 @@ coh901318_prep_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 	struct coh901318_chan *cohc = to_coh901318_chan(chan);
 	int lli_len;
 	u32 ctrl_last = cohc_chan_param(cohc)->ctrl_lli_last;
+	int ret;
 
 	spin_lock_irqsave(&cohc->lock, flg);
 
@@ -890,22 +897,19 @@ coh901318_prep_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 	if (data == NULL)
 		goto err;
 
-	cohd = coh901318_desc_get(cohc);
-	cohd->sg = NULL;
-	cohd->sg_len = 0;
-	cohd->data = data;
-
-	cohd->pending_irqs =
-		coh901318_lli_fill_memcpy(
-				&cohc->base->pool, data, src, size, dest,
-				cohc_chan_param(cohc)->ctrl_lli_chained,
-				ctrl_last);
-	cohd->flags = flags;
+	ret = coh901318_lli_fill_memcpy(
+		&cohc->base->pool, data, src, size, dest,
+		cohc_chan_param(cohc)->ctrl_lli_chained,
+		ctrl_last);
+	if (ret)
+		goto err;
 
 	COH_DBG(coh901318_list_print(cohc, data));
 
-	dma_async_tx_descriptor_init(&cohd->desc, chan);
-
+	/* Pick a descriptor to handle this transfer */
+	cohd = coh901318_desc_get(cohc);
+	cohd->data = data;
+	cohd->flags = flags;
 	cohd->desc.tx_submit = coh901318_tx_submit;
 
 	spin_unlock_irqrestore(&cohc->lock, flg);
@@ -924,6 +928,7 @@ coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 	struct coh901318_chan *cohc = to_coh901318_chan(chan);
 	struct coh901318_lli *data;
 	struct coh901318_desc *cohd;
+	const struct coh901318_params *params;
 	struct scatterlist *sg;
 	int len = 0;
 	int size;
@@ -931,7 +936,9 @@ coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 	u32 ctrl_chained = cohc_chan_param(cohc)->ctrl_lli_chained;
 	u32 ctrl = cohc_chan_param(cohc)->ctrl_lli;
 	u32 ctrl_last = cohc_chan_param(cohc)->ctrl_lli_last;
+	u32 config;
 	unsigned long flg;
+	int ret;
 
 	if (!sgl)
 		goto out;
@@ -947,15 +954,14 @@ coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		/* Trigger interrupt after last lli */
 		ctrl_last |= COH901318_CX_CTRL_TC_IRQ_ENABLE;
 
-	cohd = coh901318_desc_get(cohc);
-	cohd->sg = NULL;
-	cohd->sg_len = 0;
-	cohd->dir = direction;
+	params = cohc_chan_param(cohc);
+	config = params->config;
 
 	if (direction == DMA_TO_DEVICE) {
 		u32 tx_flags = COH901318_CX_CTRL_PRDD_SOURCE |
 			COH901318_CX_CTRL_SRC_ADDR_INC_ENABLE;
 
+		config |= COH901318_CX_CFG_RM_MEMORY_TO_PRIMARY;
 		ctrl_chained |= tx_flags;
 		ctrl_last |= tx_flags;
 		ctrl |= tx_flags;
@@ -963,16 +969,14 @@ coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		u32 rx_flags = COH901318_CX_CTRL_PRDD_DEST |
 			COH901318_CX_CTRL_DST_ADDR_INC_ENABLE;
 
+		config |= COH901318_CX_CFG_RM_PRIMARY_TO_MEMORY;
 		ctrl_chained |= rx_flags;
 		ctrl_last |= rx_flags;
 		ctrl |= rx_flags;
 	} else
 		goto err_direction;
 
-	dma_async_tx_descriptor_init(&cohd->desc, chan);
-
-	cohd->desc.tx_submit = coh901318_tx_submit;
-
+	coh901318_set_conf(cohc, config);
 
 	/* The dma only supports transmitting packages up to
 	 * MAX_DMA_PACKET_SIZE. Calculate to total number of
@@ -994,32 +998,37 @@ coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		len += factor;
 	}
 
+	pr_debug("Allocate %d lli:s for this transfer\n", len);
 	data = coh901318_lli_alloc(&cohc->base->pool, len);
 
 	if (data == NULL)
 		goto err_dma_alloc;
 
 	/* initiate allocated data list */
-	cohd->pending_irqs =
-		coh901318_lli_fill_sg(&cohc->base->pool, data, sgl, sg_len,
-				      cohc_dev_addr(cohc),
-				      ctrl_chained,
-				      ctrl,
-				      ctrl_last,
-				      direction, COH901318_CX_CTRL_TC_IRQ_ENABLE);
-	cohd->data = data;
-
-	cohd->flags = flags;
+	ret = coh901318_lli_fill_sg(&cohc->base->pool, data, sgl, sg_len,
+				    cohc_dev_addr(cohc),
+				    ctrl_chained,
+				    ctrl,
+				    ctrl_last,
+				    direction, COH901318_CX_CTRL_TC_IRQ_ENABLE);
+	if (ret)
+		goto err_lli_fill;
 
 	COH_DBG(coh901318_list_print(cohc, data));
 
+	/* Pick a descriptor to handle this transfer */
+	cohd = coh901318_desc_get(cohc);
+	cohd->dir = direction;
+	cohd->flags = flags;
+	cohd->desc.tx_submit = coh901318_tx_submit;
+	cohd->data = data;
+
 	spin_unlock_irqrestore(&cohc->lock, flg);
 
 	return &cohd->desc;
+ err_lli_fill:
  err_dma_alloc:
  err_direction:
-	coh901318_desc_remove(cohd);
-	coh901318_desc_free(cohc, cohd);
 	spin_unlock_irqrestore(&cohc->lock, flg);
  out:
 	return NULL;
@@ -1092,9 +1101,8 @@ coh901318_terminate_all(struct dma_chan *chan)
 		/* release the lli allocation*/
 		coh901318_lli_free(&cohc->base->pool, &cohd->data);
 
-		coh901318_desc_remove(cohd);
-
 		/* return desc to free-list */
+		coh901318_desc_remove(cohd);
 		coh901318_desc_free(cohc, cohd);
 	}
 
@@ -1102,16 +1110,14 @@ coh901318_terminate_all(struct dma_chan *chan)
 		/* release the lli allocation*/
 		coh901318_lli_free(&cohc->base->pool, &cohd->data);
 
-		coh901318_desc_remove(cohd);
-
 		/* return desc to free-list */
+		coh901318_desc_remove(cohd);
 		coh901318_desc_free(cohc, cohd);
 	}
 
 
 	cohc->nbr_active_done = 0;
 	cohc->busy = 0;
-	cohc->pending_irqs = 0;
 
 	spin_unlock_irqrestore(&cohc->lock, flags);
 }
@@ -1138,7 +1144,6 @@ void coh901318_base_init(struct dma_device *dma, const int *pick_chans,
 
 			spin_lock_init(&cohc->lock);
 
-			cohc->pending_irqs = 0;
 			cohc->nbr_active_done = 0;
 			cohc->busy = 0;
 			INIT_LIST_HEAD(&cohc->free);
@@ -1254,12 +1259,17 @@ static int __init coh901318_probe(struct platform_device *pdev)
 	base->dma_memcpy.device_issue_pending = coh901318_issue_pending;
 	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,
+	 * i.e. 2^2
+	 */
+	base->dma_memcpy.copy_align = 2;
 	err = dma_async_device_register(&base->dma_memcpy);
 
 	if (err)
 		goto err_register_memcpy;
 
-	dev_dbg(&pdev->dev, "Initialized COH901318 DMA on virtual base 0x%08x\n",
+	dev_info(&pdev->dev, "Initialized COH901318 DMA on virtual base 0x%08x\n",
 		(u32) base->virtbase);
 
 	return err;
diff --git a/drivers/dma/coh901318_lli.c b/drivers/dma/coh901318_lli.c
index f5120f238a4..71d58c1a1e8 100644
--- a/drivers/dma/coh901318_lli.c
+++ b/drivers/dma/coh901318_lli.c
@@ -74,6 +74,8 @@ coh901318_lli_alloc(struct coh901318_pool *pool, unsigned int len)
 
 	lli = head;
 	lli->phy_this = phy;
+	lli->link_addr = 0x00000000;
+	lli->virt_link_addr = 0x00000000U;
 
 	for (i = 1; i < len; i++) {
 		lli_prev = lli;
@@ -85,13 +87,13 @@ coh901318_lli_alloc(struct coh901318_pool *pool, unsigned int len)
 
 		DEBUGFS_POOL_COUNTER_ADD(pool, 1);
 		lli->phy_this = phy;
+		lli->link_addr = 0x00000000;
+		lli->virt_link_addr = 0x00000000U;
 
 		lli_prev->link_addr = phy;
 		lli_prev->virt_link_addr = lli;
 	}
 
-	lli->link_addr = 0x00000000U;
-
 	spin_unlock(&pool->lock);
 
 	return head;
@@ -166,8 +168,7 @@ coh901318_lli_fill_memcpy(struct coh901318_pool *pool,
 	lli->src_addr = src;
 	lli->dst_addr = dst;
 
-	/* One irq per single transfer */
-	return 1;
+	return 0;
 }
 
 int
@@ -223,8 +224,7 @@ coh901318_lli_fill_single(struct coh901318_pool *pool,
 	lli->src_addr = src;
 	lli->dst_addr = dst;
 
-	/* One irq per single transfer */
-	return 1;
+	return 0;
 }
 
 int
@@ -240,7 +240,6 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool,
 	u32 ctrl_sg;
 	dma_addr_t src = 0;
 	dma_addr_t dst = 0;
-	int nbr_of_irq = 0;
 	u32 bytes_to_transfer;
 	u32 elem_size;
 
@@ -269,15 +268,12 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool,
 			ctrl_sg = ctrl ? ctrl : ctrl_last;
 
 
-		if ((ctrl_sg & ctrl_irq_mask))
-			nbr_of_irq++;
-
 		if (dir == DMA_TO_DEVICE)
 			/* increment source address */
-			src = sg_dma_address(sg);
+			src = sg_phys(sg);
 		else
 			/* increment destination address */
-			dst =  sg_dma_address(sg);
+			dst =  sg_phys(sg);
 
 		bytes_to_transfer = sg_dma_len(sg);
 
@@ -310,8 +306,7 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool,
 	}
 	spin_unlock(&pool->lock);
 
-	/* There can be many IRQs per sg transfer */
-	return nbr_of_irq;
+	return 0;
  err:
 	spin_unlock(&pool->lock);
 	return -EINVAL;
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index e7a3230fb7d..87399cafce3 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -284,7 +284,7 @@ struct dma_chan_tbl_ent {
 /**
  * channel_table - percpu lookup table for memory-to-memory offload providers
  */
-static struct dma_chan_tbl_ent *channel_table[DMA_TX_TYPE_END];
+static struct dma_chan_tbl_ent __percpu *channel_table[DMA_TX_TYPE_END];
 
 static int __init dma_channel_table_init(void)
 {
diff --git a/drivers/dma/dmatest.c b/drivers/dma/dmatest.c
index 948d563941c..6fa55fe3dd2 100644
--- a/drivers/dma/dmatest.c
+++ b/drivers/dma/dmatest.c
@@ -237,7 +237,7 @@ static int dmatest_func(void *data)
 	dma_cookie_t		cookie;
 	enum dma_status		status;
 	enum dma_ctrl_flags 	flags;
-	u8			pq_coefs[pq_sources];
+	u8			pq_coefs[pq_sources + 1];
 	int			ret;
 	int			src_cnt;
 	int			dst_cnt;
@@ -257,7 +257,7 @@ static int dmatest_func(void *data)
 	} else if (thread->type == DMA_PQ) {
 		src_cnt = pq_sources | 1; /* force odd to ensure dst = src */
 		dst_cnt = 2;
-		for (i = 0; i < pq_sources; i++)
+		for (i = 0; i < src_cnt; i++)
 			pq_coefs[i] = 1;
 	} else
 		goto err_srcs;
@@ -347,7 +347,7 @@ static int dmatest_func(void *data)
 		else if (thread->type == DMA_XOR)
 			tx = dev->device_prep_dma_xor(chan,
 						      dma_dsts[0] + dst_off,
-						      dma_srcs, xor_sources,
+						      dma_srcs, src_cnt,
 						      len, flags);
 		else if (thread->type == DMA_PQ) {
 			dma_addr_t dma_pq[dst_cnt];
@@ -355,7 +355,7 @@ static int dmatest_func(void *data)
 			for (i = 0; i < dst_cnt; i++)
 				dma_pq[i] = dma_dsts[i] + dst_off;
 			tx = dev->device_prep_dma_pq(chan, dma_pq, dma_srcs,
-						     pq_sources, pq_coefs,
+						     src_cnt, pq_coefs,
 						     len, flags);
 		}
 
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index 296f9e747fa..bbb4be5a3ff 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -37,19 +37,19 @@
 #include <asm/fsldma.h>
 #include "fsldma.h"
 
-static void dma_init(struct fsl_dma_chan *fsl_chan)
+static void dma_init(struct fsldma_chan *chan)
 {
 	/* Reset the channel */
-	DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr, 0, 32);
+	DMA_OUT(chan, &chan->regs->mr, 0, 32);
 
-	switch (fsl_chan->feature & FSL_DMA_IP_MASK) {
+	switch (chan->feature & FSL_DMA_IP_MASK) {
 	case FSL_DMA_IP_85XX:
 		/* Set the channel to below modes:
 		 * EIE - Error interrupt enable
 		 * EOSIE - End of segments interrupt enable (basic mode)
 		 * EOLNIE - End of links interrupt enable
 		 */
-		DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr, FSL_DMA_MR_EIE
+		DMA_OUT(chan, &chan->regs->mr, FSL_DMA_MR_EIE
 				| FSL_DMA_MR_EOLNIE | FSL_DMA_MR_EOSIE, 32);
 		break;
 	case FSL_DMA_IP_83XX:
@@ -57,170 +57,146 @@ static void dma_init(struct fsl_dma_chan *fsl_chan)
 		 * EOTIE - End-of-transfer interrupt enable
 		 * PRC_RM - PCI read multiple
 		 */
-		DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr, FSL_DMA_MR_EOTIE
+		DMA_OUT(chan, &chan->regs->mr, FSL_DMA_MR_EOTIE
 				| FSL_DMA_MR_PRC_RM, 32);
 		break;
 	}
-
 }
 
-static void set_sr(struct fsl_dma_chan *fsl_chan, u32 val)
+static void set_sr(struct fsldma_chan *chan, u32 val)
 {
-	DMA_OUT(fsl_chan, &fsl_chan->reg_base->sr, val, 32);
+	DMA_OUT(chan, &chan->regs->sr, val, 32);
 }
 
-static u32 get_sr(struct fsl_dma_chan *fsl_chan)
+static u32 get_sr(struct fsldma_chan *chan)
 {
-	return DMA_IN(fsl_chan, &fsl_chan->reg_base->sr, 32);
+	return DMA_IN(chan, &chan->regs->sr, 32);
 }
 
-static void set_desc_cnt(struct fsl_dma_chan *fsl_chan,
+static void set_desc_cnt(struct fsldma_chan *chan,
 				struct fsl_dma_ld_hw *hw, u32 count)
 {
-	hw->count = CPU_TO_DMA(fsl_chan, count, 32);
+	hw->count = CPU_TO_DMA(chan, count, 32);
 }
 
-static void set_desc_src(struct fsl_dma_chan *fsl_chan,
+static void set_desc_src(struct fsldma_chan *chan,
 				struct fsl_dma_ld_hw *hw, dma_addr_t src)
 {
 	u64 snoop_bits;
 
-	snoop_bits = ((fsl_chan->feature & FSL_DMA_IP_MASK) == FSL_DMA_IP_85XX)
+	snoop_bits = ((chan->feature & FSL_DMA_IP_MASK) == FSL_DMA_IP_85XX)
 		? ((u64)FSL_DMA_SATR_SREADTYPE_SNOOP_READ << 32) : 0;
-	hw->src_addr = CPU_TO_DMA(fsl_chan, snoop_bits | src, 64);
+	hw->src_addr = CPU_TO_DMA(chan, snoop_bits | src, 64);
 }
 
-static void set_desc_dest(struct fsl_dma_chan *fsl_chan,
-				struct fsl_dma_ld_hw *hw, dma_addr_t dest)
+static void set_desc_dst(struct fsldma_chan *chan,
+				struct fsl_dma_ld_hw *hw, dma_addr_t dst)
 {
 	u64 snoop_bits;
 
-	snoop_bits = ((fsl_chan->feature & FSL_DMA_IP_MASK) == FSL_DMA_IP_85XX)
+	snoop_bits = ((chan->feature & FSL_DMA_IP_MASK) == FSL_DMA_IP_85XX)
 		? ((u64)FSL_DMA_DATR_DWRITETYPE_SNOOP_WRITE << 32) : 0;
-	hw->dst_addr = CPU_TO_DMA(fsl_chan, snoop_bits | dest, 64);
+	hw->dst_addr = CPU_TO_DMA(chan, snoop_bits | dst, 64);
 }
 
-static void set_desc_next(struct fsl_dma_chan *fsl_chan,
+static void set_desc_next(struct fsldma_chan *chan,
 				struct fsl_dma_ld_hw *hw, dma_addr_t next)
 {
 	u64 snoop_bits;
 
-	snoop_bits = ((fsl_chan->feature & FSL_DMA_IP_MASK) == FSL_DMA_IP_83XX)
+	snoop_bits = ((chan->feature & FSL_DMA_IP_MASK) == FSL_DMA_IP_83XX)
 		? FSL_DMA_SNEN : 0;
-	hw->next_ln_addr = CPU_TO_DMA(fsl_chan, snoop_bits | next, 64);
-}
-
-static void set_cdar(struct fsl_dma_chan *fsl_chan, dma_addr_t addr)
-{
-	DMA_OUT(fsl_chan, &fsl_chan->reg_base->cdar, addr | FSL_DMA_SNEN, 64);
+	hw->next_ln_addr = CPU_TO_DMA(chan, snoop_bits | next, 64);
 }
 
-static dma_addr_t get_cdar(struct fsl_dma_chan *fsl_chan)
+static void set_cdar(struct fsldma_chan *chan, dma_addr_t addr)
 {
-	return DMA_IN(fsl_chan, &fsl_chan->reg_base->cdar, 64) & ~FSL_DMA_SNEN;
+	DMA_OUT(chan, &chan->regs->cdar, addr | FSL_DMA_SNEN, 64);
 }
 
-static void set_ndar(struct fsl_dma_chan *fsl_chan, dma_addr_t addr)
+static dma_addr_t get_cdar(struct fsldma_chan *chan)
 {
-	DMA_OUT(fsl_chan, &fsl_chan->reg_base->ndar, addr, 64);
+	return DMA_IN(chan, &chan->regs->cdar, 64) & ~FSL_DMA_SNEN;
 }
 
-static dma_addr_t get_ndar(struct fsl_dma_chan *fsl_chan)
+static dma_addr_t get_ndar(struct fsldma_chan *chan)
 {
-	return DMA_IN(fsl_chan, &fsl_chan->reg_base->ndar, 64);
+	return DMA_IN(chan, &chan->regs->ndar, 64);
 }
 
-static u32 get_bcr(struct fsl_dma_chan *fsl_chan)
+static u32 get_bcr(struct fsldma_chan *chan)
 {
-	return DMA_IN(fsl_chan, &fsl_chan->reg_base->bcr, 32);
+	return DMA_IN(chan, &chan->regs->bcr, 32);
 }
 
-static int dma_is_idle(struct fsl_dma_chan *fsl_chan)
+static int dma_is_idle(struct fsldma_chan *chan)
 {
-	u32 sr = get_sr(fsl_chan);
+	u32 sr = get_sr(chan);
 	return (!(sr & FSL_DMA_SR_CB)) || (sr & FSL_DMA_SR_CH);
 }
 
-static void dma_start(struct fsl_dma_chan *fsl_chan)
+static void dma_start(struct fsldma_chan *chan)
 {
-	u32 mr_set = 0;
-
-	if (fsl_chan->feature & FSL_DMA_CHAN_PAUSE_EXT) {
-		DMA_OUT(fsl_chan, &fsl_chan->reg_base->bcr, 0, 32);
-		mr_set |= FSL_DMA_MR_EMP_EN;
-	} else if ((fsl_chan->feature & FSL_DMA_IP_MASK) == FSL_DMA_IP_85XX) {
-		DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr,
-			DMA_IN(fsl_chan, &fsl_chan->reg_base->mr, 32)
-				& ~FSL_DMA_MR_EMP_EN, 32);
+	u32 mode;
+
+	mode = DMA_IN(chan, &chan->regs->mr, 32);
+
+	if ((chan->feature & FSL_DMA_IP_MASK) == FSL_DMA_IP_85XX) {
+		if (chan->feature & FSL_DMA_CHAN_PAUSE_EXT) {
+			DMA_OUT(chan, &chan->regs->bcr, 0, 32);
+			mode |= FSL_DMA_MR_EMP_EN;
+		} else {
+			mode &= ~FSL_DMA_MR_EMP_EN;
+		}
 	}
 
-	if (fsl_chan->feature & FSL_DMA_CHAN_START_EXT)
-		mr_set |= FSL_DMA_MR_EMS_EN;
+	if (chan->feature & FSL_DMA_CHAN_START_EXT)
+		mode |= FSL_DMA_MR_EMS_EN;
 	else
-		mr_set |= FSL_DMA_MR_CS;
+		mode |= FSL_DMA_MR_CS;
 
-	DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr,
-			DMA_IN(fsl_chan, &fsl_chan->reg_base->mr, 32)
-			| mr_set, 32);
+	DMA_OUT(chan, &chan->regs->mr, mode, 32);
 }
 
-static void dma_halt(struct fsl_dma_chan *fsl_chan)
+static void dma_halt(struct fsldma_chan *chan)
 {
+	u32 mode;
 	int i;
 
-	DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr,
-		DMA_IN(fsl_chan, &fsl_chan->reg_base->mr, 32) | FSL_DMA_MR_CA,
-		32);
-	DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr,
-		DMA_IN(fsl_chan, &fsl_chan->reg_base->mr, 32) & ~(FSL_DMA_MR_CS
-		| FSL_DMA_MR_EMS_EN | FSL_DMA_MR_CA), 32);
+	mode = DMA_IN(chan, &chan->regs->mr, 32);
+	mode |= FSL_DMA_MR_CA;
+	DMA_OUT(chan, &chan->regs->mr, mode, 32);
+
+	mode &= ~(FSL_DMA_MR_CS | FSL_DMA_MR_EMS_EN | FSL_DMA_MR_CA);
+	DMA_OUT(chan, &chan->regs->mr, mode, 32);
 
 	for (i = 0; i < 100; i++) {
-		if (dma_is_idle(fsl_chan))
-			break;
+		if (dma_is_idle(chan))
+			return;
+
 		udelay(10);
 	}
-	if (i >= 100 && !dma_is_idle(fsl_chan))
-		dev_err(fsl_chan->dev, "DMA halt timeout!\n");
+
+	if (!dma_is_idle(chan))
+		dev_err(chan->dev, "DMA halt timeout!\n");
 }
 
-static void set_ld_eol(struct fsl_dma_chan *fsl_chan,
+static void set_ld_eol(struct fsldma_chan *chan,
 			struct fsl_desc_sw *desc)
 {
 	u64 snoop_bits;
 
-	snoop_bits = ((fsl_chan->feature & FSL_DMA_IP_MASK) == FSL_DMA_IP_83XX)
+	snoop_bits = ((chan->feature & FSL_DMA_IP_MASK) == FSL_DMA_IP_83XX)
 		? FSL_DMA_SNEN : 0;
 
-	desc->hw.next_ln_addr = CPU_TO_DMA(fsl_chan,
-		DMA_TO_CPU(fsl_chan, desc->hw.next_ln_addr, 64) | FSL_DMA_EOL
+	desc->hw.next_ln_addr = CPU_TO_DMA(chan,
+		DMA_TO_CPU(chan, desc->hw.next_ln_addr, 64) | FSL_DMA_EOL
 			| snoop_bits, 64);
 }
 
