Merge git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6: (26 commits)
  sh: Convert sh to use read/update_persistent_clock
  sh: Move PMB debugfs entry initialization to later stage
  sh: Fix up flush_cache_vmap() on SMP.
  sh: fix up MMU reset with variable PMB mapping sizes.
  sh: establish PMB mappings for NUMA nodes.
  sh: check for existing mappings for bolted PMB entries.
  sh: fixed virt/phys mapping helpers for PMB.
  sh: make pmb iomapping configurable.
  sh: reworked dynamic PMB mapping.
  sh: Fix up cpumask_of_pcibus() for the NUMA build.
  serial: sh-sci: Tidy up build warnings.
  sh: Fix up ctrl_read/write stragglers in migor setup.
  serial: sh-sci: Add DMA support.
  dmaengine: shdma: extend .device_terminate_all() to record partial transfer
  sh: merge sh7722 and sh7724 DMA register definitions
  sh: activate runtime PM for dmaengine on sh7722 and sh7724
  dmaengine: shdma: add runtime PM support.
  dmaengine: shdma: separate DMA headers.
  dmaengine: shdma: convert to platform device resources
  dmaengine: shdma: fix DMA error handling.
  ...

28 files changed, 1295 insertions, 478 deletions

diff --git a/arch/sh/boards/mach-migor/setup.c b/arch/sh/boards/mach-migor/setup.c
index be300aaca6fe..7da0fc94a01e 100644
--- a/arch/sh/boards/mach-migor/setup.c
+++ b/arch/sh/boards/mach-migor/setup.c
@@ -419,6 +419,9 @@ static struct i2c_board_info migor_i2c_devices[] = {
 		I2C_BOARD_INFO("migor_ts", 0x51),
 		.irq = 38, /* IRQ6 */
 	},
+	{
+		I2C_BOARD_INFO("wm8978", 0x1a),
+	},
 };
 
 static struct i2c_board_info migor_i2c_camera[] = {
@@ -619,6 +622,19 @@ static int __init migor_devices_setup(void)
 
 	platform_resource_setup_memory(&migor_ceu_device, "ceu", 4 << 20);
 
+	/* SIU: Port B */
+	gpio_request(GPIO_FN_SIUBOLR, NULL);
+	gpio_request(GPIO_FN_SIUBOBT, NULL);
+	gpio_request(GPIO_FN_SIUBISLD, NULL);
+	gpio_request(GPIO_FN_SIUBOSLD, NULL);
+	gpio_request(GPIO_FN_SIUMCKB, NULL);
+
+	/*
+	 * The original driver sets SIUB OLR/OBT, ILR/IBT, and SIUA OLR/OBT to
+	 * output. Need only SIUB, set to output for master mode (table 34.2)
+	 */
+	__raw_writew(__raw_readw(PORT_MSELCRA) | 1, PORT_MSELCRA);
+
 	i2c_register_board_info(0, migor_i2c_devices,
 				ARRAY_SIZE(migor_i2c_devices));
 
diff --git a/arch/sh/boot/compressed/cache.c b/arch/sh/boot/compressed/cache.c
index e27fc74f228c..d0b77b68a4d0 100644
--- a/arch/sh/boot/compressed/cache.c
+++ b/arch/sh/boot/compressed/cache.c
@@ -5,7 +5,7 @@ int cache_control(unsigned int command)
 
 	for (i = 0; i < (32 * 1024); i += 32) {
 		(void)*p;
-		p += (32 / sizeof (int));
+		p += (32 / sizeof(int));
 	}
 
 	return 0;
diff --git a/arch/sh/include/asm/cacheflush.h b/arch/sh/include/asm/cacheflush.h
index da3ebec921a7..1f4e562c5e8c 100644
--- a/arch/sh/include/asm/cacheflush.h
+++ b/arch/sh/include/asm/cacheflush.h
@@ -86,8 +86,8 @@ extern void copy_from_user_page(struct vm_area_struct *vma,
 	struct page *page, unsigned long vaddr, void *dst, const void *src,
 	unsigned long len);
 
-#define flush_cache_vmap(start, end)		flush_cache_all()
-#define flush_cache_vunmap(start, end)		flush_cache_all()
+#define flush_cache_vmap(start, end)		local_flush_cache_all(NULL)
+#define flush_cache_vunmap(start, end)		local_flush_cache_all(NULL)
 
 #define flush_dcache_mmap_lock(mapping)		do { } while (0)
 #define flush_dcache_mmap_unlock(mapping)	do { } while (0)
diff --git a/arch/sh/include/asm/dma-register.h b/arch/sh/include/asm/dma-register.h
new file mode 100644
index 000000000000..51cd78feacff
--- /dev/null
+++ b/arch/sh/include/asm/dma-register.h
@@ -0,0 +1,51 @@
+/*
+ * Common header for the legacy SH DMA driver and the new dmaengine driver
+ *
+ * extracted from arch/sh/include/asm/dma-sh.h:
+ *
+ * Copyright (C) 2000  Takashi YOSHII
+ * Copyright (C) 2003  Paul Mundt
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ */
+#ifndef DMA_REGISTER_H
+#define DMA_REGISTER_H
+
+/* DMA register */
+#define SAR	0x00
+#define DAR	0x04
+#define TCR	0x08
+#define CHCR	0x0C
+#define DMAOR	0x40
+
+/* DMAOR definitions */
+#define DMAOR_AE	0x00000004
+#define DMAOR_NMIF	0x00000002
+#define DMAOR_DME	0x00000001
+
+/* Definitions for the SuperH DMAC */
+#define REQ_L	0x00000000
+#define REQ_E	0x00080000
+#define RACK_H	0x00000000
+#define RACK_L	0x00040000
+#define ACK_R	0x00000000
+#define ACK_W	0x00020000
+#define ACK_H	0x00000000
+#define ACK_L	0x00010000
+#define DM_INC	0x00004000
+#define DM_DEC	0x00008000
+#define DM_FIX	0x0000c000
+#define SM_INC	0x00001000
+#define SM_DEC	0x00002000
+#define SM_FIX	0x00003000
+#define RS_IN	0x00000200
+#define RS_OUT	0x00000300
+#define TS_BLK	0x00000040
+#define TM_BUR	0x00000020
+#define CHCR_DE	0x00000001
+#define CHCR_TE	0x00000002
+#define CHCR_IE	0x00000004
+
+#endif
diff --git a/arch/sh/include/asm/dma-sh.h b/arch/sh/include/asm/dma-sh.h
index e934a2e66651..f3acb8e34c6b 100644
--- a/arch/sh/include/asm/dma-sh.h
+++ b/arch/sh/include/asm/dma-sh.h
@@ -11,7 +11,8 @@
 #ifndef __DMA_SH_H
 #define __DMA_SH_H
 
-#include <asm/dma.h>
+#include <asm/dma-register.h>
+#include <cpu/dma-register.h>
 #include <cpu/dma.h>
 
 /* DMAOR contorl: The DMAOR access size is different by CPU.*/
@@ -53,34 +54,6 @@ static int dmte_irq_map[] __maybe_unused = {
 #endif
 };
 
-/* Definitions for the SuperH DMAC */
-#define REQ_L	0x00000000
-#define REQ_E	0x00080000
-#define RACK_H	0x00000000
-#define RACK_L	0x00040000
-#define ACK_R	0x00000000
-#define ACK_W	0x00020000
-#define ACK_H	0x00000000
-#define ACK_L	0x00010000
-#define DM_INC	0x00004000
-#define DM_DEC	0x00008000
-#define DM_FIX	0x0000c000
-#define SM_INC	0x00001000
-#define SM_DEC	0x00002000
-#define SM_FIX	0x00003000
-#define RS_IN	0x00000200
-#define RS_OUT	0x00000300
-#define TS_BLK	0x00000040
-#define TM_BUR	0x00000020
-#define CHCR_DE 0x00000001
-#define CHCR_TE 0x00000002
-#define CHCR_IE 0x00000004
-
-/* DMAOR definitions */
-#define DMAOR_AE	0x00000004
-#define DMAOR_NMIF	0x00000002
-#define DMAOR_DME	0x00000001
-
 /*
  * Define the default configuration for dual address memory-memory transfer.
  * The 0x400 value represents auto-request, external->external.
@@ -111,61 +84,4 @@ static u32 dma_base_addr[] __maybe_unused = {
 #endif
 };
 
-/* DMA register */
-#define SAR     0x00
-#define DAR     0x04
-#define TCR     0x08
-#define CHCR    0x0C
-#define DMAOR	0x40
-
-/*
- * for dma engine
- *
- * SuperH DMA mode
- */
-#define SHDMA_MIX_IRQ	(1 << 1)
-#define SHDMA_DMAOR1	(1 << 2)
-#define SHDMA_DMAE1	(1 << 3)
-
-enum sh_dmae_slave_chan_id {
-	SHDMA_SLAVE_SCIF0_TX,
-	SHDMA_SLAVE_SCIF0_RX,
-	SHDMA_SLAVE_SCIF1_TX,
-	SHDMA_SLAVE_SCIF1_RX,
-	SHDMA_SLAVE_SCIF2_TX,
-	SHDMA_SLAVE_SCIF2_RX,
-	SHDMA_SLAVE_SCIF3_TX,
-	SHDMA_SLAVE_SCIF3_RX,
-	SHDMA_SLAVE_SCIF4_TX,
-	SHDMA_SLAVE_SCIF4_RX,
-	SHDMA_SLAVE_SCIF5_TX,
-	SHDMA_SLAVE_SCIF5_RX,
-	SHDMA_SLAVE_SIUA_TX,
-	SHDMA_SLAVE_SIUA_RX,
-	SHDMA_SLAVE_SIUB_TX,
-	SHDMA_SLAVE_SIUB_RX,
-	SHDMA_SLAVE_NUMBER,	/* Must stay last */
-};
-
-struct sh_dmae_slave_config {
-	enum sh_dmae_slave_chan_id	slave_id;
-	dma_addr_t			addr;
-	u32				chcr;
-	char				mid_rid;
-};
-
-struct sh_dmae_pdata {
-	unsigned int mode;
-	struct sh_dmae_slave_config *config;
-	int config_num;
-};
-
-struct device;
-
-struct sh_dmae_slave {
-	enum sh_dmae_slave_chan_id	slave_id; /* Set by the platform */
-	struct device			*dma_dev; /* Set by the platform */
-	struct sh_dmae_slave_config	*config;  /* Set by the driver */
-};
-
 #endif /* __DMA_SH_H */
diff --git a/arch/sh/include/asm/dmaengine.h b/arch/sh/include/asm/dmaengine.h
new file mode 100644
index 000000000000..bf2f30cf0a27
--- /dev/null
+++ b/arch/sh/include/asm/dmaengine.h
@@ -0,0 +1,93 @@
+/*
+ * Header for the new SH dmaengine driver
+ *
+ * Copyright (C) 2010 Guennadi Liakhovetski <g.liakhovetski@gmx.de>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#ifndef ASM_DMAENGINE_H
+#define ASM_DMAENGINE_H
+
+#include <linux/dmaengine.h>
+#include <linux/list.h>
+
+#include <asm/dma-register.h>
+
+#define SH_DMAC_MAX_CHANNELS	6
+
+enum sh_dmae_slave_chan_id {
+	SHDMA_SLAVE_SCIF0_TX,
+	SHDMA_SLAVE_SCIF0_RX,
+	SHDMA_SLAVE_SCIF1_TX,
+	SHDMA_SLAVE_SCIF1_RX,
+	SHDMA_SLAVE_SCIF2_TX,
+	SHDMA_SLAVE_SCIF2_RX,
+	SHDMA_SLAVE_SCIF3_TX,
+	SHDMA_SLAVE_SCIF3_RX,
+	SHDMA_SLAVE_SCIF4_TX,
+	SHDMA_SLAVE_SCIF4_RX,
+	SHDMA_SLAVE_SCIF5_TX,
+	SHDMA_SLAVE_SCIF5_RX,
+	SHDMA_SLAVE_SIUA_TX,
+	SHDMA_SLAVE_SIUA_RX,
+	SHDMA_SLAVE_SIUB_TX,
+	SHDMA_SLAVE_SIUB_RX,
+	SHDMA_SLAVE_NUMBER,	/* Must stay last */
+};
+
+struct sh_dmae_slave_config {
+	enum sh_dmae_slave_chan_id	slave_id;
+	dma_addr_t			addr;
+	u32				chcr;
+	char				mid_rid;
+};
+
+struct sh_dmae_channel {
+	unsigned int	offset;
+	unsigned int	dmars;
+	unsigned int	dmars_bit;
+};
+
+struct sh_dmae_pdata {
+	struct sh_dmae_slave_config *slave;
+	int slave_num;
+	struct sh_dmae_channel *channel;
+	int channel_num;
+	unsigned int ts_low_shift;
+	unsigned int ts_low_mask;
+	unsigned int ts_high_shift;
+	unsigned int ts_high_mask;
+	unsigned int *ts_shift;
+	int ts_shift_num;
+	u16 dmaor_init;
+};
+
+struct device;
+
+/* Used by slave DMA clients to request DMA to/from a specific peripheral */
+struct sh_dmae_slave {
+	enum sh_dmae_slave_chan_id	slave_id; /* Set by the platform */
+	struct device			*dma_dev; /* Set by the platform */
+	struct sh_dmae_slave_config	*config;  /* Set by the driver */
+};
+
+struct sh_dmae_regs {
+	u32 sar; /* SAR / source address */
+	u32 dar; /* DAR / destination address */
+	u32 tcr; /* TCR / transfer count */
+};
+
+struct sh_desc {
+	struct sh_dmae_regs hw;
+	struct list_head node;
+	struct dma_async_tx_descriptor async_tx;
+	enum dma_data_direction direction;
+	dma_cookie_t cookie;
+	size_t partial;
+	int chunks;
+	int mark;
+};
+
+#endif
diff --git a/arch/sh/include/asm/io.h b/arch/sh/include/asm/io.h
index 7dab7b23a5ec..f689554e17c1 100644
--- a/arch/sh/include/asm/io.h
+++ b/arch/sh/include/asm/io.h
@@ -291,21 +291,21 @@ unsigned long long poke_real_address_q(unsigned long long addr,
  * doesn't exist, so everything must go through page tables.
  */
 #ifdef CONFIG_MMU
-void __iomem *__ioremap_caller(unsigned long offset, unsigned long size,
+void __iomem *__ioremap_caller(phys_addr_t offset, unsigned long size,
 			       pgprot_t prot, void *caller);
 void __iounmap(void __iomem *addr);
 
