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

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

Fix up some fairly trivial conflicts with various other cleanups.

* 'next' of git://git.infradead.org/users/vkoul/slave-dma: (67 commits)
  dmaengine: imx: fix the build failure on x86_64
  dmaengine: i.MX: Fix merge of cookie branch.
  dmaengine: i.MX: Add support for interleaved transfers.
  dmaengine: imx-dma: use 'dev_dbg' and 'dev_warn' for messages.
  dmaengine: imx-dma: remove 'imx_dmav1_baseaddr' and 'dma_clk'.
  dmaengine: imx-dma: remove unused arg of imxdma_sg_next.
  dmaengine: imx-dma: remove internal structure.
  dmaengine: imx-dma: remove 'resbytes' field of 'internal' structure.
  dmaengine: imx-dma: remove 'in_use' field of 'internal' structure.
  dmaengine: imx-dma: remove sg member from internal structure.
  dmaengine: imx-dma: remove 'imxdma_setup_sg_hw' function.
  dmaengine: imx-dma: remove 'imxdma_config_channel_hw' function.
  dmaengine: imx-dma: remove 'imxdma_setup_mem2mem_hw' function.
  dmaengine: imx-dma: remove dma_mode member of internal structure.
  dmaengine: imx-dma: remove data member from internal structure.
  dmaengine: imx-dma: merge old dma-v1.c with imx-dma.c
  dmaengine: at_hdmac: add slave config operation
  dmaengine: add context parameter to prep_slave_sg and prep_dma_cyclic
  dmaengine/dma_slave: introduce inline wrappers
  dma: imx-sdma: Treat firmware messages as warnings instead of erros
  ...
diff --git a/arch/arm/common/Kconfig b/arch/arm/common/Kconfig
index 3bb1d75..283fa1d 100644
--- a/arch/arm/common/Kconfig
+++ b/arch/arm/common/Kconfig
@@ -24,9 +24,6 @@
 config ICST
 	bool
 
-config PL330
-	bool
-
 config SA1111
 	bool
 	select DMABOUNCE if !ARCH_PXA
diff --git a/arch/arm/common/Makefile b/arch/arm/common/Makefile
index 69feafe..215816f 100644
--- a/arch/arm/common/Makefile
+++ b/arch/arm/common/Makefile
@@ -5,7 +5,6 @@
 obj-$(CONFIG_ARM_GIC)		+= gic.o
 obj-$(CONFIG_ARM_VIC)		+= vic.o
 obj-$(CONFIG_ICST)		+= icst.o
-obj-$(CONFIG_PL330)		+= pl330.o
 obj-$(CONFIG_SA1111)		+= sa1111.o
 obj-$(CONFIG_PCI_HOST_VIA82C505) += via82c505.o
 obj-$(CONFIG_DMABOUNCE)		+= dmabounce.o
diff --git a/arch/arm/common/pl330.c b/arch/arm/common/pl330.c
deleted file mode 100644
index ff3ad22..0000000
--- a/arch/arm/common/pl330.c
+++ /dev/null
@@ -1,1960 +0,0 @@
-/* linux/arch/arm/common/pl330.c
- *
- * Copyright (C) 2010 Samsung Electronics Co Ltd.
- *	Jaswinder Singh <jassi.brar@samsung.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
- */
-
-#include <linux/kernel.h>
-#include <linux/init.h>
-#include <linux/slab.h>
-#include <linux/module.h>
-#include <linux/string.h>
-#include <linux/io.h>
-#include <linux/delay.h>
-#include <linux/interrupt.h>
-#include <linux/dma-mapping.h>
-
-#include <asm/hardware/pl330.h>
-
-/* Register and Bit field Definitions */
-#define DS		0x0
-#define DS_ST_STOP	0x0
-#define DS_ST_EXEC	0x1
-#define DS_ST_CMISS	0x2
-#define DS_ST_UPDTPC	0x3
-#define DS_ST_WFE	0x4
-#define DS_ST_ATBRR	0x5
-#define DS_ST_QBUSY	0x6
-#define DS_ST_WFP	0x7
-#define DS_ST_KILL	0x8
-#define DS_ST_CMPLT	0x9
-#define DS_ST_FLTCMP	0xe
-#define DS_ST_FAULT	0xf
-
-#define DPC		0x4
-#define INTEN		0x20
-#define ES		0x24
-#define INTSTATUS	0x28
-#define INTCLR		0x2c
-#define FSM		0x30
-#define FSC		0x34
-#define FTM		0x38
-
-#define _FTC		0x40
-#define FTC(n)		(_FTC + (n)*0x4)
-
-#define _CS		0x100
-#define CS(n)		(_CS + (n)*0x8)
-#define CS_CNS		(1 << 21)
-
-#define _CPC		0x104
-#define CPC(n)		(_CPC + (n)*0x8)
-
-#define _SA		0x400
-#define SA(n)		(_SA + (n)*0x20)
-
-#define _DA		0x404
-#define DA(n)		(_DA + (n)*0x20)
-
-#define _CC		0x408
-#define CC(n)		(_CC + (n)*0x20)
-
-#define CC_SRCINC	(1 << 0)
-#define CC_DSTINC	(1 << 14)
-#define CC_SRCPRI	(1 << 8)
-#define CC_DSTPRI	(1 << 22)
-#define CC_SRCNS	(1 << 9)
-#define CC_DSTNS	(1 << 23)
-#define CC_SRCIA	(1 << 10)
-#define CC_DSTIA	(1 << 24)
-#define CC_SRCBRSTLEN_SHFT	4
-#define CC_DSTBRSTLEN_SHFT	18
-#define CC_SRCBRSTSIZE_SHFT	1
-#define CC_DSTBRSTSIZE_SHFT	15
-#define CC_SRCCCTRL_SHFT	11
-#define CC_SRCCCTRL_MASK	0x7
-#define CC_DSTCCTRL_SHFT	25
-#define CC_DRCCCTRL_MASK	0x7
-#define CC_SWAP_SHFT	28
-
-#define _LC0		0x40c
-#define LC0(n)		(_LC0 + (n)*0x20)
-
-#define _LC1		0x410
-#define LC1(n)		(_LC1 + (n)*0x20)
-
-#define DBGSTATUS	0xd00
-#define DBG_BUSY	(1 << 0)
-
-#define DBGCMD		0xd04
-#define DBGINST0	0xd08
-#define DBGINST1	0xd0c
-
-#define CR0		0xe00
-#define CR1		0xe04
-#define CR2		0xe08
-#define CR3		0xe0c
-#define CR4		0xe10
-#define CRD		0xe14
-
-#define PERIPH_ID	0xfe0
-#define PCELL_ID	0xff0
-
-#define CR0_PERIPH_REQ_SET	(1 << 0)
-#define CR0_BOOT_EN_SET		(1 << 1)
-#define CR0_BOOT_MAN_NS		(1 << 2)
-#define CR0_NUM_CHANS_SHIFT	4
-#define CR0_NUM_CHANS_MASK	0x7
-#define CR0_NUM_PERIPH_SHIFT	12
-#define CR0_NUM_PERIPH_MASK	0x1f
-#define CR0_NUM_EVENTS_SHIFT	17
-#define CR0_NUM_EVENTS_MASK	0x1f
-
-#define CR1_ICACHE_LEN_SHIFT	0
-#define CR1_ICACHE_LEN_MASK	0x7
-#define CR1_NUM_ICACHELINES_SHIFT	4
-#define CR1_NUM_ICACHELINES_MASK	0xf
-
-#define CRD_DATA_WIDTH_SHIFT	0
-#define CRD_DATA_WIDTH_MASK	0x7
-#define CRD_WR_CAP_SHIFT	4
-#define CRD_WR_CAP_MASK		0x7
-#define CRD_WR_Q_DEP_SHIFT	8
-#define CRD_WR_Q_DEP_MASK	0xf
-#define CRD_RD_CAP_SHIFT	12
-#define CRD_RD_CAP_MASK		0x7
-#define CRD_RD_Q_DEP_SHIFT	16
-#define CRD_RD_Q_DEP_MASK	0xf
-#define CRD_DATA_BUFF_SHIFT	20
-#define CRD_DATA_BUFF_MASK	0x3ff
-
-#define	PART		0x330
-#define DESIGNER	0x41
-#define REVISION	0x0
-#define INTEG_CFG	0x0
-#define PERIPH_ID_VAL	((PART << 0) | (DESIGNER << 12))
-
-#define PCELL_ID_VAL	0xb105f00d
-
-#define PL330_STATE_STOPPED		(1 << 0)
-#define PL330_STATE_EXECUTING		(1 << 1)
-#define PL330_STATE_WFE			(1 << 2)
-#define PL330_STATE_FAULTING		(1 << 3)
-#define PL330_STATE_COMPLETING		(1 << 4)
-#define PL330_STATE_WFP			(1 << 5)
-#define PL330_STATE_KILLING		(1 << 6)
-#define PL330_STATE_FAULT_COMPLETING	(1 << 7)
-#define PL330_STATE_CACHEMISS		(1 << 8)
-#define PL330_STATE_UPDTPC		(1 << 9)
-#define PL330_STATE_ATBARRIER		(1 << 10)
-#define PL330_STATE_QUEUEBUSY		(1 << 11)
-#define PL330_STATE_INVALID		(1 << 15)
-
-#define PL330_STABLE_STATES (PL330_STATE_STOPPED | PL330_STATE_EXECUTING \
-				| PL330_STATE_WFE | PL330_STATE_FAULTING)
-
-#define CMD_DMAADDH	0x54
-#define CMD_DMAEND	0x00
-#define CMD_DMAFLUSHP	0x35
-#define CMD_DMAGO	0xa0
-#define CMD_DMALD	0x04
-#define CMD_DMALDP	0x25
-#define CMD_DMALP	0x20
-#define CMD_DMALPEND	0x28
-#define CMD_DMAKILL	0x01
-#define CMD_DMAMOV	0xbc
-#define CMD_DMANOP	0x18
-#define CMD_DMARMB	0x12
-#define CMD_DMASEV	0x34
-#define CMD_DMAST	0x08
-#define CMD_DMASTP	0x29
-#define CMD_DMASTZ	0x0c
-#define CMD_DMAWFE	0x36
-#define CMD_DMAWFP	0x30
-#define CMD_DMAWMB	0x13
-
-#define SZ_DMAADDH	3
-#define SZ_DMAEND	1
-#define SZ_DMAFLUSHP	2
-#define SZ_DMALD	1
-#define SZ_DMALDP	2
-#define SZ_DMALP	2
-#define SZ_DMALPEND	2
-#define SZ_DMAKILL	1
-#define SZ_DMAMOV	6
-#define SZ_DMANOP	1
-#define SZ_DMARMB	1
-#define SZ_DMASEV	2
-#define SZ_DMAST	1
-#define SZ_DMASTP	2
-#define SZ_DMASTZ	1
-#define SZ_DMAWFE	2
-#define SZ_DMAWFP	2
-#define SZ_DMAWMB	1
-#define SZ_DMAGO	6
-
-#define BRST_LEN(ccr)	((((ccr) >> CC_SRCBRSTLEN_SHFT) & 0xf) + 1)
-#define BRST_SIZE(ccr)	(1 << (((ccr) >> CC_SRCBRSTSIZE_SHFT) & 0x7))
-
-#define BYTE_TO_BURST(b, ccr)  ((b) / BRST_SIZE(ccr) / BRST_LEN(ccr))
-#define BURST_TO_BYTE(c, ccr)  ((c) * BRST_SIZE(ccr) * BRST_LEN(ccr))
-
-/*
- * With 256 bytes, we can do more than 2.5MB and 5MB xfers per req
- * at 1byte/burst for P<->M and M<->M respectively.
- * For typical scenario, at 1word/burst, 10MB and 20MB xfers per req
- * should be enough for P<->M and M<->M respectively.
- */
-#define MCODE_BUFF_PER_REQ	256
-
-/* If the _pl330_req is available to the client */
-#define IS_FREE(req)	(*((u8 *)((req)->mc_cpu)) == CMD_DMAEND)
-
-/* Use this _only_ to wait on transient states */
-#define UNTIL(t, s)	while (!(_state(t) & (s))) cpu_relax();
-
-#ifdef PL330_DEBUG_MCGEN
-static unsigned cmd_line;
-#define PL330_DBGCMD_DUMP(off, x...)	do { \
-						printk("%x:", cmd_line); \
-						printk(x); \
-						cmd_line += off; \
-					} while (0)
-#define PL330_DBGMC_START(addr)		(cmd_line = addr)
-#else
-#define PL330_DBGCMD_DUMP(off, x...)	do {} while (0)
-#define PL330_DBGMC_START(addr)		do {} while (0)
-#endif
-
-struct _xfer_spec {
-	u32 ccr;
-	struct pl330_req *r;
-	struct pl330_xfer *x;
-};
-
-enum dmamov_dst {
-	SAR = 0,
-	CCR,
-	DAR,
-};
-
-enum pl330_dst {
-	SRC = 0,
-	DST,
-};
-
-enum pl330_cond {
-	SINGLE,
-	BURST,
-	ALWAYS,
-};
-
-struct _pl330_req {
-	u32 mc_bus;
-	void *mc_cpu;
-	/* Number of bytes taken to setup MC for the req */
-	u32 mc_len;
-	struct pl330_req *r;
-	/* Hook to attach to DMAC's list of reqs with due callback */
-	struct list_head rqd;
-};
-
-/* ToBeDone for tasklet */
-struct _pl330_tbd {
-	bool reset_dmac;
-	bool reset_mngr;
-	u8 reset_chan;
-};
-
-/* A DMAC Thread */
-struct pl330_thread {
-	u8 id;
-	int ev;
-	/* If the channel is not yet acquired by any client */
-	bool free;
-	/* Parent DMAC */
-	struct pl330_dmac *dmac;
-	/* Only two at a time */
-	struct _pl330_req req[2];
-	/* Index of the last enqueued request */
-	unsigned lstenq;
-	/* Index of the last submitted request or -1 if the DMA is stopped */
-	int req_running;
-};
-
-enum pl330_dmac_state {
-	UNINIT,
-	INIT,
-	DYING,
-};
-
-/* A DMAC */
-struct pl330_dmac {
-	spinlock_t		lock;
-	/* Holds list of reqs with due callbacks */
-	struct list_head	req_done;
-	/* Pointer to platform specific stuff */
-	struct pl330_info	*pinfo;
-	/* Maximum possible events/irqs */
-	int			events[32];
-	/* BUS address of MicroCode buffer */
-	u32			mcode_bus;
-	/* CPU address of MicroCode buffer */
-	void			*mcode_cpu;
-	/* List of all Channel threads */
-	struct pl330_thread	*channels;
-	/* Pointer to the MANAGER thread */
-	struct pl330_thread	*manager;
-	/* To handle bad news in interrupt */
-	struct tasklet_struct	tasks;
-	struct _pl330_tbd	dmac_tbd;
-	/* State of DMAC operation */
-	enum pl330_dmac_state	state;
-};
-
-static inline void _callback(struct pl330_req *r, enum pl330_op_err err)
-{
-	if (r && r->xfer_cb)
-		r->xfer_cb(r->token, err);
-}
-
-static inline bool _queue_empty(struct pl330_thread *thrd)
-{
-	return (IS_FREE(&thrd->req[0]) && IS_FREE(&thrd->req[1]))
-		? true : false;
-}
-
-static inline bool _queue_full(struct pl330_thread *thrd)
-{
-	return (IS_FREE(&thrd->req[0]) || IS_FREE(&thrd->req[1]))
-		? false : true;
-}
-
-static inline bool is_manager(struct pl330_thread *thrd)
-{
-	struct pl330_dmac *pl330 = thrd->dmac;
-
-	/* MANAGER is indexed at the end */
-	if (thrd->id == pl330->pinfo->pcfg.num_chan)
-		return true;
-	else
-		return false;
-}
-
-/* If manager of the thread is in Non-Secure mode */
-static inline bool _manager_ns(struct pl330_thread *thrd)
-{
-	struct pl330_dmac *pl330 = thrd->dmac;
-
-	return (pl330->pinfo->pcfg.mode & DMAC_MODE_NS) ? true : false;
-}
-
-static inline u32 get_id(struct pl330_info *pi, u32 off)
-{
-	void __iomem *regs = pi->base;
-	u32 id = 0;
-
-	id |= (readb(regs + off + 0x0) << 0);
-	id |= (readb(regs + off + 0x4) << 8);
-	id |= (readb(regs + off + 0x8) << 16);
-	id |= (readb(regs + off + 0xc) << 24);
-
-	return id;
-}
-
-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);
-	*((u16 *)&buf[1]) = val;
-
-	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)
-		return SZ_DMAEND;
-
-	buf[0] = CMD_DMAEND;
-
-	PL330_DBGCMD_DUMP(SZ_DMAEND, "\tDMAEND\n");
-
-	return SZ_DMAEND;
-}
-
-static inline u32 _emit_FLUSHP(unsigned dry_run, u8 buf[], u8 peri)
-{
-	if (dry_run)
-		return SZ_DMAFLUSHP;
-
-	buf[0] = CMD_DMAFLUSHP;
-
-	peri &= 0x1f;
-	peri <<= 3;
-	buf[1] = peri;
-
-	PL330_DBGCMD_DUMP(SZ_DMAFLUSHP, "\tDMAFLUSHP %u\n", peri >> 3);
-
-	return SZ_DMAFLUSHP;
-}
-
-static inline u32 _emit_LD(unsigned dry_run, u8 buf[],	enum pl330_cond cond)
-{
-	if (dry_run)
-		return SZ_DMALD;
-
-	buf[0] = CMD_DMALD;
-
-	if (cond == SINGLE)
-		buf[0] |= (0 << 1) | (1 << 0);
-	else if (cond == BURST)
-		buf[0] |= (1 << 1) | (1 << 0);
-
-	PL330_DBGCMD_DUMP(SZ_DMALD, "\tDMALD%c\n",
-		cond == SINGLE ? 'S' : (cond == BURST ? 'B' : 'A'));
-
-	return SZ_DMALD;
-}
-
-static inline u32 _emit_LDP(unsigned dry_run, u8 buf[],
-		enum pl330_cond cond, u8 peri)
-{
-	if (dry_run)
-		return SZ_DMALDP;
-
-	buf[0] = CMD_DMALDP;
-
-	if (cond == BURST)
-		buf[0] |= (1 << 1);
-
-	peri &= 0x1f;
-	peri <<= 3;
-	buf[1] = peri;
-
-	PL330_DBGCMD_DUMP(SZ_DMALDP, "\tDMALDP%c %u\n",
-		cond == SINGLE ? 'S' : 'B', peri >> 3);
-
-	return SZ_DMALDP;
-}
-
-static inline u32 _emit_LP(unsigned dry_run, u8 buf[],
-		unsigned loop, u8 cnt)
-{
-	if (dry_run)
-		return SZ_DMALP;
-
-	buf[0] = CMD_DMALP;
-
-	if (loop)
-		buf[0] |= (1 << 1);
-
-	cnt--; /* DMAC increments by 1 internally */
-	buf[1] = cnt;
-
-	PL330_DBGCMD_DUMP(SZ_DMALP, "\tDMALP_%c %u\n", loop ? '1' : '0', cnt);
-
-	return SZ_DMALP;
-}
-
-struct _arg_LPEND {
-	enum pl330_cond cond;
-	bool forever;
-	unsigned loop;
-	u8 bjump;
-};
-
-static inline u32 _emit_LPEND(unsigned dry_run, u8 buf[],
-		const struct _arg_LPEND *arg)
-{
-	enum pl330_cond cond = arg->cond;
-	bool forever = arg->forever;
-	unsigned loop = arg->loop;
-	u8 bjump = arg->bjump;
-
-	if (dry_run)
-		return SZ_DMALPEND;
-
-	buf[0] = CMD_DMALPEND;
-
-	if (loop)
-		buf[0] |= (1 << 2);
-
-	if (!forever)
-		buf[0] |= (1 << 4);
-
-	if (cond == SINGLE)
-		buf[0] |= (0 << 1) | (1 << 0);
-	else if (cond == BURST)
-		buf[0] |= (1 << 1) | (1 << 0);
-
-	buf[1] = bjump;
-
-	PL330_DBGCMD_DUMP(SZ_DMALPEND, "\tDMALP%s%c_%c bjmpto_%x\n",
-			forever ? "FE" : "END",
-			cond == SINGLE ? 'S' : (cond == BURST ? 'B' : 'A'),
-			loop ? '1' : '0',
-			bjump);
-
-	return SZ_DMALPEND;
-}
-
-static inline u32 _emit_KILL(unsigned dry_run, u8 buf[])
-{
-	if (dry_run)
-		return SZ_DMAKILL;
-
-	buf[0] = CMD_DMAKILL;
-
-	return SZ_DMAKILL;
-}
-
-static inline u32 _emit_MOV(unsigned dry_run, u8 buf[],
-		enum dmamov_dst dst, u32 val)
-{
-	if (dry_run)
-		return SZ_DMAMOV;
-
-	buf[0] = CMD_DMAMOV;
-	buf[1] = dst;
-	*((u32 *)&buf[2]) = val;
-
-	PL330_DBGCMD_DUMP(SZ_DMAMOV, "\tDMAMOV %s 0x%x\n",
-		dst == SAR ? "SAR" : (dst == DAR ? "DAR" : "CCR"), val);
-
-	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)
-		return SZ_DMARMB;
-
-	buf[0] = CMD_DMARMB;
-
-	PL330_DBGCMD_DUMP(SZ_DMARMB, "\tDMARMB\n");
-
-	return SZ_DMARMB;
-}
-
-static inline u32 _emit_SEV(unsigned dry_run, u8 buf[], u8 ev)
-{
-	if (dry_run)
-		return SZ_DMASEV;
-
-	buf[0] = CMD_DMASEV;
-
-	ev &= 0x1f;
-	ev <<= 3;
-	buf[1] = ev;
-
-	PL330_DBGCMD_DUMP(SZ_DMASEV, "\tDMASEV %u\n", ev >> 3);
-
-	return SZ_DMASEV;
-}
-
-static inline u32 _emit_ST(unsigned dry_run, u8 buf[], enum pl330_cond cond)
-{
-	if (dry_run)
-		return SZ_DMAST;
-
-	buf[0] = CMD_DMAST;
-
-	if (cond == SINGLE)
-		buf[0] |= (0 << 1) | (1 << 0);
-	else if (cond == BURST)
-		buf[0] |= (1 << 1) | (1 << 0);
-
-	PL330_DBGCMD_DUMP(SZ_DMAST, "\tDMAST%c\n",
-		cond == SINGLE ? 'S' : (cond == BURST ? 'B' : 'A'));
-
-	return SZ_DMAST;
-}
-
-static inline u32 _emit_STP(unsigned dry_run, u8 buf[],
-		enum pl330_cond cond, u8 peri)
-{
-	if (dry_run)
-		return SZ_DMASTP;
-
-	buf[0] = CMD_DMASTP;
-
-	if (cond == BURST)
-		buf[0] |= (1 << 1);
-
-	peri &= 0x1f;
-	peri <<= 3;
-	buf[1] = peri;
-
-	PL330_DBGCMD_DUMP(SZ_DMASTP, "\tDMASTP%c %u\n",
-		cond == SINGLE ? 'S' : 'B', peri >> 3);
-
-	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)
-{
-	if (dry_run)
-		return SZ_DMAWFP;
-
-	buf[0] = CMD_DMAWFP;
-
-	if (cond == SINGLE)
-		buf[0] |= (0 << 1) | (0 << 0);
-	else if (cond == BURST)
-		buf[0] |= (1 << 1) | (0 << 0);
-	else
-		buf[0] |= (0 << 1) | (1 << 0);
-
-	peri &= 0x1f;
-	peri <<= 3;
-	buf[1] = peri;
-
-	PL330_DBGCMD_DUMP(SZ_DMAWFP, "\tDMAWFP%c %u\n",
-		cond == SINGLE ? 'S' : (cond == BURST ? 'B' : 'P'), peri >> 3);
-
-	return SZ_DMAWFP;
-}
-
-static inline u32 _emit_WMB(unsigned dry_run, u8 buf[])
-{
-	if (dry_run)
-		return SZ_DMAWMB;
-
-	buf[0] = CMD_DMAWMB;
-
-	PL330_DBGCMD_DUMP(SZ_DMAWMB, "\tDMAWMB\n");
-
-	return SZ_DMAWMB;
-}
-
-struct _arg_GO {
-	u8 chan;
-	u32 addr;
-	unsigned ns;
-};
-
-static inline u32 _emit_GO(unsigned dry_run, u8 buf[],
-		const struct _arg_GO *arg)
-{
-	u8 chan = arg->chan;
-	u32 addr = arg->addr;
-	unsigned ns = arg->ns;
-
-	if (dry_run)
-		return SZ_DMAGO;
-
-	buf[0] = CMD_DMAGO;
-	buf[0] |= (ns << 1);
-
-	buf[1] = chan & 0x7;
-
-	*((u32 *)&buf[2]) = addr;
-
-	return SZ_DMAGO;
-}
-
-#define msecs_to_loops(t) (loops_per_jiffy / 1000 * HZ * t)
-
-/* Returns Time-Out */
-static bool _until_dmac_idle(struct pl330_thread *thrd)
-{
-	void __iomem *regs = thrd->dmac->pinfo->base;
-	unsigned long loops = msecs_to_loops(5);
-
-	do {
-		/* Until Manager is Idle */
-		if (!(readl(regs + DBGSTATUS) & DBG_BUSY))
-			break;
-
-		cpu_relax();
-	} while (--loops);
-
-	if (!loops)
-		return true;
-
-	return false;
-}
-
-static inline void _execute_DBGINSN(struct pl330_thread *thrd,
-		u8 insn[], bool as_manager)
-{
-	void __iomem *regs = thrd->dmac->pinfo->base;
-	u32 val;
-
-	val = (insn[0] << 16) | (insn[1] << 24);
-	if (!as_manager) {
-		val |= (1 << 0);
-		val |= (thrd->id << 8); /* Channel Number */
-	}
-	writel(val, regs + DBGINST0);
-
-	val = *((u32 *)&insn[2]);
-	writel(val, regs + DBGINST1);
-
-	/* If timed out due to halted state-machine */
-	if (_until_dmac_idle(thrd)) {
-		dev_err(thrd->dmac->pinfo->dev, "DMAC halted!\n");
-		return;
-	}
-
-	/* Get going */
-	writel(0, regs + DBGCMD);
-}
-
-/*
- * Mark a _pl330_req as free.
- * We do it by writing DMAEND as the first instruction
- * because no valid request is going to have DMAEND as
- * its first instruction to execute.
- */
-static void mark_free(struct pl330_thread *thrd, int idx)
-{
-	struct _pl330_req *req = &thrd->req[idx];
-
-	_emit_END(0, req->mc_cpu);
-	req->mc_len = 0;
-
-	thrd->req_running = -1;
-}
-
-static inline u32 _state(struct pl330_thread *thrd)
-{
-	void __iomem *regs = thrd->dmac->pinfo->base;
-	u32 val;
-
-	if (is_manager(thrd))
-		val = readl(regs + DS) & 0xf;
-	else
-		val = readl(regs + CS(thrd->id)) & 0xf;
-
-	switch (val) {
-	case DS_ST_STOP:
-		return PL330_STATE_STOPPED;
-	case DS_ST_EXEC:
-		return PL330_STATE_EXECUTING;
-	case DS_ST_CMISS:
-		return PL330_STATE_CACHEMISS;
-	case DS_ST_UPDTPC:
-		return PL330_STATE_UPDTPC;
-	case DS_ST_WFE:
-		return PL330_STATE_WFE;
-	case DS_ST_FAULT:
-		return PL330_STATE_FAULTING;
-	case DS_ST_ATBRR:
-		if (is_manager(thrd))
-			return PL330_STATE_INVALID;
-		else
-			return PL330_STATE_ATBARRIER;
-	case DS_ST_QBUSY:
-		if (is_manager(thrd))
-			return PL330_STATE_INVALID;
-		else
-			return PL330_STATE_QUEUEBUSY;
-	case DS_ST_WFP:
-		if (is_manager(thrd))
-			return PL330_STATE_INVALID;
-		else
-			return PL330_STATE_WFP;
-	case DS_ST_KILL:
-		if (is_manager(thrd))
-			return PL330_STATE_INVALID;
-		else
-			return PL330_STATE_KILLING;
-	case DS_ST_CMPLT:
-		if (is_manager(thrd))
-			return PL330_STATE_INVALID;
-		else
-			return PL330_STATE_COMPLETING;
-	case DS_ST_FLTCMP:
-		if (is_manager(thrd))
-			return PL330_STATE_INVALID;
-		else
-			return PL330_STATE_FAULT_COMPLETING;
-	default:
-		return PL330_STATE_INVALID;
-	}
-}
-
-static void _stop(struct pl330_thread *thrd)
-{
-	void __iomem *regs = thrd->dmac->pinfo->base;
-	u8 insn[6] = {0, 0, 0, 0, 0, 0};
-
-	if (_state(thrd) == PL330_STATE_FAULT_COMPLETING)
-		UNTIL(thrd, PL330_STATE_FAULTING | PL330_STATE_KILLING);
-
-	/* Return if nothing needs to be done */
-	if (_state(thrd) == PL330_STATE_COMPLETING
-		  || _state(thrd) == PL330_STATE_KILLING
-		  || _state(thrd) == PL330_STATE_STOPPED)
-		return;
-
-	_emit_KILL(0, insn);
-
-	/* Stop generating interrupts for SEV */
-	writel(readl(regs + INTEN) & ~(1 << thrd->ev), regs + INTEN);
-
-	_execute_DBGINSN(thrd, insn, is_manager(thrd));
-}
-
-/* Start doing req 'idx' of thread 'thrd' */
-static bool _trigger(struct pl330_thread *thrd)
-{
-	void __iomem *regs = thrd->dmac->pinfo->base;
-	struct _pl330_req *req;
-	struct pl330_req *r;
-	struct _arg_GO go;
-	unsigned ns;
-	u8 insn[6] = {0, 0, 0, 0, 0, 0};
-	int idx;
-
-	/* Return if already ACTIVE */
-	if (_state(thrd) != PL330_STATE_STOPPED)
-		return true;
-
-	idx = 1 - thrd->lstenq;
-	if (!IS_FREE(&thrd->req[idx]))
-		req = &thrd->req[idx];
-	else {
-		idx = thrd->lstenq;
-		if (!IS_FREE(&thrd->req[idx]))
-			req = &thrd->req[idx];
-		else
-			req = NULL;
-	}
-
-	/* Return if no request */
-	if (!req || !req->r)
-		return true;
-
-	r = req->r;
-
-	if (r->cfg)
-		ns = r->cfg->nonsecure ? 1 : 0;
-	else if (readl(regs + CS(thrd->id)) & CS_CNS)
-		ns = 1;
-	else
-		ns = 0;
-
-	/* See 'Abort Sources' point-4 at Page 2-25 */
-	if (_manager_ns(thrd) && !ns)
-		dev_info(thrd->dmac->pinfo->dev, "%s:%d Recipe for ABORT!\n",
-			__func__, __LINE__);
-
-	go.chan = thrd->id;
-	go.addr = req->mc_bus;
-	go.ns = ns;
-	_emit_GO(0, insn, &go);
-
-	/* Set to generate interrupts for SEV */
-	writel(readl(regs + INTEN) | (1 << thrd->ev), regs + INTEN);
-
-	/* Only manager can execute GO */
-	_execute_DBGINSN(thrd, insn, true);
-
-	thrd->req_running = idx;
-
-	return true;
-}
-
-static bool _start(struct pl330_thread *thrd)
-{
-	switch (_state(thrd)) {
-	case PL330_STATE_FAULT_COMPLETING:
-		UNTIL(thrd, PL330_STATE_FAULTING | PL330_STATE_KILLING);
-
-		if (_state(thrd) == PL330_STATE_KILLING)
-			UNTIL(thrd, PL330_STATE_STOPPED)
-
-	case PL330_STATE_FAULTING:
-		_stop(thrd);
-
-	case PL330_STATE_KILLING:
-	case PL330_STATE_COMPLETING:
-		UNTIL(thrd, PL330_STATE_STOPPED)
-
-	case PL330_STATE_STOPPED:
-		return _trigger(thrd);
-
-	case PL330_STATE_WFP:
-	case PL330_STATE_QUEUEBUSY:
-	case PL330_STATE_ATBARRIER:
-	case PL330_STATE_UPDTPC:
-	case PL330_STATE_CACHEMISS:
-	case PL330_STATE_EXECUTING:
-		return true;
-
-	case PL330_STATE_WFE: /* For RESUME, nothing yet */
-	default:
-		return false;
-	}
-}
-
-static inline int _ldst_memtomem(unsigned dry_run, u8 buf[],
-		const struct _xfer_spec *pxs, int cyc)
-{
-	int off = 0;
-
-	while (cyc--) {
-		off += _emit_LD(dry_run, &buf[off], ALWAYS);
-		off += _emit_RMB(dry_run, &buf[off]);
-		off += _emit_ST(dry_run, &buf[off], ALWAYS);
-		off += _emit_WMB(dry_run, &buf[off]);
-	}
-
-	return off;
-}
-
-static inline int _ldst_devtomem(unsigned dry_run, u8 buf[],
-		const struct _xfer_spec *pxs, int cyc)
-{
-	int off = 0;
-
-	while (cyc--) {
-		off += _emit_WFP(dry_run, &buf[off], SINGLE, pxs->r->peri);
-		off += _emit_LDP(dry_run, &buf[off], SINGLE, pxs->r->peri);
-		off += _emit_ST(dry_run, &buf[off], ALWAYS);
-		off += _emit_FLUSHP(dry_run, &buf[off], pxs->r->peri);
-	}
-
-	return off;
-}
-
-static inline int _ldst_memtodev(unsigned dry_run, u8 buf[],
-		const struct _xfer_spec *pxs, int cyc)
-{
-	int off = 0;
-
-	while (cyc--) {
-		off += _emit_WFP(dry_run, &buf[off], SINGLE, pxs->r->peri);
-		off += _emit_LD(dry_run, &buf[off], ALWAYS);
-		off += _emit_STP(dry_run, &buf[off], SINGLE, pxs->r->peri);
-		off += _emit_FLUSHP(dry_run, &buf[off], pxs->r->peri);
-	}
-
-	return off;
-}
-
-static int _bursts(unsigned dry_run, u8 buf[],
-		const struct _xfer_spec *pxs, int cyc)
-{
-	int off = 0;
-
-	switch (pxs->r->rqtype) {
-	case MEMTODEV:
-		off += _ldst_memtodev(dry_run, &buf[off], pxs, cyc);
-		break;
-	case DEVTOMEM:
-		off += _ldst_devtomem(dry_run, &buf[off], pxs, cyc);
-		break;
-	case MEMTOMEM:
-		off += _ldst_memtomem(dry_run, &buf[off], pxs, cyc);
-		break;
-	default:
-		off += 0x40000000; /* Scare off the Client */
-		break;
-	}
-
-	return off;
-}
-
-/* Returns bytes consumed and updates bursts */
-static inline int _loop(unsigned dry_run, u8 buf[],
-		unsigned long *bursts, const struct _xfer_spec *pxs)
-{
-	int cyc, cycmax, szlp, szlpend, szbrst, off;
-	unsigned lcnt0, lcnt1, ljmp0, ljmp1;
-	struct _arg_LPEND lpend;
-
-	/* Max iterations possible in DMALP is 256 */
-	if (*bursts >= 256*256) {
-		lcnt1 = 256;
-		lcnt0 = 256;
-		cyc = *bursts / lcnt1 / lcnt0;
-	} else if (*bursts > 256) {
-		lcnt1 = 256;
-		lcnt0 = *bursts / lcnt1;
-		cyc = 1;
-	} else {
-		lcnt1 = *bursts;
-		lcnt0 = 0;
-		cyc = 1;
-	}
-
-	szlp = _emit_LP(1, buf, 0, 0);
-	szbrst = _bursts(1, buf, pxs, 1);
-
-	lpend.cond = ALWAYS;
-	lpend.forever = false;
-	lpend.loop = 0;
-	lpend.bjump = 0;
-	szlpend = _emit_LPEND(1, buf, &lpend);
-
-	if (lcnt0) {
-		szlp *= 2;
-		szlpend *= 2;
-	}
-
-	/*
-	 * Max bursts that we can unroll due to limit on the
-	 * size of backward jump that can be encoded in DMALPEND
-	 * which is 8-bits and hence 255
-	 */
-	cycmax = (255 - (szlp + szlpend)) / szbrst;
-
-	cyc = (cycmax < cyc) ? cycmax : cyc;
-
-	off = 0;
-
-	if (lcnt0) {
-		off += _emit_LP(dry_run, &buf[off], 0, lcnt0);
-		ljmp0 = off;
-	}
-
-	off += _emit_LP(dry_run, &buf[off], 1, lcnt1);
-	ljmp1 = off;
-
-	off += _bursts(dry_run, &buf[off], pxs, cyc);
-
-	lpend.cond = ALWAYS;
-	lpend.forever = false;
-	lpend.loop = 1;
-	lpend.bjump = off - ljmp1;
-	off += _emit_LPEND(dry_run, &buf[off], &lpend);
-
-	if (lcnt0) {
-		lpend.cond = ALWAYS;
-		lpend.forever = false;
-		lpend.loop = 0;
-		lpend.bjump = off - ljmp0;
-		off += _emit_LPEND(dry_run, &buf[off], &lpend);
-	}
-
-	*bursts = lcnt1 * cyc;
-	if (lcnt0)
-		*bursts *= lcnt0;
-
-	return off;
-}
-
-static inline int _setup_loops(unsigned dry_run, u8 buf[],
-		const struct _xfer_spec *pxs)
-{
-	struct pl330_xfer *x = pxs->x;
-	u32 ccr = pxs->ccr;
-	unsigned long c, bursts = BYTE_TO_BURST(x->bytes, ccr);
-	int off = 0;
-
-	while (bursts) {
-		c = bursts;
-		off += _loop(dry_run, &buf[off], &c, pxs);
-		bursts -= c;
-	}
-
-	return off;
-}
-
-static inline int _setup_xfer(unsigned dry_run, u8 buf[],
-		const struct _xfer_spec *pxs)
-{
-	struct pl330_xfer *x = pxs->x;
-	int off = 0;
-
-	/* DMAMOV SAR, x->src_addr */
-	off += _emit_MOV(dry_run, &buf[off], SAR, x->src_addr);
-	/* DMAMOV DAR, x->dst_addr */
-	off += _emit_MOV(dry_run, &buf[off], DAR, x->dst_addr);
-
-	/* Setup Loop(s) */
-	off += _setup_loops(dry_run, &buf[off], pxs);
-
-	return off;
-}
-
-/*
- * A req is a sequence of one or more xfer units.
- * Returns the number of bytes taken to setup the MC for the req.
- */
-static int _setup_req(unsigned dry_run, struct pl330_thread *thrd,
-		unsigned index, struct _xfer_spec *pxs)
-{
-	struct _pl330_req *req = &thrd->req[index];
-	struct pl330_xfer *x;
-	u8 *buf = req->mc_cpu;
-	int off = 0;
-
-	PL330_DBGMC_START(req->mc_bus);
-
-	/* DMAMOV CCR, ccr */
-	off += _emit_MOV(dry_run, &buf[off], CCR, pxs->ccr);
-
-	x = pxs->r->x;
-	do {
-		/* Error if xfer length is not aligned at burst size */
-		if (x->bytes % (BRST_SIZE(pxs->ccr) * BRST_LEN(pxs->ccr)))
-			return -EINVAL;
-
-		pxs->x = x;
-		off += _setup_xfer(dry_run, &buf[off], pxs);
-
-		x = x->next;
-	} while (x);
-
-	/* DMASEV peripheral/event */
-	off += _emit_SEV(dry_run, &buf[off], thrd->ev);
-	/* DMAEND */
-	off += _emit_END(dry_run, &buf[off]);
-
-	return off;
-}
-
-static inline u32 _prepare_ccr(const struct pl330_reqcfg *rqc)
-{
-	u32 ccr = 0;
-
-	if (rqc->src_inc)
-		ccr |= CC_SRCINC;
-
-	if (rqc->dst_inc)
-		ccr |= CC_DSTINC;
-
-	/* We set same protection levels for Src and DST for now */
-	if (rqc->privileged)
-		ccr |= CC_SRCPRI | CC_DSTPRI;
-	if (rqc->nonsecure)
-		ccr |= CC_SRCNS | CC_DSTNS;
-	if (rqc->insnaccess)
-		ccr |= CC_SRCIA | CC_DSTIA;
-
-	ccr |= (((rqc->brst_len - 1) & 0xf) << CC_SRCBRSTLEN_SHFT);
-	ccr |= (((rqc->brst_len - 1) & 0xf) << CC_DSTBRSTLEN_SHFT);
-
-	ccr |= (rqc->brst_size << CC_SRCBRSTSIZE_SHFT);
-	ccr |= (rqc->brst_size << CC_DSTBRSTSIZE_SHFT);
-
-	ccr |= (rqc->scctl << CC_SRCCCTRL_SHFT);
-	ccr |= (rqc->dcctl << CC_DSTCCTRL_SHFT);
-
-	ccr |= (rqc->swap << CC_SWAP_SHFT);
-
-	return ccr;
-}
-
-static inline bool _is_valid(u32 ccr)
-{
-	enum pl330_dstcachectrl dcctl;
-	enum pl330_srccachectrl scctl;
-
-	dcctl = (ccr >> CC_DSTCCTRL_SHFT) & CC_DRCCCTRL_MASK;
-	scctl = (ccr >> CC_SRCCCTRL_SHFT) & CC_SRCCCTRL_MASK;
-
-	if (dcctl == DINVALID1 || dcctl == DINVALID2
-			|| scctl == SINVALID1 || scctl == SINVALID2)
-		return false;
-	else
-		return true;
-}
-
-/*
- * Submit a list of xfers after which the client wants notification.
- * Client is not notified after each xfer unit, just once after all
- * xfer units are done or some error occurs.
- */
-int pl330_submit_req(void *ch_id, struct pl330_req *r)
-{
-	struct pl330_thread *thrd = ch_id;
-	struct pl330_dmac *pl330;
-	struct pl330_info *pi;
-	struct _xfer_spec xs;
-	unsigned long flags;
-	void __iomem *regs;
-	unsigned idx;
-	u32 ccr;
-	int ret = 0;
-
-	/* No Req or Unacquired Channel or DMAC */
-	if (!r || !thrd || thrd->free)
-		return -EINVAL;
-
-	pl330 = thrd->dmac;
-	pi = pl330->pinfo;
-	regs = pi->base;
-
-	if (pl330->state == DYING
-		|| pl330->dmac_tbd.reset_chan & (1 << thrd->id)) {
-		dev_info(thrd->dmac->pinfo->dev, "%s:%d\n",
-			__func__, __LINE__);
-		return -EAGAIN;
-	}
-
-	/* If request for non-existing peripheral */
-	if (r->rqtype != MEMTOMEM && r->peri >= pi->pcfg.num_peri) {
-		dev_info(thrd->dmac->pinfo->dev,
-				"%s:%d Invalid peripheral(%u)!\n",
-				__func__, __LINE__, r->peri);
-		return -EINVAL;
-	}
-
-	spin_lock_irqsave(&pl330->lock, flags);
-
-	if (_queue_full(thrd)) {
-		ret = -EAGAIN;
-		goto xfer_exit;
-	}
-
-	/* Prefer Secure Channel */
-	if (!_manager_ns(thrd))
-		r->cfg->nonsecure = 0;
-	else
-		r->cfg->nonsecure = 1;
-
-	/* Use last settings, if not provided */
-	if (r->cfg)
-		ccr = _prepare_ccr(r->cfg);
-	else
-		ccr = readl(regs + CC(thrd->id));
-
-	/* If this req doesn't have valid xfer settings */
-	if (!_is_valid(ccr)) {
-		ret = -EINVAL;
-		dev_info(thrd->dmac->pinfo->dev, "%s:%d Invalid CCR(%x)!\n",
-			__func__, __LINE__, ccr);
-		goto xfer_exit;
-	}
-
-	idx = IS_FREE(&thrd->req[0]) ? 0 : 1;
-
-	xs.ccr = ccr;
-	xs.r = r;
-
-	/* First dry run to check if req is acceptable */
-	ret = _setup_req(1, thrd, idx, &xs);
-	if (ret < 0)
-		goto xfer_exit;
-
-	if (ret > pi->mcbufsz / 2) {
-		dev_info(thrd->dmac->pinfo->dev,
-			"%s:%d Trying increasing mcbufsz\n",
-				__func__, __LINE__);
-		ret = -ENOMEM;
-		goto xfer_exit;
-	}
-
-	/* Hook the request */
-	thrd->lstenq = idx;
-	thrd->req[idx].mc_len = _setup_req(0, thrd, idx, &xs);
-	thrd->req[idx].r = r;
-
-	ret = 0;
-
-xfer_exit:
-	spin_unlock_irqrestore(&pl330->lock, flags);
-
-	return ret;
-}
-EXPORT_SYMBOL(pl330_submit_req);
-
-static void pl330_dotask(unsigned long data)
-{
-	struct pl330_dmac *pl330 = (struct pl330_dmac *) data;
-	struct pl330_info *pi = pl330->pinfo;
-	unsigned long flags;
-	int i;
-
-	spin_lock_irqsave(&pl330->lock, flags);
-
-	/* The DMAC itself gone nuts */
-	if (pl330->dmac_tbd.reset_dmac) {
-		pl330->state = DYING;
-		/* Reset the manager too */
-		pl330->dmac_tbd.reset_mngr = true;
-		/* Clear the reset flag */
-		pl330->dmac_tbd.reset_dmac = false;
-	}
-
-	if (pl330->dmac_tbd.reset_mngr) {
-		_stop(pl330->manager);
-		/* Reset all channels */
-		pl330->dmac_tbd.reset_chan = (1 << pi->pcfg.num_chan) - 1;
-		/* Clear the reset flag */
-		pl330->dmac_tbd.reset_mngr = false;
-	}
-
-	for (i = 0; i < pi->pcfg.num_chan; i++) {
-
-		if (pl330->dmac_tbd.reset_chan & (1 << i)) {
-			struct pl330_thread *thrd = &pl330->channels[i];
-			void __iomem *regs = pi->base;
-			enum pl330_op_err err;
-
-			_stop(thrd);
-
-			if (readl(regs + FSC) & (1 << thrd->id))
-				err = PL330_ERR_FAIL;
-			else
-				err = PL330_ERR_ABORT;
-
-			spin_unlock_irqrestore(&pl330->lock, flags);
-
-			_callback(thrd->req[1 - thrd->lstenq].r, err);
-			_callback(thrd->req[thrd->lstenq].r, err);
-
-			spin_lock_irqsave(&pl330->lock, flags);
-
-			thrd->req[0].r = NULL;
-			thrd->req[1].r = NULL;
-			mark_free(thrd, 0);
-			mark_free(thrd, 1);
-
-			/* Clear the reset flag */
-			pl330->dmac_tbd.reset_chan &= ~(1 << i);
-		}
-	}
-
-	spin_unlock_irqrestore(&pl330->lock, flags);
-
-	return;
-}
-
-/* Returns 1 if state was updated, 0 otherwise */
-int pl330_update(const struct pl330_info *pi)
-{
-	struct _pl330_req *rqdone;
-	struct pl330_dmac *pl330;
-	unsigned long flags;
-	void __iomem *regs;
-	u32 val;
-	int id, ev, ret = 0;
-
-	if (!pi || !pi->pl330_data)
-		return 0;
-
-	regs = pi->base;
-	pl330 = pi->pl330_data;
-
-	spin_lock_irqsave(&pl330->lock, flags);
-
-	val = readl(regs + FSM) & 0x1;
-	if (val)
-		pl330->dmac_tbd.reset_mngr = true;
-	else
-		pl330->dmac_tbd.reset_mngr = false;
-
-	val = readl(regs + FSC) & ((1 << pi->pcfg.num_chan) - 1);
-	pl330->dmac_tbd.reset_chan |= val;
-	if (val) {
-		int i = 0;
-		while (i < pi->pcfg.num_chan) {
-			if (val & (1 << i)) {
-				dev_info(pi->dev,
-					"Reset Channel-%d\t CS-%x FTC-%x\n",
-						i, readl(regs + CS(i)),
-						readl(regs + FTC(i)));
-				_stop(&pl330->channels[i]);
-			}
-			i++;
-		}
-	}
-
-	/* Check which event happened i.e, thread notified */
-	val = readl(regs + ES);
-	if (pi->pcfg.num_events < 32
-			&& val & ~((1 << pi->pcfg.num_events) - 1)) {
-		pl330->dmac_tbd.reset_dmac = true;
-		dev_err(pi->dev, "%s:%d Unexpected!\n", __func__, __LINE__);
-		ret = 1;
-		goto updt_exit;
-	}
-
-	for (ev = 0; ev < pi->pcfg.num_events; ev++) {
-		if (val & (1 << ev)) { /* Event occurred */
-			struct pl330_thread *thrd;
-			u32 inten = readl(regs + INTEN);
-			int active;
-
-			/* Clear the event */
-			if (inten & (1 << ev))
-				writel(1 << ev, regs + INTCLR);
-
-			ret = 1;
-
-			id = pl330->events[ev];
-
-			thrd = &pl330->channels[id];
-
-			active = thrd->req_running;
-			if (active == -1) /* Aborted */
-				continue;
-
-			rqdone = &thrd->req[active];
-			mark_free(thrd, active);
-
-			/* Get going again ASAP */
-			_start(thrd);
-
-			/* For now, just make a list of callbacks to be done */
-			list_add_tail(&rqdone->rqd, &pl330->req_done);
-		}
-	}
-
-	/* Now that we are in no hurry, do the callbacks */
-	while (!list_empty(&pl330->req_done)) {
-		struct pl330_req *r;
-
-		rqdone = container_of(pl330->req_done.next,
-					struct _pl330_req, rqd);
-
-		list_del_init(&rqdone->rqd);
-
-		/* Detach the req */
-		r = rqdone->r;
-		rqdone->r = NULL;
-
-		spin_unlock_irqrestore(&pl330->lock, flags);
-		_callback(r, PL330_ERR_NONE);
-		spin_lock_irqsave(&pl330->lock, flags);
-	}
-
-updt_exit:
-	spin_unlock_irqrestore(&pl330->lock, flags);
-
-	if (pl330->dmac_tbd.reset_dmac
-			|| pl330->dmac_tbd.reset_mngr
-			|| pl330->dmac_tbd.reset_chan) {
-		ret = 1;
-		tasklet_schedule(&pl330->tasks);
-	}
-
-	return ret;
-}
-EXPORT_SYMBOL(pl330_update);
-
-int pl330_chan_ctrl(void *ch_id, enum pl330_chan_op op)
-{
-	struct pl330_thread *thrd = ch_id;
-	struct pl330_dmac *pl330;
-	unsigned long flags;
-	int ret = 0, active;
-
-	if (!thrd || thrd->free || thrd->dmac->state == DYING)
-		return -EINVAL;
-
-	pl330 = thrd->dmac;
-	active = thrd->req_running;
-
-	spin_lock_irqsave(&pl330->lock, flags);
-
-	switch (op) {
-	case PL330_OP_FLUSH:
-		/* Make sure the channel is stopped */
-		_stop(thrd);
-
-		thrd->req[0].r = NULL;
-		thrd->req[1].r = NULL;
-		mark_free(thrd, 0);
-		mark_free(thrd, 1);
-		break;
-
-	case PL330_OP_ABORT:
-		/* Make sure the channel is stopped */
-		_stop(thrd);
-
-		/* ABORT is only for the active req */
-		if (active == -1)
-			break;
-
-		thrd->req[active].r = NULL;
-		mark_free(thrd, active);
-
-		/* Start the next */
-	case PL330_OP_START:
-		if ((active == -1) && !_start(thrd))
-			ret = -EIO;
-		break;
-
-	default:
-		ret = -EINVAL;
-	}
-
-	spin_unlock_irqrestore(&pl330->lock, flags);
-	return ret;
-}
-EXPORT_SYMBOL(pl330_chan_ctrl);
-
-int pl330_chan_status(void *ch_id, struct pl330_chanstatus *pstatus)
-{
-	struct pl330_thread *thrd = ch_id;
-	struct pl330_dmac *pl330;
-	struct pl330_info *pi;
-	void __iomem *regs;
-	int active;
-	u32 val;
-
-	if (!pstatus || !thrd || thrd->free)
-		return -EINVAL;
-
-	pl330 = thrd->dmac;
-	pi = pl330->pinfo;
-	regs = pi->base;
-
-	/* The client should remove the DMAC and add again */
-	if (pl330->state == DYING)
-		pstatus->dmac_halted = true;
-	else
-		pstatus->dmac_halted = false;
-
-	val = readl(regs + FSC);
-	if (val & (1 << thrd->id))
-		pstatus->faulting = true;
-	else
-		pstatus->faulting = false;
-
-	active = thrd->req_running;
-
-	if (active == -1) {
-		/* Indicate that the thread is not running */
-		pstatus->top_req = NULL;
-		pstatus->wait_req = NULL;
-	} else {
-		pstatus->top_req = thrd->req[active].r;
-		pstatus->wait_req = !IS_FREE(&thrd->req[1 - active])
-					? thrd->req[1 - active].r : NULL;
-	}
-
-	pstatus->src_addr = readl(regs + SA(thrd->id));
-	pstatus->dst_addr = readl(regs + DA(thrd->id));
-
-	return 0;
-}
-EXPORT_SYMBOL(pl330_chan_status);
-
-/* Reserve an event */
-static inline int _alloc_event(struct pl330_thread *thrd)
-{
-	struct pl330_dmac *pl330 = thrd->dmac;
-	struct pl330_info *pi = pl330->pinfo;
-	int ev;
-
-	for (ev = 0; ev < pi->pcfg.num_events; ev++)
-		if (pl330->events[ev] == -1) {
-			pl330->events[ev] = thrd->id;
-			return ev;
-		}
-
-	return -1;
-}
-
-static bool _chan_ns(const struct pl330_info *pi, int i)
-{
-	return pi->pcfg.irq_ns & (1 << i);
-}
-
-/* Upon success, returns IdentityToken for the
- * allocated channel, NULL otherwise.
- */
-void *pl330_request_channel(const struct pl330_info *pi)
-{
-	struct pl330_thread *thrd = NULL;
-	struct pl330_dmac *pl330;
-	unsigned long flags;
-	int chans, i;
-
-	if (!pi || !pi->pl330_data)
-		return NULL;
-
-	pl330 = pi->pl330_data;
-
-	if (pl330->state == DYING)
-		return NULL;
-
-	chans = pi->pcfg.num_chan;
-
-	spin_lock_irqsave(&pl330->lock, flags);
-
-	for (i = 0; i < chans; i++) {
-		thrd = &pl330->channels[i];
-		if ((thrd->free) && (!_manager_ns(thrd) ||
-					_chan_ns(pi, i))) {
-			thrd->ev = _alloc_event(thrd);
-			if (thrd->ev >= 0) {
-				thrd->free = false;
-				thrd->lstenq = 1;
-				thrd->req[0].r = NULL;
-				mark_free(thrd, 0);
-				thrd->req[1].r = NULL;
-				mark_free(thrd, 1);
-				break;
-			}
-		}
-		thrd = NULL;
-	}
-
-	spin_unlock_irqrestore(&pl330->lock, flags);
-
-	return thrd;
-}
-EXPORT_SYMBOL(pl330_request_channel);
-
-/* Release an event */
-static inline void _free_event(struct pl330_thread *thrd, int ev)
-{
-	struct pl330_dmac *pl330 = thrd->dmac;
-	struct pl330_info *pi = pl330->pinfo;
-
-	/* If the event is valid and was held by the thread */
-	if (ev >= 0 && ev < pi->pcfg.num_events
-			&& pl330->events[ev] == thrd->id)
-		pl330->events[ev] = -1;
-}
-
-void pl330_release_channel(void *ch_id)
-{
-	struct pl330_thread *thrd = ch_id;
-	struct pl330_dmac *pl330;
-	unsigned long flags;
-
-	if (!thrd || thrd->free)
-		return;
-
-	_stop(thrd);
-
-	_callback(thrd->req[1 - thrd->lstenq].r, PL330_ERR_ABORT);
-	_callback(thrd->req[thrd->lstenq].r, PL330_ERR_ABORT);
-
-	pl330 = thrd->dmac;
-
-	spin_lock_irqsave(&pl330->lock, flags);
-	_free_event(thrd, thrd->ev);
-	thrd->free = true;
-	spin_unlock_irqrestore(&pl330->lock, flags);
-}
-EXPORT_SYMBOL(pl330_release_channel);
-
-/* Initialize the structure for PL330 configuration, that can be used
- * by the client driver the make best use of the DMAC
- */
-static void read_dmac_config(struct pl330_info *pi)
-{
-	void __iomem *regs = pi->base;
-	u32 val;
-
-	val = readl(regs + CRD) >> CRD_DATA_WIDTH_SHIFT;
-	val &= CRD_DATA_WIDTH_MASK;
-	pi->pcfg.data_bus_width = 8 * (1 << val);
-
-	val = readl(regs + CRD) >> CRD_DATA_BUFF_SHIFT;
-	val &= CRD_DATA_BUFF_MASK;
-	pi->pcfg.data_buf_dep = val + 1;
-
-	val = readl(regs + CR0) >> CR0_NUM_CHANS_SHIFT;
-	val &= CR0_NUM_CHANS_MASK;
-	val += 1;
-	pi->pcfg.num_chan = val;
-
-	val = readl(regs + CR0);
-	if (val & CR0_PERIPH_REQ_SET) {
-		val = (val >> CR0_NUM_PERIPH_SHIFT) & CR0_NUM_PERIPH_MASK;
-		val += 1;
-		pi->pcfg.num_peri = val;
-		pi->pcfg.peri_ns = readl(regs + CR4);
-	} else {
-		pi->pcfg.num_peri = 0;
-	}
-
-	val = readl(regs + CR0);
-	if (val & CR0_BOOT_MAN_NS)
-		pi->pcfg.mode |= DMAC_MODE_NS;
-	else
-		pi->pcfg.mode &= ~DMAC_MODE_NS;
-
-	val = readl(regs + CR0) >> CR0_NUM_EVENTS_SHIFT;
-	val &= CR0_NUM_EVENTS_MASK;
-	val += 1;
-	pi->pcfg.num_events = val;
-
-	pi->pcfg.irq_ns = readl(regs + CR3);
-
-	pi->pcfg.periph_id = get_id(pi, PERIPH_ID);
-	pi->pcfg.pcell_id = get_id(pi, PCELL_ID);
-}
-
-static inline void _reset_thread(struct pl330_thread *thrd)
-{
-	struct pl330_dmac *pl330 = thrd->dmac;
-	struct pl330_info *pi = pl330->pinfo;
-
-	thrd->req[0].mc_cpu = pl330->mcode_cpu
-				+ (thrd->id * pi->mcbufsz);
-	thrd->req[0].mc_bus = pl330->mcode_bus
-				+ (thrd->id * pi->mcbufsz);
-	thrd->req[0].r = NULL;
-	mark_free(thrd, 0);
-
-	thrd->req[1].mc_cpu = thrd->req[0].mc_cpu
-				+ pi->mcbufsz / 2;
-	thrd->req[1].mc_bus = thrd->req[0].mc_bus
-				+ pi->mcbufsz / 2;
-	thrd->req[1].r = NULL;
-	mark_free(thrd, 1);
-}
-
-static int dmac_alloc_threads(struct pl330_dmac *pl330)
-{
-	struct pl330_info *pi = pl330->pinfo;
-	int chans = pi->pcfg.num_chan;
-	struct pl330_thread *thrd;
-	int i;
-
-	/* Allocate 1 Manager and 'chans' Channel threads */
-	pl330->channels = kzalloc((1 + chans) * sizeof(*thrd),
-					GFP_KERNEL);
-	if (!pl330->channels)
-		return -ENOMEM;
-
-	/* Init Channel threads */
-	for (i = 0; i < chans; i++) {
-		thrd = &pl330->channels[i];
-		thrd->id = i;
-		thrd->dmac = pl330;
-		_reset_thread(thrd);
-		thrd->free = true;
-	}
-
-	/* MANAGER is indexed at the end */
-	thrd = &pl330->channels[chans];
-	thrd->id = chans;
-	thrd->dmac = pl330;
-	thrd->free = false;
-	pl330->manager = thrd;
-
-	return 0;
-}
-
-static int dmac_alloc_resources(struct pl330_dmac *pl330)
-{
-	struct pl330_info *pi = pl330->pinfo;
-	int chans = pi->pcfg.num_chan;
-	int ret;
-
-	/*
-	 * Alloc MicroCode buffer for 'chans' Channel threads.
-	 * A channel's buffer offset is (Channel_Id * MCODE_BUFF_PERCHAN)
-	 */
-	pl330->mcode_cpu = dma_alloc_coherent(pi->dev,
-				chans * pi->mcbufsz,
-				&pl330->mcode_bus, GFP_KERNEL);
-	if (!pl330->mcode_cpu) {
-		dev_err(pi->dev, "%s:%d Can't allocate memory!\n",
-			__func__, __LINE__);
-		return -ENOMEM;
-	}
-
-	ret = dmac_alloc_threads(pl330);
-	if (ret) {
-		dev_err(pi->dev, "%s:%d Can't to create channels for DMAC!\n",
-			__func__, __LINE__);
-		dma_free_coherent(pi->dev,
-				chans * pi->mcbufsz,
-				pl330->mcode_cpu, pl330->mcode_bus);
-		return ret;
-	}
-
-	return 0;
-}
-
-int pl330_add(struct pl330_info *pi)
-{
-	struct pl330_dmac *pl330;
-	void __iomem *regs;
-	int i, ret;
-
-	if (!pi || !pi->dev)
-		return -EINVAL;
-
-	/* If already added */
-	if (pi->pl330_data)
-		return -EINVAL;
-
-	/*
-	 * If the SoC can perform reset on the DMAC, then do it
-	 * before reading its configuration.
-	 */
-	if (pi->dmac_reset)
-		pi->dmac_reset(pi);
-
-	regs = pi->base;
-
-	/* Check if we can handle this DMAC */
-	if ((get_id(pi, PERIPH_ID) & 0xfffff) != PERIPH_ID_VAL
-	   || get_id(pi, PCELL_ID) != PCELL_ID_VAL) {
-		dev_err(pi->dev, "PERIPH_ID 0x%x, PCELL_ID 0x%x !\n",
-			get_id(pi, PERIPH_ID), get_id(pi, PCELL_ID));
-		return -EINVAL;
-	}
-
-	/* Read the configuration of the DMAC */
-	read_dmac_config(pi);
-
-	if (pi->pcfg.num_events == 0) {
-		dev_err(pi->dev, "%s:%d Can't work without events!\n",
-			__func__, __LINE__);
-		return -EINVAL;
-	}
-
-	pl330 = kzalloc(sizeof(*pl330), GFP_KERNEL);
-	if (!pl330) {
-		dev_err(pi->dev, "%s:%d Can't allocate memory!\n",
-			__func__, __LINE__);
-		return -ENOMEM;
-	}
-
-	/* Assign the info structure and private data */
-	pl330->pinfo = pi;
-	pi->pl330_data = pl330;
-
-	spin_lock_init(&pl330->lock);
-
-	INIT_LIST_HEAD(&pl330->req_done);
-
-	/* Use default MC buffer size if not provided */
-	if (!pi->mcbufsz)
-		pi->mcbufsz = MCODE_BUFF_PER_REQ * 2;
-
-	/* Mark all events as free */
-	for (i = 0; i < pi->pcfg.num_events; i++)
-		pl330->events[i] = -1;
-
-	/* Allocate resources needed by the DMAC */
-	ret = dmac_alloc_resources(pl330);
-	if (ret) {
-		dev_err(pi->dev, "Unable to create channels for DMAC\n");
-		kfree(pl330);
-		return ret;
-	}
-
-	tasklet_init(&pl330->tasks, pl330_dotask, (unsigned long) pl330);
-
-	pl330->state = INIT;
-
-	return 0;
-}
-EXPORT_SYMBOL(pl330_add);
-
-static int dmac_free_threads(struct pl330_dmac *pl330)
-{
-	struct pl330_info *pi = pl330->pinfo;
-	int chans = pi->pcfg.num_chan;
-	struct pl330_thread *thrd;
-	int i;
-
-	/* Release Channel threads */
-	for (i = 0; i < chans; i++) {
-		thrd = &pl330->channels[i];
-		pl330_release_channel((void *)thrd);
-	}
-
-	/* Free memory */
-	kfree(pl330->channels);
-
-	return 0;
-}
-
-static void dmac_free_resources(struct pl330_dmac *pl330)
-{
-	struct pl330_info *pi = pl330->pinfo;
-	int chans = pi->pcfg.num_chan;
-
-	dmac_free_threads(pl330);
-
-	dma_free_coherent(pi->dev, chans * pi->mcbufsz,
-				pl330->mcode_cpu, pl330->mcode_bus);
-}
-
-void pl330_del(struct pl330_info *pi)
-{
-	struct pl330_dmac *pl330;
-
-	if (!pi || !pi->pl330_data)
-		return;
-
-	pl330 = pi->pl330_data;
-
-	pl330->state = UNINIT;
-
-	tasklet_kill(&pl330->tasks);
-
-	/* Free DMAC resources */
-	dmac_free_resources(pl330);
-
-	kfree(pl330);
-	pi->pl330_data = NULL;
-}
-EXPORT_SYMBOL(pl330_del);
diff --git a/arch/arm/include/asm/hardware/iop_adma.h b/arch/arm/include/asm/hardware/iop_adma.h
index 59b8c38..122f86d 100644
--- a/arch/arm/include/asm/hardware/iop_adma.h
+++ b/arch/arm/include/asm/hardware/iop_adma.h
@@ -49,7 +49,6 @@
 /**
  * struct iop_adma_chan - internal representation of an ADMA device
  * @pending: allows batching of hardware operations
- * @completed_cookie: identifier for the most recently completed operation
  * @lock: serializes enqueue/dequeue operations to the slot pool
  * @mmr_base: memory mapped register base
  * @chain: device chain view of the descriptors
@@ -62,7 +61,6 @@
  */
 struct iop_adma_chan {
 	int pending;
-	dma_cookie_t completed_cookie;
 	spinlock_t lock; /* protects the descriptor slot pool */
 	void __iomem *mmr_base;
 	struct list_head chain;
diff --git a/arch/arm/include/asm/hardware/pl330.h b/arch/arm/include/asm/hardware/pl330.h
deleted file mode 100644
index c182138..0000000
--- a/arch/arm/include/asm/hardware/pl330.h
+++ /dev/null
@@ -1,217 +0,0 @@
-/* linux/include/asm/hardware/pl330.h
- *
- * Copyright (C) 2010 Samsung Electronics Co. Ltd.
- *	Jaswinder Singh <jassi.brar@samsung.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
- */
-
-#ifndef __PL330_CORE_H
-#define __PL330_CORE_H
-
-#define PL330_MAX_CHAN		8
-#define PL330_MAX_IRQS		32
-#define PL330_MAX_PERI		32
-
-enum pl330_srccachectrl {
-	SCCTRL0 = 0, /* Noncacheable and nonbufferable */
-	SCCTRL1, /* Bufferable only */
-	SCCTRL2, /* Cacheable, but do not allocate */
-	SCCTRL3, /* Cacheable and bufferable, but do not allocate */
-	SINVALID1,
-	SINVALID2,
-	SCCTRL6, /* Cacheable write-through, allocate on reads only */
-	SCCTRL7, /* Cacheable write-back, allocate on reads only */
-};
-
-enum pl330_dstcachectrl {
-	DCCTRL0 = 0, /* Noncacheable and nonbufferable */
-	DCCTRL1, /* Bufferable only */
-	DCCTRL2, /* Cacheable, but do not allocate */
-	DCCTRL3, /* Cacheable and bufferable, but do not allocate */
-	DINVALID1,              /* AWCACHE = 0x1000 */
-	DINVALID2,
-	DCCTRL6, /* Cacheable write-through, allocate on writes only */
-	DCCTRL7, /* Cacheable write-back, allocate on writes only */
-};
-
-/* Populated by the PL330 core driver for DMA API driver's info */
-struct pl330_config {
-	u32	periph_id;
-	u32	pcell_id;
-#define DMAC_MODE_NS	(1 << 0)
-	unsigned int	mode;
-	unsigned int	data_bus_width:10; /* In number of bits */
-	unsigned int	data_buf_dep:10;
-	unsigned int	num_chan:4;
-	unsigned int	num_peri:6;
-	u32		peri_ns;
-	unsigned int	num_events:6;
-	u32		irq_ns;
-};
-
-/* Handle to the DMAC provided to the PL330 core */
-struct pl330_info {
-	/* Owning device */
-	struct device *dev;
-	/* Size of MicroCode buffers for each channel. */
-	unsigned mcbufsz;
-	/* ioremap'ed address of PL330 registers. */
-	void __iomem	*base;
-	/* Client can freely use it. */
-	void	*client_data;
-	/* PL330 core data, Client must not touch it. */
-	void	*pl330_data;
-	/* Populated by the PL330 core driver during pl330_add */
-	struct pl330_config	pcfg;
-	/*
-	 * If the DMAC has some reset mechanism, then the
-	 * client may want to provide pointer to the method.
-	 */
-	void (*dmac_reset)(struct pl330_info *pi);
-};
-
-enum pl330_byteswap {
-	SWAP_NO = 0,
-	SWAP_2,
-	SWAP_4,
-	SWAP_8,
-	SWAP_16,
-};
-
-/**
- * Request Configuration.
- * The PL330 core does not modify this and uses the last
- * working configuration if the request doesn't provide any.
- *
- * The Client may want to provide this info only for the
- * first request and a request with new settings.
- */
-struct pl330_reqcfg {
-	/* Address Incrementing */
-	unsigned dst_inc:1;
-	unsigned src_inc:1;
-
-	/*
-	 * For now, the SRC & DST protection levels
-	 * and burst size/length are assumed same.
-	 */
-	bool nonsecure;
-	bool privileged;
-	bool insnaccess;
-	unsigned brst_len:5;
-	unsigned brst_size:3; /* in power of 2 */
-
-	enum pl330_dstcachectrl dcctl;
-	enum pl330_srccachectrl scctl;
-	enum pl330_byteswap swap;
-};
-
-/*
- * One cycle of DMAC operation.
- * There may be more than one xfer in a request.
- */
-struct pl330_xfer {
-	u32 src_addr;
-	u32 dst_addr;
-	/* Size to xfer */
-	u32 bytes;
-	/*
-	 * Pointer to next xfer in the list.
-	 * The last xfer in the req must point to NULL.
-	 */
-	struct pl330_xfer *next;
-};
-
-/* The xfer callbacks are made with one of these arguments. */
-enum pl330_op_err {
-	/* The all xfers in the request were success. */
-	PL330_ERR_NONE,
-	/* If req aborted due to global error. */
-	PL330_ERR_ABORT,
-	/* If req failed due to problem with Channel. */
-	PL330_ERR_FAIL,
-};
-
-enum pl330_reqtype {
-	MEMTOMEM,
-	MEMTODEV,
-	DEVTOMEM,
-	DEVTODEV,
-};
-
-/* A request defining Scatter-Gather List ending with NULL xfer. */
-struct pl330_req {
-	enum pl330_reqtype rqtype;
-	/* Index of peripheral for the xfer. */
-	unsigned peri:5;
-	/* Unique token for this xfer, set by the client. */
-	void *token;
-	/* Callback to be called after xfer. */
-	void (*xfer_cb)(void *token, enum pl330_op_err err);
-	/* If NULL, req will be done at last set parameters. */
-	struct pl330_reqcfg *cfg;
-	/* Pointer to first xfer in the request. */
-	struct pl330_xfer *x;
-};
-
-/*
- * To know the status of the channel and DMAC, the client
- * provides a pointer to this structure. The PL330 core
- * fills it with current information.
- */
-struct pl330_chanstatus {
-	/*
-	 * If the DMAC engine halted due to some error,
-	 * the client should remove-add DMAC.
-	 */
-	bool dmac_halted;
-	/*
-	 * If channel is halted due to some error,
-	 * the client should ABORT/FLUSH and START the channel.
-	 */
-	bool faulting;
-	/* Location of last load */
-	u32 src_addr;
-	/* Location of last store */
-	u32 dst_addr;
-	/*
-	 * Pointer to the currently active req, NULL if channel is
-	 * inactive, even though the requests may be present.
-	 */
-	struct pl330_req *top_req;
-	/* Pointer to req waiting second in the queue if any. */
-	struct pl330_req *wait_req;
-};
-
-enum pl330_chan_op {
-	/* Start the channel */
-	PL330_OP_START,
-	/* Abort the active xfer */
-	PL330_OP_ABORT,
-	/* Stop xfer and flush queue */
-	PL330_OP_FLUSH,
-};
-
-extern int pl330_add(struct pl330_info *);
-extern void pl330_del(struct pl330_info *pi);
-extern int pl330_update(const struct pl330_info *pi);
-extern void pl330_release_channel(void *ch_id);
-extern void *pl330_request_channel(const struct pl330_info *pi);
-extern int pl330_chan_status(void *ch_id, struct pl330_chanstatus *pstatus);
-extern int pl330_chan_ctrl(void *ch_id, enum pl330_chan_op op);
-extern int pl330_submit_req(void *ch_id, struct pl330_req *r);
-
-#endif	/* __PL330_CORE_H */
diff --git a/arch/arm/mach-at91/at91sam9g45_devices.c b/arch/arm/mach-at91/at91sam9g45_devices.c
index 4320b20..698479f 100644
--- a/arch/arm/mach-at91/at91sam9g45_devices.c
+++ b/arch/arm/mach-at91/at91sam9g45_devices.c
@@ -437,7 +437,6 @@
 