-static void append_ld_queue(struct fsl_dma_chan *fsl_chan,
-		struct fsl_desc_sw *new_desc)
-{
-	struct fsl_desc_sw *queue_tail = to_fsl_desc(fsl_chan->ld_queue.prev);
-
-	if (list_empty(&fsl_chan->ld_queue))
-		return;
-
-	/* Link to the new descriptor physical address and
-	 * Enable End-of-segment interrupt for
-	 * the last link descriptor.
-	 * (the previous node's next link descriptor)
-	 *
-	 * For FSL_DMA_IP_83xx, the snoop enable bit need be set.
-	 */
-	queue_tail->hw.next_ln_addr = CPU_TO_DMA(fsl_chan,
-			new_desc->async_tx.phys | FSL_DMA_EOSIE |
-			(((fsl_chan->feature & FSL_DMA_IP_MASK)
-				== FSL_DMA_IP_83XX) ? FSL_DMA_SNEN : 0), 64);
-}
-
 /**
  * fsl_chan_set_src_loop_size - Set source address hold transfer size
- * @fsl_chan : Freescale DMA channel
+ * @chan : Freescale DMA channel
  * @size     : Address loop size, 0 for disable loop
  *
  * The set source address hold transfer size. The source
@@ -229,29 +205,30 @@ static void append_ld_queue(struct fsl_dma_chan *fsl_chan,
  * read data from SA, SA + 1, SA + 2, SA + 3, then loop back to SA,
  * SA + 1 ... and so on.
  */
-static void fsl_chan_set_src_loop_size(struct fsl_dma_chan *fsl_chan, int size)
+static void fsl_chan_set_src_loop_size(struct fsldma_chan *chan, int size)
 {
+	u32 mode;
+
+	mode = DMA_IN(chan, &chan->regs->mr, 32);
+
 	switch (size) {
 	case 0:
-		DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr,
-			DMA_IN(fsl_chan, &fsl_chan->reg_base->mr, 32) &
-			(~FSL_DMA_MR_SAHE), 32);
+		mode &= ~FSL_DMA_MR_SAHE;
 		break;
 	case 1:
 	case 2:
 	case 4:
 	case 8:
-		DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr,
-			DMA_IN(fsl_chan, &fsl_chan->reg_base->mr, 32) |
-			FSL_DMA_MR_SAHE | (__ilog2(size) << 14),
-			32);
+		mode |= FSL_DMA_MR_SAHE | (__ilog2(size) << 14);
 		break;
 	}
+
+	DMA_OUT(chan, &chan->regs->mr, mode, 32);
 }
 
 /**
- * fsl_chan_set_dest_loop_size - Set destination address hold transfer size
- * @fsl_chan : Freescale DMA channel
+ * fsl_chan_set_dst_loop_size - Set destination address hold transfer size
+ * @chan : Freescale DMA channel
  * @size     : Address loop size, 0 for disable loop
  *
  * The set destination address hold transfer size. The destination
@@ -260,29 +237,30 @@ static void fsl_chan_set_src_loop_size(struct fsl_dma_chan *fsl_chan, int size)
  * write data to TA, TA + 1, TA + 2, TA + 3, then loop back to TA,
  * TA + 1 ... and so on.
  */
-static void fsl_chan_set_dest_loop_size(struct fsl_dma_chan *fsl_chan, int size)
+static void fsl_chan_set_dst_loop_size(struct fsldma_chan *chan, int size)
 {
+	u32 mode;
+
+	mode = DMA_IN(chan, &chan->regs->mr, 32);
+
 	switch (size) {
 	case 0:
-		DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr,
-			DMA_IN(fsl_chan, &fsl_chan->reg_base->mr, 32) &
-			(~FSL_DMA_MR_DAHE), 32);
+		mode &= ~FSL_DMA_MR_DAHE;
 		break;
 	case 1:
 	case 2:
 	case 4:
 	case 8:
-		DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr,
-			DMA_IN(fsl_chan, &fsl_chan->reg_base->mr, 32) |
-			FSL_DMA_MR_DAHE | (__ilog2(size) << 16),
-			32);
+		mode |= FSL_DMA_MR_DAHE | (__ilog2(size) << 16);
 		break;
 	}
+
+	DMA_OUT(chan, &chan->regs->mr, mode, 32);
 }
 
 /**
  * fsl_chan_set_request_count - Set DMA Request Count for external control
- * @fsl_chan : Freescale DMA channel
+ * @chan : Freescale DMA channel
  * @size     : Number of bytes to transfer in a single request
  *
  * The Freescale DMA channel can be controlled by the external signal DREQ#.
@@ -292,35 +270,38 @@ static void fsl_chan_set_dest_loop_size(struct fsl_dma_chan *fsl_chan, int size)
  *
  * A size of 0 disables external pause control. The maximum size is 1024.
  */
-static void fsl_chan_set_request_count(struct fsl_dma_chan *fsl_chan, int size)
+static void fsl_chan_set_request_count(struct fsldma_chan *chan, int size)
 {
+	u32 mode;
+
 	BUG_ON(size > 1024);
-	DMA_OUT(fsl_chan, &fsl_chan->reg_base->mr,
-		DMA_IN(fsl_chan, &fsl_chan->reg_base->mr, 32)
-			| ((__ilog2(size) << 24) & 0x0f000000),
-		32);
+
+	mode = DMA_IN(chan, &chan->regs->mr, 32);
+	mode |= (__ilog2(size) << 24) & 0x0f000000;
+
+	DMA_OUT(chan, &chan->regs->mr, mode, 32);
 }
 
 /**
  * fsl_chan_toggle_ext_pause - Toggle channel external pause status
- * @fsl_chan : Freescale DMA channel
+ * @chan : Freescale DMA channel
  * @enable   : 0 is disabled, 1 is enabled.
  *
  * The Freescale DMA channel can be controlled by the external signal DREQ#.
  * The DMA Request Count feature should be used in addition to this feature
  * to set the number of bytes to transfer before pausing the channel.
  */
-static void fsl_chan_toggle_ext_pause(struct fsl_dma_chan *fsl_chan, int enable)
+static void fsl_chan_toggle_ext_pause(struct fsldma_chan *chan, int enable)
 {
 	if (enable)
-		fsl_chan->feature |= FSL_DMA_CHAN_PAUSE_EXT;
+		chan->feature |= FSL_DMA_CHAN_PAUSE_EXT;
 	else
-		fsl_chan->feature &= ~FSL_DMA_CHAN_PAUSE_EXT;
+		chan->feature &= ~FSL_DMA_CHAN_PAUSE_EXT;
 }
 
 /**
  * fsl_chan_toggle_ext_start - Toggle channel external start status
- * @fsl_chan : Freescale DMA channel
+ * @chan : Freescale DMA channel
  * @enable   : 0 is disabled, 1 is enabled.
  *
  * If enable the external start, the channel can be started by an
@@ -328,141 +309,196 @@ static void fsl_chan_toggle_ext_pause(struct fsl_dma_chan *fsl_chan, int enable)
  * transfer immediately. The DMA channel will wait for the
  * control pin asserted.
  */
-static void fsl_chan_toggle_ext_start(struct fsl_dma_chan *fsl_chan, int enable)
+static void fsl_chan_toggle_ext_start(struct fsldma_chan *chan, int enable)
 {
 	if (enable)
-		fsl_chan->feature |= FSL_DMA_CHAN_START_EXT;
+		chan->feature |= FSL_DMA_CHAN_START_EXT;
 	else
-		fsl_chan->feature &= ~FSL_DMA_CHAN_START_EXT;
+		chan->feature &= ~FSL_DMA_CHAN_START_EXT;
+}
+
+static void append_ld_queue(struct fsldma_chan *chan,
+			    struct fsl_desc_sw *desc)
+{
+	struct fsl_desc_sw *tail = to_fsl_desc(chan->ld_pending.prev);
+
+	if (list_empty(&chan->ld_pending))
+		goto out_splice;
+
+	/*
+	 * Add the hardware descriptor to the chain of hardware descriptors
+	 * that already exists in memory.
+	 *
+	 * This will un-set the EOL bit of the existing transaction, and the
+	 * last link in this transaction will become the EOL descriptor.
+	 */
+	set_desc_next(chan, &tail->hw, desc->async_tx.phys);
+
+	/*
+	 * Add the software descriptor and all children to the list
+	 * of pending transactions
+	 */
+out_splice:
+	list_splice_tail_init(&desc->tx_list, &chan->ld_pending);
 }
 
 static dma_cookie_t fsl_dma_tx_submit(struct dma_async_tx_descriptor *tx)
 {
-	struct fsl_dma_chan *fsl_chan = to_fsl_chan(tx->chan);
+	struct fsldma_chan *chan = to_fsl_chan(tx->chan);
 	struct fsl_desc_sw *desc = tx_to_fsl_desc(tx);
 	struct fsl_desc_sw *child;
 	unsigned long flags;
 	dma_cookie_t cookie;
 
-	/* cookie increment and adding to ld_queue must be atomic */
-	spin_lock_irqsave(&fsl_chan->desc_lock, flags);
+	spin_lock_irqsave(&chan->desc_lock, flags);
 
-	cookie = fsl_chan->common.cookie;
+	/*
+	 * assign cookies to all of the software descriptors
+	 * that make up this transaction
+	 */
+	cookie = chan->common.cookie;
 	list_for_each_entry(child, &desc->tx_list, node) {
 		cookie++;
 		if (cookie < 0)
 			cookie = 1;
 
-		desc->async_tx.cookie = cookie;
+		child->async_tx.cookie = cookie;
 	}
 
-	fsl_chan->common.cookie = cookie;
-	append_ld_queue(fsl_chan, desc);
-	list_splice_init(&desc->tx_list, fsl_chan->ld_queue.prev);
+	chan->common.cookie = cookie;
+
+	/* put this transaction onto the tail of the pending queue */
+	append_ld_queue(chan, desc);
 
-	spin_unlock_irqrestore(&fsl_chan->desc_lock, flags);
+	spin_unlock_irqrestore(&chan->desc_lock, flags);
 
 	return cookie;
 }
 
 /**
  * fsl_dma_alloc_descriptor - Allocate descriptor from channel's DMA pool.
- * @fsl_chan : Freescale DMA channel
+ * @chan : Freescale DMA channel
  *
  * Return - The descriptor allocated. NULL for failed.
  */
 static struct fsl_desc_sw *fsl_dma_alloc_descriptor(
-					struct fsl_dma_chan *fsl_chan)
+					struct fsldma_chan *chan)
 {
+	struct fsl_desc_sw *desc;
 	dma_addr_t pdesc;
-	struct fsl_desc_sw *desc_sw;
-
-	desc_sw = dma_pool_alloc(fsl_chan->desc_pool, GFP_ATOMIC, &pdesc);
-	if (desc_sw) {
-		memset(desc_sw, 0, sizeof(struct fsl_desc_sw));
-		INIT_LIST_HEAD(&desc_sw->tx_list);
-		dma_async_tx_descriptor_init(&desc_sw->async_tx,
-						&fsl_chan->common);
-		desc_sw->async_tx.tx_submit = fsl_dma_tx_submit;
-		desc_sw->async_tx.phys = pdesc;
+
+	desc = dma_pool_alloc(chan->desc_pool, GFP_ATOMIC, &pdesc);
+	if (!desc) {
+		dev_dbg(chan->dev, "out of memory for link desc\n");
+		return NULL;
 	}
 
-	return desc_sw;
+	memset(desc, 0, sizeof(*desc));
+	INIT_LIST_HEAD(&desc->tx_list);
+	dma_async_tx_descriptor_init(&desc->async_tx, &chan->common);
+	desc->async_tx.tx_submit = fsl_dma_tx_submit;
+	desc->async_tx.phys = pdesc;
+
+	return desc;
 }
 
 
 /**
  * fsl_dma_alloc_chan_resources - Allocate resources for DMA channel.
- * @fsl_chan : Freescale DMA channel
+ * @chan : Freescale DMA channel
  *
  * This function will create a dma pool for descriptor allocation.
  *
  * Return - The number of descriptors allocated.
  */
-static int fsl_dma_alloc_chan_resources(struct dma_chan *chan)
+static int fsl_dma_alloc_chan_resources(struct dma_chan *dchan)
 {
-	struct fsl_dma_chan *fsl_chan = to_fsl_chan(chan);
+	struct fsldma_chan *chan = to_fsl_chan(dchan);
 
 	/* Has this channel already been allocated? */
-	if (fsl_chan->desc_pool)
+	if (chan->desc_pool)
 		return 1;
 
-	/* We need the descriptor to be aligned to 32bytes
+	/*
+	 * We need the descriptor to be aligned to 32bytes
 	 * for meeting FSL DMA specification requirement.
 	 */
-	fsl_chan->desc_pool = dma_pool_create("fsl_dma_engine_desc_pool",
-			fsl_chan->dev, sizeof(struct fsl_desc_sw),
-			32, 0);
-	if (!fsl_chan->desc_pool) {
-		dev_err(fsl_chan->dev, "No memory for channel %d "
-			"descriptor dma pool.\n", fsl_chan->id);
-		return 0;
+	chan->desc_pool = dma_pool_create("fsl_dma_engine_desc_pool",
+					  chan->dev,
+					  sizeof(struct fsl_desc_sw),
+					  __alignof__(struct fsl_desc_sw), 0);
+	if (!chan->desc_pool) {
+		dev_err(chan->dev, "unable to allocate channel %d "
+				   "descriptor pool\n", chan->id);
+		return -ENOMEM;
 	}
 
+	/* there is at least one descriptor free to be allocated */
 	return 1;
 }
 
 /**
- * fsl_dma_free_chan_resources - Free all resources of the channel.
- * @fsl_chan : Freescale DMA channel
+ * fsldma_free_desc_list - Free all descriptors in a queue
+ * @chan: Freescae DMA channel
+ * @list: the list to free
+ *
+ * LOCKING: must hold chan->desc_lock
  */
-static void fsl_dma_free_chan_resources(struct dma_chan *chan)
+static void fsldma_free_desc_list(struct fsldma_chan *chan,
+				  struct list_head *list)
 {
-	struct fsl_dma_chan *fsl_chan = to_fsl_chan(chan);
 	struct fsl_desc_sw *desc, *_desc;
-	unsigned long flags;
 
-	dev_dbg(fsl_chan->dev, "Free all channel resources.\n");
-	spin_lock_irqsave(&fsl_chan->desc_lock, flags);
-	list_for_each_entry_safe(desc, _desc, &fsl_chan->ld_queue, node) {
-#ifdef FSL_DMA_LD_DEBUG
-		dev_dbg(fsl_chan->dev,
-				"LD %p will be released.\n", desc);
-#endif
+	list_for_each_entry_safe(desc, _desc, list, node) {
+		list_del(&desc->node);
+		dma_pool_free(chan->desc_pool, desc, desc->async_tx.phys);
+	}
+}
+
+static void fsldma_free_desc_list_reverse(struct fsldma_chan *chan,
+					  struct list_head *list)
+{
+	struct fsl_desc_sw *desc, *_desc;
+
+	list_for_each_entry_safe_reverse(desc, _desc, list, node) {
 		list_del(&desc->node);
-		/* free link descriptor */
-		dma_pool_free(fsl_chan->desc_pool, desc, desc->async_tx.phys);
+		dma_pool_free(chan->desc_pool, desc, desc->async_tx.phys);
 	}
-	spin_unlock_irqrestore(&fsl_chan->desc_lock, flags);
-	dma_pool_destroy(fsl_chan->desc_pool);
+}
+
+/**
+ * fsl_dma_free_chan_resources - Free all resources of the channel.
+ * @chan : Freescale DMA channel
+ */
+static void fsl_dma_free_chan_resources(struct dma_chan *dchan)
+{
+	struct fsldma_chan *chan = to_fsl_chan(dchan);
+	unsigned long flags;
+
+	dev_dbg(chan->dev, "Free all channel resources.\n");
+	spin_lock_irqsave(&chan->desc_lock, flags);
+	fsldma_free_desc_list(chan, &chan->ld_pending);
+	fsldma_free_desc_list(chan, &chan->ld_running);
+	spin_unlock_irqrestore(&chan->desc_lock, flags);
 
-	fsl_chan->desc_pool = NULL;
+	dma_pool_destroy(chan->desc_pool);
+	chan->desc_pool = NULL;
 }
 
 static struct dma_async_tx_descriptor *
-fsl_dma_prep_interrupt(struct dma_chan *chan, unsigned long flags)
+fsl_dma_prep_interrupt(struct dma_chan *dchan, unsigned long flags)
 {
-	struct fsl_dma_chan *fsl_chan;
+	struct fsldma_chan *chan;
 	struct fsl_desc_sw *new;
 
-	if (!chan)
+	if (!dchan)
 		return NULL;
 
-	fsl_chan = to_fsl_chan(chan);
+	chan = to_fsl_chan(dchan);
 
-	new = fsl_dma_alloc_descriptor(fsl_chan);
+	new = fsl_dma_alloc_descriptor(chan);
 	if (!new) {
-		dev_err(fsl_chan->dev, "No free memory for link descriptor\n");
+		dev_err(chan->dev, "No free memory for link descriptor\n");
 		return NULL;
 	}
 
@@ -473,51 +509,50 @@ fsl_dma_prep_interrupt(struct dma_chan *chan, unsigned long flags)
 	list_add_tail(&new->node, &new->tx_list);
 
 	/* Set End-of-link to the last link descriptor of new list*/
-	set_ld_eol(fsl_chan, new);
+	set_ld_eol(chan, new);
 
 	return &new->async_tx;
 }
 
 static struct dma_async_tx_descriptor *fsl_dma_prep_memcpy(
-	struct dma_chan *chan, dma_addr_t dma_dest, dma_addr_t dma_src,
+	struct dma_chan *dchan, dma_addr_t dma_dst, dma_addr_t dma_src,
 	size_t len, unsigned long flags)
 {
-	struct fsl_dma_chan *fsl_chan;
+	struct fsldma_chan *chan;
 	struct fsl_desc_sw *first = NULL, *prev = NULL, *new;
-	struct list_head *list;
 	size_t copy;
 
-	if (!chan)
+	if (!dchan)
 		return NULL;
 
 	if (!len)
 		return NULL;
 
-	fsl_chan = to_fsl_chan(chan);
+	chan = to_fsl_chan(dchan);
 
 	do {
 
 		/* Allocate the link descriptor from DMA pool */
-		new = fsl_dma_alloc_descriptor(fsl_chan);
+		new = fsl_dma_alloc_descriptor(chan);
 		if (!new) {
-			dev_err(fsl_chan->dev,
+			dev_err(chan->dev,
 					"No free memory for link descriptor\n");
 			goto fail;
 		}
 #ifdef FSL_DMA_LD_DEBUG
-		dev_dbg(fsl_chan->dev, "new link desc alloc %p\n", new);
+		dev_dbg(chan->dev, "new link desc alloc %p\n", new);
 #endif
 
 		copy = min(len, (size_t)FSL_DMA_BCR_MAX_CNT);
 
-		set_desc_cnt(fsl_chan, &new->hw, copy);
-		set_desc_src(fsl_chan, &new->hw, dma_src);
-		set_desc_dest(fsl_chan, &new->hw, dma_dest);
+		set_desc_cnt(chan, &new->hw, copy);
+		set_desc_src(chan, &new->hw, dma_src);
+		set_desc_dst(chan, &new->hw, dma_dst);
 
 		if (!first)
 			first = new;
 		else
-			set_desc_next(fsl_chan, &prev->hw, new->async_tx.phys);
+			set_desc_next(chan, &prev->hw, new->async_tx.phys);
 
 		new->async_tx.cookie = 0;
 		async_tx_ack(&new->async_tx);
@@ -525,7 +560,7 @@ static struct dma_async_tx_descriptor *fsl_dma_prep_memcpy(
 		prev = new;
 		len -= copy;
 		dma_src += copy;
-		dma_dest += copy;
+		dma_dst += copy;
 
 		/* Insert the link descriptor to the LD ring */
 		list_add_tail(&new->node, &first->tx_list);
@@ -535,7 +570,7 @@ static struct dma_async_tx_descriptor *fsl_dma_prep_memcpy(
 	new->async_tx.cookie = -EBUSY;
 
 	/* Set End-of-link to the last link descriptor of new list*/
-	set_ld_eol(fsl_chan, new);
+	set_ld_eol(chan, new);
 
 	return &first->async_tx;
 
@@ -543,12 +578,7 @@ fail:
 	if (!first)
 		return NULL;
 
-	list = &first->tx_list;
-	list_for_each_entry_safe_reverse(new, prev, list, node) {
-		list_del(&new->node);
-		dma_pool_free(fsl_chan->desc_pool, new, new->async_tx.phys);
-	}
-
+	fsldma_free_desc_list_reverse(chan, &first->tx_list);
 	return NULL;
 }
 
@@ -565,13 +595,12 @@ fail:
  * chan->private variable.
  */
 static struct dma_async_tx_descriptor *fsl_dma_prep_slave_sg(
-	struct dma_chan *chan, struct scatterlist *sgl, unsigned int sg_len,
+	struct dma_chan *dchan, struct scatterlist *sgl, unsigned int sg_len,
 	enum dma_data_direction direction, unsigned long flags)
 {
-	struct fsl_dma_chan *fsl_chan;
+	struct fsldma_chan *chan;
 	struct fsl_desc_sw *first = NULL, *prev = NULL, *new = NULL;
 	struct fsl_dma_slave *slave;
-	struct list_head *tx_list;
 	size_t copy;
 
 	int i;
@@ -581,14 +610,14 @@ static struct dma_async_tx_descriptor *fsl_dma_prep_slave_sg(
 	struct fsl_dma_hw_addr *hw;
 	dma_addr_t dma_dst, dma_src;
 
-	if (!chan)
+	if (!dchan)
 		return NULL;
 
-	if (!chan->private)
+	if (!dchan->private)
 		return NULL;
 
-	fsl_chan = to_fsl_chan(chan);
-	slave = chan->private;
+	chan = to_fsl_chan(dchan);
+	slave = dchan->private;
 
 	if (list_empty(&slave->addresses))
 		return NULL;
@@ -637,14 +666,14 @@ static struct dma_async_tx_descriptor *fsl_dma_prep_slave_sg(
 			}
 
 			/* Allocate the link descriptor from DMA pool */
-			new = fsl_dma_alloc_descriptor(fsl_chan);
+			new = fsl_dma_alloc_descriptor(chan);
 			if (!new) {
-				dev_err(fsl_chan->dev, "No free memory for "
+				dev_err(chan->dev, "No free memory for "
 						       "link descriptor\n");
 				goto fail;
 			}
 #ifdef FSL_DMA_LD_DEBUG
-			dev_dbg(fsl_chan->dev, "new link desc alloc %p\n", new);
+			dev_dbg(chan->dev, "new link desc alloc %p\n", new);
 #endif
 
 			/*
@@ -671,9 +700,9 @@ static struct dma_async_tx_descriptor *fsl_dma_prep_slave_sg(
 			}
 
 			/* Fill in the descriptor */
-			set_desc_cnt(fsl_chan, &new->hw, copy);
-			set_desc_src(fsl_chan, &new->hw, dma_src);
-			set_desc_dest(fsl_chan, &new->hw, dma_dst);
+			set_desc_cnt(chan, &new->hw, copy);
+			set_desc_src(chan, &new->hw, dma_src);
+			set_desc_dst(chan, &new->hw, dma_dst);
 
 			/*
 			 * If this is not the first descriptor, chain the
@@ -682,7 +711,7 @@ static struct dma_async_tx_descriptor *fsl_dma_prep_slave_sg(
 			if (!first) {
 				first = new;
 			} else {
-				set_desc_next(fsl_chan, &prev->hw,
+				set_desc_next(chan, &prev->hw,
 					      new->async_tx.phys);
 			}
 
@@ -708,23 +737,23 @@ finished:
 	new->async_tx.cookie = -EBUSY;
 
 	/* Set End-of-link to the last link descriptor of new list */
-	set_ld_eol(fsl_chan, new);
+	set_ld_eol(chan, new);
 
 	/* Enable extra controller features */
-	if (fsl_chan->set_src_loop_size)
-		fsl_chan->set_src_loop_size(fsl_chan, slave->src_loop_size);
+	if (chan->set_src_loop_size)
+		chan->set_src_loop_size(chan, slave->src_loop_size);
 
-	if (fsl_chan->set_dest_loop_size)
-		fsl_chan->set_dest_loop_size(fsl_chan, slave->dst_loop_size);
+	if (chan->set_dst_loop_size)
+		chan->set_dst_loop_size(chan, slave->dst_loop_size);
 
-	if (fsl_chan->toggle_ext_start)
-		fsl_chan->toggle_ext_start(fsl_chan, slave->external_start);
+	if (chan->toggle_ext_start)
+		chan->toggle_ext_start(chan, slave->external_start);
 
-	if (fsl_chan->toggle_ext_pause)
-		fsl_chan->toggle_ext_pause(fsl_chan, slave->external_pause);
+	if (chan->toggle_ext_pause)
+		chan->toggle_ext_pause(chan, slave->external_pause);
 
-	if (fsl_chan->set_request_count)
-		fsl_chan->set_request_count(fsl_chan, slave->request_count);
+	if (chan->set_request_count)
+		chan->set_request_count(chan, slave->request_count);
 
 	return &first->async_tx;
 
@@ -741,215 +770,216 @@ fail:
 	 *
 	 * We're re-using variables for the loop, oh well
 	 */
-	tx_list = &first->tx_list;
-	list_for_each_entry_safe_reverse(new, prev, tx_list, node) {
-		list_del_init(&new->node);
-		dma_pool_free(fsl_chan->desc_pool, new, new->async_tx.phys);
-	}
-
+	fsldma_free_desc_list_reverse(chan, &first->tx_list);
 	return NULL;
 }
 
-static void fsl_dma_device_terminate_all(struct dma_chan *chan)
+static void fsl_dma_device_terminate_all(struct dma_chan *dchan)
 {
-	struct fsl_dma_chan *fsl_chan;
-	struct fsl_desc_sw *desc, *tmp;
+	struct fsldma_chan *chan;
 	unsigned long flags;
 