 static inline void __iomem *
-__ioremap(unsigned long offset, unsigned long size, pgprot_t prot)
+__ioremap(phys_addr_t offset, unsigned long size, pgprot_t prot)
 {
 	return __ioremap_caller(offset, size, prot, __builtin_return_address(0));
 }
 
 static inline void __iomem *
-__ioremap_29bit(unsigned long offset, unsigned long size, pgprot_t prot)
+__ioremap_29bit(phys_addr_t offset, unsigned long size, pgprot_t prot)
 {
 #ifdef CONFIG_29BIT
-	unsigned long last_addr = offset + size - 1;
+	phys_addr_t last_addr = offset + size - 1;
 
 	/*
 	 * For P1 and P2 space this is trivial, as everything is already
@@ -329,7 +329,7 @@ __ioremap_29bit(unsigned long offset, unsigned long size, pgprot_t prot)
 }
 
 static inline void __iomem *
-__ioremap_mode(unsigned long offset, unsigned long size, pgprot_t prot)
+__ioremap_mode(phys_addr_t offset, unsigned long size, pgprot_t prot)
 {
 	void __iomem *ret;
 
@@ -349,35 +349,32 @@ __ioremap_mode(unsigned long offset, unsigned long size, pgprot_t prot)
 #define __iounmap(addr)				do { } while (0)
 #endif /* CONFIG_MMU */
 
-static inline void __iomem *
-ioremap(unsigned long offset, unsigned long size)
+static inline void __iomem *ioremap(phys_addr_t offset, unsigned long size)
 {
 	return __ioremap_mode(offset, size, PAGE_KERNEL_NOCACHE);
 }
 
 static inline void __iomem *
-ioremap_cache(unsigned long offset, unsigned long size)
+ioremap_cache(phys_addr_t offset, unsigned long size)
 {
 	return __ioremap_mode(offset, size, PAGE_KERNEL);
 }
 
 #ifdef CONFIG_HAVE_IOREMAP_PROT
 static inline void __iomem *
-ioremap_prot(resource_size_t offset, unsigned long size, unsigned long flags)
+ioremap_prot(phys_addr_t offset, unsigned long size, unsigned long flags)
 {
 	return __ioremap_mode(offset, size, __pgprot(flags));
 }
 #endif
 
 #ifdef CONFIG_IOREMAP_FIXED
-extern void __iomem *ioremap_fixed(resource_size_t, unsigned long,
-				   unsigned long, pgprot_t);
+extern void __iomem *ioremap_fixed(phys_addr_t, unsigned long, pgprot_t);
 extern int iounmap_fixed(void __iomem *);
 extern void ioremap_fixed_init(void);
 #else
 static inline void __iomem *
-ioremap_fixed(resource_size_t phys_addr, unsigned long offset,
-	      unsigned long size, pgprot_t prot)
+ioremap_fixed(phys_addr_t phys_addr, unsigned long size, pgprot_t prot)
 {
 	BUG();
 	return NULL;
diff --git a/arch/sh/include/asm/mmu.h b/arch/sh/include/asm/mmu.h
index 15a05b615ba7..19fe84550b49 100644
--- a/arch/sh/include/asm/mmu.h
+++ b/arch/sh/include/asm/mmu.h
@@ -55,19 +55,29 @@ typedef struct {
 
 #ifdef CONFIG_PMB
 /* arch/sh/mm/pmb.c */
-long pmb_remap(unsigned long virt, unsigned long phys,
-	       unsigned long size, pgprot_t prot);
-void pmb_unmap(unsigned long addr);
-void pmb_init(void);
 bool __in_29bit_mode(void);
+
+void pmb_init(void);
+int pmb_bolt_mapping(unsigned long virt, phys_addr_t phys,
+		     unsigned long size, pgprot_t prot);
+void __iomem *pmb_remap_caller(phys_addr_t phys, unsigned long size,
+			       pgprot_t prot, void *caller);
+int pmb_unmap(void __iomem *addr);
+
 #else
-static inline long pmb_remap(unsigned long virt, unsigned long phys,
-			     unsigned long size, pgprot_t prot)
+
+static inline void __iomem *
+pmb_remap_caller(phys_addr_t phys, unsigned long size,
+		 pgprot_t prot, void *caller)
+{
+	return NULL;
+}
+
+static inline int pmb_unmap(void __iomem *addr)
 {
 	return -EINVAL;
 }
 
-#define pmb_unmap(addr)		do { } while (0)
 #define pmb_init(addr)		do { } while (0)
 
 #ifdef CONFIG_29BIT
@@ -77,6 +87,13 @@ static inline long pmb_remap(unsigned long virt, unsigned long phys,
 #endif
 
 #endif /* CONFIG_PMB */
+
+static inline void __iomem *
+pmb_remap(phys_addr_t phys, unsigned long size, pgprot_t prot)
+{
+	return pmb_remap_caller(phys, size, prot, __builtin_return_address(0));
+}
+
 #endif /* __ASSEMBLY__ */
 
 #endif /* __MMU_H */
diff --git a/arch/sh/include/asm/siu.h b/arch/sh/include/asm/siu.h
index 57565a3b551f..f1b1e6944a5f 100644
--- a/arch/sh/include/asm/siu.h
+++ b/arch/sh/include/asm/siu.h
@@ -11,7 +11,7 @@
 #ifndef ASM_SIU_H
 #define ASM_SIU_H
 
-#include <asm/dma-sh.h>
+#include <asm/dmaengine.h>
 
 struct device;
 
diff --git a/arch/sh/include/asm/topology.h b/arch/sh/include/asm/topology.h
index 37cdadd975ac..88e734069fa6 100644
--- a/arch/sh/include/asm/topology.h
+++ b/arch/sh/include/asm/topology.h
@@ -35,7 +35,7 @@
 
 #define pcibus_to_node(bus)	((void)(bus), -1)
 #define cpumask_of_pcibus(bus)	(pcibus_to_node(bus) == -1 ? \
-					CPU_MASK_ALL_PTR : \
+					cpu_all_mask : \
 					cpumask_of_node(pcibus_to_node(bus)))
 
 #endif
diff --git a/arch/sh/include/cpu-sh3/cpu/dma-register.h b/arch/sh/include/cpu-sh3/cpu/dma-register.h
new file mode 100644
index 000000000000..2349e488c9a6
--- /dev/null
+++ b/arch/sh/include/cpu-sh3/cpu/dma-register.h
@@ -0,0 +1,41 @@
+/*
+ * SH3 CPU-specific DMA definitions, used by both DMA drivers
+ *
+ * Copyright (C) 2010 Guennadi Liakhovetski <g.liakhovetski@gmx.de>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#ifndef CPU_DMA_REGISTER_H
+#define CPU_DMA_REGISTER_H
+
+#define CHCR_TS_LOW_MASK	0x18
+#define CHCR_TS_LOW_SHIFT	3
+#define CHCR_TS_HIGH_MASK	0
+#define CHCR_TS_HIGH_SHIFT	0
+
+#define DMAOR_INIT	DMAOR_DME
+
+/*
+ * The SuperH DMAC supports a number of transmit sizes, we list them here,
+ * with their respective values as they appear in the CHCR registers.
+ */
+enum {
+	XMIT_SZ_8BIT,
+	XMIT_SZ_16BIT,
+	XMIT_SZ_32BIT,
+	XMIT_SZ_128BIT,
+};
+
+/* log2(size / 8) - used to calculate number of transfers */
+#define TS_SHIFT {			\
+	[XMIT_SZ_8BIT]		= 0,	\
+	[XMIT_SZ_16BIT]		= 1,	\
+	[XMIT_SZ_32BIT]		= 2,	\
+	[XMIT_SZ_128BIT]	= 4,	\
+}
+
+#define TS_INDEX2VAL(i)	(((i) & 3) << CHCR_TS_LOW_SHIFT)
+
+#endif
diff --git a/arch/sh/include/cpu-sh3/cpu/dma.h b/arch/sh/include/cpu-sh3/cpu/dma.h
index 207811a7a650..24e28b91c9d5 100644
--- a/arch/sh/include/cpu-sh3/cpu/dma.h
+++ b/arch/sh/include/cpu-sh3/cpu/dma.h
@@ -20,31 +20,4 @@
 #define TS_32		0x00000010
 #define TS_128		0x00000018
 
