Merge tag 'dmaengine-4.4-rc1' of git://git.infradead.org/users/vkoul/slave-dma

Pull dmaengine updates from Vinod Koul:
 "This time we have a very typical update which is mostly fixes and
  updates to drivers and no new drivers.

   - the biggest change is coming from Peter for edma cleanup which even
     caused some last minute regression, things seem settled now
   - idma64 and dw updates
   - iotdma updates
   - module autoload fixes for various drivers
   - scatter gather support for hdmac"

* tag 'dmaengine-4.4-rc1' of git://git.infradead.org/users/vkoul/slave-dma: (77 commits)
  dmaengine: edma: Add dummy driver skeleton for edma3-tptc
  Revert "ARM: DTS: am33xx: Use the new DT bindings for the eDMA3"
  Revert "ARM: DTS: am437x: Use the new DT bindings for the eDMA3"
  dmaengine: dw: some Intel devices has no memcpy support
  dmaengine: dw: platform: provide platform data for Intel
  dmaengine: dw: don't override platform data with autocfg
  dmaengine: hdmac: Add scatter-gathered memset support
  dmaengine: hdmac: factorise memset descriptor allocation
  dmaengine: virt-dma: Fix kernel-doc annotations
  ARM: DTS: am437x: Use the new DT bindings for the eDMA3
  ARM: DTS: am33xx: Use the new DT bindings for the eDMA3
  dmaengine: edma: New device tree binding
  dmaengine: Kconfig: edma: Select TI_DMA_CROSSBAR in case of ARCH_OMAP
  dmaengine: ti-dma-crossbar: Add support for crossbar on AM33xx/AM43xx
  dmaengine: edma: Merge the of parsing functions
  dmaengine: edma: Do not allocate memory for edma_rsv_info in case of DT boot
  dmaengine: edma: Refactor the dma device and channel struct initialization
  dmaengine: edma: Get qDMA channel information from HW also
  dmaengine: edma: Merge map_dmach_to_queue into assign_channel_eventq
  dmaengine: edma: Correct PaRAM access function names (_parm_ to _param_)
  ...

30 files changed, 2278 insertions, 537 deletions

diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index b458475..e6cd1a3 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -229,7 +229,7 @@ config IMX_SDMA
 	  Support the i.MX SDMA engine. This engine is integrated into
 	  Freescale i.MX25/31/35/51/53/6 chips.
 
-config IDMA64
+config INTEL_IDMA64
 	tristate "Intel integrated DMA 64-bit support"
 	select DMA_ENGINE
 	select DMA_VIRTUAL_CHANNELS
@@ -486,7 +486,7 @@ config TI_EDMA
 	depends on ARCH_DAVINCI || ARCH_OMAP || ARCH_KEYSTONE
 	select DMA_ENGINE
 	select DMA_VIRTUAL_CHANNELS
-	select TI_PRIV_EDMA
+	select TI_DMA_CROSSBAR if ARCH_OMAP
 	default n
 	help
 	  Enable support for the TI EDMA controller. This DMA
diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile
index 7711a71..ef9c099 100644
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@@ -34,7 +34,7 @@ obj-$(CONFIG_HSU_DMA) += hsu/
 obj-$(CONFIG_IMG_MDC_DMA) += img-mdc-dma.o
 obj-$(CONFIG_IMX_DMA) += imx-dma.o
 obj-$(CONFIG_IMX_SDMA) += imx-sdma.o
-obj-$(CONFIG_IDMA64) += idma64.o
+obj-$(CONFIG_INTEL_IDMA64) += idma64.o
 obj-$(CONFIG_INTEL_IOATDMA) += ioat/
 obj-$(CONFIG_INTEL_IOP_ADMA) += iop-adma.o
 obj-$(CONFIG_INTEL_MIC_X100_DMA) += mic_x100_dma.o
diff --git a/drivers/dma/acpi-dma.c b/drivers/dma/acpi-dma.c
index 981a38f..16d0daa 100644
--- a/drivers/dma/acpi-dma.c
+++ b/drivers/dma/acpi-dma.c
@@ -161,10 +161,8 @@ int acpi_dma_controller_register(struct device *dev,
 		return -EINVAL;
 
 	/* Check if the device was enumerated by ACPI */
-	if (!ACPI_HANDLE(dev))
-		return -EINVAL;
-
-	if (acpi_bus_get_device(ACPI_HANDLE(dev), &adev))
+	adev = ACPI_COMPANION(dev);
+	if (!adev)
 		return -EINVAL;
 
 	adma = kzalloc(sizeof(*adma), GFP_KERNEL);
@@ -359,10 +357,11 @@ struct dma_chan *acpi_dma_request_slave_chan_by_index(struct device *dev,
 	int found;
 
 	/* Check if the device was enumerated by ACPI */
-	if (!dev || !ACPI_HANDLE(dev))
+	if (!dev)
 		return ERR_PTR(-ENODEV);
 
-	if (acpi_bus_get_device(ACPI_HANDLE(dev), &adev))
+	adev = ACPI_COMPANION(dev);
+	if (!adev)
 		return ERR_PTR(-ENODEV);
 
 	memset(&pdata, 0, sizeof(pdata));
diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
index 58d4062..4e55239 100644
--- a/drivers/dma/at_hdmac.c
+++ b/drivers/dma/at_hdmac.c
@@ -458,10 +458,10 @@ atc_chain_complete(struct at_dma_chan *atchan, struct at_desc *desc)
 		dma_cookie_complete(txd);
 
 	/* If the transfer was a memset, free our temporary buffer */
-	if (desc->memset) {
+	if (desc->memset_buffer) {
 		dma_pool_free(atdma->memset_pool, desc->memset_vaddr,
 			      desc->memset_paddr);
-		desc->memset = false;
+		desc->memset_buffer = false;
 	}
 
 	/* move children to free_list */
@@ -881,6 +881,46 @@ err_desc_get:
 	return NULL;
 }
 
+static struct at_desc *atc_create_memset_desc(struct dma_chan *chan,
+					      dma_addr_t psrc,
+					      dma_addr_t pdst,
+					      size_t len)
+{
+	struct at_dma_chan *atchan = to_at_dma_chan(chan);
+	struct at_desc *desc;
+	size_t xfer_count;
+
+	u32 ctrla = ATC_SRC_WIDTH(2) | ATC_DST_WIDTH(2);
+	u32 ctrlb = ATC_DEFAULT_CTRLB | ATC_IEN |
+		ATC_SRC_ADDR_MODE_FIXED |
+		ATC_DST_ADDR_MODE_INCR |
+		ATC_FC_MEM2MEM;
+
+	xfer_count = len >> 2;
+	if (xfer_count > ATC_BTSIZE_MAX) {
+		dev_err(chan2dev(chan), "%s: buffer is too big\n",
+			__func__);
+		return NULL;
+	}
+
+	desc = atc_desc_get(atchan);
+	if (!desc) {
+		dev_err(chan2dev(chan), "%s: can't get a descriptor\n",
+			__func__);
+		return NULL;
+	}
+
+	desc->lli.saddr = psrc;
+	desc->lli.daddr = pdst;
+	desc->lli.ctrla = ctrla | xfer_count;
+	desc->lli.ctrlb = ctrlb;
+
+	desc->txd.cookie = 0;
+	desc->len = len;
+
+	return desc;
+}
+
 /**
  * atc_prep_dma_memset - prepare a memcpy operation
  * @chan: the channel to prepare operation on
@@ -893,12 +933,10 @@ static struct dma_async_tx_descriptor *
 atc_prep_dma_memset(struct dma_chan *chan, dma_addr_t dest, int value,
 		    size_t len, unsigned long flags)
 {
-	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
 	struct at_dma		*atdma = to_at_dma(chan->device);
-	struct at_desc		*desc = NULL;
-	size_t			xfer_count;
-	u32			ctrla;
-	u32			ctrlb;
+	struct at_desc		*desc;
+	void __iomem		*vaddr;
+	dma_addr_t		paddr;
 
 	dev_vdbg(chan2dev(chan), "%s: d0x%x v0x%x l0x%zx f0x%lx\n", __func__,
 		dest, value, len, flags);
@@ -914,61 +952,117 @@ atc_prep_dma_memset(struct dma_chan *chan, dma_addr_t dest, int value,
 		return NULL;
 	}
 
-	xfer_count = len >> 2;
-	if (xfer_count > ATC_BTSIZE_MAX) {
-		dev_err(chan2dev(chan), "%s: buffer is too big\n",
+	vaddr = dma_pool_alloc(atdma->memset_pool, GFP_ATOMIC, &paddr);
+	if (!vaddr) {
+		dev_err(chan2dev(chan), "%s: couldn't allocate buffer\n",
 			__func__);
 		return NULL;
 	}
+	*(u32*)vaddr = value;
 
-	ctrlb =   ATC_DEFAULT_CTRLB | ATC_IEN
-		| ATC_SRC_ADDR_MODE_FIXED
-		| ATC_DST_ADDR_MODE_INCR
-		| ATC_FC_MEM2MEM;
+	desc = atc_create_memset_desc(chan, paddr, dest, len);
+	if (!desc) {
+		dev_err(chan2dev(chan), "%s: couldn't get a descriptor\n",
+			__func__);
+		goto err_free_buffer;
+	}
 
-	ctrla = ATC_SRC_WIDTH(2) |
-		ATC_DST_WIDTH(2);
+	desc->memset_paddr = paddr;
+	desc->memset_vaddr = vaddr;
+	desc->memset_buffer = true;
 
-	desc = atc_desc_get(atchan);
-	if (!desc) {
-		dev_err(chan2dev(chan), "%s: can't get a descriptor\n",
+	desc->txd.cookie = -EBUSY;
+	desc->total_len = len;
+
+	/* set end-of-link on the descriptor */
+	set_desc_eol(desc);
+
+	desc->txd.flags = flags;
+
+	return &desc->txd;
+
+err_free_buffer:
+	dma_pool_free(atdma->memset_pool, vaddr, paddr);
+	return NULL;
+}
+
+static struct dma_async_tx_descriptor *
+atc_prep_dma_memset_sg(struct dma_chan *chan,
+		       struct scatterlist *sgl,
+		       unsigned int sg_len, int value,
+		       unsigned long flags)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	struct at_dma		*atdma = to_at_dma(chan->device);
+	struct at_desc		*desc = NULL, *first = NULL, *prev = NULL;
+	struct scatterlist	*sg;
+	void __iomem		*vaddr;
+	dma_addr_t		paddr;
+	size_t			total_len = 0;
+	int			i;
+
+	dev_vdbg(chan2dev(chan), "%s: v0x%x l0x%zx f0x%lx\n", __func__,
+		 value, sg_len, flags);
+
+	if (unlikely(!sgl || !sg_len)) {
+		dev_dbg(chan2dev(chan), "%s: scatterlist is empty!\n",
 			__func__);
 		return NULL;
 	}
 
-	desc->memset_vaddr = dma_pool_alloc(atdma->memset_pool, GFP_ATOMIC,
-					    &desc->memset_paddr);
-	if (!desc->memset_vaddr) {
+	vaddr = dma_pool_alloc(atdma->memset_pool, GFP_ATOMIC, &paddr);
+	if (!vaddr) {
 		dev_err(chan2dev(chan), "%s: couldn't allocate buffer\n",
 			__func__);
-		goto err_put_desc;
+		return NULL;
 	}
+	*(u32*)vaddr = value;
 
-	*desc->memset_vaddr = value;
-	desc->memset = true;
+	for_each_sg(sgl, sg, sg_len, i) {
+		dma_addr_t dest = sg_dma_address(sg);
+		size_t len = sg_dma_len(sg);
 
-	desc->lli.saddr = desc->memset_paddr;
-	desc->lli.daddr = dest;
-	desc->lli.ctrla = ctrla | xfer_count;
-	desc->lli.ctrlb = ctrlb;
+		dev_vdbg(chan2dev(chan), "%s: d0x%08x, l0x%zx\n",
+			 __func__, dest, len);
 
-	desc->txd.cookie = -EBUSY;
-	desc->len = len;
-	desc->total_len = len;
+		if (!is_dma_fill_aligned(chan->device, dest, 0, len)) {
+			dev_err(chan2dev(chan), "%s: buffer is not aligned\n",
+				__func__);
+			goto err_put_desc;
+		}
+
+		desc = atc_create_memset_desc(chan, paddr, dest, len);
+		if (!desc)
+			goto err_put_desc;
+
+		atc_desc_chain(&first, &prev, desc);
+
+		total_len += len;
+	}
+
+	/*
+	 * Only set the buffer pointers on the last descriptor to
+	 * avoid free'ing while we have our transfer still going
+	 */
+	desc->memset_paddr = paddr;
+	desc->memset_vaddr = vaddr;
+	desc->memset_buffer = true;
+
+	first->txd.cookie = -EBUSY;
+	first->total_len = total_len;
 
 	/* set end-of-link on the descriptor */
 	set_desc_eol(desc);
 
-	desc->txd.flags = flags;
+	first->txd.flags = flags;
 
-	return &desc->txd;
+	return &first->txd;
 
 err_put_desc:
-	atc_desc_put(atchan, desc);
+	atc_desc_put(atchan, first);
 	return NULL;
 }
 
-
 /**
  * atc_prep_slave_sg - prepare descriptors for a DMA_SLAVE transaction
  * @chan: DMA channel
@@ -1851,6 +1945,7 @@ static int __init at_dma_probe(struct platform_device *pdev)
 	dma_cap_set(DMA_INTERLEAVE, at91sam9g45_config.cap_mask);
 	dma_cap_set(DMA_MEMCPY, at91sam9g45_config.cap_mask);
 	dma_cap_set(DMA_MEMSET, at91sam9g45_config.cap_mask);
+	dma_cap_set(DMA_MEMSET_SG, at91sam9g45_config.cap_mask);
 	dma_cap_set(DMA_PRIVATE, at91sam9g45_config.cap_mask);
 	dma_cap_set(DMA_SLAVE, at91sam9g45_config.cap_mask);
 	dma_cap_set(DMA_SG, at91sam9g45_config.cap_mask);
@@ -1972,6 +2067,7 @@ static int __init at_dma_probe(struct platform_device *pdev)
 
 	if (dma_has_cap(DMA_MEMSET, atdma->dma_common.cap_mask)) {
 		atdma->dma_common.device_prep_dma_memset = atc_prep_dma_memset;
+		atdma->dma_common.device_prep_dma_memset_sg = atc_prep_dma_memset_sg;
 		atdma->dma_common.fill_align = DMAENGINE_ALIGN_4_BYTES;
 	}
 
diff --git a/drivers/dma/at_hdmac_regs.h b/drivers/dma/at_hdmac_regs.h
index c3bebbe..d1cfc8c 100644
--- a/drivers/dma/at_hdmac_regs.h
+++ b/drivers/dma/at_hdmac_regs.h
@@ -202,7 +202,7 @@ struct at_desc {
 	size_t				src_hole;
 
 	/* Memset temporary buffer */
-	bool				memset;
+	bool				memset_buffer;
 	dma_addr_t			memset_paddr;
 	int				*memset_vaddr;
 };
diff --git a/drivers/dma/at_xdmac.c b/drivers/dma/at_xdmac.c
index dd24375..b5e132d 100644
--- a/drivers/dma/at_xdmac.c
+++ b/drivers/dma/at_xdmac.c
@@ -938,13 +938,19 @@ at_xdmac_prep_interleaved(struct dma_chan *chan,
 {
 	struct at_xdmac_chan	*atchan = to_at_xdmac_chan(chan);
 	struct at_xdmac_desc	*prev = NULL, *first = NULL;
-	struct data_chunk	*chunk, *prev_chunk = NULL;
 	dma_addr_t		dst_addr, src_addr;
-	size_t			dst_skip, src_skip, len = 0;
-	size_t			prev_dst_icg = 0, prev_src_icg = 0;
+	size_t			src_skip = 0, dst_skip = 0, len = 0;
+	struct data_chunk	*chunk;
 	int			i;
 
-	if (!xt || (xt->numf != 1) || (xt->dir != DMA_MEM_TO_MEM))
+	if (!xt || !xt->numf || (xt->dir != DMA_MEM_TO_MEM))
+		return NULL;
+
+	/*
+	 * TODO: Handle the case where we have to repeat a chain of
+	 * descriptors...
+	 */
+	if ((xt->numf > 1) && (xt->frame_size > 1))
 		return NULL;
 
 	dev_dbg(chan2dev(chan), "%s: src=0x%08x, dest=0x%08x, numf=%d, frame_size=%d, flags=0x%lx\n",
@@ -954,66 +960,60 @@ at_xdmac_prep_interleaved(struct dma_chan *chan,
 	src_addr = xt->src_start;
 	dst_addr = xt->dst_start;
 
-	for (i = 0; i < xt->frame_size; i++) {
-		struct at_xdmac_desc *desc;
-		size_t src_icg, dst_icg;
+	if (xt->numf > 1) {
+		first = at_xdmac_interleaved_queue_desc(chan, atchan,
+							NULL,
+							src_addr, dst_addr,
+							xt, xt->sgl);
+		for (i = 0; i < xt->numf; i++)
+			at_xdmac_increment_block_count(chan, first);
 
-		chunk = xt->sgl + i;
+		dev_dbg(chan2dev(chan), "%s: add desc 0x%p to descs_list 0x%p\n",
+			__func__, first, first);
+		list_add_tail(&first->desc_node, &first->descs_list);
+	} else {
+		for (i = 0; i < xt->frame_size; i++) {
+			size_t src_icg = 0, dst_icg = 0;
+			struct at_xdmac_desc *desc;
 
-		dst_icg = dmaengine_get_dst_icg(xt, chunk);
-		src_icg = dmaengine_get_src_icg(xt, chunk);
+			chunk = xt->sgl + i;
 
-		src_skip = chunk->size + src_icg;
-		dst_skip = chunk->size + dst_icg;
+			dst_icg = dmaengine_get_dst_icg(xt, chunk);
+			src_icg = dmaengine_get_src_icg(xt, chunk);
 
-		dev_dbg(chan2dev(chan),
-			"%s: chunk size=%d, src icg=%d, dst icg=%d\n",
-			__func__, chunk->size, src_icg, dst_icg);
+			src_skip = chunk->size + src_icg;
+			dst_skip = chunk->size + dst_icg;
 
-		/*
-		 * Handle the case where we just have the same
-		 * transfer to setup, we can just increase the
-		 * block number and reuse the same descriptor.
-		 */
-		if (prev_chunk && prev &&
-		    (prev_chunk->size == chunk->size) &&
-		    (prev_src_icg == src_icg) &&
-		    (prev_dst_icg == dst_icg)) {
 			dev_dbg(chan2dev(chan),
-				"%s: same configuration that the previous chunk, merging the descriptors...\n",
-				__func__);
-			at_xdmac_increment_block_count(chan, prev);
-			continue;
-		}
-
-		desc = at_xdmac_interleaved_queue_desc(chan, atchan,
-						       prev,
-						       src_addr, dst_addr,
-						       xt, chunk);
-		if (!desc) {
-			list_splice_init(&first->descs_list,
-					 &atchan->free_descs_list);
-			return NULL;
-		}
+				"%s: chunk size=%d, src icg=%d, dst icg=%d\n",
+				__func__, chunk->size, src_icg, dst_icg);
+
+			desc = at_xdmac_interleaved_queue_desc(chan, atchan,
+							       prev,
+							       src_addr, dst_addr,
+							       xt, chunk);
+			if (!desc) {
+				list_splice_init(&first->descs_list,
+						 &atchan->free_descs_list);
+				return NULL;
+			}
 
-		if (!first)
-			first = desc;
+			if (!first)
+				first = desc;
 
-		dev_dbg(chan2dev(chan), "%s: add desc 0x%p to descs_list 0x%p\n",
-			__func__, desc, first);
-		list_add_tail(&desc->desc_node, &first->descs_list);
+			dev_dbg(chan2dev(chan), "%s: add desc 0x%p to descs_list 0x%p\n",
+				__func__, desc, first);
+			list_add_tail(&desc->desc_node, &first->descs_list);
 
-		if (xt->src_sgl)
-			src_addr += src_skip;
+			if (xt->src_sgl)
+				src_addr += src_skip;
 
-		if (xt->dst_sgl)
-			dst_addr += dst_skip;
+			if (xt->dst_sgl)
+				dst_addr += dst_skip;
 