-	if (!chan)
+	if (!dchan)
 		return;
 
-	fsl_chan = to_fsl_chan(chan);
+	chan = to_fsl_chan(dchan);
 
 	/* Halt the DMA engine */
-	dma_halt(fsl_chan);
+	dma_halt(chan);
 
-	spin_lock_irqsave(&fsl_chan->desc_lock, flags);
+	spin_lock_irqsave(&chan->desc_lock, flags);
 
 	/* Remove and free all of the descriptors in the LD queue */
-	list_for_each_entry_safe(desc, tmp, &fsl_chan->ld_queue, node) {
-		list_del(&desc->node);
-		dma_pool_free(fsl_chan->desc_pool, desc, desc->async_tx.phys);
-	}
+	fsldma_free_desc_list(chan, &chan->ld_pending);
+	fsldma_free_desc_list(chan, &chan->ld_running);
 
-	spin_unlock_irqrestore(&fsl_chan->desc_lock, flags);
+	spin_unlock_irqrestore(&chan->desc_lock, flags);
 }
 
 /**
  * fsl_dma_update_completed_cookie - Update the completed cookie.
- * @fsl_chan : Freescale DMA channel
+ * @chan : Freescale DMA channel
+ *
+ * CONTEXT: hardirq
  */
-static void fsl_dma_update_completed_cookie(struct fsl_dma_chan *fsl_chan)
+static void fsl_dma_update_completed_cookie(struct fsldma_chan *chan)
 {
-	struct fsl_desc_sw *cur_desc, *desc;
-	dma_addr_t ld_phy;
+	struct fsl_desc_sw *desc;
+	unsigned long flags;
+	dma_cookie_t cookie;
 
-	ld_phy = get_cdar(fsl_chan) & FSL_DMA_NLDA_MASK;
+	spin_lock_irqsave(&chan->desc_lock, flags);
 
-	if (ld_phy) {
-		cur_desc = NULL;
-		list_for_each_entry(desc, &fsl_chan->ld_queue, node)
-			if (desc->async_tx.phys == ld_phy) {
-				cur_desc = desc;
-				break;
-			}
+	if (list_empty(&chan->ld_running)) {
+		dev_dbg(chan->dev, "no running descriptors\n");
+		goto out_unlock;
+	}
 
-		if (cur_desc && cur_desc->async_tx.cookie) {
-			if (dma_is_idle(fsl_chan))
-				fsl_chan->completed_cookie =
-					cur_desc->async_tx.cookie;
-			else
-				fsl_chan->completed_cookie =
-					cur_desc->async_tx.cookie - 1;
-		}
+	/* Get the last descriptor, update the cookie to that */
+	desc = to_fsl_desc(chan->ld_running.prev);
+	if (dma_is_idle(chan))
+		cookie = desc->async_tx.cookie;
+	else {
+		cookie = desc->async_tx.cookie - 1;
+		if (unlikely(cookie < DMA_MIN_COOKIE))
+			cookie = DMA_MAX_COOKIE;
 	}
+
+	chan->completed_cookie = cookie;
+
+out_unlock:
+	spin_unlock_irqrestore(&chan->desc_lock, flags);
+}
+
+/**
+ * fsldma_desc_status - Check the status of a descriptor
+ * @chan: Freescale DMA channel
+ * @desc: DMA SW descriptor
+ *
+ * This function will return the status of the given descriptor
+ */
+static enum dma_status fsldma_desc_status(struct fsldma_chan *chan,
+					  struct fsl_desc_sw *desc)
+{
+	return dma_async_is_complete(desc->async_tx.cookie,
+				     chan->completed_cookie,
+				     chan->common.cookie);
 }
 
 /**
  * fsl_chan_ld_cleanup - Clean up link descriptors
- * @fsl_chan : Freescale DMA channel
+ * @chan : Freescale DMA channel
  *
  * This function clean up the ld_queue of DMA channel.
- * If 'in_intr' is set, the function will move the link descriptor to
- * the recycle list. Otherwise, free it directly.
  */
-static void fsl_chan_ld_cleanup(struct fsl_dma_chan *fsl_chan)
+static void fsl_chan_ld_cleanup(struct fsldma_chan *chan)
 {
 	struct fsl_desc_sw *desc, *_desc;
 	unsigned long flags;
 
-	spin_lock_irqsave(&fsl_chan->desc_lock, flags);
+	spin_lock_irqsave(&chan->desc_lock, flags);
 
-	dev_dbg(fsl_chan->dev, "chan completed_cookie = %d\n",
-			fsl_chan->completed_cookie);
-	list_for_each_entry_safe(desc, _desc, &fsl_chan->ld_queue, node) {
+	dev_dbg(chan->dev, "chan completed_cookie = %d\n", chan->completed_cookie);
+	list_for_each_entry_safe(desc, _desc, &chan->ld_running, node) {
 		dma_async_tx_callback callback;
 		void *callback_param;
 
-		if (dma_async_is_complete(desc->async_tx.cookie,
-			    fsl_chan->completed_cookie, fsl_chan->common.cookie)
-				== DMA_IN_PROGRESS)
+		if (fsldma_desc_status(chan, desc) == DMA_IN_PROGRESS)
 			break;
 
-		callback = desc->async_tx.callback;
-		callback_param = desc->async_tx.callback_param;
-
-		/* Remove from ld_queue list */
+		/* Remove from the list of running transactions */
 		list_del(&desc->node);
 
-		dev_dbg(fsl_chan->dev, "link descriptor %p will be recycle.\n",
-				desc);
-		dma_pool_free(fsl_chan->desc_pool, desc, desc->async_tx.phys);
-
 		/* Run the link descriptor callback function */
+		callback = desc->async_tx.callback;
+		callback_param = desc->async_tx.callback_param;
 		if (callback) {
-			spin_unlock_irqrestore(&fsl_chan->desc_lock, flags);
-			dev_dbg(fsl_chan->dev, "link descriptor %p callback\n",
-					desc);
+			spin_unlock_irqrestore(&chan->desc_lock, flags);
+			dev_dbg(chan->dev, "LD %p callback\n", desc);
 			callback(callback_param);
-			spin_lock_irqsave(&fsl_chan->desc_lock, flags);
+			spin_lock_irqsave(&chan->desc_lock, flags);
 		}
+
+		/* Run any dependencies, then free the descriptor */
+		dma_run_dependencies(&desc->async_tx);
+		dma_pool_free(chan->desc_pool, desc, desc->async_tx.phys);
 	}
-	spin_unlock_irqrestore(&fsl_chan->desc_lock, flags);
+
+	spin_unlock_irqrestore(&chan->desc_lock, flags);
 }
 
 /**
- * fsl_chan_xfer_ld_queue - Transfer link descriptors in channel ld_queue.
- * @fsl_chan : Freescale DMA channel
+ * fsl_chan_xfer_ld_queue - transfer any pending transactions
+ * @chan : Freescale DMA channel
+ *
+ * This will make sure that any pending transactions will be run.
+ * If the DMA controller is idle, it will be started. Otherwise,
+ * the DMA controller's interrupt handler will start any pending
+ * transactions when it becomes idle.
  */
-static void fsl_chan_xfer_ld_queue(struct fsl_dma_chan *fsl_chan)
+static void fsl_chan_xfer_ld_queue(struct fsldma_chan *chan)
 {
-	struct list_head *ld_node;
-	dma_addr_t next_dest_addr;
+	struct fsl_desc_sw *desc;
 	unsigned long flags;
 
-	spin_lock_irqsave(&fsl_chan->desc_lock, flags);
+	spin_lock_irqsave(&chan->desc_lock, flags);
 
-	if (!dma_is_idle(fsl_chan))
+	/*
+	 * If the list of pending descriptors is empty, then we
+	 * don't need to do any work at all
+	 */
+	if (list_empty(&chan->ld_pending)) {
+		dev_dbg(chan->dev, "no pending LDs\n");
 		goto out_unlock;
+	}
 
-	dma_halt(fsl_chan);
+	/*
+	 * The DMA controller is not idle, which means the interrupt
+	 * handler will start any queued transactions when it runs
+	 * at the end of the current transaction
+	 */
+	if (!dma_is_idle(chan)) {
+		dev_dbg(chan->dev, "DMA controller still busy\n");
+		goto out_unlock;
+	}
 
-	/* If there are some link descriptors
-	 * not transfered in queue. We need to start it.
+	/*
+	 * TODO:
+	 * make sure the dma_halt() function really un-wedges the
+	 * controller as much as possible
 	 */
+	dma_halt(chan);
 
-	/* Find the first un-transfer desciptor */
-	for (ld_node = fsl_chan->ld_queue.next;
-		(ld_node != &fsl_chan->ld_queue)
-			&& (dma_async_is_complete(
-				to_fsl_desc(ld_node)->async_tx.cookie,
-				fsl_chan->completed_cookie,
-				fsl_chan->common.cookie) == DMA_SUCCESS);
-		ld_node = ld_node->next);
-
-	if (ld_node != &fsl_chan->ld_queue) {
-		/* Get the ld start address from ld_queue */
-		next_dest_addr = to_fsl_desc(ld_node)->async_tx.phys;
-		dev_dbg(fsl_chan->dev, "xfer LDs staring from 0x%llx\n",
-				(unsigned long long)next_dest_addr);
-		set_cdar(fsl_chan, next_dest_addr);
-		dma_start(fsl_chan);
-	} else {
-		set_cdar(fsl_chan, 0);
-		set_ndar(fsl_chan, 0);
-	}
+	/*
+	 * If there are some link descriptors which have not been
+	 * transferred, we need to start the controller
+	 */
+
+	/*
+	 * Move all elements from the queue of pending transactions
+	 * onto the list of running transactions
+	 */
+	desc = list_first_entry(&chan->ld_pending, struct fsl_desc_sw, node);
+	list_splice_tail_init(&chan->ld_pending, &chan->ld_running);
+
+	/*
+	 * Program the descriptor's address into the DMA controller,
+	 * then start the DMA transaction
+	 */
+	set_cdar(chan, desc->async_tx.phys);
+	dma_start(chan);
 
 out_unlock:
-	spin_unlock_irqrestore(&fsl_chan->desc_lock, flags);
+	spin_unlock_irqrestore(&chan->desc_lock, flags);
 }
 
 /**
  * fsl_dma_memcpy_issue_pending - Issue the DMA start command
- * @fsl_chan : Freescale DMA channel
+ * @chan : Freescale DMA channel
  */
-static void fsl_dma_memcpy_issue_pending(struct dma_chan *chan)
+static void fsl_dma_memcpy_issue_pending(struct dma_chan *dchan)
 {
-	struct fsl_dma_chan *fsl_chan = to_fsl_chan(chan);
-
-#ifdef FSL_DMA_LD_DEBUG
-	struct fsl_desc_sw *ld;
-	unsigned long flags;
-
-	spin_lock_irqsave(&fsl_chan->desc_lock, flags);
-	if (list_empty(&fsl_chan->ld_queue)) {
-		spin_unlock_irqrestore(&fsl_chan->desc_lock, flags);
-		return;
-	}
-
-	dev_dbg(fsl_chan->dev, "--memcpy issue--\n");
-	list_for_each_entry(ld, &fsl_chan->ld_queue, node) {
-		int i;
-		dev_dbg(fsl_chan->dev, "Ch %d, LD %08x\n",
-				fsl_chan->id, ld->async_tx.phys);
-		for (i = 0; i < 8; i++)
-			dev_dbg(fsl_chan->dev, "LD offset %d: %08x\n",
-					i, *(((u32 *)&ld->hw) + i));
-	}
-	dev_dbg(fsl_chan->dev, "----------------\n");
-	spin_unlock_irqrestore(&fsl_chan->desc_lock, flags);
-#endif
-
-	fsl_chan_xfer_ld_queue(fsl_chan);
+	struct fsldma_chan *chan = to_fsl_chan(dchan);
+	fsl_chan_xfer_ld_queue(chan);
 }
 
 /**
  * fsl_dma_is_complete - Determine the DMA status
- * @fsl_chan : Freescale DMA channel
+ * @chan : Freescale DMA channel
  */
-static enum dma_status fsl_dma_is_complete(struct dma_chan *chan,
+static enum dma_status fsl_dma_is_complete(struct dma_chan *dchan,
 					dma_cookie_t cookie,
 					dma_cookie_t *done,
 					dma_cookie_t *used)
 {
-	struct fsl_dma_chan *fsl_chan = to_fsl_chan(chan);
+	struct fsldma_chan *chan = to_fsl_chan(dchan);
 	dma_cookie_t last_used;
 	dma_cookie_t last_complete;
 
-	fsl_chan_ld_cleanup(fsl_chan);
+	fsl_chan_ld_cleanup(chan);
 
-	last_used = chan->cookie;
-	last_complete = fsl_chan->completed_cookie;
+	last_used = dchan->cookie;
+	last_complete = chan->completed_cookie;
 
 	if (done)
 		*done = last_complete;
@@ -960,32 +990,37 @@ static enum dma_status fsl_dma_is_complete(struct dma_chan *chan,
 	return dma_async_is_complete(cookie, last_complete, last_used);
 }
 
-static irqreturn_t fsl_dma_chan_do_interrupt(int irq, void *data)
+/*----------------------------------------------------------------------------*/
+/* Interrupt Handling                                                         */
+/*----------------------------------------------------------------------------*/
+
+static irqreturn_t fsldma_chan_irq(int irq, void *data)
 {
-	struct fsl_dma_chan *fsl_chan = (struct fsl_dma_chan *)data;
-	u32 stat;
+	struct fsldma_chan *chan = data;
 	int update_cookie = 0;
 	int xfer_ld_q = 0;
+	u32 stat;
 
-	stat = get_sr(fsl_chan);
-	dev_dbg(fsl_chan->dev, "event: channel %d, stat = 0x%x\n",
-						fsl_chan->id, stat);
-	set_sr(fsl_chan, stat);		/* Clear the event register */
+	/* save and clear the status register */
+	stat = get_sr(chan);
+	set_sr(chan, stat);
+	dev_dbg(chan->dev, "irq: channel %d, stat = 0x%x\n", chan->id, stat);
 
 	stat &= ~(FSL_DMA_SR_CB | FSL_DMA_SR_CH);
 	if (!stat)
 		return IRQ_NONE;
 
 	if (stat & FSL_DMA_SR_TE)
-		dev_err(fsl_chan->dev, "Transfer Error!\n");
+		dev_err(chan->dev, "Transfer Error!\n");
 
-	/* Programming Error
+	/*
+	 * Programming Error
 	 * The DMA_INTERRUPT async_tx is a NULL transfer, which will
 	 * triger a PE interrupt.
 	 */
 	if (stat & FSL_DMA_SR_PE) {
-		dev_dbg(fsl_chan->dev, "event: Programming Error INT\n");
-		if (get_bcr(fsl_chan) == 0) {
+		dev_dbg(chan->dev, "irq: Programming Error INT\n");
+		if (get_bcr(chan) == 0) {
 			/* BCR register is 0, this is a DMA_INTERRUPT async_tx.
 			 * Now, update the completed cookie, and continue the
 			 * next uncompleted transfer.
@@ -996,208 +1031,296 @@ static irqreturn_t fsl_dma_chan_do_interrupt(int irq, void *data)
 		stat &= ~FSL_DMA_SR_PE;
 	}
 
-	/* If the link descriptor segment transfer finishes,
+	/*
+	 * If the link descriptor segment transfer finishes,
 	 * we will recycle the used descriptor.
 	 */
 	if (stat & FSL_DMA_SR_EOSI) {
-		dev_dbg(fsl_chan->dev, "event: End-of-segments INT\n");
-		dev_dbg(fsl_chan->dev, "event: clndar 0x%llx, nlndar 0x%llx\n",
-			(unsigned long long)get_cdar(fsl_chan),
-			(unsigned long long)get_ndar(fsl_chan));
+		dev_dbg(chan->dev, "irq: End-of-segments INT\n");
+		dev_dbg(chan->dev, "irq: clndar 0x%llx, nlndar 0x%llx\n",
+			(unsigned long long)get_cdar(chan),
+			(unsigned long long)get_ndar(chan));
 		stat &= ~FSL_DMA_SR_EOSI;
 		update_cookie = 1;
 	}
 
-	/* For MPC8349, EOCDI event need to update cookie
+	/*
+	 * For MPC8349, EOCDI event need to update cookie
 	 * and start the next transfer if it exist.
 	 */
 	if (stat & FSL_DMA_SR_EOCDI) {
-		dev_dbg(fsl_chan->dev, "event: End-of-Chain link INT\n");
+		dev_dbg(chan->dev, "irq: End-of-Chain link INT\n");
 		stat &= ~FSL_DMA_SR_EOCDI;
 		update_cookie = 1;
 		xfer_ld_q = 1;
 	}
 
-	/* If it current transfer is the end-of-transfer,
+	/*
+	 * If it current transfer is the end-of-transfer,
 	 * we should clear the Channel Start bit for
 	 * prepare next transfer.
 	 */
 	if (stat & FSL_DMA_SR_EOLNI) {
-		dev_dbg(fsl_chan->dev, "event: End-of-link INT\n");
+		dev_dbg(chan->dev, "irq: End-of-link INT\n");
 		stat &= ~FSL_DMA_SR_EOLNI;
 		xfer_ld_q = 1;
 	}
 
 	if (update_cookie)
-		fsl_dma_update_completed_cookie(fsl_chan);
+		fsl_dma_update_completed_cookie(chan);
 	if (xfer_ld_q)
-		fsl_chan_xfer_ld_queue(fsl_chan);
+		fsl_chan_xfer_ld_queue(chan);
 	if (stat)
-		dev_dbg(fsl_chan->dev, "event: unhandled sr 0x%02x\n",
-					stat);
+		dev_dbg(chan->dev, "irq: unhandled sr 0x%02x\n", stat);
 
-	dev_dbg(fsl_chan->dev, "event: Exit\n");
-	tasklet_schedule(&fsl_chan->tasklet);
+	dev_dbg(chan->dev, "irq: Exit\n");
+	tasklet_schedule(&chan->tasklet);
 	return IRQ_HANDLED;
 }
 
-static irqreturn_t fsl_dma_do_interrupt(int irq, void *data)
+static void dma_do_tasklet(unsigned long data)
 {
-	struct fsl_dma_device *fdev = (struct fsl_dma_device *)data;
-	u32 gsr;
-	int ch_nr;
+	struct fsldma_chan *chan = (struct fsldma_chan *)data;
+	fsl_chan_ld_cleanup(chan);
+}
+
+static irqreturn_t fsldma_ctrl_irq(int irq, void *data)
+{
+	struct fsldma_device *fdev = data;
+	struct fsldma_chan *chan;
+	unsigned int handled = 0;
+	u32 gsr, mask;
+	int i;
 
-	gsr = (fdev->feature & FSL_DMA_BIG_ENDIAN) ? in_be32(fdev->reg_base)
-			: in_le32(fdev->reg_base);
-	ch_nr = (32 - ffs(gsr)) / 8;
+	gsr = (fdev->feature & FSL_DMA_BIG_ENDIAN) ? in_be32(fdev->regs)
+						   : in_le32(fdev->regs);
+	mask = 0xff000000;
+	dev_dbg(fdev->dev, "IRQ: gsr 0x%.8x\n", gsr);
+
+	for (i = 0; i < FSL_DMA_MAX_CHANS_PER_DEVICE; i++) {
+		chan = fdev->chan[i];
+		if (!chan)
+			continue;
+
+		if (gsr & mask) {
+			dev_dbg(fdev->dev, "IRQ: chan %d\n", chan->id);
+			fsldma_chan_irq(irq, chan);
+			handled++;
+		}
 
-	return fdev->chan[ch_nr] ? fsl_dma_chan_do_interrupt(irq,
-			fdev->chan[ch_nr]) : IRQ_NONE;
+		gsr &= ~mask;
+		mask >>= 8;
+	}
+
+	return IRQ_RETVAL(handled);
 }
 
-static void dma_do_tasklet(unsigned long data)
+static void fsldma_free_irqs(struct fsldma_device *fdev)
 {
-	struct fsl_dma_chan *fsl_chan = (struct fsl_dma_chan *)data;
-	fsl_chan_ld_cleanup(fsl_chan);
+	struct fsldma_chan *chan;
+	int i;
+
+	if (fdev->irq != NO_IRQ) {
+		dev_dbg(fdev->dev, "free per-controller IRQ\n");
+		free_irq(fdev->irq, fdev);
+		return;
+	}
+
+	for (i = 0; i < FSL_DMA_MAX_CHANS_PER_DEVICE; i++) {
+		chan = fdev->chan[i];
+		if (chan && chan->irq != NO_IRQ) {
+			dev_dbg(fdev->dev, "free channel %d IRQ\n", chan->id);
+			free_irq(chan->irq, chan);
+		}
+	}
 }
 
-static int __devinit fsl_dma_chan_probe(struct fsl_dma_device *fdev,
+static int fsldma_request_irqs(struct fsldma_device *fdev)
+{
+	struct fsldma_chan *chan;
+	int ret;
+	int i;
+
+	/* if we have a per-controller IRQ, use that */
+	if (fdev->irq != NO_IRQ) {
+		dev_dbg(fdev->dev, "request per-controller IRQ\n");
+		ret = request_irq(fdev->irq, fsldma_ctrl_irq, IRQF_SHARED,
+				  "fsldma-controller", fdev);
+		return ret;
+	}
+
+	/* no per-controller IRQ, use the per-channel IRQs */
+	for (i = 0; i < FSL_DMA_MAX_CHANS_PER_DEVICE; i++) {
+		chan = fdev->chan[i];
+		if (!chan)
+			continue;
+
+		if (chan->irq == NO_IRQ) {
+			dev_err(fdev->dev, "no interrupts property defined for "
+					   "DMA channel %d. Please fix your "
+					   "device tree\n", chan->id);
+			ret = -ENODEV;
+			goto out_unwind;
+		}
+
+		dev_dbg(fdev->dev, "request channel %d IRQ\n", chan->id);
+		ret = request_irq(chan->irq, fsldma_chan_irq, IRQF_SHARED,
+				  "fsldma-chan", chan);
+		if (ret) {
+			dev_err(fdev->dev, "unable to request IRQ for DMA "
+					   "channel %d\n", chan->id);
+			goto out_unwind;
+		}
+	}
+
+	return 0;
+
+out_unwind:
+	for (/* none */; i >= 0; i--) {
+		chan = fdev->chan[i];
+		if (!chan)
+			continue;
+
+		if (chan->irq == NO_IRQ)
+			continue;
+
+		free_irq(chan->irq, chan);
+	}
+
+	return ret;
+}
+
+/*----------------------------------------------------------------------------*/
+/* OpenFirmware Subsystem                                                     */
+/*----------------------------------------------------------------------------*/
+
+static int __devinit fsl_dma_chan_probe(struct fsldma_device *fdev,
 	struct device_node *node, u32 feature, const char *compatible)
 {
-	struct fsl_dma_chan *new_fsl_chan;
+	struct fsldma_chan *chan;
+	struct resource res;
 	int err;
 
 	/* alloc channel */
-	new_fsl_chan = kzalloc(sizeof(struct fsl_dma_chan), GFP_KERNEL);
-	if (!new_fsl_chan) {
-		dev_err(fdev->dev, "No free memory for allocating "
-				"dma channels!\n");
-		return -ENOMEM;
+	chan = kzalloc(sizeof(*chan), GFP_KERNEL);
+	if (!chan) {
+		dev_err(fdev->dev, "no free memory for DMA channels!\n");
+		err = -ENOMEM;
+		goto out_return;
 	}
 
-	/* get dma channel register base */
-	err = of_address_to_resource(node, 0, &new_fsl_chan->reg);
-	if (err) {
-		dev_err(fdev->dev, "Can't get %s property 'reg'\n",
-				node->full_name);
-		goto err_no_reg;
+	/* ioremap registers for use */
+	chan->regs = of_iomap(node, 0);
+	if (!chan->regs) {
+		dev_err(fdev->dev, "unable to ioremap registers\n");
+		err = -ENOMEM;
+		goto out_free_chan;
 	}
 
-	new_fsl_chan->feature = feature;
+	err = of_address_to_resource(node, 0, &res);
+	if (err) {
+		dev_err(fdev->dev, "unable to find 'reg' property\n");
+		goto out_iounmap_regs;
+	}
 
+	chan->feature = feature;
 	if (!fdev->feature)
-		fdev->feature = new_fsl_chan->feature;
+		fdev->feature = chan->feature;
 
-	/* If the DMA device's feature is different than its channels',
-	 * report the bug.
+	/*
+	 * If the DMA device's feature is different than the feature
+	 * of its channels, report the bug
 	 */
-	WARN_ON(fdev->feature != new_fsl_chan->feature);
+	WARN_ON(fdev->feature != chan->feature);
 
