Merge tag 'dmaengine-4.13-rc1' of git://git.infradead.org/users/vkoul/slave-dma

Pull dmaengine updates from Vinod Koul:

 - removal of AVR32 support in dw driver as AVR32 is gone

 - new driver for Broadcom stream buffer accelerator (SBA) RAID driver

 - add support for Faraday Technology FTDMAC020 in amba-pl08x driver

 - IOMMU support in pl330 driver

 - updates to bunch of drivers

* tag 'dmaengine-4.13-rc1' of git://git.infradead.org/users/vkoul/slave-dma: (36 commits)
  dmaengine: qcom_hidma: correct API violation for submit
  dmaengine: zynqmp_dma: Remove max len check in zynqmp_dma_prep_memcpy
  dmaengine: tegra-apb: Really fix runtime-pm usage
  dmaengine: fsl_raid: make of_device_ids const.
  dmaengine: qcom_hidma: allow ACPI/DT parameters to be overridden
  dmaengine: fsldma: set BWC, DAHTS and SAHTS values correctly
  dmaengine: Kconfig: Simplify the help text for MXS_DMA
  dmaengine: pl330: Delete unused functions
  dmaengine: Replace WARN_TAINT_ONCE() with pr_warn_once()
  dmaengine: Kconfig: Extend the dependency for MXS_DMA
  dmaengine: mxs: Use %zu for printing a size_t variable
  dmaengine: ste_dma40: Cleanup scatterlist layering violations
  dmaengine: imx-dma: cleanup scatterlist layering violations
  dmaengine: use proper name for the R-Car SoC
  dmaengine: imx-sdma: Fix compilation warning.
  dmaengine: imx-sdma: Handle return value of clk_prepare_enable
  dmaengine: pl330: Add IOMMU support to slave tranfers
  dmaengine: DW DMAC: Handle return value of clk_prepare_enable
  dmaengine: pl08x: use GENMASK() to create bitmasks
  dmaengine: pl08x: Add support for Faraday Technology FTDMAC020
  ...

21 files changed, 2824 insertions, 378 deletions

diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index 24e8597b2c3e..fa8f9c07ce73 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -62,8 +62,10 @@ config AMBA_PL08X
 	select DMA_ENGINE
 	select DMA_VIRTUAL_CHANNELS
 	help
-	  Platform has a PL08x DMAC device
-	  which can provide DMA engine support
+	  Say yes if your platform has a PL08x DMAC device which can
+	  provide DMA engine support. This includes the original ARM
+	  PL080 and PL081, Samsungs PL080 derivative and Faraday
+	  Technology's FTDMAC020 PL080 derivative.
 
 config AMCC_PPC440SPE_ADMA
 	tristate "AMCC PPC440SPe ADMA support"
@@ -99,6 +101,21 @@ config AXI_DMAC
 	  controller is often used in Analog Device's reference designs for FPGA
 	  platforms.
 
+config BCM_SBA_RAID
+	tristate "Broadcom SBA RAID engine support"
+	depends on ARM64 || COMPILE_TEST
+	depends on MAILBOX && RAID6_PQ
+	select DMA_ENGINE
+	select DMA_ENGINE_RAID
+	select ASYNC_TX_DISABLE_XOR_VAL_DMA
+	select ASYNC_TX_DISABLE_PQ_VAL_DMA
+	default ARCH_BCM_IPROC
+	help
+	  Enable support for Broadcom SBA RAID Engine. The SBA RAID
+	  engine is available on most of the Broadcom iProc SoCs. It
+	  has the capability to offload memcpy, xor and pq computation
+	  for raid5/6.
+
 config COH901318
 	bool "ST-Ericsson COH901318 DMA support"
 	select DMA_ENGINE
@@ -354,13 +371,12 @@ config MV_XOR_V2
 
 config MXS_DMA
 	bool "MXS DMA support"
-	depends on SOC_IMX23 || SOC_IMX28 || SOC_IMX6Q || SOC_IMX6UL
+	depends on ARCH_MXS || ARCH_MXC || COMPILE_TEST
 	select STMP_DEVICE
 	select DMA_ENGINE
 	help
 	  Support the MXS DMA engine. This engine including APBH-DMA
-	  and APBX-DMA is integrated into Freescale
-	  i.MX23/28/MX6Q/MX6DL/MX6UL chips.
+	  and APBX-DMA is integrated into some Freescale chips.
 
 config MX3_IPU
 	bool "MX3x Image Processing Unit support"
diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile
index 0b723e94d9e6..d12ab2985ed1 100644
--- a/drivers/dma/Makefile
+++ b/drivers/dma/Makefile
@@ -17,6 +17,7 @@ obj-$(CONFIG_AMCC_PPC440SPE_ADMA) += ppc4xx/
 obj-$(CONFIG_AT_HDMAC) += at_hdmac.o
 obj-$(CONFIG_AT_XDMAC) += at_xdmac.o
 obj-$(CONFIG_AXI_DMAC) += dma-axi-dmac.o
+obj-$(CONFIG_BCM_SBA_RAID) += bcm-sba-raid.o
 obj-$(CONFIG_COH901318) += coh901318.o coh901318_lli.o
 obj-$(CONFIG_DMA_BCM2835) += bcm2835-dma.o
 obj-$(CONFIG_DMA_JZ4740) += dma-jz4740.o
diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c
index 6bb8813ca275..13cc95c0474c 100644
--- a/drivers/dma/amba-pl08x.c
+++ b/drivers/dma/amba-pl08x.c
@@ -1,9 +1,10 @@
 /*
  * Copyright (c) 2006 ARM Ltd.
  * Copyright (c) 2010 ST-Ericsson SA
+ * Copyirght (c) 2017 Linaro Ltd.
  *
  * Author: Peter Pearse <peter.pearse@arm.com>
- * Author: Linus Walleij <linus.walleij@stericsson.com>
+ * Author: Linus Walleij <linus.walleij@linaro.org>
  *
  * This program is free software; you can redistribute it and/or modify it
  * under the terms of the GNU General Public License as published by the Free
@@ -110,11 +111,12 @@ struct pl08x_driver_data;
  * @channels: the number of channels available in this variant
  * @signals: the number of request signals available from the hardware
  * @dualmaster: whether this version supports dual AHB masters or not.
- * @nomadik: whether the channels have Nomadik security extension bits
- *	that need to be checked for permission before use and some registers are
- *	missing
- * @pl080s: whether this version is a PL080S, which has separate register and
- *	LLI word for transfer size.
+ * @nomadik: whether this variant is a ST Microelectronics Nomadik, where the
+ *	channels have Nomadik security extension bits that need to be checked
+ *	for permission before use and some registers are missing
+ * @pl080s: whether this variant is a Samsung PL080S, which has separate
+ *	register and LLI word for transfer size.
+ * @ftdmac020: whether this variant is a Faraday Technology FTDMAC020
  * @max_transfer_size: the maximum single element transfer size for this
  *	PL08x variant.
  */
@@ -125,6 +127,7 @@ struct vendor_data {
 	bool dualmaster;
 	bool nomadik;
 	bool pl080s;
+	bool ftdmac020;
 	u32 max_transfer_size;
 };
 
@@ -148,19 +151,34 @@ struct pl08x_bus_data {
  * @id: physical index to this channel
  * @base: memory base address for this physical channel
  * @reg_config: configuration address for this physical channel
+ * @reg_control: control address for this physical channel
+ * @reg_src: transfer source address register
+ * @reg_dst: transfer destination address register
+ * @reg_lli: transfer LLI address register
+ * @reg_busy: if the variant has a special per-channel busy register,
+ * this contains a pointer to it
  * @lock: a lock to use when altering an instance of this struct
  * @serving: the virtual channel currently being served by this physical
  * channel
  * @locked: channel unavailable for the system, e.g. dedicated to secure
  * world
+ * @ftdmac020: channel is on a FTDMAC020
+ * @pl080s: channel is on a PL08s
  */
 struct pl08x_phy_chan {
 	unsigned int id;
 	void __iomem *base;
 	void __iomem *reg_config;
+	void __iomem *reg_control;
+	void __iomem *reg_src;
+	void __iomem *reg_dst;
+	void __iomem *reg_lli;
+	void __iomem *reg_busy;
 	spinlock_t lock;
 	struct pl08x_dma_chan *serving;
 	bool locked;
+	bool ftdmac020;
+	bool pl080s;
 };
 
 /**
@@ -253,8 +271,9 @@ struct pl08x_dma_chan {
 
 /**
  * struct pl08x_driver_data - the local state holder for the PL08x
- * @slave: slave engine for this instance
+ * @slave: optional slave engine for this instance
  * @memcpy: memcpy engine for this instance
+ * @has_slave: the PL08x has a slave engine (routed signals)
  * @base: virtual memory base (remapped) for the PL08x
  * @adev: the corresponding AMBA (PrimeCell) bus entry
  * @vd: vendor data for this PL08x variant
@@ -269,6 +288,7 @@ struct pl08x_dma_chan {
 struct pl08x_driver_data {
 	struct dma_device slave;
 	struct dma_device memcpy;
+	bool has_slave;
 	void __iomem *base;
 	struct amba_device *adev;
 	const struct vendor_data *vd;
@@ -360,10 +380,24 @@ static int pl08x_phy_channel_busy(struct pl08x_phy_chan *ch)
 {
 	unsigned int val;
 
+	/* If we have a special busy register, take a shortcut */
+	if (ch->reg_busy) {
+		val = readl(ch->reg_busy);
+		return !!(val & BIT(ch->id));
+	}
 	val = readl(ch->reg_config);
 	return val & PL080_CONFIG_ACTIVE;
 }
 
+/*
+ * pl08x_write_lli() - Write an LLI into the DMA controller.
+ *
+ * The PL08x derivatives support linked lists, but the first item of the
+ * list containing the source, destination, control word and next LLI is
+ * ignored. Instead the driver has to write those values directly into the
+ * SRC, DST, LLI and control registers. On FTDMAC020 also the SIZE
+ * register need to be set up for the first transfer.
+ */
 static void pl08x_write_lli(struct pl08x_driver_data *pl08x,
 		struct pl08x_phy_chan *phychan, const u32 *lli, u32 ccfg)
 {
@@ -381,11 +415,112 @@ static void pl08x_write_lli(struct pl08x_driver_data *pl08x,
 			phychan->id, lli[PL080_LLI_SRC], lli[PL080_LLI_DST],
 			lli[PL080_LLI_LLI], lli[PL080_LLI_CCTL], ccfg);
 
-	writel_relaxed(lli[PL080_LLI_SRC], phychan->base + PL080_CH_SRC_ADDR);
-	writel_relaxed(lli[PL080_LLI_DST], phychan->base + PL080_CH_DST_ADDR);
-	writel_relaxed(lli[PL080_LLI_LLI], phychan->base + PL080_CH_LLI);
-	writel_relaxed(lli[PL080_LLI_CCTL], phychan->base + PL080_CH_CONTROL);
+	writel_relaxed(lli[PL080_LLI_SRC], phychan->reg_src);
+	writel_relaxed(lli[PL080_LLI_DST], phychan->reg_dst);
+	writel_relaxed(lli[PL080_LLI_LLI], phychan->reg_lli);
+
+	/*
+	 * The FTMAC020 has a different layout in the CCTL word of the LLI
+	 * and the CCTL register which is split in CSR and SIZE registers.
+	 * Convert the LLI item CCTL into the proper values to write into
+	 * the CSR and SIZE registers.
+	 */
+	if (phychan->ftdmac020) {
+		u32 llictl = lli[PL080_LLI_CCTL];
+		u32 val = 0;
+
+		/* Write the transfer size (12 bits) to the size register */
+		writel_relaxed(llictl & FTDMAC020_LLI_TRANSFER_SIZE_MASK,
+			       phychan->base + FTDMAC020_CH_SIZE);
+		/*
+		 * Then write the control bits 28..16 to the control register
+		 * by shuffleing the bits around to where they are in the
+		 * main register. The mapping is as follows:
+		 * Bit 28: TC_MSK - mask on all except last LLI
+		 * Bit 27..25: SRC_WIDTH
+		 * Bit 24..22: DST_WIDTH
+		 * Bit 21..20: SRCAD_CTRL
+		 * Bit 19..17: DSTAD_CTRL
+		 * Bit 17: SRC_SEL
+		 * Bit 16: DST_SEL
+		 */
+		if (llictl & FTDMAC020_LLI_TC_MSK)
+			val |= FTDMAC020_CH_CSR_TC_MSK;
+		val |= ((llictl  & FTDMAC020_LLI_SRC_WIDTH_MSK) >>
+			(FTDMAC020_LLI_SRC_WIDTH_SHIFT -
+			 FTDMAC020_CH_CSR_SRC_WIDTH_SHIFT));
+		val |= ((llictl  & FTDMAC020_LLI_DST_WIDTH_MSK) >>
+			(FTDMAC020_LLI_DST_WIDTH_SHIFT -
+			 FTDMAC020_CH_CSR_DST_WIDTH_SHIFT));
+		val |= ((llictl  & FTDMAC020_LLI_SRCAD_CTL_MSK) >>
+			(FTDMAC020_LLI_SRCAD_CTL_SHIFT -
+			 FTDMAC020_CH_CSR_SRCAD_CTL_SHIFT));
+		val |= ((llictl  & FTDMAC020_LLI_DSTAD_CTL_MSK) >>
+			(FTDMAC020_LLI_DSTAD_CTL_SHIFT -
+			 FTDMAC020_CH_CSR_DSTAD_CTL_SHIFT));
+		if (llictl & FTDMAC020_LLI_SRC_SEL)
+			val |= FTDMAC020_CH_CSR_SRC_SEL;
+		if (llictl & FTDMAC020_LLI_DST_SEL)
+			val |= FTDMAC020_CH_CSR_DST_SEL;
+
+		/*
+		 * Set up the bits that exist in the CSR but are not
+		 * part the LLI, i.e. only gets written to the control
+		 * register right here.
+		 *
+		 * FIXME: do not just handle memcpy, also handle slave DMA.
+		 */
+		switch (pl08x->pd->memcpy_burst_size) {
+		default:
+		case PL08X_BURST_SZ_1:
+			val |= PL080_BSIZE_1 <<
+				FTDMAC020_CH_CSR_SRC_SIZE_SHIFT;
+			break;
+		case PL08X_BURST_SZ_4:
+			val |= PL080_BSIZE_4 <<
+				FTDMAC020_CH_CSR_SRC_SIZE_SHIFT;
+			break;
+		case PL08X_BURST_SZ_8:
+			val |= PL080_BSIZE_8 <<
+				FTDMAC020_CH_CSR_SRC_SIZE_SHIFT;
+			break;
+		case PL08X_BURST_SZ_16:
+			val |= PL080_BSIZE_16 <<
+				FTDMAC020_CH_CSR_SRC_SIZE_SHIFT;
+			break;
+		case PL08X_BURST_SZ_32:
+			val |= PL080_BSIZE_32 <<
+				FTDMAC020_CH_CSR_SRC_SIZE_SHIFT;
+			break;
+		case PL08X_BURST_SZ_64:
+			val |= PL080_BSIZE_64 <<
+				FTDMAC020_CH_CSR_SRC_SIZE_SHIFT;
+			break;
+		case PL08X_BURST_SZ_128:
+			val |= PL080_BSIZE_128 <<
+				FTDMAC020_CH_CSR_SRC_SIZE_SHIFT;
+			break;
+		case PL08X_BURST_SZ_256:
+			val |= PL080_BSIZE_256 <<
+				FTDMAC020_CH_CSR_SRC_SIZE_SHIFT;
+			break;
+		}
+
+		/* Protection flags */
+		if (pl08x->pd->memcpy_prot_buff)
+			val |= FTDMAC020_CH_CSR_PROT2;
+		if (pl08x->pd->memcpy_prot_cache)
+			val |= FTDMAC020_CH_CSR_PROT3;
+		/* We are the kernel, so we are in privileged mode */
+		val |= FTDMAC020_CH_CSR_PROT1;
+
+		writel_relaxed(val, phychan->reg_control);
+	} else {
+		/* Bits are just identical */
+		writel_relaxed(lli[PL080_LLI_CCTL], phychan->reg_control);
+	}
 
+	/* Second control word on the PL080s */
 	if (pl08x->vd->pl080s)
 		writel_relaxed(lli[PL080S_LLI_CCTL2],
 				phychan->base + PL080S_CH_CONTROL2);
@@ -423,11 +558,25 @@ static void pl08x_start_next_txd(struct pl08x_dma_chan *plchan)
 		cpu_relax();
 
 	/* Do not access config register until channel shows as inactive */
-	val = readl(phychan->reg_config);
-	while ((val & PL080_CONFIG_ACTIVE) || (val & PL080_CONFIG_ENABLE))
+	if (phychan->ftdmac020) {
+		val = readl(phychan->reg_config);
+		while (val & FTDMAC020_CH_CFG_BUSY)
+			val = readl(phychan->reg_config);
+
+		val = readl(phychan->reg_control);
+		while (val & FTDMAC020_CH_CSR_EN)
+			val = readl(phychan->reg_control);
+
+		writel(val | FTDMAC020_CH_CSR_EN,
+		       phychan->reg_control);
+	} else {
 		val = readl(phychan->reg_config);
+		while ((val & PL080_CONFIG_ACTIVE) ||
+		       (val & PL080_CONFIG_ENABLE))
+			val = readl(phychan->reg_config);
 
-	writel(val | PL080_CONFIG_ENABLE, phychan->reg_config);
+		writel(val | PL080_CONFIG_ENABLE, phychan->reg_config);
+	}
 }
 
 /*
@@ -445,6 +594,14 @@ static void pl08x_pause_phy_chan(struct pl08x_phy_chan *ch)
 	u32 val;
 	int timeout;
 
+	if (ch->ftdmac020) {
+		/* Use the enable bit on the FTDMAC020 */
+		val = readl(ch->reg_control);
+		val &= ~FTDMAC020_CH_CSR_EN;
+		writel(val, ch->reg_control);
+		return;
+	}
+
 	/* Set the HALT bit and wait for the FIFO to drain */
 	val = readl(ch->reg_config);
 	val |= PL080_CONFIG_HALT;
@@ -464,6 +621,14 @@ static void pl08x_resume_phy_chan(struct pl08x_phy_chan *ch)
 {
 	u32 val;
 
+	/* Use the enable bit on the FTDMAC020 */
+	if (ch->ftdmac020) {
+		val = readl(ch->reg_control);
+		val |= FTDMAC020_CH_CSR_EN;
+		writel(val, ch->reg_control);
+		return;
+	}
+
 	/* Clear the HALT bit */
 	val = readl(ch->reg_config);
 	val &= ~PL080_CONFIG_HALT;
@@ -479,25 +644,68 @@ static void pl08x_resume_phy_chan(struct pl08x_phy_chan *ch)
 static void pl08x_terminate_phy_chan(struct pl08x_driver_data *pl08x,
 	struct pl08x_phy_chan *ch)
 {
-	u32 val = readl(ch->reg_config);
+	u32 val;
 
+	/* The layout for the FTDMAC020 is different */
+	if (ch->ftdmac020) {
+		/* Disable all interrupts */
+		val = readl(ch->reg_config);
+		val |= (FTDMAC020_CH_CFG_INT_ABT_MASK |
+			FTDMAC020_CH_CFG_INT_ERR_MASK |
+			FTDMAC020_CH_CFG_INT_TC_MASK);
+		writel(val, ch->reg_config);
+
+		/* Abort and disable channel */
+		val = readl(ch->reg_control);
+		val &= ~FTDMAC020_CH_CSR_EN;
+		val |= FTDMAC020_CH_CSR_ABT;
+		writel(val, ch->reg_control);
+
+		/* Clear ABT and ERR interrupt flags */
+		writel(BIT(ch->id) | BIT(ch->id + 16),
+		       pl08x->base + PL080_ERR_CLEAR);
+		writel(BIT(ch->id), pl08x->base + PL080_TC_CLEAR);
+
+		return;
+	}
+
+	val = readl(ch->reg_config);
 	val &= ~(PL080_CONFIG_ENABLE | PL080_CONFIG_ERR_IRQ_MASK |
 		 PL080_CONFIG_TC_IRQ_MASK);
-
 	writel(val, ch->reg_config);
 
 	writel(BIT(ch->id), pl08x->base + PL080_ERR_CLEAR);
 	writel(BIT(ch->id), pl08x->base + PL080_TC_CLEAR);
 }
 
-static inline u32 get_bytes_in_cctl(u32 cctl)
+static u32 get_bytes_in_phy_channel(struct pl08x_phy_chan *ch)
 {
-	/* The source width defines the number of bytes */
-	u32 bytes = cctl & PL080_CONTROL_TRANSFER_SIZE_MASK;
+	u32 val;
+	u32 bytes;
+
+	if (ch->ftdmac020) {
+		bytes = readl(ch->base + FTDMAC020_CH_SIZE);
 
-	cctl &= PL080_CONTROL_SWIDTH_MASK;
+		val = readl(ch->reg_control);
+		val &= FTDMAC020_CH_CSR_SRC_WIDTH_MSK;
+		val >>= FTDMAC020_CH_CSR_SRC_WIDTH_SHIFT;
+	} else if (ch->pl080s) {
+		val = readl(ch->base + PL080S_CH_CONTROL2);
+		bytes = val & PL080S_CONTROL_TRANSFER_SIZE_MASK;
 
-	switch (cctl >> PL080_CONTROL_SWIDTH_SHIFT) {
+		val = readl(ch->reg_control);
+		val &= PL080_CONTROL_SWIDTH_MASK;
+		val >>= PL080_CONTROL_SWIDTH_SHIFT;
+	} else {
+		/* Plain PL08x */
+		val = readl(ch->reg_control);
+		bytes = val & PL080_CONTROL_TRANSFER_SIZE_MASK;
+
+		val &= PL080_CONTROL_SWIDTH_MASK;
+		val >>= PL080_CONTROL_SWIDTH_SHIFT;
+	}
+
+	switch (val) {
 	case PL080_WIDTH_8BIT:
 		break;
 	case PL080_WIDTH_16BIT:
@@ -510,14 +718,35 @@ static inline u32 get_bytes_in_cctl(u32 cctl)
 	return bytes;
 }
 
