Merge commit 'v2.6.30-rc5' into core/iommu

Merge reason: core/iommu was on an .30-rc1 base,
              update it to .30-rc5 to refresh.

Signed-off-by: Ingo Molnar <mingo@elte.hu>

32 files changed, 3536 insertions, 509 deletions

diff --git a/arch/arm/mach-davinci/Kconfig b/arch/arm/mach-davinci/Kconfig
index bac988e7a4c..a9c78bc72b8 100644
--- a/arch/arm/mach-davinci/Kconfig
+++ b/arch/arm/mach-davinci/Kconfig
@@ -4,19 +4,56 @@ menu "TI DaVinci Implementations"
 
 comment "DaVinci Core Type"
 
-config ARCH_DAVINCI644x
-	default y
+config ARCH_DAVINCI_DM644x
 	bool "DaVinci 644x based system"
 
 comment "DaVinci Board Type"
 
 config MACH_DAVINCI_EVM
-	bool "TI DaVinci EVM"
+	bool "TI DM644x EVM"
 	default y
-	depends on ARCH_DAVINCI644x
+	depends on ARCH_DAVINCI_DM644x
 	help
 	  Configure this option to specify the whether the board used
-	  for development is a DaVinci EVM
+	  for development is a DM644x EVM
+
+
+config DAVINCI_MUX
+	bool "DAVINCI multiplexing support"
+	depends on ARCH_DAVINCI
+	default y
+	help
+	  Pin multiplexing support for DAVINCI boards. If your bootloader
+	  sets the multiplexing correctly, say N. Otherwise, or if unsure,
+	  say Y.
+
+config DAVINCI_MUX_DEBUG
+        bool "Multiplexing debug output"
+        depends on DAVINCI_MUX
+        help
+          Makes the multiplexing functions print out a lot of debug info.
+          This is useful if you want to find out the correct values of the
+          multiplexing registers.
+
+config DAVINCI_MUX_WARNINGS
+        bool "Warn about pins the bootloader didn't set up"
+        depends on DAVINCI_MUX
+        help
+          Choose Y here to warn whenever driver initialization logic needs
+          to change the pin multiplexing setup.  When there are no warnings
+          printed, it's safe to deselect DAVINCI_MUX for your product.
+
+config DAVINCI_RESET_CLOCKS
+	bool "Reset unused clocks during boot"
+	depends on ARCH_DAVINCI
+	help
+	  Say Y if you want to reset unused clocks during boot.
+	  This option saves power, but assumes all drivers are
+	  using the clock framework. Broken drivers that do not
+	  yet use clock framework may not work with this option.
+	  If you are booting from another operating system, you
+	  probably do not want this option enabled until your
+	  device drivers work properly.
 
 endmenu
 
diff --git a/arch/arm/mach-davinci/Makefile b/arch/arm/mach-davinci/Makefile
index 4dc458597f4..1674661942f 100644
--- a/arch/arm/mach-davinci/Makefile
+++ b/arch/arm/mach-davinci/Makefile
@@ -5,7 +5,12 @@
 
 # Common objects
 obj-y 			:= time.o irq.o clock.o serial.o io.o id.o psc.o \
-			   gpio.o mux.o devices.o usb.o
+			   gpio.o devices.o dma.o usb.o
+
+obj-$(CONFIG_DAVINCI_MUX)		+= mux.o
+
+# Chip specific
+obj-$(CONFIG_ARCH_DAVINCI_DM644x)       += dm644x.o
 
 # Board specific
-obj-$(CONFIG_MACH_DAVINCI_EVM)  += board-evm.o
+obj-$(CONFIG_MACH_DAVINCI_EVM)  	+= board-dm644x-evm.o
diff --git a/arch/arm/mach-davinci/board-evm.c b/arch/arm/mach-davinci/board-dm644x-evm.c
index 0b97a528902..b2e7f9c63bc 100644
--- a/arch/arm/mach-davinci/board-evm.c
+++ b/arch/arm/mach-davinci/board-dm644x-evm.c
@@ -15,15 +15,20 @@
 #include <linux/platform_device.h>
 #include <linux/gpio.h>
 #include <linux/leds.h>
+#include <linux/memory.h>
+#include <linux/etherdevice.h>
 
 #include <linux/i2c.h>
 #include <linux/i2c/pcf857x.h>
 #include <linux/i2c/at24.h>
 
 #include <linux/mtd/mtd.h>
+#include <linux/mtd/nand.h>
 #include <linux/mtd/partitions.h>
 #include <linux/mtd/physmap.h>
 #include <linux/io.h>
+#include <linux/phy.h>
+#include <linux/clk.h>
 
 #include <asm/setup.h>
 #include <asm/mach-types.h>
@@ -32,25 +37,34 @@
 #include <asm/mach/map.h>
 #include <asm/mach/flash.h>
 
-#include <mach/hardware.h>
+#include <mach/dm644x.h>
 #include <mach/common.h>
 #include <mach/i2c.h>
+#include <mach/serial.h>
+#include <mach/mux.h>
+#include <mach/psc.h>
+#include <mach/nand.h>
 
-/* other misc. init functions */
-void __init davinci_psc_init(void);
-void __init davinci_irq_init(void);
-void __init davinci_map_common_io(void);
-void __init davinci_init_common_hw(void);
+#define DM644X_EVM_PHY_MASK		(0x2)
+#define DM644X_EVM_MDIO_FREQUENCY	(2200000) /* PHY bus frequency */
 
-#if defined(CONFIG_MTD_PHYSMAP) || \
-    defined(CONFIG_MTD_PHYSMAP_MODULE)
+#define DAVINCI_CFC_ATA_BASE		  0x01C66000
+
+#define DAVINCI_ASYNC_EMIF_CONTROL_BASE   0x01e00000
+#define DAVINCI_ASYNC_EMIF_DATA_CE0_BASE  0x02000000
+#define DAVINCI_ASYNC_EMIF_DATA_CE1_BASE  0x04000000
+#define DAVINCI_ASYNC_EMIF_DATA_CE2_BASE  0x06000000
+#define DAVINCI_ASYNC_EMIF_DATA_CE3_BASE  0x08000000
+
+#define LXT971_PHY_ID	(0x001378e2)
+#define LXT971_PHY_MASK	(0xfffffff0)
 
 static struct mtd_partition davinci_evm_norflash_partitions[] = {
-	/* bootloader (U-Boot, etc) in first 4 sectors */
+	/* bootloader (UBL, U-Boot, etc) in first 5 sectors */
 	{
 		.name		= "bootloader",
 		.offset		= 0,
-		.size		= 4 * SZ_64K,
+		.size		= 5 * SZ_64K,
 		.mask_flags	= MTD_WRITEABLE, /* force read-only */
 	},
 	/* bootloader params in the next 1 sectors */
@@ -100,10 +114,89 @@ static struct platform_device davinci_evm_norflash_device = {
 	.resource	= &davinci_evm_norflash_resource,
 };
 
-#endif
+/* DM644x EVM includes a 64 MByte small-page NAND flash (16K blocks).
+ * It may used instead of the (default) NOR chip to boot, using TI's
+ * tools to install the secondary boot loader (UBL) and U-Boot.
+ */
+struct mtd_partition davinci_evm_nandflash_partition[] = {
+	/* Bootloader layout depends on whose u-boot is installed, but we
+	 * can hide all the details.
+	 *  - block 0 for u-boot environment ... in mainline u-boot
+	 *  - block 1 for UBL (plus up to four backup copies in blocks 2..5)
+	 *  - blocks 6...? for u-boot
+	 *  - blocks 16..23 for u-boot environment ... in TI's u-boot
+	 */
+	{
+		.name		= "bootloader",
+		.offset		= 0,
+		.size		= SZ_256K + SZ_128K,
+		.mask_flags	= MTD_WRITEABLE,	/* force read-only */
+	},
+	/* Kernel */
+	{
+		.name		= "kernel",
+		.offset		= MTDPART_OFS_APPEND,
+		.size		= SZ_4M,
+		.mask_flags	= 0,
+	},
+	/* File system (older GIT kernels started this on the 5MB mark) */
+	{
+		.name		= "filesystem",
+		.offset		= MTDPART_OFS_APPEND,
+		.size		= MTDPART_SIZ_FULL,
+		.mask_flags	= 0,
+	}
+	/* A few blocks at end hold a flash BBT ... created by TI's CCS
+	 * using flashwriter_nand.out, but ignored by TI's versions of
+	 * Linux and u-boot.  We boot faster by using them.
+	 */
+};
 
-#if defined(CONFIG_BLK_DEV_PALMCHIP_BK3710) || \
-    defined(CONFIG_BLK_DEV_PALMCHIP_BK3710_MODULE)
+static struct davinci_nand_pdata davinci_evm_nandflash_data = {
+	.parts		= davinci_evm_nandflash_partition,
+	.nr_parts	= ARRAY_SIZE(davinci_evm_nandflash_partition),
+	.ecc_mode	= NAND_ECC_HW,
+	.options	= NAND_USE_FLASH_BBT,
+};
+
+static struct resource davinci_evm_nandflash_resource[] = {
+	{
+		.start		= DAVINCI_ASYNC_EMIF_DATA_CE0_BASE,
+		.end		= DAVINCI_ASYNC_EMIF_DATA_CE0_BASE + SZ_16M - 1,
+		.flags		= IORESOURCE_MEM,
+	}, {
+		.start		= DAVINCI_ASYNC_EMIF_CONTROL_BASE,
+		.end		= DAVINCI_ASYNC_EMIF_CONTROL_BASE + SZ_4K - 1,
+		.flags		= IORESOURCE_MEM,
+	},
+};
+
+static struct platform_device davinci_evm_nandflash_device = {
+	.name		= "davinci_nand",
+	.id		= 0,
+	.dev		= {
+		.platform_data	= &davinci_evm_nandflash_data,
+	},
+	.num_resources	= ARRAY_SIZE(davinci_evm_nandflash_resource),
+	.resource	= davinci_evm_nandflash_resource,
+};
+
+static u64 davinci_fb_dma_mask = DMA_BIT_MASK(32);
+
+static struct platform_device davinci_fb_device = {
+	.name		= "davincifb",
+	.id		= -1,
+	.dev = {
+		.dma_mask		= &davinci_fb_dma_mask,
+		.coherent_dma_mask      = DMA_BIT_MASK(32),
+	},
+	.num_resources = 0,
+};
+
+static struct platform_device rtc_dev = {
+	.name           = "rtc_davinci_evm",
+	.id             = -1,
+};
 
 static struct resource ide_resources[] = {
 	{
@@ -131,8 +224,6 @@ static struct platform_device ide_dev = {
 	},
 };
 
-#endif
-
 /*----------------------------------------------------------------------*/
 
 /*
@@ -311,7 +402,9 @@ evm_u35_setup(struct i2c_client *client, int gpio, unsigned ngpio, void *c)
 	gpio_request(gpio + 7, "nCF_SEL");
 	gpio_direction_output(gpio + 7, 1);
 
-	/* irlml6401 sustains over 3A, switches 5V in under 8 msec */
+	/* irlml6401 switches over 1A, in under 8 msec;
+	 * now it can be managed by nDRV_VBUS ...
+	 */
 	setup_usb(500, 8);
 
 	return 0;
@@ -343,14 +436,120 @@ static struct pcf857x_platform_data pcf_data_u35 = {
  *  - 0x0039, 1 byte NTSC vs PAL (bit 0x80 == PAL)
  *  - ... newer boards may have more
  */
+static struct memory_accessor *at24_mem_acc;
+
+static void at24_setup(struct memory_accessor *mem_acc, void *context)
+{
+	DECLARE_MAC_BUF(mac_str);
+	char mac_addr[6];
+
+	at24_mem_acc = mem_acc;
+
+	/* Read MAC addr from EEPROM */
+	if (at24_mem_acc->read(at24_mem_acc, mac_addr, 0x7f00, 6) == 6) {
+		printk(KERN_INFO "Read MAC addr from EEPROM: %s\n",
+		       print_mac(mac_str, mac_addr));
+	}
+}
+
 static struct at24_platform_data eeprom_info = {
 	.byte_len	= (256*1024) / 8,
 	.page_size	= 64,
 	.flags		= AT24_FLAG_ADDR16,
+	.setup          = at24_setup,
 };
 
+int dm6446evm_eeprom_read(void *buf, off_t off, size_t count)
+{
+	if (at24_mem_acc)
+		return at24_mem_acc->read(at24_mem_acc, buf, off, count);
+	return -ENODEV;
+}
+EXPORT_SYMBOL(dm6446evm_eeprom_read);
+
+int dm6446evm_eeprom_write(void *buf, off_t off, size_t count)
+{
+	if (at24_mem_acc)
+		return at24_mem_acc->write(at24_mem_acc, buf, off, count);
+	return -ENODEV;
+}
+EXPORT_SYMBOL(dm6446evm_eeprom_write);
+
+/*
+ * MSP430 supports RTC, card detection, input from IR remote, and
+ * a bit more.  It triggers interrupts on GPIO(7) from pressing
+ * buttons on the IR remote, and for card detect switches.
+ */
+static struct i2c_client *dm6446evm_msp;
+
+static int dm6446evm_msp_probe(struct i2c_client *client,
+		const struct i2c_device_id *id)
+{
+	dm6446evm_msp = client;
+	return 0;
+}
+
+static int dm6446evm_msp_remove(struct i2c_client *client)
+{
+	dm6446evm_msp = NULL;
+	return 0;
+}
+
+static const struct i2c_device_id dm6446evm_msp_ids[] = {
+	{ "dm6446evm_msp", 0, },
+	{ /* end of list */ },
+};
+
+static struct i2c_driver dm6446evm_msp_driver = {
+	.driver.name	= "dm6446evm_msp",
+	.id_table	= dm6446evm_msp_ids,
+	.probe		= dm6446evm_msp_probe,
+	.remove		= dm6446evm_msp_remove,
+};
+
+static int dm6444evm_msp430_get_pins(void)
+{
+	static const char txbuf[2] = { 2, 4, };
+	char buf[4];
+	struct i2c_msg msg[2] = {
+		{
+			.addr = dm6446evm_msp->addr,
+			.flags = 0,
+			.len = 2,
+			.buf = (void __force *)txbuf,
+		},
+		{
+			.addr = dm6446evm_msp->addr,
+			.flags = I2C_M_RD,
+			.len = 4,
+			.buf = buf,
+		},
+	};
+	int status;
+
+	if (!dm6446evm_msp)
+		return -ENXIO;
+
+	/* Command 4 == get input state, returns port 2 and port3 data
+	 *   S Addr W [A] len=2 [A] cmd=4 [A]
+	 *   RS Addr R [A] [len=4] A [cmd=4] A [port2] A [port3] N P
+	 */
+	status = i2c_transfer(dm6446evm_msp->adapter, msg, 2);
+	if (status < 0)
+		return status;
+
+	dev_dbg(&dm6446evm_msp->dev,
+		"PINS: %02x %02x %02x %02x\n",
+		buf[0], buf[1], buf[2], buf[3]);
+
+	return (buf[3] << 8) | buf[2];
+}
+
 static struct i2c_board_info __initdata i2c_info[] =  {
 	{
+		I2C_BOARD_INFO("dm6446evm_msp", 0x23),
+	},
+	{
 		I2C_BOARD_INFO("pcf8574", 0x38),
 		.platform_data	= &pcf_data_u2,
 	},
@@ -368,7 +567,6 @@ static struct i2c_board_info __initdata i2c_info[] =  {
 	},
 	/* ALSO:
 	 * - tvl320aic33 audio codec (0x1b)
-	 * - msp430 microcontroller (0x23)
 	 * - tvp5146 video decoder (0x5d)
 	 */
 };
@@ -384,51 +582,109 @@ static struct davinci_i2c_platform_data i2c_pdata = {
 static void __init evm_init_i2c(void)
 {
 	davinci_init_i2c(&i2c_pdata);
+	i2c_add_driver(&dm6446evm_msp_driver);
 	i2c_register_board_info(1, i2c_info, ARRAY_SIZE(i2c_info));
 }
 
 static struct platform_device *davinci_evm_devices[] __initdata = {
-#if defined(CONFIG_MTD_PHYSMAP) || \
-    defined(CONFIG_MTD_PHYSMAP_MODULE)
-	&davinci_evm_norflash_device,
-#endif
-#if defined(CONFIG_BLK_DEV_PALMCHIP_BK3710) || \
-    defined(CONFIG_BLK_DEV_PALMCHIP_BK3710_MODULE)
-	&ide_dev,
-#endif
+	&davinci_fb_device,
+	&rtc_dev,
+};
+
+static struct davinci_uart_config uart_config __initdata = {
+	.enabled_uarts = (1 << 0),
 };
 
 static void __init
 davinci_evm_map_io(void)
 {
 	davinci_map_common_io();
+	dm644x_init();
 }
 
-static __init void davinci_evm_init(void)
+static int davinci_phy_fixup(struct phy_device *phydev)
 {
-	davinci_psc_init();
+	unsigned int control;
+	/* CRITICAL: Fix for increasing PHY signal drive strength for
+	 * TX lockup issue. On DaVinci EVM, the Intel LXT971 PHY
+	 * signal strength was low causing  TX to fail randomly. The
+	 * fix is to Set bit 11 (Increased MII drive strength) of PHY
+	 * register 26 (Digital Config register) on this phy. */
+	control = phy_read(phydev, 26);
+	phy_write(phydev, 26, (control | 0x800));
+	return 0;
+}
 
 #if defined(CONFIG_BLK_DEV_PALMCHIP_BK3710) || \
     defined(CONFIG_BLK_DEV_PALMCHIP_BK3710_MODULE)
+#define HAS_ATA 1
+#else
+#define HAS_ATA 0
+#endif
+
 #if defined(CONFIG_MTD_PHYSMAP) || \
     defined(CONFIG_MTD_PHYSMAP_MODULE)
-	printk(KERN_WARNING "WARNING: both IDE and NOR flash are enabled, "
-	       "but share pins.\n\t Disable IDE for NOR support.\n");
+#define HAS_NOR 1
+#else
+#define HAS_NOR 0
 #endif
+
+#if defined(CONFIG_MTD_NAND_DAVINCI) || \
+    defined(CONFIG_MTD_NAND_DAVINCI_MODULE)
+#define HAS_NAND 1
+#else
+#define HAS_NAND 0
 #endif
 