 	/* DMA slave channel configuration */
 	atslave->dma_dev = &at_hdmac_device.dev;
-	atslave->reg_width = AT_DMA_SLAVE_WIDTH_32BIT;
 	atslave->cfg = ATC_FIFOCFG_HALFFIFO
 			| ATC_SRC_H2SEL_HW | ATC_DST_H2SEL_HW;
 	atslave->ctrla = ATC_SCSIZE_16 | ATC_DCSIZE_16;
diff --git a/arch/arm/mach-at91/include/mach/at_hdmac.h b/arch/arm/mach-at91/include/mach/at_hdmac.h
index 187cb58..fff48d1 100644
--- a/arch/arm/mach-at91/include/mach/at_hdmac.h
+++ b/arch/arm/mach-at91/include/mach/at_hdmac.h
@@ -24,18 +24,6 @@
 };
 
 /**
- * enum at_dma_slave_width - DMA slave register access width.
- * @AT_DMA_SLAVE_WIDTH_8BIT: Do 8-bit slave register accesses
- * @AT_DMA_SLAVE_WIDTH_16BIT: Do 16-bit slave register accesses
- * @AT_DMA_SLAVE_WIDTH_32BIT: Do 32-bit slave register accesses
- */
-enum at_dma_slave_width {
-	AT_DMA_SLAVE_WIDTH_8BIT = 0,
-	AT_DMA_SLAVE_WIDTH_16BIT,
-	AT_DMA_SLAVE_WIDTH_32BIT,
-};
-
-/**
  * struct at_dma_slave - Controller-specific information about a slave
  * @dma_dev: required DMA master device
  * @tx_reg: physical address of data register used for
@@ -48,9 +36,6 @@
  */
 struct at_dma_slave {
 	struct device		*dma_dev;
-	dma_addr_t		tx_reg;
-	dma_addr_t		rx_reg;
-	enum at_dma_slave_width	reg_width;
 	u32			cfg;
 	u32			ctrla;
 };
diff --git a/arch/arm/mach-imx/Kconfig b/arch/arm/mach-imx/Kconfig
index 52359f8..7561eca 100644
--- a/arch/arm/mach-imx/Kconfig
+++ b/arch/arm/mach-imx/Kconfig
@@ -1,6 +1,3 @@
-config IMX_HAVE_DMA_V1
-	bool
-
 config HAVE_IMX_GPC
 	bool
 
@@ -38,7 +35,6 @@
 	bool
 	select ARCH_MX1
 	select CPU_ARM920T
-	select IMX_HAVE_DMA_V1
 	select IMX_HAVE_IOMUX_V1
 	select MXC_AVIC
 
@@ -46,7 +42,6 @@
 	bool
 	select MACH_MX21
 	select CPU_ARM926T
-	select IMX_HAVE_DMA_V1
 	select IMX_HAVE_IOMUX_V1
 	select MXC_AVIC
 
@@ -61,7 +56,6 @@
 	bool
 	select MACH_MX27
 	select CPU_ARM926T
-	select IMX_HAVE_DMA_V1
 	select IMX_HAVE_IOMUX_V1
 	select MXC_AVIC
 
diff --git a/arch/arm/mach-imx/Makefile b/arch/arm/mach-imx/Makefile
index 35fc450..ab939c5 100644
--- a/arch/arm/mach-imx/Makefile
+++ b/arch/arm/mach-imx/Makefile
@@ -1,5 +1,3 @@
-obj-$(CONFIG_IMX_HAVE_DMA_V1) += dma-v1.o
-
 obj-$(CONFIG_SOC_IMX1) += clock-imx1.o mm-imx1.o
 obj-$(CONFIG_SOC_IMX21) += clock-imx21.o mm-imx21.o
 
diff --git a/arch/arm/mach-imx/dma-v1.c b/arch/arm/mach-imx/dma-v1.c
deleted file mode 100644
index 3189a60..0000000
--- a/arch/arm/mach-imx/dma-v1.c
+++ /dev/null
@@ -1,845 +0,0 @@
-/*
- *  linux/arch/arm/plat-mxc/dma-v1.c
- *
- *  i.MX DMA registration and IRQ dispatching
- *
- * Copyright 2006 Pavel Pisa <pisa@cmp.felk.cvut.cz>
- * Copyright 2008 Juergen Beisert, <kernel@pengutronix.de>
- * Copyright 2008 Sascha Hauer, <s.hauer@pengutronix.de>
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version 2
- * of the License, or (at your option) any later version.
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
- */
-
-#include <linux/module.h>
-#include <linux/init.h>
-#include <linux/kernel.h>
-#include <linux/interrupt.h>
-#include <linux/err.h>
-#include <linux/errno.h>
-#include <linux/clk.h>
-#include <linux/scatterlist.h>
-#include <linux/io.h>
-
-#include <asm/irq.h>
-#include <mach/hardware.h>
-#include <mach/dma-v1.h>
-
-#define DMA_DCR     0x00		/* Control Register */
-#define DMA_DISR    0x04		/* Interrupt status Register */
-#define DMA_DIMR    0x08		/* Interrupt mask Register */
-#define DMA_DBTOSR  0x0c		/* Burst timeout status Register */
-#define DMA_DRTOSR  0x10		/* Request timeout Register */
-#define DMA_DSESR   0x14		/* Transfer Error Status Register */
-#define DMA_DBOSR   0x18		/* Buffer overflow status Register */
-#define DMA_DBTOCR  0x1c		/* Burst timeout control Register */
-#define DMA_WSRA    0x40		/* W-Size Register A */
-#define DMA_XSRA    0x44		/* X-Size Register A */
-#define DMA_YSRA    0x48		/* Y-Size Register A */
-#define DMA_WSRB    0x4c		/* W-Size Register B */
-#define DMA_XSRB    0x50		/* X-Size Register B */
-#define DMA_YSRB    0x54		/* Y-Size Register B */
-#define DMA_SAR(x)  (0x80 + ((x) << 6))	/* Source Address Registers */
-#define DMA_DAR(x)  (0x84 + ((x) << 6))	/* Destination Address Registers */
-#define DMA_CNTR(x) (0x88 + ((x) << 6))	/* Count Registers */
-#define DMA_CCR(x)  (0x8c + ((x) << 6))	/* Control Registers */
-#define DMA_RSSR(x) (0x90 + ((x) << 6))	/* Request source select Registers */
-#define DMA_BLR(x)  (0x94 + ((x) << 6))	/* Burst length Registers */
-#define DMA_RTOR(x) (0x98 + ((x) << 6))	/* Request timeout Registers */
-#define DMA_BUCR(x) (0x98 + ((x) << 6))	/* Bus Utilization Registers */
-#define DMA_CCNR(x) (0x9C + ((x) << 6))	/* Channel counter Registers */
-
-#define DCR_DRST           (1<<1)
-#define DCR_DEN            (1<<0)
-#define DBTOCR_EN          (1<<15)
-#define DBTOCR_CNT(x)      ((x) & 0x7fff)
-#define CNTR_CNT(x)        ((x) & 0xffffff)
-#define CCR_ACRPT          (1<<14)
-#define CCR_DMOD_LINEAR    (0x0 << 12)
-#define CCR_DMOD_2D        (0x1 << 12)
-#define CCR_DMOD_FIFO      (0x2 << 12)
-#define CCR_DMOD_EOBFIFO   (0x3 << 12)
-#define CCR_SMOD_LINEAR    (0x0 << 10)
-#define CCR_SMOD_2D        (0x1 << 10)
-#define CCR_SMOD_FIFO      (0x2 << 10)
-#define CCR_SMOD_EOBFIFO   (0x3 << 10)
-#define CCR_MDIR_DEC       (1<<9)
-#define CCR_MSEL_B         (1<<8)
-#define CCR_DSIZ_32        (0x0 << 6)
-#define CCR_DSIZ_8         (0x1 << 6)
-#define CCR_DSIZ_16        (0x2 << 6)
-#define CCR_SSIZ_32        (0x0 << 4)
-#define CCR_SSIZ_8         (0x1 << 4)
-#define CCR_SSIZ_16        (0x2 << 4)
-#define CCR_REN            (1<<3)
-#define CCR_RPT            (1<<2)
-#define CCR_FRC            (1<<1)
-#define CCR_CEN            (1<<0)
-#define RTOR_EN            (1<<15)
-#define RTOR_CLK           (1<<14)
-#define RTOR_PSC           (1<<13)
-
-/*
- * struct imx_dma_channel - i.MX specific DMA extension
- * @name: name specified by DMA client
- * @irq_handler: client callback for end of transfer
- * @err_handler: client callback for error condition
- * @data: clients context data for callbacks
- * @dma_mode: direction of the transfer %DMA_MODE_READ or %DMA_MODE_WRITE
- * @sg: pointer to the actual read/written chunk for scatter-gather emulation
- * @resbytes: total residual number of bytes to transfer
- *            (it can be lower or same as sum of SG mapped chunk sizes)
- * @sgcount: number of chunks to be read/written
- *
- * Structure is used for IMX DMA processing. It would be probably good
- * @struct dma_struct in the future for external interfacing and use
- * @struct imx_dma_channel only as extension to it.
- */
-
-struct imx_dma_channel {
-	const char *name;
-	void (*irq_handler) (int, void *);
-	void (*err_handler) (int, void *, int errcode);
-	void (*prog_handler) (int, void *, struct scatterlist *);
-	void *data;
-	unsigned int dma_mode;
-	struct scatterlist *sg;
-	unsigned int resbytes;
-	int dma_num;
-
-	int in_use;
-
-	u32 ccr_from_device;
-	u32 ccr_to_device;
-
-	struct timer_list watchdog;
-
-	int hw_chaining;
-};
-
-static void __iomem *imx_dmav1_baseaddr;
-
-static void imx_dmav1_writel(unsigned val, unsigned offset)
-{
-	__raw_writel(val, imx_dmav1_baseaddr + offset);
-}
-
-static unsigned imx_dmav1_readl(unsigned offset)
-{
-	return __raw_readl(imx_dmav1_baseaddr + offset);
-}
-
-static struct imx_dma_channel imx_dma_channels[IMX_DMA_CHANNELS];
-
-static struct clk *dma_clk;
-
-static int imx_dma_hw_chain(struct imx_dma_channel *imxdma)
-{
-	if (cpu_is_mx27())
-		return imxdma->hw_chaining;
-	else
-		return 0;
-}
-
-/*
- * imx_dma_sg_next - prepare next chunk for scatter-gather DMA emulation
- */
-static inline int imx_dma_sg_next(int channel, struct scatterlist *sg)
-{
-	struct imx_dma_channel *imxdma = &imx_dma_channels[channel];
-	unsigned long now;
-
-	if (!imxdma->name) {
-		printk(KERN_CRIT "%s: called for  not allocated channel %d\n",
-		       __func__, channel);
-		return 0;
-	}
-
-	now = min(imxdma->resbytes, sg->length);
-	if (imxdma->resbytes != IMX_DMA_LENGTH_LOOP)
-		imxdma->resbytes -= now;
-
-	if ((imxdma->dma_mode & DMA_MODE_MASK) == DMA_MODE_READ)
-		imx_dmav1_writel(sg->dma_address, DMA_DAR(channel));
-	else
-		imx_dmav1_writel(sg->dma_address, DMA_SAR(channel));
-
-	imx_dmav1_writel(now, DMA_CNTR(channel));
-
-	pr_debug("imxdma%d: next sg chunk dst 0x%08x, src 0x%08x, "
-		"size 0x%08x\n", channel,
-		 imx_dmav1_readl(DMA_DAR(channel)),
-		 imx_dmav1_readl(DMA_SAR(channel)),
-		 imx_dmav1_readl(DMA_CNTR(channel)));
-
-	return now;
-}
-
-/**
- * imx_dma_setup_single - setup i.MX DMA channel for linear memory to/from
- * device transfer
- *
- * @channel: i.MX DMA channel number
- * @dma_address: the DMA/physical memory address of the linear data block
- *		to transfer
- * @dma_length: length of the data block in bytes
- * @dev_addr: physical device port address
- * @dmamode: DMA transfer mode, %DMA_MODE_READ from the device to the memory
- *           or %DMA_MODE_WRITE from memory to the device
- *
- * Return value: if incorrect parameters are provided -%EINVAL.
- *		Zero indicates success.
- */
-int
-imx_dma_setup_single(int channel, dma_addr_t dma_address,
-		     unsigned int dma_length, unsigned int dev_addr,
-		     unsigned int dmamode)
-{
-	struct imx_dma_channel *imxdma = &imx_dma_channels[channel];
-
-	imxdma->sg = NULL;
-	imxdma->dma_mode = dmamode;
-
-	if (!dma_address) {
-		printk(KERN_ERR "imxdma%d: imx_dma_setup_single null address\n",
-		       channel);
-		return -EINVAL;
-	}
-
-	if (!dma_length) {
-		printk(KERN_ERR "imxdma%d: imx_dma_setup_single zero length\n",
-		       channel);
-		return -EINVAL;
-	}
-
-	if ((dmamode & DMA_MODE_MASK) == DMA_MODE_READ) {
-		pr_debug("imxdma%d: %s dma_addressg=0x%08x dma_length=%d "
-			"dev_addr=0x%08x for read\n",
-			channel, __func__, (unsigned int)dma_address,
-			dma_length, dev_addr);
-
-		imx_dmav1_writel(dev_addr, DMA_SAR(channel));
-		imx_dmav1_writel(dma_address, DMA_DAR(channel));
-		imx_dmav1_writel(imxdma->ccr_from_device, DMA_CCR(channel));
-	} else if ((dmamode & DMA_MODE_MASK) == DMA_MODE_WRITE) {
-		pr_debug("imxdma%d: %s dma_addressg=0x%08x dma_length=%d "
-			"dev_addr=0x%08x for write\n",
-			channel, __func__, (unsigned int)dma_address,
-			dma_length, dev_addr);
-
-		imx_dmav1_writel(dma_address, DMA_SAR(channel));
-		imx_dmav1_writel(dev_addr, DMA_DAR(channel));
-		imx_dmav1_writel(imxdma->ccr_to_device,
-				DMA_CCR(channel));
-	} else {
-		printk(KERN_ERR "imxdma%d: imx_dma_setup_single bad dmamode\n",
-		       channel);
-		return -EINVAL;
-	}
-
-	imx_dmav1_writel(dma_length, DMA_CNTR(channel));
-
-	return 0;
-}
-EXPORT_SYMBOL(imx_dma_setup_single);
-
-/**
- * imx_dma_setup_sg - setup i.MX DMA channel SG list to/from device transfer
- * @channel: i.MX DMA channel number
- * @sg: pointer to the scatter-gather list/vector
- * @sgcount: scatter-gather list hungs count
- * @dma_length: total length of the transfer request in bytes
- * @dev_addr: physical device port address
- * @dmamode: DMA transfer mode, %DMA_MODE_READ from the device to the memory
- *           or %DMA_MODE_WRITE from memory to the device
- *
- * The function sets up DMA channel state and registers to be ready for
- * transfer specified by provided parameters. The scatter-gather emulation
- * is set up according to the parameters.
- *
- * The full preparation of the transfer requires setup of more register
- * by the caller before imx_dma_enable() can be called.
- *
- * %BLR(channel) holds transfer burst length in bytes, 0 means 64 bytes
- *
- * %RSSR(channel) has to be set to the DMA request line source %DMA_REQ_xxx
- *
- * %CCR(channel) has to specify transfer parameters, the next settings is
- * typical for linear or simple scatter-gather transfers if %DMA_MODE_READ is
- * specified
- *
- * %CCR_DMOD_LINEAR | %CCR_DSIZ_32 | %CCR_SMOD_FIFO | %CCR_SSIZ_x
- *
- * The typical setup for %DMA_MODE_WRITE is specified by next options
- * combination
- *
- * %CCR_SMOD_LINEAR | %CCR_SSIZ_32 | %CCR_DMOD_FIFO | %CCR_DSIZ_x
- *
- * Be careful here and do not mistakenly mix source and target device
- * port sizes constants, they are really different:
- * %CCR_SSIZ_8, %CCR_SSIZ_16, %CCR_SSIZ_32,
- * %CCR_DSIZ_8, %CCR_DSIZ_16, %CCR_DSIZ_32
- *
- * Return value: if incorrect parameters are provided -%EINVAL.
- * Zero indicates success.
- */
-int
-imx_dma_setup_sg(int channel,
-		 struct scatterlist *sg, unsigned int sgcount,
-		 unsigned int dma_length, unsigned int dev_addr,
-		 unsigned int dmamode)
-{
-	struct imx_dma_channel *imxdma = &imx_dma_channels[channel];
-
-	if (imxdma->in_use)
-		return -EBUSY;
-
-	imxdma->sg = sg;
-	imxdma->dma_mode = dmamode;
-	imxdma->resbytes = dma_length;
-
-	if (!sg || !sgcount) {
-		printk(KERN_ERR "imxdma%d: imx_dma_setup_sg empty sg list\n",
-		       channel);
-		return -EINVAL;
-	}
-
-	if (!sg->length) {
-		printk(KERN_ERR "imxdma%d: imx_dma_setup_sg zero length\n",
-		       channel);
-		return -EINVAL;
-	}
-
-	if ((dmamode & DMA_MODE_MASK) == DMA_MODE_READ) {
-		pr_debug("imxdma%d: %s sg=%p sgcount=%d total length=%d "
-			"dev_addr=0x%08x for read\n",
-			channel, __func__, sg, sgcount, dma_length, dev_addr);
-
-		imx_dmav1_writel(dev_addr, DMA_SAR(channel));
-		imx_dmav1_writel(imxdma->ccr_from_device, DMA_CCR(channel));
-	} else if ((dmamode & DMA_MODE_MASK) == DMA_MODE_WRITE) {
-		pr_debug("imxdma%d: %s sg=%p sgcount=%d total length=%d "
-			"dev_addr=0x%08x for write\n",
-			channel, __func__, sg, sgcount, dma_length, dev_addr);
-
-		imx_dmav1_writel(dev_addr, DMA_DAR(channel));
-		imx_dmav1_writel(imxdma->ccr_to_device, DMA_CCR(channel));
-	} else {
-		printk(KERN_ERR "imxdma%d: imx_dma_setup_sg bad dmamode\n",
-		       channel);
-		return -EINVAL;
-	}
-
-	imx_dma_sg_next(channel, sg);
-
-	return 0;
-}
-EXPORT_SYMBOL(imx_dma_setup_sg);
-
-int
-imx_dma_config_channel(int channel, unsigned int config_port,
-	unsigned int config_mem, unsigned int dmareq, int hw_chaining)
-{
-	struct imx_dma_channel *imxdma = &imx_dma_channels[channel];
-	u32 dreq = 0;
-
-	imxdma->hw_chaining = 0;
-
-	if (hw_chaining) {
-		imxdma->hw_chaining = 1;
-		if (!imx_dma_hw_chain(imxdma))
-			return -EINVAL;
-	}
-
-	if (dmareq)
-		dreq = CCR_REN;
-
-	imxdma->ccr_from_device = config_port | (config_mem << 2) | dreq;
-	imxdma->ccr_to_device = config_mem | (config_port << 2) | dreq;
-
-	imx_dmav1_writel(dmareq, DMA_RSSR(channel));
-
-	return 0;
-}
-EXPORT_SYMBOL(imx_dma_config_channel);
-
-void imx_dma_config_burstlen(int channel, unsigned int burstlen)
-{
-	imx_dmav1_writel(burstlen, DMA_BLR(channel));
-}
-EXPORT_SYMBOL(imx_dma_config_burstlen);
-
-/**
- * imx_dma_setup_handlers - setup i.MX DMA channel end and error notification
- * handlers
- * @channel: i.MX DMA channel number
- * @irq_handler: the pointer to the function called if the transfer
- *		ends successfully
- * @err_handler: the pointer to the function called if the premature
- *		end caused by error occurs
- * @data: user specified value to be passed to the handlers
- */
-int
-imx_dma_setup_handlers(int channel,
-		       void (*irq_handler) (int, void *),
-		       void (*err_handler) (int, void *, int),
-		       void *data)
-{
-	struct imx_dma_channel *imxdma = &imx_dma_channels[channel];
-	unsigned long flags;
-
-	if (!imxdma->name) {
-		printk(KERN_CRIT "%s: called for  not allocated channel %d\n",
-		       __func__, channel);
-		return -ENODEV;
-	}
-
-	local_irq_save(flags);
-	imx_dmav1_writel(1 << channel, DMA_DISR);
-	imxdma->irq_handler = irq_handler;
-	imxdma->err_handler = err_handler;
-	imxdma->data = data;
-	local_irq_restore(flags);
-	return 0;
-}
-EXPORT_SYMBOL(imx_dma_setup_handlers);
-
-/**
- * imx_dma_setup_progression_handler - setup i.MX DMA channel progression
- * handlers
- * @channel: i.MX DMA channel number
- * @prog_handler: the pointer to the function called if the transfer progresses
- */
-int
-imx_dma_setup_progression_handler(int channel,
-			void (*prog_handler) (int, void*, struct scatterlist*))
-{
-	struct imx_dma_channel *imxdma = &imx_dma_channels[channel];
-	unsigned long flags;
-
-	if (!imxdma->name) {
-		printk(KERN_CRIT "%s: called for  not allocated channel %d\n",
-		       __func__, channel);
-		return -ENODEV;
-	}
-
-	local_irq_save(flags);
-	imxdma->prog_handler = prog_handler;
-	local_irq_restore(flags);
-	return 0;
-}
-EXPORT_SYMBOL(imx_dma_setup_progression_handler);
-
-/**
- * imx_dma_enable - function to start i.MX DMA channel operation
- * @channel: i.MX DMA channel number
- *
- * The channel has to be allocated by driver through imx_dma_request()
- * or imx_dma_request_by_prio() function.
- * The transfer parameters has to be set to the channel registers through
- * call of the imx_dma_setup_single() or imx_dma_setup_sg() function
- * and registers %BLR(channel), %RSSR(channel) and %CCR(channel) has to
- * be set prior this function call by the channel user.
- */
-void imx_dma_enable(int channel)
-{
-	struct imx_dma_channel *imxdma = &imx_dma_channels[channel];
-	unsigned long flags;
-
-	pr_debug("imxdma%d: imx_dma_enable\n", channel);
-
-	if (!imxdma->name) {
-		printk(KERN_CRIT "%s: called for  not allocated channel %d\n",
-		       __func__, channel);
-		return;
-	}
-
-	if (imxdma->in_use)
-		return;
-
-	local_irq_save(flags);
-
-	imx_dmav1_writel(1 << channel, DMA_DISR);
-	imx_dmav1_writel(imx_dmav1_readl(DMA_DIMR) & ~(1 << channel), DMA_DIMR);
-	imx_dmav1_writel(imx_dmav1_readl(DMA_CCR(channel)) | CCR_CEN |
-		CCR_ACRPT, DMA_CCR(channel));
-
-	if ((cpu_is_mx21() || cpu_is_mx27()) &&
-			imxdma->sg && imx_dma_hw_chain(imxdma)) {
-		imxdma->sg = sg_next(imxdma->sg);
-		if (imxdma->sg) {
-			u32 tmp;
-			imx_dma_sg_next(channel, imxdma->sg);
-			tmp = imx_dmav1_readl(DMA_CCR(channel));
-			imx_dmav1_writel(tmp | CCR_RPT | CCR_ACRPT,
-				DMA_CCR(channel));
-		}
-	}
-	imxdma->in_use = 1;
-
-	local_irq_restore(flags);
-}
-EXPORT_SYMBOL(imx_dma_enable);
-
-/**
- * imx_dma_disable - stop, finish i.MX DMA channel operatin
- * @channel: i.MX DMA channel number
- */
-void imx_dma_disable(int channel)
-{
-	struct imx_dma_channel *imxdma = &imx_dma_channels[channel];
-	unsigned long flags;
-
-	pr_debug("imxdma%d: imx_dma_disable\n", channel);
-
-	if (imx_dma_hw_chain(imxdma))
-		del_timer(&imxdma->watchdog);
-
-	local_irq_save(flags);
-	imx_dmav1_writel(imx_dmav1_readl(DMA_DIMR) | (1 << channel), DMA_DIMR);
-	imx_dmav1_writel(imx_dmav1_readl(DMA_CCR(channel)) & ~CCR_CEN,
-			DMA_CCR(channel));
-	imx_dmav1_writel(1 << channel, DMA_DISR);
-	imxdma->in_use = 0;
-	local_irq_restore(flags);
-}
-EXPORT_SYMBOL(imx_dma_disable);
-
-static void imx_dma_watchdog(unsigned long chno)
-{
-	struct imx_dma_channel *imxdma = &imx_dma_channels[chno];
-
-	imx_dmav1_writel(0, DMA_CCR(chno));
-	imxdma->in_use = 0;
-	imxdma->sg = NULL;
-
-	if (imxdma->err_handler)
-		imxdma->err_handler(chno, imxdma->data, IMX_DMA_ERR_TIMEOUT);
-}
-
-static irqreturn_t dma_err_handler(int irq, void *dev_id)
-{
-	int i, disr;
-	struct imx_dma_channel *imxdma;
-	unsigned int err_mask;
-	int errcode;
-
-	disr = imx_dmav1_readl(DMA_DISR);
-
-	err_mask = imx_dmav1_readl(DMA_DBTOSR) |
-		   imx_dmav1_readl(DMA_DRTOSR) |
-		   imx_dmav1_readl(DMA_DSESR)  |
-		   imx_dmav1_readl(DMA_DBOSR);
-
-	if (!err_mask)
-		return IRQ_HANDLED;
-
-	imx_dmav1_writel(disr & err_mask, DMA_DISR);
-
-	for (i = 0; i < IMX_DMA_CHANNELS; i++) {
-		if (!(err_mask & (1 << i)))
-			continue;
-		imxdma = &imx_dma_channels[i];
-		errcode = 0;
-
-		if (imx_dmav1_readl(DMA_DBTOSR) & (1 << i)) {
-			imx_dmav1_writel(1 << i, DMA_DBTOSR);
-			errcode |= IMX_DMA_ERR_BURST;
-		}
-		if (imx_dmav1_readl(DMA_DRTOSR) & (1 << i)) {
-			imx_dmav1_writel(1 << i, DMA_DRTOSR);
-			errcode |= IMX_DMA_ERR_REQUEST;
-		}
-		if (imx_dmav1_readl(DMA_DSESR) & (1 << i)) {
-			imx_dmav1_writel(1 << i, DMA_DSESR);
-			errcode |= IMX_DMA_ERR_TRANSFER;
-		}
-		if (imx_dmav1_readl(DMA_DBOSR) & (1 << i)) {
-			imx_dmav1_writel(1 << i, DMA_DBOSR);
-			errcode |= IMX_DMA_ERR_BUFFER;
-		}
-		if (imxdma->name && imxdma->err_handler) {
-			imxdma->err_handler(i, imxdma->data, errcode);
-			continue;
-		}
-
-		imx_dma_channels[i].sg = NULL;
-
-		printk(KERN_WARNING
-		       "DMA timeout on channel %d (%s) -%s%s%s%s\n",
-		       i, imxdma->name,
-		       errcode & IMX_DMA_ERR_BURST ?    " burst" : "",
-		       errcode & IMX_DMA_ERR_REQUEST ?  " request" : "",
-		       errcode & IMX_DMA_ERR_TRANSFER ? " transfer" : "",
-		       errcode & IMX_DMA_ERR_BUFFER ?   " buffer" : "");
-	}
-	return IRQ_HANDLED;
-}
-
-static void dma_irq_handle_channel(int chno)
-{
-	struct imx_dma_channel *imxdma = &imx_dma_channels[chno];
-
-	if (!imxdma->name) {
-		/*
-		 * IRQ for an unregistered DMA channel:
-		 * let's clear the interrupts and disable it.
-		 */
-		printk(KERN_WARNING
-		       "spurious IRQ for DMA channel %d\n", chno);
-		return;
-	}
-
-	if (imxdma->sg) {
-		u32 tmp;
-		struct scatterlist *current_sg = imxdma->sg;
-		imxdma->sg = sg_next(imxdma->sg);
-
-		if (imxdma->sg) {
-			imx_dma_sg_next(chno, imxdma->sg);
-
-			tmp = imx_dmav1_readl(DMA_CCR(chno));
-
-			if (imx_dma_hw_chain(imxdma)) {
-				/* FIXME: The timeout should probably be
-				 * configurable
-				 */
-				mod_timer(&imxdma->watchdog,
-					jiffies + msecs_to_jiffies(500));
-
-				tmp |= CCR_CEN | CCR_RPT | CCR_ACRPT;
-				imx_dmav1_writel(tmp, DMA_CCR(chno));
-			} else {
-				imx_dmav1_writel(tmp & ~CCR_CEN, DMA_CCR(chno));
-				tmp |= CCR_CEN;
-			}
-
-			imx_dmav1_writel(tmp, DMA_CCR(chno));
-
-			if (imxdma->prog_handler)
-				imxdma->prog_handler(chno, imxdma->data,
-						current_sg);
-
-			return;
-		}
-
-		if (imx_dma_hw_chain(imxdma)) {
-			del_timer(&imxdma->watchdog);
-			return;
-		}
-	}
-
-	imx_dmav1_writel(0, DMA_CCR(chno));
-	imxdma->in_use = 0;
-	if (imxdma->irq_handler)
-		imxdma->irq_handler(chno, imxdma->data);
-}
-
-static irqreturn_t dma_irq_handler(int irq, void *dev_id)
-{
-	int i, disr;
-
-	if (cpu_is_mx21() || cpu_is_mx27())
-		dma_err_handler(irq, dev_id);
-
-	disr = imx_dmav1_readl(DMA_DISR);
-
-	pr_debug("imxdma: dma_irq_handler called, disr=0x%08x\n",
-		     disr);
-
-	imx_dmav1_writel(disr, DMA_DISR);
-	for (i = 0; i < IMX_DMA_CHANNELS; i++) {
-		if (disr & (1 << i))
-			dma_irq_handle_channel(i);
-	}
-
-	return IRQ_HANDLED;
-}
-
-/**
- * imx_dma_request - request/allocate specified channel number
- * @channel: i.MX DMA channel number
- * @name: the driver/caller own non-%NULL identification
- */
-int imx_dma_request(int channel, const char *name)
-{
-	struct imx_dma_channel *imxdma = &imx_dma_channels[channel];
-	unsigned long flags;
-	int ret = 0;
-
-	/* basic sanity checks */
-	if (!name)
-		return -EINVAL;
-
-	if (channel >= IMX_DMA_CHANNELS) {
-		printk(KERN_CRIT "%s: called for  non-existed channel %d\n",
-		       __func__, channel);
-		return -EINVAL;
-	}
-
-	local_irq_save(flags);
-	if (imxdma->name) {
-		local_irq_restore(flags);
-		return -EBUSY;
-	}
-	memset(imxdma, 0, sizeof(*imxdma));
-	imxdma->name = name;
-	local_irq_restore(flags); /* request_irq() can block */
-
-	if (cpu_is_mx21() || cpu_is_mx27()) {
-		ret = request_irq(MX2x_INT_DMACH0 + channel,
-				dma_irq_handler, 0, "DMA", NULL);
-		if (ret) {
-			imxdma->name = NULL;
-			pr_crit("Can't register IRQ %d for DMA channel %d\n",
-					MX2x_INT_DMACH0 + channel, channel);
-			return ret;
-		}
-		init_timer(&imxdma->watchdog);
-		imxdma->watchdog.function = &imx_dma_watchdog;
-		imxdma->watchdog.data = channel;
-	}
-
-	return ret;
-}
-EXPORT_SYMBOL(imx_dma_request);
-
-/**
- * imx_dma_free - release previously acquired channel
- * @channel: i.MX DMA channel number
- */
-void imx_dma_free(int channel)
-{
-	unsigned long flags;
-	struct imx_dma_channel *imxdma = &imx_dma_channels[channel];
-
-	if (!imxdma->name) {
-		printk(KERN_CRIT
-		       "%s: trying to free free channel %d\n",
-		       __func__, channel);
-		return;
-	}
-
-	local_irq_save(flags);
-	/* Disable interrupts */
-	imx_dma_disable(channel);
-	imxdma->name = NULL;
-
-	if (cpu_is_mx21() || cpu_is_mx27())
-		free_irq(MX2x_INT_DMACH0 + channel, NULL);
-
-	local_irq_restore(flags);
-}
-EXPORT_SYMBOL(imx_dma_free);
-
-/**
- * imx_dma_request_by_prio - find and request some of free channels best
- * suiting requested priority
- * @channel: i.MX DMA channel number
- * @name: the driver/caller own non-%NULL identification
- *
- * This function tries to find a free channel in the specified priority group
- * if the priority cannot be achieved it tries to look for free channel
- * in the higher and then even lower priority groups.
- *
- * Return value: If there is no free channel to allocate, -%ENODEV is returned.
- *               On successful allocation channel is returned.
- */
-int imx_dma_request_by_prio(const char *name, enum imx_dma_prio prio)
-{
-	int i;
-	int best;
-
-	switch (prio) {
-	case (DMA_PRIO_HIGH):
-		best = 8;
-		break;
-	case (DMA_PRIO_MEDIUM):
-		best = 4;
-		break;
-	case (DMA_PRIO_LOW):
-	default:
-		best = 0;
-		break;
-	}
-
-	for (i = best; i < IMX_DMA_CHANNELS; i++)
-		if (!imx_dma_request(i, name))
-			return i;
-
-	for (i = best - 1; i >= 0; i--)
-		if (!imx_dma_request(i, name))
-			return i;
-
-	printk(KERN_ERR "%s: no free DMA channel found\n", __func__);
-
-	return -ENODEV;
-}
-EXPORT_SYMBOL(imx_dma_request_by_prio);
-
-static int __init imx_dma_init(void)
-{
-	int ret = 0;
-	int i;
-
-	if (cpu_is_mx1())
-		imx_dmav1_baseaddr = MX1_IO_ADDRESS(MX1_DMA_BASE_ADDR);
-	else if (cpu_is_mx21())
-		imx_dmav1_baseaddr = MX21_IO_ADDRESS(MX21_DMA_BASE_ADDR);
-	else if (cpu_is_mx27())
-		imx_dmav1_baseaddr = MX27_IO_ADDRESS(MX27_DMA_BASE_ADDR);
-	else
-		return 0;
-
-	dma_clk = clk_get(NULL, "dma");
-	if (IS_ERR(dma_clk))
-		return PTR_ERR(dma_clk);
-	clk_enable(dma_clk);
-
-	/* reset DMA module */
-	imx_dmav1_writel(DCR_DRST, DMA_DCR);
-
-	if (cpu_is_mx1()) {
-		ret = request_irq(MX1_DMA_INT, dma_irq_handler, 0, "DMA", NULL);
-		if (ret) {
-			pr_crit("Wow!  Can't register IRQ for DMA\n");
-			return ret;
-		}
-
-		ret = request_irq(MX1_DMA_ERR, dma_err_handler, 0, "DMA", NULL);
-		if (ret) {
-			pr_crit("Wow!  Can't register ERRIRQ for DMA\n");
-			free_irq(MX1_DMA_INT, NULL);
-			return ret;
-		}
-	}
-
-	/* enable DMA module */
-	imx_dmav1_writel(DCR_DEN, DMA_DCR);
-
-	/* clear all interrupts */
-	imx_dmav1_writel((1 << IMX_DMA_CHANNELS) - 1, DMA_DISR);
-
-	/* disable interrupts */
-	imx_dmav1_writel((1 << IMX_DMA_CHANNELS) - 1, DMA_DIMR);
-
-	for (i = 0; i < IMX_DMA_CHANNELS; i++) {
-		imx_dma_channels[i].sg = NULL;
-		imx_dma_channels[i].dma_num = i;
-	}
-
-	return ret;
-}
-
-arch_initcall(imx_dma_init);
diff --git a/arch/arm/mach-imx/include/mach/dma-v1.h b/arch/arm/mach-imx/include/mach/dma-v1.h
deleted file mode 100644
index ac6fd71..0000000
--- a/arch/arm/mach-imx/include/mach/dma-v1.h
+++ /dev/null
@@ -1,103 +0,0 @@
-/*
- *  linux/arch/arm/mach-imx/include/mach/dma-v1.h
- *
- *  i.MX DMA registration and IRQ dispatching
- *
- * Copyright 2006 Pavel Pisa <pisa@cmp.felk.cvut.cz>
- * Copyright 2008 Juergen Beisert, <kernel@pengutronix.de>
- * Copyright 2008 Sascha Hauer, <s.hauer@pengutronix.de>
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version 2
- * of the License, or (at your option) any later version.
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
- */
-
-#ifndef __MACH_DMA_V1_H__
-#define __MACH_DMA_V1_H__
-
-#define imx_has_dma_v1()	(cpu_is_mx1() || cpu_is_mx21() || cpu_is_mx27())
-
-#include <mach/dma.h>
-
-#define IMX_DMA_CHANNELS  16
-
-#define DMA_MODE_READ		0
-#define DMA_MODE_WRITE		1
-#define DMA_MODE_MASK		1
-
-#define MX1_DMA_REG(offset)	MX1_IO_ADDRESS(MX1_DMA_BASE_ADDR + (offset))
-
-/* DMA Interrupt Mask Register */
-#define MX1_DMA_DIMR		MX1_DMA_REG(0x08)
-
-/* Channel Control Register */
-#define MX1_DMA_CCR(x)		MX1_DMA_REG(0x8c + ((x) << 6))
-
-#define IMX_DMA_MEMSIZE_32	(0 << 4)
-#define IMX_DMA_MEMSIZE_8	(1 << 4)
-#define IMX_DMA_MEMSIZE_16	(2 << 4)
-#define IMX_DMA_TYPE_LINEAR	(0 << 10)
-#define IMX_DMA_TYPE_2D		(1 << 10)
-#define IMX_DMA_TYPE_FIFO	(2 << 10)
-
-#define IMX_DMA_ERR_BURST     (1 << 0)
-#define IMX_DMA_ERR_REQUEST   (1 << 1)
-#define IMX_DMA_ERR_TRANSFER  (1 << 2)
-#define IMX_DMA_ERR_BUFFER    (1 << 3)
-#define IMX_DMA_ERR_TIMEOUT   (1 << 4)
-
-int
-imx_dma_config_channel(int channel, unsigned int config_port,
-	unsigned int config_mem, unsigned int dmareq, int hw_chaining);
-
-void
-imx_dma_config_burstlen(int channel, unsigned int burstlen);
-
-int
-imx_dma_setup_single(int channel, dma_addr_t dma_address,
-		unsigned int dma_length, unsigned int dev_addr,
-		unsigned int dmamode);
-
-
-/*
- * Use this flag as the dma_length argument to imx_dma_setup_sg()
- * to create an endless running dma loop. The end of the scatterlist
- * must be linked to the beginning for this to work.
- */
-#define IMX_DMA_LENGTH_LOOP	((unsigned int)-1)
-
-int
-imx_dma_setup_sg(int channel, struct scatterlist *sg,
-		unsigned int sgcount, unsigned int dma_length,
-		unsigned int dev_addr, unsigned int dmamode);
-
-int
-imx_dma_setup_handlers(int channel,
-		void (*irq_handler) (int, void *),
-		void (*err_handler) (int, void *, int), void *data);
-
-int
-imx_dma_setup_progression_handler(int channel,
-		void (*prog_handler) (int, void*, struct scatterlist*));
-
-void imx_dma_enable(int channel);
-
-void imx_dma_disable(int channel);
-
-int imx_dma_request(int channel, const char *name);
-
-void imx_dma_free(int channel);
-
-int imx_dma_request_by_prio(const char *name, enum imx_dma_prio prio);
-
-#endif	/* __MACH_DMA_V1_H__ */
diff --git a/arch/arm/plat-nomadik/include/plat/ste_dma40.h b/arch/arm/plat-nomadik/include/plat/ste_dma40.h
index fd0ee84..9ff93b0 100644
--- a/arch/arm/plat-nomadik/include/plat/ste_dma40.h
+++ b/arch/arm/plat-nomadik/include/plat/ste_dma40.h
@@ -200,8 +200,7 @@
 	sg.dma_address = addr;
 	sg.length = size;
 