-	new_fsl_chan->dev = fdev->dev;
-	new_fsl_chan->reg_base = ioremap(new_fsl_chan->reg.start,
-			new_fsl_chan->reg.end - new_fsl_chan->reg.start + 1);
-
-	new_fsl_chan->id = ((new_fsl_chan->reg.start - 0x100) & 0xfff) >> 7;
-	if (new_fsl_chan->id >= FSL_DMA_MAX_CHANS_PER_DEVICE) {
-		dev_err(fdev->dev, "There is no %d channel!\n",
-				new_fsl_chan->id);
+	chan->dev = fdev->dev;
+	chan->id = ((res.start - 0x100) & 0xfff) >> 7;
+	if (chan->id >= FSL_DMA_MAX_CHANS_PER_DEVICE) {
+		dev_err(fdev->dev, "too many channels for device\n");
 		err = -EINVAL;
-		goto err_no_chan;
+		goto out_iounmap_regs;
 	}
-	fdev->chan[new_fsl_chan->id] = new_fsl_chan;
-	tasklet_init(&new_fsl_chan->tasklet, dma_do_tasklet,
-			(unsigned long)new_fsl_chan);
 
-	/* Init the channel */
-	dma_init(new_fsl_chan);
+	fdev->chan[chan->id] = chan;
+	tasklet_init(&chan->tasklet, dma_do_tasklet, (unsigned long)chan);
+
+	/* Initialize the channel */
+	dma_init(chan);
 
 	/* Clear cdar registers */
-	set_cdar(new_fsl_chan, 0);
+	set_cdar(chan, 0);
 
-	switch (new_fsl_chan->feature & FSL_DMA_IP_MASK) {
+	switch (chan->feature & FSL_DMA_IP_MASK) {
 	case FSL_DMA_IP_85XX:
-		new_fsl_chan->toggle_ext_pause = fsl_chan_toggle_ext_pause;
+		chan->toggle_ext_pause = fsl_chan_toggle_ext_pause;
 	case FSL_DMA_IP_83XX:
-		new_fsl_chan->toggle_ext_start = fsl_chan_toggle_ext_start;
-		new_fsl_chan->set_src_loop_size = fsl_chan_set_src_loop_size;
-		new_fsl_chan->set_dest_loop_size = fsl_chan_set_dest_loop_size;
-		new_fsl_chan->set_request_count = fsl_chan_set_request_count;
+		chan->toggle_ext_start = fsl_chan_toggle_ext_start;
+		chan->set_src_loop_size = fsl_chan_set_src_loop_size;
+		chan->set_dst_loop_size = fsl_chan_set_dst_loop_size;
+		chan->set_request_count = fsl_chan_set_request_count;
 	}
 
-	spin_lock_init(&new_fsl_chan->desc_lock);
-	INIT_LIST_HEAD(&new_fsl_chan->ld_queue);
+	spin_lock_init(&chan->desc_lock);
+	INIT_LIST_HEAD(&chan->ld_pending);
+	INIT_LIST_HEAD(&chan->ld_running);
+
+	chan->common.device = &fdev->common;
 
-	new_fsl_chan->common.device = &fdev->common;
+	/* find the IRQ line, if it exists in the device tree */
+	chan->irq = irq_of_parse_and_map(node, 0);
 
 	/* Add the channel to DMA device channel list */
-	list_add_tail(&new_fsl_chan->common.device_node,
-			&fdev->common.channels);
+	list_add_tail(&chan->common.device_node, &fdev->common.channels);
 	fdev->common.chancnt++;
 
-	new_fsl_chan->irq = irq_of_parse_and_map(node, 0);
-	if (new_fsl_chan->irq != NO_IRQ) {
-		err = request_irq(new_fsl_chan->irq,
-					&fsl_dma_chan_do_interrupt, IRQF_SHARED,
-					"fsldma-channel", new_fsl_chan);
-		if (err) {
-			dev_err(fdev->dev, "DMA channel %s request_irq error "
-				"with return %d\n", node->full_name, err);
-			goto err_no_irq;
-		}
-	}
-
-	dev_info(fdev->dev, "#%d (%s), irq %d\n", new_fsl_chan->id,
-		 compatible,
-		 new_fsl_chan->irq != NO_IRQ ? new_fsl_chan->irq : fdev->irq);
+	dev_info(fdev->dev, "#%d (%s), irq %d\n", chan->id, compatible,
+		 chan->irq != NO_IRQ ? chan->irq : fdev->irq);
 
 	return 0;
 
-err_no_irq:
-	list_del(&new_fsl_chan->common.device_node);
-err_no_chan:
-	iounmap(new_fsl_chan->reg_base);
-err_no_reg:
-	kfree(new_fsl_chan);
+out_iounmap_regs:
+	iounmap(chan->regs);
+out_free_chan:
+	kfree(chan);
+out_return:
 	return err;
 }
 
-static void fsl_dma_chan_remove(struct fsl_dma_chan *fchan)
+static void fsl_dma_chan_remove(struct fsldma_chan *chan)
 {
-	if (fchan->irq != NO_IRQ)
-		free_irq(fchan->irq, fchan);
-	list_del(&fchan->common.device_node);
-	iounmap(fchan->reg_base);
-	kfree(fchan);
+	irq_dispose_mapping(chan->irq);
+	list_del(&chan->common.device_node);
+	iounmap(chan->regs);
+	kfree(chan);
 }
 
-static int __devinit of_fsl_dma_probe(struct of_device *dev,
+static int __devinit fsldma_of_probe(struct of_device *op,
 			const struct of_device_id *match)
 {
-	int err;
-	struct fsl_dma_device *fdev;
+	struct fsldma_device *fdev;
 	struct device_node *child;
+	int err;
 
-	fdev = kzalloc(sizeof(struct fsl_dma_device), GFP_KERNEL);
+	fdev = kzalloc(sizeof(*fdev), GFP_KERNEL);
 	if (!fdev) {
-		dev_err(&dev->dev, "No enough memory for 'priv'\n");
-		return -ENOMEM;
+		dev_err(&op->dev, "No enough memory for 'priv'\n");
+		err = -ENOMEM;
+		goto out_return;
 	}
-	fdev->dev = &dev->dev;
+
+	fdev->dev = &op->dev;
 	INIT_LIST_HEAD(&fdev->common.channels);
 
-	/* get DMA controller register base */
-	err = of_address_to_resource(dev->node, 0, &fdev->reg);
-	if (err) {
-		dev_err(&dev->dev, "Can't get %s property 'reg'\n",
-				dev->node->full_name);
-		goto err_no_reg;
+	/* ioremap the registers for use */
+	fdev->regs = of_iomap(op->node, 0);
+	if (!fdev->regs) {
+		dev_err(&op->dev, "unable to ioremap registers\n");
+		err = -ENOMEM;
+		goto out_free_fdev;
 	}
 
-	dev_info(&dev->dev, "Probe the Freescale DMA driver for %s "
-			"controller at 0x%llx...\n",
-			match->compatible, (unsigned long long)fdev->reg.start);
-	fdev->reg_base = ioremap(fdev->reg.start, fdev->reg.end
-						- fdev->reg.start + 1);
+	/* map the channel IRQ if it exists, but don't hookup the handler yet */
+	fdev->irq = irq_of_parse_and_map(op->node, 0);
 
 	dma_cap_set(DMA_MEMCPY, fdev->common.cap_mask);
 	dma_cap_set(DMA_INTERRUPT, fdev->common.cap_mask);
@@ -1210,103 +1333,111 @@ static int __devinit of_fsl_dma_probe(struct of_device *dev,
 	fdev->common.device_issue_pending = fsl_dma_memcpy_issue_pending;
 	fdev->common.device_prep_slave_sg = fsl_dma_prep_slave_sg;
 	fdev->common.device_terminate_all = fsl_dma_device_terminate_all;
-	fdev->common.dev = &dev->dev;
+	fdev->common.dev = &op->dev;
 
-	fdev->irq = irq_of_parse_and_map(dev->node, 0);
-	if (fdev->irq != NO_IRQ) {
-		err = request_irq(fdev->irq, &fsl_dma_do_interrupt, IRQF_SHARED,
-					"fsldma-device", fdev);
-		if (err) {
-			dev_err(&dev->dev, "DMA device request_irq error "
-				"with return %d\n", err);
-			goto err;
-		}
-	}
-
-	dev_set_drvdata(&(dev->dev), fdev);
+	dev_set_drvdata(&op->dev, fdev);
 
-	/* We cannot use of_platform_bus_probe() because there is no
-	 * of_platform_bus_remove.  Instead, we manually instantiate every DMA
+	/*
+	 * We cannot use of_platform_bus_probe() because there is no
+	 * of_platform_bus_remove(). Instead, we manually instantiate every DMA
 	 * channel object.
 	 */
-	for_each_child_of_node(dev->node, child) {
-		if (of_device_is_compatible(child, "fsl,eloplus-dma-channel"))
+	for_each_child_of_node(op->node, child) {
+		if (of_device_is_compatible(child, "fsl,eloplus-dma-channel")) {
 			fsl_dma_chan_probe(fdev, child,
 				FSL_DMA_IP_85XX | FSL_DMA_BIG_ENDIAN,
 				"fsl,eloplus-dma-channel");
-		if (of_device_is_compatible(child, "fsl,elo-dma-channel"))
+		}
+
+		if (of_device_is_compatible(child, "fsl,elo-dma-channel")) {
 			fsl_dma_chan_probe(fdev, child,
 				FSL_DMA_IP_83XX | FSL_DMA_LITTLE_ENDIAN,
 				"fsl,elo-dma-channel");
+		}
+	}
+
+	/*
+	 * Hookup the IRQ handler(s)
+	 *
+	 * If we have a per-controller interrupt, we prefer that to the
+	 * per-channel interrupts to reduce the number of shared interrupt
+	 * handlers on the same IRQ line
+	 */
+	err = fsldma_request_irqs(fdev);
+	if (err) {
+		dev_err(fdev->dev, "unable to request IRQs\n");
+		goto out_free_fdev;
 	}
 
 	dma_async_device_register(&fdev->common);
 	return 0;
 
-err:
-	iounmap(fdev->reg_base);
-err_no_reg:
+out_free_fdev:
+	irq_dispose_mapping(fdev->irq);
 	kfree(fdev);
+out_return:
 	return err;
 }
 
-static int of_fsl_dma_remove(struct of_device *of_dev)
+static int fsldma_of_remove(struct of_device *op)
 {
-	struct fsl_dma_device *fdev;
+	struct fsldma_device *fdev;
 	unsigned int i;
 
-	fdev = dev_get_drvdata(&of_dev->dev);
-
+	fdev = dev_get_drvdata(&op->dev);
 	dma_async_device_unregister(&fdev->common);
 
-	for (i = 0; i < FSL_DMA_MAX_CHANS_PER_DEVICE; i++)
+	fsldma_free_irqs(fdev);
+
+	for (i = 0; i < FSL_DMA_MAX_CHANS_PER_DEVICE; i++) {
 		if (fdev->chan[i])
 			fsl_dma_chan_remove(fdev->chan[i]);
+	}
 
-	if (fdev->irq != NO_IRQ)
-		free_irq(fdev->irq, fdev);
-
-	iounmap(fdev->reg_base);
-
+	iounmap(fdev->regs);
+	dev_set_drvdata(&op->dev, NULL);
 	kfree(fdev);
-	dev_set_drvdata(&of_dev->dev, NULL);
 
 	return 0;
 }
 
-static struct of_device_id of_fsl_dma_ids[] = {
+static const struct of_device_id fsldma_of_ids[] = {
 	{ .compatible = "fsl,eloplus-dma", },
 	{ .compatible = "fsl,elo-dma", },
 	{}
 };
 
-static struct of_platform_driver of_fsl_dma_driver = {
-	.name = "fsl-elo-dma",
-	.match_table = of_fsl_dma_ids,
-	.probe = of_fsl_dma_probe,
-	.remove = of_fsl_dma_remove,
+static struct of_platform_driver fsldma_of_driver = {
+	.name		= "fsl-elo-dma",
+	.match_table	= fsldma_of_ids,
+	.probe		= fsldma_of_probe,
+	.remove		= fsldma_of_remove,
 };
 
-static __init int of_fsl_dma_init(void)
+/*----------------------------------------------------------------------------*/
+/* Module Init / Exit                                                         */
+/*----------------------------------------------------------------------------*/
+
+static __init int fsldma_init(void)
 {
 	int ret;
 
 	pr_info("Freescale Elo / Elo Plus DMA driver\n");
 
-	ret = of_register_platform_driver(&of_fsl_dma_driver);
+	ret = of_register_platform_driver(&fsldma_of_driver);
 	if (ret)
 		pr_err("fsldma: failed to register platform driver\n");
 
 	return ret;
 }
 
-static void __exit of_fsl_dma_exit(void)
+static void __exit fsldma_exit(void)
 {
-	of_unregister_platform_driver(&of_fsl_dma_driver);
+	of_unregister_platform_driver(&fsldma_of_driver);
 }
 
-subsys_initcall(of_fsl_dma_init);
-module_exit(of_fsl_dma_exit);
+subsys_initcall(fsldma_init);
+module_exit(fsldma_exit);
 
 MODULE_DESCRIPTION("Freescale Elo / Elo Plus DMA driver");
 MODULE_LICENSE("GPL");
diff --git a/drivers/dma/fsldma.h b/drivers/dma/fsldma.h
index 0df14cbb8ca..cb4d6ff5159 100644
--- a/drivers/dma/fsldma.h
+++ b/drivers/dma/fsldma.h
@@ -92,11 +92,9 @@ struct fsl_desc_sw {
 	struct list_head node;
 	struct list_head tx_list;
 	struct dma_async_tx_descriptor async_tx;
-	struct list_head *ld;
-	void *priv;
 } __attribute__((aligned(32)));
 
-struct fsl_dma_chan_regs {
+struct fsldma_chan_regs {
 	u32 mr;	/* 0x00 - Mode Register */
 	u32 sr;	/* 0x04 - Status Register */
 	u64 cdar;	/* 0x08 - Current descriptor address register */
@@ -106,20 +104,19 @@ struct fsl_dma_chan_regs {
 	u64 ndar;	/* 0x24 - Next Descriptor Address Register */
 };
 
-struct fsl_dma_chan;
+struct fsldma_chan;
 #define FSL_DMA_MAX_CHANS_PER_DEVICE 4
 
-struct fsl_dma_device {
-	void __iomem *reg_base;	/* DGSR register base */
-	struct resource reg;	/* Resource for register */
+struct fsldma_device {
+	void __iomem *regs;	/* DGSR register base */
 	struct device *dev;
 	struct dma_device common;
-	struct fsl_dma_chan *chan[FSL_DMA_MAX_CHANS_PER_DEVICE];
+	struct fsldma_chan *chan[FSL_DMA_MAX_CHANS_PER_DEVICE];
 	u32 feature;		/* The same as DMA channels */
 	int irq;		/* Channel IRQ */
 };
 
-/* Define macros for fsl_dma_chan->feature property */
+/* Define macros for fsldma_chan->feature property */
 #define FSL_DMA_LITTLE_ENDIAN	0x00000000
 #define FSL_DMA_BIG_ENDIAN	0x00000001
 
@@ -130,28 +127,28 @@ struct fsl_dma_device {
 #define FSL_DMA_CHAN_PAUSE_EXT	0x00001000
 #define FSL_DMA_CHAN_START_EXT	0x00002000
 
-struct fsl_dma_chan {
-	struct fsl_dma_chan_regs __iomem *reg_base;
+struct fsldma_chan {
+	struct fsldma_chan_regs __iomem *regs;
 	dma_cookie_t completed_cookie;	/* The maximum cookie completed */
 	spinlock_t desc_lock;		/* Descriptor operation lock */
-	struct list_head ld_queue;	/* Link descriptors queue */
+	struct list_head ld_pending;	/* Link descriptors queue */
+	struct list_head ld_running;	/* Link descriptors queue */
 	struct dma_chan common;		/* DMA common channel */
 	struct dma_pool *desc_pool;	/* Descriptors pool */
 	struct device *dev;		/* Channel device */
-	struct resource reg;		/* Resource for register */
 	int irq;			/* Channel IRQ */
 	int id;				/* Raw id of this channel */
 	struct tasklet_struct tasklet;
 	u32 feature;
 
-	void (*toggle_ext_pause)(struct fsl_dma_chan *fsl_chan, int enable);
-	void (*toggle_ext_start)(struct fsl_dma_chan *fsl_chan, int enable);
-	void (*set_src_loop_size)(struct fsl_dma_chan *fsl_chan, int size);
-	void (*set_dest_loop_size)(struct fsl_dma_chan *fsl_chan, int size);
-	void (*set_request_count)(struct fsl_dma_chan *fsl_chan, int size);
+	void (*toggle_ext_pause)(struct fsldma_chan *fsl_chan, int enable);
+	void (*toggle_ext_start)(struct fsldma_chan *fsl_chan, int enable);
+	void (*set_src_loop_size)(struct fsldma_chan *fsl_chan, int size);
+	void (*set_dst_loop_size)(struct fsldma_chan *fsl_chan, int size);
+	void (*set_request_count)(struct fsldma_chan *fsl_chan, int size);
 };
 
-#define to_fsl_chan(chan) container_of(chan, struct fsl_dma_chan, common)
+#define to_fsl_chan(chan) container_of(chan, struct fsldma_chan, common)
 #define to_fsl_desc(lh) container_of(lh, struct fsl_desc_sw, node)
 #define tx_to_fsl_desc(tx) container_of(tx, struct fsl_desc_sw, async_tx)
 
diff --git a/drivers/dma/ioat/dma.c b/drivers/dma/ioat/dma.c
index dcc4ab78b32..af14c9a5b8d 100644
--- a/drivers/dma/ioat/dma.c
+++ b/drivers/dma/ioat/dma.c
@@ -71,7 +71,7 @@ static irqreturn_t ioat_dma_do_interrupt(int irq, void *data)
 	}
 
 	attnstatus = readl(instance->reg_base + IOAT_ATTNSTATUS_OFFSET);
-	for_each_bit(bit, &attnstatus, BITS_PER_LONG) {
+	for_each_set_bit(bit, &attnstatus, BITS_PER_LONG) {
 		chan = ioat_chan_by_index(instance, bit);
 		tasklet_schedule(&chan->cleanup_task);
 	}
@@ -94,16 +94,12 @@ static irqreturn_t ioat_dma_do_interrupt_msix(int irq, void *data)
 	return IRQ_HANDLED;
 }
 
-static void ioat1_cleanup_tasklet(unsigned long data);
-
 /* common channel initialization */
-void ioat_init_channel(struct ioatdma_device *device,
-		       struct ioat_chan_common *chan, int idx,
-		       void (*timer_fn)(unsigned long),
-		       void (*tasklet)(unsigned long),
-		       unsigned long ioat)
+void ioat_init_channel(struct ioatdma_device *device, struct ioat_chan_common *chan, int idx)
 {
 	struct dma_device *dma = &device->common;
+	struct dma_chan *c = &chan->common;
+	unsigned long data = (unsigned long) c;
 
 	chan->device = device;
 	chan->reg_base = device->reg_base + (0x80 * (idx + 1));
@@ -112,14 +108,12 @@ void ioat_init_channel(struct ioatdma_device *device,
 	list_add_tail(&chan->common.device_node, &dma->channels);
 	device->idx[idx] = chan;
 	init_timer(&chan->timer);
-	chan->timer.function = timer_fn;
-	chan->timer.data = ioat;
-	tasklet_init(&chan->cleanup_task, tasklet, ioat);
+	chan->timer.function = device->timer_fn;
+	chan->timer.data = data;
+	tasklet_init(&chan->cleanup_task, device->cleanup_fn, data);
 	tasklet_disable(&chan->cleanup_task);
 }
 
-static void ioat1_timer_event(unsigned long data);
-
 /**
  * ioat1_dma_enumerate_channels - find and initialize the device's channels
  * @device: the device to be enumerated
@@ -155,10 +149,7 @@ static int ioat1_enumerate_channels(struct ioatdma_device *device)
 		if (!ioat)
 			break;
 
-		ioat_init_channel(device, &ioat->base, i,
-				  ioat1_timer_event,
-				  ioat1_cleanup_tasklet,
-				  (unsigned long) ioat);
+		ioat_init_channel(device, &ioat->base, i);
 		ioat->xfercap = xfercap;
 		spin_lock_init(&ioat->desc_lock);
 		INIT_LIST_HEAD(&ioat->free_desc);
@@ -532,12 +523,12 @@ ioat1_dma_prep_memcpy(struct dma_chan *c, dma_addr_t dma_dest,
 	return &desc->txd;
 }
 
-static void ioat1_cleanup_tasklet(unsigned long data)
+static void ioat1_cleanup_event(unsigned long data)
 {
-	struct ioat_dma_chan *chan = (void *)data;
+	struct ioat_dma_chan *ioat = to_ioat_chan((void *) data);
 
-	ioat1_cleanup(chan);
-	writew(IOAT_CHANCTRL_RUN, chan->base.reg_base + IOAT_CHANCTRL_OFFSET);
+	ioat1_cleanup(ioat);
+	writew(IOAT_CHANCTRL_RUN, ioat->base.reg_base + IOAT_CHANCTRL_OFFSET);
 }
 
 void ioat_dma_unmap(struct ioat_chan_common *chan, enum dma_ctrl_flags flags,
@@ -687,7 +678,7 @@ static void ioat1_cleanup(struct ioat_dma_chan *ioat)
 
 static void ioat1_timer_event(unsigned long data)
 {
-	struct ioat_dma_chan *ioat = (void *) data;
+	struct ioat_dma_chan *ioat = to_ioat_chan((void *) data);
 	struct ioat_chan_common *chan = &ioat->base;
 
 	dev_dbg(to_dev(chan), "%s: state: %lx\n", __func__, chan->state);
@@ -734,16 +725,17 @@ static void ioat1_timer_event(unsigned long data)
 	spin_unlock_bh(&chan->cleanup_lock);
 }
 
-static enum dma_status
-ioat1_dma_is_complete(struct dma_chan *c, dma_cookie_t cookie,
+enum dma_status
+ioat_is_dma_complete(struct dma_chan *c, dma_cookie_t cookie,
 		      dma_cookie_t *done, dma_cookie_t *used)
 {
-	struct ioat_dma_chan *ioat = to_ioat_chan(c);
+	struct ioat_chan_common *chan = to_chan_common(c);
+	struct ioatdma_device *device = chan->device;
 
 	if (ioat_is_complete(c, cookie, done, used) == DMA_SUCCESS)
 		return DMA_SUCCESS;
 
-	ioat1_cleanup(ioat);
+	device->cleanup_fn((unsigned long) c);
 
 	return ioat_is_complete(c, cookie, done, used);
 }
@@ -1199,12 +1191,14 @@ int __devinit ioat1_dma_probe(struct ioatdma_device *device, int dca)
 	device->intr_quirk = ioat1_intr_quirk;
 	device->enumerate_channels = ioat1_enumerate_channels;
 	device->self_test = ioat_dma_self_test;
+	device->timer_fn = ioat1_timer_event;
+	device->cleanup_fn = ioat1_cleanup_event;
 	dma = &device->common;
 	dma->device_prep_dma_memcpy = ioat1_dma_prep_memcpy;
 	dma->device_issue_pending = ioat1_dma_memcpy_issue_pending;
 	dma->device_alloc_chan_resources = ioat1_dma_alloc_chan_resources;
 	dma->device_free_chan_resources = ioat1_dma_free_chan_resources;
-	dma->device_is_tx_complete = ioat1_dma_is_complete;
+	dma->device_is_tx_complete = ioat_is_dma_complete;
 
 	err = ioat_probe(device);
 	if (err)
diff --git a/drivers/dma/ioat/dma.h b/drivers/dma/ioat/dma.h
index bbc3e78ef33..4f747a25407 100644
--- a/drivers/dma/ioat/dma.h
+++ b/drivers/dma/ioat/dma.h
@@ -61,7 +61,7 @@
  * @intr_quirk: interrupt setup quirk (for ioat_v1 devices)
  * @enumerate_channels: hw version specific channel enumeration
  * @reset_hw: hw version specific channel (re)initialization
- * @cleanup_tasklet: select between the v2 and v3 cleanup routines
+ * @cleanup_fn: select between the v2 and v3 cleanup routines
  * @timer_fn: select between the v2 and v3 timer watchdog routines
  * @self_test: hardware version specific self test for each supported op type
  *
@@ -80,7 +80,7 @@ struct ioatdma_device {
 	void (*intr_quirk)(struct ioatdma_device *device);
 	int (*enumerate_channels)(struct ioatdma_device *device);
 	int (*reset_hw)(struct ioat_chan_common *chan);
-	void (*cleanup_tasklet)(unsigned long data);
+	void (*cleanup_fn)(unsigned long data);
 	void (*timer_fn)(unsigned long data);
 	int (*self_test)(struct ioatdma_device *device);
 };
@@ -337,10 +337,9 @@ struct dca_provider * __devinit ioat_dca_init(struct pci_dev *pdev,
 					      void __iomem *iobase);
 unsigned long ioat_get_current_completion(struct ioat_chan_common *chan);
 void ioat_init_channel(struct ioatdma_device *device,
-		       struct ioat_chan_common *chan, int idx,
-		       void (*timer_fn)(unsigned long),
-		       void (*tasklet)(unsigned long),
-		       unsigned long ioat);
+		       struct ioat_chan_common *chan, int idx);
+enum dma_status ioat_is_dma_complete(struct dma_chan *c, dma_cookie_t cookie,
+				     dma_cookie_t *done, dma_cookie_t *used);
 void ioat_dma_unmap(struct ioat_chan_common *chan, enum dma_ctrl_flags flags,
 		    size_t len, struct ioat_dma_descriptor *hw);
 bool ioat_cleanup_preamble(struct ioat_chan_common *chan,
diff --git a/drivers/dma/ioat/dma_v2.c b/drivers/dma/ioat/dma_v2.c
index 5cc37afe2bc..1ed5d66d7dc 100644
--- a/drivers/dma/ioat/dma_v2.c
+++ b/drivers/dma/ioat/dma_v2.c
@@ -51,48 +51,40 @@ MODULE_PARM_DESC(ioat_ring_max_alloc_order,
 
 void __ioat2_issue_pending(struct ioat2_dma_chan *ioat)
 {
-	void * __iomem reg_base = ioat->base.reg_base;
+	struct ioat_chan_common *chan = &ioat->base;
 