+static __init void davinci_evm_init(void)
+{
+	struct clk *aemif_clk;
+
+	aemif_clk = clk_get(NULL, "aemif");
+	clk_enable(aemif_clk);
+
+	if (HAS_ATA) {
+		if (HAS_NAND || HAS_NOR)
+			pr_warning("WARNING: both IDE and Flash are "
+				"enabled, but they share AEMIF pins.\n"
+				"\tDisable IDE for NAND/NOR support.\n");
+		davinci_cfg_reg(DM644X_HPIEN_DISABLE);
+		davinci_cfg_reg(DM644X_ATAEN);
+		davinci_cfg_reg(DM644X_HDIREN);
+		platform_device_register(&ide_dev);
+	} else if (HAS_NAND || HAS_NOR) {
+		davinci_cfg_reg(DM644X_HPIEN_DISABLE);
+		davinci_cfg_reg(DM644X_ATAEN_DISABLE);
+
+		/* only one device will be jumpered and detected */
+		if (HAS_NAND) {
+			platform_device_register(&davinci_evm_nandflash_device);
+			evm_leds[7].default_trigger = "nand-disk";
+			if (HAS_NOR)
+				pr_warning("WARNING: both NAND and NOR flash "
+					"are enabled; disable one of them.\n");
+		} else if (HAS_NOR)
+			platform_device_register(&davinci_evm_norflash_device);
+	}
+
 	platform_add_devices(davinci_evm_devices,
 			     ARRAY_SIZE(davinci_evm_devices));
 	evm_init_i2c();
+
+	davinci_serial_init(&uart_config);
+
+	/* Register the fixup for PHY on DaVinci */
+	phy_register_fixup_for_uid(LXT971_PHY_ID, LXT971_PHY_MASK,
+					davinci_phy_fixup);
+
 }
 
 static __init void davinci_evm_irq_init(void)
 {
-	davinci_init_common_hw();
 	davinci_irq_init();
 }
 
-MACHINE_START(DAVINCI_EVM, "DaVinci EVM")
+MACHINE_START(DAVINCI_EVM, "DaVinci DM644x EVM")
 	/* Maintainer: MontaVista Software <source@mvista.com> */
 	.phys_io      = IO_PHYS,
 	.io_pg_offst  = (__IO_ADDRESS(IO_PHYS) >> 18) & 0xfffc,
diff --git a/arch/arm/mach-davinci/clock.c b/arch/arm/mach-davinci/clock.c
index abb92b7eca0..f0baaa15a57 100644
--- a/arch/arm/mach-davinci/clock.c
+++ b/arch/arm/mach-davinci/clock.c
@@ -1,7 +1,8 @@
 /*
- * TI DaVinci clock config file
+ * Clock and PLL control for DaVinci devices
  *
- * Copyright (C) 2006 Texas Instruments.
+ * Copyright (C) 2006-2007 Texas Instruments.
+ * Copyright (C) 2008-2009 Deep Root Systems, LLC
  *
  * 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
@@ -13,6 +14,7 @@
 #include <linux/kernel.h>
 #include <linux/init.h>
 #include <linux/errno.h>
+#include <linux/clk.h>
 #include <linux/err.h>
 #include <linux/mutex.h>
 #include <linux/platform_device.h>
@@ -21,98 +23,50 @@
 #include <mach/hardware.h>
 
 #include <mach/psc.h>
+#include <mach/cputype.h>
 #include "clock.h"
 
-/* PLL/Reset register offsets */
-#define PLLM		0x110
-
 static LIST_HEAD(clocks);
 static DEFINE_MUTEX(clocks_mutex);
 static DEFINE_SPINLOCK(clockfw_lock);
 
-static unsigned int commonrate;
-static unsigned int armrate;
-static unsigned int fixedrate = 27000000;	/* 27 MHZ */
-
-extern void davinci_psc_config(unsigned int domain, unsigned int id, char enable);
-
-/*
- * Returns a clock. Note that we first try to use device id on the bus
- * and clock name. If this fails, we try to use clock name only.
- */
-struct clk *clk_get(struct device *dev, const char *id)
+static unsigned psc_domain(struct clk *clk)
 {
-	struct clk *p, *clk = ERR_PTR(-ENOENT);
-	int idno;
-
-	if (dev == NULL || dev->bus != &platform_bus_type)
-		idno = -1;
-	else
-		idno = to_platform_device(dev)->id;
-
-	mutex_lock(&clocks_mutex);
-
-	list_for_each_entry(p, &clocks, node) {
-		if (p->id == idno &&
-		    strcmp(id, p->name) == 0 && try_module_get(p->owner)) {
-			clk = p;
-			goto found;
-		}
-	}
-
-	list_for_each_entry(p, &clocks, node) {
-		if (strcmp(id, p->name) == 0 && try_module_get(p->owner)) {
-			clk = p;
-			break;
-		}
-	}
-
-found:
-	mutex_unlock(&clocks_mutex);
-
-	return clk;
+	return (clk->flags & PSC_DSP)
+		? DAVINCI_GPSC_DSPDOMAIN
+		: DAVINCI_GPSC_ARMDOMAIN;
 }
-EXPORT_SYMBOL(clk_get);
 
-void clk_put(struct clk *clk)
+static void __clk_enable(struct clk *clk)
 {
-	if (clk && !IS_ERR(clk))
-		module_put(clk->owner);
-}
-EXPORT_SYMBOL(clk_put);
-
-static int __clk_enable(struct clk *clk)
-{
-	if (clk->flags & ALWAYS_ENABLED)
-		return 0;
-
-	davinci_psc_config(DAVINCI_GPSC_ARMDOMAIN, clk->lpsc, 1);
-	return 0;
+	if (clk->parent)
+		__clk_enable(clk->parent);
+	if (clk->usecount++ == 0 && (clk->flags & CLK_PSC))
+		davinci_psc_config(psc_domain(clk), clk->lpsc, 1);
 }
 
 static void __clk_disable(struct clk *clk)
 {
-	if (clk->usecount)
+	if (WARN_ON(clk->usecount == 0))
 		return;
-
-	davinci_psc_config(DAVINCI_GPSC_ARMDOMAIN, clk->lpsc, 0);
+	if (--clk->usecount == 0 && !(clk->flags & CLK_PLL))
+		davinci_psc_config(psc_domain(clk), clk->lpsc, 0);
+	if (clk->parent)
+		__clk_disable(clk->parent);
 }
 
 int clk_enable(struct clk *clk)
 {
 	unsigned long flags;
-	int ret = 0;
 
 	if (clk == NULL || IS_ERR(clk))
 		return -EINVAL;
 
-	if (clk->usecount++ == 0) {
-		spin_lock_irqsave(&clockfw_lock, flags);
-		ret = __clk_enable(clk);
-		spin_unlock_irqrestore(&clockfw_lock, flags);
-	}
+	spin_lock_irqsave(&clockfw_lock, flags);
+	__clk_enable(clk);
+	spin_unlock_irqrestore(&clockfw_lock, flags);
 
-	return ret;
+	return 0;
 }
 EXPORT_SYMBOL(clk_enable);
 
@@ -123,11 +77,9 @@ void clk_disable(struct clk *clk)
 	if (clk == NULL || IS_ERR(clk))
 		return;
 
-	if (clk->usecount > 0 && !(--clk->usecount)) {
-		spin_lock_irqsave(&clockfw_lock, flags);
-		__clk_disable(clk);
-		spin_unlock_irqrestore(&clockfw_lock, flags);
-	}
+	spin_lock_irqsave(&clockfw_lock, flags);
+	__clk_disable(clk);
+	spin_unlock_irqrestore(&clockfw_lock, flags);
 }
 EXPORT_SYMBOL(clk_disable);
 
@@ -136,7 +88,7 @@ unsigned long clk_get_rate(struct clk *clk)
 	if (clk == NULL || IS_ERR(clk))
 		return -EINVAL;
 
-	return *(clk->rate);
+	return clk->rate;
 }
 EXPORT_SYMBOL(clk_get_rate);
 
@@ -145,7 +97,7 @@ long clk_round_rate(struct clk *clk, unsigned long rate)
 	if (clk == NULL || IS_ERR(clk))
 		return -EINVAL;
 
-	return *(clk->rate);
+	return clk->rate;
 }
 EXPORT_SYMBOL(clk_round_rate);
 
@@ -164,10 +116,23 @@ int clk_register(struct clk *clk)
 	if (clk == NULL || IS_ERR(clk))
 		return -EINVAL;
 
+	if (WARN(clk->parent && !clk->parent->rate,
+			"CLK: %s parent %s has no rate!\n",
+			clk->name, clk->parent->name))
+		return -EINVAL;
+
 	mutex_lock(&clocks_mutex);
-	list_add(&clk->node, &clocks);
+	list_add_tail(&clk->node, &clocks);
 	mutex_unlock(&clocks_mutex);
 
+	/* If rate is already set, use it */
+	if (clk->rate)
+		return 0;
+
+	/* Otherwise, default to parent rate */
+	if (clk->parent)
+		clk->rate = clk->parent->rate;
+
 	return 0;
 }
 EXPORT_SYMBOL(clk_register);
@@ -183,84 +148,150 @@ void clk_unregister(struct clk *clk)
 }
 EXPORT_SYMBOL(clk_unregister);
 
-static struct clk davinci_clks[] = {
-	{
-		.name = "ARMCLK",
-		.rate = &armrate,
-		.lpsc = -1,
-		.flags = ALWAYS_ENABLED,
-	},
-	{
-		.name = "UART",
-		.rate = &fixedrate,
-		.lpsc = DAVINCI_LPSC_UART0,
-	},
-	{
-		.name = "EMACCLK",
-		.rate = &commonrate,
-		.lpsc = DAVINCI_LPSC_EMAC_WRAPPER,
-	},
-	{
-		.name = "I2CCLK",
-		.rate = &fixedrate,
-		.lpsc = DAVINCI_LPSC_I2C,
-	},
-	{
-		.name = "IDECLK",
-		.rate = &commonrate,
-		.lpsc = DAVINCI_LPSC_ATA,
-	},
-	{
-		.name = "McBSPCLK",
-		.rate = &commonrate,
-		.lpsc = DAVINCI_LPSC_McBSP,
-	},
-	{
-		.name = "MMCSDCLK",
-		.rate = &commonrate,
-		.lpsc = DAVINCI_LPSC_MMC_SD,
-	},
-	{
-		.name = "SPICLK",
-		.rate = &commonrate,
-		.lpsc = DAVINCI_LPSC_SPI,
-	},
-	{
-		.name = "gpio",
-		.rate = &commonrate,
-		.lpsc = DAVINCI_LPSC_GPIO,
-	},
-	{
-		.name = "usb",
-		.rate = &commonrate,
-		.lpsc = DAVINCI_LPSC_USB,
-	},
-	{
-		.name = "AEMIFCLK",
-		.rate = &commonrate,
-		.lpsc = DAVINCI_LPSC_AEMIF,
-		.usecount = 1,
+#ifdef CONFIG_DAVINCI_RESET_CLOCKS
+/*
+ * Disable any unused clocks left on by the bootloader
+ */
+static int __init clk_disable_unused(void)
+{
+	struct clk *ck;
+
+	spin_lock_irq(&clockfw_lock);
+	list_for_each_entry(ck, &clocks, node) {
+		if (ck->usecount > 0)
+			continue;
+		if (!(ck->flags & CLK_PSC))
+			continue;
+
+		/* ignore if in Disabled or SwRstDisable states */
+		if (!davinci_psc_is_clk_active(ck->lpsc))
+			continue;
+
+		pr_info("Clocks: disable unused %s\n", ck->name);
+		davinci_psc_config(psc_domain(ck), ck->lpsc, 0);
 	}
-};
+	spin_unlock_irq(&clockfw_lock);
+
+	return 0;
+}
+late_initcall(clk_disable_unused);
+#endif
 
-int __init davinci_clk_init(void)
+static void clk_sysclk_recalc(struct clk *clk)
 {
-	struct clk *clkp;
-	int count = 0;
-	u32 pll_mult;
-
-	pll_mult = davinci_readl(DAVINCI_PLL_CNTRL0_BASE + PLLM);
-	commonrate = ((pll_mult + 1) * 27000000) / 6;
-	armrate = ((pll_mult + 1) * 27000000) / 2;
-
-	for (clkp = davinci_clks; count < ARRAY_SIZE(davinci_clks);
-	     count++, clkp++) {
-		clk_register(clkp);
-
-		/* Turn on clocks that have been enabled in the
-		 * table above */
-		if (clkp->usecount)
-			clk_enable(clkp);
+	u32 v, plldiv;
+	struct pll_data *pll;
+
+	/* If this is the PLL base clock, no more calculations needed */
+	if (clk->pll_data)
+		return;
+
+	if (WARN_ON(!clk->parent))
+		return;
+
+	clk->rate = clk->parent->rate;
+
+	/* Otherwise, the parent must be a PLL */
+	if (WARN_ON(!clk->parent->pll_data))
+		return;
+
+	pll = clk->parent->pll_data;
+
+	/* If pre-PLL, source clock is before the multiplier and divider(s) */
+	if (clk->flags & PRE_PLL)
+		clk->rate = pll->input_rate;
+
+	if (!clk->div_reg)
+		return;
+
+	v = __raw_readl(pll->base + clk->div_reg);
+	if (v & PLLDIV_EN) {
+		plldiv = (v & PLLDIV_RATIO_MASK) + 1;
+		if (plldiv)
+			clk->rate /= plldiv;
+	}
+}
+
+static void __init clk_pll_init(struct clk *clk)
+{
+	u32 ctrl, mult = 1, prediv = 1, postdiv = 1;
+	u8 bypass;
+	struct pll_data *pll = clk->pll_data;
+
+	pll->base = IO_ADDRESS(pll->phys_base);
+	ctrl = __raw_readl(pll->base + PLLCTL);
+	clk->rate = pll->input_rate = clk->parent->rate;
+
+	if (ctrl & PLLCTL_PLLEN) {
+		bypass = 0;
+		mult = __raw_readl(pll->base + PLLM);
+		mult = (mult & PLLM_PLLM_MASK) + 1;
+	} else
+		bypass = 1;
+
+	if (pll->flags & PLL_HAS_PREDIV) {
+		prediv = __raw_readl(pll->base + PREDIV);
+		if (prediv & PLLDIV_EN)
+			prediv = (prediv & PLLDIV_RATIO_MASK) + 1;
+		else
+			prediv = 1;
+	}
+
+	/* pre-divider is fixed, but (some?) chips won't report that */
+	if (cpu_is_davinci_dm355() && pll->num == 1)
+		prediv = 8;
+
+	if (pll->flags & PLL_HAS_POSTDIV) {
+		postdiv = __raw_readl(pll->base + POSTDIV);
+		if (postdiv & PLLDIV_EN)
+			postdiv = (postdiv & PLLDIV_RATIO_MASK) + 1;
+		else
+			postdiv = 1;
+	}
+
+	if (!bypass) {
+		clk->rate /= prediv;
+		clk->rate *= mult;
+		clk->rate /= postdiv;
+	}
+
+	pr_debug("PLL%d: input = %lu MHz [ ",
+		 pll->num, clk->parent->rate / 1000000);
+	if (bypass)
+		pr_debug("bypass ");
+	if (prediv > 1)
+		pr_debug("/ %d ", prediv);
+	if (mult > 1)
+		pr_debug("* %d ", mult);
+	if (postdiv > 1)
+		pr_debug("/ %d ", postdiv);
+	pr_debug("] --> %lu MHz output.\n", clk->rate / 1000000);
+}
+
+int __init davinci_clk_init(struct davinci_clk *clocks)
+  {
+	struct davinci_clk *c;
+	struct clk *clk;
+
+	for (c = clocks; c->lk.clk; c++) {
+		clk = c->lk.clk;
+
+		if (clk->pll_data)
+			clk_pll_init(clk);
+
+		/* Calculate rates for PLL-derived clocks */
+		else if (clk->flags & CLK_PLL)
+			clk_sysclk_recalc(clk);
+
+		if (clk->lpsc)
+			clk->flags |= CLK_PSC;
+
+		clkdev_add(&c->lk);
+		clk_register(clk);
+
+		/* Turn on clocks that Linux doesn't otherwise manage */
+		if (clk->flags & ALWAYS_ENABLED)
+			clk_enable(clk);
 	}
 
 	return 0;
@@ -285,12 +316,52 @@ static void davinci_ck_stop(struct seq_file *m, void *v)
 {
 }
 
-static int davinci_ck_show(struct seq_file *m, void *v)
+#define CLKNAME_MAX	10		/* longest clock name */
+#define NEST_DELTA	2
+#define NEST_MAX	4
+
+static void
+dump_clock(struct seq_file *s, unsigned nest, struct clk *parent)
 {
-	struct clk *cp;
+	char		*state;
+	char		buf[CLKNAME_MAX + NEST_DELTA * NEST_MAX];
+	struct clk	*clk;
+	unsigned	i;
+
+	if (parent->flags & CLK_PLL)
+		state = "pll";
+	else if (parent->flags & CLK_PSC)
+		state = "psc";
+	else
+		state = "";
+
+	/* <nest spaces> name <pad to end> */
+	memset(buf, ' ', sizeof(buf) - 1);
+	buf[sizeof(buf) - 1] = 0;
+	i = strlen(parent->name);
+	memcpy(buf + nest, parent->name,
+			min(i, (unsigned)(sizeof(buf) - 1 - nest)));
+
+	seq_printf(s, "%s users=%2d %-3s %9ld Hz\n",
+		   buf, parent->usecount, state, clk_get_rate(parent));
+	/* REVISIT show device associations too */
+
+	/* cost is now small, but not linear... */
+	list_for_each_entry(clk, &clocks, node) {
+		if (clk->parent == parent)
+			dump_clock(s, nest + NEST_DELTA, clk);
+	}
+}
 
-	list_for_each_entry(cp, &clocks, node)
-		seq_printf(m,"%s %d %d\n", cp->name, *(cp->rate), cp->usecount);
+static int davinci_ck_show(struct seq_file *m, void *v)
+{
+	/* Show clock tree; we know the main oscillator is first.
+	 * We trust nonzero usecounts equate to PSC enables...
+	 */
+	mutex_lock(&clocks_mutex);
+	if (!list_empty(&clocks))
+		dump_clock(m, 0, list_first_entry(&clocks, struct clk, node));
+	mutex_unlock(&clocks_mutex);
 
 	return 0;
 }
@@ -321,4 +392,4 @@ static int __init davinci_ck_proc_init(void)
 