-		len += chunk->size;
-		prev_chunk = chunk;
-		prev_dst_icg = dst_icg;
-		prev_src_icg = src_icg;
-		prev = desc;
+			len += chunk->size;
+			prev = desc;
+		}
 	}
 
 	first->tx_dma_desc.cookie = -EBUSY;
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index 09479d4..3ecec14 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -1074,11 +1074,9 @@ static void dmaengine_destroy_unmap_pool(void)
 	for (i = 0; i < ARRAY_SIZE(unmap_pool); i++) {
 		struct dmaengine_unmap_pool *p = &unmap_pool[i];
 
-		if (p->pool)
-			mempool_destroy(p->pool);
+		mempool_destroy(p->pool);
 		p->pool = NULL;
-		if (p->cache)
-			kmem_cache_destroy(p->cache);
+		kmem_cache_destroy(p->cache);
 		p->cache = NULL;
 	}
 }
diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index bedce03..7067b6d 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -163,7 +163,7 @@ static void dwc_initialize(struct dw_dma_chan *dwc)
 
 /*----------------------------------------------------------------------*/
 
-static inline unsigned int dwc_fast_fls(unsigned long long v)
+static inline unsigned int dwc_fast_ffs(unsigned long long v)
 {
 	/*
 	 * We can be a lot more clever here, but this should take care
@@ -712,7 +712,7 @@ dwc_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 			   dw->data_width[dwc->dst_master]);
 
 	src_width = dst_width = min_t(unsigned int, data_width,
-				      dwc_fast_fls(src | dest | len));
+				      dwc_fast_ffs(src | dest | len));
 
 	ctllo = DWC_DEFAULT_CTLLO(chan)
 			| DWC_CTLL_DST_WIDTH(dst_width)
@@ -791,7 +791,7 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 
 	switch (direction) {
 	case DMA_MEM_TO_DEV:
-		reg_width = __fls(sconfig->dst_addr_width);
+		reg_width = __ffs(sconfig->dst_addr_width);
 		reg = sconfig->dst_addr;
 		ctllo = (DWC_DEFAULT_CTLLO(chan)
 				| DWC_CTLL_DST_WIDTH(reg_width)
@@ -811,7 +811,7 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 			len = sg_dma_len(sg);
 
 			mem_width = min_t(unsigned int,
-					  data_width, dwc_fast_fls(mem | len));
+					  data_width, dwc_fast_ffs(mem | len));
 
 slave_sg_todev_fill_desc:
 			desc = dwc_desc_get(dwc);
@@ -848,7 +848,7 @@ slave_sg_todev_fill_desc:
 		}
 		break;
 	case DMA_DEV_TO_MEM:
-		reg_width = __fls(sconfig->src_addr_width);
+		reg_width = __ffs(sconfig->src_addr_width);
 		reg = sconfig->src_addr;
 		ctllo = (DWC_DEFAULT_CTLLO(chan)
 				| DWC_CTLL_SRC_WIDTH(reg_width)
@@ -868,7 +868,7 @@ slave_sg_todev_fill_desc:
 			len = sg_dma_len(sg);
 
 			mem_width = min_t(unsigned int,
-					  data_width, dwc_fast_fls(mem | len));
+					  data_width, dwc_fast_ffs(mem | len));
 
 slave_sg_fromdev_fill_desc:
 			desc = dwc_desc_get(dwc);
@@ -1499,9 +1499,8 @@ EXPORT_SYMBOL(dw_dma_cyclic_free);
 int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 {
 	struct dw_dma		*dw;
-	bool			autocfg;
+	bool			autocfg = false;
 	unsigned int		dw_params;
-	unsigned int		nr_channels;
 	unsigned int		max_blk_size = 0;
 	int			err;
 	int			i;
@@ -1515,33 +1514,42 @@ int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 
 	pm_runtime_get_sync(chip->dev);
 
-	dw_params = dma_read_byaddr(chip->regs, DW_PARAMS);
-	autocfg = dw_params >> DW_PARAMS_EN & 0x1;
+	if (!pdata) {
+		dw_params = dma_read_byaddr(chip->regs, DW_PARAMS);
+		dev_dbg(chip->dev, "DW_PARAMS: 0x%08x\n", dw_params);
 
-	dev_dbg(chip->dev, "DW_PARAMS: 0x%08x\n", dw_params);
+		autocfg = dw_params >> DW_PARAMS_EN & 1;
+		if (!autocfg) {
+			err = -EINVAL;
+			goto err_pdata;
+		}
 
-	if (!pdata && autocfg) {
 		pdata = devm_kzalloc(chip->dev, sizeof(*pdata), GFP_KERNEL);
 		if (!pdata) {
 			err = -ENOMEM;
 			goto err_pdata;
 		}
 
+		/* Get hardware configuration parameters */
+		pdata->nr_channels = (dw_params >> DW_PARAMS_NR_CHAN & 7) + 1;
+		pdata->nr_masters = (dw_params >> DW_PARAMS_NR_MASTER & 3) + 1;
+		for (i = 0; i < pdata->nr_masters; i++) {
+			pdata->data_width[i] =
+				(dw_params >> DW_PARAMS_DATA_WIDTH(i) & 3) + 2;
+		}
+		max_blk_size = dma_readl(dw, MAX_BLK_SIZE);
+
 		/* Fill platform data with the default values */
 		pdata->is_private = true;
+		pdata->is_memcpy = true;
 		pdata->chan_allocation_order = CHAN_ALLOCATION_ASCENDING;
 		pdata->chan_priority = CHAN_PRIORITY_ASCENDING;
-	} else if (!pdata || pdata->nr_channels > DW_DMA_MAX_NR_CHANNELS) {
+	} else if (pdata->nr_channels > DW_DMA_MAX_NR_CHANNELS) {
 		err = -EINVAL;
 		goto err_pdata;
 	}
 
-	if (autocfg)
-		nr_channels = (dw_params >> DW_PARAMS_NR_CHAN & 0x7) + 1;
-	else
-		nr_channels = pdata->nr_channels;
-
-	dw->chan = devm_kcalloc(chip->dev, nr_channels, sizeof(*dw->chan),
+	dw->chan = devm_kcalloc(chip->dev, pdata->nr_channels, sizeof(*dw->chan),
 				GFP_KERNEL);
 	if (!dw->chan) {
 		err = -ENOMEM;
@@ -1549,22 +1557,12 @@ int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 	}
 
 	/* Get hardware configuration parameters */
-	if (autocfg) {
-		max_blk_size = dma_readl(dw, MAX_BLK_SIZE);
-
-		dw->nr_masters = (dw_params >> DW_PARAMS_NR_MASTER & 3) + 1;
-		for (i = 0; i < dw->nr_masters; i++) {
-			dw->data_width[i] =
-				(dw_params >> DW_PARAMS_DATA_WIDTH(i) & 3) + 2;
-		}
-	} else {
-		dw->nr_masters = pdata->nr_masters;
-		for (i = 0; i < dw->nr_masters; i++)
-			dw->data_width[i] = pdata->data_width[i];
-	}
+	dw->nr_masters = pdata->nr_masters;
+	for (i = 0; i < dw->nr_masters; i++)
+		dw->data_width[i] = pdata->data_width[i];
 
 	/* Calculate all channel mask before DMA setup */
-	dw->all_chan_mask = (1 << nr_channels) - 1;
+	dw->all_chan_mask = (1 << pdata->nr_channels) - 1;
 
 	/* Force dma off, just in case */
 	dw_dma_off(dw);
@@ -1589,7 +1587,7 @@ int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 		goto err_pdata;
 
 	INIT_LIST_HEAD(&dw->dma.channels);
-	for (i = 0; i < nr_channels; i++) {
+	for (i = 0; i < pdata->nr_channels; i++) {
 		struct dw_dma_chan	*dwc = &dw->chan[i];
 
 		dwc->chan.device = &dw->dma;
@@ -1602,7 +1600,7 @@ int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 
 		/* 7 is highest priority & 0 is lowest. */
 		if (pdata->chan_priority == CHAN_PRIORITY_ASCENDING)
-			dwc->priority = nr_channels - i - 1;
+			dwc->priority = pdata->nr_channels - i - 1;
 		else
 			dwc->priority = i;
 
@@ -1656,10 +1654,13 @@ int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 	dma_writel(dw, CLEAR.DST_TRAN, dw->all_chan_mask);
 	dma_writel(dw, CLEAR.ERROR, dw->all_chan_mask);
 
-	dma_cap_set(DMA_MEMCPY, dw->dma.cap_mask);
+	/* Set capabilities */
 	dma_cap_set(DMA_SLAVE, dw->dma.cap_mask);
 	if (pdata->is_private)
 		dma_cap_set(DMA_PRIVATE, dw->dma.cap_mask);
+	if (pdata->is_memcpy)
+		dma_cap_set(DMA_MEMCPY, dw->dma.cap_mask);
+
 	dw->dma.dev = chip->dev;
 	dw->dma.device_alloc_chan_resources = dwc_alloc_chan_resources;
 	dw->dma.device_free_chan_resources = dwc_free_chan_resources;
@@ -1687,7 +1688,7 @@ int dw_dma_probe(struct dw_dma_chip *chip, struct dw_dma_platform_data *pdata)
 		goto err_dma_register;
 
 	dev_info(chip->dev, "DesignWare DMA Controller, %d channels\n",
-		 nr_channels);
+		 pdata->nr_channels);
 
 	pm_runtime_put_sync_suspend(chip->dev);
 
diff --git a/drivers/dma/dw/pci.c b/drivers/dma/dw/pci.c
index b144706..4c30fdd 100644
--- a/drivers/dma/dw/pci.c
+++ b/drivers/dma/dw/pci.c
@@ -15,12 +15,6 @@
 
 #include "internal.h"
 
-static struct dw_dma_platform_data dw_pci_pdata = {
-	.is_private = 1,
-	.chan_allocation_order = CHAN_ALLOCATION_ASCENDING,
-	.chan_priority = CHAN_PRIORITY_ASCENDING,
-};
-
 static int dw_pci_probe(struct pci_dev *pdev, const struct pci_device_id *pid)
 {
 	struct dw_dma_chip *chip;
@@ -101,19 +95,19 @@ static const struct dev_pm_ops dw_pci_dev_pm_ops = {
 
 static const struct pci_device_id dw_pci_id_table[] = {
 	/* Medfield */
-	{ PCI_VDEVICE(INTEL, 0x0827), (kernel_ulong_t)&dw_pci_pdata },
-	{ PCI_VDEVICE(INTEL, 0x0830), (kernel_ulong_t)&dw_pci_pdata },
+	{ PCI_VDEVICE(INTEL, 0x0827) },
+	{ PCI_VDEVICE(INTEL, 0x0830) },
 
 	/* BayTrail */
-	{ PCI_VDEVICE(INTEL, 0x0f06), (kernel_ulong_t)&dw_pci_pdata },
-	{ PCI_VDEVICE(INTEL, 0x0f40), (kernel_ulong_t)&dw_pci_pdata },
+	{ PCI_VDEVICE(INTEL, 0x0f06) },
+	{ PCI_VDEVICE(INTEL, 0x0f40) },
 
 	/* Braswell */
-	{ PCI_VDEVICE(INTEL, 0x2286), (kernel_ulong_t)&dw_pci_pdata },
-	{ PCI_VDEVICE(INTEL, 0x22c0), (kernel_ulong_t)&dw_pci_pdata },
+	{ PCI_VDEVICE(INTEL, 0x2286) },
+	{ PCI_VDEVICE(INTEL, 0x22c0) },
 
 	/* Haswell */
-	{ PCI_VDEVICE(INTEL, 0x9c60), (kernel_ulong_t)&dw_pci_pdata },
+	{ PCI_VDEVICE(INTEL, 0x9c60) },
 	{ }
 };
 MODULE_DEVICE_TABLE(pci, dw_pci_id_table);
diff --git a/drivers/dma/dw/platform.c b/drivers/dma/dw/platform.c
index b2c3ae0..68a4815 100644
--- a/drivers/dma/dw/platform.c
+++ b/drivers/dma/dw/platform.c
@@ -155,6 +155,7 @@ static int dw_probe(struct platform_device *pdev)
 	struct dw_dma_chip *chip;
 	struct device *dev = &pdev->dev;
 	struct resource *mem;
+	const struct acpi_device_id *id;
 	struct dw_dma_platform_data *pdata;
 	int err;
 
@@ -178,6 +179,11 @@ static int dw_probe(struct platform_device *pdev)
 	pdata = dev_get_platdata(dev);
 	if (!pdata)
 		pdata = dw_dma_parse_dt(pdev);
+	if (!pdata && has_acpi_companion(dev)) {
+		id = acpi_match_device(dev->driver->acpi_match_table, dev);
+		if (id)
+			pdata = (struct dw_dma_platform_data *)id->driver_data;
+	}
 
 	chip->dev = dev;
 
@@ -246,8 +252,17 @@ MODULE_DEVICE_TABLE(of, dw_dma_of_id_table);
 #endif
 
 #ifdef CONFIG_ACPI
+static struct dw_dma_platform_data dw_dma_acpi_pdata = {
+	.nr_channels = 8,
+	.is_private = true,
+	.chan_allocation_order = CHAN_ALLOCATION_ASCENDING,
+	.chan_priority = CHAN_PRIORITY_ASCENDING,
+	.block_size = 4095,
+	.nr_masters = 2,
+};
+
 static const struct acpi_device_id dw_dma_acpi_id_table[] = {
-	{ "INTL9C60", 0 },
+	{ "INTL9C60", (kernel_ulong_t)&dw_dma_acpi_pdata },
 	{ }
 };
 MODULE_DEVICE_TABLE(acpi, dw_dma_acpi_id_table);
diff --git a/drivers/dma/edma.c b/drivers/dma/edma.c
index 3e5d4f1..6b03e4e 100644
--- a/drivers/dma/edma.c
+++ b/drivers/dma/edma.c
@@ -25,28 +25,93 @@
 #include <linux/slab.h>
 #include <linux/spinlock.h>
 #include <linux/of.h>
+#include <linux/of_dma.h>
+#include <linux/of_irq.h>
+#include <linux/of_address.h>
+#include <linux/of_device.h>
+#include <linux/pm_runtime.h>
 
 #include <linux/platform_data/edma.h>
 
 #include "dmaengine.h"
 #include "virt-dma.h"
 
-/*
- * This will go away when the private EDMA API is folded
- * into this driver and the platform device(s) are
- * instantiated in the arch code. We can only get away
- * with this simplification because DA8XX may not be built
- * in the same kernel image with other DaVinci parts. This
- * avoids having to sprinkle dmaengine driver platform devices
- * and data throughout all the existing board files.
- */
-#ifdef CONFIG_ARCH_DAVINCI_DA8XX
-#define EDMA_CTLRS	2
-#define EDMA_CHANS	32
-#else
-#define EDMA_CTLRS	1
-#define EDMA_CHANS	64
-#endif /* CONFIG_ARCH_DAVINCI_DA8XX */
+/* Offsets matching "struct edmacc_param" */
+#define PARM_OPT		0x00
+#define PARM_SRC		0x04
+#define PARM_A_B_CNT		0x08
+#define PARM_DST		0x0c
+#define PARM_SRC_DST_BIDX	0x10
+#define PARM_LINK_BCNTRLD	0x14
+#define PARM_SRC_DST_CIDX	0x18
+#define PARM_CCNT		0x1c
+
+#define PARM_SIZE		0x20
+
+/* Offsets for EDMA CC global channel registers and their shadows */
+#define SH_ER			0x00	/* 64 bits */
+#define SH_ECR			0x08	/* 64 bits */
+#define SH_ESR			0x10	/* 64 bits */
+#define SH_CER			0x18	/* 64 bits */
+#define SH_EER			0x20	/* 64 bits */
+#define SH_EECR			0x28	/* 64 bits */
+#define SH_EESR			0x30	/* 64 bits */
+#define SH_SER			0x38	/* 64 bits */
+#define SH_SECR			0x40	/* 64 bits */
+#define SH_IER			0x50	/* 64 bits */
+#define SH_IECR			0x58	/* 64 bits */
+#define SH_IESR			0x60	/* 64 bits */
+#define SH_IPR			0x68	/* 64 bits */
+#define SH_ICR			0x70	/* 64 bits */
+#define SH_IEVAL		0x78
+#define SH_QER			0x80
+#define SH_QEER			0x84
+#define SH_QEECR		0x88
+#define SH_QEESR		0x8c
+#define SH_QSER			0x90
+#define SH_QSECR		0x94
+#define SH_SIZE			0x200
+
+/* Offsets for EDMA CC global registers */
+#define EDMA_REV		0x0000
+#define EDMA_CCCFG		0x0004
+#define EDMA_QCHMAP		0x0200	/* 8 registers */
+#define EDMA_DMAQNUM		0x0240	/* 8 registers (4 on OMAP-L1xx) */
+#define EDMA_QDMAQNUM		0x0260
+#define EDMA_QUETCMAP		0x0280
+#define EDMA_QUEPRI		0x0284
+#define EDMA_EMR		0x0300	/* 64 bits */
+#define EDMA_EMCR		0x0308	/* 64 bits */
+#define EDMA_QEMR		0x0310
+#define EDMA_QEMCR		0x0314
+#define EDMA_CCERR		0x0318
+#define EDMA_CCERRCLR		0x031c
+#define EDMA_EEVAL		0x0320
+#define EDMA_DRAE		0x0340	/* 4 x 64 bits*/
+#define EDMA_QRAE		0x0380	/* 4 registers */
+#define EDMA_QUEEVTENTRY	0x0400	/* 2 x 16 registers */
+#define EDMA_QSTAT		0x0600	/* 2 registers */
+#define EDMA_QWMTHRA		0x0620
+#define EDMA_QWMTHRB		0x0624
+#define EDMA_CCSTAT		0x0640
+
+#define EDMA_M			0x1000	/* global channel registers */
+#define EDMA_ECR		0x1008
+#define EDMA_ECRH		0x100C
+#define EDMA_SHADOW0		0x2000	/* 4 shadow regions */
+#define EDMA_PARM		0x4000	/* PaRAM entries */
+
+#define PARM_OFFSET(param_no)	(EDMA_PARM + ((param_no) << 5))
+
+#define EDMA_DCHMAP		0x0100  /* 64 registers */
+
+/* CCCFG register */
+#define GET_NUM_DMACH(x)	(x & 0x7) /* bits 0-2 */
+#define GET_NUM_QDMACH(x)	(x & 0x70 >> 4) /* bits 4-6 */
+#define GET_NUM_PAENTRY(x)	((x & 0x7000) >> 12) /* bits 12-14 */
+#define GET_NUM_EVQUE(x)	((x & 0x70000) >> 16) /* bits 16-18 */
+#define GET_NUM_REGN(x)		((x & 0x300000) >> 20) /* bits 20-21 */
+#define CHMAP_EXIST		BIT(24)
 
 /*
  * Max of 20 segments per channel to conserve PaRAM slots
@@ -59,6 +124,37 @@
 #define EDMA_MAX_SLOTS		MAX_NR_SG
 #define EDMA_DESCRIPTORS	16
 
+#define EDMA_CHANNEL_ANY		-1	/* for edma_alloc_channel() */
+#define EDMA_SLOT_ANY			-1	/* for edma_alloc_slot() */
+#define EDMA_CONT_PARAMS_ANY		 1001
+#define EDMA_CONT_PARAMS_FIXED_EXACT	 1002
+#define EDMA_CONT_PARAMS_FIXED_NOT_EXACT 1003
+
+/* PaRAM slots are laid out like this */
+struct edmacc_param {
+	u32 opt;
+	u32 src;
+	u32 a_b_cnt;
+	u32 dst;
+	u32 src_dst_bidx;
+	u32 link_bcntrld;
+	u32 src_dst_cidx;
+	u32 ccnt;
+} __packed;
+
+/* fields in edmacc_param.opt */
+#define SAM		BIT(0)
+#define DAM		BIT(1)
+#define SYNCDIM		BIT(2)
+#define STATIC		BIT(3)
+#define EDMA_FWID	(0x07 << 8)
+#define TCCMODE		BIT(11)
+#define EDMA_TCC(t)	((t) << 12)
+#define TCINTEN		BIT(20)
+#define ITCINTEN	BIT(21)
+#define TCCHEN		BIT(22)
+#define ITCCHEN		BIT(23)
+
 struct edma_pset {
 	u32				len;
 	dma_addr_t			addr;
@@ -105,26 +201,524 @@ struct edma_desc {
 
 struct edma_cc;
 