-static inline u32 get_bytes_in_cctl_pl080s(u32 cctl, u32 cctl1)
+static u32 get_bytes_in_lli(struct pl08x_phy_chan *ch, const u32 *llis_va)
 {
-	/* The source width defines the number of bytes */
-	u32 bytes = cctl1 & PL080S_CONTROL_TRANSFER_SIZE_MASK;
+	u32 val;
+	u32 bytes;
+
+	if (ch->ftdmac020) {
+		val = llis_va[PL080_LLI_CCTL];
+		bytes = val & FTDMAC020_LLI_TRANSFER_SIZE_MASK;
+
+		val = llis_va[PL080_LLI_CCTL];
+		val &= FTDMAC020_LLI_SRC_WIDTH_MSK;
+		val >>= FTDMAC020_LLI_SRC_WIDTH_SHIFT;
+	} else if (ch->pl080s) {
+		val = llis_va[PL080S_LLI_CCTL2];
+		bytes = val & PL080S_CONTROL_TRANSFER_SIZE_MASK;
+
+		val = llis_va[PL080_LLI_CCTL];
+		val &= PL080_CONTROL_SWIDTH_MASK;
+		val >>= PL080_CONTROL_SWIDTH_SHIFT;
+	} else {
+		/* Plain PL08x */
+		val = llis_va[PL080_LLI_CCTL];
+		bytes = val & PL080_CONTROL_TRANSFER_SIZE_MASK;
 
-	cctl &= PL080_CONTROL_SWIDTH_MASK;
+		val &= PL080_CONTROL_SWIDTH_MASK;
+		val >>= PL080_CONTROL_SWIDTH_SHIFT;
+	}
 
-	switch (cctl >> PL080_CONTROL_SWIDTH_SHIFT) {
+	switch (val) {
 	case PL080_WIDTH_8BIT:
 		break;
 	case PL080_WIDTH_16BIT:
@@ -552,15 +781,10 @@ static u32 pl08x_getbytes_chan(struct pl08x_dma_chan *plchan)
 	 * Follow the LLIs to get the number of remaining
 	 * bytes in the currently active transaction.
 	 */
-	clli = readl(ch->base + PL080_CH_LLI) & ~PL080_LLI_LM_AHB2;
+	clli = readl(ch->reg_lli) & ~PL080_LLI_LM_AHB2;
 
 	/* First get the remaining bytes in the active transfer */
-	if (pl08x->vd->pl080s)
-		bytes = get_bytes_in_cctl_pl080s(
-				readl(ch->base + PL080_CH_CONTROL),
-				readl(ch->base + PL080S_CH_CONTROL2));
-	else
-		bytes = get_bytes_in_cctl(readl(ch->base + PL080_CH_CONTROL));
+	bytes = get_bytes_in_phy_channel(ch);
 
 	if (!clli)
 		return bytes;
@@ -581,12 +805,7 @@ static u32 pl08x_getbytes_chan(struct pl08x_dma_chan *plchan)
 	llis_va_limit = llis_va + llis_max_words;
 
 	for (; llis_va < llis_va_limit; llis_va += pl08x->lli_words) {
-		if (pl08x->vd->pl080s)
-			bytes += get_bytes_in_cctl_pl080s(
-						llis_va[PL080_LLI_CCTL],
-						llis_va[PL080S_LLI_CCTL2]);
-		else
-			bytes += get_bytes_in_cctl(llis_va[PL080_LLI_CCTL]);
+		bytes += get_bytes_in_lli(ch, llis_va);
 
 		/*
 		 * A LLI pointer going backward terminates the LLI list
@@ -705,7 +924,7 @@ static void pl08x_phy_free(struct pl08x_dma_chan *plchan)
 			break;
 		}
 
-	if (!next) {
+	if (!next && pl08x->has_slave) {
 		list_for_each_entry(p, &pl08x->slave.channels, vc.chan.device_node)
 			if (p->state == PL08X_CHAN_WAITING) {
 				next = p;
@@ -746,9 +965,30 @@ static void pl08x_phy_free(struct pl08x_dma_chan *plchan)
  * LLI handling
  */
 
-static inline unsigned int pl08x_get_bytes_for_cctl(unsigned int coded)
+static inline unsigned int
+pl08x_get_bytes_for_lli(struct pl08x_driver_data *pl08x,
+			u32 cctl,
+			bool source)
 {
-	switch (coded) {
+	u32 val;
+
+	if (pl08x->vd->ftdmac020) {
+		if (source)
+			val = (cctl & FTDMAC020_LLI_SRC_WIDTH_MSK) >>
+				FTDMAC020_LLI_SRC_WIDTH_SHIFT;
+		else
+			val = (cctl & FTDMAC020_LLI_DST_WIDTH_MSK) >>
+				FTDMAC020_LLI_DST_WIDTH_SHIFT;
+	} else {
+		if (source)
+			val = (cctl & PL080_CONTROL_SWIDTH_MASK) >>
+				PL080_CONTROL_SWIDTH_SHIFT;
+		else
+			val = (cctl & PL080_CONTROL_DWIDTH_MASK) >>
+				PL080_CONTROL_DWIDTH_SHIFT;
+	}
+
+	switch (val) {
 	case PL080_WIDTH_8BIT:
 		return 1;
 	case PL080_WIDTH_16BIT:
@@ -762,49 +1002,106 @@ static inline unsigned int pl08x_get_bytes_for_cctl(unsigned int coded)
 	return 0;
 }
 
-static inline u32 pl08x_cctl_bits(u32 cctl, u8 srcwidth, u8 dstwidth,
-				  size_t tsize)
+static inline u32 pl08x_lli_control_bits(struct pl08x_driver_data *pl08x,
+					 u32 cctl,
+					 u8 srcwidth, u8 dstwidth,
+					 size_t tsize)
 {
 	u32 retbits = cctl;
 
-	/* Remove all src, dst and transfer size bits */
-	retbits &= ~PL080_CONTROL_DWIDTH_MASK;
-	retbits &= ~PL080_CONTROL_SWIDTH_MASK;
-	retbits &= ~PL080_CONTROL_TRANSFER_SIZE_MASK;
+	/*
+	 * Remove all src, dst and transfer size bits, then set the
+	 * width and size according to the parameters. The bit offsets
+	 * are different in the FTDMAC020 so we need to accound for this.
+	 */
+	if (pl08x->vd->ftdmac020) {
+		retbits &= ~FTDMAC020_LLI_DST_WIDTH_MSK;
+		retbits &= ~FTDMAC020_LLI_SRC_WIDTH_MSK;
+		retbits &= ~FTDMAC020_LLI_TRANSFER_SIZE_MASK;
+
+		switch (srcwidth) {
+		case 1:
+			retbits |= PL080_WIDTH_8BIT <<
+				FTDMAC020_LLI_SRC_WIDTH_SHIFT;
+			break;
+		case 2:
+			retbits |= PL080_WIDTH_16BIT <<
+				FTDMAC020_LLI_SRC_WIDTH_SHIFT;
+			break;
+		case 4:
+			retbits |= PL080_WIDTH_32BIT <<
+				FTDMAC020_LLI_SRC_WIDTH_SHIFT;
+			break;
+		default:
+			BUG();
+			break;
+		}
 
-	/* Then set the bits according to the parameters */
-	switch (srcwidth) {
-	case 1:
-		retbits |= PL080_WIDTH_8BIT << PL080_CONTROL_SWIDTH_SHIFT;
-		break;
-	case 2:
-		retbits |= PL080_WIDTH_16BIT << PL080_CONTROL_SWIDTH_SHIFT;
-		break;
-	case 4:
-		retbits |= PL080_WIDTH_32BIT << PL080_CONTROL_SWIDTH_SHIFT;
-		break;
-	default:
-		BUG();
-		break;
-	}
+		switch (dstwidth) {
+		case 1:
+			retbits |= PL080_WIDTH_8BIT <<
+				FTDMAC020_LLI_DST_WIDTH_SHIFT;
+			break;
+		case 2:
+			retbits |= PL080_WIDTH_16BIT <<
+				FTDMAC020_LLI_DST_WIDTH_SHIFT;
+			break;
+		case 4:
+			retbits |= PL080_WIDTH_32BIT <<
+				FTDMAC020_LLI_DST_WIDTH_SHIFT;
+			break;
+		default:
+			BUG();
+			break;
+		}
 
-	switch (dstwidth) {
-	case 1:
-		retbits |= PL080_WIDTH_8BIT << PL080_CONTROL_DWIDTH_SHIFT;
-		break;
-	case 2:
-		retbits |= PL080_WIDTH_16BIT << PL080_CONTROL_DWIDTH_SHIFT;
-		break;
-	case 4:
-		retbits |= PL080_WIDTH_32BIT << PL080_CONTROL_DWIDTH_SHIFT;
-		break;
-	default:
-		BUG();
-		break;
+		tsize &= FTDMAC020_LLI_TRANSFER_SIZE_MASK;
+		retbits |= tsize << FTDMAC020_LLI_TRANSFER_SIZE_SHIFT;
+	} else {
+		retbits &= ~PL080_CONTROL_DWIDTH_MASK;
+		retbits &= ~PL080_CONTROL_SWIDTH_MASK;
+		retbits &= ~PL080_CONTROL_TRANSFER_SIZE_MASK;
+
+		switch (srcwidth) {
+		case 1:
+			retbits |= PL080_WIDTH_8BIT <<
+				PL080_CONTROL_SWIDTH_SHIFT;
+			break;
+		case 2:
+			retbits |= PL080_WIDTH_16BIT <<
+				PL080_CONTROL_SWIDTH_SHIFT;
+			break;
+		case 4:
+			retbits |= PL080_WIDTH_32BIT <<
+				PL080_CONTROL_SWIDTH_SHIFT;
+			break;
+		default:
+			BUG();
+			break;
+		}
+
+		switch (dstwidth) {
+		case 1:
+			retbits |= PL080_WIDTH_8BIT <<
+				PL080_CONTROL_DWIDTH_SHIFT;
+			break;
+		case 2:
+			retbits |= PL080_WIDTH_16BIT <<
+				PL080_CONTROL_DWIDTH_SHIFT;
+			break;
+		case 4:
+			retbits |= PL080_WIDTH_32BIT <<
+				PL080_CONTROL_DWIDTH_SHIFT;
+			break;
+		default:
+			BUG();
+			break;
+		}
+
+		tsize &= PL080_CONTROL_TRANSFER_SIZE_MASK;
+		retbits |= tsize << PL080_CONTROL_TRANSFER_SIZE_SHIFT;
 	}
 
-	tsize &= PL080_CONTROL_TRANSFER_SIZE_MASK;
-	retbits |= tsize << PL080_CONTROL_TRANSFER_SIZE_SHIFT;
 	return retbits;
 }
 
@@ -825,13 +1122,35 @@ struct pl08x_lli_build_data {
  * - prefers the destination bus if both available
  * - prefers bus with fixed address (i.e. peripheral)
  */
-static void pl08x_choose_master_bus(struct pl08x_lli_build_data *bd,
-	struct pl08x_bus_data **mbus, struct pl08x_bus_data **sbus, u32 cctl)
+static void pl08x_choose_master_bus(struct pl08x_driver_data *pl08x,
+				    struct pl08x_lli_build_data *bd,
+				    struct pl08x_bus_data **mbus,
+				    struct pl08x_bus_data **sbus,
+				    u32 cctl)
 {
-	if (!(cctl & PL080_CONTROL_DST_INCR)) {
+	bool dst_incr;
+	bool src_incr;
+
+	/*
+	 * The FTDMAC020 only supports memory-to-memory transfer, so
+	 * source and destination always increase.
+	 */
+	if (pl08x->vd->ftdmac020) {
+		dst_incr = true;
+		src_incr = true;
+	} else {
+		dst_incr = !!(cctl & PL080_CONTROL_DST_INCR);
+		src_incr = !!(cctl & PL080_CONTROL_SRC_INCR);
+	}
+
+	/*
+	 * If either bus is not advancing, i.e. it is a peripheral, that
+	 * one becomes master
+	 */
+	if (!dst_incr) {
 		*mbus = &bd->dstbus;
 		*sbus = &bd->srcbus;
-	} else if (!(cctl & PL080_CONTROL_SRC_INCR)) {
+	} else if (!src_incr) {
 		*mbus = &bd->srcbus;
 		*sbus = &bd->dstbus;
 	} else {
@@ -869,10 +1188,16 @@ static void pl08x_fill_lli_for_desc(struct pl08x_driver_data *pl08x,
 	if (pl08x->vd->pl080s)
 		llis_va[PL080S_LLI_CCTL2] = cctl2;
 
-	if (cctl & PL080_CONTROL_SRC_INCR)
+	if (pl08x->vd->ftdmac020) {
+		/* FIXME: only memcpy so far so both increase */
 		bd->srcbus.addr += len;
-	if (cctl & PL080_CONTROL_DST_INCR)
 		bd->dstbus.addr += len;
+	} else {
+		if (cctl & PL080_CONTROL_SRC_INCR)
+			bd->srcbus.addr += len;
+		if (cctl & PL080_CONTROL_DST_INCR)
+			bd->dstbus.addr += len;
+	}
 
 	BUG_ON(bd->remainder < len);
 
@@ -883,12 +1208,12 @@ static inline void prep_byte_width_lli(struct pl08x_driver_data *pl08x,
 			struct pl08x_lli_build_data *bd, u32 *cctl, u32 len,
 			int num_llis, size_t *total_bytes)
 {
-	*cctl = pl08x_cctl_bits(*cctl, 1, 1, len);
+	*cctl = pl08x_lli_control_bits(pl08x, *cctl, 1, 1, len);
 	pl08x_fill_lli_for_desc(pl08x, bd, num_llis, len, *cctl, len);
 	(*total_bytes) += len;
 }
 
-#ifdef VERBOSE_DEBUG
+#if 1
 static void pl08x_dump_lli(struct pl08x_driver_data *pl08x,
 			   const u32 *llis_va, int num_llis)
 {
@@ -953,14 +1278,10 @@ static int pl08x_fill_llis_for_desc(struct pl08x_driver_data *pl08x,
 	cctl = txd->cctl;
 
 	/* Find maximum width of the source bus */
-	bd.srcbus.maxwidth =
-		pl08x_get_bytes_for_cctl((cctl & PL080_CONTROL_SWIDTH_MASK) >>
-				       PL080_CONTROL_SWIDTH_SHIFT);
+	bd.srcbus.maxwidth = pl08x_get_bytes_for_lli(pl08x, cctl, true);
 
 	/* Find maximum width of the destination bus */
-	bd.dstbus.maxwidth =
-		pl08x_get_bytes_for_cctl((cctl & PL080_CONTROL_DWIDTH_MASK) >>
-				       PL080_CONTROL_DWIDTH_SHIFT);
+	bd.dstbus.maxwidth = pl08x_get_bytes_for_lli(pl08x, cctl, false);
 
 	list_for_each_entry(dsg, &txd->dsg_list, node) {
 		total_bytes = 0;
@@ -972,7 +1293,7 @@ static int pl08x_fill_llis_for_desc(struct pl08x_driver_data *pl08x,
 		bd.srcbus.buswidth = bd.srcbus.maxwidth;
 		bd.dstbus.buswidth = bd.dstbus.maxwidth;
 
-		pl08x_choose_master_bus(&bd, &mbus, &sbus, cctl);
+		pl08x_choose_master_bus(pl08x, &bd, &mbus, &sbus, cctl);
 
 		dev_vdbg(&pl08x->adev->dev,
 			"src=0x%08llx%s/%u dst=0x%08llx%s/%u len=%zu\n",
@@ -1009,8 +1330,14 @@ static int pl08x_fill_llis_for_desc(struct pl08x_driver_data *pl08x,
 		 *   supported. Thus, we can't have scattered addresses.
 		 */
 		if (!bd.remainder) {
-			u32 fc = (txd->ccfg & PL080_CONFIG_FLOW_CONTROL_MASK) >>
-				PL080_CONFIG_FLOW_CONTROL_SHIFT;
+			u32 fc;
+
+			/* FTDMAC020 only does memory-to-memory */
+			if (pl08x->vd->ftdmac020)
+				fc = PL080_FLOW_MEM2MEM;
+			else
+				fc = (txd->ccfg & PL080_CONFIG_FLOW_CONTROL_MASK) >>
+					PL080_CONFIG_FLOW_CONTROL_SHIFT;
 			if (!((fc >= PL080_FLOW_SRC2DST_DST) &&
 					(fc <= PL080_FLOW_SRC2DST_SRC))) {
 				dev_err(&pl08x->adev->dev, "%s sg len can't be zero",
@@ -1027,8 +1354,9 @@ static int pl08x_fill_llis_for_desc(struct pl08x_driver_data *pl08x,
 				return 0;
 			}
 
-			cctl = pl08x_cctl_bits(cctl, bd.srcbus.buswidth,
-					bd.dstbus.buswidth, 0);
+			cctl = pl08x_lli_control_bits(pl08x, cctl,
+					bd.srcbus.buswidth, bd.dstbus.buswidth,
+					0);
 			pl08x_fill_lli_for_desc(pl08x, &bd, num_llis++,
 					0, cctl, 0);
 			break;
@@ -1107,8 +1435,9 @@ static int pl08x_fill_llis_for_desc(struct pl08x_driver_data *pl08x,
 					"size 0x%08zx (remainder 0x%08zx)\n",
 					__func__, lli_len, bd.remainder);
 
-				cctl = pl08x_cctl_bits(cctl, bd.srcbus.buswidth,
-					bd.dstbus.buswidth, tsize);
+				cctl = pl08x_lli_control_bits(pl08x, cctl,
+					bd.srcbus.buswidth, bd.dstbus.buswidth,
+					tsize);
 				pl08x_fill_lli_for_desc(pl08x, &bd, num_llis++,
 						lli_len, cctl, tsize);
 				total_bytes += lli_len;
@@ -1151,7 +1480,10 @@ static int pl08x_fill_llis_for_desc(struct pl08x_driver_data *pl08x,
 		/* The final LLI terminates the LLI. */
 		last_lli[PL080_LLI_LLI] = 0;
 		/* The final LLI element shall also fire an interrupt. */
-		last_lli[PL080_LLI_CCTL] |= PL080_CONTROL_TC_IRQ_EN;
+		if (pl08x->vd->ftdmac020)
+			last_lli[PL080_LLI_CCTL] &= ~FTDMAC020_LLI_TC_MSK;
+		else
+			last_lli[PL080_LLI_CCTL] |= PL080_CONTROL_TC_IRQ_EN;
 	}
 
 	pl08x_dump_lli(pl08x, llis_va, num_llis);
@@ -1317,14 +1649,25 @@ static const struct burst_table burst_sizes[] = {
  * will be routed to each port.  We try to have source and destination
  * on separate ports, but always respect the allowable settings.
  */
-static u32 pl08x_select_bus(u8 src, u8 dst)
+static u32 pl08x_select_bus(bool ftdmac020, u8 src, u8 dst)
 {
 	u32 cctl = 0;
+	u32 dst_ahb2;
+	u32 src_ahb2;
+
+	/* The FTDMAC020 use different bits to indicate src/dst bus */
+	if (ftdmac020) {
+		dst_ahb2 = FTDMAC020_LLI_DST_SEL;
+		src_ahb2 = FTDMAC020_LLI_SRC_SEL;
+	} else {
+		dst_ahb2 = PL080_CONTROL_DST_AHB2;
+		src_ahb2 = PL080_CONTROL_SRC_AHB2;
+	}
 
 	if (!(dst & PL08X_AHB1) || ((dst & PL08X_AHB2) && (src & PL08X_AHB1)))
-		cctl |= PL080_CONTROL_DST_AHB2;
+		cctl |= dst_ahb2;
 	if (!(src & PL08X_AHB1) || ((src & PL08X_AHB2) && !(dst & PL08X_AHB2)))
-		cctl |= PL080_CONTROL_SRC_AHB2;
+		cctl |= src_ahb2;
 
 	return cctl;
 }
@@ -1412,14 +1755,134 @@ static struct pl08x_txd *pl08x_get_txd(struct pl08x_dma_chan *plchan)
 {
 	struct pl08x_txd *txd = kzalloc(sizeof(*txd), GFP_NOWAIT);
 