-	ioat->pending = 0;
 	ioat->dmacount += ioat2_ring_pending(ioat);
 	ioat->issued = ioat->head;
 	/* make descriptor updates globally visible before notifying channel */
 	wmb();
-	writew(ioat->dmacount, reg_base + IOAT_CHAN_DMACOUNT_OFFSET);
-	dev_dbg(to_dev(&ioat->base),
+	writew(ioat->dmacount, chan->reg_base + IOAT_CHAN_DMACOUNT_OFFSET);
+	dev_dbg(to_dev(chan),
 		"%s: head: %#x tail: %#x issued: %#x count: %#x\n",
 		__func__, ioat->head, ioat->tail, ioat->issued, ioat->dmacount);
 }
 
-void ioat2_issue_pending(struct dma_chan *chan)
+void ioat2_issue_pending(struct dma_chan *c)
 {
-	struct ioat2_dma_chan *ioat = to_ioat2_chan(chan);
+	struct ioat2_dma_chan *ioat = to_ioat2_chan(c);
 
-	spin_lock_bh(&ioat->ring_lock);
-	if (ioat->pending == 1)
+	if (ioat2_ring_pending(ioat)) {
+		spin_lock_bh(&ioat->ring_lock);
 		__ioat2_issue_pending(ioat);
-	spin_unlock_bh(&ioat->ring_lock);
+		spin_unlock_bh(&ioat->ring_lock);
+	}
 }
 
 /**
  * ioat2_update_pending - log pending descriptors
  * @ioat: ioat2+ channel
  *
- * set pending to '1' unless pending is already set to '2', pending == 2
- * indicates that submission is temporarily blocked due to an in-flight
- * reset.  If we are already above the ioat_pending_level threshold then
- * just issue pending.
- *
- * called with ring_lock held
+ * Check if the number of unsubmitted descriptors has exceeded the
+ * watermark.  Called with ring_lock held
  */
 static void ioat2_update_pending(struct ioat2_dma_chan *ioat)
 {
-	if (unlikely(ioat->pending == 2))
-		return;
-	else if (ioat2_ring_pending(ioat) > ioat_pending_level)
+	if (ioat2_ring_pending(ioat) > ioat_pending_level)
 		__ioat2_issue_pending(ioat);
-	else
-		ioat->pending = 1;
 }
 
 static void __ioat2_start_null_desc(struct ioat2_dma_chan *ioat)
@@ -166,7 +158,7 @@ static void __cleanup(struct ioat2_dma_chan *ioat, unsigned long phys_complete)
 			seen_current = true;
 	}
 	ioat->tail += i;
-	BUG_ON(!seen_current); /* no active descs have written a completion? */
+	BUG_ON(active && !seen_current); /* no active descs have written a completion? */
 
 	chan->last_completion = phys_complete;
 	if (ioat->head == ioat->tail) {
@@ -207,9 +199,9 @@ static void ioat2_cleanup(struct ioat2_dma_chan *ioat)
 	spin_unlock_bh(&chan->cleanup_lock);
 }
 
-void ioat2_cleanup_tasklet(unsigned long data)
+void ioat2_cleanup_event(unsigned long data)
 {
-	struct ioat2_dma_chan *ioat = (void *) data;
+	struct ioat2_dma_chan *ioat = to_ioat2_chan((void *) data);
 
 	ioat2_cleanup(ioat);
 	writew(IOAT_CHANCTRL_RUN, ioat->base.reg_base + IOAT_CHANCTRL_OFFSET);
@@ -291,7 +283,7 @@ static void ioat2_restart_channel(struct ioat2_dma_chan *ioat)
 
 void ioat2_timer_event(unsigned long data)
 {
-	struct ioat2_dma_chan *ioat = (void *) data;
+	struct ioat2_dma_chan *ioat = to_ioat2_chan((void *) data);
 	struct ioat_chan_common *chan = &ioat->base;
 
 	spin_lock_bh(&chan->cleanup_lock);
@@ -397,10 +389,7 @@ int ioat2_enumerate_channels(struct ioatdma_device *device)
 		if (!ioat)
 			break;
 
-		ioat_init_channel(device, &ioat->base, i,
-				  device->timer_fn,
-				  device->cleanup_tasklet,
-				  (unsigned long) ioat);
+		ioat_init_channel(device, &ioat->base, i);
 		ioat->xfercap_log = xfercap_log;
 		spin_lock_init(&ioat->ring_lock);
 		if (device->reset_hw(&ioat->base)) {
@@ -546,7 +535,6 @@ int ioat2_alloc_chan_resources(struct dma_chan *c)
 	ioat->head = 0;
 	ioat->issued = 0;
 	ioat->tail = 0;
-	ioat->pending = 0;
 	ioat->alloc_order = order;
 	spin_unlock_bh(&ioat->ring_lock);
 
@@ -701,7 +689,7 @@ int ioat2_alloc_and_lock(u16 *idx, struct ioat2_dma_chan *ioat, int num_descs)
 
 			mod_timer(&chan->timer, jiffies + COMPLETION_TIMEOUT);
 			spin_unlock_bh(&chan->cleanup_lock);
-			device->timer_fn((unsigned long) ioat);
+			device->timer_fn((unsigned long) &chan->common);
 		} else
 			spin_unlock_bh(&chan->cleanup_lock);
 		return -ENOMEM;
@@ -785,7 +773,7 @@ void ioat2_free_chan_resources(struct dma_chan *c)
 
 	tasklet_disable(&chan->cleanup_task);
 	del_timer_sync(&chan->timer);
-	device->cleanup_tasklet((unsigned long) ioat);
+	device->cleanup_fn((unsigned long) c);
 	device->reset_hw(chan);
 
 	spin_lock_bh(&ioat->ring_lock);
@@ -815,25 +803,9 @@ void ioat2_free_chan_resources(struct dma_chan *c)
 
 	chan->last_completion = 0;
 	chan->completion_dma = 0;
-	ioat->pending = 0;
 	ioat->dmacount = 0;
 }
 
-enum dma_status
-ioat2_is_complete(struct dma_chan *c, dma_cookie_t cookie,
-		     dma_cookie_t *done, dma_cookie_t *used)
-{
-	struct ioat2_dma_chan *ioat = to_ioat2_chan(c);
-	struct ioatdma_device *device = ioat->base.device;
-
-	if (ioat_is_complete(c, cookie, done, used) == DMA_SUCCESS)
-		return DMA_SUCCESS;
-
-	device->cleanup_tasklet((unsigned long) ioat);
-
-	return ioat_is_complete(c, cookie, done, used);
-}
-
 static ssize_t ring_size_show(struct dma_chan *c, char *page)
 {
 	struct ioat2_dma_chan *ioat = to_ioat2_chan(c);
@@ -874,7 +846,7 @@ int __devinit ioat2_dma_probe(struct ioatdma_device *device, int dca)
 
 	device->enumerate_channels = ioat2_enumerate_channels;
 	device->reset_hw = ioat2_reset_hw;
-	device->cleanup_tasklet = ioat2_cleanup_tasklet;
+	device->cleanup_fn = ioat2_cleanup_event;
 	device->timer_fn = ioat2_timer_event;
 	device->self_test = ioat_dma_self_test;
 	dma = &device->common;
@@ -882,7 +854,7 @@ int __devinit ioat2_dma_probe(struct ioatdma_device *device, int dca)
 	dma->device_issue_pending = ioat2_issue_pending;
 	dma->device_alloc_chan_resources = ioat2_alloc_chan_resources;
 	dma->device_free_chan_resources = ioat2_free_chan_resources;
-	dma->device_is_tx_complete = ioat2_is_complete;
+	dma->device_is_tx_complete = ioat_is_dma_complete;
 
 	err = ioat_probe(device);
 	if (err)
diff --git a/drivers/dma/ioat/dma_v2.h b/drivers/dma/ioat/dma_v2.h
index 3afad8da43c..ef2871fd786 100644
--- a/drivers/dma/ioat/dma_v2.h
+++ b/drivers/dma/ioat/dma_v2.h
@@ -47,7 +47,6 @@ extern int ioat_ring_alloc_order;
  * @head: allocated index
  * @issued: hardware notification point
  * @tail: cleanup index
- * @pending: lock free indicator for issued != head
  * @dmacount: identical to 'head' except for occasionally resetting to zero
  * @alloc_order: log2 of the number of allocated descriptors
  * @ring: software ring buffer implementation of hardware ring
@@ -61,7 +60,6 @@ struct ioat2_dma_chan {
 	u16 tail;
 	u16 dmacount;
 	u16 alloc_order;
-	int pending;
 	struct ioat_ring_ent **ring;
 	spinlock_t ring_lock;
 };
@@ -178,12 +176,10 @@ ioat2_dma_prep_memcpy_lock(struct dma_chan *c, dma_addr_t dma_dest,
 void ioat2_issue_pending(struct dma_chan *chan);
 int ioat2_alloc_chan_resources(struct dma_chan *c);
 void ioat2_free_chan_resources(struct dma_chan *c);
-enum dma_status ioat2_is_complete(struct dma_chan *c, dma_cookie_t cookie,
-				  dma_cookie_t *done, dma_cookie_t *used);
 void __ioat2_restart_chan(struct ioat2_dma_chan *ioat);
 bool reshape_ring(struct ioat2_dma_chan *ioat, int order);
 void __ioat2_issue_pending(struct ioat2_dma_chan *ioat);
-void ioat2_cleanup_tasklet(unsigned long data);
+void ioat2_cleanup_event(unsigned long data);
 void ioat2_timer_event(unsigned long data);
 int ioat2_quiesce(struct ioat_chan_common *chan, unsigned long tmo);
 int ioat2_reset_sync(struct ioat_chan_common *chan, unsigned long tmo);
diff --git a/drivers/dma/ioat/dma_v3.c b/drivers/dma/ioat/dma_v3.c
index 9908c9e94b2..26febc56dab 100644
--- a/drivers/dma/ioat/dma_v3.c
+++ b/drivers/dma/ioat/dma_v3.c
@@ -293,17 +293,25 @@ static void __cleanup(struct ioat2_dma_chan *ioat, unsigned long phys_complete)
 		}
 	}
 	ioat->tail += i;
-	BUG_ON(!seen_current); /* no active descs have written a completion? */
+	BUG_ON(active && !seen_current); /* no active descs have written a completion? */
 	chan->last_completion = phys_complete;
-	if (ioat->head == ioat->tail) {
+
+	active = ioat2_ring_active(ioat);
+	if (active == 0) {
 		dev_dbg(to_dev(chan), "%s: cancel completion timeout\n",
 			__func__);
 		clear_bit(IOAT_COMPLETION_PENDING, &chan->state);
 		mod_timer(&chan->timer, jiffies + IDLE_TIMEOUT);
 	}
+	/* 5 microsecond delay per pending descriptor */
+	writew(min((5 * active), IOAT_INTRDELAY_MASK),
+	       chan->device->reg_base + IOAT_INTRDELAY_OFFSET);
 }
 
-static void ioat3_cleanup(struct ioat2_dma_chan *ioat)
+/* try to cleanup, but yield (via spin_trylock) to incoming submissions
+ * with the expectation that we will immediately poll again shortly
+ */
+static void ioat3_cleanup_poll(struct ioat2_dma_chan *ioat)
 {
 	struct ioat_chan_common *chan = &ioat->base;
 	unsigned long phys_complete;
@@ -329,29 +337,41 @@ static void ioat3_cleanup(struct ioat2_dma_chan *ioat)
 	spin_unlock_bh(&chan->cleanup_lock);
 }
 
-static void ioat3_cleanup_tasklet(unsigned long data)
+/* run cleanup now because we already delayed the interrupt via INTRDELAY */
+static void ioat3_cleanup_sync(struct ioat2_dma_chan *ioat)
+{
+	struct ioat_chan_common *chan = &ioat->base;
+	unsigned long phys_complete;
+
+	prefetch(chan->completion);
+
+	spin_lock_bh(&chan->cleanup_lock);
+	if (!ioat_cleanup_preamble(chan, &phys_complete)) {
+		spin_unlock_bh(&chan->cleanup_lock);
+		return;
+	}
+	spin_lock_bh(&ioat->ring_lock);
+
+	__cleanup(ioat, phys_complete);
+
+	spin_unlock_bh(&ioat->ring_lock);
+	spin_unlock_bh(&chan->cleanup_lock);
+}
+
+static void ioat3_cleanup_event(unsigned long data)
 {
-	struct ioat2_dma_chan *ioat = (void *) data;
+	struct ioat2_dma_chan *ioat = to_ioat2_chan((void *) data);
 
-	ioat3_cleanup(ioat);
-	writew(IOAT_CHANCTRL_RUN | IOAT3_CHANCTRL_COMPL_DCA_EN,
-	       ioat->base.reg_base + IOAT_CHANCTRL_OFFSET);
+	ioat3_cleanup_sync(ioat);
+	writew(IOAT_CHANCTRL_RUN, ioat->base.reg_base + IOAT_CHANCTRL_OFFSET);
 }
 
 static void ioat3_restart_channel(struct ioat2_dma_chan *ioat)
 {
 	struct ioat_chan_common *chan = &ioat->base;
 	unsigned long phys_complete;
-	u32 status;
-
-	status = ioat_chansts(chan);
-	if (is_ioat_active(status) || is_ioat_idle(status))
-		ioat_suspend(chan);
-	while (is_ioat_active(status) || is_ioat_idle(status)) {
-		status = ioat_chansts(chan);
-		cpu_relax();
-	}
 
+	ioat2_quiesce(chan, 0);
 	if (ioat_cleanup_preamble(chan, &phys_complete))
 		__cleanup(ioat, phys_complete);
 
@@ -360,7 +380,7 @@ static void ioat3_restart_channel(struct ioat2_dma_chan *ioat)
 
 static void ioat3_timer_event(unsigned long data)
 {
-	struct ioat2_dma_chan *ioat = (void *) data;
+	struct ioat2_dma_chan *ioat = to_ioat2_chan((void *) data);
 	struct ioat_chan_common *chan = &ioat->base;
 
 	spin_lock_bh(&chan->cleanup_lock);
@@ -426,7 +446,7 @@ ioat3_is_complete(struct dma_chan *c, dma_cookie_t cookie,
 	if (ioat_is_complete(c, cookie, done, used) == DMA_SUCCESS)
 		return DMA_SUCCESS;
 
-	ioat3_cleanup(ioat);
+	ioat3_cleanup_poll(ioat);
 
 	return ioat_is_complete(c, cookie, done, used);
 }
@@ -1239,11 +1259,11 @@ int __devinit ioat3_dma_probe(struct ioatdma_device *device, int dca)
 
 	if (is_raid_device) {
 		dma->device_is_tx_complete = ioat3_is_complete;
-		device->cleanup_tasklet = ioat3_cleanup_tasklet;
+		device->cleanup_fn = ioat3_cleanup_event;
 		device->timer_fn = ioat3_timer_event;
 	} else {
-		dma->device_is_tx_complete = ioat2_is_complete;
-		device->cleanup_tasklet = ioat2_cleanup_tasklet;
+		dma->device_is_tx_complete = ioat_is_dma_complete;
+		device->cleanup_fn = ioat2_cleanup_event;
 		device->timer_fn = ioat2_timer_event;
 	}
 
diff --git a/drivers/dma/ioat/registers.h b/drivers/dma/ioat/registers.h
index e8ae63baf58..1391798542b 100644
--- a/drivers/dma/ioat/registers.h
+++ b/drivers/dma/ioat/registers.h
@@ -60,7 +60,7 @@
 #define IOAT_PERPORTOFFSET_OFFSET		0x0A	/* 16-bit */
 
 #define IOAT_INTRDELAY_OFFSET			0x0C	/* 16-bit */
-#define IOAT_INTRDELAY_INT_DELAY_MASK		0x3FFF	/* Interrupt Delay Time */
+#define IOAT_INTRDELAY_MASK			0x3FFF	/* Interrupt Delay Time */
 #define IOAT_INTRDELAY_COALESE_SUPPORT		0x8000	/* Interrupt Coalescing Supported */
 