-#define CHCR_TS_LOW_MASK	0x18
-#define CHCR_TS_LOW_SHIFT	3
-#define CHCR_TS_HIGH_MASK	0
-#define CHCR_TS_HIGH_SHIFT	0
-
-#define DMAOR_INIT	DMAOR_DME
-
-/*
- * The SuperH DMAC supports a number of transmit sizes, we list them here,
- * with their respective values as they appear in the CHCR registers.
- */
-enum {
-	XMIT_SZ_8BIT,
-	XMIT_SZ_16BIT,
-	XMIT_SZ_32BIT,
-	XMIT_SZ_128BIT,
-};
-
-#define TS_SHIFT {			\
-	[XMIT_SZ_8BIT]		= 0,	\
-	[XMIT_SZ_16BIT]		= 1,	\
-	[XMIT_SZ_32BIT]		= 2,	\
-	[XMIT_SZ_128BIT]	= 4,	\
-}
-
-#define TS_INDEX2VAL(i)	(((i) & 3) << CHCR_TS_LOW_SHIFT)
-
 #endif /* __ASM_CPU_SH3_DMA_H */
diff --git a/arch/sh/include/cpu-sh4/cpu/dma-register.h b/arch/sh/include/cpu-sh4/cpu/dma-register.h
new file mode 100644
index 000000000000..55f9fec082d4
--- /dev/null
+++ b/arch/sh/include/cpu-sh4/cpu/dma-register.h
@@ -0,0 +1,112 @@
+/*
+ * SH4 CPU-specific DMA definitions, used by both DMA drivers
+ *
+ * Copyright (C) 2010 Guennadi Liakhovetski <g.liakhovetski@gmx.de>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#ifndef CPU_DMA_REGISTER_H
+#define CPU_DMA_REGISTER_H
+
+/* SH7751/7760/7780 DMA IRQ sources */
+
+#ifdef CONFIG_CPU_SH4A
+
+#define DMAOR_INIT	DMAOR_DME
+
+#if defined(CONFIG_CPU_SUBTYPE_SH7343) || \
+	defined(CONFIG_CPU_SUBTYPE_SH7730)
+#define CHCR_TS_LOW_MASK	0x00000018
+#define CHCR_TS_LOW_SHIFT	3
+#define CHCR_TS_HIGH_MASK	0
+#define CHCR_TS_HIGH_SHIFT	0
+#elif defined(CONFIG_CPU_SUBTYPE_SH7722) || \
+	defined(CONFIG_CPU_SUBTYPE_SH7724)
+#define CHCR_TS_LOW_MASK	0x00000018
+#define CHCR_TS_LOW_SHIFT	3
+#define CHCR_TS_HIGH_MASK	0x00300000
+#define CHCR_TS_HIGH_SHIFT	(20 - 2)	/* 2 bits for shifted low TS */
+#elif defined(CONFIG_CPU_SUBTYPE_SH7763) || \
+	defined(CONFIG_CPU_SUBTYPE_SH7764)
+#define CHCR_TS_LOW_MASK	0x00000018
+#define CHCR_TS_LOW_SHIFT	3
+#define CHCR_TS_HIGH_MASK	0
+#define CHCR_TS_HIGH_SHIFT	0
+#elif defined(CONFIG_CPU_SUBTYPE_SH7723)
+#define CHCR_TS_LOW_MASK	0x00000018
+#define CHCR_TS_LOW_SHIFT	3
+#define CHCR_TS_HIGH_MASK	0
+#define CHCR_TS_HIGH_SHIFT	0
+#elif defined(CONFIG_CPU_SUBTYPE_SH7780)
+#define CHCR_TS_LOW_MASK	0x00000018
+#define CHCR_TS_LOW_SHIFT	3
+#define CHCR_TS_HIGH_MASK	0
+#define CHCR_TS_HIGH_SHIFT	0
+#else /* SH7785 */
+#define CHCR_TS_LOW_MASK	0x00000018
+#define CHCR_TS_LOW_SHIFT	3
+#define CHCR_TS_HIGH_MASK	0
+#define CHCR_TS_HIGH_SHIFT	0
+#endif
+
+/* Transmit sizes and respective CHCR register values */
+enum {
+	XMIT_SZ_8BIT		= 0,
+	XMIT_SZ_16BIT		= 1,
+	XMIT_SZ_32BIT		= 2,
+	XMIT_SZ_64BIT		= 7,
+	XMIT_SZ_128BIT		= 3,
+	XMIT_SZ_256BIT		= 4,
+	XMIT_SZ_128BIT_BLK	= 0xb,
+	XMIT_SZ_256BIT_BLK	= 0xc,
+};
+
+/* log2(size / 8) - used to calculate number of transfers */
+#define TS_SHIFT {			\
+	[XMIT_SZ_8BIT]		= 0,	\
+	[XMIT_SZ_16BIT]		= 1,	\
+	[XMIT_SZ_32BIT]		= 2,	\
+	[XMIT_SZ_64BIT]		= 3,	\
+	[XMIT_SZ_128BIT]	= 4,	\
+	[XMIT_SZ_256BIT]	= 5,	\
+	[XMIT_SZ_128BIT_BLK]	= 4,	\
+	[XMIT_SZ_256BIT_BLK]	= 5,	\
+}
+
+#define TS_INDEX2VAL(i)	((((i) & 3) << CHCR_TS_LOW_SHIFT) | \
+			 ((((i) >> 2) & 3) << CHCR_TS_HIGH_SHIFT))
+
+#else /* CONFIG_CPU_SH4A */
+
+#define DMAOR_INIT	(0x8000 | DMAOR_DME)
+
+#define CHCR_TS_LOW_MASK	0x70
+#define CHCR_TS_LOW_SHIFT	4
+#define CHCR_TS_HIGH_MASK	0
+#define CHCR_TS_HIGH_SHIFT	0
+
+/* Transmit sizes and respective CHCR register values */
+enum {
+	XMIT_SZ_8BIT	= 1,
+	XMIT_SZ_16BIT	= 2,
+	XMIT_SZ_32BIT	= 3,
+	XMIT_SZ_64BIT	= 0,
+	XMIT_SZ_256BIT	= 4,
+};
+
+/* log2(size / 8) - used to calculate number of transfers */
+#define TS_SHIFT {			\
+	[XMIT_SZ_8BIT]		= 0,	\
+	[XMIT_SZ_16BIT]		= 1,	\
+	[XMIT_SZ_32BIT]		= 2,	\
+	[XMIT_SZ_64BIT]		= 3,	\
+	[XMIT_SZ_256BIT]	= 5,	\
+}
+
+#define TS_INDEX2VAL(i)	(((i) & 7) << CHCR_TS_LOW_SHIFT)
+
+#endif /* CONFIG_CPU_SH4A */
+
+#endif
diff --git a/arch/sh/include/cpu-sh4/cpu/dma-sh4a.h b/arch/sh/include/cpu-sh4/cpu/dma-sh4a.h
index e734ea47d8a0..9647e681fd27 100644
--- a/arch/sh/include/cpu-sh4/cpu/dma-sh4a.h
+++ b/arch/sh/include/cpu-sh4/cpu/dma-sh4a.h
@@ -8,20 +8,12 @@
 #define DMAE0_IRQ	78	/* DMA Error IRQ*/
 #define SH_DMAC_BASE0	0xFE008020
 #define SH_DMARS_BASE0	0xFE009000
-#define CHCR_TS_LOW_MASK	0x00000018
-#define CHCR_TS_LOW_SHIFT	3
-#define CHCR_TS_HIGH_MASK	0
-#define CHCR_TS_HIGH_SHIFT	0
 #elif defined(CONFIG_CPU_SUBTYPE_SH7722)
 #define DMTE0_IRQ	48
 #define DMTE4_IRQ	76
 #define DMAE0_IRQ	78	/* DMA Error IRQ*/
 #define SH_DMAC_BASE0	0xFE008020
 #define SH_DMARS_BASE0	0xFE009000
-#define CHCR_TS_LOW_MASK	0x00000018
-#define CHCR_TS_LOW_SHIFT	3
-#define CHCR_TS_HIGH_MASK	0x00300000
-#define CHCR_TS_HIGH_SHIFT	20
 #elif defined(CONFIG_CPU_SUBTYPE_SH7763) || \
 	defined(CONFIG_CPU_SUBTYPE_SH7764)
 #define DMTE0_IRQ	34
@@ -29,10 +21,6 @@
 #define DMAE0_IRQ	38
 #define SH_DMAC_BASE0	0xFF608020
 #define SH_DMARS_BASE0	0xFF609000
-#define CHCR_TS_LOW_MASK	0x00000018
-#define CHCR_TS_LOW_SHIFT	3
-#define CHCR_TS_HIGH_MASK	0
-#define CHCR_TS_HIGH_SHIFT	0
 #elif defined(CONFIG_CPU_SUBTYPE_SH7723)
 #define DMTE0_IRQ	48	/* DMAC0A*/
 #define DMTE4_IRQ	76	/* DMAC0B */
@@ -46,10 +34,6 @@
 #define SH_DMAC_BASE0	0xFE008020
 #define SH_DMAC_BASE1	0xFDC08020
 #define SH_DMARS_BASE0	0xFDC09000
-#define CHCR_TS_LOW_MASK	0x00000018
-#define CHCR_TS_LOW_SHIFT	3
-#define CHCR_TS_HIGH_MASK	0
-#define CHCR_TS_HIGH_SHIFT	0
 #elif defined(CONFIG_CPU_SUBTYPE_SH7724)
 #define DMTE0_IRQ	48	/* DMAC0A*/
 #define DMTE4_IRQ	76	/* DMAC0B */
@@ -64,10 +48,6 @@
 #define SH_DMAC_BASE1	0xFDC08020
 #define SH_DMARS_BASE0	0xFE009000
 #define SH_DMARS_BASE1	0xFDC09000
-#define CHCR_TS_LOW_MASK	0x00000018
-#define CHCR_TS_LOW_SHIFT	3
-#define CHCR_TS_HIGH_MASK	0x00600000
-#define CHCR_TS_HIGH_SHIFT	21
 #elif defined(CONFIG_CPU_SUBTYPE_SH7780)
 #define DMTE0_IRQ	34
 #define DMTE4_IRQ	44
@@ -80,10 +60,6 @@
 #define SH_DMAC_BASE0	0xFC808020
 #define SH_DMAC_BASE1	0xFC818020
 #define SH_DMARS_BASE0	0xFC809000
-#define CHCR_TS_LOW_MASK	0x00000018
-#define CHCR_TS_LOW_SHIFT	3
-#define CHCR_TS_HIGH_MASK	0
-#define CHCR_TS_HIGH_SHIFT	0
 #else /* SH7785 */
 #define DMTE0_IRQ	33
 #define DMTE4_IRQ	37
@@ -97,10 +73,6 @@
 #define SH_DMAC_BASE0	0xFC808020
 #define SH_DMAC_BASE1	0xFCC08020
 #define SH_DMARS_BASE0	0xFC809000
-#define CHCR_TS_LOW_MASK	0x00000018
-#define CHCR_TS_LOW_SHIFT	3
-#define CHCR_TS_HIGH_MASK	0
-#define CHCR_TS_HIGH_SHIFT	0
 #endif
 
 #define REQ_HE		0x000000C0
@@ -108,38 +80,4 @@
 #define REQ_LE		0x00000040
 #define TM_BURST	0x00000020
 