-	if (txd) {
+	if (txd)
 		INIT_LIST_HEAD(&txd->dsg_list);
+	return txd;
+}
 
-		/* Always enable error and terminal interrupts */
-		txd->ccfg = PL080_CONFIG_ERR_IRQ_MASK |
-			    PL080_CONFIG_TC_IRQ_MASK;
+static u32 pl08x_memcpy_cctl(struct pl08x_driver_data *pl08x)
+{
+	u32 cctl = 0;
+
+	/* Conjure cctl */
+	switch (pl08x->pd->memcpy_burst_size) {
+	default:
+		dev_err(&pl08x->adev->dev,
+			"illegal burst size for memcpy, set to 1\n");
+		/* Fall through */
+	case PL08X_BURST_SZ_1:
+		cctl |= PL080_BSIZE_1 << PL080_CONTROL_SB_SIZE_SHIFT |
+			PL080_BSIZE_1 << PL080_CONTROL_DB_SIZE_SHIFT;
+		break;
+	case PL08X_BURST_SZ_4:
+		cctl |= PL080_BSIZE_4 << PL080_CONTROL_SB_SIZE_SHIFT |
+			PL080_BSIZE_4 << PL080_CONTROL_DB_SIZE_SHIFT;
+		break;
+	case PL08X_BURST_SZ_8:
+		cctl |= PL080_BSIZE_8 << PL080_CONTROL_SB_SIZE_SHIFT |
+			PL080_BSIZE_8 << PL080_CONTROL_DB_SIZE_SHIFT;
+		break;
+	case PL08X_BURST_SZ_16:
+		cctl |= PL080_BSIZE_16 << PL080_CONTROL_SB_SIZE_SHIFT |
+			PL080_BSIZE_16 << PL080_CONTROL_DB_SIZE_SHIFT;
+		break;
+	case PL08X_BURST_SZ_32:
+		cctl |= PL080_BSIZE_32 << PL080_CONTROL_SB_SIZE_SHIFT |
+			PL080_BSIZE_32 << PL080_CONTROL_DB_SIZE_SHIFT;
+		break;
+	case PL08X_BURST_SZ_64:
+		cctl |= PL080_BSIZE_64 << PL080_CONTROL_SB_SIZE_SHIFT |
+			PL080_BSIZE_64 << PL080_CONTROL_DB_SIZE_SHIFT;
+		break;
+	case PL08X_BURST_SZ_128:
+		cctl |= PL080_BSIZE_128 << PL080_CONTROL_SB_SIZE_SHIFT |
+			PL080_BSIZE_128 << PL080_CONTROL_DB_SIZE_SHIFT;
+		break;
+	case PL08X_BURST_SZ_256:
+		cctl |= PL080_BSIZE_256 << PL080_CONTROL_SB_SIZE_SHIFT |
+			PL080_BSIZE_256 << PL080_CONTROL_DB_SIZE_SHIFT;
+		break;
 	}
-	return txd;
+
+	switch (pl08x->pd->memcpy_bus_width) {
+	default:
+		dev_err(&pl08x->adev->dev,
+			"illegal bus width for memcpy, set to 8 bits\n");
+		/* Fall through */
+	case PL08X_BUS_WIDTH_8_BITS:
+		cctl |= PL080_WIDTH_8BIT << PL080_CONTROL_SWIDTH_SHIFT |
+			PL080_WIDTH_8BIT << PL080_CONTROL_DWIDTH_SHIFT;
+		break;
+	case PL08X_BUS_WIDTH_16_BITS:
+		cctl |= PL080_WIDTH_16BIT << PL080_CONTROL_SWIDTH_SHIFT |
+			PL080_WIDTH_16BIT << PL080_CONTROL_DWIDTH_SHIFT;
+		break;
+	case PL08X_BUS_WIDTH_32_BITS:
+		cctl |= PL080_WIDTH_32BIT << PL080_CONTROL_SWIDTH_SHIFT |
+			PL080_WIDTH_32BIT << PL080_CONTROL_DWIDTH_SHIFT;
+		break;
+	}
+
+	/* Protection flags */
+	if (pl08x->pd->memcpy_prot_buff)
+		cctl |= PL080_CONTROL_PROT_BUFF;
+	if (pl08x->pd->memcpy_prot_cache)
+		cctl |= PL080_CONTROL_PROT_CACHE;
+
+	/* We are the kernel, so we are in privileged mode */
+	cctl |= PL080_CONTROL_PROT_SYS;
+
+	/* Both to be incremented or the code will break */
+	cctl |= PL080_CONTROL_SRC_INCR | PL080_CONTROL_DST_INCR;
+
+	if (pl08x->vd->dualmaster)
+		cctl |= pl08x_select_bus(false,
+					 pl08x->mem_buses,
+					 pl08x->mem_buses);
+
+	return cctl;
+}
+
+static u32 pl08x_ftdmac020_memcpy_cctl(struct pl08x_driver_data *pl08x)
+{
+	u32 cctl = 0;
+
+	/* Conjure cctl */
+	switch (pl08x->pd->memcpy_bus_width) {
+	default:
+		dev_err(&pl08x->adev->dev,
+			"illegal bus width for memcpy, set to 8 bits\n");
+		/* Fall through */
+	case PL08X_BUS_WIDTH_8_BITS:
+		cctl |= PL080_WIDTH_8BIT << FTDMAC020_LLI_SRC_WIDTH_SHIFT |
+			PL080_WIDTH_8BIT << FTDMAC020_LLI_DST_WIDTH_SHIFT;
+		break;
+	case PL08X_BUS_WIDTH_16_BITS:
+		cctl |= PL080_WIDTH_16BIT << FTDMAC020_LLI_SRC_WIDTH_SHIFT |
+			PL080_WIDTH_16BIT << FTDMAC020_LLI_DST_WIDTH_SHIFT;
+		break;
+	case PL08X_BUS_WIDTH_32_BITS:
+		cctl |= PL080_WIDTH_32BIT << FTDMAC020_LLI_SRC_WIDTH_SHIFT |
+			PL080_WIDTH_32BIT << FTDMAC020_LLI_DST_WIDTH_SHIFT;
+		break;
+	}
+
+	/*
+	 * By default mask the TC IRQ on all LLIs, it will be unmasked on
+	 * the last LLI item by other code.
+	 */
+	cctl |= FTDMAC020_LLI_TC_MSK;
+
+	/*
+	 * Both to be incremented so leave bits FTDMAC020_LLI_SRCAD_CTL
+	 * and FTDMAC020_LLI_DSTAD_CTL as zero
+	 */
+	if (pl08x->vd->dualmaster)
+		cctl |= pl08x_select_bus(true,
+					 pl08x->mem_buses,
+					 pl08x->mem_buses);
+
+	return cctl;
 }
 
 /*
@@ -1452,18 +1915,16 @@ static struct dma_async_tx_descriptor *pl08x_prep_dma_memcpy(
 	dsg->src_addr = src;
 	dsg->dst_addr = dest;
 	dsg->len = len;
-
-	/* Set platform data for m2m */
-	txd->ccfg |= PL080_FLOW_MEM2MEM << PL080_CONFIG_FLOW_CONTROL_SHIFT;
-	txd->cctl = pl08x->pd->memcpy_channel.cctl_memcpy &
-			~(PL080_CONTROL_DST_AHB2 | PL080_CONTROL_SRC_AHB2);
-
-	/* Both to be incremented or the code will break */
-	txd->cctl |= PL080_CONTROL_SRC_INCR | PL080_CONTROL_DST_INCR;
-
-	if (pl08x->vd->dualmaster)
-		txd->cctl |= pl08x_select_bus(pl08x->mem_buses,
-					      pl08x->mem_buses);
+	if (pl08x->vd->ftdmac020) {
+		/* Writing CCFG zero ENABLES all interrupts */
+		txd->ccfg = 0;
+		txd->cctl = pl08x_ftdmac020_memcpy_cctl(pl08x);
+	} else {
+		txd->ccfg = PL080_CONFIG_ERR_IRQ_MASK |
+			PL080_CONFIG_TC_IRQ_MASK |
+			PL080_FLOW_MEM2MEM << PL080_CONFIG_FLOW_CONTROL_SHIFT;
+		txd->cctl = pl08x_memcpy_cctl(pl08x);
+	}
 
 	ret = pl08x_fill_llis_for_desc(plchan->host, txd);
 	if (!ret) {
@@ -1527,7 +1988,7 @@ static struct pl08x_txd *pl08x_init_txd(
 		return NULL;
 	}
 
-	txd->cctl = cctl | pl08x_select_bus(src_buses, dst_buses);
+	txd->cctl = cctl | pl08x_select_bus(false, src_buses, dst_buses);
 
 	if (plchan->cfg.device_fc)
 		tmp = (direction == DMA_MEM_TO_DEV) ? PL080_FLOW_MEM2PER_PER :
@@ -1536,7 +1997,9 @@ static struct pl08x_txd *pl08x_init_txd(
 		tmp = (direction == DMA_MEM_TO_DEV) ? PL080_FLOW_MEM2PER :
 			PL080_FLOW_PER2MEM;
 
-	txd->ccfg |= tmp << PL080_CONFIG_FLOW_CONTROL_SHIFT;
+	txd->ccfg = PL080_CONFIG_ERR_IRQ_MASK |
+		PL080_CONFIG_TC_IRQ_MASK |
+		tmp << PL080_CONFIG_FLOW_CONTROL_SHIFT;
 
 	ret = pl08x_request_mux(plchan);
 	if (ret < 0) {
@@ -1813,6 +2276,11 @@ static void pl08x_ensure_on(struct pl08x_driver_data *pl08x)
 	/* The Nomadik variant does not have the config register */
 	if (pl08x->vd->nomadik)
 		return;
+	/* The FTDMAC020 variant does this in another register */
+	if (pl08x->vd->ftdmac020) {
+		writel(PL080_CONFIG_ENABLE, pl08x->base + FTDMAC020_CSR);
+		return;
+	}
 	writel(PL080_CONFIG_ENABLE, pl08x->base + PL080_CONFIG);
 }
 
@@ -1925,9 +2393,16 @@ static int pl08x_dma_init_virtual_channels(struct pl08x_driver_data *pl08x,
 			chan->signal = i;
 			pl08x_dma_slave_init(chan);
 		} else {
-			chan->cd = &pl08x->pd->memcpy_channel;
+			chan->cd = kzalloc(sizeof(*chan->cd), GFP_KERNEL);
+			if (!chan->cd) {
+				kfree(chan);
+				return -ENOMEM;
+			}
+			chan->cd->bus_id = "memcpy";
+			chan->cd->periph_buses = pl08x->pd->mem_buses;
 			chan->name = kasprintf(GFP_KERNEL, "memcpy%d", i);
 			if (!chan->name) {
+				kfree(chan->cd);
 				kfree(chan);
 				return -ENOMEM;
 			}
@@ -2009,12 +2484,15 @@ static int pl08x_debugfs_show(struct seq_file *s, void *data)
 			   pl08x_state_str(chan->state));
 	}
 
-	seq_printf(s, "\nPL08x virtual slave channels:\n");
-	seq_printf(s, "CHANNEL:\tSTATE:\n");
-	seq_printf(s, "--------\t------\n");
-	list_for_each_entry(chan, &pl08x->slave.channels, vc.chan.device_node) {
-		seq_printf(s, "%s\t\t%s\n", chan->name,
-			   pl08x_state_str(chan->state));
+	if (pl08x->has_slave) {
+		seq_printf(s, "\nPL08x virtual slave channels:\n");
+		seq_printf(s, "CHANNEL:\tSTATE:\n");
+		seq_printf(s, "--------\t------\n");
+		list_for_each_entry(chan, &pl08x->slave.channels,
+				    vc.chan.device_node) {
+			seq_printf(s, "%s\t\t%s\n", chan->name,
+				   pl08x_state_str(chan->state));
+		}
 	}
 
 	return 0;
@@ -2052,6 +2530,10 @@ static struct dma_chan *pl08x_find_chan_id(struct pl08x_driver_data *pl08x,
 {
 	struct pl08x_dma_chan *chan;
 
+	/* Trying to get a slave channel from something with no slave support */
+	if (!pl08x->has_slave)
+		return NULL;
+
 	list_for_each_entry(chan, &pl08x->slave.channels, vc.chan.device_node) {
 		if (chan->signal == id)
 			return &chan->vc.chan;
@@ -2099,7 +2581,6 @@ static int pl08x_of_probe(struct amba_device *adev,
 {
 	struct pl08x_platform_data *pd;
 	struct pl08x_channel_data *chanp = NULL;
-	u32 cctl_memcpy = 0;
 	u32 val;
 	int ret;
 	int i;
@@ -2139,36 +2620,28 @@ static int pl08x_of_probe(struct amba_device *adev,
 		dev_err(&adev->dev, "illegal burst size for memcpy, set to 1\n");
 		/* Fall through */
 	case 1:
-		cctl_memcpy |= PL080_BSIZE_1 << PL080_CONTROL_SB_SIZE_SHIFT |
-			       PL080_BSIZE_1 << PL080_CONTROL_DB_SIZE_SHIFT;
+		pd->memcpy_burst_size = PL08X_BURST_SZ_1;
 		break;
 	case 4:
-		cctl_memcpy |= PL080_BSIZE_4 << PL080_CONTROL_SB_SIZE_SHIFT |
-			       PL080_BSIZE_4 << PL080_CONTROL_DB_SIZE_SHIFT;
+		pd->memcpy_burst_size = PL08X_BURST_SZ_4;
 		break;
 	case 8:
-		cctl_memcpy |= PL080_BSIZE_8 << PL080_CONTROL_SB_SIZE_SHIFT |
-			       PL080_BSIZE_8 << PL080_CONTROL_DB_SIZE_SHIFT;
+		pd->memcpy_burst_size = PL08X_BURST_SZ_8;
 		break;
 	case 16:
-		cctl_memcpy |= PL080_BSIZE_16 << PL080_CONTROL_SB_SIZE_SHIFT |
-			       PL080_BSIZE_16 << PL080_CONTROL_DB_SIZE_SHIFT;
+		pd->memcpy_burst_size = PL08X_BURST_SZ_16;
 		break;
 	case 32:
-		cctl_memcpy |= PL080_BSIZE_32 << PL080_CONTROL_SB_SIZE_SHIFT |
-			       PL080_BSIZE_32 << PL080_CONTROL_DB_SIZE_SHIFT;
+		pd->memcpy_burst_size = PL08X_BURST_SZ_32;
 		break;
 	case 64:
-		cctl_memcpy |= PL080_BSIZE_64 << PL080_CONTROL_SB_SIZE_SHIFT |
-			       PL080_BSIZE_64 << PL080_CONTROL_DB_SIZE_SHIFT;
+		pd->memcpy_burst_size = PL08X_BURST_SZ_64;
 		break;
 	case 128:
-		cctl_memcpy |= PL080_BSIZE_128 << PL080_CONTROL_SB_SIZE_SHIFT |
-			       PL080_BSIZE_128 << PL080_CONTROL_DB_SIZE_SHIFT;
+		pd->memcpy_burst_size = PL08X_BURST_SZ_128;
 		break;
 	case 256:
-		cctl_memcpy |= PL080_BSIZE_256 << PL080_CONTROL_SB_SIZE_SHIFT |
-			       PL080_BSIZE_256 << PL080_CONTROL_DB_SIZE_SHIFT;
+		pd->memcpy_burst_size = PL08X_BURST_SZ_256;
 		break;
 	}
 
@@ -2182,48 +2655,40 @@ static int pl08x_of_probe(struct amba_device *adev,
 		dev_err(&adev->dev, "illegal bus width for memcpy, set to 8 bits\n");
 		/* Fall through */
 	case 8:
-		cctl_memcpy |= PL080_WIDTH_8BIT << PL080_CONTROL_SWIDTH_SHIFT |
-			       PL080_WIDTH_8BIT << PL080_CONTROL_DWIDTH_SHIFT;
+		pd->memcpy_bus_width = PL08X_BUS_WIDTH_8_BITS;
 		break;
 	case 16:
-		cctl_memcpy |= PL080_WIDTH_16BIT << PL080_CONTROL_SWIDTH_SHIFT |
-			       PL080_WIDTH_16BIT << PL080_CONTROL_DWIDTH_SHIFT;
+		pd->memcpy_bus_width = PL08X_BUS_WIDTH_16_BITS;
 		break;
 	case 32:
-		cctl_memcpy |= PL080_WIDTH_32BIT << PL080_CONTROL_SWIDTH_SHIFT |
-			       PL080_WIDTH_32BIT << PL080_CONTROL_DWIDTH_SHIFT;
+		pd->memcpy_bus_width = PL08X_BUS_WIDTH_32_BITS;
 		break;
 	}
 
-	/* This is currently the only thing making sense */
-	cctl_memcpy |= PL080_CONTROL_PROT_SYS;
-
-	/* Set up memcpy channel */
-	pd->memcpy_channel.bus_id = "memcpy";
-	pd->memcpy_channel.cctl_memcpy = cctl_memcpy;
-	/* Use the buses that can access memory, obviously */
-	pd->memcpy_channel.periph_buses = pd->mem_buses;
-
 	/*
 	 * Allocate channel data for all possible slave channels (one
 	 * for each possible signal), channels will then be allocated
 	 * for a device and have it's AHB interfaces set up at
 	 * translation time.
 	 */
-	chanp = devm_kcalloc(&adev->dev,
-			pl08x->vd->signals,
-			sizeof(struct pl08x_channel_data),
-			GFP_KERNEL);
-	if (!chanp)
-		return -ENOMEM;
+	if (pl08x->vd->signals) {
+		chanp = devm_kcalloc(&adev->dev,
+				     pl08x->vd->signals,
+				     sizeof(struct pl08x_channel_data),
+				     GFP_KERNEL);
+		if (!chanp)
+			return -ENOMEM;
 
-	pd->slave_channels = chanp;
-	for (i = 0; i < pl08x->vd->signals; i++) {
-		/* chanp->periph_buses will be assigned at translation */
-		chanp->bus_id = kasprintf(GFP_KERNEL, "slave%d", i);
-		chanp++;
+		pd->slave_channels = chanp;
+		for (i = 0; i < pl08x->vd->signals; i++) {
+			/*
+			 * chanp->periph_buses will be assigned at translation
+			 */
+			chanp->bus_id = kasprintf(GFP_KERNEL, "slave%d", i);
+			chanp++;
+		}
+		pd->num_slave_channels = pl08x->vd->signals;
 	}
-	pd->num_slave_channels = pl08x->vd->signals;
 
 	pl08x->pd = pd;
 
@@ -2242,7 +2707,7 @@ static inline int pl08x_of_probe(struct amba_device *adev,
 static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 {
 	struct pl08x_driver_data *pl08x;
-	const struct vendor_data *vd = id->data;
+	struct vendor_data *vd = id->data;
 	struct device_node *np = adev->dev.of_node;
 	u32 tsfr_size;
 	int ret = 0;
@@ -2268,6 +2733,34 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 	pl08x->adev = adev;
 	pl08x->vd = vd;
 
+	pl08x->base = ioremap(adev->res.start, resource_size(&adev->res));
+	if (!pl08x->base) {
+		ret = -ENOMEM;
+		goto out_no_ioremap;
+	}
+
+	if (vd->ftdmac020) {
+		u32 val;
+
+		val = readl(pl08x->base + FTDMAC020_REVISION);
+		dev_info(&pl08x->adev->dev, "FTDMAC020 %d.%d rel %d\n",
+			 (val >> 16) & 0xff, (val >> 8) & 0xff, val & 0xff);
+		val = readl(pl08x->base + FTDMAC020_FEATURE);
+		dev_info(&pl08x->adev->dev, "FTDMAC020 %d channels, "
+			 "%s built-in bridge, %s, %s linked lists\n",
+			 (val >> 12) & 0x0f,
+			 (val & BIT(10)) ? "no" : "has",
+			 (val & BIT(9)) ? "AHB0 and AHB1" : "AHB0",
+			 (val & BIT(8)) ? "supports" : "does not support");
+
+		/* Vendor data from feature register */
+		if (!(val & BIT(8)))
+			dev_warn(&pl08x->adev->dev,
+				 "linked lists not supported, required\n");
+		vd->channels = (val >> 12) & 0x0f;
+		vd->dualmaster = !!(val & BIT(9));
+	}
+
 	/* Initialize memcpy engine */
 	dma_cap_set(DMA_MEMCPY, pl08x->memcpy.cap_mask);
 	pl08x->memcpy.dev = &adev->dev;
@@ -2284,25 +2777,38 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 	pl08x->memcpy.dst_addr_widths = PL80X_DMA_BUSWIDTHS;
 	pl08x->memcpy.directions = BIT(DMA_MEM_TO_MEM);
 	pl08x->memcpy.residue_granularity = DMA_RESIDUE_GRANULARITY_SEGMENT;
+	if (vd->ftdmac020)
+		pl08x->memcpy.copy_align = DMAENGINE_ALIGN_4_BYTES;
 
-	/* Initialize slave engine */
-	dma_cap_set(DMA_SLAVE, pl08x->slave.cap_mask);
-	dma_cap_set(DMA_CYCLIC, pl08x->slave.cap_mask);
-	pl08x->slave.dev = &adev->dev;
-	pl08x->slave.device_free_chan_resources = pl08x_free_chan_resources;
-	pl08x->slave.device_prep_dma_interrupt = pl08x_prep_dma_interrupt;
-	pl08x->slave.device_tx_status = pl08x_dma_tx_status;
-	pl08x->slave.device_issue_pending = pl08x_issue_pending;
-	pl08x->slave.device_prep_slave_sg = pl08x_prep_slave_sg;
-	pl08x->slave.device_prep_dma_cyclic = pl08x_prep_dma_cyclic;
-	pl08x->slave.device_config = pl08x_config;
-	pl08x->slave.device_pause = pl08x_pause;
-	pl08x->slave.device_resume = pl08x_resume;
-	pl08x->slave.device_terminate_all = pl08x_terminate_all;
-	pl08x->slave.src_addr_widths = PL80X_DMA_BUSWIDTHS;
-	pl08x->slave.dst_addr_widths = PL80X_DMA_BUSWIDTHS;
-	pl08x->slave.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
-	pl08x->slave.residue_granularity = DMA_RESIDUE_GRANULARITY_SEGMENT;
+
+	/*
+	 * Initialize slave engine, if the block has no signals, that means
+	 * we have no slave support.
+	 */
+	if (vd->signals) {
+		pl08x->has_slave = true;
+		dma_cap_set(DMA_SLAVE, pl08x->slave.cap_mask);
+		dma_cap_set(DMA_CYCLIC, pl08x->slave.cap_mask);
+		pl08x->slave.dev = &adev->dev;
+		pl08x->slave.device_free_chan_resources =
+			pl08x_free_chan_resources;
+		pl08x->slave.device_prep_dma_interrupt =
+			pl08x_prep_dma_interrupt;
+		pl08x->slave.device_tx_status = pl08x_dma_tx_status;
+		pl08x->slave.device_issue_pending = pl08x_issue_pending;
+		pl08x->slave.device_prep_slave_sg = pl08x_prep_slave_sg;
+		pl08x->slave.device_prep_dma_cyclic = pl08x_prep_dma_cyclic;
+		pl08x->slave.device_config = pl08x_config;
+		pl08x->slave.device_pause = pl08x_pause;
+		pl08x->slave.device_resume = pl08x_resume;
+		pl08x->slave.device_terminate_all = pl08x_terminate_all;
+		pl08x->slave.src_addr_widths = PL80X_DMA_BUSWIDTHS;
+		pl08x->slave.dst_addr_widths = PL80X_DMA_BUSWIDTHS;
+		pl08x->slave.directions =
+			BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV);
+		pl08x->slave.residue_granularity =
+			DMA_RESIDUE_GRANULARITY_SEGMENT;
+	}
 
 	/* Get the platform data */
 	pl08x->pd = dev_get_platdata(&adev->dev);
@@ -2344,19 +2850,18 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 		goto out_no_lli_pool;
 	}
 
-	pl08x->base = ioremap(adev->res.start, resource_size(&adev->res));
-	if (!pl08x->base) {
-		ret = -ENOMEM;
-		goto out_no_ioremap;
-	}
-
 	/* Turn on the PL08x */
 	pl08x_ensure_on(pl08x);
 
-	/* Attach the interrupt handler */
-	writel(0x000000FF, pl08x->base + PL080_ERR_CLEAR);
+	/* Clear any pending interrupts */
+	if (vd->ftdmac020)
+		/* This variant has error IRQs in bits 16-19 */
+		writel(0x0000FFFF, pl08x->base + PL080_ERR_CLEAR);
+	else
+		writel(0x000000FF, pl08x->base + PL080_ERR_CLEAR);
 	writel(0x000000FF, pl08x->base + PL080_TC_CLEAR);
 
+	/* Attach the interrupt handler */
 	ret = request_irq(adev->irq[0], pl08x_irq, 0, DRIVER_NAME, pl08x);
 	if (ret) {
 		dev_err(&adev->dev, "%s failed to request interrupt %d\n",
@@ -2377,7 +2882,25 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 
 		ch->id = i;
 		ch->base = pl08x->base + PL080_Cx_BASE(i);
-		ch->reg_config = ch->base + vd->config_offset;
+		if (vd->ftdmac020) {
+			/* FTDMA020 has a special channel busy register */
+			ch->reg_busy = ch->base + FTDMAC020_CH_BUSY;
+			ch->reg_config = ch->base + FTDMAC020_CH_CFG;
+			ch->reg_control = ch->base + FTDMAC020_CH_CSR;
+			ch->reg_src = ch->base + FTDMAC020_CH_SRC_ADDR;
+			ch->reg_dst = ch->base + FTDMAC020_CH_DST_ADDR;
+			ch->reg_lli = ch->base + FTDMAC020_CH_LLP;
+			ch->ftdmac020 = true;
+		} else {
+			ch->reg_config = ch->base + vd->config_offset;
+			ch->reg_control = ch->base + PL080_CH_CONTROL;
+			ch->reg_src = ch->base + PL080_CH_SRC_ADDR;
+			ch->reg_dst = ch->base + PL080_CH_DST_ADDR;
+			ch->reg_lli = ch->base + PL080_CH_LLI;
+		}
+		if (vd->pl080s)
+			ch->pl080s = true;
+
 		spin_lock_init(&ch->lock);
 