+struct edma_tc {
+	struct device_node		*node;
+	u16				id;
+};
+
 struct edma_chan {
 	struct virt_dma_chan		vchan;
 	struct list_head		node;
 	struct edma_desc		*edesc;
 	struct edma_cc			*ecc;
+	struct edma_tc			*tc;
 	int				ch_num;
 	bool				alloced;
+	bool				hw_triggered;
 	int				slot[EDMA_MAX_SLOTS];
 	int				missed;
 	struct dma_slave_config		cfg;
 };
 
 struct edma_cc {
-	int				ctlr;
+	struct device			*dev;
+	struct edma_soc_info		*info;
+	void __iomem			*base;
+	int				id;
+	bool				legacy_mode;
+
+	/* eDMA3 resource information */
+	unsigned			num_channels;
+	unsigned			num_qchannels;
+	unsigned			num_region;
+	unsigned			num_slots;
+	unsigned			num_tc;
+	bool				chmap_exist;
+	enum dma_event_q		default_queue;
+
+	/*
+	 * The slot_inuse bit for each PaRAM slot is clear unless the slot is
+	 * in use by Linux or if it is allocated to be used by DSP.
+	 */
+	unsigned long *slot_inuse;
+
 	struct dma_device		dma_slave;
-	struct edma_chan		slave_chans[EDMA_CHANS];
-	int				num_slave_chans;
+	struct dma_device		*dma_memcpy;
+	struct edma_chan		*slave_chans;
+	struct edma_tc			*tc_list;
 	int				dummy_slot;
 };
 
+/* dummy param set used to (re)initialize parameter RAM slots */
+static const struct edmacc_param dummy_paramset = {
+	.link_bcntrld = 0xffff,
+	.ccnt = 1,
+};
+
+#define EDMA_BINDING_LEGACY	0
+#define EDMA_BINDING_TPCC	1
+static const struct of_device_id edma_of_ids[] = {
+	{
+		.compatible = "ti,edma3",
+		.data = (void *)EDMA_BINDING_LEGACY,
+	},
+	{
+		.compatible = "ti,edma3-tpcc",
+		.data = (void *)EDMA_BINDING_TPCC,
+	},
+	{}
+};
+
+static const struct of_device_id edma_tptc_of_ids[] = {
+	{ .compatible = "ti,edma3-tptc", },
+	{}
+};
+
+static inline unsigned int edma_read(struct edma_cc *ecc, int offset)
+{
+	return (unsigned int)__raw_readl(ecc->base + offset);
+}
+
+static inline void edma_write(struct edma_cc *ecc, int offset, int val)
+{
+	__raw_writel(val, ecc->base + offset);
+}
+
+static inline void edma_modify(struct edma_cc *ecc, int offset, unsigned and,
+			       unsigned or)
+{
+	unsigned val = edma_read(ecc, offset);
+
+	val &= and;
+	val |= or;
+	edma_write(ecc, offset, val);
+}
+
+static inline void edma_and(struct edma_cc *ecc, int offset, unsigned and)
+{
+	unsigned val = edma_read(ecc, offset);
+
+	val &= and;
+	edma_write(ecc, offset, val);
+}
+
+static inline void edma_or(struct edma_cc *ecc, int offset, unsigned or)
+{
+	unsigned val = edma_read(ecc, offset);
+
+	val |= or;
+	edma_write(ecc, offset, val);
+}
+
+static inline unsigned int edma_read_array(struct edma_cc *ecc, int offset,
+					   int i)
+{
+	return edma_read(ecc, offset + (i << 2));
+}
+
+static inline void edma_write_array(struct edma_cc *ecc, int offset, int i,
+				    unsigned val)
+{
+	edma_write(ecc, offset + (i << 2), val);
+}
+
+static inline void edma_modify_array(struct edma_cc *ecc, int offset, int i,
+				     unsigned and, unsigned or)
+{
+	edma_modify(ecc, offset + (i << 2), and, or);
+}
+
+static inline void edma_or_array(struct edma_cc *ecc, int offset, int i,
+				 unsigned or)
+{
+	edma_or(ecc, offset + (i << 2), or);
+}
+
+static inline void edma_or_array2(struct edma_cc *ecc, int offset, int i, int j,
+				  unsigned or)
+{
+	edma_or(ecc, offset + ((i * 2 + j) << 2), or);
+}
+
+static inline void edma_write_array2(struct edma_cc *ecc, int offset, int i,
+				     int j, unsigned val)
+{
+	edma_write(ecc, offset + ((i * 2 + j) << 2), val);
+}
+
+static inline unsigned int edma_shadow0_read(struct edma_cc *ecc, int offset)
+{
+	return edma_read(ecc, EDMA_SHADOW0 + offset);
+}
+
+static inline unsigned int edma_shadow0_read_array(struct edma_cc *ecc,
+						   int offset, int i)
+{
+	return edma_read(ecc, EDMA_SHADOW0 + offset + (i << 2));
+}
+
+static inline void edma_shadow0_write(struct edma_cc *ecc, int offset,
+				      unsigned val)
+{
+	edma_write(ecc, EDMA_SHADOW0 + offset, val);
+}
+
+static inline void edma_shadow0_write_array(struct edma_cc *ecc, int offset,
+					    int i, unsigned val)
+{
+	edma_write(ecc, EDMA_SHADOW0 + offset + (i << 2), val);
+}
+
+static inline unsigned int edma_param_read(struct edma_cc *ecc, int offset,
+					   int param_no)
+{
+	return edma_read(ecc, EDMA_PARM + offset + (param_no << 5));
+}
+
+static inline void edma_param_write(struct edma_cc *ecc, int offset,
+				    int param_no, unsigned val)
+{
+	edma_write(ecc, EDMA_PARM + offset + (param_no << 5), val);
+}
+
+static inline void edma_param_modify(struct edma_cc *ecc, int offset,
+				     int param_no, unsigned and, unsigned or)
+{
+	edma_modify(ecc, EDMA_PARM + offset + (param_no << 5), and, or);
+}
+
+static inline void edma_param_and(struct edma_cc *ecc, int offset, int param_no,
+				  unsigned and)
+{
+	edma_and(ecc, EDMA_PARM + offset + (param_no << 5), and);
+}
+
+static inline void edma_param_or(struct edma_cc *ecc, int offset, int param_no,
+				 unsigned or)
+{
+	edma_or(ecc, EDMA_PARM + offset + (param_no << 5), or);
+}
+
+static inline void set_bits(int offset, int len, unsigned long *p)
+{
+	for (; len > 0; len--)
+		set_bit(offset + (len - 1), p);
+}
+
+static inline void clear_bits(int offset, int len, unsigned long *p)
+{
+	for (; len > 0; len--)
+		clear_bit(offset + (len - 1), p);
+}
+
+static void edma_assign_priority_to_queue(struct edma_cc *ecc, int queue_no,
+					  int priority)
+{
+	int bit = queue_no * 4;
+
+	edma_modify(ecc, EDMA_QUEPRI, ~(0x7 << bit), ((priority & 0x7) << bit));
+}
+
+static void edma_set_chmap(struct edma_chan *echan, int slot)
+{
+	struct edma_cc *ecc = echan->ecc;
+	int channel = EDMA_CHAN_SLOT(echan->ch_num);
+
+	if (ecc->chmap_exist) {
+		slot = EDMA_CHAN_SLOT(slot);
+		edma_write_array(ecc, EDMA_DCHMAP, channel, (slot << 5));
+	}
+}
+
+static void edma_setup_interrupt(struct edma_chan *echan, bool enable)
+{
+	struct edma_cc *ecc = echan->ecc;
+	int channel = EDMA_CHAN_SLOT(echan->ch_num);
+
+	if (enable) {
+		edma_shadow0_write_array(ecc, SH_ICR, channel >> 5,
+					 BIT(channel & 0x1f));
+		edma_shadow0_write_array(ecc, SH_IESR, channel >> 5,
+					 BIT(channel & 0x1f));
+	} else {
+		edma_shadow0_write_array(ecc, SH_IECR, channel >> 5,
+					 BIT(channel & 0x1f));
+	}
+}
+
+/*
+ * paRAM slot management functions
+ */
+static void edma_write_slot(struct edma_cc *ecc, unsigned slot,
+			    const struct edmacc_param *param)
+{
+	slot = EDMA_CHAN_SLOT(slot);
+	if (slot >= ecc->num_slots)
+		return;
+	memcpy_toio(ecc->base + PARM_OFFSET(slot), param, PARM_SIZE);
+}
+
+static void edma_read_slot(struct edma_cc *ecc, unsigned slot,
+			   struct edmacc_param *param)
+{
+	slot = EDMA_CHAN_SLOT(slot);
+	if (slot >= ecc->num_slots)
+		return;
+	memcpy_fromio(param, ecc->base + PARM_OFFSET(slot), PARM_SIZE);
+}
+
+/**
+ * edma_alloc_slot - allocate DMA parameter RAM
+ * @ecc: pointer to edma_cc struct
+ * @slot: specific slot to allocate; negative for "any unused slot"
+ *
+ * This allocates a parameter RAM slot, initializing it to hold a
+ * dummy transfer.  Slots allocated using this routine have not been
+ * mapped to a hardware DMA channel, and will normally be used by
+ * linking to them from a slot associated with a DMA channel.
+ *
+ * Normal use is to pass EDMA_SLOT_ANY as the @slot, but specific
+ * slots may be allocated on behalf of DSP firmware.
+ *
+ * Returns the number of the slot, else negative errno.
+ */
+static int edma_alloc_slot(struct edma_cc *ecc, int slot)
+{
+	if (slot > 0) {
+		slot = EDMA_CHAN_SLOT(slot);
+		/* Requesting entry paRAM slot for a HW triggered channel. */
+		if (ecc->chmap_exist && slot < ecc->num_channels)
+			slot = EDMA_SLOT_ANY;
+	}
+
+	if (slot < 0) {
+		if (ecc->chmap_exist)
+			slot = 0;
+		else
+			slot = ecc->num_channels;
+		for (;;) {
+			slot = find_next_zero_bit(ecc->slot_inuse,
+						  ecc->num_slots,
+						  slot);
+			if (slot == ecc->num_slots)
+				return -ENOMEM;
+			if (!test_and_set_bit(slot, ecc->slot_inuse))
+				break;
+		}
+	} else if (slot >= ecc->num_slots) {
+		return -EINVAL;
+	} else if (test_and_set_bit(slot, ecc->slot_inuse)) {
+		return -EBUSY;
+	}
+
+	edma_write_slot(ecc, slot, &dummy_paramset);
+
+	return EDMA_CTLR_CHAN(ecc->id, slot);
+}
+
+static void edma_free_slot(struct edma_cc *ecc, unsigned slot)
+{
+	slot = EDMA_CHAN_SLOT(slot);
+	if (slot >= ecc->num_slots)
+		return;
+
+	edma_write_slot(ecc, slot, &dummy_paramset);
+	clear_bit(slot, ecc->slot_inuse);
+}
+
+/**
+ * edma_link - link one parameter RAM slot to another
+ * @ecc: pointer to edma_cc struct
+ * @from: parameter RAM slot originating the link
+ * @to: parameter RAM slot which is the link target
+ *
+ * The originating slot should not be part of any active DMA transfer.
+ */
+static void edma_link(struct edma_cc *ecc, unsigned from, unsigned to)
+{
+	if (unlikely(EDMA_CTLR(from) != EDMA_CTLR(to)))
+		dev_warn(ecc->dev, "Ignoring eDMA instance for linking\n");
+
+	from = EDMA_CHAN_SLOT(from);
+	to = EDMA_CHAN_SLOT(to);
+	if (from >= ecc->num_slots || to >= ecc->num_slots)
+		return;
+
+	edma_param_modify(ecc, PARM_LINK_BCNTRLD, from, 0xffff0000,
+			  PARM_OFFSET(to));
+}
+
+/**
+ * edma_get_position - returns the current transfer point
+ * @ecc: pointer to edma_cc struct
+ * @slot: parameter RAM slot being examined
+ * @dst:  true selects the dest position, false the source
+ *
+ * Returns the position of the current active slot
+ */
+static dma_addr_t edma_get_position(struct edma_cc *ecc, unsigned slot,
+				    bool dst)
+{
+	u32 offs;
+
+	slot = EDMA_CHAN_SLOT(slot);
+	offs = PARM_OFFSET(slot);
+	offs += dst ? PARM_DST : PARM_SRC;
+
+	return edma_read(ecc, offs);
+}
+
+/*
+ * Channels with event associations will be triggered by their hardware
+ * events, and channels without such associations will be triggered by
+ * software.  (At this writing there is no interface for using software
+ * triggers except with channels that don't support hardware triggers.)
+ */
+static void edma_start(struct edma_chan *echan)
+{
+	struct edma_cc *ecc = echan->ecc;
+	int channel = EDMA_CHAN_SLOT(echan->ch_num);
+	int j = (channel >> 5);
+	unsigned int mask = BIT(channel & 0x1f);
+
+	if (!echan->hw_triggered) {
+		/* EDMA channels without event association */
+		dev_dbg(ecc->dev, "ESR%d %08x\n", j,
+			edma_shadow0_read_array(ecc, SH_ESR, j));
+		edma_shadow0_write_array(ecc, SH_ESR, j, mask);
+	} else {
+		/* EDMA channel with event association */
+		dev_dbg(ecc->dev, "ER%d %08x\n", j,
+			edma_shadow0_read_array(ecc, SH_ER, j));
+		/* Clear any pending event or error */
+		edma_write_array(ecc, EDMA_ECR, j, mask);
+		edma_write_array(ecc, EDMA_EMCR, j, mask);
+		/* Clear any SER */
+		edma_shadow0_write_array(ecc, SH_SECR, j, mask);
+		edma_shadow0_write_array(ecc, SH_EESR, j, mask);
+		dev_dbg(ecc->dev, "EER%d %08x\n", j,
+			edma_shadow0_read_array(ecc, SH_EER, j));
+	}
+}
+
+static void edma_stop(struct edma_chan *echan)
+{
+	struct edma_cc *ecc = echan->ecc;
+	int channel = EDMA_CHAN_SLOT(echan->ch_num);
+	int j = (channel >> 5);
+	unsigned int mask = BIT(channel & 0x1f);
+
+	edma_shadow0_write_array(ecc, SH_EECR, j, mask);
+	edma_shadow0_write_array(ecc, SH_ECR, j, mask);
+	edma_shadow0_write_array(ecc, SH_SECR, j, mask);
+	edma_write_array(ecc, EDMA_EMCR, j, mask);
+
+	/* clear possibly pending completion interrupt */
+	edma_shadow0_write_array(ecc, SH_ICR, j, mask);
+
+	dev_dbg(ecc->dev, "EER%d %08x\n", j,
+		edma_shadow0_read_array(ecc, SH_EER, j));
+
+	/* REVISIT:  consider guarding against inappropriate event
+	 * chaining by overwriting with dummy_paramset.
+	 */
+}
+
+/*
+ * Temporarily disable EDMA hardware events on the specified channel,
+ * preventing them from triggering new transfers
+ */
+static void edma_pause(struct edma_chan *echan)
+{
+	int channel = EDMA_CHAN_SLOT(echan->ch_num);
+	unsigned int mask = BIT(channel & 0x1f);
+
+	edma_shadow0_write_array(echan->ecc, SH_EECR, channel >> 5, mask);
+}
+
+/* Re-enable EDMA hardware events on the specified channel.  */
+static void edma_resume(struct edma_chan *echan)
+{
+	int channel = EDMA_CHAN_SLOT(echan->ch_num);
+	unsigned int mask = BIT(channel & 0x1f);
+
+	edma_shadow0_write_array(echan->ecc, SH_EESR, channel >> 5, mask);
+}
+
+static void edma_trigger_channel(struct edma_chan *echan)
+{
+	struct edma_cc *ecc = echan->ecc;
+	int channel = EDMA_CHAN_SLOT(echan->ch_num);
+	unsigned int mask = BIT(channel & 0x1f);
+
+	edma_shadow0_write_array(ecc, SH_ESR, (channel >> 5), mask);
+
+	dev_dbg(ecc->dev, "ESR%d %08x\n", (channel >> 5),
+		edma_shadow0_read_array(ecc, SH_ESR, (channel >> 5)));
+}
+
+static void edma_clean_channel(struct edma_chan *echan)
+{
+	struct edma_cc *ecc = echan->ecc;
+	int channel = EDMA_CHAN_SLOT(echan->ch_num);
+	int j = (channel >> 5);
+	unsigned int mask = BIT(channel & 0x1f);
+
+	dev_dbg(ecc->dev, "EMR%d %08x\n", j, edma_read_array(ecc, EDMA_EMR, j));
+	edma_shadow0_write_array(ecc, SH_ECR, j, mask);
+	/* Clear the corresponding EMR bits */
+	edma_write_array(ecc, EDMA_EMCR, j, mask);
+	/* Clear any SER */
+	edma_shadow0_write_array(ecc, SH_SECR, j, mask);
+	edma_write(ecc, EDMA_CCERRCLR, BIT(16) | BIT(1) | BIT(0));
+}
+
+/* Move channel to a specific event queue */
+static void edma_assign_channel_eventq(struct edma_chan *echan,
+				       enum dma_event_q eventq_no)
+{
+	struct edma_cc *ecc = echan->ecc;
+	int channel = EDMA_CHAN_SLOT(echan->ch_num);
+	int bit = (channel & 0x7) * 4;
+
+	/* default to low priority queue */
+	if (eventq_no == EVENTQ_DEFAULT)
+		eventq_no = ecc->default_queue;
+	if (eventq_no >= ecc->num_tc)
+		return;
+
+	eventq_no &= 7;
+	edma_modify_array(ecc, EDMA_DMAQNUM, (channel >> 3), ~(0x7 << bit),
+			  eventq_no << bit);
+}
+
+static int edma_alloc_channel(struct edma_chan *echan,
+			      enum dma_event_q eventq_no)
+{
+	struct edma_cc *ecc = echan->ecc;
+	int channel = EDMA_CHAN_SLOT(echan->ch_num);
+
+	/* ensure access through shadow region 0 */
+	edma_or_array2(ecc, EDMA_DRAE, 0, channel >> 5, BIT(channel & 0x1f));
+
+	/* ensure no events are pending */
+	edma_stop(echan);
+
+	edma_setup_interrupt(echan, true);
+
+	edma_assign_channel_eventq(echan, eventq_no);
+
+	return 0;
+}
+
+static void edma_free_channel(struct edma_chan *echan)
+{
+	/* ensure no events are pending */
+	edma_stop(echan);
+	/* REVISIT should probably take out of shadow region 0 */
+	edma_setup_interrupt(echan, false);
+}
+
 static inline struct edma_cc *to_edma_cc(struct dma_device *d)
 {
 	return container_of(d, struct edma_cc, dma_slave);
@@ -135,8 +729,7 @@ static inline struct edma_chan *to_edma_chan(struct dma_chan *c)
 	return container_of(c, struct edma_chan, vchan.chan);
 }
 
-static inline struct edma_desc
-*to_edma_desc(struct dma_async_tx_descriptor *tx)
+static inline struct edma_desc *to_edma_desc(struct dma_async_tx_descriptor *tx)
 {
 	return container_of(tx, struct edma_desc, vdesc.tx);
 }
@@ -149,20 +742,17 @@ static void edma_desc_free(struct virt_dma_desc *vdesc)
 /* Dispatch a queued descriptor to the controller (caller holds lock) */
 static void edma_execute(struct edma_chan *echan)
 {
+	struct edma_cc *ecc = echan->ecc;
 	struct virt_dma_desc *vdesc;
 	struct edma_desc *edesc;
 	struct device *dev = echan->vchan.chan.device->dev;
 	int i, j, left, nslots;
 