-/*
- * The SuperH DMAC supports a number of transmit sizes, we list them here,
- * with their respective values as they appear in the CHCR registers.
- *
- * Defaults to a 64-bit transfer size.
- */
-enum {
-	XMIT_SZ_8BIT		= 0,
-	XMIT_SZ_16BIT		= 1,
-	XMIT_SZ_32BIT		= 2,
-	XMIT_SZ_64BIT		= 7,
-	XMIT_SZ_128BIT		= 3,
-	XMIT_SZ_256BIT		= 4,
-	XMIT_SZ_128BIT_BLK	= 0xb,
-	XMIT_SZ_256BIT_BLK	= 0xc,
-};
-
-/*
- * The DMA count is defined as the number of bytes to transfer.
- */
-#define TS_SHIFT {			\
-	[XMIT_SZ_8BIT]		= 0,	\
-	[XMIT_SZ_16BIT]		= 1,	\
-	[XMIT_SZ_32BIT]		= 2,	\
-	[XMIT_SZ_64BIT]		= 3,	\
-	[XMIT_SZ_128BIT]	= 4,	\
-	[XMIT_SZ_256BIT]	= 5,	\
-	[XMIT_SZ_128BIT_BLK]	= 4,	\
-	[XMIT_SZ_256BIT_BLK]	= 5,	\
-}
-
-#define TS_INDEX2VAL(i)	((((i) & 3) << CHCR_TS_LOW_SHIFT) | \
-			 ((((i) >> 2) & 3) << CHCR_TS_HIGH_SHIFT))
-
 #endif /* __ASM_SH_CPU_SH4_DMA_SH7780_H */
diff --git a/arch/sh/include/cpu-sh4/cpu/dma.h b/arch/sh/include/cpu-sh4/cpu/dma.h
index 114a369705bc..ca747e93c2ed 100644
--- a/arch/sh/include/cpu-sh4/cpu/dma.h
+++ b/arch/sh/include/cpu-sh4/cpu/dma.h
@@ -5,9 +5,8 @@
 
 #ifdef CONFIG_CPU_SH4A
 
-#define DMAOR_INIT	(DMAOR_DME)
-
 #include <cpu/dma-sh4a.h>
+
 #else /* CONFIG_CPU_SH4A */
 /*
  * SH7750/SH7751/SH7760
@@ -17,7 +16,6 @@
 #define DMTE6_IRQ	46
 #define DMAE0_IRQ	38
 
-#define DMAOR_INIT	(0x8000|DMAOR_DME)
 #define SH_DMAC_BASE0	0xffa00000
 #define SH_DMAC_BASE1	0xffa00070
 /* Definitions for the SuperH DMAC */
@@ -27,40 +25,8 @@
 #define TS_32		0x00000030
 #define TS_64		0x00000000
 
-#define CHCR_TS_LOW_MASK	0x70
-#define CHCR_TS_LOW_SHIFT	4
-#define CHCR_TS_HIGH_MASK	0
-#define CHCR_TS_HIGH_SHIFT	0
-
 #define DMAOR_COD	0x00000008
 
-/*
- * The SuperH DMAC supports a number of transmit sizes, we list them here,
- * with their respective values as they appear in the CHCR registers.
- *
- * Defaults to a 64-bit transfer size.
- */
-enum {
-	XMIT_SZ_8BIT	= 1,
-	XMIT_SZ_16BIT	= 2,
-	XMIT_SZ_32BIT	= 3,
-	XMIT_SZ_64BIT	= 0,
-	XMIT_SZ_256BIT	= 4,
-};
-
-/*
- * The DMA count is defined as the number of bytes to transfer.
- */
-#define TS_SHIFT {			\
-	[XMIT_SZ_8BIT]		= 0,	\
-	[XMIT_SZ_16BIT]		= 1,	\
-	[XMIT_SZ_32BIT]		= 2,	\
-	[XMIT_SZ_64BIT]		= 3,	\
-	[XMIT_SZ_256BIT]	= 5,	\
-}
-
-#define TS_INDEX2VAL(i)	(((i) & 7) << CHCR_TS_LOW_SHIFT)
-
 #endif
 
 #endif /* __ASM_CPU_SH4_DMA_H */
diff --git a/arch/sh/include/mach-migor/mach/migor.h b/arch/sh/include/mach-migor/mach/migor.h
index cee6cb88e020..42fccf93412e 100644
--- a/arch/sh/include/mach-migor/mach/migor.h
+++ b/arch/sh/include/mach-migor/mach/migor.h
@@ -1,6 +1,7 @@
 #ifndef __ASM_SH_MIGOR_H
 #define __ASM_SH_MIGOR_H
 
+#define PORT_MSELCRA 0xa4050180
 #define PORT_MSELCRB 0xa4050182
 #define BSC_CS4BCR 0xfec10010
 #define BSC_CS6ABCR 0xfec1001c
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c
index ef3f97827808..fd7e3639e845 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c
@@ -7,19 +7,167 @@
  * License.  See the file "COPYING" in the main directory of this archive
  * for more details.
  */
-#include <linux/platform_device.h>
 #include <linux/init.h>
+#include <linux/mm.h>
+#include <linux/platform_device.h>
 #include <linux/serial.h>
 #include <linux/serial_sci.h>
-#include <linux/mm.h>
+#include <linux/sh_timer.h>
 #include <linux/uio_driver.h>
 #include <linux/usb/m66592.h>
-#include <linux/sh_timer.h>
+
 #include <asm/clock.h>
+#include <asm/dmaengine.h>
 #include <asm/mmzone.h>
-#include <asm/dma-sh.h>
+#include <asm/siu.h>
+
+#include <cpu/dma-register.h>
 #include <cpu/sh7722.h>
 
+static struct sh_dmae_slave_config sh7722_dmae_slaves[] = {
+	{
+		.slave_id	= SHDMA_SLAVE_SCIF0_TX,
+		.addr		= 0xffe0000c,
+		.chcr		= DM_FIX | SM_INC | 0x800 | TS_INDEX2VAL(XMIT_SZ_8BIT),
+		.mid_rid	= 0x21,
+	}, {
+		.slave_id	= SHDMA_SLAVE_SCIF0_RX,
+		.addr		= 0xffe00014,
+		.chcr		= DM_INC | SM_FIX | 0x800 | TS_INDEX2VAL(XMIT_SZ_8BIT),
+		.mid_rid	= 0x22,
+	}, {
+		.slave_id	= SHDMA_SLAVE_SCIF1_TX,
+		.addr		= 0xffe1000c,
+		.chcr		= DM_FIX | SM_INC | 0x800 | TS_INDEX2VAL(XMIT_SZ_8BIT),
+		.mid_rid	= 0x25,
+	}, {
+		.slave_id	= SHDMA_SLAVE_SCIF1_RX,
+		.addr		= 0xffe10014,
+		.chcr		= DM_INC | SM_FIX | 0x800 | TS_INDEX2VAL(XMIT_SZ_8BIT),
+		.mid_rid	= 0x26,
+	}, {
+		.slave_id	= SHDMA_SLAVE_SCIF2_TX,
+		.addr		= 0xffe2000c,
+		.chcr		= DM_FIX | SM_INC | 0x800 | TS_INDEX2VAL(XMIT_SZ_8BIT),
+		.mid_rid	= 0x29,
+	}, {
+		.slave_id	= SHDMA_SLAVE_SCIF2_RX,
+		.addr		= 0xffe20014,
+		.chcr		= DM_INC | SM_FIX | 0x800 | TS_INDEX2VAL(XMIT_SZ_8BIT),
+		.mid_rid	= 0x2a,
+	}, {
+		.slave_id	= SHDMA_SLAVE_SIUA_TX,
+		.addr		= 0xa454c098,
+		.chcr		= DM_FIX | SM_INC | 0x800 | TS_INDEX2VAL(XMIT_SZ_32BIT),
+		.mid_rid	= 0xb1,
+	}, {
+		.slave_id	= SHDMA_SLAVE_SIUA_RX,
+		.addr		= 0xa454c090,
+		.chcr		= DM_INC | SM_FIX | 0x800 | TS_INDEX2VAL(XMIT_SZ_32BIT),
+		.mid_rid	= 0xb2,
+	}, {
+		.slave_id	= SHDMA_SLAVE_SIUB_TX,
+		.addr		= 0xa454c09c,
+		.chcr		= DM_FIX | SM_INC | 0x800 | TS_INDEX2VAL(XMIT_SZ_32BIT),
+		.mid_rid	= 0xb5,
+	}, {
+		.slave_id	= SHDMA_SLAVE_SIUB_RX,
+		.addr		= 0xa454c094,
+		.chcr		= DM_INC | SM_FIX | 0x800 | TS_INDEX2VAL(XMIT_SZ_32BIT),
+		.mid_rid	= 0xb6,
+	},
+};
+
+static struct sh_dmae_channel sh7722_dmae_channels[] = {
+	{
+		.offset = 0,
+		.dmars = 0,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x10,
+		.dmars = 0,
+		.dmars_bit = 8,
+	}, {
+		.offset = 0x20,
+		.dmars = 4,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x30,
+		.dmars = 4,
+		.dmars_bit = 8,
+	}, {
+		.offset = 0x50,
+		.dmars = 8,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x60,
+		.dmars = 8,
+		.dmars_bit = 8,
+	}
+};
+
+static unsigned int ts_shift[] = TS_SHIFT;
+
+static struct sh_dmae_pdata dma_platform_data = {
+	.slave		= sh7722_dmae_slaves,
+	.slave_num	= ARRAY_SIZE(sh7722_dmae_slaves),
+	.channel	= sh7722_dmae_channels,
+	.channel_num	= ARRAY_SIZE(sh7722_dmae_channels),
+	.ts_low_shift	= CHCR_TS_LOW_SHIFT,
+	.ts_low_mask	= CHCR_TS_LOW_MASK,
+	.ts_high_shift	= CHCR_TS_HIGH_SHIFT,
+	.ts_high_mask	= CHCR_TS_HIGH_MASK,
+	.ts_shift	= ts_shift,
+	.ts_shift_num	= ARRAY_SIZE(ts_shift),
+	.dmaor_init	= DMAOR_INIT,
+};
+
+static struct resource sh7722_dmae_resources[] = {
+	[0] = {
+		/* Channel registers and DMAOR */
+		.start	= 0xfe008020,
+		.end	= 0xfe00808f,
+		.flags	= IORESOURCE_MEM,
+	},
+	[1] = {
+		/* DMARSx */
+		.start	= 0xfe009000,
+		.end	= 0xfe00900b,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		/* DMA error IRQ */
+		.start	= 78,
+		.end	= 78,
+		.flags	= IORESOURCE_IRQ,
+	},
+	{
+		/* IRQ for channels 0-3 */
+		.start	= 48,
+		.end	= 51,
+		.flags	= IORESOURCE_IRQ,
+	},
+	{
+		/* IRQ for channels 4-5 */
+		.start	= 76,
+		.end	= 77,
+		.flags	= IORESOURCE_IRQ,
+	},
+};
+
+struct platform_device dma_device = {
+	.name		= "sh-dma-engine",
+	.id		= -1,
+	.resource	= sh7722_dmae_resources,
+	.num_resources	= ARRAY_SIZE(sh7722_dmae_resources),
+	.dev		= {
+		.platform_data	= &dma_platform_data,
+	},
+	.archdata = {
+		.hwblk_id = HWBLK_DMAC,
+	},
+};
+
 /* Serial */
 static struct plat_sci_port scif0_platform_data = {
 	.mapbase        = 0xffe00000,
@@ -388,15 +536,36 @@ static struct platform_device tmu2_device = {
 	},
 };
 
-static struct sh_dmae_pdata dma_platform_data = {
-	.mode = 0,
+static struct siu_platform siu_platform_data = {
+	.dma_dev	= &dma_device.dev,
+	.dma_slave_tx_a	= SHDMA_SLAVE_SIUA_TX,
+	.dma_slave_rx_a	= SHDMA_SLAVE_SIUA_RX,
+	.dma_slave_tx_b	= SHDMA_SLAVE_SIUB_TX,
+	.dma_slave_rx_b	= SHDMA_SLAVE_SIUB_RX,
 };
 
