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

Pull slave-dma updates from Vinod Koul:
 "Some notable changes are:
   - new driver for AMBA AXI NBPF by Guennadi
   - new driver for sun6i controller by Maxime
   - pl330 drivers fixes from Lar's
   - sh-dma updates and fixes from Laurent, Geert and Kuninori
   - Documentation updates from Geert
   - drivers fixes and updates spread over dw, edma, freescale, mpc512x
     etc.."

* 'for-linus' of git://git.infradead.org/users/vkoul/slave-dma: (72 commits)
  dmaengine: sun6i: depends on RESET_CONTROLLER
  dma: at_hdmac: fix invalid remaining bytes detection
  dmaengine: nbpfaxi: don't build this driver where it cannot be used
  dmaengine: nbpf_error_get_channel() can be static
  dma: pl08x: Use correct specifier for size_t values
  dmaengine: Remove the context argument to the prep_dma_cyclic operation
  dmaengine: nbpfaxi: convert to tasklet
  dmaengine: nbpfaxi: fix a theoretical race
  dmaengine: add a driver for AMBA AXI NBPF DMAC IP cores
  dmaengine: add device tree binding documentation for the nbpfaxi driver
  dmaengine: edma: Do not register second device when booted with DT
  dmaengine: edma: Do not change the error code returned from edma_alloc_slot
  dmaengine: rcar-dmac: Add device tree bindings documentation
  dmaengine: shdma: Allocate cyclic sg list dynamically
  dmaengine: shdma: Make channel filter ignore unrelated devices
  dmaengine: sh: Rework Kconfig and Makefile
  dmaengine: sun6i: Fix memory leaks
  dmaengine: sun6i: Free the interrupt before killing the tasklet
  dmaengine: sun6i: Remove switch statement from buswidth convertion routine
  dmaengine: of: kconfig: select DMA_ENGINE when DMA_OF is selected
  ...

38 files changed, 3526 insertions, 873 deletions

diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index 8f6afbf9ba54..9b1ea0ef59af 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -393,6 +393,22 @@ config XILINX_VDMA
 	  channels, Memory Mapped to Stream (MM2S) and Stream to
 	  Memory Mapped (S2MM) for the data transfers.
 
+config DMA_SUN6I
+	tristate "Allwinner A31 SoCs DMA support"
+	depends on MACH_SUN6I || COMPILE_TEST
+	depends on RESET_CONTROLLER
+	select DMA_ENGINE
+	select DMA_VIRTUAL_CHANNELS
+	help
+	  Support for the DMA engine for Allwinner A31 SoCs.
+
+config NBPFAXI_DMA
+	tristate "Renesas Type-AXI NBPF DMA support"
+	select DMA_ENGINE
+	depends on ARM || COMPILE_TEST
+	help
+	  Support for "Type-AXI" NBPF DMA IPs from Renesas
+
 config DMA_ENGINE
 	bool
 
@@ -406,6 +422,7 @@ config DMA_ACPI
 config DMA_OF
 	def_bool y
 	depends on OF
+	select DMA_ENGINE
 
 comment "DMA Clients"
 	depends on DMA_ENGINE
diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile
index bd9e7fa928bd..c6adb925f0b9 100644
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@@ -1,5 +1,5 @@
-ccflags-$(CONFIG_DMADEVICES_DEBUG)  := -DDEBUG
-ccflags-$(CONFIG_DMADEVICES_VDEBUG) += -DVERBOSE_DEBUG
+subdir-ccflags-$(CONFIG_DMADEVICES_DEBUG)  := -DDEBUG
+subdir-ccflags-$(CONFIG_DMADEVICES_VDEBUG) += -DVERBOSE_DEBUG
 
 obj-$(CONFIG_DMA_ENGINE) += dmaengine.o
 obj-$(CONFIG_DMA_VIRTUAL_CHANNELS) += virt-dma.o
@@ -48,3 +48,5 @@ obj-$(CONFIG_FSL_EDMA) += fsl-edma.o
 obj-$(CONFIG_QCOM_BAM_DMA) += qcom_bam_dma.o
 obj-y += xilinx/
 obj-$(CONFIG_INTEL_MIC_X100_DMA) += mic_x100_dma.o
+obj-$(CONFIG_NBPFAXI_DMA) += nbpfaxi.o
+obj-$(CONFIG_DMA_SUN6I) += sun6i-dma.o
diff --git a/drivers/dma/TODO b/drivers/dma/TODO
index 734ed0206cd5..b8045cd42ee1 100644
--- a/drivers/dma/TODO
+++ b/drivers/dma/TODO
@@ -7,7 +7,6 @@ TODO for slave dma
 	- imx-dma
 	- imx-sdma
 	- mxs-dma.c
-	- dw_dmac
 	- intel_mid_dma
 4. Check other subsystems for dma drivers and merge/move to dmaengine
 5. Remove dma_slave_config's dma direction.
diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c
index 8114731a1c62..e34024b000a4 100644
--- a/drivers/dma/amba-pl08x.c
+++ b/drivers/dma/amba-pl08x.c
@@ -1040,7 +1040,7 @@ static int pl08x_fill_llis_for_desc(struct pl08x_driver_data *pl08x,
 
 		if (early_bytes) {
 			dev_vdbg(&pl08x->adev->dev,
-				"%s byte width LLIs (remain 0x%08x)\n",
+				"%s byte width LLIs (remain 0x%08zx)\n",
 				__func__, bd.remainder);
 			prep_byte_width_lli(pl08x, &bd, &cctl, early_bytes,
 				num_llis++, &total_bytes);
@@ -1653,7 +1653,7 @@ static struct dma_async_tx_descriptor *pl08x_prep_slave_sg(
 static struct dma_async_tx_descriptor *pl08x_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
 		size_t period_len, enum dma_transfer_direction direction,
-		unsigned long flags, void *context)
+		unsigned long flags)
 {
 	struct pl08x_dma_chan *plchan = to_pl08x_chan(chan);
 	struct pl08x_driver_data *pl08x = plchan->host;
@@ -1662,7 +1662,7 @@ static struct dma_async_tx_descriptor *pl08x_prep_dma_cyclic(
 	dma_addr_t slave_addr;
 
 	dev_dbg(&pl08x->adev->dev,
-		"%s prepare cyclic transaction of %d/%d bytes %s %s\n",
+		"%s prepare cyclic transaction of %zd/%zd bytes %s %s\n",
 		__func__, period_len, buf_len,
 		direction == DMA_MEM_TO_DEV ? "to" : "from",
 		plchan->name);
diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
index c13a3bb0f594..ca9dd2613283 100644
--- a/drivers/dma/at_hdmac.c
+++ b/drivers/dma/at_hdmac.c
@@ -294,14 +294,16 @@ static int atc_get_bytes_left(struct dma_chan *chan)
 			ret = -EINVAL;
 			goto out;
 		}
-		atchan->remain_desc -= (desc_cur->lli.ctrla & ATC_BTSIZE_MAX)
-						<< (desc_first->tx_width);
-		if (atchan->remain_desc < 0) {
+
+		count = (desc_cur->lli.ctrla & ATC_BTSIZE_MAX)
+			<< desc_first->tx_width;
+		if (atchan->remain_desc < count) {
 			ret = -EINVAL;
 			goto out;
-		} else {
-			ret = atchan->remain_desc;
 		}
+
+		atchan->remain_desc -= count;
+		ret = atchan->remain_desc;
 	} else {
 		/*
 		 * Get residual bytes when current
@@ -893,12 +895,11 @@ atc_dma_cyclic_fill_desc(struct dma_chan *chan, struct at_desc *desc,
  * @period_len: number of bytes for each period
  * @direction: transfer direction, to or from device
  * @flags: tx descriptor status flags
- * @context: transfer context (ignored)
  */
 static struct dma_async_tx_descriptor *
 atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
 		size_t period_len, enum dma_transfer_direction direction,
-		unsigned long flags, void *context)
+		unsigned long flags)
 {
 	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
 	struct at_dma_slave	*atslave = chan->private;
diff --git a/drivers/dma/bcm2835-dma.c b/drivers/dma/bcm2835-dma.c
index a03602164e3e..68007974961a 100644
--- a/drivers/dma/bcm2835-dma.c
+++ b/drivers/dma/bcm2835-dma.c
@@ -335,7 +335,7 @@ static void bcm2835_dma_issue_pending(struct dma_chan *chan)
 static struct dma_async_tx_descriptor *bcm2835_dma_prep_dma_cyclic(
 	struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
 	size_t period_len, enum dma_transfer_direction direction,
-	unsigned long flags, void *context)
+	unsigned long flags)
 {
 	struct bcm2835_chan *c = to_bcm2835_dma_chan(chan);
 	enum dma_slave_buswidth dev_width;
diff --git a/drivers/dma/dma-jz4740.c b/drivers/dma/dma-jz4740.c
index 94c380f07538..6a9d89c93b1f 100644
--- a/drivers/dma/dma-jz4740.c
+++ b/drivers/dma/dma-jz4740.c
@@ -433,7 +433,7 @@ static struct dma_async_tx_descriptor *jz4740_dma_prep_slave_sg(
 static struct dma_async_tx_descriptor *jz4740_dma_prep_dma_cyclic(
 	struct dma_chan *c, dma_addr_t buf_addr, size_t buf_len,
 	size_t period_len, enum dma_transfer_direction direction,
-	unsigned long flags, void *context)
+	unsigned long flags)
 {
 	struct jz4740_dmaengine_chan *chan = to_jz4740_dma_chan(c);
 	struct jz4740_dma_desc *desc;
@@ -614,4 +614,4 @@ module_platform_driver(jz4740_dma_driver);
 
 MODULE_AUTHOR("Lars-Peter Clausen <lars@metafoo.de>");
 MODULE_DESCRIPTION("JZ4740 DMA driver");
-MODULE_LICENSE("GPLv2");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index a27ded53ab4f..1af731b83b3f 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -279,6 +279,19 @@ static void dwc_dostart(struct dw_dma_chan *dwc, struct dw_desc *first)
 	channel_set_bit(dw, CH_EN, dwc->mask);
 }
 
+static void dwc_dostart_first_queued(struct dw_dma_chan *dwc)
+{
+	struct dw_desc *desc;
+
+	if (list_empty(&dwc->queue))
+		return;
+
+	list_move(dwc->queue.next, &dwc->active_list);
+	desc = dwc_first_active(dwc);
+	dev_vdbg(chan2dev(&dwc->chan), "%s: started %u\n", __func__, desc->txd.cookie);
+	dwc_dostart(dwc, desc);
+}
+
 /*----------------------------------------------------------------------*/
 
 static void
@@ -335,10 +348,7 @@ static void dwc_complete_all(struct dw_dma *dw, struct dw_dma_chan *dwc)
 	 * the completed ones.
 	 */
 	list_splice_init(&dwc->active_list, &list);
-	if (!list_empty(&dwc->queue)) {
-		list_move(dwc->queue.next, &dwc->active_list);
-		dwc_dostart(dwc, dwc_first_active(dwc));
-	}
+	dwc_dostart_first_queued(dwc);
 
 	spin_unlock_irqrestore(&dwc->lock, flags);
 
@@ -467,10 +477,7 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 	/* Try to continue after resetting the channel... */
 	dwc_chan_disable(dw, dwc);
 
-	if (!list_empty(&dwc->queue)) {
-		list_move(dwc->queue.next, &dwc->active_list);
-		dwc_dostart(dwc, dwc_first_active(dwc));
-	}
+	dwc_dostart_first_queued(dwc);
 	spin_unlock_irqrestore(&dwc->lock, flags);
 }
 
@@ -677,17 +684,9 @@ static dma_cookie_t dwc_tx_submit(struct dma_async_tx_descriptor *tx)
 	 * possible, perhaps even appending to those already submitted
 	 * for DMA. But this is hard to do in a race-free manner.
 	 */
-	if (list_empty(&dwc->active_list)) {
-		dev_vdbg(chan2dev(tx->chan), "%s: started %u\n", __func__,
-				desc->txd.cookie);
-		list_add_tail(&desc->desc_node, &dwc->active_list);
-		dwc_dostart(dwc, dwc_first_active(dwc));
-	} else {
-		dev_vdbg(chan2dev(tx->chan), "%s: queued %u\n", __func__,
-				desc->txd.cookie);
 
-		list_add_tail(&desc->desc_node, &dwc->queue);
-	}
+	dev_vdbg(chan2dev(tx->chan), "%s: queued %u\n", __func__, desc->txd.cookie);
+	list_add_tail(&desc->desc_node, &dwc->queue);
 
 	spin_unlock_irqrestore(&dwc->lock, flags);
 
@@ -1092,9 +1091,12 @@ dwc_tx_status(struct dma_chan *chan,
 static void dwc_issue_pending(struct dma_chan *chan)
 {
 	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
+	unsigned long		flags;
 
-	if (!list_empty(&dwc->queue))
-		dwc_scan_descriptors(to_dw_dma(chan->device), dwc);
+	spin_lock_irqsave(&dwc->lock, flags);
+	if (list_empty(&dwc->active_list))
+		dwc_dostart_first_queued(dwc);
+	spin_unlock_irqrestore(&dwc->lock, flags);
 }
 
 static int dwc_alloc_chan_resources(struct dma_chan *chan)
diff --git a/drivers/dma/edma.c b/drivers/dma/edma.c
index b512caf46944..7b65633f495e 100644
--- a/drivers/dma/edma.c
+++ b/drivers/dma/edma.c
@@ -23,6 +23,7 @@
 #include <linux/platform_device.h>
 #include <linux/slab.h>
 #include <linux/spinlock.h>
+#include <linux/of.h>
 
 #include <linux/platform_data/edma.h>
 
@@ -256,8 +257,13 @@ static int edma_terminate_all(struct edma_chan *echan)
 	 * echan->edesc is NULL and exit.)
 	 */
 	if (echan->edesc) {
+		int cyclic = echan->edesc->cyclic;
 		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);
@@ -592,7 +598,7 @@ struct dma_async_tx_descriptor *edma_prep_dma_memcpy(
 static struct dma_async_tx_descriptor *edma_prep_dma_cyclic(
 	struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
 	size_t period_len, enum dma_transfer_direction direction,
-	unsigned long tx_flags, void *context)
+	unsigned long tx_flags)
 {
 	struct edma_chan *echan = to_edma_chan(chan);
 	struct device *dev = chan->device->dev;
@@ -718,12 +724,15 @@ static struct dma_async_tx_descriptor *edma_prep_dma_cyclic(
 		edesc->absync = ret;
 
 		/*
-		 * Enable interrupts for every period because callback
-		 * has to be called for every period.
+		 * Enable period interrupt only if it is requested
 		 */
-		edesc->pset[i].param.opt |= TCINTEN;
+		if (tx_flags & DMA_PREP_INTERRUPT)
+			edesc->pset[i].param.opt |= TCINTEN;
 	}
 
+	/* Place the cyclic channel to highest priority queue */
+	edma_assign_channel_eventq(echan->ch_num, EVENTQ_0);
+
 	return vchan_tx_prep(&echan->vchan, &edesc->vdesc, tx_flags);
 }
 
@@ -993,7 +1002,7 @@ static int edma_dma_device_slave_caps(struct dma_chan *dchan,
 	caps->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
 	caps->cmd_pause = true;
 	caps->cmd_terminate = true;
-	caps->residue_granularity = DMA_RESIDUE_GRANULARITY_DESCRIPTOR;
+	caps->residue_granularity = DMA_RESIDUE_GRANULARITY_BURST;
 
 	return 0;
 }
@@ -1040,7 +1049,7 @@ static int edma_probe(struct platform_device *pdev)
 	ecc->dummy_slot = edma_alloc_slot(ecc->ctlr, EDMA_SLOT_ANY);
 	if (ecc->dummy_slot < 0) {
 		dev_err(&pdev->dev, "Can't allocate PaRAM dummy slot\n");
-		return -EIO;
+		return ecc->dummy_slot;
 	}
 
 	dma_cap_zero(ecc->dma_slave.cap_mask);
@@ -1125,7 +1134,7 @@ static int edma_init(void)
 		}
 	}
 
-	if (EDMA_CTLRS == 2) {
+	if (!of_have_populated_dt() && EDMA_CTLRS == 2) {
 		pdev1 = platform_device_register_full(&edma_dev_info1);
 		if (IS_ERR(pdev1)) {
 			platform_driver_unregister(&edma_driver);
diff --git a/drivers/dma/ep93xx_dma.c b/drivers/dma/ep93xx_dma.c
index cb4bf682a708..7650470196c4 100644
--- a/drivers/dma/ep93xx_dma.c
+++ b/drivers/dma/ep93xx_dma.c
@@ -1092,7 +1092,6 @@ fail:
  * @period_len: length of a single period
  * @dir: direction of the operation
  * @flags: tx descriptor status flags
- * @context: operation context (ignored)
  *
  * Prepares a descriptor for cyclic DMA operation. This means that once the
  * descriptor is submitted, we will be submitting in a @period_len sized
@@ -1105,8 +1104,7 @@ fail:
 static struct dma_async_tx_descriptor *
 ep93xx_dma_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t dma_addr,
 			   size_t buf_len, size_t period_len,
-			   enum dma_transfer_direction dir, unsigned long flags,
-			   void *context)
+			   enum dma_transfer_direction dir, unsigned long flags)
 {
 	struct ep93xx_dma_chan *edmac = to_ep93xx_dma_chan(chan);
 	struct ep93xx_dma_desc *desc, *first;
diff --git a/drivers/dma/fsl-edma.c b/drivers/dma/fsl-edma.c
index b396a7fb53ab..3c5711d5fe97 100644
--- a/drivers/dma/fsl-edma.c
+++ b/drivers/dma/fsl-edma.c
@@ -248,11 +248,12 @@ static void fsl_edma_chan_mux(struct fsl_edma_chan *fsl_chan,
 			unsigned int slot, bool enable)
 {
 	u32 ch = fsl_chan->vchan.chan.chan_id;
-	void __iomem *muxaddr = fsl_chan->edma->muxbase[ch / DMAMUX_NR];
+	void __iomem *muxaddr;
 	unsigned chans_per_mux, ch_off;
 
 	chans_per_mux = fsl_chan->edma->n_chans / DMAMUX_NR;
 	ch_off = fsl_chan->vchan.chan.chan_id % chans_per_mux;
+	muxaddr = fsl_chan->edma->muxbase[ch / chans_per_mux];
 
 	if (enable)
 		edma_writeb(fsl_chan->edma,
@@ -516,7 +517,7 @@ err:
 static struct dma_async_tx_descriptor *fsl_edma_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len,
 		size_t period_len, enum dma_transfer_direction direction,
-		unsigned long flags, void *context)
+		unsigned long flags)
 {
 	struct fsl_edma_chan *fsl_chan = to_fsl_edma_chan(chan);
 	struct fsl_edma_desc *fsl_desc;
@@ -724,6 +725,7 @@ static struct dma_chan *fsl_edma_xlate(struct of_phandle_args *dma_spec,
 {
 	struct fsl_edma_engine *fsl_edma = ofdma->of_dma_data;
 	struct dma_chan *chan, *_chan;
+	unsigned long chans_per_mux = fsl_edma->n_chans / DMAMUX_NR;
 
 	if (dma_spec->args_count != 2)
 		return NULL;
@@ -732,7 +734,7 @@ static struct dma_chan *fsl_edma_xlate(struct of_phandle_args *dma_spec,
 	list_for_each_entry_safe(chan, _chan, &fsl_edma->dma_dev.channels, device_node) {
 		if (chan->client_count)
 			continue;
-		if ((chan->chan_id / DMAMUX_NR) == dma_spec->args[0]) {
+		if ((chan->chan_id / chans_per_mux) == dma_spec->args[0]) {
 			chan = dma_get_slave_channel(chan);
 			if (chan) {
 				chan->device->privatecnt++;
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index e0fec68aed25..d5d6885ab341 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -396,10 +396,17 @@ static dma_cookie_t fsl_dma_tx_submit(struct dma_async_tx_descriptor *tx)
 	struct fsldma_chan *chan = to_fsl_chan(tx->chan);
 	struct fsl_desc_sw *desc = tx_to_fsl_desc(tx);
 	struct fsl_desc_sw *child;
-	unsigned long flags;
 	dma_cookie_t cookie = -EINVAL;
 
-	spin_lock_irqsave(&chan->desc_lock, flags);
+	spin_lock_bh(&chan->desc_lock);
+
+#ifdef CONFIG_PM
+	if (unlikely(chan->pm_state != RUNNING)) {
+		chan_dbg(chan, "cannot submit due to suspend\n");
+		spin_unlock_bh(&chan->desc_lock);
+		return -1;
+	}
+#endif
 
 	/*
 	 * assign cookies to all of the software descriptors
@@ -412,7 +419,7 @@ static dma_cookie_t fsl_dma_tx_submit(struct dma_async_tx_descriptor *tx)
 	/* put this transaction onto the tail of the pending queue */
 	append_ld_queue(chan, desc);
 
-	spin_unlock_irqrestore(&chan->desc_lock, flags);
+	spin_unlock_bh(&chan->desc_lock);
 
 	return cookie;
 }
@@ -459,6 +466,88 @@ static struct fsl_desc_sw *fsl_dma_alloc_descriptor(struct fsldma_chan *chan)
 }
 
 /**
+ * fsldma_clean_completed_descriptor - free all descriptors which
+ * has been completed and acked
+ * @chan: Freescale DMA channel
+ *
+ * This function is used on all completed and acked descriptors.
+ * All descriptors should only be freed in this function.
+ */
+static void fsldma_clean_completed_descriptor(struct fsldma_chan *chan)
+{
+	struct fsl_desc_sw *desc, *_desc;
+
+	/* Run the callback for each descriptor, in order */
+	list_for_each_entry_safe(desc, _desc, &chan->ld_completed, node)
+		if (async_tx_test_ack(&desc->async_tx))
+			fsl_dma_free_descriptor(chan, desc);
+}
+
+/**
+ * fsldma_run_tx_complete_actions - cleanup a single link descriptor
+ * @chan: Freescale DMA channel
+ * @desc: descriptor to cleanup and free
+ * @cookie: Freescale DMA transaction identifier
+ *
+ * This function is used on a descriptor which has been executed by the DMA
+ * controller. It will run any callbacks, submit any dependencies.
+ */
+static dma_cookie_t fsldma_run_tx_complete_actions(struct fsldma_chan *chan,
+		struct fsl_desc_sw *desc, dma_cookie_t cookie)
+{
+	struct dma_async_tx_descriptor *txd = &desc->async_tx;
+	dma_cookie_t ret = cookie;
+
+	BUG_ON(txd->cookie < 0);
+
+	if (txd->cookie > 0) {
+		ret = txd->cookie;
+
+		/* Run the link descriptor callback function */
+		if (txd->callback) {
+			chan_dbg(chan, "LD %p callback\n", desc);
+			txd->callback(txd->callback_param);
+		}
+	}
+
+	/* Run any dependencies */
+	dma_run_dependencies(txd);
+
+	return ret;
+}
+
+/**
+ * fsldma_clean_running_descriptor - move the completed descriptor from
+ * ld_running to ld_completed
+ * @chan: Freescale DMA channel
+ * @desc: the descriptor which is completed
+ *
+ * Free the descriptor directly if acked by async_tx api, or move it to
+ * queue ld_completed.
+ */
+static void fsldma_clean_running_descriptor(struct fsldma_chan *chan,
+		struct fsl_desc_sw *desc)
+{
+	/* Remove from the list of transactions */
+	list_del(&desc->node);
+
+	/*
+	 * the client is allowed to attach dependent operations
+	 * until 'ack' is set
+	 */
+	if (!async_tx_test_ack(&desc->async_tx)) {
+		/*
+		 * Move this descriptor to the list of descriptors which is
+		 * completed, but still awaiting the 'ack' bit to be set.
+		 */
+		list_add_tail(&desc->node, &chan->ld_completed);
+		return;
+	}
+
+	dma_pool_free(chan->desc_pool, desc, desc->async_tx.phys);
+}
+
+/**
  * fsl_chan_xfer_ld_queue - transfer any pending transactions
  * @chan : Freescale DMA channel
  *
@@ -526,31 +615,58 @@ static void fsl_chan_xfer_ld_queue(struct fsldma_chan *chan)
 }
 
 /**
- * fsldma_cleanup_descriptor - cleanup and free a single link descriptor
+ * fsldma_cleanup_descriptors - cleanup link descriptors which are completed
+ * and move them to ld_completed to free until flag 'ack' is set
  * @chan: Freescale DMA channel
- * @desc: descriptor to cleanup and free
  *
- * This function is used on a descriptor which has been executed by the DMA
- * controller. It will run any callbacks, submit any dependencies, and then
- * free the descriptor.
+ * This function is used on descriptors which have been executed by the DMA
+ * controller. It will run any callbacks, submit any dependencies, then
+ * free these descriptors if flag 'ack' is set.
  */
-static void fsldma_cleanup_descriptor(struct fsldma_chan *chan,
-				      struct fsl_desc_sw *desc)
+static void fsldma_cleanup_descriptors(struct fsldma_chan *chan)
 {
-	struct dma_async_tx_descriptor *txd = &desc->async_tx;
+	struct fsl_desc_sw *desc, *_desc;
+	dma_cookie_t cookie = 0;
+	dma_addr_t curr_phys = get_cdar(chan);
+	int seen_current = 0;
+
+	fsldma_clean_completed_descriptor(chan);
+
+	/* Run the callback for each descriptor, in order */
+	list_for_each_entry_safe(desc, _desc, &chan->ld_running, node) {
+		/*
+		 * do not advance past the current descriptor loaded into the
+		 * hardware channel, subsequent descriptors are either in
+		 * process or have not been submitted
+		 */
+		if (seen_current)
+			break;
+
+		/*
+		 * stop the search if we reach the current descriptor and the
+		 * channel is busy
+		 */
+		if (desc->async_tx.phys == curr_phys) {
+			seen_current = 1;
+			if (!dma_is_idle(chan))
+				break;
+		}
+
+		cookie = fsldma_run_tx_complete_actions(chan, desc, cookie);
 
-	/* Run the link descriptor callback function */
-	if (txd->callback) {
-		chan_dbg(chan, "LD %p callback\n", desc);
-		txd->callback(txd->callback_param);
+		fsldma_clean_running_descriptor(chan, desc);
 	}
 
-	/* Run any dependencies */
-	dma_run_dependencies(txd);
+	/*
+	 * Start any pending transactions automatically
+	 *
+	 * In the ideal case, we keep the DMA controller busy while we go
+	 * ahead and free the descriptors below.
+	 */
+	fsl_chan_xfer_ld_queue(chan);
 
-	dma_descriptor_unmap(txd);
-	chan_dbg(chan, "LD %p free\n", desc);
-	dma_pool_free(chan->desc_pool, desc, txd->phys);
+	if (cookie > 0)
+		chan->common.completed_cookie = cookie;
 }
 
 /**
@@ -617,13 +733,14 @@ static void fsldma_free_desc_list_reverse(struct fsldma_chan *chan,
 static void fsl_dma_free_chan_resources(struct dma_chan *dchan)
 {
 	struct fsldma_chan *chan = to_fsl_chan(dchan);
-	unsigned long flags;
 
 	chan_dbg(chan, "free all channel resources\n");
-	spin_lock_irqsave(&chan->desc_lock, flags);
+	spin_lock_bh(&chan->desc_lock);
+	fsldma_cleanup_descriptors(chan);
 	fsldma_free_desc_list(chan, &chan->ld_pending);
 	fsldma_free_desc_list(chan, &chan->ld_running);
-	spin_unlock_irqrestore(&chan->desc_lock, flags);
+	fsldma_free_desc_list(chan, &chan->ld_completed);
+	spin_unlock_bh(&chan->desc_lock);
 
 	dma_pool_destroy(chan->desc_pool);
 	chan->desc_pool = NULL;
@@ -842,7 +959,6 @@ static int fsl_dma_device_control(struct dma_chan *dchan,
 {
 	struct dma_slave_config *config;
 	struct fsldma_chan *chan;
-	unsigned long flags;
 	int size;
 
 	if (!dchan)
@@ -852,7 +968,7 @@ static int fsl_dma_device_control(struct dma_chan *dchan,
 
 	switch (cmd) {
 	case DMA_TERMINATE_ALL:
-		spin_lock_irqsave(&chan->desc_lock, flags);
+		spin_lock_bh(&chan->desc_lock);
 
 		/* Halt the DMA engine */
 		dma_halt(chan);
@@ -860,9 +976,10 @@ static int fsl_dma_device_control(struct dma_chan *dchan,
 		/* Remove and free all of the descriptors in the LD queue */
 		fsldma_free_desc_list(chan, &chan->ld_pending);
 		fsldma_free_desc_list(chan, &chan->ld_running);
+		fsldma_free_desc_list(chan, &chan->ld_completed);
 		chan->idle = true;
 
-		spin_unlock_irqrestore(&chan->desc_lock, flags);
+		spin_unlock_bh(&chan->desc_lock);
 		return 0;
 
 	case DMA_SLAVE_CONFIG:
@@ -904,11 +1021,10 @@ static int fsl_dma_device_control(struct dma_chan *dchan,
 static void fsl_dma_memcpy_issue_pending(struct dma_chan *dchan)
 {
 	struct fsldma_chan *chan = to_fsl_chan(dchan);
-	unsigned long flags;
 
-	spin_lock_irqsave(&chan->desc_lock, flags);
+	spin_lock_bh(&chan->desc_lock);
 	fsl_chan_xfer_ld_queue(chan);
-	spin_unlock_irqrestore(&chan->desc_lock, flags);
+	spin_unlock_bh(&chan->desc_lock);
 }
 
 /**
@@ -919,6 +1035,17 @@ static enum dma_status fsl_tx_status(struct dma_chan *dchan,
 					dma_cookie_t cookie,
 					struct dma_tx_state *txstate)
 {
+	struct fsldma_chan *chan = to_fsl_chan(dchan);
+	enum dma_status ret;
+
+	ret = dma_cookie_status(dchan, cookie, txstate);
+	if (ret == DMA_COMPLETE)
+		return ret;
+
+	spin_lock_bh(&chan->desc_lock);
+	fsldma_cleanup_descriptors(chan);
+	spin_unlock_bh(&chan->desc_lock);
+
 	return dma_cookie_status(dchan, cookie, txstate);
 }
 
@@ -996,52 +1123,18 @@ static irqreturn_t fsldma_chan_irq(int irq, void *data)
 static void dma_do_tasklet(unsigned long data)
 {
 	struct fsldma_chan *chan = (struct fsldma_chan *)data;
-	struct fsl_desc_sw *desc, *_desc;
-	LIST_HEAD(ld_cleanup);
-	unsigned long flags;
 
 	chan_dbg(chan, "tasklet entry\n");
 
-	spin_lock_irqsave(&chan->desc_lock, flags);
-
-	/* update the cookie if we have some descriptors to cleanup */
-	if (!list_empty(&chan->ld_running)) {
-		dma_cookie_t cookie;
-
-		desc = to_fsl_desc(chan->ld_running.prev);
-		cookie = desc->async_tx.cookie;
-		dma_cookie_complete(&desc->async_tx);
-
-		chan_dbg(chan, "completed_cookie=%d\n", cookie);
-	}
-
-	/*
-	 * move the descriptors to a temporary list so we can drop the lock
-	 * during the entire cleanup operation
-	 */
-	list_splice_tail_init(&chan->ld_running, &ld_cleanup);
+	spin_lock_bh(&chan->desc_lock);
 