 }
 __initcall(davinci_ck_proc_init);
-#endif	/* CONFIG_DEBUG_PROC_FS */
+#endif /* CONFIG_DEBUG_PROC_FS */
diff --git a/arch/arm/mach-davinci/clock.h b/arch/arm/mach-davinci/clock.h
index ed47079a52e..35736ec202f 100644
--- a/arch/arm/mach-davinci/clock.h
+++ b/arch/arm/mach-davinci/clock.h
@@ -1,7 +1,8 @@
 /*
  * TI DaVinci clock definitions
  *
- * Copyright (C) 2006 Texas Instruments.
+ * Copyright (C) 2006-2007 Texas Instruments.
+ * Copyright (C) 2008-2009 Deep Root Systems, LLC
  *
  * 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
@@ -11,23 +12,85 @@
 #ifndef __ARCH_ARM_DAVINCI_CLOCK_H
 #define __ARCH_ARM_DAVINCI_CLOCK_H
 
+#include <linux/list.h>
+#include <asm/clkdev.h>
+
+#define DAVINCI_PLL1_BASE 0x01c40800
+#define DAVINCI_PLL2_BASE 0x01c40c00
+#define MAX_PLL 2
+
+/* PLL/Reset register offsets */
+#define PLLCTL          0x100
+#define PLLCTL_PLLEN    BIT(0)
+#define PLLCTL_CLKMODE  BIT(8)
+
+#define PLLM		0x110
+#define PLLM_PLLM_MASK  0xff
+
+#define PREDIV          0x114
+#define PLLDIV1         0x118
+#define PLLDIV2         0x11c
+#define PLLDIV3         0x120
+#define POSTDIV         0x128
+#define BPDIV           0x12c
+#define PLLCMD		0x138
+#define PLLSTAT		0x13c
+#define PLLALNCTL	0x140
+#define PLLDCHANGE	0x144
+#define PLLCKEN		0x148
+#define PLLCKSTAT	0x14c
+#define PLLSYSTAT	0x150
+#define PLLDIV4         0x160
+#define PLLDIV5         0x164
+#define PLLDIV6         0x168
+#define PLLDIV7         0x16c
+#define PLLDIV8         0x170
+#define PLLDIV9         0x174
+#define PLLDIV_EN       BIT(15)
+#define PLLDIV_RATIO_MASK 0x1f
+
+struct pll_data {
+	u32 phys_base;
+	void __iomem *base;
+	u32 num;
+	u32 flags;
+	u32 input_rate;
+};
+#define PLL_HAS_PREDIV          0x01
+#define PLL_HAS_POSTDIV         0x02
+
 struct clk {
 	struct list_head	node;
 	struct module		*owner;
 	const char		*name;
-	unsigned int		*rate;
-	int			id;
-	__s8			usecount;
-	__u8			flags;
-	__u8			lpsc;
+	unsigned long		rate;
+	u8			usecount;
+	u8			flags;
+	u8			lpsc;
+	struct clk              *parent;
+	struct pll_data         *pll_data;
+	u32                     div_reg;
 };
 
 /* Clock flags */
-#define RATE_CKCTL		1
-#define RATE_FIXED		2
-#define RATE_PROPAGATES		4
-#define VIRTUAL_CLOCK		8
-#define ALWAYS_ENABLED		16
-#define ENABLE_REG_32BIT	32
+#define ALWAYS_ENABLED		BIT(1)
+#define CLK_PSC                 BIT(2)
+#define PSC_DSP                 BIT(3) /* PSC uses DSP domain, not ARM */
+#define CLK_PLL			BIT(4) /* PLL-derived clock */
+#define PRE_PLL                 BIT(5) /* source is before PLL mult/div */
+
+struct davinci_clk {
+	struct clk_lookup lk;
+};
+
+#define CLK(dev, con, ck) 		\
+	{				\
+		.lk = {			\
+			.dev_id = dev,	\
+			.con_id = con,	\
+			.clk = ck,	\
+		},			\
+	}
 
+int davinci_clk_init(struct davinci_clk *clocks);
 #endif
diff --git a/arch/arm/mach-davinci/devices.c b/arch/arm/mach-davinci/devices.c
index 808633f9f03..a31370b93dd 100644
--- a/arch/arm/mach-davinci/devices.c
+++ b/arch/arm/mach-davinci/devices.c
@@ -21,6 +21,10 @@
 #include <mach/hardware.h>
 #include <mach/i2c.h>
 #include <mach/irqs.h>
+#include <mach/cputype.h>
+#include <mach/mux.h>
+
+#define DAVINCI_I2C_BASE	     0x01C21000
 
 static struct resource i2c_resources[] = {
 	{
@@ -43,6 +47,9 @@ static struct platform_device davinci_i2c_device = {
 