-static struct platform_device dma_device = {
-	.name		= "sh-dma-engine",
+static struct resource siu_resources[] = {
+	[0] = {
+		.start	= 0xa4540000,
+		.end	= 0xa454c10f,
+		.flags	= IORESOURCE_MEM,
+	},
+	[1] = {
+		.start	= 108,
+		.flags	= IORESOURCE_IRQ,
+	},
+};
+
+static struct platform_device siu_device = {
+	.name		= "sh_siu",
 	.id		= -1,
-	.dev		= {
-		.platform_data	= &dma_platform_data,
+	.dev = {
+		.platform_data	= &siu_platform_data,
+	},
+	.resource	= siu_resources,
+	.num_resources	= ARRAY_SIZE(siu_resources),
+	.archdata = {
+		.hwblk_id = HWBLK_SIU,
 	},
 };
 
@@ -414,6 +583,7 @@ static struct platform_device *sh7722_devices[] __initdata = {
 	&vpu_device,
 	&veu_device,
 	&jpu_device,
+	&siu_device,
 	&dma_device,
 };
 
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7724.c b/arch/sh/kernel/cpu/sh4a/setup-sh7724.c
index 31e3451f7e3d..e7fa2a92fc1f 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7724.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7724.c
@@ -21,22 +21,189 @@
 #include <linux/sh_timer.h>
 #include <linux/io.h>
 #include <linux/notifier.h>
+
 #include <asm/suspend.h>
 #include <asm/clock.h>
-#include <asm/dma-sh.h>
+#include <asm/dmaengine.h>
 #include <asm/mmzone.h>
+
+#include <cpu/dma-register.h>
 #include <cpu/sh7724.h>
 
 /* DMA */
-static struct sh_dmae_pdata dma_platform_data = {
-	.mode = SHDMA_DMAOR1,
+static struct sh_dmae_channel sh7724_dmae0_channels[] = {
+	{
+		.offset = 0,
+		.dmars = 0,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x10,
+		.dmars = 0,
+		.dmars_bit = 8,
+	}, {
+		.offset = 0x20,
+		.dmars = 4,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x30,
+		.dmars = 4,
+		.dmars_bit = 8,
+	}, {
+		.offset = 0x50,
+		.dmars = 8,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x60,
+		.dmars = 8,
+		.dmars_bit = 8,
+	}
+};
+
+static struct sh_dmae_channel sh7724_dmae1_channels[] = {
+	{
+		.offset = 0,
+		.dmars = 0,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x10,
+		.dmars = 0,
+		.dmars_bit = 8,
+	}, {
+		.offset = 0x20,
+		.dmars = 4,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x30,
+		.dmars = 4,
+		.dmars_bit = 8,
+	}, {
+		.offset = 0x50,
+		.dmars = 8,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x60,
+		.dmars = 8,
+		.dmars_bit = 8,
+	}
+};
+
+static unsigned int ts_shift[] = TS_SHIFT;
+
+static struct sh_dmae_pdata dma0_platform_data = {
+	.channel	= sh7724_dmae0_channels,
+	.channel_num	= ARRAY_SIZE(sh7724_dmae0_channels),
+	.ts_low_shift	= CHCR_TS_LOW_SHIFT,
+	.ts_low_mask	= CHCR_TS_LOW_MASK,
+	.ts_high_shift	= CHCR_TS_HIGH_SHIFT,
+	.ts_high_mask	= CHCR_TS_HIGH_MASK,
+	.ts_shift	= ts_shift,
+	.ts_shift_num	= ARRAY_SIZE(ts_shift),
+	.dmaor_init	= DMAOR_INIT,
+};
+
+static struct sh_dmae_pdata dma1_platform_data = {
+	.channel	= sh7724_dmae1_channels,
+	.channel_num	= ARRAY_SIZE(sh7724_dmae1_channels),
+	.ts_low_shift	= CHCR_TS_LOW_SHIFT,
+	.ts_low_mask	= CHCR_TS_LOW_MASK,
+	.ts_high_shift	= CHCR_TS_HIGH_SHIFT,
+	.ts_high_mask	= CHCR_TS_HIGH_MASK,
+	.ts_shift	= ts_shift,
+	.ts_shift_num	= ARRAY_SIZE(ts_shift),
+	.dmaor_init	= DMAOR_INIT,
+};
+
+/* Resource order important! */
+static struct resource sh7724_dmae0_resources[] = {
+	{
+		/* Channel registers and DMAOR */
+		.start	= 0xfe008020,
+		.end	= 0xfe00808f,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		/* DMARSx */
+		.start	= 0xfe009000,
+		.end	= 0xfe00900b,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		/* DMA error IRQ */
+		.start	= 78,
+		.end	= 78,
+		.flags	= IORESOURCE_IRQ,
+	},
+	{
+		/* IRQ for channels 0-3 */
+		.start	= 48,
+		.end	= 51,
+		.flags	= IORESOURCE_IRQ,
+	},
+	{
+		/* IRQ for channels 4-5 */
+		.start	= 76,
+		.end	= 77,
+		.flags	= IORESOURCE_IRQ,
+	},
 };
 
-static struct platform_device dma_device = {
-	.name	= "sh-dma-engine",
-	.id		= -1,
-	.dev	= {
-		.platform_data	= &dma_platform_data,
+/* Resource order important! */
+static struct resource sh7724_dmae1_resources[] = {
+	{
+		/* Channel registers and DMAOR */
+		.start	= 0xfdc08020,
+		.end	= 0xfdc0808f,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		/* DMARSx */
+		.start	= 0xfdc09000,
+		.end	= 0xfdc0900b,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		/* DMA error IRQ */
+		.start	= 74,
+		.end	= 74,
+		.flags	= IORESOURCE_IRQ,
+	},
+	{
+		/* IRQ for channels 0-3 */
+		.start	= 40,
+		.end	= 43,
+		.flags	= IORESOURCE_IRQ,
+	},
+	{
+		/* IRQ for channels 4-5 */
+		.start	= 72,
+		.end	= 73,
+		.flags	= IORESOURCE_IRQ,
+	},
+};
+
+static struct platform_device dma0_device = {
+	.name		= "sh-dma-engine",
+	.id		= 0,
+	.resource	= sh7724_dmae0_resources,
+	.num_resources	= ARRAY_SIZE(sh7724_dmae0_resources),
+	.dev		= {
+		.platform_data	= &dma0_platform_data,
+	},
+	.archdata = {
+		.hwblk_id = HWBLK_DMAC0,
+	},
+};
+
+static struct platform_device dma1_device = {
+	.name		= "sh-dma-engine",
+	.id		= 1,
+	.resource	= sh7724_dmae1_resources,
+	.num_resources	= ARRAY_SIZE(sh7724_dmae1_resources),
+	.dev		= {
+		.platform_data	= &dma1_platform_data,
+	},
+	.archdata = {
+		.hwblk_id = HWBLK_DMAC1,
 	},
 };
 
@@ -663,7 +830,8 @@ static struct platform_device *sh7724_devices[] __initdata = {
 	&tmu3_device,
 	&tmu4_device,
 	&tmu5_device,
-	&dma_device,
+	&dma0_device,
+	&dma1_device,
 	&rtc_device,
 	&iic0_device,
 	&iic1_device,
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7780.c b/arch/sh/kernel/cpu/sh4a/setup-sh7780.c
index f8f21618d785..02e792c90de6 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7780.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7780.c
@@ -13,7 +13,10 @@
 #include <linux/io.h>
 #include <linux/serial_sci.h>
 #include <linux/sh_timer.h>
-#include <asm/dma-sh.h>
+
+#include <asm/dmaengine.h>
+
+#include <cpu/dma-register.h>
 
 static struct plat_sci_port scif0_platform_data = {
 	.mapbase	= 0xffe00000,
@@ -247,15 +250,131 @@ static struct platform_device rtc_device = {
 	.resource	= rtc_resources,
 };
 
-static struct sh_dmae_pdata dma_platform_data = {
-	.mode = (SHDMA_MIX_IRQ | SHDMA_DMAOR1),
+/* DMA */
+static struct sh_dmae_channel sh7780_dmae0_channels[] = {
+	{
+		.offset = 0,
+		.dmars = 0,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x10,
+		.dmars = 0,
+		.dmars_bit = 8,
+	}, {
+		.offset = 0x20,
+		.dmars = 4,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x30,
+		.dmars = 4,
+		.dmars_bit = 8,
+	}, {
+		.offset = 0x50,
+		.dmars = 8,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x60,
+		.dmars = 8,
+		.dmars_bit = 8,
+	}
+};
+
+static struct sh_dmae_channel sh7780_dmae1_channels[] = {
+	{
+		.offset = 0,
+	}, {
+		.offset = 0x10,
+	}, {
+		.offset = 0x20,
+	}, {
+		.offset = 0x30,
+	}, {
+		.offset = 0x50,
+	}, {
+		.offset = 0x60,
+	}
+};
+
+static unsigned int ts_shift[] = TS_SHIFT;
+
+static struct sh_dmae_pdata dma0_platform_data = {
+	.channel	= sh7780_dmae0_channels,
+	.channel_num	= ARRAY_SIZE(sh7780_dmae0_channels),
+	.ts_low_shift	= CHCR_TS_LOW_SHIFT,
+	.ts_low_mask	= CHCR_TS_LOW_MASK,
+	.ts_high_shift	= CHCR_TS_HIGH_SHIFT,
+	.ts_high_mask	= CHCR_TS_HIGH_MASK,
+	.ts_shift	= ts_shift,
+	.ts_shift_num	= ARRAY_SIZE(ts_shift),
+	.dmaor_init	= DMAOR_INIT,
+};
+
+static struct sh_dmae_pdata dma1_platform_data = {
+	.channel	= sh7780_dmae1_channels,
+	.channel_num	= ARRAY_SIZE(sh7780_dmae1_channels),
+	.ts_low_shift	= CHCR_TS_LOW_SHIFT,
+	.ts_low_mask	= CHCR_TS_LOW_MASK,
+	.ts_high_shift	= CHCR_TS_HIGH_SHIFT,
+	.ts_high_mask	= CHCR_TS_HIGH_MASK,
+	.ts_shift	= ts_shift,
+	.ts_shift_num	= ARRAY_SIZE(ts_shift),
+	.dmaor_init	= DMAOR_INIT,
 };
 
-static struct platform_device dma_device = {
+static struct resource sh7780_dmae0_resources[] = {
+	[0] = {
+		/* Channel registers and DMAOR */
+		.start	= 0xfc808020,
+		.end	= 0xfc80808f,
+		.flags	= IORESOURCE_MEM,
+	},
+	[1] = {
+		/* DMARSx */
+		.start	= 0xfc809000,
+		.end	= 0xfc80900b,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		/* Real DMA error IRQ is 38, and channel IRQs are 34-37, 44-45 */
+		.start	= 34,
+		.end	= 34,
+		.flags	= IORESOURCE_IRQ | IORESOURCE_IRQ_SHAREABLE,
+	},
+};
+
+static struct resource sh7780_dmae1_resources[] = {
+	[0] = {
+		/* Channel registers and DMAOR */
+		.start	= 0xfc818020,
+		.end	= 0xfc81808f,
+		.flags	= IORESOURCE_MEM,
+	},
+	/* DMAC1 has no DMARS */
+	{
+		/* Real DMA error IRQ is 38, and channel IRQs are 46-47, 92-95 */
+		.start	= 46,
+		.end	= 46,
+		.flags	= IORESOURCE_IRQ | IORESOURCE_IRQ_SHAREABLE,
+	},
+};
+
+static struct platform_device dma0_device = {
 	.name           = "sh-dma-engine",
-	.id             = -1,
+	.id             = 0,
+	.resource	= sh7780_dmae0_resources,
+	.num_resources	= ARRAY_SIZE(sh7780_dmae0_resources),
 	.dev            = {
-		.platform_data  = &dma_platform_data,
+		.platform_data	= &dma0_platform_data,
+	},
+};
+
+static struct platform_device dma1_device = {
+	.name		= "sh-dma-engine",
+	.id		= 1,
+	.resource	= sh7780_dmae1_resources,
+	.num_resources	= ARRAY_SIZE(sh7780_dmae1_resources),
+	.dev		= {
+		.platform_data	= &dma1_platform_data,
 	},
 };
 