-	return chan->device->device_prep_slave_sg(chan, &sg, 1,
-						  direction, flags);
+	return dmaengine_prep_slave_sg(chan, &sg, 1, direction, flags);
 }
 
 #else
diff --git a/arch/arm/plat-samsung/dma-ops.c b/arch/arm/plat-samsung/dma-ops.c
index 301d9c3..eb9f4f5 100644
--- a/arch/arm/plat-samsung/dma-ops.c
+++ b/arch/arm/plat-samsung/dma-ops.c
@@ -79,11 +79,11 @@
 			    info->len, offset_in_page(info->buf));
 		sg_dma_address(&sg) = info->buf;
 
-		desc = chan->device->device_prep_slave_sg(chan,
+		desc = dmaengine_prep_slave_sg(chan,
 			&sg, 1, info->direction, DMA_PREP_INTERRUPT);
 		break;
 	case DMA_CYCLIC:
-		desc = chan->device->device_prep_dma_cyclic(chan,
+		desc = dmaengine_prep_dma_cyclic(chan,
 			info->buf, info->len, info->period, info->direction);
 		break;
 	default:
diff --git a/arch/avr32/mach-at32ap/at32ap700x.c b/arch/avr32/mach-at32ap/at32ap700x.c
index 889c544..0445c4f 100644
--- a/arch/avr32/mach-at32ap/at32ap700x.c
+++ b/arch/avr32/mach-at32ap/at32ap700x.c
@@ -1351,7 +1351,6 @@
 		goto fail;
 
 	slave->sdata.dma_dev = &dw_dmac0_device.dev;
-	slave->sdata.reg_width = DW_DMA_SLAVE_WIDTH_32BIT;
 	slave->sdata.cfg_hi = (DWC_CFGH_SRC_PER(0)
 				| DWC_CFGH_DST_PER(1));
 	slave->sdata.cfg_lo &= ~(DWC_CFGL_HS_DST_POL
@@ -2046,27 +2045,19 @@
 	/* Check if DMA slave interface for capture should be configured. */
 	if (flags & AC97C_CAPTURE) {
 		rx_dws->dma_dev = &dw_dmac0_device.dev;
-		rx_dws->reg_width = DW_DMA_SLAVE_WIDTH_16BIT;
 		rx_dws->cfg_hi = DWC_CFGH_SRC_PER(3);
 		rx_dws->cfg_lo &= ~(DWC_CFGL_HS_DST_POL | DWC_CFGL_HS_SRC_POL);
 		rx_dws->src_master = 0;
 		rx_dws->dst_master = 1;
-		rx_dws->src_msize = DW_DMA_MSIZE_1;
-		rx_dws->dst_msize = DW_DMA_MSIZE_1;
-		rx_dws->fc = DW_DMA_FC_D_P2M;
 	}
 
 	/* Check if DMA slave interface for playback should be configured. */
 	if (flags & AC97C_PLAYBACK) {
 		tx_dws->dma_dev = &dw_dmac0_device.dev;
-		tx_dws->reg_width = DW_DMA_SLAVE_WIDTH_16BIT;
 		tx_dws->cfg_hi = DWC_CFGH_DST_PER(4);
 		tx_dws->cfg_lo &= ~(DWC_CFGL_HS_DST_POL | DWC_CFGL_HS_SRC_POL);
 		tx_dws->src_master = 0;
 		tx_dws->dst_master = 1;
-		tx_dws->src_msize = DW_DMA_MSIZE_1;
-		tx_dws->dst_msize = DW_DMA_MSIZE_1;
-		tx_dws->fc = DW_DMA_FC_D_M2P;
 	}
 
 	if (platform_device_add_data(pdev, data,
@@ -2136,14 +2127,10 @@
 	dws = &data->dws;
 
 	dws->dma_dev = &dw_dmac0_device.dev;
-	dws->reg_width = DW_DMA_SLAVE_WIDTH_32BIT;
 	dws->cfg_hi = DWC_CFGH_DST_PER(2);
 	dws->cfg_lo &= ~(DWC_CFGL_HS_DST_POL | DWC_CFGL_HS_SRC_POL);
 	dws->src_master = 0;
 	dws->dst_master = 1;
-	dws->src_msize = DW_DMA_MSIZE_1;
-	dws->dst_msize = DW_DMA_MSIZE_1;
-	dws->fc = DW_DMA_FC_D_M2P;
 
 	if (platform_device_add_data(pdev, data,
 				sizeof(struct atmel_abdac_pdata)))
diff --git a/arch/avr32/mach-at32ap/include/mach/atmel-mci.h b/arch/avr32/mach-at32ap/include/mach/atmel-mci.h
index a9b3896..4bba585 100644
--- a/arch/avr32/mach-at32ap/include/mach/atmel-mci.h
+++ b/arch/avr32/mach-at32ap/include/mach/atmel-mci.h
@@ -14,11 +14,4 @@
 #define	slave_data_ptr(s)	(&(s)->sdata)
 #define find_slave_dev(s)	((s)->sdata.dma_dev)
 
-#define	setup_dma_addr(s, t, r)	do {		\
-	if (s) {				\
-		(s)->sdata.tx_reg = (t);	\
-		(s)->sdata.rx_reg = (r);	\
-	}					\
-} while (0)
-
 #endif /* __MACH_ATMEL_MCI_H */
diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig
index 4a6c46d..cf9da36 100644
--- a/drivers/dma/Kconfig
+++ b/drivers/dma/Kconfig
@@ -201,7 +201,6 @@
 	tristate "DMA API Driver for PL330"
 	select DMA_ENGINE
 	depends on ARM_AMBA
-	select PL330
 	help
 	  Select if your platform has one or more PL330 DMACs.
 	  You need to provide platform specific settings via
@@ -231,7 +230,7 @@
 
 config IMX_DMA
 	tristate "i.MX DMA support"
-	depends on IMX_HAVE_DMA_V1
+	depends on ARCH_MXC
 	select DMA_ENGINE
 	help
 	  Support the i.MX DMA engine. This engine is integrated into
diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c
index 8a28158..c301a8e 100644
--- a/drivers/dma/amba-pl08x.c
+++ b/drivers/dma/amba-pl08x.c
@@ -85,6 +85,8 @@
 #include <linux/slab.h>
 #include <asm/hardware/pl080.h>
 
+#include "dmaengine.h"
+
 #define DRIVER_NAME	"pl08xdmac"
 
 static struct amba_driver pl08x_amba_driver;
@@ -649,7 +651,7 @@
 			}
 
 			if ((bd.srcbus.addr % bd.srcbus.buswidth) ||
-					(bd.srcbus.addr % bd.srcbus.buswidth)) {
+					(bd.dstbus.addr % bd.dstbus.buswidth)) {
 				dev_err(&pl08x->adev->dev,
 					"%s src & dst address must be aligned to src"
 					" & dst width if peripheral is flow controller",
@@ -919,13 +921,10 @@
 	struct pl08x_dma_chan *plchan = to_pl08x_chan(tx->chan);
 	struct pl08x_txd *txd = to_pl08x_txd(tx);
 	unsigned long flags;
+	dma_cookie_t cookie;
 
 	spin_lock_irqsave(&plchan->lock, flags);
-
-	plchan->chan.cookie += 1;
-	if (plchan->chan.cookie < 0)
-		plchan->chan.cookie = 1;
-	tx->cookie = plchan->chan.cookie;
+	cookie = dma_cookie_assign(tx);
 
 	/* Put this onto the pending list */
 	list_add_tail(&txd->node, &plchan->pend_list);
@@ -945,7 +944,7 @@
 
 	spin_unlock_irqrestore(&plchan->lock, flags);
 
-	return tx->cookie;
+	return cookie;
 }
 
 static struct dma_async_tx_descriptor *pl08x_prep_dma_interrupt(
@@ -965,31 +964,17 @@
 		dma_cookie_t cookie, struct dma_tx_state *txstate)
 {
 	struct pl08x_dma_chan *plchan = to_pl08x_chan(chan);
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
 	enum dma_status ret;
-	u32 bytesleft = 0;
 
-	last_used = plchan->chan.cookie;
-	last_complete = plchan->lc;
-
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
-	if (ret == DMA_SUCCESS) {
-		dma_set_tx_state(txstate, last_complete, last_used, 0);
+	ret = dma_cookie_status(chan, cookie, txstate);
+	if (ret == DMA_SUCCESS)
 		return ret;
-	}
 
 	/*
 	 * This cookie not complete yet
+	 * Get number of bytes left in the active transactions and queue
 	 */
-	last_used = plchan->chan.cookie;
-	last_complete = plchan->lc;
-
-	/* Get number of bytes left in the active transactions and queue */
-	bytesleft = pl08x_getbytes_chan(plchan);
-
-	dma_set_tx_state(txstate, last_complete, last_used,
-			 bytesleft);
+	dma_set_residue(txstate, pl08x_getbytes_chan(plchan));
 
 	if (plchan->state == PL08X_CHAN_PAUSED)
 		return DMA_PAUSED;
@@ -1139,6 +1124,8 @@
 	cctl |= burst << PL080_CONTROL_SB_SIZE_SHIFT;
 	cctl |= burst << PL080_CONTROL_DB_SIZE_SHIFT;
 
+	plchan->device_fc = config->device_fc;
+
 	if (plchan->runtime_direction == DMA_DEV_TO_MEM) {
 		plchan->src_addr = config->src_addr;
 		plchan->src_cctl = pl08x_cctl(cctl) | PL080_CONTROL_DST_INCR |
@@ -1326,7 +1313,7 @@
 static struct dma_async_tx_descriptor *pl08x_prep_slave_sg(
 		struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned int sg_len, enum dma_transfer_direction direction,
-		unsigned long flags)
+		unsigned long flags, void *context)
 {
 	struct pl08x_dma_chan *plchan = to_pl08x_chan(chan);
 	struct pl08x_driver_data *pl08x = plchan->host;
@@ -1370,7 +1357,7 @@
 		return NULL;
 	}
 
-	if (plchan->cd->device_fc)
+	if (plchan->device_fc)
 		tmp = (direction == DMA_MEM_TO_DEV) ? PL080_FLOW_MEM2PER_PER :
 			PL080_FLOW_PER2MEM_PER;
 	else
@@ -1541,7 +1528,7 @@
 
 	if (txd) {
 		/* Update last completed */
-		plchan->lc = txd->tx.cookie;
+		dma_cookie_complete(&txd->tx);
 	}
 
 	/* If a new descriptor is queued, set it up plchan->at is NULL here */
@@ -1722,8 +1709,7 @@
 			 chan->name);
 
 		chan->chan.device = dmadev;
-		chan->chan.cookie = 0;
-		chan->lc = 0;
+		dma_cookie_init(&chan->chan);
 
 		spin_lock_init(&chan->lock);
 		INIT_LIST_HEAD(&chan->pend_list);
diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
index f4aed5f..7aa58d2 100644
--- a/drivers/dma/at_hdmac.c
+++ b/drivers/dma/at_hdmac.c
@@ -27,6 +27,7 @@
 #include <linux/of_device.h>
 
 #include "at_hdmac_regs.h"
+#include "dmaengine.h"
 
 /*
  * Glossary
@@ -192,27 +193,6 @@
 }
 
 /**
- * atc_assign_cookie - compute and assign new cookie
- * @atchan: channel we work on
- * @desc: descriptor to assign cookie for
- *
- * Called with atchan->lock held and bh disabled
- */
-static dma_cookie_t
-atc_assign_cookie(struct at_dma_chan *atchan, struct at_desc *desc)
-{
-	dma_cookie_t cookie = atchan->chan_common.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	atchan->chan_common.cookie = cookie;
-	desc->txd.cookie = cookie;
-
-	return cookie;
-}
-
-/**
  * atc_dostart - starts the DMA engine for real
  * @atchan: the channel we want to start
  * @first: first descriptor in the list we want to begin with
@@ -269,7 +249,7 @@
 	dev_vdbg(chan2dev(&atchan->chan_common),
 		"descriptor %u complete\n", txd->cookie);
 
-	atchan->completed_cookie = txd->cookie;
+	dma_cookie_complete(txd);
 
 	/* move children to free_list */
 	list_splice_init(&desc->tx_list, &atchan->free_list);
@@ -547,7 +527,7 @@
 	unsigned long		flags;
 
 	spin_lock_irqsave(&atchan->lock, flags);
-	cookie = atc_assign_cookie(atchan, desc);
+	cookie = dma_cookie_assign(tx);
 
 	if (list_empty(&atchan->active_list)) {
 		dev_vdbg(chan2dev(tx->chan), "tx_submit: started %u\n",
@@ -659,14 +639,16 @@
  * @sg_len: number of entries in @scatterlist
  * @direction: DMA direction
  * @flags: tx descriptor status flags
+ * @context: transaction context (ignored)
  */
 static struct dma_async_tx_descriptor *
 atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned int sg_len, enum dma_transfer_direction direction,
-		unsigned long flags)
+		unsigned long flags, void *context)
 {
 	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
 	struct at_dma_slave	*atslave = chan->private;
+	struct dma_slave_config	*sconfig = &atchan->dma_sconfig;
 	struct at_desc		*first = NULL;
 	struct at_desc		*prev = NULL;
 	u32			ctrla;
@@ -688,19 +670,18 @@
 		return NULL;
 	}
 
-	reg_width = atslave->reg_width;
-
 	ctrla = ATC_DEFAULT_CTRLA | atslave->ctrla;
 	ctrlb = ATC_IEN;
 
 	switch (direction) {
 	case DMA_MEM_TO_DEV:
+		reg_width = convert_buswidth(sconfig->dst_addr_width);
 		ctrla |=  ATC_DST_WIDTH(reg_width);
 		ctrlb |=  ATC_DST_ADDR_MODE_FIXED
 			| ATC_SRC_ADDR_MODE_INCR
 			| ATC_FC_MEM2PER
 			| ATC_SIF(AT_DMA_MEM_IF) | ATC_DIF(AT_DMA_PER_IF);
-		reg = atslave->tx_reg;
+		reg = sconfig->dst_addr;
 		for_each_sg(sgl, sg, sg_len, i) {
 			struct at_desc	*desc;
 			u32		len;
@@ -728,13 +709,14 @@
 		}
 		break;
 	case DMA_DEV_TO_MEM:
+		reg_width = convert_buswidth(sconfig->src_addr_width);
 		ctrla |=  ATC_SRC_WIDTH(reg_width);
 		ctrlb |=  ATC_DST_ADDR_MODE_INCR
 			| ATC_SRC_ADDR_MODE_FIXED
 			| ATC_FC_PER2MEM
 			| ATC_SIF(AT_DMA_PER_IF) | ATC_DIF(AT_DMA_MEM_IF);
 
-		reg = atslave->rx_reg;
+		reg = sconfig->src_addr;
 		for_each_sg(sgl, sg, sg_len, i) {
 			struct at_desc	*desc;
 			u32		len;
@@ -810,12 +792,15 @@
  * atc_dma_cyclic_fill_desc - Fill one period decriptor
  */
 static int
-atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc,
+atc_dma_cyclic_fill_desc(struct dma_chan *chan, struct at_desc *desc,
 		unsigned int period_index, dma_addr_t buf_addr,
-		size_t period_len, enum dma_transfer_direction direction)
+		unsigned int reg_width, size_t period_len,
+		enum dma_transfer_direction direction)
 {
-	u32		ctrla;
-	unsigned int	reg_width = atslave->reg_width;
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+	struct at_dma_slave	*atslave = chan->private;
+	struct dma_slave_config	*sconfig = &atchan->dma_sconfig;
+	u32			ctrla;
 
 	/* prepare common CRTLA value */
 	ctrla =   ATC_DEFAULT_CTRLA | atslave->ctrla
@@ -826,7 +811,7 @@
 	switch (direction) {
 	case DMA_MEM_TO_DEV:
 		desc->lli.saddr = buf_addr + (period_len * period_index);
-		desc->lli.daddr = atslave->tx_reg;
+		desc->lli.daddr = sconfig->dst_addr;
 		desc->lli.ctrla = ctrla;
 		desc->lli.ctrlb = ATC_DST_ADDR_MODE_FIXED
 				| ATC_SRC_ADDR_MODE_INCR
@@ -836,7 +821,7 @@
 		break;
 
 	case DMA_DEV_TO_MEM:
-		desc->lli.saddr = atslave->rx_reg;
+		desc->lli.saddr = sconfig->src_addr;
 		desc->lli.daddr = buf_addr + (period_len * period_index);
 		desc->lli.ctrla = ctrla;
 		desc->lli.ctrlb = ATC_DST_ADDR_MODE_INCR
@@ -860,16 +845,20 @@
  * @buf_len: total number of bytes for the entire buffer
  * @period_len: number of bytes for each period
  * @direction: transfer direction, to or from device
+ * @context: transfer context (ignored)
  */
 static struct dma_async_tx_descriptor *
 atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
-		size_t period_len, enum dma_transfer_direction direction)
+		size_t period_len, enum dma_transfer_direction direction,
+		void *context)
 {
 	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
 	struct at_dma_slave	*atslave = chan->private;
+	struct dma_slave_config	*sconfig = &atchan->dma_sconfig;
 	struct at_desc		*first = NULL;
 	struct at_desc		*prev = NULL;
 	unsigned long		was_cyclic;
+	unsigned int		reg_width;
 	unsigned int		periods = buf_len / period_len;
 	unsigned int		i;
 
@@ -889,8 +878,13 @@
 		return NULL;
 	}
 
+	if (sconfig->direction == DMA_MEM_TO_DEV)
+		reg_width = convert_buswidth(sconfig->dst_addr_width);
+	else
+		reg_width = convert_buswidth(sconfig->src_addr_width);
+
 	/* Check for too big/unaligned periods and unaligned DMA buffer */
-	if (atc_dma_cyclic_check_values(atslave->reg_width, buf_addr,
+	if (atc_dma_cyclic_check_values(reg_width, buf_addr,
 					period_len, direction))
 		goto err_out;
 
@@ -902,8 +896,8 @@
 		if (!desc)
 			goto err_desc_get;
 
-		if (atc_dma_cyclic_fill_desc(atslave, desc, i, buf_addr,
-						period_len, direction))
+		if (atc_dma_cyclic_fill_desc(chan, desc, i, buf_addr,
+					     reg_width, period_len, direction))
 			goto err_desc_get;
 
 		atc_desc_chain(&first, &prev, desc);
@@ -926,6 +920,23 @@
 	return NULL;
 }
 
+static int set_runtime_config(struct dma_chan *chan,
+			      struct dma_slave_config *sconfig)
+{
+	struct at_dma_chan	*atchan = to_at_dma_chan(chan);
+
+	/* Check if it is chan is configured for slave transfers */
+	if (!chan->private)
+		return -EINVAL;
+
+	memcpy(&atchan->dma_sconfig, sconfig, sizeof(*sconfig));
+
+	convert_burst(&atchan->dma_sconfig.src_maxburst);
+	convert_burst(&atchan->dma_sconfig.dst_maxburst);
+
+	return 0;
+}
+
 
 static int atc_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 		       unsigned long arg)
@@ -986,6 +997,8 @@
 		clear_bit(ATC_IS_CYCLIC, &atchan->status);
 
 		spin_unlock_irqrestore(&atchan->lock, flags);
+	} else if (cmd == DMA_SLAVE_CONFIG) {
+		return set_runtime_config(chan, (struct dma_slave_config *)arg);
 	} else {
 		return -ENXIO;
 	}
@@ -1016,26 +1029,20 @@
 
 	spin_lock_irqsave(&atchan->lock, flags);
 
-	last_complete = atchan->completed_cookie;
-	last_used = chan->cookie;
-
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
+	ret = dma_cookie_status(chan, cookie, txstate);
 	if (ret != DMA_SUCCESS) {
 		atc_cleanup_descriptors(atchan);
 
-		last_complete = atchan->completed_cookie;
-		last_used = chan->cookie;
-
-		ret = dma_async_is_complete(cookie, last_complete, last_used);
+		ret = dma_cookie_status(chan, cookie, txstate);
 	}
 
+	last_complete = chan->completed_cookie;
+	last_used = chan->cookie;
+
 	spin_unlock_irqrestore(&atchan->lock, flags);
 
 	if (ret != DMA_SUCCESS)
-		dma_set_tx_state(txstate, last_complete, last_used,
-			atc_first_active(atchan)->len);
-	else
-		dma_set_tx_state(txstate, last_complete, last_used, 0);
+		dma_set_residue(txstate, atc_first_active(atchan)->len);
 
 	if (atc_chan_is_paused(atchan))
 		ret = DMA_PAUSED;
@@ -1129,7 +1136,7 @@
 	spin_lock_irqsave(&atchan->lock, flags);
 	atchan->descs_allocated = i;
 	list_splice(&tmp_list, &atchan->free_list);
-	atchan->completed_cookie = chan->cookie = 1;
+	dma_cookie_init(chan);
 	spin_unlock_irqrestore(&atchan->lock, flags);
 
 	/* channel parameters */
@@ -1329,7 +1336,7 @@
 		struct at_dma_chan	*atchan = &atdma->chan[i];
 
 		atchan->chan_common.device = &atdma->dma_common;
-		atchan->chan_common.cookie = atchan->completed_cookie = 1;
+		dma_cookie_init(&atchan->chan_common);
 		list_add_tail(&atchan->chan_common.device_node,
 				&atdma->dma_common.channels);
 
diff --git a/drivers/dma/at_hdmac_regs.h b/drivers/dma/at_hdmac_regs.h
index a8d3277..897a8bc 100644
--- a/drivers/dma/at_hdmac_regs.h
+++ b/drivers/dma/at_hdmac_regs.h
@@ -207,8 +207,8 @@
  * @save_cfg: configuration register that is saved on suspend/resume cycle
  * @save_dscr: for cyclic operations, preserve next descriptor address in
  *             the cyclic list on suspend/resume cycle
+ * @dma_sconfig: configuration for slave transfers, passed via DMA_SLAVE_CONFIG
  * @lock: serializes enqueue/dequeue operations to descriptors lists
- * @completed_cookie: identifier for the most recently completed operation
  * @active_list: list of descriptors dmaengine is being running on
  * @queue: list of descriptors ready to be submitted to engine
  * @free_list: list of descriptors usable by the channel
@@ -223,11 +223,11 @@
 	struct tasklet_struct	tasklet;
 	u32			save_cfg;
 	u32			save_dscr;
+	struct dma_slave_config dma_sconfig;
 
 	spinlock_t		lock;
 
 	/* these other elements are all protected by lock */
-	dma_cookie_t		completed_cookie;
 	struct list_head	active_list;
 	struct list_head	queue;
 	struct list_head	free_list;
@@ -245,6 +245,36 @@
 	return container_of(dchan, struct at_dma_chan, chan_common);
 }
 
+/*
+ * Fix sconfig's burst size according to at_hdmac. We need to convert them as:
+ * 1 -> 0, 4 -> 1, 8 -> 2, 16 -> 3, 32 -> 4, 64 -> 5, 128 -> 6, 256 -> 7.
+ *
+ * This can be done by finding most significant bit set.
+ */
+static inline void convert_burst(u32 *maxburst)
+{
+	if (*maxburst > 1)
+		*maxburst = fls(*maxburst) - 2;
+	else
+		*maxburst = 0;
+}
+
+/*
+ * Fix sconfig's bus width according to at_hdmac.
+ * 1 byte -> 0, 2 bytes -> 1, 4 bytes -> 2.
+ */
+static inline u8 convert_buswidth(enum dma_slave_buswidth addr_width)
+{
+	switch (addr_width) {
+	case DMA_SLAVE_BUSWIDTH_2_BYTES:
+		return 1;
+	case DMA_SLAVE_BUSWIDTH_4_BYTES:
+		return 2;
+	default:
+		/* For 1 byte width or fallback */
+		return 0;
+	}
+}
 
 /*--  Controller  ------------------------------------------------------*/
 
diff --git a/drivers/dma/coh901318.c b/drivers/dma/coh901318.c
index d65a718..dc89455 100644
--- a/drivers/dma/coh901318.c
+++ b/drivers/dma/coh901318.c
@@ -24,6 +24,7 @@
 #include <mach/coh901318.h>
 
 #include "coh901318_lli.h"
+#include "dmaengine.h"
 
 #define COHC_2_DEV(cohc) (&cohc->chan.dev->device)
 
@@ -59,7 +60,6 @@
 struct coh901318_chan {
 	spinlock_t lock;
 	int allocated;
-	int completed;
 	int id;
 	int stopped;
 
@@ -318,20 +318,6 @@
 
 	return 0;
 }
-static dma_cookie_t
-coh901318_assign_cookie(struct coh901318_chan *cohc,
-			struct coh901318_desc *cohd)
-{
-	dma_cookie_t cookie = cohc->chan.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	cohc->chan.cookie = cookie;
-	cohd->desc.cookie = cookie;
-
-	return cookie;
-}
 
 static struct coh901318_desc *
 coh901318_desc_get(struct coh901318_chan *cohc)
@@ -705,7 +691,7 @@
 	callback_param = cohd_fin->desc.callback_param;
 
 	/* sign this job as completed on the channel */
-	cohc->completed = cohd_fin->desc.cookie;
+	dma_cookie_complete(&cohd_fin->desc);
 
 	/* release the lli allocation and remove the descriptor */
 	coh901318_lli_free(&cohc->base->pool, &cohd_fin->lli);
@@ -929,7 +915,7 @@
 	coh901318_config(cohc, NULL);
 
 	cohc->allocated = 1;
-	cohc->completed = chan->cookie = 1;
+	dma_cookie_init(chan);
 
 	spin_unlock_irqrestore(&cohc->lock, flags);
 
@@ -966,16 +952,16 @@
 						   desc);
 	struct coh901318_chan *cohc = to_coh901318_chan(tx->chan);
 	unsigned long flags;
+	dma_cookie_t cookie;
 
 	spin_lock_irqsave(&cohc->lock, flags);
-
-	tx->cookie = coh901318_assign_cookie(cohc, cohd);
+	cookie = dma_cookie_assign(tx);
 
 	coh901318_desc_queue(cohc, cohd);
 
 	spin_unlock_irqrestore(&cohc->lock, flags);
 
-	return tx->cookie;
+	return cookie;
 }
 
 static struct dma_async_tx_descriptor *
@@ -1035,7 +1021,7 @@
 static struct dma_async_tx_descriptor *
 coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 			unsigned int sg_len, enum dma_transfer_direction direction,
-			unsigned long flags)
+			unsigned long flags, void *context)
 {
 	struct coh901318_chan *cohc = to_coh901318_chan(chan);
 	struct coh901318_lli *lli;
@@ -1165,17 +1151,12 @@
 		 struct dma_tx_state *txstate)
 {
 	struct coh901318_chan *cohc = to_coh901318_chan(chan);
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
-	int ret;
+	enum dma_status ret;
 
-	last_complete = cohc->completed;
-	last_used = chan->cookie;
+	ret = dma_cookie_status(chan, cookie, txstate);
+	/* FIXME: should be conditional on ret != DMA_SUCCESS? */
+	dma_set_residue(txstate, coh901318_get_bytes_left(chan));
 
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
-
-	dma_set_tx_state(txstate, last_complete, last_used,
-			 coh901318_get_bytes_left(chan));
 	if (ret == DMA_IN_PROGRESS && cohc->stopped)
 		ret = DMA_PAUSED;
 
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index a6c6051..767bcc3 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -510,8 +510,8 @@
 					 dma_chan_name(chan));
 				list_del_rcu(&device->global_node);
 			} else if (err)
-				pr_debug("dmaengine: failed to get %s: (%d)\n",
-					 dma_chan_name(chan), err);
+				pr_debug("%s: failed to get %s: (%d)\n",
+					__func__, dma_chan_name(chan), err);
 			else
 				break;
 			if (--device->privatecnt == 0)
@@ -564,8 +564,8 @@
 				list_del_rcu(&device->global_node);
 				break;
 			} else if (err)
-				pr_err("dmaengine: failed to get %s: (%d)\n",
-				       dma_chan_name(chan), err);
+				pr_err("%s: failed to get %s: (%d)\n",
+					__func__, dma_chan_name(chan), err);
 		}
 	}
 
diff --git a/drivers/dma/dmaengine.h b/drivers/dma/dmaengine.h
new file mode 100644
index 0000000..17f983a
--- /dev/null
+++ b/drivers/dma/dmaengine.h
@@ -0,0 +1,89 @@
+/*
+ * The contents of this file are private to DMA engine drivers, and is not
+ * part of the API to be used by DMA engine users.
+ */
+#ifndef DMAENGINE_H
+#define DMAENGINE_H
+
+#include <linux/bug.h>
+#include <linux/dmaengine.h>
+
+/**
+ * dma_cookie_init - initialize the cookies for a DMA channel
+ * @chan: dma channel to initialize
+ */
+static inline void dma_cookie_init(struct dma_chan *chan)
+{
+	chan->cookie = DMA_MIN_COOKIE;
+	chan->completed_cookie = DMA_MIN_COOKIE;
+}
+
+/**
+ * dma_cookie_assign - assign a DMA engine cookie to the descriptor
+ * @tx: descriptor needing cookie
+ *
+ * Assign a unique non-zero per-channel cookie to the descriptor.
+ * Note: caller is expected to hold a lock to prevent concurrency.
+ */
+static inline dma_cookie_t dma_cookie_assign(struct dma_async_tx_descriptor *tx)
+{
+	struct dma_chan *chan = tx->chan;
+	dma_cookie_t cookie;
+
+	cookie = chan->cookie + 1;
+	if (cookie < DMA_MIN_COOKIE)
+		cookie = DMA_MIN_COOKIE;
+	tx->cookie = chan->cookie = cookie;
+
+	return cookie;
+}
+
+/**
+ * dma_cookie_complete - complete a descriptor
+ * @tx: descriptor to complete
+ *
+ * Mark this descriptor complete by updating the channels completed
+ * cookie marker.  Zero the descriptors cookie to prevent accidental
+ * repeated completions.
+ *
+ * Note: caller is expected to hold a lock to prevent concurrency.
+ */
+static inline void dma_cookie_complete(struct dma_async_tx_descriptor *tx)
+{
+	BUG_ON(tx->cookie < DMA_MIN_COOKIE);
+	tx->chan->completed_cookie = tx->cookie;
+	tx->cookie = 0;
+}
+
+/**
+ * dma_cookie_status - report cookie status
+ * @chan: dma channel
+ * @cookie: cookie we are interested in
+ * @state: dma_tx_state structure to return last/used cookies
+ *
+ * Report the status of the cookie, filling in the state structure if
+ * non-NULL.  No locking is required.
+ */
+static inline enum dma_status dma_cookie_status(struct dma_chan *chan,
+	dma_cookie_t cookie, struct dma_tx_state *state)
+{
+	dma_cookie_t used, complete;
+
+	used = chan->cookie;
+	complete = chan->completed_cookie;
+	barrier();
+	if (state) {
+		state->last = complete;
+		state->used = used;
+		state->residue = 0;
+	}
+	return dma_async_is_complete(cookie, complete, used);
+}
+
+static inline void dma_set_residue(struct dma_tx_state *state, u32 residue)
+{
+	if (state)
+		state->residue = residue;
+}
+
+#endif
diff --git a/drivers/dma/dw_dmac.c b/drivers/dma/dw_dmac.c
index 9b592b0..7439079 100644
--- a/drivers/dma/dw_dmac.c
+++ b/drivers/dma/dw_dmac.c
@@ -9,6 +9,7 @@
  * it under the terms of the GNU General Public License version 2 as
  * published by the Free Software Foundation.
  */
+#include <linux/bitops.h>
 #include <linux/clk.h>
 #include <linux/delay.h>
 #include <linux/dmaengine.h>
@@ -22,6 +23,7 @@
 #include <linux/slab.h>
 
 #include "dw_dmac_regs.h"
+#include "dmaengine.h"
 
 /*
  * This supports the Synopsys "DesignWare AHB Central DMA Controller",
@@ -33,19 +35,23 @@
  * which does not support descriptor writeback.
  */
 
-#define DWC_DEFAULT_CTLLO(private) ({				\
-		struct dw_dma_slave *__slave = (private);	\
-		int dms = __slave ? __slave->dst_master : 0;	\
-		int sms = __slave ? __slave->src_master : 1;	\
-		u8 smsize = __slave ? __slave->src_msize : DW_DMA_MSIZE_16; \
-		u8 dmsize = __slave ? __slave->dst_msize : DW_DMA_MSIZE_16; \
+#define DWC_DEFAULT_CTLLO(_chan) ({				\
+		struct dw_dma_slave *__slave = (_chan->private);	\
+		struct dw_dma_chan *_dwc = to_dw_dma_chan(_chan);	\
+		struct dma_slave_config	*_sconfig = &_dwc->dma_sconfig;	\
+		int _dms = __slave ? __slave->dst_master : 0;	\
+		int _sms = __slave ? __slave->src_master : 1;	\
+		u8 _smsize = __slave ? _sconfig->src_maxburst :	\
+			DW_DMA_MSIZE_16;			\
+		u8 _dmsize = __slave ? _sconfig->dst_maxburst :	\
+			DW_DMA_MSIZE_16;			\
 								\
-		(DWC_CTLL_DST_MSIZE(dmsize)			\
-		 | DWC_CTLL_SRC_MSIZE(smsize)			\
+		(DWC_CTLL_DST_MSIZE(_dmsize)			\
+		 | DWC_CTLL_SRC_MSIZE(_smsize)			\
 		 | DWC_CTLL_LLP_D_EN				\
 		 | DWC_CTLL_LLP_S_EN				\
-		 | DWC_CTLL_DMS(dms)				\
-		 | DWC_CTLL_SMS(sms));				\
+		 | DWC_CTLL_DMS(_dms)				\
+		 | DWC_CTLL_SMS(_sms));				\
 	})
 
 /*
@@ -151,21 +157,6 @@
 	}
 }
 
-/* Called with dwc->lock held and bh disabled */
-static dma_cookie_t
-dwc_assign_cookie(struct dw_dma_chan *dwc, struct dw_desc *desc)
-{
-	dma_cookie_t cookie = dwc->chan.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	dwc->chan.cookie = cookie;
-	desc->txd.cookie = cookie;
-
-	return cookie;
-}
-
 static void dwc_initialize(struct dw_dma_chan *dwc)
 {
 	struct dw_dma *dw = to_dw_dma(dwc->chan.device);
@@ -192,7 +183,6 @@
 
 	/* Enable interrupts */
 	channel_set_bit(dw, MASK.XFER, dwc->mask);
-	channel_set_bit(dw, MASK.BLOCK, dwc->mask);
 	channel_set_bit(dw, MASK.ERROR, dwc->mask);
 
 	dwc->initialized = true;
@@ -245,7 +235,7 @@
 	dev_vdbg(chan2dev(&dwc->chan), "descriptor %u complete\n", txd->cookie);
 
 	spin_lock_irqsave(&dwc->lock, flags);
-	dwc->completed = txd->cookie;
+	dma_cookie_complete(txd);
 	if (callback_required) {
 		callback = txd->callback;
 		param = txd->callback_param;
@@ -329,12 +319,6 @@
 	unsigned long flags;
 
 	spin_lock_irqsave(&dwc->lock, flags);
-	/*
-	 * Clear block interrupt flag before scanning so that we don't
-	 * miss any, and read LLP before RAW_XFER to ensure it is
-	 * valid if we decide to scan the list.
-	 */
-	dma_writel(dw, CLEAR.BLOCK, dwc->mask);
 	llp = channel_readl(dwc, LLP);
 	status_xfer = dma_readl(dw, RAW.XFER);
 
@@ -470,17 +454,16 @@
 
 /* called with dwc->lock held and all DMAC interrupts disabled */
 static void dwc_handle_cyclic(struct dw_dma *dw, struct dw_dma_chan *dwc,
-		u32 status_block, u32 status_err, u32 status_xfer)
+		u32 status_err, u32 status_xfer)
 {
 	unsigned long flags;
 
-	if (status_block & dwc->mask) {
+	if (dwc->mask) {
 		void (*callback)(void *param);
 		void *callback_param;
 
 		dev_vdbg(chan2dev(&dwc->chan), "new cyclic period llp 0x%08x\n",
 				channel_readl(dwc, LLP));
-		dma_writel(dw, CLEAR.BLOCK, dwc->mask);
 
 		callback = dwc->cdesc->period_callback;
 		callback_param = dwc->cdesc->period_callback_param;
@@ -520,7 +503,6 @@
 		channel_writel(dwc, CTL_LO, 0);
 		channel_writel(dwc, CTL_HI, 0);
 
-		dma_writel(dw, CLEAR.BLOCK, dwc->mask);
 		dma_writel(dw, CLEAR.ERROR, dwc->mask);
 		dma_writel(dw, CLEAR.XFER, dwc->mask);
 
@@ -537,36 +519,29 @@
 {
 	struct dw_dma *dw = (struct dw_dma *)data;
 	struct dw_dma_chan *dwc;
-	u32 status_block;
 	u32 status_xfer;
 	u32 status_err;
 	int i;
 
-	status_block = dma_readl(dw, RAW.BLOCK);
 	status_xfer = dma_readl(dw, RAW.XFER);
 	status_err = dma_readl(dw, RAW.ERROR);
 
-	dev_vdbg(dw->dma.dev, "tasklet: status_block=%x status_err=%x\n",
-			status_block, status_err);
+	dev_vdbg(dw->dma.dev, "tasklet: status_err=%x\n", status_err);
 
 	for (i = 0; i < dw->dma.chancnt; i++) {
 		dwc = &dw->chan[i];
 		if (test_bit(DW_DMA_IS_CYCLIC, &dwc->flags))
-			dwc_handle_cyclic(dw, dwc, status_block, status_err,
-					status_xfer);
+			dwc_handle_cyclic(dw, dwc, status_err, status_xfer);
 		else if (status_err & (1 << i))
 			dwc_handle_error(dw, dwc);
-		else if ((status_block | status_xfer) & (1 << i))
+		else if (status_xfer & (1 << i))
 			dwc_scan_descriptors(dw, dwc);
 	}
 
 	/*
-	 * Re-enable interrupts. Block Complete interrupts are only
-	 * enabled if the INT_EN bit in the descriptor is set. This
-	 * will trigger a scan before the whole list is done.
+	 * Re-enable interrupts.
 	 */
 	channel_set_bit(dw, MASK.XFER, dw->all_chan_mask);
-	channel_set_bit(dw, MASK.BLOCK, dw->all_chan_mask);
 	channel_set_bit(dw, MASK.ERROR, dw->all_chan_mask);
 }
 
@@ -583,7 +558,6 @@
 	 * softirq handler.
 	 */
 	channel_clear_bit(dw, MASK.XFER, dw->all_chan_mask);
-	channel_clear_bit(dw, MASK.BLOCK, dw->all_chan_mask);
 	channel_clear_bit(dw, MASK.ERROR, dw->all_chan_mask);
 
 	status = dma_readl(dw, STATUS_INT);
@@ -594,7 +568,6 @@
 
 		/* Try to recover */
 		channel_clear_bit(dw, MASK.XFER, (1 << 8) - 1);
-		channel_clear_bit(dw, MASK.BLOCK, (1 << 8) - 1);
 		channel_clear_bit(dw, MASK.SRC_TRAN, (1 << 8) - 1);
 		channel_clear_bit(dw, MASK.DST_TRAN, (1 << 8) - 1);
 		channel_clear_bit(dw, MASK.ERROR, (1 << 8) - 1);
@@ -615,7 +588,7 @@
 	unsigned long		flags;
 
 	spin_lock_irqsave(&dwc->lock, flags);
-	cookie = dwc_assign_cookie(dwc, desc);
+	cookie = dma_cookie_assign(tx);
 
 	/*
 	 * REVISIT: We should attempt to chain as many descriptors as
@@ -674,7 +647,7 @@
 	else
 		src_width = dst_width = 0;
 
-	ctllo = DWC_DEFAULT_CTLLO(chan->private)
+	ctllo = DWC_DEFAULT_CTLLO(chan)
 			| DWC_CTLL_DST_WIDTH(dst_width)
 			| DWC_CTLL_SRC_WIDTH(src_width)
 			| DWC_CTLL_DST_INC
@@ -731,10 +704,11 @@
 static struct dma_async_tx_descriptor *
 dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned int sg_len, enum dma_transfer_direction direction,
-		unsigned long flags)
+		unsigned long flags, void *context)
 {
 	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
 	struct dw_dma_slave	*dws = chan->private;
+	struct dma_slave_config	*sconfig = &dwc->dma_sconfig;
 	struct dw_desc		*prev;
 	struct dw_desc		*first;
 	u32			ctllo;
@@ -750,25 +724,34 @@
 	if (unlikely(!dws || !sg_len))
 		return NULL;
 
-	reg_width = dws->reg_width;
 	prev = first = NULL;
 
 	switch (direction) {
 	case DMA_MEM_TO_DEV:
-		ctllo = (DWC_DEFAULT_CTLLO(chan->private)
+		reg_width = __fls(sconfig->dst_addr_width);
+		reg = sconfig->dst_addr;
+		ctllo = (DWC_DEFAULT_CTLLO(chan)
 				| DWC_CTLL_DST_WIDTH(reg_width)
 				| DWC_CTLL_DST_FIX
-				| DWC_CTLL_SRC_INC
-				| DWC_CTLL_FC(dws->fc));
-		reg = dws->tx_reg;
+				| DWC_CTLL_SRC_INC);
+
+		ctllo |= sconfig->device_fc ? DWC_CTLL_FC(DW_DMA_FC_P_M2P) :
+			DWC_CTLL_FC(DW_DMA_FC_D_M2P);
+
 		for_each_sg(sgl, sg, sg_len, i) {
 			struct dw_desc	*desc;
 			u32		len, dlen, mem;
 
 			mem = sg_phys(sg);
 			len = sg_dma_len(sg);
-			mem_width = 2;
-			if (unlikely(mem & 3 || len & 3))
+
+			if (!((mem | len) & 7))
+				mem_width = 3;
+			else if (!((mem | len) & 3))
+				mem_width = 2;
+			else if (!((mem | len) & 1))
+				mem_width = 1;
+			else
 				mem_width = 0;
 
 slave_sg_todev_fill_desc:
@@ -812,21 +795,30 @@
 		}
 		break;
 	case DMA_DEV_TO_MEM:
-		ctllo = (DWC_DEFAULT_CTLLO(chan->private)
+		reg_width = __fls(sconfig->src_addr_width);
+		reg = sconfig->src_addr;
+		ctllo = (DWC_DEFAULT_CTLLO(chan)
 				| DWC_CTLL_SRC_WIDTH(reg_width)
 				| DWC_CTLL_DST_INC
-				| DWC_CTLL_SRC_FIX
-				| DWC_CTLL_FC(dws->fc));
+				| DWC_CTLL_SRC_FIX);
 