 void __init davinci_init_i2c(struct davinci_i2c_platform_data *pdata)
 {
+	if (cpu_is_davinci_dm644x())
+		davinci_cfg_reg(DM644X_I2C);
+
 	davinci_i2c_device.dev.platform_data = pdata;
 	(void) platform_device_register(&davinci_i2c_device);
 }
diff --git a/arch/arm/mach-davinci/dm644x.c b/arch/arm/mach-davinci/dm644x.c
new file mode 100644
index 00000000000..d428ef192ea
--- /dev/null
+++ b/arch/arm/mach-davinci/dm644x.c
@@ -0,0 +1,461 @@
+/*
+ * TI DaVinci DM644x chip specific setup
+ *
+ * Author: Kevin Hilman, Deep Root Systems, LLC
+ *
+ * 2007 (c) Deep Root Systems, LLC. This file is licensed under
+ * the terms of the GNU General Public License version 2. This program
+ * is licensed "as is" without any warranty of any kind, whether express
+ * or implied.
+ */
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/clk.h>
+#include <linux/platform_device.h>
+
+#include <mach/dm644x.h>
+#include <mach/clock.h>
+#include <mach/cputype.h>
+#include <mach/edma.h>
+#include <mach/irqs.h>
+#include <mach/psc.h>
+#include <mach/mux.h>
+
+#include "clock.h"
+#include "mux.h"
+
+/*
+ * Device specific clocks
+ */
+#define DM644X_REF_FREQ		27000000
+
+static struct pll_data pll1_data = {
+	.num       = 1,
+	.phys_base = DAVINCI_PLL1_BASE,
+};
+
+static struct pll_data pll2_data = {
+	.num       = 2,
+	.phys_base = DAVINCI_PLL2_BASE,
+};
+
+static struct clk ref_clk = {
+	.name = "ref_clk",
+	.rate = DM644X_REF_FREQ,
+};
+
+static struct clk pll1_clk = {
+	.name = "pll1",
+	.parent = &ref_clk,
+	.pll_data = &pll1_data,
+	.flags = CLK_PLL,
+};
+
+static struct clk pll1_sysclk1 = {
+	.name = "pll1_sysclk1",
+	.parent = &pll1_clk,
+	.flags = CLK_PLL,
+	.div_reg = PLLDIV1,
+};
+
+static struct clk pll1_sysclk2 = {
+	.name = "pll1_sysclk2",
+	.parent = &pll1_clk,
+	.flags = CLK_PLL,
+	.div_reg = PLLDIV2,
+};
+
+static struct clk pll1_sysclk3 = {
+	.name = "pll1_sysclk3",
+	.parent = &pll1_clk,
+	.flags = CLK_PLL,
+	.div_reg = PLLDIV3,
+};
+
+static struct clk pll1_sysclk5 = {
+	.name = "pll1_sysclk5",
+	.parent = &pll1_clk,
+	.flags = CLK_PLL,
+	.div_reg = PLLDIV5,
+};
+
+static struct clk pll1_aux_clk = {
+	.name = "pll1_aux_clk",
+	.parent = &pll1_clk,
+	.flags = CLK_PLL | PRE_PLL,
+};
+
+static struct clk pll1_sysclkbp = {
+	.name = "pll1_sysclkbp",
+	.parent = &pll1_clk,
+	.flags = CLK_PLL | PRE_PLL,
+	.div_reg = BPDIV
+};
+
+static struct clk pll2_clk = {
+	.name = "pll2",
+	.parent = &ref_clk,
+	.pll_data = &pll2_data,
+	.flags = CLK_PLL,
+};
+
+static struct clk pll2_sysclk1 = {
+	.name = "pll2_sysclk1",
+	.parent = &pll2_clk,
+	.flags = CLK_PLL,
+	.div_reg = PLLDIV1,
+};
+
+static struct clk pll2_sysclk2 = {
+	.name = "pll2_sysclk2",
+	.parent = &pll2_clk,
+	.flags = CLK_PLL,
+	.div_reg = PLLDIV2,
+};
+
+static struct clk pll2_sysclkbp = {
+	.name = "pll2_sysclkbp",
+	.parent = &pll2_clk,
+	.flags = CLK_PLL | PRE_PLL,
+	.div_reg = BPDIV
+};
+
+static struct clk dsp_clk = {
+	.name = "dsp",
+	.parent = &pll1_sysclk1,
+	.lpsc = DAVINCI_LPSC_GEM,
+	.flags = PSC_DSP,
+	.usecount = 1,			/* REVISIT how to disable? */
+};
+
+static struct clk arm_clk = {
+	.name = "arm",
+	.parent = &pll1_sysclk2,
+	.lpsc = DAVINCI_LPSC_ARM,
+	.flags = ALWAYS_ENABLED,
+};
+
+static struct clk vicp_clk = {
+	.name = "vicp",
+	.parent = &pll1_sysclk2,
+	.lpsc = DAVINCI_LPSC_IMCOP,
+	.flags = PSC_DSP,
+	.usecount = 1,			/* REVISIT how to disable? */
+};
+
+static struct clk vpss_master_clk = {
+	.name = "vpss_master",
+	.parent = &pll1_sysclk3,
+	.lpsc = DAVINCI_LPSC_VPSSMSTR,
+	.flags = CLK_PSC,
+};
+
+static struct clk vpss_slave_clk = {
+	.name = "vpss_slave",
+	.parent = &pll1_sysclk3,
+	.lpsc = DAVINCI_LPSC_VPSSSLV,
+};
+
+static struct clk uart0_clk = {
+	.name = "uart0",
+	.parent = &pll1_aux_clk,
+	.lpsc = DAVINCI_LPSC_UART0,
+};
+
+static struct clk uart1_clk = {
+	.name = "uart1",
+	.parent = &pll1_aux_clk,
+	.lpsc = DAVINCI_LPSC_UART1,
+};
+
+static struct clk uart2_clk = {
+	.name = "uart2",
+	.parent = &pll1_aux_clk,
+	.lpsc = DAVINCI_LPSC_UART2,
+};
+
+static struct clk emac_clk = {
+	.name = "emac",
+	.parent = &pll1_sysclk5,
+	.lpsc = DAVINCI_LPSC_EMAC_WRAPPER,
+};
+
+static struct clk i2c_clk = {
+	.name = "i2c",
+	.parent = &pll1_aux_clk,
+	.lpsc = DAVINCI_LPSC_I2C,
+};
+
+static struct clk ide_clk = {
+	.name = "ide",
+	.parent = &pll1_sysclk5,
+	.lpsc = DAVINCI_LPSC_ATA,
+};
+
+static struct clk asp_clk = {
+	.name = "asp0",
+	.parent = &pll1_sysclk5,
+	.lpsc = DAVINCI_LPSC_McBSP,
+};
+
+static struct clk mmcsd_clk = {
+	.name = "mmcsd",
+	.parent = &pll1_sysclk5,
+	.lpsc = DAVINCI_LPSC_MMC_SD,
+};
+
+static struct clk spi_clk = {
+	.name = "spi",
+	.parent = &pll1_sysclk5,
+	.lpsc = DAVINCI_LPSC_SPI,
+};
+
+static struct clk gpio_clk = {
+	.name = "gpio",
+	.parent = &pll1_sysclk5,
+	.lpsc = DAVINCI_LPSC_GPIO,
+};
+
+static struct clk usb_clk = {
+	.name = "usb",
+	.parent = &pll1_sysclk5,
+	.lpsc = DAVINCI_LPSC_USB,
+};
+
+static struct clk vlynq_clk = {
+	.name = "vlynq",
+	.parent = &pll1_sysclk5,
+	.lpsc = DAVINCI_LPSC_VLYNQ,
+};
+
+static struct clk aemif_clk = {
+	.name = "aemif",
+	.parent = &pll1_sysclk5,
+	.lpsc = DAVINCI_LPSC_AEMIF,
+};
+
+static struct clk pwm0_clk = {
+	.name = "pwm0",
+	.parent = &pll1_aux_clk,
+	.lpsc = DAVINCI_LPSC_PWM0,
+};
+
+static struct clk pwm1_clk = {
+	.name = "pwm1",
+	.parent = &pll1_aux_clk,
+	.lpsc = DAVINCI_LPSC_PWM1,
+};
+
+static struct clk pwm2_clk = {
+	.name = "pwm2",
+	.parent = &pll1_aux_clk,
+	.lpsc = DAVINCI_LPSC_PWM2,
+};
+
+static struct clk timer0_clk = {
+	.name = "timer0",
+	.parent = &pll1_aux_clk,
+	.lpsc = DAVINCI_LPSC_TIMER0,
+};
+
+static struct clk timer1_clk = {
+	.name = "timer1",
+	.parent = &pll1_aux_clk,
+	.lpsc = DAVINCI_LPSC_TIMER1,
+};
+
+static struct clk timer2_clk = {
+	.name = "timer2",
+	.parent = &pll1_aux_clk,
+	.lpsc = DAVINCI_LPSC_TIMER2,
+	.usecount = 1,              /* REVISIT: why cant' this be disabled? */
+};
+
+struct davinci_clk dm644x_clks[] = {
+	CLK(NULL, "ref", &ref_clk),
+	CLK(NULL, "pll1", &pll1_clk),
+	CLK(NULL, "pll1_sysclk1", &pll1_sysclk1),
+	CLK(NULL, "pll1_sysclk2", &pll1_sysclk2),
+	CLK(NULL, "pll1_sysclk3", &pll1_sysclk3),
+	CLK(NULL, "pll1_sysclk5", &pll1_sysclk5),
+	CLK(NULL, "pll1_aux", &pll1_aux_clk),
+	CLK(NULL, "pll1_sysclkbp", &pll1_sysclkbp),
+	CLK(NULL, "pll2", &pll2_clk),
+	CLK(NULL, "pll2_sysclk1", &pll2_sysclk1),
+	CLK(NULL, "pll2_sysclk2", &pll2_sysclk2),
+	CLK(NULL, "pll2_sysclkbp", &pll2_sysclkbp),
+	CLK(NULL, "dsp", &dsp_clk),
+	CLK(NULL, "arm", &arm_clk),
+	CLK(NULL, "vicp", &vicp_clk),
+	CLK(NULL, "vpss_master", &vpss_master_clk),
+	CLK(NULL, "vpss_slave", &vpss_slave_clk),
+	CLK(NULL, "arm", &arm_clk),
+	CLK(NULL, "uart0", &uart0_clk),
+	CLK(NULL, "uart1", &uart1_clk),
+	CLK(NULL, "uart2", &uart2_clk),
+	CLK("davinci_emac.1", NULL, &emac_clk),
+	CLK("i2c_davinci.1", NULL, &i2c_clk),
+	CLK("palm_bk3710", NULL, &ide_clk),
+	CLK("soc-audio.0", NULL, &asp_clk),
+	CLK("davinci_mmc.0", NULL, &mmcsd_clk),
+	CLK(NULL, "spi", &spi_clk),
+	CLK(NULL, "gpio", &gpio_clk),
+	CLK(NULL, "usb", &usb_clk),
+	CLK(NULL, "vlynq", &vlynq_clk),
+	CLK(NULL, "aemif", &aemif_clk),
+	CLK(NULL, "pwm0", &pwm0_clk),
+	CLK(NULL, "pwm1", &pwm1_clk),
+	CLK(NULL, "pwm2", &pwm2_clk),
+	CLK(NULL, "timer0", &timer0_clk),
+	CLK(NULL, "timer1", &timer1_clk),
+	CLK("watchdog", NULL, &timer2_clk),
+	CLK(NULL, NULL, NULL),
+};
+
+#if defined(CONFIG_TI_DAVINCI_EMAC) || defined(CONFIG_TI_DAVINCI_EMAC_MODULE)
+
+static struct resource dm644x_emac_resources[] = {
+	{
+		.start	= DM644X_EMAC_BASE,
+		.end	= DM644X_EMAC_BASE + 0x47ff,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		.start = IRQ_EMACINT,
+		.end   = IRQ_EMACINT,
+		.flags = IORESOURCE_IRQ,
+	},
+};
+
+static struct platform_device dm644x_emac_device = {
+       .name		= "davinci_emac",
+       .id		= 1,
+       .num_resources	= ARRAY_SIZE(dm644x_emac_resources),
+       .resource	= dm644x_emac_resources,
+};
+
+#endif
+
+/*
+ * Device specific mux setup
+ *
+ *	soc	description	mux  mode   mode  mux	 dbg
+ *				reg  offset mask  mode
+ */
+static const struct mux_config dm644x_pins[] = {
+MUX_CFG(DM644X, HDIREN,		0,   16,    1,	  1,	 true)
+MUX_CFG(DM644X, ATAEN,		0,   17,    1,	  1,	 true)
+MUX_CFG(DM644X, ATAEN_DISABLE,	0,   17,    1,	  0,	 true)
+
+MUX_CFG(DM644X, HPIEN_DISABLE,	0,   29,    1,	  0,	 true)
+
+MUX_CFG(DM644X, AEAW,		0,   0,     31,	  31,	 true)
+
+MUX_CFG(DM644X, MSTK,		1,   9,     1,	  0,	 false)
+
+MUX_CFG(DM644X, I2C,		1,   7,     1,	  1,	 false)
+
+MUX_CFG(DM644X, MCBSP,		1,   10,    1,	  1,	 false)
+
+MUX_CFG(DM644X, UART1,		1,   1,     1,	  1,	 true)
+MUX_CFG(DM644X, UART2,		1,   2,     1,	  1,	 true)
+
+MUX_CFG(DM644X, PWM0,		1,   4,     1,	  1,	 false)
+
+MUX_CFG(DM644X, PWM1,		1,   5,     1,	  1,	 false)
+
+MUX_CFG(DM644X, PWM2,		1,   6,     1,	  1,	 false)
+
+MUX_CFG(DM644X, VLYNQEN,	0,   15,    1,	  1,	 false)
+MUX_CFG(DM644X, VLSCREN,	0,   14,    1,	  1,	 false)
+MUX_CFG(DM644X, VLYNQWD,	0,   12,    3,	  3,	 false)
+
+MUX_CFG(DM644X, EMACEN,		0,   31,    1,	  1,	 true)
+
+MUX_CFG(DM644X, GPIO3V,		0,   31,    1,	  0,	 true)
+
+MUX_CFG(DM644X, GPIO0,		0,   24,    1,	  0,	 true)
+MUX_CFG(DM644X, GPIO3,		0,   25,    1,	  0,	 false)
+MUX_CFG(DM644X, GPIO43_44,	1,   7,     1,	  0,	 false)
+MUX_CFG(DM644X, GPIO46_47,	0,   22,    1,	  0,	 true)
+
+MUX_CFG(DM644X, RGB666,		0,   22,    1,	  1,	 true)
+
+MUX_CFG(DM644X, LOEEN,		0,   24,    1,	  1,	 true)
+MUX_CFG(DM644X, LFLDEN,		0,   25,    1,	  1,	 false)
+};
+
+
+/*----------------------------------------------------------------------*/
+
+static const s8 dma_chan_dm644x_no_event[] = {
+	 0,  1, 12, 13, 14,
+	15, 25, 30, 31, 45,
+	46, 47, 55, 56, 57,
+	58, 59, 60, 61, 62,
+	63,
+	-1
+};
+
+static struct edma_soc_info dm644x_edma_info = {
+	.n_channel	= 64,
+	.n_region	= 4,
+	.n_slot		= 128,
+	.n_tc		= 2,
+	.noevent	= dma_chan_dm644x_no_event,
+};
+
+static struct resource edma_resources[] = {
+	{
+		.name	= "edma_cc",
+		.start	= 0x01c00000,
+		.end	= 0x01c00000 + SZ_64K - 1,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		.name	= "edma_tc0",
+		.start	= 0x01c10000,
+		.end	= 0x01c10000 + SZ_1K - 1,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		.name	= "edma_tc1",
+		.start	= 0x01c10400,
+		.end	= 0x01c10400 + SZ_1K - 1,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		.start	= IRQ_CCINT0,
+		.flags	= IORESOURCE_IRQ,
+	},
+	{
+		.start	= IRQ_CCERRINT,
+		.flags	= IORESOURCE_IRQ,
+	},
+	/* not using TC*_ERR */
+};
+
+static struct platform_device dm644x_edma_device = {
+	.name			= "edma",
+	.id			= -1,
+	.dev.platform_data	= &dm644x_edma_info,
+	.num_resources		= ARRAY_SIZE(edma_resources),
+	.resource		= edma_resources,
+};
+
+/*----------------------------------------------------------------------*/
+void __init dm644x_init(void)
+{
+	davinci_clk_init(dm644x_clks);
+	davinci_mux_register(dm644x_pins, ARRAY_SIZE(dm644x_pins));
+}
+
+static int __init dm644x_init_devices(void)
+{
+	if (!cpu_is_davinci_dm644x())
+		return 0;
+
+	platform_device_register(&dm644x_edma_device);
+	return 0;
+}
+postcore_initcall(dm644x_init_devices);
diff --git a/arch/arm/mach-davinci/dma.c b/arch/arm/mach-davinci/dma.c
new file mode 100644
index 00000000000..15e9eb158bb
--- /dev/null
+++ b/arch/arm/mach-davinci/dma.c
@@ -0,0 +1,1135 @@
+/*
+ * EDMA3 support for DaVinci
+ *
+ * Copyright (C) 2006-2009 Texas Instruments.
+ *
+ * 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/sched.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/spinlock.h>
+#include <linux/compiler.h>
+#include <linux/io.h>
+
+#include <mach/cputype.h>
+#include <mach/memory.h>
+#include <mach/hardware.h>
+#include <mach/irqs.h>
+#include <mach/edma.h>
+#include <mach/mux.h>
+
+
+/* Offsets matching "struct edmacc_param" */
+#define PARM_OPT		0x00
+#define PARM_SRC		0x04
+#define PARM_A_B_CNT		0x08
+#define PARM_DST		0x0c
+#define PARM_SRC_DST_BIDX	0x10
+#define PARM_LINK_BCNTRLD	0x14
+#define PARM_SRC_DST_CIDX	0x18
+#define PARM_CCNT		0x1c
+
+#define PARM_SIZE		0x20
+
+/* Offsets for EDMA CC global channel registers and their shadows */
+#define SH_ER		0x00	/* 64 bits */
+#define SH_ECR		0x08	/* 64 bits */
+#define SH_ESR		0x10	/* 64 bits */
+#define SH_CER		0x18	/* 64 bits */
+#define SH_EER		0x20	/* 64 bits */
+#define SH_EECR		0x28	/* 64 bits */
+#define SH_EESR		0x30	/* 64 bits */
+#define SH_SER		0x38	/* 64 bits */
+#define SH_SECR		0x40	/* 64 bits */
+#define SH_IER		0x50	/* 64 bits */
+#define SH_IECR		0x58	/* 64 bits */
+#define SH_IESR		0x60	/* 64 bits */
+#define SH_IPR		0x68	/* 64 bits */
+#define SH_ICR		0x70	/* 64 bits */
+#define SH_IEVAL	0x78
+#define SH_QER		0x80
+#define SH_QEER		0x84
+#define SH_QEECR	0x88
+#define SH_QEESR	0x8c
+#define SH_QSER		0x90
+#define SH_QSECR	0x94
+#define SH_SIZE		0x200
+
+/* Offsets for EDMA CC global registers */
+#define EDMA_REV	0x0000
+#define EDMA_CCCFG	0x0004
+#define EDMA_QCHMAP	0x0200	/* 8 registers */
+#define EDMA_DMAQNUM	0x0240	/* 8 registers (4 on OMAP-L1xx) */
+#define EDMA_QDMAQNUM	0x0260
+#define EDMA_QUETCMAP	0x0280
+#define EDMA_QUEPRI	0x0284
+#define EDMA_EMR	0x0300	/* 64 bits */
+#define EDMA_EMCR	0x0308	/* 64 bits */
+#define EDMA_QEMR	0x0310
+#define EDMA_QEMCR	0x0314
+#define EDMA_CCERR	0x0318
+#define EDMA_CCERRCLR	0x031c
+#define EDMA_EEVAL	0x0320
+#define EDMA_DRAE	0x0340	/* 4 x 64 bits*/
+#define EDMA_QRAE	0x0380	/* 4 registers */
+#define EDMA_QUEEVTENTRY	0x0400	/* 2 x 16 registers */
+#define EDMA_QSTAT	0x0600	/* 2 registers */
+#define EDMA_QWMTHRA	0x0620
+#define EDMA_QWMTHRB	0x0624
+#define EDMA_CCSTAT	0x0640
+
+#define EDMA_M		0x1000	/* global channel registers */
+#define EDMA_ECR	0x1008
+#define EDMA_ECRH	0x100C
+#define EDMA_SHADOW0	0x2000	/* 4 regions shadowing global channels */
+#define EDMA_PARM	0x4000	/* 128 param entries */
+
+#define DAVINCI_DMA_3PCC_BASE	0x01C00000
+
+#define PARM_OFFSET(param_no)	(EDMA_PARM + ((param_no) << 5))
+
+#define EDMA_MAX_DMACH           64
+#define EDMA_MAX_PARAMENTRY     512
+#define EDMA_MAX_EVQUE            2	/* FIXME too small */
+
+
+/*****************************************************************************/
+
+static void __iomem *edmacc_regs_base;
+
+static inline unsigned int edma_read(int offset)
+{
+	return (unsigned int)__raw_readl(edmacc_regs_base + offset);
+}
+
+static inline void edma_write(int offset, int val)
+{
+	__raw_writel(val, edmacc_regs_base + offset);
+}
+static inline void edma_modify(int offset, unsigned and, unsigned or)
+{
+	unsigned val = edma_read(offset);
+	val &= and;
+	val |= or;
+	edma_write(offset, val);
+}
+static inline void edma_and(int offset, unsigned and)
+{
+	unsigned val = edma_read(offset);
+	val &= and;
+	edma_write(offset, val);
+}
+static inline void edma_or(int offset, unsigned or)
+{
+	unsigned val = edma_read(offset);
+	val |= or;
+	edma_write(offset, val);
+}
+static inline unsigned int edma_read_array(int offset, int i)
+{
+	return edma_read(offset + (i << 2));
+}
+static inline void edma_write_array(int offset, int i, unsigned val)
+{
+	edma_write(offset + (i << 2), val);
+}
+static inline void edma_modify_array(int offset, int i,
+		unsigned and, unsigned or)
+{
+	edma_modify(offset + (i << 2), and, or);
+}
+static inline void edma_or_array(int offset, int i, unsigned or)
+{
+	edma_or(offset + (i << 2), or);
+}
+static inline void edma_or_array2(int offset, int i, int j, unsigned or)
+{
+	edma_or(offset + ((i*2 + j) << 2), or);
+}
+static inline void edma_write_array2(int offset, int i, int j, unsigned val)
+{
+	edma_write(offset + ((i*2 + j) << 2), val);
+}
+static inline unsigned int edma_shadow0_read(int offset)
+{
+	return edma_read(EDMA_SHADOW0 + offset);
+}
+static inline unsigned int edma_shadow0_read_array(int offset, int i)
+{
+	return edma_read(EDMA_SHADOW0 + offset + (i << 2));
+}
+static inline void edma_shadow0_write(int offset, unsigned val)
+{
+	edma_write(EDMA_SHADOW0 + offset, val);
+}
+static inline void edma_shadow0_write_array(int offset, int i, unsigned val)
+{
+	edma_write(EDMA_SHADOW0 + offset + (i << 2), val);
+}
+static inline unsigned int edma_parm_read(int offset, int param_no)
+{
+	return edma_read(EDMA_PARM + offset + (param_no << 5));
+}
+static inline void edma_parm_write(int offset, int param_no, unsigned val)
+{
+	edma_write(EDMA_PARM + offset + (param_no << 5), val);
+}
+static inline void edma_parm_modify(int offset, int param_no,
+		unsigned and, unsigned or)
+{
+	edma_modify(EDMA_PARM + offset + (param_no << 5), and, or);
+}
+static inline void edma_parm_and(int offset, int param_no, unsigned and)
+{
+	edma_and(EDMA_PARM + offset + (param_no << 5), and);
+}
+static inline void edma_parm_or(int offset, int param_no, unsigned or)
+{
+	edma_or(EDMA_PARM + offset + (param_no << 5), or);
+}
+
+/*****************************************************************************/
+
+/* actual number of DMA channels and slots on this silicon */
+static unsigned num_channels;
+static unsigned num_slots;
+
+static struct dma_interrupt_data {
+	void (*callback)(unsigned channel, unsigned short ch_status,
+			 void *data);
+	void *data;
+} intr_data[EDMA_MAX_DMACH];
+
+/* The edma_inuse bit for each PaRAM slot is clear unless the
+ * channel is in use ... by ARM or DSP, for QDMA, or whatever.
+ */
+static DECLARE_BITMAP(edma_inuse, EDMA_MAX_PARAMENTRY);
+
+/* The edma_noevent bit for each channel is clear unless
+ * it doesn't trigger DMA events on this platform.  It uses a
+ * bit of SOC-specific initialization code.
+ */
+static DECLARE_BITMAP(edma_noevent, EDMA_MAX_DMACH);
+
+/* dummy param set used to (re)initialize parameter RAM slots */
+static const struct edmacc_param dummy_paramset = {
+	.link_bcntrld = 0xffff,
+	.ccnt = 1,
+};
+
+static const int __initconst
+queue_tc_mapping[EDMA_MAX_EVQUE + 1][2] = {
+/* {event queue no, TC no} */
+	{0, 0},
+	{1, 1},
+	{-1, -1}
+};
+
+static const int __initconst
+queue_priority_mapping[EDMA_MAX_EVQUE + 1][2] = {
+	/* {event queue no, Priority} */
+	{0, 3},
+	{1, 7},
+	{-1, -1}
+};
+
+/*****************************************************************************/
+
+static void map_dmach_queue(unsigned ch_no, enum dma_event_q queue_no)
+{
+	int bit = (ch_no & 0x7) * 4;
+
+	/* default to low priority queue */
+	if (queue_no == EVENTQ_DEFAULT)
+		queue_no = EVENTQ_1;
+
+	queue_no &= 7;
+	edma_modify_array(EDMA_DMAQNUM, (ch_no >> 3),
+			~(0x7 << bit), queue_no << bit);
+}
+
+static void __init map_queue_tc(int queue_no, int tc_no)
+{
+	int bit = queue_no * 4;
+	edma_modify(EDMA_QUETCMAP, ~(0x7 << bit), ((tc_no & 0x7) << bit));
+}
+
+static void __init assign_priority_to_queue(int queue_no, int priority)
+{
+	int bit = queue_no * 4;
+	edma_modify(EDMA_QUEPRI, ~(0x7 << bit), ((priority & 0x7) << bit));
+}
+
+static inline void
+setup_dma_interrupt(unsigned lch,
+	void (*callback)(unsigned channel, u16 ch_status, void *data),
+	void *data)
+{
+	if (!callback) {
+		edma_shadow0_write_array(SH_IECR, lch >> 5,
+				(1 << (lch & 0x1f)));
+	}
+
+	intr_data[lch].callback = callback;
+	intr_data[lch].data = data;
+
+	if (callback) {
+		edma_shadow0_write_array(SH_ICR, lch >> 5,
+				(1 << (lch & 0x1f)));
+		edma_shadow0_write_array(SH_IESR, lch >> 5,
+				(1 << (lch & 0x1f)));
+	}
+}
+
+/******************************************************************************
+ *
+ * DMA interrupt handler
+ *
+ *****************************************************************************/
+static irqreturn_t dma_irq_handler(int irq, void *data)
+{
+	int i;
+	unsigned int cnt = 0;
+
+	dev_dbg(data, "dma_irq_handler\n");
+
+	if ((edma_shadow0_read_array(SH_IPR, 0) == 0)
+	    && (edma_shadow0_read_array(SH_IPR, 1) == 0))
+		return IRQ_NONE;
+
+	while (1) {
+		int j;
+		if (edma_shadow0_read_array(SH_IPR, 0))
+			j = 0;
+		else if (edma_shadow0_read_array(SH_IPR, 1))
+			j = 1;
+		else
+			break;
+		dev_dbg(data, "IPR%d %08x\n", j,
+				edma_shadow0_read_array(SH_IPR, j));
+		for (i = 0; i < 32; i++) {
+			int k = (j << 5) + i;
+			if (edma_shadow0_read_array(SH_IPR, j) & (1 << i)) {
+				/* Clear the corresponding IPR bits */
+				edma_shadow0_write_array(SH_ICR, j, (1 << i));
+				if (intr_data[k].callback) {
+					intr_data[k].callback(k, DMA_COMPLETE,
+						intr_data[k].data);
+				}
+			}
+		}
+		cnt++;
+		if (cnt > 10)
+			break;
+	}
+	edma_shadow0_write(SH_IEVAL, 1);
+	return IRQ_HANDLED;
+}
+
+/******************************************************************************
+ *
+ * DMA error interrupt handler
+ *
+ *****************************************************************************/
+static irqreturn_t dma_ccerr_handler(int irq, void *data)
+{
+	int i;
+	unsigned int cnt = 0;
+
+	dev_dbg(data, "dma_ccerr_handler\n");
+
+	if ((edma_read_array(EDMA_EMR, 0) == 0) &&
+	    (edma_read_array(EDMA_EMR, 1) == 0) &&
+	    (edma_read(EDMA_QEMR) == 0) && (edma_read(EDMA_CCERR) == 0))
+		return IRQ_NONE;
+
+	while (1) {
+		int j = -1;
+		if (edma_read_array(EDMA_EMR, 0))
+			j = 0;
+		else if (edma_read_array(EDMA_EMR, 1))
+			j = 1;
+		if (j >= 0) {
+			dev_dbg(data, "EMR%d %08x\n", j,
+					edma_read_array(EDMA_EMR, j));
+			for (i = 0; i < 32; i++) {
+				int k = (j << 5) + i;
+				if (edma_read_array(EDMA_EMR, j) & (1 << i)) {
+					/* Clear the corresponding EMR bits */
+					edma_write_array(EDMA_EMCR, j, 1 << i);
+					/* Clear any SER */
+					edma_shadow0_write_array(SH_SECR, j,
+							(1 << i));
+					if (intr_data[k].callback) {
+						intr_data[k].callback(k,
+								DMA_CC_ERROR,
+								intr_data
+								[k].data);
+					}
+				}
+			}
+		} else if (edma_read(EDMA_QEMR)) {
+			dev_dbg(data, "QEMR %02x\n",
+				edma_read(EDMA_QEMR));
+			for (i = 0; i < 8; i++) {
+				if (edma_read(EDMA_QEMR) & (1 << i)) {
+					/* Clear the corresponding IPR bits */
+					edma_write(EDMA_QEMCR, 1 << i);
+					edma_shadow0_write(SH_QSECR, (1 << i));
+
+					/* NOTE:  not reported!! */
+				}
+			}
+		} else if (edma_read(EDMA_CCERR)) {
+			dev_dbg(data, "CCERR %08x\n",
+				edma_read(EDMA_CCERR));
+			/* FIXME:  CCERR.BIT(16) ignored!  much better
+			 * to just write CCERRCLR with CCERR value...
+			 */
+			for (i = 0; i < 8; i++) {
+				if (edma_read(EDMA_CCERR) & (1 << i)) {
+					/* Clear the corresponding IPR bits */
+					edma_write(EDMA_CCERRCLR, 1 << i);
+
+					/* NOTE:  not reported!! */
+				}
+			}
+		}
+		if ((edma_read_array(EDMA_EMR, 0) == 0)
+		    && (edma_read_array(EDMA_EMR, 1) == 0)
+		    && (edma_read(EDMA_QEMR) == 0)
+		    && (edma_read(EDMA_CCERR) == 0)) {
+			break;
+		}
+		cnt++;
+		if (cnt > 10)
+			break;
+	}
+	edma_write(EDMA_EEVAL, 1);