@@ -269,7 +388,8 @@ static struct platform_device *sh7780_devices[] __initdata = {
 	&tmu4_device,
 	&tmu5_device,
 	&rtc_device,
-	&dma_device,
+	&dma0_device,
+	&dma1_device,
 };
 
 static int __init sh7780_devices_setup(void)
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7785.c b/arch/sh/kernel/cpu/sh4a/setup-sh7785.c
index 23448d8c6711..1fcd88b1671e 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7785.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7785.c
@@ -14,9 +14,12 @@
 #include <linux/io.h>
 #include <linux/mm.h>
 #include <linux/sh_timer.h>
-#include <asm/dma-sh.h>
+
+#include <asm/dmaengine.h>
 #include <asm/mmzone.h>
 
+#include <cpu/dma-register.h>
+
 static struct plat_sci_port scif0_platform_data = {
 	.mapbase	= 0xffea0000,
 	.flags		= UPF_BOOT_AUTOCONF,
@@ -295,15 +298,131 @@ static struct platform_device tmu5_device = {
 	.num_resources	= ARRAY_SIZE(tmu5_resources),
 };
 
-static struct sh_dmae_pdata dma_platform_data = {
-	.mode = (SHDMA_MIX_IRQ | SHDMA_DMAOR1),
+/* DMA */
+static struct sh_dmae_channel sh7785_dmae0_channels[] = {
+	{
+		.offset = 0,
+		.dmars = 0,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x10,
+		.dmars = 0,
+		.dmars_bit = 8,
+	}, {
+		.offset = 0x20,
+		.dmars = 4,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x30,
+		.dmars = 4,
+		.dmars_bit = 8,
+	}, {
+		.offset = 0x50,
+		.dmars = 8,
+		.dmars_bit = 0,
+	}, {
+		.offset = 0x60,
+		.dmars = 8,
+		.dmars_bit = 8,
+	}
+};
+
+static struct sh_dmae_channel sh7785_dmae1_channels[] = {
+	{
+		.offset = 0,
+	}, {
+		.offset = 0x10,
+	}, {
+		.offset = 0x20,
+	}, {
+		.offset = 0x30,
+	}, {
+		.offset = 0x50,
+	}, {
+		.offset = 0x60,
+	}
+};
+
+static unsigned int ts_shift[] = TS_SHIFT;
+
+static struct sh_dmae_pdata dma0_platform_data = {
+	.channel	= sh7785_dmae0_channels,
+	.channel_num	= ARRAY_SIZE(sh7785_dmae0_channels),
+	.ts_low_shift	= CHCR_TS_LOW_SHIFT,
+	.ts_low_mask	= CHCR_TS_LOW_MASK,
+	.ts_high_shift	= CHCR_TS_HIGH_SHIFT,
+	.ts_high_mask	= CHCR_TS_HIGH_MASK,
+	.ts_shift	= ts_shift,
+	.ts_shift_num	= ARRAY_SIZE(ts_shift),
+	.dmaor_init	= DMAOR_INIT,
+};
+
+static struct sh_dmae_pdata dma1_platform_data = {
+	.channel	= sh7785_dmae1_channels,
+	.channel_num	= ARRAY_SIZE(sh7785_dmae1_channels),
+	.ts_low_shift	= CHCR_TS_LOW_SHIFT,
+	.ts_low_mask	= CHCR_TS_LOW_MASK,
+	.ts_high_shift	= CHCR_TS_HIGH_SHIFT,
+	.ts_high_mask	= CHCR_TS_HIGH_MASK,
+	.ts_shift	= ts_shift,
+	.ts_shift_num	= ARRAY_SIZE(ts_shift),
+	.dmaor_init	= DMAOR_INIT,
 };
 
-static struct platform_device dma_device = {
+static struct resource sh7785_dmae0_resources[] = {
+	[0] = {
+		/* Channel registers and DMAOR */
+		.start	= 0xfc808020,
+		.end	= 0xfc80808f,
+		.flags	= IORESOURCE_MEM,
+	},
+	[1] = {
+		/* DMARSx */
+		.start	= 0xfc809000,
+		.end	= 0xfc80900b,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		/* Real DMA error IRQ is 39, and channel IRQs are 33-38 */
+		.start	= 33,
+		.end	= 33,
+		.flags	= IORESOURCE_IRQ | IORESOURCE_IRQ_SHAREABLE,
+	},
+};
+
+static struct resource sh7785_dmae1_resources[] = {
+	[0] = {
+		/* Channel registers and DMAOR */
+		.start	= 0xfcc08020,
+		.end	= 0xfcc0808f,
+		.flags	= IORESOURCE_MEM,
+	},
+	/* DMAC1 has no DMARS */
+	{
+		/* Real DMA error IRQ is 58, and channel IRQs are 52-57 */
+		.start	= 52,
+		.end	= 52,
+		.flags	= IORESOURCE_IRQ | IORESOURCE_IRQ_SHAREABLE,
+	},
+};
+
+static struct platform_device dma0_device = {
 	.name           = "sh-dma-engine",
-	.id             = -1,
+	.id             = 0,
+	.resource	= sh7785_dmae0_resources,
+	.num_resources	= ARRAY_SIZE(sh7785_dmae0_resources),
 	.dev            = {
-		.platform_data  = &dma_platform_data,
+		.platform_data	= &dma0_platform_data,
+	},
+};
+
+static struct platform_device dma1_device = {
+	.name		= "sh-dma-engine",
+	.id		= 1,
+	.resource	= sh7785_dmae1_resources,
+	.num_resources	= ARRAY_SIZE(sh7785_dmae1_resources),
+	.dev		= {
+		.platform_data	= &dma1_platform_data,
 	},
 };
 
@@ -320,7 +439,8 @@ static struct platform_device *sh7785_devices[] __initdata = {
 	&tmu3_device,
 	&tmu4_device,
 	&tmu5_device,
-	&dma_device,
+	&dma0_device,
+	&dma1_device,
 };
 
 static int __init sh7785_devices_setup(void)
diff --git a/arch/sh/kernel/hw_breakpoint.c b/arch/sh/kernel/hw_breakpoint.c
index e2f1753d275c..675eea7785d9 100644
--- a/arch/sh/kernel/hw_breakpoint.c
+++ b/arch/sh/kernel/hw_breakpoint.c
@@ -143,26 +143,6 @@ static int arch_check_va_in_kernelspace(unsigned long va, u8 hbp_len)
 	return (va >= TASK_SIZE) && ((va + len - 1) >= TASK_SIZE);
 }
 
-/*
- * Store a breakpoint's encoded address, length, and type.
- */
-static int arch_store_info(struct perf_event *bp)
-{
-	struct arch_hw_breakpoint *info = counter_arch_bp(bp);
-
-	/*
-	 * User-space requests will always have the address field populated
-	 * For kernel-addresses, either the address or symbol name can be
-	 * specified.
-	 */
-	if (info->name)
-		info->address = (unsigned long)kallsyms_lookup_name(info->name);
-	if (info->address)
-		return 0;
-
-	return -EINVAL;
-}
-
 int arch_bp_generic_fields(int sh_len, int sh_type,
 			   int *gen_len, int *gen_type)
 {
@@ -276,10 +256,12 @@ int arch_validate_hwbkpt_settings(struct perf_event *bp,
 		return ret;
 	}
 
-	ret = arch_store_info(bp);
-
-	if (ret < 0)
-		return ret;
+	/*
+	 * For kernel-addresses, either the address or symbol name can be
+	 * specified.
+	 */
+	if (info->name)
+		info->address = (unsigned long)kallsyms_lookup_name(info->name);
 
 	/*
 	 * Check that the low-order bits of the address are appropriate
diff --git a/arch/sh/kernel/setup.c b/arch/sh/kernel/setup.c
index 3459e70eed72..8870d6ba64bf 100644
--- a/arch/sh/kernel/setup.c
+++ b/arch/sh/kernel/setup.c
@@ -443,7 +443,7 @@ void __init setup_arch(char **cmdline_p)
 
 	nodes_clear(node_online_map);
 
-	/* Setup bootmem with available RAM */
+	pmb_init();
 	lmb_init();
 	setup_memory();
 	sparse_init();
@@ -452,7 +452,6 @@ void __init setup_arch(char **cmdline_p)
 	conswitchp = &dummy_con;
 #endif
 	paging_init();
-	pmb_init();
 
 	ioremap_fixed_init();
 
diff --git a/arch/sh/kernel/time.c b/arch/sh/kernel/time.c
index 953fa1613312..8a0072de2bcc 100644
--- a/arch/sh/kernel/time.c
+++ b/arch/sh/kernel/time.c
@@ -39,12 +39,12 @@ static int null_rtc_set_time(const time_t secs)
 void (*rtc_sh_get_time)(struct timespec *) = null_rtc_get_time;
 int (*rtc_sh_set_time)(const time_t) = null_rtc_set_time;
 
-#ifdef CONFIG_GENERIC_CMOS_UPDATE
 void read_persistent_clock(struct timespec *ts)
 {
 	rtc_sh_get_time(ts);
 }
 
+#ifdef CONFIG_GENERIC_CMOS_UPDATE
 int update_persistent_clock(struct timespec now)
 {
 	return rtc_sh_set_time(now.tv_sec);
@@ -113,9 +113,5 @@ void __init time_init(void)
 	hwblk_init();
 	clk_init();
 
-	rtc_sh_get_time(&xtime);
-	set_normalized_timespec(&wall_to_monotonic,
-				-xtime.tv_sec, -xtime.tv_nsec);
-
 	late_time_init = sh_late_time_init;
 }
diff --git a/arch/sh/lib/libgcc.h b/arch/sh/lib/libgcc.h
index 3f19d1c5d942..05909d58e2fe 100644
--- a/arch/sh/lib/libgcc.h
+++ b/arch/sh/lib/libgcc.h
@@ -17,8 +17,7 @@ struct DWstruct {
 #error I feel sick.
 #endif
 
-typedef union
-{
+typedef union {
 	struct DWstruct s;
 	long long ll;
 } DWunion;
diff --git a/arch/sh/mm/ioremap.c b/arch/sh/mm/ioremap.c
index c68d2d7d00a9..1ab2385ecefe 100644
--- a/arch/sh/mm/ioremap.c
+++ b/arch/sh/mm/ioremap.c
@@ -34,11 +34,12 @@
  * caller shouldn't need to know that small detail.
  */
 void __iomem * __init_refok
-__ioremap_caller(unsigned long phys_addr, unsigned long size,
+__ioremap_caller(phys_addr_t phys_addr, unsigned long size,
 		 pgprot_t pgprot, void *caller)
 {
 	struct vm_struct *area;
 	unsigned long offset, last_addr, addr, orig_addr;
+	void __iomem *mapped;
 