-		reg = dws->rx_reg;
+		ctllo |= sconfig->device_fc ? DWC_CTLL_FC(DW_DMA_FC_P_P2M) :
+			DWC_CTLL_FC(DW_DMA_FC_D_P2M);
+
 		for_each_sg(sgl, sg, sg_len, i) {
 			struct dw_desc	*desc;
 			u32		len, dlen, mem;
 
 			mem = sg_phys(sg);
 			len = sg_dma_len(sg);
-			mem_width = 2;
-			if (unlikely(mem & 3 || len & 3))
+
+			if (!((mem | len) & 7))
+				mem_width = 3;
+			else if (!((mem | len) & 3))
+				mem_width = 2;
+			else if (!((mem | len) & 1))
+				mem_width = 1;
+			else
 				mem_width = 0;
 
 slave_sg_fromdev_fill_desc:
@@ -890,6 +882,39 @@
 	return NULL;
 }
 
+/*
+ * Fix sconfig's burst size according to dw_dmac. We need to convert them as:
+ * 1 -> 0, 4 -> 1, 8 -> 2, 16 -> 3.
+ *
+ * NOTE: burst size 2 is not supported by controller.
+ *
+ * This can be done by finding least significant bit set: n & (n - 1)
+ */
+static inline void convert_burst(u32 *maxburst)
+{
+	if (*maxburst > 1)
+		*maxburst = fls(*maxburst) - 2;
+	else
+		*maxburst = 0;
+}
+
+static int
+set_runtime_config(struct dma_chan *chan, struct dma_slave_config *sconfig)
+{
+	struct dw_dma_chan *dwc = to_dw_dma_chan(chan);
+
+	/* Check if it is chan is configured for slave transfers */
+	if (!chan->private)
+		return -EINVAL;
+
+	memcpy(&dwc->dma_sconfig, sconfig, sizeof(*sconfig));
+
+	convert_burst(&dwc->dma_sconfig.src_maxburst);
+	convert_burst(&dwc->dma_sconfig.dst_maxburst);
+
+	return 0;
+}
+
 static int dwc_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 		       unsigned long arg)
 {
@@ -939,8 +964,11 @@
 		/* Flush all pending and queued descriptors */
 		list_for_each_entry_safe(desc, _desc, &list, desc_node)
 			dwc_descriptor_complete(dwc, desc, false);
-	} else
+	} else if (cmd == DMA_SLAVE_CONFIG) {
+		return set_runtime_config(chan, (struct dma_slave_config *)arg);
+	} else {
 		return -ENXIO;
+	}
 
 	return 0;
 }
@@ -951,28 +979,17 @@
 	      struct dma_tx_state *txstate)
 {
 	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
-	dma_cookie_t		last_used;
-	dma_cookie_t		last_complete;
-	int			ret;
+	enum dma_status		ret;
 
-	last_complete = dwc->completed;
-	last_used = chan->cookie;
-
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
+	ret = dma_cookie_status(chan, cookie, txstate);
 	if (ret != DMA_SUCCESS) {
 		dwc_scan_descriptors(to_dw_dma(chan->device), dwc);
 
-		last_complete = dwc->completed;
-		last_used = chan->cookie;
-
-		ret = dma_async_is_complete(cookie, last_complete, last_used);
+		ret = dma_cookie_status(chan, cookie, txstate);
 	}
 
 	if (ret != DMA_SUCCESS)
-		dma_set_tx_state(txstate, last_complete, last_used,
-				dwc_first_active(dwc)->len);
-	else
-		dma_set_tx_state(txstate, last_complete, last_used, 0);
+		dma_set_residue(txstate, dwc_first_active(dwc)->len);
 
 	if (dwc->paused)
 		return DMA_PAUSED;
@@ -1004,7 +1021,7 @@
 		return -EIO;
 	}
 
-	dwc->completed = chan->cookie = 1;
+	dma_cookie_init(chan);
 
 	/*
 	 * NOTE: some controllers may have additional features that we
@@ -1068,7 +1085,6 @@
 
 	/* Disable interrupts */
 	channel_clear_bit(dw, MASK.XFER, dwc->mask);
-	channel_clear_bit(dw, MASK.BLOCK, dwc->mask);
 	channel_clear_bit(dw, MASK.ERROR, dwc->mask);
 
 	spin_unlock_irqrestore(&dwc->lock, flags);
@@ -1120,7 +1136,6 @@
 		return -EBUSY;
 	}
 
-	dma_writel(dw, CLEAR.BLOCK, dwc->mask);
 	dma_writel(dw, CLEAR.ERROR, dwc->mask);
 	dma_writel(dw, CLEAR.XFER, dwc->mask);
 
@@ -1175,11 +1190,11 @@
 		enum dma_transfer_direction direction)
 {
 	struct dw_dma_chan		*dwc = to_dw_dma_chan(chan);
+	struct dma_slave_config		*sconfig = &dwc->dma_sconfig;
 	struct dw_cyclic_desc		*cdesc;
 	struct dw_cyclic_desc		*retval = NULL;
 	struct dw_desc			*desc;
 	struct dw_desc			*last = NULL;
-	struct dw_dma_slave		*dws = chan->private;
 	unsigned long			was_cyclic;
 	unsigned int			reg_width;
 	unsigned int			periods;
@@ -1203,7 +1218,12 @@
 	}
 
 	retval = ERR_PTR(-EINVAL);
-	reg_width = dws->reg_width;
+
+	if (direction == DMA_MEM_TO_DEV)
+		reg_width = __ffs(sconfig->dst_addr_width);
+	else
+		reg_width = __ffs(sconfig->src_addr_width);
+
 	periods = buf_len / period_len;
 
 	/* Check for too big/unaligned periods and unaligned DMA buffer. */
@@ -1236,26 +1256,34 @@
 
 		switch (direction) {
 		case DMA_MEM_TO_DEV:
-			desc->lli.dar = dws->tx_reg;
+			desc->lli.dar = sconfig->dst_addr;
 			desc->lli.sar = buf_addr + (period_len * i);
-			desc->lli.ctllo = (DWC_DEFAULT_CTLLO(chan->private)
+			desc->lli.ctllo = (DWC_DEFAULT_CTLLO(chan)
 					| DWC_CTLL_DST_WIDTH(reg_width)
 					| DWC_CTLL_SRC_WIDTH(reg_width)
 					| DWC_CTLL_DST_FIX
 					| DWC_CTLL_SRC_INC
-					| DWC_CTLL_FC(dws->fc)
 					| DWC_CTLL_INT_EN);
+
+			desc->lli.ctllo |= sconfig->device_fc ?
+				DWC_CTLL_FC(DW_DMA_FC_P_M2P) :
+				DWC_CTLL_FC(DW_DMA_FC_D_M2P);
+
 			break;
 		case DMA_DEV_TO_MEM:
 			desc->lli.dar = buf_addr + (period_len * i);
-			desc->lli.sar = dws->rx_reg;
-			desc->lli.ctllo = (DWC_DEFAULT_CTLLO(chan->private)
+			desc->lli.sar = sconfig->src_addr;
+			desc->lli.ctllo = (DWC_DEFAULT_CTLLO(chan)
 					| DWC_CTLL_SRC_WIDTH(reg_width)
 					| DWC_CTLL_DST_WIDTH(reg_width)
 					| DWC_CTLL_DST_INC
 					| DWC_CTLL_SRC_FIX
-					| DWC_CTLL_FC(dws->fc)
 					| DWC_CTLL_INT_EN);
+
+			desc->lli.ctllo |= sconfig->device_fc ?
+				DWC_CTLL_FC(DW_DMA_FC_P_P2M) :
+				DWC_CTLL_FC(DW_DMA_FC_D_P2M);
+
 			break;
 		default:
 			break;
@@ -1322,7 +1350,6 @@
 	while (dma_readl(dw, CH_EN) & dwc->mask)
 		cpu_relax();
 
-	dma_writel(dw, CLEAR.BLOCK, dwc->mask);
 	dma_writel(dw, CLEAR.ERROR, dwc->mask);
 	dma_writel(dw, CLEAR.XFER, dwc->mask);
 
@@ -1347,7 +1374,6 @@
 	dma_writel(dw, CFG, 0);
 
 	channel_clear_bit(dw, MASK.XFER, dw->all_chan_mask);
-	channel_clear_bit(dw, MASK.BLOCK, dw->all_chan_mask);
 	channel_clear_bit(dw, MASK.SRC_TRAN, dw->all_chan_mask);
 	channel_clear_bit(dw, MASK.DST_TRAN, dw->all_chan_mask);
 	channel_clear_bit(dw, MASK.ERROR, dw->all_chan_mask);
@@ -1369,7 +1395,7 @@
 	int			err;
 	int			i;
 
-	pdata = pdev->dev.platform_data;
+	pdata = dev_get_platdata(&pdev->dev);
 	if (!pdata || pdata->nr_channels > DW_DMA_MAX_NR_CHANNELS)
 		return -EINVAL;
 
@@ -1423,7 +1449,7 @@
 		struct dw_dma_chan	*dwc = &dw->chan[i];
 
 		dwc->chan.device = &dw->dma;
-		dwc->chan.cookie = dwc->completed = 1;
+		dma_cookie_init(&dwc->chan);
 		if (pdata->chan_allocation_order == CHAN_ALLOCATION_ASCENDING)
 			list_add_tail(&dwc->chan.device_node,
 					&dw->dma.channels);
@@ -1432,7 +1458,7 @@
 
 		/* 7 is highest priority & 0 is lowest. */
 		if (pdata->chan_priority == CHAN_PRIORITY_ASCENDING)
-			dwc->priority = 7 - i;
+			dwc->priority = pdata->nr_channels - i - 1;
 		else
 			dwc->priority = i;
 
@@ -1449,13 +1475,11 @@
 
 	/* Clear/disable all interrupts on all channels. */
 	dma_writel(dw, CLEAR.XFER, dw->all_chan_mask);
-	dma_writel(dw, CLEAR.BLOCK, dw->all_chan_mask);
 	dma_writel(dw, CLEAR.SRC_TRAN, dw->all_chan_mask);
 	dma_writel(dw, CLEAR.DST_TRAN, dw->all_chan_mask);
 	dma_writel(dw, CLEAR.ERROR, dw->all_chan_mask);
 
 	channel_clear_bit(dw, MASK.XFER, dw->all_chan_mask);
-	channel_clear_bit(dw, MASK.BLOCK, dw->all_chan_mask);
 	channel_clear_bit(dw, MASK.SRC_TRAN, dw->all_chan_mask);
 	channel_clear_bit(dw, MASK.DST_TRAN, dw->all_chan_mask);
 	channel_clear_bit(dw, MASK.ERROR, dw->all_chan_mask);
@@ -1562,6 +1586,10 @@
 static const struct dev_pm_ops dw_dev_pm_ops = {
 	.suspend_noirq = dw_suspend_noirq,
 	.resume_noirq = dw_resume_noirq,
+	.freeze_noirq = dw_suspend_noirq,
+	.thaw_noirq = dw_resume_noirq,
+	.restore_noirq = dw_resume_noirq,
+	.poweroff_noirq = dw_suspend_noirq,
 };
 
 static struct platform_driver dw_driver = {
diff --git a/drivers/dma/dw_dmac_regs.h b/drivers/dma/dw_dmac_regs.h
index 5eef694..f298f69 100644
--- a/drivers/dma/dw_dmac_regs.h
+++ b/drivers/dma/dw_dmac_regs.h
@@ -13,6 +13,18 @@
 
 #define DW_DMA_MAX_NR_CHANNELS	8
 
+/* flow controller */
+enum dw_dma_fc {
+	DW_DMA_FC_D_M2M,
+	DW_DMA_FC_D_M2P,
+	DW_DMA_FC_D_P2M,
+	DW_DMA_FC_D_P2P,
+	DW_DMA_FC_P_P2M,
+	DW_DMA_FC_SP_P2P,
+	DW_DMA_FC_P_M2P,
+	DW_DMA_FC_DP_P2P,
+};
+
 /*
  * Redefine this macro to handle differences between 32- and 64-bit
  * addressing, big vs. little endian, etc.
@@ -146,13 +158,15 @@
 
 	/* these other elements are all protected by lock */
 	unsigned long		flags;
-	dma_cookie_t		completed;
 	struct list_head	active_list;
 	struct list_head	queue;
 	struct list_head	free_list;
 	struct dw_cyclic_desc	*cdesc;
 
 	unsigned int		descs_allocated;
+
+	/* configuration passed via DMA_SLAVE_CONFIG */
+	struct dma_slave_config dma_sconfig;
 };
 
 static inline struct dw_dma_chan_regs __iomem *
diff --git a/drivers/dma/ep93xx_dma.c b/drivers/dma/ep93xx_dma.c
index 59e7a96..e6f133b 100644
--- a/drivers/dma/ep93xx_dma.c
+++ b/drivers/dma/ep93xx_dma.c
@@ -28,6 +28,8 @@
 
 #include <mach/dma.h>
 
+#include "dmaengine.h"
+
 /* M2P registers */
 #define M2P_CONTROL			0x0000
 #define M2P_CONTROL_STALLINT		BIT(0)
@@ -122,7 +124,6 @@
  * @lock: lock protecting the fields following
  * @flags: flags for the channel
  * @buffer: which buffer to use next (0/1)
- * @last_completed: last completed cookie value
  * @active: flattened chain of descriptors currently being processed
  * @queue: pending descriptors which are handled next
  * @free_list: list of free descriptors which can be used
@@ -157,7 +158,6 @@
 #define EP93XX_DMA_IS_CYCLIC		0
 
 	int				buffer;
-	dma_cookie_t			last_completed;
 	struct list_head		active;
 	struct list_head		queue;
 	struct list_head		free_list;
@@ -703,7 +703,7 @@
 	desc = ep93xx_dma_get_active(edmac);
 	if (desc) {
 		if (desc->complete) {
-			edmac->last_completed = desc->txd.cookie;
+			dma_cookie_complete(&desc->txd);
 			list_splice_init(&edmac->active, &list);
 		}
 		callback = desc->txd.callback;
@@ -783,17 +783,10 @@
 	unsigned long flags;
 
 	spin_lock_irqsave(&edmac->lock, flags);
-
-	cookie = edmac->chan.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
+	cookie = dma_cookie_assign(tx);
 
 	desc = container_of(tx, struct ep93xx_dma_desc, txd);
 
-	edmac->chan.cookie = cookie;
-	desc->txd.cookie = cookie;
-
 	/*
 	 * If nothing is currently prosessed, we push this descriptor
 	 * directly to the hardware. Otherwise we put the descriptor
@@ -861,8 +854,7 @@
 		goto fail_clk_disable;
 
 	spin_lock_irq(&edmac->lock);
-	edmac->last_completed = 1;
-	edmac->chan.cookie = 1;
+	dma_cookie_init(&edmac->chan);
 	ret = edmac->edma->hw_setup(edmac);
 	spin_unlock_irq(&edmac->lock);
 
@@ -983,13 +975,14 @@
  * @sg_len: number of entries in @sgl
  * @dir: direction of tha DMA transfer
  * @flags: flags for the descriptor
+ * @context: operation context (ignored)
  *
  * Returns a valid DMA descriptor or %NULL in case of failure.
  */
 static struct dma_async_tx_descriptor *
 ep93xx_dma_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 			 unsigned int sg_len, enum dma_transfer_direction dir,
-			 unsigned long flags)
+			 unsigned long flags, void *context)
 {
 	struct ep93xx_dma_chan *edmac = to_ep93xx_dma_chan(chan);
 	struct ep93xx_dma_desc *desc, *first;
@@ -1056,6 +1049,7 @@
  * @buf_len: length of the buffer (in bytes)
  * @period_len: lenght of a single period
  * @dir: direction of the operation
+ * @context: operation context (ignored)
  *
  * Prepares a descriptor for cyclic DMA operation. This means that once the
  * descriptor is submitted, we will be submitting in a @period_len sized
@@ -1068,7 +1062,7 @@
 static struct dma_async_tx_descriptor *
 ep93xx_dma_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t dma_addr,
 			   size_t buf_len, size_t period_len,
-			   enum dma_transfer_direction dir)
+			   enum dma_transfer_direction dir, void *context)
 {
 	struct ep93xx_dma_chan *edmac = to_ep93xx_dma_chan(chan);
 	struct ep93xx_dma_desc *desc, *first;
@@ -1248,18 +1242,13 @@
 					    struct dma_tx_state *state)
 {
 	struct ep93xx_dma_chan *edmac = to_ep93xx_dma_chan(chan);
-	dma_cookie_t last_used, last_completed;
 	enum dma_status ret;
 	unsigned long flags;
 
 	spin_lock_irqsave(&edmac->lock, flags);
-	last_used = chan->cookie;
-	last_completed = edmac->last_completed;
+	ret = dma_cookie_status(chan, cookie, state);
 	spin_unlock_irqrestore(&edmac->lock, flags);
 
-	ret = dma_async_is_complete(cookie, last_completed, last_used);
-	dma_set_tx_state(state, last_completed, last_used, 0);
-
 	return ret;
 }
 
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index b98070c..8f84761 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -35,6 +35,7 @@
 #include <linux/dmapool.h>
 #include <linux/of_platform.h>
 
+#include "dmaengine.h"
 #include "fsldma.h"
 
 #define chan_dbg(chan, fmt, arg...)					\
@@ -413,17 +414,10 @@
 	 * assign cookies to all of the software descriptors
 	 * that make up this transaction
 	 */
-	cookie = chan->common.cookie;
 	list_for_each_entry(child, &desc->tx_list, node) {
-		cookie++;
-		if (cookie < DMA_MIN_COOKIE)
-			cookie = DMA_MIN_COOKIE;
-
-		child->async_tx.cookie = cookie;
+		cookie = dma_cookie_assign(&child->async_tx);
 	}
 
-	chan->common.cookie = cookie;
-
 	/* put this transaction onto the tail of the pending queue */
 	append_ld_queue(chan, desc);
 
@@ -765,6 +759,7 @@
  * @sg_len: number of entries in @scatterlist
  * @direction: DMA direction
  * @flags: DMAEngine flags
+ * @context: transaction context (ignored)
  *
  * Prepare a set of descriptors for a DMA_SLAVE transaction. Following the
  * DMA_SLAVE API, this gets the device-specific information from the
@@ -772,7 +767,8 @@
  */
 static struct dma_async_tx_descriptor *fsl_dma_prep_slave_sg(
 	struct dma_chan *dchan, struct scatterlist *sgl, unsigned int sg_len,
-	enum dma_transfer_direction direction, unsigned long flags)
+	enum dma_transfer_direction direction, unsigned long flags,
+	void *context)
 {
 	/*
 	 * This operation is not supported on the Freescale DMA controller
@@ -984,19 +980,14 @@
 					struct dma_tx_state *txstate)
 {
 	struct fsldma_chan *chan = to_fsl_chan(dchan);
-	dma_cookie_t last_complete;
-	dma_cookie_t last_used;
+	enum dma_status ret;
 	unsigned long flags;
 
 	spin_lock_irqsave(&chan->desc_lock, flags);
-
-	last_complete = chan->completed_cookie;
-	last_used = dchan->cookie;
-
+	ret = dma_cookie_status(dchan, cookie, txstate);
 	spin_unlock_irqrestore(&chan->desc_lock, flags);
 
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-	return dma_async_is_complete(cookie, last_complete, last_used);
+	return ret;
 }
 
 /*----------------------------------------------------------------------------*/
@@ -1087,8 +1078,8 @@
 
 		desc = to_fsl_desc(chan->ld_running.prev);
 		cookie = desc->async_tx.cookie;
+		dma_cookie_complete(&desc->async_tx);
 
-		chan->completed_cookie = cookie;
 		chan_dbg(chan, "completed_cookie=%d\n", cookie);
 	}
 
@@ -1303,6 +1294,7 @@
 	chan->idle = true;
 
 	chan->common.device = &fdev->common;
+	dma_cookie_init(&chan->common);
 
 	/* find the IRQ line, if it exists in the device tree */
 	chan->irq = irq_of_parse_and_map(node, 0);
diff --git a/drivers/dma/fsldma.h b/drivers/dma/fsldma.h
index 9cb5aa5..f5c3879 100644
--- a/drivers/dma/fsldma.h
+++ b/drivers/dma/fsldma.h
@@ -137,7 +137,6 @@
 struct fsldma_chan {
 	char name[8];			/* Channel name */
 	struct fsldma_chan_regs __iomem *regs;
-	dma_cookie_t completed_cookie;	/* The maximum cookie completed */
 	spinlock_t desc_lock;		/* Descriptor operation lock */
 	struct list_head ld_pending;	/* Link descriptors queue */
 	struct list_head ld_running;	/* Link descriptors queue */
diff --git a/drivers/dma/imx-dma.c b/drivers/dma/imx-dma.c
index 38586ba..a45b5d2 100644
--- a/drivers/dma/imx-dma.c
+++ b/drivers/dma/imx-dma.c
@@ -5,6 +5,7 @@
  * found on i.MX1/21/27
  *
  * Copyright 2010 Sascha Hauer, Pengutronix <s.hauer@pengutronix.de>
+ * Copyright 2012 Javier Martin, Vista Silicon <javier.martin@vista-silicon.com>
  *
  * The code contained herein is licensed under the GNU General Public
  * License. You may obtain a copy of the GNU General Public License
@@ -22,37 +23,159 @@
 #include <linux/dma-mapping.h>
 #include <linux/slab.h>
 #include <linux/platform_device.h>
+#include <linux/clk.h>
 #include <linux/dmaengine.h>
 #include <linux/module.h>
 
 #include <asm/irq.h>
-#include <mach/dma-v1.h>
+#include <mach/dma.h>
 #include <mach/hardware.h>
 
+#include "dmaengine.h"
+#define IMXDMA_MAX_CHAN_DESCRIPTORS	16
+#define IMX_DMA_CHANNELS  16
+
+#define IMX_DMA_2D_SLOTS	2
+#define IMX_DMA_2D_SLOT_A	0
+#define IMX_DMA_2D_SLOT_B	1
+
+#define IMX_DMA_LENGTH_LOOP	((unsigned int)-1)
+#define IMX_DMA_MEMSIZE_32	(0 << 4)
+#define IMX_DMA_MEMSIZE_8	(1 << 4)
+#define IMX_DMA_MEMSIZE_16	(2 << 4)
+#define IMX_DMA_TYPE_LINEAR	(0 << 10)
+#define IMX_DMA_TYPE_2D		(1 << 10)
+#define IMX_DMA_TYPE_FIFO	(2 << 10)
+
+#define IMX_DMA_ERR_BURST     (1 << 0)
+#define IMX_DMA_ERR_REQUEST   (1 << 1)
+#define IMX_DMA_ERR_TRANSFER  (1 << 2)
+#define IMX_DMA_ERR_BUFFER    (1 << 3)
+#define IMX_DMA_ERR_TIMEOUT   (1 << 4)
+
+#define DMA_DCR     0x00		/* Control Register */
+#define DMA_DISR    0x04		/* Interrupt status Register */
+#define DMA_DIMR    0x08		/* Interrupt mask Register */
+#define DMA_DBTOSR  0x0c		/* Burst timeout status Register */
+#define DMA_DRTOSR  0x10		/* Request timeout Register */
+#define DMA_DSESR   0x14		/* Transfer Error Status Register */
+#define DMA_DBOSR   0x18		/* Buffer overflow status Register */
+#define DMA_DBTOCR  0x1c		/* Burst timeout control Register */
+#define DMA_WSRA    0x40		/* W-Size Register A */
+#define DMA_XSRA    0x44		/* X-Size Register A */
+#define DMA_YSRA    0x48		/* Y-Size Register A */
+#define DMA_WSRB    0x4c		/* W-Size Register B */
+#define DMA_XSRB    0x50		/* X-Size Register B */
+#define DMA_YSRB    0x54		/* Y-Size Register B */
+#define DMA_SAR(x)  (0x80 + ((x) << 6))	/* Source Address Registers */
+#define DMA_DAR(x)  (0x84 + ((x) << 6))	/* Destination Address Registers */
+#define DMA_CNTR(x) (0x88 + ((x) << 6))	/* Count Registers */
+#define DMA_CCR(x)  (0x8c + ((x) << 6))	/* Control Registers */
+#define DMA_RSSR(x) (0x90 + ((x) << 6))	/* Request source select Registers */
+#define DMA_BLR(x)  (0x94 + ((x) << 6))	/* Burst length Registers */
+#define DMA_RTOR(x) (0x98 + ((x) << 6))	/* Request timeout Registers */
+#define DMA_BUCR(x) (0x98 + ((x) << 6))	/* Bus Utilization Registers */
+#define DMA_CCNR(x) (0x9C + ((x) << 6))	/* Channel counter Registers */
+
+#define DCR_DRST           (1<<1)
+#define DCR_DEN            (1<<0)
+#define DBTOCR_EN          (1<<15)
+#define DBTOCR_CNT(x)      ((x) & 0x7fff)
+#define CNTR_CNT(x)        ((x) & 0xffffff)
+#define CCR_ACRPT          (1<<14)
+#define CCR_DMOD_LINEAR    (0x0 << 12)
+#define CCR_DMOD_2D        (0x1 << 12)
+#define CCR_DMOD_FIFO      (0x2 << 12)
+#define CCR_DMOD_EOBFIFO   (0x3 << 12)
+#define CCR_SMOD_LINEAR    (0x0 << 10)
+#define CCR_SMOD_2D        (0x1 << 10)
+#define CCR_SMOD_FIFO      (0x2 << 10)
+#define CCR_SMOD_EOBFIFO   (0x3 << 10)
+#define CCR_MDIR_DEC       (1<<9)
+#define CCR_MSEL_B         (1<<8)
+#define CCR_DSIZ_32        (0x0 << 6)
+#define CCR_DSIZ_8         (0x1 << 6)
+#define CCR_DSIZ_16        (0x2 << 6)
+#define CCR_SSIZ_32        (0x0 << 4)
+#define CCR_SSIZ_8         (0x1 << 4)
+#define CCR_SSIZ_16        (0x2 << 4)
+#define CCR_REN            (1<<3)
+#define CCR_RPT            (1<<2)
+#define CCR_FRC            (1<<1)
+#define CCR_CEN            (1<<0)
+#define RTOR_EN            (1<<15)
+#define RTOR_CLK           (1<<14)
+#define RTOR_PSC           (1<<13)
+
+enum  imxdma_prep_type {
+	IMXDMA_DESC_MEMCPY,
+	IMXDMA_DESC_INTERLEAVED,
+	IMXDMA_DESC_SLAVE_SG,
+	IMXDMA_DESC_CYCLIC,
+};
+
+struct imx_dma_2d_config {
+	u16		xsr;
+	u16		ysr;
+	u16		wsr;
+	int		count;
+};
+
+struct imxdma_desc {
+	struct list_head		node;
+	struct dma_async_tx_descriptor	desc;
+	enum dma_status			status;
+	dma_addr_t			src;
+	dma_addr_t			dest;
+	size_t				len;
+	enum dma_transfer_direction	direction;
+	enum imxdma_prep_type		type;
+	/* For memcpy and interleaved */
+	unsigned int			config_port;
+	unsigned int			config_mem;
+	/* For interleaved transfers */
+	unsigned int			x;
+	unsigned int			y;
+	unsigned int			w;
+	/* For slave sg and cyclic */
+	struct scatterlist		*sg;
+	unsigned int			sgcount;
+};
+
 struct imxdma_channel {
+	int				hw_chaining;
+	struct timer_list		watchdog;
 	struct imxdma_engine		*imxdma;
 	unsigned int			channel;
-	unsigned int			imxdma_channel;
 
+	struct tasklet_struct		dma_tasklet;
+	struct list_head		ld_free;
+	struct list_head		ld_queue;
+	struct list_head		ld_active;
+	int				descs_allocated;
 	enum dma_slave_buswidth		word_size;
 	dma_addr_t			per_address;
 	u32				watermark_level;
 	struct dma_chan			chan;
-	spinlock_t			lock;
 	struct dma_async_tx_descriptor	desc;
-	dma_cookie_t			last_completed;
 	enum dma_status			status;
 	int				dma_request;
 	struct scatterlist		*sg_list;
+	u32				ccr_from_device;
+	u32				ccr_to_device;
+	bool				enabled_2d;
+	int				slot_2d;
 };
 
-#define MAX_DMA_CHANNELS 8
-
 struct imxdma_engine {
 	struct device			*dev;
 	struct device_dma_parameters	dma_parms;
 	struct dma_device		dma_device;
-	struct imxdma_channel		channel[MAX_DMA_CHANNELS];
+	void __iomem			*base;
+	struct clk			*dma_clk;
+	spinlock_t			lock;
+	struct imx_dma_2d_config	slots_2d[IMX_DMA_2D_SLOTS];
+	struct imxdma_channel		channel[IMX_DMA_CHANNELS];
 };
 
 static struct imxdma_channel *to_imxdma_chan(struct dma_chan *chan)
@@ -60,36 +183,418 @@
 	return container_of(chan, struct imxdma_channel, chan);
 }
 
-static void imxdma_handle(struct imxdma_channel *imxdmac)
+static inline bool imxdma_chan_is_doing_cyclic(struct imxdma_channel *imxdmac)
 {
-	if (imxdmac->desc.callback)
-		imxdmac->desc.callback(imxdmac->desc.callback_param);
-	imxdmac->last_completed = imxdmac->desc.cookie;
+	struct imxdma_desc *desc;
+
+	if (!list_empty(&imxdmac->ld_active)) {
+		desc = list_first_entry(&imxdmac->ld_active, struct imxdma_desc,
+					node);
+		if (desc->type == IMXDMA_DESC_CYCLIC)
+			return true;
+	}
+	return false;
 }
 
-static void imxdma_irq_handler(int channel, void *data)
-{
-	struct imxdma_channel *imxdmac = data;
 
-	imxdmac->status = DMA_SUCCESS;
-	imxdma_handle(imxdmac);
+
+static void imx_dmav1_writel(struct imxdma_engine *imxdma, unsigned val,
+			     unsigned offset)
+{
+	__raw_writel(val, imxdma->base + offset);
 }
 
-static void imxdma_err_handler(int channel, void *data, int error)
+static unsigned imx_dmav1_readl(struct imxdma_engine *imxdma, unsigned offset)
 {
-	struct imxdma_channel *imxdmac = data;
-
-	imxdmac->status = DMA_ERROR;
-	imxdma_handle(imxdmac);
+	return __raw_readl(imxdma->base + offset);
 }
 
-static void imxdma_progression(int channel, void *data,
-		struct scatterlist *sg)
+static int imxdma_hw_chain(struct imxdma_channel *imxdmac)
 {
-	struct imxdma_channel *imxdmac = data;
+	if (cpu_is_mx27())
+		return imxdmac->hw_chaining;
+	else
+		return 0;
+}
 
-	imxdmac->status = DMA_SUCCESS;
-	imxdma_handle(imxdmac);
+/*
+ * imxdma_sg_next - prepare next chunk for scatter-gather DMA emulation
+ */
+static inline int imxdma_sg_next(struct imxdma_desc *d)
+{
+	struct imxdma_channel *imxdmac = to_imxdma_chan(d->desc.chan);
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	struct scatterlist *sg = d->sg;
+	unsigned long now;
+
+	now = min(d->len, sg->length);
+	if (d->len != IMX_DMA_LENGTH_LOOP)
+		d->len -= now;
+
+	if (d->direction == DMA_DEV_TO_MEM)
+		imx_dmav1_writel(imxdma, sg->dma_address,
+				 DMA_DAR(imxdmac->channel));
+	else
+		imx_dmav1_writel(imxdma, sg->dma_address,
+				 DMA_SAR(imxdmac->channel));
+
+	imx_dmav1_writel(imxdma, now, DMA_CNTR(imxdmac->channel));
+
+	dev_dbg(imxdma->dev, " %s channel: %d dst 0x%08x, src 0x%08x, "
+		"size 0x%08x\n", __func__, imxdmac->channel,
+		 imx_dmav1_readl(imxdma, DMA_DAR(imxdmac->channel)),
+		 imx_dmav1_readl(imxdma, DMA_SAR(imxdmac->channel)),
+		 imx_dmav1_readl(imxdma, DMA_CNTR(imxdmac->channel)));
+
+	return now;
+}
+
+static void imxdma_enable_hw(struct imxdma_desc *d)
+{
+	struct imxdma_channel *imxdmac = to_imxdma_chan(d->desc.chan);
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	int channel = imxdmac->channel;
+	unsigned long flags;
+
+	dev_dbg(imxdma->dev, "%s channel %d\n", __func__, channel);
+
+	local_irq_save(flags);
+
+	imx_dmav1_writel(imxdma, 1 << channel, DMA_DISR);
+	imx_dmav1_writel(imxdma, imx_dmav1_readl(imxdma, DMA_DIMR) &
+			 ~(1 << channel), DMA_DIMR);
+	imx_dmav1_writel(imxdma, imx_dmav1_readl(imxdma, DMA_CCR(channel)) |
+			 CCR_CEN | CCR_ACRPT, DMA_CCR(channel));
+
+	if ((cpu_is_mx21() || cpu_is_mx27()) &&
+			d->sg && imxdma_hw_chain(imxdmac)) {
+		d->sg = sg_next(d->sg);
+		if (d->sg) {
+			u32 tmp;
+			imxdma_sg_next(d);
+			tmp = imx_dmav1_readl(imxdma, DMA_CCR(channel));
+			imx_dmav1_writel(imxdma, tmp | CCR_RPT | CCR_ACRPT,
+					 DMA_CCR(channel));
+		}
+	}
+
+	local_irq_restore(flags);
+}
+
+static void imxdma_disable_hw(struct imxdma_channel *imxdmac)
+{
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	int channel = imxdmac->channel;
+	unsigned long flags;
+
+	dev_dbg(imxdma->dev, "%s channel %d\n", __func__, channel);
+
+	if (imxdma_hw_chain(imxdmac))
+		del_timer(&imxdmac->watchdog);
+
+	local_irq_save(flags);
+	imx_dmav1_writel(imxdma, imx_dmav1_readl(imxdma, DMA_DIMR) |
+			 (1 << channel), DMA_DIMR);
+	imx_dmav1_writel(imxdma, imx_dmav1_readl(imxdma, DMA_CCR(channel)) &
+			 ~CCR_CEN, DMA_CCR(channel));
+	imx_dmav1_writel(imxdma, 1 << channel, DMA_DISR);
+	local_irq_restore(flags);
+}
+
+static void imxdma_watchdog(unsigned long data)
+{
+	struct imxdma_channel *imxdmac = (struct imxdma_channel *)data;
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	int channel = imxdmac->channel;
+
+	imx_dmav1_writel(imxdma, 0, DMA_CCR(channel));
+
+	/* Tasklet watchdog error handler */
+	tasklet_schedule(&imxdmac->dma_tasklet);
+	dev_dbg(imxdma->dev, "channel %d: watchdog timeout!\n",
+		imxdmac->channel);
+}
+
+static irqreturn_t imxdma_err_handler(int irq, void *dev_id)
+{
+	struct imxdma_engine *imxdma = dev_id;
+	unsigned int err_mask;
+	int i, disr;
+	int errcode;
+
+	disr = imx_dmav1_readl(imxdma, DMA_DISR);
+
+	err_mask = imx_dmav1_readl(imxdma, DMA_DBTOSR) |
+		   imx_dmav1_readl(imxdma, DMA_DRTOSR) |
+		   imx_dmav1_readl(imxdma, DMA_DSESR)  |
+		   imx_dmav1_readl(imxdma, DMA_DBOSR);
+
+	if (!err_mask)
+		return IRQ_HANDLED;
+
+	imx_dmav1_writel(imxdma, disr & err_mask, DMA_DISR);
+
+	for (i = 0; i < IMX_DMA_CHANNELS; i++) {
+		if (!(err_mask & (1 << i)))
+			continue;
+		errcode = 0;
+
+		if (imx_dmav1_readl(imxdma, DMA_DBTOSR) & (1 << i)) {
+			imx_dmav1_writel(imxdma, 1 << i, DMA_DBTOSR);
+			errcode |= IMX_DMA_ERR_BURST;
+		}
+		if (imx_dmav1_readl(imxdma, DMA_DRTOSR) & (1 << i)) {
+			imx_dmav1_writel(imxdma, 1 << i, DMA_DRTOSR);
+			errcode |= IMX_DMA_ERR_REQUEST;
+		}
+		if (imx_dmav1_readl(imxdma, DMA_DSESR) & (1 << i)) {
+			imx_dmav1_writel(imxdma, 1 << i, DMA_DSESR);
+			errcode |= IMX_DMA_ERR_TRANSFER;
+		}
+		if (imx_dmav1_readl(imxdma, DMA_DBOSR) & (1 << i)) {
+			imx_dmav1_writel(imxdma, 1 << i, DMA_DBOSR);
+			errcode |= IMX_DMA_ERR_BUFFER;
+		}
+		/* Tasklet error handler */
+		tasklet_schedule(&imxdma->channel[i].dma_tasklet);
+
+		printk(KERN_WARNING
+		       "DMA timeout on channel %d -%s%s%s%s\n", i,
+		       errcode & IMX_DMA_ERR_BURST ?    " burst" : "",
+		       errcode & IMX_DMA_ERR_REQUEST ?  " request" : "",
+		       errcode & IMX_DMA_ERR_TRANSFER ? " transfer" : "",
+		       errcode & IMX_DMA_ERR_BUFFER ?   " buffer" : "");
+	}
+	return IRQ_HANDLED;
+}
+
+static void dma_irq_handle_channel(struct imxdma_channel *imxdmac)
+{
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	int chno = imxdmac->channel;
+	struct imxdma_desc *desc;
+
+	spin_lock(&imxdma->lock);
+	if (list_empty(&imxdmac->ld_active)) {
+		spin_unlock(&imxdma->lock);
+		goto out;
+	}
+
+	desc = list_first_entry(&imxdmac->ld_active,
+				struct imxdma_desc,
+				node);
+	spin_unlock(&imxdma->lock);
+
+	if (desc->sg) {
+		u32 tmp;
+		desc->sg = sg_next(desc->sg);
+
+		if (desc->sg) {
+			imxdma_sg_next(desc);
+
+			tmp = imx_dmav1_readl(imxdma, DMA_CCR(chno));
+
+			if (imxdma_hw_chain(imxdmac)) {
+				/* FIXME: The timeout should probably be
+				 * configurable
+				 */
+				mod_timer(&imxdmac->watchdog,
+					jiffies + msecs_to_jiffies(500));
+
+				tmp |= CCR_CEN | CCR_RPT | CCR_ACRPT;
+				imx_dmav1_writel(imxdma, tmp, DMA_CCR(chno));
+			} else {
+				imx_dmav1_writel(imxdma, tmp & ~CCR_CEN,
+						 DMA_CCR(chno));
+				tmp |= CCR_CEN;
+			}
+
+			imx_dmav1_writel(imxdma, tmp, DMA_CCR(chno));
+
+			if (imxdma_chan_is_doing_cyclic(imxdmac))
+				/* Tasklet progression */
+				tasklet_schedule(&imxdmac->dma_tasklet);
+
+			return;
+		}
+
+		if (imxdma_hw_chain(imxdmac)) {
+			del_timer(&imxdmac->watchdog);
+			return;
+		}
+	}
+
+out:
+	imx_dmav1_writel(imxdma, 0, DMA_CCR(chno));
+	/* Tasklet irq */
+	tasklet_schedule(&imxdmac->dma_tasklet);
+}
+
+static irqreturn_t dma_irq_handler(int irq, void *dev_id)
+{
+	struct imxdma_engine *imxdma = dev_id;
+	int i, disr;
+
+	if (cpu_is_mx21() || cpu_is_mx27())
+		imxdma_err_handler(irq, dev_id);
+
+	disr = imx_dmav1_readl(imxdma, DMA_DISR);
+
+	dev_dbg(imxdma->dev, "%s called, disr=0x%08x\n", __func__, disr);
+
+	imx_dmav1_writel(imxdma, disr, DMA_DISR);
+	for (i = 0; i < IMX_DMA_CHANNELS; i++) {
+		if (disr & (1 << i))
+			dma_irq_handle_channel(&imxdma->channel[i]);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int imxdma_xfer_desc(struct imxdma_desc *d)
+{
+	struct imxdma_channel *imxdmac = to_imxdma_chan(d->desc.chan);
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	unsigned long flags;
+	int slot = -1;
+	int i;
+
+	/* Configure and enable */
+	switch (d->type) {
+	case IMXDMA_DESC_INTERLEAVED:
+		/* Try to get a free 2D slot */
+		spin_lock_irqsave(&imxdma->lock, flags);
+		for (i = 0; i < IMX_DMA_2D_SLOTS; i++) {
+			if ((imxdma->slots_2d[i].count > 0) &&
+			((imxdma->slots_2d[i].xsr != d->x) ||
+			(imxdma->slots_2d[i].ysr != d->y) ||
+			(imxdma->slots_2d[i].wsr != d->w)))
+				continue;
+			slot = i;
+			break;
+		}
+		if (slot < 0)
+			return -EBUSY;
+
+		imxdma->slots_2d[slot].xsr = d->x;
+		imxdma->slots_2d[slot].ysr = d->y;
+		imxdma->slots_2d[slot].wsr = d->w;
+		imxdma->slots_2d[slot].count++;
+
+		imxdmac->slot_2d = slot;
+		imxdmac->enabled_2d = true;
+		spin_unlock_irqrestore(&imxdma->lock, flags);
+
+		if (slot == IMX_DMA_2D_SLOT_A) {
+			d->config_mem &= ~CCR_MSEL_B;
+			d->config_port &= ~CCR_MSEL_B;
+			imx_dmav1_writel(imxdma, d->x, DMA_XSRA);
+			imx_dmav1_writel(imxdma, d->y, DMA_YSRA);
+			imx_dmav1_writel(imxdma, d->w, DMA_WSRA);
+		} else {
+			d->config_mem |= CCR_MSEL_B;
+			d->config_port |= CCR_MSEL_B;
+			imx_dmav1_writel(imxdma, d->x, DMA_XSRB);
+			imx_dmav1_writel(imxdma, d->y, DMA_YSRB);
+			imx_dmav1_writel(imxdma, d->w, DMA_WSRB);
+		}
+		/*
+		 * We fall-through here intentionally, since a 2D transfer is
+		 * similar to MEMCPY just adding the 2D slot configuration.
+		 */
+	case IMXDMA_DESC_MEMCPY:
+		imx_dmav1_writel(imxdma, d->src, DMA_SAR(imxdmac->channel));
+		imx_dmav1_writel(imxdma, d->dest, DMA_DAR(imxdmac->channel));
+		imx_dmav1_writel(imxdma, d->config_mem | (d->config_port << 2),
+			 DMA_CCR(imxdmac->channel));
+
+		imx_dmav1_writel(imxdma, d->len, DMA_CNTR(imxdmac->channel));
+
+		dev_dbg(imxdma->dev, "%s channel: %d dest=0x%08x src=0x%08x "
+			"dma_length=%d\n", __func__, imxdmac->channel,
+			d->dest, d->src, d->len);
+
+		break;
+	/* Cyclic transfer is the same as slave_sg with special sg configuration. */
+	case IMXDMA_DESC_CYCLIC:
+	case IMXDMA_DESC_SLAVE_SG:
+		if (d->direction == DMA_DEV_TO_MEM) {
+			imx_dmav1_writel(imxdma, imxdmac->per_address,
+					 DMA_SAR(imxdmac->channel));
+			imx_dmav1_writel(imxdma, imxdmac->ccr_from_device,
+					 DMA_CCR(imxdmac->channel));
+
+			dev_dbg(imxdma->dev, "%s channel: %d sg=%p sgcount=%d "
+				"total length=%d dev_addr=0x%08x (dev2mem)\n",
+				__func__, imxdmac->channel, d->sg, d->sgcount,
+				d->len, imxdmac->per_address);
+		} else if (d->direction == DMA_MEM_TO_DEV) {
+			imx_dmav1_writel(imxdma, imxdmac->per_address,
+					 DMA_DAR(imxdmac->channel));
+			imx_dmav1_writel(imxdma, imxdmac->ccr_to_device,
+					 DMA_CCR(imxdmac->channel));
+
+			dev_dbg(imxdma->dev, "%s channel: %d sg=%p sgcount=%d "
+				"total length=%d dev_addr=0x%08x (mem2dev)\n",
+				__func__, imxdmac->channel, d->sg, d->sgcount,
+				d->len, imxdmac->per_address);
+		} else {
+			dev_err(imxdma->dev, "%s channel: %d bad dma mode\n",
+				__func__, imxdmac->channel);
+			return -EINVAL;
+		}
+
+		imxdma_sg_next(d);
+
+		break;
+	default:
+		return -EINVAL;
+	}
+	imxdma_enable_hw(d);
+	return 0;
+}
+
+static void imxdma_tasklet(unsigned long data)
+{
+	struct imxdma_channel *imxdmac = (void *)data;
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	struct imxdma_desc *desc;
+
+	spin_lock(&imxdma->lock);
+
+	if (list_empty(&imxdmac->ld_active)) {
+		/* Someone might have called terminate all */
+		goto out;
+	}
+	desc = list_first_entry(&imxdmac->ld_active, struct imxdma_desc, node);
+
+	if (desc->desc.callback)
+		desc->desc.callback(desc->desc.callback_param);
+
+	dma_cookie_complete(&desc->desc);
+
+	/* If we are dealing with a cyclic descriptor keep it on ld_active */
+	if (imxdma_chan_is_doing_cyclic(imxdmac))
+		goto out;
+
+	/* Free 2D slot if it was an interleaved transfer */
+	if (imxdmac->enabled_2d) {
+		imxdma->slots_2d[imxdmac->slot_2d].count--;
+		imxdmac->enabled_2d = false;
+	}
+
+	list_move_tail(imxdmac->ld_active.next, &imxdmac->ld_free);
+
+	if (!list_empty(&imxdmac->ld_queue)) {
+		desc = list_first_entry(&imxdmac->ld_queue, struct imxdma_desc,
+					node);
+		list_move_tail(imxdmac->ld_queue.next, &imxdmac->ld_active);
+		if (imxdma_xfer_desc(desc) < 0)
+			dev_warn(imxdma->dev, "%s: channel: %d couldn't xfer desc\n",
+				 __func__, imxdmac->channel);
+	}
+out:
+	spin_unlock(&imxdma->lock);
 }
 
 static int imxdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
@@ -97,13 +602,18 @@
 {
 	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
 	struct dma_slave_config *dmaengine_cfg = (void *)arg;
-	int ret;
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	unsigned long flags;
 	unsigned int mode = 0;
 
 	switch (cmd) {
 	case DMA_TERMINATE_ALL:
-		imxdmac->status = DMA_ERROR;
-		imx_dma_disable(imxdmac->imxdma_channel);
+		imxdma_disable_hw(imxdmac);
+
+		spin_lock_irqsave(&imxdma->lock, flags);
+		list_splice_tail_init(&imxdmac->ld_active, &imxdmac->ld_free);
+		list_splice_tail_init(&imxdmac->ld_queue, &imxdmac->ld_free);
+		spin_unlock_irqrestore(&imxdma->lock, flags);
 		return 0;
 	case DMA_SLAVE_CONFIG:
 		if (dmaengine_cfg->direction == DMA_DEV_TO_MEM) {
@@ -128,16 +638,22 @@
 			mode = IMX_DMA_MEMSIZE_32;
 			break;
 		}
-		ret = imx_dma_config_channel(imxdmac->imxdma_channel,
-				mode | IMX_DMA_TYPE_FIFO,
-				IMX_DMA_MEMSIZE_32 | IMX_DMA_TYPE_LINEAR,
-				imxdmac->dma_request, 1);
 
-		if (ret)
-			return ret;
+		imxdmac->hw_chaining = 1;
+		if (!imxdma_hw_chain(imxdmac))
+			return -EINVAL;
+		imxdmac->ccr_from_device = (mode | IMX_DMA_TYPE_FIFO) |
+			((IMX_DMA_MEMSIZE_32 | IMX_DMA_TYPE_LINEAR) << 2) |
+			CCR_REN;
+		imxdmac->ccr_to_device =
+			(IMX_DMA_MEMSIZE_32 | IMX_DMA_TYPE_LINEAR) |
+			((mode | IMX_DMA_TYPE_FIFO) << 2) | CCR_REN;
+		imx_dmav1_writel(imxdma, imxdmac->dma_request,
+				 DMA_RSSR(imxdmac->channel));
 
-		imx_dma_config_burstlen(imxdmac->imxdma_channel,
-				imxdmac->watermark_level * imxdmac->word_size);
+		/* Set burst length */
+		imx_dmav1_writel(imxdma, imxdmac->watermark_level *
+				imxdmac->word_size, DMA_BLR(imxdmac->channel));
 
 		return 0;
 	default:
@@ -151,43 +667,20 @@
 					    dma_cookie_t cookie,
 					    struct dma_tx_state *txstate)
 {
-	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
-	dma_cookie_t last_used;
-	enum dma_status ret;
-
-	last_used = chan->cookie;
-
-	ret = dma_async_is_complete(cookie, imxdmac->last_completed, last_used);
-	dma_set_tx_state(txstate, imxdmac->last_completed, last_used, 0);
-
-	return ret;
-}
-
-static dma_cookie_t imxdma_assign_cookie(struct imxdma_channel *imxdma)
-{
-	dma_cookie_t cookie = imxdma->chan.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	imxdma->chan.cookie = cookie;
-	imxdma->desc.cookie = cookie;
-
-	return cookie;
+	return dma_cookie_status(chan, cookie, txstate);
 }
 
 static dma_cookie_t imxdma_tx_submit(struct dma_async_tx_descriptor *tx)
 {
 	struct imxdma_channel *imxdmac = to_imxdma_chan(tx->chan);
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
 	dma_cookie_t cookie;
+	unsigned long flags;
 
-	spin_lock_irq(&imxdmac->lock);
-
-	cookie = imxdma_assign_cookie(imxdmac);
-
-	imx_dma_enable(imxdmac->imxdma_channel);
-
-	spin_unlock_irq(&imxdmac->lock);
+	spin_lock_irqsave(&imxdma->lock, flags);
+	list_move_tail(imxdmac->ld_free.next, &imxdmac->ld_queue);
+	cookie = dma_cookie_assign(tx);
+	spin_unlock_irqrestore(&imxdma->lock, flags);
 
 	return cookie;
 }
@@ -197,23 +690,52 @@
 	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
 	struct imx_dma_data *data = chan->private;
 
-	imxdmac->dma_request = data->dma_request;
+	if (data != NULL)
+		imxdmac->dma_request = data->dma_request;
 
-	dma_async_tx_descriptor_init(&imxdmac->desc, chan);
-	imxdmac->desc.tx_submit = imxdma_tx_submit;
-	/* txd.flags will be overwritten in prep funcs */
-	imxdmac->desc.flags = DMA_CTRL_ACK;
+	while (imxdmac->descs_allocated < IMXDMA_MAX_CHAN_DESCRIPTORS) {
+		struct imxdma_desc *desc;
 
-	imxdmac->status = DMA_SUCCESS;
+		desc = kzalloc(sizeof(*desc), GFP_KERNEL);
+		if (!desc)
+			break;
+		__memzero(&desc->desc, sizeof(struct dma_async_tx_descriptor));
+		dma_async_tx_descriptor_init(&desc->desc, chan);
+		desc->desc.tx_submit = imxdma_tx_submit;
+		/* txd.flags will be overwritten in prep funcs */
+		desc->desc.flags = DMA_CTRL_ACK;
+		desc->status = DMA_SUCCESS;
 
-	return 0;
+		list_add_tail(&desc->node, &imxdmac->ld_free);
+		imxdmac->descs_allocated++;
+	}
+
+	if (!imxdmac->descs_allocated)
+		return -ENOMEM;
+
+	return imxdmac->descs_allocated;
 }
 
 static void imxdma_free_chan_resources(struct dma_chan *chan)
 {
 	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	struct imxdma_desc *desc, *_desc;
+	unsigned long flags;
 
-	imx_dma_disable(imxdmac->imxdma_channel);
+	spin_lock_irqsave(&imxdma->lock, flags);
+
+	imxdma_disable_hw(imxdmac);
+	list_splice_tail_init(&imxdmac->ld_active, &imxdmac->ld_free);
+	list_splice_tail_init(&imxdmac->ld_queue, &imxdmac->ld_free);
+
+	spin_unlock_irqrestore(&imxdma->lock, flags);
+
+	list_for_each_entry_safe(desc, _desc, &imxdmac->ld_free, node) {
+		kfree(desc);
+		imxdmac->descs_allocated--;
+	}
+	INIT_LIST_HEAD(&imxdmac->ld_free);
 
 	if (imxdmac->sg_list) {
 		kfree(imxdmac->sg_list);
@@ -224,27 +746,23 @@
 static struct dma_async_tx_descriptor *imxdma_prep_slave_sg(
 		struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned int sg_len, enum dma_transfer_direction direction,
-		unsigned long flags)
+		unsigned long flags, void *context)
 {
 	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
 	struct scatterlist *sg;
-	int i, ret, dma_length = 0;
-	unsigned int dmamode;
+	int i, dma_length = 0;
+	struct imxdma_desc *desc;
 
-	if (imxdmac->status == DMA_IN_PROGRESS)
+	if (list_empty(&imxdmac->ld_free) ||
+	    imxdma_chan_is_doing_cyclic(imxdmac))
 		return NULL;
 
-	imxdmac->status = DMA_IN_PROGRESS;
+	desc = list_first_entry(&imxdmac->ld_free, struct imxdma_desc, node);
 
 	for_each_sg(sgl, sg, sg_len, i) {
 		dma_length += sg->length;
 	}
 
-	if (direction == DMA_DEV_TO_MEM)
-		dmamode = DMA_MODE_READ;
-	else
-		dmamode = DMA_MODE_WRITE;
-
 	switch (imxdmac->word_size) {
 	case DMA_SLAVE_BUSWIDTH_4_BYTES:
 		if (sgl->length & 3 || sgl->dma_address & 3)
@@ -260,37 +778,41 @@
 		return NULL;
 	}
 