-	/* If either we processed all psets or we're still not started */
-	if (!echan->edesc ||
-	    echan->edesc->pset_nr == echan->edesc->processed) {
-		/* Get next vdesc */
+	if (!echan->edesc) {
+		/* Setup is needed for the first transfer */
 		vdesc = vchan_next_desc(&echan->vchan);
-		if (!vdesc) {
-			echan->edesc = NULL;
+		if (!vdesc)
 			return;
-		}
 		list_del(&vdesc->node);
 		echan->edesc = to_edma_desc(&vdesc->tx);
 	}
@@ -177,32 +767,32 @@ static void edma_execute(struct edma_chan *echan)
 	/* Write descriptor PaRAM set(s) */
 	for (i = 0; i < nslots; i++) {
 		j = i + edesc->processed;
-		edma_write_slot(echan->slot[i], &edesc->pset[j].param);
+		edma_write_slot(ecc, echan->slot[i], &edesc->pset[j].param);
 		edesc->sg_len += edesc->pset[j].len;
-		dev_vdbg(echan->vchan.chan.device->dev,
-			"\n pset[%d]:\n"
-			"  chnum\t%d\n"
-			"  slot\t%d\n"
-			"  opt\t%08x\n"
-			"  src\t%08x\n"
-			"  dst\t%08x\n"
-			"  abcnt\t%08x\n"
-			"  ccnt\t%08x\n"
-			"  bidx\t%08x\n"
-			"  cidx\t%08x\n"
-			"  lkrld\t%08x\n",
-			j, echan->ch_num, echan->slot[i],
-			edesc->pset[j].param.opt,
-			edesc->pset[j].param.src,
-			edesc->pset[j].param.dst,
-			edesc->pset[j].param.a_b_cnt,
-			edesc->pset[j].param.ccnt,
-			edesc->pset[j].param.src_dst_bidx,
-			edesc->pset[j].param.src_dst_cidx,
-			edesc->pset[j].param.link_bcntrld);
+		dev_vdbg(dev,
+			 "\n pset[%d]:\n"
+			 "  chnum\t%d\n"
+			 "  slot\t%d\n"
+			 "  opt\t%08x\n"
+			 "  src\t%08x\n"
+			 "  dst\t%08x\n"
+			 "  abcnt\t%08x\n"
+			 "  ccnt\t%08x\n"
+			 "  bidx\t%08x\n"
+			 "  cidx\t%08x\n"
+			 "  lkrld\t%08x\n",
+			 j, echan->ch_num, echan->slot[i],
+			 edesc->pset[j].param.opt,
+			 edesc->pset[j].param.src,
+			 edesc->pset[j].param.dst,
+			 edesc->pset[j].param.a_b_cnt,
+			 edesc->pset[j].param.ccnt,
+			 edesc->pset[j].param.src_dst_bidx,
+			 edesc->pset[j].param.src_dst_cidx,
+			 edesc->pset[j].param.link_bcntrld);
 		/* Link to the previous slot if not the last set */
 		if (i != (nslots - 1))
-			edma_link(echan->slot[i], echan->slot[i+1]);
+			edma_link(ecc, echan->slot[i], echan->slot[i + 1]);
 	}
 
 	edesc->processed += nslots;
@@ -214,34 +804,32 @@ static void edma_execute(struct edma_chan *echan)
 	 */
 	if (edesc->processed == edesc->pset_nr) {
 		if (edesc->cyclic)
-			edma_link(echan->slot[nslots-1], echan->slot[1]);
+			edma_link(ecc, echan->slot[nslots - 1], echan->slot[1]);
 		else
-			edma_link(echan->slot[nslots-1],
+			edma_link(ecc, echan->slot[nslots - 1],
 				  echan->ecc->dummy_slot);
 	}
 
-	if (edesc->processed <= MAX_NR_SG) {
+	if (echan->missed) {
+		/*
+		 * This happens due to setup times between intermediate
+		 * transfers in long SG lists which have to be broken up into
+		 * transfers of MAX_NR_SG
+		 */
+		dev_dbg(dev, "missed event on channel %d\n", echan->ch_num);
+		edma_clean_channel(echan);
+		edma_stop(echan);
+		edma_start(echan);
+		edma_trigger_channel(echan);
+		echan->missed = 0;
+	} else if (edesc->processed <= MAX_NR_SG) {
 		dev_dbg(dev, "first transfer starting on channel %d\n",
 			echan->ch_num);
-		edma_start(echan->ch_num);
+		edma_start(echan);
 	} else {
 		dev_dbg(dev, "chan: %d: completed %d elements, resuming\n",
 			echan->ch_num, edesc->processed);
-		edma_resume(echan->ch_num);
-	}
-
-	/*
-	 * This happens due to setup times between intermediate transfers
-	 * in long SG lists which have to be broken up into transfers of
-	 * MAX_NR_SG
-	 */
-	if (echan->missed) {
-		dev_dbg(dev, "missed event on channel %d\n", echan->ch_num);
-		edma_clean_channel(echan->ch_num);
-		edma_stop(echan->ch_num);
-		edma_start(echan->ch_num);
-		edma_trigger_channel(echan->ch_num);
-		echan->missed = 0;
+		edma_resume(echan);
 	}
 }
 
@@ -259,20 +847,16 @@ static int edma_terminate_all(struct dma_chan *chan)
 	 * echan->edesc is NULL and exit.)
 	 */
 	if (echan->edesc) {
-		int cyclic = echan->edesc->cyclic;
-
+		edma_stop(echan);
+		/* Move the cyclic channel back to default queue */
+		if (!echan->tc && echan->edesc->cyclic)
+			edma_assign_channel_eventq(echan, EVENTQ_DEFAULT);
 		/*
 		 * free the running request descriptor
 		 * since it is not in any of the vdesc lists
 		 */
 		edma_desc_free(&echan->edesc->vdesc);
-
 		echan->edesc = NULL;
-		edma_stop(echan->ch_num);
-		/* Move the cyclic channel back to default queue */
-		if (cyclic)
-			edma_assign_channel_eventq(echan->ch_num,
-						   EVENTQ_DEFAULT);
 	}
 
 	vchan_get_all_descriptors(&echan->vchan, &head);
@@ -303,7 +887,7 @@ static int edma_dma_pause(struct dma_chan *chan)
 	if (!echan->edesc)
 		return -EINVAL;
 
-	edma_pause(echan->ch_num);
+	edma_pause(echan);
 	return 0;
 }
 
@@ -311,7 +895,7 @@ static int edma_dma_resume(struct dma_chan *chan)
 {
 	struct edma_chan *echan = to_edma_chan(chan);
 
-	edma_resume(echan->ch_num);
+	edma_resume(echan);
 	return 0;
 }
 
@@ -327,19 +911,17 @@ static int edma_dma_resume(struct dma_chan *chan)
  * @direction: Direction of the transfer
  */
 static int edma_config_pset(struct dma_chan *chan, struct edma_pset *epset,
-	dma_addr_t src_addr, dma_addr_t dst_addr, u32 burst,
-	enum dma_slave_buswidth dev_width, unsigned int dma_length,
-	enum dma_transfer_direction direction)
+			    dma_addr_t src_addr, dma_addr_t dst_addr, u32 burst,
+			    unsigned int acnt, unsigned int dma_length,
+			    enum dma_transfer_direction direction)
 {
 	struct edma_chan *echan = to_edma_chan(chan);
 	struct device *dev = chan->device->dev;
 	struct edmacc_param *param = &epset->param;
-	int acnt, bcnt, ccnt, cidx;
+	int bcnt, ccnt, cidx;
 	int src_bidx, dst_bidx, src_cidx, dst_cidx;
 	int absync;
 
-	acnt = dev_width;
-
 	/* src/dst_maxburst == 0 is the same case as src/dst_maxburst == 1 */
 	if (!burst)
 		burst = 1;
@@ -475,8 +1057,8 @@ static struct dma_async_tx_descriptor *edma_prep_slave_sg(
 		return NULL;
 	}
 
-	edesc = kzalloc(sizeof(*edesc) + sg_len *
-		sizeof(edesc->pset[0]), GFP_ATOMIC);
+	edesc = kzalloc(sizeof(*edesc) + sg_len * sizeof(edesc->pset[0]),
+			GFP_ATOMIC);
 	if (!edesc) {
 		dev_err(dev, "%s: Failed to allocate a descriptor\n", __func__);
 		return NULL;
@@ -493,8 +1075,7 @@ static struct dma_async_tx_descriptor *edma_prep_slave_sg(
 	for (i = 0; i < nslots; i++) {
 		if (echan->slot[i] < 0) {
 			echan->slot[i] =
-				edma_alloc_slot(EDMA_CTLR(echan->ch_num),
-						EDMA_SLOT_ANY);
+				edma_alloc_slot(echan->ecc, EDMA_SLOT_ANY);
 			if (echan->slot[i] < 0) {
 				kfree(edesc);
 				dev_err(dev, "%s: Failed to allocate slot\n",
@@ -541,36 +1122,98 @@ static struct dma_async_tx_descriptor *edma_prep_dma_memcpy(
 	struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 	size_t len, unsigned long tx_flags)
 {
-	int ret;
+	int ret, nslots;
 	struct edma_desc *edesc;
 	struct device *dev = chan->device->dev;
 	struct edma_chan *echan = to_edma_chan(chan);
+	unsigned int width, pset_len;
 
 	if (unlikely(!echan || !len))
 		return NULL;
 
-	edesc = kzalloc(sizeof(*edesc) + sizeof(edesc->pset[0]), GFP_ATOMIC);
+	if (len < SZ_64K) {
+		/*
+		 * Transfer size less than 64K can be handled with one paRAM
+		 * slot and with one burst.
+		 * ACNT = length
+		 */
+		width = len;
+		pset_len = len;
+		nslots = 1;
+	} else {
+		/*
+		 * Transfer size bigger than 64K will be handled with maximum of
+		 * two paRAM slots.
+		 * slot1: (full_length / 32767) times 32767 bytes bursts.
+		 *	  ACNT = 32767, length1: (full_length / 32767) * 32767
+		 * slot2: the remaining amount of data after slot1.
+		 *	  ACNT = full_length - length1, length2 = ACNT
+		 *
+		 * When the full_length is multibple of 32767 one slot can be
+		 * used to complete the transfer.
+		 */
+		width = SZ_32K - 1;
+		pset_len = rounddown(len, width);
+		/* One slot is enough for lengths multiple of (SZ_32K -1) */
+		if (unlikely(pset_len == len))
+			nslots = 1;
+		else
+			nslots = 2;
+	}
+
+	edesc = kzalloc(sizeof(*edesc) + nslots * sizeof(edesc->pset[0]),
+			GFP_ATOMIC);
 	if (!edesc) {
 		dev_dbg(dev, "Failed to allocate a descriptor\n");
 		return NULL;
 	}
 
-	edesc->pset_nr = 1;
+	edesc->pset_nr = nslots;
+	edesc->residue = edesc->residue_stat = len;
+	edesc->direction = DMA_MEM_TO_MEM;
+	edesc->echan = echan;
 
 	ret = edma_config_pset(chan, &edesc->pset[0], src, dest, 1,
-			       DMA_SLAVE_BUSWIDTH_4_BYTES, len, DMA_MEM_TO_MEM);
-	if (ret < 0)
+			       width, pset_len, DMA_MEM_TO_MEM);
+	if (ret < 0) {
+		kfree(edesc);
 		return NULL;
+	}
 
 	edesc->absync = ret;
 
-	/*
-	 * Enable intermediate transfer chaining to re-trigger channel
-	 * on completion of every TR, and enable transfer-completion
-	 * interrupt on completion of the whole transfer.
-	 */
 	edesc->pset[0].param.opt |= ITCCHEN;
-	edesc->pset[0].param.opt |= TCINTEN;
+	if (nslots == 1) {
+		/* Enable transfer complete interrupt */
+		edesc->pset[0].param.opt |= TCINTEN;
+	} else {
+		/* Enable transfer complete chaining for the first slot */
+		edesc->pset[0].param.opt |= TCCHEN;
+
+		if (echan->slot[1] < 0) {
+			echan->slot[1] = edma_alloc_slot(echan->ecc,
+							 EDMA_SLOT_ANY);
+			if (echan->slot[1] < 0) {
+				kfree(edesc);
+				dev_err(dev, "%s: Failed to allocate slot\n",
+					__func__);
+				return NULL;
+			}
+		}
+		dest += pset_len;
+		src += pset_len;
+		pset_len = width = len % (SZ_32K - 1);
+
+		ret = edma_config_pset(chan, &edesc->pset[1], src, dest, 1,
+				       width, pset_len, DMA_MEM_TO_MEM);
+		if (ret < 0) {
+			kfree(edesc);
+			return NULL;
+		}
+
+		edesc->pset[1].param.opt |= ITCCHEN;
+		edesc->pset[1].param.opt |= TCINTEN;
+	}
 
 	return vchan_tx_prep(&echan->vchan, &edesc->vdesc, tx_flags);
 }
@@ -629,8 +1272,8 @@ static struct dma_async_tx_descriptor *edma_prep_dma_cyclic(
 	if (nslots > MAX_NR_SG)
 		return NULL;
 
-	edesc = kzalloc(sizeof(*edesc) + nslots *
-		sizeof(edesc->pset[0]), GFP_ATOMIC);
+	edesc = kzalloc(sizeof(*edesc) + nslots * sizeof(edesc->pset[0]),
+			GFP_ATOMIC);
 	if (!edesc) {
 		dev_err(dev, "%s: Failed to allocate a descriptor\n", __func__);
 		return NULL;
@@ -649,8 +1292,7 @@ static struct dma_async_tx_descriptor *edma_prep_dma_cyclic(
 		/* Allocate a PaRAM slot, if needed */
 		if (echan->slot[i] < 0) {
 			echan->slot[i] =
-				edma_alloc_slot(EDMA_CTLR(echan->ch_num),
-						EDMA_SLOT_ANY);
+				edma_alloc_slot(echan->ecc, EDMA_SLOT_ANY);
 			if (echan->slot[i] < 0) {
 				kfree(edesc);
 				dev_err(dev, "%s: Failed to allocate slot\n",
@@ -711,128 +1353,281 @@ static struct dma_async_tx_descriptor *edma_prep_dma_cyclic(
 	}
 
 	/* Place the cyclic channel to highest priority queue */
-	edma_assign_channel_eventq(echan->ch_num, EVENTQ_0);
+	if (!echan->tc)
+		edma_assign_channel_eventq(echan, EVENTQ_0);
 
 	return vchan_tx_prep(&echan->vchan, &edesc->vdesc, tx_flags);
 }
 
-static void edma_callback(unsigned ch_num, u16 ch_status, void *data)
+static void edma_completion_handler(struct edma_chan *echan)
 {
-	struct edma_chan *echan = data;
 	struct device *dev = echan->vchan.chan.device->dev;
-	struct edma_desc *edesc;
-	struct edmacc_param p;
+	struct edma_desc *edesc = echan->edesc;
 
-	edesc = echan->edesc;
+	if (!edesc)
+		return;
 
-	/* Pause the channel for non-cyclic */
-	if (!edesc || (edesc && !edesc->cyclic))
-		edma_pause(echan->ch_num);
-
-	switch (ch_status) {
-	case EDMA_DMA_COMPLETE:
-		spin_lock(&echan->vchan.lock);
-
-		if (edesc) {
-			if (edesc->cyclic) {
-				vchan_cyclic_callback(&edesc->vdesc);
-			} else if (edesc->processed == edesc->pset_nr) {
-				dev_dbg(dev, "Transfer complete, stopping channel %d\n", ch_num);
-				edesc->residue = 0;
-				edma_stop(echan->ch_num);
-				vchan_cookie_complete(&edesc->vdesc);
-				edma_execute(echan);
-			} else {
-				dev_dbg(dev, "Intermediate transfer complete on channel %d\n", ch_num);
-
-				/* Update statistics for tx_status */
-				edesc->residue -= edesc->sg_len;
-				edesc->residue_stat = edesc->residue;
-				edesc->processed_stat = edesc->processed;
-
-				edma_execute(echan);
-			}
+	spin_lock(&echan->vchan.lock);
+	if (edesc->cyclic) {
+		vchan_cyclic_callback(&edesc->vdesc);
+		spin_unlock(&echan->vchan.lock);
+		return;
+	} else if (edesc->processed == edesc->pset_nr) {
+		edesc->residue = 0;
+		edma_stop(echan);
+		vchan_cookie_complete(&edesc->vdesc);
+		echan->edesc = NULL;
+
+		dev_dbg(dev, "Transfer completed on channel %d\n",
+			echan->ch_num);
+	} else {
+		dev_dbg(dev, "Sub transfer completed on channel %d\n",
+			echan->ch_num);
+
+		edma_pause(echan);
+
+		/* Update statistics for tx_status */
+		edesc->residue -= edesc->sg_len;
+		edesc->residue_stat = edesc->residue;
+		edesc->processed_stat = edesc->processed;
+	}
+	edma_execute(echan);
+
+	spin_unlock(&echan->vchan.lock);
+}
+
+/* eDMA interrupt handler */
+static irqreturn_t dma_irq_handler(int irq, void *data)
+{
+	struct edma_cc *ecc = data;
+	int ctlr;
+	u32 sh_ier;
+	u32 sh_ipr;
+	u32 bank;
+
+	ctlr = ecc->id;
+	if (ctlr < 0)
+		return IRQ_NONE;
+
+	dev_vdbg(ecc->dev, "dma_irq_handler\n");
+
+	sh_ipr = edma_shadow0_read_array(ecc, SH_IPR, 0);
+	if (!sh_ipr) {
+		sh_ipr = edma_shadow0_read_array(ecc, SH_IPR, 1);
+		if (!sh_ipr)
+			return IRQ_NONE;
+		sh_ier = edma_shadow0_read_array(ecc, SH_IER, 1);
+		bank = 1;
+	} else {
+		sh_ier = edma_shadow0_read_array(ecc, SH_IER, 0);
+		bank = 0;
+	}
+
+	do {
+		u32 slot;
+		u32 channel;
+
+		slot = __ffs(sh_ipr);
+		sh_ipr &= ~(BIT(slot));
+
+		if (sh_ier & BIT(slot)) {
+			channel = (bank << 5) | slot;
+			/* Clear the corresponding IPR bits */
+			edma_shadow0_write_array(ecc, SH_ICR, bank, BIT(slot));
+			edma_completion_handler(&ecc->slave_chans[channel]);
 		}
+	} while (sh_ipr);
 
-		spin_unlock(&echan->vchan.lock);
+	edma_shadow0_write(ecc, SH_IEVAL, 1);
+	return IRQ_HANDLED;
+}
+
+static void edma_error_handler(struct edma_chan *echan)
+{
+	struct edma_cc *ecc = echan->ecc;
+	struct device *dev = echan->vchan.chan.device->dev;
+	struct edmacc_param p;
 
-		break;
-	case EDMA_DMA_CC_ERROR:
-		spin_lock(&echan->vchan.lock);
+	if (!echan->edesc)
+		return;
 
-		edma_read_slot(EDMA_CHAN_SLOT(echan->slot[0]), &p);
+	spin_lock(&echan->vchan.lock);
 
+	edma_read_slot(ecc, echan->slot[0], &p);
+	/*
+	 * Issue later based on missed flag which will be sure
+	 * to happen as:
+	 * (1) we finished transmitting an intermediate slot and
+	 *     edma_execute is coming up.
+	 * (2) or we finished current transfer and issue will
+	 *     call edma_execute.
+	 *
+	 * Important note: issuing can be dangerous here and
+	 * lead to some nasty recursion when we are in a NULL
+	 * slot. So we avoid doing so and set the missed flag.
+	 */
+	if (p.a_b_cnt == 0 && p.ccnt == 0) {
+		dev_dbg(dev, "Error on null slot, setting miss\n");
+		echan->missed = 1;
+	} else {
 		/*
-		 * Issue later based on missed flag which will be sure
-		 * to happen as:
-		 * (1) we finished transmitting an intermediate slot and
-		 *     edma_execute is coming up.
-		 * (2) or we finished current transfer and issue will
-		 *     call edma_execute.
-		 *
-		 * Important note: issuing can be dangerous here and
-		 * lead to some nasty recursion when we are in a NULL
-		 * slot. So we avoid doing so and set the missed flag.
+		 * The slot is already programmed but the event got
+		 * missed, so its safe to issue it here.
 		 */
-		if (p.a_b_cnt == 0 && p.ccnt == 0) {
-			dev_dbg(dev, "Error occurred, looks like slot is null, just setting miss\n");
-			echan->missed = 1;
-		} else {
-			/*
-			 * The slot is already programmed but the event got
-			 * missed, so its safe to issue it here.
-			 */
-			dev_dbg(dev, "Error occurred but slot is non-null, TRIGGERING\n");
-			edma_clean_channel(echan->ch_num);
-			edma_stop(echan->ch_num);
-			edma_start(echan->ch_num);
-			edma_trigger_channel(echan->ch_num);
+		dev_dbg(dev, "Missed event, TRIGGERING\n");
+		edma_clean_channel(echan);
+		edma_stop(echan);
+		edma_start(echan);
+		edma_trigger_channel(echan);
+	}
+	spin_unlock(&echan->vchan.lock);
+}
+
+static inline bool edma_error_pending(struct edma_cc *ecc)
+{
+	if (edma_read_array(ecc, EDMA_EMR, 0) ||
+	    edma_read_array(ecc, EDMA_EMR, 1) ||
+	    edma_read(ecc, EDMA_QEMR) || edma_read(ecc, EDMA_CCERR))
+		return true;
+
+	return false;
+}
+
+/* eDMA error interrupt handler */
+static irqreturn_t dma_ccerr_handler(int irq, void *data)
+{
+	struct edma_cc *ecc = data;
+	int i, j;
+	int ctlr;
+	unsigned int cnt = 0;
+	unsigned int val;
+
+	ctlr = ecc->id;
+	if (ctlr < 0)
+		return IRQ_NONE;
+
+	dev_vdbg(ecc->dev, "dma_ccerr_handler\n");
+
+	if (!edma_error_pending(ecc))
+		return IRQ_NONE;
+
+	while (1) {
+		/* Event missed register(s) */
+		for (j = 0; j < 2; j++) {
+			unsigned long emr;
+
+			val = edma_read_array(ecc, EDMA_EMR, j);
+			if (!val)
+				continue;
+
+			dev_dbg(ecc->dev, "EMR%d 0x%08x\n", j, val);
+			emr = val;
+			for (i = find_next_bit(&emr, 32, 0); i < 32;
+			     i = find_next_bit(&emr, 32, i + 1)) {
+				int k = (j << 5) + i;
+
+				/* Clear the corresponding EMR bits */
+				edma_write_array(ecc, EDMA_EMCR, j, BIT(i));
+				/* Clear any SER */
+				edma_shadow0_write_array(ecc, SH_SECR, j,
+							 BIT(i));
+				edma_error_handler(&ecc->slave_chans[k]);
+			}
 		}
 