 	/* the hardware is now idle and ready for more */
 	chan->idle = true;
 
-	/*
-	 * Start any pending transactions automatically
-	 *
-	 * In the ideal case, we keep the DMA controller busy while we go
-	 * ahead and free the descriptors below.
-	 */
-	fsl_chan_xfer_ld_queue(chan);
-	spin_unlock_irqrestore(&chan->desc_lock, flags);
-
-	/* Run the callback for each descriptor, in order */
-	list_for_each_entry_safe(desc, _desc, &ld_cleanup, node) {
-
-		/* Remove from the list of transactions */
-		list_del(&desc->node);
+	/* Run all cleanup for descriptors which have been completed */
+	fsldma_cleanup_descriptors(chan);
 
-		/* Run all cleanup for this descriptor */
-		fsldma_cleanup_descriptor(chan, desc);
-	}
+	spin_unlock_bh(&chan->desc_lock);
 
 	chan_dbg(chan, "tasklet exit\n");
 }
@@ -1225,7 +1318,11 @@ static int fsl_dma_chan_probe(struct fsldma_device *fdev,
 	spin_lock_init(&chan->desc_lock);
 	INIT_LIST_HEAD(&chan->ld_pending);
 	INIT_LIST_HEAD(&chan->ld_running);
+	INIT_LIST_HEAD(&chan->ld_completed);
 	chan->idle = true;
+#ifdef CONFIG_PM
+	chan->pm_state = RUNNING;
+#endif
 
 	chan->common.device = &fdev->common;
 	dma_cookie_init(&chan->common);
@@ -1365,6 +1462,69 @@ static int fsldma_of_remove(struct platform_device *op)
 	return 0;
 }
 
+#ifdef CONFIG_PM
+static int fsldma_suspend_late(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct fsldma_device *fdev = platform_get_drvdata(pdev);
+	struct fsldma_chan *chan;
+	int i;
+
+	for (i = 0; i < FSL_DMA_MAX_CHANS_PER_DEVICE; i++) {
+		chan = fdev->chan[i];
+		if (!chan)
+			continue;
+
+		spin_lock_bh(&chan->desc_lock);
+		if (unlikely(!chan->idle))
+			goto out;
+		chan->regs_save.mr = get_mr(chan);
+		chan->pm_state = SUSPENDED;
+		spin_unlock_bh(&chan->desc_lock);
+	}
+	return 0;
+
+out:
+	for (; i >= 0; i--) {
+		chan = fdev->chan[i];
+		if (!chan)
+			continue;
+		chan->pm_state = RUNNING;
+		spin_unlock_bh(&chan->desc_lock);
+	}
+	return -EBUSY;
+}
+
+static int fsldma_resume_early(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct fsldma_device *fdev = platform_get_drvdata(pdev);
+	struct fsldma_chan *chan;
+	u32 mode;
+	int i;
+
+	for (i = 0; i < FSL_DMA_MAX_CHANS_PER_DEVICE; i++) {
+		chan = fdev->chan[i];
+		if (!chan)
+			continue;
+
+		spin_lock_bh(&chan->desc_lock);
+		mode = chan->regs_save.mr
+			& ~FSL_DMA_MR_CS & ~FSL_DMA_MR_CC & ~FSL_DMA_MR_CA;
+		set_mr(chan, mode);
+		chan->pm_state = RUNNING;
+		spin_unlock_bh(&chan->desc_lock);
+	}
+
+	return 0;
+}
+
+static const struct dev_pm_ops fsldma_pm_ops = {
+	.suspend_late	= fsldma_suspend_late,
+	.resume_early	= fsldma_resume_early,
+};
+#endif
+
 static const struct of_device_id fsldma_of_ids[] = {
 	{ .compatible = "fsl,elo3-dma", },
 	{ .compatible = "fsl,eloplus-dma", },
@@ -1377,6 +1537,9 @@ static struct platform_driver fsldma_of_driver = {
 		.name = "fsl-elo-dma",
 		.owner = THIS_MODULE,
 		.of_match_table = fsldma_of_ids,
+#ifdef CONFIG_PM
+		.pm = &fsldma_pm_ops,
+#endif
 	},
 	.probe = fsldma_of_probe,
 	.remove = fsldma_of_remove,
diff --git a/drivers/dma/fsldma.h b/drivers/dma/fsldma.h
index d56e83599825..239c20c84382 100644
--- a/drivers/dma/fsldma.h
+++ b/drivers/dma/fsldma.h
@@ -134,12 +134,36 @@ struct fsldma_device {
 #define FSL_DMA_CHAN_PAUSE_EXT	0x00001000
 #define FSL_DMA_CHAN_START_EXT	0x00002000
 
+#ifdef CONFIG_PM
+struct fsldma_chan_regs_save {
+	u32 mr;
+};
+
+enum fsldma_pm_state {
+	RUNNING = 0,
+	SUSPENDED,
+};
+#endif
+
 struct fsldma_chan {
 	char name[8];			/* Channel name */
 	struct fsldma_chan_regs __iomem *regs;
 	spinlock_t desc_lock;		/* Descriptor operation lock */
-	struct list_head ld_pending;	/* Link descriptors queue */
-	struct list_head ld_running;	/* Link descriptors queue */
+	/*
+	 * Descriptors which are queued to run, but have not yet been
+	 * submitted to the hardware for execution
+	 */
+	struct list_head ld_pending;
+	/*
+	 * Descriptors which are currently being executed by the hardware
+	 */
+	struct list_head ld_running;
+	/*
+	 * Descriptors which have finished execution by the hardware. These
+	 * descriptors have already had their cleanup actions run. They are
+	 * waiting for the ACK bit to be set by the async_tx API.
+	 */
+	struct list_head ld_completed;	/* Link descriptors queue */
 	struct dma_chan common;		/* DMA common channel */
 	struct dma_pool *desc_pool;	/* Descriptors pool */
 	struct device *dev;		/* Channel device */
@@ -148,6 +172,10 @@ struct fsldma_chan {
 	struct tasklet_struct tasklet;
 	u32 feature;
 	bool idle;			/* DMA controller is idle */
+#ifdef CONFIG_PM
+	struct fsldma_chan_regs_save regs_save;
+	enum fsldma_pm_state pm_state;
+#endif
 
 	void (*toggle_ext_pause)(struct fsldma_chan *fsl_chan, int enable);
 	void (*toggle_ext_start)(struct fsldma_chan *fsl_chan, int enable);
diff --git a/drivers/dma/imx-dma.c b/drivers/dma/imx-dma.c
index 286660a12cc6..9d2c9e7374dc 100644
--- a/drivers/dma/imx-dma.c
+++ b/drivers/dma/imx-dma.c
@@ -866,7 +866,7 @@ static struct dma_async_tx_descriptor *imxdma_prep_slave_sg(
 static struct dma_async_tx_descriptor *imxdma_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len,
 		size_t period_len, enum dma_transfer_direction direction,
-		unsigned long flags, void *context)
+		unsigned long flags)
 {
 	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
 	struct imxdma_engine *imxdma = imxdmac->imxdma;
diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c
index 14867e3ac8ff..f7626e37d0b8 100644
--- a/drivers/dma/imx-sdma.c
+++ b/drivers/dma/imx-sdma.c
@@ -271,6 +271,7 @@ struct sdma_channel {
 	unsigned int			chn_count;
 	unsigned int			chn_real_count;
 	struct tasklet_struct		tasklet;
+	struct imx_dma_data		data;
 };
 
 #define IMX_DMA_SG_LOOP		BIT(0)
@@ -749,6 +750,11 @@ static void sdma_get_pc(struct sdma_channel *sdmac,
 		emi_2_per = sdma->script_addrs->asrc_2_mcu_addr;
 		per_2_per = sdma->script_addrs->per_2_per_addr;
 		break;
+	case IMX_DMATYPE_ASRC_SP:
+		per_2_emi = sdma->script_addrs->shp_2_mcu_addr;
+		emi_2_per = sdma->script_addrs->mcu_2_shp_addr;
+		per_2_per = sdma->script_addrs->per_2_per_addr;
+		break;
 	case IMX_DMATYPE_MSHC:
 		per_2_emi = sdma->script_addrs->mshc_2_mcu_addr;
 		emi_2_per = sdma->script_addrs->mcu_2_mshc_addr;
@@ -911,14 +917,13 @@ static int sdma_request_channel(struct sdma_channel *sdmac)
 	int channel = sdmac->channel;
 	int ret = -EBUSY;
 
-	sdmac->bd = dma_alloc_coherent(NULL, PAGE_SIZE, &sdmac->bd_phys, GFP_KERNEL);
+	sdmac->bd = dma_zalloc_coherent(NULL, PAGE_SIZE, &sdmac->bd_phys,
+					GFP_KERNEL);
 	if (!sdmac->bd) {
 		ret = -ENOMEM;
 		goto out;
 	}
 
-	memset(sdmac->bd, 0, PAGE_SIZE);
-
 	sdma->channel_control[channel].base_bd_ptr = sdmac->bd_phys;
 	sdma->channel_control[channel].current_bd_ptr = sdmac->bd_phys;
 
@@ -1120,7 +1125,7 @@ err_out:
 static struct dma_async_tx_descriptor *sdma_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len,
 		size_t period_len, enum dma_transfer_direction direction,
-		unsigned long flags, void *context)
+		unsigned long flags)
 {
 	struct sdma_channel *sdmac = to_sdma_chan(chan);
 	struct sdma_engine *sdma = sdmac->sdma;
@@ -1414,12 +1419,14 @@ err_dma_alloc:
 
 static bool sdma_filter_fn(struct dma_chan *chan, void *fn_param)
 {
+	struct sdma_channel *sdmac = to_sdma_chan(chan);
 	struct imx_dma_data *data = fn_param;
 
 	if (!imx_dma_is_general_purpose(chan))
 		return false;
 
-	chan->private = data;
+	sdmac->data = *data;
+	chan->private = &sdmac->data;
 
 	return true;
 }
diff --git a/drivers/dma/ipu/ipu_idmac.c b/drivers/dma/ipu/ipu_idmac.c
index 128ca143486d..bbf62927bd72 100644
--- a/drivers/dma/ipu/ipu_idmac.c
+++ b/drivers/dma/ipu/ipu_idmac.c
@@ -1532,11 +1532,17 @@ static int idmac_alloc_chan_resources(struct dma_chan *chan)
 #ifdef DEBUG
 	if (chan->chan_id == IDMAC_IC_7) {
 		ic_sof = ipu_irq_map(69);
-		if (ic_sof > 0)
-			request_irq(ic_sof, ic_sof_irq, 0, "IC SOF", ichan);
+		if (ic_sof > 0) {
+			ret = request_irq(ic_sof, ic_sof_irq, 0, "IC SOF", ichan);
+			if (ret)
+				dev_err(&chan->dev->device, "request irq failed for IC SOF");
+		}
 		ic_eof = ipu_irq_map(70);
-		if (ic_eof > 0)
-			request_irq(ic_eof, ic_eof_irq, 0, "IC EOF", ichan);
+		if (ic_eof > 0) {
+			ret = request_irq(ic_eof, ic_eof_irq, 0, "IC EOF", ichan);
+			if (ret)
+				dev_err(&chan->dev->device, "request irq failed for IC EOF");
+		}
 	}
 #endif
 
diff --git a/drivers/dma/mmp_pdma.c b/drivers/dma/mmp_pdma.c
index a7b186d536b3..a1a4db5721b8 100644
--- a/drivers/dma/mmp_pdma.c
+++ b/drivers/dma/mmp_pdma.c
@@ -601,7 +601,7 @@ static struct dma_async_tx_descriptor *
 mmp_pdma_prep_dma_cyclic(struct dma_chan *dchan,
 			 dma_addr_t buf_addr, size_t len, size_t period_len,
 			 enum dma_transfer_direction direction,
-			 unsigned long flags, void *context)
+			 unsigned long flags)
 {
 	struct mmp_pdma_chan *chan;
 	struct mmp_pdma_desc_sw *first = NULL, *prev = NULL, *new;
diff --git a/drivers/dma/mmp_tdma.c b/drivers/dma/mmp_tdma.c
index 724f7f4c9720..6ad30e2c5038 100644
--- a/drivers/dma/mmp_tdma.c
+++ b/drivers/dma/mmp_tdma.c
@@ -389,7 +389,7 @@ struct mmp_tdma_desc *mmp_tdma_alloc_descriptor(struct mmp_tdma_chan *tdmac)
 static struct dma_async_tx_descriptor *mmp_tdma_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len,
 		size_t period_len, enum dma_transfer_direction direction,
-		unsigned long flags, void *context)
+		unsigned long flags)
 {
 	struct mmp_tdma_chan *tdmac = to_mmp_tdma_chan(chan);
 	struct mmp_tdma_desc *desc;
diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c
index 2ad43738ac8b..881db2bcb48b 100644
--- a/drivers/dma/mpc512x_dma.c
+++ b/drivers/dma/mpc512x_dma.c
@@ -53,6 +53,7 @@
 #include <linux/of_address.h>
 #include <linux/of_device.h>
 #include <linux/of_irq.h>
+#include <linux/of_dma.h>
 #include <linux/of_platform.h>
 
 #include <linux/random.h>
@@ -1036,7 +1037,15 @@ static int mpc_dma_probe(struct platform_device *op)
 	if (retval)
 		goto err_free2;
 
-	return retval;
+	/* Register with OF helpers for DMA lookups (nonfatal) */
+	if (dev->of_node) {
+		retval = of_dma_controller_register(dev->of_node,
+						of_dma_xlate_by_chan_id, mdma);
+		if (retval)
+			dev_warn(dev, "Could not register for OF lookup\n");
+	}
+
+	return 0;
 
 err_free2:
 	if (mdma->is_mpc8308)
@@ -1057,6 +1066,8 @@ static int mpc_dma_remove(struct platform_device *op)
 	struct device *dev = &op->dev;
 	struct mpc_dma *mdma = dev_get_drvdata(dev);
 
+	if (dev->of_node)
+		of_dma_controller_free(dev->of_node);
 	dma_async_device_unregister(&mdma->dma);
 	if (mdma->is_mpc8308) {
 		free_irq(mdma->irq2, mdma);
diff --git a/drivers/dma/mxs-dma.c b/drivers/dma/mxs-dma.c
index ead491346da7..5ea61201dbf0 100644
--- a/drivers/dma/mxs-dma.c
+++ b/drivers/dma/mxs-dma.c
@@ -413,16 +413,14 @@ static int mxs_dma_alloc_chan_resources(struct dma_chan *chan)
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
 	int ret;
 
-	mxs_chan->ccw = dma_alloc_coherent(mxs_dma->dma_device.dev,
-				CCW_BLOCK_SIZE, &mxs_chan->ccw_phys,
-				GFP_KERNEL);
+	mxs_chan->ccw = dma_zalloc_coherent(mxs_dma->dma_device.dev,
+					    CCW_BLOCK_SIZE,
+					    &mxs_chan->ccw_phys, GFP_KERNEL);
 	if (!mxs_chan->ccw) {
 		ret = -ENOMEM;
 		goto err_alloc;
 	}
 