-	ret = imx_dma_setup_sg(imxdmac->imxdma_channel, sgl, sg_len,
-		 dma_length, imxdmac->per_address, dmamode);
-	if (ret)
-		return NULL;
+	desc->type = IMXDMA_DESC_SLAVE_SG;
+	desc->sg = sgl;
+	desc->sgcount = sg_len;
+	desc->len = dma_length;
+	desc->direction = direction;
+	if (direction == DMA_DEV_TO_MEM) {
+		desc->src = imxdmac->per_address;
+	} else {
+		desc->dest = imxdmac->per_address;
+	}
+	desc->desc.callback = NULL;
+	desc->desc.callback_param = NULL;
 
-	return &imxdmac->desc;
+	return &desc->desc;
 }
 
 static struct dma_async_tx_descriptor *imxdma_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len,
-		size_t period_len, enum dma_transfer_direction direction)
+		size_t period_len, enum dma_transfer_direction direction,
+		void *context)
 {
 	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
 	struct imxdma_engine *imxdma = imxdmac->imxdma;
-	int i, ret;
+	struct imxdma_desc *desc;
+	int i;
 	unsigned int periods = buf_len / period_len;
-	unsigned int dmamode;
 
 	dev_dbg(imxdma->dev, "%s channel: %d buf_len=%d period_len=%d\n",
 			__func__, imxdmac->channel, buf_len, period_len);
 
-	if (imxdmac->status == DMA_IN_PROGRESS)
+	if (list_empty(&imxdmac->ld_free) ||
+	    imxdma_chan_is_doing_cyclic(imxdmac))
 		return NULL;
-	imxdmac->status = DMA_IN_PROGRESS;
 
-	ret = imx_dma_setup_progression_handler(imxdmac->imxdma_channel,
-			imxdma_progression);
-	if (ret) {
-		dev_err(imxdma->dev, "Failed to setup the DMA handler\n");
-		return NULL;
-	}
+	desc = list_first_entry(&imxdmac->ld_free, struct imxdma_desc, node);
 
 	if (imxdmac->sg_list)
 		kfree(imxdmac->sg_list);
@@ -316,62 +838,221 @@
 	imxdmac->sg_list[periods].page_link =
 		((unsigned long)imxdmac->sg_list | 0x01) & ~0x02;
 
-	if (direction == DMA_DEV_TO_MEM)
-		dmamode = DMA_MODE_READ;
-	else
-		dmamode = DMA_MODE_WRITE;
+	desc->type = IMXDMA_DESC_CYCLIC;
+	desc->sg = imxdmac->sg_list;
+	desc->sgcount = periods;
+	desc->len = IMX_DMA_LENGTH_LOOP;
+	desc->direction = direction;
+	if (direction == DMA_DEV_TO_MEM) {
+		desc->src = imxdmac->per_address;
+	} else {
+		desc->dest = imxdmac->per_address;
+	}
+	desc->desc.callback = NULL;
+	desc->desc.callback_param = NULL;
 
-	ret = imx_dma_setup_sg(imxdmac->imxdma_channel, imxdmac->sg_list, periods,
-		 IMX_DMA_LENGTH_LOOP, imxdmac->per_address, dmamode);
-	if (ret)
+	return &desc->desc;
+}
+
+static struct dma_async_tx_descriptor *imxdma_prep_dma_memcpy(
+	struct dma_chan *chan, dma_addr_t dest,
+	dma_addr_t src, size_t len, unsigned long flags)
+{
+	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	struct imxdma_desc *desc;
+
+	dev_dbg(imxdma->dev, "%s channel: %d src=0x%x dst=0x%x len=%d\n",
+			__func__, imxdmac->channel, src, dest, len);
+
+	if (list_empty(&imxdmac->ld_free) ||
+	    imxdma_chan_is_doing_cyclic(imxdmac))
 		return NULL;
 
-	return &imxdmac->desc;
+	desc = list_first_entry(&imxdmac->ld_free, struct imxdma_desc, node);
+
+	desc->type = IMXDMA_DESC_MEMCPY;
+	desc->src = src;
+	desc->dest = dest;
+	desc->len = len;
+	desc->direction = DMA_MEM_TO_MEM;
+	desc->config_port = IMX_DMA_MEMSIZE_32 | IMX_DMA_TYPE_LINEAR;
+	desc->config_mem = IMX_DMA_MEMSIZE_32 | IMX_DMA_TYPE_LINEAR;
+	desc->desc.callback = NULL;
+	desc->desc.callback_param = NULL;
+
+	return &desc->desc;
+}
+
+static struct dma_async_tx_descriptor *imxdma_prep_dma_interleaved(
+	struct dma_chan *chan, struct dma_interleaved_template *xt,
+	unsigned long flags)
+{
+	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	struct imxdma_desc *desc;
+
+	dev_dbg(imxdma->dev, "%s channel: %d src_start=0x%x dst_start=0x%x\n"
+		"   src_sgl=%s dst_sgl=%s numf=%d frame_size=%d\n", __func__,
+		imxdmac->channel, xt->src_start, xt->dst_start,
+		xt->src_sgl ? "true" : "false", xt->dst_sgl ? "true" : "false",
+		xt->numf, xt->frame_size);
+
+	if (list_empty(&imxdmac->ld_free) ||
+	    imxdma_chan_is_doing_cyclic(imxdmac))
+		return NULL;
+
+	if (xt->frame_size != 1 || xt->numf <= 0 || xt->dir != DMA_MEM_TO_MEM)
+		return NULL;
+
+	desc = list_first_entry(&imxdmac->ld_free, struct imxdma_desc, node);
+
+	desc->type = IMXDMA_DESC_INTERLEAVED;
+	desc->src = xt->src_start;
+	desc->dest = xt->dst_start;
+	desc->x = xt->sgl[0].size;
+	desc->y = xt->numf;
+	desc->w = xt->sgl[0].icg + desc->x;
+	desc->len = desc->x * desc->y;
+	desc->direction = DMA_MEM_TO_MEM;
+	desc->config_port = IMX_DMA_MEMSIZE_32;
+	desc->config_mem = IMX_DMA_MEMSIZE_32;
+	if (xt->src_sgl)
+		desc->config_mem |= IMX_DMA_TYPE_2D;
+	if (xt->dst_sgl)
+		desc->config_port |= IMX_DMA_TYPE_2D;
+	desc->desc.callback = NULL;
+	desc->desc.callback_param = NULL;
+
+	return &desc->desc;
 }
 
 static void imxdma_issue_pending(struct dma_chan *chan)
 {
-	/*
-	 * Nothing to do. We only have a single descriptor
-	 */
+	struct imxdma_channel *imxdmac = to_imxdma_chan(chan);
+	struct imxdma_engine *imxdma = imxdmac->imxdma;
+	struct imxdma_desc *desc;
+	unsigned long flags;
+
+	spin_lock_irqsave(&imxdma->lock, flags);
+	if (list_empty(&imxdmac->ld_active) &&
+	    !list_empty(&imxdmac->ld_queue)) {
+		desc = list_first_entry(&imxdmac->ld_queue,
+					struct imxdma_desc, node);
+
+		if (imxdma_xfer_desc(desc) < 0) {
+			dev_warn(imxdma->dev,
+				 "%s: channel: %d couldn't issue DMA xfer\n",
+				 __func__, imxdmac->channel);
+		} else {
+			list_move_tail(imxdmac->ld_queue.next,
+				       &imxdmac->ld_active);
+		}
+	}
+	spin_unlock_irqrestore(&imxdma->lock, flags);
 }
 
 static int __init imxdma_probe(struct platform_device *pdev)
-{
+	{
 	struct imxdma_engine *imxdma;
 	int ret, i;
 
+
 	imxdma = kzalloc(sizeof(*imxdma), GFP_KERNEL);
 	if (!imxdma)
 		return -ENOMEM;
 
+	if (cpu_is_mx1()) {
+		imxdma->base = MX1_IO_ADDRESS(MX1_DMA_BASE_ADDR);
+	} else if (cpu_is_mx21()) {
+		imxdma->base = MX21_IO_ADDRESS(MX21_DMA_BASE_ADDR);
+	} else if (cpu_is_mx27()) {
+		imxdma->base = MX27_IO_ADDRESS(MX27_DMA_BASE_ADDR);
+	} else {
+		kfree(imxdma);
+		return 0;
+	}
+
+	imxdma->dma_clk = clk_get(NULL, "dma");
+	if (IS_ERR(imxdma->dma_clk))
+		return PTR_ERR(imxdma->dma_clk);
+	clk_enable(imxdma->dma_clk);
+
+	/* reset DMA module */
+	imx_dmav1_writel(imxdma, DCR_DRST, DMA_DCR);
+
+	if (cpu_is_mx1()) {
+		ret = request_irq(MX1_DMA_INT, dma_irq_handler, 0, "DMA", imxdma);
+		if (ret) {
+			dev_warn(imxdma->dev, "Can't register IRQ for DMA\n");
+			kfree(imxdma);
+			return ret;
+		}
+
+		ret = request_irq(MX1_DMA_ERR, imxdma_err_handler, 0, "DMA", imxdma);
+		if (ret) {
+			dev_warn(imxdma->dev, "Can't register ERRIRQ for DMA\n");
+			free_irq(MX1_DMA_INT, NULL);
+			kfree(imxdma);
+			return ret;
+		}
+	}
+
+	/* enable DMA module */
+	imx_dmav1_writel(imxdma, DCR_DEN, DMA_DCR);
+
+	/* clear all interrupts */
+	imx_dmav1_writel(imxdma, (1 << IMX_DMA_CHANNELS) - 1, DMA_DISR);
+
+	/* disable interrupts */
+	imx_dmav1_writel(imxdma, (1 << IMX_DMA_CHANNELS) - 1, DMA_DIMR);
+
 	INIT_LIST_HEAD(&imxdma->dma_device.channels);
 
 	dma_cap_set(DMA_SLAVE, imxdma->dma_device.cap_mask);
 	dma_cap_set(DMA_CYCLIC, imxdma->dma_device.cap_mask);
+	dma_cap_set(DMA_MEMCPY, imxdma->dma_device.cap_mask);
+	dma_cap_set(DMA_INTERLEAVE, imxdma->dma_device.cap_mask);
+
+	/* Initialize 2D global parameters */
+	for (i = 0; i < IMX_DMA_2D_SLOTS; i++)
+		imxdma->slots_2d[i].count = 0;
+
+	spin_lock_init(&imxdma->lock);
 
 	/* Initialize channel parameters */
-	for (i = 0; i < MAX_DMA_CHANNELS; i++) {
+	for (i = 0; i < IMX_DMA_CHANNELS; i++) {
 		struct imxdma_channel *imxdmac = &imxdma->channel[i];
 
-		imxdmac->imxdma_channel = imx_dma_request_by_prio("dmaengine",
-				DMA_PRIO_MEDIUM);
-		if ((int)imxdmac->channel < 0) {
-			ret = -ENODEV;
-			goto err_init;
+		if (cpu_is_mx21() || cpu_is_mx27()) {
+			ret = request_irq(MX2x_INT_DMACH0 + i,
+					dma_irq_handler, 0, "DMA", imxdma);
+			if (ret) {
+				dev_warn(imxdma->dev, "Can't register IRQ %d "
+					 "for DMA channel %d\n",
+					 MX2x_INT_DMACH0 + i, i);
+				goto err_init;
+			}
+			init_timer(&imxdmac->watchdog);
+			imxdmac->watchdog.function = &imxdma_watchdog;
+			imxdmac->watchdog.data = (unsigned long)imxdmac;
 		}
 
-		imx_dma_setup_handlers(imxdmac->imxdma_channel,
-		       imxdma_irq_handler, imxdma_err_handler, imxdmac);
-
 		imxdmac->imxdma = imxdma;
-		spin_lock_init(&imxdmac->lock);
 
+		INIT_LIST_HEAD(&imxdmac->ld_queue);
+		INIT_LIST_HEAD(&imxdmac->ld_free);
+		INIT_LIST_HEAD(&imxdmac->ld_active);
+
+		tasklet_init(&imxdmac->dma_tasklet, imxdma_tasklet,
+			     (unsigned long)imxdmac);
 		imxdmac->chan.device = &imxdma->dma_device;
+		dma_cookie_init(&imxdmac->chan);
 		imxdmac->channel = i;
 
 		/* Add the channel to the DMAC list */
-		list_add_tail(&imxdmac->chan.device_node, &imxdma->dma_device.channels);
+		list_add_tail(&imxdmac->chan.device_node,
+			      &imxdma->dma_device.channels);
 	}
 
 	imxdma->dev = &pdev->dev;
@@ -382,11 +1063,14 @@
 	imxdma->dma_device.device_tx_status = imxdma_tx_status;
 	imxdma->dma_device.device_prep_slave_sg = imxdma_prep_slave_sg;
 	imxdma->dma_device.device_prep_dma_cyclic = imxdma_prep_dma_cyclic;
+	imxdma->dma_device.device_prep_dma_memcpy = imxdma_prep_dma_memcpy;
+	imxdma->dma_device.device_prep_interleaved_dma = imxdma_prep_dma_interleaved;
 	imxdma->dma_device.device_control = imxdma_control;
 	imxdma->dma_device.device_issue_pending = imxdma_issue_pending;
 
 	platform_set_drvdata(pdev, imxdma);
 
+	imxdma->dma_device.copy_align = 2; /* 2^2 = 4 bytes alignment */
 	imxdma->dma_device.dev->dma_parms = &imxdma->dma_parms;
 	dma_set_max_seg_size(imxdma->dma_device.dev, 0xffffff);
 
@@ -399,9 +1083,13 @@
 	return 0;
 
 err_init:
-	while (--i >= 0) {
-		struct imxdma_channel *imxdmac = &imxdma->channel[i];
-		imx_dma_free(imxdmac->imxdma_channel);
+
+	if (cpu_is_mx21() || cpu_is_mx27()) {
+		while (--i >= 0)
+			free_irq(MX2x_INT_DMACH0 + i, NULL);
+	} else if cpu_is_mx1() {
+		free_irq(MX1_DMA_INT, NULL);
+		free_irq(MX1_DMA_ERR, NULL);
 	}
 
 	kfree(imxdma);
@@ -415,10 +1103,12 @@
 
         dma_async_device_unregister(&imxdma->dma_device);
 
-	for (i = 0; i < MAX_DMA_CHANNELS; i++) {
-		struct imxdma_channel *imxdmac = &imxdma->channel[i];
-
-		 imx_dma_free(imxdmac->imxdma_channel);
+	if (cpu_is_mx21() || cpu_is_mx27()) {
+		for (i = 0; i < IMX_DMA_CHANNELS; i++)
+			free_irq(MX2x_INT_DMACH0 + i, NULL);
+	} else if cpu_is_mx1() {
+		free_irq(MX1_DMA_INT, NULL);
+		free_irq(MX1_DMA_ERR, NULL);
 	}
 
         kfree(imxdma);
diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c
index 63540d3..d3e38e2 100644
--- a/drivers/dma/imx-sdma.c
+++ b/drivers/dma/imx-sdma.c
@@ -20,6 +20,7 @@
 #include <linux/init.h>
 #include <linux/module.h>
 #include <linux/types.h>
+#include <linux/bitops.h>
 #include <linux/mm.h>
 #include <linux/interrupt.h>
 #include <linux/clk.h>
@@ -41,6 +42,8 @@
 #include <mach/dma.h>
 #include <mach/hardware.h>
 
+#include "dmaengine.h"
+
 /* SDMA registers */
 #define SDMA_H_C0PTR		0x000
 #define SDMA_H_INTR		0x004
@@ -259,19 +262,18 @@
 	unsigned int			pc_from_device, pc_to_device;
 	unsigned long			flags;
 	dma_addr_t			per_address;
-	u32				event_mask0, event_mask1;
-	u32				watermark_level;
+	unsigned long			event_mask[2];
+	unsigned long			watermark_level;
 	u32				shp_addr, per_addr;
 	struct dma_chan			chan;
 	spinlock_t			lock;
 	struct dma_async_tx_descriptor	desc;
-	dma_cookie_t			last_completed;
 	enum dma_status			status;
 	unsigned int			chn_count;
 	unsigned int			chn_real_count;
 };
 
-#define IMX_DMA_SG_LOOP		(1 << 0)
+#define IMX_DMA_SG_LOOP		BIT(0)
 
 #define MAX_DMA_CHANNELS 32
 #define MXC_SDMA_DEFAULT_PRIORITY 1
@@ -345,9 +347,9 @@
 };
 MODULE_DEVICE_TABLE(of, sdma_dt_ids);
 
-#define SDMA_H_CONFIG_DSPDMA	(1 << 12) /* indicates if the DSPDMA is used */
-#define SDMA_H_CONFIG_RTD_PINS	(1 << 11) /* indicates if Real-Time Debug pins are enabled */
-#define SDMA_H_CONFIG_ACR	(1 << 4)  /* indicates if AHB freq /core freq = 2 or 1 */
+#define SDMA_H_CONFIG_DSPDMA	BIT(12) /* indicates if the DSPDMA is used */
+#define SDMA_H_CONFIG_RTD_PINS	BIT(11) /* indicates if Real-Time Debug pins are enabled */
+#define SDMA_H_CONFIG_ACR	BIT(4)  /* indicates if AHB freq /core freq = 2 or 1 */
 #define SDMA_H_CONFIG_CSM	(3)       /* indicates which context switch mode is selected*/
 
 static inline u32 chnenbl_ofs(struct sdma_engine *sdma, unsigned int event)
@@ -362,37 +364,42 @@
 {
 	struct sdma_engine *sdma = sdmac->sdma;
 	int channel = sdmac->channel;
-	u32 evt, mcu, dsp;
+	unsigned long evt, mcu, dsp;
 
 	if (event_override && mcu_override && dsp_override)
 		return -EINVAL;
 
-	evt = __raw_readl(sdma->regs + SDMA_H_EVTOVR);
-	mcu = __raw_readl(sdma->regs + SDMA_H_HOSTOVR);
-	dsp = __raw_readl(sdma->regs + SDMA_H_DSPOVR);
+	evt = readl_relaxed(sdma->regs + SDMA_H_EVTOVR);
+	mcu = readl_relaxed(sdma->regs + SDMA_H_HOSTOVR);
+	dsp = readl_relaxed(sdma->regs + SDMA_H_DSPOVR);
 
 	if (dsp_override)
-		dsp &= ~(1 << channel);
+		__clear_bit(channel, &dsp);
 	else
-		dsp |= (1 << channel);
+		__set_bit(channel, &dsp);
 
 	if (event_override)
-		evt &= ~(1 << channel);
+		__clear_bit(channel, &evt);
 	else
-		evt |= (1 << channel);
+		__set_bit(channel, &evt);
 
 	if (mcu_override)
-		mcu &= ~(1 << channel);
+		__clear_bit(channel, &mcu);
 	else
-		mcu |= (1 << channel);
+		__set_bit(channel, &mcu);
 
-	__raw_writel(evt, sdma->regs + SDMA_H_EVTOVR);
-	__raw_writel(mcu, sdma->regs + SDMA_H_HOSTOVR);
-	__raw_writel(dsp, sdma->regs + SDMA_H_DSPOVR);
+	writel_relaxed(evt, sdma->regs + SDMA_H_EVTOVR);
+	writel_relaxed(mcu, sdma->regs + SDMA_H_HOSTOVR);
+	writel_relaxed(dsp, sdma->regs + SDMA_H_DSPOVR);
 
 	return 0;
 }
 
+static void sdma_enable_channel(struct sdma_engine *sdma, int channel)
+{
+	writel(BIT(channel), sdma->regs + SDMA_H_START);
+}
+
 /*
  * sdma_run_channel - run a channel and wait till it's done
  */
@@ -404,7 +411,7 @@
 
 	init_completion(&sdmac->done);
 
-	__raw_writel(1 << channel, sdma->regs + SDMA_H_START);
+	sdma_enable_channel(sdma, channel);
 
 	ret = wait_for_completion_timeout(&sdmac->done, HZ);
 
@@ -451,12 +458,12 @@
 {
 	struct sdma_engine *sdma = sdmac->sdma;
 	int channel = sdmac->channel;
-	u32 val;
+	unsigned long val;
 	u32 chnenbl = chnenbl_ofs(sdma, event);
 
-	val = __raw_readl(sdma->regs + chnenbl);
-	val |= (1 << channel);
-	__raw_writel(val, sdma->regs + chnenbl);
+	val = readl_relaxed(sdma->regs + chnenbl);
+	__set_bit(channel, &val);
+	writel_relaxed(val, sdma->regs + chnenbl);
 }
 
 static void sdma_event_disable(struct sdma_channel *sdmac, unsigned int event)
@@ -464,11 +471,11 @@
 	struct sdma_engine *sdma = sdmac->sdma;
 	int channel = sdmac->channel;
 	u32 chnenbl = chnenbl_ofs(sdma, event);
-	u32 val;
+	unsigned long val;
 
-	val = __raw_readl(sdma->regs + chnenbl);
-	val &= ~(1 << channel);
-	__raw_writel(val, sdma->regs + chnenbl);
+	val = readl_relaxed(sdma->regs + chnenbl);
+	__clear_bit(channel, &val);
+	writel_relaxed(val, sdma->regs + chnenbl);
 }
 
 static void sdma_handle_channel_loop(struct sdma_channel *sdmac)
@@ -522,7 +529,7 @@
 	else
 		sdmac->status = DMA_SUCCESS;
 
-	sdmac->last_completed = sdmac->desc.cookie;
+	dma_cookie_complete(&sdmac->desc);
 	if (sdmac->desc.callback)
 		sdmac->desc.callback(sdmac->desc.callback_param);
 }
@@ -544,10 +551,10 @@
 static irqreturn_t sdma_int_handler(int irq, void *dev_id)
 {
 	struct sdma_engine *sdma = dev_id;
-	u32 stat;
+	unsigned long stat;
 
-	stat = __raw_readl(sdma->regs + SDMA_H_INTR);
-	__raw_writel(stat, sdma->regs + SDMA_H_INTR);
+	stat = readl_relaxed(sdma->regs + SDMA_H_INTR);
+	writel_relaxed(stat, sdma->regs + SDMA_H_INTR);
 
 	while (stat) {
 		int channel = fls(stat) - 1;
@@ -555,7 +562,7 @@
 
 		mxc_sdma_handle_channel(sdmac);
 
-		stat &= ~(1 << channel);
+		__clear_bit(channel, &stat);
 	}
 
 	return IRQ_HANDLED;
@@ -663,11 +670,11 @@
 		return load_address;
 
 	dev_dbg(sdma->dev, "load_address = %d\n", load_address);
-	dev_dbg(sdma->dev, "wml = 0x%08x\n", sdmac->watermark_level);
+	dev_dbg(sdma->dev, "wml = 0x%08x\n", (u32)sdmac->watermark_level);
 	dev_dbg(sdma->dev, "shp_addr = 0x%08x\n", sdmac->shp_addr);
 	dev_dbg(sdma->dev, "per_addr = 0x%08x\n", sdmac->per_addr);
-	dev_dbg(sdma->dev, "event_mask0 = 0x%08x\n", sdmac->event_mask0);
-	dev_dbg(sdma->dev, "event_mask1 = 0x%08x\n", sdmac->event_mask1);
+	dev_dbg(sdma->dev, "event_mask0 = 0x%08x\n", (u32)sdmac->event_mask[0]);
+	dev_dbg(sdma->dev, "event_mask1 = 0x%08x\n", (u32)sdmac->event_mask[1]);
 
 	mutex_lock(&sdma->channel_0_lock);
 
@@ -677,8 +684,8 @@
 	/* Send by context the event mask,base address for peripheral
 	 * and watermark level
 	 */
-	context->gReg[0] = sdmac->event_mask1;
-	context->gReg[1] = sdmac->event_mask0;
+	context->gReg[0] = sdmac->event_mask[1];
+	context->gReg[1] = sdmac->event_mask[0];
 	context->gReg[2] = sdmac->per_addr;
 	context->gReg[6] = sdmac->shp_addr;
 	context->gReg[7] = sdmac->watermark_level;
@@ -701,7 +708,7 @@
 	struct sdma_engine *sdma = sdmac->sdma;
 	int channel = sdmac->channel;
 
-	__raw_writel(1 << channel, sdma->regs + SDMA_H_STATSTOP);
+	writel_relaxed(BIT(channel), sdma->regs + SDMA_H_STATSTOP);
 	sdmac->status = DMA_ERROR;
 }
 
@@ -711,13 +718,13 @@
 
 	sdma_disable_channel(sdmac);
 
-	sdmac->event_mask0 = 0;
-	sdmac->event_mask1 = 0;
+	sdmac->event_mask[0] = 0;
+	sdmac->event_mask[1] = 0;
 	sdmac->shp_addr = 0;
 	sdmac->per_addr = 0;
 
 	if (sdmac->event_id0) {
-		if (sdmac->event_id0 > 32)
+		if (sdmac->event_id0 >= sdmac->sdma->num_events)
 			return -EINVAL;
 		sdma_event_enable(sdmac, sdmac->event_id0);
 	}
@@ -740,15 +747,14 @@
 			(sdmac->peripheral_type != IMX_DMATYPE_DSP)) {
 		/* Handle multiple event channels differently */
 		if (sdmac->event_id1) {
-			sdmac->event_mask1 = 1 << (sdmac->event_id1 % 32);
+			sdmac->event_mask[1] = BIT(sdmac->event_id1 % 32);
 			if (sdmac->event_id1 > 31)
-				sdmac->watermark_level |= 1 << 31;
-			sdmac->event_mask0 = 1 << (sdmac->event_id0 % 32);
+				__set_bit(31, &sdmac->watermark_level);
+			sdmac->event_mask[0] = BIT(sdmac->event_id0 % 32);
 			if (sdmac->event_id0 > 31)
-				sdmac->watermark_level |= 1 << 30;
+				__set_bit(30, &sdmac->watermark_level);
 		} else {
-			sdmac->event_mask0 = 1 << sdmac->event_id0;
-			sdmac->event_mask1 = 1 << (sdmac->event_id0 - 32);
+			__set_bit(sdmac->event_id0, sdmac->event_mask);
 		}
 		/* Watermark Level */
 		sdmac->watermark_level |= sdmac->watermark_level;
@@ -774,7 +780,7 @@
 		return -EINVAL;
 	}
 
-	__raw_writel(priority, sdma->regs + SDMA_CHNPRI_0 + 4 * channel);
+	writel_relaxed(priority, sdma->regs + SDMA_CHNPRI_0 + 4 * channel);
 
 	return 0;
 }
@@ -796,8 +802,6 @@
 	sdma->channel_control[channel].base_bd_ptr = sdmac->bd_phys;
 	sdma->channel_control[channel].current_bd_ptr = sdmac->bd_phys;
 
-	clk_enable(sdma->clk);
-
 	sdma_set_channel_priority(sdmac, MXC_SDMA_DEFAULT_PRIORITY);
 
 	init_completion(&sdmac->done);
@@ -810,24 +814,6 @@
 	return ret;
 }
 
-static void sdma_enable_channel(struct sdma_engine *sdma, int channel)
-{
-	__raw_writel(1 << channel, sdma->regs + SDMA_H_START);
-}
-
-static dma_cookie_t sdma_assign_cookie(struct sdma_channel *sdmac)
-{
-	dma_cookie_t cookie = sdmac->chan.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	sdmac->chan.cookie = cookie;
-	sdmac->desc.cookie = cookie;
-
-	return cookie;
-}
-
 static struct sdma_channel *to_sdma_chan(struct dma_chan *chan)
 {
 	return container_of(chan, struct sdma_channel, chan);
@@ -837,14 +823,11 @@
 {
 	unsigned long flags;
 	struct sdma_channel *sdmac = to_sdma_chan(tx->chan);
-	struct sdma_engine *sdma = sdmac->sdma;
 	dma_cookie_t cookie;
 
 	spin_lock_irqsave(&sdmac->lock, flags);
 
-	cookie = sdma_assign_cookie(sdmac);
-
-	sdma_enable_channel(sdma, sdmac->channel);
+	cookie = dma_cookie_assign(tx);
 
 	spin_unlock_irqrestore(&sdmac->lock, flags);
 
@@ -875,11 +858,14 @@
 
 	sdmac->peripheral_type = data->peripheral_type;
 	sdmac->event_id0 = data->dma_request;
-	ret = sdma_set_channel_priority(sdmac, prio);
+
+	clk_enable(sdmac->sdma->clk);
+
+	ret = sdma_request_channel(sdmac);
 	if (ret)
 		return ret;
 
-	ret = sdma_request_channel(sdmac);
+	ret = sdma_set_channel_priority(sdmac, prio);
 	if (ret)
 		return ret;
 
@@ -916,7 +902,7 @@
 static struct dma_async_tx_descriptor *sdma_prep_slave_sg(
 		struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned int sg_len, enum dma_transfer_direction direction,
-		unsigned long flags)
+		unsigned long flags, void *context)
 {
 	struct sdma_channel *sdmac = to_sdma_chan(chan);
 	struct sdma_engine *sdma = sdmac->sdma;
@@ -1014,7 +1000,8 @@
 
 static struct dma_async_tx_descriptor *sdma_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len,
-		size_t period_len, enum dma_transfer_direction direction)
+		size_t period_len, enum dma_transfer_direction direction,
+		void *context)
 {
 	struct sdma_channel *sdmac = to_sdma_chan(chan);
 	struct sdma_engine *sdma = sdmac->sdma;
@@ -1128,7 +1115,7 @@
 
 	last_used = chan->cookie;
 
-	dma_set_tx_state(txstate, sdmac->last_completed, last_used,
+	dma_set_tx_state(txstate, chan->completed_cookie, last_used,
 			sdmac->chn_count - sdmac->chn_real_count);
 
 	return sdmac->status;
@@ -1136,9 +1123,11 @@
 
 static void sdma_issue_pending(struct dma_chan *chan)
 {
-	/*
-	 * Nothing to do. We only have a single descriptor
-	 */
+	struct sdma_channel *sdmac = to_sdma_chan(chan);
+	struct sdma_engine *sdma = sdmac->sdma;
+
+	if (sdmac->status == DMA_IN_PROGRESS)
+		sdma_enable_channel(sdma, sdmac->channel);
 }
 
 #define SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V1	34
@@ -1230,7 +1219,7 @@
 	clk_enable(sdma->clk);
 
 	/* Be sure SDMA has not started yet */
-	__raw_writel(0, sdma->regs + SDMA_H_C0PTR);
+	writel_relaxed(0, sdma->regs + SDMA_H_C0PTR);
 
 	sdma->channel_control = dma_alloc_coherent(NULL,
 			MAX_DMA_CHANNELS * sizeof (struct sdma_channel_control) +
@@ -1253,11 +1242,11 @@
 
 	/* disable all channels */
 	for (i = 0; i < sdma->num_events; i++)
-		__raw_writel(0, sdma->regs + chnenbl_ofs(sdma, i));
+		writel_relaxed(0, sdma->regs + chnenbl_ofs(sdma, i));
 
 	/* All channels have priority 0 */
 	for (i = 0; i < MAX_DMA_CHANNELS; i++)
-		__raw_writel(0, sdma->regs + SDMA_CHNPRI_0 + i * 4);
+		writel_relaxed(0, sdma->regs + SDMA_CHNPRI_0 + i * 4);
 
 	ret = sdma_request_channel(&sdma->channel[0]);
 	if (ret)
@@ -1266,16 +1255,16 @@
 	sdma_config_ownership(&sdma->channel[0], false, true, false);
 
 	/* Set Command Channel (Channel Zero) */
-	__raw_writel(0x4050, sdma->regs + SDMA_CHN0ADDR);
+	writel_relaxed(0x4050, sdma->regs + SDMA_CHN0ADDR);
 
 	/* Set bits of CONFIG register but with static context switching */
 	/* FIXME: Check whether to set ACR bit depending on clock ratios */
-	__raw_writel(0, sdma->regs + SDMA_H_CONFIG);
+	writel_relaxed(0, sdma->regs + SDMA_H_CONFIG);
 
-	__raw_writel(ccb_phys, sdma->regs + SDMA_H_C0PTR);
+	writel_relaxed(ccb_phys, sdma->regs + SDMA_H_C0PTR);
 
 	/* Set bits of CONFIG register with given context switching mode */
-	__raw_writel(SDMA_H_CONFIG_CSM, sdma->regs + SDMA_H_CONFIG);
+	writel_relaxed(SDMA_H_CONFIG_CSM, sdma->regs + SDMA_H_CONFIG);
 
 	/* Initializes channel's priorities */
 	sdma_set_channel_priority(&sdma->channel[0], 7);
@@ -1367,6 +1356,7 @@
 		spin_lock_init(&sdmac->lock);
 
 		sdmac->chan.device = &sdma->dma_device;
+		dma_cookie_init(&sdmac->chan);
 		sdmac->channel = i;
 
 		/*
@@ -1387,7 +1377,9 @@
 		sdma_add_scripts(sdma, pdata->script_addrs);
 
 	if (pdata) {
-		sdma_get_firmware(sdma, pdata->fw_name);
+		ret = sdma_get_firmware(sdma, pdata->fw_name);
+		if (ret)
+			dev_warn(&pdev->dev, "failed to get firmware from platform data\n");
 	} else {
 		/*
 		 * Because that device tree does not encode ROM script address,
@@ -1396,15 +1388,12 @@
 		 */
 		ret = of_property_read_string(np, "fsl,sdma-ram-script-name",
 					      &fw_name);
-		if (ret) {
-			dev_err(&pdev->dev, "failed to get firmware name\n");
-			goto err_init;
-		}
-
-		ret = sdma_get_firmware(sdma, fw_name);
-		if (ret) {
-			dev_err(&pdev->dev, "failed to get firmware\n");
-			goto err_init;
+		if (ret)
+			dev_warn(&pdev->dev, "failed to get firmware name\n");
+		else {
+			ret = sdma_get_firmware(sdma, fw_name);
+			if (ret)
+				dev_warn(&pdev->dev, "failed to get firmware from device tree\n");
 		}
 	}
 
diff --git a/drivers/dma/intel_mid_dma.c b/drivers/dma/intel_mid_dma.c
index 74f70aa..c900ca7 100644
--- a/drivers/dma/intel_mid_dma.c
+++ b/drivers/dma/intel_mid_dma.c
@@ -29,6 +29,8 @@
 #include <linux/intel_mid_dma.h>
 #include <linux/module.h>
 
+#include "dmaengine.h"
+
 #define MAX_CHAN	4 /*max ch across controllers*/
 #include "intel_mid_dma_regs.h"
 
@@ -288,7 +290,7 @@
 	struct intel_mid_dma_lli	*llitem;
 	void *param_txd = NULL;
 
-	midc->completed = txd->cookie;
+	dma_cookie_complete(txd);
 	callback_txd = txd->callback;
 	param_txd = txd->callback_param;
 
@@ -434,14 +436,7 @@
 	dma_cookie_t		cookie;
 
 	spin_lock_bh(&midc->lock);
-	cookie = midc->chan.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	midc->chan.cookie = cookie;
-	desc->txd.cookie = cookie;
-
+	cookie = dma_cookie_assign(tx);
 
 	if (list_empty(&midc->active_list))
 		list_add_tail(&desc->desc_node, &midc->active_list);
@@ -482,31 +477,18 @@
 						dma_cookie_t cookie,
 						struct dma_tx_state *txstate)
 {
-	struct intel_mid_dma_chan	*midc = to_intel_mid_dma_chan(chan);
-	dma_cookie_t		last_used;
-	dma_cookie_t		last_complete;
-	int				ret;
+	struct intel_mid_dma_chan *midc = to_intel_mid_dma_chan(chan);
+	enum dma_status ret;
 
-	last_complete = midc->completed;
-	last_used = chan->cookie;
-
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
+	ret = dma_cookie_status(chan, cookie, txstate);
 	if (ret != DMA_SUCCESS) {
 		spin_lock_bh(&midc->lock);
 		midc_scan_descriptors(to_middma_device(chan->device), midc);
 		spin_unlock_bh(&midc->lock);
 
-		last_complete = midc->completed;
-		last_used = chan->cookie;
-
-		ret = dma_async_is_complete(cookie, last_complete, last_used);
+		ret = dma_cookie_status(chan, cookie, txstate);
 	}
 
-	if (txstate) {
-		txstate->last = last_complete;
-		txstate->used = last_used;
-		txstate->residue = 0;
-	}
 	return ret;
 }
 
@@ -732,13 +714,14 @@
  * @sg_len: length of sg txn
  * @direction: DMA transfer dirtn
  * @flags: DMA flags
+ * @context: transfer context (ignored)
  *
  * Prepares LLI based periphral transfer
  */
 static struct dma_async_tx_descriptor *intel_mid_dma_prep_slave_sg(
 			struct dma_chan *chan, struct scatterlist *sgl,
 			unsigned int sg_len, enum dma_transfer_direction direction,
-			unsigned long flags)
+			unsigned long flags, void *context)
 {
 	struct intel_mid_dma_chan *midc = NULL;
 	struct intel_mid_dma_slave *mids = NULL;
@@ -832,7 +815,6 @@
 		/*trying to free ch in use!!!!!*/
 		pr_err("ERR_MDMA: trying to free ch in use\n");
 	}
-	pm_runtime_put(&mid->pdev->dev);
 	spin_lock_bh(&midc->lock);
 	midc->descs_allocated = 0;
 	list_for_each_entry_safe(desc, _desc, &midc->active_list, desc_node) {
@@ -853,6 +835,7 @@
 	/* Disable CH interrupts */
 	iowrite32(MASK_INTR_REG(midc->ch_id), mid->dma_base + MASK_BLOCK);
 	iowrite32(MASK_INTR_REG(midc->ch_id), mid->dma_base + MASK_ERR);
+	pm_runtime_put(&mid->pdev->dev);
 }
 
 /**
@@ -886,7 +869,7 @@
 		pm_runtime_put(&mid->pdev->dev);
 		return -EIO;
 	}
-	midc->completed = chan->cookie = 1;
+	dma_cookie_init(chan);
 
 	spin_lock_bh(&midc->lock);
 	while (midc->descs_allocated < DESCS_PER_CHANNEL) {
@@ -1056,7 +1039,8 @@
 	}
 	err_status &= mid->intr_mask;
 	if (err_status) {
-		iowrite32(MASK_INTR_REG(err_status), mid->dma_base + MASK_ERR);
+		iowrite32((err_status << INT_MASK_WE),
+			  mid->dma_base + MASK_ERR);
 		call_tasklet = 1;
 	}
 	if (call_tasklet)
@@ -1118,7 +1102,7 @@
 		struct intel_mid_dma_chan *midch = &dma->ch[i];
 
 		midch->chan.device = &dma->common;
-		midch->chan.cookie =  1;
+		dma_cookie_init(&midch->chan);
 		midch->ch_id = dma->chan_base + i;
 		pr_debug("MDMA:Init CH %d, ID %d\n", i, midch->ch_id);
 
diff --git a/drivers/dma/intel_mid_dma_regs.h b/drivers/dma/intel_mid_dma_regs.h
index c83d35b..1bfa926 100644
--- a/drivers/dma/intel_mid_dma_regs.h
+++ b/drivers/dma/intel_mid_dma_regs.h
@@ -165,7 +165,6 @@
  * @dma_base: MMIO register space DMA engine base pointer
  * @ch_id: DMA channel id
  * @lock: channel spinlock
- * @completed: DMA cookie
  * @active_list: current active descriptors
  * @queue: current queued up descriptors
  * @free_list: current free descriptors
@@ -183,7 +182,6 @@
 	void __iomem		*dma_base;
 	int			ch_id;
 	spinlock_t		lock;
-	dma_cookie_t		completed;
 	struct list_head	active_list;
 	struct list_head	queue;
 	struct list_head	free_list;
diff --git a/drivers/dma/ioat/dma.c b/drivers/dma/ioat/dma.c
index a4d6cb0..31493d8 100644
--- a/drivers/dma/ioat/dma.c
+++ b/drivers/dma/ioat/dma.c
@@ -40,6 +40,8 @@
 #include "registers.h"
 #include "hw.h"
 
+#include "../dmaengine.h"
+
 int ioat_pending_level = 4;
 module_param(ioat_pending_level, int, 0644);
 MODULE_PARM_DESC(ioat_pending_level,
@@ -107,6 +109,7 @@
 	chan->reg_base = device->reg_base + (0x80 * (idx + 1));
 	spin_lock_init(&chan->cleanup_lock);
 	chan->common.device = dma;
+	dma_cookie_init(&chan->common);
 	list_add_tail(&chan->common.device_node, &dma->channels);
 	device->idx[idx] = chan;
 	init_timer(&chan->timer);
@@ -235,12 +238,7 @@
 
 	spin_lock_bh(&ioat->desc_lock);
 	/* cookie incr and addition to used_list must be atomic */
-	cookie = c->cookie;
-	cookie++;
-	if (cookie < 0)
-		cookie = 1;
-	c->cookie = cookie;
-	tx->cookie = cookie;
+	cookie = dma_cookie_assign(tx);
 	dev_dbg(to_dev(&ioat->base), "%s: cookie: %d\n", __func__, cookie);
 
 	/* write address into NextDescriptor field of last desc in chain */
@@ -603,8 +601,7 @@
 		 */
 		dump_desc_dbg(ioat, desc);
 		if (tx->cookie) {
-			chan->completed_cookie = tx->cookie;
-			tx->cookie = 0;
+			dma_cookie_complete(tx);
 			ioat_dma_unmap(chan, tx->flags, desc->len, desc->hw);
 			ioat->active -= desc->hw->tx_cnt;
 			if (tx->callback) {
@@ -733,13 +730,15 @@
 {
 	struct ioat_chan_common *chan = to_chan_common(c);
 	struct ioatdma_device *device = chan->device;
+	enum dma_status ret;
 
-	if (ioat_tx_status(c, cookie, txstate) == DMA_SUCCESS)
-		return DMA_SUCCESS;
+	ret = dma_cookie_status(c, cookie, txstate);
+	if (ret == DMA_SUCCESS)
+		return ret;
 
 	device->cleanup_fn((unsigned long) c);
 
-	return ioat_tx_status(c, cookie, txstate);
+	return dma_cookie_status(c, cookie, txstate);
 }
 
 static void ioat1_dma_start_null_desc(struct ioat_dma_chan *ioat)
diff --git a/drivers/dma/ioat/dma.h b/drivers/dma/ioat/dma.h
index 5216c8a..c7888bc 100644
--- a/drivers/dma/ioat/dma.h
+++ b/drivers/dma/ioat/dma.h
@@ -90,7 +90,6 @@
 	void __iomem *reg_base;
 	unsigned long last_completion;
 	spinlock_t cleanup_lock;
-	dma_cookie_t completed_cookie;
 	unsigned long state;
 	#define IOAT_COMPLETION_PENDING 0
 	#define IOAT_COMPLETION_ACK 1
@@ -143,28 +142,6 @@
 	return container_of(chan, struct ioat_dma_chan, base);
 }
 
-/**
- * ioat_tx_status - poll the status of an ioat transaction
- * @c: channel handle
- * @cookie: transaction identifier
- * @txstate: if set, updated with the transaction state
- */
-static inline enum dma_status
-ioat_tx_status(struct dma_chan *c, dma_cookie_t cookie,
-		 struct dma_tx_state *txstate)
-{
-	struct ioat_chan_common *chan = to_chan_common(c);
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
-
-	last_used = c->cookie;
-	last_complete = chan->completed_cookie;
-
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-
-	return dma_async_is_complete(cookie, last_complete, last_used);
-}
-
 /* wrapper around hardware descriptor format + additional software fields */
 
 /**
diff --git a/drivers/dma/ioat/dma_v2.c b/drivers/dma/ioat/dma_v2.c
index 5d65f83..e8e110f 100644
--- a/drivers/dma/ioat/dma_v2.c
+++ b/drivers/dma/ioat/dma_v2.c
@@ -41,6 +41,8 @@
 #include "registers.h"
 #include "hw.h"
 
+#include "../dmaengine.h"
+
 int ioat_ring_alloc_order = 8;
 module_param(ioat_ring_alloc_order, int, 0644);
 MODULE_PARM_DESC(ioat_ring_alloc_order,
@@ -147,8 +149,7 @@
 		dump_desc_dbg(ioat, desc);
 		if (tx->cookie) {
 			ioat_dma_unmap(chan, tx->flags, desc->len, desc->hw);
-			chan->completed_cookie = tx->cookie;
-			tx->cookie = 0;
+			dma_cookie_complete(tx);
 			if (tx->callback) {
 				tx->callback(tx->callback_param);
 				tx->callback = NULL;
@@ -398,13 +399,9 @@
 	struct dma_chan *c = tx->chan;
 	struct ioat2_dma_chan *ioat = to_ioat2_chan(c);
 	struct ioat_chan_common *chan = &ioat->base;
-	dma_cookie_t cookie = c->cookie;
+	dma_cookie_t cookie;
 
-	cookie++;
-	if (cookie < 0)
-		cookie = 1;
-	tx->cookie = cookie;
-	c->cookie = cookie;
+	cookie = dma_cookie_assign(tx);
 	dev_dbg(to_dev(&ioat->base), "%s: cookie: %d\n", __func__, cookie);
 
 	if (!test_and_set_bit(IOAT_COMPLETION_PENDING, &chan->state))
diff --git a/drivers/dma/ioat/dma_v3.c b/drivers/dma/ioat/dma_v3.c
index f519c93..2c4476c 100644
--- a/drivers/dma/ioat/dma_v3.c
+++ b/drivers/dma/ioat/dma_v3.c
@@ -61,6 +61,7 @@
 #include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
 #include <linux/prefetch.h>
+#include "../dmaengine.h"
 #include "registers.h"
 #include "hw.h"
 #include "dma.h"
@@ -277,9 +278,8 @@
 		dump_desc_dbg(ioat, desc);
 		tx = &desc->txd;
 		if (tx->cookie) {
-			chan->completed_cookie = tx->cookie;
+			dma_cookie_complete(tx);
 			ioat3_dma_unmap(ioat, desc, idx + i);
-			tx->cookie = 0;
 			if (tx->callback) {
 				tx->callback(tx->callback_param);
 				tx->callback = NULL;
@@ -411,13 +411,15 @@
 		struct dma_tx_state *txstate)
 {
 	struct ioat2_dma_chan *ioat = to_ioat2_chan(c);
+	enum dma_status ret;
 
-	if (ioat_tx_status(c, cookie, txstate) == DMA_SUCCESS)
-		return DMA_SUCCESS;
+	ret = dma_cookie_status(c, cookie, txstate);
+	if (ret == DMA_SUCCESS)
+		return ret;
 
 	ioat3_cleanup(ioat);
 
-	return ioat_tx_status(c, cookie, txstate);
+	return dma_cookie_status(c, cookie, txstate);
 }
 
 static struct dma_async_tx_descriptor *
diff --git a/drivers/dma/iop-adma.c b/drivers/dma/iop-adma.c
index faf88b7..da6c4c2 100644
--- a/drivers/dma/iop-adma.c
+++ b/drivers/dma/iop-adma.c
@@ -36,6 +36,8 @@
 
 #include <mach/adma.h>
 
+#include "dmaengine.h"
+
 #define to_iop_adma_chan(chan) container_of(chan, struct iop_adma_chan, common)
 #define to_iop_adma_device(dev) \
 	container_of(dev, struct iop_adma_device, common)
@@ -317,7 +319,7 @@
 	}
 
 	if (cookie > 0) {
-		iop_chan->completed_cookie = cookie;
+		iop_chan->common.completed_cookie = cookie;
 		pr_debug("\tcompleted cookie %d\n", cookie);
 	}
 }
@@ -438,18 +440,6 @@
 	return NULL;
 }
 
-static dma_cookie_t
-iop_desc_assign_cookie(struct iop_adma_chan *iop_chan,
-	struct iop_adma_desc_slot *desc)
-{
-	dma_cookie_t cookie = iop_chan->common.cookie;
-	cookie++;
-	if (cookie < 0)
-		cookie = 1;
-	iop_chan->common.cookie = desc->async_tx.cookie = cookie;
-	return cookie;
-}
-
 static void iop_adma_check_threshold(struct iop_adma_chan *iop_chan)
 {
 	dev_dbg(iop_chan->device->common.dev, "pending: %d\n",
@@ -477,7 +467,7 @@
 	slots_per_op = grp_start->slots_per_op;
 
 	spin_lock_bh(&iop_chan->lock);
-	cookie = iop_desc_assign_cookie(iop_chan, sw_desc);
+	cookie = dma_cookie_assign(tx);
 
 	old_chain_tail = list_entry(iop_chan->chain.prev,
 		struct iop_adma_desc_slot, chain_node);
@@ -904,24 +894,15 @@
 					struct dma_tx_state *txstate)
 {
 	struct iop_adma_chan *iop_chan = to_iop_adma_chan(chan);
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
-	enum dma_status ret;
+	int ret;
 
-	last_used = chan->cookie;
-	last_complete = iop_chan->completed_cookie;
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
+	ret = dma_cookie_status(chan, cookie, txstate);
 	if (ret == DMA_SUCCESS)
 		return ret;
 
 	iop_adma_slot_cleanup(iop_chan);
 
-	last_used = chan->cookie;
-	last_complete = iop_chan->completed_cookie;
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-
-	return dma_async_is_complete(cookie, last_complete, last_used);
+	return dma_cookie_status(chan, cookie, txstate);
 }
 
 static irqreturn_t iop_adma_eot_handler(int irq, void *data)
@@ -1565,6 +1546,7 @@
 	INIT_LIST_HEAD(&iop_chan->chain);
 	INIT_LIST_HEAD(&iop_chan->all_slots);
 	iop_chan->common.device = dma_dev;
+	dma_cookie_init(&iop_chan->common);
 	list_add_tail(&iop_chan->common.device_node, &dma_dev->channels);
 
 	if (dma_has_cap(DMA_MEMCPY, dma_dev->cap_mask)) {
@@ -1642,16 +1624,12 @@
 		iop_desc_set_dest_addr(grp_start, iop_chan, 0);
 		iop_desc_set_memcpy_src_addr(grp_start, 0);
 
-		cookie = iop_chan->common.cookie;
-		cookie++;
-		if (cookie <= 1)
-			cookie = 2;
+		cookie = dma_cookie_assign(&sw_desc->async_tx);
 
 		/* initialize the completed cookie to be less than
 		 * the most recently used cookie
 		 */
-		iop_chan->completed_cookie = cookie - 1;
-		iop_chan->common.cookie = sw_desc->async_tx.cookie = cookie;
+		iop_chan->common.completed_cookie = cookie - 1;
 
 		/* channel should not be busy */
 		BUG_ON(iop_chan_is_busy(iop_chan));
@@ -1699,16 +1677,12 @@
 		iop_desc_set_xor_src_addr(grp_start, 0, 0);
 		iop_desc_set_xor_src_addr(grp_start, 1, 0);
 
-		cookie = iop_chan->common.cookie;
-		cookie++;
-		if (cookie <= 1)
-			cookie = 2;
+		cookie = dma_cookie_assign(&sw_desc->async_tx);
 
 		/* initialize the completed cookie to be less than
 		 * the most recently used cookie
 		 */
-		iop_chan->completed_cookie = cookie - 1;
-		iop_chan->common.cookie = sw_desc->async_tx.cookie = cookie;
+		iop_chan->common.completed_cookie = cookie - 1;
 
 		/* channel should not be busy */
 		BUG_ON(iop_chan_is_busy(iop_chan));
diff --git a/drivers/dma/ipu/ipu_idmac.c b/drivers/dma/ipu/ipu_idmac.c
index 6212b16..62e3f8e 100644
--- a/drivers/dma/ipu/ipu_idmac.c
+++ b/drivers/dma/ipu/ipu_idmac.c
@@ -25,6 +25,7 @@
 
 #include <mach/ipu.h>
 
+#include "../dmaengine.h"
 #include "ipu_intern.h"
 
 #define FS_VF_IN_VALID	0x00000002
@@ -866,14 +867,7 @@
 
 	dev_dbg(dev, "Submitting sg %p\n", &desc->sg[0]);
 
-	cookie = ichan->dma_chan.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	/* from dmaengine.h: "last cookie value returned to client" */
-	ichan->dma_chan.cookie = cookie;
-	tx->cookie = cookie;
+	cookie = dma_cookie_assign(tx);
 