 #define IOAT_DEVICE_STATUS_OFFSET		0x0E	/* 16-bit */
diff --git a/drivers/dma/ipu/ipu_idmac.c b/drivers/dma/ipu/ipu_idmac.c
index e80bae1673f..2a446397c88 100644
--- a/drivers/dma/ipu/ipu_idmac.c
+++ b/drivers/dma/ipu/ipu_idmac.c
@@ -348,6 +348,7 @@ static void ipu_ch_param_set_size(union chan_param_mem *params,
 		break;
 	case IPU_PIX_FMT_BGRA32:
 	case IPU_PIX_FMT_BGR32:
+	case IPU_PIX_FMT_ABGR32:
 		params->ip.bpp	= 0;
 		params->ip.pfs	= 4;
 		params->ip.npb	= 7;
@@ -376,20 +377,6 @@ static void ipu_ch_param_set_size(union chan_param_mem *params,
 		params->ip.wid2	= 7;		/* Blue bit width - 1 */
 		params->ip.wid3	= 7;		/* Alpha bit width - 1 */
 		break;
-	case IPU_PIX_FMT_ABGR32:
-		params->ip.bpp	= 0;
-		params->ip.pfs	= 4;
-		params->ip.npb	= 7;
-		params->ip.sat	= 2;		/* SAT = 32-bit access */
-		params->ip.ofs0	= 8;		/* Red bit offset */
-		params->ip.ofs1	= 16;		/* Green bit offset */
-		params->ip.ofs2	= 24;		/* Blue bit offset */
-		params->ip.ofs3	= 0;		/* Alpha bit offset */
-		params->ip.wid0	= 7;		/* Red bit width - 1 */
-		params->ip.wid1	= 7;		/* Green bit width - 1 */
-		params->ip.wid2	= 7;		/* Blue bit width - 1 */
-		params->ip.wid3	= 7;		/* Alpha bit width - 1 */
-		break;
 	case IPU_PIX_FMT_UYVY:
 		params->ip.bpp	= 2;
 		params->ip.pfs	= 6;
diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c
new file mode 100644
index 00000000000..3fdf1f46bd6
--- /dev/null
+++ b/drivers/dma/mpc512x_dma.c
@@ -0,0 +1,800 @@
+/*
+ * Copyright (C) Freescale Semicondutor, Inc. 2007, 2008.
+ * Copyright (C) Semihalf 2009
+ *
+ * Written by Piotr Ziecik <kosmo@semihalf.com>. Hardware description
+ * (defines, structures and comments) was taken from MPC5121 DMA driver
+ * written by Hongjun Chen <hong-jun.chen@freescale.com>.
+ *
+ * Approved as OSADL project by a majority of OSADL members and funded
+ * by OSADL membership fees in 2009;  for details see www.osadl.org.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option)
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program; if not, write to the Free Software Foundation, Inc., 59
+ * Temple Place - Suite 330, Boston, MA  02111-1307, USA.
+ *
+ * The full GNU General Public License is included in this distribution in the
+ * file called COPYING.
+ */
+
+/*
+ * This is initial version of MPC5121 DMA driver. Only memory to memory
+ * transfers are supported (tested using dmatest module).
+ */
+
+#include <linux/module.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/of_device.h>
+#include <linux/of_platform.h>
+
+#include <linux/random.h>
+
+/* Number of DMA Transfer descriptors allocated per channel */
+#define MPC_DMA_DESCRIPTORS	64
+
+/* Macro definitions */
+#define MPC_DMA_CHANNELS	64
+#define MPC_DMA_TCD_OFFSET	0x1000
+
+/* Arbitration mode of group and channel */
+#define MPC_DMA_DMACR_EDCG	(1 << 31)
+#define MPC_DMA_DMACR_ERGA	(1 << 3)
+#define MPC_DMA_DMACR_ERCA	(1 << 2)
+
+/* Error codes */
+#define MPC_DMA_DMAES_VLD	(1 << 31)
+#define MPC_DMA_DMAES_GPE	(1 << 15)
+#define MPC_DMA_DMAES_CPE	(1 << 14)
+#define MPC_DMA_DMAES_ERRCHN(err) \
+				(((err) >> 8) & 0x3f)
+#define MPC_DMA_DMAES_SAE	(1 << 7)
+#define MPC_DMA_DMAES_SOE	(1 << 6)
+#define MPC_DMA_DMAES_DAE	(1 << 5)
+#define MPC_DMA_DMAES_DOE	(1 << 4)
+#define MPC_DMA_DMAES_NCE	(1 << 3)
+#define MPC_DMA_DMAES_SGE	(1 << 2)
+#define MPC_DMA_DMAES_SBE	(1 << 1)
+#define MPC_DMA_DMAES_DBE	(1 << 0)
+
+#define MPC_DMA_TSIZE_1		0x00
+#define MPC_DMA_TSIZE_2		0x01
+#define MPC_DMA_TSIZE_4		0x02
+#define MPC_DMA_TSIZE_16	0x04
+#define MPC_DMA_TSIZE_32	0x05
+
+/* MPC5121 DMA engine registers */
+struct __attribute__ ((__packed__)) mpc_dma_regs {
+	/* 0x00 */
+	u32 dmacr;		/* DMA control register */
+	u32 dmaes;		/* DMA error status */
+	/* 0x08 */
+	u32 dmaerqh;		/* DMA enable request high(channels 63~32) */
+	u32 dmaerql;		/* DMA enable request low(channels 31~0) */
+	u32 dmaeeih;		/* DMA enable error interrupt high(ch63~32) */
+	u32 dmaeeil;		/* DMA enable error interrupt low(ch31~0) */
+	/* 0x18 */
+	u8 dmaserq;		/* DMA set enable request */
+	u8 dmacerq;		/* DMA clear enable request */
+	u8 dmaseei;		/* DMA set enable error interrupt */
+	u8 dmaceei;		/* DMA clear enable error interrupt */
+	/* 0x1c */
+	u8 dmacint;		/* DMA clear interrupt request */
+	u8 dmacerr;		/* DMA clear error */
+	u8 dmassrt;		/* DMA set start bit */
+	u8 dmacdne;		/* DMA clear DONE status bit */
+	/* 0x20 */
+	u32 dmainth;		/* DMA interrupt request high(ch63~32) */
+	u32 dmaintl;		/* DMA interrupt request low(ch31~0) */
+	u32 dmaerrh;		/* DMA error high(ch63~32) */
+	u32 dmaerrl;		/* DMA error low(ch31~0) */
+	/* 0x30 */
+	u32 dmahrsh;		/* DMA hw request status high(ch63~32) */
+	u32 dmahrsl;		/* DMA hardware request status low(ch31~0) */
+	u32 dmaihsa;		/* DMA interrupt high select AXE(ch63~32) */
+	u32 dmailsa;		/* DMA interrupt low select AXE(ch31~0) */
+	/* 0x40 ~ 0xff */
+	u32 reserve0[48];	/* Reserved */
+	/* 0x100 */
+	u8 dchpri[MPC_DMA_CHANNELS];
+	/* DMA channels(0~63) priority */
+};
+
+struct __attribute__ ((__packed__)) mpc_dma_tcd {
+	/* 0x00 */
+	u32 saddr;		/* Source address */
+
+	u32 smod:5;		/* Source address modulo */
+	u32 ssize:3;		/* Source data transfer size */
+	u32 dmod:5;		/* Destination address modulo */
+	u32 dsize:3;		/* Destination data transfer size */
+	u32 soff:16;		/* Signed source address offset */
+
+	/* 0x08 */
+	u32 nbytes;		/* Inner "minor" byte count */
+	u32 slast;		/* Last source address adjustment */
+	u32 daddr;		/* Destination address */
+
+	/* 0x14 */
+	u32 citer_elink:1;	/* Enable channel-to-channel linking on
+				 * minor loop complete
+				 */
+	u32 citer_linkch:6;	/* Link channel for minor loop complete */
+	u32 citer:9;		/* Current "major" iteration count */
+	u32 doff:16;		/* Signed destination address offset */
+
+	/* 0x18 */
+	u32 dlast_sga;		/* Last Destination address adjustment/scatter
+				 * gather address
+				 */
+
+	/* 0x1c */
+	u32 biter_elink:1;	/* Enable channel-to-channel linking on major
+				 * loop complete
+				 */
+	u32 biter_linkch:6;
+	u32 biter:9;		/* Beginning "major" iteration count */
+	u32 bwc:2;		/* Bandwidth control */
+	u32 major_linkch:6;	/* Link channel number */
+	u32 done:1;		/* Channel done */
+	u32 active:1;		/* Channel active */
+	u32 major_elink:1;	/* Enable channel-to-channel linking on major
+				 * loop complete
+				 */
+	u32 e_sg:1;		/* Enable scatter/gather processing */
+	u32 d_req:1;		/* Disable request */
+	u32 int_half:1;		/* Enable an interrupt when major counter is
+				 * half complete
+				 */
+	u32 int_maj:1;		/* Enable an interrupt when major iteration
+				 * count completes
+				 */
+	u32 start:1;		/* Channel start */
+};
+
+struct mpc_dma_desc {
+	struct dma_async_tx_descriptor	desc;
+	struct mpc_dma_tcd		*tcd;
+	dma_addr_t			tcd_paddr;
+	int				error;
+	struct list_head		node;
+};
+
+struct mpc_dma_chan {
+	struct dma_chan			chan;
+	struct list_head		free;
+	struct list_head		prepared;
+	struct list_head		queued;
+	struct list_head		active;
+	struct list_head		completed;
+	struct mpc_dma_tcd		*tcd;
+	dma_addr_t			tcd_paddr;
+	dma_cookie_t			completed_cookie;
+
+	/* Lock for this structure */
+	spinlock_t			lock;
+};
+
+struct mpc_dma {
+	struct dma_device		dma;
+	struct tasklet_struct		tasklet;
+	struct mpc_dma_chan		channels[MPC_DMA_CHANNELS];
+	struct mpc_dma_regs __iomem	*regs;
+	struct mpc_dma_tcd __iomem	*tcd;
+	int				irq;
+	uint				error_status;
+
+	/* Lock for error_status field in this structure */
+	spinlock_t			error_status_lock;
+};
+
+#define DRV_NAME	"mpc512x_dma"
+
+/* Convert struct dma_chan to struct mpc_dma_chan */
+static inline struct mpc_dma_chan *dma_chan_to_mpc_dma_chan(struct dma_chan *c)
+{
+	return container_of(c, struct mpc_dma_chan, chan);
+}
+
+/* Convert struct dma_chan to struct mpc_dma */
+static inline struct mpc_dma *dma_chan_to_mpc_dma(struct dma_chan *c)
+{
+	struct mpc_dma_chan *mchan = dma_chan_to_mpc_dma_chan(c);
+	return container_of(mchan, struct mpc_dma, channels[c->chan_id]);
+}
+
+/*
+ * Execute all queued DMA descriptors.
+ *
+ * Following requirements must be met while calling mpc_dma_execute():
+ * 	a) mchan->lock is acquired,
+ * 	b) mchan->active list is empty,
+ * 	c) mchan->queued list contains at least one entry.
+ */
+static void mpc_dma_execute(struct mpc_dma_chan *mchan)
+{
+	struct mpc_dma *mdma = dma_chan_to_mpc_dma(&mchan->chan);
+	struct mpc_dma_desc *first = NULL;
+	struct mpc_dma_desc *prev = NULL;
+	struct mpc_dma_desc *mdesc;
+	int cid = mchan->chan.chan_id;
+
+	/* Move all queued descriptors to active list */
+	list_splice_tail_init(&mchan->queued, &mchan->active);
+
+	/* Chain descriptors into one transaction */
+	list_for_each_entry(mdesc, &mchan->active, node) {
+		if (!first)
+			first = mdesc;
+
+		if (!prev) {
+			prev = mdesc;
+			continue;
+		}
+
+		prev->tcd->dlast_sga = mdesc->tcd_paddr;
+		prev->tcd->e_sg = 1;
+		mdesc->tcd->start = 1;
+
+		prev = mdesc;
+	}
+
+	prev->tcd->start = 0;
+	prev->tcd->int_maj = 1;
+
+	/* Send first descriptor in chain into hardware */
+	memcpy_toio(&mdma->tcd[cid], first->tcd, sizeof(struct mpc_dma_tcd));
+	out_8(&mdma->regs->dmassrt, cid);
+}
+
+/* Handle interrupt on one half of DMA controller (32 channels) */
+static void mpc_dma_irq_process(struct mpc_dma *mdma, u32 is, u32 es, int off)
+{
+	struct mpc_dma_chan *mchan;
+	struct mpc_dma_desc *mdesc;
+	u32 status = is | es;
+	int ch;
+
+	while ((ch = fls(status) - 1) >= 0) {
+		status &= ~(1 << ch);
+		mchan = &mdma->channels[ch + off];
+
+		spin_lock(&mchan->lock);
+
+		/* Check error status */
+		if (es & (1 << ch))
+			list_for_each_entry(mdesc, &mchan->active, node)
+				mdesc->error = -EIO;
+
+		/* Execute queued descriptors */
+		list_splice_tail_init(&mchan->active, &mchan->completed);
+		if (!list_empty(&mchan->queued))
+			mpc_dma_execute(mchan);
+
+		spin_unlock(&mchan->lock);
+	}
+}
+
+/* Interrupt handler */
+static irqreturn_t mpc_dma_irq(int irq, void *data)
+{
+	struct mpc_dma *mdma = data;
+	uint es;
+
+	/* Save error status register */
+	es = in_be32(&mdma->regs->dmaes);
+	spin_lock(&mdma->error_status_lock);
+	if ((es & MPC_DMA_DMAES_VLD) && mdma->error_status == 0)
+		mdma->error_status = es;
+	spin_unlock(&mdma->error_status_lock);
+
+	/* Handle interrupt on each channel */
+	mpc_dma_irq_process(mdma, in_be32(&mdma->regs->dmainth),
+					in_be32(&mdma->regs->dmaerrh), 32);
+	mpc_dma_irq_process(mdma, in_be32(&mdma->regs->dmaintl),
+					in_be32(&mdma->regs->dmaerrl), 0);
+
+	/* Ack interrupt on all channels */
+	out_be32(&mdma->regs->dmainth, 0xFFFFFFFF);
+	out_be32(&mdma->regs->dmaintl, 0xFFFFFFFF);
+	out_be32(&mdma->regs->dmaerrh, 0xFFFFFFFF);
+	out_be32(&mdma->regs->dmaerrl, 0xFFFFFFFF);
+
+	/* Schedule tasklet */
+	tasklet_schedule(&mdma->tasklet);
+
+	return IRQ_HANDLED;
+}
+
+/* DMA Tasklet */
+static void mpc_dma_tasklet(unsigned long data)
+{
+	struct mpc_dma *mdma = (void *)data;
+	dma_cookie_t last_cookie = 0;
+	struct mpc_dma_chan *mchan;
+	struct mpc_dma_desc *mdesc;
+	struct dma_async_tx_descriptor *desc;
+	unsigned long flags;
+	LIST_HEAD(list);
+	uint es;
+	int i;
+
+	spin_lock_irqsave(&mdma->error_status_lock, flags);
+	es = mdma->error_status;
+	mdma->error_status = 0;
+	spin_unlock_irqrestore(&mdma->error_status_lock, flags);
+
+	/* Print nice error report */
+	if (es) {
+		dev_err(mdma->dma.dev,
+			"Hardware reported following error(s) on channel %u:\n",
+						      MPC_DMA_DMAES_ERRCHN(es));
+
+		if (es & MPC_DMA_DMAES_GPE)
+			dev_err(mdma->dma.dev, "- Group Priority Error\n");
+		if (es & MPC_DMA_DMAES_CPE)
+			dev_err(mdma->dma.dev, "- Channel Priority Error\n");
+		if (es & MPC_DMA_DMAES_SAE)
+			dev_err(mdma->dma.dev, "- Source Address Error\n");
+		if (es & MPC_DMA_DMAES_SOE)
+			dev_err(mdma->dma.dev, "- Source Offset"
+						" Configuration Error\n");
+		if (es & MPC_DMA_DMAES_DAE)
+			dev_err(mdma->dma.dev, "- Destination Address"
+								" Error\n");
+		if (es & MPC_DMA_DMAES_DOE)
+			dev_err(mdma->dma.dev, "- Destination Offset"
+						" Configuration Error\n");
+		if (es & MPC_DMA_DMAES_NCE)
+			dev_err(mdma->dma.dev, "- NBytes/Citter"
+						" Configuration Error\n");
+		if (es & MPC_DMA_DMAES_SGE)
+			dev_err(mdma->dma.dev, "- Scatter/Gather"
+						" Configuration Error\n");
+		if (es & MPC_DMA_DMAES_SBE)
+			dev_err(mdma->dma.dev, "- Source Bus Error\n");
+		if (es & MPC_DMA_DMAES_DBE)
+			dev_err(mdma->dma.dev, "- Destination Bus Error\n");
+	}
+
+	for (i = 0; i < mdma->dma.chancnt; i++) {
+		mchan = &mdma->channels[i];
+
+		/* Get all completed descriptors */
+		spin_lock_irqsave(&mchan->lock, flags);
+		if (!list_empty(&mchan->completed))
+			list_splice_tail_init(&mchan->completed, &list);
+		spin_unlock_irqrestore(&mchan->lock, flags);
+
+		if (list_empty(&list))
+			continue;
+
+		/* Execute callbacks and run dependencies */
+		list_for_each_entry(mdesc, &list, node) {
+			desc = &mdesc->desc;
+
+			if (desc->callback)
+				desc->callback(desc->callback_param);
+
+			last_cookie = desc->cookie;
+			dma_run_dependencies(desc);
+		}
+
+		/* Free descriptors */
+		spin_lock_irqsave(&mchan->lock, flags);
+		list_splice_tail_init(&list, &mchan->free);
+		mchan->completed_cookie = last_cookie;
+		spin_unlock_irqrestore(&mchan->lock, flags);
+	}
+}
+
+/* Submit descriptor to hardware */
+static dma_cookie_t mpc_dma_tx_submit(struct dma_async_tx_descriptor *txd)
+{
+	struct mpc_dma_chan *mchan = dma_chan_to_mpc_dma_chan(txd->chan);
+	struct mpc_dma_desc *mdesc;
+	unsigned long flags;
+	dma_cookie_t cookie;
+
+	mdesc = container_of(txd, struct mpc_dma_desc, desc);
+
+	spin_lock_irqsave(&mchan->lock, flags);
+
+	/* Move descriptor to queue */
+	list_move_tail(&mdesc->node, &mchan->queued);
+
+	/* If channel is idle, execute all queued descriptors */
+	if (list_empty(&mchan->active))
+		mpc_dma_execute(mchan);
+
+	/* Update cookie */
+	cookie = mchan->chan.cookie + 1;
+	if (cookie <= 0)
+		cookie = 1;
+
+	mchan->chan.cookie = cookie;
+	mdesc->desc.cookie = cookie;
+
+	spin_unlock_irqrestore(&mchan->lock, flags);
+
+	return cookie;
+}
+
+/* Alloc channel resources */
+static int mpc_dma_alloc_chan_resources(struct dma_chan *chan)
+{
+	struct mpc_dma *mdma = dma_chan_to_mpc_dma(chan);
+	struct mpc_dma_chan *mchan = dma_chan_to_mpc_dma_chan(chan);
+	struct mpc_dma_desc *mdesc;
+	struct mpc_dma_tcd *tcd;
+	dma_addr_t tcd_paddr;
+	unsigned long flags;
+	LIST_HEAD(descs);
+	int i;
+
+	/* Alloc DMA memory for Transfer Control Descriptors */
+	tcd = dma_alloc_coherent(mdma->dma.dev,
+			MPC_DMA_DESCRIPTORS * sizeof(struct mpc_dma_tcd),
+							&tcd_paddr, GFP_KERNEL);
+	if (!tcd)
+		return -ENOMEM;
+
+	/* Alloc descriptors for this channel */
+	for (i = 0; i < MPC_DMA_DESCRIPTORS; i++) {
+		mdesc = kzalloc(sizeof(struct mpc_dma_desc), GFP_KERNEL);
+		if (!mdesc) {
+			dev_notice(mdma->dma.dev, "Memory allocation error. "
+					"Allocated only %u descriptors\n", i);
+			break;
+		}
+
+		dma_async_tx_descriptor_init(&mdesc->desc, chan);
+		mdesc->desc.flags = DMA_CTRL_ACK;
+		mdesc->desc.tx_submit = mpc_dma_tx_submit;
+
+		mdesc->tcd = &tcd[i];
+		mdesc->tcd_paddr = tcd_paddr + (i * sizeof(struct mpc_dma_tcd));
+
+		list_add_tail(&mdesc->node, &descs);
+	}
+
+	/* Return error only if no descriptors were allocated */
+	if (i == 0) {
+		dma_free_coherent(mdma->dma.dev,
+			MPC_DMA_DESCRIPTORS * sizeof(struct mpc_dma_tcd),
+								tcd, tcd_paddr);
+		return -ENOMEM;
+	}
+
+	spin_lock_irqsave(&mchan->lock, flags);
+	mchan->tcd = tcd;
+	mchan->tcd_paddr = tcd_paddr;
+	list_splice_tail_init(&descs, &mchan->free);
+	spin_unlock_irqrestore(&mchan->lock, flags);
+
+	/* Enable Error Interrupt */
+	out_8(&mdma->regs->dmaseei, chan->chan_id);
+
+	return 0;
+}
+
+/* Free channel resources */
+static void mpc_dma_free_chan_resources(struct dma_chan *chan)
+{
+	struct mpc_dma *mdma = dma_chan_to_mpc_dma(chan);
+	struct mpc_dma_chan *mchan = dma_chan_to_mpc_dma_chan(chan);
+	struct mpc_dma_desc *mdesc, *tmp;
+	struct mpc_dma_tcd *tcd;
+	dma_addr_t tcd_paddr;
+	unsigned long flags;
+	LIST_HEAD(descs);
+
+	spin_lock_irqsave(&mchan->lock, flags);
+
+	/* Channel must be idle */
+	BUG_ON(!list_empty(&mchan->prepared));
+	BUG_ON(!list_empty(&mchan->queued));
+	BUG_ON(!list_empty(&mchan->active));
+	BUG_ON(!list_empty(&mchan->completed));
+
+	/* Move data */
+	list_splice_tail_init(&mchan->free, &descs);
+	tcd = mchan->tcd;
+	tcd_paddr = mchan->tcd_paddr;
+
+	spin_unlock_irqrestore(&mchan->lock, flags);
+
+	/* Free DMA memory used by descriptors */
+	dma_free_coherent(mdma->dma.dev,
+			MPC_DMA_DESCRIPTORS * sizeof(struct mpc_dma_tcd),
+								tcd, tcd_paddr);
+
+	/* Free descriptors */
+	list_for_each_entry_safe(mdesc, tmp, &descs, node)
+		kfree(mdesc);
+
+	/* Disable Error Interrupt */
+	out_8(&mdma->regs->dmaceei, chan->chan_id);
+}
+
+/* Send all pending descriptor to hardware */
+static void mpc_dma_issue_pending(struct dma_chan *chan)
+{
+	/*
+	 * We are posting descriptors to the hardware as soon as
+	 * they are ready, so this function does nothing.
+	 */
+}
+
+/* Check request completion status */
+static enum dma_status
+mpc_dma_is_tx_complete(struct dma_chan *chan, dma_cookie_t cookie,
+					dma_cookie_t *done, dma_cookie_t *used)
+{
+	struct mpc_dma_chan *mchan = dma_chan_to_mpc_dma_chan(chan);
+	unsigned long flags;
+	dma_cookie_t last_used;
+	dma_cookie_t last_complete;
+
+	spin_lock_irqsave(&mchan->lock, flags);
+	last_used = mchan->chan.cookie;
+	last_complete = mchan->completed_cookie;
+	spin_unlock_irqrestore(&mchan->lock, flags);
+
+	if (done)
+		*done = last_complete;
+
+	if (used)
+		*used = last_used;
+
+	return dma_async_is_complete(cookie, last_complete, last_used);
+}
+
+/* Prepare descriptor for memory to memory copy */
+static struct dma_async_tx_descriptor *
+mpc_dma_prep_memcpy(struct dma_chan *chan, dma_addr_t dst, dma_addr_t src,
+					size_t len, unsigned long flags)
+{
+	struct mpc_dma_chan *mchan = dma_chan_to_mpc_dma_chan(chan);
+	struct mpc_dma_desc *mdesc = NULL;
+	struct mpc_dma_tcd *tcd;
+	unsigned long iflags;
+
+	/* Get free descriptor */
+	spin_lock_irqsave(&mchan->lock, iflags);
+	if (!list_empty(&mchan->free)) {
+		mdesc = list_first_entry(&mchan->free, struct mpc_dma_desc,
+									node);
+		list_del(&mdesc->node);
+	}
+	spin_unlock_irqrestore(&mchan->lock, iflags);
+
+	if (!mdesc)
+		return NULL;
+
+	mdesc->error = 0;
+	tcd = mdesc->tcd;
+
+	/* Prepare Transfer Control Descriptor for this transaction */
+	memset(tcd, 0, sizeof(struct mpc_dma_tcd));
+
+	if (IS_ALIGNED(src | dst | len, 32)) {
+		tcd->ssize = MPC_DMA_TSIZE_32;
+		tcd->dsize = MPC_DMA_TSIZE_32;
+		tcd->soff = 32;
+		tcd->doff = 32;
+	} else if (IS_ALIGNED(src | dst | len, 16)) {
+		tcd->ssize = MPC_DMA_TSIZE_16;
+		tcd->dsize = MPC_DMA_TSIZE_16;
+		tcd->soff = 16;
+		tcd->doff = 16;
+	} else if (IS_ALIGNED(src | dst | len, 4)) {
+		tcd->ssize = MPC_DMA_TSIZE_4;
+		tcd->dsize = MPC_DMA_TSIZE_4;
+		tcd->soff = 4;
+		tcd->doff = 4;
+	} else if (IS_ALIGNED(src | dst | len, 2)) {
+		tcd->ssize = MPC_DMA_TSIZE_2;
+		tcd->dsize = MPC_DMA_TSIZE_2;
+		tcd->soff = 2;
+		tcd->doff = 2;
+	} else {
+		tcd->ssize = MPC_DMA_TSIZE_1;
+		tcd->dsize = MPC_DMA_TSIZE_1;
+		tcd->soff = 1;
+		tcd->doff = 1;
+	}
+
+	tcd->saddr = src;
+	tcd->daddr = dst;
+	tcd->nbytes = len;
+	tcd->biter = 1;
+	tcd->citer = 1;
+
+	/* Place descriptor in prepared list */
+	spin_lock_irqsave(&mchan->lock, iflags);
+	list_add_tail(&mdesc->node, &mchan->prepared);
+	spin_unlock_irqrestore(&mchan->lock, iflags);
+
+	return &mdesc->desc;
+}
+
+static int __devinit mpc_dma_probe(struct of_device *op,
+					const struct of_device_id *match)
+{
+	struct device_node *dn = op->node;
+	struct device *dev = &op->dev;
+	struct dma_device *dma;
+	struct mpc_dma *mdma;
+	struct mpc_dma_chan *mchan;
+	struct resource res;
+	ulong regs_start, regs_size;
+	int retval, i;
+
+	mdma = devm_kzalloc(dev, sizeof(struct mpc_dma), GFP_KERNEL);
+	if (!mdma) {
+		dev_err(dev, "Memory exhausted!\n");
+		return -ENOMEM;
+	}
+
+	mdma->irq = irq_of_parse_and_map(dn, 0);
+	if (mdma->irq == NO_IRQ) {
+		dev_err(dev, "Error mapping IRQ!\n");
+		return -EINVAL;
+	}
+
+	retval = of_address_to_resource(dn, 0, &res);
+	if (retval) {
+		dev_err(dev, "Error parsing memory region!\n");
+		return retval;
+	}
+
+	regs_start = res.start;
+	regs_size = res.end - res.start + 1;
+
+	if (!devm_request_mem_region(dev, regs_start, regs_size, DRV_NAME)) {
+		dev_err(dev, "Error requesting memory region!\n");
+		return -EBUSY;
+	}
+
+	mdma->regs = devm_ioremap(dev, regs_start, regs_size);
+	if (!mdma->regs) {
+		dev_err(dev, "Error mapping memory region!\n");
+		return -ENOMEM;
+	}
+
+	mdma->tcd = (struct mpc_dma_tcd *)((u8 *)(mdma->regs)
+							+ MPC_DMA_TCD_OFFSET);
+
+	retval = devm_request_irq(dev, mdma->irq, &mpc_dma_irq, 0, DRV_NAME,
+									mdma);
+	if (retval) {
+		dev_err(dev, "Error requesting IRQ!\n");
+		return -EINVAL;
+	}
+
+	spin_lock_init(&mdma->error_status_lock);
+
+	dma = &mdma->dma;
+	dma->dev = dev;
+	dma->chancnt = MPC_DMA_CHANNELS;
+	dma->device_alloc_chan_resources = mpc_dma_alloc_chan_resources;
+	dma->device_free_chan_resources = mpc_dma_free_chan_resources;
+	dma->device_issue_pending = mpc_dma_issue_pending;
+	dma->device_is_tx_complete = mpc_dma_is_tx_complete;
+	dma->device_prep_dma_memcpy = mpc_dma_prep_memcpy;
+
+	INIT_LIST_HEAD(&dma->channels);
+	dma_cap_set(DMA_MEMCPY, dma->cap_mask);
+
+	for (i = 0; i < dma->chancnt; i++) {
+		mchan = &mdma->channels[i];
+
+		mchan->chan.device = dma;
+		mchan->chan.chan_id = i;
+		mchan->chan.cookie = 1;
+		mchan->completed_cookie = mchan->chan.cookie;
+
+		INIT_LIST_HEAD(&mchan->free);
+		INIT_LIST_HEAD(&mchan->prepared);
+		INIT_LIST_HEAD(&mchan->queued);
+		INIT_LIST_HEAD(&mchan->active);
+		INIT_LIST_HEAD(&mchan->completed);
+
+		spin_lock_init(&mchan->lock);
+		list_add_tail(&mchan->chan.device_node, &dma->channels);
+	}
+
+	tasklet_init(&mdma->tasklet, mpc_dma_tasklet, (unsigned long)mdma);
+
+	/*
+	 * Configure DMA Engine:
+	 * - Dynamic clock,
+	 * - Round-robin group arbitration,
+	 * - Round-robin channel arbitration.
+	 */
+	out_be32(&mdma->regs->dmacr, MPC_DMA_DMACR_EDCG |
+				MPC_DMA_DMACR_ERGA | MPC_DMA_DMACR_ERCA);
+
+	/* Disable hardware DMA requests */
+	out_be32(&mdma->regs->dmaerqh, 0);
+	out_be32(&mdma->regs->dmaerql, 0);
+
+	/* Disable error interrupts */
+	out_be32(&mdma->regs->dmaeeih, 0);
+	out_be32(&mdma->regs->dmaeeil, 0);
+
+	/* Clear interrupts status */
+	out_be32(&mdma->regs->dmainth, 0xFFFFFFFF);
+	out_be32(&mdma->regs->dmaintl, 0xFFFFFFFF);
+	out_be32(&mdma->regs->dmaerrh, 0xFFFFFFFF);
+	out_be32(&mdma->regs->dmaerrl, 0xFFFFFFFF);
+
+	/* Route interrupts to IPIC */
+	out_be32(&mdma->regs->dmaihsa, 0);
+	out_be32(&mdma->regs->dmailsa, 0);
+
+	/* Register DMA engine */
+	dev_set_drvdata(dev, mdma);
+	retval = dma_async_device_register(dma);
+	if (retval) {
+		devm_free_irq(dev, mdma->irq, mdma);
+		irq_dispose_mapping(mdma->irq);
+	}
+
+	return retval;
+}
+
+static int __devexit mpc_dma_remove(struct of_device *op)
+{
+	struct device *dev = &op->dev;
+	struct mpc_dma *mdma = dev_get_drvdata(dev);
+
+	dma_async_device_unregister(&mdma->dma);
+	devm_free_irq(dev, mdma->irq, mdma);
+	irq_dispose_mapping(mdma->irq);
+
+	return 0;
+}
+
+static struct of_device_id mpc_dma_match[] = {
+	{ .compatible = "fsl,mpc5121-dma", },
+	{},
+};
+
+static struct of_platform_driver mpc_dma_driver = {
+	.match_table	= mpc_dma_match,
+	.probe		= mpc_dma_probe,
+	.remove		= __devexit_p(mpc_dma_remove),
+	.driver		= {
+		.name	= DRV_NAME,
+		.owner	= THIS_MODULE,
+	},
+};
+
+static int __init mpc_dma_init(void)
+{
+	return of_register_platform_driver(&mpc_dma_driver);
+}
+module_init(mpc_dma_init);
+
+static void __exit mpc_dma_exit(void)
+{
+	of_unregister_platform_driver(&mpc_dma_driver);
+}
+module_exit(mpc_dma_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Piotr Ziecik <kosmo@semihalf.com>");
diff --git a/drivers/dma/ppc4xx/adma.c b/drivers/dma/ppc4xx/adma.c
index 0a3478e910f..e69d87f24a2 100644
--- a/drivers/dma/ppc4xx/adma.c
+++ b/drivers/dma/ppc4xx/adma.c
@@ -4940,7 +4940,7 @@ out_free:
 	return ret;
 }
 