-	memset(mxs_chan->ccw, 0, CCW_BLOCK_SIZE);
-
 	if (mxs_chan->chan_irq != NO_IRQ) {
 		ret = request_irq(mxs_chan->chan_irq, mxs_dma_int_handler,
 					0, "mxs-dma", mxs_dma);
@@ -591,7 +589,7 @@ err_out:
 static struct dma_async_tx_descriptor *mxs_dma_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len,
 		size_t period_len, enum dma_transfer_direction direction,
-		unsigned long flags, void *context)
+		unsigned long flags)
 {
 	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
diff --git a/drivers/dma/nbpfaxi.c b/drivers/dma/nbpfaxi.c
new file mode 100644
index 000000000000..5aeada56a442
--- /dev/null
+++ b/drivers/dma/nbpfaxi.c
@@ -0,0 +1,1517 @@
+/*
+ * Copyright (C) 2013-2014 Renesas Electronics Europe Ltd.
+ * Author: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/bitmap.h>
+#include <linux/bitops.h>
+#include <linux/clk.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/log2.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/of_dma.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+
+#include <dt-bindings/dma/nbpfaxi.h>
+
+#include "dmaengine.h"
+
+#define NBPF_REG_CHAN_OFFSET	0
+#define NBPF_REG_CHAN_SIZE	0x40
+
+/* Channel Current Transaction Byte register */
+#define NBPF_CHAN_CUR_TR_BYTE	0x20
+
+/* Channel Status register */
+#define NBPF_CHAN_STAT	0x24
+#define NBPF_CHAN_STAT_EN	1
+#define NBPF_CHAN_STAT_TACT	4
+#define NBPF_CHAN_STAT_ERR	0x10
+#define NBPF_CHAN_STAT_END	0x20
+#define NBPF_CHAN_STAT_TC	0x40
+#define NBPF_CHAN_STAT_DER	0x400
+
+/* Channel Control register */
+#define NBPF_CHAN_CTRL	0x28
+#define NBPF_CHAN_CTRL_SETEN	1
+#define NBPF_CHAN_CTRL_CLREN	2
+#define NBPF_CHAN_CTRL_STG	4
+#define NBPF_CHAN_CTRL_SWRST	8
+#define NBPF_CHAN_CTRL_CLRRQ	0x10
+#define NBPF_CHAN_CTRL_CLREND	0x20
+#define NBPF_CHAN_CTRL_CLRTC	0x40
+#define NBPF_CHAN_CTRL_SETSUS	0x100
+#define NBPF_CHAN_CTRL_CLRSUS	0x200
+
+/* Channel Configuration register */
+#define NBPF_CHAN_CFG	0x2c
+#define NBPF_CHAN_CFG_SEL	7		/* terminal SELect: 0..7 */
+#define NBPF_CHAN_CFG_REQD	8		/* REQuest Direction: DMAREQ is 0: input, 1: output */
+#define NBPF_CHAN_CFG_LOEN	0x10		/* LOw ENable: low DMA request line is: 0: inactive, 1: active */
+#define NBPF_CHAN_CFG_HIEN	0x20		/* HIgh ENable: high DMA request line is: 0: inactive, 1: active */
+#define NBPF_CHAN_CFG_LVL	0x40		/* LeVeL: DMA request line is sensed as 0: edge, 1: level */
+#define NBPF_CHAN_CFG_AM	0x700		/* ACK Mode: 0: Pulse mode, 1: Level mode, b'1x: Bus Cycle */
+#define NBPF_CHAN_CFG_SDS	0xf000		/* Source Data Size: 0: 8 bits,... , 7: 1024 bits */
+#define NBPF_CHAN_CFG_DDS	0xf0000		/* Destination Data Size: as above */
+#define NBPF_CHAN_CFG_SAD	0x100000	/* Source ADdress counting: 0: increment, 1: fixed */
+#define NBPF_CHAN_CFG_DAD	0x200000	/* Destination ADdress counting: 0: increment, 1: fixed */
+#define NBPF_CHAN_CFG_TM	0x400000	/* Transfer Mode: 0: single, 1: block TM */
+#define NBPF_CHAN_CFG_DEM	0x1000000	/* DMAEND interrupt Mask */
+#define NBPF_CHAN_CFG_TCM	0x2000000	/* DMATCO interrupt Mask */
+#define NBPF_CHAN_CFG_SBE	0x8000000	/* Sweep Buffer Enable */
+#define NBPF_CHAN_CFG_RSEL	0x10000000	/* RM: Register Set sELect */
+#define NBPF_CHAN_CFG_RSW	0x20000000	/* RM: Register Select sWitch */
+#define NBPF_CHAN_CFG_REN	0x40000000	/* RM: Register Set Enable */
+#define NBPF_CHAN_CFG_DMS	0x80000000	/* 0: register mode (RM), 1: link mode (LM) */
+
+#define NBPF_CHAN_NXLA	0x38
+#define NBPF_CHAN_CRLA	0x3c
+
+/* Link Header field */
+#define NBPF_HEADER_LV	1
+#define NBPF_HEADER_LE	2
+#define NBPF_HEADER_WBD	4
+#define NBPF_HEADER_DIM	8
+
+#define NBPF_CTRL	0x300
+#define NBPF_CTRL_PR	1		/* 0: fixed priority, 1: round robin */
+#define NBPF_CTRL_LVINT	2		/* DMAEND and DMAERR signalling: 0: pulse, 1: level */
+
+#define NBPF_DSTAT_ER	0x314
+#define NBPF_DSTAT_END	0x318
+
+#define NBPF_DMA_BUSWIDTHS \
+	(BIT(DMA_SLAVE_BUSWIDTH_UNDEFINED) | \
+	 BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) | \
+	 BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) | \
+	 BIT(DMA_SLAVE_BUSWIDTH_4_BYTES) | \
+	 BIT(DMA_SLAVE_BUSWIDTH_8_BYTES))
+
+struct nbpf_config {
+	int num_channels;
+	int buffer_size;
+};
+
+/*
+ * We've got 3 types of objects, used to describe DMA transfers:
+ * 1. high-level descriptor, containing a struct dma_async_tx_descriptor object
+ *	in it, used to communicate with the user
+ * 2. hardware DMA link descriptors, that we pass to DMAC for DMA transfer
+ *	queuing, these must be DMAable, using either the streaming DMA API or
+ *	allocated from coherent memory - one per SG segment
+ * 3. one per SG segment descriptors, used to manage HW link descriptors from
+ *	(2). They do not have to be DMAable. They can either be (a) allocated
+ *	together with link descriptors as mixed (DMA / CPU) objects, or (b)
+ *	separately. Even if allocated separately it would be best to link them
+ *	to link descriptors once during channel resource allocation and always
+ *	use them as a single object.
+ * Therefore for both cases (a) and (b) at run-time objects (2) and (3) shall be
+ * treated as a single SG segment descriptor.
+ */
+
+struct nbpf_link_reg {
+	u32	header;
+	u32	src_addr;
+	u32	dst_addr;
+	u32	transaction_size;
+	u32	config;
+	u32	interval;
+	u32	extension;
+	u32	next;
+} __packed;
+
+struct nbpf_device;
+struct nbpf_channel;
+struct nbpf_desc;
+
+struct nbpf_link_desc {
+	struct nbpf_link_reg *hwdesc;
+	dma_addr_t hwdesc_dma_addr;
+	struct nbpf_desc *desc;
+	struct list_head node;
+};
+
+/**
+ * struct nbpf_desc - DMA transfer descriptor
+ * @async_tx:	dmaengine object
+ * @user_wait:	waiting for a user ack
+ * @length:	total transfer length
+ * @sg:		list of hardware descriptors, represented by struct nbpf_link_desc
+ * @node:	member in channel descriptor lists
+ */
+struct nbpf_desc {
+	struct dma_async_tx_descriptor async_tx;
+	bool user_wait;
+	size_t length;
+	struct nbpf_channel *chan;
+	struct list_head sg;
+	struct list_head node;
+};
+
+/* Take a wild guess: allocate 4 segments per descriptor */
+#define NBPF_SEGMENTS_PER_DESC 4
+#define NBPF_DESCS_PER_PAGE ((PAGE_SIZE - sizeof(struct list_head)) /	\
+	(sizeof(struct nbpf_desc) +					\
+	 NBPF_SEGMENTS_PER_DESC *					\
+	 (sizeof(struct nbpf_link_desc) + sizeof(struct nbpf_link_reg))))
+#define NBPF_SEGMENTS_PER_PAGE (NBPF_SEGMENTS_PER_DESC * NBPF_DESCS_PER_PAGE)
+
+struct nbpf_desc_page {
+	struct list_head node;
+	struct nbpf_desc desc[NBPF_DESCS_PER_PAGE];
+	struct nbpf_link_desc ldesc[NBPF_SEGMENTS_PER_PAGE];
+	struct nbpf_link_reg hwdesc[NBPF_SEGMENTS_PER_PAGE];
+};
+
+/**
+ * struct nbpf_channel - one DMAC channel
+ * @dma_chan:	standard dmaengine channel object
+ * @base:	register address base
+ * @nbpf:	DMAC
+ * @name:	IRQ name
+ * @irq:	IRQ number
+ * @slave_addr:	address for slave DMA
+ * @slave_width:slave data size in bytes
+ * @slave_burst:maximum slave burst size in bytes
+ * @terminal:	DMA terminal, assigned to this channel
+ * @dmarq_cfg:	DMA request line configuration - high / low, edge / level for NBPF_CHAN_CFG
+ * @flags:	configuration flags from DT
+ * @lock:	protect descriptor lists
+ * @free_links:	list of free link descriptors
+ * @free:	list of free descriptors
+ * @queued:	list of queued descriptors
+ * @active:	list of descriptors, scheduled for processing
+ * @done:	list of completed descriptors, waiting post-processing
+ * @desc_page:	list of additionally allocated descriptor pages - if any
+ */
+struct nbpf_channel {
+	struct dma_chan dma_chan;
+	struct tasklet_struct tasklet;
+	void __iomem *base;
+	struct nbpf_device *nbpf;
+	char name[16];
+	int irq;
+	dma_addr_t slave_src_addr;
+	size_t slave_src_width;
+	size_t slave_src_burst;
+	dma_addr_t slave_dst_addr;
+	size_t slave_dst_width;
+	size_t slave_dst_burst;
+	unsigned int terminal;
+	u32 dmarq_cfg;
+	unsigned long flags;
+	spinlock_t lock;
+	struct list_head free_links;
+	struct list_head free;
+	struct list_head queued;
+	struct list_head active;
+	struct list_head done;
+	struct list_head desc_page;
+	struct nbpf_desc *running;
+	bool paused;
+};
+
+struct nbpf_device {
+	struct dma_device dma_dev;
+	void __iomem *base;
+	struct clk *clk;
+	const struct nbpf_config *config;
+	struct nbpf_channel chan[];
+};
+
+enum nbpf_model {
+	NBPF1B4,
+	NBPF1B8,
+	NBPF1B16,
+	NBPF4B4,
+	NBPF4B8,
+	NBPF4B16,
+	NBPF8B4,
+	NBPF8B8,
+	NBPF8B16,
+};
+
+static struct nbpf_config nbpf_cfg[] = {
+	[NBPF1B4] = {
+		.num_channels = 1,
+		.buffer_size = 4,
+	},
+	[NBPF1B8] = {
+		.num_channels = 1,
+		.buffer_size = 8,
+	},
+	[NBPF1B16] = {
+		.num_channels = 1,
+		.buffer_size = 16,
+	},
+	[NBPF4B4] = {
+		.num_channels = 4,
+		.buffer_size = 4,
+	},
+	[NBPF4B8] = {
+		.num_channels = 4,
+		.buffer_size = 8,
+	},
+	[NBPF4B16] = {
+		.num_channels = 4,
+		.buffer_size = 16,
+	},
+	[NBPF8B4] = {
+		.num_channels = 8,
+		.buffer_size = 4,
+	},
+	[NBPF8B8] = {
+		.num_channels = 8,
+		.buffer_size = 8,
+	},
+	[NBPF8B16] = {
+		.num_channels = 8,
+		.buffer_size = 16,
+	},
+};
+
+#define nbpf_to_chan(d) container_of(d, struct nbpf_channel, dma_chan)
+
+/*
+ * dmaengine drivers seem to have a lot in common and instead of sharing more
+ * code, they reimplement those common algorithms independently. In this driver
+ * we try to separate the hardware-specific part from the (largely) generic
+ * part. This improves code readability and makes it possible in the future to
+ * reuse the generic code in form of a helper library. That generic code should
+ * be suitable for various DMA controllers, using transfer descriptors in RAM
+ * and pushing one SG list at a time to the DMA controller.
+ */
+
+/*		Hardware-specific part		*/
+
+static inline u32 nbpf_chan_read(struct nbpf_channel *chan,
+				 unsigned int offset)
+{
+	u32 data = ioread32(chan->base + offset);
+	dev_dbg(chan->dma_chan.device->dev, "%s(0x%p + 0x%x) = 0x%x\n",
+		__func__, chan->base, offset, data);
+	return data;
+}
+
+static inline void nbpf_chan_write(struct nbpf_channel *chan,
+				   unsigned int offset, u32 data)
+{
+	iowrite32(data, chan->base + offset);
+	dev_dbg(chan->dma_chan.device->dev, "%s(0x%p + 0x%x) = 0x%x\n",
+		__func__, chan->base, offset, data);
+}
+
+static inline u32 nbpf_read(struct nbpf_device *nbpf,
+			    unsigned int offset)
+{
+	u32 data = ioread32(nbpf->base + offset);
+	dev_dbg(nbpf->dma_dev.dev, "%s(0x%p + 0x%x) = 0x%x\n",
+		__func__, nbpf->base, offset, data);
+	return data;
+}
+
+static inline void nbpf_write(struct nbpf_device *nbpf,
+			      unsigned int offset, u32 data)
+{
+	iowrite32(data, nbpf->base + offset);
+	dev_dbg(nbpf->dma_dev.dev, "%s(0x%p + 0x%x) = 0x%x\n",
+		__func__, nbpf->base, offset, data);
+}
+
+static void nbpf_chan_halt(struct nbpf_channel *chan)
+{
+	nbpf_chan_write(chan, NBPF_CHAN_CTRL, NBPF_CHAN_CTRL_CLREN);
+}
+
+static bool nbpf_status_get(struct nbpf_channel *chan)
+{
+	u32 status = nbpf_read(chan->nbpf, NBPF_DSTAT_END);
+
+	return status & BIT(chan - chan->nbpf->chan);
+}
+
+static void nbpf_status_ack(struct nbpf_channel *chan)
+{
+	nbpf_chan_write(chan, NBPF_CHAN_CTRL, NBPF_CHAN_CTRL_CLREND);
+}
+
+static u32 nbpf_error_get(struct nbpf_device *nbpf)
+{
+	return nbpf_read(nbpf, NBPF_DSTAT_ER);
+}
+
+static struct nbpf_channel *nbpf_error_get_channel(struct nbpf_device *nbpf, u32 error)
+{
+	return nbpf->chan + __ffs(error);
+}
+
+static void nbpf_error_clear(struct nbpf_channel *chan)
+{
+	u32 status;
+	int i;
+
+	/* Stop the channel, make sure DMA has been aborted */
+	nbpf_chan_halt(chan);
+
+	for (i = 1000; i; i--) {
+		status = nbpf_chan_read(chan, NBPF_CHAN_STAT);
+		if (!(status & NBPF_CHAN_STAT_TACT))
+			break;
+		cpu_relax();
+	}
+
+	if (!i)
+		dev_err(chan->dma_chan.device->dev,
+			"%s(): abort timeout, channel status 0x%x\n", __func__, status);
+
+	nbpf_chan_write(chan, NBPF_CHAN_CTRL, NBPF_CHAN_CTRL_SWRST);
+}
+
+static int nbpf_start(struct nbpf_desc *desc)
+{
+	struct nbpf_channel *chan = desc->chan;
+	struct nbpf_link_desc *ldesc = list_first_entry(&desc->sg, struct nbpf_link_desc, node);
+
+	nbpf_chan_write(chan, NBPF_CHAN_NXLA, (u32)ldesc->hwdesc_dma_addr);
+	nbpf_chan_write(chan, NBPF_CHAN_CTRL, NBPF_CHAN_CTRL_SETEN | NBPF_CHAN_CTRL_CLRSUS);
+	chan->paused = false;
+
+	/* Software trigger MEMCPY - only MEMCPY uses the block mode */
+	if (ldesc->hwdesc->config & NBPF_CHAN_CFG_TM)
+		nbpf_chan_write(chan, NBPF_CHAN_CTRL, NBPF_CHAN_CTRL_STG);
+
+	dev_dbg(chan->nbpf->dma_dev.dev, "%s(): next 0x%x, cur 0x%x\n", __func__,
+		nbpf_chan_read(chan, NBPF_CHAN_NXLA), nbpf_chan_read(chan, NBPF_CHAN_CRLA));
+
+	return 0;
+}
+
+static void nbpf_chan_prepare(struct nbpf_channel *chan)
+{
+	chan->dmarq_cfg = (chan->flags & NBPF_SLAVE_RQ_HIGH ? NBPF_CHAN_CFG_HIEN : 0) |
+		(chan->flags & NBPF_SLAVE_RQ_LOW ? NBPF_CHAN_CFG_LOEN : 0) |
+		(chan->flags & NBPF_SLAVE_RQ_LEVEL ?
+		 NBPF_CHAN_CFG_LVL | (NBPF_CHAN_CFG_AM & 0x200) : 0) |
+		chan->terminal;
+}
+
+static void nbpf_chan_prepare_default(struct nbpf_channel *chan)
+{
+	/* Don't output DMAACK */
+	chan->dmarq_cfg = NBPF_CHAN_CFG_AM & 0x400;
+	chan->terminal = 0;
+	chan->flags = 0;
+}
+
+static void nbpf_chan_configure(struct nbpf_channel *chan)
+{
+	/*
+	 * We assume, that only the link mode and DMA request line configuration
+	 * have to be set in the configuration register manually. Dynamic
+	 * per-transfer configuration will be loaded from transfer descriptors.
+	 */
+	nbpf_chan_write(chan, NBPF_CHAN_CFG, NBPF_CHAN_CFG_DMS | chan->dmarq_cfg);
+}
+
+static u32 nbpf_xfer_ds(struct nbpf_device *nbpf, size_t size)
+{
+	/* Maximum supported bursts depend on the buffer size */
+	return min_t(int, __ffs(size), ilog2(nbpf->config->buffer_size * 8));
+}
+
+static size_t nbpf_xfer_size(struct nbpf_device *nbpf,
+			     enum dma_slave_buswidth width, u32 burst)
+{
+	size_t size;
+
+	if (!burst)
+		burst = 1;
+
+	switch (width) {
+	case DMA_SLAVE_BUSWIDTH_8_BYTES:
+		size = 8 * burst;
+		break;
+
+	case DMA_SLAVE_BUSWIDTH_4_BYTES:
+		size = 4 * burst;
+		break;
+
+	case DMA_SLAVE_BUSWIDTH_2_BYTES:
+		size = 2 * burst;
+		break;
+
+	default:
+		pr_warn("%s(): invalid bus width %u\n", __func__, width);
+	case DMA_SLAVE_BUSWIDTH_1_BYTE:
+		size = burst;
+	}
+
+	return nbpf_xfer_ds(nbpf, size);
+}
+
+/*
+ * We need a way to recognise slaves, whose data is sent "raw" over the bus,
+ * i.e. it isn't known in advance how many bytes will be received. Therefore
+ * the slave driver has to provide a "large enough" buffer and either read the
+ * buffer, when it is full, or detect, that some data has arrived, then wait for
+ * a timeout, if no more data arrives - receive what's already there. We want to
+ * handle such slaves in a special way to allow an optimised mode for other
+ * users, for whom the amount of data is known in advance. So far there's no way
+ * to recognise such slaves. We use a data-width check to distinguish between
+ * the SD host and the PL011 UART.
+ */
+
+static int nbpf_prep_one(struct nbpf_link_desc *ldesc,
+			 enum dma_transfer_direction direction,
+			 dma_addr_t src, dma_addr_t dst, size_t size, bool last)
+{
+	struct nbpf_link_reg *hwdesc = ldesc->hwdesc;
+	struct nbpf_desc *desc = ldesc->desc;
+	struct nbpf_channel *chan = desc->chan;
+	struct device *dev = chan->dma_chan.device->dev;
+	size_t mem_xfer, slave_xfer;
+	bool can_burst;
+
+	hwdesc->header = NBPF_HEADER_WBD | NBPF_HEADER_LV |
+		(last ? NBPF_HEADER_LE : 0);
+
+	hwdesc->src_addr = src;
+	hwdesc->dst_addr = dst;
+	hwdesc->transaction_size = size;
+
+	/*
+	 * set config: SAD, DAD, DDS, SDS, etc.
+	 * Note on transfer sizes: the DMAC can perform unaligned DMA transfers,
+	 * but it is important to have transaction size a multiple of both
+	 * receiver and transmitter transfer sizes. It is also possible to use
+	 * different RAM and device transfer sizes, and it does work well with
+	 * some devices, e.g. with V08R07S01E SD host controllers, which can use
+	 * 128 byte transfers. But this doesn't work with other devices,
+	 * especially when the transaction size is unknown. This is the case,
+	 * e.g. with serial drivers like amba-pl011.c. For reception it sets up
+	 * the transaction size of 4K and if fewer bytes are received, it
+	 * pauses DMA and reads out data received via DMA as well as those left
+	 * in the Rx FIFO. For this to work with the RAM side using burst
+	 * transfers we enable the SBE bit and terminate the transfer in our
+	 * DMA_PAUSE handler.
+	 */
+	mem_xfer = nbpf_xfer_ds(chan->nbpf, size);
+
+	switch (direction) {
+	case DMA_DEV_TO_MEM:
+		can_burst = chan->slave_src_width >= 3;
+		slave_xfer = min(mem_xfer, can_burst ?
+				 chan->slave_src_burst : chan->slave_src_width);
+		/*
+		 * Is the slave narrower than 64 bits, i.e. isn't using the full
+		 * bus width and cannot use bursts?
+		 */
+		if (mem_xfer > chan->slave_src_burst && !can_burst)
+			mem_xfer = chan->slave_src_burst;
+		/* Device-to-RAM DMA is unreliable without REQD set */
+		hwdesc->config = NBPF_CHAN_CFG_SAD | (NBPF_CHAN_CFG_DDS & (mem_xfer << 16)) |
+			(NBPF_CHAN_CFG_SDS & (slave_xfer << 12)) | NBPF_CHAN_CFG_REQD |
+			NBPF_CHAN_CFG_SBE;
+		break;
+
+	case DMA_MEM_TO_DEV:
+		slave_xfer = min(mem_xfer, chan->slave_dst_width >= 3 ?
+				 chan->slave_dst_burst : chan->slave_dst_width);
+		hwdesc->config = NBPF_CHAN_CFG_DAD | (NBPF_CHAN_CFG_SDS & (mem_xfer << 12)) |
+			(NBPF_CHAN_CFG_DDS & (slave_xfer << 16)) | NBPF_CHAN_CFG_REQD;
+		break;
+
+	case DMA_MEM_TO_MEM:
+		hwdesc->config = NBPF_CHAN_CFG_TCM | NBPF_CHAN_CFG_TM |
+			(NBPF_CHAN_CFG_SDS & (mem_xfer << 12)) |
+			(NBPF_CHAN_CFG_DDS & (mem_xfer << 16));
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	hwdesc->config |= chan->dmarq_cfg | (last ? 0 : NBPF_CHAN_CFG_DEM) |
+		NBPF_CHAN_CFG_DMS;
+
+	dev_dbg(dev, "%s(): desc @ %pad: hdr 0x%x, cfg 0x%x, %zu @ %pad -> %pad\n",
+		__func__, &ldesc->hwdesc_dma_addr, hwdesc->header,
+		hwdesc->config, size, &src, &dst);
+
+	dma_sync_single_for_device(dev, ldesc->hwdesc_dma_addr, sizeof(*hwdesc),
+				   DMA_TO_DEVICE);
+
+	return 0;
+}
+
+static size_t nbpf_bytes_left(struct nbpf_channel *chan)
+{
+	return nbpf_chan_read(chan, NBPF_CHAN_CUR_TR_BYTE);
+}
+
+static void nbpf_configure(struct nbpf_device *nbpf)
+{
+	nbpf_write(nbpf, NBPF_CTRL, NBPF_CTRL_LVINT);
+}
+
+static void nbpf_pause(struct nbpf_channel *chan)
+{
+	nbpf_chan_write(chan, NBPF_CHAN_CTRL, NBPF_CHAN_CTRL_SETSUS);
+	/* See comment in nbpf_prep_one() */
+	nbpf_chan_write(chan, NBPF_CHAN_CTRL, NBPF_CHAN_CTRL_CLREN);
+}
+
+/*		Generic part			*/
+
+/* DMA ENGINE functions */
+static void nbpf_issue_pending(struct dma_chan *dchan)
+{
+	struct nbpf_channel *chan = nbpf_to_chan(dchan);
+	unsigned long flags;
+
+	dev_dbg(dchan->device->dev, "Entry %s()\n", __func__);
+
+	spin_lock_irqsave(&chan->lock, flags);
+	if (list_empty(&chan->queued))
+		goto unlock;
+
+	list_splice_tail_init(&chan->queued, &chan->active);
+
+	if (!chan->running) {
+		struct nbpf_desc *desc = list_first_entry(&chan->active,
+						struct nbpf_desc, node);
+		if (!nbpf_start(desc))
+			chan->running = desc;
+	}
+
+unlock:
+	spin_unlock_irqrestore(&chan->lock, flags);
+}
+
+static enum dma_status nbpf_tx_status(struct dma_chan *dchan,
+		dma_cookie_t cookie, struct dma_tx_state *state)
+{
+	struct nbpf_channel *chan = nbpf_to_chan(dchan);
+	enum dma_status status = dma_cookie_status(dchan, cookie, state);
+
+	if (state) {
+		dma_cookie_t running;
+		unsigned long flags;
+
+		spin_lock_irqsave(&chan->lock, flags);
+		running = chan->running ? chan->running->async_tx.cookie : -EINVAL;
+
+		if (cookie == running) {
+			state->residue = nbpf_bytes_left(chan);
+			dev_dbg(dchan->device->dev, "%s(): residue %u\n", __func__,
+				state->residue);
+		} else if (status == DMA_IN_PROGRESS) {
+			struct nbpf_desc *desc;
+			bool found = false;
+
+			list_for_each_entry(desc, &chan->active, node)
+				if (desc->async_tx.cookie == cookie) {
+					found = true;
+					break;
+				}
+
+			if (!found)
+				list_for_each_entry(desc, &chan->queued, node)
+					if (desc->async_tx.cookie == cookie) {
+						found = true;
+						break;
+
+					}
+
+			state->residue = found ? desc->length : 0;
+		}
+
+		spin_unlock_irqrestore(&chan->lock, flags);
+	}
+
+	if (chan->paused)
+		status = DMA_PAUSED;
+
+	return status;
+}
+
+static dma_cookie_t nbpf_tx_submit(struct dma_async_tx_descriptor *tx)
+{
+	struct nbpf_desc *desc = container_of(tx, struct nbpf_desc, async_tx);
+	struct nbpf_channel *chan = desc->chan;
+	unsigned long flags;
+	dma_cookie_t cookie;
+
+	spin_lock_irqsave(&chan->lock, flags);
+	cookie = dma_cookie_assign(tx);
+	list_add_tail(&desc->node, &chan->queued);
+	spin_unlock_irqrestore(&chan->lock, flags);
+
+	dev_dbg(chan->dma_chan.device->dev, "Entry %s(%d)\n", __func__, cookie);
+
+	return cookie;
+}
+
+static int nbpf_desc_page_alloc(struct nbpf_channel *chan)
+{
+	struct dma_chan *dchan = &chan->dma_chan;
+	struct nbpf_desc_page *dpage = (void *)get_zeroed_page(GFP_KERNEL | GFP_DMA);
+	struct nbpf_link_desc *ldesc;
+	struct nbpf_link_reg *hwdesc;
+	struct nbpf_desc *desc;
+	LIST_HEAD(head);
+	LIST_HEAD(lhead);
+	int i;
+	struct device *dev = dchan->device->dev;
+
+	if (!dpage)
+		return -ENOMEM;
+
+	dev_dbg(dev, "%s(): alloc %lu descriptors, %lu segments, total alloc %zu\n",
+		__func__, NBPF_DESCS_PER_PAGE, NBPF_SEGMENTS_PER_PAGE, sizeof(*dpage));
+
+	for (i = 0, ldesc = dpage->ldesc, hwdesc = dpage->hwdesc;
+	     i < ARRAY_SIZE(dpage->ldesc);
+	     i++, ldesc++, hwdesc++) {
+		ldesc->hwdesc = hwdesc;
+		list_add_tail(&ldesc->node, &lhead);
+		ldesc->hwdesc_dma_addr = dma_map_single(dchan->device->dev,
+					hwdesc, sizeof(*hwdesc), DMA_TO_DEVICE);
+
+		dev_dbg(dev, "%s(): mapped 0x%p to %pad\n", __func__,
+			hwdesc, &ldesc->hwdesc_dma_addr);
+	}
+
+	for (i = 0, desc = dpage->desc;
+	     i < ARRAY_SIZE(dpage->desc);
+	     i++, desc++) {
+		dma_async_tx_descriptor_init(&desc->async_tx, dchan);
+		desc->async_tx.tx_submit = nbpf_tx_submit;
+		desc->chan = chan;
+		INIT_LIST_HEAD(&desc->sg);
+		list_add_tail(&desc->node, &head);
+	}
+
+	/*
+	 * This function cannot be called from interrupt context, so, no need to
+	 * save flags
+	 */
+	spin_lock_irq(&chan->lock);
+	list_splice_tail(&lhead, &chan->free_links);
+	list_splice_tail(&head, &chan->free);
+	list_add(&dpage->node, &chan->desc_page);
+	spin_unlock_irq(&chan->lock);
+
+	return ARRAY_SIZE(dpage->desc);
+}
+
+static void nbpf_desc_put(struct nbpf_desc *desc)
+{
+	struct nbpf_channel *chan = desc->chan;
+	struct nbpf_link_desc *ldesc, *tmp;
+	unsigned long flags;
+
+	spin_lock_irqsave(&chan->lock, flags);
+	list_for_each_entry_safe(ldesc, tmp, &desc->sg, node)
+		list_move(&ldesc->node, &chan->free_links);
+
+	list_add(&desc->node, &chan->free);
+	spin_unlock_irqrestore(&chan->lock, flags);
+}
+
+static void nbpf_scan_acked(struct nbpf_channel *chan)
+{
+	struct nbpf_desc *desc, *tmp;
+	unsigned long flags;
+	LIST_HEAD(head);
+
+	spin_lock_irqsave(&chan->lock, flags);
+	list_for_each_entry_safe(desc, tmp, &chan->done, node)
+		if (async_tx_test_ack(&desc->async_tx) && desc->user_wait) {
+			list_move(&desc->node, &head);
+			desc->user_wait = false;
+		}
+	spin_unlock_irqrestore(&chan->lock, flags);
+
+	list_for_each_entry_safe(desc, tmp, &head, node) {
+		list_del(&desc->node);
+		nbpf_desc_put(desc);
+	}
+}
+
+/*
+ * We have to allocate descriptors with the channel lock dropped. This means,
+ * before we re-acquire the lock buffers can be taken already, so we have to
+ * re-check after re-acquiring the lock and possibly retry, if buffers are gone
+ * again.
+ */
+static struct nbpf_desc *nbpf_desc_get(struct nbpf_channel *chan, size_t len)
+{
+	struct nbpf_desc *desc = NULL;
+	struct nbpf_link_desc *ldesc, *prev = NULL;
+
+	nbpf_scan_acked(chan);
+
+	spin_lock_irq(&chan->lock);
+
+	do {
+		int i = 0, ret;
+
+		if (list_empty(&chan->free)) {
+			/* No more free descriptors */
+			spin_unlock_irq(&chan->lock);
+			ret = nbpf_desc_page_alloc(chan);
+			if (ret < 0)
+				return NULL;
+			spin_lock_irq(&chan->lock);
+			continue;
+		}
+		desc = list_first_entry(&chan->free, struct nbpf_desc, node);
+		list_del(&desc->node);
+
+		do {
+			if (list_empty(&chan->free_links)) {
+				/* No more free link descriptors */
+				spin_unlock_irq(&chan->lock);
+				ret = nbpf_desc_page_alloc(chan);
+				if (ret < 0) {
+					nbpf_desc_put(desc);
+					return NULL;
+				}
+				spin_lock_irq(&chan->lock);
+				continue;
+			}
+
+			ldesc = list_first_entry(&chan->free_links,
+						 struct nbpf_link_desc, node);
+			ldesc->desc = desc;
+			if (prev)
+				prev->hwdesc->next = (u32)ldesc->hwdesc_dma_addr;
+
+			prev = ldesc;
+			list_move_tail(&ldesc->node, &desc->sg);
+
+			i++;
+		} while (i < len);
+	} while (!desc);
+
+	prev->hwdesc->next = 0;
+
+	spin_unlock_irq(&chan->lock);
+
+	return desc;
+}
+
+static void nbpf_chan_idle(struct nbpf_channel *chan)
+{
+	struct nbpf_desc *desc, *tmp;
+	unsigned long flags;
+	LIST_HEAD(head);
+
+	spin_lock_irqsave(&chan->lock, flags);
+
+	list_splice_init(&chan->done, &head);
+	list_splice_init(&chan->active, &head);
+	list_splice_init(&chan->queued, &head);
+
+	chan->running = NULL;
+
+	spin_unlock_irqrestore(&chan->lock, flags);
+
+	list_for_each_entry_safe(desc, tmp, &head, node) {
+		dev_dbg(chan->nbpf->dma_dev.dev, "%s(): force-free desc %p cookie %d\n",
+			__func__, desc, desc->async_tx.cookie);
+		list_del(&desc->node);
+		nbpf_desc_put(desc);
+	}
+}
+
+static int nbpf_control(struct dma_chan *dchan, enum dma_ctrl_cmd cmd,
+			unsigned long arg)
+{
+	struct nbpf_channel *chan = nbpf_to_chan(dchan);
+	struct dma_slave_config *config;
+
+	dev_dbg(dchan->device->dev, "Entry %s(%d)\n", __func__, cmd);
+
+	switch (cmd) {
+	case DMA_TERMINATE_ALL:
+		dev_dbg(dchan->device->dev, "Terminating\n");
+		nbpf_chan_halt(chan);
+		nbpf_chan_idle(chan);
+		break;
+
+	case DMA_SLAVE_CONFIG:
+		if (!arg)
+			return -EINVAL;
+		config = (struct dma_slave_config *)arg;
+
+		/*
+		 * We could check config->slave_id to match chan->terminal here,
+		 * but with DT they would be coming from the same source, so
+		 * such a check would be superflous
+		 */
+
+		chan->slave_dst_addr = config->dst_addr;
+		chan->slave_dst_width = nbpf_xfer_size(chan->nbpf,
+						       config->dst_addr_width, 1);
+		chan->slave_dst_burst = nbpf_xfer_size(chan->nbpf,
+						       config->dst_addr_width,
+						       config->dst_maxburst);
+		chan->slave_src_addr = config->src_addr;
+		chan->slave_src_width = nbpf_xfer_size(chan->nbpf,
+						       config->src_addr_width, 1);
+		chan->slave_src_burst = nbpf_xfer_size(chan->nbpf,
+						       config->src_addr_width,
+						       config->src_maxburst);
+		break;
+
+	case DMA_PAUSE:
+		chan->paused = true;
+		nbpf_pause(chan);
+		break;
+
+	default:
+		return -ENXIO;
+	}
+
+	return 0;
+}
+
+static struct dma_async_tx_descriptor *nbpf_prep_sg(struct nbpf_channel *chan,
+		struct scatterlist *src_sg, struct scatterlist *dst_sg,
+		size_t len, enum dma_transfer_direction direction,
+		unsigned long flags)
+{
+	struct nbpf_link_desc *ldesc;
+	struct scatterlist *mem_sg;
+	struct nbpf_desc *desc;
+	bool inc_src, inc_dst;
+	size_t data_len = 0;
+	int i = 0;
+
+	switch (direction) {
+	case DMA_DEV_TO_MEM:
+		mem_sg = dst_sg;
+		inc_src = false;
+		inc_dst = true;
+		break;
+
+	case DMA_MEM_TO_DEV:
+		mem_sg = src_sg;
+		inc_src = true;
+		inc_dst = false;
+		break;
+
+	default:
+	case DMA_MEM_TO_MEM:
+		mem_sg = src_sg;
+		inc_src = true;
+		inc_dst = true;
+	}
+
+	desc = nbpf_desc_get(chan, len);
+	if (!desc)
+		return NULL;
+
+	desc->async_tx.flags = flags;
+	desc->async_tx.cookie = -EBUSY;
+	desc->user_wait = false;
+
+	/*
+	 * This is a private descriptor list, and we own the descriptor. No need
+	 * to lock.
+	 */
+	list_for_each_entry(ldesc, &desc->sg, node) {
+		int ret = nbpf_prep_one(ldesc, direction,
+					sg_dma_address(src_sg),
+					sg_dma_address(dst_sg),
+					sg_dma_len(mem_sg),
+					i == len - 1);
+		if (ret < 0) {
+			nbpf_desc_put(desc);
+			return NULL;
+		}
+		data_len += sg_dma_len(mem_sg);
+		if (inc_src)
+			src_sg = sg_next(src_sg);
+		if (inc_dst)
+			dst_sg = sg_next(dst_sg);
+		mem_sg = direction == DMA_DEV_TO_MEM ? dst_sg : src_sg;
+		i++;
+	}
+
+	desc->length = data_len;
+
+	/* The user has to return the descriptor to us ASAP via .tx_submit() */
+	return &desc->async_tx;
+}
+
+static struct dma_async_tx_descriptor *nbpf_prep_memcpy(
+	struct dma_chan *dchan, dma_addr_t dst, dma_addr_t src,
+	size_t len, unsigned long flags)
+{
+	struct nbpf_channel *chan = nbpf_to_chan(dchan);
+	struct scatterlist dst_sg;
+	struct scatterlist src_sg;
+
+	sg_init_table(&dst_sg, 1);
+	sg_init_table(&src_sg, 1);
+
+	sg_dma_address(&dst_sg) = dst;
+	sg_dma_address(&src_sg) = src;
+
+	sg_dma_len(&dst_sg) = len;
+	sg_dma_len(&src_sg) = len;
+
+	dev_dbg(dchan->device->dev, "%s(): %zu @ %pad -> %pad\n",
+		__func__, len, &src, &dst);
+
+	return nbpf_prep_sg(chan, &src_sg, &dst_sg, 1,
+			    DMA_MEM_TO_MEM, flags);
+}
+
+static struct dma_async_tx_descriptor *nbpf_prep_memcpy_sg(
+	struct dma_chan *dchan,
+	struct scatterlist *dst_sg, unsigned int dst_nents,
+	struct scatterlist *src_sg, unsigned int src_nents,
+	unsigned long flags)
+{
+	struct nbpf_channel *chan = nbpf_to_chan(dchan);
+
+	if (dst_nents != src_nents)
+		return NULL;
+
+	return nbpf_prep_sg(chan, src_sg, dst_sg, src_nents,
+			    DMA_MEM_TO_MEM, flags);
+}
+
+static struct dma_async_tx_descriptor *nbpf_prep_slave_sg(
+	struct dma_chan *dchan, struct scatterlist *sgl, unsigned int sg_len,
+	enum dma_transfer_direction direction, unsigned long flags, void *context)
+{
+	struct nbpf_channel *chan = nbpf_to_chan(dchan);
+	struct scatterlist slave_sg;
+
+	dev_dbg(dchan->device->dev, "Entry %s()\n", __func__);
+
+	sg_init_table(&slave_sg, 1);
+
+	switch (direction) {
+	case DMA_MEM_TO_DEV:
+		sg_dma_address(&slave_sg) = chan->slave_dst_addr;
+		return nbpf_prep_sg(chan, sgl, &slave_sg, sg_len,
+				    direction, flags);
+
+	case DMA_DEV_TO_MEM:
+		sg_dma_address(&slave_sg) = chan->slave_src_addr;
+		return nbpf_prep_sg(chan, &slave_sg, sgl, sg_len,
+				    direction, flags);
+
+	default:
+		return NULL;
+	}
+}
+
+static int nbpf_alloc_chan_resources(struct dma_chan *dchan)
+{
+	struct nbpf_channel *chan = nbpf_to_chan(dchan);
+	int ret;
+
+	INIT_LIST_HEAD(&chan->free);
+	INIT_LIST_HEAD(&chan->free_links);
+	INIT_LIST_HEAD(&chan->queued);
+	INIT_LIST_HEAD(&chan->active);
+	INIT_LIST_HEAD(&chan->done);
+
+	ret = nbpf_desc_page_alloc(chan);
+	if (ret < 0)
+		return ret;
+
+	dev_dbg(dchan->device->dev, "Entry %s(): terminal %u\n", __func__,
+		chan->terminal);
+
+	nbpf_chan_configure(chan);
+
+	return ret;
+}
+
+static void nbpf_free_chan_resources(struct dma_chan *dchan)
+{
+	struct nbpf_channel *chan = nbpf_to_chan(dchan);
+	struct nbpf_desc_page *dpage, *tmp;
+
+	dev_dbg(dchan->device->dev, "Entry %s()\n", __func__);
+
+	nbpf_chan_halt(chan);
+	nbpf_chan_idle(chan);
+	/* Clean up for if a channel is re-used for MEMCPY after slave DMA */
+	nbpf_chan_prepare_default(chan);
+
+	list_for_each_entry_safe(dpage, tmp, &chan->desc_page, node) {
+		struct nbpf_link_desc *ldesc;
+		int i;
+		list_del(&dpage->node);
+		for (i = 0, ldesc = dpage->ldesc;
+		     i < ARRAY_SIZE(dpage->ldesc);
+		     i++, ldesc++)
+			dma_unmap_single(dchan->device->dev, ldesc->hwdesc_dma_addr,
+					 sizeof(*ldesc->hwdesc), DMA_TO_DEVICE);
+		free_page((unsigned long)dpage);
+	}
+}
+
+static int nbpf_slave_caps(struct dma_chan *dchan,
+			   struct dma_slave_caps *caps)
+{
+	caps->src_addr_widths = NBPF_DMA_BUSWIDTHS;
+	caps->dstn_addr_widths = NBPF_DMA_BUSWIDTHS;
+	caps->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+	caps->cmd_pause = false;
+	caps->cmd_terminate = true;
+
+	return 0;
+}
+
+static struct dma_chan *nbpf_of_xlate(struct of_phandle_args *dma_spec,
+				      struct of_dma *ofdma)
+{
+	struct nbpf_device *nbpf = ofdma->of_dma_data;
+	struct dma_chan *dchan;
+	struct nbpf_channel *chan;
+
+	if (dma_spec->args_count != 2)
+		return NULL;
+
+	dchan = dma_get_any_slave_channel(&nbpf->dma_dev);
+	if (!dchan)
+		return NULL;
+
+	dev_dbg(dchan->device->dev, "Entry %s(%s)\n", __func__,
+		dma_spec->np->name);
+
+	chan = nbpf_to_chan(dchan);
+
+	chan->terminal = dma_spec->args[0];
+	chan->flags = dma_spec->args[1];
+
+	nbpf_chan_prepare(chan);
+	nbpf_chan_configure(chan);
+
+	return dchan;
+}
+
+static void nbpf_chan_tasklet(unsigned long data)
+{
+	struct nbpf_channel *chan = (struct nbpf_channel *)data;
+	struct nbpf_desc *desc, *tmp;
+	dma_async_tx_callback callback;
+	void *param;
+
+	while (!list_empty(&chan->done)) {
+		bool found = false, must_put, recycling = false;
+
+		spin_lock_irq(&chan->lock);
+
+		list_for_each_entry_safe(desc, tmp, &chan->done, node) {
+			if (!desc->user_wait) {
+				/* Newly completed descriptor, have to process */
+				found = true;
+				break;
+			} else if (async_tx_test_ack(&desc->async_tx)) {
+				/*
+				 * This descriptor was waiting for a user ACK,
+				 * it can be recycled now.
+				 */
+				list_del(&desc->node);
+				spin_unlock_irq(&chan->lock);
+				nbpf_desc_put(desc);
+				recycling = true;
+				break;
+			}
+		}
+
+		if (recycling)
+			continue;
+
+		if (!found) {
+			/* This can happen if TERMINATE_ALL has been called */
+			spin_unlock_irq(&chan->lock);
+			break;
+		}
+
+		dma_cookie_complete(&desc->async_tx);
+
+		/*
+		 * With released lock we cannot dereference desc, maybe it's
+		 * still on the "done" list
+		 */
+		if (async_tx_test_ack(&desc->async_tx)) {
+			list_del(&desc->node);
+			must_put = true;
+		} else {
+			desc->user_wait = true;
+			must_put = false;
+		}
+
+		callback = desc->async_tx.callback;
+		param = desc->async_tx.callback_param;
+
+		/* ack and callback completed descriptor */
+		spin_unlock_irq(&chan->lock);
+
+		if (callback)
+			callback(param);
+
+		if (must_put)
+			nbpf_desc_put(desc);
+	}
+}
+
+static irqreturn_t nbpf_chan_irq(int irq, void *dev)
+{
+	struct nbpf_channel *chan = dev;
+	bool done = nbpf_status_get(chan);
+	struct nbpf_desc *desc;
+	irqreturn_t ret;
+	bool bh = false;
+
+	if (!done)
+		return IRQ_NONE;
+
+	nbpf_status_ack(chan);
+
+	dev_dbg(&chan->dma_chan.dev->device, "%s()\n", __func__);
+
+	spin_lock(&chan->lock);
+	desc = chan->running;
+	if (WARN_ON(!desc)) {
+		ret = IRQ_NONE;
+		goto unlock;
+	} else {
+		ret = IRQ_HANDLED;
+		bh = true;
+	}
+
+	list_move_tail(&desc->node, &chan->done);
+	chan->running = NULL;
+
+	if (!list_empty(&chan->active)) {
+		desc = list_first_entry(&chan->active,
+					struct nbpf_desc, node);
+		if (!nbpf_start(desc))
+			chan->running = desc;
+	}
+
+unlock:
+	spin_unlock(&chan->lock);
+
+	if (bh)
+		tasklet_schedule(&chan->tasklet);
+
+	return ret;
+}
+
+static irqreturn_t nbpf_err_irq(int irq, void *dev)
+{
+	struct nbpf_device *nbpf = dev;
+	u32 error = nbpf_error_get(nbpf);
+
+	dev_warn(nbpf->dma_dev.dev, "DMA error IRQ %u\n", irq);
+
+	if (!error)
+		return IRQ_NONE;
+
+	do {
+		struct nbpf_channel *chan = nbpf_error_get_channel(nbpf, error);
+		/* On error: abort all queued transfers, no callback */
+		nbpf_error_clear(chan);
+		nbpf_chan_idle(chan);
+		error = nbpf_error_get(nbpf);
+	} while (error);
+
+	return IRQ_HANDLED;
+}
+
+static int nbpf_chan_probe(struct nbpf_device *nbpf, int n)
+{
+	struct dma_device *dma_dev = &nbpf->dma_dev;
+	struct nbpf_channel *chan = nbpf->chan + n;
+	int ret;
+
+	chan->nbpf = nbpf;
+	chan->base = nbpf->base + NBPF_REG_CHAN_OFFSET + NBPF_REG_CHAN_SIZE * n;
+	INIT_LIST_HEAD(&chan->desc_page);
+	spin_lock_init(&chan->lock);
+	chan->dma_chan.device = dma_dev;
+	dma_cookie_init(&chan->dma_chan);
+	nbpf_chan_prepare_default(chan);
+
+	dev_dbg(dma_dev->dev, "%s(): channel %d: -> %p\n", __func__, n, chan->base);
+
+	snprintf(chan->name, sizeof(chan->name), "nbpf %d", n);
+
+	tasklet_init(&chan->tasklet, nbpf_chan_tasklet, (unsigned long)chan);
+	ret = devm_request_irq(dma_dev->dev, chan->irq,
+			nbpf_chan_irq, IRQF_SHARED,
+			chan->name, chan);
+	if (ret < 0)
+		return ret;
+
+	/* Add the channel to DMA device channel list */
+	list_add_tail(&chan->dma_chan.device_node,
+		      &dma_dev->channels);
+
+	return 0;
+}
+
+static const struct of_device_id nbpf_match[] = {
+	{.compatible = "renesas,nbpfaxi64dmac1b4",	.data = &nbpf_cfg[NBPF1B4]},
+	{.compatible = "renesas,nbpfaxi64dmac1b8",	.data = &nbpf_cfg[NBPF1B8]},
+	{.compatible = "renesas,nbpfaxi64dmac1b16",	.data = &nbpf_cfg[NBPF1B16]},
+	{.compatible = "renesas,nbpfaxi64dmac4b4",	.data = &nbpf_cfg[NBPF4B4]},
+	{.compatible = "renesas,nbpfaxi64dmac4b8",	.data = &nbpf_cfg[NBPF4B8]},
+	{.compatible = "renesas,nbpfaxi64dmac4b16",	.data = &nbpf_cfg[NBPF4B16]},
+	{.compatible = "renesas,nbpfaxi64dmac8b4",	.data = &nbpf_cfg[NBPF8B4]},
+	{.compatible = "renesas,nbpfaxi64dmac8b8",	.data = &nbpf_cfg[NBPF8B8]},
+	{.compatible = "renesas,nbpfaxi64dmac8b16",	.data = &nbpf_cfg[NBPF8B16]},
+	{}
+};
+MODULE_DEVICE_TABLE(of, nbpf_match);
+
+static int nbpf_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	const struct of_device_id *of_id = of_match_device(nbpf_match, dev);
+	struct device_node *np = dev->of_node;
+	struct nbpf_device *nbpf;
+	struct dma_device *dma_dev;
+	struct resource *iomem, *irq_res;
+	const struct nbpf_config *cfg;
+	int num_channels;
+	int ret, irq, eirq, i;
+	int irqbuf[9] /* maximum 8 channels + error IRQ */;
+	unsigned int irqs = 0;
+
+	BUILD_BUG_ON(sizeof(struct nbpf_desc_page) > PAGE_SIZE);
+
+	/* DT only */
+	if (!np || !of_id || !of_id->data)
+		return -ENODEV;
+
+	cfg = of_id->data;
+	num_channels = cfg->num_channels;
+
+	nbpf = devm_kzalloc(dev, sizeof(*nbpf) + num_channels *
+			    sizeof(nbpf->chan[0]), GFP_KERNEL);
+	if (!nbpf) {
+		dev_err(dev, "Memory allocation failed\n");
+		return -ENOMEM;
+	}
+	dma_dev = &nbpf->dma_dev;
+	dma_dev->dev = dev;
+
+	iomem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	nbpf->base = devm_ioremap_resource(dev, iomem);
+	if (IS_ERR(nbpf->base))
+		return PTR_ERR(nbpf->base);
+
+	nbpf->clk = devm_clk_get(dev, NULL);
+	if (IS_ERR(nbpf->clk))
+		return PTR_ERR(nbpf->clk);
+
+	nbpf->config = cfg;
+
+	for (i = 0; irqs < ARRAY_SIZE(irqbuf); i++) {
+		irq_res = platform_get_resource(pdev, IORESOURCE_IRQ, i);
+		if (!irq_res)
+			break;
+
+		for (irq = irq_res->start; irq <= irq_res->end;
+		     irq++, irqs++)
+			irqbuf[irqs] = irq;
+	}
+
+	/*
+	 * 3 IRQ resource schemes are supported:
+	 * 1. 1 shared IRQ for error and all channels
+	 * 2. 2 IRQs: one for error and one shared for all channels
+	 * 3. 1 IRQ for error and an own IRQ for each channel
+	 */
+	if (irqs != 1 && irqs != 2 && irqs != num_channels + 1)
+		return -ENXIO;
+
+	if (irqs == 1) {
+		eirq = irqbuf[0];
+
+		for (i = 0; i <= num_channels; i++)
+			nbpf->chan[i].irq = irqbuf[0];
+	} else {
+		eirq = platform_get_irq_byname(pdev, "error");
+		if (eirq < 0)
+			return eirq;
+
+		if (irqs == num_channels + 1) {
+			struct nbpf_channel *chan;
+
+			for (i = 0, chan = nbpf->chan; i <= num_channels;
+			     i++, chan++) {
+				/* Skip the error IRQ */
+				if (irqbuf[i] == eirq)
+					i++;
+				chan->irq = irqbuf[i];
+			}
+
+			if (chan != nbpf->chan + num_channels)
+				return -EINVAL;
+		} else {
+			/* 2 IRQs and more than one channel */
+			if (irqbuf[0] == eirq)
+				irq = irqbuf[1];
+			else
+				irq = irqbuf[0];
+
+			for (i = 0; i <= num_channels; i++)
+				nbpf->chan[i].irq = irq;
+		}
+	}
+
+	ret = devm_request_irq(dev, eirq, nbpf_err_irq,
+			       IRQF_SHARED, "dma error", nbpf);
+	if (ret < 0)
+		return ret;
+
+	INIT_LIST_HEAD(&dma_dev->channels);
+
+	/* Create DMA Channel */
+	for (i = 0; i < num_channels; i++) {
+		ret = nbpf_chan_probe(nbpf, i);
+		if (ret < 0)
+			return ret;
+	}
+
+	dma_cap_set(DMA_MEMCPY, dma_dev->cap_mask);
+	dma_cap_set(DMA_SLAVE, dma_dev->cap_mask);
+	dma_cap_set(DMA_PRIVATE, dma_dev->cap_mask);
+	dma_cap_set(DMA_SG, dma_dev->cap_mask);
+
+	/* Common and MEMCPY operations */
+	dma_dev->device_alloc_chan_resources
+		= nbpf_alloc_chan_resources;
+	dma_dev->device_free_chan_resources = nbpf_free_chan_resources;
+	dma_dev->device_prep_dma_sg = nbpf_prep_memcpy_sg;
+	dma_dev->device_prep_dma_memcpy = nbpf_prep_memcpy;
+	dma_dev->device_tx_status = nbpf_tx_status;
+	dma_dev->device_issue_pending = nbpf_issue_pending;
+	dma_dev->device_slave_caps = nbpf_slave_caps;
+
+	/*
+	 * If we drop support for unaligned MEMCPY buffer addresses and / or
+	 * lengths by setting
+	 * dma_dev->copy_align = 4;
+	 * then we can set transfer length to 4 bytes in nbpf_prep_one() for
+	 * DMA_MEM_TO_MEM
+	 */
+
+	/* Compulsory for DMA_SLAVE fields */
+	dma_dev->device_prep_slave_sg = nbpf_prep_slave_sg;
+	dma_dev->device_control = nbpf_control;
+
+	platform_set_drvdata(pdev, nbpf);
+
+	ret = clk_prepare_enable(nbpf->clk);
+	if (ret < 0)
+		return ret;
+
+	nbpf_configure(nbpf);
+
+	ret = dma_async_device_register(dma_dev);
+	if (ret < 0)
+		goto e_clk_off;
+
+	ret = of_dma_controller_register(np, nbpf_of_xlate, nbpf);
+	if (ret < 0)
+		goto e_dma_dev_unreg;
+
+	return 0;
+
+e_dma_dev_unreg:
+	dma_async_device_unregister(dma_dev);
+e_clk_off:
+	clk_disable_unprepare(nbpf->clk);
+
+	return ret;
+}
+
+static int nbpf_remove(struct platform_device *pdev)
+{
+	struct nbpf_device *nbpf = platform_get_drvdata(pdev);
+
+	of_dma_controller_free(pdev->dev.of_node);
+	dma_async_device_unregister(&nbpf->dma_dev);
+	clk_disable_unprepare(nbpf->clk);
+
+	return 0;
+}
+
+static struct platform_device_id nbpf_ids[] = {
+	{"nbpfaxi64dmac1b4",	(kernel_ulong_t)&nbpf_cfg[NBPF1B4]},
+	{"nbpfaxi64dmac1b8",	(kernel_ulong_t)&nbpf_cfg[NBPF1B8]},
+	{"nbpfaxi64dmac1b16",	(kernel_ulong_t)&nbpf_cfg[NBPF1B16]},
+	{"nbpfaxi64dmac4b4",	(kernel_ulong_t)&nbpf_cfg[NBPF4B4]},
+	{"nbpfaxi64dmac4b8",	(kernel_ulong_t)&nbpf_cfg[NBPF4B8]},
+	{"nbpfaxi64dmac4b16",	(kernel_ulong_t)&nbpf_cfg[NBPF4B16]},
+	{"nbpfaxi64dmac8b4",	(kernel_ulong_t)&nbpf_cfg[NBPF8B4]},
+	{"nbpfaxi64dmac8b8",	(kernel_ulong_t)&nbpf_cfg[NBPF8B8]},
+	{"nbpfaxi64dmac8b16",	(kernel_ulong_t)&nbpf_cfg[NBPF8B16]},
+	{},
+};
+MODULE_DEVICE_TABLE(platform, nbpf_ids);
+
+#ifdef CONFIG_PM_RUNTIME
+static int nbpf_runtime_suspend(struct device *dev)
+{
+	struct nbpf_device *nbpf = platform_get_drvdata(to_platform_device(dev));
+	clk_disable_unprepare(nbpf->clk);
+	return 0;
+}
+
+static int nbpf_runtime_resume(struct device *dev)
+{
+	struct nbpf_device *nbpf = platform_get_drvdata(to_platform_device(dev));
+	return clk_prepare_enable(nbpf->clk);
+}
+#endif
+
+static const struct dev_pm_ops nbpf_pm_ops = {
+	SET_RUNTIME_PM_OPS(nbpf_runtime_suspend, nbpf_runtime_resume, NULL)
+};
+
+static struct platform_driver nbpf_driver = {
+	.driver = {
+		.owner = THIS_MODULE,
+		.name = "dma-nbpf",
+		.of_match_table = nbpf_match,
+		.pm = &nbpf_pm_ops,
+	},
+	.id_table = nbpf_ids,
+	.probe = nbpf_probe,
+	.remove = nbpf_remove,
+};
+
+module_platform_driver(nbpf_driver);
+
+MODULE_AUTHOR("Guennadi Liakhovetski <g.liakhovetski@gmx.de>");
+MODULE_DESCRIPTION("dmaengine driver for NBPFAXI64* DMACs");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/dma/of-dma.c b/drivers/dma/of-dma.c
index e8fe9dc455f4..d5fbeaa1e7ba 100644
--- a/drivers/dma/of-dma.c
+++ b/drivers/dma/of-dma.c
@@ -218,3 +218,38 @@ struct dma_chan *of_dma_simple_xlate(struct of_phandle_args *dma_spec,
 			&dma_spec->args[0]);
 }
 EXPORT_SYMBOL_GPL(of_dma_simple_xlate);
+
+/**
+ * of_dma_xlate_by_chan_id - Translate dt property to DMA channel by channel id
+ * @dma_spec:	pointer to DMA specifier as found in the device tree
+ * @of_dma:	pointer to DMA controller data
+ *
+ * This function can be used as the of xlate callback for DMA driver which wants
+ * to match the channel based on the channel id. When using this xlate function
+ * the #dma-cells propety of the DMA controller dt node needs to be set to 1.
+ * The data parameter of of_dma_controller_register must be a pointer to the
+ * dma_device struct the function should match upon.
+ *
+ * Returns pointer to appropriate dma channel on success or NULL on error.
+ */
+struct dma_chan *of_dma_xlate_by_chan_id(struct of_phandle_args *dma_spec,
+					 struct of_dma *ofdma)
+{
+	struct dma_device *dev = ofdma->of_dma_data;
+	struct dma_chan *chan, *candidate = NULL;
+
+	if (!dev || dma_spec->args_count != 1)
+		return NULL;
+
+	list_for_each_entry(chan, &dev->channels, device_node)
+		if (chan->chan_id == dma_spec->args[0]) {
+			candidate = chan;
+			break;
+		}
+
+	if (!candidate)
+		return NULL;
+
+	return dma_get_slave_channel(candidate);
+}
+EXPORT_SYMBOL_GPL(of_dma_xlate_by_chan_id);
diff --git a/drivers/dma/omap-dma.c b/drivers/dma/omap-dma.c
index b19f04f4390b..4cf7d9a950d7 100644
--- a/drivers/dma/omap-dma.c
+++ b/drivers/dma/omap-dma.c
@@ -853,8 +853,7 @@ static struct dma_async_tx_descriptor *omap_dma_prep_slave_sg(
 