 		/*
@@ -2410,13 +2933,15 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 	}
 
 	/* Register slave channels */
-	ret = pl08x_dma_init_virtual_channels(pl08x, &pl08x->slave,
-			pl08x->pd->num_slave_channels, true);
-	if (ret < 0) {
-		dev_warn(&pl08x->adev->dev,
-			"%s failed to enumerate slave channels - %d\n",
-				__func__, ret);
-		goto out_no_slave;
+	if (pl08x->has_slave) {
+		ret = pl08x_dma_init_virtual_channels(pl08x, &pl08x->slave,
+					pl08x->pd->num_slave_channels, true);
+		if (ret < 0) {
+			dev_warn(&pl08x->adev->dev,
+				 "%s failed to enumerate slave channels - %d\n",
+				 __func__, ret);
+			goto out_no_slave;
+		}
 	}
 
 	ret = dma_async_device_register(&pl08x->memcpy);
@@ -2427,12 +2952,14 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 		goto out_no_memcpy_reg;
 	}
 
-	ret = dma_async_device_register(&pl08x->slave);
-	if (ret) {
-		dev_warn(&pl08x->adev->dev,
+	if (pl08x->has_slave) {
+		ret = dma_async_device_register(&pl08x->slave);
+		if (ret) {
+			dev_warn(&pl08x->adev->dev,
 			"%s failed to register slave as an async device - %d\n",
 			__func__, ret);
-		goto out_no_slave_reg;
+			goto out_no_slave_reg;
+		}
 	}
 
 	amba_set_drvdata(adev, pl08x);
@@ -2446,7 +2973,8 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
 out_no_slave_reg:
 	dma_async_device_unregister(&pl08x->memcpy);
 out_no_memcpy_reg:
-	pl08x_free_virtual_channels(&pl08x->slave);
+	if (pl08x->has_slave)
+		pl08x_free_virtual_channels(&pl08x->slave);
 out_no_slave:
 	pl08x_free_virtual_channels(&pl08x->memcpy);
 out_no_memcpy:
@@ -2454,11 +2982,11 @@ out_no_memcpy:
 out_no_phychans:
 	free_irq(adev->irq[0], pl08x);
 out_no_irq:
-	iounmap(pl08x->base);
-out_no_ioremap:
 	dma_pool_destroy(pl08x->pool);
 out_no_lli_pool:
 out_no_platdata:
+	iounmap(pl08x->base);
+out_no_ioremap:
 	kfree(pl08x);
 out_no_pl08x:
 	amba_release_regions(adev);
@@ -2499,6 +3027,12 @@ static struct vendor_data vendor_pl081 = {
 	.max_transfer_size = PL080_CONTROL_TRANSFER_SIZE_MASK,
 };
 
+static struct vendor_data vendor_ftdmac020 = {
+	.config_offset = PL080_CH_CONFIG,
+	.ftdmac020 = true,
+	.max_transfer_size = PL080_CONTROL_TRANSFER_SIZE_MASK,
+};
+
 static struct amba_id pl08x_ids[] = {
 	/* Samsung PL080S variant */
 	{
@@ -2524,6 +3058,12 @@ static struct amba_id pl08x_ids[] = {
 		.mask	= 0x00ffffff,
 		.data	= &vendor_nomadik,
 	},
+	/* Faraday Technology FTDMAC020 */
+	{
+		.id	= 0x0003b080,
+		.mask	= 0x000fffff,
+		.data	= &vendor_ftdmac020,
+	},
 	{ 0, 0 },
 };
 