+	return IRQ_HANDLED;
+}
+
+/******************************************************************************
+ *
+ * Transfer controller error interrupt handlers
+ *
+ *****************************************************************************/
+
+#define tc_errs_handled	false	/* disabled as long as they're NOPs */
+
+static irqreturn_t dma_tc0err_handler(int irq, void *data)
+{
+	dev_dbg(data, "dma_tc0err_handler\n");
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t dma_tc1err_handler(int irq, void *data)
+{
+	dev_dbg(data, "dma_tc1err_handler\n");
+	return IRQ_HANDLED;
+}
+
+/*-----------------------------------------------------------------------*/
+
+/* Resource alloc/free:  dma channels, parameter RAM slots */
+
+/**
+ * edma_alloc_channel - allocate DMA channel and paired parameter RAM
+ * @channel: specific channel to allocate; negative for "any unmapped channel"
+ * @callback: optional; to be issued on DMA completion or errors
+ * @data: passed to callback
+ * @eventq_no: an EVENTQ_* constant, used to choose which Transfer
+ *	Controller (TC) executes requests using this channel.  Use
+ *	EVENTQ_DEFAULT unless you really need a high priority queue.
+ *
+ * This allocates a DMA channel and its associated parameter RAM slot.
+ * The parameter RAM is initialized to hold a dummy transfer.
+ *
+ * Normal use is to pass a specific channel number as @channel, to make
+ * use of hardware events mapped to that channel.  When the channel will
+ * be used only for software triggering or event chaining, channels not
+ * mapped to hardware events (or mapped to unused events) are preferable.
+ *
+ * DMA transfers start from a channel using edma_start(), or by
+ * chaining.  When the transfer described in that channel's parameter RAM
+ * slot completes, that slot's data may be reloaded through a link.
+ *
+ * DMA errors are only reported to the @callback associated with the
+ * channel driving that transfer, but transfer completion callbacks can
+ * be sent to another channel under control of the TCC field in
+ * the option word of the transfer's parameter RAM set.  Drivers must not
+ * use DMA transfer completion callbacks for channels they did not allocate.
+ * (The same applies to TCC codes used in transfer chaining.)
+ *
+ * Returns the number of the channel, else negative errno.
+ */
+int edma_alloc_channel(int channel,
+		void (*callback)(unsigned channel, u16 ch_status, void *data),
+		void *data,
+		enum dma_event_q eventq_no)
+{
+	if (channel < 0) {
+		channel = 0;
+		for (;;) {
+			channel = find_next_bit(edma_noevent,
+					num_channels, channel);
+			if (channel == num_channels)
+				return -ENOMEM;
+			if (!test_and_set_bit(channel, edma_inuse))
+				break;
+			channel++;
+		}
+	} else if (channel >= num_channels) {
+		return -EINVAL;
+	} else if (test_and_set_bit(channel, edma_inuse)) {
+		return -EBUSY;
+	}
+
+	/* ensure access through shadow region 0 */
+	edma_or_array2(EDMA_DRAE, 0, channel >> 5, 1 << (channel & 0x1f));
+
+	/* ensure no events are pending */
+	edma_stop(channel);
+	memcpy_toio(edmacc_regs_base + PARM_OFFSET(channel),
+			&dummy_paramset, PARM_SIZE);
+
+	if (callback)
+		setup_dma_interrupt(channel, callback, data);
+
+	map_dmach_queue(channel, eventq_no);
+
+	return channel;
+}
+EXPORT_SYMBOL(edma_alloc_channel);
+
+
+/**
+ * edma_free_channel - deallocate DMA channel
+ * @channel: dma channel returned from edma_alloc_channel()
+ *
+ * This deallocates the DMA channel and associated parameter RAM slot
+ * allocated by edma_alloc_channel().
+ *
+ * Callers are responsible for ensuring the channel is inactive, and
+ * will not be reactivated by linking, chaining, or software calls to
+ * edma_start().
+ */
+void edma_free_channel(unsigned channel)
+{
+	if (channel >= num_channels)
+		return;
+
+	setup_dma_interrupt(channel, NULL, NULL);
+	/* REVISIT should probably take out of shadow region 0 */
+
+	memcpy_toio(edmacc_regs_base + PARM_OFFSET(channel),
+			&dummy_paramset, PARM_SIZE);
+	clear_bit(channel, edma_inuse);
+}
+EXPORT_SYMBOL(edma_free_channel);
+
+/**
+ * edma_alloc_slot - allocate DMA parameter RAM
+ * @slot: specific slot to allocate; negative for "any unused slot"
+ *
+ * This allocates a parameter RAM slot, initializing it to hold a
+ * dummy transfer.  Slots allocated using this routine have not been
+ * mapped to a hardware DMA channel, and will normally be used by
+ * linking to them from a slot associated with a DMA channel.
+ *
+ * Normal use is to pass EDMA_SLOT_ANY as the @slot, but specific
+ * slots may be allocated on behalf of DSP firmware.
+ *
+ * Returns the number of the slot, else negative errno.
+ */
+int edma_alloc_slot(int slot)
+{
+	if (slot < 0) {
+		slot = num_channels;
+		for (;;) {
+			slot = find_next_zero_bit(edma_inuse,
+					num_slots, slot);
+			if (slot == num_slots)
+				return -ENOMEM;
+			if (!test_and_set_bit(slot, edma_inuse))
+				break;
+		}
+	} else if (slot < num_channels || slot >= num_slots) {
+		return -EINVAL;
+	} else if (test_and_set_bit(slot, edma_inuse)) {
+		return -EBUSY;
+	}
+
+	memcpy_toio(edmacc_regs_base + PARM_OFFSET(slot),
+			&dummy_paramset, PARM_SIZE);
+
+	return slot;
+}
+EXPORT_SYMBOL(edma_alloc_slot);
+
+/**
+ * edma_free_slot - deallocate DMA parameter RAM
+ * @slot: parameter RAM slot returned from edma_alloc_slot()
+ *
+ * This deallocates the parameter RAM slot allocated by edma_alloc_slot().
+ * Callers are responsible for ensuring the slot is inactive, and will
+ * not be activated.
+ */
+void edma_free_slot(unsigned slot)
+{
+	if (slot < num_channels || slot >= num_slots)
+		return;
+
+	memcpy_toio(edmacc_regs_base + PARM_OFFSET(slot),
+			&dummy_paramset, PARM_SIZE);
+	clear_bit(slot, edma_inuse);
+}
+EXPORT_SYMBOL(edma_free_slot);
+
+/*-----------------------------------------------------------------------*/
+
+/* Parameter RAM operations (i) -- read/write partial slots */
+
+/**
+ * edma_set_src - set initial DMA source address in parameter RAM slot
+ * @slot: parameter RAM slot being configured
+ * @src_port: physical address of source (memory, controller FIFO, etc)
+ * @addressMode: INCR, except in very rare cases
+ * @fifoWidth: ignored unless @addressMode is FIFO, else specifies the
+ *	width to use when addressing the fifo (e.g. W8BIT, W32BIT)
+ *
+ * Note that the source address is modified during the DMA transfer
+ * according to edma_set_src_index().
+ */
+void edma_set_src(unsigned slot, dma_addr_t src_port,
+				enum address_mode mode, enum fifo_width width)
+{
+	if (slot < num_slots) {
+		unsigned int i = edma_parm_read(PARM_OPT, slot);
+
+		if (mode) {
+			/* set SAM and program FWID */
+			i = (i & ~(EDMA_FWID)) | (SAM | ((width & 0x7) << 8));
+		} else {
+			/* clear SAM */
+			i &= ~SAM;
+		}
+		edma_parm_write(PARM_OPT, slot, i);
+
+		/* set the source port address
+		   in source register of param structure */
+		edma_parm_write(PARM_SRC, slot, src_port);
+	}
+}
+EXPORT_SYMBOL(edma_set_src);
+
+/**
+ * edma_set_dest - set initial DMA destination address in parameter RAM slot
+ * @slot: parameter RAM slot being configured
+ * @dest_port: physical address of destination (memory, controller FIFO, etc)
+ * @addressMode: INCR, except in very rare cases
+ * @fifoWidth: ignored unless @addressMode is FIFO, else specifies the
+ *	width to use when addressing the fifo (e.g. W8BIT, W32BIT)
+ *
+ * Note that the destination address is modified during the DMA transfer
+ * according to edma_set_dest_index().
+ */
+void edma_set_dest(unsigned slot, dma_addr_t dest_port,
+				 enum address_mode mode, enum fifo_width width)
+{
+	if (slot < num_slots) {
+		unsigned int i = edma_parm_read(PARM_OPT, slot);
+
+		if (mode) {
+			/* set DAM and program FWID */
+			i = (i & ~(EDMA_FWID)) | (DAM | ((width & 0x7) << 8));
+		} else {
+			/* clear DAM */
+			i &= ~DAM;
+		}
+		edma_parm_write(PARM_OPT, slot, i);
+		/* set the destination port address
+		   in dest register of param structure */
+		edma_parm_write(PARM_DST, slot, dest_port);
+	}
+}
+EXPORT_SYMBOL(edma_set_dest);
+
+/**
+ * edma_get_position - returns the current transfer points
+ * @slot: parameter RAM slot being examined
+ * @src: pointer to source port position
+ * @dst: pointer to destination port position
+ *
+ * Returns current source and destination addresses for a particular
+ * parameter RAM slot.  Its channel should not be active when this is called.
+ */
+void edma_get_position(unsigned slot, dma_addr_t *src, dma_addr_t *dst)
+{
+	struct edmacc_param temp;
+
+	edma_read_slot(slot, &temp);
+	if (src != NULL)
+		*src = temp.src;
+	if (dst != NULL)
+		*dst = temp.dst;
+}
+EXPORT_SYMBOL(edma_get_position);
+
+/**
+ * edma_set_src_index - configure DMA source address indexing
+ * @slot: parameter RAM slot being configured
+ * @src_bidx: byte offset between source arrays in a frame
+ * @src_cidx: byte offset between source frames in a block
+ *
+ * Offsets are specified to support either contiguous or discontiguous
+ * memory transfers, or repeated access to a hardware register, as needed.
+ * When accessing hardware registers, both offsets are normally zero.
+ */
+void edma_set_src_index(unsigned slot, s16 src_bidx, s16 src_cidx)
+{
+	if (slot < num_slots) {
+		edma_parm_modify(PARM_SRC_DST_BIDX, slot,
+				0xffff0000, src_bidx);
+		edma_parm_modify(PARM_SRC_DST_CIDX, slot,
+				0xffff0000, src_cidx);
+	}
+}
+EXPORT_SYMBOL(edma_set_src_index);
+
+/**
+ * edma_set_dest_index - configure DMA destination address indexing
+ * @slot: parameter RAM slot being configured
+ * @dest_bidx: byte offset between destination arrays in a frame
+ * @dest_cidx: byte offset between destination frames in a block
+ *
+ * Offsets are specified to support either contiguous or discontiguous
+ * memory transfers, or repeated access to a hardware register, as needed.
+ * When accessing hardware registers, both offsets are normally zero.
+ */
+void edma_set_dest_index(unsigned slot, s16 dest_bidx, s16 dest_cidx)
+{
+	if (slot < num_slots) {
+		edma_parm_modify(PARM_SRC_DST_BIDX, slot,
+				0x0000ffff, dest_bidx << 16);
+		edma_parm_modify(PARM_SRC_DST_CIDX, slot,
+				0x0000ffff, dest_cidx << 16);
+	}
+}
+EXPORT_SYMBOL(edma_set_dest_index);
+
+/**
+ * edma_set_transfer_params - configure DMA transfer parameters
+ * @slot: parameter RAM slot being configured
+ * @acnt: how many bytes per array (at least one)
+ * @bcnt: how many arrays per frame (at least one)
+ * @ccnt: how many frames per block (at least one)
+ * @bcnt_rld: used only for A-Synchronized transfers; this specifies
+ *	the value to reload into bcnt when it decrements to zero
+ * @sync_mode: ASYNC or ABSYNC
+ *
+ * See the EDMA3 documentation to understand how to configure and link
+ * transfers using the fields in PaRAM slots.  If you are not doing it
+ * all at once with edma_write_slot(), you will use this routine
+ * plus two calls each for source and destination, setting the initial
+ * address and saying how to index that address.
+ *
+ * An example of an A-Synchronized transfer is a serial link using a
+ * single word shift register.  In that case, @acnt would be equal to
+ * that word size; the serial controller issues a DMA synchronization
+ * event to transfer each word, and memory access by the DMA transfer
+ * controller will be word-at-a-time.
+ *
+ * An example of an AB-Synchronized transfer is a device using a FIFO.
+ * In that case, @acnt equals the FIFO width and @bcnt equals its depth.
+ * The controller with the FIFO issues DMA synchronization events when
+ * the FIFO threshold is reached, and the DMA transfer controller will
+ * transfer one frame to (or from) the FIFO.  It will probably use
+ * efficient burst modes to access memory.
+ */
+void edma_set_transfer_params(unsigned slot,
+		u16 acnt, u16 bcnt, u16 ccnt,
+		u16 bcnt_rld, enum sync_dimension sync_mode)
+{
+	if (slot < num_slots) {
+		edma_parm_modify(PARM_LINK_BCNTRLD, slot,
+				0x0000ffff, bcnt_rld << 16);
+		if (sync_mode == ASYNC)
+			edma_parm_and(PARM_OPT, slot, ~SYNCDIM);
+		else
+			edma_parm_or(PARM_OPT, slot, SYNCDIM);
+		/* Set the acount, bcount, ccount registers */
+		edma_parm_write(PARM_A_B_CNT, slot, (bcnt << 16) | acnt);
+		edma_parm_write(PARM_CCNT, slot, ccnt);
+	}
+}
+EXPORT_SYMBOL(edma_set_transfer_params);
+
+/**
+ * edma_link - link one parameter RAM slot to another
+ * @from: parameter RAM slot originating the link
+ * @to: parameter RAM slot which is the link target
+ *
+ * The originating slot should not be part of any active DMA transfer.
+ */
+void edma_link(unsigned from, unsigned to)
+{
+	if (from >= num_slots)
+		return;
+	if (to >= num_slots)
+		return;
+	edma_parm_modify(PARM_LINK_BCNTRLD, from, 0xffff0000, PARM_OFFSET(to));
+}
+EXPORT_SYMBOL(edma_link);
+
+/**
+ * edma_unlink - cut link from one parameter RAM slot
+ * @from: parameter RAM slot originating the link
+ *
+ * The originating slot should not be part of any active DMA transfer.
+ * Its link is set to 0xffff.
+ */
+void edma_unlink(unsigned from)
+{
+	if (from >= num_slots)
+		return;
+	edma_parm_or(PARM_LINK_BCNTRLD, from, 0xffff);
+}
+EXPORT_SYMBOL(edma_unlink);
+
+/*-----------------------------------------------------------------------*/
+
+/* Parameter RAM operations (ii) -- read/write whole parameter sets */
+
+/**
+ * edma_write_slot - write parameter RAM data for slot
+ * @slot: number of parameter RAM slot being modified
+ * @param: data to be written into parameter RAM slot
+ *
+ * Use this to assign all parameters of a transfer at once.  This
+ * allows more efficient setup of transfers than issuing multiple
+ * calls to set up those parameters in small pieces, and provides
+ * complete control over all transfer options.
+ */
+void edma_write_slot(unsigned slot, const struct edmacc_param *param)
+{
+	if (slot >= num_slots)
+		return;
+	memcpy_toio(edmacc_regs_base + PARM_OFFSET(slot), param, PARM_SIZE);
+}
+EXPORT_SYMBOL(edma_write_slot);
+
+/**
+ * edma_read_slot - read parameter RAM data from slot
+ * @slot: number of parameter RAM slot being copied
+ * @param: where to store copy of parameter RAM data
+ *
+ * Use this to read data from a parameter RAM slot, perhaps to
+ * save them as a template for later reuse.
+ */
+void edma_read_slot(unsigned slot, struct edmacc_param *param)
+{
+	if (slot >= num_slots)
+		return;
+	memcpy_fromio(param, edmacc_regs_base + PARM_OFFSET(slot), PARM_SIZE);
+}
+EXPORT_SYMBOL(edma_read_slot);
+
+/*-----------------------------------------------------------------------*/
+
+/* Various EDMA channel control operations */
+
+/**
+ * edma_pause - pause dma on a channel
+ * @channel: on which edma_start() has been called
+ *
+ * This temporarily disables EDMA hardware events on the specified channel,
+ * preventing them from triggering new transfers on its behalf
+ */
+void edma_pause(unsigned channel)
+{
+	if (channel < num_channels) {
+		unsigned int mask = (1 << (channel & 0x1f));
+
+		edma_shadow0_write_array(SH_EECR, channel >> 5, mask);
+	}
+}
+EXPORT_SYMBOL(edma_pause);
+
+/**
+ * edma_resume - resumes dma on a paused channel
+ * @channel: on which edma_pause() has been called
+ *
+ * This re-enables EDMA hardware events on the specified channel.
+ */
+void edma_resume(unsigned channel)
+{
+	if (channel < num_channels) {
+		unsigned int mask = (1 << (channel & 0x1f));
+
+		edma_shadow0_write_array(SH_EESR, channel >> 5, mask);
+	}
+}
+EXPORT_SYMBOL(edma_resume);
+
+/**
+ * edma_start - start dma on a channel
+ * @channel: channel being activated
+ *
+ * Channels with event associations will be triggered by their hardware
+ * events, and channels without such associations will be triggered by
+ * software.  (At this writing there is no interface for using software
+ * triggers except with channels that don't support hardware triggers.)
+ *
+ * Returns zero on success, else negative errno.
+ */
+int edma_start(unsigned channel)
+{
+	if (channel < num_channels) {
+		int j = channel >> 5;
+		unsigned int mask = (1 << (channel & 0x1f));
+
+		/* EDMA channels without event association */
+		if (test_bit(channel, edma_noevent)) {
+			pr_debug("EDMA: ESR%d %08x\n", j,
+				edma_shadow0_read_array(SH_ESR, j));
+			edma_shadow0_write_array(SH_ESR, j, mask);
+			return 0;
+		}
+
+		/* EDMA channel with event association */
+		pr_debug("EDMA: ER%d %08x\n", j,
+			edma_shadow0_read_array(SH_ER, j));
+		/* Clear any pending error */
+		edma_write_array(EDMA_EMCR, j, mask);
+		/* Clear any SER */
+		edma_shadow0_write_array(SH_SECR, j, mask);
+		edma_shadow0_write_array(SH_EESR, j, mask);
+		pr_debug("EDMA: EER%d %08x\n", j,
+			edma_shadow0_read_array(SH_EER, j));
+		return 0;
+	}
+
+	return -EINVAL;
+}
+EXPORT_SYMBOL(edma_start);
+
+/**
+ * edma_stop - stops dma on the channel passed
+ * @channel: channel being deactivated
+ *
+ * When @lch is a channel, any active transfer is paused and
+ * all pending hardware events are cleared.  The current transfer
+ * may not be resumed, and the channel's Parameter RAM should be
+ * reinitialized before being reused.
+ */
+void edma_stop(unsigned channel)
+{
+	if (channel < num_channels) {
+		int j = channel >> 5;
+		unsigned int mask = (1 << (channel & 0x1f));
+
+		edma_shadow0_write_array(SH_EECR, j, mask);
+		edma_shadow0_write_array(SH_ECR, j, mask);
+		edma_shadow0_write_array(SH_SECR, j, mask);
+		edma_write_array(EDMA_EMCR, j, mask);
+
+		pr_debug("EDMA: EER%d %08x\n", j,
+				edma_shadow0_read_array(SH_EER, j));
+
+		/* REVISIT:  consider guarding against inappropriate event
+		 * chaining by overwriting with dummy_paramset.
+		 */
+	}
+}
+EXPORT_SYMBOL(edma_stop);
+
+/******************************************************************************
+ *
+ * It cleans ParamEntry qand bring back EDMA to initial state if media has
+ * been removed before EDMA has finished.It is usedful for removable media.
+ * Arguments:
+ *      ch_no     - channel no
+ *
+ * Return: zero on success, or corresponding error no on failure
+ *
+ * FIXME this should not be needed ... edma_stop() should suffice.
+ *
+ *****************************************************************************/
+
+void edma_clean_channel(unsigned channel)
+{
+	if (channel < num_channels) {
+		int j = (channel >> 5);
+		unsigned int mask = 1 << (channel & 0x1f);
+
+		pr_debug("EDMA: EMR%d %08x\n", j,
+				edma_read_array(EDMA_EMR, j));
+		edma_shadow0_write_array(SH_ECR, j, mask);
+		/* Clear the corresponding EMR bits */
+		edma_write_array(EDMA_EMCR, j, mask);
+		/* Clear any SER */
+		edma_shadow0_write_array(SH_SECR, j, mask);
+		edma_write(EDMA_CCERRCLR, (1 << 16) | 0x3);
+	}
+}
+EXPORT_SYMBOL(edma_clean_channel);
+
+/*
+ * edma_clear_event - clear an outstanding event on the DMA channel
+ * Arguments:
+ *	channel - channel number
+ */
+void edma_clear_event(unsigned channel)
+{
+	if (channel >= num_channels)
+		return;
+	if (channel < 32)
+		edma_write(EDMA_ECR, 1 << channel);
+	else
+		edma_write(EDMA_ECRH, 1 << (channel - 32));
+}
+EXPORT_SYMBOL(edma_clear_event);
+
+/*-----------------------------------------------------------------------*/
+
+static int __init edma_probe(struct platform_device *pdev)
+{
+	struct edma_soc_info	*info = pdev->dev.platform_data;
+	int			i;
+	int			status;
+	const s8		*noevent;
+	int			irq = 0, err_irq = 0;
+	struct resource		*r;
+	resource_size_t		len;
+
+	if (!info)
+		return -ENODEV;
+
+	r = platform_get_resource_byname(pdev, IORESOURCE_MEM, "edma_cc");
+	if (!r)
+		return -ENODEV;
+
+	len = r->end - r->start + 1;
+
+	r = request_mem_region(r->start, len, r->name);
+	if (!r)
+		return -EBUSY;
+
+	edmacc_regs_base = ioremap(r->start, len);
+	if (!edmacc_regs_base) {
+		status = -EBUSY;
+		goto fail1;
+	}
+
+	num_channels = min_t(unsigned, info->n_channel, EDMA_MAX_DMACH);
+	num_slots = min_t(unsigned, info->n_slot, EDMA_MAX_PARAMENTRY);
+
+	dev_dbg(&pdev->dev, "DMA REG BASE ADDR=%p\n", edmacc_regs_base);
+
+	for (i = 0; i < num_slots; i++)
+		memcpy_toio(edmacc_regs_base + PARM_OFFSET(i),
+				&dummy_paramset, PARM_SIZE);
+
+	noevent = info->noevent;
+	if (noevent) {
+		while (*noevent != -1)
+			set_bit(*noevent++, edma_noevent);
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	status = request_irq(irq, dma_irq_handler, 0, "edma", &pdev->dev);
+	if (status < 0) {
+		dev_dbg(&pdev->dev, "request_irq %d failed --> %d\n",
+			irq, status);
+		goto fail;
+	}
+
+	err_irq = platform_get_irq(pdev, 1);
+	status = request_irq(err_irq, dma_ccerr_handler, 0,
+				"edma_error", &pdev->dev);
+	if (status < 0) {
+		dev_dbg(&pdev->dev, "request_irq %d failed --> %d\n",
+			err_irq, status);
+		goto fail;
+	}
+
+	if (tc_errs_handled) {
+		status = request_irq(IRQ_TCERRINT0, dma_tc0err_handler, 0,
+					"edma_tc0", &pdev->dev);
+		if (status < 0) {
+			dev_dbg(&pdev->dev, "request_irq %d failed --> %d\n",
+				IRQ_TCERRINT0, status);
+			return status;
+		}
+		status = request_irq(IRQ_TCERRINT, dma_tc1err_handler, 0,
+					"edma_tc1", &pdev->dev);
+		if (status < 0) {
+			dev_dbg(&pdev->dev, "request_irq %d --> %d\n",
+				IRQ_TCERRINT, status);
+			return status;
+		}
+	}
+
+	/* Everything lives on transfer controller 1 until otherwise specified.
+	 * This way, long transfers on the low priority queue
+	 * started by the codec engine will not cause audio defects.
+	 */
+	for (i = 0; i < num_channels; i++)
+		map_dmach_queue(i, EVENTQ_1);
+
+	/* Event queue to TC mapping */
+	for (i = 0; queue_tc_mapping[i][0] != -1; i++)
+		map_queue_tc(queue_tc_mapping[i][0], queue_tc_mapping[i][1]);
+
+	/* Event queue priority mapping */
+	for (i = 0; queue_priority_mapping[i][0] != -1; i++)
+		assign_priority_to_queue(queue_priority_mapping[i][0],
+					 queue_priority_mapping[i][1]);
+
+	for (i = 0; i < info->n_region; i++) {
+		edma_write_array2(EDMA_DRAE, i, 0, 0x0);
+		edma_write_array2(EDMA_DRAE, i, 1, 0x0);
+		edma_write_array(EDMA_QRAE, i, 0x0);
+	}
+
+	return 0;
+
+fail:
+	if (err_irq)
+		free_irq(err_irq, NULL);
+	if (irq)
+		free_irq(irq, NULL);
+	iounmap(edmacc_regs_base);
+fail1:
+	release_mem_region(r->start, len);
+	return status;
+}
+
+
+static struct platform_driver edma_driver = {
+	.driver.name	= "edma",
+};
+
+static int __init edma_init(void)
+{
+	return platform_driver_probe(&edma_driver, edma_probe);
+}
+arch_initcall(edma_init);
+
diff --git a/arch/arm/mach-davinci/gpio.c b/arch/arm/mach-davinci/gpio.c
index b49e9d092aa..1aba41c6351 100644
--- a/arch/arm/mach-davinci/gpio.c
+++ b/arch/arm/mach-davinci/gpio.c
@@ -20,6 +20,7 @@
 #include <linux/irq.h>
 #include <linux/bitops.h>
 