-static struct of_device_id __devinitdata ppc440spe_adma_of_match[] = {
+static const struct of_device_id ppc440spe_adma_of_match[] __devinitconst = {
 	{ .compatible	= "ibm,dma-440spe", },
 	{ .compatible	= "amcc,xor-accelerator", },
 	{},
diff --git a/drivers/dma/shdma.c b/drivers/dma/shdma.c
index d10cc899c46..5d17e09cb62 100644
--- a/drivers/dma/shdma.c
+++ b/drivers/dma/shdma.c
@@ -24,8 +24,10 @@
 #include <linux/delay.h>
 #include <linux/dma-mapping.h>
 #include <linux/platform_device.h>
-#include <cpu/dma.h>
-#include <asm/dma-sh.h>
+#include <linux/pm_runtime.h>
+
+#include <asm/dmaengine.h>
+
 #include "shdma.h"
 
 /* DMA descriptor control */
@@ -38,33 +40,32 @@ enum sh_dmae_desc_status {
 };
 
 #define NR_DESCS_PER_CHANNEL 32
-/*
- * Define the default configuration for dual address memory-memory transfer.
- * The 0x400 value represents auto-request, external->external.
- *
- * And this driver set 4byte burst mode.
- * If you want to change mode, you need to change RS_DEFAULT of value.
- * (ex 1byte burst mode -> (RS_DUAL & ~TS_32)
- */
-#define RS_DEFAULT  (RS_DUAL)
+/* Default MEMCPY transfer size = 2^2 = 4 bytes */
+#define LOG2_DEFAULT_XFER_SIZE	2
+
+/* A bitmask with bits enough for enum sh_dmae_slave_chan_id */
+static unsigned long sh_dmae_slave_used[BITS_TO_LONGS(SHDMA_SLAVE_NUMBER)];
 
 static void sh_dmae_chan_ld_cleanup(struct sh_dmae_chan *sh_chan, bool all);
 
-#define SH_DMAC_CHAN_BASE(id) (dma_base_addr[id])
 static void sh_dmae_writel(struct sh_dmae_chan *sh_dc, u32 data, u32 reg)
 {
-	ctrl_outl(data, (SH_DMAC_CHAN_BASE(sh_dc->id) + reg));
+	__raw_writel(data, sh_dc->base + reg / sizeof(u32));
 }
 
 static u32 sh_dmae_readl(struct sh_dmae_chan *sh_dc, u32 reg)
 {
-	return ctrl_inl((SH_DMAC_CHAN_BASE(sh_dc->id) + reg));
+	return __raw_readl(sh_dc->base + reg / sizeof(u32));
 }
 
-static void dmae_init(struct sh_dmae_chan *sh_chan)
+static u16 dmaor_read(struct sh_dmae_device *shdev)
 {
-	u32 chcr = RS_DEFAULT; /* default is DUAL mode */
-	sh_dmae_writel(sh_chan, chcr, CHCR);
+	return __raw_readw(shdev->chan_reg + DMAOR / sizeof(u32));
+}
+
+static void dmaor_write(struct sh_dmae_device *shdev, u16 data)
+{
+	__raw_writew(data, shdev->chan_reg + DMAOR / sizeof(u32));
 }
 
 /*
@@ -72,50 +73,75 @@ static void dmae_init(struct sh_dmae_chan *sh_chan)
  *
  * SH7780 has two DMAOR register
  */
-static void sh_dmae_ctl_stop(int id)
+static void sh_dmae_ctl_stop(struct sh_dmae_device *shdev)
 {
-	unsigned short dmaor = dmaor_read_reg(id);
+	unsigned short dmaor = dmaor_read(shdev);
 
-	dmaor &= ~(DMAOR_NMIF | DMAOR_AE);
-	dmaor_write_reg(id, dmaor);
+	dmaor_write(shdev, dmaor & ~(DMAOR_NMIF | DMAOR_AE | DMAOR_DME));
 }
 
-static int sh_dmae_rst(int id)
+static int sh_dmae_rst(struct sh_dmae_device *shdev)
 {
 	unsigned short dmaor;
 
-	sh_dmae_ctl_stop(id);
-	dmaor = dmaor_read_reg(id) | DMAOR_INIT;
+	sh_dmae_ctl_stop(shdev);
+	dmaor = dmaor_read(shdev) | shdev->pdata->dmaor_init;
 
-	dmaor_write_reg(id, dmaor);
-	if (dmaor_read_reg(id) & (DMAOR_AE | DMAOR_NMIF)) {
-		pr_warning(KERN_ERR "dma-sh: Can't initialize DMAOR.\n");
+	dmaor_write(shdev, dmaor);
+	if (dmaor_read(shdev) & (DMAOR_AE | DMAOR_NMIF)) {
+		pr_warning("dma-sh: Can't initialize DMAOR.\n");
 		return -EINVAL;
 	}
 	return 0;
 }
 
-static int dmae_is_busy(struct sh_dmae_chan *sh_chan)
+static bool dmae_is_busy(struct sh_dmae_chan *sh_chan)
 {
 	u32 chcr = sh_dmae_readl(sh_chan, CHCR);
-	if (chcr & CHCR_DE) {
-		if (!(chcr & CHCR_TE))
-			return -EBUSY; /* working */
-	}
-	return 0; /* waiting */
+
+	if ((chcr & (CHCR_DE | CHCR_TE)) == CHCR_DE)
+		return true; /* working */
+
+	return false; /* waiting */
 }
 
-static inline unsigned int calc_xmit_shift(struct sh_dmae_chan *sh_chan)
+static unsigned int calc_xmit_shift(struct sh_dmae_chan *sh_chan, u32 chcr)
 {
-	u32 chcr = sh_dmae_readl(sh_chan, CHCR);
-	return ts_shift[(chcr & CHCR_TS_MASK) >> CHCR_TS_SHIFT];
+	struct sh_dmae_device *shdev = container_of(sh_chan->common.device,
+						struct sh_dmae_device, common);
+	struct sh_dmae_pdata *pdata = shdev->pdata;
+	int cnt = ((chcr & pdata->ts_low_mask) >> pdata->ts_low_shift) |
+		((chcr & pdata->ts_high_mask) >> pdata->ts_high_shift);
+
+	if (cnt >= pdata->ts_shift_num)
+		cnt = 0;
+
+	return pdata->ts_shift[cnt];
+}
+
+static u32 log2size_to_chcr(struct sh_dmae_chan *sh_chan, int l2size)
+{
+	struct sh_dmae_device *shdev = container_of(sh_chan->common.device,
+						struct sh_dmae_device, common);
+	struct sh_dmae_pdata *pdata = shdev->pdata;
+	int i;
+
+	for (i = 0; i < pdata->ts_shift_num; i++)
+		if (pdata->ts_shift[i] == l2size)
+			break;
+
+	if (i == pdata->ts_shift_num)
+		i = 0;
+
+	return ((i << pdata->ts_low_shift) & pdata->ts_low_mask) |
+		((i << pdata->ts_high_shift) & pdata->ts_high_mask);
 }
 
 static void dmae_set_reg(struct sh_dmae_chan *sh_chan, struct sh_dmae_regs *hw)
 {
 	sh_dmae_writel(sh_chan, hw->sar, SAR);
 	sh_dmae_writel(sh_chan, hw->dar, DAR);
-	sh_dmae_writel(sh_chan, hw->tcr >> calc_xmit_shift(sh_chan), TCR);
+	sh_dmae_writel(sh_chan, hw->tcr >> sh_chan->xmit_shift, TCR);
 }
 
 static void dmae_start(struct sh_dmae_chan *sh_chan)
@@ -123,7 +149,7 @@ static void dmae_start(struct sh_dmae_chan *sh_chan)
 	u32 chcr = sh_dmae_readl(sh_chan, CHCR);
 
 	chcr |= CHCR_DE | CHCR_IE;
-	sh_dmae_writel(sh_chan, chcr, CHCR);
+	sh_dmae_writel(sh_chan, chcr & ~CHCR_TE, CHCR);
 }
 
 static void dmae_halt(struct sh_dmae_chan *sh_chan)
@@ -134,55 +160,44 @@ static void dmae_halt(struct sh_dmae_chan *sh_chan)
 	sh_dmae_writel(sh_chan, chcr, CHCR);
 }
 
+static void dmae_init(struct sh_dmae_chan *sh_chan)
+{
+	/*
+	 * Default configuration for dual address memory-memory transfer.
+	 * 0x400 represents auto-request.
+	 */
+	u32 chcr = DM_INC | SM_INC | 0x400 | log2size_to_chcr(sh_chan,
+						   LOG2_DEFAULT_XFER_SIZE);
+	sh_chan->xmit_shift = calc_xmit_shift(sh_chan, chcr);
+	sh_dmae_writel(sh_chan, chcr, CHCR);
+}
+
 static int dmae_set_chcr(struct sh_dmae_chan *sh_chan, u32 val)
 {
-	int ret = dmae_is_busy(sh_chan);
 	/* When DMA was working, can not set data to CHCR */
-	if (ret)
-		return ret;
+	if (dmae_is_busy(sh_chan))
+		return -EBUSY;
 
+	sh_chan->xmit_shift = calc_xmit_shift(sh_chan, val);
 	sh_dmae_writel(sh_chan, val, CHCR);
+
 	return 0;
 }
 
-#define DMARS1_ADDR	0x04
-#define DMARS2_ADDR	0x08
-#define DMARS_SHIFT 8
-#define DMARS_CHAN_MSK 0x01
 static int dmae_set_dmars(struct sh_dmae_chan *sh_chan, u16 val)
 {
-	u32 addr;
-	int shift = 0;
-	int ret = dmae_is_busy(sh_chan);
-	if (ret)
-		return ret;
-
-	if (sh_chan->id & DMARS_CHAN_MSK)
-		shift = DMARS_SHIFT;
-
-	switch (sh_chan->id) {
-	/* DMARS0 */
-	case 0:
-	case 1:
-		addr = SH_DMARS_BASE;
-		break;
-	/* DMARS1 */
-	case 2:
-	case 3:
-		addr = (SH_DMARS_BASE + DMARS1_ADDR);
-		break;
-	/* DMARS2 */
-	case 4:
-	case 5:
-		addr = (SH_DMARS_BASE + DMARS2_ADDR);
-		break;
-	default:
-		return -EINVAL;
-	}
+	struct sh_dmae_device *shdev = container_of(sh_chan->common.device,
+						struct sh_dmae_device, common);
+	struct sh_dmae_pdata *pdata = shdev->pdata;
+	struct sh_dmae_channel *chan_pdata = &pdata->channel[sh_chan->id];
+	u16 __iomem *addr = shdev->dmars + chan_pdata->dmars / sizeof(u16);
+	int shift = chan_pdata->dmars_bit;
+
+	if (dmae_is_busy(sh_chan))
+		return -EBUSY;
 
-	ctrl_outw((val << shift) |
-		(ctrl_inw(addr) & (shift ? 0xFF00 : 0x00FF)),
-		addr);
+	__raw_writew((__raw_readw(addr) & (0xff00 >> shift)) | (val << shift),
+		     addr);
 
 	return 0;
 }
@@ -250,10 +265,54 @@ static struct sh_desc *sh_dmae_get_desc(struct sh_dmae_chan *sh_chan)
 	return NULL;
 }
 
+static struct sh_dmae_slave_config *sh_dmae_find_slave(
+	struct sh_dmae_chan *sh_chan, enum sh_dmae_slave_chan_id slave_id)
+{
+	struct dma_device *dma_dev = sh_chan->common.device;
+	struct sh_dmae_device *shdev = container_of(dma_dev,
+					struct sh_dmae_device, common);
+	struct sh_dmae_pdata *pdata = shdev->pdata;
+	int i;
+
+	if ((unsigned)slave_id >= SHDMA_SLAVE_NUMBER)
+		return NULL;
+
+	for (i = 0; i < pdata->slave_num; i++)
+		if (pdata->slave[i].slave_id == slave_id)
+			return pdata->slave + i;
+
+	return NULL;
+}
+
 static int sh_dmae_alloc_chan_resources(struct dma_chan *chan)
 {
 	struct sh_dmae_chan *sh_chan = to_sh_chan(chan);
 	struct sh_desc *desc;
+	struct sh_dmae_slave *param = chan->private;
+
+	pm_runtime_get_sync(sh_chan->dev);
+
+	/*
+	 * This relies on the guarantee from dmaengine that alloc_chan_resources
+	 * never runs concurrently with itself or free_chan_resources.
+	 */
+	if (param) {
+		struct sh_dmae_slave_config *cfg;
+
+		cfg = sh_dmae_find_slave(sh_chan, param->slave_id);
+		if (!cfg)
+			return -EINVAL;
+
+		if (test_and_set_bit(param->slave_id, sh_dmae_slave_used))
+			return -EBUSY;
+
+		param->config = cfg;
+
+		dmae_set_dmars(sh_chan, cfg->mid_rid);
+		dmae_set_chcr(sh_chan, cfg->chcr);
+	} else if ((sh_dmae_readl(sh_chan, CHCR) & 0xf00) != 0x400) {
+		dmae_init(sh_chan);
+	}
 
 	spin_lock_bh(&sh_chan->desc_lock);
 	while (sh_chan->descs_allocated < NR_DESCS_PER_CHANNEL) {
@@ -274,6 +333,9 @@ static int sh_dmae_alloc_chan_resources(struct dma_chan *chan)
 	}
 	spin_unlock_bh(&sh_chan->desc_lock);
 
+	if (!sh_chan->descs_allocated)
+		pm_runtime_put(sh_chan->dev);
+
 	return sh_chan->descs_allocated;
 }
 
@@ -285,11 +347,20 @@ static void sh_dmae_free_chan_resources(struct dma_chan *chan)
 	struct sh_dmae_chan *sh_chan = to_sh_chan(chan);
 	struct sh_desc *desc, *_desc;
 	LIST_HEAD(list);
+	int descs = sh_chan->descs_allocated;
+
+	dmae_halt(sh_chan);
 
 	/* Prepared and not submitted descriptors can still be on the queue */
 	if (!list_empty(&sh_chan->ld_queue))
 		sh_dmae_chan_ld_cleanup(sh_chan, true);
 
+	if (chan->private) {
+		/* The caller is holding dma_list_mutex */
+		struct sh_dmae_slave *param = chan->private;
+		clear_bit(param->slave_id, sh_dmae_slave_used);
+	}
+
 	spin_lock_bh(&sh_chan->desc_lock);
 
 	list_splice_init(&sh_chan->ld_free, &list);
@@ -297,27 +368,104 @@ static void sh_dmae_free_chan_resources(struct dma_chan *chan)
 
 	spin_unlock_bh(&sh_chan->desc_lock);
 
+	if (descs > 0)
+		pm_runtime_put(sh_chan->dev);
+
 	list_for_each_entry_safe(desc, _desc, &list, node)
 		kfree(desc);
 }
 
-static struct dma_async_tx_descriptor *sh_dmae_prep_memcpy(
-	struct dma_chan *chan, dma_addr_t dma_dest, dma_addr_t dma_src,
-	size_t len, unsigned long flags)
+/**
+ * sh_dmae_add_desc - get, set up and return one transfer descriptor
+ * @sh_chan:	DMA channel
+ * @flags:	DMA transfer flags
+ * @dest:	destination DMA address, incremented when direction equals
+ *		DMA_FROM_DEVICE or DMA_BIDIRECTIONAL
+ * @src:	source DMA address, incremented when direction equals
+ *		DMA_TO_DEVICE or DMA_BIDIRECTIONAL
+ * @len:	DMA transfer length
+ * @first:	if NULL, set to the current descriptor and cookie set to -EBUSY
+ * @direction:	needed for slave DMA to decide which address to keep constant,
+ *		equals DMA_BIDIRECTIONAL for MEMCPY
+ * Returns 0 or an error
+ * Locks: called with desc_lock held
+ */
+static struct sh_desc *sh_dmae_add_desc(struct sh_dmae_chan *sh_chan,
+	unsigned long flags, dma_addr_t *dest, dma_addr_t *src, size_t *len,
+	struct sh_desc **first, enum dma_data_direction direction)
 {
-	struct sh_dmae_chan *sh_chan;
-	struct sh_desc *first = NULL, *prev = NULL, *new;
+	struct sh_desc *new;
 	size_t copy_size;
-	LIST_HEAD(tx_list);
-	int chunks = (len + SH_DMA_TCR_MAX) / (SH_DMA_TCR_MAX + 1);
 
-	if (!chan)
+	if (!*len)
 		return NULL;
 
-	if (!len)
+	/* Allocate the link descriptor from the free list */
+	new = sh_dmae_get_desc(sh_chan);
+	if (!new) {
+		dev_err(sh_chan->dev, "No free link descriptor available\n");
 		return NULL;
+	}
 
-	sh_chan = to_sh_chan(chan);
+	copy_size = min(*len, (size_t)SH_DMA_TCR_MAX + 1);
+
+	new->hw.sar = *src;
+	new->hw.dar = *dest;
+	new->hw.tcr = copy_size;
+
+	if (!*first) {
+		/* First desc */
+		new->async_tx.cookie = -EBUSY;
+		*first = new;
+	} else {
+		/* Other desc - invisible to the user */
+		new->async_tx.cookie = -EINVAL;
+	}
+
+	dev_dbg(sh_chan->dev,
+		"chaining (%u/%u)@%x -> %x with %p, cookie %d, shift %d\n",
+		copy_size, *len, *src, *dest, &new->async_tx,
+		new->async_tx.cookie, sh_chan->xmit_shift);
+
+	new->mark = DESC_PREPARED;
+	new->async_tx.flags = flags;
+	new->direction = direction;
+
+	*len -= copy_size;
+	if (direction == DMA_BIDIRECTIONAL || direction == DMA_TO_DEVICE)
+		*src += copy_size;
+	if (direction == DMA_BIDIRECTIONAL || direction == DMA_FROM_DEVICE)
+		*dest += copy_size;
+
+	return new;
+}
+
+/*
+ * sh_dmae_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_BIDIRECTIONAL
+ * and the SG list contains only one element and points at the source buffer.
+ */
+static struct dma_async_tx_descriptor *sh_dmae_prep_sg(struct sh_dmae_chan *sh_chan,
+	struct scatterlist *sgl, unsigned int sg_len, dma_addr_t *addr,
+	enum dma_data_direction direction, unsigned long flags)
+{
+	struct scatterlist *sg;
+	struct sh_desc *first = NULL, *new = NULL /* compiler... */;
+	LIST_HEAD(tx_list);
+	int chunks = 0;
+	int i;
+
+	if (!sg_len)
+		return NULL;
+
+	for_each_sg(sgl, sg, sg_len, i)
+		chunks += (sg_dma_len(sg) + SH_DMA_TCR_MAX) /
+			(SH_DMA_TCR_MAX + 1);
 
 	/* Have to lock the whole loop to protect against concurrent release */
 	spin_lock_bh(&sh_chan->desc_lock);
@@ -333,49 +481,32 @@ static struct dma_async_tx_descriptor *sh_dmae_prep_memcpy(
 	 *	only during this function, then they are immediately spliced
 	 *	back onto the free list in form of a chain
 	 */
-	do {
-		/* Allocate the link descriptor from the free list */
-		new = sh_dmae_get_desc(sh_chan);
-		if (!new) {
-			dev_err(sh_chan->dev,
-				"No free memory for link descriptor\n");
-			list_for_each_entry(new, &tx_list, node)
-				new->mark = DESC_IDLE;
-			list_splice(&tx_list, &sh_chan->ld_free);
-			spin_unlock_bh(&sh_chan->desc_lock);
-			return NULL;
-		}
-
-		copy_size = min(len, (size_t)SH_DMA_TCR_MAX + 1);
-
-		new->hw.sar = dma_src;
-		new->hw.dar = dma_dest;
-		new->hw.tcr = copy_size;
-		if (!first) {
-			/* First desc */
-			new->async_tx.cookie = -EBUSY;
-			first = new;
-		} else {
-			/* Other desc - invisible to the user */
-			new->async_tx.cookie = -EINVAL;
-		}
-
-		dev_dbg(sh_chan->dev,
-			"chaining %u of %u with %p, dst %x, cookie %d\n",
-			copy_size, len, &new->async_tx, dma_dest,
-			new->async_tx.cookie);
-
-		new->mark = DESC_PREPARED;
-		new->async_tx.flags = flags;
-		new->chunks = chunks--;
-
-		prev = new;
-		len -= copy_size;
-		dma_src += copy_size;
-		dma_dest += copy_size;
-		/* Insert the link descriptor to the LD ring */
-		list_add_tail(&new->node, &tx_list);
-	} while (len);
+	for_each_sg(sgl, sg, sg_len, i) {
+		dma_addr_t sg_addr = sg_dma_address(sg);
+		size_t len = sg_dma_len(sg);
+
+		if (!len)
+			goto err_get_desc;
+
+		do {
+			dev_dbg(sh_chan->dev, "Add SG #%d@%p[%d], dma %llx\n",
+				i, sg, len, (unsigned long long)sg_addr);
+
+			if (direction == DMA_FROM_DEVICE)
+				new = sh_dmae_add_desc(sh_chan, flags,
+						&sg_addr, addr, &len, &first,
+						direction);
+			else
+				new = sh_dmae_add_desc(sh_chan, flags,
+						addr, &sg_addr, &len, &first,
+						direction);
+			if (!new)
+				goto err_get_desc;
+
+			new->chunks = chunks--;
+			list_add_tail(&new->node, &tx_list);
+		} while (len);
+	}
 
 	if (new != first)
 		new->async_tx.cookie = -ENOSPC;
@@ -386,6 +517,90 @@ static struct dma_async_tx_descriptor *sh_dmae_prep_memcpy(
 	spin_unlock_bh(&sh_chan->desc_lock);
 
 	return &first->async_tx;
+
+err_get_desc:
+	list_for_each_entry(new, &tx_list, node)
+		new->mark = DESC_IDLE;
+	list_splice(&tx_list, &sh_chan->ld_free);
+
+	spin_unlock_bh(&sh_chan->desc_lock);
+
+	return NULL;
+}
+
+static struct dma_async_tx_descriptor *sh_dmae_prep_memcpy(
+	struct dma_chan *chan, dma_addr_t dma_dest, dma_addr_t dma_src,
+	size_t len, unsigned long flags)
+{
+	struct sh_dmae_chan *sh_chan;
+	struct scatterlist sg;
+
+	if (!chan || !len)
+		return NULL;
+
+	chan->private = NULL;
+
+	sh_chan = to_sh_chan(chan);
+
+	sg_init_table(&sg, 1);
+	sg_set_page(&sg, pfn_to_page(PFN_DOWN(dma_src)), len,
+		    offset_in_page(dma_src));
+	sg_dma_address(&sg) = dma_src;
+	sg_dma_len(&sg) = len;
+
+	return sh_dmae_prep_sg(sh_chan, &sg, 1, &dma_dest, DMA_BIDIRECTIONAL,
+			       flags);
+}
+
+static struct dma_async_tx_descriptor *sh_dmae_prep_slave_sg(
+	struct dma_chan *chan, struct scatterlist *sgl, unsigned int sg_len,
+	enum dma_data_direction direction, unsigned long flags)
+{
+	struct sh_dmae_slave *param;
+	struct sh_dmae_chan *sh_chan;
+
+	if (!chan)
+		return NULL;
+
+	sh_chan = to_sh_chan(chan);
+	param = chan->private;
+
+	/* Someone calling slave DMA on a public channel? */
+	if (!param || !sg_len) {
+		dev_warn(sh_chan->dev, "%s: bad parameter: %p, %d, %d\n",
+			 __func__, param, sg_len, param ? param->slave_id : -1);
+		return NULL;
+	}
+
+	/*
+	 * if (param != NULL), this is a successfully requested slave channel,
+	 * therefore param->config != NULL too.
+	 */
+	return sh_dmae_prep_sg(sh_chan, sgl, sg_len, &param->config->addr,
+			       direction, flags);
+}
+
+static void sh_dmae_terminate_all(struct dma_chan *chan)
+{
+	struct sh_dmae_chan *sh_chan = to_sh_chan(chan);
+
+	if (!chan)
+		return;
+
+	dmae_halt(sh_chan);
+
+	spin_lock_bh(&sh_chan->desc_lock);
+	if (!list_empty(&sh_chan->ld_queue)) {
+		/* Record partial transfer */
+		struct sh_desc *desc = list_entry(sh_chan->ld_queue.next,
+						  struct sh_desc, node);
+		desc->partial = (desc->hw.tcr - sh_dmae_readl(sh_chan, TCR)) <<
+			sh_chan->xmit_shift;
+
+	}
+	spin_unlock_bh(&sh_chan->desc_lock);
+
+	sh_dmae_chan_ld_cleanup(sh_chan, true);
 }
 
 static dma_async_tx_callback __ld_cleanup(struct sh_dmae_chan *sh_chan, bool all)
@@ -419,7 +634,11 @@ static dma_async_tx_callback __ld_cleanup(struct sh_dmae_chan *sh_chan, bool all
 			cookie = tx->cookie;
 
 		if (desc->mark == DESC_COMPLETED && desc->chunks == 1) {
-			BUG_ON(sh_chan->completed_cookie != desc->cookie - 1);
+			if (sh_chan->completed_cookie != desc->cookie - 1)
+				dev_dbg(sh_chan->dev,
+					"Completing cookie %d, expected %d\n",
+					desc->cookie,
+					sh_chan->completed_cookie + 1);
 			sh_chan->completed_cookie = desc->cookie;
 		}
 