 static struct dma_async_tx_descriptor *omap_dma_prep_dma_cyclic(
 	struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
-	size_t period_len, enum dma_transfer_direction dir, unsigned long flags,
-	void *context)
+	size_t period_len, enum dma_transfer_direction dir, unsigned long flags)
 {
 	struct omap_dmadev *od = to_omap_dma_dev(chan->device);
 	struct omap_chan *c = to_omap_dma_chan(chan);
diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c
index 73fa9b7a10ab..d5149aacd2fe 100644
--- a/drivers/dma/pl330.c
+++ b/drivers/dma/pl330.c
@@ -33,26 +33,15 @@
 #define PL330_MAX_IRQS		32
 #define PL330_MAX_PERI		32
 
-enum pl330_srccachectrl {
-	SCCTRL0,	/* Noncacheable and nonbufferable */
-	SCCTRL1,	/* Bufferable only */
-	SCCTRL2,	/* Cacheable, but do not allocate */
-	SCCTRL3,	/* Cacheable and bufferable, but do not allocate */
-	SINVALID1,
-	SINVALID2,
-	SCCTRL6,	/* Cacheable write-through, allocate on reads only */
-	SCCTRL7,	/* Cacheable write-back, allocate on reads only */
-};
-
-enum pl330_dstcachectrl {
-	DCCTRL0,	/* Noncacheable and nonbufferable */
-	DCCTRL1,	/* Bufferable only */
-	DCCTRL2,	/* Cacheable, but do not allocate */
-	DCCTRL3,	/* Cacheable and bufferable, but do not allocate */
-	DINVALID1,	/* AWCACHE = 0x1000 */
-	DINVALID2,
-	DCCTRL6,	/* Cacheable write-through, allocate on writes only */
-	DCCTRL7,	/* Cacheable write-back, allocate on writes only */
+enum pl330_cachectrl {
+	CCTRL0,		/* Noncacheable and nonbufferable */
+	CCTRL1,		/* Bufferable only */
+	CCTRL2,		/* Cacheable, but do not allocate */
+	CCTRL3,		/* Cacheable and bufferable, but do not allocate */
+	INVALID1,	/* AWCACHE = 0x1000 */
+	INVALID2,
+	CCTRL6,		/* Cacheable write-through, allocate on writes only */
+	CCTRL7,		/* Cacheable write-back, allocate on writes only */
 };
 
 enum pl330_byteswap {
@@ -63,13 +52,6 @@ enum pl330_byteswap {
 	SWAP_16,
 };
 
-enum pl330_reqtype {
-	MEMTOMEM,
-	MEMTODEV,
-	DEVTOMEM,
-	DEVTODEV,
-};
-
 /* Register and Bit field Definitions */
 #define DS			0x0
 #define DS_ST_STOP		0x0
@@ -263,9 +245,6 @@ enum pl330_reqtype {
  */
 #define MCODE_BUFF_PER_REQ	256
 
-/* If the _pl330_req is available to the client */
-#define IS_FREE(req)	(*((u8 *)((req)->mc_cpu)) == CMD_DMAEND)
-
 /* Use this _only_ to wait on transient states */
 #define UNTIL(t, s)	while (!(_state(t) & (s))) cpu_relax();
 
@@ -300,27 +279,6 @@ struct pl330_config {
 	u32		irq_ns;
 };
 
-/* Handle to the DMAC provided to the PL330 core */
-struct pl330_info {
-	/* Owning device */
-	struct device *dev;
-	/* Size of MicroCode buffers for each channel. */
-	unsigned mcbufsz;
-	/* ioremap'ed address of PL330 registers. */
-	void __iomem	*base;
-	/* Client can freely use it. */
-	void	*client_data;
-	/* PL330 core data, Client must not touch it. */
-	void	*pl330_data;
-	/* Populated by the PL330 core driver during pl330_add */
-	struct pl330_config	pcfg;
-	/*
-	 * If the DMAC has some reset mechanism, then the
-	 * client may want to provide pointer to the method.
-	 */
-	void (*dmac_reset)(struct pl330_info *pi);
-};
-
 /**
  * Request Configuration.
  * The PL330 core does not modify this and uses the last
@@ -344,8 +302,8 @@ struct pl330_reqcfg {
 	unsigned brst_len:5;
 	unsigned brst_size:3; /* in power of 2 */
 
-	enum pl330_dstcachectrl dcctl;
-	enum pl330_srccachectrl scctl;
+	enum pl330_cachectrl dcctl;
+	enum pl330_cachectrl scctl;
 	enum pl330_byteswap swap;
 	struct pl330_config *pcfg;
 };
@@ -359,11 +317,6 @@ struct pl330_xfer {
 	u32 dst_addr;
 	/* Size to xfer */
 	u32 bytes;
-	/*
-	 * Pointer to next xfer in the list.
-	 * The last xfer in the req must point to NULL.
-	 */
-	struct pl330_xfer *next;
 };
 
 /* The xfer callbacks are made with one of these arguments. */
@@ -376,67 +329,6 @@ enum pl330_op_err {
 	PL330_ERR_FAIL,
 };
 
-/* A request defining Scatter-Gather List ending with NULL xfer. */
-struct pl330_req {
-	enum pl330_reqtype rqtype;
-	/* Index of peripheral for the xfer. */
-	unsigned peri:5;
-	/* Unique token for this xfer, set by the client. */
-	void *token;
-	/* Callback to be called after xfer. */
-	void (*xfer_cb)(void *token, enum pl330_op_err err);
-	/* If NULL, req will be done at last set parameters. */
-	struct pl330_reqcfg *cfg;
-	/* Pointer to first xfer in the request. */
-	struct pl330_xfer *x;
-	/* Hook to attach to DMAC's list of reqs with due callback */
-	struct list_head rqd;
-};
-
-/*
- * To know the status of the channel and DMAC, the client
- * provides a pointer to this structure. The PL330 core
- * fills it with current information.
- */
-struct pl330_chanstatus {
-	/*
-	 * If the DMAC engine halted due to some error,
-	 * the client should remove-add DMAC.
-	 */
-	bool dmac_halted;
-	/*
-	 * If channel is halted due to some error,
-	 * the client should ABORT/FLUSH and START the channel.
-	 */
-	bool faulting;
-	/* Location of last load */
-	u32 src_addr;
-	/* Location of last store */
-	u32 dst_addr;
-	/*
-	 * Pointer to the currently active req, NULL if channel is
-	 * inactive, even though the requests may be present.
-	 */
-	struct pl330_req *top_req;
-	/* Pointer to req waiting second in the queue if any. */
-	struct pl330_req *wait_req;
-};
-
-enum pl330_chan_op {
-	/* Start the channel */
-	PL330_OP_START,
-	/* Abort the active xfer */
-	PL330_OP_ABORT,
-	/* Stop xfer and flush queue */
-	PL330_OP_FLUSH,
-};
-
-struct _xfer_spec {
-	u32 ccr;
-	struct pl330_req *r;
-	struct pl330_xfer *x;
-};
-
 enum dmamov_dst {
 	SAR = 0,
 	CCR,
@@ -454,12 +346,12 @@ enum pl330_cond {
 	ALWAYS,
 };
 
+struct dma_pl330_desc;
+
 struct _pl330_req {
 	u32 mc_bus;
 	void *mc_cpu;
-	/* Number of bytes taken to setup MC for the req */
-	u32 mc_len;
-	struct pl330_req *r;
+	struct dma_pl330_desc *desc;
 };
 
 /* ToBeDone for tasklet */
@@ -491,30 +383,6 @@ enum pl330_dmac_state {
 	DYING,
 };
 
-/* A DMAC */
-struct pl330_dmac {
-	spinlock_t		lock;
-	/* Holds list of reqs with due callbacks */
-	struct list_head	req_done;
-	/* Pointer to platform specific stuff */
-	struct pl330_info	*pinfo;
-	/* Maximum possible events/irqs */
-	int			events[32];
-	/* BUS address of MicroCode buffer */
-	dma_addr_t		mcode_bus;
-	/* CPU address of MicroCode buffer */
-	void			*mcode_cpu;
-	/* List of all Channel threads */
-	struct pl330_thread	*channels;
-	/* Pointer to the MANAGER thread */
-	struct pl330_thread	*manager;
-	/* To handle bad news in interrupt */
-	struct tasklet_struct	tasks;
-	struct _pl330_tbd	dmac_tbd;
-	/* State of DMAC operation */
-	enum pl330_dmac_state	state;
-};
-
 enum desc_status {
 	/* In the DMAC pool */
 	FREE,
@@ -555,15 +423,16 @@ struct dma_pl330_chan {
 	 * As the parent, this DMAC also provides descriptors
 	 * to the channel.
 	 */
-	struct dma_pl330_dmac *dmac;
+	struct pl330_dmac *dmac;
 
 	/* To protect channel manipulation */
 	spinlock_t lock;
 
-	/* Token of a hardware channel thread of PL330 DMAC
-	 * NULL if the channel is available to be acquired.
+	/*
+	 * Hardware channel thread of PL330 DMAC. NULL if the channel is
+	 * available.
 	 */
-	void *pl330_chid;
+	struct pl330_thread *thread;
 
 	/* For D-to-M and M-to-D channels */
 	int burst_sz; /* the peripheral fifo width */
@@ -574,9 +443,7 @@ struct dma_pl330_chan {
 	bool cyclic;
 };
 
-struct dma_pl330_dmac {
-	struct pl330_info pif;
-
+struct pl330_dmac {
 	/* DMA-Engine Device */
 	struct dma_device ddma;
 
@@ -588,6 +455,32 @@ struct dma_pl330_dmac {
 	/* To protect desc_pool manipulation */
 	spinlock_t pool_lock;
 
+	/* Size of MicroCode buffers for each channel. */
+	unsigned mcbufsz;
+	/* ioremap'ed address of PL330 registers. */
+	void __iomem	*base;
+	/* Populated by the PL330 core driver during pl330_add */
+	struct pl330_config	pcfg;
+
+	spinlock_t		lock;
+	/* Maximum possible events/irqs */
+	int			events[32];
+	/* BUS address of MicroCode buffer */
+	dma_addr_t		mcode_bus;
+	/* CPU address of MicroCode buffer */
+	void			*mcode_cpu;
+	/* List of all Channel threads */
+	struct pl330_thread	*channels;
+	/* Pointer to the MANAGER thread */
+	struct pl330_thread	*manager;
+	/* To handle bad news in interrupt */
+	struct tasklet_struct	tasks;
+	struct _pl330_tbd	dmac_tbd;
+	/* State of DMAC operation */
+	enum pl330_dmac_state	state;
+	/* Holds list of reqs with due callbacks */
+	struct list_head        req_done;
+
 	/* Peripheral channels connected to this DMAC */
 	unsigned int num_peripherals;
 	struct dma_pl330_chan *peripherals; /* keep at end */
@@ -604,49 +497,43 @@ struct dma_pl330_desc {
 	struct pl330_xfer px;
 
 	struct pl330_reqcfg rqcfg;
-	struct pl330_req req;
 
 	enum desc_status status;
 
 	/* The channel which currently holds this desc */
 	struct dma_pl330_chan *pchan;
+
+	enum dma_transfer_direction rqtype;
+	/* Index of peripheral for the xfer. */
+	unsigned peri:5;
+	/* Hook to attach to DMAC's list of reqs with due callback */
+	struct list_head rqd;
 };
 
-static inline void _callback(struct pl330_req *r, enum pl330_op_err err)
-{
-	if (r && r->xfer_cb)
-		r->xfer_cb(r->token, err);
-}
+struct _xfer_spec {
+	u32 ccr;
+	struct dma_pl330_desc *desc;
+};
 
 static inline bool _queue_empty(struct pl330_thread *thrd)
 {
-	return (IS_FREE(&thrd->req[0]) && IS_FREE(&thrd->req[1]))
-		? true : false;
+	return thrd->req[0].desc == NULL && thrd->req[1].desc == NULL;
 }
 
 static inline bool _queue_full(struct pl330_thread *thrd)
 {
-	return (IS_FREE(&thrd->req[0]) || IS_FREE(&thrd->req[1]))
-		? false : true;
+	return thrd->req[0].desc != NULL && thrd->req[1].desc != NULL;
 }
 
 static inline bool is_manager(struct pl330_thread *thrd)
 {
-	struct pl330_dmac *pl330 = thrd->dmac;
-
-	/* MANAGER is indexed at the end */
-	if (thrd->id == pl330->pinfo->pcfg.num_chan)
-		return true;
-	else
-		return false;
+	return thrd->dmac->manager == thrd;
 }
 
 /* If manager of the thread is in Non-Secure mode */
 static inline bool _manager_ns(struct pl330_thread *thrd)
 {
-	struct pl330_dmac *pl330 = thrd->dmac;
-
-	return (pl330->pinfo->pcfg.mode & DMAC_MODE_NS) ? true : false;
+	return (thrd->dmac->pcfg.mode & DMAC_MODE_NS) ? true : false;
 }
 
 static inline u32 get_revision(u32 periph_id)
@@ -1004,7 +891,7 @@ static inline u32 _emit_GO(unsigned dry_run, u8 buf[],
 /* Returns Time-Out */
 static bool _until_dmac_idle(struct pl330_thread *thrd)
 {
-	void __iomem *regs = thrd->dmac->pinfo->base;
+	void __iomem *regs = thrd->dmac->base;
 	unsigned long loops = msecs_to_loops(5);
 
 	do {
@@ -1024,7 +911,7 @@ static bool _until_dmac_idle(struct pl330_thread *thrd)
 static inline void _execute_DBGINSN(struct pl330_thread *thrd,
 		u8 insn[], bool as_manager)
 {
-	void __iomem *regs = thrd->dmac->pinfo->base;
+	void __iomem *regs = thrd->dmac->base;
 	u32 val;
 
 	val = (insn[0] << 16) | (insn[1] << 24);
@@ -1039,7 +926,7 @@ static inline void _execute_DBGINSN(struct pl330_thread *thrd,
 
 	/* If timed out due to halted state-machine */
 	if (_until_dmac_idle(thrd)) {
-		dev_err(thrd->dmac->pinfo->dev, "DMAC halted!\n");
+		dev_err(thrd->dmac->ddma.dev, "DMAC halted!\n");
 		return;
 	}
 
@@ -1047,25 +934,9 @@ static inline void _execute_DBGINSN(struct pl330_thread *thrd,
 	writel(0, regs + DBGCMD);
 }
 
-/*
- * Mark a _pl330_req as free.
- * We do it by writing DMAEND as the first instruction
- * because no valid request is going to have DMAEND as
- * its first instruction to execute.
- */
-static void mark_free(struct pl330_thread *thrd, int idx)
-{
-	struct _pl330_req *req = &thrd->req[idx];
-
-	_emit_END(0, req->mc_cpu);
-	req->mc_len = 0;
-
-	thrd->req_running = -1;
-}
-
 static inline u32 _state(struct pl330_thread *thrd)
 {
-	void __iomem *regs = thrd->dmac->pinfo->base;
+	void __iomem *regs = thrd->dmac->base;
 	u32 val;
 
 	if (is_manager(thrd))
@@ -1123,7 +994,7 @@ static inline u32 _state(struct pl330_thread *thrd)
 
 static void _stop(struct pl330_thread *thrd)
 {
-	void __iomem *regs = thrd->dmac->pinfo->base;
+	void __iomem *regs = thrd->dmac->base;
 	u8 insn[6] = {0, 0, 0, 0, 0, 0};
 
 	if (_state(thrd) == PL330_STATE_FAULT_COMPLETING)
@@ -1146,9 +1017,9 @@ static void _stop(struct pl330_thread *thrd)
 /* Start doing req 'idx' of thread 'thrd' */
 static bool _trigger(struct pl330_thread *thrd)
 {
-	void __iomem *regs = thrd->dmac->pinfo->base;
+	void __iomem *regs = thrd->dmac->base;
 	struct _pl330_req *req;
-	struct pl330_req *r;
+	struct dma_pl330_desc *desc;
 	struct _arg_GO go;
 	unsigned ns;
 	u8 insn[6] = {0, 0, 0, 0, 0, 0};
@@ -1159,32 +1030,27 @@ static bool _trigger(struct pl330_thread *thrd)
 		return true;
 
 	idx = 1 - thrd->lstenq;
-	if (!IS_FREE(&thrd->req[idx]))
+	if (thrd->req[idx].desc != NULL) {
 		req = &thrd->req[idx];
-	else {
+	} else {
 		idx = thrd->lstenq;
-		if (!IS_FREE(&thrd->req[idx]))
+		if (thrd->req[idx].desc != NULL)
 			req = &thrd->req[idx];
 		else
 			req = NULL;
 	}
 
 	/* Return if no request */
-	if (!req || !req->r)
+	if (!req)
 		return true;
 
-	r = req->r;
+	desc = req->desc;
 
-	if (r->cfg)
-		ns = r->cfg->nonsecure ? 1 : 0;
-	else if (readl(regs + CS(thrd->id)) & CS_CNS)
-		ns = 1;
-	else
-		ns = 0;
+	ns = desc->rqcfg.nonsecure ? 1 : 0;
 
 	/* See 'Abort Sources' point-4 at Page 2-25 */
 	if (_manager_ns(thrd) && !ns)
-		dev_info(thrd->dmac->pinfo->dev, "%s:%d Recipe for ABORT!\n",
+		dev_info(thrd->dmac->ddma.dev, "%s:%d Recipe for ABORT!\n",
 			__func__, __LINE__);
 
 	go.chan = thrd->id;
@@ -1240,7 +1106,7 @@ static inline int _ldst_memtomem(unsigned dry_run, u8 buf[],
 		const struct _xfer_spec *pxs, int cyc)
 {
 	int off = 0;
-	struct pl330_config *pcfg = pxs->r->cfg->pcfg;
+	struct pl330_config *pcfg = pxs->desc->rqcfg.pcfg;
 