 	/* ipu->lock can be taken under ichan->lock, but not v.v. */
 	spin_lock_irqsave(&ichan->lock, flags);
@@ -1295,7 +1289,7 @@
 	/* Flip the active buffer - even if update above failed */
 	ichan->active_buffer = !ichan->active_buffer;
 	if (done)
-		ichan->completed = desc->txd.cookie;
+		dma_cookie_complete(&desc->txd);
 
 	callback = desc->txd.callback;
 	callback_param = desc->txd.callback_param;
@@ -1341,7 +1335,8 @@
 /* Allocate and initialise a transfer descriptor. */
 static struct dma_async_tx_descriptor *idmac_prep_slave_sg(struct dma_chan *chan,
 		struct scatterlist *sgl, unsigned int sg_len,
-		enum dma_transfer_direction direction, unsigned long tx_flags)
+		enum dma_transfer_direction direction, unsigned long tx_flags,
+		void *context)
 {
 	struct idmac_channel *ichan = to_idmac_chan(chan);
 	struct idmac_tx_desc *desc = NULL;
@@ -1510,8 +1505,7 @@
 	BUG_ON(chan->client_count > 1);
 	WARN_ON(ichan->status != IPU_CHANNEL_FREE);
 
-	chan->cookie		= 1;
-	ichan->completed	= -ENXIO;
+	dma_cookie_init(chan);
 
 	ret = ipu_irq_map(chan->chan_id);
 	if (ret < 0)
@@ -1600,9 +1594,7 @@
 static enum dma_status idmac_tx_status(struct dma_chan *chan,
 		       dma_cookie_t cookie, struct dma_tx_state *txstate)
 {
-	struct idmac_channel *ichan = to_idmac_chan(chan);
-
-	dma_set_tx_state(txstate, ichan->completed, chan->cookie, 0);
+	dma_set_tx_state(txstate, chan->completed_cookie, chan->cookie, 0);
 	if (cookie != chan->cookie)
 		return DMA_ERROR;
 	return DMA_SUCCESS;
@@ -1638,11 +1630,10 @@
 
 		ichan->status		= IPU_CHANNEL_FREE;
 		ichan->sec_chan_en	= false;
-		ichan->completed	= -ENXIO;
 		snprintf(ichan->eof_name, sizeof(ichan->eof_name), "IDMAC EOF %d", i);
 
 		dma_chan->device	= &idmac->dma;
-		dma_chan->cookie	= 1;
+		dma_cookie_init(dma_chan);
 		dma_chan->chan_id	= i;
 		list_add_tail(&dma_chan->device_node, &dma->channels);
 	}
diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c
index 4d6d4cf..2ab0a3d 100644
--- a/drivers/dma/mpc512x_dma.c
+++ b/drivers/dma/mpc512x_dma.c
@@ -44,6 +44,8 @@
 
 #include <linux/random.h>
 
+#include "dmaengine.h"
+
 /* Number of DMA Transfer descriptors allocated per channel */
 #define MPC_DMA_DESCRIPTORS	64
 
@@ -188,7 +190,6 @@
 	struct list_head		completed;
 	struct mpc_dma_tcd		*tcd;
 	dma_addr_t			tcd_paddr;
-	dma_cookie_t			completed_cookie;
 
 	/* Lock for this structure */
 	spinlock_t			lock;
@@ -365,7 +366,7 @@
 		/* Free descriptors */
 		spin_lock_irqsave(&mchan->lock, flags);
 		list_splice_tail_init(&list, &mchan->free);
-		mchan->completed_cookie = last_cookie;
+		mchan->chan.completed_cookie = last_cookie;
 		spin_unlock_irqrestore(&mchan->lock, flags);
 	}
 }
@@ -438,13 +439,7 @@
 		mpc_dma_execute(mchan);
 
 	/* Update cookie */
-	cookie = mchan->chan.cookie + 1;
-	if (cookie <= 0)
-		cookie = 1;
-
-	mchan->chan.cookie = cookie;
-	mdesc->desc.cookie = cookie;
-
+	cookie = dma_cookie_assign(txd);
 	spin_unlock_irqrestore(&mchan->lock, flags);
 
 	return cookie;
@@ -562,17 +557,14 @@
 	       struct dma_tx_state *txstate)
 {
 	struct mpc_dma_chan *mchan = dma_chan_to_mpc_dma_chan(chan);
+	enum dma_status ret;
 	unsigned long flags;
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
 
 	spin_lock_irqsave(&mchan->lock, flags);
-	last_used = mchan->chan.cookie;
-	last_complete = mchan->completed_cookie;
+	ret = dma_cookie_status(chan, cookie, txstate);
 	spin_unlock_irqrestore(&mchan->lock, flags);
 
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-	return dma_async_is_complete(cookie, last_complete, last_used);
+	return ret;
 }
 
 /* Prepare descriptor for memory to memory copy */
@@ -741,8 +733,7 @@
 		mchan = &mdma->channels[i];
 
 		mchan->chan.device = dma;
-		mchan->chan.cookie = 1;
-		mchan->completed_cookie = mchan->chan.cookie;
+		dma_cookie_init(&mchan->chan);
 
 		INIT_LIST_HEAD(&mchan->free);
 		INIT_LIST_HEAD(&mchan->prepared);
diff --git a/drivers/dma/mv_xor.c b/drivers/dma/mv_xor.c
index e779b43..fa5d55f 100644
--- a/drivers/dma/mv_xor.c
+++ b/drivers/dma/mv_xor.c
@@ -26,6 +26,8 @@
 #include <linux/platform_device.h>
 #include <linux/memory.h>
 #include <plat/mv_xor.h>
+
+#include "dmaengine.h"
 #include "mv_xor.h"
 
 static void mv_xor_issue_pending(struct dma_chan *chan);
@@ -435,7 +437,7 @@
 	}
 
 	if (cookie > 0)
-		mv_chan->completed_cookie = cookie;
+		mv_chan->common.completed_cookie = cookie;
 }
 
 static void
@@ -534,18 +536,6 @@
 	return NULL;
 }
 
-static dma_cookie_t
-mv_desc_assign_cookie(struct mv_xor_chan *mv_chan,
-		      struct mv_xor_desc_slot *desc)
-{
-	dma_cookie_t cookie = mv_chan->common.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-	mv_chan->common.cookie = desc->async_tx.cookie = cookie;
-	return cookie;
-}
-
 /************************ DMA engine API functions ****************************/
 static dma_cookie_t
 mv_xor_tx_submit(struct dma_async_tx_descriptor *tx)
@@ -563,7 +553,7 @@
 	grp_start = sw_desc->group_head;
 
 	spin_lock_bh(&mv_chan->lock);
-	cookie = mv_desc_assign_cookie(mv_chan, sw_desc);
+	cookie = dma_cookie_assign(tx);
 
 	if (list_empty(&mv_chan->chain))
 		list_splice_init(&sw_desc->tx_list, &mv_chan->chain);
@@ -820,27 +810,16 @@
 					  struct dma_tx_state *txstate)
 {
 	struct mv_xor_chan *mv_chan = to_mv_xor_chan(chan);
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
 	enum dma_status ret;
 
-	last_used = chan->cookie;
-	last_complete = mv_chan->completed_cookie;
-	mv_chan->is_complete_cookie = cookie;
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
+	ret = dma_cookie_status(chan, cookie, txstate);
 	if (ret == DMA_SUCCESS) {
 		mv_xor_clean_completed_slots(mv_chan);
 		return ret;
 	}
 	mv_xor_slot_cleanup(mv_chan);
 
-	last_used = chan->cookie;
-	last_complete = mv_chan->completed_cookie;
-
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-	return dma_async_is_complete(cookie, last_complete, last_used);
+	return dma_cookie_status(chan, cookie, txstate);
 }
 
 static void mv_dump_xor_regs(struct mv_xor_chan *chan)
@@ -1214,6 +1193,7 @@
 	INIT_LIST_HEAD(&mv_chan->completed_slots);
 	INIT_LIST_HEAD(&mv_chan->all_slots);
 	mv_chan->common.device = dma_dev;
+	dma_cookie_init(&mv_chan->common);
 
 	list_add_tail(&mv_chan->common.device_node, &dma_dev->channels);
 
diff --git a/drivers/dma/mv_xor.h b/drivers/dma/mv_xor.h
index 977b592..654876b 100644
--- a/drivers/dma/mv_xor.h
+++ b/drivers/dma/mv_xor.h
@@ -78,7 +78,6 @@
 /**
  * struct mv_xor_chan - internal representation of a XOR channel
  * @pending: allows batching of hardware operations
- * @completed_cookie: identifier for the most recently completed operation
  * @lock: serializes enqueue/dequeue operations to the descriptors pool
  * @mmr_base: memory mapped register base
  * @idx: the index of the xor channel
@@ -93,7 +92,6 @@
  */
 struct mv_xor_chan {
 	int			pending;
-	dma_cookie_t		completed_cookie;
 	spinlock_t		lock; /* protects the descriptor slot pool */
 	void __iomem		*mmr_base;
 	unsigned int		idx;
@@ -109,7 +107,6 @@
 #ifdef USE_TIMER
 	unsigned long		cleanup_time;
 	u32			current_on_last_cleanup;
-	dma_cookie_t		is_complete_cookie;
 #endif
 };
 
diff --git a/drivers/dma/mxs-dma.c b/drivers/dma/mxs-dma.c
index b06cd4c..65334c4 100644
--- a/drivers/dma/mxs-dma.c
+++ b/drivers/dma/mxs-dma.c
@@ -28,6 +28,8 @@
 #include <mach/dma.h>
 #include <mach/common.h>
 
+#include "dmaengine.h"
+
 /*
  * NOTE: The term "PIO" throughout the mxs-dma implementation means
  * PIO mode of mxs apbh-dma and apbx-dma.  With this working mode,
@@ -111,7 +113,6 @@
 	struct mxs_dma_ccw		*ccw;
 	dma_addr_t			ccw_phys;
 	int				desc_count;
-	dma_cookie_t			last_completed;
 	enum dma_status			status;
 	unsigned int			flags;
 #define MXS_DMA_SG_LOOP			(1 << 0)
@@ -193,19 +194,6 @@
 	mxs_chan->status = DMA_IN_PROGRESS;
 }
 
-static dma_cookie_t mxs_dma_assign_cookie(struct mxs_dma_chan *mxs_chan)
-{
-	dma_cookie_t cookie = mxs_chan->chan.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	mxs_chan->chan.cookie = cookie;
-	mxs_chan->desc.cookie = cookie;
-
-	return cookie;
-}
-
 static struct mxs_dma_chan *to_mxs_dma_chan(struct dma_chan *chan)
 {
 	return container_of(chan, struct mxs_dma_chan, chan);
@@ -217,7 +205,7 @@
 
 	mxs_dma_enable_chan(mxs_chan);
 
-	return mxs_dma_assign_cookie(mxs_chan);
+	return dma_cookie_assign(tx);
 }
 
 static void mxs_dma_tasklet(unsigned long data)
@@ -274,7 +262,7 @@
 		stat1 &= ~(1 << channel);
 
 		if (mxs_chan->status == DMA_SUCCESS)
-			mxs_chan->last_completed = mxs_chan->desc.cookie;
+			dma_cookie_complete(&mxs_chan->desc);
 
 		/* schedule tasklet on this channel */
 		tasklet_schedule(&mxs_chan->tasklet);
@@ -352,7 +340,7 @@
 static struct dma_async_tx_descriptor *mxs_dma_prep_slave_sg(
 		struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned int sg_len, enum dma_transfer_direction direction,
-		unsigned long append)
+		unsigned long append, void *context)
 {
 	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
@@ -447,7 +435,8 @@
 
 static struct dma_async_tx_descriptor *mxs_dma_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len,
-		size_t period_len, enum dma_transfer_direction direction)
+		size_t period_len, enum dma_transfer_direction direction,
+		void *context)
 {
 	struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan);
 	struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma;
@@ -538,7 +527,7 @@
 	dma_cookie_t last_used;
 
 	last_used = chan->cookie;
-	dma_set_tx_state(txstate, mxs_chan->last_completed, last_used, 0);
+	dma_set_tx_state(txstate, chan->completed_cookie, last_used, 0);
 
 	return mxs_chan->status;
 }
@@ -630,6 +619,7 @@
 
 		mxs_chan->mxs_dma = mxs_dma;
 		mxs_chan->chan.device = &mxs_dma->dma_device;
+		dma_cookie_init(&mxs_chan->chan);
 
 		tasklet_init(&mxs_chan->tasklet, mxs_dma_tasklet,
 			     (unsigned long) mxs_chan);
diff --git a/drivers/dma/pch_dma.c b/drivers/dma/pch_dma.c
index 823f581..65c0495 100644
--- a/drivers/dma/pch_dma.c
+++ b/drivers/dma/pch_dma.c
@@ -25,6 +25,8 @@
 #include <linux/module.h>
 #include <linux/pch_dma.h>
 
+#include "dmaengine.h"
+
 #define DRV_NAME "pch-dma"
 
 #define DMA_CTL0_DISABLE		0x0
@@ -105,7 +107,6 @@
 
 	spinlock_t		lock;
 
-	dma_cookie_t		completed_cookie;
 	struct list_head	active_list;
 	struct list_head	queue;
 	struct list_head	free_list;
@@ -416,20 +417,6 @@
 	}
 }
 
-static dma_cookie_t pdc_assign_cookie(struct pch_dma_chan *pd_chan,
-				      struct pch_dma_desc *desc)
-{
-	dma_cookie_t cookie = pd_chan->chan.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	pd_chan->chan.cookie = cookie;
-	desc->txd.cookie = cookie;
-
-	return cookie;
-}
-
 static dma_cookie_t pd_tx_submit(struct dma_async_tx_descriptor *txd)
 {
 	struct pch_dma_desc *desc = to_pd_desc(txd);
@@ -437,7 +424,7 @@
 	dma_cookie_t cookie;
 
 	spin_lock(&pd_chan->lock);
-	cookie = pdc_assign_cookie(pd_chan, desc);
+	cookie = dma_cookie_assign(txd);
 
 	if (list_empty(&pd_chan->active_list)) {
 		list_add_tail(&desc->desc_node, &pd_chan->active_list);
@@ -544,7 +531,7 @@
 	spin_lock_irq(&pd_chan->lock);
 	list_splice(&tmp_list, &pd_chan->free_list);
 	pd_chan->descs_allocated = i;
-	pd_chan->completed_cookie = chan->cookie = 1;
+	dma_cookie_init(chan);
 	spin_unlock_irq(&pd_chan->lock);
 
 	pdc_enable_irq(chan, 1);
@@ -578,19 +565,12 @@
 				    struct dma_tx_state *txstate)
 {
 	struct pch_dma_chan *pd_chan = to_pd_chan(chan);
-	dma_cookie_t last_used;
-	dma_cookie_t last_completed;
-	int ret;
+	enum dma_status ret;
 
 	spin_lock_irq(&pd_chan->lock);
-	last_completed = pd_chan->completed_cookie;
-	last_used = chan->cookie;
+	ret = dma_cookie_status(chan, cookie, txstate);
 	spin_unlock_irq(&pd_chan->lock);
 
-	ret = dma_async_is_complete(cookie, last_completed, last_used);
-
-	dma_set_tx_state(txstate, last_completed, last_used, 0);
-
 	return ret;
 }
 
@@ -607,7 +587,8 @@
 
 static struct dma_async_tx_descriptor *pd_prep_slave_sg(struct dma_chan *chan,
 			struct scatterlist *sgl, unsigned int sg_len,
-			enum dma_transfer_direction direction, unsigned long flags)
+			enum dma_transfer_direction direction, unsigned long flags,
+			void *context)
 {
 	struct pch_dma_chan *pd_chan = to_pd_chan(chan);
 	struct pch_dma_slave *pd_slave = chan->private;
@@ -932,7 +913,7 @@
 		struct pch_dma_chan *pd_chan = &pd->channels[i];
 
 		pd_chan->chan.device = &pd->dma;
-		pd_chan->chan.cookie = 1;
+		dma_cookie_init(&pd_chan->chan);
 
 		pd_chan->membase = &regs->desc[i];
 
diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c
index 16b66c8..282caf1 100644
--- a/drivers/dma/pl330.c
+++ b/drivers/dma/pl330.c
@@ -1,4 +1,6 @@
-/* linux/drivers/dma/pl330.c
+/*
+ * Copyright (c) 2012 Samsung Electronics Co., Ltd.
+ *		http://www.samsung.com
  *
  * Copyright (C) 2010 Samsung Electronics Co. Ltd.
  *	Jaswinder Singh <jassi.brar@samsung.com>
@@ -9,10 +11,15 @@
  * (at your option) any later version.
  */
 
+#include <linux/kernel.h>
 #include <linux/io.h>
 #include <linux/init.h>
 #include <linux/slab.h>
 #include <linux/module.h>
+#include <linux/string.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/dma-mapping.h>
 #include <linux/dmaengine.h>
 #include <linux/interrupt.h>
 #include <linux/amba/bus.h>
@@ -21,8 +28,497 @@
 #include <linux/scatterlist.h>
 #include <linux/of.h>
 
+#include "dmaengine.h"
+#define PL330_MAX_CHAN		8
+#define PL330_MAX_IRQS		32
+#define PL330_MAX_PERI		32
+
+enum pl330_srccachectrl {
+	SCCTRL0,	/* Noncacheable and nonbufferable */
+	SCCTRL1,	/* Bufferable only */
+	SCCTRL2,	/* Cacheable, but do not allocate */
+	SCCTRL3,	/* Cacheable and bufferable, but do not allocate */
+	SINVALID1,
+	SINVALID2,
+	SCCTRL6,	/* Cacheable write-through, allocate on reads only */
+	SCCTRL7,	/* Cacheable write-back, allocate on reads only */
+};
+
+enum pl330_dstcachectrl {
+	DCCTRL0,	/* Noncacheable and nonbufferable */
+	DCCTRL1,	/* Bufferable only */
+	DCCTRL2,	/* Cacheable, but do not allocate */
+	DCCTRL3,	/* Cacheable and bufferable, but do not allocate */
+	DINVALID1,	/* AWCACHE = 0x1000 */
+	DINVALID2,
+	DCCTRL6,	/* Cacheable write-through, allocate on writes only */
+	DCCTRL7,	/* Cacheable write-back, allocate on writes only */
+};
+
+enum pl330_byteswap {
+	SWAP_NO,
+	SWAP_2,
+	SWAP_4,
+	SWAP_8,
+	SWAP_16,
+};
+
+enum pl330_reqtype {
+	MEMTOMEM,
+	MEMTODEV,
+	DEVTOMEM,
+	DEVTODEV,
+};
+
+/* Register and Bit field Definitions */
+#define DS			0x0
+#define DS_ST_STOP		0x0
+#define DS_ST_EXEC		0x1
+#define DS_ST_CMISS		0x2
+#define DS_ST_UPDTPC		0x3
+#define DS_ST_WFE		0x4
+#define DS_ST_ATBRR		0x5
+#define DS_ST_QBUSY		0x6
+#define DS_ST_WFP		0x7
+#define DS_ST_KILL		0x8
+#define DS_ST_CMPLT		0x9
+#define DS_ST_FLTCMP		0xe
+#define DS_ST_FAULT		0xf
+
+#define DPC			0x4
+#define INTEN			0x20
+#define ES			0x24
+#define INTSTATUS		0x28
+#define INTCLR			0x2c
+#define FSM			0x30
+#define FSC			0x34
+#define FTM			0x38
+
+#define _FTC			0x40
+#define FTC(n)			(_FTC + (n)*0x4)
+
+#define _CS			0x100
+#define CS(n)			(_CS + (n)*0x8)
+#define CS_CNS			(1 << 21)
+
+#define _CPC			0x104
+#define CPC(n)			(_CPC + (n)*0x8)
+
+#define _SA			0x400
+#define SA(n)			(_SA + (n)*0x20)
+
+#define _DA			0x404
+#define DA(n)			(_DA + (n)*0x20)
+
+#define _CC			0x408
+#define CC(n)			(_CC + (n)*0x20)
+
+#define CC_SRCINC		(1 << 0)
+#define CC_DSTINC		(1 << 14)
+#define CC_SRCPRI		(1 << 8)
+#define CC_DSTPRI		(1 << 22)
+#define CC_SRCNS		(1 << 9)
+#define CC_DSTNS		(1 << 23)
+#define CC_SRCIA		(1 << 10)
+#define CC_DSTIA		(1 << 24)
+#define CC_SRCBRSTLEN_SHFT	4
+#define CC_DSTBRSTLEN_SHFT	18
+#define CC_SRCBRSTSIZE_SHFT	1
+#define CC_DSTBRSTSIZE_SHFT	15
+#define CC_SRCCCTRL_SHFT	11
+#define CC_SRCCCTRL_MASK	0x7
+#define CC_DSTCCTRL_SHFT	25
+#define CC_DRCCCTRL_MASK	0x7
+#define CC_SWAP_SHFT		28
+
+#define _LC0			0x40c
+#define LC0(n)			(_LC0 + (n)*0x20)
+
+#define _LC1			0x410
+#define LC1(n)			(_LC1 + (n)*0x20)
+
+#define DBGSTATUS		0xd00
+#define DBG_BUSY		(1 << 0)
+
+#define DBGCMD			0xd04
+#define DBGINST0		0xd08
+#define DBGINST1		0xd0c
+
+#define CR0			0xe00
+#define CR1			0xe04
+#define CR2			0xe08
+#define CR3			0xe0c
+#define CR4			0xe10
+#define CRD			0xe14
+
+#define PERIPH_ID		0xfe0
+#define PERIPH_REV_SHIFT	20
+#define PERIPH_REV_MASK		0xf
+#define PERIPH_REV_R0P0		0
+#define PERIPH_REV_R1P0		1
+#define PERIPH_REV_R1P1		2
+#define PCELL_ID		0xff0
+
+#define CR0_PERIPH_REQ_SET	(1 << 0)
+#define CR0_BOOT_EN_SET		(1 << 1)
+#define CR0_BOOT_MAN_NS		(1 << 2)
+#define CR0_NUM_CHANS_SHIFT	4
+#define CR0_NUM_CHANS_MASK	0x7
+#define CR0_NUM_PERIPH_SHIFT	12
+#define CR0_NUM_PERIPH_MASK	0x1f
+#define CR0_NUM_EVENTS_SHIFT	17
+#define CR0_NUM_EVENTS_MASK	0x1f
+
+#define CR1_ICACHE_LEN_SHIFT	0
+#define CR1_ICACHE_LEN_MASK	0x7
+#define CR1_NUM_ICACHELINES_SHIFT	4
+#define CR1_NUM_ICACHELINES_MASK	0xf
+
+#define CRD_DATA_WIDTH_SHIFT	0
+#define CRD_DATA_WIDTH_MASK	0x7
+#define CRD_WR_CAP_SHIFT	4
+#define CRD_WR_CAP_MASK		0x7
+#define CRD_WR_Q_DEP_SHIFT	8
+#define CRD_WR_Q_DEP_MASK	0xf
+#define CRD_RD_CAP_SHIFT	12
+#define CRD_RD_CAP_MASK		0x7
+#define CRD_RD_Q_DEP_SHIFT	16
+#define CRD_RD_Q_DEP_MASK	0xf
+#define CRD_DATA_BUFF_SHIFT	20
+#define CRD_DATA_BUFF_MASK	0x3ff
+
+#define PART			0x330
+#define DESIGNER		0x41
+#define REVISION		0x0
+#define INTEG_CFG		0x0
+#define PERIPH_ID_VAL		((PART << 0) | (DESIGNER << 12))
+
+#define PCELL_ID_VAL		0xb105f00d
+
+#define PL330_STATE_STOPPED		(1 << 0)
+#define PL330_STATE_EXECUTING		(1 << 1)
+#define PL330_STATE_WFE			(1 << 2)
+#define PL330_STATE_FAULTING		(1 << 3)
+#define PL330_STATE_COMPLETING		(1 << 4)
+#define PL330_STATE_WFP			(1 << 5)
+#define PL330_STATE_KILLING		(1 << 6)
+#define PL330_STATE_FAULT_COMPLETING	(1 << 7)
+#define PL330_STATE_CACHEMISS		(1 << 8)
+#define PL330_STATE_UPDTPC		(1 << 9)
+#define PL330_STATE_ATBARRIER		(1 << 10)
+#define PL330_STATE_QUEUEBUSY		(1 << 11)
+#define PL330_STATE_INVALID		(1 << 15)
+
+#define PL330_STABLE_STATES (PL330_STATE_STOPPED | PL330_STATE_EXECUTING \
+				| PL330_STATE_WFE | PL330_STATE_FAULTING)
+
+#define CMD_DMAADDH		0x54
+#define CMD_DMAEND		0x00
+#define CMD_DMAFLUSHP		0x35
+#define CMD_DMAGO		0xa0
+#define CMD_DMALD		0x04
+#define CMD_DMALDP		0x25
+#define CMD_DMALP		0x20
+#define CMD_DMALPEND		0x28
+#define CMD_DMAKILL		0x01
+#define CMD_DMAMOV		0xbc
+#define CMD_DMANOP		0x18
+#define CMD_DMARMB		0x12
+#define CMD_DMASEV		0x34
+#define CMD_DMAST		0x08
+#define CMD_DMASTP		0x29
+#define CMD_DMASTZ		0x0c
+#define CMD_DMAWFE		0x36
+#define CMD_DMAWFP		0x30
+#define CMD_DMAWMB		0x13
+
+#define SZ_DMAADDH		3
+#define SZ_DMAEND		1
+#define SZ_DMAFLUSHP		2
+#define SZ_DMALD		1
+#define SZ_DMALDP		2
+#define SZ_DMALP		2
+#define SZ_DMALPEND		2
+#define SZ_DMAKILL		1
+#define SZ_DMAMOV		6
+#define SZ_DMANOP		1
+#define SZ_DMARMB		1
+#define SZ_DMASEV		2
+#define SZ_DMAST		1
+#define SZ_DMASTP		2
+#define SZ_DMASTZ		1
+#define SZ_DMAWFE		2
+#define SZ_DMAWFP		2
+#define SZ_DMAWMB		1
+#define SZ_DMAGO		6
+
+#define BRST_LEN(ccr)		((((ccr) >> CC_SRCBRSTLEN_SHFT) & 0xf) + 1)
+#define BRST_SIZE(ccr)		(1 << (((ccr) >> CC_SRCBRSTSIZE_SHFT) & 0x7))
+
+#define BYTE_TO_BURST(b, ccr)	((b) / BRST_SIZE(ccr) / BRST_LEN(ccr))
+#define BURST_TO_BYTE(c, ccr)	((c) * BRST_SIZE(ccr) * BRST_LEN(ccr))
+
+/*
+ * With 256 bytes, we can do more than 2.5MB and 5MB xfers per req
+ * at 1byte/burst for P<->M and M<->M respectively.
+ * For typical scenario, at 1word/burst, 10MB and 20MB xfers per req
+ * should be enough for P<->M and M<->M respectively.
+ */
+#define MCODE_BUFF_PER_REQ	256
+
+/* If the _pl330_req is available to the client */
+#define IS_FREE(req)	(*((u8 *)((req)->mc_cpu)) == CMD_DMAEND)
+
+/* Use this _only_ to wait on transient states */
+#define UNTIL(t, s)	while (!(_state(t) & (s))) cpu_relax();
+
+#ifdef PL330_DEBUG_MCGEN
+static unsigned cmd_line;
+#define PL330_DBGCMD_DUMP(off, x...)	do { \
+						printk("%x:", cmd_line); \
+						printk(x); \
+						cmd_line += off; \
+					} while (0)
+#define PL330_DBGMC_START(addr)		(cmd_line = addr)
+#else
+#define PL330_DBGCMD_DUMP(off, x...)	do {} while (0)
+#define PL330_DBGMC_START(addr)		do {} while (0)
+#endif
+
+/* The number of default descriptors */
+
 #define NR_DEFAULT_DESC	16
 
+/* Populated by the PL330 core driver for DMA API driver's info */
+struct pl330_config {
+	u32	periph_id;
+	u32	pcell_id;
+#define DMAC_MODE_NS	(1 << 0)
+	unsigned int	mode;
+	unsigned int	data_bus_width:10; /* In number of bits */
+	unsigned int	data_buf_dep:10;
+	unsigned int	num_chan:4;
+	unsigned int	num_peri:6;
+	u32		peri_ns;
+	unsigned int	num_events:6;
+	u32		irq_ns;
+};
+
+/* Handle to the DMAC provided to the PL330 core */
+struct pl330_info {
+	/* Owning device */
+	struct device *dev;
+	/* Size of MicroCode buffers for each channel. */
+	unsigned mcbufsz;
+	/* ioremap'ed address of PL330 registers. */
+	void __iomem	*base;
+	/* Client can freely use it. */
+	void	*client_data;
+	/* PL330 core data, Client must not touch it. */
+	void	*pl330_data;
+	/* Populated by the PL330 core driver during pl330_add */
+	struct pl330_config	pcfg;
+	/*
+	 * If the DMAC has some reset mechanism, then the
+	 * client may want to provide pointer to the method.
+	 */
+	void (*dmac_reset)(struct pl330_info *pi);
+};
+
+/**
+ * Request Configuration.
+ * The PL330 core does not modify this and uses the last
+ * working configuration if the request doesn't provide any.
+ *
+ * The Client may want to provide this info only for the
+ * first request and a request with new settings.
+ */
+struct pl330_reqcfg {
+	/* Address Incrementing */
+	unsigned dst_inc:1;
+	unsigned src_inc:1;
+
+	/*
+	 * For now, the SRC & DST protection levels
+	 * and burst size/length are assumed same.
+	 */
+	bool nonsecure;
+	bool privileged;
+	bool insnaccess;
+	unsigned brst_len:5;
+	unsigned brst_size:3; /* in power of 2 */
+
+	enum pl330_dstcachectrl dcctl;
+	enum pl330_srccachectrl scctl;
+	enum pl330_byteswap swap;
+	struct pl330_config *pcfg;
+};
+
+/*
+ * One cycle of DMAC operation.
+ * There may be more than one xfer in a request.
+ */
+struct pl330_xfer {
+	u32 src_addr;
+	u32 dst_addr;
+	/* Size to xfer */
+	u32 bytes;
+	/*
+	 * Pointer to next xfer in the list.
+	 * The last xfer in the req must point to NULL.
+	 */
+	struct pl330_xfer *next;
+};
+
+/* The xfer callbacks are made with one of these arguments. */
+enum pl330_op_err {
+	/* The all xfers in the request were success. */
+	PL330_ERR_NONE,
+	/* If req aborted due to global error. */
+	PL330_ERR_ABORT,
+	/* If req failed due to problem with Channel. */
+	PL330_ERR_FAIL,
+};
+
+/* A request defining Scatter-Gather List ending with NULL xfer. */
+struct pl330_req {
+	enum pl330_reqtype rqtype;
+	/* Index of peripheral for the xfer. */
+	unsigned peri:5;
+	/* Unique token for this xfer, set by the client. */
+	void *token;
+	/* Callback to be called after xfer. */
+	void (*xfer_cb)(void *token, enum pl330_op_err err);
+	/* If NULL, req will be done at last set parameters. */
+	struct pl330_reqcfg *cfg;
+	/* Pointer to first xfer in the request. */
+	struct pl330_xfer *x;
+};
+
+/*
+ * To know the status of the channel and DMAC, the client
+ * provides a pointer to this structure. The PL330 core
+ * fills it with current information.
+ */
+struct pl330_chanstatus {
+	/*
+	 * If the DMAC engine halted due to some error,
+	 * the client should remove-add DMAC.
+	 */
+	bool dmac_halted;
+	/*
+	 * If channel is halted due to some error,
+	 * the client should ABORT/FLUSH and START the channel.
+	 */
+	bool faulting;
+	/* Location of last load */
+	u32 src_addr;
+	/* Location of last store */
+	u32 dst_addr;
+	/*
+	 * Pointer to the currently active req, NULL if channel is
+	 * inactive, even though the requests may be present.
+	 */
+	struct pl330_req *top_req;
+	/* Pointer to req waiting second in the queue if any. */
+	struct pl330_req *wait_req;
+};
+
+enum pl330_chan_op {
+	/* Start the channel */
+	PL330_OP_START,
+	/* Abort the active xfer */
+	PL330_OP_ABORT,
+	/* Stop xfer and flush queue */
+	PL330_OP_FLUSH,
+};
+
+struct _xfer_spec {
+	u32 ccr;
+	struct pl330_req *r;
+	struct pl330_xfer *x;
+};
+
+enum dmamov_dst {
+	SAR = 0,
+	CCR,
+	DAR,
+};
+
+enum pl330_dst {
+	SRC = 0,
+	DST,
+};
+
+enum pl330_cond {
+	SINGLE,
+	BURST,
+	ALWAYS,
+};
+
+struct _pl330_req {
+	u32 mc_bus;
+	void *mc_cpu;
+	/* Number of bytes taken to setup MC for the req */
+	u32 mc_len;
+	struct pl330_req *r;
+	/* Hook to attach to DMAC's list of reqs with due callback */
+	struct list_head rqd;
+};
+
+/* ToBeDone for tasklet */
+struct _pl330_tbd {
+	bool reset_dmac;
+	bool reset_mngr;
+	u8 reset_chan;
+};
+
+/* A DMAC Thread */
+struct pl330_thread {
+	u8 id;
+	int ev;
+	/* If the channel is not yet acquired by any client */
+	bool free;
+	/* Parent DMAC */
+	struct pl330_dmac *dmac;
+	/* Only two at a time */
+	struct _pl330_req req[2];
+	/* Index of the last enqueued request */
+	unsigned lstenq;
+	/* Index of the last submitted request or -1 if the DMA is stopped */
+	int req_running;
+};
+
+enum pl330_dmac_state {
+	UNINIT,
+	INIT,
+	DYING,
+};
+
+/* A DMAC */
+struct pl330_dmac {
+	spinlock_t		lock;
+	/* Holds list of reqs with due callbacks */
+	struct list_head	req_done;
+	/* Pointer to platform specific stuff */
+	struct pl330_info	*pinfo;
+	/* Maximum possible events/irqs */
+	int			events[32];
+	/* BUS address of MicroCode buffer */
+	u32			mcode_bus;
+	/* CPU address of MicroCode buffer */
+	void			*mcode_cpu;
+	/* List of all Channel threads */
+	struct pl330_thread	*channels;
+	/* Pointer to the MANAGER thread */
+	struct pl330_thread	*manager;
+	/* To handle bad news in interrupt */
+	struct tasklet_struct	tasks;
+	struct _pl330_tbd	dmac_tbd;
+	/* State of DMAC operation */
+	enum pl330_dmac_state	state;
+};
+
 enum desc_status {
 	/* In the DMAC pool */
 	FREE,
@@ -51,9 +547,6 @@
 	/* DMA-Engine Channel */
 	struct dma_chan chan;
 
-	/* Last completed cookie */
-	dma_cookie_t completed;
-
 	/* List of to be xfered descriptors */
 	struct list_head work_list;
 
@@ -117,6 +610,1599 @@
 	struct dma_pl330_chan *pchan;
 };
 