diff --git a/drivers/dma/bcm-sba-raid.c b/drivers/dma/bcm-sba-raid.c
new file mode 100644
index 000000000000..e41bbc7cb094
--- /dev/null
+++ b/drivers/dma/bcm-sba-raid.c
@@ -0,0 +1,1785 @@
+/*
+ * Copyright (C) 2017 Broadcom
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+/*
+ * Broadcom SBA RAID Driver
+ *
+ * The Broadcom stream buffer accelerator (SBA) provides offloading
+ * capabilities for RAID operations. The SBA offload engine is accessible
+ * via Broadcom SoC specific ring manager. Two or more offload engines
+ * can share same Broadcom SoC specific ring manager due to this Broadcom
+ * SoC specific ring manager driver is implemented as a mailbox controller
+ * driver and offload engine drivers are implemented as mallbox clients.
+ *
+ * Typically, Broadcom SoC specific ring manager will implement larger
+ * number of hardware rings over one or more SBA hardware devices. By
+ * design, the internal buffer size of SBA hardware device is limited
+ * but all offload operations supported by SBA can be broken down into
+ * multiple small size requests and executed parallely on multiple SBA
+ * hardware devices for achieving high through-put.
+ *
+ * The Broadcom SBA RAID driver does not require any register programming
+ * except submitting request to SBA hardware device via mailbox channels.
+ * This driver implements a DMA device with one DMA channel using a set
+ * of mailbox channels provided by Broadcom SoC specific ring manager
+ * driver. To exploit parallelism (as described above), all DMA request
+ * coming to SBA RAID DMA channel are broken down to smaller requests
+ * and submitted to multiple mailbox channels in round-robin fashion.
+ * For having more SBA DMA channels, we can create more SBA device nodes
+ * in Broadcom SoC specific DTS based on number of hardware rings supported
+ * by Broadcom SoC ring manager.
+ */
+
+#include <linux/bitops.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/list.h>
+#include <linux/mailbox_client.h>
+#include <linux/mailbox/brcm-message.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/slab.h>
+#include <linux/raid/pq.h>
+
+#include "dmaengine.h"
+
+/* SBA command related defines */
+#define SBA_TYPE_SHIFT					48
+#define SBA_TYPE_MASK					GENMASK(1, 0)
+#define SBA_TYPE_A					0x0
+#define SBA_TYPE_B					0x2
+#define SBA_TYPE_C					0x3
+#define SBA_USER_DEF_SHIFT				32
+#define SBA_USER_DEF_MASK				GENMASK(15, 0)
+#define SBA_R_MDATA_SHIFT				24
+#define SBA_R_MDATA_MASK				GENMASK(7, 0)
+#define SBA_C_MDATA_MS_SHIFT				18
+#define SBA_C_MDATA_MS_MASK				GENMASK(1, 0)
+#define SBA_INT_SHIFT					17
+#define SBA_INT_MASK					BIT(0)
+#define SBA_RESP_SHIFT					16
+#define SBA_RESP_MASK					BIT(0)
+#define SBA_C_MDATA_SHIFT				8
+#define SBA_C_MDATA_MASK				GENMASK(7, 0)
+#define SBA_C_MDATA_BNUMx_SHIFT(__bnum)			(2 * (__bnum))
+#define SBA_C_MDATA_BNUMx_MASK				GENMASK(1, 0)
+#define SBA_C_MDATA_DNUM_SHIFT				5
+#define SBA_C_MDATA_DNUM_MASK				GENMASK(4, 0)
+#define SBA_C_MDATA_LS(__v)				((__v) & 0xff)
+#define SBA_C_MDATA_MS(__v)				(((__v) >> 8) & 0x3)
+#define SBA_CMD_SHIFT					0
+#define SBA_CMD_MASK					GENMASK(3, 0)
+#define SBA_CMD_ZERO_BUFFER				0x4
+#define SBA_CMD_ZERO_ALL_BUFFERS			0x8
+#define SBA_CMD_LOAD_BUFFER				0x9
+#define SBA_CMD_XOR					0xa
+#define SBA_CMD_GALOIS_XOR				0xb
+#define SBA_CMD_WRITE_BUFFER				0xc
+#define SBA_CMD_GALOIS					0xe
+
+/* Driver helper macros */
+#define to_sba_request(tx)		\
+	container_of(tx, struct sba_request, tx)
+#define to_sba_device(dchan)		\
+	container_of(dchan, struct sba_device, dma_chan)
+
+enum sba_request_state {
+	SBA_REQUEST_STATE_FREE = 1,
+	SBA_REQUEST_STATE_ALLOCED = 2,
+	SBA_REQUEST_STATE_PENDING = 3,
+	SBA_REQUEST_STATE_ACTIVE = 4,
+	SBA_REQUEST_STATE_RECEIVED = 5,
+	SBA_REQUEST_STATE_COMPLETED = 6,
+	SBA_REQUEST_STATE_ABORTED = 7,
+};
+
+struct sba_request {
+	/* Global state */
+	struct list_head node;
+	struct sba_device *sba;
+	enum sba_request_state state;
+	bool fence;
+	/* Chained requests management */
+	struct sba_request *first;
+	struct list_head next;
+	unsigned int next_count;
+	atomic_t next_pending_count;
+	/* BRCM message data */
+	void *resp;
+	dma_addr_t resp_dma;
+	struct brcm_sba_command *cmds;
+	struct brcm_message msg;
+	struct dma_async_tx_descriptor tx;
+};
+
+enum sba_version {
+	SBA_VER_1 = 0,
+	SBA_VER_2
+};
+
+struct sba_device {
+	/* Underlying device */
+	struct device *dev;
+	/* DT configuration parameters */
+	enum sba_version ver;
+	/* Derived configuration parameters */
+	u32 max_req;
+	u32 hw_buf_size;
+	u32 hw_resp_size;
+	u32 max_pq_coefs;
+	u32 max_pq_srcs;
+	u32 max_cmd_per_req;
+	u32 max_xor_srcs;
+	u32 max_resp_pool_size;
+	u32 max_cmds_pool_size;
+	/* Maibox client and Mailbox channels */
+	struct mbox_client client;
+	int mchans_count;
+	atomic_t mchans_current;
+	struct mbox_chan **mchans;
+	struct device *mbox_dev;
+	/* DMA device and DMA channel */
+	struct dma_device dma_dev;
+	struct dma_chan dma_chan;
+	/* DMA channel resources */
+	void *resp_base;
+	dma_addr_t resp_dma_base;
+	void *cmds_base;
+	dma_addr_t cmds_dma_base;
+	spinlock_t reqs_lock;
+	struct sba_request *reqs;
+	bool reqs_fence;
+	struct list_head reqs_alloc_list;
+	struct list_head reqs_pending_list;
+	struct list_head reqs_active_list;
+	struct list_head reqs_received_list;
+	struct list_head reqs_completed_list;
+	struct list_head reqs_aborted_list;
+	struct list_head reqs_free_list;
+	int reqs_free_count;
+};
+
+/* ====== SBA command helper routines ===== */
+
+static inline u64 __pure sba_cmd_enc(u64 cmd, u32 val, u32 shift, u32 mask)
+{
+	cmd &= ~((u64)mask << shift);
+	cmd |= ((u64)(val & mask) << shift);
+	return cmd;
+}
+
+static inline u32 __pure sba_cmd_load_c_mdata(u32 b0)
+{
+	return b0 & SBA_C_MDATA_BNUMx_MASK;
+}
+
+static inline u32 __pure sba_cmd_write_c_mdata(u32 b0)
+{
+	return b0 & SBA_C_MDATA_BNUMx_MASK;
+}
+
+static inline u32 __pure sba_cmd_xor_c_mdata(u32 b1, u32 b0)
+{
+	return (b0 & SBA_C_MDATA_BNUMx_MASK) |
+	       ((b1 & SBA_C_MDATA_BNUMx_MASK) << SBA_C_MDATA_BNUMx_SHIFT(1));
+}
+
+static inline u32 __pure sba_cmd_pq_c_mdata(u32 d, u32 b1, u32 b0)
+{
+	return (b0 & SBA_C_MDATA_BNUMx_MASK) |
+	       ((b1 & SBA_C_MDATA_BNUMx_MASK) << SBA_C_MDATA_BNUMx_SHIFT(1)) |
+	       ((d & SBA_C_MDATA_DNUM_MASK) << SBA_C_MDATA_DNUM_SHIFT);
+}
+
+/* ====== Channel resource management routines ===== */
+
+static struct sba_request *sba_alloc_request(struct sba_device *sba)
+{
+	unsigned long flags;
+	struct sba_request *req = NULL;
+
+	spin_lock_irqsave(&sba->reqs_lock, flags);
+
+	req = list_first_entry_or_null(&sba->reqs_free_list,
+				       struct sba_request, node);
+	if (req) {
+		list_move_tail(&req->node, &sba->reqs_alloc_list);
+		req->state = SBA_REQUEST_STATE_ALLOCED;
+		req->fence = false;
+		req->first = req;
+		INIT_LIST_HEAD(&req->next);
+		req->next_count = 1;
+		atomic_set(&req->next_pending_count, 1);
+
+		sba->reqs_free_count--;
+
+		dma_async_tx_descriptor_init(&req->tx, &sba->dma_chan);
+	}
+
+	spin_unlock_irqrestore(&sba->reqs_lock, flags);
+
+	return req;
+}
+
+/* Note: Must be called with sba->reqs_lock held */
+static void _sba_pending_request(struct sba_device *sba,
+				 struct sba_request *req)
+{
+	lockdep_assert_held(&sba->reqs_lock);
+	req->state = SBA_REQUEST_STATE_PENDING;
+	list_move_tail(&req->node, &sba->reqs_pending_list);
+	if (list_empty(&sba->reqs_active_list))
+		sba->reqs_fence = false;
+}
+
+/* Note: Must be called with sba->reqs_lock held */
+static bool _sba_active_request(struct sba_device *sba,
+				struct sba_request *req)
+{
+	lockdep_assert_held(&sba->reqs_lock);
+	if (list_empty(&sba->reqs_active_list))
+		sba->reqs_fence = false;
+	if (sba->reqs_fence)
+		return false;
+	req->state = SBA_REQUEST_STATE_ACTIVE;
+	list_move_tail(&req->node, &sba->reqs_active_list);
+	if (req->fence)
+		sba->reqs_fence = true;
+	return true;
+}
+
+/* Note: Must be called with sba->reqs_lock held */
+static void _sba_abort_request(struct sba_device *sba,
+			       struct sba_request *req)
+{
+	lockdep_assert_held(&sba->reqs_lock);
+	req->state = SBA_REQUEST_STATE_ABORTED;
+	list_move_tail(&req->node, &sba->reqs_aborted_list);
+	if (list_empty(&sba->reqs_active_list))
+		sba->reqs_fence = false;
+}
+
+/* Note: Must be called with sba->reqs_lock held */
+static void _sba_free_request(struct sba_device *sba,
+			      struct sba_request *req)
+{
+	lockdep_assert_held(&sba->reqs_lock);
+	req->state = SBA_REQUEST_STATE_FREE;
+	list_move_tail(&req->node, &sba->reqs_free_list);
+	if (list_empty(&sba->reqs_active_list))
+		sba->reqs_fence = false;
+	sba->reqs_free_count++;
+}
+
+static void sba_received_request(struct sba_request *req)
+{
+	unsigned long flags;
+	struct sba_device *sba = req->sba;
+
+	spin_lock_irqsave(&sba->reqs_lock, flags);
+	req->state = SBA_REQUEST_STATE_RECEIVED;
+	list_move_tail(&req->node, &sba->reqs_received_list);
+	spin_unlock_irqrestore(&sba->reqs_lock, flags);
+}
+
+static void sba_complete_chained_requests(struct sba_request *req)
+{
+	unsigned long flags;
+	struct sba_request *nreq;
+	struct sba_device *sba = req->sba;
+
+	spin_lock_irqsave(&sba->reqs_lock, flags);
+
+	req->state = SBA_REQUEST_STATE_COMPLETED;
+	list_move_tail(&req->node, &sba->reqs_completed_list);
+	list_for_each_entry(nreq, &req->next, next) {
+		nreq->state = SBA_REQUEST_STATE_COMPLETED;
+		list_move_tail(&nreq->node, &sba->reqs_completed_list);
+	}
+	if (list_empty(&sba->reqs_active_list))
+		sba->reqs_fence = false;
+
+	spin_unlock_irqrestore(&sba->reqs_lock, flags);
+}
+
+static void sba_free_chained_requests(struct sba_request *req)
+{
+	unsigned long flags;
+	struct sba_request *nreq;
+	struct sba_device *sba = req->sba;
+
+	spin_lock_irqsave(&sba->reqs_lock, flags);
+
+	_sba_free_request(sba, req);
+	list_for_each_entry(nreq, &req->next, next)
+		_sba_free_request(sba, nreq);
+
+	spin_unlock_irqrestore(&sba->reqs_lock, flags);
+}
+
+static void sba_chain_request(struct sba_request *first,
+			      struct sba_request *req)
+{
+	unsigned long flags;
+	struct sba_device *sba = req->sba;
+
+	spin_lock_irqsave(&sba->reqs_lock, flags);
+
+	list_add_tail(&req->next, &first->next);
+	req->first = first;
+	first->next_count++;
+	atomic_set(&first->next_pending_count, first->next_count);
+
+	spin_unlock_irqrestore(&sba->reqs_lock, flags);
+}
+
+static void sba_cleanup_nonpending_requests(struct sba_device *sba)
+{
+	unsigned long flags;
+	struct sba_request *req, *req1;
+
+	spin_lock_irqsave(&sba->reqs_lock, flags);
+
+	/* Freeup all alloced request */
+	list_for_each_entry_safe(req, req1, &sba->reqs_alloc_list, node)
+		_sba_free_request(sba, req);
+
+	/* Freeup all received request */
+	list_for_each_entry_safe(req, req1, &sba->reqs_received_list, node)
+		_sba_free_request(sba, req);
+
+	/* Freeup all completed request */
+	list_for_each_entry_safe(req, req1, &sba->reqs_completed_list, node)
+		_sba_free_request(sba, req);
+
+	/* Set all active requests as aborted */
+	list_for_each_entry_safe(req, req1, &sba->reqs_active_list, node)
+		_sba_abort_request(sba, req);
+
+	/*
+	 * Note: We expect that aborted request will be eventually
+	 * freed by sba_receive_message()
+	 */
+
+	spin_unlock_irqrestore(&sba->reqs_lock, flags);
+}
+
+static void sba_cleanup_pending_requests(struct sba_device *sba)
+{
+	unsigned long flags;
+	struct sba_request *req, *req1;
+
+	spin_lock_irqsave(&sba->reqs_lock, flags);
+
+	/* Freeup all pending request */
+	list_for_each_entry_safe(req, req1, &sba->reqs_pending_list, node)
+		_sba_free_request(sba, req);
+
+	spin_unlock_irqrestore(&sba->reqs_lock, flags);
+}
+
+/* ====== DMAENGINE callbacks ===== */
+
+static void sba_free_chan_resources(struct dma_chan *dchan)
+{
+	/*
+	 * Channel resources are pre-alloced so we just free-up
+	 * whatever we can so that we can re-use pre-alloced
+	 * channel resources next time.
+	 */
+	sba_cleanup_nonpending_requests(to_sba_device(dchan));
+}
+
+static int sba_device_terminate_all(struct dma_chan *dchan)
+{
+	/* Cleanup all pending requests */
+	sba_cleanup_pending_requests(to_sba_device(dchan));
+
+	return 0;
+}
+
+static int sba_send_mbox_request(struct sba_device *sba,
+				 struct sba_request *req)
+{
+	int mchans_idx, ret = 0;
+
+	/* Select mailbox channel in round-robin fashion */
+	mchans_idx = atomic_inc_return(&sba->mchans_current);
+	mchans_idx = mchans_idx % sba->mchans_count;
+
+	/* Send message for the request */
+	req->msg.error = 0;
+	ret = mbox_send_message(sba->mchans[mchans_idx], &req->msg);
+	if (ret < 0) {
+		dev_err(sba->dev, "send message failed with error %d", ret);
+		return ret;
+	}
+	ret = req->msg.error;
+	if (ret < 0) {
+		dev_err(sba->dev, "message error %d", ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static void sba_issue_pending(struct dma_chan *dchan)
+{
+	int ret;
+	unsigned long flags;
+	struct sba_request *req, *req1;
+	struct sba_device *sba = to_sba_device(dchan);
+
+	spin_lock_irqsave(&sba->reqs_lock, flags);
+
+	/* Process all pending request */
+	list_for_each_entry_safe(req, req1, &sba->reqs_pending_list, node) {
+		/* Try to make request active */
+		if (!_sba_active_request(sba, req))
+			break;
+
+		/* Send request to mailbox channel */
+		spin_unlock_irqrestore(&sba->reqs_lock, flags);
+		ret = sba_send_mbox_request(sba, req);
+		spin_lock_irqsave(&sba->reqs_lock, flags);
+
+		/* If something went wrong then keep request pending */
+		if (ret < 0) {
+			_sba_pending_request(sba, req);
+			break;
+		}
+	}
+
+	spin_unlock_irqrestore(&sba->reqs_lock, flags);
+}
+
+static dma_cookie_t sba_tx_submit(struct dma_async_tx_descriptor *tx)
+{
+	unsigned long flags;
+	dma_cookie_t cookie;
+	struct sba_device *sba;
+	struct sba_request *req, *nreq;
+
+	if (unlikely(!tx))
+		return -EINVAL;
+
+	sba = to_sba_device(tx->chan);
+	req = to_sba_request(tx);
+
+	/* Assign cookie and mark all chained requests pending */
+	spin_lock_irqsave(&sba->reqs_lock, flags);
+	cookie = dma_cookie_assign(tx);
+	_sba_pending_request(sba, req);
+	list_for_each_entry(nreq, &req->next, next)
+		_sba_pending_request(sba, nreq);
+	spin_unlock_irqrestore(&sba->reqs_lock, flags);
+
+	return cookie;
+}
+
+static enum dma_status sba_tx_status(struct dma_chan *dchan,
+				     dma_cookie_t cookie,
+				     struct dma_tx_state *txstate)
+{
+	int mchan_idx;
+	enum dma_status ret;
+	struct sba_device *sba = to_sba_device(dchan);
+
+	for (mchan_idx = 0; mchan_idx < sba->mchans_count; mchan_idx++)
+		mbox_client_peek_data(sba->mchans[mchan_idx]);
+
+	ret = dma_cookie_status(dchan, cookie, txstate);
+	if (ret == DMA_COMPLETE)
+		return ret;
+
+	return dma_cookie_status(dchan, cookie, txstate);
+}
+
+static void sba_fillup_interrupt_msg(struct sba_request *req,
+				     struct brcm_sba_command *cmds,
+				     struct brcm_message *msg)
+{
+	u64 cmd;
+	u32 c_mdata;
+	struct brcm_sba_command *cmdsp = cmds;
+
+	/* Type-B command to load dummy data into buf0 */
+	cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+			  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+	cmd = sba_cmd_enc(cmd, req->sba->hw_resp_size,
+			  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+	c_mdata = sba_cmd_load_c_mdata(0);
+	cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+			  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+	cmd = sba_cmd_enc(cmd, SBA_CMD_LOAD_BUFFER,
+			  SBA_CMD_SHIFT, SBA_CMD_MASK);
+	cmdsp->cmd = cmd;
+	*cmdsp->cmd_dma = cpu_to_le64(cmd);
+	cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+	cmdsp->data = req->resp_dma;
+	cmdsp->data_len = req->sba->hw_resp_size;
+	cmdsp++;
+
+	/* Type-A command to write buf0 to dummy location */
+	cmd = sba_cmd_enc(0x0, SBA_TYPE_A,
+			  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+	cmd = sba_cmd_enc(cmd, req->sba->hw_resp_size,
+			  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+	cmd = sba_cmd_enc(cmd, 0x1,
+			  SBA_RESP_SHIFT, SBA_RESP_MASK);
+	c_mdata = sba_cmd_write_c_mdata(0);
+	cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+			  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+	cmd = sba_cmd_enc(cmd, SBA_CMD_WRITE_BUFFER,
+			  SBA_CMD_SHIFT, SBA_CMD_MASK);
+	cmdsp->cmd = cmd;
+	*cmdsp->cmd_dma = cpu_to_le64(cmd);
+	cmdsp->flags = BRCM_SBA_CMD_TYPE_A;
+	if (req->sba->hw_resp_size) {
+		cmdsp->flags |= BRCM_SBA_CMD_HAS_RESP;
+		cmdsp->resp = req->resp_dma;
+		cmdsp->resp_len = req->sba->hw_resp_size;
+	}
+	cmdsp->flags |= BRCM_SBA_CMD_HAS_OUTPUT;
+	cmdsp->data = req->resp_dma;
+	cmdsp->data_len = req->sba->hw_resp_size;
+	cmdsp++;
+
+	/* Fillup brcm_message */
+	msg->type = BRCM_MESSAGE_SBA;
+	msg->sba.cmds = cmds;
+	msg->sba.cmds_count = cmdsp - cmds;
+	msg->ctx = req;
+	msg->error = 0;
+}
+
+static struct dma_async_tx_descriptor *
+sba_prep_dma_interrupt(struct dma_chan *dchan, unsigned long flags)
+{
+	struct sba_request *req = NULL;
+	struct sba_device *sba = to_sba_device(dchan);
+
+	/* Alloc new request */
+	req = sba_alloc_request(sba);
+	if (!req)
+		return NULL;
+
+	/*
+	 * Force fence so that no requests are submitted
+	 * until DMA callback for this request is invoked.
+	 */
+	req->fence = true;
+
+	/* Fillup request message */
+	sba_fillup_interrupt_msg(req, req->cmds, &req->msg);
+
+	/* Init async_tx descriptor */
+	req->tx.flags = flags;
+	req->tx.cookie = -EBUSY;
+
+	return &req->tx;
+}
+
+static void sba_fillup_memcpy_msg(struct sba_request *req,
+				  struct brcm_sba_command *cmds,
+				  struct brcm_message *msg,
+				  dma_addr_t msg_offset, size_t msg_len,
+				  dma_addr_t dst, dma_addr_t src)
+{
+	u64 cmd;
+	u32 c_mdata;
+	struct brcm_sba_command *cmdsp = cmds;
+
+	/* Type-B command to load data into buf0 */
+	cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+			  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+	cmd = sba_cmd_enc(cmd, msg_len,
+			  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+	c_mdata = sba_cmd_load_c_mdata(0);
+	cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+			  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+	cmd = sba_cmd_enc(cmd, SBA_CMD_LOAD_BUFFER,
+			  SBA_CMD_SHIFT, SBA_CMD_MASK);
+	cmdsp->cmd = cmd;
+	*cmdsp->cmd_dma = cpu_to_le64(cmd);
+	cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+	cmdsp->data = src + msg_offset;
+	cmdsp->data_len = msg_len;
+	cmdsp++;
+
+	/* Type-A command to write buf0 */
+	cmd = sba_cmd_enc(0x0, SBA_TYPE_A,
+			  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+	cmd = sba_cmd_enc(cmd, msg_len,
+			  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+	cmd = sba_cmd_enc(cmd, 0x1,
+			  SBA_RESP_SHIFT, SBA_RESP_MASK);
+	c_mdata = sba_cmd_write_c_mdata(0);
+	cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+			  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+	cmd = sba_cmd_enc(cmd, SBA_CMD_WRITE_BUFFER,
+			  SBA_CMD_SHIFT, SBA_CMD_MASK);
+	cmdsp->cmd = cmd;
+	*cmdsp->cmd_dma = cpu_to_le64(cmd);
+	cmdsp->flags = BRCM_SBA_CMD_TYPE_A;
+	if (req->sba->hw_resp_size) {
+		cmdsp->flags |= BRCM_SBA_CMD_HAS_RESP;
+		cmdsp->resp = req->resp_dma;
+		cmdsp->resp_len = req->sba->hw_resp_size;
+	}
+	cmdsp->flags |= BRCM_SBA_CMD_HAS_OUTPUT;
+	cmdsp->data = dst + msg_offset;
+	cmdsp->data_len = msg_len;
+	cmdsp++;
+
+	/* Fillup brcm_message */
+	msg->type = BRCM_MESSAGE_SBA;
+	msg->sba.cmds = cmds;
+	msg->sba.cmds_count = cmdsp - cmds;
+	msg->ctx = req;
+	msg->error = 0;
+}
+
+static struct sba_request *
+sba_prep_dma_memcpy_req(struct sba_device *sba,
+			dma_addr_t off, dma_addr_t dst, dma_addr_t src,
+			size_t len, unsigned long flags)
+{
+	struct sba_request *req = NULL;
+
+	/* Alloc new request */
+	req = sba_alloc_request(sba);
+	if (!req)
+		return NULL;
+	req->fence = (flags & DMA_PREP_FENCE) ? true : false;
+
+	/* Fillup request message */
+	sba_fillup_memcpy_msg(req, req->cmds, &req->msg,
+			      off, len, dst, src);
+
+	/* Init async_tx descriptor */
+	req->tx.flags = flags;
+	req->tx.cookie = -EBUSY;
+
+	return req;
+}
+
+static struct dma_async_tx_descriptor *
+sba_prep_dma_memcpy(struct dma_chan *dchan, dma_addr_t dst, dma_addr_t src,
+		    size_t len, unsigned long flags)
+{
+	size_t req_len;
+	dma_addr_t off = 0;
+	struct sba_device *sba = to_sba_device(dchan);
+	struct sba_request *first = NULL, *req;
+
+	/* Create chained requests where each request is upto hw_buf_size */
+	while (len) {
+		req_len = (len < sba->hw_buf_size) ? len : sba->hw_buf_size;
+
+		req = sba_prep_dma_memcpy_req(sba, off, dst, src,
+					      req_len, flags);
+		if (!req) {
+			if (first)
+				sba_free_chained_requests(first);
+			return NULL;
+		}
+
+		if (first)
+			sba_chain_request(first, req);
+		else
+			first = req;
+
+		off += req_len;
+		len -= req_len;
+	}
+
+	return (first) ? &first->tx : NULL;
+}
+
+static void sba_fillup_xor_msg(struct sba_request *req,
+				struct brcm_sba_command *cmds,
+				struct brcm_message *msg,
+				dma_addr_t msg_offset, size_t msg_len,
+				dma_addr_t dst, dma_addr_t *src, u32 src_cnt)
+{
+	u64 cmd;
+	u32 c_mdata;
+	unsigned int i;
+	struct brcm_sba_command *cmdsp = cmds;
+
+	/* Type-B command to load data into buf0 */
+	cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+			  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+	cmd = sba_cmd_enc(cmd, msg_len,
+			  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+	c_mdata = sba_cmd_load_c_mdata(0);
+	cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+			  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+	cmd = sba_cmd_enc(cmd, SBA_CMD_LOAD_BUFFER,
+			  SBA_CMD_SHIFT, SBA_CMD_MASK);
+	cmdsp->cmd = cmd;
+	*cmdsp->cmd_dma = cpu_to_le64(cmd);
+	cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+	cmdsp->data = src[0] + msg_offset;
+	cmdsp->data_len = msg_len;
+	cmdsp++;
+
+	/* Type-B commands to xor data with buf0 and put it back in buf0 */
+	for (i = 1; i < src_cnt; i++) {
+		cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+				  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+		cmd = sba_cmd_enc(cmd, msg_len,
+				  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+		c_mdata = sba_cmd_xor_c_mdata(0, 0);
+		cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+				  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_CMD_XOR,
+				  SBA_CMD_SHIFT, SBA_CMD_MASK);
+		cmdsp->cmd = cmd;
+		*cmdsp->cmd_dma = cpu_to_le64(cmd);
+		cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+		cmdsp->data = src[i] + msg_offset;
+		cmdsp->data_len = msg_len;
+		cmdsp++;
+	}
+
+	/* Type-A command to write buf0 */
+	cmd = sba_cmd_enc(0x0, SBA_TYPE_A,
+			  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+	cmd = sba_cmd_enc(cmd, msg_len,
+			  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+	cmd = sba_cmd_enc(cmd, 0x1,
+			  SBA_RESP_SHIFT, SBA_RESP_MASK);
+	c_mdata = sba_cmd_write_c_mdata(0);
+	cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+			  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+	cmd = sba_cmd_enc(cmd, SBA_CMD_WRITE_BUFFER,
+			  SBA_CMD_SHIFT, SBA_CMD_MASK);
+	cmdsp->cmd = cmd;
+	*cmdsp->cmd_dma = cpu_to_le64(cmd);
+	cmdsp->flags = BRCM_SBA_CMD_TYPE_A;
+	if (req->sba->hw_resp_size) {
+		cmdsp->flags |= BRCM_SBA_CMD_HAS_RESP;
+		cmdsp->resp = req->resp_dma;
+		cmdsp->resp_len = req->sba->hw_resp_size;
+	}
+	cmdsp->flags |= BRCM_SBA_CMD_HAS_OUTPUT;
+	cmdsp->data = dst + msg_offset;
+	cmdsp->data_len = msg_len;
+	cmdsp++;
+
+	/* Fillup brcm_message */
+	msg->type = BRCM_MESSAGE_SBA;
+	msg->sba.cmds = cmds;
+	msg->sba.cmds_count = cmdsp - cmds;
+	msg->ctx = req;
+	msg->error = 0;
+}
+
+struct sba_request *
+sba_prep_dma_xor_req(struct sba_device *sba,
+		     dma_addr_t off, dma_addr_t dst, dma_addr_t *src,
+		     u32 src_cnt, size_t len, unsigned long flags)
+{
+	struct sba_request *req = NULL;
+
+	/* Alloc new request */
+	req = sba_alloc_request(sba);
+	if (!req)
+		return NULL;
+	req->fence = (flags & DMA_PREP_FENCE) ? true : false;
+
+	/* Fillup request message */
+	sba_fillup_xor_msg(req, req->cmds, &req->msg,
+			   off, len, dst, src, src_cnt);
+
+	/* Init async_tx descriptor */
+	req->tx.flags = flags;
+	req->tx.cookie = -EBUSY;
+
+	return req;
+}
+
+static struct dma_async_tx_descriptor *
+sba_prep_dma_xor(struct dma_chan *dchan, dma_addr_t dst, dma_addr_t *src,
+		 u32 src_cnt, size_t len, unsigned long flags)
+{
+	size_t req_len;
+	dma_addr_t off = 0;
+	struct sba_device *sba = to_sba_device(dchan);
+	struct sba_request *first = NULL, *req;
+
+	/* Sanity checks */
+	if (unlikely(src_cnt > sba->max_xor_srcs))
+		return NULL;
+
+	/* Create chained requests where each request is upto hw_buf_size */
+	while (len) {
+		req_len = (len < sba->hw_buf_size) ? len : sba->hw_buf_size;
+
+		req = sba_prep_dma_xor_req(sba, off, dst, src, src_cnt,
+					   req_len, flags);
+		if (!req) {
+			if (first)
+				sba_free_chained_requests(first);
+			return NULL;
+		}
+
+		if (first)
+			sba_chain_request(first, req);
+		else
+			first = req;
+
+		off += req_len;
+		len -= req_len;
+	}
+
+	return (first) ? &first->tx : NULL;
+}
+
+static void sba_fillup_pq_msg(struct sba_request *req,
+				bool pq_continue,
+				struct brcm_sba_command *cmds,
+				struct brcm_message *msg,
+				dma_addr_t msg_offset, size_t msg_len,
+				dma_addr_t *dst_p, dma_addr_t *dst_q,
+				const u8 *scf, dma_addr_t *src, u32 src_cnt)
+{
+	u64 cmd;
+	u32 c_mdata;
+	unsigned int i;
+	struct brcm_sba_command *cmdsp = cmds;
+
+	if (pq_continue) {
+		/* Type-B command to load old P into buf0 */
+		if (dst_p) {
+			cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+				SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+			cmd = sba_cmd_enc(cmd, msg_len,
+				SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+			c_mdata = sba_cmd_load_c_mdata(0);
+			cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+				SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+			cmd = sba_cmd_enc(cmd, SBA_CMD_LOAD_BUFFER,
+				SBA_CMD_SHIFT, SBA_CMD_MASK);
+			cmdsp->cmd = cmd;
+			*cmdsp->cmd_dma = cpu_to_le64(cmd);
+			cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+			cmdsp->data = *dst_p + msg_offset;
+			cmdsp->data_len = msg_len;
+			cmdsp++;
+		}
+
+		/* Type-B command to load old Q into buf1 */
+		if (dst_q) {
+			cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+				SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+			cmd = sba_cmd_enc(cmd, msg_len,
+				SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+			c_mdata = sba_cmd_load_c_mdata(1);
+			cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+				SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+			cmd = sba_cmd_enc(cmd, SBA_CMD_LOAD_BUFFER,
+				SBA_CMD_SHIFT, SBA_CMD_MASK);
+			cmdsp->cmd = cmd;
+			*cmdsp->cmd_dma = cpu_to_le64(cmd);