 	/* check lock-up free version */
 	if (get_revision(pcfg->periph_id) >= PERIPH_REV_R1P0) {
@@ -1266,10 +1132,10 @@ static inline int _ldst_devtomem(unsigned dry_run, u8 buf[],
 	int off = 0;
 
 	while (cyc--) {
-		off += _emit_WFP(dry_run, &buf[off], SINGLE, pxs->r->peri);
-		off += _emit_LDP(dry_run, &buf[off], SINGLE, pxs->r->peri);
+		off += _emit_WFP(dry_run, &buf[off], SINGLE, pxs->desc->peri);
+		off += _emit_LDP(dry_run, &buf[off], SINGLE, pxs->desc->peri);
 		off += _emit_ST(dry_run, &buf[off], ALWAYS);
-		off += _emit_FLUSHP(dry_run, &buf[off], pxs->r->peri);
+		off += _emit_FLUSHP(dry_run, &buf[off], pxs->desc->peri);
 	}
 
 	return off;
@@ -1281,10 +1147,10 @@ static inline int _ldst_memtodev(unsigned dry_run, u8 buf[],
 	int off = 0;
 
 	while (cyc--) {
-		off += _emit_WFP(dry_run, &buf[off], SINGLE, pxs->r->peri);
+		off += _emit_WFP(dry_run, &buf[off], SINGLE, pxs->desc->peri);
 		off += _emit_LD(dry_run, &buf[off], ALWAYS);
-		off += _emit_STP(dry_run, &buf[off], SINGLE, pxs->r->peri);
-		off += _emit_FLUSHP(dry_run, &buf[off], pxs->r->peri);
+		off += _emit_STP(dry_run, &buf[off], SINGLE, pxs->desc->peri);
+		off += _emit_FLUSHP(dry_run, &buf[off], pxs->desc->peri);
 	}
 
 	return off;
@@ -1295,14 +1161,14 @@ static int _bursts(unsigned dry_run, u8 buf[],
 {
 	int off = 0;
 
-	switch (pxs->r->rqtype) {
-	case MEMTODEV:
+	switch (pxs->desc->rqtype) {
+	case DMA_MEM_TO_DEV:
 		off += _ldst_memtodev(dry_run, &buf[off], pxs, cyc);
 		break;
-	case DEVTOMEM:
+	case DMA_DEV_TO_MEM:
 		off += _ldst_devtomem(dry_run, &buf[off], pxs, cyc);
 		break;
-	case MEMTOMEM:
+	case DMA_MEM_TO_MEM:
 		off += _ldst_memtomem(dry_run, &buf[off], pxs, cyc);
 		break;
 	default:
@@ -1395,7 +1261,7 @@ static inline int _loop(unsigned dry_run, u8 buf[],
 static inline int _setup_loops(unsigned dry_run, u8 buf[],
 		const struct _xfer_spec *pxs)
 {
-	struct pl330_xfer *x = pxs->x;
+	struct pl330_xfer *x = &pxs->desc->px;
 	u32 ccr = pxs->ccr;
 	unsigned long c, bursts = BYTE_TO_BURST(x->bytes, ccr);
 	int off = 0;
@@ -1412,7 +1278,7 @@ static inline int _setup_loops(unsigned dry_run, u8 buf[],
 static inline int _setup_xfer(unsigned dry_run, u8 buf[],
 		const struct _xfer_spec *pxs)
 {
-	struct pl330_xfer *x = pxs->x;
+	struct pl330_xfer *x = &pxs->desc->px;
 	int off = 0;
 
 	/* DMAMOV SAR, x->src_addr */
@@ -1443,17 +1309,12 @@ static int _setup_req(unsigned dry_run, struct pl330_thread *thrd,
 	/* DMAMOV CCR, ccr */
 	off += _emit_MOV(dry_run, &buf[off], CCR, pxs->ccr);
 
-	x = pxs->r->x;
-	do {
-		/* Error if xfer length is not aligned at burst size */
-		if (x->bytes % (BRST_SIZE(pxs->ccr) * BRST_LEN(pxs->ccr)))
-			return -EINVAL;
-
-		pxs->x = x;
-		off += _setup_xfer(dry_run, &buf[off], pxs);
+	x = &pxs->desc->px;
+	/* Error if xfer length is not aligned at burst size */
+	if (x->bytes % (BRST_SIZE(pxs->ccr) * BRST_LEN(pxs->ccr)))
+		return -EINVAL;
 
-		x = x->next;
-	} while (x);
+	off += _setup_xfer(dry_run, &buf[off], pxs);
 
 	/* DMASEV peripheral/event */
 	off += _emit_SEV(dry_run, &buf[off], thrd->ev);
@@ -1495,31 +1356,15 @@ static inline u32 _prepare_ccr(const struct pl330_reqcfg *rqc)
 	return ccr;
 }
 
-static inline bool _is_valid(u32 ccr)
-{
-	enum pl330_dstcachectrl dcctl;
-	enum pl330_srccachectrl scctl;
-
-	dcctl = (ccr >> CC_DSTCCTRL_SHFT) & CC_DRCCCTRL_MASK;
-	scctl = (ccr >> CC_SRCCCTRL_SHFT) & CC_SRCCCTRL_MASK;
-
-	if (dcctl == DINVALID1 || dcctl == DINVALID2
-			|| scctl == SINVALID1 || scctl == SINVALID2)
-		return false;
-	else
-		return true;
-}
-
 /*
  * Submit a list of xfers after which the client wants notification.
  * Client is not notified after each xfer unit, just once after all
  * xfer units are done or some error occurs.
  */
-static int pl330_submit_req(void *ch_id, struct pl330_req *r)
+static int pl330_submit_req(struct pl330_thread *thrd,
+	struct dma_pl330_desc *desc)
 {
-	struct pl330_thread *thrd = ch_id;
-	struct pl330_dmac *pl330;
-	struct pl330_info *pi;
+	struct pl330_dmac *pl330 = thrd->dmac;
 	struct _xfer_spec xs;
 	unsigned long flags;
 	void __iomem *regs;
@@ -1528,25 +1373,24 @@ static int pl330_submit_req(void *ch_id, struct pl330_req *r)
 	int ret = 0;
 
 	/* No Req or Unacquired Channel or DMAC */
-	if (!r || !thrd || thrd->free)
+	if (!desc || !thrd || thrd->free)
 		return -EINVAL;
 
-	pl330 = thrd->dmac;
-	pi = pl330->pinfo;
-	regs = pi->base;
+	regs = thrd->dmac->base;
 
 	if (pl330->state == DYING
 		|| pl330->dmac_tbd.reset_chan & (1 << thrd->id)) {
-		dev_info(thrd->dmac->pinfo->dev, "%s:%d\n",
+		dev_info(thrd->dmac->ddma.dev, "%s:%d\n",
 			__func__, __LINE__);
 		return -EAGAIN;
 	}
 
 	/* If request for non-existing peripheral */
-	if (r->rqtype != MEMTOMEM && r->peri >= pi->pcfg.num_peri) {
-		dev_info(thrd->dmac->pinfo->dev,
+	if (desc->rqtype != DMA_MEM_TO_MEM &&
+	    desc->peri >= pl330->pcfg.num_peri) {
+		dev_info(thrd->dmac->ddma.dev,
 				"%s:%d Invalid peripheral(%u)!\n",
-				__func__, __LINE__, r->peri);
+				__func__, __LINE__, desc->peri);
 		return -EINVAL;
 	}
 
@@ -1557,41 +1401,26 @@ static int pl330_submit_req(void *ch_id, struct pl330_req *r)
 		goto xfer_exit;
 	}
 
+	/* Prefer Secure Channel */
+	if (!_manager_ns(thrd))
+		desc->rqcfg.nonsecure = 0;
+	else
+		desc->rqcfg.nonsecure = 1;
 
-	/* Use last settings, if not provided */
-	if (r->cfg) {
-		/* Prefer Secure Channel */
-		if (!_manager_ns(thrd))
-			r->cfg->nonsecure = 0;
-		else
-			r->cfg->nonsecure = 1;
-
-		ccr = _prepare_ccr(r->cfg);
-	} else {
-		ccr = readl(regs + CC(thrd->id));
-	}
-
-	/* If this req doesn't have valid xfer settings */
-	if (!_is_valid(ccr)) {
-		ret = -EINVAL;
-		dev_info(thrd->dmac->pinfo->dev, "%s:%d Invalid CCR(%x)!\n",
-			__func__, __LINE__, ccr);
-		goto xfer_exit;
-	}
+	ccr = _prepare_ccr(&desc->rqcfg);
 
-	idx = IS_FREE(&thrd->req[0]) ? 0 : 1;
+	idx = thrd->req[0].desc == NULL ? 0 : 1;
 
 	xs.ccr = ccr;
-	xs.r = r;
+	xs.desc = desc;
 
 	/* First dry run to check if req is acceptable */
 	ret = _setup_req(1, thrd, idx, &xs);
 	if (ret < 0)
 		goto xfer_exit;
 
-	if (ret > pi->mcbufsz / 2) {
-		dev_info(thrd->dmac->pinfo->dev,
-			"%s:%d Trying increasing mcbufsz\n",
+	if (ret > pl330->mcbufsz / 2) {
+		dev_info(pl330->ddma.dev, "%s:%d Trying increasing mcbufsz\n",
 				__func__, __LINE__);
 		ret = -ENOMEM;
 		goto xfer_exit;
@@ -1599,8 +1428,8 @@ static int pl330_submit_req(void *ch_id, struct pl330_req *r)
 
 	/* Hook the request */
 	thrd->lstenq = idx;
-	thrd->req[idx].mc_len = _setup_req(0, thrd, idx, &xs);
-	thrd->req[idx].r = r;
+	thrd->req[idx].desc = desc;
+	_setup_req(0, thrd, idx, &xs);
 
 	ret = 0;
 
@@ -1610,10 +1439,32 @@ xfer_exit:
 	return ret;
 }
 
+static void dma_pl330_rqcb(struct dma_pl330_desc *desc, enum pl330_op_err err)
+{
+	struct dma_pl330_chan *pch;
+	unsigned long flags;
+
+	if (!desc)
+		return;
+
+	pch = desc->pchan;
+
+	/* If desc aborted */
+	if (!pch)
+		return;
+
+	spin_lock_irqsave(&pch->lock, flags);
+
+	desc->status = DONE;
+
+	spin_unlock_irqrestore(&pch->lock, flags);
+
+	tasklet_schedule(&pch->task);
+}
+
 static void pl330_dotask(unsigned long data)
 {
 	struct pl330_dmac *pl330 = (struct pl330_dmac *) data;
-	struct pl330_info *pi = pl330->pinfo;
 	unsigned long flags;
 	int i;
 
@@ -1631,16 +1482,16 @@ static void pl330_dotask(unsigned long data)
 	if (pl330->dmac_tbd.reset_mngr) {
 		_stop(pl330->manager);
 		/* Reset all channels */
-		pl330->dmac_tbd.reset_chan = (1 << pi->pcfg.num_chan) - 1;
+		pl330->dmac_tbd.reset_chan = (1 << pl330->pcfg.num_chan) - 1;
 		/* Clear the reset flag */
 		pl330->dmac_tbd.reset_mngr = false;
 	}
 
-	for (i = 0; i < pi->pcfg.num_chan; i++) {
+	for (i = 0; i < pl330->pcfg.num_chan; i++) {
 
 		if (pl330->dmac_tbd.reset_chan & (1 << i)) {
 			struct pl330_thread *thrd = &pl330->channels[i];
-			void __iomem *regs = pi->base;
+			void __iomem *regs = pl330->base;
 			enum pl330_op_err err;
 
 			_stop(thrd);
@@ -1651,16 +1502,13 @@ static void pl330_dotask(unsigned long data)
 				err = PL330_ERR_ABORT;
 
 			spin_unlock_irqrestore(&pl330->lock, flags);
-
-			_callback(thrd->req[1 - thrd->lstenq].r, err);
-			_callback(thrd->req[thrd->lstenq].r, err);
-
+			dma_pl330_rqcb(thrd->req[1 - thrd->lstenq].desc, err);
+			dma_pl330_rqcb(thrd->req[thrd->lstenq].desc, err);
 			spin_lock_irqsave(&pl330->lock, flags);
 
-			thrd->req[0].r = NULL;
-			thrd->req[1].r = NULL;
-			mark_free(thrd, 0);
-			mark_free(thrd, 1);
+			thrd->req[0].desc = NULL;
+			thrd->req[1].desc = NULL;
+			thrd->req_running = -1;
 
 			/* Clear the reset flag */
 			pl330->dmac_tbd.reset_chan &= ~(1 << i);
@@ -1673,20 +1521,15 @@ static void pl330_dotask(unsigned long data)
 }
 
 /* Returns 1 if state was updated, 0 otherwise */
-static int pl330_update(const struct pl330_info *pi)
+static int pl330_update(struct pl330_dmac *pl330)
 {
-	struct pl330_req *rqdone, *tmp;
-	struct pl330_dmac *pl330;
+	struct dma_pl330_desc *descdone, *tmp;
 	unsigned long flags;
 	void __iomem *regs;
 	u32 val;
 	int id, ev, ret = 0;
 
-	if (!pi || !pi->pl330_data)
-		return 0;
-
-	regs = pi->base;
-	pl330 = pi->pl330_data;
+	regs = pl330->base;
 
 	spin_lock_irqsave(&pl330->lock, flags);
 
@@ -1696,13 +1539,13 @@ static int pl330_update(const struct pl330_info *pi)
 	else
 		pl330->dmac_tbd.reset_mngr = false;
 
-	val = readl(regs + FSC) & ((1 << pi->pcfg.num_chan) - 1);
+	val = readl(regs + FSC) & ((1 << pl330->pcfg.num_chan) - 1);
 	pl330->dmac_tbd.reset_chan |= val;
 	if (val) {
 		int i = 0;
-		while (i < pi->pcfg.num_chan) {
+		while (i < pl330->pcfg.num_chan) {
 			if (val & (1 << i)) {
-				dev_info(pi->dev,
+				dev_info(pl330->ddma.dev,
 					"Reset Channel-%d\t CS-%x FTC-%x\n",
 						i, readl(regs + CS(i)),
 						readl(regs + FTC(i)));
@@ -1714,15 +1557,16 @@ static int pl330_update(const struct pl330_info *pi)
 
 	/* Check which event happened i.e, thread notified */
 	val = readl(regs + ES);
-	if (pi->pcfg.num_events < 32
-			&& val & ~((1 << pi->pcfg.num_events) - 1)) {
+	if (pl330->pcfg.num_events < 32
+			&& val & ~((1 << pl330->pcfg.num_events) - 1)) {
 		pl330->dmac_tbd.reset_dmac = true;
-		dev_err(pi->dev, "%s:%d Unexpected!\n", __func__, __LINE__);
+		dev_err(pl330->ddma.dev, "%s:%d Unexpected!\n", __func__,
+			__LINE__);
 		ret = 1;
 		goto updt_exit;
 	}
 
-	for (ev = 0; ev < pi->pcfg.num_events; ev++) {
+	for (ev = 0; ev < pl330->pcfg.num_events; ev++) {
 		if (val & (1 << ev)) { /* Event occurred */
 			struct pl330_thread *thrd;
 			u32 inten = readl(regs + INTEN);
@@ -1743,25 +1587,22 @@ static int pl330_update(const struct pl330_info *pi)
 				continue;
 
 			/* Detach the req */
-			rqdone = thrd->req[active].r;
-			thrd->req[active].r = NULL;
-
-			mark_free(thrd, active);
+			descdone = thrd->req[active].desc;
+			thrd->req[active].desc = NULL;
 
 			/* Get going again ASAP */
 			_start(thrd);
 
 			/* For now, just make a list of callbacks to be done */
-			list_add_tail(&rqdone->rqd, &pl330->req_done);
+			list_add_tail(&descdone->rqd, &pl330->req_done);
 		}
 	}
 
 	/* Now that we are in no hurry, do the callbacks */
-	list_for_each_entry_safe(rqdone, tmp, &pl330->req_done, rqd) {
-		list_del(&rqdone->rqd);
-
+	list_for_each_entry_safe(descdone, tmp, &pl330->req_done, rqd) {
+		list_del(&descdone->rqd);
 		spin_unlock_irqrestore(&pl330->lock, flags);
-		_callback(rqdone, PL330_ERR_NONE);
+		dma_pl330_rqcb(descdone, PL330_ERR_NONE);
 		spin_lock_irqsave(&pl330->lock, flags);
 	}
 
@@ -1778,65 +1619,13 @@ updt_exit:
 	return ret;
 }
 
-static int pl330_chan_ctrl(void *ch_id, enum pl330_chan_op op)
-{
-	struct pl330_thread *thrd = ch_id;
-	struct pl330_dmac *pl330;
-	unsigned long flags;
-	int ret = 0, active;
-
-	if (!thrd || thrd->free || thrd->dmac->state == DYING)
-		return -EINVAL;
-
-	pl330 = thrd->dmac;
-	active = thrd->req_running;
-
-	spin_lock_irqsave(&pl330->lock, flags);
-
-	switch (op) {
-	case PL330_OP_FLUSH:
-		/* Make sure the channel is stopped */
-		_stop(thrd);
-
-		thrd->req[0].r = NULL;
-		thrd->req[1].r = NULL;
-		mark_free(thrd, 0);
-		mark_free(thrd, 1);
-		break;
-
-	case PL330_OP_ABORT:
-		/* Make sure the channel is stopped */
-		_stop(thrd);
-
-		/* ABORT is only for the active req */
-		if (active == -1)
-			break;
-
-		thrd->req[active].r = NULL;
-		mark_free(thrd, active);
-
-		/* Start the next */
-	case PL330_OP_START:
-		if ((active == -1) && !_start(thrd))
-			ret = -EIO;
-		break;
-
-	default:
-		ret = -EINVAL;
-	}
-
-	spin_unlock_irqrestore(&pl330->lock, flags);
-	return ret;
-}
-
 /* Reserve an event */
 static inline int _alloc_event(struct pl330_thread *thrd)
 {
 	struct pl330_dmac *pl330 = thrd->dmac;
-	struct pl330_info *pi = pl330->pinfo;
 	int ev;
 
-	for (ev = 0; ev < pi->pcfg.num_events; ev++)
+	for (ev = 0; ev < pl330->pcfg.num_events; ev++)
 		if (pl330->events[ev] == -1) {
 			pl330->events[ev] = thrd->id;
 			return ev;
@@ -1845,45 +1634,38 @@ static inline int _alloc_event(struct pl330_thread *thrd)
 	return -1;
 }
 
-static bool _chan_ns(const struct pl330_info *pi, int i)
+static bool _chan_ns(const struct pl330_dmac *pl330, int i)
 {
-	return pi->pcfg.irq_ns & (1 << i);
+	return pl330->pcfg.irq_ns & (1 << i);
 }
 
 /* Upon success, returns IdentityToken for the
  * allocated channel, NULL otherwise.
  */
-static void *pl330_request_channel(const struct pl330_info *pi)
+static struct pl330_thread *pl330_request_channel(struct pl330_dmac *pl330)
 {
 	struct pl330_thread *thrd = NULL;
-	struct pl330_dmac *pl330;
 	unsigned long flags;
 	int chans, i;
 
-	if (!pi || !pi->pl330_data)
-		return NULL;
-
-	pl330 = pi->pl330_data;
-
 	if (pl330->state == DYING)
 		return NULL;
 
-	chans = pi->pcfg.num_chan;
+	chans = pl330->pcfg.num_chan;
 
 	spin_lock_irqsave(&pl330->lock, flags);
 
 	for (i = 0; i < chans; i++) {
 		thrd = &pl330->channels[i];
 		if ((thrd->free) && (!_manager_ns(thrd) ||
-					_chan_ns(pi, i))) {
+					_chan_ns(pl330, i))) {
 			thrd->ev = _alloc_event(thrd);
 			if (thrd->ev >= 0) {
 				thrd->free = false;
 				thrd->lstenq = 1;
-				thrd->req[0].r = NULL;
-				mark_free(thrd, 0);
-				thrd->req[1].r = NULL;
-				mark_free(thrd, 1);
+				thrd->req[0].desc = NULL;
+				thrd->req[1].desc = NULL;
+				thrd->req_running = -1;
 				break;
 			}
 		}
@@ -1899,17 +1681,15 @@ static void *pl330_request_channel(const struct pl330_info *pi)
 static inline void _free_event(struct pl330_thread *thrd, int ev)
 {
 	struct pl330_dmac *pl330 = thrd->dmac;
-	struct pl330_info *pi = pl330->pinfo;
 
 	/* If the event is valid and was held by the thread */
-	if (ev >= 0 && ev < pi->pcfg.num_events
+	if (ev >= 0 && ev < pl330->pcfg.num_events
 			&& pl330->events[ev] == thrd->id)
 		pl330->events[ev] = -1;
 }
 
-static void pl330_release_channel(void *ch_id)
+static void pl330_release_channel(struct pl330_thread *thrd)
 {
-	struct pl330_thread *thrd = ch_id;
 	struct pl330_dmac *pl330;
 	unsigned long flags;
 
@@ -1918,8 +1698,8 @@ static void pl330_release_channel(void *ch_id)
 
 	_stop(thrd);
 
-	_callback(thrd->req[1 - thrd->lstenq].r, PL330_ERR_ABORT);
-	_callback(thrd->req[thrd->lstenq].r, PL330_ERR_ABORT);
+	dma_pl330_rqcb(thrd->req[1 - thrd->lstenq].desc, PL330_ERR_ABORT);
+	dma_pl330_rqcb(thrd->req[thrd->lstenq].desc, PL330_ERR_ABORT);
 
 	pl330 = thrd->dmac;
 
@@ -1932,72 +1712,70 @@ static void pl330_release_channel(void *ch_id)
 /* Initialize the structure for PL330 configuration, that can be used
  * by the client driver the make best use of the DMAC
  */
-static void read_dmac_config(struct pl330_info *pi)
+static void read_dmac_config(struct pl330_dmac *pl330)
 {
-	void __iomem *regs = pi->base;
+	void __iomem *regs = pl330->base;
 	u32 val;
 
 	val = readl(regs + CRD) >> CRD_DATA_WIDTH_SHIFT;
 	val &= CRD_DATA_WIDTH_MASK;
-	pi->pcfg.data_bus_width = 8 * (1 << val);
+	pl330->pcfg.data_bus_width = 8 * (1 << val);
 
 	val = readl(regs + CRD) >> CRD_DATA_BUFF_SHIFT;
 	val &= CRD_DATA_BUFF_MASK;
-	pi->pcfg.data_buf_dep = val + 1;
+	pl330->pcfg.data_buf_dep = val + 1;
 
 	val = readl(regs + CR0) >> CR0_NUM_CHANS_SHIFT;
 	val &= CR0_NUM_CHANS_MASK;
 	val += 1;
-	pi->pcfg.num_chan = val;
+	pl330->pcfg.num_chan = val;
 
 	val = readl(regs + CR0);
 	if (val & CR0_PERIPH_REQ_SET) {
 		val = (val >> CR0_NUM_PERIPH_SHIFT) & CR0_NUM_PERIPH_MASK;
 		val += 1;
-		pi->pcfg.num_peri = val;
-		pi->pcfg.peri_ns = readl(regs + CR4);
+		pl330->pcfg.num_peri = val;
+		pl330->pcfg.peri_ns = readl(regs + CR4);
 	} else {
-		pi->pcfg.num_peri = 0;
+		pl330->pcfg.num_peri = 0;
 	}
 
 	val = readl(regs + CR0);
 	if (val & CR0_BOOT_MAN_NS)
-		pi->pcfg.mode |= DMAC_MODE_NS;
+		pl330->pcfg.mode |= DMAC_MODE_NS;
 	else
-		pi->pcfg.mode &= ~DMAC_MODE_NS;
+		pl330->pcfg.mode &= ~DMAC_MODE_NS;
 
 	val = readl(regs + CR0) >> CR0_NUM_EVENTS_SHIFT;
 	val &= CR0_NUM_EVENTS_MASK;
 	val += 1;
-	pi->pcfg.num_events = val;
+	pl330->pcfg.num_events = val;
 
-	pi->pcfg.irq_ns = readl(regs + CR3);
+	pl330->pcfg.irq_ns = readl(regs + CR3);
 }
 
 static inline void _reset_thread(struct pl330_thread *thrd)
 {
 	struct pl330_dmac *pl330 = thrd->dmac;
-	struct pl330_info *pi = pl330->pinfo;
 
 	thrd->req[0].mc_cpu = pl330->mcode_cpu
-				+ (thrd->id * pi->mcbufsz);
+				+ (thrd->id * pl330->mcbufsz);
 	thrd->req[0].mc_bus = pl330->mcode_bus
-				+ (thrd->id * pi->mcbufsz);
-	thrd->req[0].r = NULL;
-	mark_free(thrd, 0);
+				+ (thrd->id * pl330->mcbufsz);
+	thrd->req[0].desc = NULL;
 
 	thrd->req[1].mc_cpu = thrd->req[0].mc_cpu
-				+ pi->mcbufsz / 2;
+				+ pl330->mcbufsz / 2;
 	thrd->req[1].mc_bus = thrd->req[0].mc_bus
-				+ pi->mcbufsz / 2;
-	thrd->req[1].r = NULL;
-	mark_free(thrd, 1);
+				+ pl330->mcbufsz / 2;
+	thrd->req[1].desc = NULL;
+
+	thrd->req_running = -1;
 }
 
 static int dmac_alloc_threads(struct pl330_dmac *pl330)
 {
-	struct pl330_info *pi = pl330->pinfo;
-	int chans = pi->pcfg.num_chan;
+	int chans = pl330->pcfg.num_chan;
 	struct pl330_thread *thrd;
 	int i;
 
@@ -2028,29 +1806,28 @@ static int dmac_alloc_threads(struct pl330_dmac *pl330)
 
 static int dmac_alloc_resources(struct pl330_dmac *pl330)
 {
-	struct pl330_info *pi = pl330->pinfo;
-	int chans = pi->pcfg.num_chan;
+	int chans = pl330->pcfg.num_chan;
 	int ret;
 
 	/*
 	 * Alloc MicroCode buffer for 'chans' Channel threads.
 	 * A channel's buffer offset is (Channel_Id * MCODE_BUFF_PERCHAN)
 	 */
-	pl330->mcode_cpu = dma_alloc_coherent(pi->dev,
-				chans * pi->mcbufsz,
+	pl330->mcode_cpu = dma_alloc_coherent(pl330->ddma.dev,
+				chans * pl330->mcbufsz,
 				&pl330->mcode_bus, GFP_KERNEL);
 	if (!pl330->mcode_cpu) {
-		dev_err(pi->dev, "%s:%d Can't allocate memory!\n",
+		dev_err(pl330->ddma.dev, "%s:%d Can't allocate memory!\n",
 			__func__, __LINE__);
 		return -ENOMEM;
 	}
 
 	ret = dmac_alloc_threads(pl330);
 	if (ret) {
-		dev_err(pi->dev, "%s:%d Can't to create channels for DMAC!\n",
+		dev_err(pl330->ddma.dev, "%s:%d Can't to create channels for DMAC!\n",
 			__func__, __LINE__);
-		dma_free_coherent(pi->dev,
-				chans * pi->mcbufsz,
+		dma_free_coherent(pl330->ddma.dev,
+				chans * pl330->mcbufsz,
 				pl330->mcode_cpu, pl330->mcode_bus);
 		return ret;
 	}
@@ -2058,71 +1835,45 @@ static int dmac_alloc_resources(struct pl330_dmac *pl330)
 	return 0;
 }
 
-static int pl330_add(struct pl330_info *pi)
+static int pl330_add(struct pl330_dmac *pl330)
 {
-	struct pl330_dmac *pl330;
 	void __iomem *regs;
 	int i, ret;
 
-	if (!pi || !pi->dev)
-		return -EINVAL;
-
-	/* If already added */
-	if (pi->pl330_data)
-		return -EINVAL;
-
-	/*
-	 * If the SoC can perform reset on the DMAC, then do it
-	 * before reading its configuration.
-	 */
-	if (pi->dmac_reset)
-		pi->dmac_reset(pi);
-
-	regs = pi->base;
+	regs = pl330->base;
 
 	/* Check if we can handle this DMAC */
-	if ((pi->pcfg.periph_id & 0xfffff) != PERIPH_ID_VAL) {
-		dev_err(pi->dev, "PERIPH_ID 0x%x !\n", pi->pcfg.periph_id);
+	if ((pl330->pcfg.periph_id & 0xfffff) != PERIPH_ID_VAL) {
+		dev_err(pl330->ddma.dev, "PERIPH_ID 0x%x !\n",
+			pl330->pcfg.periph_id);
 		return -EINVAL;
 	}
 
 	/* Read the configuration of the DMAC */
-	read_dmac_config(pi);
+	read_dmac_config(pl330);
 
-	if (pi->pcfg.num_events == 0) {
-		dev_err(pi->dev, "%s:%d Can't work without events!\n",
+	if (pl330->pcfg.num_events == 0) {
+		dev_err(pl330->ddma.dev, "%s:%d Can't work without events!\n",
 			__func__, __LINE__);
 		return -EINVAL;
 	}
 
-	pl330 = kzalloc(sizeof(*pl330), GFP_KERNEL);
-	if (!pl330) {
-		dev_err(pi->dev, "%s:%d Can't allocate memory!\n",
-			__func__, __LINE__);
-		return -ENOMEM;
-	}
-
-	/* Assign the info structure and private data */
-	pl330->pinfo = pi;
-	pi->pl330_data = pl330;
-
 	spin_lock_init(&pl330->lock);
 
 	INIT_LIST_HEAD(&pl330->req_done);
 
 	/* Use default MC buffer size if not provided */
-	if (!pi->mcbufsz)
-		pi->mcbufsz = MCODE_BUFF_PER_REQ * 2;
+	if (!pl330->mcbufsz)
+		pl330->mcbufsz = MCODE_BUFF_PER_REQ * 2;
 
 	/* Mark all events as free */
-	for (i = 0; i < pi->pcfg.num_events; i++)
+	for (i = 0; i < pl330->pcfg.num_events; i++)
 		pl330->events[i] = -1;
 
 	/* Allocate resources needed by the DMAC */
 	ret = dmac_alloc_resources(pl330);
 	if (ret) {
-		dev_err(pi->dev, "Unable to create channels for DMAC\n");
-		kfree(pl330);
+		dev_err(pl330->ddma.dev, "Unable to create channels for DMAC\n");
 		return ret;
 	}
 
@@ -2135,15 +1886,13 @@ static int pl330_add(struct pl330_info *pi)
 
 static int dmac_free_threads(struct pl330_dmac *pl330)
 {
-	struct pl330_info *pi = pl330->pinfo;
-	int chans = pi->pcfg.num_chan;
 	struct pl330_thread *thrd;
 	int i;
 
 	/* Release Channel threads */
-	for (i = 0; i < chans; i++) {
+	for (i = 0; i < pl330->pcfg.num_chan; i++) {
 		thrd = &pl330->channels[i];
-		pl330_release_channel((void *)thrd);
+		pl330_release_channel(thrd);
 	}
 
 	/* Free memory */
@@ -2152,35 +1901,18 @@ static int dmac_free_threads(struct pl330_dmac *pl330)
 	return 0;
 }
 
-static void dmac_free_resources(struct pl330_dmac *pl330)
+static void pl330_del(struct pl330_dmac *pl330)
 {
-	struct pl330_info *pi = pl330->pinfo;
-	int chans = pi->pcfg.num_chan;
-
-	dmac_free_threads(pl330);
-
-	dma_free_coherent(pi->dev, chans * pi->mcbufsz,
-				pl330->mcode_cpu, pl330->mcode_bus);
-}
-
-static void pl330_del(struct pl330_info *pi)
-{
-	struct pl330_dmac *pl330;
-
-	if (!pi || !pi->pl330_data)
-		return;
-
-	pl330 = pi->pl330_data;
-
 	pl330->state = UNINIT;
 
 	tasklet_kill(&pl330->tasks);
 