 	/* Don't allow wraparound or zero size */
 	last_addr = phys_addr + size - 1;
@@ -46,6 +47,20 @@ __ioremap_caller(unsigned long phys_addr, unsigned long size,
 		return NULL;
 
 	/*
+	 * If we can't yet use the regular approach, go the fixmap route.
+	 */
+	if (!mem_init_done)
+		return ioremap_fixed(phys_addr, size, pgprot);
+
+	/*
+	 * First try to remap through the PMB.
+	 * PMB entries are all pre-faulted.
+	 */
+	mapped = pmb_remap_caller(phys_addr, size, pgprot, caller);
+	if (mapped && !IS_ERR(mapped))
+		return mapped;
+
+	/*
 	 * Mappings have to be page-aligned
 	 */
 	offset = phys_addr & ~PAGE_MASK;
@@ -53,12 +68,6 @@ __ioremap_caller(unsigned long phys_addr, unsigned long size,
 	size = PAGE_ALIGN(last_addr+1) - phys_addr;
 
 	/*
-	 * If we can't yet use the regular approach, go the fixmap route.
-	 */
-	if (!mem_init_done)
-		return ioremap_fixed(phys_addr, offset, size, pgprot);
-
-	/*
 	 * Ok, go for it..
 	 */
 	area = get_vm_area_caller(size, VM_IOREMAP, caller);
@@ -67,33 +76,10 @@ __ioremap_caller(unsigned long phys_addr, unsigned long size,
 	area->phys_addr = phys_addr;
 	orig_addr = addr = (unsigned long)area->addr;
 
-#ifdef CONFIG_PMB
-	/*
-	 * First try to remap through the PMB once a valid VMA has been
-	 * established. Smaller allocations (or the rest of the size
-	 * remaining after a PMB mapping due to the size not being
-	 * perfectly aligned on a PMB size boundary) are then mapped
-	 * through the UTLB using conventional page tables.
-	 *
-	 * PMB entries are all pre-faulted.
-	 */
-	if (unlikely(phys_addr >= P1SEG)) {
-		unsigned long mapped;
-
-		mapped = pmb_remap(addr, phys_addr, size, pgprot);
-		if (likely(mapped)) {
-			addr		+= mapped;
-			phys_addr	+= mapped;
-			size		-= mapped;
-		}
+	if (ioremap_page_range(addr, addr + size, phys_addr, pgprot)) {
+		vunmap((void *)orig_addr);
+		return NULL;
 	}
-#endif
-
-	if (likely(size))
-		if (ioremap_page_range(addr, addr + size, phys_addr, pgprot)) {
-			vunmap((void *)orig_addr);
-			return NULL;
-		}
 
 	return (void __iomem *)(offset + (char *)orig_addr);
 }
@@ -133,23 +119,11 @@ void __iounmap(void __iomem *addr)
 	if (iounmap_fixed(addr) == 0)
 		return;
 
-#ifdef CONFIG_PMB
 	/*
-	 * Purge any PMB entries that may have been established for this
-	 * mapping, then proceed with conventional VMA teardown.
-	 *
-	 * XXX: Note that due to the way that remove_vm_area() does
-	 * matching of the resultant VMA, we aren't able to fast-forward
-	 * the address past the PMB space until the end of the VMA where
-	 * the page tables reside. As such, unmap_vm_area() will be
-	 * forced to linearly scan over the area until it finds the page
-	 * tables where PTEs that need to be unmapped actually reside,
-	 * which is far from optimal. Perhaps we need to use a separate
-	 * VMA for the PMB mappings?
-	 *					-- PFM.
+	 * If the PMB handled it, there's nothing else to do.
 	 */
-	pmb_unmap(vaddr);
-#endif
+	if (pmb_unmap(addr) == 0)
+		return;
 
 	p = remove_vm_area((void *)(vaddr & PAGE_MASK));
 	if (!p) {
diff --git a/arch/sh/mm/ioremap_fixed.c b/arch/sh/mm/ioremap_fixed.c
index 0b78b1e20ef1..7f682e5dafcf 100644
--- a/arch/sh/mm/ioremap_fixed.c
+++ b/arch/sh/mm/ioremap_fixed.c
@@ -45,14 +45,21 @@ void __init ioremap_fixed_init(void)
 }
 
 void __init __iomem *
-ioremap_fixed(resource_size_t phys_addr, unsigned long offset,
-	      unsigned long size, pgprot_t prot)
+ioremap_fixed(phys_addr_t phys_addr, unsigned long size, pgprot_t prot)
 {
 	enum fixed_addresses idx0, idx;
 	struct ioremap_map *map;
 	unsigned int nrpages;
+	unsigned long offset;
 	int i, slot;
 
+	/*
+	 * Mappings have to be page-aligned
+	 */
+	offset = phys_addr & ~PAGE_MASK;
+	phys_addr &= PAGE_MASK;
+	size = PAGE_ALIGN(phys_addr + size) - phys_addr;
+
 	slot = -1;
 	for (i = 0; i < FIX_N_IOREMAPS; i++) {
 		map = &ioremap_maps[i];
diff --git a/arch/sh/mm/numa.c b/arch/sh/mm/numa.c
index 422e92721878..961b34085e3b 100644
--- a/arch/sh/mm/numa.c
+++ b/arch/sh/mm/numa.c
@@ -74,6 +74,9 @@ void __init setup_bootmem_node(int nid, unsigned long start, unsigned long end)
 	start_pfn = start >> PAGE_SHIFT;
 	end_pfn = end >> PAGE_SHIFT;
 
+	pmb_bolt_mapping((unsigned long)__va(start), start, end - start,
+			 PAGE_KERNEL);
+
 	lmb_add(start, end - start);
 
 	__add_active_range(nid, start_pfn, end_pfn);
diff --git a/arch/sh/mm/pmb.c b/arch/sh/mm/pmb.c
index 198bcff5e96f..a4662e2782c3 100644
--- a/arch/sh/mm/pmb.c
+++ b/arch/sh/mm/pmb.c
@@ -23,7 +23,8 @@
 #include <linux/err.h>
 #include <linux/io.h>
 #include <linux/spinlock.h>
-#include <linux/rwlock.h>
+#include <linux/vmalloc.h>
+#include <asm/cacheflush.h>
 #include <asm/sizes.h>
 #include <asm/system.h>
 #include <asm/uaccess.h>
@@ -52,12 +53,24 @@ struct pmb_entry {
 	struct pmb_entry *link;
 };
 
+static struct {
+	unsigned long size;
+	int flag;
+} pmb_sizes[] = {
+	{ .size	= SZ_512M, .flag = PMB_SZ_512M, },
+	{ .size = SZ_128M, .flag = PMB_SZ_128M, },
+	{ .size = SZ_64M,  .flag = PMB_SZ_64M,  },
+	{ .size = SZ_16M,  .flag = PMB_SZ_16M,  },
+};
+
 static void pmb_unmap_entry(struct pmb_entry *, int depth);
 
 static DEFINE_RWLOCK(pmb_rwlock);
 static struct pmb_entry pmb_entry_list[NR_PMB_ENTRIES];
 static DECLARE_BITMAP(pmb_map, NR_PMB_ENTRIES);
 
+static unsigned int pmb_iomapping_enabled;
+
 static __always_inline unsigned long mk_pmb_entry(unsigned int entry)
 {
 	return (entry & PMB_E_MASK) << PMB_E_SHIFT;
@@ -73,6 +86,142 @@ static __always_inline unsigned long mk_pmb_data(unsigned int entry)
 	return mk_pmb_entry(entry) | PMB_DATA;
 }
 
+static __always_inline unsigned int pmb_ppn_in_range(unsigned long ppn)
+{
+	return ppn >= __pa(memory_start) && ppn < __pa(memory_end);
+}
+
+/*
+ * Ensure that the PMB entries match our cache configuration.
+ *
+ * When we are in 32-bit address extended mode, CCR.CB becomes
+ * invalid, so care must be taken to manually adjust cacheable
+ * translations.
+ */
+static __always_inline unsigned long pmb_cache_flags(void)
+{
+	unsigned long flags = 0;
+
+#if defined(CONFIG_CACHE_OFF)
+	flags |= PMB_WT | PMB_UB;
+#elif defined(CONFIG_CACHE_WRITETHROUGH)
+	flags |= PMB_C | PMB_WT | PMB_UB;
+#elif defined(CONFIG_CACHE_WRITEBACK)
+	flags |= PMB_C;
+#endif
+
+	return flags;
+}
+
+/*
+ * Convert typical pgprot value to the PMB equivalent
+ */
+static inline unsigned long pgprot_to_pmb_flags(pgprot_t prot)
+{
+	unsigned long pmb_flags = 0;
+	u64 flags = pgprot_val(prot);
+
+	if (flags & _PAGE_CACHABLE)
+		pmb_flags |= PMB_C;
+	if (flags & _PAGE_WT)
+		pmb_flags |= PMB_WT | PMB_UB;
+
+	return pmb_flags;
+}
+
+static inline bool pmb_can_merge(struct pmb_entry *a, struct pmb_entry *b)
+{
+	return (b->vpn == (a->vpn + a->size)) &&
+	       (b->ppn == (a->ppn + a->size)) &&
+	       (b->flags == a->flags);
+}
+
+static bool pmb_mapping_exists(unsigned long vaddr, phys_addr_t phys,
+			       unsigned long size)
+{
+	int i;
+
+	read_lock(&pmb_rwlock);
+
+	for (i = 0; i < ARRAY_SIZE(pmb_entry_list); i++) {
+		struct pmb_entry *pmbe, *iter;
+		unsigned long span;
+
+		if (!test_bit(i, pmb_map))
+			continue;
+
+		pmbe = &pmb_entry_list[i];
+
+		/*
+		 * See if VPN and PPN are bounded by an existing mapping.
+		 */
+		if ((vaddr < pmbe->vpn) || (vaddr >= (pmbe->vpn + pmbe->size)))
+			continue;
+		if ((phys < pmbe->ppn) || (phys >= (pmbe->ppn + pmbe->size)))
+			continue;
+
+		/*
+		 * Now see if we're in range of a simple mapping.
+		 */
+		if (size <= pmbe->size) {
+			read_unlock(&pmb_rwlock);
+			return true;
+		}
+
+		span = pmbe->size;
+
+		/*
+		 * Finally for sizes that involve compound mappings, walk
+		 * the chain.
+		 */
+		for (iter = pmbe->link; iter; iter = iter->link)
+			span += iter->size;
+
+		/*
+		 * Nothing else to do if the range requirements are met.
+		 */
+		if (size <= span) {
+			read_unlock(&pmb_rwlock);
+			return true;
+		}
+	}
+
+	read_unlock(&pmb_rwlock);
+	return false;
+}
+
+static bool pmb_size_valid(unsigned long size)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(pmb_sizes); i++)
+		if (pmb_sizes[i].size == size)
+			return true;
+
+	return false;
+}
+
+static inline bool pmb_addr_valid(unsigned long addr, unsigned long size)
+{
+	return (addr >= P1SEG && (addr + size - 1) < P3SEG);
+}
+
+static inline bool pmb_prot_valid(pgprot_t prot)
+{
+	return (pgprot_val(prot) & _PAGE_USER) == 0;
+}
+
+static int pmb_size_to_flags(unsigned long size)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(pmb_sizes); i++)
+		if (pmb_sizes[i].size == size)
+			return pmb_sizes[i].flag;
+
+	return 0;
+}
+
 static int pmb_alloc_entry(void)
 {
 	int pos;
@@ -140,33 +289,22 @@ static void pmb_free(struct pmb_entry *pmbe)
 }
 