+			cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+			cmdsp->data = *dst_q + msg_offset;
+			cmdsp->data_len = msg_len;
+			cmdsp++;
+		}
+	} else {
+		/* Type-A command to zero all buffers */
+		cmd = sba_cmd_enc(0x0, SBA_TYPE_A,
+				  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+		cmd = sba_cmd_enc(cmd, msg_len,
+				  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_CMD_ZERO_ALL_BUFFERS,
+				  SBA_CMD_SHIFT, SBA_CMD_MASK);
+		cmdsp->cmd = cmd;
+		*cmdsp->cmd_dma = cpu_to_le64(cmd);
+		cmdsp->flags = BRCM_SBA_CMD_TYPE_A;
+		cmdsp++;
+	}
+
+	/* Type-B commands for generate P onto buf0 and Q onto buf1 */
+	for (i = 0; i < src_cnt; i++) {
+		cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+				  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+		cmd = sba_cmd_enc(cmd, msg_len,
+				  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+		c_mdata = sba_cmd_pq_c_mdata(raid6_gflog[scf[i]], 1, 0);
+		cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+				  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_C_MDATA_MS(c_mdata),
+				  SBA_C_MDATA_MS_SHIFT, SBA_C_MDATA_MS_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_CMD_GALOIS_XOR,
+				  SBA_CMD_SHIFT, SBA_CMD_MASK);
+		cmdsp->cmd = cmd;
+		*cmdsp->cmd_dma = cpu_to_le64(cmd);
+		cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+		cmdsp->data = src[i] + msg_offset;
+		cmdsp->data_len = msg_len;
+		cmdsp++;
+	}
+
+	/* Type-A command to write buf0 */
+	if (dst_p) {
+		cmd = sba_cmd_enc(0x0, SBA_TYPE_A,
+				  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+		cmd = sba_cmd_enc(cmd, msg_len,
+				  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+		cmd = sba_cmd_enc(cmd, 0x1,
+				  SBA_RESP_SHIFT, SBA_RESP_MASK);
+		c_mdata = sba_cmd_write_c_mdata(0);
+		cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+				  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_CMD_WRITE_BUFFER,
+				  SBA_CMD_SHIFT, SBA_CMD_MASK);
+		cmdsp->cmd = cmd;
+		*cmdsp->cmd_dma = cpu_to_le64(cmd);
+		cmdsp->flags = BRCM_SBA_CMD_TYPE_A;
+		if (req->sba->hw_resp_size) {
+			cmdsp->flags |= BRCM_SBA_CMD_HAS_RESP;
+			cmdsp->resp = req->resp_dma;
+			cmdsp->resp_len = req->sba->hw_resp_size;
+		}
+		cmdsp->flags |= BRCM_SBA_CMD_HAS_OUTPUT;
+		cmdsp->data = *dst_p + msg_offset;
+		cmdsp->data_len = msg_len;
+		cmdsp++;
+	}
+
+	/* Type-A command to write buf1 */
+	if (dst_q) {
+		cmd = sba_cmd_enc(0x0, SBA_TYPE_A,
+				  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+		cmd = sba_cmd_enc(cmd, msg_len,
+				  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+		cmd = sba_cmd_enc(cmd, 0x1,
+				  SBA_RESP_SHIFT, SBA_RESP_MASK);
+		c_mdata = sba_cmd_write_c_mdata(1);
+		cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+				  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_CMD_WRITE_BUFFER,
+				  SBA_CMD_SHIFT, SBA_CMD_MASK);
+		cmdsp->cmd = cmd;
+		*cmdsp->cmd_dma = cpu_to_le64(cmd);
+		cmdsp->flags = BRCM_SBA_CMD_TYPE_A;
+		if (req->sba->hw_resp_size) {
+			cmdsp->flags |= BRCM_SBA_CMD_HAS_RESP;
+			cmdsp->resp = req->resp_dma;
+			cmdsp->resp_len = req->sba->hw_resp_size;
+		}
+		cmdsp->flags |= BRCM_SBA_CMD_HAS_OUTPUT;
+		cmdsp->data = *dst_q + msg_offset;
+		cmdsp->data_len = msg_len;
+		cmdsp++;
+	}
+
+	/* Fillup brcm_message */
+	msg->type = BRCM_MESSAGE_SBA;
+	msg->sba.cmds = cmds;
+	msg->sba.cmds_count = cmdsp - cmds;
+	msg->ctx = req;
+	msg->error = 0;
+}
+
+struct sba_request *
+sba_prep_dma_pq_req(struct sba_device *sba, dma_addr_t off,
+		    dma_addr_t *dst_p, dma_addr_t *dst_q, dma_addr_t *src,
+		    u32 src_cnt, const u8 *scf, size_t len, unsigned long flags)
+{
+	struct sba_request *req = NULL;
+
+	/* Alloc new request */
+	req = sba_alloc_request(sba);
+	if (!req)
+		return NULL;
+	req->fence = (flags & DMA_PREP_FENCE) ? true : false;
+
+	/* Fillup request messages */
+	sba_fillup_pq_msg(req, dmaf_continue(flags),
+			  req->cmds, &req->msg,
+			  off, len, dst_p, dst_q, scf, src, src_cnt);
+
+	/* Init async_tx descriptor */
+	req->tx.flags = flags;
+	req->tx.cookie = -EBUSY;
+
+	return req;
+}
+
+static void sba_fillup_pq_single_msg(struct sba_request *req,
+				bool pq_continue,
+				struct brcm_sba_command *cmds,
+				struct brcm_message *msg,
+				dma_addr_t msg_offset, size_t msg_len,
+				dma_addr_t *dst_p, dma_addr_t *dst_q,
+				dma_addr_t src, u8 scf)
+{
+	u64 cmd;
+	u32 c_mdata;
+	u8 pos, dpos = raid6_gflog[scf];
+	struct brcm_sba_command *cmdsp = cmds;
+
+	if (!dst_p)
+		goto skip_p;
+
+	if (pq_continue) {
+		/* Type-B command to load old P into buf0 */
+		cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+				  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+		cmd = sba_cmd_enc(cmd, msg_len,
+				  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+		c_mdata = sba_cmd_load_c_mdata(0);
+		cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+				  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_CMD_LOAD_BUFFER,
+				  SBA_CMD_SHIFT, SBA_CMD_MASK);
+		cmdsp->cmd = cmd;
+		*cmdsp->cmd_dma = cpu_to_le64(cmd);
+		cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+		cmdsp->data = *dst_p + msg_offset;
+		cmdsp->data_len = msg_len;
+		cmdsp++;
+
+		/*
+		 * Type-B commands to xor data with buf0 and put it
+		 * back in buf0
+		 */
+		cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+				  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+		cmd = sba_cmd_enc(cmd, msg_len,
+				  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+		c_mdata = sba_cmd_xor_c_mdata(0, 0);
+		cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+				  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_CMD_XOR,
+				  SBA_CMD_SHIFT, SBA_CMD_MASK);
+		cmdsp->cmd = cmd;
+		*cmdsp->cmd_dma = cpu_to_le64(cmd);
+		cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+		cmdsp->data = src + msg_offset;
+		cmdsp->data_len = msg_len;
+		cmdsp++;
+	} else {
+		/* Type-B command to load old P into buf0 */
+		cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+				  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+		cmd = sba_cmd_enc(cmd, msg_len,
+				  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+		c_mdata = sba_cmd_load_c_mdata(0);
+		cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+				  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_CMD_LOAD_BUFFER,
+				  SBA_CMD_SHIFT, SBA_CMD_MASK);
+		cmdsp->cmd = cmd;
+		*cmdsp->cmd_dma = cpu_to_le64(cmd);
+		cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+		cmdsp->data = src + msg_offset;
+		cmdsp->data_len = msg_len;
+		cmdsp++;
+	}
+
+	/* Type-A command to write buf0 */
+	cmd = sba_cmd_enc(0x0, SBA_TYPE_A,
+			  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+	cmd = sba_cmd_enc(cmd, msg_len,
+			  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+	cmd = sba_cmd_enc(cmd, 0x1,
+			  SBA_RESP_SHIFT, SBA_RESP_MASK);
+	c_mdata = sba_cmd_write_c_mdata(0);
+	cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+			  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+	cmd = sba_cmd_enc(cmd, SBA_CMD_WRITE_BUFFER,
+			  SBA_CMD_SHIFT, SBA_CMD_MASK);
+	cmdsp->cmd = cmd;
+	*cmdsp->cmd_dma = cpu_to_le64(cmd);
+	cmdsp->flags = BRCM_SBA_CMD_TYPE_A;
+	if (req->sba->hw_resp_size) {
+		cmdsp->flags |= BRCM_SBA_CMD_HAS_RESP;
+		cmdsp->resp = req->resp_dma;
+		cmdsp->resp_len = req->sba->hw_resp_size;
+	}
+	cmdsp->flags |= BRCM_SBA_CMD_HAS_OUTPUT;
+	cmdsp->data = *dst_p + msg_offset;
+	cmdsp->data_len = msg_len;
+	cmdsp++;
+
+skip_p:
+	if (!dst_q)
+		goto skip_q;
+
+	/* Type-A command to zero all buffers */
+	cmd = sba_cmd_enc(0x0, SBA_TYPE_A,
+			  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+	cmd = sba_cmd_enc(cmd, msg_len,
+			  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+	cmd = sba_cmd_enc(cmd, SBA_CMD_ZERO_ALL_BUFFERS,
+			  SBA_CMD_SHIFT, SBA_CMD_MASK);
+	cmdsp->cmd = cmd;
+	*cmdsp->cmd_dma = cpu_to_le64(cmd);
+	cmdsp->flags = BRCM_SBA_CMD_TYPE_A;
+	cmdsp++;
+
+	if (dpos == 255)
+		goto skip_q_computation;
+	pos = (dpos < req->sba->max_pq_coefs) ?
+		dpos : (req->sba->max_pq_coefs - 1);
+
+	/*
+	 * Type-B command to generate initial Q from data
+	 * and store output into buf0
+	 */
+	cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+			  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+	cmd = sba_cmd_enc(cmd, msg_len,
+			  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+	c_mdata = sba_cmd_pq_c_mdata(pos, 0, 0);
+	cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+			  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+	cmd = sba_cmd_enc(cmd, SBA_C_MDATA_MS(c_mdata),
+			  SBA_C_MDATA_MS_SHIFT, SBA_C_MDATA_MS_MASK);
+	cmd = sba_cmd_enc(cmd, SBA_CMD_GALOIS,
+			  SBA_CMD_SHIFT, SBA_CMD_MASK);
+	cmdsp->cmd = cmd;
+	*cmdsp->cmd_dma = cpu_to_le64(cmd);
+	cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+	cmdsp->data = src + msg_offset;
+	cmdsp->data_len = msg_len;
+	cmdsp++;
+
+	dpos -= pos;
+
+	/* Multiple Type-A command to generate final Q */
+	while (dpos) {
+		pos = (dpos < req->sba->max_pq_coefs) ?
+			dpos : (req->sba->max_pq_coefs - 1);
+
+		/*
+		 * Type-A command to generate Q with buf0 and
+		 * buf1 store result in buf0
+		 */
+		cmd = sba_cmd_enc(0x0, SBA_TYPE_A,
+				  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+		cmd = sba_cmd_enc(cmd, msg_len,
+				  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+		c_mdata = sba_cmd_pq_c_mdata(pos, 0, 1);
+		cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+				  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_C_MDATA_MS(c_mdata),
+				  SBA_C_MDATA_MS_SHIFT, SBA_C_MDATA_MS_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_CMD_GALOIS,
+				  SBA_CMD_SHIFT, SBA_CMD_MASK);
+		cmdsp->cmd = cmd;
+		*cmdsp->cmd_dma = cpu_to_le64(cmd);
+		cmdsp->flags = BRCM_SBA_CMD_TYPE_A;
+		cmdsp++;
+
+		dpos -= pos;
+	}
+
+skip_q_computation:
+	if (pq_continue) {
+		/*
+		 * Type-B command to XOR previous output with
+		 * buf0 and write it into buf0
+		 */
+		cmd = sba_cmd_enc(0x0, SBA_TYPE_B,
+				  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+		cmd = sba_cmd_enc(cmd, msg_len,
+				  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+		c_mdata = sba_cmd_xor_c_mdata(0, 0);
+		cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+				  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+		cmd = sba_cmd_enc(cmd, SBA_CMD_XOR,
+				  SBA_CMD_SHIFT, SBA_CMD_MASK);
+		cmdsp->cmd = cmd;
+		*cmdsp->cmd_dma = cpu_to_le64(cmd);
+		cmdsp->flags = BRCM_SBA_CMD_TYPE_B;
+		cmdsp->data = *dst_q + msg_offset;
+		cmdsp->data_len = msg_len;
+		cmdsp++;
+	}
+
+	/* Type-A command to write buf0 */
+	cmd = sba_cmd_enc(0x0, SBA_TYPE_A,
+			  SBA_TYPE_SHIFT, SBA_TYPE_MASK);
+	cmd = sba_cmd_enc(cmd, msg_len,
+			  SBA_USER_DEF_SHIFT, SBA_USER_DEF_MASK);
+	cmd = sba_cmd_enc(cmd, 0x1,
+			  SBA_RESP_SHIFT, SBA_RESP_MASK);
+	c_mdata = sba_cmd_write_c_mdata(0);
+	cmd = sba_cmd_enc(cmd, SBA_C_MDATA_LS(c_mdata),
+			  SBA_C_MDATA_SHIFT, SBA_C_MDATA_MASK);
+	cmd = sba_cmd_enc(cmd, SBA_CMD_WRITE_BUFFER,
+			  SBA_CMD_SHIFT, SBA_CMD_MASK);
+	cmdsp->cmd = cmd;
+	*cmdsp->cmd_dma = cpu_to_le64(cmd);
+	cmdsp->flags = BRCM_SBA_CMD_TYPE_A;
+	if (req->sba->hw_resp_size) {
+		cmdsp->flags |= BRCM_SBA_CMD_HAS_RESP;
+		cmdsp->resp = req->resp_dma;
+		cmdsp->resp_len = req->sba->hw_resp_size;
+	}
+	cmdsp->flags |= BRCM_SBA_CMD_HAS_OUTPUT;
+	cmdsp->data = *dst_q + msg_offset;
+	cmdsp->data_len = msg_len;
+	cmdsp++;
+
+skip_q:
+	/* Fillup brcm_message */
+	msg->type = BRCM_MESSAGE_SBA;
+	msg->sba.cmds = cmds;
+	msg->sba.cmds_count = cmdsp - cmds;
+	msg->ctx = req;
+	msg->error = 0;
+}
+
+struct sba_request *
+sba_prep_dma_pq_single_req(struct sba_device *sba, dma_addr_t off,
+			   dma_addr_t *dst_p, dma_addr_t *dst_q,
+			   dma_addr_t src, u8 scf, size_t len,
+			   unsigned long flags)
+{
+	struct sba_request *req = NULL;
+
+	/* Alloc new request */
+	req = sba_alloc_request(sba);
+	if (!req)
+		return NULL;
+	req->fence = (flags & DMA_PREP_FENCE) ? true : false;
+
+	/* Fillup request messages */
+	sba_fillup_pq_single_msg(req,  dmaf_continue(flags),
+				 req->cmds, &req->msg, off, len,
+				 dst_p, dst_q, src, scf);
+
+	/* Init async_tx descriptor */
+	req->tx.flags = flags;
+	req->tx.cookie = -EBUSY;
+
+	return req;
+}
+
+static struct dma_async_tx_descriptor *
+sba_prep_dma_pq(struct dma_chan *dchan, dma_addr_t *dst, dma_addr_t *src,
+		u32 src_cnt, const u8 *scf, size_t len, unsigned long flags)
+{
+	u32 i, dst_q_index;
+	size_t req_len;
+	bool slow = false;
+	dma_addr_t off = 0;
+	dma_addr_t *dst_p = NULL, *dst_q = NULL;
+	struct sba_device *sba = to_sba_device(dchan);
+	struct sba_request *first = NULL, *req;
+
+	/* Sanity checks */
+	if (unlikely(src_cnt > sba->max_pq_srcs))
+		return NULL;
+	for (i = 0; i < src_cnt; i++)
+		if (sba->max_pq_coefs <= raid6_gflog[scf[i]])
+			slow = true;
+
+	/* Figure-out P and Q destination addresses */
+	if (!(flags & DMA_PREP_PQ_DISABLE_P))
+		dst_p = &dst[0];
+	if (!(flags & DMA_PREP_PQ_DISABLE_Q))
+		dst_q = &dst[1];
+
+	/* Create chained requests where each request is upto hw_buf_size */
+	while (len) {
+		req_len = (len < sba->hw_buf_size) ? len : sba->hw_buf_size;
+
+		if (slow) {
+			dst_q_index = src_cnt;
+
+			if (dst_q) {
+				for (i = 0; i < src_cnt; i++) {
+					if (*dst_q == src[i]) {
+						dst_q_index = i;
+						break;
+					}
+				}
+			}
+
+			if (dst_q_index < src_cnt) {
+				i = dst_q_index;
+				req = sba_prep_dma_pq_single_req(sba,
+					off, dst_p, dst_q, src[i], scf[i],
+					req_len, flags | DMA_PREP_FENCE);
+				if (!req)
+					goto fail;
+
+				if (first)
+					sba_chain_request(first, req);
+				else
+					first = req;
+
+				flags |= DMA_PREP_CONTINUE;
+			}
+
+			for (i = 0; i < src_cnt; i++) {
+				if (dst_q_index == i)
+					continue;
+
+				req = sba_prep_dma_pq_single_req(sba,
+					off, dst_p, dst_q, src[i], scf[i],
+					req_len, flags | DMA_PREP_FENCE);
+				if (!req)
+					goto fail;
+
+				if (first)
+					sba_chain_request(first, req);
+				else
+					first = req;
+
+				flags |= DMA_PREP_CONTINUE;
+			}
+		} else {
+			req = sba_prep_dma_pq_req(sba, off,
+						  dst_p, dst_q, src, src_cnt,
+						  scf, req_len, flags);
+			if (!req)
+				goto fail;
+
+			if (first)
+				sba_chain_request(first, req);
+			else
+				first = req;
+		}
+
+		off += req_len;
+		len -= req_len;
+	}
+
+	return (first) ? &first->tx : NULL;
+
+fail:
+	if (first)
+		sba_free_chained_requests(first);
+	return NULL;
+}
+
+/* ====== Mailbox callbacks ===== */
+
+static void sba_dma_tx_actions(struct sba_request *req)
+{
+	struct dma_async_tx_descriptor *tx = &req->tx;
+
+	WARN_ON(tx->cookie < 0);
+
+	if (tx->cookie > 0) {
+		dma_cookie_complete(tx);
+
+		/*
+		 * Call the callback (must not sleep or submit new
+		 * operations to this channel)
+		 */
+		if (tx->callback)
+			tx->callback(tx->callback_param);
+
+		dma_descriptor_unmap(tx);
+	}
+
+	/* Run dependent operations */
+	dma_run_dependencies(tx);
+
+	/* If waiting for 'ack' then move to completed list */
+	if (!async_tx_test_ack(&req->tx))
+		sba_complete_chained_requests(req);
+	else
+		sba_free_chained_requests(req);
+}
+
+static void sba_receive_message(struct mbox_client *cl, void *msg)
+{
+	unsigned long flags;
+	struct brcm_message *m = msg;
+	struct sba_request *req = m->ctx, *req1;
+	struct sba_device *sba = req->sba;
+
+	/* Error count if message has error */
+	if (m->error < 0)
+		dev_err(sba->dev, "%s got message with error %d",
+			dma_chan_name(&sba->dma_chan), m->error);
+
+	/* Mark request as received */
+	sba_received_request(req);
+
+	/* Wait for all chained requests to be completed */
+	if (atomic_dec_return(&req->first->next_pending_count))
+		goto done;
+
+	/* Point to first request */
+	req = req->first;
+
+	/* Update request */
+	if (req->state == SBA_REQUEST_STATE_RECEIVED)
+		sba_dma_tx_actions(req);
+	else
+		sba_free_chained_requests(req);
+
+	spin_lock_irqsave(&sba->reqs_lock, flags);
+
+	/* Re-check all completed request waiting for 'ack' */
+	list_for_each_entry_safe(req, req1, &sba->reqs_completed_list, node) {
+		spin_unlock_irqrestore(&sba->reqs_lock, flags);
+		sba_dma_tx_actions(req);
+		spin_lock_irqsave(&sba->reqs_lock, flags);
+	}
+
+	spin_unlock_irqrestore(&sba->reqs_lock, flags);
+
+done:
+	/* Try to submit pending request */
+	sba_issue_pending(&sba->dma_chan);
+}
+
+/* ====== Platform driver routines ===== */
+
+static int sba_prealloc_channel_resources(struct sba_device *sba)
+{
+	int i, j, p, ret = 0;
+	struct sba_request *req = NULL;
+
+	sba->resp_base = dma_alloc_coherent(sba->dma_dev.dev,
+					    sba->max_resp_pool_size,
+					    &sba->resp_dma_base, GFP_KERNEL);
+	if (!sba->resp_base)
+		return -ENOMEM;
+
+	sba->cmds_base = dma_alloc_coherent(sba->dma_dev.dev,
+					    sba->max_cmds_pool_size,
+					    &sba->cmds_dma_base, GFP_KERNEL);
+	if (!sba->cmds_base) {
+		ret = -ENOMEM;
+		goto fail_free_resp_pool;
+	}
+
+	spin_lock_init(&sba->reqs_lock);
+	sba->reqs_fence = false;
+	INIT_LIST_HEAD(&sba->reqs_alloc_list);
+	INIT_LIST_HEAD(&sba->reqs_pending_list);
+	INIT_LIST_HEAD(&sba->reqs_active_list);
+	INIT_LIST_HEAD(&sba->reqs_received_list);
+	INIT_LIST_HEAD(&sba->reqs_completed_list);
+	INIT_LIST_HEAD(&sba->reqs_aborted_list);
+	INIT_LIST_HEAD(&sba->reqs_free_list);
+
+	sba->reqs = devm_kcalloc(sba->dev, sba->max_req,
+				 sizeof(*req), GFP_KERNEL);
+	if (!sba->reqs) {
+		ret = -ENOMEM;
+		goto fail_free_cmds_pool;
+	}
+
+	for (i = 0, p = 0; i < sba->max_req; i++) {
+		req = &sba->reqs[i];
+		INIT_LIST_HEAD(&req->node);
+		req->sba = sba;
+		req->state = SBA_REQUEST_STATE_FREE;
+		INIT_LIST_HEAD(&req->next);
+		req->next_count = 1;
+		atomic_set(&req->next_pending_count, 0);
+		req->fence = false;
+		req->resp = sba->resp_base + p;
+		req->resp_dma = sba->resp_dma_base + p;
+		p += sba->hw_resp_size;
+		req->cmds = devm_kcalloc(sba->dev, sba->max_cmd_per_req,
+					 sizeof(*req->cmds), GFP_KERNEL);
+		if (!req->cmds) {
+			ret = -ENOMEM;
+			goto fail_free_cmds_pool;
+		}
+		for (j = 0; j < sba->max_cmd_per_req; j++) {
+			req->cmds[j].cmd = 0;
+			req->cmds[j].cmd_dma = sba->cmds_base +
+				(i * sba->max_cmd_per_req + j) * sizeof(u64);
+			req->cmds[j].cmd_dma_addr = sba->cmds_dma_base +
+				(i * sba->max_cmd_per_req + j) * sizeof(u64);
+			req->cmds[j].flags = 0;
+		}
+		memset(&req->msg, 0, sizeof(req->msg));
+		dma_async_tx_descriptor_init(&req->tx, &sba->dma_chan);
+		req->tx.tx_submit = sba_tx_submit;
+		req->tx.phys = req->resp_dma;
+		list_add_tail(&req->node, &sba->reqs_free_list);
+	}
+
+	sba->reqs_free_count = sba->max_req;
+
+	return 0;
+
+fail_free_cmds_pool:
+	dma_free_coherent(sba->dma_dev.dev,
+			  sba->max_cmds_pool_size,
+			  sba->cmds_base, sba->cmds_dma_base);
+fail_free_resp_pool:
+	dma_free_coherent(sba->dma_dev.dev,
+			  sba->max_resp_pool_size,
+			  sba->resp_base, sba->resp_dma_base);
+	return ret;
+}
+
+static void sba_freeup_channel_resources(struct sba_device *sba)
+{
+	dmaengine_terminate_all(&sba->dma_chan);
+	dma_free_coherent(sba->dma_dev.dev, sba->max_cmds_pool_size,
+			  sba->cmds_base, sba->cmds_dma_base);
+	dma_free_coherent(sba->dma_dev.dev, sba->max_resp_pool_size,
+			  sba->resp_base, sba->resp_dma_base);
+	sba->resp_base = NULL;
+	sba->resp_dma_base = 0;
+}
+
+static int sba_async_register(struct sba_device *sba)
+{
+	int ret;
+	struct dma_device *dma_dev = &sba->dma_dev;
+
+	/* Initialize DMA channel cookie */
+	sba->dma_chan.device = dma_dev;
+	dma_cookie_init(&sba->dma_chan);
+
+	/* Initialize DMA device capability mask */
+	dma_cap_zero(dma_dev->cap_mask);
+	dma_cap_set(DMA_INTERRUPT, dma_dev->cap_mask);
+	dma_cap_set(DMA_MEMCPY, dma_dev->cap_mask);
+	dma_cap_set(DMA_XOR, dma_dev->cap_mask);
+	dma_cap_set(DMA_PQ, dma_dev->cap_mask);
+
+	/*
+	 * Set mailbox channel device as the base device of
+	 * our dma_device because the actual memory accesses
+	 * will be done by mailbox controller
+	 */
+	dma_dev->dev = sba->mbox_dev;
+
+	/* Set base prep routines */
+	dma_dev->device_free_chan_resources = sba_free_chan_resources;
+	dma_dev->device_terminate_all = sba_device_terminate_all;
+	dma_dev->device_issue_pending = sba_issue_pending;
+	dma_dev->device_tx_status = sba_tx_status;
+
+	/* Set interrupt routine */
+	if (dma_has_cap(DMA_INTERRUPT, dma_dev->cap_mask))
+		dma_dev->device_prep_dma_interrupt = sba_prep_dma_interrupt;
+
+	/* Set memcpy routine */
+	if (dma_has_cap(DMA_MEMCPY, dma_dev->cap_mask))
+		dma_dev->device_prep_dma_memcpy = sba_prep_dma_memcpy;
+
+	/* Set xor routine and capability */
+	if (dma_has_cap(DMA_XOR, dma_dev->cap_mask)) {
+		dma_dev->device_prep_dma_xor = sba_prep_dma_xor;
+		dma_dev->max_xor = sba->max_xor_srcs;
+	}
+
+	/* Set pq routine and capability */
+	if (dma_has_cap(DMA_PQ, dma_dev->cap_mask)) {
+		dma_dev->device_prep_dma_pq = sba_prep_dma_pq;
+		dma_set_maxpq(dma_dev, sba->max_pq_srcs, 0);
+	}
+
+	/* Initialize DMA device channel list */
+	INIT_LIST_HEAD(&dma_dev->channels);
+	list_add_tail(&sba->dma_chan.device_node, &dma_dev->channels);
+
+	/* Register with Linux async DMA framework*/
+	ret = dma_async_device_register(dma_dev);
+	if (ret) {
+		dev_err(sba->dev, "async device register error %d", ret);
+		return ret;
+	}
+
+	dev_info(sba->dev, "%s capabilities: %s%s%s%s\n",
+	dma_chan_name(&sba->dma_chan),
+	dma_has_cap(DMA_INTERRUPT, dma_dev->cap_mask) ? "interrupt " : "",
+	dma_has_cap(DMA_MEMCPY, dma_dev->cap_mask) ? "memcpy " : "",
+	dma_has_cap(DMA_XOR, dma_dev->cap_mask) ? "xor " : "",
+	dma_has_cap(DMA_PQ, dma_dev->cap_mask) ? "pq " : "");
+
+	return 0;
+}
+
+static int sba_probe(struct platform_device *pdev)
+{
+	int i, ret = 0, mchans_count;
+	struct sba_device *sba;
+	struct platform_device *mbox_pdev;
+	struct of_phandle_args args;
+
+	/* Allocate main SBA struct */
+	sba = devm_kzalloc(&pdev->dev, sizeof(*sba), GFP_KERNEL);
+	if (!sba)
+		return -ENOMEM;
+
+	sba->dev = &pdev->dev;
+	platform_set_drvdata(pdev, sba);
+
+	/* Determine SBA version from DT compatible string */
+	if (of_device_is_compatible(sba->dev->of_node, "brcm,iproc-sba"))
+		sba->ver = SBA_VER_1;
+	else if (of_device_is_compatible(sba->dev->of_node,
+					 "brcm,iproc-sba-v2"))
+		sba->ver = SBA_VER_2;
+	else
+		return -ENODEV;
+
+	/* Derived Configuration parameters */
+	switch (sba->ver) {
+	case SBA_VER_1:
+		sba->max_req = 1024;
+		sba->hw_buf_size = 4096;
+		sba->hw_resp_size = 8;
+		sba->max_pq_coefs = 6;
+		sba->max_pq_srcs = 6;
+		break;
+	case SBA_VER_2:
+		sba->max_req = 1024;
+		sba->hw_buf_size = 4096;
+		sba->hw_resp_size = 8;
+		sba->max_pq_coefs = 30;
+		/*
+		 * We can support max_pq_srcs == max_pq_coefs because
+		 * we are limited by number of SBA commands that we can
+		 * fit in one message for underlying ring manager HW.
+		 */
+		sba->max_pq_srcs = 12;
+		break;
+	default:
+		return -EINVAL;
+	}
+	sba->max_cmd_per_req = sba->max_pq_srcs + 3;
+	sba->max_xor_srcs = sba->max_cmd_per_req - 1;
+	sba->max_resp_pool_size = sba->max_req * sba->hw_resp_size;
+	sba->max_cmds_pool_size = sba->max_req *
+				  sba->max_cmd_per_req * sizeof(u64);
+
+	/* Setup mailbox client */
+	sba->client.dev			= &pdev->dev;
+	sba->client.rx_callback		= sba_receive_message;
+	sba->client.tx_block		= false;
+	sba->client.knows_txdone	= false;
+	sba->client.tx_tout		= 0;
+
+	/* Number of channels equals number of mailbox channels */
+	ret = of_count_phandle_with_args(pdev->dev.of_node,
+					 "mboxes", "#mbox-cells");
+	if (ret <= 0)
+		return -ENODEV;
+	mchans_count = ret;
+	sba->mchans_count = 0;
+	atomic_set(&sba->mchans_current, 0);
+
+	/* Allocate mailbox channel array */
+	sba->mchans = devm_kcalloc(&pdev->dev, sba->mchans_count,
+				   sizeof(*sba->mchans), GFP_KERNEL);
+	if (!sba->mchans)
+		return -ENOMEM;
+
+	/* Request mailbox channels */
+	for (i = 0; i < mchans_count; i++) {
+		sba->mchans[i] = mbox_request_channel(&sba->client, i);
+		if (IS_ERR(sba->mchans[i])) {
+			ret = PTR_ERR(sba->mchans[i]);
+			goto fail_free_mchans;
+		}
+		sba->mchans_count++;
+	}
+
+	/* Find-out underlying mailbox device */
+	ret = of_parse_phandle_with_args(pdev->dev.of_node,
+					 "mboxes", "#mbox-cells", 0, &args);
+	if (ret)
+		goto fail_free_mchans;
+	mbox_pdev = of_find_device_by_node(args.np);
+	of_node_put(args.np);
+	if (!mbox_pdev) {
+		ret = -ENODEV;
+		goto fail_free_mchans;
+	}
+	sba->mbox_dev = &mbox_pdev->dev;
+
+	/* All mailbox channels should be of same ring manager device */
+	for (i = 1; i < mchans_count; i++) {
+		ret = of_parse_phandle_with_args(pdev->dev.of_node,
+					 "mboxes", "#mbox-cells", i, &args);
+		if (ret)
+			goto fail_free_mchans;
+		mbox_pdev = of_find_device_by_node(args.np);
+		of_node_put(args.np);
+		if (sba->mbox_dev != &mbox_pdev->dev) {
+			ret = -EINVAL;
+			goto fail_free_mchans;
+		}
+	}
+
+	/* Register DMA device with linux async framework */
+	ret = sba_async_register(sba);
+	if (ret)
+		goto fail_free_mchans;
+
+	/* Prealloc channel resource */
+	ret = sba_prealloc_channel_resources(sba);
+	if (ret)
+		goto fail_async_dev_unreg;
+
+	/* Print device info */
+	dev_info(sba->dev, "%s using SBAv%d and %d mailbox channels",
+		 dma_chan_name(&sba->dma_chan), sba->ver+1,
+		 sba->mchans_count);
+
+	return 0;
+
+fail_async_dev_unreg:
+	dma_async_device_unregister(&sba->dma_dev);
+fail_free_mchans:
+	for (i = 0; i < sba->mchans_count; i++)
+		mbox_free_channel(sba->mchans[i]);
+	return ret;
+}
+
+static int sba_remove(struct platform_device *pdev)
+{
+	int i;
+	struct sba_device *sba = platform_get_drvdata(pdev);
+
+	sba_freeup_channel_resources(sba);
+
+	dma_async_device_unregister(&sba->dma_dev);
+
+	for (i = 0; i < sba->mchans_count; i++)
+		mbox_free_channel(sba->mchans[i]);
+
+	return 0;
+}
+
+static const struct of_device_id sba_of_match[] = {
+	{ .compatible = "brcm,iproc-sba", },
+	{ .compatible = "brcm,iproc-sba-v2", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, sba_of_match);
+
+static struct platform_driver sba_driver = {
+	.probe = sba_probe,
+	.remove = sba_remove,
+	.driver = {
+		.name = "bcm-sba-raid",
+		.of_match_table = sba_of_match,
+	},
+};
+module_platform_driver(sba_driver);
+
+MODULE_DESCRIPTION("Broadcom SBA RAID driver");
+MODULE_AUTHOR("Anup Patel <anup.patel@broadcom.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/dma/dw/platform.c b/drivers/dma/dw/platform.c
index c639c60b825a..bc31fe802061 100644
--- a/drivers/dma/dw/platform.c
+++ b/drivers/dma/dw/platform.c
@@ -306,8 +306,12 @@ static int dw_resume_early(struct device *dev)
 {
 	struct platform_device *pdev = to_platform_device(dev);
 	struct dw_dma_chip *chip = platform_get_drvdata(pdev);
+	int ret;
+
+	ret = clk_prepare_enable(chip->clk);
+	if (ret)
+		return ret;
 