-		spin_unlock(&echan->vchan.lock);
+		val = edma_read(ecc, EDMA_QEMR);
+		if (val) {
+			dev_dbg(ecc->dev, "QEMR 0x%02x\n", val);
+			/* Not reported, just clear the interrupt reason. */
+			edma_write(ecc, EDMA_QEMCR, val);
+			edma_shadow0_write(ecc, SH_QSECR, val);
+		}
+
+		val = edma_read(ecc, EDMA_CCERR);
+		if (val) {
+			dev_warn(ecc->dev, "CCERR 0x%08x\n", val);
+			/* Not reported, just clear the interrupt reason. */
+			edma_write(ecc, EDMA_CCERRCLR, val);
+		}
 
-		break;
-	default:
-		break;
+		if (!edma_error_pending(ecc))
+			break;
+		cnt++;
+		if (cnt > 10)
+			break;
 	}
+	edma_write(ecc, EDMA_EEVAL, 1);
+	return IRQ_HANDLED;
+}
+
+static void edma_tc_set_pm_state(struct edma_tc *tc, bool enable)
+{
+	struct platform_device *tc_pdev;
+	int ret;
+
+	if (!tc)
+		return;
+
+	tc_pdev = of_find_device_by_node(tc->node);
+	if (!tc_pdev) {
+		pr_err("%s: TPTC device is not found\n", __func__);
+		return;
+	}
+	if (!pm_runtime_enabled(&tc_pdev->dev))
+		pm_runtime_enable(&tc_pdev->dev);
+
+	if (enable)
+		ret = pm_runtime_get_sync(&tc_pdev->dev);
+	else
+		ret = pm_runtime_put_sync(&tc_pdev->dev);
+
+	if (ret < 0)
+		pr_err("%s: pm_runtime_%s_sync() failed for %s\n", __func__,
+		       enable ? "get" : "put", dev_name(&tc_pdev->dev));
 }
 
 /* Alloc channel resources */
 static int edma_alloc_chan_resources(struct dma_chan *chan)
 {
 	struct edma_chan *echan = to_edma_chan(chan);
-	struct device *dev = chan->device->dev;
+	struct edma_cc *ecc = echan->ecc;
+	struct device *dev = ecc->dev;
+	enum dma_event_q eventq_no = EVENTQ_DEFAULT;
 	int ret;
-	int a_ch_num;
-	LIST_HEAD(descs);
 
-	a_ch_num = edma_alloc_channel(echan->ch_num, edma_callback,
-					echan, EVENTQ_DEFAULT);
-
-	if (a_ch_num < 0) {
-		ret = -ENODEV;
-		goto err_no_chan;
+	if (echan->tc) {
+		eventq_no = echan->tc->id;
+	} else if (ecc->tc_list) {
+		/* memcpy channel */
+		echan->tc = &ecc->tc_list[ecc->info->default_queue];
+		eventq_no = echan->tc->id;
 	}
 
-	if (a_ch_num != echan->ch_num) {
-		dev_err(dev, "failed to allocate requested channel %u:%u\n",
-			EDMA_CTLR(echan->ch_num),
+	ret = edma_alloc_channel(echan, eventq_no);
+	if (ret)
+		return ret;
+
+	echan->slot[0] = edma_alloc_slot(ecc, echan->ch_num);
+	if (echan->slot[0] < 0) {
+		dev_err(dev, "Entry slot allocation failed for channel %u\n",
 			EDMA_CHAN_SLOT(echan->ch_num));
-		ret = -ENODEV;
-		goto err_wrong_chan;
+		goto err_slot;
 	}
 
+	/* Set up channel -> slot mapping for the entry slot */
+	edma_set_chmap(echan, echan->slot[0]);
 	echan->alloced = true;
-	echan->slot[0] = echan->ch_num;
 
-	dev_dbg(dev, "allocated channel %d for %u:%u\n", echan->ch_num,
-		EDMA_CTLR(echan->ch_num), EDMA_CHAN_SLOT(echan->ch_num));
+	dev_dbg(dev, "Got eDMA channel %d for virt channel %d (%s trigger)\n",
+		EDMA_CHAN_SLOT(echan->ch_num), chan->chan_id,
+		echan->hw_triggered ? "HW" : "SW");
+
+	edma_tc_set_pm_state(echan->tc, true);
 
 	return 0;
 
-err_wrong_chan:
-	edma_free_channel(a_ch_num);
-err_no_chan:
+err_slot:
+	edma_free_channel(echan);
 	return ret;
 }
 
@@ -840,29 +1635,37 @@ err_no_chan:
 static void edma_free_chan_resources(struct dma_chan *chan)
 {
 	struct edma_chan *echan = to_edma_chan(chan);
-	struct device *dev = chan->device->dev;
+	struct device *dev = echan->ecc->dev;
 	int i;
 
 	/* Terminate transfers */
-	edma_stop(echan->ch_num);
+	edma_stop(echan);
 
 	vchan_free_chan_resources(&echan->vchan);
 
 	/* Free EDMA PaRAM slots */
-	for (i = 1; i < EDMA_MAX_SLOTS; i++) {
+	for (i = 0; i < EDMA_MAX_SLOTS; i++) {
 		if (echan->slot[i] >= 0) {
-			edma_free_slot(echan->slot[i]);
+			edma_free_slot(echan->ecc, echan->slot[i]);
 			echan->slot[i] = -1;
 		}
 	}
 
+	/* Set entry slot to the dummy slot */
+	edma_set_chmap(echan, echan->ecc->dummy_slot);
+
 	/* Free EDMA channel */
 	if (echan->alloced) {
-		edma_free_channel(echan->ch_num);
+		edma_free_channel(echan);
 		echan->alloced = false;
 	}
 
-	dev_dbg(dev, "freeing channel for %u\n", echan->ch_num);
+	edma_tc_set_pm_state(echan->tc, false);
+	echan->tc = NULL;
+	echan->hw_triggered = false;
+
+	dev_dbg(dev, "Free eDMA channel %d for virt channel %d\n",
+		EDMA_CHAN_SLOT(echan->ch_num), chan->chan_id);
 }
 
 /* Send pending descriptor to hardware */
@@ -888,7 +1691,7 @@ static u32 edma_residue(struct edma_desc *edesc)
 	 * We always read the dst/src position from the first RamPar
 	 * pset. That's the one which is active now.
 	 */
-	pos = edma_get_position(edesc->echan->slot[0], dst);
+	pos = edma_get_position(edesc->echan->ecc, edesc->echan->slot[0], dst);
 
 	/*
 	 * Cyclic is simple. Just subtract pset[0].addr from pos.
@@ -949,19 +1752,101 @@ static enum dma_status edma_tx_status(struct dma_chan *chan,
 	return ret;
 }
 
-static void __init edma_chan_init(struct edma_cc *ecc,
-				  struct dma_device *dma,
-				  struct edma_chan *echans)
+static bool edma_is_memcpy_channel(int ch_num, u16 *memcpy_channels)
 {
+	s16 *memcpy_ch = memcpy_channels;
+
+	if (!memcpy_channels)
+		return false;
+	while (*memcpy_ch != -1) {
+		if (*memcpy_ch == ch_num)
+			return true;
+		memcpy_ch++;
+	}
+	return false;
+}
+
+#define EDMA_DMA_BUSWIDTHS	(BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) | \
+				 BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) | \
+				 BIT(DMA_SLAVE_BUSWIDTH_3_BYTES) | \
+				 BIT(DMA_SLAVE_BUSWIDTH_4_BYTES))
+
+static void edma_dma_init(struct edma_cc *ecc, bool legacy_mode)
+{
+	struct dma_device *s_ddev = &ecc->dma_slave;
+	struct dma_device *m_ddev = NULL;
+	s16 *memcpy_channels = ecc->info->memcpy_channels;
 	int i, j;
 
-	for (i = 0; i < EDMA_CHANS; i++) {
-		struct edma_chan *echan = &echans[i];
-		echan->ch_num = EDMA_CTLR_CHAN(ecc->ctlr, i);
+	dma_cap_zero(s_ddev->cap_mask);
+	dma_cap_set(DMA_SLAVE, s_ddev->cap_mask);
+	dma_cap_set(DMA_CYCLIC, s_ddev->cap_mask);
+	if (ecc->legacy_mode && !memcpy_channels) {
+		dev_warn(ecc->dev,
+			 "Legacy memcpy is enabled, things might not work\n");
+
+		dma_cap_set(DMA_MEMCPY, s_ddev->cap_mask);
+		s_ddev->device_prep_dma_memcpy = edma_prep_dma_memcpy;
+		s_ddev->directions = BIT(DMA_MEM_TO_MEM);
+	}
+
+	s_ddev->device_prep_slave_sg = edma_prep_slave_sg;
+	s_ddev->device_prep_dma_cyclic = edma_prep_dma_cyclic;
+	s_ddev->device_alloc_chan_resources = edma_alloc_chan_resources;
+	s_ddev->device_free_chan_resources = edma_free_chan_resources;
+	s_ddev->device_issue_pending = edma_issue_pending;
+	s_ddev->device_tx_status = edma_tx_status;
+	s_ddev->device_config = edma_slave_config;
+	s_ddev->device_pause = edma_dma_pause;
+	s_ddev->device_resume = edma_dma_resume;
+	s_ddev->device_terminate_all = edma_terminate_all;
+
+	s_ddev->src_addr_widths = EDMA_DMA_BUSWIDTHS;
+	s_ddev->dst_addr_widths = EDMA_DMA_BUSWIDTHS;
+	s_ddev->directions |= (BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV));
+	s_ddev->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+
+	s_ddev->dev = ecc->dev;
+	INIT_LIST_HEAD(&s_ddev->channels);
+
+	if (memcpy_channels) {
+		m_ddev = devm_kzalloc(ecc->dev, sizeof(*m_ddev), GFP_KERNEL);
+		ecc->dma_memcpy = m_ddev;
+
+		dma_cap_zero(m_ddev->cap_mask);
+		dma_cap_set(DMA_MEMCPY, m_ddev->cap_mask);
+
+		m_ddev->device_prep_dma_memcpy = edma_prep_dma_memcpy;
+		m_ddev->device_alloc_chan_resources = edma_alloc_chan_resources;
+		m_ddev->device_free_chan_resources = edma_free_chan_resources;
+		m_ddev->device_issue_pending = edma_issue_pending;
+		m_ddev->device_tx_status = edma_tx_status;
+		m_ddev->device_config = edma_slave_config;
+		m_ddev->device_pause = edma_dma_pause;
+		m_ddev->device_resume = edma_dma_resume;
+		m_ddev->device_terminate_all = edma_terminate_all;
+
+		m_ddev->src_addr_widths = EDMA_DMA_BUSWIDTHS;
+		m_ddev->dst_addr_widths = EDMA_DMA_BUSWIDTHS;
+		m_ddev->directions = BIT(DMA_MEM_TO_MEM);
+		m_ddev->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+
+		m_ddev->dev = ecc->dev;
+		INIT_LIST_HEAD(&m_ddev->channels);
+	} else if (!ecc->legacy_mode) {
+		dev_info(ecc->dev, "memcpy is disabled\n");
+	}
+
+	for (i = 0; i < ecc->num_channels; i++) {
+		struct edma_chan *echan = &ecc->slave_chans[i];
+		echan->ch_num = EDMA_CTLR_CHAN(ecc->id, i);
 		echan->ecc = ecc;
 		echan->vchan.desc_free = edma_desc_free;
 
-		vchan_init(&echan->vchan, dma);
+		if (m_ddev && edma_is_memcpy_channel(i, memcpy_channels))
+			vchan_init(&echan->vchan, m_ddev);
+		else
+			vchan_init(&echan->vchan, s_ddev);
 
 		INIT_LIST_HEAD(&echan->node);
 		for (j = 0; j < EDMA_MAX_SLOTS; j++)
@@ -969,85 +1854,474 @@ static void __init edma_chan_init(struct edma_cc *ecc,
 	}
 }
 
-#define EDMA_DMA_BUSWIDTHS	(BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) | \
-				 BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) | \
-				 BIT(DMA_SLAVE_BUSWIDTH_3_BYTES) | \
-				 BIT(DMA_SLAVE_BUSWIDTH_4_BYTES))
-
-static void edma_dma_init(struct edma_cc *ecc, struct dma_device *dma,
-			  struct device *dev)
+static int edma_setup_from_hw(struct device *dev, struct edma_soc_info *pdata,
+			      struct edma_cc *ecc)
 {
-	dma->device_prep_slave_sg = edma_prep_slave_sg;
-	dma->device_prep_dma_cyclic = edma_prep_dma_cyclic;
-	dma->device_prep_dma_memcpy = edma_prep_dma_memcpy;
-	dma->device_alloc_chan_resources = edma_alloc_chan_resources;
-	dma->device_free_chan_resources = edma_free_chan_resources;
-	dma->device_issue_pending = edma_issue_pending;
-	dma->device_tx_status = edma_tx_status;
-	dma->device_config = edma_slave_config;
-	dma->device_pause = edma_dma_pause;
-	dma->device_resume = edma_dma_resume;
-	dma->device_terminate_all = edma_terminate_all;
+	int i;
+	u32 value, cccfg;
+	s8 (*queue_priority_map)[2];
+
+	/* Decode the eDMA3 configuration from CCCFG register */
+	cccfg = edma_read(ecc, EDMA_CCCFG);
+
+	value = GET_NUM_REGN(cccfg);
+	ecc->num_region = BIT(value);
+
+	value = GET_NUM_DMACH(cccfg);
+	ecc->num_channels = BIT(value + 1);
 
-	dma->src_addr_widths = EDMA_DMA_BUSWIDTHS;
-	dma->dst_addr_widths = EDMA_DMA_BUSWIDTHS;
-	dma->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
-	dma->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
+	value = GET_NUM_QDMACH(cccfg);
+	ecc->num_qchannels = value * 2;
 
-	dma->dev = dev;
+	value = GET_NUM_PAENTRY(cccfg);
+	ecc->num_slots = BIT(value + 4);
+
+	value = GET_NUM_EVQUE(cccfg);
+	ecc->num_tc = value + 1;
+
+	ecc->chmap_exist = (cccfg & CHMAP_EXIST) ? true : false;
+
+	dev_dbg(dev, "eDMA3 CC HW configuration (cccfg: 0x%08x):\n", cccfg);
+	dev_dbg(dev, "num_region: %u\n", ecc->num_region);
+	dev_dbg(dev, "num_channels: %u\n", ecc->num_channels);
+	dev_dbg(dev, "num_qchannels: %u\n", ecc->num_qchannels);
+	dev_dbg(dev, "num_slots: %u\n", ecc->num_slots);
+	dev_dbg(dev, "num_tc: %u\n", ecc->num_tc);
+	dev_dbg(dev, "chmap_exist: %s\n", ecc->chmap_exist ? "yes" : "no");
+
+	/* Nothing need to be done if queue priority is provided */
+	if (pdata->queue_priority_mapping)
+		return 0;
 
 	/*
-	 * code using dma memcpy must make sure alignment of
-	 * length is at dma->copy_align boundary.
+	 * Configure TC/queue priority as follows:
+	 * Q0 - priority 0
+	 * Q1 - priority 1
+	 * Q2 - priority 2
+	 * ...
+	 * The meaning of priority numbers: 0 highest priority, 7 lowest
+	 * priority. So Q0 is the highest priority queue and the last queue has
+	 * the lowest priority.
 	 */
-	dma->copy_align = DMAENGINE_ALIGN_4_BYTES;
+	queue_priority_map = devm_kcalloc(dev, ecc->num_tc + 1, sizeof(s8),
+					  GFP_KERNEL);
+	if (!queue_priority_map)
+		return -ENOMEM;
+
+	for (i = 0; i < ecc->num_tc; i++) {
+		queue_priority_map[i][0] = i;
+		queue_priority_map[i][1] = i;
+	}
+	queue_priority_map[i][0] = -1;
+	queue_priority_map[i][1] = -1;
+
+	pdata->queue_priority_mapping = queue_priority_map;
+	/* Default queue has the lowest priority */
+	pdata->default_queue = i - 1;
+
+	return 0;
+}
+
+#if IS_ENABLED(CONFIG_OF)
+static int edma_xbar_event_map(struct device *dev, struct edma_soc_info *pdata,
+			       size_t sz)
+{
+	const char pname[] = "ti,edma-xbar-event-map";
+	struct resource res;
+	void __iomem *xbar;
+	s16 (*xbar_chans)[2];
+	size_t nelm = sz / sizeof(s16);
+	u32 shift, offset, mux;
+	int ret, i;
+
+	xbar_chans = devm_kcalloc(dev, nelm + 2, sizeof(s16), GFP_KERNEL);
+	if (!xbar_chans)
+		return -ENOMEM;
+
+	ret = of_address_to_resource(dev->of_node, 1, &res);
+	if (ret)
+		return -ENOMEM;
+
+	xbar = devm_ioremap(dev, res.start, resource_size(&res));
+	if (!xbar)
+		return -ENOMEM;
+
+	ret = of_property_read_u16_array(dev->of_node, pname, (u16 *)xbar_chans,
+					 nelm);
+	if (ret)
+		return -EIO;
+
+	/* Invalidate last entry for the other user of this mess */
+	nelm >>= 1;
+	xbar_chans[nelm][0] = -1;
+	xbar_chans[nelm][1] = -1;
+
+	for (i = 0; i < nelm; i++) {
+		shift = (xbar_chans[i][1] & 0x03) << 3;
+		offset = xbar_chans[i][1] & 0xfffffffc;
+		mux = readl(xbar + offset);
+		mux &= ~(0xff << shift);
+		mux |= xbar_chans[i][0] << shift;
+		writel(mux, (xbar + offset));
+	}
 