 /*
- * Ensure that the PMB entries match our cache configuration.
- *
- * When we are in 32-bit address extended mode, CCR.CB becomes
- * invalid, so care must be taken to manually adjust cacheable
- * translations.
+ * Must be run uncached.
  */
-static __always_inline unsigned long pmb_cache_flags(void)
+static void __set_pmb_entry(struct pmb_entry *pmbe)
 {
-	unsigned long flags = 0;
+	unsigned long addr, data;
 
-#if defined(CONFIG_CACHE_WRITETHROUGH)
-	flags |= PMB_C | PMB_WT | PMB_UB;
-#elif defined(CONFIG_CACHE_WRITEBACK)
-	flags |= PMB_C;
-#endif
+	addr = mk_pmb_addr(pmbe->entry);
+	data = mk_pmb_data(pmbe->entry);
 
-	return flags;
-}
+	jump_to_uncached();
 
-/*
- * Must be run uncached.
- */
-static void __set_pmb_entry(struct pmb_entry *pmbe)
-{
-	writel_uncached(pmbe->vpn | PMB_V, mk_pmb_addr(pmbe->entry));
-	writel_uncached(pmbe->ppn | pmbe->flags | PMB_V,
-			mk_pmb_data(pmbe->entry));
+	/* Set V-bit */
+	__raw_writel(pmbe->vpn | PMB_V, addr);
+	__raw_writel(pmbe->ppn | pmbe->flags | PMB_V, data);
+
+	back_to_cached();
 }
 
 static void __clear_pmb_entry(struct pmb_entry *pmbe)
@@ -194,144 +332,155 @@ static void set_pmb_entry(struct pmb_entry *pmbe)
 	spin_unlock_irqrestore(&pmbe->lock, flags);
 }
 
-static struct {
-	unsigned long size;
-	int flag;
-} pmb_sizes[] = {
-	{ .size	= SZ_512M, .flag = PMB_SZ_512M, },
-	{ .size = SZ_128M, .flag = PMB_SZ_128M, },
-	{ .size = SZ_64M,  .flag = PMB_SZ_64M,  },
-	{ .size = SZ_16M,  .flag = PMB_SZ_16M,  },
-};
-
-long pmb_remap(unsigned long vaddr, unsigned long phys,
-	       unsigned long size, pgprot_t prot)
+int pmb_bolt_mapping(unsigned long vaddr, phys_addr_t phys,
+		     unsigned long size, pgprot_t prot)
 {
 	struct pmb_entry *pmbp, *pmbe;
-	unsigned long wanted;
-	int pmb_flags, i;
-	long err;
-	u64 flags;
+	unsigned long orig_addr, orig_size;
+	unsigned long flags, pmb_flags;
+	int i, mapped;
 
-	flags = pgprot_val(prot);
+	if (!pmb_addr_valid(vaddr, size))
+		return -EFAULT;
+	if (pmb_mapping_exists(vaddr, phys, size))
+		return 0;
 
-	pmb_flags = PMB_WT | PMB_UB;
-
-	/* Convert typical pgprot value to the PMB equivalent */
-	if (flags & _PAGE_CACHABLE) {
-		pmb_flags |= PMB_C;
+	orig_addr = vaddr;
+	orig_size = size;
 
-		if ((flags & _PAGE_WT) == 0)
-			pmb_flags &= ~(PMB_WT | PMB_UB);
-	}
+	flush_tlb_kernel_range(vaddr, vaddr + size);
 
+	pmb_flags = pgprot_to_pmb_flags(prot);
 	pmbp = NULL;
-	wanted = size;
 
-again:
-	for (i = 0; i < ARRAY_SIZE(pmb_sizes); i++) {
-		unsigned long flags;
+	do {
+		for (i = mapped = 0; i < ARRAY_SIZE(pmb_sizes); i++) {
+			if (size < pmb_sizes[i].size)
+				continue;
+
+			pmbe = pmb_alloc(vaddr, phys, pmb_flags |
+					 pmb_sizes[i].flag, PMB_NO_ENTRY);
+			if (IS_ERR(pmbe)) {
+				pmb_unmap_entry(pmbp, mapped);
+				return PTR_ERR(pmbe);
+			}
 
-		if (size < pmb_sizes[i].size)
-			continue;
+			spin_lock_irqsave(&pmbe->lock, flags);
 
-		pmbe = pmb_alloc(vaddr, phys, pmb_flags | pmb_sizes[i].flag,
-				 PMB_NO_ENTRY);
-		if (IS_ERR(pmbe)) {
-			err = PTR_ERR(pmbe);
-			goto out;
-		}
+			pmbe->size = pmb_sizes[i].size;
 
-		spin_lock_irqsave(&pmbe->lock, flags);
+			__set_pmb_entry(pmbe);
 
-		__set_pmb_entry(pmbe);
+			phys	+= pmbe->size;
+			vaddr	+= pmbe->size;
+			size	-= pmbe->size;
 
-		phys	+= pmb_sizes[i].size;
-		vaddr	+= pmb_sizes[i].size;
-		size	-= pmb_sizes[i].size;
+			/*
+			 * Link adjacent entries that span multiple PMB
+			 * entries for easier tear-down.
+			 */
+			if (likely(pmbp)) {
+				spin_lock(&pmbp->lock);
+				pmbp->link = pmbe;
+				spin_unlock(&pmbp->lock);
+			}
 
-		pmbe->size = pmb_sizes[i].size;
+			pmbp = pmbe;
 
-		/*
-		 * Link adjacent entries that span multiple PMB entries
-		 * for easier tear-down.
-		 */
-		if (likely(pmbp)) {
-			spin_lock(&pmbp->lock);
-			pmbp->link = pmbe;
-			spin_unlock(&pmbp->lock);
+			/*
+			 * Instead of trying smaller sizes on every
+			 * iteration (even if we succeed in allocating
+			 * space), try using pmb_sizes[i].size again.
+			 */
+			i--;
+			mapped++;
+
+			spin_unlock_irqrestore(&pmbe->lock, flags);
 		}
+	} while (size >= SZ_16M);
 
-		pmbp = pmbe;
+	flush_cache_vmap(orig_addr, orig_addr + orig_size);
 
-		/*
-		 * Instead of trying smaller sizes on every iteration
-		 * (even if we succeed in allocating space), try using
-		 * pmb_sizes[i].size again.
-		 */
-		i--;
+	return 0;
+}
 
-		spin_unlock_irqrestore(&pmbe->lock, flags);
-	}
+void __iomem *pmb_remap_caller(phys_addr_t phys, unsigned long size,
+			       pgprot_t prot, void *caller)
+{
+	unsigned long vaddr;
+	phys_addr_t offset, last_addr;
+	phys_addr_t align_mask;
+	unsigned long aligned;
+	struct vm_struct *area;
+	int i, ret;
 
-	if (size >= SZ_16M)
-		goto again;
+	if (!pmb_iomapping_enabled)
+		return NULL;
 
-	return wanted - size;
+	/*
+	 * Small mappings need to go through the TLB.
+	 */
+	if (size < SZ_16M)
+		return ERR_PTR(-EINVAL);
+	if (!pmb_prot_valid(prot))
+		return ERR_PTR(-EINVAL);
 
-out:
-	pmb_unmap_entry(pmbp, NR_PMB_ENTRIES);
+	for (i = 0; i < ARRAY_SIZE(pmb_sizes); i++)
+		if (size >= pmb_sizes[i].size)
+			break;
+
+	last_addr = phys + size;
+	align_mask = ~(pmb_sizes[i].size - 1);
+	offset = phys & ~align_mask;
+	phys &= align_mask;
+	aligned = ALIGN(last_addr, pmb_sizes[i].size) - phys;
+
+	/*
+	 * XXX: This should really start from uncached_end, but this
+	 * causes the MMU to reset, so for now we restrict it to the
+	 * 0xb000...0xc000 range.
+	 */
+	area = __get_vm_area_caller(aligned, VM_IOREMAP, 0xb0000000,
+				    P3SEG, caller);
+	if (!area)
+		return NULL;
+
+	area->phys_addr = phys;
+	vaddr = (unsigned long)area->addr;
+
+	ret = pmb_bolt_mapping(vaddr, phys, size, prot);
+	if (unlikely(ret != 0))
+		return ERR_PTR(ret);
 
-	return err;
+	return (void __iomem *)(offset + (char *)vaddr);
 }
 
-void pmb_unmap(unsigned long addr)
+int pmb_unmap(void __iomem *addr)
 {
 	struct pmb_entry *pmbe = NULL;
-	int i;
+	unsigned long vaddr = (unsigned long __force)addr;
+	int i, found = 0;
 
 	read_lock(&pmb_rwlock);
 
 	for (i = 0; i < ARRAY_SIZE(pmb_entry_list); i++) {
 		if (test_bit(i, pmb_map)) {
 			pmbe = &pmb_entry_list[i];
-			if (pmbe->vpn == addr)
+			if (pmbe->vpn == vaddr) {
+				found = 1;
 				break;
+			}
 		}
 	}
 
 	read_unlock(&pmb_rwlock);
 
-	pmb_unmap_entry(pmbe, NR_PMB_ENTRIES);
-}
-
-static bool pmb_can_merge(struct pmb_entry *a, struct pmb_entry *b)
-{
-	return (b->vpn == (a->vpn + a->size)) &&
-	       (b->ppn == (a->ppn + a->size)) &&
-	       (b->flags == a->flags);
-}
-
-static bool pmb_size_valid(unsigned long size)
-{
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(pmb_sizes); i++)
-		if (pmb_sizes[i].size == size)
-			return true;
-
-	return false;
-}
-
-static int pmb_size_to_flags(unsigned long size)
-{
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(pmb_sizes); i++)
-		if (pmb_sizes[i].size == size)
-			return pmb_sizes[i].flag;
+	if (found) {
+		pmb_unmap_entry(pmbe, NR_PMB_ENTRIES);
+		return 0;
+	}
 
-	return 0;
+	return -EINVAL;
 }
 
 static void __pmb_unmap_entry(struct pmb_entry *pmbe, int depth)
@@ -351,6 +500,8 @@ static void __pmb_unmap_entry(struct pmb_entry *pmbe, int depth)
 		 */
 		__clear_pmb_entry(pmbe);
 
+		flush_cache_vunmap(pmbe->vpn, pmbe->vpn + pmbe->size);
+
 		pmbe = pmblink->link;
 
 		pmb_free(pmblink);
@@ -369,11 +520,6 @@ static void pmb_unmap_entry(struct pmb_entry *pmbe, int depth)
 	write_unlock_irqrestore(&pmb_rwlock, flags);
 }
 
-static __always_inline unsigned int pmb_ppn_in_range(unsigned long ppn)
-{
-	return ppn >= __pa(memory_start) && ppn < __pa(memory_end);
-}
-
 static void __init pmb_notify(void)
 {
 	int i;
@@ -625,6 +771,18 @@ static void __init pmb_resize(void)
 }
 #endif
 
+static int __init early_pmb(char *p)
+{
+	if (!p)
+		return 0;
+
+	if (strstr(p, "iomap"))
+		pmb_iomapping_enabled = 1;
+
+	return 0;
+}
+early_param("pmb", early_pmb);
+
 void __init pmb_init(void)
 {
 	/* Synchronize software state */
@@ -713,7 +871,7 @@ static int __init pmb_debugfs_init(void)
 
 	return 0;
 }
-postcore_initcall(pmb_debugfs_init);
+subsys_initcall(pmb_debugfs_init);
 
 #ifdef CONFIG_PM
 static int pmb_sysdev_suspend(struct sys_device *dev, pm_message_t state)