-	clk_prepare_enable(chip->clk);
 	return dw_dma_enable(chip);
 }
 
diff --git a/drivers/dma/fsl_raid.c b/drivers/dma/fsl_raid.c
index 90d29f90acfb..493dc6c59d1d 100644
--- a/drivers/dma/fsl_raid.c
+++ b/drivers/dma/fsl_raid.c
@@ -877,7 +877,7 @@ static int fsl_re_remove(struct platform_device *ofdev)
 	return 0;
 }
 
-static struct of_device_id fsl_re_ids[] = {
+static const struct of_device_id fsl_re_ids[] = {
 	{ .compatible = "fsl,raideng-v1.0", },
 	{}
 };
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index 51c75bf2b9b6..3b8b752ede2d 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -269,6 +269,7 @@ static void fsl_chan_set_src_loop_size(struct fsldma_chan *chan, int size)
 	case 2:
 	case 4:
 	case 8:
+		mode &= ~FSL_DMA_MR_SAHTS_MASK;
 		mode |= FSL_DMA_MR_SAHE | (__ilog2(size) << 14);
 		break;
 	}
@@ -301,6 +302,7 @@ static void fsl_chan_set_dst_loop_size(struct fsldma_chan *chan, int size)
 	case 2:
 	case 4:
 	case 8:
+		mode &= ~FSL_DMA_MR_DAHTS_MASK;
 		mode |= FSL_DMA_MR_DAHE | (__ilog2(size) << 16);
 		break;
 	}
@@ -327,7 +329,8 @@ static void fsl_chan_set_request_count(struct fsldma_chan *chan, int size)
 	BUG_ON(size > 1024);
 
 	mode = get_mr(chan);
-	mode |= (__ilog2(size) << 24) & 0x0f000000;
+	mode &= ~FSL_DMA_MR_BWC_MASK;
+	mode |= (__ilog2(size) << 24) & FSL_DMA_MR_BWC_MASK;
 
 	set_mr(chan, mode);
 }
diff --git a/drivers/dma/fsldma.h b/drivers/dma/fsldma.h
index 31bffccdcc75..4787d485dd76 100644
--- a/drivers/dma/fsldma.h
+++ b/drivers/dma/fsldma.h
@@ -36,6 +36,10 @@
 #define FSL_DMA_MR_DAHE		0x00002000
 #define FSL_DMA_MR_SAHE		0x00001000
 
+#define FSL_DMA_MR_SAHTS_MASK	0x0000C000
+#define FSL_DMA_MR_DAHTS_MASK	0x00030000
+#define FSL_DMA_MR_BWC_MASK	0x0f000000
+
 /*
  * Bandwidth/pause control determines how many bytes a given
  * channel is allowed to transfer before the DMA engine pauses
diff --git a/drivers/dma/imx-dma.c b/drivers/dma/imx-dma.c
index ab0fb804fb1e..f681df8f0ed3 100644
--- a/drivers/dma/imx-dma.c
+++ b/drivers/dma/imx-dma.c
@@ -888,7 +888,7 @@ static struct dma_async_tx_descriptor *imxdma_prep_dma_cyclic(
 	sg_init_table(imxdmac->sg_list, periods);
 
 	for (i = 0; i < periods; i++) {
-		imxdmac->sg_list[i].page_link = 0;
+		sg_assign_page(&imxdmac->sg_list[i], NULL);
 		imxdmac->sg_list[i].offset = 0;
 		imxdmac->sg_list[i].dma_address = dma_addr;
 		sg_dma_len(&imxdmac->sg_list[i]) = period_len;
@@ -896,10 +896,7 @@ static struct dma_async_tx_descriptor *imxdma_prep_dma_cyclic(
 	}
 
 	/* close the loop */
-	imxdmac->sg_list[periods].offset = 0;
-	sg_dma_len(&imxdmac->sg_list[periods]) = 0;
-	imxdmac->sg_list[periods].page_link =
-		((unsigned long)imxdmac->sg_list | 0x01) & ~0x02;
+	sg_chain(imxdmac->sg_list, periods + 1, imxdmac->sg_list);
 
 	desc->type = IMXDMA_DESC_CYCLIC;
 	desc->sg = imxdmac->sg_list;
diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c
index 085993cb2ccc..a67ec1bdc4e0 100644
--- a/drivers/dma/imx-sdma.c
+++ b/drivers/dma/imx-sdma.c
@@ -1323,7 +1323,7 @@ static struct dma_async_tx_descriptor *sdma_prep_dma_cyclic(
 	}
 
 	if (period_len > 0xffff) {
-		dev_err(sdma->dev, "SDMA channel %d: maximum period size exceeded: %d > %d\n",
+		dev_err(sdma->dev, "SDMA channel %d: maximum period size exceeded: %zu > %d\n",
 				channel, period_len, 0xffff);
 		goto err_out;
 	}
@@ -1347,7 +1347,7 @@ static struct dma_async_tx_descriptor *sdma_prep_dma_cyclic(
 		if (i + 1 == num_periods)
 			param |= BD_WRAP;
 
-		dev_dbg(sdma->dev, "entry %d: count: %d dma: %#llx %s%s\n",
+		dev_dbg(sdma->dev, "entry %d: count: %zu dma: %#llx %s%s\n",
 				i, period_len, (u64)dma_addr,
 				param & BD_WRAP ? "wrap" : "",
 				param & BD_INTR ? " intr" : "");
@@ -1755,19 +1755,26 @@ static int sdma_probe(struct platform_device *pdev)
 	if (IS_ERR(sdma->clk_ahb))
 		return PTR_ERR(sdma->clk_ahb);
 
-	clk_prepare(sdma->clk_ipg);
-	clk_prepare(sdma->clk_ahb);
+	ret = clk_prepare(sdma->clk_ipg);
+	if (ret)
+		return ret;
+
+	ret = clk_prepare(sdma->clk_ahb);
+	if (ret)
+		goto err_clk;
 
 	ret = devm_request_irq(&pdev->dev, irq, sdma_int_handler, 0, "sdma",
 			       sdma);
 	if (ret)
-		return ret;
+		goto err_irq;
 
 	sdma->irq = irq;
 
 	sdma->script_addrs = kzalloc(sizeof(*sdma->script_addrs), GFP_KERNEL);
-	if (!sdma->script_addrs)
-		return -ENOMEM;
+	if (!sdma->script_addrs) {
+		ret = -ENOMEM;
+		goto err_irq;
+	}
 
 	/* initially no scripts available */
 	saddr_arr = (s32 *)sdma->script_addrs;
@@ -1882,6 +1889,10 @@ err_register:
 	dma_async_device_unregister(&sdma->dma_device);
 err_init:
 	kfree(sdma->script_addrs);
+err_irq:
+	clk_unprepare(sdma->clk_ahb);
+err_clk:
+	clk_unprepare(sdma->clk_ipg);
 	return ret;
 }
 
@@ -1893,6 +1904,8 @@ static int sdma_remove(struct platform_device *pdev)
 	devm_free_irq(&pdev->dev, sdma->irq, sdma);
 	dma_async_device_unregister(&sdma->dma_device);
 	kfree(sdma->script_addrs);
+	clk_unprepare(sdma->clk_ahb);
+	clk_unprepare(sdma->clk_ipg);
 	/* Kill the tasklet */
 	for (i = 0; i < MAX_DMA_CHANNELS; i++) {
 		struct sdma_channel *sdmac = &sdma->channel[i];
diff --git a/drivers/dma/ioat/dca.c b/drivers/dma/ioat/dca.c
index 0b9b6b07db9e..eab2fdda29ec 100644
--- a/drivers/dma/ioat/dca.c
+++ b/drivers/dma/ioat/dca.c
@@ -336,10 +336,10 @@ struct dca_provider *ioat_dca_init(struct pci_dev *pdev, void __iomem *iobase)
 	}
 
 	if (dca3_tag_map_invalid(ioatdca->tag_map)) {
-		WARN_TAINT_ONCE(1, TAINT_FIRMWARE_WORKAROUND,
-				"%s %s: APICID_TAG_MAP set incorrectly by BIOS, disabling DCA\n",
-				dev_driver_string(&pdev->dev),
-				dev_name(&pdev->dev));
+		add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
+		pr_warn_once("%s %s: APICID_TAG_MAP set incorrectly by BIOS, disabling DCA\n",
+			     dev_driver_string(&pdev->dev),
+			     dev_name(&pdev->dev));
 		free_dca_provider(dca);
 		return NULL;
 	}
diff --git a/drivers/dma/mv_xor_v2.c b/drivers/dma/mv_xor_v2.c
index f3e211f8f6c5..f652a0e0f5a2 100644
--- a/drivers/dma/mv_xor_v2.c
+++ b/drivers/dma/mv_xor_v2.c
@@ -42,6 +42,7 @@
 #define MV_XOR_V2_DMA_IMSG_THRD_OFF			0x018
 #define   MV_XOR_V2_DMA_IMSG_THRD_MASK			0x7FFF
 #define   MV_XOR_V2_DMA_IMSG_THRD_SHIFT			0x0
+#define   MV_XOR_V2_DMA_IMSG_TIMER_EN			BIT(18)
 #define MV_XOR_V2_DMA_DESQ_AWATTR_OFF			0x01C
   /* Same flags as MV_XOR_V2_DMA_DESQ_ARATTR_OFF */
 #define MV_XOR_V2_DMA_DESQ_ALLOC_OFF			0x04C
@@ -55,6 +56,9 @@
 #define MV_XOR_V2_DMA_DESQ_STOP_OFF			0x800
 #define MV_XOR_V2_DMA_DESQ_DEALLOC_OFF			0x804
 #define MV_XOR_V2_DMA_DESQ_ADD_OFF			0x808
+#define MV_XOR_V2_DMA_IMSG_TMOT				0x810
+#define   MV_XOR_V2_DMA_IMSG_TIMER_THRD_MASK		0x1FFF
+#define   MV_XOR_V2_DMA_IMSG_TIMER_THRD_SHIFT		0
 
 /* XOR Global registers */
 #define MV_XOR_V2_GLOB_BW_CTRL				0x4
@@ -90,6 +94,13 @@
  */
 #define MV_XOR_V2_DESC_NUM				1024
 
+/*
+ * Threshold values for descriptors and timeout, determined by
+ * experimentation as giving a good level of performance.
+ */
+#define MV_XOR_V2_DONE_IMSG_THRD  0x14
+#define MV_XOR_V2_TIMER_THRD      0xB0
+
 /**
  * struct mv_xor_v2_descriptor - DMA HW descriptor
  * @desc_id: used by S/W and is not affected by H/W.
@@ -246,6 +257,29 @@ static int mv_xor_v2_set_desc_size(struct mv_xor_v2_device *xor_dev)
 	return MV_XOR_V2_EXT_DESC_SIZE;
 }
 
+/*
+ * Set the IMSG threshold
+ */
+static inline
+void mv_xor_v2_enable_imsg_thrd(struct mv_xor_v2_device *xor_dev)
+{
+	u32 reg;
+
+	/* Configure threshold of number of descriptors, and enable timer */
+	reg = readl(xor_dev->dma_base + MV_XOR_V2_DMA_IMSG_THRD_OFF);
+	reg &= (~MV_XOR_V2_DMA_IMSG_THRD_MASK << MV_XOR_V2_DMA_IMSG_THRD_SHIFT);
+	reg |= (MV_XOR_V2_DONE_IMSG_THRD << MV_XOR_V2_DMA_IMSG_THRD_SHIFT);
+	reg |= MV_XOR_V2_DMA_IMSG_TIMER_EN;
+	writel(reg, xor_dev->dma_base + MV_XOR_V2_DMA_IMSG_THRD_OFF);
+
+	/* Configure Timer Threshold */
+	reg = readl(xor_dev->dma_base + MV_XOR_V2_DMA_IMSG_TMOT);
+	reg &= (~MV_XOR_V2_DMA_IMSG_TIMER_THRD_MASK <<
+		MV_XOR_V2_DMA_IMSG_TIMER_THRD_SHIFT);
+	reg |= (MV_XOR_V2_TIMER_THRD << MV_XOR_V2_DMA_IMSG_TIMER_THRD_SHIFT);
+	writel(reg, xor_dev->dma_base + MV_XOR_V2_DMA_IMSG_TMOT);
+}
+
 static irqreturn_t mv_xor_v2_interrupt_handler(int irq, void *data)
 {
 	struct mv_xor_v2_device *xor_dev = data;
@@ -501,9 +535,6 @@ static void mv_xor_v2_issue_pending(struct dma_chan *chan)
 	mv_xor_v2_add_desc_to_desq(xor_dev, xor_dev->npendings);
 	xor_dev->npendings = 0;
 
-	/* Activate the channel */
-	writel(0, xor_dev->dma_base + MV_XOR_V2_DMA_DESQ_STOP_OFF);
-
 	spin_unlock_bh(&xor_dev->lock);
 }
 
@@ -665,6 +696,27 @@ static int mv_xor_v2_descq_init(struct mv_xor_v2_device *xor_dev)
 	return 0;
 }
 
+static int mv_xor_v2_suspend(struct platform_device *dev, pm_message_t state)
+{
+	struct mv_xor_v2_device *xor_dev = platform_get_drvdata(dev);
+
+	/* Set this bit to disable to stop the XOR unit. */
+	writel(0x1, xor_dev->dma_base + MV_XOR_V2_DMA_DESQ_STOP_OFF);
+
+	return 0;
+}
+
+static int mv_xor_v2_resume(struct platform_device *dev)
+{
+	struct mv_xor_v2_device *xor_dev = platform_get_drvdata(dev);
+
+	mv_xor_v2_set_desc_size(xor_dev);
+	mv_xor_v2_enable_imsg_thrd(xor_dev);
+	mv_xor_v2_descq_init(xor_dev);
+
+	return 0;
+}
+
 static int mv_xor_v2_probe(struct platform_device *pdev)
 {
 	struct mv_xor_v2_device *xor_dev;
@@ -795,6 +847,8 @@ static int mv_xor_v2_probe(struct platform_device *pdev)
 	list_add_tail(&xor_dev->dmachan.device_node,
 		      &dma_dev->channels);
 
+	mv_xor_v2_enable_imsg_thrd(xor_dev);
+
 	mv_xor_v2_descq_init(xor_dev);
 
 	ret = dma_async_device_register(dma_dev);
@@ -844,6 +898,8 @@ MODULE_DEVICE_TABLE(of, mv_xor_v2_dt_ids);
 
 static struct platform_driver mv_xor_v2_driver = {
 	.probe		= mv_xor_v2_probe,
+	.suspend	= mv_xor_v2_suspend,
+	.resume		= mv_xor_v2_resume,
 	.remove		= mv_xor_v2_remove,
 	.driver		= {
 		.name	= "mv_xor_v2",
diff --git a/drivers/dma/mxs-dma.c b/drivers/dma/mxs-dma.c
index e217268c7098..41d167921fab 100644
--- a/drivers/dma/mxs-dma.c
+++ b/drivers/dma/mxs-dma.c
@@ -617,7 +617,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_dma_cyclic(
 
 	if (period_len > MAX_XFER_BYTES) {
 		dev_err(mxs_dma->dma_device.dev,
-				"maximum period size exceeded: %d > %d\n",
+				"maximum period size exceeded: %zu > %d\n",
 				period_len, MAX_XFER_BYTES);
 		goto err_out;
 	}
diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c
index e90a7a0d760a..b19ee04567b5 100644
--- a/drivers/dma/pl330.c
+++ b/drivers/dma/pl330.c
@@ -443,7 +443,10 @@ struct dma_pl330_chan {
 	/* For D-to-M and M-to-D channels */
 	int burst_sz; /* the peripheral fifo width */
 	int burst_len; /* the number of burst */
-	dma_addr_t fifo_addr;
+	phys_addr_t fifo_addr;
+	/* DMA-mapped view of the FIFO; may differ if an IOMMU is present */
+	dma_addr_t fifo_dma;
+	enum dma_data_direction dir;
 
 	/* for cyclic capability */
 	bool cyclic;
@@ -538,11 +541,6 @@ struct _xfer_spec {
 	struct dma_pl330_desc *desc;
 };
 
-static inline bool _queue_empty(struct pl330_thread *thrd)
-{
-	return thrd->req[0].desc == NULL && thrd->req[1].desc == NULL;
-}
-
 static inline bool _queue_full(struct pl330_thread *thrd)
 {
 	return thrd->req[0].desc != NULL && thrd->req[1].desc != NULL;
@@ -564,23 +562,6 @@ static inline u32 get_revision(u32 periph_id)
 	return (periph_id >> PERIPH_REV_SHIFT) & PERIPH_REV_MASK;
 }
 
-static inline u32 _emit_ADDH(unsigned dry_run, u8 buf[],
-		enum pl330_dst da, u16 val)
-{
-	if (dry_run)
-		return SZ_DMAADDH;
-
-	buf[0] = CMD_DMAADDH;
-	buf[0] |= (da << 1);
-	buf[1] = val;
-	buf[2] = val >> 8;
-
-	PL330_DBGCMD_DUMP(SZ_DMAADDH, "\tDMAADDH %s %u\n",
-		da == 1 ? "DA" : "SA", val);
-
-	return SZ_DMAADDH;
-}
-
 static inline u32 _emit_END(unsigned dry_run, u8 buf[])
 {
 	if (dry_run)
@@ -738,18 +719,6 @@ static inline u32 _emit_MOV(unsigned dry_run, u8 buf[],
 	return SZ_DMAMOV;
 }
 
-static inline u32 _emit_NOP(unsigned dry_run, u8 buf[])
-{
-	if (dry_run)
-		return SZ_DMANOP;
-
-	buf[0] = CMD_DMANOP;
-
-	PL330_DBGCMD_DUMP(SZ_DMANOP, "\tDMANOP\n");
-
-	return SZ_DMANOP;
-}
-
 static inline u32 _emit_RMB(unsigned dry_run, u8 buf[])
 {
 	if (dry_run)
@@ -817,39 +786,6 @@ static inline u32 _emit_STP(unsigned dry_run, u8 buf[],
 	return SZ_DMASTP;
 }
 