-	INIT_LIST_HEAD(&dma->channels);
+	pdata->xbar_chans = (const s16 (*)[2]) xbar_chans;
+	return 0;
+}
+
+static struct edma_soc_info *edma_setup_info_from_dt(struct device *dev,
+						     bool legacy_mode)
+{
+	struct edma_soc_info *info;
+	struct property *prop;
+	size_t sz;
+	int ret;
+
+	info = devm_kzalloc(dev, sizeof(struct edma_soc_info), GFP_KERNEL);
+	if (!info)
+		return ERR_PTR(-ENOMEM);
+
+	if (legacy_mode) {
+		prop = of_find_property(dev->of_node, "ti,edma-xbar-event-map",
+					&sz);
+		if (prop) {
+			ret = edma_xbar_event_map(dev, info, sz);
+			if (ret)
+				return ERR_PTR(ret);
+		}
+		return info;
+	}
+
+	/* Get the list of channels allocated to be used for memcpy */
+	prop = of_find_property(dev->of_node, "ti,edma-memcpy-channels", &sz);
+	if (prop) {
+		const char pname[] = "ti,edma-memcpy-channels";
+		size_t nelm = sz / sizeof(s16);
+		s16 *memcpy_ch;
+
+		memcpy_ch = devm_kcalloc(dev, nelm + 1, sizeof(s16),
+					 GFP_KERNEL);
+		if (!memcpy_ch)
+			return ERR_PTR(-ENOMEM);
+
+		ret = of_property_read_u16_array(dev->of_node, pname,
+						 (u16 *)memcpy_ch, nelm);
+		if (ret)
+			return ERR_PTR(ret);
+
+		memcpy_ch[nelm] = -1;
+		info->memcpy_channels = memcpy_ch;
+	}
+
+	prop = of_find_property(dev->of_node, "ti,edma-reserved-slot-ranges",
+				&sz);
+	if (prop) {
+		const char pname[] = "ti,edma-reserved-slot-ranges";
+		s16 (*rsv_slots)[2];
+		size_t nelm = sz / sizeof(*rsv_slots);
+		struct edma_rsv_info *rsv_info;
+
+		if (!nelm)
+			return info;
+
+		rsv_info = devm_kzalloc(dev, sizeof(*rsv_info), GFP_KERNEL);
+		if (!rsv_info)
+			return ERR_PTR(-ENOMEM);
+
+		rsv_slots = devm_kcalloc(dev, nelm + 1, sizeof(*rsv_slots),
+					 GFP_KERNEL);
+		if (!rsv_slots)
+			return ERR_PTR(-ENOMEM);
+
+		ret = of_property_read_u16_array(dev->of_node, pname,
+						 (u16 *)rsv_slots, nelm * 2);
+		if (ret)
+			return ERR_PTR(ret);
+
+		rsv_slots[nelm][0] = -1;
+		rsv_slots[nelm][1] = -1;
+		info->rsv = rsv_info;
+		info->rsv->rsv_slots = (const s16 (*)[2])rsv_slots;
+	}
+
+	return info;
+}
+
+static struct dma_chan *of_edma_xlate(struct of_phandle_args *dma_spec,
+				      struct of_dma *ofdma)
+{
+	struct edma_cc *ecc = ofdma->of_dma_data;
+	struct dma_chan *chan = NULL;
+	struct edma_chan *echan;
+	int i;
+
+	if (!ecc || dma_spec->args_count < 1)
+		return NULL;
+
+	for (i = 0; i < ecc->num_channels; i++) {
+		echan = &ecc->slave_chans[i];
+		if (echan->ch_num == dma_spec->args[0]) {
+			chan = &echan->vchan.chan;
+			break;
+		}
+	}
+
+	if (!chan)
+		return NULL;
+
+	if (echan->ecc->legacy_mode && dma_spec->args_count == 1)
+		goto out;
+
+	if (!echan->ecc->legacy_mode && dma_spec->args_count == 2 &&
+	    dma_spec->args[1] < echan->ecc->num_tc) {
+		echan->tc = &echan->ecc->tc_list[dma_spec->args[1]];
+		goto out;
+	}
+
+	return NULL;
+out:
+	/* The channel is going to be used as HW synchronized */
+	echan->hw_triggered = true;
+	return dma_get_slave_channel(chan);
+}
+#else
+static struct edma_soc_info *edma_setup_info_from_dt(struct device *dev,
+						     bool legacy_mode)
+{
+	return ERR_PTR(-EINVAL);
+}
+
+static struct dma_chan *of_edma_xlate(struct of_phandle_args *dma_spec,
+				      struct of_dma *ofdma)
+{
+	return NULL;
 }
+#endif
 
 static int edma_probe(struct platform_device *pdev)
 {
-	struct edma_cc *ecc;
+	struct edma_soc_info	*info = pdev->dev.platform_data;
+	s8			(*queue_priority_mapping)[2];
+	int			i, off, ln;
+	const s16		(*rsv_slots)[2];
+	const s16		(*xbar_chans)[2];
+	int			irq;
+	char			*irq_name;
+	struct resource		*mem;
+	struct device_node	*node = pdev->dev.of_node;
+	struct device		*dev = &pdev->dev;
+	struct edma_cc		*ecc;
+	bool			legacy_mode = true;
 	int ret;
 
-	ret = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32));
+	if (node) {
+		const struct of_device_id *match;
+
+		match = of_match_node(edma_of_ids, node);
+		if (match && (u32)match->data == EDMA_BINDING_TPCC)
+			legacy_mode = false;
+
+		info = edma_setup_info_from_dt(dev, legacy_mode);
+		if (IS_ERR(info)) {
+			dev_err(dev, "failed to get DT data\n");
+			return PTR_ERR(info);
+		}
+	}
+
+	if (!info)
+		return -ENODEV;
+
+	pm_runtime_enable(dev);
+	ret = pm_runtime_get_sync(dev);
+	if (ret < 0) {
+		dev_err(dev, "pm_runtime_get_sync() failed\n");
+		return ret;
+	}
+
+	ret = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(32));
 	if (ret)
 		return ret;
 
-	ecc = devm_kzalloc(&pdev->dev, sizeof(*ecc), GFP_KERNEL);
+	ecc = devm_kzalloc(dev, sizeof(*ecc), GFP_KERNEL);
 	if (!ecc) {
-		dev_err(&pdev->dev, "Can't allocate controller\n");
+		dev_err(dev, "Can't allocate controller\n");
 		return -ENOMEM;
 	}
 
-	ecc->ctlr = pdev->id;
-	ecc->dummy_slot = edma_alloc_slot(ecc->ctlr, EDMA_SLOT_ANY);
+	ecc->dev = dev;
+	ecc->id = pdev->id;
+	ecc->legacy_mode = legacy_mode;
+	/* When booting with DT the pdev->id is -1 */
+	if (ecc->id < 0)
+		ecc->id = 0;
+
+	mem = platform_get_resource_byname(pdev, IORESOURCE_MEM, "edma3_cc");
+	if (!mem) {
+		dev_dbg(dev, "mem resource not found, using index 0\n");
+		mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+		if (!mem) {
+			dev_err(dev, "no mem resource?\n");
+			return -ENODEV;
+		}
+	}
+	ecc->base = devm_ioremap_resource(dev, mem);
+	if (IS_ERR(ecc->base))
+		return PTR_ERR(ecc->base);
+
+	platform_set_drvdata(pdev, ecc);
+
+	/* Get eDMA3 configuration from IP */
+	ret = edma_setup_from_hw(dev, info, ecc);
+	if (ret)
+		return ret;
+
+	/* Allocate memory based on the information we got from the IP */
+	ecc->slave_chans = devm_kcalloc(dev, ecc->num_channels,
+					sizeof(*ecc->slave_chans), GFP_KERNEL);
+	if (!ecc->slave_chans)
+		return -ENOMEM;
+
+	ecc->slot_inuse = devm_kcalloc(dev, BITS_TO_LONGS(ecc->num_slots),
+				       sizeof(unsigned long), GFP_KERNEL);
+	if (!ecc->slot_inuse)
+		return -ENOMEM;
+
+	ecc->default_queue = info->default_queue;
+
+	for (i = 0; i < ecc->num_slots; i++)
+		edma_write_slot(ecc, i, &dummy_paramset);
+
+	if (info->rsv) {
+		/* Set the reserved slots in inuse list */
+		rsv_slots = info->rsv->rsv_slots;
+		if (rsv_slots) {
+			for (i = 0; rsv_slots[i][0] != -1; i++) {
+				off = rsv_slots[i][0];
+				ln = rsv_slots[i][1];
+				set_bits(off, ln, ecc->slot_inuse);
+			}
+		}
+	}
+
+	/* Clear the xbar mapped channels in unused list */
+	xbar_chans = info->xbar_chans;
+	if (xbar_chans) {
+		for (i = 0; xbar_chans[i][1] != -1; i++) {
+			off = xbar_chans[i][1];
+		}
+	}
+
+	irq = platform_get_irq_byname(pdev, "edma3_ccint");
+	if (irq < 0 && node)
+		irq = irq_of_parse_and_map(node, 0);
+
+	if (irq >= 0) {
+		irq_name = devm_kasprintf(dev, GFP_KERNEL, "%s_ccint",
+					  dev_name(dev));
+		ret = devm_request_irq(dev, irq, dma_irq_handler, 0, irq_name,
+				       ecc);
+		if (ret) {
+			dev_err(dev, "CCINT (%d) failed --> %d\n", irq, ret);
+			return ret;
+		}
+	}
+
+	irq = platform_get_irq_byname(pdev, "edma3_ccerrint");
+	if (irq < 0 && node)
+		irq = irq_of_parse_and_map(node, 2);
+
+	if (irq >= 0) {
+		irq_name = devm_kasprintf(dev, GFP_KERNEL, "%s_ccerrint",
+					  dev_name(dev));
+		ret = devm_request_irq(dev, irq, dma_ccerr_handler, 0, irq_name,
+				       ecc);
+		if (ret) {
+			dev_err(dev, "CCERRINT (%d) failed --> %d\n", irq, ret);
+			return ret;
+		}
+	}
+
+	ecc->dummy_slot = edma_alloc_slot(ecc, EDMA_SLOT_ANY);
 	if (ecc->dummy_slot < 0) {
-		dev_err(&pdev->dev, "Can't allocate PaRAM dummy slot\n");
+		dev_err(dev, "Can't allocate PaRAM dummy slot\n");
 		return ecc->dummy_slot;
 	}
 
-	dma_cap_zero(ecc->dma_slave.cap_mask);
-	dma_cap_set(DMA_SLAVE, ecc->dma_slave.cap_mask);
-	dma_cap_set(DMA_CYCLIC, ecc->dma_slave.cap_mask);
-	dma_cap_set(DMA_MEMCPY, ecc->dma_slave.cap_mask);
+	queue_priority_mapping = info->queue_priority_mapping;
+
+	if (!ecc->legacy_mode) {
+		int lowest_priority = 0;
+		struct of_phandle_args tc_args;
+
+		ecc->tc_list = devm_kcalloc(dev, ecc->num_tc,
+					    sizeof(*ecc->tc_list), GFP_KERNEL);
+		if (!ecc->tc_list)
+			return -ENOMEM;
+
+		for (i = 0;; i++) {
+			ret = of_parse_phandle_with_fixed_args(node, "ti,tptcs",
+							       1, i, &tc_args);
+			if (ret || i == ecc->num_tc)
+				break;
+
+			ecc->tc_list[i].node = tc_args.np;
+			ecc->tc_list[i].id = i;
+			queue_priority_mapping[i][1] = tc_args.args[0];
+			if (queue_priority_mapping[i][1] > lowest_priority) {
+				lowest_priority = queue_priority_mapping[i][1];
+				info->default_queue = i;
+			}
+		}
+	}
+
+	/* Event queue priority mapping */
+	for (i = 0; queue_priority_mapping[i][0] != -1; i++)
+		edma_assign_priority_to_queue(ecc, queue_priority_mapping[i][0],
+					      queue_priority_mapping[i][1]);
+
+	for (i = 0; i < ecc->num_region; i++) {
+		edma_write_array2(ecc, EDMA_DRAE, i, 0, 0x0);
+		edma_write_array2(ecc, EDMA_DRAE, i, 1, 0x0);
+		edma_write_array(ecc, EDMA_QRAE, i, 0x0);
+	}
+	ecc->info = info;
 
-	edma_dma_init(ecc, &ecc->dma_slave, &pdev->dev);
+	/* Init the dma device and channels */
+	edma_dma_init(ecc, legacy_mode);
 
-	edma_chan_init(ecc, &ecc->dma_slave, ecc->slave_chans);
+	for (i = 0; i < ecc->num_channels; i++) {
+		/* Assign all channels to the default queue */
+		edma_assign_channel_eventq(&ecc->slave_chans[i],
+					   info->default_queue);
+		/* Set entry slot to the dummy slot */
+		edma_set_chmap(&ecc->slave_chans[i], ecc->dummy_slot);
+	}
 
 	ret = dma_async_device_register(&ecc->dma_slave);
-	if (ret)
+	if (ret) {
+		dev_err(dev, "slave ddev registration failed (%d)\n", ret);
 		goto err_reg1;
+	}
 
-	platform_set_drvdata(pdev, ecc);
+	if (ecc->dma_memcpy) {
+		ret = dma_async_device_register(ecc->dma_memcpy);
+		if (ret) {
+			dev_err(dev, "memcpy ddev registration failed (%d)\n",
+				ret);
+			dma_async_device_unregister(&ecc->dma_slave);
+			goto err_reg1;
+		}
+	}
+
+	if (node)
+		of_dma_controller_register(node, of_edma_xlate, ecc);
 
-	dev_info(&pdev->dev, "TI EDMA DMA engine driver\n");
+	dev_info(dev, "TI EDMA DMA engine driver\n");
 
 	return 0;
 
 err_reg1:
-	edma_free_slot(ecc->dummy_slot);
+	edma_free_slot(ecc, ecc->dummy_slot);
 	return ret;
 }
 
@@ -1056,33 +2330,112 @@ static int edma_remove(struct platform_device *pdev)
 	struct device *dev = &pdev->dev;
 	struct edma_cc *ecc = dev_get_drvdata(dev);
 
+	if (dev->of_node)
+		of_dma_controller_free(dev->of_node);
 	dma_async_device_unregister(&ecc->dma_slave);
-	edma_free_slot(ecc->dummy_slot);
+	if (ecc->dma_memcpy)
+		dma_async_device_unregister(ecc->dma_memcpy);
+	edma_free_slot(ecc, ecc->dummy_slot);
 
 	return 0;
 }
 
+#ifdef CONFIG_PM_SLEEP
+static int edma_pm_suspend(struct device *dev)
+{
+	struct edma_cc *ecc = dev_get_drvdata(dev);
+	struct edma_chan *echan = ecc->slave_chans;
+	int i;
+
+	for (i = 0; i < ecc->num_channels; i++) {
+		if (echan[i].alloced) {
+			edma_setup_interrupt(&echan[i], false);
+			edma_tc_set_pm_state(echan[i].tc, false);
+		}
+	}
+
+	return 0;
+}
+
+static int edma_pm_resume(struct device *dev)
+{
+	struct edma_cc *ecc = dev_get_drvdata(dev);
+	struct edma_chan *echan = ecc->slave_chans;
+	int i;
+	s8 (*queue_priority_mapping)[2];
+
+	queue_priority_mapping = ecc->info->queue_priority_mapping;
+
+	/* Event queue priority mapping */
+	for (i = 0; queue_priority_mapping[i][0] != -1; i++)
+		edma_assign_priority_to_queue(ecc, queue_priority_mapping[i][0],
+					      queue_priority_mapping[i][1]);
+
+	for (i = 0; i < ecc->num_channels; i++) {
+		if (echan[i].alloced) {
+			/* ensure access through shadow region 0 */
+			edma_or_array2(ecc, EDMA_DRAE, 0, i >> 5,
+				       BIT(i & 0x1f));
+
+			edma_setup_interrupt(&echan[i], true);
+
+			/* Set up channel -> slot mapping for the entry slot */
+			edma_set_chmap(&echan[i], echan[i].slot[0]);
+
+			edma_tc_set_pm_state(echan[i].tc, true);
+		}
+	}
+
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops edma_pm_ops = {
+	SET_LATE_SYSTEM_SLEEP_PM_OPS(edma_pm_suspend, edma_pm_resume)
+};
+
 static struct platform_driver edma_driver = {
 	.probe		= edma_probe,
 	.remove		= edma_remove,
 	.driver = {
-		.name = "edma-dma-engine",
+		.name	= "edma",
+		.pm	= &edma_pm_ops,
+		.of_match_table = edma_of_ids,
+	},
+};
+
+static struct platform_driver edma_tptc_driver = {
+	.driver = {
+		.name	= "edma3-tptc",
+		.of_match_table = edma_tptc_of_ids,
 	},
 };
 
 bool edma_filter_fn(struct dma_chan *chan, void *param)
 {
+	bool match = false;
+
 	if (chan->device->dev->driver == &edma_driver.driver) {
 		struct edma_chan *echan = to_edma_chan(chan);
 		unsigned ch_req = *(unsigned *)param;
-		return ch_req == echan->ch_num;
+		if (ch_req == echan->ch_num) {
+			/* The channel is going to be used as HW synchronized */
+			echan->hw_triggered = true;
+			match = true;
+		}
 	}
-	return false;
+	return match;
 }
 EXPORT_SYMBOL(edma_filter_fn);
 
 static int edma_init(void)
 {
+	int ret;
+
+	ret = platform_driver_register(&edma_tptc_driver);
+	if (ret)
+		return ret;
+
 	return platform_driver_register(&edma_driver);
 }
 subsys_initcall(edma_init);
@@ -1090,6 +2443,7 @@ subsys_initcall(edma_init);
 static void __exit edma_exit(void)
 {
 	platform_driver_unregister(&edma_driver);
+	platform_driver_unregister(&edma_tptc_driver);
 }
 module_exit(edma_exit);
 
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index 300f821..2209f75 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -1512,6 +1512,7 @@ static const struct of_device_id fsldma_of_ids[] = {
 	{ .compatible = "fsl,elo-dma", },
 	{}
 };
+MODULE_DEVICE_TABLE(of, fsldma_of_ids);
 
 static struct platform_driver fsldma_of_driver = {
 	.driver = {
diff --git a/drivers/dma/idma64.c b/drivers/dma/idma64.c
index 48d6d9e..7d56b47 100644
--- a/drivers/dma/idma64.c
+++ b/drivers/dma/idma64.c
@@ -65,9 +65,6 @@ static void idma64_chan_init(struct idma64 *idma64, struct idma64_chan *idma64c)
 	u32 cfghi = IDMA64C_CFGH_SRC_PER(1) | IDMA64C_CFGH_DST_PER(0);
 	u32 cfglo = 0;
 
-	/* Enforce FIFO drain when channel is suspended */
-	cfglo |= IDMA64C_CFGL_CH_DRAIN;
-
 	/* Set default burst alignment */
 	cfglo |= IDMA64C_CFGL_DST_BURST_ALIGN | IDMA64C_CFGL_SRC_BURST_ALIGN;
 
@@ -257,15 +254,15 @@ static u64 idma64_hw_desc_fill(struct idma64_hw_desc *hw,
 		dar = config->dst_addr;
 		ctllo |= IDMA64C_CTLL_DST_FIX | IDMA64C_CTLL_SRC_INC |
 			 IDMA64C_CTLL_FC_M2P;
-		src_width = min_t(u32, 2, __fls(sar | hw->len));
-		dst_width = __fls(config->dst_addr_width);
+		src_width = __ffs(sar | hw->len | 4);
+		dst_width = __ffs(config->dst_addr_width);
 	} else {	/* DMA_DEV_TO_MEM */
 		sar = config->src_addr;
 		dar = hw->phys;
 		ctllo |= IDMA64C_CTLL_DST_INC | IDMA64C_CTLL_SRC_FIX |
 			 IDMA64C_CTLL_FC_P2M;
-		src_width = __fls(config->src_addr_width);
-		dst_width = min_t(u32, 2, __fls(dar | hw->len));
+		src_width = __ffs(config->src_addr_width);
+		dst_width = __ffs(dar | hw->len | 4);
 	}
 
 	lli->sar = sar;
@@ -428,12 +425,17 @@ static int idma64_slave_config(struct dma_chan *chan,
 	return 0;
 }
 
-static void idma64_chan_deactivate(struct idma64_chan *idma64c)
+static void idma64_chan_deactivate(struct idma64_chan *idma64c, bool drain)
 {
 	unsigned short count = 100;
 	u32 cfglo;
 
 	cfglo = channel_readl(idma64c, CFG_LO);
+	if (drain)
+		cfglo |= IDMA64C_CFGL_CH_DRAIN;
+	else
+		cfglo &= ~IDMA64C_CFGL_CH_DRAIN;
+
 	channel_writel(idma64c, CFG_LO, cfglo | IDMA64C_CFGL_CH_SUSP);
 	do {
 		udelay(1);
@@ -456,7 +458,7 @@ static int idma64_pause(struct dma_chan *chan)
 
 	spin_lock_irqsave(&idma64c->vchan.lock, flags);
 	if (idma64c->desc && idma64c->desc->status == DMA_IN_PROGRESS) {
-		idma64_chan_deactivate(idma64c);
+		idma64_chan_deactivate(idma64c, false);
 		idma64c->desc->status = DMA_PAUSED;
 	}
 	spin_unlock_irqrestore(&idma64c->vchan.lock, flags);
@@ -486,7 +488,7 @@ static int idma64_terminate_all(struct dma_chan *chan)
 	LIST_HEAD(head);
 
 	spin_lock_irqsave(&idma64c->vchan.lock, flags);
-	idma64_chan_deactivate(idma64c);
+	idma64_chan_deactivate(idma64c, true);
 	idma64_stop_transfer(idma64c);
 	if (idma64c->desc) {
 		idma64_vdesc_free(&idma64c->desc->vdesc);
diff --git a/drivers/dma/idma64.h b/drivers/dma/idma64.h
index a4d9968..f6aeff0 100644
--- a/drivers/dma/idma64.h
+++ b/drivers/dma/idma64.h
@@ -16,6 +16,8 @@
 #include <linux/spinlock.h>
 #include <linux/types.h>
 