+#include <mach/cputype.h>
 #include <mach/irqs.h>
 #include <mach/hardware.h>
 #include <mach/gpio.h>
@@ -36,9 +37,10 @@ struct davinci_gpio {
 
 static struct davinci_gpio chips[DIV_ROUND_UP(DAVINCI_N_GPIO, 32)];
 
+static unsigned __initdata ngpio;
 
 /* create a non-inlined version */
-static struct gpio_controller *__iomem __init gpio2controller(unsigned gpio)
+static struct gpio_controller __iomem * __init gpio2controller(unsigned gpio)
 {
 	return __gpio_to_controller(gpio);
 }
@@ -114,9 +116,30 @@ static int __init davinci_gpio_setup(void)
 {
 	int i, base;
 
-	for (i = 0, base = 0;
-			i < ARRAY_SIZE(chips);
-			i++, base += 32) {
+	/* The gpio banks conceptually expose a segmented bitmap,
+	 * and "ngpio" is one more than the largest zero-based
+	 * bit index that's valid.
+	 */
+	if (cpu_is_davinci_dm355()) {		/* or dm335() */
+		ngpio = 104;
+	} else if (cpu_is_davinci_dm644x()) {	/* or dm337() */
+		ngpio = 71;
+	} else if (cpu_is_davinci_dm646x()) {
+		/* NOTE:  each bank has several "reserved" bits,
+		 * unusable as GPIOs.  Only 33 of the GPIO numbers
+		 * are usable, and we're not rejecting the others.
+		 */
+		ngpio = 43;
+	} else {
+		/* if cpu_is_davinci_dm643x() ngpio = 111 */
+		pr_err("GPIO setup:  how many GPIOs?\n");
+		return -EINVAL;
+	}
+
+	if (WARN_ON(DAVINCI_N_GPIO < ngpio))
+		ngpio = DAVINCI_N_GPIO;
+
+	for (i = 0, base = 0; base < ngpio; i++, base += 32) {
 		chips[i].chip.label = "DaVinci";
 
 		chips[i].chip.direction_input = davinci_direction_in;
@@ -125,7 +148,7 @@ static int __init davinci_gpio_setup(void)
 		chips[i].chip.set = davinci_gpio_set;
 
 		chips[i].chip.base = base;
-		chips[i].chip.ngpio = DAVINCI_N_GPIO - base;
+		chips[i].chip.ngpio = ngpio - base;
 		if (chips[i].chip.ngpio > 32)
 			chips[i].chip.ngpio = 32;
 
@@ -143,11 +166,11 @@ pure_initcall(davinci_gpio_setup);
  * We expect irqs will normally be set up as input pins, but they can also be
  * used as output pins ... which is convenient for testing.
  *
- * NOTE:  GPIO0..GPIO7 also have direct INTC hookups, which work in addition
- * to their GPIOBNK0 irq (but with a bit less overhead).  But we don't have
- * a good way to hook those up ...
+ * NOTE:  The first few GPIOs also have direct INTC hookups in addition
+ * to their GPIOBNK0 irq, with a bit less overhead but less flexibility
+ * on triggering (e.g. no edge options).  We don't try to use those.
  *
- * All those INTC hookups (GPIO0..GPIO7 plus five IRQ banks) can also
+ * All those INTC hookups (direct, plus several IRQ banks) can also
  * serve as EDMA event triggers.
  */
 
@@ -235,29 +258,42 @@ gpio_irq_handler(unsigned irq, struct irq_desc *desc)
 }
 
 /*
- * NOTE:  for suspend/resume, probably best to make a sysdev (and class)
- * with its suspend/resume calls hooking into the results of the set_wake()
+ * NOTE:  for suspend/resume, probably best to make a platform_device with
+ * suspend_late/resume_resume calls hooking into results of the set_wake()
  * calls ... so if no gpios are wakeup events the clock can be disabled,
  * with outputs left at previously set levels, and so that VDD3P3V.IOPWDN0
- * can be set appropriately for GPIOV33 pins.
+ * (dm6446) can be set appropriately for GPIOV33 pins.
  */
 
 static int __init davinci_gpio_irq_setup(void)
 {
 	unsigned	gpio, irq, bank;
+	unsigned	bank_irq;
 	struct clk	*clk;
+	u32		binten = 0;
+
+	if (cpu_is_davinci_dm355()) {		/* or dm335() */
+		bank_irq = IRQ_DM355_GPIOBNK0;
+	} else if (cpu_is_davinci_dm644x()) {
+		bank_irq = IRQ_GPIOBNK0;
+	} else if (cpu_is_davinci_dm646x()) {
+		bank_irq = IRQ_DM646X_GPIOBNK0;
+	} else {
+		printk(KERN_ERR "Don't know first GPIO bank IRQ.\n");
+		return -EINVAL;
+	}
 
 	clk = clk_get(NULL, "gpio");
 	if (IS_ERR(clk)) {
 		printk(KERN_ERR "Error %ld getting gpio clock?\n",
 		       PTR_ERR(clk));
-		return 0;
+		return PTR_ERR(clk);
 	}
-
 	clk_enable(clk);
 
-	for (gpio = 0, irq = gpio_to_irq(0), bank = IRQ_GPIOBNK0;
-	     gpio < DAVINCI_N_GPIO; bank++) {
+	for (gpio = 0, irq = gpio_to_irq(0), bank = 0;
+			gpio < ngpio;
+			bank++, bank_irq++) {
 		struct gpio_controller	*__iomem g = gpio2controller(gpio);
 		unsigned		i;
 
@@ -265,28 +301,28 @@ static int __init davinci_gpio_irq_setup(void)
 		__raw_writel(~0, &g->clr_rising);
 
 		/* set up all irqs in this bank */
-		set_irq_chained_handler(bank, gpio_irq_handler);
-		set_irq_chip_data(bank, g);
-		set_irq_data(bank, (void *)irq);
+		set_irq_chained_handler(bank_irq, gpio_irq_handler);
+		set_irq_chip_data(bank_irq, g);
+		set_irq_data(bank_irq, (void *)irq);
 
-		for (i = 0; i < 16 && gpio < DAVINCI_N_GPIO;
-		     i++, irq++, gpio++) {
+		for (i = 0; i < 16 && gpio < ngpio; i++, irq++, gpio++) {
 			set_irq_chip(irq, &gpio_irqchip);
 			set_irq_chip_data(irq, g);
 			set_irq_handler(irq, handle_simple_irq);
 			set_irq_flags(irq, IRQF_VALID);
 		}
+
+		binten |= BIT(bank);
 	}
 
 	/* BINTEN -- per-bank interrupt enable. genirq would also let these
 	 * bits be set/cleared dynamically.
 	 */
-	__raw_writel(0x1f, (void *__iomem)
+	__raw_writel(binten, (void *__iomem)
 		     IO_ADDRESS(DAVINCI_GPIO_BASE + 0x08));
 
 	printk(KERN_INFO "DaVinci: %d gpio irqs\n", irq - gpio_to_irq(0));
 
 	return 0;
 }
-
 arch_initcall(davinci_gpio_irq_setup);
diff --git a/arch/arm/mach-davinci/id.c b/arch/arm/mach-davinci/id.c
index bf067d60491..018b994cd79 100644
--- a/arch/arm/mach-davinci/id.c
+++ b/arch/arm/mach-davinci/id.c
@@ -15,7 +15,9 @@
 #include <linux/init.h>
 #include <linux/io.h>
 
-#define JTAG_ID_BASE		0x01c40028
+#define JTAG_ID_BASE		IO_ADDRESS(0x01c40028)
+
+static unsigned int davinci_revision;
 
 struct davinci_id {
 	u8	variant;	/* JTAG ID bits 31:28 */
@@ -33,6 +35,20 @@ static struct davinci_id davinci_ids[] __initdata = {
 		.manufacturer = 0x017,
 		.type	      = 0x64460000,
 	},
+	{
+		/* DM646X */
+		.part_no      = 0xb770,
+		.variant      = 0x0,
+		.manufacturer = 0x017,
+		.type         = 0x64670000,
+	},
+	{
+		/* DM355 */
+		.part_no	= 0xb73b,
+		.variant	= 0x0,
+		.manufacturer	= 0x00f,
+		.type		= 0x03550000,
+	},
 };
 
 /*
@@ -42,7 +58,7 @@ static u16 __init davinci_get_part_no(void)
 {
 	u32 dev_id, part_no;
 
-	dev_id = davinci_readl(JTAG_ID_BASE);
+	dev_id = __raw_readl(JTAG_ID_BASE);
 
 	part_no = ((dev_id >> 12) & 0xffff);
 
@@ -56,13 +72,19 @@ static u8 __init davinci_get_variant(void)
 {
 	u32 variant;
 
-	variant = davinci_readl(JTAG_ID_BASE);
+	variant = __raw_readl(JTAG_ID_BASE);
 
 	variant = (variant >> 28) & 0xf;
 
 	return variant;
 }
 
+unsigned int davinci_rev(void)
+{
+	return davinci_revision >> 16;
+}
+EXPORT_SYMBOL(davinci_rev);
+
 void __init davinci_check_revision(void)
 {
 	int i;
@@ -75,7 +97,7 @@ void __init davinci_check_revision(void)
 	/* First check only the major version in a safe way */
 	for (i = 0; i < ARRAY_SIZE(davinci_ids); i++) {
 		if (part_no == (davinci_ids[i].part_no)) {
-			system_rev = davinci_ids[i].type;
+			davinci_revision = davinci_ids[i].type;
 			break;
 		}
 	}
@@ -84,10 +106,11 @@ void __init davinci_check_revision(void)
 	for (i = 0; i < ARRAY_SIZE(davinci_ids); i++) {
 		if (part_no == davinci_ids[i].part_no &&
 		    variant == davinci_ids[i].variant) {
-			system_rev = davinci_ids[i].type;
+			davinci_revision = davinci_ids[i].type;
 			break;
 		}
 	}
 