-static inline u32 _emit_STZ(unsigned dry_run, u8 buf[])
-{
-	if (dry_run)
-		return SZ_DMASTZ;
-
-	buf[0] = CMD_DMASTZ;
-
-	PL330_DBGCMD_DUMP(SZ_DMASTZ, "\tDMASTZ\n");
-
-	return SZ_DMASTZ;
-}
-
-static inline u32 _emit_WFE(unsigned dry_run, u8 buf[], u8 ev,
-		unsigned invalidate)
-{
-	if (dry_run)
-		return SZ_DMAWFE;
-
-	buf[0] = CMD_DMAWFE;
-
-	ev &= 0x1f;
-	ev <<= 3;
-	buf[1] = ev;
-
-	if (invalidate)
-		buf[1] |= (1 << 1);
-
-	PL330_DBGCMD_DUMP(SZ_DMAWFE, "\tDMAWFE %u%s\n",
-		ev >> 3, invalidate ? ", I" : "");
-
-	return SZ_DMAWFE;
-}
-
 static inline u32 _emit_WFP(unsigned dry_run, u8 buf[],
 		enum pl330_cond cond, u8 peri)
 {
@@ -2120,11 +2056,60 @@ static int pl330_alloc_chan_resources(struct dma_chan *chan)
 	return 1;
 }
 
+/*
+ * We need the data direction between the DMAC (the dma-mapping "device") and
+ * the FIFO (the dmaengine "dev"), from the FIFO's point of view. Confusing!
+ */
+static enum dma_data_direction
+pl330_dma_slave_map_dir(enum dma_transfer_direction dir)
+{
+	switch (dir) {
+	case DMA_MEM_TO_DEV:
+		return DMA_FROM_DEVICE;
+	case DMA_DEV_TO_MEM:
+		return DMA_TO_DEVICE;
+	case DMA_DEV_TO_DEV:
+		return DMA_BIDIRECTIONAL;
+	default:
+		return DMA_NONE;
+	}
+}
+
+static void pl330_unprep_slave_fifo(struct dma_pl330_chan *pch)
+{
+	if (pch->dir != DMA_NONE)
+		dma_unmap_resource(pch->chan.device->dev, pch->fifo_dma,
+				   1 << pch->burst_sz, pch->dir, 0);
+	pch->dir = DMA_NONE;
+}
+
+
+static bool pl330_prep_slave_fifo(struct dma_pl330_chan *pch,
+				  enum dma_transfer_direction dir)
+{
+	struct device *dev = pch->chan.device->dev;
+	enum dma_data_direction dma_dir = pl330_dma_slave_map_dir(dir);
+
+	/* Already mapped for this config? */
+	if (pch->dir == dma_dir)
+		return true;
+
+	pl330_unprep_slave_fifo(pch);
+	pch->fifo_dma = dma_map_resource(dev, pch->fifo_addr,
+					 1 << pch->burst_sz, dma_dir, 0);
+	if (dma_mapping_error(dev, pch->fifo_dma))
+		return false;
+
+	pch->dir = dma_dir;
+	return true;
+}
+
 static int pl330_config(struct dma_chan *chan,
 			struct dma_slave_config *slave_config)
 {
 	struct dma_pl330_chan *pch = to_pchan(chan);
 
+	pl330_unprep_slave_fifo(pch);
 	if (slave_config->direction == DMA_MEM_TO_DEV) {
 		if (slave_config->dst_addr)
 			pch->fifo_addr = slave_config->dst_addr;
@@ -2235,6 +2220,7 @@ static void pl330_free_chan_resources(struct dma_chan *chan)
 	spin_unlock_irqrestore(&pl330->lock, flags);
 	pm_runtime_mark_last_busy(pch->dmac->ddma.dev);
 	pm_runtime_put_autosuspend(pch->dmac->ddma.dev);
+	pl330_unprep_slave_fifo(pch);
 }
 
 static int pl330_get_current_xferred_count(struct dma_pl330_chan *pch,
@@ -2564,6 +2550,9 @@ static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic(
 		return NULL;
 	}
 
+	if (!pl330_prep_slave_fifo(pch, direction))
+		return NULL;
+
 	for (i = 0; i < len / period_len; i++) {
 		desc = pl330_get_desc(pch);
 		if (!desc) {
@@ -2593,12 +2582,12 @@ static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic(
 			desc->rqcfg.src_inc = 1;
 			desc->rqcfg.dst_inc = 0;
 			src = dma_addr;
-			dst = pch->fifo_addr;
+			dst = pch->fifo_dma;
 			break;
 		case DMA_DEV_TO_MEM:
 			desc->rqcfg.src_inc = 0;
 			desc->rqcfg.dst_inc = 1;
-			src = pch->fifo_addr;
+			src = pch->fifo_dma;
 			dst = dma_addr;
 			break;
 		default:
@@ -2711,12 +2700,12 @@ pl330_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 	struct dma_pl330_chan *pch = to_pchan(chan);
 	struct scatterlist *sg;
 	int i;
-	dma_addr_t addr;
 
 	if (unlikely(!pch || !sgl || !sg_len))
 		return NULL;
 
-	addr = pch->fifo_addr;
+	if (!pl330_prep_slave_fifo(pch, direction))
+		return NULL;
 
 	first = NULL;
 
@@ -2742,13 +2731,13 @@ 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;
-			fill_px(&desc->px,
-				addr, sg_dma_address(sg), sg_dma_len(sg));
+			fill_px(&desc->px, pch->fifo_dma, sg_dma_address(sg),
+				sg_dma_len(sg));
 		} else {
 			desc->rqcfg.src_inc = 0;
 			desc->rqcfg.dst_inc = 1;
-			fill_px(&desc->px,
-				sg_dma_address(sg), addr, sg_dma_len(sg));
+			fill_px(&desc->px, sg_dma_address(sg), pch->fifo_dma,
+				sg_dma_len(sg));
 		}
 
 		desc->rqcfg.brst_size = pch->burst_sz;
@@ -2906,6 +2895,7 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id)
 		pch->thread = NULL;
 		pch->chan.device = pd;
 		pch->dmac = pl330;
+		pch->dir = DMA_NONE;
 
 		/* Add the channel to the DMAC list */
 		list_add_tail(&pch->chan.device_node, &pd->channels);
diff --git a/drivers/dma/qcom/hidma.c b/drivers/dma/qcom/hidma.c
index 5072a7d306d4..34fb6afd229b 100644
--- a/drivers/dma/qcom/hidma.c
+++ b/drivers/dma/qcom/hidma.c
@@ -1,7 +1,7 @@
 /*
  * Qualcomm Technologies HIDMA DMA engine interface
  *
- * Copyright (c) 2015-2016, The Linux Foundation. All rights reserved.
+ * Copyright (c) 2015-2017, The Linux Foundation. All rights reserved.
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 and
@@ -210,6 +210,7 @@ static int hidma_chan_init(struct hidma_dev *dmadev, u32 dma_sig)
 	INIT_LIST_HEAD(&mchan->prepared);
 	INIT_LIST_HEAD(&mchan->active);
 	INIT_LIST_HEAD(&mchan->completed);
+	INIT_LIST_HEAD(&mchan->queued);
 
 	spin_lock_init(&mchan->lock);
 	list_add_tail(&mchan->chan.device_node, &ddev->channels);
@@ -230,9 +231,15 @@ static void hidma_issue_pending(struct dma_chan *dmach)
 	struct hidma_chan *mchan = to_hidma_chan(dmach);
 	struct hidma_dev *dmadev = mchan->dmadev;
 	unsigned long flags;
+	struct hidma_desc *qdesc, *next;
 	int status;
 
 	spin_lock_irqsave(&mchan->lock, flags);
+	list_for_each_entry_safe(qdesc, next, &mchan->queued, node) {
+		hidma_ll_queue_request(dmadev->lldev, qdesc->tre_ch);
+		list_move_tail(&qdesc->node, &mchan->active);
+	}
+
 	if (!mchan->running) {
 		struct hidma_desc *desc = list_first_entry(&mchan->active,
 							   struct hidma_desc,
@@ -315,17 +322,18 @@ static dma_cookie_t hidma_tx_submit(struct dma_async_tx_descriptor *txd)
 		pm_runtime_put_autosuspend(dmadev->ddev.dev);
 		return -ENODEV;
 	}
+	pm_runtime_mark_last_busy(dmadev->ddev.dev);
+	pm_runtime_put_autosuspend(dmadev->ddev.dev);
 
 	mdesc = container_of(txd, struct hidma_desc, desc);
 	spin_lock_irqsave(&mchan->lock, irqflags);
 
-	/* Move descriptor to active */
-	list_move_tail(&mdesc->node, &mchan->active);
+	/* Move descriptor to queued */
+	list_move_tail(&mdesc->node, &mchan->queued);
 
 	/* Update cookie */
 	cookie = dma_cookie_assign(txd);
 
-	hidma_ll_queue_request(dmadev->lldev, mdesc->tre_ch);
 	spin_unlock_irqrestore(&mchan->lock, irqflags);
 
 	return cookie;
@@ -431,6 +439,7 @@ static int hidma_terminate_channel(struct dma_chan *chan)
 	list_splice_init(&mchan->active, &list);
 	list_splice_init(&mchan->prepared, &list);
 	list_splice_init(&mchan->completed, &list);
+	list_splice_init(&mchan->queued, &list);
 	spin_unlock_irqrestore(&mchan->lock, irqflags);
 
 	/* this suspends the existing transfer */
@@ -795,8 +804,11 @@ static int hidma_probe(struct platform_device *pdev)
 	device_property_read_u32(&pdev->dev, "desc-count",
 				 &dmadev->nr_descriptors);
 
-	if (!dmadev->nr_descriptors && nr_desc_prm)
+	if (nr_desc_prm) {
+		dev_info(&pdev->dev, "overriding number of descriptors as %d\n",
+			 nr_desc_prm);
 		dmadev->nr_descriptors = nr_desc_prm;
+	}
 
 	if (!dmadev->nr_descriptors)
 		dmadev->nr_descriptors = HIDMA_NR_DEFAULT_DESC;
diff --git a/drivers/dma/qcom/hidma.h b/drivers/dma/qcom/hidma.h
index c7d014235c32..41e0aa283828 100644
--- a/drivers/dma/qcom/hidma.h
+++ b/drivers/dma/qcom/hidma.h
@@ -104,6 +104,7 @@ struct hidma_chan {
 	struct dma_chan			chan;
 	struct list_head		free;
 	struct list_head		prepared;
+	struct list_head		queued;
 	struct list_head		active;
 	struct list_head		completed;
 
diff --git a/drivers/dma/qcom/hidma_mgmt.c b/drivers/dma/qcom/hidma_mgmt.c
index f847d32cc4b5..5a0991bc4787 100644
--- a/drivers/dma/qcom/hidma_mgmt.c
+++ b/drivers/dma/qcom/hidma_mgmt.c
@@ -1,7 +1,7 @@
 /*
  * Qualcomm Technologies HIDMA DMA engine Management interface
  *
- * Copyright (c) 2015-2016, The Linux Foundation. All rights reserved.
+ * Copyright (c) 2015-2017, The Linux Foundation. All rights reserved.
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 and
@@ -49,6 +49,26 @@
 #define HIDMA_AUTOSUSPEND_TIMEOUT	2000
 #define HIDMA_MAX_CHANNEL_WEIGHT	15
 
+static unsigned int max_write_request;
+module_param(max_write_request, uint, 0644);
+MODULE_PARM_DESC(max_write_request,
+		"maximum write burst (default: ACPI/DT value)");
+
+static unsigned int max_read_request;
+module_param(max_read_request, uint, 0644);
+MODULE_PARM_DESC(max_read_request,
+		"maximum read burst (default: ACPI/DT value)");
+
+static unsigned int max_wr_xactions;
+module_param(max_wr_xactions, uint, 0644);
+MODULE_PARM_DESC(max_wr_xactions,
+	"maximum number of write transactions (default: ACPI/DT value)");
+
+static unsigned int max_rd_xactions;
+module_param(max_rd_xactions, uint, 0644);
+MODULE_PARM_DESC(max_rd_xactions,
+	"maximum number of read transactions (default: ACPI/DT value)");
+
 int hidma_mgmt_setup(struct hidma_mgmt_dev *mgmtdev)
 {
 	unsigned int i;
@@ -207,12 +227,25 @@ static int hidma_mgmt_probe(struct platform_device *pdev)
 		goto out;
 	}
 
+	if (max_write_request) {
+		dev_info(&pdev->dev, "overriding max-write-burst-bytes: %d\n",
+			max_write_request);
+		mgmtdev->max_write_request = max_write_request;
+	} else
+		max_write_request = mgmtdev->max_write_request;
+
 	rc = device_property_read_u32(&pdev->dev, "max-read-burst-bytes",
 				      &mgmtdev->max_read_request);
 	if (rc) {
 		dev_err(&pdev->dev, "max-read-burst-bytes missing\n");
 		goto out;
 	}
+	if (max_read_request) {
+		dev_info(&pdev->dev, "overriding max-read-burst-bytes: %d\n",
+			max_read_request);
+		mgmtdev->max_read_request = max_read_request;
+	} else
+		max_read_request = mgmtdev->max_read_request;
 
 	rc = device_property_read_u32(&pdev->dev, "max-write-transactions",
 				      &mgmtdev->max_wr_xactions);
@@ -220,6 +253,12 @@ static int hidma_mgmt_probe(struct platform_device *pdev)
 		dev_err(&pdev->dev, "max-write-transactions missing\n");
 		goto out;
 	}
+	if (max_wr_xactions) {
+		dev_info(&pdev->dev, "overriding max-write-transactions: %d\n",
+			max_wr_xactions);
+		mgmtdev->max_wr_xactions = max_wr_xactions;
+	} else
+		max_wr_xactions = mgmtdev->max_wr_xactions;
 
 	rc = device_property_read_u32(&pdev->dev, "max-read-transactions",
 				      &mgmtdev->max_rd_xactions);
@@ -227,6 +266,12 @@ static int hidma_mgmt_probe(struct platform_device *pdev)
 		dev_err(&pdev->dev, "max-read-transactions missing\n");
 		goto out;
 	}
+	if (max_rd_xactions) {
+		dev_info(&pdev->dev, "overriding max-read-transactions: %d\n",
+			max_rd_xactions);
+		mgmtdev->max_rd_xactions = max_rd_xactions;
+	} else
+		max_rd_xactions = mgmtdev->max_rd_xactions;
 
 	mgmtdev->priority = devm_kcalloc(&pdev->dev,
 					 mgmtdev->dma_channels,
diff --git a/drivers/dma/sh/rcar-dmac.c b/drivers/dma/sh/rcar-dmac.c
index bd261c9e9664..ffcadca53243 100644
--- a/drivers/dma/sh/rcar-dmac.c
+++ b/drivers/dma/sh/rcar-dmac.c
@@ -144,6 +144,7 @@ struct rcar_dmac_chan_map {
  * @chan: base DMA channel object
  * @iomem: channel I/O memory base
  * @index: index of this channel in the controller
+ * @irq: channel IRQ
  * @src: slave memory address and size on the source side
  * @dst: slave memory address and size on the destination side
  * @mid_rid: hardware MID/RID for the DMA client using this channel
@@ -161,6 +162,7 @@ struct rcar_dmac_chan {
 	struct dma_chan chan;
 	void __iomem *iomem;
 	unsigned int index;
+	int irq;
 
 	struct rcar_dmac_chan_slave src;
 	struct rcar_dmac_chan_slave dst;
@@ -1008,7 +1010,11 @@ static void rcar_dmac_free_chan_resources(struct dma_chan *chan)
 	rcar_dmac_chan_halt(rchan);
 	spin_unlock_irq(&rchan->lock);
 
-	/* Now no new interrupts will occur */
+	/*
+	 * Now no new interrupts will occur, but one might already be
+	 * running. Wait for it to finish before freeing resources.
+	 */
+	synchronize_irq(rchan->irq);
 
 	if (rchan->mid_rid >= 0) {
 		/* The caller is holding dma_list_mutex */
@@ -1366,6 +1372,13 @@ done:
 	spin_unlock_irqrestore(&rchan->lock, flags);
 }
 
+static void rcar_dmac_device_synchronize(struct dma_chan *chan)
+{
+	struct rcar_dmac_chan *rchan = to_rcar_dmac_chan(chan);
+
+	synchronize_irq(rchan->irq);
+}
+
 /* -----------------------------------------------------------------------------
  * IRQ handling
  */
@@ -1650,7 +1663,6 @@ static int rcar_dmac_chan_probe(struct rcar_dmac *dmac,
 	struct dma_chan *chan = &rchan->chan;
 	char pdev_irqname[5];
 	char *irqname;
-	int irq;
 	int ret;
 
 	rchan->index = index;
@@ -1667,8 +1679,8 @@ static int rcar_dmac_chan_probe(struct rcar_dmac *dmac,
 
 	/* Request the channel interrupt. */
 	sprintf(pdev_irqname, "ch%u", index);
-	irq = platform_get_irq_byname(pdev, pdev_irqname);
-	if (irq < 0) {
+	rchan->irq = platform_get_irq_byname(pdev, pdev_irqname);
+	if (rchan->irq < 0) {
 		dev_err(dmac->dev, "no IRQ specified for channel %u\n", index);
 		return -ENODEV;
 	}
@@ -1678,11 +1690,13 @@ static int rcar_dmac_chan_probe(struct rcar_dmac *dmac,
 	if (!irqname)
 		return -ENOMEM;
 
-	ret = devm_request_threaded_irq(dmac->dev, irq, rcar_dmac_isr_channel,
+	ret = devm_request_threaded_irq(dmac->dev, rchan->irq,
+					rcar_dmac_isr_channel,
 					rcar_dmac_isr_channel_thread, 0,
 					irqname, rchan);
 	if (ret) {
-		dev_err(dmac->dev, "failed to request IRQ %u (%d)\n", irq, ret);
+		dev_err(dmac->dev, "failed to request IRQ %u (%d)\n",
+			rchan->irq, ret);
 		return ret;
 	}
 
@@ -1846,6 +1860,7 @@ static int rcar_dmac_probe(struct platform_device *pdev)
 	engine->device_terminate_all = rcar_dmac_chan_terminate_all;
 	engine->device_tx_status = rcar_dmac_tx_status;
 	engine->device_issue_pending = rcar_dmac_issue_pending;
+	engine->device_synchronize = rcar_dmac_device_synchronize;
 
 	ret = dma_async_device_register(engine);
 	if (ret < 0)
diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c
index a6620b671d1d..c3052fbfd092 100644
--- a/drivers/dma/ste_dma40.c
+++ b/drivers/dma/ste_dma40.c
@@ -2528,10 +2528,7 @@ dma40_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t dma_addr,
 		dma_addr += period_len;
 	}
 
-	sg[periods].offset = 0;
-	sg_dma_len(&sg[periods]) = 0;
-	sg[periods].page_link =
-		((unsigned long)sg | 0x01) & ~0x02;
+	sg_chain(sg, periods + 1, sg);
 
 	txd = d40_prep_sg(chan, sg, sg, periods, direction,
 			  DMA_PREP_INTERRUPT);
diff --git a/drivers/dma/tegra20-apb-dma.c b/drivers/dma/tegra20-apb-dma.c
index 3722b9d8d9fe..b9d75a54c896 100644
--- a/drivers/dma/tegra20-apb-dma.c
+++ b/drivers/dma/tegra20-apb-dma.c
@@ -1494,35 +1494,7 @@ static int tegra_dma_remove(struct platform_device *pdev)
 static int tegra_dma_runtime_suspend(struct device *dev)
 {
 	struct tegra_dma *tdma = dev_get_drvdata(dev);
-
-	clk_disable_unprepare(tdma->dma_clk);
-	return 0;
-}
-
-static int tegra_dma_runtime_resume(struct device *dev)
-{
-	struct tegra_dma *tdma = dev_get_drvdata(dev);
-	int ret;
-
-	ret = clk_prepare_enable(tdma->dma_clk);
-	if (ret < 0) {
-		dev_err(dev, "clk_enable failed: %d\n", ret);
-		return ret;
-	}
-	return 0;
-}
-
-#ifdef CONFIG_PM_SLEEP
-static int tegra_dma_pm_suspend(struct device *dev)
-{
-	struct tegra_dma *tdma = dev_get_drvdata(dev);
 	int i;
-	int ret;
-
-	/* Enable clock before accessing register */
-	ret = pm_runtime_get_sync(dev);
-	if (ret < 0)
-		return ret;
 
 	tdma->reg_gen = tdma_read(tdma, TEGRA_APBDMA_GENERAL);
 	for (i = 0; i < tdma->chip_data->nr_channels; i++) {
@@ -1543,21 +1515,21 @@ static int tegra_dma_pm_suspend(struct device *dev)
 						  TEGRA_APBDMA_CHAN_WCOUNT);
 	}
 
-	/* Disable clock */
-	pm_runtime_put(dev);
+	clk_disable_unprepare(tdma->dma_clk);
+
 	return 0;
 }
 
-static int tegra_dma_pm_resume(struct device *dev)
+static int tegra_dma_runtime_resume(struct device *dev)
 {
 	struct tegra_dma *tdma = dev_get_drvdata(dev);
-	int i;
-	int ret;
+	int i, ret;
 
-	/* Enable clock before accessing register */
-	ret = pm_runtime_get_sync(dev);
-	if (ret < 0)
+	ret = clk_prepare_enable(tdma->dma_clk);
+	if (ret < 0) {
+		dev_err(dev, "clk_enable failed: %d\n", ret);
 		return ret;
+	}
 
 	tdma_write(tdma, TEGRA_APBDMA_GENERAL, tdma->reg_gen);
 	tdma_write(tdma, TEGRA_APBDMA_CONTROL, 0);
@@ -1582,16 +1554,14 @@ static int tegra_dma_pm_resume(struct device *dev)
 			(ch_reg->csr & ~TEGRA_APBDMA_CSR_ENB));
 	}
 
-	/* Disable clock */
-	pm_runtime_put(dev);
 	return 0;
 }
-#endif
 
 static const struct dev_pm_ops tegra_dma_dev_pm_ops = {
 	SET_RUNTIME_PM_OPS(tegra_dma_runtime_suspend, tegra_dma_runtime_resume,
 			   NULL)
-	SET_SYSTEM_SLEEP_PM_OPS(tegra_dma_pm_suspend, tegra_dma_pm_resume)
+	SET_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
+				pm_runtime_force_resume)
 };
 
 static const struct of_device_id tegra_dma_of_match[] = {
diff --git a/drivers/dma/xilinx/zynqmp_dma.c b/drivers/dma/xilinx/zynqmp_dma.c
index 6d221e5c72ee..47f64192d2fd 100644
--- a/drivers/dma/xilinx/zynqmp_dma.c
+++ b/drivers/dma/xilinx/zynqmp_dma.c
@@ -794,9 +794,6 @@ static struct dma_async_tx_descriptor *zynqmp_dma_prep_memcpy(
 
 	chan = to_chan(dchan);
 
-	if (len > ZYNQMP_DMA_MAX_TRANS_LEN)
-		return NULL;
-
 	desc_cnt = DIV_ROUND_UP(len, ZYNQMP_DMA_MAX_TRANS_LEN);
 
 	spin_lock_bh(&chan->lock);