@@ -483,7 +702,7 @@ static void sh_dmae_chan_ld_cleanup(struct sh_dmae_chan *sh_chan, bool all)
 
 static void sh_chan_xfer_ld_queue(struct sh_dmae_chan *sh_chan)
 {
-	struct sh_desc *sd;
+	struct sh_desc *desc;
 
 	spin_lock_bh(&sh_chan->desc_lock);
 	/* DMA work check */
@@ -492,11 +711,14 @@ static void sh_chan_xfer_ld_queue(struct sh_dmae_chan *sh_chan)
 		return;
 	}
 
-	/* Find the first un-transfer desciptor */
-	list_for_each_entry(sd, &sh_chan->ld_queue, node)
-		if (sd->mark == DESC_SUBMITTED) {
+	/* Find the first not transferred desciptor */
+	list_for_each_entry(desc, &sh_chan->ld_queue, node)
+		if (desc->mark == DESC_SUBMITTED) {
+			dev_dbg(sh_chan->dev, "Queue #%d to %d: %u@%x -> %x\n",
+				desc->async_tx.cookie, sh_chan->id,
+				desc->hw.tcr, desc->hw.sar, desc->hw.dar);
 			/* Get the ld start address from ld_queue */
-			dmae_set_reg(sh_chan, &sd->hw);
+			dmae_set_reg(sh_chan, &desc->hw);
 			dmae_start(sh_chan);
 			break;
 		}
@@ -518,6 +740,7 @@ static enum dma_status sh_dmae_is_complete(struct dma_chan *chan,
 	struct sh_dmae_chan *sh_chan = to_sh_chan(chan);
 	dma_cookie_t last_used;
 	dma_cookie_t last_complete;
+	enum dma_status status;
 
 	sh_dmae_chan_ld_cleanup(sh_chan, false);
 
@@ -531,7 +754,27 @@ static enum dma_status sh_dmae_is_complete(struct dma_chan *chan,
 	if (used)
 		*used = last_used;
 
-	return dma_async_is_complete(cookie, last_complete, last_used);
+	spin_lock_bh(&sh_chan->desc_lock);
+
+	status = dma_async_is_complete(cookie, last_complete, last_used);
+
+	/*
+	 * If we don't find cookie on the queue, it has been aborted and we have
+	 * to report error
+	 */
+	if (status != DMA_SUCCESS) {
+		struct sh_desc *desc;
+		status = DMA_ERROR;
+		list_for_each_entry(desc, &sh_chan->ld_queue, node)
+			if (desc->cookie == cookie) {
+				status = DMA_IN_PROGRESS;
+				break;
+			}
+	}
+
+	spin_unlock_bh(&sh_chan->desc_lock);
+
+	return status;
 }
 
 static irqreturn_t sh_dmae_interrupt(int irq, void *data)
@@ -554,40 +797,32 @@ static irqreturn_t sh_dmae_interrupt(int irq, void *data)
 #if defined(CONFIG_CPU_SH4)
 static irqreturn_t sh_dmae_err(int irq, void *data)
 {
-	int err = 0;
 	struct sh_dmae_device *shdev = (struct sh_dmae_device *)data;
+	int i;
 
-	/* IRQ Multi */
-	if (shdev->pdata.mode & SHDMA_MIX_IRQ) {
-		int cnt = 0;
-		switch (irq) {
-#if defined(DMTE6_IRQ) && defined(DMAE1_IRQ)
-		case DMTE6_IRQ:
-			cnt++;
-#endif
-		case DMTE0_IRQ:
-			if (dmaor_read_reg(cnt) & (DMAOR_NMIF | DMAOR_AE)) {
-				disable_irq(irq);
-				return IRQ_HANDLED;
+	/* halt the dma controller */
+	sh_dmae_ctl_stop(shdev);
+
+	/* We cannot detect, which channel caused the error, have to reset all */
+	for (i = 0; i < SH_DMAC_MAX_CHANNELS; i++) {
+		struct sh_dmae_chan *sh_chan = shdev->chan[i];
+		if (sh_chan) {
+			struct sh_desc *desc;
+			/* Stop the channel */
+			dmae_halt(sh_chan);
+			/* Complete all  */
+			list_for_each_entry(desc, &sh_chan->ld_queue, node) {
+				struct dma_async_tx_descriptor *tx = &desc->async_tx;
+				desc->mark = DESC_IDLE;
+				if (tx->callback)
+					tx->callback(tx->callback_param);
 			}
-		default:
-			return IRQ_NONE;
-		}
-	} else {
-		/* reset dma controller */
-		err = sh_dmae_rst(0);
-		if (err)
-			return err;
-#ifdef SH_DMAC_BASE1
-		if (shdev->pdata.mode & SHDMA_DMAOR1) {
-			err = sh_dmae_rst(1);
-			if (err)
-				return err;
+			list_splice_init(&sh_chan->ld_queue, &sh_chan->ld_free);
 		}
-#endif
-		disable_irq(irq);
-		return IRQ_HANDLED;
 	}
+	sh_dmae_rst(shdev);
+
+	return IRQ_HANDLED;
 }
 #endif
 
@@ -596,11 +831,14 @@ static void dmae_do_tasklet(unsigned long data)
 	struct sh_dmae_chan *sh_chan = (struct sh_dmae_chan *)data;
 	struct sh_desc *desc;
 	u32 sar_buf = sh_dmae_readl(sh_chan, SAR);
+	u32 dar_buf = sh_dmae_readl(sh_chan, DAR);
 
 	spin_lock(&sh_chan->desc_lock);
 	list_for_each_entry(desc, &sh_chan->ld_queue, node) {
-		if ((desc->hw.sar + desc->hw.tcr) == sar_buf &&
-		    desc->mark == DESC_SUBMITTED) {
+		if (desc->mark == DESC_SUBMITTED &&
+		    ((desc->direction == DMA_FROM_DEVICE &&
+		      (desc->hw.dar + desc->hw.tcr) == dar_buf) ||
+		     (desc->hw.sar + desc->hw.tcr) == sar_buf)) {
 			dev_dbg(sh_chan->dev, "done #%d@%p dst %u\n",
 				desc->async_tx.cookie, &desc->async_tx,
 				desc->hw.dar);
@@ -615,19 +853,12 @@ static void dmae_do_tasklet(unsigned long data)
 	sh_dmae_chan_ld_cleanup(sh_chan, false);
 }
 
-static unsigned int get_dmae_irq(unsigned int id)
-{
-	unsigned int irq = 0;
-	if (id < ARRAY_SIZE(dmte_irq_map))
-		irq = dmte_irq_map[id];
-	return irq;
-}
-
-static int __devinit sh_dmae_chan_probe(struct sh_dmae_device *shdev, int id)
+static int __devinit sh_dmae_chan_probe(struct sh_dmae_device *shdev, int id,
+					int irq, unsigned long flags)
 {
 	int err;
-	unsigned int irq = get_dmae_irq(id);
-	unsigned long irqflags = IRQF_DISABLED;
+	struct sh_dmae_channel *chan_pdata = &shdev->pdata->channel[id];
+	struct platform_device *pdev = to_platform_device(shdev->common.dev);
 	struct sh_dmae_chan *new_sh_chan;
 
 	/* alloc channel */
@@ -638,8 +869,13 @@ static int __devinit sh_dmae_chan_probe(struct sh_dmae_device *shdev, int id)
 		return -ENOMEM;
 	}
 
+	/* copy struct dma_device */
+	new_sh_chan->common.device = &shdev->common;
+
 	new_sh_chan->dev = shdev->common.dev;
 	new_sh_chan->id = id;
+	new_sh_chan->irq = irq;
+	new_sh_chan->base = shdev->chan_reg + chan_pdata->offset / sizeof(u32);
 
 	/* Init DMA tasklet */
 	tasklet_init(&new_sh_chan->tasklet, dmae_do_tasklet,
@@ -654,29 +890,20 @@ static int __devinit sh_dmae_chan_probe(struct sh_dmae_device *shdev, int id)
 	INIT_LIST_HEAD(&new_sh_chan->ld_queue);
 	INIT_LIST_HEAD(&new_sh_chan->ld_free);
 
-	/* copy struct dma_device */
-	new_sh_chan->common.device = &shdev->common;
-
 	/* Add the channel to DMA device channel list */
 	list_add_tail(&new_sh_chan->common.device_node,
 			&shdev->common.channels);
 	shdev->common.chancnt++;
 
-	if (shdev->pdata.mode & SHDMA_MIX_IRQ) {
-		irqflags = IRQF_SHARED;
-#if defined(DMTE6_IRQ)
-		if (irq >= DMTE6_IRQ)
-			irq = DMTE6_IRQ;
-		else
-#endif
-			irq = DMTE0_IRQ;
-	}
-
-	snprintf(new_sh_chan->dev_id, sizeof(new_sh_chan->dev_id),
-			"sh-dmae%d", new_sh_chan->id);
+	if (pdev->id >= 0)
+		snprintf(new_sh_chan->dev_id, sizeof(new_sh_chan->dev_id),
+			 "sh-dmae%d.%d", pdev->id, new_sh_chan->id);
+	else
+		snprintf(new_sh_chan->dev_id, sizeof(new_sh_chan->dev_id),
+			 "sh-dma%d", new_sh_chan->id);
 
 	/* set up channel irq */
-	err = request_irq(irq, &sh_dmae_interrupt, irqflags,
+	err = request_irq(irq, &sh_dmae_interrupt, flags,
 			  new_sh_chan->dev_id, new_sh_chan);
 	if (err) {
 		dev_err(shdev->common.dev, "DMA channel %d request_irq error "
@@ -684,11 +911,6 @@ static int __devinit sh_dmae_chan_probe(struct sh_dmae_device *shdev, int id)
 		goto err_no_irq;
 	}
 
-	/* CHCR register control function */
-	new_sh_chan->set_chcr = dmae_set_chcr;
-	/* DMARS register control function */
-	new_sh_chan->set_dmars = dmae_set_dmars;
-
 	shdev->chan[id] = new_sh_chan;
 	return 0;
 
@@ -705,12 +927,12 @@ static void sh_dmae_chan_remove(struct sh_dmae_device *shdev)
 
 	for (i = shdev->common.chancnt - 1 ; i >= 0 ; i--) {
 		if (shdev->chan[i]) {
-			struct sh_dmae_chan *shchan = shdev->chan[i];
-			if (!(shdev->pdata.mode & SHDMA_MIX_IRQ))
-				free_irq(dmte_irq_map[i], shchan);
+			struct sh_dmae_chan *sh_chan = shdev->chan[i];
 
-			list_del(&shchan->common.device_node);
-			kfree(shchan);
+			free_irq(sh_chan->irq, sh_chan);
+
+			list_del(&sh_chan->common.device_node);
+			kfree(sh_chan);
 			shdev->chan[i] = NULL;
 		}
 	}
@@ -719,85 +941,164 @@ static void sh_dmae_chan_remove(struct sh_dmae_device *shdev)
 
 static int __init sh_dmae_probe(struct platform_device *pdev)
 {
-	int err = 0, cnt, ecnt;
-	unsigned long irqflags = IRQF_DISABLED;
-#if defined(CONFIG_CPU_SH4)
-	int eirq[] = { DMAE0_IRQ,
-#if defined(DMAE1_IRQ)
-			DMAE1_IRQ
-#endif
-		};
-#endif
+	struct sh_dmae_pdata *pdata = pdev->dev.platform_data;
+	unsigned long irqflags = IRQF_DISABLED,
+		chan_flag[SH_DMAC_MAX_CHANNELS] = {};
+	int errirq, chan_irq[SH_DMAC_MAX_CHANNELS];
+	int err, i, irq_cnt = 0, irqres = 0;
 	struct sh_dmae_device *shdev;
+	struct resource *chan, *dmars, *errirq_res, *chanirq_res;
 
 	/* get platform data */
-	if (!pdev->dev.platform_data)
+	if (!pdata || !pdata->channel_num)
+		return -ENODEV;
+
+	chan = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	/* DMARS area is optional, if absent, this controller cannot do slave DMA */
+	dmars = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	/*
+	 * IRQ resources:
+	 * 1. there always must be at least one IRQ IO-resource. On SH4 it is
+	 *    the error IRQ, in which case it is the only IRQ in this resource:
+	 *    start == end. If it is the only IRQ resource, all channels also
+	 *    use the same IRQ.
+	 * 2. DMA channel IRQ resources can be specified one per resource or in
+	 *    ranges (start != end)
+	 * 3. iff all events (channels and, optionally, error) on this
+	 *    controller use the same IRQ, only one IRQ resource can be
+	 *    specified, otherwise there must be one IRQ per channel, even if
+	 *    some of them are equal
+	 * 4. if all IRQs on this controller are equal or if some specific IRQs
+	 *    specify IORESOURCE_IRQ_SHAREABLE in their resources, they will be
+	 *    requested with the IRQF_SHARED flag
+	 */
+	errirq_res = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
+	if (!chan || !errirq_res)
 		return -ENODEV;
 
+	if (!request_mem_region(chan->start, resource_size(chan), pdev->name)) {
+		dev_err(&pdev->dev, "DMAC register region already claimed\n");
+		return -EBUSY;
+	}
+
+	if (dmars && !request_mem_region(dmars->start, resource_size(dmars), pdev->name)) {
+		dev_err(&pdev->dev, "DMAC DMARS region already claimed\n");
+		err = -EBUSY;
+		goto ermrdmars;
+	}
+
+	err = -ENOMEM;
 	shdev = kzalloc(sizeof(struct sh_dmae_device), GFP_KERNEL);
 	if (!shdev) {
-		dev_err(&pdev->dev, "No enough memory\n");
-		return -ENOMEM;
+		dev_err(&pdev->dev, "Not enough memory\n");
+		goto ealloc;
+	}
+
+	shdev->chan_reg = ioremap(chan->start, resource_size(chan));
+	if (!shdev->chan_reg)
+		goto emapchan;
+	if (dmars) {
+		shdev->dmars = ioremap(dmars->start, resource_size(dmars));
+		if (!shdev->dmars)
+			goto emapdmars;
 	}
 
 	/* platform data */
-	memcpy(&shdev->pdata, pdev->dev.platform_data,
-			sizeof(struct sh_dmae_pdata));
+	shdev->pdata = pdata;
+
+	pm_runtime_enable(&pdev->dev);
+	pm_runtime_get_sync(&pdev->dev);
 
 	/* reset dma controller */
-	err = sh_dmae_rst(0);
+	err = sh_dmae_rst(shdev);
 	if (err)
 		goto rst_err;
 
-	/* SH7780/85/23 has DMAOR1 */
-	if (shdev->pdata.mode & SHDMA_DMAOR1) {
-		err = sh_dmae_rst(1);
-		if (err)
-			goto rst_err;
-	}
-
 	INIT_LIST_HEAD(&shdev->common.channels);
 
 	dma_cap_set(DMA_MEMCPY, shdev->common.cap_mask);
+	if (dmars)
+		dma_cap_set(DMA_SLAVE, shdev->common.cap_mask);
+
 	shdev->common.device_alloc_chan_resources
 		= sh_dmae_alloc_chan_resources;
 	shdev->common.device_free_chan_resources = sh_dmae_free_chan_resources;
 	shdev->common.device_prep_dma_memcpy = sh_dmae_prep_memcpy;
 	shdev->common.device_is_tx_complete = sh_dmae_is_complete;
 	shdev->common.device_issue_pending = sh_dmae_memcpy_issue_pending;
+
+	/* Compulsory for DMA_SLAVE fields */
+	shdev->common.device_prep_slave_sg = sh_dmae_prep_slave_sg;
+	shdev->common.device_terminate_all = sh_dmae_terminate_all;
+
 	shdev->common.dev = &pdev->dev;
 	/* Default transfer size of 32 bytes requires 32-byte alignment */
-	shdev->common.copy_align = 5;
+	shdev->common.copy_align = LOG2_DEFAULT_XFER_SIZE;
 
 #if defined(CONFIG_CPU_SH4)
-	/* Non Mix IRQ mode SH7722/SH7730 etc... */
-	if (shdev->pdata.mode & SHDMA_MIX_IRQ) {
+	chanirq_res = platform_get_resource(pdev, IORESOURCE_IRQ, 1);
+
+	if (!chanirq_res)
+		chanirq_res = errirq_res;
+	else
+		irqres++;
+
+	if (chanirq_res == errirq_res ||
+	    (errirq_res->flags & IORESOURCE_BITS) == IORESOURCE_IRQ_SHAREABLE)
 		irqflags = IRQF_SHARED;
-		eirq[0] = DMTE0_IRQ;
-#if defined(DMTE6_IRQ) && defined(DMAE1_IRQ)
-		eirq[1] = DMTE6_IRQ;
-#endif
+
+	errirq = errirq_res->start;
+
+	err = request_irq(errirq, sh_dmae_err, irqflags,
+			  "DMAC Address Error", shdev);
+	if (err) {
+		dev_err(&pdev->dev,
+			"DMA failed requesting irq #%d, error %d\n",
+			errirq, err);
+		goto eirq_err;
 	}
 
-	for (ecnt = 0 ; ecnt < ARRAY_SIZE(eirq); ecnt++) {
-		err = request_irq(eirq[ecnt], sh_dmae_err, irqflags,
-				  "DMAC Address Error", shdev);
-		if (err) {
-			dev_err(&pdev->dev, "DMA device request_irq"
-				"error (irq %d) with return %d\n",
-				eirq[ecnt], err);
-			goto eirq_err;
+#else
+	chanirq_res = errirq_res;
+#endif /* CONFIG_CPU_SH4 */
+
+	if (chanirq_res->start == chanirq_res->end &&
+	    !platform_get_resource(pdev, IORESOURCE_IRQ, 1)) {
+		/* Special case - all multiplexed */
+		for (; irq_cnt < pdata->channel_num; irq_cnt++) {
+			chan_irq[irq_cnt] = chanirq_res->start;
+			chan_flag[irq_cnt] = IRQF_SHARED;
 		}
+	} else {
+		do {
+			for (i = chanirq_res->start; i <= chanirq_res->end; i++) {
+				if ((errirq_res->flags & IORESOURCE_BITS) ==
+				    IORESOURCE_IRQ_SHAREABLE)
+					chan_flag[irq_cnt] = IRQF_SHARED;
+				else
+					chan_flag[irq_cnt] = IRQF_DISABLED;
+				dev_dbg(&pdev->dev,
+					"Found IRQ %d for channel %d\n",
+					i, irq_cnt);
+				chan_irq[irq_cnt++] = i;
+			}
+			chanirq_res = platform_get_resource(pdev,
+						IORESOURCE_IRQ, ++irqres);
+		} while (irq_cnt < pdata->channel_num && chanirq_res);
 	}
-#endif /* CONFIG_CPU_SH4 */
+
+	if (irq_cnt < pdata->channel_num)
+		goto eirqres;
 
 	/* Create DMA Channel */
-	for (cnt = 0 ; cnt < MAX_DMA_CHANNELS ; cnt++) {
-		err = sh_dmae_chan_probe(shdev, cnt);
+	for (i = 0; i < pdata->channel_num; i++) {
+		err = sh_dmae_chan_probe(shdev, i, chan_irq[i], chan_flag[i]);
 		if (err)
 			goto chan_probe_err;
 	}
 
+	pm_runtime_put(&pdev->dev);
+
 	platform_set_drvdata(pdev, shdev);
 	dma_async_device_register(&shdev->common);
 
@@ -805,13 +1106,24 @@ static int __init sh_dmae_probe(struct platform_device *pdev)
 
 chan_probe_err:
 	sh_dmae_chan_remove(shdev);
-
+eirqres:
+#if defined(CONFIG_CPU_SH4)
+	free_irq(errirq, shdev);
 eirq_err:
-	for (ecnt-- ; ecnt >= 0; ecnt--)
-		free_irq(eirq[ecnt], shdev);
-
+#endif
 rst_err:
+	pm_runtime_put(&pdev->dev);
+	if (dmars)
+		iounmap(shdev->dmars);
+emapdmars:
+	iounmap(shdev->chan_reg);
+emapchan:
 	kfree(shdev);
+ealloc:
+	if (dmars)
+		release_mem_region(dmars->start, resource_size(dmars));
+ermrdmars:
+	release_mem_region(chan->start, resource_size(chan));
 
 	return err;
 }
@@ -819,36 +1131,39 @@ rst_err:
 static int __exit sh_dmae_remove(struct platform_device *pdev)
 {
 	struct sh_dmae_device *shdev = platform_get_drvdata(pdev);
+	struct resource *res;
+	int errirq = platform_get_irq(pdev, 0);
 
 	dma_async_device_unregister(&shdev->common);
 
-	if (shdev->pdata.mode & SHDMA_MIX_IRQ) {
-		free_irq(DMTE0_IRQ, shdev);
-#if defined(DMTE6_IRQ)
-		free_irq(DMTE6_IRQ, shdev);
-#endif
-	}
+	if (errirq > 0)
+		free_irq(errirq, shdev);
 
 	/* channel data remove */
 	sh_dmae_chan_remove(shdev);
 
-	if (!(shdev->pdata.mode & SHDMA_MIX_IRQ)) {
-		free_irq(DMAE0_IRQ, shdev);
-#if defined(DMAE1_IRQ)
-		free_irq(DMAE1_IRQ, shdev);
-#endif
-	}
+	pm_runtime_disable(&pdev->dev);
+
+	if (shdev->dmars)
+		iounmap(shdev->dmars);
+	iounmap(shdev->chan_reg);
+
 	kfree(shdev);
 
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (res)
+		release_mem_region(res->start, resource_size(res));
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	if (res)
+		release_mem_region(res->start, resource_size(res));
+
 	return 0;
 }
 
 static void sh_dmae_shutdown(struct platform_device *pdev)
 {
 	struct sh_dmae_device *shdev = platform_get_drvdata(pdev);
-	sh_dmae_ctl_stop(0);
-	if (shdev->pdata.mode & SHDMA_DMAOR1)
-		sh_dmae_ctl_stop(1);
+	sh_dmae_ctl_stop(shdev);
 }
 
 static struct platform_driver sh_dmae_driver = {
diff --git a/drivers/dma/shdma.h b/drivers/dma/shdma.h
index 108f1cffb6f..153609a1e96 100644
--- a/drivers/dma/shdma.h
+++ b/drivers/dma/shdma.h
@@ -17,22 +17,9 @@
 #include <linux/interrupt.h>
 #include <linux/list.h>
 
-#define SH_DMA_TCR_MAX 0x00FFFFFF	/* 16MB */
-
-struct sh_dmae_regs {
-	u32 sar; /* SAR / source address */
-	u32 dar; /* DAR / destination address */
-	u32 tcr; /* TCR / transfer count */
-};
+#include <asm/dmaengine.h>
 
-struct sh_desc {
-	struct sh_dmae_regs hw;
-	struct list_head node;
-	struct dma_async_tx_descriptor async_tx;
-	dma_cookie_t cookie;
-	int chunks;
-	int mark;
-};
+#define SH_DMA_TCR_MAX 0x00FFFFFF	/* 16MB */
 
 struct device;
 
@@ -45,19 +32,19 @@ struct sh_dmae_chan {
 	struct device *dev;		/* Channel device */
 	struct tasklet_struct tasklet;	/* Tasklet */
 	int descs_allocated;		/* desc count */
+	int xmit_shift;			/* log_2(bytes_per_xfer) */
+	int irq;
 	int id;				/* Raw id of this channel */
+	u32 __iomem *base;
 	char dev_id[16];		/* unique name per DMAC of channel */
-
-	/* Set chcr */
-	int (*set_chcr)(struct sh_dmae_chan *sh_chan, u32 regs);
-	/* Set DMA resource */
-	int (*set_dmars)(struct sh_dmae_chan *sh_chan, u16 res);
 };
 
 struct sh_dmae_device {
 	struct dma_device common;
-	struct sh_dmae_chan *chan[MAX_DMA_CHANNELS];
-	struct sh_dmae_pdata pdata;
+	struct sh_dmae_chan *chan[SH_DMAC_MAX_CHANNELS];
+	struct sh_dmae_pdata *pdata;
+	u32 __iomem *chan_reg;
+	u16 __iomem *dmars;
 };
 
 #define to_sh_chan(chan) container_of(chan, struct sh_dmae_chan, common)