+#include <asm-generic/io-64-nonatomic-lo-hi.h>
+
 #include "virt-dma.h"
 
 /* Channel registers */
@@ -166,19 +168,13 @@ static inline void idma64c_writel(struct idma64_chan *idma64c, int offset,
 
 static inline u64 idma64c_readq(struct idma64_chan *idma64c, int offset)
 {
-	u64 l, h;
-
-	l = idma64c_readl(idma64c, offset);
-	h = idma64c_readl(idma64c, offset + 4);
-
-	return l | (h << 32);
+	return lo_hi_readq(idma64c->regs + offset);
 }
 
 static inline void idma64c_writeq(struct idma64_chan *idma64c, int offset,
 				  u64 value)
 {
-	idma64c_writel(idma64c, offset, value);
-	idma64c_writel(idma64c, offset + 4, value >> 32);
+	lo_hi_writeq(value, idma64c->regs + offset);
 }
 
 #define channel_readq(idma64c, reg)		\
@@ -217,7 +213,7 @@ static inline void idma64_writel(struct idma64 *idma64, int offset, u32 value)
 	idma64_writel(idma64, IDMA64_##reg, (value))
 
 /**
- * struct idma64_chip - representation of DesignWare DMA controller hardware
+ * struct idma64_chip - representation of iDMA 64-bit controller hardware
  * @dev:		struct device of the DMA controller
  * @irq:		irq line
  * @regs:		memory mapped I/O space
diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c
index 9d375bc..7058d58 100644
--- a/drivers/dma/imx-sdma.c
+++ b/drivers/dma/imx-sdma.c
@@ -1478,7 +1478,7 @@ static int __init sdma_event_remap(struct sdma_engine *sdma)
 	event_remap = of_find_property(np, propname, NULL);
 	num_map = event_remap ? (event_remap->length / sizeof(u32)) : 0;
 	if (!num_map) {
-		dev_warn(sdma->dev, "no event needs to be remapped\n");
+		dev_dbg(sdma->dev, "no event needs to be remapped\n");
 		goto out;
 	} else if (num_map % EVENT_REMAP_CELLS) {
 		dev_err(sdma->dev, "the property %s must modulo %d\n",
@@ -1826,8 +1826,6 @@ static int sdma_probe(struct platform_device *pdev)
 		of_node_put(spba_bus);
 	}
 
-	dev_info(sdma->dev, "initialized\n");
-
 	return 0;
 
 err_register:
@@ -1852,7 +1850,6 @@ static int sdma_remove(struct platform_device *pdev)
 	}
 
 	platform_set_drvdata(pdev, NULL);
-	dev_info(&pdev->dev, "Removed...\n");
 	return 0;
 }
 
diff --git a/drivers/dma/ioat/dma.c b/drivers/dma/ioat/dma.c
index f66b7e6..1d5df2e 100644
--- a/drivers/dma/ioat/dma.c
+++ b/drivers/dma/ioat/dma.c
@@ -197,7 +197,8 @@ static void __ioat_start_null_desc(struct ioatdma_chan *ioat_chan)
 void ioat_start_null_desc(struct ioatdma_chan *ioat_chan)
 {
 	spin_lock_bh(&ioat_chan->prep_lock);
-	__ioat_start_null_desc(ioat_chan);
+	if (!test_bit(IOAT_CHAN_DOWN, &ioat_chan->state))
+		__ioat_start_null_desc(ioat_chan);
 	spin_unlock_bh(&ioat_chan->prep_lock);
 }
 
diff --git a/drivers/dma/ioat/dma.h b/drivers/dma/ioat/dma.h
index 1bc08498..8f4e607 100644
--- a/drivers/dma/ioat/dma.h
+++ b/drivers/dma/ioat/dma.h
@@ -82,8 +82,9 @@ struct ioatdma_device {
 	struct dma_pool *sed_hw_pool[MAX_SED_POOLS];
 	struct dma_device dma_dev;
 	u8 version;
-	struct msix_entry msix_entries[4];
-	struct ioatdma_chan *idx[4];
+#define IOAT_MAX_CHANS 4
+	struct msix_entry msix_entries[IOAT_MAX_CHANS];
+	struct ioatdma_chan *idx[IOAT_MAX_CHANS];
 	struct dca_provider *dca;
 	enum ioat_irq_mode irq_mode;
 	u32 cap;
@@ -95,6 +96,7 @@ struct ioatdma_chan {
 	dma_addr_t last_completion;
 	spinlock_t cleanup_lock;
 	unsigned long state;
+	#define IOAT_CHAN_DOWN 0
 	#define IOAT_COMPLETION_ACK 1
 	#define IOAT_RESET_PENDING 2
 	#define IOAT_KOBJ_INIT_FAIL 3
diff --git a/drivers/dma/ioat/init.c b/drivers/dma/ioat/init.c
index 1c3c9b0..4ef0c5e 100644
--- a/drivers/dma/ioat/init.c
+++ b/drivers/dma/ioat/init.c
@@ -27,6 +27,7 @@
 #include <linux/workqueue.h>
 #include <linux/prefetch.h>
 #include <linux/dca.h>
+#include <linux/aer.h>
 #include "dma.h"
 #include "registers.h"
 #include "hw.h"
@@ -1186,13 +1187,116 @@ static int ioat3_dma_probe(struct ioatdma_device *ioat_dma, int dca)
 	return 0;
 }
 
+static void ioat_shutdown(struct pci_dev *pdev)
+{
+	struct ioatdma_device *ioat_dma = pci_get_drvdata(pdev);
+	struct ioatdma_chan *ioat_chan;
+	int i;
+
+	if (!ioat_dma)
+		return;
+
+	for (i = 0; i < IOAT_MAX_CHANS; i++) {
+		ioat_chan = ioat_dma->idx[i];
+		if (!ioat_chan)
+			continue;
+
+		spin_lock_bh(&ioat_chan->prep_lock);
+		set_bit(IOAT_CHAN_DOWN, &ioat_chan->state);
+		del_timer_sync(&ioat_chan->timer);
+		spin_unlock_bh(&ioat_chan->prep_lock);
+		/* this should quiesce then reset */
+		ioat_reset_hw(ioat_chan);
+	}
+
+	ioat_disable_interrupts(ioat_dma);
+}
+
+void ioat_resume(struct ioatdma_device *ioat_dma)
+{
+	struct ioatdma_chan *ioat_chan;
+	u32 chanerr;
+	int i;
+
+	for (i = 0; i < IOAT_MAX_CHANS; i++) {
+		ioat_chan = ioat_dma->idx[i];
+		if (!ioat_chan)
+			continue;
+
+		spin_lock_bh(&ioat_chan->prep_lock);
+		clear_bit(IOAT_CHAN_DOWN, &ioat_chan->state);
+		spin_unlock_bh(&ioat_chan->prep_lock);
+
+		chanerr = readl(ioat_chan->reg_base + IOAT_CHANERR_OFFSET);
+		writel(chanerr, ioat_chan->reg_base + IOAT_CHANERR_OFFSET);
+
+		/* no need to reset as shutdown already did that */
+	}
+}
+
 #define DRV_NAME "ioatdma"
 
+static pci_ers_result_t ioat_pcie_error_detected(struct pci_dev *pdev,
+						 enum pci_channel_state error)
+{
+	dev_dbg(&pdev->dev, "%s: PCIe AER error %d\n", DRV_NAME, error);
+
+	/* quiesce and block I/O */
+	ioat_shutdown(pdev);
+
+	return PCI_ERS_RESULT_NEED_RESET;
+}
+
+static pci_ers_result_t ioat_pcie_error_slot_reset(struct pci_dev *pdev)
+{
+	pci_ers_result_t result = PCI_ERS_RESULT_RECOVERED;
+	int err;
+
+	dev_dbg(&pdev->dev, "%s post reset handling\n", DRV_NAME);
+
+	if (pci_enable_device_mem(pdev) < 0) {
+		dev_err(&pdev->dev,
+			"Failed to enable PCIe device after reset.\n");
+		result = PCI_ERS_RESULT_DISCONNECT;
+	} else {
+		pci_set_master(pdev);
+		pci_restore_state(pdev);
+		pci_save_state(pdev);
+		pci_wake_from_d3(pdev, false);
+	}
+
+	err = pci_cleanup_aer_uncorrect_error_status(pdev);
+	if (err) {
+		dev_err(&pdev->dev,
+			"AER uncorrect error status clear failed: %#x\n", err);
+	}
+
+	return result;
+}
+
+static void ioat_pcie_error_resume(struct pci_dev *pdev)
+{
+	struct ioatdma_device *ioat_dma = pci_get_drvdata(pdev);
+
+	dev_dbg(&pdev->dev, "%s: AER handling resuming\n", DRV_NAME);
+
+	/* initialize and bring everything back */
+	ioat_resume(ioat_dma);
+}
+
+static const struct pci_error_handlers ioat_err_handler = {
+	.error_detected = ioat_pcie_error_detected,
+	.slot_reset = ioat_pcie_error_slot_reset,
+	.resume = ioat_pcie_error_resume,
+};
+
 static struct pci_driver ioat_pci_driver = {
 	.name		= DRV_NAME,
 	.id_table	= ioat_pci_tbl,
 	.probe		= ioat_pci_probe,
 	.remove		= ioat_remove,
+	.shutdown	= ioat_shutdown,
+	.err_handler	= &ioat_err_handler,
 };
 
 static struct ioatdma_device *
@@ -1245,13 +1349,17 @@ static int ioat_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	pci_set_drvdata(pdev, device);
 
 	device->version = readb(device->reg_base + IOAT_VER_OFFSET);
-	if (device->version >= IOAT_VER_3_0)
+	if (device->version >= IOAT_VER_3_0) {
 		err = ioat3_dma_probe(device, ioat_dca_enabled);
-	else
+
+		if (device->version >= IOAT_VER_3_3)
+			pci_enable_pcie_error_reporting(pdev);
+	} else
 		return -ENODEV;
 
 	if (err) {
 		dev_err(dev, "Intel(R) I/OAT DMA Engine init failed\n");
+		pci_disable_pcie_error_reporting(pdev);
 		return -ENODEV;
 	}
 
@@ -1271,6 +1379,8 @@ static void ioat_remove(struct pci_dev *pdev)
 		free_dca_provider(device->dca);
 		device->dca = NULL;
 	}
+
+	pci_disable_pcie_error_reporting(pdev);
 	ioat_dma_remove(device);
 }
 
diff --git a/drivers/dma/ioat/prep.c b/drivers/dma/ioat/prep.c
index ad4fb41..6bb4a13 100644
--- a/drivers/dma/ioat/prep.c
+++ b/drivers/dma/ioat/prep.c
@@ -121,6 +121,9 @@ ioat_dma_prep_memcpy_lock(struct dma_chan *c, dma_addr_t dma_dest,
 	size_t total_len = len;
 	int num_descs, idx, i;
 
+	if (test_bit(IOAT_CHAN_DOWN, &ioat_chan->state))
+		return NULL;
+
 	num_descs = ioat_xferlen_to_descs(ioat_chan, len);
 	if (likely(num_descs) &&
 	    ioat_check_space_lock(ioat_chan, num_descs) == 0)
@@ -254,6 +257,11 @@ struct dma_async_tx_descriptor *
 ioat_prep_xor(struct dma_chan *chan, dma_addr_t dest, dma_addr_t *src,
 	       unsigned int src_cnt, size_t len, unsigned long flags)
 {
+	struct ioatdma_chan *ioat_chan = to_ioat_chan(chan);
+
+	if (test_bit(IOAT_CHAN_DOWN, &ioat_chan->state))
+		return NULL;
+
 	return __ioat_prep_xor_lock(chan, NULL, dest, src, src_cnt, len, flags);
 }
 
@@ -262,6 +270,11 @@ ioat_prep_xor_val(struct dma_chan *chan, dma_addr_t *src,
 		    unsigned int src_cnt, size_t len,
 		    enum sum_check_flags *result, unsigned long flags)
 {
+	struct ioatdma_chan *ioat_chan = to_ioat_chan(chan);
+
+	if (test_bit(IOAT_CHAN_DOWN, &ioat_chan->state))
+		return NULL;
+
 	/* the cleanup routine only sets bits on validate failure, it
 	 * does not clear bits on validate success... so clear it here
 	 */
@@ -574,6 +587,11 @@ ioat_prep_pq(struct dma_chan *chan, dma_addr_t *dst, dma_addr_t *src,
 	      unsigned int src_cnt, const unsigned char *scf, size_t len,
 	      unsigned long flags)
 {
+	struct ioatdma_chan *ioat_chan = to_ioat_chan(chan);
+
+	if (test_bit(IOAT_CHAN_DOWN, &ioat_chan->state))
+		return NULL;
+
 	/* specify valid address for disabled result */
 	if (flags & DMA_PREP_PQ_DISABLE_P)
 		dst[0] = dst[1];
@@ -614,6 +632,11 @@ ioat_prep_pq_val(struct dma_chan *chan, dma_addr_t *pq, dma_addr_t *src,
 		  unsigned int src_cnt, const unsigned char *scf, size_t len,
 		  enum sum_check_flags *pqres, unsigned long flags)
 {
+	struct ioatdma_chan *ioat_chan = to_ioat_chan(chan);
+
+	if (test_bit(IOAT_CHAN_DOWN, &ioat_chan->state))
+		return NULL;
+
 	/* specify valid address for disabled result */
 	if (flags & DMA_PREP_PQ_DISABLE_P)
 		pq[0] = pq[1];
@@ -638,6 +661,10 @@ ioat_prep_pqxor(struct dma_chan *chan, dma_addr_t dst, dma_addr_t *src,
 {
 	unsigned char scf[MAX_SCF];
 	dma_addr_t pq[2];
+	struct ioatdma_chan *ioat_chan = to_ioat_chan(chan);
+
+	if (test_bit(IOAT_CHAN_DOWN, &ioat_chan->state))
+		return NULL;
 
 	if (src_cnt > MAX_SCF)
 		return NULL;
@@ -661,6 +688,10 @@ ioat_prep_pqxor_val(struct dma_chan *chan, dma_addr_t *src,
 {
 	unsigned char scf[MAX_SCF];
 	dma_addr_t pq[2];
+	struct ioatdma_chan *ioat_chan = to_ioat_chan(chan);
+
+	if (test_bit(IOAT_CHAN_DOWN, &ioat_chan->state))
+		return NULL;
 
 	if (src_cnt > MAX_SCF)
 		return NULL;
@@ -689,6 +720,9 @@ ioat_prep_interrupt_lock(struct dma_chan *c, unsigned long flags)
 	struct ioat_ring_ent *desc;
 	struct ioat_dma_descriptor *hw;
 
+	if (test_bit(IOAT_CHAN_DOWN, &ioat_chan->state))
+		return NULL;
+
 	if (ioat_check_space_lock(ioat_chan, 1) == 0)
 		desc = ioat_get_ring_ent(ioat_chan, ioat_chan->head);
 	else
diff --git a/drivers/dma/moxart-dma.c b/drivers/dma/moxart-dma.c
index b463410..631c443 100644
--- a/drivers/dma/moxart-dma.c
+++ b/drivers/dma/moxart-dma.c
@@ -652,6 +652,7 @@ static const struct of_device_id moxart_dma_match[] = {
 	{ .compatible = "moxa,moxart-dma" },
 	{ }
 };
+MODULE_DEVICE_TABLE(of, moxart_dma_match);
 
 static struct platform_driver moxart_driver = {
 	.probe	= moxart_probe,
diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c
index e6281e7..aae76fb 100644
--- a/drivers/dma/mpc512x_dma.c
+++ b/drivers/dma/mpc512x_dma.c
@@ -1073,6 +1073,7 @@ static const struct of_device_id mpc_dma_match[] = {
 	{ .compatible = "fsl,mpc8308-dma", },
 	{},
 };
+MODULE_DEVICE_TABLE(of, mpc_dma_match);
 
 static struct platform_driver mpc_dma_driver = {
 	.probe		= mpc_dma_probe,
diff --git a/drivers/dma/omap-dma.c b/drivers/dma/omap-dma.c
index 249445c..1dfc71c 100644
--- a/drivers/dma/omap-dma.c
+++ b/drivers/dma/omap-dma.c
@@ -935,8 +935,12 @@ static struct dma_async_tx_descriptor *omap_dma_prep_dma_cyclic(
 		else
 			d->ccr |= CCR_SYNC_ELEMENT;
 
-		if (dir == DMA_DEV_TO_MEM)
+		if (dir == DMA_DEV_TO_MEM) {
 			d->ccr |= CCR_TRIGGER_SRC;
+			d->csdp |= CSDP_DST_PACKED;
+		} else {
+			d->csdp |= CSDP_SRC_PACKED;
+		}
 
 		d->cicr |= CICR_MISALIGNED_ERR_IE | CICR_TRANS_ERR_IE;
 
diff --git a/drivers/dma/sirf-dma.c b/drivers/dma/sirf-dma.c
index 7d5598d..22ea241 100644
--- a/drivers/dma/sirf-dma.c
+++ b/drivers/dma/sirf-dma.c
@@ -1149,6 +1149,7 @@ static const struct of_device_id sirfsoc_dma_match[] = {
 	{ .compatible = "sirf,atlas7-dmac-v2", .data = &sirfsoc_dmadata_a7v2,},
 	{},
 };
+MODULE_DEVICE_TABLE(of, sirfsoc_dma_match);
 
 static struct platform_driver sirfsoc_dma_driver = {
 	.probe		= sirfsoc_dma_probe,
diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c
index 750d1b3..dd3e7ba 100644
--- a/drivers/dma/ste_dma40.c
+++ b/drivers/dma/ste_dma40.c
@@ -2907,7 +2907,7 @@ static int __init d40_dmaengine_init(struct d40_base *base,
 
 	if (err) {
 		d40_err(base->dev,
-			"Failed to regsiter memcpy only channels\n");
+			"Failed to register memcpy only channels\n");
 		goto failure2;
 	}
 
diff --git a/drivers/dma/sun6i-dma.c b/drivers/dma/sun6i-dma.c
index 73e0be6..2db12e4 100644
--- a/drivers/dma/sun6i-dma.c
+++ b/drivers/dma/sun6i-dma.c
@@ -908,6 +908,7 @@ static const struct of_device_id sun6i_dma_match[] = {
 	{ .compatible = "allwinner,sun8i-h3-dma", .data = &sun8i_h3_dma_cfg },
 	{ /* sentinel */ }
 };
+MODULE_DEVICE_TABLE(of, sun6i_dma_match);
 
 static int sun6i_dma_probe(struct platform_device *pdev)
 {
diff --git a/drivers/dma/ti-dma-crossbar.c b/drivers/dma/ti-dma-crossbar.c
index 5cce8c9..a415edb 100644
--- a/drivers/dma/ti-dma-crossbar.c
+++ b/drivers/dma/ti-dma-crossbar.c
@@ -17,13 +17,184 @@
 #include <linux/of_device.h>
 #include <linux/of_dma.h>
 