+static inline void _callback(struct pl330_req *r, enum pl330_op_err err)
+{
+	if (r && r->xfer_cb)
+		r->xfer_cb(r->token, err);
+}
+
+static inline bool _queue_empty(struct pl330_thread *thrd)
+{
+	return (IS_FREE(&thrd->req[0]) && IS_FREE(&thrd->req[1]))
+		? true : false;
+}
+
+static inline bool _queue_full(struct pl330_thread *thrd)
+{
+	return (IS_FREE(&thrd->req[0]) || IS_FREE(&thrd->req[1]))
+		? false : true;
+}
+
+static inline bool is_manager(struct pl330_thread *thrd)
+{
+	struct pl330_dmac *pl330 = thrd->dmac;
+
+	/* MANAGER is indexed at the end */
+	if (thrd->id == pl330->pinfo->pcfg.num_chan)
+		return true;
+	else
+		return false;
+}
+
+/* If manager of the thread is in Non-Secure mode */
+static inline bool _manager_ns(struct pl330_thread *thrd)
+{
+	struct pl330_dmac *pl330 = thrd->dmac;
+
+	return (pl330->pinfo->pcfg.mode & DMAC_MODE_NS) ? true : false;
+}
+
+static inline u32 get_id(struct pl330_info *pi, u32 off)
+{
+	void __iomem *regs = pi->base;
+	u32 id = 0;
+
+	id |= (readb(regs + off + 0x0) << 0);
+	id |= (readb(regs + off + 0x4) << 8);
+	id |= (readb(regs + off + 0x8) << 16);
+	id |= (readb(regs + off + 0xc) << 24);
+
+	return id;
+}
+
+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);
+	*((u16 *)&buf[1]) = val;
+
+	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)
+		return SZ_DMAEND;
+
+	buf[0] = CMD_DMAEND;
+
+	PL330_DBGCMD_DUMP(SZ_DMAEND, "\tDMAEND\n");
+
+	return SZ_DMAEND;
+}
+
+static inline u32 _emit_FLUSHP(unsigned dry_run, u8 buf[], u8 peri)
+{
+	if (dry_run)
+		return SZ_DMAFLUSHP;
+
+	buf[0] = CMD_DMAFLUSHP;
+
+	peri &= 0x1f;
+	peri <<= 3;
+	buf[1] = peri;
+
+	PL330_DBGCMD_DUMP(SZ_DMAFLUSHP, "\tDMAFLUSHP %u\n", peri >> 3);
+
+	return SZ_DMAFLUSHP;
+}
+
+static inline u32 _emit_LD(unsigned dry_run, u8 buf[],	enum pl330_cond cond)
+{
+	if (dry_run)
+		return SZ_DMALD;
+
+	buf[0] = CMD_DMALD;
+
+	if (cond == SINGLE)
+		buf[0] |= (0 << 1) | (1 << 0);
+	else if (cond == BURST)
+		buf[0] |= (1 << 1) | (1 << 0);
+
+	PL330_DBGCMD_DUMP(SZ_DMALD, "\tDMALD%c\n",
+		cond == SINGLE ? 'S' : (cond == BURST ? 'B' : 'A'));
+
+	return SZ_DMALD;
+}
+
+static inline u32 _emit_LDP(unsigned dry_run, u8 buf[],
+		enum pl330_cond cond, u8 peri)
+{
+	if (dry_run)
+		return SZ_DMALDP;
+
+	buf[0] = CMD_DMALDP;
+
+	if (cond == BURST)
+		buf[0] |= (1 << 1);
+
+	peri &= 0x1f;
+	peri <<= 3;
+	buf[1] = peri;
+
+	PL330_DBGCMD_DUMP(SZ_DMALDP, "\tDMALDP%c %u\n",
+		cond == SINGLE ? 'S' : 'B', peri >> 3);
+
+	return SZ_DMALDP;
+}
+
+static inline u32 _emit_LP(unsigned dry_run, u8 buf[],
+		unsigned loop, u8 cnt)
+{
+	if (dry_run)
+		return SZ_DMALP;
+
+	buf[0] = CMD_DMALP;
+
+	if (loop)
+		buf[0] |= (1 << 1);
+
+	cnt--; /* DMAC increments by 1 internally */
+	buf[1] = cnt;
+
+	PL330_DBGCMD_DUMP(SZ_DMALP, "\tDMALP_%c %u\n", loop ? '1' : '0', cnt);
+
+	return SZ_DMALP;
+}
+
+struct _arg_LPEND {
+	enum pl330_cond cond;
+	bool forever;
+	unsigned loop;
+	u8 bjump;
+};
+
+static inline u32 _emit_LPEND(unsigned dry_run, u8 buf[],
+		const struct _arg_LPEND *arg)
+{
+	enum pl330_cond cond = arg->cond;
+	bool forever = arg->forever;
+	unsigned loop = arg->loop;
+	u8 bjump = arg->bjump;
+
+	if (dry_run)
+		return SZ_DMALPEND;
+
+	buf[0] = CMD_DMALPEND;
+
+	if (loop)
+		buf[0] |= (1 << 2);
+
+	if (!forever)
+		buf[0] |= (1 << 4);
+
+	if (cond == SINGLE)
+		buf[0] |= (0 << 1) | (1 << 0);
+	else if (cond == BURST)
+		buf[0] |= (1 << 1) | (1 << 0);
+
+	buf[1] = bjump;
+
+	PL330_DBGCMD_DUMP(SZ_DMALPEND, "\tDMALP%s%c_%c bjmpto_%x\n",
+			forever ? "FE" : "END",
+			cond == SINGLE ? 'S' : (cond == BURST ? 'B' : 'A'),
+			loop ? '1' : '0',
+			bjump);
+
+	return SZ_DMALPEND;
+}
+
+static inline u32 _emit_KILL(unsigned dry_run, u8 buf[])
+{
+	if (dry_run)
+		return SZ_DMAKILL;
+
+	buf[0] = CMD_DMAKILL;
+
+	return SZ_DMAKILL;
+}
+
+static inline u32 _emit_MOV(unsigned dry_run, u8 buf[],
+		enum dmamov_dst dst, u32 val)
+{
+	if (dry_run)
+		return SZ_DMAMOV;
+
+	buf[0] = CMD_DMAMOV;
+	buf[1] = dst;
+	*((u32 *)&buf[2]) = val;
+
+	PL330_DBGCMD_DUMP(SZ_DMAMOV, "\tDMAMOV %s 0x%x\n",
+		dst == SAR ? "SAR" : (dst == DAR ? "DAR" : "CCR"), val);
+
+	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)
+		return SZ_DMARMB;
+
+	buf[0] = CMD_DMARMB;
+
+	PL330_DBGCMD_DUMP(SZ_DMARMB, "\tDMARMB\n");
+
+	return SZ_DMARMB;
+}
+
+static inline u32 _emit_SEV(unsigned dry_run, u8 buf[], u8 ev)
+{
+	if (dry_run)
+		return SZ_DMASEV;
+
+	buf[0] = CMD_DMASEV;
+
+	ev &= 0x1f;
+	ev <<= 3;
+	buf[1] = ev;
+
+	PL330_DBGCMD_DUMP(SZ_DMASEV, "\tDMASEV %u\n", ev >> 3);
+
+	return SZ_DMASEV;
+}
+
+static inline u32 _emit_ST(unsigned dry_run, u8 buf[], enum pl330_cond cond)
+{
+	if (dry_run)
+		return SZ_DMAST;
+
+	buf[0] = CMD_DMAST;
+
+	if (cond == SINGLE)
+		buf[0] |= (0 << 1) | (1 << 0);
+	else if (cond == BURST)
+		buf[0] |= (1 << 1) | (1 << 0);
+
+	PL330_DBGCMD_DUMP(SZ_DMAST, "\tDMAST%c\n",
+		cond == SINGLE ? 'S' : (cond == BURST ? 'B' : 'A'));
+
+	return SZ_DMAST;
+}
+
+static inline u32 _emit_STP(unsigned dry_run, u8 buf[],
+		enum pl330_cond cond, u8 peri)
+{
+	if (dry_run)
+		return SZ_DMASTP;
+
+	buf[0] = CMD_DMASTP;
+
+	if (cond == BURST)
+		buf[0] |= (1 << 1);
+
+	peri &= 0x1f;
+	peri <<= 3;
+	buf[1] = peri;
+
+	PL330_DBGCMD_DUMP(SZ_DMASTP, "\tDMASTP%c %u\n",
+		cond == SINGLE ? 'S' : 'B', peri >> 3);
+
+	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)
+{
+	if (dry_run)
+		return SZ_DMAWFP;
+
+	buf[0] = CMD_DMAWFP;
+
+	if (cond == SINGLE)
+		buf[0] |= (0 << 1) | (0 << 0);
+	else if (cond == BURST)
+		buf[0] |= (1 << 1) | (0 << 0);
+	else
+		buf[0] |= (0 << 1) | (1 << 0);
+
+	peri &= 0x1f;
+	peri <<= 3;
+	buf[1] = peri;
+
+	PL330_DBGCMD_DUMP(SZ_DMAWFP, "\tDMAWFP%c %u\n",
+		cond == SINGLE ? 'S' : (cond == BURST ? 'B' : 'P'), peri >> 3);
+
+	return SZ_DMAWFP;
+}
+
+static inline u32 _emit_WMB(unsigned dry_run, u8 buf[])
+{
+	if (dry_run)
+		return SZ_DMAWMB;
+
+	buf[0] = CMD_DMAWMB;
+
+	PL330_DBGCMD_DUMP(SZ_DMAWMB, "\tDMAWMB\n");
+
+	return SZ_DMAWMB;
+}
+
+struct _arg_GO {
+	u8 chan;
+	u32 addr;
+	unsigned ns;
+};
+
+static inline u32 _emit_GO(unsigned dry_run, u8 buf[],
+		const struct _arg_GO *arg)
+{
+	u8 chan = arg->chan;
+	u32 addr = arg->addr;
+	unsigned ns = arg->ns;
+
+	if (dry_run)
+		return SZ_DMAGO;
+
+	buf[0] = CMD_DMAGO;
+	buf[0] |= (ns << 1);
+
+	buf[1] = chan & 0x7;
+
+	*((u32 *)&buf[2]) = addr;
+
+	return SZ_DMAGO;
+}
+
+#define msecs_to_loops(t) (loops_per_jiffy / 1000 * HZ * t)
+
+/* Returns Time-Out */
+static bool _until_dmac_idle(struct pl330_thread *thrd)
+{
+	void __iomem *regs = thrd->dmac->pinfo->base;
+	unsigned long loops = msecs_to_loops(5);
+
+	do {
+		/* Until Manager is Idle */
+		if (!(readl(regs + DBGSTATUS) & DBG_BUSY))
+			break;
+
+		cpu_relax();
+	} while (--loops);
+
+	if (!loops)
+		return true;
+
+	return false;
+}
+
+static inline void _execute_DBGINSN(struct pl330_thread *thrd,
+		u8 insn[], bool as_manager)
+{
+	void __iomem *regs = thrd->dmac->pinfo->base;
+	u32 val;
+
+	val = (insn[0] << 16) | (insn[1] << 24);
+	if (!as_manager) {
+		val |= (1 << 0);
+		val |= (thrd->id << 8); /* Channel Number */
+	}
+	writel(val, regs + DBGINST0);
+
+	val = *((u32 *)&insn[2]);
+	writel(val, regs + DBGINST1);
+
+	/* If timed out due to halted state-machine */
+	if (_until_dmac_idle(thrd)) {
+		dev_err(thrd->dmac->pinfo->dev, "DMAC halted!\n");
+		return;
+	}
+
+	/* Get going */
+	writel(0, regs + DBGCMD);
+}
+
+/*
+ * Mark a _pl330_req as free.
+ * We do it by writing DMAEND as the first instruction
+ * because no valid request is going to have DMAEND as
+ * its first instruction to execute.
+ */
+static void mark_free(struct pl330_thread *thrd, int idx)
+{
+	struct _pl330_req *req = &thrd->req[idx];
+
+	_emit_END(0, req->mc_cpu);
+	req->mc_len = 0;
+
+	thrd->req_running = -1;
+}
+
+static inline u32 _state(struct pl330_thread *thrd)
+{
+	void __iomem *regs = thrd->dmac->pinfo->base;
+	u32 val;
+
+	if (is_manager(thrd))
+		val = readl(regs + DS) & 0xf;
+	else
+		val = readl(regs + CS(thrd->id)) & 0xf;
+
+	switch (val) {
+	case DS_ST_STOP:
+		return PL330_STATE_STOPPED;
+	case DS_ST_EXEC:
+		return PL330_STATE_EXECUTING;
+	case DS_ST_CMISS:
+		return PL330_STATE_CACHEMISS;
+	case DS_ST_UPDTPC:
+		return PL330_STATE_UPDTPC;
+	case DS_ST_WFE:
+		return PL330_STATE_WFE;
+	case DS_ST_FAULT:
+		return PL330_STATE_FAULTING;
+	case DS_ST_ATBRR:
+		if (is_manager(thrd))
+			return PL330_STATE_INVALID;
+		else
+			return PL330_STATE_ATBARRIER;
+	case DS_ST_QBUSY:
+		if (is_manager(thrd))
+			return PL330_STATE_INVALID;
+		else
+			return PL330_STATE_QUEUEBUSY;
+	case DS_ST_WFP:
+		if (is_manager(thrd))
+			return PL330_STATE_INVALID;
+		else
+			return PL330_STATE_WFP;
+	case DS_ST_KILL:
+		if (is_manager(thrd))
+			return PL330_STATE_INVALID;
+		else
+			return PL330_STATE_KILLING;
+	case DS_ST_CMPLT:
+		if (is_manager(thrd))
+			return PL330_STATE_INVALID;
+		else
+			return PL330_STATE_COMPLETING;
+	case DS_ST_FLTCMP:
+		if (is_manager(thrd))
+			return PL330_STATE_INVALID;
+		else
+			return PL330_STATE_FAULT_COMPLETING;
+	default:
+		return PL330_STATE_INVALID;
+	}
+}
+
+static void _stop(struct pl330_thread *thrd)
+{
+	void __iomem *regs = thrd->dmac->pinfo->base;
+	u8 insn[6] = {0, 0, 0, 0, 0, 0};
+
+	if (_state(thrd) == PL330_STATE_FAULT_COMPLETING)
+		UNTIL(thrd, PL330_STATE_FAULTING | PL330_STATE_KILLING);
+
+	/* Return if nothing needs to be done */
+	if (_state(thrd) == PL330_STATE_COMPLETING
+		  || _state(thrd) == PL330_STATE_KILLING
+		  || _state(thrd) == PL330_STATE_STOPPED)
+		return;
+
+	_emit_KILL(0, insn);
+
+	/* Stop generating interrupts for SEV */
+	writel(readl(regs + INTEN) & ~(1 << thrd->ev), regs + INTEN);
+
+	_execute_DBGINSN(thrd, insn, is_manager(thrd));
+}
+
+/* Start doing req 'idx' of thread 'thrd' */
+static bool _trigger(struct pl330_thread *thrd)
+{
+	void __iomem *regs = thrd->dmac->pinfo->base;
+	struct _pl330_req *req;
+	struct pl330_req *r;
+	struct _arg_GO go;
+	unsigned ns;
+	u8 insn[6] = {0, 0, 0, 0, 0, 0};
+	int idx;
+
+	/* Return if already ACTIVE */
+	if (_state(thrd) != PL330_STATE_STOPPED)
+		return true;
+
+	idx = 1 - thrd->lstenq;
+	if (!IS_FREE(&thrd->req[idx]))
+		req = &thrd->req[idx];
+	else {
+		idx = thrd->lstenq;
+		if (!IS_FREE(&thrd->req[idx]))
+			req = &thrd->req[idx];
+		else
+			req = NULL;
+	}
+
+	/* Return if no request */
+	if (!req || !req->r)
+		return true;
+
+	r = req->r;
+
+	if (r->cfg)
+		ns = r->cfg->nonsecure ? 1 : 0;
+	else if (readl(regs + CS(thrd->id)) & CS_CNS)
+		ns = 1;
+	else
+		ns = 0;
+
+	/* See 'Abort Sources' point-4 at Page 2-25 */
+	if (_manager_ns(thrd) && !ns)
+		dev_info(thrd->dmac->pinfo->dev, "%s:%d Recipe for ABORT!\n",
+			__func__, __LINE__);
+
+	go.chan = thrd->id;
+	go.addr = req->mc_bus;
+	go.ns = ns;
+	_emit_GO(0, insn, &go);
+
+	/* Set to generate interrupts for SEV */
+	writel(readl(regs + INTEN) | (1 << thrd->ev), regs + INTEN);
+
+	/* Only manager can execute GO */
+	_execute_DBGINSN(thrd, insn, true);
+
+	thrd->req_running = idx;
+
+	return true;
+}
+
+static bool _start(struct pl330_thread *thrd)
+{
+	switch (_state(thrd)) {
+	case PL330_STATE_FAULT_COMPLETING:
+		UNTIL(thrd, PL330_STATE_FAULTING | PL330_STATE_KILLING);
+
+		if (_state(thrd) == PL330_STATE_KILLING)
+			UNTIL(thrd, PL330_STATE_STOPPED)
+
+	case PL330_STATE_FAULTING:
+		_stop(thrd);
+
+	case PL330_STATE_KILLING:
+	case PL330_STATE_COMPLETING:
+		UNTIL(thrd, PL330_STATE_STOPPED)
+
+	case PL330_STATE_STOPPED:
+		return _trigger(thrd);
+
+	case PL330_STATE_WFP:
+	case PL330_STATE_QUEUEBUSY:
+	case PL330_STATE_ATBARRIER:
+	case PL330_STATE_UPDTPC:
+	case PL330_STATE_CACHEMISS:
+	case PL330_STATE_EXECUTING:
+		return true;
+
+	case PL330_STATE_WFE: /* For RESUME, nothing yet */
+	default:
+		return false;
+	}
+}
+
+static inline int _ldst_memtomem(unsigned dry_run, u8 buf[],
+		const struct _xfer_spec *pxs, int cyc)
+{
+	int off = 0;
+	struct pl330_config *pcfg = pxs->r->cfg->pcfg;
+
+	/* check lock-up free version */
+	if (get_revision(pcfg->periph_id) >= PERIPH_REV_R1P0) {
+		while (cyc--) {
+			off += _emit_LD(dry_run, &buf[off], ALWAYS);
+			off += _emit_ST(dry_run, &buf[off], ALWAYS);
+		}
+	} else {
+		while (cyc--) {
+			off += _emit_LD(dry_run, &buf[off], ALWAYS);
+			off += _emit_RMB(dry_run, &buf[off]);
+			off += _emit_ST(dry_run, &buf[off], ALWAYS);
+			off += _emit_WMB(dry_run, &buf[off]);
+		}
+	}
+
+	return off;
+}
+
+static inline int _ldst_devtomem(unsigned dry_run, u8 buf[],
+		const struct _xfer_spec *pxs, int cyc)
+{
+	int off = 0;
+
+	while (cyc--) {
+		off += _emit_WFP(dry_run, &buf[off], SINGLE, pxs->r->peri);
+		off += _emit_LDP(dry_run, &buf[off], SINGLE, pxs->r->peri);
+		off += _emit_ST(dry_run, &buf[off], ALWAYS);
+		off += _emit_FLUSHP(dry_run, &buf[off], pxs->r->peri);
+	}
+
+	return off;
+}
+
+static inline int _ldst_memtodev(unsigned dry_run, u8 buf[],
+		const struct _xfer_spec *pxs, int cyc)
+{
+	int off = 0;
+
+	while (cyc--) {
+		off += _emit_WFP(dry_run, &buf[off], SINGLE, pxs->r->peri);
+		off += _emit_LD(dry_run, &buf[off], ALWAYS);
+		off += _emit_STP(dry_run, &buf[off], SINGLE, pxs->r->peri);
+		off += _emit_FLUSHP(dry_run, &buf[off], pxs->r->peri);
+	}
+
+	return off;
+}
+
+static int _bursts(unsigned dry_run, u8 buf[],
+		const struct _xfer_spec *pxs, int cyc)
+{
+	int off = 0;
+
+	switch (pxs->r->rqtype) {
+	case MEMTODEV:
+		off += _ldst_memtodev(dry_run, &buf[off], pxs, cyc);
+		break;
+	case DEVTOMEM:
+		off += _ldst_devtomem(dry_run, &buf[off], pxs, cyc);
+		break;
+	case MEMTOMEM:
+		off += _ldst_memtomem(dry_run, &buf[off], pxs, cyc);
+		break;
+	default:
+		off += 0x40000000; /* Scare off the Client */
+		break;
+	}
+
+	return off;
+}
+
+/* Returns bytes consumed and updates bursts */
+static inline int _loop(unsigned dry_run, u8 buf[],
+		unsigned long *bursts, const struct _xfer_spec *pxs)
+{
+	int cyc, cycmax, szlp, szlpend, szbrst, off;
+	unsigned lcnt0, lcnt1, ljmp0, ljmp1;
+	struct _arg_LPEND lpend;
+
+	/* Max iterations possible in DMALP is 256 */
+	if (*bursts >= 256*256) {
+		lcnt1 = 256;
+		lcnt0 = 256;
+		cyc = *bursts / lcnt1 / lcnt0;
+	} else if (*bursts > 256) {
+		lcnt1 = 256;
+		lcnt0 = *bursts / lcnt1;
+		cyc = 1;
+	} else {
+		lcnt1 = *bursts;
+		lcnt0 = 0;
+		cyc = 1;
+	}
+
+	szlp = _emit_LP(1, buf, 0, 0);
+	szbrst = _bursts(1, buf, pxs, 1);
+
+	lpend.cond = ALWAYS;
+	lpend.forever = false;
+	lpend.loop = 0;
+	lpend.bjump = 0;
+	szlpend = _emit_LPEND(1, buf, &lpend);
+
+	if (lcnt0) {
+		szlp *= 2;
+		szlpend *= 2;
+	}
+
+	/*
+	 * Max bursts that we can unroll due to limit on the
+	 * size of backward jump that can be encoded in DMALPEND
+	 * which is 8-bits and hence 255
+	 */
+	cycmax = (255 - (szlp + szlpend)) / szbrst;
+
+	cyc = (cycmax < cyc) ? cycmax : cyc;
+
+	off = 0;
+
+	if (lcnt0) {
+		off += _emit_LP(dry_run, &buf[off], 0, lcnt0);
+		ljmp0 = off;
+	}
+
+	off += _emit_LP(dry_run, &buf[off], 1, lcnt1);
+	ljmp1 = off;
+
+	off += _bursts(dry_run, &buf[off], pxs, cyc);
+
+	lpend.cond = ALWAYS;
+	lpend.forever = false;
+	lpend.loop = 1;
+	lpend.bjump = off - ljmp1;
+	off += _emit_LPEND(dry_run, &buf[off], &lpend);
+
+	if (lcnt0) {
+		lpend.cond = ALWAYS;
+		lpend.forever = false;
+		lpend.loop = 0;
+		lpend.bjump = off - ljmp0;
+		off += _emit_LPEND(dry_run, &buf[off], &lpend);
+	}
+
+	*bursts = lcnt1 * cyc;
+	if (lcnt0)
+		*bursts *= lcnt0;
+
+	return off;
+}
+
+static inline int _setup_loops(unsigned dry_run, u8 buf[],
+		const struct _xfer_spec *pxs)
+{
+	struct pl330_xfer *x = pxs->x;
+	u32 ccr = pxs->ccr;
+	unsigned long c, bursts = BYTE_TO_BURST(x->bytes, ccr);
+	int off = 0;
+
+	while (bursts) {
+		c = bursts;
+		off += _loop(dry_run, &buf[off], &c, pxs);
+		bursts -= c;
+	}
+
+	return off;
+}
+
+static inline int _setup_xfer(unsigned dry_run, u8 buf[],
+		const struct _xfer_spec *pxs)
+{
+	struct pl330_xfer *x = pxs->x;
+	int off = 0;
+
+	/* DMAMOV SAR, x->src_addr */
+	off += _emit_MOV(dry_run, &buf[off], SAR, x->src_addr);
+	/* DMAMOV DAR, x->dst_addr */
+	off += _emit_MOV(dry_run, &buf[off], DAR, x->dst_addr);
+
+	/* Setup Loop(s) */
+	off += _setup_loops(dry_run, &buf[off], pxs);
+
+	return off;
+}
+
+/*
+ * A req is a sequence of one or more xfer units.
+ * Returns the number of bytes taken to setup the MC for the req.
+ */
+static int _setup_req(unsigned dry_run, struct pl330_thread *thrd,
+		unsigned index, struct _xfer_spec *pxs)
+{
+	struct _pl330_req *req = &thrd->req[index];
+	struct pl330_xfer *x;
+	u8 *buf = req->mc_cpu;
+	int off = 0;
+
+	PL330_DBGMC_START(req->mc_bus);
+
+	/* DMAMOV CCR, ccr */
+	off += _emit_MOV(dry_run, &buf[off], CCR, pxs->ccr);
+
+	x = pxs->r->x;
+	do {
+		/* Error if xfer length is not aligned at burst size */
+		if (x->bytes % (BRST_SIZE(pxs->ccr) * BRST_LEN(pxs->ccr)))
+			return -EINVAL;
+
+		pxs->x = x;
+		off += _setup_xfer(dry_run, &buf[off], pxs);
+
+		x = x->next;
+	} while (x);
+
+	/* DMASEV peripheral/event */
+	off += _emit_SEV(dry_run, &buf[off], thrd->ev);
+	/* DMAEND */
+	off += _emit_END(dry_run, &buf[off]);
+
+	return off;
+}
+
+static inline u32 _prepare_ccr(const struct pl330_reqcfg *rqc)
+{
+	u32 ccr = 0;
+
+	if (rqc->src_inc)
+		ccr |= CC_SRCINC;
+
+	if (rqc->dst_inc)
+		ccr |= CC_DSTINC;
+
+	/* We set same protection levels for Src and DST for now */
+	if (rqc->privileged)
+		ccr |= CC_SRCPRI | CC_DSTPRI;
+	if (rqc->nonsecure)
+		ccr |= CC_SRCNS | CC_DSTNS;
+	if (rqc->insnaccess)
+		ccr |= CC_SRCIA | CC_DSTIA;
+
+	ccr |= (((rqc->brst_len - 1) & 0xf) << CC_SRCBRSTLEN_SHFT);
+	ccr |= (((rqc->brst_len - 1) & 0xf) << CC_DSTBRSTLEN_SHFT);
+
+	ccr |= (rqc->brst_size << CC_SRCBRSTSIZE_SHFT);
+	ccr |= (rqc->brst_size << CC_DSTBRSTSIZE_SHFT);
+
+	ccr |= (rqc->scctl << CC_SRCCCTRL_SHFT);
+	ccr |= (rqc->dcctl << CC_DSTCCTRL_SHFT);
+
+	ccr |= (rqc->swap << CC_SWAP_SHFT);
+
+	return ccr;
+}
+
+static inline bool _is_valid(u32 ccr)
+{
+	enum pl330_dstcachectrl dcctl;
+	enum pl330_srccachectrl scctl;
+
+	dcctl = (ccr >> CC_DSTCCTRL_SHFT) & CC_DRCCCTRL_MASK;
+	scctl = (ccr >> CC_SRCCCTRL_SHFT) & CC_SRCCCTRL_MASK;
+
+	if (dcctl == DINVALID1 || dcctl == DINVALID2
+			|| scctl == SINVALID1 || scctl == SINVALID2)
+		return false;
+	else
+		return true;
+}
+
+/*
+ * Submit a list of xfers after which the client wants notification.
+ * Client is not notified after each xfer unit, just once after all
+ * xfer units are done or some error occurs.
+ */
+static int pl330_submit_req(void *ch_id, struct pl330_req *r)
+{
+	struct pl330_thread *thrd = ch_id;
+	struct pl330_dmac *pl330;
+	struct pl330_info *pi;
+	struct _xfer_spec xs;
+	unsigned long flags;
+	void __iomem *regs;
+	unsigned idx;
+	u32 ccr;
+	int ret = 0;
+
+	/* No Req or Unacquired Channel or DMAC */
+	if (!r || !thrd || thrd->free)
+		return -EINVAL;
+
+	pl330 = thrd->dmac;
+	pi = pl330->pinfo;
+	regs = pi->base;
+
+	if (pl330->state == DYING
+		|| pl330->dmac_tbd.reset_chan & (1 << thrd->id)) {
+		dev_info(thrd->dmac->pinfo->dev, "%s:%d\n",
+			__func__, __LINE__);
+		return -EAGAIN;
+	}
+
+	/* If request for non-existing peripheral */
+	if (r->rqtype != MEMTOMEM && r->peri >= pi->pcfg.num_peri) {
+		dev_info(thrd->dmac->pinfo->dev,
+				"%s:%d Invalid peripheral(%u)!\n",
+				__func__, __LINE__, r->peri);
+		return -EINVAL;
+	}
+
+	spin_lock_irqsave(&pl330->lock, flags);
+
+	if (_queue_full(thrd)) {
+		ret = -EAGAIN;
+		goto xfer_exit;
+	}
+
+	/* Prefer Secure Channel */
+	if (!_manager_ns(thrd))
+		r->cfg->nonsecure = 0;
+	else
+		r->cfg->nonsecure = 1;
+
+	/* Use last settings, if not provided */
+	if (r->cfg)
+		ccr = _prepare_ccr(r->cfg);
+	else
+		ccr = readl(regs + CC(thrd->id));
+
+	/* If this req doesn't have valid xfer settings */
+	if (!_is_valid(ccr)) {
+		ret = -EINVAL;
+		dev_info(thrd->dmac->pinfo->dev, "%s:%d Invalid CCR(%x)!\n",
+			__func__, __LINE__, ccr);
+		goto xfer_exit;
+	}
+
+	idx = IS_FREE(&thrd->req[0]) ? 0 : 1;
+
+	xs.ccr = ccr;
+	xs.r = r;
+
+	/* First dry run to check if req is acceptable */
+	ret = _setup_req(1, thrd, idx, &xs);
+	if (ret < 0)
+		goto xfer_exit;
+
+	if (ret > pi->mcbufsz / 2) {
+		dev_info(thrd->dmac->pinfo->dev,
+			"%s:%d Trying increasing mcbufsz\n",
+				__func__, __LINE__);
+		ret = -ENOMEM;
+		goto xfer_exit;
+	}
+
+	/* Hook the request */
+	thrd->lstenq = idx;
+	thrd->req[idx].mc_len = _setup_req(0, thrd, idx, &xs);
+	thrd->req[idx].r = r;
+
+	ret = 0;
+
+xfer_exit:
+	spin_unlock_irqrestore(&pl330->lock, flags);
+
+	return ret;
+}
+
+static void pl330_dotask(unsigned long data)
+{
+	struct pl330_dmac *pl330 = (struct pl330_dmac *) data;
+	struct pl330_info *pi = pl330->pinfo;
+	unsigned long flags;
+	int i;
+
+	spin_lock_irqsave(&pl330->lock, flags);
+
+	/* The DMAC itself gone nuts */
+	if (pl330->dmac_tbd.reset_dmac) {
+		pl330->state = DYING;
+		/* Reset the manager too */
+		pl330->dmac_tbd.reset_mngr = true;
+		/* Clear the reset flag */
+		pl330->dmac_tbd.reset_dmac = false;
+	}
+
+	if (pl330->dmac_tbd.reset_mngr) {
+		_stop(pl330->manager);
+		/* Reset all channels */
+		pl330->dmac_tbd.reset_chan = (1 << pi->pcfg.num_chan) - 1;
+		/* Clear the reset flag */
+		pl330->dmac_tbd.reset_mngr = false;
+	}
+
+	for (i = 0; i < pi->pcfg.num_chan; i++) {
+
+		if (pl330->dmac_tbd.reset_chan & (1 << i)) {
+			struct pl330_thread *thrd = &pl330->channels[i];
+			void __iomem *regs = pi->base;
+			enum pl330_op_err err;
+
+			_stop(thrd);
+
+			if (readl(regs + FSC) & (1 << thrd->id))
+				err = PL330_ERR_FAIL;
+			else
+				err = PL330_ERR_ABORT;
+
+			spin_unlock_irqrestore(&pl330->lock, flags);
+
+			_callback(thrd->req[1 - thrd->lstenq].r, err);
+			_callback(thrd->req[thrd->lstenq].r, err);
+
+			spin_lock_irqsave(&pl330->lock, flags);
+
+			thrd->req[0].r = NULL;
+			thrd->req[1].r = NULL;
+			mark_free(thrd, 0);
+			mark_free(thrd, 1);
+
+			/* Clear the reset flag */
+			pl330->dmac_tbd.reset_chan &= ~(1 << i);
+		}
+	}
+
+	spin_unlock_irqrestore(&pl330->lock, flags);
+
+	return;
+}
+
+/* Returns 1 if state was updated, 0 otherwise */
+static int pl330_update(const struct pl330_info *pi)
+{
+	struct _pl330_req *rqdone;
+	struct pl330_dmac *pl330;
+	unsigned long flags;
+	void __iomem *regs;
+	u32 val;
+	int id, ev, ret = 0;
+
+	if (!pi || !pi->pl330_data)
+		return 0;
+
+	regs = pi->base;
+	pl330 = pi->pl330_data;
+
+	spin_lock_irqsave(&pl330->lock, flags);
+
+	val = readl(regs + FSM) & 0x1;
+	if (val)
+		pl330->dmac_tbd.reset_mngr = true;
+	else
+		pl330->dmac_tbd.reset_mngr = false;
+
+	val = readl(regs + FSC) & ((1 << pi->pcfg.num_chan) - 1);
+	pl330->dmac_tbd.reset_chan |= val;
+	if (val) {
+		int i = 0;
+		while (i < pi->pcfg.num_chan) {
+			if (val & (1 << i)) {
+				dev_info(pi->dev,
+					"Reset Channel-%d\t CS-%x FTC-%x\n",
+						i, readl(regs + CS(i)),
+						readl(regs + FTC(i)));
+				_stop(&pl330->channels[i]);
+			}
+			i++;
+		}
+	}
+
+	/* Check which event happened i.e, thread notified */
+	val = readl(regs + ES);
+	if (pi->pcfg.num_events < 32
+			&& val & ~((1 << pi->pcfg.num_events) - 1)) {
+		pl330->dmac_tbd.reset_dmac = true;
+		dev_err(pi->dev, "%s:%d Unexpected!\n", __func__, __LINE__);
+		ret = 1;
+		goto updt_exit;
+	}
+
+	for (ev = 0; ev < pi->pcfg.num_events; ev++) {
+		if (val & (1 << ev)) { /* Event occurred */
+			struct pl330_thread *thrd;
+			u32 inten = readl(regs + INTEN);
+			int active;
+
+			/* Clear the event */
+			if (inten & (1 << ev))
+				writel(1 << ev, regs + INTCLR);
+
+			ret = 1;
+
+			id = pl330->events[ev];
+
+			thrd = &pl330->channels[id];
+
+			active = thrd->req_running;
+			if (active == -1) /* Aborted */
+				continue;
+
+			rqdone = &thrd->req[active];
+			mark_free(thrd, active);
+
+			/* Get going again ASAP */
+			_start(thrd);
+
+			/* For now, just make a list of callbacks to be done */
+			list_add_tail(&rqdone->rqd, &pl330->req_done);
+		}
+	}
+
+	/* Now that we are in no hurry, do the callbacks */
+	while (!list_empty(&pl330->req_done)) {
+		struct pl330_req *r;
+
+		rqdone = container_of(pl330->req_done.next,
+					struct _pl330_req, rqd);
+
+		list_del_init(&rqdone->rqd);
+
+		/* Detach the req */
+		r = rqdone->r;
+		rqdone->r = NULL;
+
+		spin_unlock_irqrestore(&pl330->lock, flags);
+		_callback(r, PL330_ERR_NONE);
+		spin_lock_irqsave(&pl330->lock, flags);
+	}
+
+updt_exit:
+	spin_unlock_irqrestore(&pl330->lock, flags);
+
+	if (pl330->dmac_tbd.reset_dmac
+			|| pl330->dmac_tbd.reset_mngr
+			|| pl330->dmac_tbd.reset_chan) {
+		ret = 1;
+		tasklet_schedule(&pl330->tasks);
+	}
+
+	return ret;
+}
+
+static int pl330_chan_ctrl(void *ch_id, enum pl330_chan_op op)
+{
+	struct pl330_thread *thrd = ch_id;
+	struct pl330_dmac *pl330;
+	unsigned long flags;
+	int ret = 0, active;
+
+	if (!thrd || thrd->free || thrd->dmac->state == DYING)
+		return -EINVAL;
+
+	pl330 = thrd->dmac;
+	active = thrd->req_running;
+
+	spin_lock_irqsave(&pl330->lock, flags);
+
+	switch (op) {
+	case PL330_OP_FLUSH:
+		/* Make sure the channel is stopped */
+		_stop(thrd);
+
+		thrd->req[0].r = NULL;
+		thrd->req[1].r = NULL;
+		mark_free(thrd, 0);
+		mark_free(thrd, 1);
+		break;
+
+	case PL330_OP_ABORT:
+		/* Make sure the channel is stopped */
+		_stop(thrd);
+
+		/* ABORT is only for the active req */
+		if (active == -1)
+			break;
+
+		thrd->req[active].r = NULL;
+		mark_free(thrd, active);
+
+		/* Start the next */
+	case PL330_OP_START:
+		if ((active == -1) && !_start(thrd))
+			ret = -EIO;
+		break;
+
+	default:
+		ret = -EINVAL;
+	}
+
+	spin_unlock_irqrestore(&pl330->lock, flags);
+	return ret;
+}
+
+/* Reserve an event */
+static inline int _alloc_event(struct pl330_thread *thrd)
+{
+	struct pl330_dmac *pl330 = thrd->dmac;
+	struct pl330_info *pi = pl330->pinfo;
+	int ev;
+
+	for (ev = 0; ev < pi->pcfg.num_events; ev++)
+		if (pl330->events[ev] == -1) {
+			pl330->events[ev] = thrd->id;
+			return ev;
+		}
+
+	return -1;
+}
+
+static bool _chan_ns(const struct pl330_info *pi, int i)
+{
+	return pi->pcfg.irq_ns & (1 << i);
+}
+
+/* Upon success, returns IdentityToken for the
+ * allocated channel, NULL otherwise.
+ */
+static void *pl330_request_channel(const struct pl330_info *pi)
+{
+	struct pl330_thread *thrd = NULL;
+	struct pl330_dmac *pl330;
+	unsigned long flags;
+	int chans, i;
+
+	if (!pi || !pi->pl330_data)
+		return NULL;
+
+	pl330 = pi->pl330_data;
+
+	if (pl330->state == DYING)
+		return NULL;
+
+	chans = pi->pcfg.num_chan;
+
+	spin_lock_irqsave(&pl330->lock, flags);
+
+	for (i = 0; i < chans; i++) {
+		thrd = &pl330->channels[i];
+		if ((thrd->free) && (!_manager_ns(thrd) ||
+					_chan_ns(pi, i))) {
+			thrd->ev = _alloc_event(thrd);
+			if (thrd->ev >= 0) {
+				thrd->free = false;
+				thrd->lstenq = 1;
+				thrd->req[0].r = NULL;
+				mark_free(thrd, 0);
+				thrd->req[1].r = NULL;
+				mark_free(thrd, 1);
+				break;
+			}
+		}
+		thrd = NULL;
+	}
+
+	spin_unlock_irqrestore(&pl330->lock, flags);
+
+	return thrd;
+}
+
+/* Release an event */
+static inline void _free_event(struct pl330_thread *thrd, int ev)
+{
+	struct pl330_dmac *pl330 = thrd->dmac;
+	struct pl330_info *pi = pl330->pinfo;
+
+	/* If the event is valid and was held by the thread */
+	if (ev >= 0 && ev < pi->pcfg.num_events
+			&& pl330->events[ev] == thrd->id)
+		pl330->events[ev] = -1;
+}
+
+static void pl330_release_channel(void *ch_id)
+{
+	struct pl330_thread *thrd = ch_id;
+	struct pl330_dmac *pl330;
+	unsigned long flags;
+
+	if (!thrd || thrd->free)
+		return;
+
+	_stop(thrd);
+
+	_callback(thrd->req[1 - thrd->lstenq].r, PL330_ERR_ABORT);
+	_callback(thrd->req[thrd->lstenq].r, PL330_ERR_ABORT);
+
+	pl330 = thrd->dmac;
+
+	spin_lock_irqsave(&pl330->lock, flags);
+	_free_event(thrd, thrd->ev);
+	thrd->free = true;
+	spin_unlock_irqrestore(&pl330->lock, flags);
+}
+
+/* Initialize the structure for PL330 configuration, that can be used
+ * by the client driver the make best use of the DMAC
+ */
+static void read_dmac_config(struct pl330_info *pi)
+{
+	void __iomem *regs = pi->base;
+	u32 val;
+
+	val = readl(regs + CRD) >> CRD_DATA_WIDTH_SHIFT;
+	val &= CRD_DATA_WIDTH_MASK;
+	pi->pcfg.data_bus_width = 8 * (1 << val);
+
+	val = readl(regs + CRD) >> CRD_DATA_BUFF_SHIFT;
+	val &= CRD_DATA_BUFF_MASK;
+	pi->pcfg.data_buf_dep = val + 1;
+
+	val = readl(regs + CR0) >> CR0_NUM_CHANS_SHIFT;
+	val &= CR0_NUM_CHANS_MASK;
+	val += 1;
+	pi->pcfg.num_chan = val;
+
+	val = readl(regs + CR0);
+	if (val & CR0_PERIPH_REQ_SET) {
+		val = (val >> CR0_NUM_PERIPH_SHIFT) & CR0_NUM_PERIPH_MASK;
+		val += 1;
+		pi->pcfg.num_peri = val;
+		pi->pcfg.peri_ns = readl(regs + CR4);
+	} else {
+		pi->pcfg.num_peri = 0;
+	}
+
+	val = readl(regs + CR0);
+	if (val & CR0_BOOT_MAN_NS)
+		pi->pcfg.mode |= DMAC_MODE_NS;
+	else
+		pi->pcfg.mode &= ~DMAC_MODE_NS;
+
+	val = readl(regs + CR0) >> CR0_NUM_EVENTS_SHIFT;
+	val &= CR0_NUM_EVENTS_MASK;
+	val += 1;
+	pi->pcfg.num_events = val;
+
+	pi->pcfg.irq_ns = readl(regs + CR3);
+
+	pi->pcfg.periph_id = get_id(pi, PERIPH_ID);
+	pi->pcfg.pcell_id = get_id(pi, PCELL_ID);
+}
+
+static inline void _reset_thread(struct pl330_thread *thrd)
+{
+	struct pl330_dmac *pl330 = thrd->dmac;
+	struct pl330_info *pi = pl330->pinfo;
+
+	thrd->req[0].mc_cpu = pl330->mcode_cpu
+				+ (thrd->id * pi->mcbufsz);
+	thrd->req[0].mc_bus = pl330->mcode_bus
+				+ (thrd->id * pi->mcbufsz);
+	thrd->req[0].r = NULL;
+	mark_free(thrd, 0);
+
+	thrd->req[1].mc_cpu = thrd->req[0].mc_cpu
+				+ pi->mcbufsz / 2;
+	thrd->req[1].mc_bus = thrd->req[0].mc_bus
+				+ pi->mcbufsz / 2;
+	thrd->req[1].r = NULL;
+	mark_free(thrd, 1);
+}
+
+static int dmac_alloc_threads(struct pl330_dmac *pl330)
+{
+	struct pl330_info *pi = pl330->pinfo;
+	int chans = pi->pcfg.num_chan;
+	struct pl330_thread *thrd;
+	int i;
+
+	/* Allocate 1 Manager and 'chans' Channel threads */
+	pl330->channels = kzalloc((1 + chans) * sizeof(*thrd),
+					GFP_KERNEL);
+	if (!pl330->channels)
+		return -ENOMEM;
+
+	/* Init Channel threads */
+	for (i = 0; i < chans; i++) {
+		thrd = &pl330->channels[i];
+		thrd->id = i;
+		thrd->dmac = pl330;
+		_reset_thread(thrd);
+		thrd->free = true;
+	}
+
+	/* MANAGER is indexed at the end */
+	thrd = &pl330->channels[chans];
+	thrd->id = chans;
+	thrd->dmac = pl330;
+	thrd->free = false;
+	pl330->manager = thrd;
+
+	return 0;
+}
+
+static int dmac_alloc_resources(struct pl330_dmac *pl330)
+{
+	struct pl330_info *pi = pl330->pinfo;
+	int chans = pi->pcfg.num_chan;
+	int ret;
+
+	/*
+	 * Alloc MicroCode buffer for 'chans' Channel threads.
+	 * A channel's buffer offset is (Channel_Id * MCODE_BUFF_PERCHAN)
+	 */
+	pl330->mcode_cpu = dma_alloc_coherent(pi->dev,
+				chans * pi->mcbufsz,
+				&pl330->mcode_bus, GFP_KERNEL);
+	if (!pl330->mcode_cpu) {
+		dev_err(pi->dev, "%s:%d Can't allocate memory!\n",
+			__func__, __LINE__);
+		return -ENOMEM;
+	}
+
+	ret = dmac_alloc_threads(pl330);
+	if (ret) {
+		dev_err(pi->dev, "%s:%d Can't to create channels for DMAC!\n",
+			__func__, __LINE__);
+		dma_free_coherent(pi->dev,
+				chans * pi->mcbufsz,
+				pl330->mcode_cpu, pl330->mcode_bus);
+		return ret;
+	}
+
+	return 0;
+}
+
+static int pl330_add(struct pl330_info *pi)
+{
+	struct pl330_dmac *pl330;
+	void __iomem *regs;
+	int i, ret;
+
+	if (!pi || !pi->dev)
+		return -EINVAL;
+
+	/* If already added */
+	if (pi->pl330_data)
+		return -EINVAL;
+
+	/*
+	 * If the SoC can perform reset on the DMAC, then do it
+	 * before reading its configuration.
+	 */
+	if (pi->dmac_reset)
+		pi->dmac_reset(pi);
+
+	regs = pi->base;
+
+	/* Check if we can handle this DMAC */
+	if ((get_id(pi, PERIPH_ID) & 0xfffff) != PERIPH_ID_VAL
+	   || get_id(pi, PCELL_ID) != PCELL_ID_VAL) {
+		dev_err(pi->dev, "PERIPH_ID 0x%x, PCELL_ID 0x%x !\n",
+			get_id(pi, PERIPH_ID), get_id(pi, PCELL_ID));
+		return -EINVAL;
+	}
+
+	/* Read the configuration of the DMAC */
+	read_dmac_config(pi);
+
+	if (pi->pcfg.num_events == 0) {
+		dev_err(pi->dev, "%s:%d Can't work without events!\n",
+			__func__, __LINE__);
+		return -EINVAL;
+	}
+
+	pl330 = kzalloc(sizeof(*pl330), GFP_KERNEL);
+	if (!pl330) {
+		dev_err(pi->dev, "%s:%d Can't allocate memory!\n",
+			__func__, __LINE__);
+		return -ENOMEM;
+	}
+
+	/* Assign the info structure and private data */
+	pl330->pinfo = pi;
+	pi->pl330_data = pl330;
+
+	spin_lock_init(&pl330->lock);
+
+	INIT_LIST_HEAD(&pl330->req_done);
+
+	/* Use default MC buffer size if not provided */
+	if (!pi->mcbufsz)
+		pi->mcbufsz = MCODE_BUFF_PER_REQ * 2;
+
+	/* Mark all events as free */
+	for (i = 0; i < pi->pcfg.num_events; i++)
+		pl330->events[i] = -1;
+
+	/* Allocate resources needed by the DMAC */
+	ret = dmac_alloc_resources(pl330);
+	if (ret) {
+		dev_err(pi->dev, "Unable to create channels for DMAC\n");
+		kfree(pl330);
+		return ret;
+	}
+
+	tasklet_init(&pl330->tasks, pl330_dotask, (unsigned long) pl330);
+
+	pl330->state = INIT;
+
+	return 0;
+}
+
+static int dmac_free_threads(struct pl330_dmac *pl330)
+{
+	struct pl330_info *pi = pl330->pinfo;
+	int chans = pi->pcfg.num_chan;
+	struct pl330_thread *thrd;
+	int i;
+
+	/* Release Channel threads */
+	for (i = 0; i < chans; i++) {
+		thrd = &pl330->channels[i];
+		pl330_release_channel((void *)thrd);
+	}
+
+	/* Free memory */
+	kfree(pl330->channels);
+
+	return 0;
+}
+
+static void dmac_free_resources(struct pl330_dmac *pl330)
+{
+	struct pl330_info *pi = pl330->pinfo;
+	int chans = pi->pcfg.num_chan;
+
+	dmac_free_threads(pl330);
+
+	dma_free_coherent(pi->dev, chans * pi->mcbufsz,
+				pl330->mcode_cpu, pl330->mcode_bus);
+}
+
+static void pl330_del(struct pl330_info *pi)
+{
+	struct pl330_dmac *pl330;
+
+	if (!pi || !pi->pl330_data)
+		return;
+
+	pl330 = pi->pl330_data;
+
+	pl330->state = UNINIT;
+
+	tasklet_kill(&pl330->tasks);
+
+	/* Free DMAC resources */
+	dmac_free_resources(pl330);
+
+	kfree(pl330);
+	pi->pl330_data = NULL;
+}
+
 /* forward declaration */
 static struct amba_driver pl330_driver;
 
@@ -234,7 +2320,7 @@
 	/* Pick up ripe tomatoes */
 	list_for_each_entry_safe(desc, _dt, &pch->work_list, node)
 		if (desc->status == DONE) {
-			pch->completed = desc->txd.cookie;
+			dma_cookie_complete(&desc->txd);
 			list_move_tail(&desc->node, &list);
 		}
 
@@ -305,7 +2391,7 @@
 
 	spin_lock_irqsave(&pch->lock, flags);
 
-	pch->completed = chan->cookie = 1;
+	dma_cookie_init(chan);
 	pch->cyclic = false;
 
 	pch->pl330_chid = pl330_request_channel(&pdmac->pif);
@@ -340,7 +2426,6 @@
 		/* Mark all desc done */
 		list_for_each_entry_safe(desc, _dt, &pch->work_list , node) {
 			desc->status = DONE;
-			pch->completed = desc->txd.cookie;
 			list_move_tail(&desc->node, &list);
 		}
 
@@ -396,18 +2481,7 @@
 pl330_tx_status(struct dma_chan *chan, dma_cookie_t cookie,
 		 struct dma_tx_state *txstate)
 {
-	struct dma_pl330_chan *pch = to_pchan(chan);
-	dma_cookie_t last_done, last_used;
-	int ret;
-
-	last_done = pch->completed;
-	last_used = chan->cookie;
-
-	ret = dma_async_is_complete(cookie, last_done, last_used);
-
-	dma_set_tx_state(txstate, last_done, last_used, 0);
-
-	return ret;
+	return dma_cookie_status(chan, cookie, txstate);
 }
 
 static void pl330_issue_pending(struct dma_chan *chan)
@@ -430,26 +2504,16 @@
 	spin_lock_irqsave(&pch->lock, flags);
 
 	/* Assign cookies to all nodes */
-	cookie = tx->chan->cookie;
-
 	while (!list_empty(&last->node)) {
 		desc = list_entry(last->node.next, struct dma_pl330_desc, node);
 
-		if (++cookie < 0)
-			cookie = 1;
-		desc->txd.cookie = cookie;
+		dma_cookie_assign(&desc->txd);
 
 		list_move_tail(&desc->node, &pch->work_list);
 	}
 
-	if (++cookie < 0)
-		cookie = 1;
-	last->txd.cookie = cookie;
-
+	cookie = dma_cookie_assign(&last->txd);
 	list_add_tail(&last->node, &pch->work_list);
-
-	tx->chan->cookie = cookie;
-
 	spin_unlock_irqrestore(&pch->lock, flags);
 
 	return cookie;
@@ -553,6 +2617,7 @@
 	async_tx_ack(&desc->txd);
 
 	desc->req.peri = peri_id ? pch->chan.chan_id : 0;
+	desc->rqcfg.pcfg = &pch->dmac->pif.pcfg;
 
 	dma_async_tx_descriptor_init(&desc->txd, &pch->chan);
 
@@ -621,7 +2686,8 @@
 
 static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic(
 		struct dma_chan *chan, dma_addr_t dma_addr, size_t len,
-		size_t period_len, enum dma_transfer_direction direction)
+		size_t period_len, enum dma_transfer_direction direction,
+		void *context)
 {
 	struct dma_pl330_desc *desc;
 	struct dma_pl330_chan *pch = to_pchan(chan);
@@ -711,7 +2777,7 @@
 static struct dma_async_tx_descriptor *
 pl330_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned int sg_len, enum dma_transfer_direction direction,
-		unsigned long flg)
+		unsigned long flg, void *context)
 {
 	struct dma_pl330_desc *first, *desc = NULL;
 	struct dma_pl330_chan *pch = to_pchan(chan);
@@ -829,7 +2895,7 @@
 	if (IS_ERR(pdmac->clk)) {
 		dev_err(&adev->dev, "Cannot get operation clock.\n");
 		ret = -EINVAL;
-		goto probe_err1;
+		goto probe_err2;
 	}
 
 	amba_set_drvdata(adev, pdmac);
@@ -843,11 +2909,11 @@
 	ret = request_irq(irq, pl330_irq_handler, 0,
 			dev_name(&adev->dev), pi);
 	if (ret)
-		goto probe_err2;
+		goto probe_err3;
 
 	ret = pl330_add(pi);
 	if (ret)
-		goto probe_err3;
+		goto probe_err4;
 
 	INIT_LIST_HEAD(&pdmac->desc_pool);
 	spin_lock_init(&pdmac->pool_lock);
@@ -904,7 +2970,7 @@
 	ret = dma_async_device_register(pd);
 	if (ret) {
 		dev_err(&adev->dev, "unable to register DMAC\n");
-		goto probe_err4;
+		goto probe_err5;
 	}
 
 	dev_info(&adev->dev,
@@ -917,10 +2983,15 @@
 
 	return 0;
 
-probe_err4:
+probe_err5:
 	pl330_del(pi);
-probe_err3:
+probe_err4:
 	free_irq(irq, pi);
+probe_err3:
+#ifndef CONFIG_PM_RUNTIME
+	clk_disable(pdmac->clk);
+#endif
+	clk_put(pdmac->clk);
 probe_err2:
 	iounmap(pi->base);
 probe_err1:
diff --git a/drivers/dma/ppc4xx/adma.c b/drivers/dma/ppc4xx/adma.c
index fc457a7..ced9882 100644
--- a/drivers/dma/ppc4xx/adma.c
+++ b/drivers/dma/ppc4xx/adma.c
@@ -46,6 +46,7 @@
 #include <asm/dcr.h>
 #include <asm/dcr-regs.h>
 #include "adma.h"
+#include "../dmaengine.h"
 
 enum ppc_adma_init_code {
 	PPC_ADMA_INIT_OK = 0,
@@ -1930,7 +1931,7 @@
 				if (end_of_chain && slot_cnt) {
 					/* Should wait for ZeroSum completion */
 					if (cookie > 0)
-						chan->completed_cookie = cookie;
+						chan->common.completed_cookie = cookie;
 					return;
 				}
 
@@ -1960,7 +1961,7 @@
 	BUG_ON(!seen_current);
 
 	if (cookie > 0) {
-		chan->completed_cookie = cookie;
+		chan->common.completed_cookie = cookie;
 		pr_debug("\tcompleted cookie %d\n", cookie);
 	}
 
@@ -2150,22 +2151,6 @@
 }
 
 /**
- * ppc440spe_desc_assign_cookie - assign a cookie
- */
-static dma_cookie_t ppc440spe_desc_assign_cookie(
-		struct ppc440spe_adma_chan *chan,
-		struct ppc440spe_adma_desc_slot *desc)
-{
-	dma_cookie_t cookie = chan->common.cookie;
-
-	cookie++;
-	if (cookie < 0)
-		cookie = 1;
-	chan->common.cookie = desc->async_tx.cookie = cookie;
-	return cookie;
-}
-
-/**
  * ppc440spe_rxor_set_region_data -
  */
 static void ppc440spe_rxor_set_region(struct ppc440spe_adma_desc_slot *desc,
@@ -2235,8 +2220,7 @@
 	slots_per_op = group_start->slots_per_op;
 
 	spin_lock_bh(&chan->lock);
-
-	cookie = ppc440spe_desc_assign_cookie(chan, sw_desc);
+	cookie = dma_cookie_assign(tx);
 
 	if (unlikely(list_empty(&chan->chain))) {
 		/* first peer */
@@ -3944,28 +3928,16 @@
 			dma_cookie_t cookie, struct dma_tx_state *txstate)
 {
 	struct ppc440spe_adma_chan *ppc440spe_chan;
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
 	enum dma_status ret;
 
 	ppc440spe_chan = to_ppc440spe_adma_chan(chan);
-	last_used = chan->cookie;
-	last_complete = ppc440spe_chan->completed_cookie;
-
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
+	ret = dma_cookie_status(chan, cookie, txstate);
 	if (ret == DMA_SUCCESS)
 		return ret;
 
 	ppc440spe_adma_slot_cleanup(ppc440spe_chan);
 
-	last_used = chan->cookie;
-	last_complete = ppc440spe_chan->completed_cookie;
-
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-
-	return dma_async_is_complete(cookie, last_complete, last_used);
+	return dma_cookie_status(chan, cookie, txstate);
 }
 
 /**
@@ -4050,16 +4022,12 @@
 		async_tx_ack(&sw_desc->async_tx);
 		ppc440spe_desc_init_null_xor(group_start);
 
-		cookie = chan->common.cookie;
-		cookie++;
-		if (cookie <= 1)
-			cookie = 2;
+		cookie = dma_cookie_assign(&sw_desc->async_tx);
 
 		/* initialize the completed cookie to be less than
 		 * the most recently used cookie
 		 */
-		chan->completed_cookie = cookie - 1;
-		chan->common.cookie = sw_desc->async_tx.cookie = cookie;
+		chan->common.completed_cookie = cookie - 1;
 
 		/* channel should not be busy */
 		BUG_ON(ppc440spe_chan_is_busy(chan));
@@ -4529,6 +4497,7 @@
 	INIT_LIST_HEAD(&chan->all_slots);
 	chan->device = adev;
 	chan->common.device = &adev->common;
+	dma_cookie_init(&chan->common);
 	list_add_tail(&chan->common.device_node, &adev->common.channels);
 	tasklet_init(&chan->irq_tasklet, ppc440spe_adma_tasklet,
 		     (unsigned long)chan);
diff --git a/drivers/dma/ppc4xx/adma.h b/drivers/dma/ppc4xx/adma.h
index 8ada5a81..26b7a5e 100644
--- a/drivers/dma/ppc4xx/adma.h
+++ b/drivers/dma/ppc4xx/adma.h
@@ -81,7 +81,6 @@
  * @common: common dmaengine channel object members
  * @all_slots: complete domain of slots usable by the channel
  * @pending: allows batching of hardware operations
- * @completed_cookie: identifier for the most recently completed operation
  * @slots_allocated: records the actual size of the descriptor slot pool
  * @hw_chain_inited: h/w descriptor chain initialization flag
  * @irq_tasklet: bottom half where ppc440spe_adma_slot_cleanup runs
@@ -99,7 +98,6 @@
 	struct list_head all_slots;
 	struct ppc440spe_adma_desc_slot *last_used;
 	int pending;
-	dma_cookie_t completed_cookie;
 	int slots_allocated;
 	int hw_chain_inited;
 	struct tasklet_struct irq_tasklet;
diff --git a/drivers/dma/shdma.c b/drivers/dma/shdma.c
index 812fd76..19d7a8d 100644
--- a/drivers/dma/shdma.c
+++ b/drivers/dma/shdma.c
@@ -30,6 +30,8 @@
 #include <linux/kdebug.h>
 #include <linux/spinlock.h>
 #include <linux/rculist.h>
+
+#include "dmaengine.h"
 #include "shdma.h"
 
 /* DMA descriptor control */
@@ -296,13 +298,7 @@
 	else
 		power_up = false;
 
-	cookie = sh_chan->common.cookie;
-	cookie++;
-	if (cookie < 0)
-		cookie = 1;
-
-	sh_chan->common.cookie = cookie;
-	tx->cookie = cookie;
+	cookie = dma_cookie_assign(tx);
 
 	/* Mark all chunks of this descriptor as submitted, move to the queue */
 	list_for_each_entry_safe(chunk, c, desc->node.prev, node) {
@@ -673,7 +669,8 @@
 
 static struct dma_async_tx_descriptor *sh_dmae_prep_slave_sg(
 	struct dma_chan *chan, struct scatterlist *sgl, unsigned int sg_len,
-	enum dma_transfer_direction direction, unsigned long flags)
+	enum dma_transfer_direction direction, unsigned long flags,
+	void *context)
 {
 	struct sh_dmae_slave *param;
 	struct sh_dmae_chan *sh_chan;
@@ -764,12 +761,12 @@
 			cookie = tx->cookie;
 
 		if (desc->mark == DESC_COMPLETED && desc->chunks == 1) {
-			if (sh_chan->completed_cookie != desc->cookie - 1)
+			if (sh_chan->common.completed_cookie != desc->cookie - 1)
 				dev_dbg(sh_chan->dev,
 					"Completing cookie %d, expected %d\n",
 					desc->cookie,
-					sh_chan->completed_cookie + 1);
-			sh_chan->completed_cookie = desc->cookie;
+					sh_chan->common.completed_cookie + 1);
+			sh_chan->common.completed_cookie = desc->cookie;
 		}
 
 		/* Call callback on the last chunk */
@@ -823,7 +820,7 @@
 		 * Terminating and the loop completed normally: forgive
 		 * uncompleted cookies
 		 */
-		sh_chan->completed_cookie = sh_chan->common.cookie;
+		sh_chan->common.completed_cookie = sh_chan->common.cookie;
 
 	spin_unlock_irqrestore(&sh_chan->desc_lock, flags);
 
@@ -883,23 +880,14 @@
 					struct dma_tx_state *txstate)
 {
 	struct sh_dmae_chan *sh_chan = to_sh_chan(chan);
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
 	enum dma_status status;
 	unsigned long flags;
 
 	sh_dmae_chan_ld_cleanup(sh_chan, false);
 
-	/* First read completed cookie to avoid a skew */
-	last_complete = sh_chan->completed_cookie;
-	rmb();
-	last_used = chan->cookie;
-	BUG_ON(last_complete < 0);
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-
 	spin_lock_irqsave(&sh_chan->desc_lock, flags);
 
-	status = dma_async_is_complete(cookie, last_complete, last_used);
+	status = dma_cookie_status(chan, cookie, txstate);
 
 	/*
 	 * If we don't find cookie on the queue, it has been aborted and we have
@@ -1102,6 +1090,7 @@
 
 	/* reference struct dma_device */
 	new_sh_chan->common.device = &shdev->common;
+	dma_cookie_init(&new_sh_chan->common);
 
 	new_sh_chan->dev = shdev->common.dev;
 	new_sh_chan->id = id;
diff --git a/drivers/dma/shdma.h b/drivers/dma/shdma.h
index 2b55a27..0b1d2c1 100644
--- a/drivers/dma/shdma.h
+++ b/drivers/dma/shdma.h
@@ -30,7 +30,6 @@
 };
 
 struct sh_dmae_chan {
-	dma_cookie_t completed_cookie;	/* The maximum cookie completed */
 	spinlock_t desc_lock;		/* Descriptor operation lock */
 	struct list_head ld_queue;	/* Link descriptors queue */
 	struct list_head ld_free;	/* Link descriptors free */
diff --git a/drivers/dma/sirf-dma.c b/drivers/dma/sirf-dma.c
index 2333810..434ad31 100644
--- a/drivers/dma/sirf-dma.c
+++ b/drivers/dma/sirf-dma.c
@@ -18,6 +18,8 @@
 #include <linux/of_platform.h>
 #include <linux/sirfsoc_dma.h>
 
+#include "dmaengine.h"
+
 #define SIRFSOC_DMA_DESCRIPTORS                 16
 #define SIRFSOC_DMA_CHANNELS                    16
 
@@ -59,7 +61,6 @@
 	struct list_head		queued;
 	struct list_head		active;
 	struct list_head		completed;
-	dma_cookie_t			completed_cookie;
 	unsigned long			happened_cyclic;
 	unsigned long			completed_cyclic;
 
@@ -208,7 +209,7 @@
 			/* Free descriptors */
 			spin_lock_irqsave(&schan->lock, flags);
 			list_splice_tail_init(&list, &schan->free);
-			schan->completed_cookie = last_cookie;
+			schan->chan.completed_cookie = last_cookie;
 			spin_unlock_irqrestore(&schan->lock, flags);
 		} else {
 			/* for cyclic channel, desc is always in active list */
@@ -258,13 +259,7 @@
 	/* Move descriptor to queue */
 	list_move_tail(&sdesc->node, &schan->queued);
 
-	/* Update cookie */
-	cookie = schan->chan.cookie + 1;
-	if (cookie <= 0)
-		cookie = 1;
-
-	schan->chan.cookie = cookie;
-	sdesc->desc.cookie = cookie;
+	cookie = dma_cookie_assign(txd);
 
 	spin_unlock_irqrestore(&schan->lock, flags);
 
@@ -414,16 +409,13 @@
 {
 	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
 	unsigned long flags;
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
+	enum dma_status ret;
 
 	spin_lock_irqsave(&schan->lock, flags);
-	last_used = schan->chan.cookie;
-	last_complete = schan->completed_cookie;
+	ret = dma_cookie_status(chan, cookie, txstate);
 	spin_unlock_irqrestore(&schan->lock, flags);
 
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-	return dma_async_is_complete(cookie, last_complete, last_used);
+	return ret;
 }
 
 static struct dma_async_tx_descriptor *sirfsoc_dma_prep_interleaved(
@@ -497,7 +489,7 @@
 static struct dma_async_tx_descriptor *
 sirfsoc_dma_prep_cyclic(struct dma_chan *chan, dma_addr_t addr,
 	size_t buf_len, size_t period_len,
-	enum dma_transfer_direction direction)
+	enum dma_transfer_direction direction, void *context)
 {
 	struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan);
 	struct sirfsoc_dma_desc *sdesc = NULL;
@@ -635,8 +627,7 @@
 		schan = &sdma->channels[i];
 
 		schan->chan.device = dma;
-		schan->chan.cookie = 1;
-		schan->completed_cookie = schan->chan.cookie;
+		dma_cookie_init(&schan->chan);
 
 		INIT_LIST_HEAD(&schan->free);
 		INIT_LIST_HEAD(&schan->prepared);
diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c
index cc5ecbc..bdd41d4 100644
--- a/drivers/dma/ste_dma40.c
+++ b/drivers/dma/ste_dma40.c
@@ -21,6 +21,7 @@
 
 #include <plat/ste_dma40.h>
 
+#include "dmaengine.h"
 #include "ste_dma40_ll.h"
 
 #define D40_NAME "dma40"
@@ -220,8 +221,6 @@
  *
  * @lock: A spinlock to protect this struct.
  * @log_num: The logical number, if any of this channel.
- * @completed: Starts with 1, after first interrupt it is set to dma engine's
- * current cookie.
  * @pending_tx: The number of pending transfers. Used between interrupt handler
  * and tasklet.
  * @busy: Set to true when transfer is ongoing on this channel.
@@ -250,8 +249,6 @@
 struct d40_chan {
 	spinlock_t			 lock;
 	int				 log_num;
-	/* ID of the most recent completed transfer */
-	int				 completed;
 	int				 pending_tx;
 	bool				 busy;
 	struct d40_phy_res		*phy_chan;
@@ -1223,21 +1220,14 @@
 					     chan);
 	struct d40_desc *d40d = container_of(tx, struct d40_desc, txd);
 	unsigned long flags;
+	dma_cookie_t cookie;
 
 	spin_lock_irqsave(&d40c->lock, flags);
-
-	d40c->chan.cookie++;
-
-	if (d40c->chan.cookie < 0)
-		d40c->chan.cookie = 1;
-
-	d40d->txd.cookie = d40c->chan.cookie;
-
+	cookie = dma_cookie_assign(tx);
 	d40_desc_queue(d40c, d40d);
-
 	spin_unlock_irqrestore(&d40c->lock, flags);
 
-	return tx->cookie;
+	return cookie;
 }
 
 static int d40_start(struct d40_chan *d40c)
@@ -1357,7 +1347,7 @@
 		goto err;
 
 	if (!d40d->cyclic)
-		d40c->completed = d40d->txd.cookie;
+		dma_cookie_complete(&d40d->txd);
 
 	/*
 	 * If terminating a channel pending_tx is set to zero.
@@ -2182,7 +2172,7 @@
 	bool is_free_phy;
 	spin_lock_irqsave(&d40c->lock, flags);
 
-	d40c->completed = chan->cookie = 1;
+	dma_cookie_init(chan);
 
 	/* If no dma configuration is set use default configuration (memcpy) */
 	if (!d40c->configured) {
@@ -2299,7 +2289,8 @@
 							 struct scatterlist *sgl,
 							 unsigned int sg_len,
 							 enum dma_transfer_direction direction,
-							 unsigned long dma_flags)
+							 unsigned long dma_flags,
+							 void *context)
 {
 	if (direction != DMA_DEV_TO_MEM && direction != DMA_MEM_TO_DEV)
 		return NULL;
@@ -2310,7 +2301,7 @@
 static struct dma_async_tx_descriptor *
 dma40_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t dma_addr,
 		     size_t buf_len, size_t period_len,
-		     enum dma_transfer_direction direction)
+		     enum dma_transfer_direction direction, void *context)
 {
 	unsigned int periods = buf_len / period_len;
 	struct dma_async_tx_descriptor *txd;
@@ -2342,25 +2333,19 @@
 				     struct dma_tx_state *txstate)
 {
 	struct d40_chan *d40c = container_of(chan, struct d40_chan, chan);
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
-	int ret;
+	enum dma_status ret;
 
 	if (d40c->phy_chan == NULL) {
 		chan_err(d40c, "Cannot read status of unallocated channel\n");
 		return -EINVAL;
 	}
 