 	/* Free DMAC resources */
-	dmac_free_resources(pl330);
+	dmac_free_threads(pl330);
 
-	kfree(pl330);
-	pi->pl330_data = NULL;
+	dma_free_coherent(pl330->ddma.dev,
+		pl330->pcfg.num_chan * pl330->mcbufsz, pl330->mcode_cpu,
+		pl330->mcode_bus);
 }
 
 /* forward declaration */
@@ -2212,8 +1944,7 @@ static inline void fill_queue(struct dma_pl330_chan *pch)
 		if (desc->status == BUSY)
 			continue;
 
-		ret = pl330_submit_req(pch->pl330_chid,
-						&desc->req);
+		ret = pl330_submit_req(pch->thread, desc);
 		if (!ret) {
 			desc->status = BUSY;
 		} else if (ret == -EAGAIN) {
@@ -2222,7 +1953,7 @@ static inline void fill_queue(struct dma_pl330_chan *pch)
 		} else {
 			/* Unacceptable request */
 			desc->status = DONE;
-			dev_err(pch->dmac->pif.dev, "%s:%d Bad Desc(%d)\n",
+			dev_err(pch->dmac->ddma.dev, "%s:%d Bad Desc(%d)\n",
 					__func__, __LINE__, desc->txd.cookie);
 			tasklet_schedule(&pch->task);
 		}
@@ -2249,7 +1980,9 @@ static void pl330_tasklet(unsigned long data)
 	fill_queue(pch);
 
 	/* Make sure the PL330 Channel thread is active */
-	pl330_chan_ctrl(pch->pl330_chid, PL330_OP_START);
+	spin_lock(&pch->thread->dmac->lock);
+	_start(pch->thread);
+	spin_unlock(&pch->thread->dmac->lock);
 
 	while (!list_empty(&pch->completed_list)) {
 		dma_async_tx_callback callback;
@@ -2280,25 +2013,6 @@ static void pl330_tasklet(unsigned long data)
 	spin_unlock_irqrestore(&pch->lock, flags);
 }
 
-static void dma_pl330_rqcb(void *token, enum pl330_op_err err)
-{
-	struct dma_pl330_desc *desc = token;
-	struct dma_pl330_chan *pch = desc->pchan;
-	unsigned long flags;
-
-	/* If desc aborted */
-	if (!pch)
-		return;
-
-	spin_lock_irqsave(&pch->lock, flags);
-
-	desc->status = DONE;
-
-	spin_unlock_irqrestore(&pch->lock, flags);
-
-	tasklet_schedule(&pch->task);
-}
-
 bool pl330_filter(struct dma_chan *chan, void *param)
 {
 	u8 *peri_id;
@@ -2315,23 +2029,26 @@ static struct dma_chan *of_dma_pl330_xlate(struct of_phandle_args *dma_spec,
 						struct of_dma *ofdma)
 {
 	int count = dma_spec->args_count;
-	struct dma_pl330_dmac *pdmac = ofdma->of_dma_data;
+	struct pl330_dmac *pl330 = ofdma->of_dma_data;
 	unsigned int chan_id;
 
+	if (!pl330)
+		return NULL;
+
 	if (count != 1)
 		return NULL;
 
 	chan_id = dma_spec->args[0];
-	if (chan_id >= pdmac->num_peripherals)
+	if (chan_id >= pl330->num_peripherals)
 		return NULL;
 
-	return dma_get_slave_channel(&pdmac->peripherals[chan_id].chan);
+	return dma_get_slave_channel(&pl330->peripherals[chan_id].chan);
 }
 
 static int pl330_alloc_chan_resources(struct dma_chan *chan)
 {
 	struct dma_pl330_chan *pch = to_pchan(chan);
-	struct dma_pl330_dmac *pdmac = pch->dmac;
+	struct pl330_dmac *pl330 = pch->dmac;
 	unsigned long flags;
 
 	spin_lock_irqsave(&pch->lock, flags);
@@ -2339,8 +2056,8 @@ static int pl330_alloc_chan_resources(struct dma_chan *chan)
 	dma_cookie_init(chan);
 	pch->cyclic = false;
 
-	pch->pl330_chid = pl330_request_channel(&pdmac->pif);
-	if (!pch->pl330_chid) {
+	pch->thread = pl330_request_channel(pl330);
+	if (!pch->thread) {
 		spin_unlock_irqrestore(&pch->lock, flags);
 		return -ENOMEM;
 	}
@@ -2357,7 +2074,7 @@ static int pl330_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, unsigned
 	struct dma_pl330_chan *pch = to_pchan(chan);
 	struct dma_pl330_desc *desc;
 	unsigned long flags;
-	struct dma_pl330_dmac *pdmac = pch->dmac;
+	struct pl330_dmac *pl330 = pch->dmac;
 	struct dma_slave_config *slave_config;
 	LIST_HEAD(list);
 
@@ -2365,8 +2082,13 @@ static int pl330_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, unsigned
 	case DMA_TERMINATE_ALL:
 		spin_lock_irqsave(&pch->lock, flags);
 
-		/* FLUSH the PL330 Channel thread */
-		pl330_chan_ctrl(pch->pl330_chid, PL330_OP_FLUSH);
+		spin_lock(&pl330->lock);
+		_stop(pch->thread);
+		spin_unlock(&pl330->lock);
+
+		pch->thread->req[0].desc = NULL;
+		pch->thread->req[1].desc = NULL;
+		pch->thread->req_running = -1;
 
 		/* Mark all desc done */
 		list_for_each_entry(desc, &pch->submitted_list, node) {
@@ -2384,9 +2106,9 @@ static int pl330_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, unsigned
 			dma_cookie_complete(&desc->txd);
 		}
 
-		list_splice_tail_init(&pch->submitted_list, &pdmac->desc_pool);
-		list_splice_tail_init(&pch->work_list, &pdmac->desc_pool);
-		list_splice_tail_init(&pch->completed_list, &pdmac->desc_pool);
+		list_splice_tail_init(&pch->submitted_list, &pl330->desc_pool);
+		list_splice_tail_init(&pch->work_list, &pl330->desc_pool);
+		list_splice_tail_init(&pch->completed_list, &pl330->desc_pool);
 		spin_unlock_irqrestore(&pch->lock, flags);
 		break;
 	case DMA_SLAVE_CONFIG:
@@ -2409,7 +2131,7 @@ static int pl330_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, unsigned
 		}
 		break;
 	default:
-		dev_err(pch->dmac->pif.dev, "Not supported command.\n");
+		dev_err(pch->dmac->ddma.dev, "Not supported command.\n");
 		return -ENXIO;
 	}
 
@@ -2425,8 +2147,8 @@ static void pl330_free_chan_resources(struct dma_chan *chan)
 
 	spin_lock_irqsave(&pch->lock, flags);
 
-	pl330_release_channel(pch->pl330_chid);
-	pch->pl330_chid = NULL;
+	pl330_release_channel(pch->thread);
+	pch->thread = NULL;
 
 	if (pch->cyclic)
 		list_splice_tail_init(&pch->work_list, &pch->dmac->desc_pool);
@@ -2489,57 +2211,46 @@ static dma_cookie_t pl330_tx_submit(struct dma_async_tx_descriptor *tx)
 
 static inline void _init_desc(struct dma_pl330_desc *desc)
 {
-	desc->req.x = &desc->px;
-	desc->req.token = desc;
 	desc->rqcfg.swap = SWAP_NO;
-	desc->rqcfg.scctl = SCCTRL0;
-	desc->rqcfg.dcctl = DCCTRL0;
-	desc->req.cfg = &desc->rqcfg;
-	desc->req.xfer_cb = dma_pl330_rqcb;
+	desc->rqcfg.scctl = CCTRL0;
+	desc->rqcfg.dcctl = CCTRL0;
 	desc->txd.tx_submit = pl330_tx_submit;
 
 	INIT_LIST_HEAD(&desc->node);
 }
 
 /* Returns the number of descriptors added to the DMAC pool */
-static int add_desc(struct dma_pl330_dmac *pdmac, gfp_t flg, int count)
+static int add_desc(struct pl330_dmac *pl330, gfp_t flg, int count)
 {
 	struct dma_pl330_desc *desc;
 	unsigned long flags;
 	int i;
 
-	if (!pdmac)
-		return 0;
-
 	desc = kcalloc(count, sizeof(*desc), flg);
 	if (!desc)
 		return 0;
 
-	spin_lock_irqsave(&pdmac->pool_lock, flags);
+	spin_lock_irqsave(&pl330->pool_lock, flags);
 
 	for (i = 0; i < count; i++) {
 		_init_desc(&desc[i]);
-		list_add_tail(&desc[i].node, &pdmac->desc_pool);
+		list_add_tail(&desc[i].node, &pl330->desc_pool);
 	}
 
-	spin_unlock_irqrestore(&pdmac->pool_lock, flags);
+	spin_unlock_irqrestore(&pl330->pool_lock, flags);
 
 	return count;
 }
 
-static struct dma_pl330_desc *
-pluck_desc(struct dma_pl330_dmac *pdmac)
+static struct dma_pl330_desc *pluck_desc(struct pl330_dmac *pl330)
 {
 	struct dma_pl330_desc *desc = NULL;
 	unsigned long flags;
 
-	if (!pdmac)
-		return NULL;
-
-	spin_lock_irqsave(&pdmac->pool_lock, flags);
+	spin_lock_irqsave(&pl330->pool_lock, flags);
 
-	if (!list_empty(&pdmac->desc_pool)) {
-		desc = list_entry(pdmac->desc_pool.next,
+	if (!list_empty(&pl330->desc_pool)) {
+		desc = list_entry(pl330->desc_pool.next,
 				struct dma_pl330_desc, node);
 
 		list_del_init(&desc->node);
@@ -2548,29 +2259,29 @@ pluck_desc(struct dma_pl330_dmac *pdmac)
 		desc->txd.callback = NULL;
 	}
 
-	spin_unlock_irqrestore(&pdmac->pool_lock, flags);
+	spin_unlock_irqrestore(&pl330->pool_lock, flags);
 
 	return desc;
 }
 
 static struct dma_pl330_desc *pl330_get_desc(struct dma_pl330_chan *pch)
 {
-	struct dma_pl330_dmac *pdmac = pch->dmac;
+	struct pl330_dmac *pl330 = pch->dmac;
 	u8 *peri_id = pch->chan.private;
 	struct dma_pl330_desc *desc;
 
 	/* Pluck one desc from the pool of DMAC */
-	desc = pluck_desc(pdmac);
+	desc = pluck_desc(pl330);
 
 	/* If the DMAC pool is empty, alloc new */
 	if (!desc) {
-		if (!add_desc(pdmac, GFP_ATOMIC, 1))
+		if (!add_desc(pl330, GFP_ATOMIC, 1))
 			return NULL;
 
 		/* Try again */
-		desc = pluck_desc(pdmac);
+		desc = pluck_desc(pl330);
 		if (!desc) {
-			dev_err(pch->dmac->pif.dev,
+			dev_err(pch->dmac->ddma.dev,
 				"%s:%d ALERT!\n", __func__, __LINE__);
 			return NULL;
 		}
@@ -2581,8 +2292,8 @@ static struct dma_pl330_desc *pl330_get_desc(struct dma_pl330_chan *pch)
 	desc->txd.cookie = 0;
 	async_tx_ack(&desc->txd);
 
-	desc->req.peri = peri_id ? pch->chan.chan_id : 0;
-	desc->rqcfg.pcfg = &pch->dmac->pif.pcfg;
+	desc->peri = peri_id ? pch->chan.chan_id : 0;
+	desc->rqcfg.pcfg = &pch->dmac->pcfg;
 
 	dma_async_tx_descriptor_init(&desc->txd, &pch->chan);
 
@@ -2592,7 +2303,6 @@ static struct dma_pl330_desc *pl330_get_desc(struct dma_pl330_chan *pch)
 static inline void fill_px(struct pl330_xfer *px,
 		dma_addr_t dst, dma_addr_t src, size_t len)
 {
-	px->next = NULL;
 	px->bytes = len;
 	px->dst_addr = dst;
 	px->src_addr = src;
@@ -2605,7 +2315,7 @@ __pl330_prep_dma_memcpy(struct dma_pl330_chan *pch, dma_addr_t dst,
 	struct dma_pl330_desc *desc = pl330_get_desc(pch);
 
 	if (!desc) {
-		dev_err(pch->dmac->pif.dev, "%s:%d Unable to fetch desc\n",
+		dev_err(pch->dmac->ddma.dev, "%s:%d Unable to fetch desc\n",
 			__func__, __LINE__);
 		return NULL;
 	}
@@ -2629,11 +2339,11 @@ __pl330_prep_dma_memcpy(struct dma_pl330_chan *pch, dma_addr_t dst,
 static inline int get_burst_len(struct dma_pl330_desc *desc, size_t len)
 {
 	struct dma_pl330_chan *pch = desc->pchan;
-	struct pl330_info *pi = &pch->dmac->pif;
+	struct pl330_dmac *pl330 = pch->dmac;
 	int burst_len;
 
-	burst_len = pi->pcfg.data_bus_width / 8;
-	burst_len *= pi->pcfg.data_buf_dep;
+	burst_len = pl330->pcfg.data_bus_width / 8;
+	burst_len *= pl330->pcfg.data_buf_dep;
 	burst_len >>= desc->rqcfg.brst_size;
 
 	/* src/dst_burst_len can't be more than 16 */
@@ -2652,11 +2362,11 @@ static inline int get_burst_len(struct dma_pl330_desc *desc, size_t len)
 static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t len,
 		size_t period_len, enum dma_transfer_direction direction,
-		unsigned long flags, void *context)
+		unsigned long flags)
 {
 	struct dma_pl330_desc *desc = NULL, *first = NULL;
 	struct dma_pl330_chan *pch = to_pchan(chan);
-	struct dma_pl330_dmac *pdmac = pch->dmac;
+	struct pl330_dmac *pl330 = pch->dmac;
 	unsigned int i;
 	dma_addr_t dst;
 	dma_addr_t src;
@@ -2665,7 +2375,7 @@ static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic(
 		return NULL;
 
 	if (!is_slave_direction(direction)) {
-		dev_err(pch->dmac->pif.dev, "%s:%d Invalid dma direction\n",
+		dev_err(pch->dmac->ddma.dev, "%s:%d Invalid dma direction\n",
 		__func__, __LINE__);
 		return NULL;
 	}
@@ -2673,23 +2383,23 @@ static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic(
 	for (i = 0; i < len / period_len; i++) {
 		desc = pl330_get_desc(pch);
 		if (!desc) {
-			dev_err(pch->dmac->pif.dev, "%s:%d Unable to fetch desc\n",
+			dev_err(pch->dmac->ddma.dev, "%s:%d Unable to fetch desc\n",
 				__func__, __LINE__);
 
 			if (!first)
 				return NULL;
 
-			spin_lock_irqsave(&pdmac->pool_lock, flags);
+			spin_lock_irqsave(&pl330->pool_lock, flags);
 
 			while (!list_empty(&first->node)) {
 				desc = list_entry(first->node.next,
 						struct dma_pl330_desc, node);
-				list_move_tail(&desc->node, &pdmac->desc_pool);
+				list_move_tail(&desc->node, &pl330->desc_pool);
 			}
 
-			list_move_tail(&first->node, &pdmac->desc_pool);
+			list_move_tail(&first->node, &pl330->desc_pool);
 
-			spin_unlock_irqrestore(&pdmac->pool_lock, flags);
+			spin_unlock_irqrestore(&pl330->pool_lock, flags);
 
 			return NULL;
 		}
@@ -2698,14 +2408,12 @@ static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic(
 		case DMA_MEM_TO_DEV:
 			desc->rqcfg.src_inc = 1;
 			desc->rqcfg.dst_inc = 0;
-			desc->req.rqtype = MEMTODEV;
 			src = dma_addr;
 			dst = pch->fifo_addr;
 			break;
 		case DMA_DEV_TO_MEM:
 			desc->rqcfg.src_inc = 0;
 			desc->rqcfg.dst_inc = 1;
-			desc->req.rqtype = DEVTOMEM;
 			src = pch->fifo_addr;
 			dst = dma_addr;
 			break;
@@ -2713,6 +2421,7 @@ static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic(
 			break;
 		}
 
+		desc->rqtype = direction;
 		desc->rqcfg.brst_size = pch->burst_sz;
 		desc->rqcfg.brst_len = 1;
 		fill_px(&desc->px, dst, src, period_len);
@@ -2740,24 +2449,22 @@ pl330_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dst,
 {
 	struct dma_pl330_desc *desc;
 	struct dma_pl330_chan *pch = to_pchan(chan);
-	struct pl330_info *pi;
+	struct pl330_dmac *pl330 = pch->dmac;
 	int burst;
 
 	if (unlikely(!pch || !len))
 		return NULL;
 
-	pi = &pch->dmac->pif;
-
 	desc = __pl330_prep_dma_memcpy(pch, dst, src, len);
 	if (!desc)
 		return NULL;
 
 	desc->rqcfg.src_inc = 1;
 	desc->rqcfg.dst_inc = 1;
-	desc->req.rqtype = MEMTOMEM;
+	desc->rqtype = DMA_MEM_TO_MEM;
 
 	/* Select max possible burst size */
-	burst = pi->pcfg.data_bus_width / 8;
+	burst = pl330->pcfg.data_bus_width / 8;
 
 	while (burst > 1) {
 		if (!(len % burst))
@@ -2776,7 +2483,7 @@ pl330_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dst,
 	return &desc->txd;
 }
 
-static void __pl330_giveback_desc(struct dma_pl330_dmac *pdmac,
+static void __pl330_giveback_desc(struct pl330_dmac *pl330,
 				  struct dma_pl330_desc *first)
 {
 	unsigned long flags;
@@ -2785,17 +2492,17 @@ static void __pl330_giveback_desc(struct dma_pl330_dmac *pdmac,
 	if (!first)
 		return;
 
-	spin_lock_irqsave(&pdmac->pool_lock, flags);
+	spin_lock_irqsave(&pl330->pool_lock, flags);
 
 	while (!list_empty(&first->node)) {
 		desc = list_entry(first->node.next,
 				struct dma_pl330_desc, node);
-		list_move_tail(&desc->node, &pdmac->desc_pool);
+		list_move_tail(&desc->node, &pl330->desc_pool);
 	}
 
-	list_move_tail(&first->node, &pdmac->desc_pool);
+	list_move_tail(&first->node, &pl330->desc_pool);
 
-	spin_unlock_irqrestore(&pdmac->pool_lock, flags);
+	spin_unlock_irqrestore(&pl330->pool_lock, flags);
 }
 
 static struct dma_async_tx_descriptor *
@@ -2820,12 +2527,12 @@ pl330_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 
 		desc = pl330_get_desc(pch);
 		if (!desc) {
-			struct dma_pl330_dmac *pdmac = pch->dmac;
+			struct pl330_dmac *pl330 = pch->dmac;
 
-			dev_err(pch->dmac->pif.dev,
+			dev_err(pch->dmac->ddma.dev,
 				"%s:%d Unable to fetch desc\n",
 				__func__, __LINE__);
-			__pl330_giveback_desc(pdmac, first);
+			__pl330_giveback_desc(pl330, first);
 
 			return NULL;
 		}
@@ -2838,19 +2545,18 @@ pl330_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		if (direction == DMA_MEM_TO_DEV) {
 			desc->rqcfg.src_inc = 1;
 			desc->rqcfg.dst_inc = 0;
-			desc->req.rqtype = MEMTODEV;
 			fill_px(&desc->px,
 				addr, sg_dma_address(sg), sg_dma_len(sg));
 		} else {
 			desc->rqcfg.src_inc = 0;
 			desc->rqcfg.dst_inc = 1;
-			desc->req.rqtype = DEVTOMEM;
 			fill_px(&desc->px,
 				sg_dma_address(sg), addr, sg_dma_len(sg));
 		}
 
 		desc->rqcfg.brst_size = pch->burst_sz;
 		desc->rqcfg.brst_len = 1;
+		desc->rqtype = direction;
 	}
 
 	/* Return the last desc in the chain */
@@ -2890,9 +2596,9 @@ static int
 pl330_probe(struct amba_device *adev, const struct amba_id *id)
 {
 	struct dma_pl330_platdata *pdat;
-	struct dma_pl330_dmac *pdmac;
+	struct pl330_config *pcfg;
+	struct pl330_dmac *pl330;
 	struct dma_pl330_chan *pch, *_p;
-	struct pl330_info *pi;
 	struct dma_device *pd;
 	struct resource *res;
 	int i, ret, irq;
@@ -2905,30 +2611,27 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 		return ret;
 
 	/* Allocate a new DMAC and its Channels */
-	pdmac = devm_kzalloc(&adev->dev, sizeof(*pdmac), GFP_KERNEL);
-	if (!pdmac) {
+	pl330 = devm_kzalloc(&adev->dev, sizeof(*pl330), GFP_KERNEL);
+	if (!pl330) {
 		dev_err(&adev->dev, "unable to allocate mem\n");
 		return -ENOMEM;
 	}
 
-	pi = &pdmac->pif;
-	pi->dev = &adev->dev;
-	pi->pl330_data = NULL;
-	pi->mcbufsz = pdat ? pdat->mcbuf_sz : 0;
+	pl330->mcbufsz = pdat ? pdat->mcbuf_sz : 0;
 
 	res = &adev->res;
-	pi->base = devm_ioremap_resource(&adev->dev, res);
-	if (IS_ERR(pi->base))
-		return PTR_ERR(pi->base);
+	pl330->base = devm_ioremap_resource(&adev->dev, res);
+	if (IS_ERR(pl330->base))
+		return PTR_ERR(pl330->base);
 
-	amba_set_drvdata(adev, pdmac);
+	amba_set_drvdata(adev, pl330);
 
 	for (i = 0; i < AMBA_NR_IRQS; i++) {
 		irq = adev->irq[i];
 		if (irq) {
 			ret = devm_request_irq(&adev->dev, irq,
 					       pl330_irq_handler, 0,
-					       dev_name(&adev->dev), pi);
+					       dev_name(&adev->dev), pl330);
 			if (ret)
 				return ret;
 		} else {
@@ -2936,38 +2639,40 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 		}
 	}
 
-	pi->pcfg.periph_id = adev->periphid;
-	ret = pl330_add(pi);
+	pcfg = &pl330->pcfg;
+
+	pcfg->periph_id = adev->periphid;
+	ret = pl330_add(pl330);
 	if (ret)
 		return ret;
 
-	INIT_LIST_HEAD(&pdmac->desc_pool);
-	spin_lock_init(&pdmac->pool_lock);
+	INIT_LIST_HEAD(&pl330->desc_pool);
+	spin_lock_init(&pl330->pool_lock);
 
 	/* Create a descriptor pool of default size */
-	if (!add_desc(pdmac, GFP_KERNEL, NR_DEFAULT_DESC))
+	if (!add_desc(pl330, GFP_KERNEL, NR_DEFAULT_DESC))
 		dev_warn(&adev->dev, "unable to allocate desc\n");
 
-	pd = &pdmac->ddma;
+	pd = &pl330->ddma;
 	INIT_LIST_HEAD(&pd->channels);
 
 	/* Initialize channel parameters */
 	if (pdat)
-		num_chan = max_t(int, pdat->nr_valid_peri, pi->pcfg.num_chan);
+		num_chan = max_t(int, pdat->nr_valid_peri, pcfg->num_chan);
 	else
-		num_chan = max_t(int, pi->pcfg.num_peri, pi->pcfg.num_chan);
+		num_chan = max_t(int, pcfg->num_peri, pcfg->num_chan);
 
-	pdmac->num_peripherals = num_chan;
+	pl330->num_peripherals = num_chan;
 
-	pdmac->peripherals = kzalloc(num_chan * sizeof(*pch), GFP_KERNEL);
-	if (!pdmac->peripherals) {
+	pl330->peripherals = kzalloc(num_chan * sizeof(*pch), GFP_KERNEL);
+	if (!pl330->peripherals) {
 		ret = -ENOMEM;
-		dev_err(&adev->dev, "unable to allocate pdmac->peripherals\n");
+		dev_err(&adev->dev, "unable to allocate pl330->peripherals\n");
 		goto probe_err2;
 	}
 
 	for (i = 0; i < num_chan; i++) {
-		pch = &pdmac->peripherals[i];
+		pch = &pl330->peripherals[i];
 		if (!adev->dev.of_node)
 			pch->chan.private = pdat ? &pdat->peri_id[i] : NULL;
 		else
@@ -2977,9 +2682,9 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 		INIT_LIST_HEAD(&pch->work_list);
 		INIT_LIST_HEAD(&pch->completed_list);
 		spin_lock_init(&pch->lock);
-		pch->pl330_chid = NULL;
+		pch->thread = NULL;
 		pch->chan.device = pd;
-		pch->dmac = pdmac;
+		pch->dmac = pl330;
 
 		/* Add the channel to the DMAC list */
 		list_add_tail(&pch->chan.device_node, &pd->channels);
@@ -2990,7 +2695,7 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 		pd->cap_mask = pdat->cap_mask;
 	} else {
 		dma_cap_set(DMA_MEMCPY, pd->cap_mask);
-		if (pi->pcfg.num_peri) {
+		if (pcfg->num_peri) {
 			dma_cap_set(DMA_SLAVE, pd->cap_mask);
 			dma_cap_set(DMA_CYCLIC, pd->cap_mask);
 			dma_cap_set(DMA_PRIVATE, pd->cap_mask);
@@ -3015,14 +2720,14 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 
 	if (adev->dev.of_node) {
 		ret = of_dma_controller_register(adev->dev.of_node,
-					 of_dma_pl330_xlate, pdmac);
+					 of_dma_pl330_xlate, pl330);
 		if (ret) {
 			dev_err(&adev->dev,
 			"unable to register DMA to the generic DT DMA helpers\n");
 		}
 	}
 
-	adev->dev.dma_parms = &pdmac->dma_parms;
+	adev->dev.dma_parms = &pl330->dma_parms;
 
 	/*
 	 * This is the limit for transfers with a buswidth of 1, larger
@@ -3037,14 +2742,13 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 		"Loaded driver for PL330 DMAC-%d\n", adev->periphid);
 	dev_info(&adev->dev,
 		"\tDBUFF-%ux%ubytes Num_Chans-%u Num_Peri-%u Num_Events-%u\n",
-		pi->pcfg.data_buf_dep,
-		pi->pcfg.data_bus_width / 8, pi->pcfg.num_chan,
-		pi->pcfg.num_peri, pi->pcfg.num_events);
+		pcfg->data_buf_dep, pcfg->data_bus_width / 8, pcfg->num_chan,
+		pcfg->num_peri, pcfg->num_events);
 
 	return 0;
 probe_err3:
 	/* Idle the DMAC */
-	list_for_each_entry_safe(pch, _p, &pdmac->ddma.channels,
+	list_for_each_entry_safe(pch, _p, &pl330->ddma.channels,
 			chan.device_node) {
 
 		/* Remove the channel */
@@ -3055,27 +2759,23 @@ probe_err3:
 		pl330_free_chan_resources(&pch->chan);
 	}
 probe_err2:
-	pl330_del(pi);
+	pl330_del(pl330);
 
 	return ret;
 }
 
 static int pl330_remove(struct amba_device *adev)
 {
-	struct dma_pl330_dmac *pdmac = amba_get_drvdata(adev);
+	struct pl330_dmac *pl330 = amba_get_drvdata(adev);
 	struct dma_pl330_chan *pch, *_p;
-	struct pl330_info *pi;
-
-	if (!pdmac)
-		return 0;
 
 	if (adev->dev.of_node)
 		of_dma_controller_free(adev->dev.of_node);
 
-	dma_async_device_unregister(&pdmac->ddma);
+	dma_async_device_unregister(&pl330->ddma);
 
 	/* Idle the DMAC */
-	list_for_each_entry_safe(pch, _p, &pdmac->ddma.channels,
+	list_for_each_entry_safe(pch, _p, &pl330->ddma.channels,
 			chan.device_node) {
 
 		/* Remove the channel */
@@ -3086,9 +2786,7 @@ static int pl330_remove(struct amba_device *adev)
 		pl330_free_chan_resources(&pch->chan);
 	}
 
-	pi = &pdmac->pif;
-
-	pl330_del(pi);
+	pl330_del(pl330);
 
 	return 0;
 }
diff --git a/drivers/dma/qcom_bam_dma.c b/drivers/dma/qcom_bam_dma.c
index 82c923146e49..7a4bbb0f80a5 100644
--- a/drivers/dma/qcom_bam_dma.c
+++ b/drivers/dma/qcom_bam_dma.c
@@ -61,12 +61,17 @@ struct bam_desc_hw {
 #define DESC_FLAG_INT BIT(15)
 #define DESC_FLAG_EOT BIT(14)
 #define DESC_FLAG_EOB BIT(13)
+#define DESC_FLAG_NWD BIT(12)
 
 struct bam_async_desc {
 	struct virt_dma_desc vd;
 
 	u32 num_desc;
 	u32 xfer_len;
+
+	/* transaction flags, EOT|EOB|NWD */
+	u16 flags;
+
 	struct bam_desc_hw *curr_desc;
 
 	enum dma_transfer_direction dir;
@@ -490,6 +495,14 @@ static struct dma_async_tx_descriptor *bam_prep_slave_sg(struct dma_chan *chan,
 	if (!async_desc)
 		goto err_out;
 
+	if (flags & DMA_PREP_FENCE)
+		async_desc->flags |= DESC_FLAG_NWD;
+
+	if (flags & DMA_PREP_INTERRUPT)
+		async_desc->flags |= DESC_FLAG_EOT;
+	else
+		async_desc->flags |= DESC_FLAG_INT;
+
 	async_desc->num_desc = num_alloc;
 	async_desc->curr_desc = async_desc->desc;
 	async_desc->dir = direction;
@@ -793,8 +806,11 @@ static void bam_start_dma(struct bam_chan *bchan)
 	else
 		async_desc->xfer_len = async_desc->num_desc;
 
-	/* set INT on last descriptor */
-	desc[async_desc->xfer_len - 1].flags |= DESC_FLAG_INT;
+	/* set any special flags on the last descriptor */
+	if (async_desc->num_desc == async_desc->xfer_len)
+		desc[async_desc->xfer_len - 1].flags = async_desc->flags;
+	else
+		desc[async_desc->xfer_len - 1].flags |= DESC_FLAG_INT;
 
 	if (bchan->tail + async_desc->xfer_len > MAX_DESCRIPTORS) {
 		u32 partial = MAX_DESCRIPTORS - bchan->tail;
diff --git a/drivers/dma/s3c24xx-dma.c b/drivers/dma/s3c24xx-dma.c
index 012520c9fd79..7416572d1e40 100644
--- a/drivers/dma/s3c24xx-dma.c
+++ b/drivers/dma/s3c24xx-dma.c
@@ -889,8 +889,7 @@ static struct dma_async_tx_descriptor *s3c24xx_dma_prep_memcpy(
 
 static struct dma_async_tx_descriptor *s3c24xx_dma_prep_dma_cyclic(
 	struct dma_chan *chan, dma_addr_t addr, size_t size, size_t period,
-	enum dma_transfer_direction direction, unsigned long flags,
-	void *context)
+	enum dma_transfer_direction direction, unsigned long flags)
 {
 	struct s3c24xx_dma_chan *s3cchan = to_s3c24xx_dma_chan(chan);
 	struct s3c24xx_dma_engine *s3cdma = s3cchan->host;
diff --git a/drivers/dma/sa11x0-dma.c b/drivers/dma/sa11x0-dma.c
index 5ebdfbc1051e..4b0ef043729a 100644
--- a/drivers/dma/sa11x0-dma.c
+++ b/drivers/dma/sa11x0-dma.c
@@ -612,7 +612,7 @@ static struct dma_async_tx_descriptor *sa11x0_dma_prep_slave_sg(
 
 static struct dma_async_tx_descriptor *sa11x0_dma_prep_dma_cyclic(
 	struct dma_chan *chan, dma_addr_t addr, size_t size, size_t period,
-	enum dma_transfer_direction dir, unsigned long flags, void *context)
+	enum dma_transfer_direction dir, unsigned long flags)
 {
 	struct sa11x0_dma_chan *c = to_sa11x0_dma_chan(chan);
 	struct sa11x0_dma_desc *txd;
diff --git a/drivers/dma/sh/Kconfig b/drivers/dma/sh/Kconfig
index 0f719816c91b..0349125a2e20 100644
--- a/drivers/dma/sh/Kconfig
+++ b/drivers/dma/sh/Kconfig
@@ -2,21 +2,39 @@
 # DMA engine configuration for sh
 #
 
+#
+# DMA Engine Helpers
+#
+
 config SH_DMAE_BASE
 	bool "Renesas SuperH DMA Engine support"
-	depends on (SUPERH && SH_DMA) || ARCH_SHMOBILE || COMPILE_TEST
+	depends on SUPERH || ARCH_SHMOBILE || COMPILE_TEST
+	depends on !SUPERH || SH_DMA
 	depends on !SH_DMA_API
 	default y
 	select DMA_ENGINE
 	help
 	  Enable support for the Renesas SuperH DMA controllers.
 