-	printk("DaVinci DM%04x variant 0x%x\n", system_rev >> 16, variant);
+	printk(KERN_INFO "DaVinci DM%04x variant 0x%x\n",
+	       davinci_rev(), variant);
 }
diff --git a/arch/arm/mach-davinci/include/mach/board-dm6446evm.h b/arch/arm/mach-davinci/include/mach/board-dm6446evm.h
new file mode 100644
index 00000000000..3216f21c123
--- /dev/null
+++ b/arch/arm/mach-davinci/include/mach/board-dm6446evm.h
@@ -0,0 +1,20 @@
+/*
+ * DaVinci DM6446 EVM board specific headers
+ *
+ * Author: Kevin Hilman, Deep Root Systems, LLC
+ *
+ * 2007 (c) Deep Root Systems, LLC. This file is licensed under
+ * the terms of the GNU General Public License version 2. This program
+ * is licensed "as is" without any warranty of any kind, whether express
+ * or ifndef.
+ */
+
+#ifndef _MACH_DAVINCI_DM6446EVM_H
+#define _MACH_DAVINCI_DM6446EVM_H
+
+#include <linux/types.h>
+
+int dm6446evm_eeprom_read(char *buf, off_t off, size_t count);
+int dm6446evm_eeprom_write(char *buf, off_t off, size_t count);
+
+#endif
diff --git a/arch/arm/mach-davinci/include/mach/clkdev.h b/arch/arm/mach-davinci/include/mach/clkdev.h
new file mode 100644
index 00000000000..730c49d1ebd
--- /dev/null
+++ b/arch/arm/mach-davinci/include/mach/clkdev.h
@@ -0,0 +1,13 @@
+#ifndef __MACH_CLKDEV_H
+#define __MACH_CLKDEV_H
+
+static inline int __clk_get(struct clk *clk)
+{
+	return 1;
+}
+
+static inline void __clk_put(struct clk *clk)
+{
+}
+
+#endif
diff --git a/arch/arm/mach-davinci/include/mach/clock.h b/arch/arm/mach-davinci/include/mach/clock.h
index 38bdd49bc18..a3b04021987 100644
--- a/arch/arm/mach-davinci/include/mach/clock.h
+++ b/arch/arm/mach-davinci/include/mach/clock.h
@@ -17,6 +17,5 @@ struct clk;
 
 extern int clk_register(struct clk *clk);
 extern void clk_unregister(struct clk *clk);
-extern int davinci_clk_init(void);
 
 #endif
diff --git a/arch/arm/mach-davinci/include/mach/common.h b/arch/arm/mach-davinci/include/mach/common.h
index 4b522e5c70e..19177097625 100644
--- a/arch/arm/mach-davinci/include/mach/common.h
+++ b/arch/arm/mach-davinci/include/mach/common.h
@@ -16,6 +16,12 @@ struct sys_timer;
 
 extern struct sys_timer davinci_timer;
 
+extern void davinci_irq_init(void);
+extern void davinci_map_common_io(void);
+
+/* parameters describe VBUS sourcing for host mode */
+extern void setup_usb(unsigned mA, unsigned potpgt_msec);
+
 /* parameters describe VBUS sourcing for host mode */
 extern void setup_usb(unsigned mA, unsigned potpgt_msec);
 
diff --git a/arch/arm/mach-davinci/include/mach/cputype.h b/arch/arm/mach-davinci/include/mach/cputype.h
new file mode 100644
index 00000000000..27cfb1b3a66
--- /dev/null
+++ b/arch/arm/mach-davinci/include/mach/cputype.h
@@ -0,0 +1,49 @@
+/*
+ * DaVinci CPU type detection
+ *
+ * Author: Kevin Hilman, Deep Root Systems, LLC
+ *
+ * Defines the cpu_is_*() macros for runtime detection of DaVinci
+ * device type.  In addtion, if support for a given device is not
+ * compiled in to the kernel, the macros return 0 so that
+ * resulting code can be optimized out.
+ *
+ * 2009 (c) Deep Root Systems, LLC. This file is licensed under
+ * the terms of the GNU General Public License version 2. This program
+ * is licensed "as is" without any warranty of any kind, whether express
+ * or implied.
+ */
+#ifndef _ASM_ARCH_CPU_H
+#define _ASM_ARCH_CPU_H
+
+extern unsigned int davinci_rev(void);
+
+#define IS_DAVINCI_CPU(type, id)			\
+static inline int is_davinci_dm ##type(void)	        \
+{							\
+	return (davinci_rev() == (id)) ? 1 : 0;	        \
+}
+
+IS_DAVINCI_CPU(644x, 0x6446)
+IS_DAVINCI_CPU(646x, 0x6467)
+IS_DAVINCI_CPU(355, 0x355)
+
+#ifdef CONFIG_ARCH_DAVINCI_DM644x
+#define cpu_is_davinci_dm644x() is_davinci_dm644x()
+#else
+#define cpu_is_davinci_dm644x() 0
+#endif
+
+#ifdef CONFIG_ARCH_DAVINCI_DM646x
+#define cpu_is_davinci_dm646x() is_davinci_dm646x()
+#else
+#define cpu_is_davinci_dm646x() 0
+#endif
+
+#ifdef CONFIG_ARCH_DAVINCI_DM355
+#define cpu_is_davinci_dm355() is_davinci_dm355()
+#else
+#define cpu_is_davinci_dm355() 0
+#endif
+
+#endif
diff --git a/arch/arm/mach-davinci/include/mach/dm644x.h b/arch/arm/mach-davinci/include/mach/dm644x.h
new file mode 100644
index 00000000000..3dcb9f4e58b
--- /dev/null
+++ b/arch/arm/mach-davinci/include/mach/dm644x.h
@@ -0,0 +1,37 @@
+/*
+ * This file contains the processor specific definitions
+ * of the TI DM644x.
+ *
+ * Copyright (C) 2008 Texas Instruments.
+ *
+ * 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., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ */
+#ifndef __ASM_ARCH_DM644X_H
+#define __ASM_ARCH_DM644X_H
+
+#include <linux/platform_device.h>
+#include <mach/hardware.h>
+
+#define DM644X_EMAC_BASE		(0x01C80000)
+#define DM644X_EMAC_CNTRL_OFFSET	(0x0000)
+#define DM644X_EMAC_CNTRL_MOD_OFFSET	(0x1000)
+#define DM644X_EMAC_CNTRL_RAM_OFFSET	(0x2000)
+#define DM644X_EMAC_MDIO_OFFSET		(0x4000)
+#define DM644X_EMAC_CNTRL_RAM_SIZE	(0x2000)
+
+void __init dm644x_init(void);
+
+#endif /* __ASM_ARCH_DM644X_H */
diff --git a/arch/arm/mach-davinci/include/mach/edma.h b/arch/arm/mach-davinci/include/mach/edma.h
new file mode 100644
index 00000000000..f6fc5396daf
--- /dev/null
+++ b/arch/arm/mach-davinci/include/mach/edma.h
@@ -0,0 +1,228 @@
+/*
+ *  TI DAVINCI dma definitions
+ *
+ *  Copyright (C) 2006-2009 Texas Instruments.
+ *
+ *  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  SOFTWARE  IS PROVIDED   ``AS  IS'' AND   ANY  EXPRESS OR IMPLIED
+ *  WARRANTIES,   INCLUDING, BUT NOT  LIMITED  TO, THE IMPLIED WARRANTIES OF
+ *  MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN
+ *  NO  EVENT  SHALL   THE AUTHOR  BE    LIABLE FOR ANY   DIRECT, INDIRECT,
+ *  INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ *  NOT LIMITED   TO, PROCUREMENT OF  SUBSTITUTE GOODS  OR SERVICES; LOSS OF
+ *  USE, DATA,  OR PROFITS; OR  BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ *  ANY THEORY OF LIABILITY, WHETHER IN  CONTRACT, STRICT LIABILITY, OR TORT
+ *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
+ *  THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ *  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.
+ *
+ */
+
+/*
+ * This EDMA3 programming framework exposes two basic kinds of resource:
+ *
+ *  Channel	Triggers transfers, usually from a hardware event but
+ *		also manually or by "chaining" from DMA completions.
+ *		Each channel is coupled to a Parameter RAM (PaRAM) slot.
+ *
+ *  Slot	Each PaRAM slot holds a DMA transfer descriptor (PaRAM
+ *		"set"), source and destination addresses, a link to a
+ *		next PaRAM slot (if any), options for the transfer, and
+ *		instructions for updating those addresses.  There are
+ *		more than twice as many slots as event channels.
+ *
+ * Each PaRAM set describes a sequence of transfers, either for one large
+ * buffer or for several discontiguous smaller buffers.  An EDMA transfer
+ * is driven only from a channel, which performs the transfers specified
+ * in its PaRAM slot until there are no more transfers.  When that last
+ * transfer completes, the "link" field may be used to reload the channel's
+ * PaRAM slot with a new transfer descriptor.
+ *
+ * The EDMA Channel Controller (CC) maps requests from channels into physical
+ * Transfer Controller (TC) requests when the channel triggers (by hardware
+ * or software events, or by chaining).  The two physical DMA channels provided
+ * by the TCs are thus shared by many logical channels.
+ *
+ * DaVinci hardware also has a "QDMA" mechanism which is not currently
+ * supported through this interface.  (DSP firmware uses it though.)
+ */
+
+#ifndef EDMA_H_
+#define EDMA_H_
+
+/* PaRAM slots are laid out like this */
+struct edmacc_param {
+	unsigned int opt;
+	unsigned int src;
+	unsigned int a_b_cnt;
+	unsigned int dst;
+	unsigned int src_dst_bidx;
+	unsigned int link_bcntrld;
+	unsigned int src_dst_cidx;
+	unsigned int ccnt;
+};
+
+#define CCINT0_INTERRUPT     16
+#define CCERRINT_INTERRUPT   17
+#define TCERRINT0_INTERRUPT   18
+#define TCERRINT1_INTERRUPT   19
+
+/* fields in edmacc_param.opt */
+#define SAM		BIT(0)
+#define DAM		BIT(1)
+#define SYNCDIM		BIT(2)
+#define STATIC		BIT(3)
+#define EDMA_FWID	(0x07 << 8)
+#define TCCMODE		BIT(11)
+#define EDMA_TCC(t)	((t) << 12)
+#define TCINTEN		BIT(20)
+#define ITCINTEN	BIT(21)
+#define TCCHEN		BIT(22)
+#define ITCCHEN		BIT(23)
+
+#define TRWORD (0x7<<2)
+#define PAENTRY (0x1ff<<5)
+
+/* Drivers should avoid using these symbolic names for dm644x
+ * channels, and use platform_device IORESOURCE_DMA resources
+ * instead.  (Other DaVinci chips have different peripherals
+ * and thus have different DMA channel mappings.)
+ */
+#define DAVINCI_DMA_MCBSP_TX              2
+#define DAVINCI_DMA_MCBSP_RX              3
+#define DAVINCI_DMA_VPSS_HIST             4
+#define DAVINCI_DMA_VPSS_H3A              5
+#define DAVINCI_DMA_VPSS_PRVU             6
+#define DAVINCI_DMA_VPSS_RSZ              7
+#define DAVINCI_DMA_IMCOP_IMXINT          8
+#define DAVINCI_DMA_IMCOP_VLCDINT         9
+#define DAVINCI_DMA_IMCO_PASQINT         10
+#define DAVINCI_DMA_IMCOP_DSQINT         11
+#define DAVINCI_DMA_SPI_SPIX             16
+#define DAVINCI_DMA_SPI_SPIR             17
+#define DAVINCI_DMA_UART0_URXEVT0        18
+#define DAVINCI_DMA_UART0_UTXEVT0        19
+#define DAVINCI_DMA_UART1_URXEVT1        20
+#define DAVINCI_DMA_UART1_UTXEVT1        21
+#define DAVINCI_DMA_UART2_URXEVT2        22
+#define DAVINCI_DMA_UART2_UTXEVT2        23
+#define DAVINCI_DMA_MEMSTK_MSEVT         24
+#define DAVINCI_DMA_MMCRXEVT             26
+#define DAVINCI_DMA_MMCTXEVT             27
+#define DAVINCI_DMA_I2C_ICREVT           28
+#define DAVINCI_DMA_I2C_ICXEVT           29
+#define DAVINCI_DMA_GPIO_GPINT0          32
+#define DAVINCI_DMA_GPIO_GPINT1          33
+#define DAVINCI_DMA_GPIO_GPINT2          34
+#define DAVINCI_DMA_GPIO_GPINT3          35
+#define DAVINCI_DMA_GPIO_GPINT4          36
+#define DAVINCI_DMA_GPIO_GPINT5          37
+#define DAVINCI_DMA_GPIO_GPINT6          38
+#define DAVINCI_DMA_GPIO_GPINT7          39
+#define DAVINCI_DMA_GPIO_GPBNKINT0       40
+#define DAVINCI_DMA_GPIO_GPBNKINT1       41
+#define DAVINCI_DMA_GPIO_GPBNKINT2       42
+#define DAVINCI_DMA_GPIO_GPBNKINT3       43
+#define DAVINCI_DMA_GPIO_GPBNKINT4       44
+#define DAVINCI_DMA_TIMER0_TINT0         48
+#define DAVINCI_DMA_TIMER1_TINT1         49
+#define DAVINCI_DMA_TIMER2_TINT2         50
+#define DAVINCI_DMA_TIMER3_TINT3         51
+#define DAVINCI_DMA_PWM0                 52
+#define DAVINCI_DMA_PWM1                 53
+#define DAVINCI_DMA_PWM2                 54
+
+/*ch_status paramater of callback function possible values*/
+#define DMA_COMPLETE 1
+#define DMA_CC_ERROR 2
+#define DMA_TC1_ERROR 3
+#define DMA_TC2_ERROR 4
+
+enum address_mode {
+	INCR = 0,
+	FIFO = 1
+};
+
+enum fifo_width {
+	W8BIT = 0,
+	W16BIT = 1,
+	W32BIT = 2,
+	W64BIT = 3,
+	W128BIT = 4,
+	W256BIT = 5
+};
+
+enum dma_event_q {
+	EVENTQ_0 = 0,
+	EVENTQ_1 = 1,
+	EVENTQ_DEFAULT = -1
+};
+
+enum sync_dimension {
+	ASYNC = 0,
+	ABSYNC = 1
+};
+
+#define EDMA_CHANNEL_ANY		-1	/* for edma_alloc_channel() */
+#define EDMA_SLOT_ANY			-1	/* for edma_alloc_slot() */
+
+/* alloc/free DMA channels and their dedicated parameter RAM slots */
+int edma_alloc_channel(int channel,
+	void (*callback)(unsigned channel, u16 ch_status, void *data),
+	void *data, enum dma_event_q);
+void edma_free_channel(unsigned channel);
+
+/* alloc/free parameter RAM slots */
+int edma_alloc_slot(int slot);
+void edma_free_slot(unsigned slot);
+
+/* calls that operate on part of a parameter RAM slot */
+void edma_set_src(unsigned slot, dma_addr_t src_port,
+				enum address_mode mode, enum fifo_width);
+void edma_set_dest(unsigned slot, dma_addr_t dest_port,
+				 enum address_mode mode, enum fifo_width);
+void edma_get_position(unsigned slot, dma_addr_t *src, dma_addr_t *dst);
+void edma_set_src_index(unsigned slot, s16 src_bidx, s16 src_cidx);
+void edma_set_dest_index(unsigned slot, s16 dest_bidx, s16 dest_cidx);
+void edma_set_transfer_params(unsigned slot, u16 acnt, u16 bcnt, u16 ccnt,
+		u16 bcnt_rld, enum sync_dimension sync_mode);
+void edma_link(unsigned from, unsigned to);
+void edma_unlink(unsigned from);
+
+/* calls that operate on an entire parameter RAM slot */
+void edma_write_slot(unsigned slot, const struct edmacc_param *params);
+void edma_read_slot(unsigned slot, struct edmacc_param *params);
+
+/* channel control operations */
+int edma_start(unsigned channel);
+void edma_stop(unsigned channel);
+void edma_clean_channel(unsigned channel);
+void edma_clear_event(unsigned channel);
+void edma_pause(unsigned channel);
+void edma_resume(unsigned channel);
+
+/* UNRELATED TO DMA */
+int davinci_alloc_iram(unsigned size);
+void davinci_free_iram(unsigned addr, unsigned size);
+
+/* platform_data for EDMA driver */
+struct edma_soc_info {
+
+	/* how many dma resources of each type */
+	unsigned	n_channel;
+	unsigned	n_region;
+	unsigned	n_slot;
+	unsigned	n_tc;
+
+	/* list of channels with no even trigger; terminated by "-1" */
+	const s8	*noevent;
+};
+
+#endif
diff --git a/arch/arm/mach-davinci/include/mach/gpio.h b/arch/arm/mach-davinci/include/mach/gpio.h
index b456f079f43..efe3281364e 100644
--- a/arch/arm/mach-davinci/include/mach/gpio.h
+++ b/arch/arm/mach-davinci/include/mach/gpio.h
@@ -15,9 +15,11 @@
 
 #include <linux/io.h>
 #include <asm-generic/gpio.h>
-#include <mach/hardware.h>
+
 #include <mach/irqs.h>
 
+#define DAVINCI_GPIO_BASE 0x01C67000
+
 /*
  * basic gpio routines
  *
@@ -26,23 +28,18 @@
  * go through boot loaders.
  *
  * the gpio clock will be turned on when gpios are used, and you may also
- * need to pay attention to PINMUX0 and PINMUX1 to be sure those pins are
+ * need to pay attention to PINMUX registers to be sure those pins are
  * used as gpios, not with other peripherals.
  *
  * On-chip GPIOs are numbered 0..(DAVINCI_N_GPIO-1).  For documentation,
- * and maybe for later updates, code should write GPIO(N) or:
- *  - GPIOV18(N) for 1.8V pins, N in 0..53; same as GPIO(0)..GPIO(53)
- *  - GPIOV33(N) for 3.3V pins, N in 0..17; same as GPIO(54)..GPIO(70)
- *
- * For GPIO IRQs use gpio_to_irq(GPIO(N)) or gpio_to_irq(GPIOV33(N)) etc
- * for now, that's != GPIO(N)
+ * and maybe for later updates, code may write GPIO(N).  These may be
+ * all 1.8V signals, all 3.3V ones, or a mix of the two.  A given chip
+ * may not support all the GPIOs in that range.
  *
  * GPIOs can also be on external chips, numbered after the ones built-in
  * to the DaVinci chip.  For now, they won't be usable as IRQ sources.
  */
-#define	GPIO(X)		(X)		/* 0 <= X <= 70 */
-#define	GPIOV18(X)	(X)		/* 1.8V i/o; 0 <= X <= 53 */
-#define	GPIOV33(X)	((X)+54)	/* 3.3V i/o; 0 <= X <= 17 */
+#define	GPIO(X)		(X)		/* 0 <= X <= (DAVINCI_N_GPIO - 1) */
 
 struct gpio_controller {
 	u32	dir;
@@ -71,12 +68,14 @@ __gpio_to_controller(unsigned gpio)
 {
 	void *__iomem ptr;
 