-	last_complete = d40c->completed;
-	last_used = chan->cookie;
+	ret = dma_cookie_status(chan, cookie, txstate);
+	if (ret != DMA_SUCCESS)
+		dma_set_residue(txstate, stedma40_residue(chan));
 
 	if (d40_is_paused(d40c))
 		ret = DMA_PAUSED;
-	else
-		ret = dma_async_is_complete(cookie, last_complete, last_used);
-
-	dma_set_tx_state(txstate, last_complete, last_used,
-			 stedma40_residue(chan));
 
 	return ret;
 }
diff --git a/drivers/dma/timb_dma.c b/drivers/dma/timb_dma.c
index a6f9c16..4e0dff5 100644
--- a/drivers/dma/timb_dma.c
+++ b/drivers/dma/timb_dma.c
@@ -31,6 +31,8 @@
 
 #include <linux/timb_dma.h>
 
+#include "dmaengine.h"
+
 #define DRIVER_NAME "timb-dma"
 
 /* Global DMA registers */
@@ -84,7 +86,6 @@
 					especially the lists and descriptors,
 					from races between the tasklet and calls
 					from above */
-	dma_cookie_t		last_completed_cookie;
 	bool			ongoing;
 	struct list_head	active_list;
 	struct list_head	queue;
@@ -284,7 +285,7 @@
 	else
 		iowrite32(0, td_chan->membase + TIMBDMA_OFFS_TX_DLAR);
 */
-	td_chan->last_completed_cookie = txd->cookie;
+	dma_cookie_complete(txd);
 	td_chan->ongoing = false;
 
 	callback = txd->callback;
@@ -349,12 +350,7 @@
 	dma_cookie_t cookie;
 
 	spin_lock_bh(&td_chan->lock);
-
-	cookie = txd->chan->cookie;
-	if (++cookie < 0)
-		cookie = 1;
-	txd->chan->cookie = cookie;
-	txd->cookie = cookie;
+	cookie = dma_cookie_assign(txd);
 
 	if (list_empty(&td_chan->active_list)) {
 		dev_dbg(chan2dev(txd->chan), "%s: started %u\n", __func__,
@@ -481,8 +477,7 @@
 	}
 
 	spin_lock_bh(&td_chan->lock);
-	td_chan->last_completed_cookie = 1;
-	chan->cookie = 1;
+	dma_cookie_init(chan);
 	spin_unlock_bh(&td_chan->lock);
 
 	return 0;
@@ -515,24 +510,13 @@
 static enum dma_status td_tx_status(struct dma_chan *chan, dma_cookie_t cookie,
 				    struct dma_tx_state *txstate)
 {
-	struct timb_dma_chan *td_chan =
-		container_of(chan, struct timb_dma_chan, chan);
-	dma_cookie_t		last_used;
-	dma_cookie_t		last_complete;
-	int			ret;
+	enum dma_status ret;
 
 	dev_dbg(chan2dev(chan), "%s: Entry\n", __func__);
 
-	last_complete = td_chan->last_completed_cookie;
-	last_used = chan->cookie;
+	ret = dma_cookie_status(chan, cookie, txstate);
 
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
-
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-
-	dev_dbg(chan2dev(chan),
-		"%s: exit, ret: %d, last_complete: %d, last_used: %d\n",
-		__func__, ret, last_complete, last_used);
+	dev_dbg(chan2dev(chan), "%s: exit, ret: %d\n", 	__func__, ret);
 
 	return ret;
 }
@@ -558,7 +542,8 @@
 
 static struct dma_async_tx_descriptor *td_prep_slave_sg(struct dma_chan *chan,
 	struct scatterlist *sgl, unsigned int sg_len,
-	enum dma_transfer_direction direction, unsigned long flags)
+	enum dma_transfer_direction direction, unsigned long flags,
+	void *context)
 {
 	struct timb_dma_chan *td_chan =
 		container_of(chan, struct timb_dma_chan, chan);
@@ -766,7 +751,7 @@
 		}
 
 		td_chan->chan.device = &td->dma;
-		td_chan->chan.cookie = 1;
+		dma_cookie_init(&td_chan->chan);
 		spin_lock_init(&td_chan->lock);
 		INIT_LIST_HEAD(&td_chan->active_list);
 		INIT_LIST_HEAD(&td_chan->queue);
diff --git a/drivers/dma/txx9dmac.c b/drivers/dma/txx9dmac.c
index 6122c36..913f55c 100644
--- a/drivers/dma/txx9dmac.c
+++ b/drivers/dma/txx9dmac.c
@@ -15,6 +15,8 @@
 #include <linux/platform_device.h>
 #include <linux/slab.h>
 #include <linux/scatterlist.h>
+
+#include "dmaengine.h"
 #include "txx9dmac.h"
 
 static struct txx9dmac_chan *to_txx9dmac_chan(struct dma_chan *chan)
@@ -279,21 +281,6 @@
 	}
 }
 
-/* Called with dc->lock held and bh disabled */
-static dma_cookie_t
-txx9dmac_assign_cookie(struct txx9dmac_chan *dc, struct txx9dmac_desc *desc)
-{
-	dma_cookie_t cookie = dc->chan.cookie;
-
-	if (++cookie < 0)
-		cookie = 1;
-
-	dc->chan.cookie = cookie;
-	desc->txd.cookie = cookie;
-
-	return cookie;
-}
-
 /*----------------------------------------------------------------------*/
 
 static void txx9dmac_dump_regs(struct txx9dmac_chan *dc)
@@ -424,7 +411,7 @@
 	dev_vdbg(chan2dev(&dc->chan), "descriptor %u %p complete\n",
 		 txd->cookie, desc);
 
-	dc->completed = txd->cookie;
+	dma_cookie_complete(txd);
 	callback = txd->callback;
 	param = txd->callback_param;
 
@@ -738,7 +725,7 @@
 	dma_cookie_t cookie;
 
 	spin_lock_bh(&dc->lock);
-	cookie = txx9dmac_assign_cookie(dc, desc);
+	cookie = dma_cookie_assign(tx);
 
 	dev_vdbg(chan2dev(tx->chan), "tx_submit: queued %u %p\n",
 		 desc->txd.cookie, desc);
@@ -846,7 +833,7 @@
 static struct dma_async_tx_descriptor *
 txx9dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned int sg_len, enum dma_transfer_direction direction,
-		unsigned long flags)
+		unsigned long flags, void *context)
 {
 	struct txx9dmac_chan *dc = to_txx9dmac_chan(chan);
 	struct txx9dmac_dev *ddev = dc->ddev;
@@ -972,27 +959,17 @@
 		   struct dma_tx_state *txstate)
 {
 	struct txx9dmac_chan *dc = to_txx9dmac_chan(chan);
-	dma_cookie_t last_used;
-	dma_cookie_t last_complete;
-	int ret;
+	enum dma_status ret;
 
-	last_complete = dc->completed;
-	last_used = chan->cookie;
-
-	ret = dma_async_is_complete(cookie, last_complete, last_used);
+	ret = dma_cookie_status(chan, cookie, txstate);
 	if (ret != DMA_SUCCESS) {
 		spin_lock_bh(&dc->lock);
 		txx9dmac_scan_descriptors(dc);
 		spin_unlock_bh(&dc->lock);
 
-		last_complete = dc->completed;
-		last_used = chan->cookie;
-
-		ret = dma_async_is_complete(cookie, last_complete, last_used);
+		ret = dma_cookie_status(chan, cookie, txstate);
 	}
 
-	dma_set_tx_state(txstate, last_complete, last_used, 0);
-
 	return ret;
 }
 
@@ -1057,7 +1034,7 @@
 		return -EIO;
 	}
 
-	dc->completed = chan->cookie = 1;
+	dma_cookie_init(chan);
 
 	dc->ccr = TXX9_DMA_CCR_IMMCHN | TXX9_DMA_CCR_INTENE | CCR_LE;
 	txx9dmac_chan_set_SMPCHN(dc);
@@ -1186,7 +1163,7 @@
 	dc->ddev->chan[ch] = dc;
 	dc->chan.device = &dc->dma;
 	list_add_tail(&dc->chan.device_node, &dc->chan.device->channels);
-	dc->chan.cookie = dc->completed = 1;
+	dma_cookie_init(&dc->chan);
 
 	if (is_dmac64(dc))
 		dc->ch_regs = &__txx9dmac_regs(dc->ddev)->CHAN[ch];
diff --git a/drivers/dma/txx9dmac.h b/drivers/dma/txx9dmac.h
index 365d423..f5a7605 100644
--- a/drivers/dma/txx9dmac.h
+++ b/drivers/dma/txx9dmac.h
@@ -172,7 +172,6 @@
 	spinlock_t		lock;
 
 	/* these other elements are all protected by lock */
-	dma_cookie_t		completed;
 	struct list_head	active_list;
 	struct list_head	queue;
 	struct list_head	free_list;
diff --git a/drivers/media/video/mx3_camera.c b/drivers/media/video/mx3_camera.c
index 7452277..93c35ef 100644
--- a/drivers/media/video/mx3_camera.c
+++ b/drivers/media/video/mx3_camera.c
@@ -286,7 +286,7 @@
 		sg_dma_address(sg)	= vb2_dma_contig_plane_dma_addr(vb, 0);
 		sg_dma_len(sg)		= new_size;
 
-		txd = ichan->dma_chan.device->device_prep_slave_sg(
+		txd = dmaengine_prep_slave_sg(
 			&ichan->dma_chan, sg, 1, DMA_DEV_TO_MEM,
 			DMA_PREP_INTERRUPT);
 		if (!txd)
diff --git a/drivers/media/video/timblogiw.c b/drivers/media/video/timblogiw.c
index 4ed1c7c2..02194c0 100644
--- a/drivers/media/video/timblogiw.c
+++ b/drivers/media/video/timblogiw.c
@@ -564,7 +564,7 @@
 
 	spin_unlock_irq(&fh->queue_lock);
 
-	desc = fh->chan->device->device_prep_slave_sg(fh->chan,
+	desc = dmaengine_prep_slave_sg(fh->chan,
 		buf->sg, sg_elems, DMA_DEV_TO_MEM,
 		DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_SRC_UNMAP);
 	if (!desc) {
diff --git a/drivers/mmc/host/atmel-mci.c b/drivers/mmc/host/atmel-mci.c
index 390863e..9819dc0 100644
--- a/drivers/mmc/host/atmel-mci.c
+++ b/drivers/mmc/host/atmel-mci.c
@@ -24,6 +24,7 @@
 #include <linux/seq_file.h>
 #include <linux/slab.h>
 #include <linux/stat.h>
+#include <linux/types.h>
 
 #include <linux/mmc/host.h>
 #include <linux/mmc/sdio.h>
@@ -173,6 +174,7 @@
 
 	struct atmel_mci_dma	dma;
 	struct dma_chan		*data_chan;
+	struct dma_slave_config	dma_conf;
 
 	u32			cmd_status;
 	u32			data_status;
@@ -863,16 +865,17 @@
 
 	if (data->flags & MMC_DATA_READ) {
 		direction = DMA_FROM_DEVICE;
-		slave_dirn = DMA_DEV_TO_MEM;
+		host->dma_conf.direction = slave_dirn = DMA_DEV_TO_MEM;
 	} else {
 		direction = DMA_TO_DEVICE;
-		slave_dirn = DMA_MEM_TO_DEV;
+		host->dma_conf.direction = slave_dirn = DMA_MEM_TO_DEV;
 	}
 
 	sglen = dma_map_sg(chan->device->dev, data->sg,
 			data->sg_len, direction);
 
-	desc = chan->device->device_prep_slave_sg(chan,
+	dmaengine_slave_config(chan, &host->dma_conf);
+	desc = dmaengine_prep_slave_sg(chan,
 			data->sg, sglen, slave_dirn,
 			DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc)
@@ -1960,10 +1963,6 @@
 	if (pdata && find_slave_dev(pdata->dma_slave)) {
 		dma_cap_mask_t mask;
 
-		setup_dma_addr(pdata->dma_slave,
-			       host->mapbase + ATMCI_TDR,
-			       host->mapbase + ATMCI_RDR);
-
 		/* Try to grab a DMA channel */
 		dma_cap_zero(mask);
 		dma_cap_set(DMA_SLAVE, mask);
@@ -1977,6 +1976,14 @@
 		dev_info(&host->pdev->dev,
 					"using %s for DMA transfers\n",
 					dma_chan_name(host->dma.chan));
+
+		host->dma_conf.src_addr = host->mapbase + ATMCI_RDR;
+		host->dma_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		host->dma_conf.src_maxburst = 1;
+		host->dma_conf.dst_addr = host->mapbase + ATMCI_TDR;
+		host->dma_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		host->dma_conf.dst_maxburst = 1;
+		host->dma_conf.device_fc = false;
 		return true;
 	}
 }
diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c
index 983e244e..032b847 100644
--- a/drivers/mmc/host/mmci.c
+++ b/drivers/mmc/host/mmci.c
@@ -30,6 +30,7 @@
 #include <linux/dma-mapping.h>
 #include <linux/amba/mmci.h>
 #include <linux/pm_runtime.h>
+#include <linux/types.h>
 
 #include <asm/div64.h>
 #include <asm/io.h>
@@ -400,6 +401,7 @@
 		.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES,
 		.src_maxburst = variant->fifohalfsize >> 2, /* # of words */
 		.dst_maxburst = variant->fifohalfsize >> 2, /* # of words */
+		.device_fc = false,
 	};
 	struct dma_chan *chan;
 	struct dma_device *device;
@@ -441,7 +443,7 @@
 		return -EINVAL;
 
 	dmaengine_slave_config(chan, &conf);
-	desc = device->device_prep_slave_sg(chan, data->sg, nr_sg,
+	desc = dmaengine_prep_slave_sg(chan, data->sg, nr_sg,
 					    conf.direction, DMA_CTRL_ACK);
 	if (!desc)
 		goto unmap_exit;
diff --git a/drivers/mmc/host/mxcmmc.c b/drivers/mmc/host/mxcmmc.c
index 4184b79..b2058b4 100644
--- a/drivers/mmc/host/mxcmmc.c
+++ b/drivers/mmc/host/mxcmmc.c
@@ -33,6 +33,7 @@
 #include <linux/gpio.h>
 #include <linux/regulator/consumer.h>
 #include <linux/dmaengine.h>
+#include <linux/types.h>
 
 #include <asm/dma.h>
 #include <asm/irq.h>
@@ -254,7 +255,7 @@
 	if (nents != data->sg_len)
 		return -EINVAL;
 
-	host->desc = host->dma->device->device_prep_slave_sg(host->dma,
+	host->desc = dmaengine_prep_slave_sg(host->dma,
 		data->sg, data->sg_len, slave_dirn,
 		DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 
@@ -267,6 +268,7 @@
 	wmb();
 
 	dmaengine_submit(host->desc);
+	dma_async_issue_pending(host->dma);
 
 	return 0;
 }
@@ -710,6 +712,7 @@
 	config->src_addr_width = 4;
 	config->dst_maxburst = host->burstlen;
 	config->src_maxburst = host->burstlen;
+	config->device_fc = false;
 
 	return dmaengine_slave_config(host->dma, config);
 }
diff --git a/drivers/mmc/host/mxs-mmc.c b/drivers/mmc/host/mxs-mmc.c
index 382c835..65f36cf 100644
--- a/drivers/mmc/host/mxs-mmc.c
+++ b/drivers/mmc/host/mxs-mmc.c
@@ -324,7 +324,7 @@
 		sg_len = SSP_PIO_NUM;
 	}
 
-	desc = host->dmach->device->device_prep_slave_sg(host->dmach,
+	desc = dmaengine_prep_slave_sg(host->dmach,
 				sgl, sg_len, host->slave_dirn, append);
 	if (desc) {
 		desc->callback = mxs_mmc_dma_irq_callback;
diff --git a/drivers/mmc/host/sh_mmcif.c b/drivers/mmc/host/sh_mmcif.c
index 60f2057..aafaf0b 100644
--- a/drivers/mmc/host/sh_mmcif.c
+++ b/drivers/mmc/host/sh_mmcif.c
@@ -286,7 +286,7 @@
 			 DMA_FROM_DEVICE);
 	if (ret > 0) {
 		host->dma_active = true;
-		desc = chan->device->device_prep_slave_sg(chan, sg, ret,
+		desc = dmaengine_prep_slave_sg(chan, sg, ret,
 			DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	}
 
@@ -335,7 +335,7 @@
 			 DMA_TO_DEVICE);
 	if (ret > 0) {
 		host->dma_active = true;
-		desc = chan->device->device_prep_slave_sg(chan, sg, ret,
+		desc = dmaengine_prep_slave_sg(chan, sg, ret,
 			DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	}
 
diff --git a/drivers/mmc/host/tmio_mmc_dma.c b/drivers/mmc/host/tmio_mmc_dma.c
index 8253ec1..fff9286 100644
--- a/drivers/mmc/host/tmio_mmc_dma.c
+++ b/drivers/mmc/host/tmio_mmc_dma.c
@@ -88,7 +88,7 @@
 
 	ret = dma_map_sg(chan->device->dev, sg, host->sg_len, DMA_FROM_DEVICE);
 	if (ret > 0)
-		desc = chan->device->device_prep_slave_sg(chan, sg, ret,
+		desc = dmaengine_prep_slave_sg(chan, sg, ret,
 			DMA_DEV_TO_MEM, DMA_CTRL_ACK);
 
 	if (desc) {
@@ -169,7 +169,7 @@
 
 	ret = dma_map_sg(chan->device->dev, sg, host->sg_len, DMA_TO_DEVICE);
 	if (ret > 0)
-		desc = chan->device->device_prep_slave_sg(chan, sg, ret,
+		desc = dmaengine_prep_slave_sg(chan, sg, ret,
 			DMA_MEM_TO_DEV, DMA_CTRL_ACK);
 
 	if (desc) {
diff --git a/drivers/mtd/nand/gpmi-nand/gpmi-lib.c b/drivers/mtd/nand/gpmi-nand/gpmi-lib.c
index 7db6555..590dd5c 100644
--- a/drivers/mtd/nand/gpmi-nand/gpmi-lib.c
+++ b/drivers/mtd/nand/gpmi-nand/gpmi-lib.c
@@ -835,7 +835,7 @@
 		| BM_GPMI_CTRL0_ADDRESS_INCREMENT
 		| BF_GPMI_CTRL0_XFER_COUNT(this->command_length);
 	pio[1] = pio[2] = 0;
-	desc = channel->device->device_prep_slave_sg(channel,
+	desc = dmaengine_prep_slave_sg(channel,
 					(struct scatterlist *)pio,
 					ARRAY_SIZE(pio), DMA_TRANS_NONE, 0);
 	if (!desc) {
@@ -848,8 +848,7 @@
 
 	sg_init_one(sgl, this->cmd_buffer, this->command_length);
 	dma_map_sg(this->dev, sgl, 1, DMA_TO_DEVICE);
-	desc = channel->device->device_prep_slave_sg(channel,
-					sgl, 1, DMA_MEM_TO_DEV, 1);
+	desc = dmaengine_prep_slave_sg(channel, sgl, 1, DMA_MEM_TO_DEV, 1);
 	if (!desc) {
 		pr_err("step 2 error\n");
 		return -1;
@@ -880,8 +879,7 @@
 		| BF_GPMI_CTRL0_ADDRESS(address)
 		| BF_GPMI_CTRL0_XFER_COUNT(this->upper_len);
 	pio[1] = 0;
-	desc = channel->device->device_prep_slave_sg(channel,
-					(struct scatterlist *)pio,
+	desc = dmaengine_prep_slave_sg(channel, (struct scatterlist *)pio,
 					ARRAY_SIZE(pio), DMA_TRANS_NONE, 0);
 	if (!desc) {
 		pr_err("step 1 error\n");
@@ -890,7 +888,7 @@
 
 	/* [2] send DMA request */
 	prepare_data_dma(this, DMA_TO_DEVICE);
-	desc = channel->device->device_prep_slave_sg(channel, &this->data_sgl,
+	desc = dmaengine_prep_slave_sg(channel, &this->data_sgl,
 						1, DMA_MEM_TO_DEV, 1);
 	if (!desc) {
 		pr_err("step 2 error\n");
@@ -916,7 +914,7 @@
 		| BF_GPMI_CTRL0_ADDRESS(BV_GPMI_CTRL0_ADDRESS__NAND_DATA)
 		| BF_GPMI_CTRL0_XFER_COUNT(this->upper_len);
 	pio[1] = 0;
-	desc = channel->device->device_prep_slave_sg(channel,
+	desc = dmaengine_prep_slave_sg(channel,
 					(struct scatterlist *)pio,
 					ARRAY_SIZE(pio), DMA_TRANS_NONE, 0);
 	if (!desc) {
@@ -926,8 +924,8 @@
 
 	/* [2] : send DMA request */
 	prepare_data_dma(this, DMA_FROM_DEVICE);
-	desc = channel->device->device_prep_slave_sg(channel, &this->data_sgl,
-						1, DMA_DEV_TO_MEM, 1);
+	desc = dmaengine_prep_slave_sg(channel, &this->data_sgl,
+					1, DMA_DEV_TO_MEM, 1);
 	if (!desc) {
 		pr_err("step 2 error\n");
 		return -1;
@@ -972,8 +970,7 @@
 	pio[4] = payload;
 	pio[5] = auxiliary;
 
-	desc = channel->device->device_prep_slave_sg(channel,
-					(struct scatterlist *)pio,
+	desc = dmaengine_prep_slave_sg(channel, (struct scatterlist *)pio,
 					ARRAY_SIZE(pio), DMA_TRANS_NONE, 0);
 	if (!desc) {
 		pr_err("step 2 error\n");
@@ -1007,7 +1004,7 @@
 		| BF_GPMI_CTRL0_ADDRESS(address)
 		| BF_GPMI_CTRL0_XFER_COUNT(0);
 	pio[1] = 0;
-	desc = channel->device->device_prep_slave_sg(channel,
+	desc = dmaengine_prep_slave_sg(channel,
 				(struct scatterlist *)pio, 2,
 				DMA_TRANS_NONE, 0);
 	if (!desc) {
@@ -1036,7 +1033,7 @@
 	pio[3] = geo->page_size;
 	pio[4] = payload;
 	pio[5] = auxiliary;
-	desc = channel->device->device_prep_slave_sg(channel,
+	desc = dmaengine_prep_slave_sg(channel,
 					(struct scatterlist *)pio,
 					ARRAY_SIZE(pio), DMA_TRANS_NONE, 1);
 	if (!desc) {
@@ -1055,7 +1052,7 @@
 		| BF_GPMI_CTRL0_ADDRESS(address)
 		| BF_GPMI_CTRL0_XFER_COUNT(geo->page_size);
 	pio[1] = 0;
-	desc = channel->device->device_prep_slave_sg(channel,
+	desc = dmaengine_prep_slave_sg(channel,
 				(struct scatterlist *)pio, 2,
 				DMA_TRANS_NONE, 1);
 	if (!desc) {
diff --git a/drivers/net/ethernet/micrel/ks8842.c b/drivers/net/ethernet/micrel/ks8842.c
index 0686b93..f84dd2d 100644
--- a/drivers/net/ethernet/micrel/ks8842.c
+++ b/drivers/net/ethernet/micrel/ks8842.c
@@ -458,7 +458,7 @@
 	if (sg_dma_len(&ctl->sg) % 4)
 		sg_dma_len(&ctl->sg) += 4 - sg_dma_len(&ctl->sg) % 4;
 
-	ctl->adesc = ctl->chan->device->device_prep_slave_sg(ctl->chan,
+	ctl->adesc = dmaengine_prep_slave_sg(ctl->chan,
 		&ctl->sg, 1, DMA_MEM_TO_DEV,
 		DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_SRC_UNMAP);
 	if (!ctl->adesc)
@@ -570,7 +570,7 @@
 
 		sg_dma_len(sg) = DMA_BUFFER_SIZE;
 
-		ctl->adesc = ctl->chan->device->device_prep_slave_sg(ctl->chan,
+		ctl->adesc = dmaengine_prep_slave_sg(ctl->chan,
 			sg, 1, DMA_DEV_TO_MEM,
 			DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_SRC_UNMAP);
 
diff --git a/drivers/spi/spi-dw-mid.c b/drivers/spi/spi-dw-mid.c
index 8418eb0..b9f0192 100644
--- a/drivers/spi/spi-dw-mid.c
+++ b/drivers/spi/spi-dw-mid.c
@@ -22,6 +22,7 @@
 #include <linux/interrupt.h>
 #include <linux/slab.h>
 #include <linux/spi/spi.h>
+#include <linux/types.h>
 
 #include "spi-dw.h"
 
@@ -136,6 +137,7 @@
 	txconf.dst_maxburst = LNW_DMA_MSIZE_16;
 	txconf.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
 	txconf.dst_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+	txconf.device_fc = false;
 
 	txchan->device->device_control(txchan, DMA_SLAVE_CONFIG,
 				       (unsigned long) &txconf);
@@ -144,7 +146,7 @@
 	dws->tx_sgl.dma_address = dws->tx_dma;
 	dws->tx_sgl.length = dws->len;
 
-	txdesc = txchan->device->device_prep_slave_sg(txchan,
+	txdesc = dmaengine_prep_slave_sg(txchan,
 				&dws->tx_sgl,
 				1,
 				DMA_MEM_TO_DEV,
@@ -158,6 +160,7 @@
 	rxconf.src_maxburst = LNW_DMA_MSIZE_16;
 	rxconf.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
 	rxconf.src_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+	rxconf.device_fc = false;
 
 	rxchan->device->device_control(rxchan, DMA_SLAVE_CONFIG,
 				       (unsigned long) &rxconf);
@@ -166,7 +169,7 @@
 	dws->rx_sgl.dma_address = dws->rx_dma;
 	dws->rx_sgl.length = dws->len;
 
-	rxdesc = rxchan->device->device_prep_slave_sg(rxchan,
+	rxdesc = dmaengine_prep_slave_sg(rxchan,
 				&dws->rx_sgl,
 				1,
 				DMA_DEV_TO_MEM,
diff --git a/drivers/spi/spi-ep93xx.c b/drivers/spi/spi-ep93xx.c
index d46e55c..6db2887 100644
--- a/drivers/spi/spi-ep93xx.c
+++ b/drivers/spi/spi-ep93xx.c
@@ -633,8 +633,8 @@
 	if (!nents)
 		return ERR_PTR(-ENOMEM);
 
-	txd = chan->device->device_prep_slave_sg(chan, sgt->sgl, nents,
-						 slave_dirn, DMA_CTRL_ACK);
+	txd = dmaengine_prep_slave_sg(chan, sgt->sgl, nents,
+					slave_dirn, DMA_CTRL_ACK);
 	if (!txd) {
 		dma_unmap_sg(chan->device->dev, sgt->sgl, sgt->nents, dir);
 		return ERR_PTR(-ENOMEM);
diff --git a/drivers/spi/spi-pl022.c b/drivers/spi/spi-pl022.c
index dc8485d..96f0da6 100644
--- a/drivers/spi/spi-pl022.c
+++ b/drivers/spi/spi-pl022.c
@@ -880,10 +880,12 @@
 	struct dma_slave_config rx_conf = {
 		.src_addr = SSP_DR(pl022->phybase),
 		.direction = DMA_DEV_TO_MEM,
+		.device_fc = false,
 	};
 	struct dma_slave_config tx_conf = {
 		.dst_addr = SSP_DR(pl022->phybase),
 		.direction = DMA_MEM_TO_DEV,
+		.device_fc = false,
 	};
 	unsigned int pages;
 	int ret;
@@ -1017,7 +1019,7 @@
 		goto err_tx_sgmap;
 
 	/* Send both scatterlists */
-	rxdesc = rxchan->device->device_prep_slave_sg(rxchan,
+	rxdesc = dmaengine_prep_slave_sg(rxchan,
 				      pl022->sgt_rx.sgl,
 				      rx_sglen,
 				      DMA_DEV_TO_MEM,
@@ -1025,7 +1027,7 @@
 	if (!rxdesc)
 		goto err_rxdesc;
 
-	txdesc = txchan->device->device_prep_slave_sg(txchan,
+	txdesc = dmaengine_prep_slave_sg(txchan,
 				      pl022->sgt_tx.sgl,
 				      tx_sglen,
 				      DMA_MEM_TO_DEV,
diff --git a/drivers/spi/spi-topcliff-pch.c b/drivers/spi/spi-topcliff-pch.c
index 5c6fa5e..ec47d3b 100644
--- a/drivers/spi/spi-topcliff-pch.c
+++ b/drivers/spi/spi-topcliff-pch.c
@@ -1099,7 +1099,7 @@
 		sg_dma_address(sg) = dma->rx_buf_dma + sg->offset;
 	}
 	sg = dma->sg_rx_p;
-	desc_rx = dma->chan_rx->device->device_prep_slave_sg(dma->chan_rx, sg,
+	desc_rx = dmaengine_prep_slave_sg(dma->chan_rx, sg,
 					num, DMA_DEV_TO_MEM,
 					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc_rx) {
@@ -1158,7 +1158,7 @@
 		sg_dma_address(sg) = dma->tx_buf_dma + sg->offset;
 	}
 	sg = dma->sg_tx_p;
-	desc_tx = dma->chan_tx->device->device_prep_slave_sg(dma->chan_tx,
+	desc_tx = dmaengine_prep_slave_sg(dma->chan_tx,
 					sg, num, DMA_MEM_TO_DEV,
 					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc_tx) {
diff --git a/drivers/tty/serial/amba-pl011.c b/drivers/tty/serial/amba-pl011.c
index 20d795d..0c65c9e 100644
--- a/drivers/tty/serial/amba-pl011.c
+++ b/drivers/tty/serial/amba-pl011.c
@@ -51,6 +51,7 @@
 #include <linux/dma-mapping.h>
 #include <linux/scatterlist.h>
 #include <linux/delay.h>
+#include <linux/types.h>
 
 #include <asm/io.h>
 #include <asm/sizes.h>
@@ -271,6 +272,7 @@
 		.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE,
 		.direction = DMA_MEM_TO_DEV,
 		.dst_maxburst = uap->fifosize >> 1,
+		.device_fc = false,
 	};
 	struct dma_chan *chan;
 	dma_cap_mask_t mask;
@@ -304,6 +306,7 @@
 			.src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE,
 			.direction = DMA_DEV_TO_MEM,
 			.src_maxburst = uap->fifosize >> 1,
+			.device_fc = false,
 		};
 
 		chan = dma_request_channel(mask, plat->dma_filter, plat->dma_rx_param);
@@ -481,7 +484,7 @@
 		return -EBUSY;
 	}
 
-	desc = dma_dev->device_prep_slave_sg(chan, &dmatx->sg, 1, DMA_MEM_TO_DEV,
+	desc = dmaengine_prep_slave_sg(chan, &dmatx->sg, 1, DMA_MEM_TO_DEV,
 					     DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc) {
 		dma_unmap_sg(dma_dev->dev, &dmatx->sg, 1, DMA_TO_DEVICE);
@@ -664,7 +667,6 @@
 static int pl011_dma_rx_trigger_dma(struct uart_amba_port *uap)
 {
 	struct dma_chan *rxchan = uap->dmarx.chan;
-	struct dma_device *dma_dev;
 	struct pl011_dmarx_data *dmarx = &uap->dmarx;
 	struct dma_async_tx_descriptor *desc;
 	struct pl011_sgbuf *sgbuf;
@@ -675,8 +677,7 @@
 	/* Start the RX DMA job */
 	sgbuf = uap->dmarx.use_buf_b ?
 		&uap->dmarx.sgbuf_b : &uap->dmarx.sgbuf_a;
-	dma_dev = rxchan->device;
-	desc = rxchan->device->device_prep_slave_sg(rxchan, &sgbuf->sg, 1,
+	desc = dmaengine_prep_slave_sg(rxchan, &sgbuf->sg, 1,
 					DMA_DEV_TO_MEM,
 					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	/*
diff --git a/drivers/tty/serial/pch_uart.c b/drivers/tty/serial/pch_uart.c
index 332f2eb..e825460 100644
--- a/drivers/tty/serial/pch_uart.c
+++ b/drivers/tty/serial/pch_uart.c
@@ -844,7 +844,7 @@
 
 	sg_dma_address(sg) = priv->rx_buf_dma;
 
-	desc = priv->chan_rx->device->device_prep_slave_sg(priv->chan_rx,
+	desc = dmaengine_prep_slave_sg(priv->chan_rx,
 			sg, 1, DMA_DEV_TO_MEM,
 			DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 
@@ -1003,7 +1003,7 @@
 			sg_dma_len(sg) = size;
 	}
 
-	desc = priv->chan_tx->device->device_prep_slave_sg(priv->chan_tx,
+	desc = dmaengine_prep_slave_sg(priv->chan_tx,
 					priv->sg_tx_p, nent, DMA_MEM_TO_DEV,
 					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc) {
diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c
index 61b7fd2..f8db8a7 100644
--- a/drivers/tty/serial/sh-sci.c
+++ b/drivers/tty/serial/sh-sci.c
@@ -1338,7 +1338,7 @@
 		struct scatterlist *sg = &s->sg_rx[i];
 		struct dma_async_tx_descriptor *desc;
 
-		desc = chan->device->device_prep_slave_sg(chan,
+		desc = dmaengine_prep_slave_sg(chan,
 			sg, 1, DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT);
 
 		if (desc) {
@@ -1453,7 +1453,7 @@
 
 	BUG_ON(!sg_dma_len(sg));
 
-	desc = chan->device->device_prep_slave_sg(chan,
+	desc = dmaengine_prep_slave_sg(chan,
 			sg, s->sg_len_tx, DMA_MEM_TO_DEV,
 			DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc) {
diff --git a/drivers/usb/musb/ux500_dma.c b/drivers/usb/musb/ux500_dma.c
index 97cb459..d05c7fb 100644
--- a/drivers/usb/musb/ux500_dma.c
+++ b/drivers/usb/musb/ux500_dma.c
@@ -115,12 +115,12 @@
 	slave_conf.dst_addr = usb_fifo_addr;
 	slave_conf.dst_addr_width = addr_width;
 	slave_conf.dst_maxburst = 16;
+	slave_conf.device_fc = false;
 
 	dma_chan->device->device_control(dma_chan, DMA_SLAVE_CONFIG,
 					     (unsigned long) &slave_conf);
 
-	dma_desc = dma_chan->device->
-			device_prep_slave_sg(dma_chan, &sg, 1, direction,
+	dma_desc = dmaengine_prep_slave_sg(dma_chan, &sg, 1, direction,
 					     DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!dma_desc)
 		return false;
diff --git a/drivers/usb/renesas_usbhs/fifo.c b/drivers/usb/renesas_usbhs/fifo.c
index 3648c82..6ec7f83 100644
--- a/drivers/usb/renesas_usbhs/fifo.c
+++ b/drivers/usb/renesas_usbhs/fifo.c
@@ -786,9 +786,8 @@
 	sg_dma_address(&sg) = pkt->dma + pkt->actual;
 	sg_dma_len(&sg) = pkt->trans;
 
-	desc = chan->device->device_prep_slave_sg(chan, &sg, 1, dir,
-						  DMA_PREP_INTERRUPT |
-						  DMA_CTRL_ACK);
+	desc = dmaengine_prep_slave_sg(chan, &sg, 1, dir,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc)
 		return;
 
diff --git a/drivers/video/mx3fb.c b/drivers/video/mx3fb.c
index 727a514..eec0d7b 100644
--- a/drivers/video/mx3fb.c
+++ b/drivers/video/mx3fb.c
@@ -337,7 +337,7 @@
 
 	/* This enables the channel */
 	if (mx3_fbi->cookie < 0) {
-		mx3_fbi->txd = dma_chan->device->device_prep_slave_sg(dma_chan,
+		mx3_fbi->txd = dmaengine_prep_slave_sg(dma_chan,
 		      &mx3_fbi->sg[0], 1, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT);
 		if (!mx3_fbi->txd) {
 			dev_err(mx3fb->dev, "Cannot allocate descriptor on %d\n",
@@ -1091,7 +1091,7 @@
 	if (mx3_fbi->txd)
 		async_tx_ack(mx3_fbi->txd);
 
-	txd = dma_chan->device->device_prep_slave_sg(dma_chan, sg +
+	txd = dmaengine_prep_slave_sg(dma_chan, sg +
 		mx3_fbi->cur_ipu_buf, 1, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT);
 	if (!txd) {
 		dev_err(fbi->device,
diff --git a/include/linux/amba/pl08x.h b/include/linux/amba/pl08x.h
index 033f6aa..e64ce2c 100644
--- a/include/linux/amba/pl08x.h
+++ b/include/linux/amba/pl08x.h
@@ -47,9 +47,6 @@
  * @muxval: a number usually used to poke into some mux regiser to
  * mux in the signal to this channel
  * @cctl_opt: default options for the channel control register
- * @device_fc: Flow Controller Settings for ccfg register. Only valid for slave
- * channels. Fill with 'true' if peripheral should be flow controller. Direction
- * will be selected at Runtime.
  * @addr: source/target address in physical memory for this DMA channel,
  * can be the address of a FIFO register for burst requests for example.
  * This can be left undefined if the PrimeCell API is used for configuring
@@ -68,7 +65,6 @@
 	int max_signal;
 	u32 muxval;
 	u32 cctl;
-	bool device_fc;
 	dma_addr_t addr;
 	bool circular_buffer;
 	bool single;
@@ -176,13 +172,15 @@
  * @runtime_addr: address for RX/TX according to the runtime config
  * @runtime_direction: current direction of this channel according to
  * runtime config
- * @lc: last completed transaction on this channel
  * @pend_list: queued transactions pending on this channel
  * @at: active transaction on this channel
  * @lock: a lock for this channel data
  * @host: a pointer to the host (internal use)
  * @state: whether the channel is idle, paused, running etc
  * @slave: whether this channel is a device (slave) or for memcpy
+ * @device_fc: Flow Controller Settings for ccfg register. Only valid for slave
+ * channels. Fill with 'true' if peripheral should be flow controller. Direction
+ * will be selected at Runtime.
  * @waiting: a TX descriptor on this channel which is waiting for a physical
  * channel to become available
  */
@@ -198,13 +196,13 @@
 	u32 src_cctl;
 	u32 dst_cctl;
 	enum dma_transfer_direction runtime_direction;
-	dma_cookie_t lc;
 	struct list_head pend_list;
 	struct pl08x_txd *at;
 	spinlock_t lock;
 	struct pl08x_driver_data *host;
 	enum pl08x_dma_chan_state state;
 	bool slave;
+	bool device_fc;
 	struct pl08x_txd *waiting;
 };
 
diff --git a/include/linux/amba/pl330.h b/include/linux/amba/pl330.h
index 12e023c..fe93758 100644
--- a/include/linux/amba/pl330.h
+++ b/include/linux/amba/pl330.h
@@ -13,7 +13,6 @@
 #define	__AMBA_PL330_H_
 
 #include <linux/dmaengine.h>
-#include <asm/hardware/pl330.h>
 
 struct dma_pl330_platdata {
 	/*
diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index a5966f6..676f967 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -18,14 +18,15 @@
  * The full GNU General Public License is included in this distribution in the
  * file called COPYING.
  */
-#ifndef DMAENGINE_H
-#define DMAENGINE_H
+#ifndef LINUX_DMAENGINE_H
+#define LINUX_DMAENGINE_H
 
 #include <linux/device.h>
 #include <linux/uio.h>
 #include <linux/bug.h>
 #include <linux/scatterlist.h>
 #include <linux/bitmap.h>
+#include <linux/types.h>
 #include <asm/page.h>
 
 /**
@@ -258,6 +259,7 @@
  * struct dma_chan - devices supply DMA channels, clients use them
  * @device: ptr to the dma device who supplies this channel, always !%NULL
  * @cookie: last cookie value returned to client
+ * @completed_cookie: last completed cookie for this channel
  * @chan_id: channel ID for sysfs
  * @dev: class device for sysfs
  * @device_node: used to add this to the device chan list
@@ -269,6 +271,7 @@
 struct dma_chan {
 	struct dma_device *device;
 	dma_cookie_t cookie;
+	dma_cookie_t completed_cookie;
 
 	/* sysfs */
 	int chan_id;
@@ -332,6 +335,9 @@
  * may or may not be applicable on memory sources.
  * @dst_maxburst: same as src_maxburst but for destination target
  * mutatis mutandis.
+ * @device_fc: Flow Controller Settings. Only valid for slave channels. Fill
+ * with 'true' if peripheral should be flow controller. Direction will be
+ * selected at Runtime.
  *
  * This struct is passed in as configuration data to a DMA engine
  * in order to set up a certain channel for DMA transport at runtime.
@@ -358,6 +364,7 @@
 	enum dma_slave_buswidth dst_addr_width;
 	u32 src_maxburst;
 	u32 dst_maxburst;
+	bool device_fc;
 };
 
 static inline const char *dma_chan_name(struct dma_chan *chan)
@@ -576,10 +583,11 @@
 	struct dma_async_tx_descriptor *(*device_prep_slave_sg)(
 		struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned int sg_len, enum dma_transfer_direction direction,
-		unsigned long flags);
+		unsigned long flags, void *context);
 	struct dma_async_tx_descriptor *(*device_prep_dma_cyclic)(
 		struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
-		size_t period_len, enum dma_transfer_direction direction);
+		size_t period_len, enum dma_transfer_direction direction,
+		void *context);
 	struct dma_async_tx_descriptor *(*device_prep_interleaved_dma)(
 		struct dma_chan *chan, struct dma_interleaved_template *xt,
 		unsigned long flags);
@@ -613,7 +621,24 @@
 	struct scatterlist sg;
 	sg_init_one(&sg, buf, len);
 
-	return chan->device->device_prep_slave_sg(chan, &sg, 1, dir, flags);
+	return chan->device->device_prep_slave_sg(chan, &sg, 1,
+						  dir, flags, NULL);
+}
+
+static inline struct dma_async_tx_descriptor *dmaengine_prep_slave_sg(
+	struct dma_chan *chan, struct scatterlist *sgl,	unsigned int sg_len,
+	enum dma_transfer_direction dir, unsigned long flags)
+{
+	return chan->device->device_prep_slave_sg(chan, sgl, sg_len,
+						  dir, flags, NULL);
+}
+
+static inline struct dma_async_tx_descriptor *dmaengine_prep_dma_cyclic(
+		struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
+		size_t period_len, enum dma_transfer_direction dir)
+{
+	return chan->device->device_prep_dma_cyclic(chan, buf_addr, buf_len,
+						period_len, dir, NULL);
 }
 
 static inline int dmaengine_terminate_all(struct dma_chan *chan)
diff --git a/include/linux/dw_dmac.h b/include/linux/dw_dmac.h
index f2c64f9..2412e02 100644
--- a/include/linux/dw_dmac.h
+++ b/include/linux/dw_dmac.h
@@ -31,18 +31,6 @@
 	unsigned char	chan_priority;
 };
 
-/**
- * enum dw_dma_slave_width - DMA slave register access width.
- * @DMA_SLAVE_WIDTH_8BIT: Do 8-bit slave register accesses
- * @DMA_SLAVE_WIDTH_16BIT: Do 16-bit slave register accesses
- * @DMA_SLAVE_WIDTH_32BIT: Do 32-bit slave register accesses
- */
-enum dw_dma_slave_width {
-	DW_DMA_SLAVE_WIDTH_8BIT,
-	DW_DMA_SLAVE_WIDTH_16BIT,
-	DW_DMA_SLAVE_WIDTH_32BIT,
-};
-
 /* bursts size */
 enum dw_dma_msize {
 	DW_DMA_MSIZE_1,
@@ -55,47 +43,21 @@
 	DW_DMA_MSIZE_256,
 };
 
-/* flow controller */
-enum dw_dma_fc {
-	DW_DMA_FC_D_M2M,
-	DW_DMA_FC_D_M2P,
-	DW_DMA_FC_D_P2M,
-	DW_DMA_FC_D_P2P,
-	DW_DMA_FC_P_P2M,
-	DW_DMA_FC_SP_P2P,
-	DW_DMA_FC_P_M2P,
-	DW_DMA_FC_DP_P2P,
-};
-
 /**
  * struct dw_dma_slave - Controller-specific information about a slave
  *
  * @dma_dev: required DMA master device
- * @tx_reg: physical address of data register used for
- *	memory-to-peripheral transfers
- * @rx_reg: physical address of data register used for
- *	peripheral-to-memory transfers
- * @reg_width: peripheral register width
  * @cfg_hi: Platform-specific initializer for the CFG_HI register
  * @cfg_lo: Platform-specific initializer for the CFG_LO register
  * @src_master: src master for transfers on allocated channel.
  * @dst_master: dest master for transfers on allocated channel.
- * @src_msize: src burst size.
- * @dst_msize: dest burst size.
- * @fc: flow controller for DMA transfer
  */
 struct dw_dma_slave {
 	struct device		*dma_dev;
-	dma_addr_t		tx_reg;
-	dma_addr_t		rx_reg;
-	enum dw_dma_slave_width	reg_width;
 	u32			cfg_hi;
 	u32			cfg_lo;
 	u8			src_master;
 	u8			dst_master;
-	u8			src_msize;
-	u8			dst_msize;
-	u8			fc;
 };
 
 /* Platform-configurable bits in CFG_HI */
diff --git a/sound/atmel/abdac.c b/sound/atmel/abdac.c
index 4fa1dbd..f7c2bb0 100644
--- a/sound/atmel/abdac.c
+++ b/sound/atmel/abdac.c
@@ -16,6 +16,7 @@
 #include <linux/interrupt.h>
 #include <linux/module.h>
 #include <linux/platform_device.h>
+#include <linux/types.h>
 #include <linux/io.h>
 
 #include <sound/core.h>
@@ -467,15 +468,24 @@
 	snd_card_set_dev(card, &pdev->dev);
 
 	if (pdata->dws.dma_dev) {
-		struct dw_dma_slave *dws = &pdata->dws;
 		dma_cap_mask_t mask;
 
-		dws->tx_reg = regs->start + DAC_DATA;
-
 		dma_cap_zero(mask);
 		dma_cap_set(DMA_SLAVE, mask);
 
-		dac->dma.chan = dma_request_channel(mask, filter, dws);
+		dac->dma.chan = dma_request_channel(mask, filter, &pdata->dws);
+		if (dac->dma.chan) {
+			struct dma_slave_config dma_conf = {
+				.dst_addr = regs->start + DAC_DATA,
+				.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES,
+				.src_maxburst = 1,
+				.dst_maxburst = 1,
+				.direction = DMA_MEM_TO_DEV,
+				.device_fc = false,
+			};
+
+			dmaengine_slave_config(dac->dma.chan, &dma_conf);
+		}
 	}
 	if (!pdata->dws.dma_dev || !dac->dma.chan) {
 		dev_dbg(&pdev->dev, "DMA not available\n");
diff --git a/sound/atmel/ac97c.c b/sound/atmel/ac97c.c
index 61dade6..115313e 100644
--- a/sound/atmel/ac97c.c
+++ b/sound/atmel/ac97c.c
@@ -20,6 +20,7 @@
 #include <linux/platform_device.h>
 #include <linux/mutex.h>
 #include <linux/gpio.h>
+#include <linux/types.h>
 #include <linux/io.h>
 
 #include <sound/core.h>
@@ -1014,16 +1015,28 @@
 
 	if (cpu_is_at32ap7000()) {
 		if (pdata->rx_dws.dma_dev) {
-			struct dw_dma_slave *dws = &pdata->rx_dws;
 			dma_cap_mask_t mask;
 
-			dws->rx_reg = regs->start + AC97C_CARHR + 2;
-
 			dma_cap_zero(mask);
 			dma_cap_set(DMA_SLAVE, mask);
 
 			chip->dma.rx_chan = dma_request_channel(mask, filter,
-								dws);
+								&pdata->rx_dws);
+			if (chip->dma.rx_chan) {
+				struct dma_slave_config dma_conf = {
+					.src_addr = regs->start + AC97C_CARHR +
+						2,
+					.src_addr_width =
+						DMA_SLAVE_BUSWIDTH_2_BYTES,
+					.src_maxburst = 1,
+					.dst_maxburst = 1,
+					.direction = DMA_DEV_TO_MEM,
+					.device_fc = false,
+				};
+
+				dmaengine_slave_config(chip->dma.rx_chan,
+						&dma_conf);
+			}
 
 			dev_info(&chip->pdev->dev, "using %s for DMA RX\n",
 				dev_name(&chip->dma.rx_chan->dev->device));
@@ -1031,16 +1044,28 @@
 		}
 
 		if (pdata->tx_dws.dma_dev) {
-			struct dw_dma_slave *dws = &pdata->tx_dws;
 			dma_cap_mask_t mask;
 
-			dws->tx_reg = regs->start + AC97C_CATHR + 2;
-
 			dma_cap_zero(mask);
 			dma_cap_set(DMA_SLAVE, mask);
 
 			chip->dma.tx_chan = dma_request_channel(mask, filter,
-								dws);
+								&pdata->tx_dws);
+			if (chip->dma.tx_chan) {
+				struct dma_slave_config dma_conf = {
+					.dst_addr = regs->start + AC97C_CATHR +
+						2,
+					.dst_addr_width =
+						DMA_SLAVE_BUSWIDTH_2_BYTES,
+					.src_maxburst = 1,
+					.dst_maxburst = 1,
+					.direction = DMA_MEM_TO_DEV,
+					.device_fc = false,
+				};
+
+				dmaengine_slave_config(chip->dma.tx_chan,
+						&dma_conf);
+			}
 
 			dev_info(&chip->pdev->dev, "using %s for DMA TX\n",
 				dev_name(&chip->dma.tx_chan->dev->device));
diff --git a/sound/soc/imx/imx-pcm-dma-mx2.c b/sound/soc/imx/imx-pcm-dma-mx2.c
index e43c8fa..6b818de 100644
--- a/sound/soc/imx/imx-pcm-dma-mx2.c
+++ b/sound/soc/imx/imx-pcm-dma-mx2.c
@@ -21,6 +21,7 @@
 #include <linux/platform_device.h>
 #include <linux/slab.h>
 #include <linux/dmaengine.h>
+#include <linux/types.h>
 
 #include <sound/core.h>
 #include <sound/initval.h>
@@ -58,6 +59,8 @@
 	if (ret)
 		return ret;
 
+	slave_config.device_fc = false;
+
 	if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) {
 		slave_config.dst_addr = dma_params->dma_addr;
 		slave_config.dst_maxburst = dma_params->burstsize;
diff --git a/sound/soc/sh/siu_pcm.c b/sound/soc/sh/siu_pcm.c
index 0193e59..5cfcc65 100644
--- a/sound/soc/sh/siu_pcm.c
+++ b/sound/soc/sh/siu_pcm.c
@@ -130,7 +130,7 @@
 	sg_dma_len(&sg) = size;
 	sg_dma_address(&sg) = buff;
 
-	desc = siu_stream->chan->device->device_prep_slave_sg(siu_stream->chan,
+	desc = dmaengine_prep_slave_sg(siu_stream->chan,
 		&sg, 1, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc) {
 		dev_err(dev, "Failed to allocate a dma descriptor\n");
@@ -180,7 +180,7 @@
 	sg_dma_len(&sg) = size;
 	sg_dma_address(&sg) = buff;
 
-	desc = siu_stream->chan->device->device_prep_slave_sg(siu_stream->chan,
+	desc = dmaengine_prep_slave_sg(siu_stream->chan,
 		&sg, 1, DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc) {
 		dev_err(dev, "Failed to allocate dma descriptor\n");
diff --git a/sound/soc/txx9/txx9aclc.c b/sound/soc/txx9/txx9aclc.c
index 2155461..b609d2c 100644
--- a/sound/soc/txx9/txx9aclc.c
+++ b/sound/soc/txx9/txx9aclc.c
@@ -132,7 +132,7 @@
 	sg_set_page(&sg, pfn_to_page(PFN_DOWN(buf_dma_addr)),
 		    dmadata->frag_bytes, buf_dma_addr & (PAGE_SIZE - 1));
 	sg_dma_address(&sg) = buf_dma_addr;
-	desc = chan->device->device_prep_slave_sg(chan, &sg, 1,
+	desc = dmaengine_prep_slave_sg(chan, &sg, 1,
 		dmadata->substream->stream == SNDRV_PCM_STREAM_PLAYBACK ?
 		DMA_MEM_TO_DEV : DMA_DEV_TO_MEM,
 		DMA_PREP_INTERRUPT | DMA_CTRL_ACK);