+#
+# DMA Controllers
+#
+
 config SH_DMAE
 	tristate "Renesas SuperH DMAC support"
 	depends on SH_DMAE_BASE
 	help
 	  Enable support for the Renesas SuperH DMA controllers.
 
+if SH_DMAE
+
+config SH_DMAE_R8A73A4
+	def_bool y
+	depends on ARCH_R8A73A4
+	depends on OF
+
+endif
+
 config SUDMAC
 	tristate "Renesas SUDMAC support"
 	depends on SH_DMAE_BASE
@@ -34,7 +52,3 @@ config RCAR_AUDMAC_PP
 	depends on SH_DMAE_BASE
 	help
 	  Enable support for the Renesas R-Car Audio DMAC Peripheral Peripheral controllers.
-
-config SHDMA_R8A73A4
-	def_bool y
-	depends on ARCH_R8A73A4 && SH_DMAE != n
diff --git a/drivers/dma/sh/Makefile b/drivers/dma/sh/Makefile
index 1ce88b28cfc6..0a5cfdb76e45 100644
--- a/drivers/dma/sh/Makefile
+++ b/drivers/dma/sh/Makefile
@@ -1,10 +1,18 @@
+#
+# DMA Engine Helpers
+#
+
 obj-$(CONFIG_SH_DMAE_BASE) += shdma-base.o shdma-of.o
-obj-$(CONFIG_SH_DMAE) += shdma.o
+
+#
+# DMA Controllers
+#
+
 shdma-y := shdmac.o
-ifeq ($(CONFIG_OF),y)
-shdma-$(CONFIG_SHDMA_R8A73A4) += shdma-r8a73a4.o
-endif
+shdma-$(CONFIG_SH_DMAE_R8A73A4) += shdma-r8a73a4.o
 shdma-objs := $(shdma-y)
+obj-$(CONFIG_SH_DMAE) += shdma.o
+
 obj-$(CONFIG_SUDMAC) += sudmac.o
 obj-$(CONFIG_RCAR_HPB_DMAE) += rcar-hpbdma.o
 obj-$(CONFIG_RCAR_AUDMAC_PP) += rcar-audmapp.o
diff --git a/drivers/dma/sh/rcar-audmapp.c b/drivers/dma/sh/rcar-audmapp.c
index 2de77289a2e9..dabbf0aba2e9 100644
--- a/drivers/dma/sh/rcar-audmapp.c
+++ b/drivers/dma/sh/rcar-audmapp.c
@@ -22,6 +22,7 @@
 #include <linux/module.h>
 #include <linux/slab.h>
 #include <linux/dmaengine.h>
+#include <linux/of_dma.h>
 #include <linux/platform_data/dma-rcar-audmapp.h>
 #include <linux/platform_device.h>
 #include <linux/shdma-base.h>
@@ -45,8 +46,9 @@
 
 struct audmapp_chan {
 	struct shdma_chan shdma_chan;
-	struct audmapp_slave_config *config;
 	void __iomem *base;
+	dma_addr_t slave_addr;
+	u32 chcr;
 };
 
 struct audmapp_device {
@@ -56,7 +58,16 @@ struct audmapp_device {
 	void __iomem *chan_reg;
 };
 
+struct audmapp_desc {
+	struct shdma_desc shdma_desc;
+	dma_addr_t src;
+	dma_addr_t dst;
+};
+
+#define to_shdma_chan(c) container_of(c, struct shdma_chan, dma_chan)
+
 #define to_chan(chan) container_of(chan, struct audmapp_chan, shdma_chan)
+#define to_desc(sdesc) container_of(sdesc, struct audmapp_desc, shdma_desc)
 #define to_dev(chan) container_of(chan->shdma_chan.dma_chan.device,	\
 				  struct audmapp_device, shdma_dev.dma_dev)
 
@@ -90,70 +101,82 @@ static void audmapp_halt(struct shdma_chan *schan)
 }
 
 static void audmapp_start_xfer(struct shdma_chan *schan,
-			       struct shdma_desc *sdecs)
+			       struct shdma_desc *sdesc)
 {
 	struct audmapp_chan *auchan = to_chan(schan);
 	struct audmapp_device *audev = to_dev(auchan);
-	struct audmapp_slave_config *cfg = auchan->config;
+	struct audmapp_desc *desc = to_desc(sdesc);
 	struct device *dev = audev->dev;
-	u32 chcr = cfg->chcr | PDMACHCR_DE;
+	u32 chcr = auchan->chcr | PDMACHCR_DE;
 
-	dev_dbg(dev, "src/dst/chcr = %pad/%pad/%x\n",
-		&cfg->src, &cfg->dst, cfg->chcr);
+	dev_dbg(dev, "src/dst/chcr = %pad/%pad/%08x\n",
+		&desc->src, &desc->dst, chcr);
 
-	audmapp_write(auchan, cfg->src,	PDMASAR);
-	audmapp_write(auchan, cfg->dst,	PDMADAR);
+	audmapp_write(auchan, desc->src,	PDMASAR);
+	audmapp_write(auchan, desc->dst,	PDMADAR);
 	audmapp_write(auchan, chcr,	PDMACHCR);
 }
 
-static struct audmapp_slave_config *
-audmapp_find_slave(struct audmapp_chan *auchan, int slave_id)
+static void audmapp_get_config(struct audmapp_chan *auchan, int slave_id,
+			      u32 *chcr, dma_addr_t *dst)
 {
 	struct audmapp_device *audev = to_dev(auchan);
 	struct audmapp_pdata *pdata = audev->pdata;
 	struct audmapp_slave_config *cfg;
 	int i;
 
+	*chcr	= 0;
+	*dst	= 0;
+
+	if (!pdata) { /* DT */
+		*chcr = ((u32)slave_id) << 16;
+		auchan->shdma_chan.slave_id = (slave_id) >> 8;
+		return;
+	}
+
+	/* non-DT */
+
 	if (slave_id >= AUDMAPP_SLAVE_NUMBER)
-		return NULL;
+		return;
 
 	for (i = 0, cfg = pdata->slave; i < pdata->slave_num; i++, cfg++)
-		if (cfg->slave_id == slave_id)
-			return cfg;
-
-	return NULL;
+		if (cfg->slave_id == slave_id) {
+			*chcr	= cfg->chcr;
+			*dst	= cfg->dst;
+			break;
+		}
 }
 
 static int audmapp_set_slave(struct shdma_chan *schan, int slave_id,
 			     dma_addr_t slave_addr, bool try)
 {
 	struct audmapp_chan *auchan = to_chan(schan);
-	struct audmapp_slave_config *cfg =
-		audmapp_find_slave(auchan, slave_id);
+	u32 chcr;
+	dma_addr_t dst;
+
+	audmapp_get_config(auchan, slave_id, &chcr, &dst);
 
-	if (!cfg)
-		return -ENODEV;
 	if (try)
 		return 0;
 
-	auchan->config	= cfg;
+	auchan->chcr		= chcr;
+	auchan->slave_addr	= slave_addr ? : dst;
 
 	return 0;
 }
 
 static int audmapp_desc_setup(struct shdma_chan *schan,
-			      struct shdma_desc *sdecs,
+			      struct shdma_desc *sdesc,
 			      dma_addr_t src, dma_addr_t dst, size_t *len)
 {
-	struct audmapp_chan *auchan = to_chan(schan);
-	struct audmapp_slave_config *cfg = auchan->config;
-
-	if (!cfg)
-		return -ENODEV;
+	struct audmapp_desc *desc = to_desc(sdesc);
 
 	if (*len > (size_t)AUDMAPP_LEN_MAX)
 		*len = (size_t)AUDMAPP_LEN_MAX;
 
+	desc->src = src;
+	desc->dst = dst;
+
 	return 0;
 }
 
@@ -164,7 +187,9 @@ static void audmapp_setup_xfer(struct shdma_chan *schan,
 
 static dma_addr_t audmapp_slave_addr(struct shdma_chan *schan)
 {
-	return 0; /* always fixed address */
+	struct audmapp_chan *auchan = to_chan(schan);
+
+	return auchan->slave_addr;
 }
 
 static bool audmapp_channel_busy(struct shdma_chan *schan)
@@ -183,7 +208,7 @@ static bool audmapp_desc_completed(struct shdma_chan *schan,
 
 static struct shdma_desc *audmapp_embedded_desc(void *buf, int i)
 {
-	return &((struct shdma_desc *)buf)[i];
+	return &((struct audmapp_desc *)buf)[i].shdma_desc;
 }
 
 static const struct shdma_ops audmapp_shdma_ops = {
@@ -234,16 +259,39 @@ static void audmapp_chan_remove(struct audmapp_device *audev)
 	dma_dev->chancnt = 0;
 }
 
+static struct dma_chan *audmapp_of_xlate(struct of_phandle_args *dma_spec,
+					 struct of_dma *ofdma)
+{
+	dma_cap_mask_t mask;
+	struct dma_chan *chan;
+	u32 chcr = dma_spec->args[0];
+
+	if (dma_spec->args_count != 1)
+		return NULL;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	chan = dma_request_channel(mask, shdma_chan_filter, NULL);
+	if (chan)
+		to_shdma_chan(chan)->hw_req = chcr;
+
+	return chan;
+}
+
 static int audmapp_probe(struct platform_device *pdev)
 {
 	struct audmapp_pdata *pdata = pdev->dev.platform_data;
+	struct device_node *np = pdev->dev.of_node;
 	struct audmapp_device *audev;
 	struct shdma_dev *sdev;
 	struct dma_device *dma_dev;
 	struct resource *res;
 	int err, i;
 
-	if (!pdata)
+	if (np)
+		of_dma_controller_register(np, audmapp_of_xlate, pdev);
+	else if (!pdata)
 		return -ENODEV;
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
@@ -260,7 +308,7 @@ static int audmapp_probe(struct platform_device *pdev)
 
 	sdev		= &audev->shdma_dev;
 	sdev->ops	= &audmapp_shdma_ops;
-	sdev->desc_size	= sizeof(struct shdma_desc);
+	sdev->desc_size	= sizeof(struct audmapp_desc);
 
 	dma_dev			= &sdev->dma_dev;
 	dma_dev->copy_align	= LOG2_DEFAULT_XFER_SIZE;
@@ -305,12 +353,18 @@ static int audmapp_remove(struct platform_device *pdev)
 	return 0;
 }
 
+static const struct of_device_id audmapp_of_match[] = {
+	{ .compatible = "renesas,rcar-audmapp", },
+	{},
+};
+
 static struct platform_driver audmapp_driver = {
 	.probe		= audmapp_probe,
 	.remove		= audmapp_remove,
 	.driver		= {
 		.owner	= THIS_MODULE,
 		.name	= "rcar-audmapp-engine",
+		.of_match_table = audmapp_of_match,
 	},
 };
 module_platform_driver(audmapp_driver);
diff --git a/drivers/dma/sh/shdma-arm.h b/drivers/dma/sh/shdma-arm.h
index a2b8258426c9..a1b0ef45d6a2 100644
--- a/drivers/dma/sh/shdma-arm.h
+++ b/drivers/dma/sh/shdma-arm.h
@@ -45,7 +45,7 @@ enum {
 	((((i) & TS_LOW_BIT) << TS_LOW_SHIFT) |\
 	 (((i) & TS_HI_BIT)  << TS_HI_SHIFT))
 
-#define CHCR_TX(xmit_sz) (DM_FIX | SM_INC | 0x800 | TS_INDEX2VAL((xmit_sz)))
-#define CHCR_RX(xmit_sz) (DM_INC | SM_FIX | 0x800 | TS_INDEX2VAL((xmit_sz)))
+#define CHCR_TX(xmit_sz) (DM_FIX | SM_INC | RS_ERS | TS_INDEX2VAL((xmit_sz)))
+#define CHCR_RX(xmit_sz) (DM_INC | SM_FIX | RS_ERS | TS_INDEX2VAL((xmit_sz)))
 
 #endif
diff --git a/drivers/dma/sh/shdma-base.c b/drivers/dma/sh/shdma-base.c
index b35007e21e6b..42d497416196 100644
--- a/drivers/dma/sh/shdma-base.c
+++ b/drivers/dma/sh/shdma-base.c
@@ -206,45 +206,6 @@ static int shdma_setup_slave(struct shdma_chan *schan, int slave_id,
 	return 0;
 }
 
-/*
- * This is the standard shdma filter function to be used as a replacement to the
- * "old" method, using the .private pointer. If for some reason you allocate a
- * channel without slave data, use something like ERR_PTR(-EINVAL) as a filter
- * parameter. If this filter is used, the slave driver, after calling
- * dma_request_channel(), will also have to call dmaengine_slave_config() with
- * .slave_id, .direction, and either .src_addr or .dst_addr set.
- * NOTE: this filter doesn't support multiple DMAC drivers with the DMA_SLAVE
- * capability! If this becomes a requirement, hardware glue drivers, using this
- * services would have to provide their own filters, which first would check
- * the device driver, similar to how other DMAC drivers, e.g., sa11x0-dma.c, do
- * this, and only then, in case of a match, call this common filter.
- * NOTE 2: This filter function is also used in the DT case by shdma_of_xlate().
- * In that case the MID-RID value is used for slave channel filtering and is
- * passed to this function in the "arg" parameter.
- */
-bool shdma_chan_filter(struct dma_chan *chan, void *arg)
-{
-	struct shdma_chan *schan = to_shdma_chan(chan);
-	struct shdma_dev *sdev = to_shdma_dev(schan->dma_chan.device);
-	const struct shdma_ops *ops = sdev->ops;
-	int match = (long)arg;
-	int ret;
-
-	if (match < 0)
-		/* No slave requested - arbitrary channel */
-		return true;
-
-	if (!schan->dev->of_node && match >= slave_num)
-		return false;
-
-	ret = ops->set_slave(schan, match, 0, true);
-	if (ret < 0)
-		return false;
-
-	return true;
-}
-EXPORT_SYMBOL(shdma_chan_filter);
-
 static int shdma_alloc_chan_resources(struct dma_chan *chan)
 {
 	struct shdma_chan *schan = to_shdma_chan(chan);
@@ -295,6 +256,51 @@ esetslave:
 	return ret;
 }
 
+/*
+ * This is the standard shdma filter function to be used as a replacement to the
+ * "old" method, using the .private pointer. If for some reason you allocate a
+ * channel without slave data, use something like ERR_PTR(-EINVAL) as a filter
+ * parameter. If this filter is used, the slave driver, after calling
+ * dma_request_channel(), will also have to call dmaengine_slave_config() with
+ * .slave_id, .direction, and either .src_addr or .dst_addr set.
+ * NOTE: this filter doesn't support multiple DMAC drivers with the DMA_SLAVE
+ * capability! If this becomes a requirement, hardware glue drivers, using this
+ * services would have to provide their own filters, which first would check
+ * the device driver, similar to how other DMAC drivers, e.g., sa11x0-dma.c, do
+ * this, and only then, in case of a match, call this common filter.
+ * NOTE 2: This filter function is also used in the DT case by shdma_of_xlate().
+ * In that case the MID-RID value is used for slave channel filtering and is
+ * passed to this function in the "arg" parameter.
+ */
+bool shdma_chan_filter(struct dma_chan *chan, void *arg)
+{
+	struct shdma_chan *schan;
+	struct shdma_dev *sdev;
+	int match = (long)arg;
+	int ret;
+
+	/* Only support channels handled by this driver. */
+	if (chan->device->device_alloc_chan_resources !=
+	    shdma_alloc_chan_resources)
+		return false;
+
+	if (match < 0)
+		/* No slave requested - arbitrary channel */
+		return true;
+
+	schan = to_shdma_chan(chan);
+	if (!schan->dev->of_node && match >= slave_num)
+		return false;
+
+	sdev = to_shdma_dev(schan->dma_chan.device);
+	ret = sdev->ops->set_slave(schan, match, 0, true);
+	if (ret < 0)
+		return false;
+
+	return true;
+}
+EXPORT_SYMBOL(shdma_chan_filter);
+
 static dma_async_tx_callback __ld_cleanup(struct shdma_chan *schan, bool all)
 {
 	struct shdma_desc *desc, *_desc;
@@ -662,15 +668,16 @@ static struct dma_async_tx_descriptor *shdma_prep_slave_sg(
 static struct dma_async_tx_descriptor *shdma_prep_dma_cyclic(
 	struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
 	size_t period_len, enum dma_transfer_direction direction,
-	unsigned long flags, void *context)
+	unsigned long flags)
 {
 	struct shdma_chan *schan = to_shdma_chan(chan);
 	struct shdma_dev *sdev = to_shdma_dev(schan->dma_chan.device);
+	struct dma_async_tx_descriptor *desc;
 	const struct shdma_ops *ops = sdev->ops;
 	unsigned int sg_len = buf_len / period_len;
 	int slave_id = schan->slave_id;
 	dma_addr_t slave_addr;
-	struct scatterlist sgl[SHDMA_MAX_SG_LEN];
+	struct scatterlist *sgl;
 	int i;
 
 	if (!chan)
@@ -694,7 +701,16 @@ static struct dma_async_tx_descriptor *shdma_prep_dma_cyclic(
 
 	slave_addr = ops->slave_addr(schan);
 
+	/*
+	 * Allocate the sg list dynamically as it would consumer too much stack
+	 * space.
+	 */
+	sgl = kcalloc(sg_len, sizeof(*sgl), GFP_KERNEL);
+	if (!sgl)
+		return NULL;
+
 	sg_init_table(sgl, sg_len);
+
 	for (i = 0; i < sg_len; i++) {
 		dma_addr_t src = buf_addr + (period_len * i);
 
@@ -704,8 +720,11 @@ static struct dma_async_tx_descriptor *shdma_prep_dma_cyclic(
 		sg_dma_len(&sgl[i]) = period_len;
 	}
 
-	return shdma_prep_sg(schan, sgl, sg_len, &slave_addr,
+	desc = shdma_prep_sg(schan, sgl, sg_len, &slave_addr,
 			     direction, flags, true);
+
+	kfree(sgl);
+	return desc;
 }
 
 static int shdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
diff --git a/drivers/dma/sh/shdma.h b/drivers/dma/sh/shdma.h
index 758a57b51875..2c0a969adc9f 100644
--- a/drivers/dma/sh/shdma.h
+++ b/drivers/dma/sh/shdma.h
@@ -62,7 +62,7 @@ struct sh_dmae_desc {
 #define to_sh_dev(chan) container_of(chan->shdma_chan.dma_chan.device,\
 				     struct sh_dmae_device, shdma_dev.dma_dev)
 
-#ifdef CONFIG_SHDMA_R8A73A4
+#ifdef CONFIG_SH_DMAE_R8A73A4
 extern const struct sh_dmae_pdata r8a73a4_dma_pdata;
 #define r8a73a4_shdma_devid (&r8a73a4_dma_pdata)
 #else
diff --git a/drivers/dma/sh/shdmac.c b/drivers/dma/sh/shdmac.c
index 146d5df926db..58eb85770eba 100644
--- a/drivers/dma/sh/shdmac.c
+++ b/drivers/dma/sh/shdmac.c
@@ -38,12 +38,12 @@
 #include "../dmaengine.h"
 #include "shdma.h"
 
-/* DMA register */
-#define SAR	0x00
-#define DAR	0x04
-#define TCR	0x08
-#define CHCR	0x0C
-#define DMAOR	0x40
+/* DMA registers */
+#define SAR	0x00	/* Source Address Register */
+#define DAR	0x04	/* Destination Address Register */
+#define TCR	0x08	/* Transfer Count Register */
+#define CHCR	0x0C	/* Channel Control Register */
+#define DMAOR	0x40	/* DMA Operation Register */
 