-#define TI_XBAR_OUTPUTS	127
-#define TI_XBAR_INPUTS	256
+#define TI_XBAR_DRA7		0
+#define TI_XBAR_AM335X		1
+
+static const struct of_device_id ti_dma_xbar_match[] = {
+	{
+		.compatible = "ti,dra7-dma-crossbar",
+		.data = (void *)TI_XBAR_DRA7,
+	},
+	{
+		.compatible = "ti,am335x-edma-crossbar",
+		.data = (void *)TI_XBAR_AM335X,
+	},
+	{},
+};
+
+/* Crossbar on AM335x/AM437x family */
+#define TI_AM335X_XBAR_LINES	64
+
+struct ti_am335x_xbar_data {
+	void __iomem *iomem;
+
+	struct dma_router dmarouter;
+
+	u32 xbar_events; /* maximum number of events to select in xbar */
+	u32 dma_requests; /* number of DMA requests on eDMA */
+};
+
+struct ti_am335x_xbar_map {
+	u16 dma_line;
+	u16 mux_val;
+};
+
+static inline void ti_am335x_xbar_write(void __iomem *iomem, int event, u16 val)
+{
+	writeb_relaxed(val & 0x1f, iomem + event);
+}
+
+static void ti_am335x_xbar_free(struct device *dev, void *route_data)
+{
+	struct ti_am335x_xbar_data *xbar = dev_get_drvdata(dev);
+	struct ti_am335x_xbar_map *map = route_data;
+
+	dev_dbg(dev, "Unmapping XBAR event %u on channel %u\n",
+		map->mux_val, map->dma_line);
+
+	ti_am335x_xbar_write(xbar->iomem, map->dma_line, 0);
+	kfree(map);
+}
+
+static void *ti_am335x_xbar_route_allocate(struct of_phandle_args *dma_spec,
+					   struct of_dma *ofdma)
+{
+	struct platform_device *pdev = of_find_device_by_node(ofdma->of_node);
+	struct ti_am335x_xbar_data *xbar = platform_get_drvdata(pdev);
+	struct ti_am335x_xbar_map *map;
+
+	if (dma_spec->args_count != 3)
+		return ERR_PTR(-EINVAL);
+
+	if (dma_spec->args[2] >= xbar->xbar_events) {
+		dev_err(&pdev->dev, "Invalid XBAR event number: %d\n",
+			dma_spec->args[2]);
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (dma_spec->args[0] >= xbar->dma_requests) {
+		dev_err(&pdev->dev, "Invalid DMA request line number: %d\n",
+			dma_spec->args[0]);
+		return ERR_PTR(-EINVAL);
+	}
+
+	/* The of_node_put() will be done in the core for the node */
+	dma_spec->np = of_parse_phandle(ofdma->of_node, "dma-masters", 0);
+	if (!dma_spec->np) {
+		dev_err(&pdev->dev, "Can't get DMA master\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	map = kzalloc(sizeof(*map), GFP_KERNEL);
+	if (!map) {
+		of_node_put(dma_spec->np);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	map->dma_line = (u16)dma_spec->args[0];
+	map->mux_val = (u16)dma_spec->args[2];
+
+	dma_spec->args[2] = 0;
+	dma_spec->args_count = 2;
+
+	dev_dbg(&pdev->dev, "Mapping XBAR event%u to DMA%u\n",
+		map->mux_val, map->dma_line);
+
+	ti_am335x_xbar_write(xbar->iomem, map->dma_line, map->mux_val);
+
+	return map;
+}
+
+static const struct of_device_id ti_am335x_master_match[] = {
+	{ .compatible = "ti,edma3-tpcc", },
+	{},
+};
+
+static int ti_am335x_xbar_probe(struct platform_device *pdev)
+{
+	struct device_node *node = pdev->dev.of_node;
+	const struct of_device_id *match;
+	struct device_node *dma_node;
+	struct ti_am335x_xbar_data *xbar;
+	struct resource *res;
+	void __iomem *iomem;
+	int i, ret;
+
+	if (!node)
+		return -ENODEV;
+
+	xbar = devm_kzalloc(&pdev->dev, sizeof(*xbar), GFP_KERNEL);
+	if (!xbar)
+		return -ENOMEM;
+
+	dma_node = of_parse_phandle(node, "dma-masters", 0);
+	if (!dma_node) {
+		dev_err(&pdev->dev, "Can't get DMA master node\n");
+		return -ENODEV;
+	}
+
+	match = of_match_node(ti_am335x_master_match, dma_node);
+	if (!match) {
+		dev_err(&pdev->dev, "DMA master is not supported\n");
+		return -EINVAL;
+	}
+
+	if (of_property_read_u32(dma_node, "dma-requests",
+				 &xbar->dma_requests)) {
+		dev_info(&pdev->dev,
+			 "Missing XBAR output information, using %u.\n",
+			 TI_AM335X_XBAR_LINES);
+		xbar->dma_requests = TI_AM335X_XBAR_LINES;
+	}
+	of_node_put(dma_node);
+
+	if (of_property_read_u32(node, "dma-requests", &xbar->xbar_events)) {
+		dev_info(&pdev->dev,
+			 "Missing XBAR input information, using %u.\n",
+			 TI_AM335X_XBAR_LINES);
+		xbar->xbar_events = TI_AM335X_XBAR_LINES;
+	}
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	iomem = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(iomem))
+		return PTR_ERR(iomem);
+
+	xbar->iomem = iomem;
+
+	xbar->dmarouter.dev = &pdev->dev;
+	xbar->dmarouter.route_free = ti_am335x_xbar_free;
+
+	platform_set_drvdata(pdev, xbar);
+
+	/* Reset the crossbar */
+	for (i = 0; i < xbar->dma_requests; i++)
+		ti_am335x_xbar_write(xbar->iomem, i, 0);
+
+	ret = of_dma_router_register(node, ti_am335x_xbar_route_allocate,
+				     &xbar->dmarouter);
+
+	return ret;
+}
+
+/* Crossbar on DRA7xx family */
+#define TI_DRA7_XBAR_OUTPUTS	127
+#define TI_DRA7_XBAR_INPUTS	256
 
 #define TI_XBAR_EDMA_OFFSET	0
 #define TI_XBAR_SDMA_OFFSET	1
 
-struct ti_dma_xbar_data {
+struct ti_dra7_xbar_data {
 	void __iomem *iomem;
 
 	struct dma_router dmarouter;
@@ -35,35 +206,35 @@ struct ti_dma_xbar_data {
 	u32 dma_offset;
 };
 
-struct ti_dma_xbar_map {
+struct ti_dra7_xbar_map {
 	u16 xbar_in;
 	int xbar_out;
 };
 
-static inline void ti_dma_xbar_write(void __iomem *iomem, int xbar, u16 val)
+static inline void ti_dra7_xbar_write(void __iomem *iomem, int xbar, u16 val)
 {
 	writew_relaxed(val, iomem + (xbar * 2));
 }
 
-static void ti_dma_xbar_free(struct device *dev, void *route_data)
+static void ti_dra7_xbar_free(struct device *dev, void *route_data)
 {
-	struct ti_dma_xbar_data *xbar = dev_get_drvdata(dev);
-	struct ti_dma_xbar_map *map = route_data;
+	struct ti_dra7_xbar_data *xbar = dev_get_drvdata(dev);
+	struct ti_dra7_xbar_map *map = route_data;
 
 	dev_dbg(dev, "Unmapping XBAR%u (was routed to %d)\n",
 		map->xbar_in, map->xbar_out);
 
-	ti_dma_xbar_write(xbar->iomem, map->xbar_out, xbar->safe_val);
+	ti_dra7_xbar_write(xbar->iomem, map->xbar_out, xbar->safe_val);
 	idr_remove(&xbar->map_idr, map->xbar_out);
 	kfree(map);
 }
 
-static void *ti_dma_xbar_route_allocate(struct of_phandle_args *dma_spec,
-					struct of_dma *ofdma)
+static void *ti_dra7_xbar_route_allocate(struct of_phandle_args *dma_spec,
+					 struct of_dma *ofdma)
 {
 	struct platform_device *pdev = of_find_device_by_node(ofdma->of_node);
-	struct ti_dma_xbar_data *xbar = platform_get_drvdata(pdev);
-	struct ti_dma_xbar_map *map;
+	struct ti_dra7_xbar_data *xbar = platform_get_drvdata(pdev);
+	struct ti_dra7_xbar_map *map;
 
 	if (dma_spec->args[0] >= xbar->xbar_requests) {
 		dev_err(&pdev->dev, "Invalid XBAR request number: %d\n",
@@ -93,12 +264,12 @@ static void *ti_dma_xbar_route_allocate(struct of_phandle_args *dma_spec,
 	dev_dbg(&pdev->dev, "Mapping XBAR%u to DMA%d\n",
 		map->xbar_in, map->xbar_out);
 
-	ti_dma_xbar_write(xbar->iomem, map->xbar_out, map->xbar_in);
+	ti_dra7_xbar_write(xbar->iomem, map->xbar_out, map->xbar_in);
 
 	return map;
 }
 
-static const struct of_device_id ti_dma_master_match[] = {
+static const struct of_device_id ti_dra7_master_match[] = {
 	{
 		.compatible = "ti,omap4430-sdma",
 		.data = (void *)TI_XBAR_SDMA_OFFSET,
@@ -110,12 +281,12 @@ static const struct of_device_id ti_dma_master_match[] = {
 	{},
 };
 
-static int ti_dma_xbar_probe(struct platform_device *pdev)
+static int ti_dra7_xbar_probe(struct platform_device *pdev)
 {
 	struct device_node *node = pdev->dev.of_node;
 	const struct of_device_id *match;
 	struct device_node *dma_node;
-	struct ti_dma_xbar_data *xbar;
+	struct ti_dra7_xbar_data *xbar;
 	struct resource *res;
 	u32 safe_val;
 	void __iomem *iomem;
@@ -136,7 +307,7 @@ static int ti_dma_xbar_probe(struct platform_device *pdev)
 		return -ENODEV;
 	}
 
-	match = of_match_node(ti_dma_master_match, dma_node);
+	match = of_match_node(ti_dra7_master_match, dma_node);
 	if (!match) {
 		dev_err(&pdev->dev, "DMA master is not supported\n");
 		return -EINVAL;
@@ -146,16 +317,16 @@ static int ti_dma_xbar_probe(struct platform_device *pdev)
 				 &xbar->dma_requests)) {
 		dev_info(&pdev->dev,
 			 "Missing XBAR output information, using %u.\n",
-			 TI_XBAR_OUTPUTS);
-		xbar->dma_requests = TI_XBAR_OUTPUTS;
+			 TI_DRA7_XBAR_OUTPUTS);
+		xbar->dma_requests = TI_DRA7_XBAR_OUTPUTS;
 	}
 	of_node_put(dma_node);
 
 	if (of_property_read_u32(node, "dma-requests", &xbar->xbar_requests)) {
 		dev_info(&pdev->dev,
 			 "Missing XBAR input information, using %u.\n",
-			 TI_XBAR_INPUTS);
-		xbar->xbar_requests = TI_XBAR_INPUTS;
+			 TI_DRA7_XBAR_INPUTS);
+		xbar->xbar_requests = TI_DRA7_XBAR_INPUTS;
 	}
 
 	if (!of_property_read_u32(node, "ti,dma-safe-map", &safe_val))
@@ -169,30 +340,50 @@ static int ti_dma_xbar_probe(struct platform_device *pdev)
 	xbar->iomem = iomem;
 
 	xbar->dmarouter.dev = &pdev->dev;
-	xbar->dmarouter.route_free = ti_dma_xbar_free;
+	xbar->dmarouter.route_free = ti_dra7_xbar_free;
 	xbar->dma_offset = (u32)match->data;
 
 	platform_set_drvdata(pdev, xbar);
 
 	/* Reset the crossbar */
 	for (i = 0; i < xbar->dma_requests; i++)
-		ti_dma_xbar_write(xbar->iomem, i, xbar->safe_val);
+		ti_dra7_xbar_write(xbar->iomem, i, xbar->safe_val);
 
-	ret = of_dma_router_register(node, ti_dma_xbar_route_allocate,
+	ret = of_dma_router_register(node, ti_dra7_xbar_route_allocate,
 				     &xbar->dmarouter);
 	if (ret) {
 		/* Restore the defaults for the crossbar */
 		for (i = 0; i < xbar->dma_requests; i++)
-			ti_dma_xbar_write(xbar->iomem, i, i);
+			ti_dra7_xbar_write(xbar->iomem, i, i);
 	}
 
 	return ret;
 }
 
-static const struct of_device_id ti_dma_xbar_match[] = {
-	{ .compatible = "ti,dra7-dma-crossbar" },
-	{},
-};
+static int ti_dma_xbar_probe(struct platform_device *pdev)
+{
+	const struct of_device_id *match;
+	int ret;
+
+	match = of_match_node(ti_dma_xbar_match, pdev->dev.of_node);
+	if (unlikely(!match))
+		return -EINVAL;
+
+	switch ((u32)match->data) {
+	case TI_XBAR_DRA7:
+		ret = ti_dra7_xbar_probe(pdev);
+		break;
+	case TI_XBAR_AM335X:
+		ret = ti_am335x_xbar_probe(pdev);
+		break;
+	default:
+		dev_err(&pdev->dev, "Unsupported crossbar\n");
+		ret = -ENODEV;
+		break;
+	}
+
+	return ret;
+}
 
 static struct platform_driver ti_dma_xbar_driver = {
 	.driver = {
diff --git a/drivers/dma/virt-dma.h b/drivers/dma/virt-dma.h
index 181b952..2fa4774 100644
--- a/drivers/dma/virt-dma.h
+++ b/drivers/dma/virt-dma.h
@@ -47,9 +47,9 @@ struct virt_dma_desc *vchan_find_desc(struct virt_dma_chan *, dma_cookie_t);
 
 /**
  * vchan_tx_prep - prepare a descriptor
- * vc: virtual channel allocating this descriptor
- * vd: virtual descriptor to prepare
- * tx_flags: flags argument passed in to prepare function
+ * @vc: virtual channel allocating this descriptor
+ * @vd: virtual descriptor to prepare
+ * @tx_flags: flags argument passed in to prepare function
  */
 static inline struct dma_async_tx_descriptor *vchan_tx_prep(struct virt_dma_chan *vc,
 	struct virt_dma_desc *vd, unsigned long tx_flags)
@@ -65,7 +65,7 @@ static inline struct dma_async_tx_descriptor *vchan_tx_prep(struct virt_dma_chan
 
 /**
  * vchan_issue_pending - move submitted descriptors to issued list
- * vc: virtual channel to update
+ * @vc: virtual channel to update
  *
  * vc.lock must be held by caller
  */
@@ -77,7 +77,7 @@ static inline bool vchan_issue_pending(struct virt_dma_chan *vc)
 
 /**
  * vchan_cookie_complete - report completion of a descriptor
- * vd: virtual descriptor to update
+ * @vd: virtual descriptor to update
  *
  * vc.lock must be held by caller
  */
@@ -97,7 +97,7 @@ static inline void vchan_cookie_complete(struct virt_dma_desc *vd)
 
 /**
  * vchan_cyclic_callback - report the completion of a period
- * vd: virtual descriptor
+ * @vd: virtual descriptor
  */
 static inline void vchan_cyclic_callback(struct virt_dma_desc *vd)
 {
@@ -109,7 +109,7 @@ static inline void vchan_cyclic_callback(struct virt_dma_desc *vd)
 
 /**
  * vchan_next_desc - peek at the next descriptor to be processed
- * vc: virtual channel to obtain descriptor from
+ * @vc: virtual channel to obtain descriptor from
  *
  * vc.lock must be held by caller
  */
@@ -123,8 +123,8 @@ static inline struct virt_dma_desc *vchan_next_desc(struct virt_dma_chan *vc)
 
 /**
  * vchan_get_all_descriptors - obtain all submitted and issued descriptors
- * vc: virtual channel to get descriptors from
- * head: list of descriptors found
+ * @vc: virtual channel to get descriptors from
+ * @head: list of descriptors found
  *
  * vc.lock must be held by caller
  *
diff --git a/drivers/dma/xgene-dma.c b/drivers/dma/xgene-dma.c
index 8d57b1b..9dfa2b0 100644
--- a/drivers/dma/xgene-dma.c
+++ b/drivers/dma/xgene-dma.c
@@ -547,14 +547,12 @@ static struct xgene_dma_desc_sw *xgene_dma_alloc_descriptor(
 	struct xgene_dma_desc_sw *desc;
 	dma_addr_t phys;
 
-	desc = dma_pool_alloc(chan->desc_pool, GFP_NOWAIT, &phys);
+	desc = dma_pool_zalloc(chan->desc_pool, GFP_NOWAIT, &phys);
 	if (!desc) {
 		chan_err(chan, "Failed to allocate LDs\n");
 		return NULL;
 	}
 
-	memset(desc, 0, sizeof(*desc));
-
 	INIT_LIST_HEAD(&desc->tx_list);
 	desc->tx.phys = phys;
 	desc->tx.tx_submit = xgene_dma_tx_submit;
@@ -894,60 +892,6 @@ static void xgene_dma_free_chan_resources(struct dma_chan *dchan)
 	chan->desc_pool = NULL;
 }
 
-static struct dma_async_tx_descriptor *xgene_dma_prep_memcpy(
-	struct dma_chan *dchan, dma_addr_t dst, dma_addr_t src,
-	size_t len, unsigned long flags)
-{
-	struct xgene_dma_desc_sw *first = NULL, *new;
-	struct xgene_dma_chan *chan;
-	size_t copy;
-
-	if (unlikely(!dchan || !len))
-		return NULL;
-
-	chan = to_dma_chan(dchan);
-
-	do {
-		/* Allocate the link descriptor from DMA pool */
-		new = xgene_dma_alloc_descriptor(chan);
-		if (!new)
-			goto fail;
-
-		/* Create the largest transaction possible */
-		copy = min_t(size_t, len, XGENE_DMA_MAX_64B_DESC_BYTE_CNT);
-
-		/* Prepare DMA descriptor */
-		xgene_dma_prep_cpy_desc(chan, new, dst, src, copy);
-
-		if (!first)
-			first = new;
-
-		new->tx.cookie = 0;
-		async_tx_ack(&new->tx);
-
-		/* Update metadata */
-		len -= copy;
-		dst += copy;
-		src += copy;
-
-		/* Insert the link descriptor to the LD ring */
-		list_add_tail(&new->node, &first->tx_list);
-	} while (len);
-
-	new->tx.flags = flags; /* client is in control of this ack */
-	new->tx.cookie = -EBUSY;
-	list_splice(&first->tx_list, &new->tx_list);
-
-	return &new->tx;
-
-fail:
-	if (!first)
-		return NULL;
-
-	xgene_dma_free_desc_list(chan, &first->tx_list);
-	return NULL;
-}
-
 static struct dma_async_tx_descriptor *xgene_dma_prep_sg(
 	struct dma_chan *dchan, struct scatterlist *dst_sg,
 	u32 dst_nents, struct scatterlist *src_sg,
@@ -1707,7 +1651,6 @@ static void xgene_dma_set_caps(struct xgene_dma_chan *chan,
 	dma_cap_zero(dma_dev->cap_mask);
 
 	/* Set DMA device capability */
-	dma_cap_set(DMA_MEMCPY, dma_dev->cap_mask);
 	dma_cap_set(DMA_SG, dma_dev->cap_mask);
 
 	/* Basically here, the X-Gene SoC DMA engine channel 0 supports XOR
@@ -1734,7 +1677,6 @@ static void xgene_dma_set_caps(struct xgene_dma_chan *chan,
 	dma_dev->device_free_chan_resources = xgene_dma_free_chan_resources;
 	dma_dev->device_issue_pending = xgene_dma_issue_pending;
 	dma_dev->device_tx_status = xgene_dma_tx_status;
-	dma_dev->device_prep_dma_memcpy = xgene_dma_prep_memcpy;
 	dma_dev->device_prep_dma_sg = xgene_dma_prep_sg;
 
 	if (dma_has_cap(DMA_XOR, dma_dev->cap_mask)) {
@@ -1787,8 +1729,7 @@ static int xgene_dma_async_register(struct xgene_dma *pdma, int id)
 
 	/* DMA capability info */
 	dev_info(pdma->dev,
-		 "%s: CAPABILITY ( %s%s%s%s)\n", dma_chan_name(&chan->dma_chan),
-		 dma_has_cap(DMA_MEMCPY, dma_dev->cap_mask) ? "MEMCPY " : "",
+		 "%s: CAPABILITY ( %s%s%s)\n", dma_chan_name(&chan->dma_chan),
 		 dma_has_cap(DMA_SG, dma_dev->cap_mask) ? "SGCPY " : "",
 		 dma_has_cap(DMA_XOR, dma_dev->cap_mask) ? "XOR " : "",
 		 dma_has_cap(DMA_PQ, dma_dev->cap_mask) ? "PQ " : "");
diff --git a/drivers/dma/xilinx/xilinx_vdma.c b/drivers/dma/xilinx/xilinx_vdma.c
index d8434d4..6f4b501 100644
--- a/drivers/dma/xilinx/xilinx_vdma.c
+++ b/drivers/dma/xilinx/xilinx_vdma.c
@@ -1349,6 +1349,7 @@ static const struct of_device_id xilinx_vdma_of_ids[] = {
 	{ .compatible = "xlnx,axi-vdma-1.00.a",},
 	{}
 };
+MODULE_DEVICE_TABLE(of, xilinx_vdma_of_ids);
 
 static struct platform_driver xilinx_vdma_driver = {
 	.driver = {
diff --git a/drivers/dma/zx296702_dma.c b/drivers/dma/zx296702_dma.c
index c017fcd..245d759 100644
--- a/drivers/dma/zx296702_dma.c
+++ b/drivers/dma/zx296702_dma.c
@@ -441,7 +441,7 @@ static struct zx_dma_desc_sw *zx_alloc_desc_resource(int num,
 		kfree(ds);
 		return NULL;
 	}
-	memset(ds->desc_hw, sizeof(struct zx_desc_hw) * num, 0);
+	memset(ds->desc_hw, 0, sizeof(struct zx_desc_hw) * num);
 	ds->desc_num = num;
 	return ds;
 }