 #define TEND	0x18 /* USB-DMAC */
 
@@ -239,9 +239,8 @@ static void dmae_init(struct sh_dmae_chan *sh_chan)
 {
 	/*
 	 * Default configuration for dual address memory-memory transfer.
-	 * 0x400 represents auto-request.
 	 */
-	u32 chcr = DM_INC | SM_INC | 0x400 | log2size_to_chcr(sh_chan,
+	u32 chcr = DM_INC | SM_INC | RS_AUTO | log2size_to_chcr(sh_chan,
 						   LOG2_DEFAULT_XFER_SIZE);
 	sh_chan->xmit_shift = calc_xmit_shift(sh_chan, chcr);
 	chcr_write(sh_chan, chcr);
diff --git a/drivers/dma/sirf-dma.c b/drivers/dma/sirf-dma.c
index 03f7820fa333..aac03ab10c54 100644
--- a/drivers/dma/sirf-dma.c
+++ b/drivers/dma/sirf-dma.c
@@ -580,7 +580,7 @@ err_dir:
 static struct dma_async_tx_descriptor *
 sirfsoc_dma_prep_cyclic(struct dma_chan *chan, dma_addr_t addr,
 	size_t buf_len, size_t period_len,
-	enum dma_transfer_direction direction, unsigned long flags, void *context)
+	enum dma_transfer_direction direction, unsigned long flags)
 {
 	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
 	struct sirfsoc_dma_desc *sdesc = NULL;
diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c
index c7984459ede7..5fe59335e247 100644
--- a/drivers/dma/ste_dma40.c
+++ b/drivers/dma/ste_dma40.c
@@ -2531,8 +2531,7 @@ d40_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 static struct dma_async_tx_descriptor *
 dma40_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t dma_addr,
 		     size_t buf_len, size_t period_len,
-		     enum dma_transfer_direction direction, unsigned long flags,
-		     void *context)
+		     enum dma_transfer_direction direction, unsigned long flags)
 {
 	unsigned int periods = buf_len / period_len;
 	struct dma_async_tx_descriptor *txd;
diff --git a/drivers/dma/sun6i-dma.c b/drivers/dma/sun6i-dma.c
new file mode 100644
index 000000000000..1f92a56fd2b6
--- /dev/null
+++ b/drivers/dma/sun6i-dma.c
@@ -0,0 +1,1053 @@
+/*
+ * Copyright (C) 2013-2014 Allwinner Tech Co., Ltd
+ * Author: Sugar <shuge@allwinnertech.com>
+ *
+ * Copyright (C) 2014 Maxime Ripard
+ * Maxime Ripard <maxime.ripard@free-electrons.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/dmaengine.h>
+#include <linux/dmapool.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/of_dma.h>
+#include <linux/platform_device.h>
+#include <linux/reset.h>
+#include <linux/slab.h>
+#include <linux/types.h>
+
+#include "virt-dma.h"
+
+/*
+ * There's 16 physical channels that can work in parallel.
+ *
+ * However we have 30 different endpoints for our requests.
+ *
+ * Since the channels are able to handle only an unidirectional
+ * transfer, we need to allocate more virtual channels so that
+ * everyone can grab one channel.
+ *
+ * Some devices can't work in both direction (mostly because it
+ * wouldn't make sense), so we have a bit fewer virtual channels than
+ * 2 channels per endpoints.
+ */
+
+#define NR_MAX_CHANNELS		16
+#define NR_MAX_REQUESTS		30
+#define NR_MAX_VCHANS		53
+
+/*
+ * Common registers
+ */
+#define DMA_IRQ_EN(x)		((x) * 0x04)
+#define DMA_IRQ_HALF			BIT(0)
+#define DMA_IRQ_PKG			BIT(1)
+#define DMA_IRQ_QUEUE			BIT(2)
+
+#define DMA_IRQ_CHAN_NR			8
+#define DMA_IRQ_CHAN_WIDTH		4
+
+
+#define DMA_IRQ_STAT(x)		((x) * 0x04 + 0x10)
+
+#define DMA_STAT		0x30
+
+/*
+ * Channels specific registers
+ */
+#define DMA_CHAN_ENABLE		0x00
+#define DMA_CHAN_ENABLE_START		BIT(0)
+#define DMA_CHAN_ENABLE_STOP		0
+
+#define DMA_CHAN_PAUSE		0x04
+#define DMA_CHAN_PAUSE_PAUSE		BIT(1)
+#define DMA_CHAN_PAUSE_RESUME		0
+
+#define DMA_CHAN_LLI_ADDR	0x08
+
+#define DMA_CHAN_CUR_CFG	0x0c
+#define DMA_CHAN_CFG_SRC_DRQ(x)		((x) & 0x1f)
+#define DMA_CHAN_CFG_SRC_IO_MODE	BIT(5)
+#define DMA_CHAN_CFG_SRC_LINEAR_MODE	(0 << 5)
+#define DMA_CHAN_CFG_SRC_BURST(x)	(((x) & 0x3) << 7)
+#define DMA_CHAN_CFG_SRC_WIDTH(x)	(((x) & 0x3) << 9)
+
+#define DMA_CHAN_CFG_DST_DRQ(x)		(DMA_CHAN_CFG_SRC_DRQ(x) << 16)
+#define DMA_CHAN_CFG_DST_IO_MODE	(DMA_CHAN_CFG_SRC_IO_MODE << 16)
+#define DMA_CHAN_CFG_DST_LINEAR_MODE	(DMA_CHAN_CFG_SRC_LINEAR_MODE << 16)
+#define DMA_CHAN_CFG_DST_BURST(x)	(DMA_CHAN_CFG_SRC_BURST(x) << 16)
+#define DMA_CHAN_CFG_DST_WIDTH(x)	(DMA_CHAN_CFG_SRC_WIDTH(x) << 16)
+
+#define DMA_CHAN_CUR_SRC	0x10
+
+#define DMA_CHAN_CUR_DST	0x14
+
+#define DMA_CHAN_CUR_CNT	0x18
+
+#define DMA_CHAN_CUR_PARA	0x1c
+
+
+/*
+ * Various hardware related defines
+ */
+#define LLI_LAST_ITEM	0xfffff800
+#define NORMAL_WAIT	8
+#define DRQ_SDRAM	1
+
+/*
+ * Hardware representation of the LLI
+ *
+ * The hardware will be fed the physical address of this structure,
+ * and read its content in order to start the transfer.
+ */
+struct sun6i_dma_lli {
+	u32			cfg;
+	u32			src;
+	u32			dst;
+	u32			len;
+	u32			para;
+	u32			p_lli_next;
+
+	/*
+	 * This field is not used by the DMA controller, but will be
+	 * used by the CPU to go through the list (mostly for dumping
+	 * or freeing it).
+	 */
+	struct sun6i_dma_lli	*v_lli_next;
+};
+
+
+struct sun6i_desc {
+	struct virt_dma_desc	vd;
+	dma_addr_t		p_lli;
+	struct sun6i_dma_lli	*v_lli;
+};
+
+struct sun6i_pchan {
+	u32			idx;
+	void __iomem		*base;
+	struct sun6i_vchan	*vchan;
+	struct sun6i_desc	*desc;
+	struct sun6i_desc	*done;
+};
+
+struct sun6i_vchan {
+	struct virt_dma_chan	vc;
+	struct list_head	node;
+	struct dma_slave_config	cfg;
+	struct sun6i_pchan	*phy;
+	u8			port;
+};
+
+struct sun6i_dma_dev {
+	struct dma_device	slave;
+	void __iomem		*base;
+	struct clk		*clk;
+	int			irq;
+	spinlock_t		lock;
+	struct reset_control	*rstc;
+	struct tasklet_struct	task;
+	atomic_t		tasklet_shutdown;
+	struct list_head	pending;
+	struct dma_pool		*pool;
+	struct sun6i_pchan	*pchans;
+	struct sun6i_vchan	*vchans;
+};
+
+static struct device *chan2dev(struct dma_chan *chan)
+{
+	return &chan->dev->device;
+}
+
+static inline struct sun6i_dma_dev *to_sun6i_dma_dev(struct dma_device *d)
+{
+	return container_of(d, struct sun6i_dma_dev, slave);
+}
+
+static inline struct sun6i_vchan *to_sun6i_vchan(struct dma_chan *chan)
+{
+	return container_of(chan, struct sun6i_vchan, vc.chan);
+}
+
+static inline struct sun6i_desc *
+to_sun6i_desc(struct dma_async_tx_descriptor *tx)
+{
+	return container_of(tx, struct sun6i_desc, vd.tx);
+}
+
+static inline void sun6i_dma_dump_com_regs(struct sun6i_dma_dev *sdev)
+{
+	dev_dbg(sdev->slave.dev, "Common register:\n"
+		"\tmask0(%04x): 0x%08x\n"
+		"\tmask1(%04x): 0x%08x\n"
+		"\tpend0(%04x): 0x%08x\n"
+		"\tpend1(%04x): 0x%08x\n"
+		"\tstats(%04x): 0x%08x\n",
+		DMA_IRQ_EN(0), readl(sdev->base + DMA_IRQ_EN(0)),
+		DMA_IRQ_EN(1), readl(sdev->base + DMA_IRQ_EN(1)),
+		DMA_IRQ_STAT(0), readl(sdev->base + DMA_IRQ_STAT(0)),
+		DMA_IRQ_STAT(1), readl(sdev->base + DMA_IRQ_STAT(1)),
+		DMA_STAT, readl(sdev->base + DMA_STAT));
+}
+
+static inline void sun6i_dma_dump_chan_regs(struct sun6i_dma_dev *sdev,
+					    struct sun6i_pchan *pchan)
+{
+	phys_addr_t reg = virt_to_phys(pchan->base);
+
+	dev_dbg(sdev->slave.dev, "Chan %d reg: %pa\n"
+		"\t___en(%04x): \t0x%08x\n"
+		"\tpause(%04x): \t0x%08x\n"
+		"\tstart(%04x): \t0x%08x\n"
+		"\t__cfg(%04x): \t0x%08x\n"
+		"\t__src(%04x): \t0x%08x\n"
+		"\t__dst(%04x): \t0x%08x\n"
+		"\tcount(%04x): \t0x%08x\n"
+		"\t_para(%04x): \t0x%08x\n\n",
+		pchan->idx, &reg,
+		DMA_CHAN_ENABLE,
+		readl(pchan->base + DMA_CHAN_ENABLE),
+		DMA_CHAN_PAUSE,
+		readl(pchan->base + DMA_CHAN_PAUSE),
+		DMA_CHAN_LLI_ADDR,
+		readl(pchan->base + DMA_CHAN_LLI_ADDR),
+		DMA_CHAN_CUR_CFG,
+		readl(pchan->base + DMA_CHAN_CUR_CFG),
+		DMA_CHAN_CUR_SRC,
+		readl(pchan->base + DMA_CHAN_CUR_SRC),
+		DMA_CHAN_CUR_DST,
+		readl(pchan->base + DMA_CHAN_CUR_DST),
+		DMA_CHAN_CUR_CNT,
+		readl(pchan->base + DMA_CHAN_CUR_CNT),
+		DMA_CHAN_CUR_PARA,
+		readl(pchan->base + DMA_CHAN_CUR_PARA));
+}
+
+static inline int convert_burst(u32 maxburst, u8 *burst)
+{
+	switch (maxburst) {
+	case 1:
+		*burst = 0;
+		break;
+	case 8:
+		*burst = 2;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static inline int convert_buswidth(enum dma_slave_buswidth addr_width, u8 *width)
+{
+	if ((addr_width < DMA_SLAVE_BUSWIDTH_1_BYTE) ||
+	    (addr_width > DMA_SLAVE_BUSWIDTH_4_BYTES))
+		return -EINVAL;
+
+	*width = addr_width >> 1;
+	return 0;
+}
+
+static void *sun6i_dma_lli_add(struct sun6i_dma_lli *prev,
+			       struct sun6i_dma_lli *next,
+			       dma_addr_t next_phy,
+			       struct sun6i_desc *txd)
+{
+	if ((!prev && !txd) || !next)
+		return NULL;
+
+	if (!prev) {
+		txd->p_lli = next_phy;
+		txd->v_lli = next;
+	} else {
+		prev->p_lli_next = next_phy;
+		prev->v_lli_next = next;
+	}
+
+	next->p_lli_next = LLI_LAST_ITEM;
+	next->v_lli_next = NULL;
+
+	return next;
+}
+
+static inline int sun6i_dma_cfg_lli(struct sun6i_dma_lli *lli,
+				    dma_addr_t src,
+				    dma_addr_t dst, u32 len,
+				    struct dma_slave_config *config)
+{
+	u8 src_width, dst_width, src_burst, dst_burst;
+	int ret;
+
+	if (!config)
+		return -EINVAL;
+
+	ret = convert_burst(config->src_maxburst, &src_burst);
+	if (ret)
+		return ret;
+
+	ret = convert_burst(config->dst_maxburst, &dst_burst);
+	if (ret)
+		return ret;
+
+	ret = convert_buswidth(config->src_addr_width, &src_width);
+	if (ret)
+		return ret;
+
+	ret = convert_buswidth(config->dst_addr_width, &dst_width);
+	if (ret)
+		return ret;
+
+	lli->cfg = DMA_CHAN_CFG_SRC_BURST(src_burst) |
+		DMA_CHAN_CFG_SRC_WIDTH(src_width) |
+		DMA_CHAN_CFG_DST_BURST(dst_burst) |
+		DMA_CHAN_CFG_DST_WIDTH(dst_width);
+
+	lli->src = src;
+	lli->dst = dst;
+	lli->len = len;
+	lli->para = NORMAL_WAIT;
+
+	return 0;
+}
+
+static inline void sun6i_dma_dump_lli(struct sun6i_vchan *vchan,
+				      struct sun6i_dma_lli *lli)
+{
+	phys_addr_t p_lli = virt_to_phys(lli);
+
+	dev_dbg(chan2dev(&vchan->vc.chan),
+		"\n\tdesc:   p - %pa v - 0x%p\n"
+		"\t\tc - 0x%08x s - 0x%08x d - 0x%08x\n"
+		"\t\tl - 0x%08x p - 0x%08x n - 0x%08x\n",
+		&p_lli, lli,
+		lli->cfg, lli->src, lli->dst,
+		lli->len, lli->para, lli->p_lli_next);
+}
+
+static void sun6i_dma_free_desc(struct virt_dma_desc *vd)
+{
+	struct sun6i_desc *txd = to_sun6i_desc(&vd->tx);
+	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(vd->tx.chan->device);
+	struct sun6i_dma_lli *v_lli, *v_next;
+	dma_addr_t p_lli, p_next;
+
+	if (unlikely(!txd))
+		return;
+
+	p_lli = txd->p_lli;
+	v_lli = txd->v_lli;
+
+	while (v_lli) {
+		v_next = v_lli->v_lli_next;
+		p_next = v_lli->p_lli_next;
+
+		dma_pool_free(sdev->pool, v_lli, p_lli);
+
+		v_lli = v_next;
+		p_lli = p_next;
+	}
+
+	kfree(txd);
+}
+
+static int sun6i_dma_terminate_all(struct sun6i_vchan *vchan)
+{
+	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(vchan->vc.chan.device);
+	struct sun6i_pchan *pchan = vchan->phy;
+	unsigned long flags;
+	LIST_HEAD(head);
+
+	spin_lock(&sdev->lock);
+	list_del_init(&vchan->node);
+	spin_unlock(&sdev->lock);
+
+	spin_lock_irqsave(&vchan->vc.lock, flags);
+
+	vchan_get_all_descriptors(&vchan->vc, &head);
+
+	if (pchan) {
+		writel(DMA_CHAN_ENABLE_STOP, pchan->base + DMA_CHAN_ENABLE);
+		writel(DMA_CHAN_PAUSE_RESUME, pchan->base + DMA_CHAN_PAUSE);
+
+		vchan->phy = NULL;
+		pchan->vchan = NULL;
+		pchan->desc = NULL;
+		pchan->done = NULL;
+	}
+
+	spin_unlock_irqrestore(&vchan->vc.lock, flags);
+
+	vchan_dma_desc_free_list(&vchan->vc, &head);
+
+	return 0;
+}
+
+static int sun6i_dma_start_desc(struct sun6i_vchan *vchan)
+{
+	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(vchan->vc.chan.device);
+	struct virt_dma_desc *desc = vchan_next_desc(&vchan->vc);
+	struct sun6i_pchan *pchan = vchan->phy;
+	u32 irq_val, irq_reg, irq_offset;
+
+	if (!pchan)
+		return -EAGAIN;
+
+	if (!desc) {
+		pchan->desc = NULL;
+		pchan->done = NULL;
+		return -EAGAIN;
+	}
+
+	list_del(&desc->node);
+
+	pchan->desc = to_sun6i_desc(&desc->tx);
+	pchan->done = NULL;
+
+	sun6i_dma_dump_lli(vchan, pchan->desc->v_lli);
+
+	irq_reg = pchan->idx / DMA_IRQ_CHAN_NR;
+	irq_offset = pchan->idx % DMA_IRQ_CHAN_NR;
+
+	irq_val = readl(sdev->base + DMA_IRQ_EN(irq_offset));
+	irq_val |= DMA_IRQ_QUEUE << (irq_offset * DMA_IRQ_CHAN_WIDTH);
+	writel(irq_val, sdev->base + DMA_IRQ_EN(irq_offset));
+
+	writel(pchan->desc->p_lli, pchan->base + DMA_CHAN_LLI_ADDR);
+	writel(DMA_CHAN_ENABLE_START, pchan->base + DMA_CHAN_ENABLE);
+
+	sun6i_dma_dump_com_regs(sdev);
+	sun6i_dma_dump_chan_regs(sdev, pchan);
+
+	return 0;
+}
+
+static void sun6i_dma_tasklet(unsigned long data)
+{
+	struct sun6i_dma_dev *sdev = (struct sun6i_dma_dev *)data;
+	struct sun6i_vchan *vchan;
+	struct sun6i_pchan *pchan;
+	unsigned int pchan_alloc = 0;
+	unsigned int pchan_idx;
+
+	list_for_each_entry(vchan, &sdev->slave.channels, vc.chan.device_node) {
+		spin_lock_irq(&vchan->vc.lock);
+
+		pchan = vchan->phy;
+
+		if (pchan && pchan->done) {
+			if (sun6i_dma_start_desc(vchan)) {
+				/*
+				 * No current txd associated with this channel
+				 */
+				dev_dbg(sdev->slave.dev, "pchan %u: free\n",
+					pchan->idx);
+
+				/* Mark this channel free */
+				vchan->phy = NULL;
+				pchan->vchan = NULL;
+			}
+		}
+		spin_unlock_irq(&vchan->vc.lock);
+	}
+
+	spin_lock_irq(&sdev->lock);
+	for (pchan_idx = 0; pchan_idx < NR_MAX_CHANNELS; pchan_idx++) {
+		pchan = &sdev->pchans[pchan_idx];
+
+		if (pchan->vchan || list_empty(&sdev->pending))
+			continue;
+
+		vchan = list_first_entry(&sdev->pending,
+					 struct sun6i_vchan, node);
+
+		/* Remove from pending channels */
+		list_del_init(&vchan->node);
+		pchan_alloc |= BIT(pchan_idx);
+
+		/* Mark this channel allocated */
+		pchan->vchan = vchan;
+		vchan->phy = pchan;
+		dev_dbg(sdev->slave.dev, "pchan %u: alloc vchan %p\n",
+			pchan->idx, &vchan->vc);
+	}
+	spin_unlock_irq(&sdev->lock);
+
+	for (pchan_idx = 0; pchan_idx < NR_MAX_CHANNELS; pchan_idx++) {
+		if (!(pchan_alloc & BIT(pchan_idx)))
+			continue;
+
+		pchan = sdev->pchans + pchan_idx;
+		vchan = pchan->vchan;
+		if (vchan) {
+			spin_lock_irq(&vchan->vc.lock);
+			sun6i_dma_start_desc(vchan);
+			spin_unlock_irq(&vchan->vc.lock);
+		}
+	}
+}
+
+static irqreturn_t sun6i_dma_interrupt(int irq, void *dev_id)
+{
+	struct sun6i_dma_dev *sdev = dev_id;
+	struct sun6i_vchan *vchan;
+	struct sun6i_pchan *pchan;
+	int i, j, ret = IRQ_NONE;
+	u32 status;
+
+	for (i = 0; i < 2; i++) {
+		status = readl(sdev->base + DMA_IRQ_STAT(i));
+		if (!status)
+			continue;
+
+		dev_dbg(sdev->slave.dev, "DMA irq status %s: 0x%x\n",
+			i ? "high" : "low", status);
+
+		writel(status, sdev->base + DMA_IRQ_STAT(i));
+
+		for (j = 0; (j < 8) && status; j++) {
+			if (status & DMA_IRQ_QUEUE) {
+				pchan = sdev->pchans + j;
+				vchan = pchan->vchan;
+
+				if (vchan) {
+					spin_lock(&vchan->vc.lock);
+					vchan_cookie_complete(&pchan->desc->vd);
+					pchan->done = pchan->desc;
+					spin_unlock(&vchan->vc.lock);
+				}
+			}
+
+			status = status >> 4;
+		}
+
+		if (!atomic_read(&sdev->tasklet_shutdown))
+			tasklet_schedule(&sdev->task);
+		ret = IRQ_HANDLED;
+	}
+
+	return ret;
+}
+
+static struct dma_async_tx_descriptor *sun6i_dma_prep_dma_memcpy(
+		struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
+		size_t len, unsigned long flags)
+{
+	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(chan->device);
+	struct sun6i_vchan *vchan = to_sun6i_vchan(chan);
+	struct dma_slave_config *sconfig = &vchan->cfg;
+	struct sun6i_dma_lli *v_lli;
+	struct sun6i_desc *txd;
+	dma_addr_t p_lli;
+	int ret;
+
+	dev_dbg(chan2dev(chan),
+		"%s; chan: %d, dest: %pad, src: %pad, len: %zu. flags: 0x%08lx\n",
+		__func__, vchan->vc.chan.chan_id, &dest, &src, len, flags);
+
+	if (!len)
+		return NULL;
+
+	txd = kzalloc(sizeof(*txd), GFP_NOWAIT);
+	if (!txd)
+		return NULL;
+
+	v_lli = dma_pool_alloc(sdev->pool, GFP_NOWAIT, &p_lli);
+	if (!v_lli) {
+		dev_err(sdev->slave.dev, "Failed to alloc lli memory\n");
+		goto err_txd_free;
+	}
+
+	ret = sun6i_dma_cfg_lli(v_lli, src, dest, len, sconfig);
+	if (ret)
+		goto err_dma_free;
+
+	v_lli->cfg |= DMA_CHAN_CFG_SRC_DRQ(DRQ_SDRAM) |
+		DMA_CHAN_CFG_DST_DRQ(DRQ_SDRAM) |
+		DMA_CHAN_CFG_DST_LINEAR_MODE |
+		DMA_CHAN_CFG_SRC_LINEAR_MODE;
+
+	sun6i_dma_lli_add(NULL, v_lli, p_lli, txd);
+
+	sun6i_dma_dump_lli(vchan, v_lli);
+
+	return vchan_tx_prep(&vchan->vc, &txd->vd, flags);
+
+err_dma_free:
+	dma_pool_free(sdev->pool, v_lli, p_lli);
+err_txd_free:
+	kfree(txd);
+	return NULL;
+}
+
+static struct dma_async_tx_descriptor *sun6i_dma_prep_slave_sg(
+		struct dma_chan *chan, struct scatterlist *sgl,
+		unsigned int sg_len, enum dma_transfer_direction dir,
+		unsigned long flags, void *context)
+{
+	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(chan->device);
+	struct sun6i_vchan *vchan = to_sun6i_vchan(chan);
+	struct dma_slave_config *sconfig = &vchan->cfg;
+	struct sun6i_dma_lli *v_lli, *prev = NULL;
+	struct sun6i_desc *txd;
+	struct scatterlist *sg;
+	dma_addr_t p_lli;
+	int i, ret;
+
+	if (!sgl)
+		return NULL;
+
+	if (!is_slave_direction(dir)) {
+		dev_err(chan2dev(chan), "Invalid DMA direction\n");
+		return NULL;
+	}
+
+	txd = kzalloc(sizeof(*txd), GFP_NOWAIT);
+	if (!txd)
+		return NULL;
+
+	for_each_sg(sgl, sg, sg_len, i) {
+		v_lli = dma_pool_alloc(sdev->pool, GFP_NOWAIT, &p_lli);
+		if (!v_lli)
+			goto err_lli_free;
+
+		if (dir == DMA_MEM_TO_DEV) {
+			ret = sun6i_dma_cfg_lli(v_lli, sg_dma_address(sg),
+						sconfig->dst_addr, sg_dma_len(sg),
+						sconfig);
+			if (ret)
+				goto err_cur_lli_free;
+
+			v_lli->cfg |= DMA_CHAN_CFG_DST_IO_MODE |
+				DMA_CHAN_CFG_SRC_LINEAR_MODE |
+				DMA_CHAN_CFG_SRC_DRQ(DRQ_SDRAM) |
+				DMA_CHAN_CFG_DST_DRQ(vchan->port);
+
+			dev_dbg(chan2dev(chan),
+				"%s; chan: %d, dest: %pad, src: %pad, len: %u. flags: 0x%08lx\n",
+				__func__, vchan->vc.chan.chan_id,
+				&sconfig->dst_addr, &sg_dma_address(sg),
+				sg_dma_len(sg), flags);
+
+		} else {
+			ret = sun6i_dma_cfg_lli(v_lli, sconfig->src_addr,
+						sg_dma_address(sg), sg_dma_len(sg),
+						sconfig);
+			if (ret)
+				goto err_cur_lli_free;
+
+			v_lli->cfg |= DMA_CHAN_CFG_DST_LINEAR_MODE |
+				DMA_CHAN_CFG_SRC_IO_MODE |
+				DMA_CHAN_CFG_DST_DRQ(DRQ_SDRAM) |
+				DMA_CHAN_CFG_SRC_DRQ(vchan->port);
+
+			dev_dbg(chan2dev(chan),
+				"%s; chan: %d, dest: %pad, src: %pad, len: %u. flags: 0x%08lx\n",
+				__func__, vchan->vc.chan.chan_id,
+				&sg_dma_address(sg), &sconfig->src_addr,
+				sg_dma_len(sg), flags);
+		}
+
+		prev = sun6i_dma_lli_add(prev, v_lli, p_lli, txd);
+	}
+
+	dev_dbg(chan2dev(chan), "First: %pad\n", &txd->p_lli);
+	for (prev = txd->v_lli; prev; prev = prev->v_lli_next)
+		sun6i_dma_dump_lli(vchan, prev);
+
+	return vchan_tx_prep(&vchan->vc, &txd->vd, flags);
+
+err_cur_lli_free:
+	dma_pool_free(sdev->pool, v_lli, p_lli);
+err_lli_free:
+	for (prev = txd->v_lli; prev; prev = prev->v_lli_next)
+		dma_pool_free(sdev->pool, prev, virt_to_phys(prev));
+	kfree(txd);
+	return NULL;
+}
+
+static int sun6i_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
+		       unsigned long arg)
+{
+	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(chan->device);
+	struct sun6i_vchan *vchan = to_sun6i_vchan(chan);
+	struct sun6i_pchan *pchan = vchan->phy;
+	unsigned long flags;
+	int ret = 0;
+
+	switch (cmd) {
+	case DMA_RESUME:
+		dev_dbg(chan2dev(chan), "vchan %p: resume\n", &vchan->vc);
+
+		spin_lock_irqsave(&vchan->vc.lock, flags);
+
+		if (pchan) {
+			writel(DMA_CHAN_PAUSE_RESUME,
+			       pchan->base + DMA_CHAN_PAUSE);
+		} else if (!list_empty(&vchan->vc.desc_issued)) {
+			spin_lock(&sdev->lock);
+			list_add_tail(&vchan->node, &sdev->pending);
+			spin_unlock(&sdev->lock);
+		}
+
+		spin_unlock_irqrestore(&vchan->vc.lock, flags);
+		break;
+
+	case DMA_PAUSE:
+		dev_dbg(chan2dev(chan), "vchan %p: pause\n", &vchan->vc);
+
+		if (pchan) {
+			writel(DMA_CHAN_PAUSE_PAUSE,
+			       pchan->base + DMA_CHAN_PAUSE);
+		} else {
+			spin_lock(&sdev->lock);
+			list_del_init(&vchan->node);
+			spin_unlock(&sdev->lock);
+		}
+		break;
+
+	case DMA_TERMINATE_ALL:
+		ret = sun6i_dma_terminate_all(vchan);
+		break;
+	case DMA_SLAVE_CONFIG:
+		memcpy(&vchan->cfg, (void *)arg, sizeof(struct dma_slave_config));
+		break;
+	default:
+		ret = -ENXIO;
+		break;
+	}
+	return ret;
+}
+
+static enum dma_status sun6i_dma_tx_status(struct dma_chan *chan,
+					   dma_cookie_t cookie,
+					   struct dma_tx_state *state)
+{
+	struct sun6i_vchan *vchan = to_sun6i_vchan(chan);
+	struct sun6i_pchan *pchan = vchan->phy;
+	struct sun6i_dma_lli *lli;
+	struct virt_dma_desc *vd;
+	struct sun6i_desc *txd;
+	enum dma_status ret;
+	unsigned long flags;
+	size_t bytes = 0;
+
+	ret = dma_cookie_status(chan, cookie, state);
+	if (ret == DMA_COMPLETE)
+		return ret;
+
+	spin_lock_irqsave(&vchan->vc.lock, flags);
+
+	vd = vchan_find_desc(&vchan->vc, cookie);
+	txd = to_sun6i_desc(&vd->tx);
+
+	if (vd) {
+		for (lli = txd->v_lli; lli != NULL; lli = lli->v_lli_next)
+			bytes += lli->len;
+	} else if (!pchan || !pchan->desc) {
+		bytes = 0;
+	} else {
+		bytes = readl(pchan->base + DMA_CHAN_CUR_CNT);
+	}
+
+	spin_unlock_irqrestore(&vchan->vc.lock, flags);
+
+	dma_set_residue(state, bytes);
+
+	return ret;
+}
+
+static void sun6i_dma_issue_pending(struct dma_chan *chan)
+{
+	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(chan->device);
+	struct sun6i_vchan *vchan = to_sun6i_vchan(chan);
+	unsigned long flags;
+
+	spin_lock_irqsave(&vchan->vc.lock, flags);
+
+	if (vchan_issue_pending(&vchan->vc)) {
+		spin_lock(&sdev->lock);
+
+		if (!vchan->phy && list_empty(&vchan->node)) {
+			list_add_tail(&vchan->node, &sdev->pending);
+			tasklet_schedule(&sdev->task);
+			dev_dbg(chan2dev(chan), "vchan %p: issued\n",
+				&vchan->vc);
+		}
+
+		spin_unlock(&sdev->lock);
+	} else {
+		dev_dbg(chan2dev(chan), "vchan %p: nothing to issue\n",
+			&vchan->vc);
+	}
+
+	spin_unlock_irqrestore(&vchan->vc.lock, flags);
+}
+
+static int sun6i_dma_alloc_chan_resources(struct dma_chan *chan)
+{
+	return 0;
+}
+
+static void sun6i_dma_free_chan_resources(struct dma_chan *chan)
+{
+	struct sun6i_dma_dev *sdev = to_sun6i_dma_dev(chan->device);
+	struct sun6i_vchan *vchan = to_sun6i_vchan(chan);
+	unsigned long flags;
+
+	spin_lock_irqsave(&sdev->lock, flags);
+	list_del_init(&vchan->node);
+	spin_unlock_irqrestore(&sdev->lock, flags);
+
+	vchan_free_chan_resources(&vchan->vc);
+}
+
+static struct dma_chan *sun6i_dma_of_xlate(struct of_phandle_args *dma_spec,
+					   struct of_dma *ofdma)
+{
+	struct sun6i_dma_dev *sdev = ofdma->of_dma_data;
+	struct sun6i_vchan *vchan;
+	struct dma_chan *chan;
+	u8 port = dma_spec->args[0];
+
+	if (port > NR_MAX_REQUESTS)
+		return NULL;
+
+	chan = dma_get_any_slave_channel(&sdev->slave);
+	if (!chan)
+		return NULL;
+
+	vchan = to_sun6i_vchan(chan);
+	vchan->port = port;
+
+	return chan;
+}
+
+static inline void sun6i_kill_tasklet(struct sun6i_dma_dev *sdev)
+{
+	/* Disable all interrupts from DMA */
+	writel(0, sdev->base + DMA_IRQ_EN(0));
+	writel(0, sdev->base + DMA_IRQ_EN(1));
+
+	/* Prevent spurious interrupts from scheduling the tasklet */
+	atomic_inc(&sdev->tasklet_shutdown);
+
+	/* Make sure we won't have any further interrupts */
+	devm_free_irq(sdev->slave.dev, sdev->irq, sdev);
+
+	/* Actually prevent the tasklet from being scheduled */
+	tasklet_kill(&sdev->task);
+}
+
+static inline void sun6i_dma_free(struct sun6i_dma_dev *sdev)
+{
+	int i;
+
+	for (i = 0; i < NR_MAX_VCHANS; i++) {
+		struct sun6i_vchan *vchan = &sdev->vchans[i];
+
+		list_del(&vchan->vc.chan.device_node);
+		tasklet_kill(&vchan->vc.task);
+	}
+}
+
+static int sun6i_dma_probe(struct platform_device *pdev)
+{
+	struct sun6i_dma_dev *sdc;
+	struct resource *res;
+	struct clk *mux, *pll6;
+	int ret, i;
+
+	sdc = devm_kzalloc(&pdev->dev, sizeof(*sdc), GFP_KERNEL);
+	if (!sdc)
+		return -ENOMEM;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	sdc->base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(sdc->base))
+		return PTR_ERR(sdc->base);
+
+	sdc->irq = platform_get_irq(pdev, 0);
+	if (sdc->irq < 0) {
+		dev_err(&pdev->dev, "Cannot claim IRQ\n");
+		return sdc->irq;
+	}
+
+	sdc->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(sdc->clk)) {
+		dev_err(&pdev->dev, "No clock specified\n");
+		return PTR_ERR(sdc->clk);
+	}
+
+	mux = clk_get(NULL, "ahb1_mux");
+	if (IS_ERR(mux)) {
+		dev_err(&pdev->dev, "Couldn't get AHB1 Mux\n");
+		return PTR_ERR(mux);
+	}
+
+	pll6 = clk_get(NULL, "pll6");
+	if (IS_ERR(pll6)) {
+		dev_err(&pdev->dev, "Couldn't get PLL6\n");
+		clk_put(mux);
+		return PTR_ERR(pll6);
+	}
+
+	ret = clk_set_parent(mux, pll6);
+	clk_put(pll6);
+	clk_put(mux);
+
+	if (ret) {
+		dev_err(&pdev->dev, "Couldn't reparent AHB1 on PLL6\n");
+		return ret;
+	}
+
+	sdc->rstc = devm_reset_control_get(&pdev->dev, NULL);
+	if (IS_ERR(sdc->rstc)) {
+		dev_err(&pdev->dev, "No reset controller specified\n");
+		return PTR_ERR(sdc->rstc);
+	}
+
+	sdc->pool = dmam_pool_create(dev_name(&pdev->dev), &pdev->dev,
+				     sizeof(struct sun6i_dma_lli), 4, 0);
+	if (!sdc->pool) {
+		dev_err(&pdev->dev, "No memory for descriptors dma pool\n");
+		return -ENOMEM;
+	}
+
+	platform_set_drvdata(pdev, sdc);
+	INIT_LIST_HEAD(&sdc->pending);
+	spin_lock_init(&sdc->lock);
+
+	dma_cap_set(DMA_PRIVATE, sdc->slave.cap_mask);
+	dma_cap_set(DMA_MEMCPY, sdc->slave.cap_mask);
+	dma_cap_set(DMA_SLAVE, sdc->slave.cap_mask);
+
+	INIT_LIST_HEAD(&sdc->slave.channels);
+	sdc->slave.device_alloc_chan_resources	= sun6i_dma_alloc_chan_resources;
+	sdc->slave.device_free_chan_resources	= sun6i_dma_free_chan_resources;
+	sdc->slave.device_tx_status		= sun6i_dma_tx_status;
+	sdc->slave.device_issue_pending		= sun6i_dma_issue_pending;
+	sdc->slave.device_prep_slave_sg		= sun6i_dma_prep_slave_sg;
+	sdc->slave.device_prep_dma_memcpy	= sun6i_dma_prep_dma_memcpy;
+	sdc->slave.device_control		= sun6i_dma_control;
+	sdc->slave.chancnt			= NR_MAX_VCHANS;
+
+	sdc->slave.dev = &pdev->dev;
+
+	sdc->pchans = devm_kcalloc(&pdev->dev, NR_MAX_CHANNELS,
+				   sizeof(struct sun6i_pchan), GFP_KERNEL);
+	if (!sdc->pchans)
+		return -ENOMEM;
+
+	sdc->vchans = devm_kcalloc(&pdev->dev, NR_MAX_VCHANS,
+				   sizeof(struct sun6i_vchan), GFP_KERNEL);
+	if (!sdc->vchans)
+		return -ENOMEM;
+
+	tasklet_init(&sdc->task, sun6i_dma_tasklet, (unsigned long)sdc);
+
+	for (i = 0; i < NR_MAX_CHANNELS; i++) {
+		struct sun6i_pchan *pchan = &sdc->pchans[i];
+
+		pchan->idx = i;
+		pchan->base = sdc->base + 0x100 + i * 0x40;
+	}
+
+	for (i = 0; i < NR_MAX_VCHANS; i++) {
+		struct sun6i_vchan *vchan = &sdc->vchans[i];
+
+		INIT_LIST_HEAD(&vchan->node);
+		vchan->vc.desc_free = sun6i_dma_free_desc;
+		vchan_init(&vchan->vc, &sdc->slave);
+	}
+
+	ret = reset_control_deassert(sdc->rstc);
+	if (ret) {
+		dev_err(&pdev->dev, "Couldn't deassert the device from reset\n");
+		goto err_chan_free;
+	}
+
+	ret = clk_prepare_enable(sdc->clk);
+	if (ret) {
+		dev_err(&pdev->dev, "Couldn't enable the clock\n");
+		goto err_reset_assert;
+	}
+
+	ret = devm_request_irq(&pdev->dev, sdc->irq, sun6i_dma_interrupt, 0,
+			       dev_name(&pdev->dev), sdc);
+	if (ret) {
+		dev_err(&pdev->dev, "Cannot request IRQ\n");
+		goto err_clk_disable;
+	}
+
+	ret = dma_async_device_register(&sdc->slave);
+	if (ret) {
+		dev_warn(&pdev->dev, "Failed to register DMA engine device\n");
+		goto err_irq_disable;
+	}
+
+	ret = of_dma_controller_register(pdev->dev.of_node, sun6i_dma_of_xlate,
+					 sdc);
+	if (ret) {
+		dev_err(&pdev->dev, "of_dma_controller_register failed\n");
+		goto err_dma_unregister;
+	}
+
+	return 0;
+
+err_dma_unregister:
+	dma_async_device_unregister(&sdc->slave);
+err_irq_disable:
+	sun6i_kill_tasklet(sdc);
+err_clk_disable:
+	clk_disable_unprepare(sdc->clk);
+err_reset_assert:
+	reset_control_assert(sdc->rstc);
+err_chan_free:
+	sun6i_dma_free(sdc);
+	return ret;
+}
+
+static int sun6i_dma_remove(struct platform_device *pdev)
+{
+	struct sun6i_dma_dev *sdc = platform_get_drvdata(pdev);
+
+	of_dma_controller_free(pdev->dev.of_node);
+	dma_async_device_unregister(&sdc->slave);
+
+	sun6i_kill_tasklet(sdc);
+
+	clk_disable_unprepare(sdc->clk);
+	reset_control_assert(sdc->rstc);
+
+	sun6i_dma_free(sdc);
+
+	return 0;
+}
+
+static struct of_device_id sun6i_dma_match[] = {
+	{ .compatible = "allwinner,sun6i-a31-dma" },
+	{ /* sentinel */ }
+};
+
+static struct platform_driver sun6i_dma_driver = {
+	.probe		= sun6i_dma_probe,
+	.remove		= sun6i_dma_remove,
+	.driver = {
+		.name		= "sun6i-dma",
+		.of_match_table	= sun6i_dma_match,
+	},
+};
+module_platform_driver(sun6i_dma_driver);
+
+MODULE_DESCRIPTION("Allwinner A31 DMA Controller Driver");
+MODULE_AUTHOR("Sugar <shuge@allwinnertech.com>");
+MODULE_AUTHOR("Maxime Ripard <maxime.ripard@free-electrons.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/dma/tegra20-apb-dma.c b/drivers/dma/tegra20-apb-dma.c
index 03ad64ecaaf0..16efa603ff65 100644
--- a/drivers/dma/tegra20-apb-dma.c
+++ b/drivers/dma/tegra20-apb-dma.c
@@ -1055,7 +1055,7 @@ static struct dma_async_tx_descriptor *tegra_dma_prep_slave_sg(
 static struct dma_async_tx_descriptor *tegra_dma_prep_dma_cyclic(
 	struct dma_chan *dc, dma_addr_t buf_addr, size_t buf_len,
 	size_t period_len, enum dma_transfer_direction direction,
-	unsigned long flags, void *context)
+	unsigned long flags)
 {
 	struct tegra_dma_channel *tdc = to_tegra_dma_chan(dc);
 	struct tegra_dma_desc *dma_desc = NULL;