-	if (gpio < 32)
+	if (gpio < 32 * 1)
 		ptr = IO_ADDRESS(DAVINCI_GPIO_BASE + 0x10);
-	else if (gpio < 64)
+	else if (gpio < 32 * 2)
 		ptr = IO_ADDRESS(DAVINCI_GPIO_BASE + 0x38);
-	else if (gpio < DAVINCI_N_GPIO)
+	else if (gpio < 32 * 3)
 		ptr = IO_ADDRESS(DAVINCI_GPIO_BASE + 0x60);
+	else if (gpio < 32 * 4)
+		ptr = IO_ADDRESS(DAVINCI_GPIO_BASE + 0x88);
 	else
 		ptr = NULL;
 	return ptr;
diff --git a/arch/arm/mach-davinci/include/mach/hardware.h b/arch/arm/mach-davinci/include/mach/hardware.h
index a2e8969afac..48c77934d51 100644
--- a/arch/arm/mach-davinci/include/mach/hardware.h
+++ b/arch/arm/mach-davinci/include/mach/hardware.h
@@ -1,9 +1,9 @@
 /*
- * Common hardware definitions
+ * Hardware definitions common to all DaVinci family processors
  *
- * Author: Kevin Hilman, MontaVista Software, Inc. <source@mvista.com>
+ * Author: Kevin Hilman, Deep Root Systems, LLC
  *
- * 2007 (c) MontaVista Software, Inc. This file is licensed under
+ * 2007 (c) Deep Root Systems, LLC. This file is licensed under
  * the terms of the GNU General Public License version 2. This program
  * is licensed "as is" without any warranty of any kind, whether express
  * or implied.
@@ -12,41 +12,16 @@
 #define __ASM_ARCH_HARDWARE_H
 
 /*
- * Base register addresses
+ * Before you add anything to ths file:
+ *
+ * This header is for defines common to ALL DaVinci family chips.
+ * Anything that is chip specific should go in <chipname>.h,
+ * and the chip/board init code should then explicitly include
+ * <chipname>.h
  */
-#define DAVINCI_DMA_3PCC_BASE			(0x01C00000)
-#define DAVINCI_DMA_3PTC0_BASE			(0x01C10000)
-#define DAVINCI_DMA_3PTC1_BASE			(0x01C10400)
-#define DAVINCI_I2C_BASE			(0x01C21000)
-#define DAVINCI_PWM0_BASE			(0x01C22000)
-#define DAVINCI_PWM1_BASE			(0x01C22400)
-#define DAVINCI_PWM2_BASE			(0x01C22800)
-#define DAVINCI_SYSTEM_MODULE_BASE		(0x01C40000)
-#define DAVINCI_PLL_CNTRL0_BASE			(0x01C40800)
-#define DAVINCI_PLL_CNTRL1_BASE			(0x01C40C00)
-#define DAVINCI_PWR_SLEEP_CNTRL_BASE		(0x01C41000)
-#define DAVINCI_SYSTEM_DFT_BASE			(0x01C42000)
-#define DAVINCI_IEEE1394_BASE			(0x01C60000)
-#define DAVINCI_USB_OTG_BASE			(0x01C64000)
-#define DAVINCI_CFC_ATA_BASE			(0x01C66000)
-#define DAVINCI_SPI_BASE			(0x01C66800)
-#define DAVINCI_GPIO_BASE			(0x01C67000)
-#define DAVINCI_UHPI_BASE			(0x01C67800)
-#define DAVINCI_VPSS_REGS_BASE			(0x01C70000)
-#define DAVINCI_EMAC_CNTRL_REGS_BASE		(0x01C80000)
-#define DAVINCI_EMAC_WRAPPER_CNTRL_REGS_BASE	(0x01C81000)
-#define DAVINCI_EMAC_WRAPPER_RAM_BASE		(0x01C82000)
-#define DAVINCI_MDIO_CNTRL_REGS_BASE		(0x01C84000)
-#define DAVINCI_IMCOP_BASE			(0x01CC0000)
-#define DAVINCI_ASYNC_EMIF_CNTRL_BASE		(0x01E00000)
-#define DAVINCI_VLYNQ_BASE			(0x01E01000)
-#define DAVINCI_MCBSP_BASE			(0x01E02000)
-#define DAVINCI_MMC_SD_BASE			(0x01E10000)
-#define DAVINCI_MS_BASE				(0x01E20000)
-#define DAVINCI_ASYNC_EMIF_DATA_CE0_BASE	(0x02000000)
-#define DAVINCI_ASYNC_EMIF_DATA_CE1_BASE	(0x04000000)
-#define DAVINCI_ASYNC_EMIF_DATA_CE2_BASE	(0x06000000)
-#define DAVINCI_ASYNC_EMIF_DATA_CE3_BASE	(0x08000000)
-#define DAVINCI_VLYNQ_REMOTE_BASE		(0x0C000000)
+#define DAVINCI_SYSTEM_MODULE_BASE        0x01C40000
+
+/* System control register offsets */
+#define DM64XX_VDD3P3V_PWDN	0x48
 
 #endif /* __ASM_ARCH_HARDWARE_H */
diff --git a/arch/arm/mach-davinci/include/mach/io.h b/arch/arm/mach-davinci/include/mach/io.h
index a48795fd241..2479785405a 100644
--- a/arch/arm/mach-davinci/include/mach/io.h
+++ b/arch/arm/mach-davinci/include/mach/io.h
@@ -40,22 +40,12 @@
 #else
 #define IOMEM(x)                ((void __force __iomem *)(x))
 
-/*
- * Functions to access the DaVinci IO region
- *
- * NOTE: - Use davinci_read/write[bwl] for physical register addresses
- *	 - Use __raw_read/write[bwl]() for virtual register addresses
- *	 - Use IO_ADDRESS(phys_addr) to convert registers to virtual addresses
- *	 - DO NOT use hardcoded virtual addresses to allow changing the
- *	   IO address space again if needed
- */
-#define davinci_readb(a)	__raw_readb(IO_ADDRESS(a))
-#define davinci_readw(a)	__raw_readw(IO_ADDRESS(a))
-#define davinci_readl(a)	__raw_readl(IO_ADDRESS(a))
+#define __arch_ioremap(p, s, t)	davinci_ioremap(p, s, t)
+#define __arch_iounmap(v)	davinci_iounmap(v)
 
-#define davinci_writeb(v, a)	__raw_writeb(v, IO_ADDRESS(a))
-#define davinci_writew(v, a)	__raw_writew(v, IO_ADDRESS(a))
-#define davinci_writel(v, a)	__raw_writel(v, IO_ADDRESS(a))
+void __iomem *davinci_ioremap(unsigned long phys, size_t size,
+			      unsigned int type);
+void davinci_iounmap(volatile void __iomem *addr);
 
 #endif /* __ASSEMBLER__ */
 #endif /* __ASM_ARCH_IO_H */
diff --git a/arch/arm/mach-davinci/include/mach/irqs.h b/arch/arm/mach-davinci/include/mach/irqs.h
index f4c5ca6da9f..18066074c99 100644
--- a/arch/arm/mach-davinci/include/mach/irqs.h
+++ b/arch/arm/mach-davinci/include/mach/irqs.h
@@ -96,10 +96,111 @@
 #define IRQ_EMUINT       63
 
 #define DAVINCI_N_AINTC_IRQ	64
-#define DAVINCI_N_GPIO		71
+#define DAVINCI_N_GPIO		104
 
 #define NR_IRQS			(DAVINCI_N_AINTC_IRQ + DAVINCI_N_GPIO)
 
 #define ARCH_TIMER_IRQ IRQ_TINT1_TINT34
 
+/* DaVinci DM6467-specific Interrupts */
+#define IRQ_DM646X_VP_VERTINT0  0
+#define IRQ_DM646X_VP_VERTINT1  1
+#define IRQ_DM646X_VP_VERTINT2  2
+#define IRQ_DM646X_VP_VERTINT3  3
+#define IRQ_DM646X_VP_ERRINT    4
+#define IRQ_DM646X_RESERVED_1   5
+#define IRQ_DM646X_RESERVED_2   6
+#define IRQ_DM646X_WDINT        7
+#define IRQ_DM646X_CRGENINT0    8
+#define IRQ_DM646X_CRGENINT1    9
+#define IRQ_DM646X_TSIFINT0     10
+#define IRQ_DM646X_TSIFINT1     11
+#define IRQ_DM646X_VDCEINT      12
+#define IRQ_DM646X_USBINT       13
+#define IRQ_DM646X_USBDMAINT    14
+#define IRQ_DM646X_PCIINT       15
+#define IRQ_DM646X_TCERRINT2    20
+#define IRQ_DM646X_TCERRINT3    21
+#define IRQ_DM646X_IDE          22
+#define IRQ_DM646X_HPIINT       23
+#define IRQ_DM646X_EMACRXTHINT  24
+#define IRQ_DM646X_EMACRXINT    25
+#define IRQ_DM646X_EMACTXINT    26
+#define IRQ_DM646X_EMACMISCINT  27
+#define IRQ_DM646X_MCASP0TXINT  28
+#define IRQ_DM646X_MCASP0RXINT  29
+#define IRQ_DM646X_RESERVED_3   31
+#define IRQ_DM646X_MCASP1TXINT  32
+#define IRQ_DM646X_VLQINT       38
+#define IRQ_DM646X_UARTINT2     42
+#define IRQ_DM646X_SPINT0       43
+#define IRQ_DM646X_SPINT1       44
+#define IRQ_DM646X_DSP2ARMINT   45
+#define IRQ_DM646X_RESERVED_4   46
+#define IRQ_DM646X_PSCINT       47
+#define IRQ_DM646X_GPIO0        48
+#define IRQ_DM646X_GPIO1        49
+#define IRQ_DM646X_GPIO2        50
+#define IRQ_DM646X_GPIO3        51
+#define IRQ_DM646X_GPIO4        52
+#define IRQ_DM646X_GPIO5        53
+#define IRQ_DM646X_GPIO6        54
+#define IRQ_DM646X_GPIO7        55
+#define IRQ_DM646X_GPIOBNK0     56
+#define IRQ_DM646X_GPIOBNK1     57
+#define IRQ_DM646X_GPIOBNK2     58
+#define IRQ_DM646X_DDRINT       59
+#define IRQ_DM646X_AEMIFINT     60
+
+/* DaVinci DM355-specific Interrupts */
+#define IRQ_DM355_CCDC_VDINT0	0
+#define IRQ_DM355_CCDC_VDINT1	1
+#define IRQ_DM355_CCDC_VDINT2	2
+#define IRQ_DM355_IPIPE_HST	3
+#define IRQ_DM355_H3AINT	4
+#define IRQ_DM355_IPIPE_SDR	5
+#define IRQ_DM355_IPIPEIFINT	6
+#define IRQ_DM355_OSDINT	7
+#define IRQ_DM355_VENCINT	8
+#define IRQ_DM355_IMCOPINT	11
+#define IRQ_DM355_RTOINT	13
+#define IRQ_DM355_TINT4		13
+#define IRQ_DM355_TINT2_TINT12	13
+#define IRQ_DM355_UARTINT2	14
+#define IRQ_DM355_TINT5		14
+#define IRQ_DM355_TINT2_TINT34	14
+#define IRQ_DM355_TINT6		15
+#define IRQ_DM355_TINT3_TINT12	15
+#define IRQ_DM355_SPINT1_0	17
+#define IRQ_DM355_SPINT1_1	18
+#define IRQ_DM355_SPINT2_0	19
+#define IRQ_DM355_SPINT2_1	21
+#define IRQ_DM355_TINT7		22
+#define IRQ_DM355_TINT3_TINT34	22
+#define IRQ_DM355_SDIOINT0	23
+#define IRQ_DM355_MMCINT0	26
+#define IRQ_DM355_MSINT		26
+#define IRQ_DM355_MMCINT1	27
+#define IRQ_DM355_PWMINT3	28
+#define IRQ_DM355_SDIOINT1	31
+#define IRQ_DM355_SPINT0_0	42
+#define IRQ_DM355_SPINT0_1	43
+#define IRQ_DM355_GPIO0		44
+#define IRQ_DM355_GPIO1		45
+#define IRQ_DM355_GPIO2		46
+#define IRQ_DM355_GPIO3		47
+#define IRQ_DM355_GPIO4		48
+#define IRQ_DM355_GPIO5		49
+#define IRQ_DM355_GPIO6		50
+#define IRQ_DM355_GPIO7		51
+#define IRQ_DM355_GPIO8		52
+#define IRQ_DM355_GPIO9		53
+#define IRQ_DM355_GPIOBNK0	54
+#define IRQ_DM355_GPIOBNK1	55
+#define IRQ_DM355_GPIOBNK2	56
+#define IRQ_DM355_GPIOBNK3	57
+#define IRQ_DM355_GPIOBNK4	58
+#define IRQ_DM355_GPIOBNK5	59
+#define IRQ_DM355_GPIOBNK6	60
+
 #endif /* __ASM_ARCH_IRQS_H */
diff --git a/arch/arm/mach-davinci/include/mach/mux.h b/arch/arm/mach-davinci/include/mach/mux.h
index c24b6782804..bae22cb3e27 100644
--- a/arch/arm/mach-davinci/include/mach/mux.h
+++ b/arch/arm/mach-davinci/include/mach/mux.h
@@ -1,55 +1,183 @@
 /*
- * DaVinci pin multiplexing defines
+ * Table of the DAVINCI register configurations for the PINMUX combinations
  *
  * Author: Vladimir Barinov, MontaVista Software, Inc. <source@mvista.com>
  *
+ * Based on linux/include/asm-arm/arch-omap/mux.h:
+ * Copyright (C) 2003 - 2005 Nokia Corporation
+ *
+ * Written by Tony Lindgren
+ *
  * 2007 (c) MontaVista Software, Inc. This file is licensed under
  * the terms of the GNU General Public License version 2. This program
  * is licensed "as is" without any warranty of any kind, whether express
  * or implied.
+ *
+ * Copyright (C) 2008 Texas Instruments.
  */
-#ifndef __ASM_ARCH_MUX_H
-#define __ASM_ARCH_MUX_H
-
-#define DAVINCI_MUX_AEAW0	0
-#define DAVINCI_MUX_AEAW1	1
-#define DAVINCI_MUX_AEAW2	2
-#define DAVINCI_MUX_AEAW3	3
-#define DAVINCI_MUX_AEAW4	4
-#define DAVINCI_MUX_AECS4	10
-#define DAVINCI_MUX_AECS5	11
-#define DAVINCI_MUX_VLYNQWD0	12
-#define DAVINCI_MUX_VLYNQWD1	13
-#define DAVINCI_MUX_VLSCREN	14
-#define DAVINCI_MUX_VLYNQEN	15
-#define DAVINCI_MUX_HDIREN	16
-#define DAVINCI_MUX_ATAEN	17
-#define DAVINCI_MUX_RGB666	22
-#define DAVINCI_MUX_RGB888	23
-#define DAVINCI_MUX_LOEEN	24
-#define DAVINCI_MUX_LFLDEN	25
-#define DAVINCI_MUX_CWEN	26
-#define DAVINCI_MUX_CFLDEN	27
-#define DAVINCI_MUX_HPIEN	29
-#define DAVINCI_MUX_1394EN	30
-#define DAVINCI_MUX_EMACEN	31
-
-#define DAVINCI_MUX_LEVEL2	32
-#define DAVINCI_MUX_UART0	(DAVINCI_MUX_LEVEL2 + 0)
-#define DAVINCI_MUX_UART1	(DAVINCI_MUX_LEVEL2 + 1)
-#define DAVINCI_MUX_UART2	(DAVINCI_MUX_LEVEL2 + 2)
-#define DAVINCI_MUX_U2FLO	(DAVINCI_MUX_LEVEL2 + 3)
-#define DAVINCI_MUX_PWM0	(DAVINCI_MUX_LEVEL2 + 4)
-#define DAVINCI_MUX_PWM1	(DAVINCI_MUX_LEVEL2 + 5)
-#define DAVINCI_MUX_PWM2	(DAVINCI_MUX_LEVEL2 + 6)
-#define DAVINCI_MUX_I2C		(DAVINCI_MUX_LEVEL2 + 7)
-#define DAVINCI_MUX_SPI		(DAVINCI_MUX_LEVEL2 + 8)
-#define DAVINCI_MUX_MSTK	(DAVINCI_MUX_LEVEL2 + 9)
-#define DAVINCI_MUX_ASP		(DAVINCI_MUX_LEVEL2 + 10)
-#define DAVINCI_MUX_CLK0	(DAVINCI_MUX_LEVEL2 + 16)
-#define DAVINCI_MUX_CLK1	(DAVINCI_MUX_LEVEL2 + 17)
-#define DAVINCI_MUX_TIMIN	(DAVINCI_MUX_LEVEL2 + 18)
-
-extern void davinci_mux_peripheral(unsigned int mux, unsigned int enable);
-
-#endif /* __ASM_ARCH_MUX_H */
+
+#ifndef __INC_MACH_MUX_H
+#define __INC_MACH_MUX_H
+
+/* System module registers */
+#define PINMUX0			0x00
+#define PINMUX1			0x04
+/* dm355 only */
+#define PINMUX2			0x08
+#define PINMUX3			0x0c
+#define PINMUX4			0x10
+#define INTMUX			0x18
+#define EVTMUX			0x1c
+
+struct mux_config {
+	const char *name;
+	const char *mux_reg_name;
+	const unsigned char mux_reg;
+	const unsigned char mask_offset;
+	const unsigned char mask;
+	const unsigned char mode;
+	bool debug;
+};
+
+enum davinci_dm644x_index {
+	/* ATA and HDDIR functions */
+	DM644X_HDIREN,
+	DM644X_ATAEN,
+	DM644X_ATAEN_DISABLE,
+
+	/* HPI functions */
+	DM644X_HPIEN_DISABLE,
+
+	/* AEAW functions */
+	DM644X_AEAW,
+
+	/* Memory Stick */
+	DM644X_MSTK,
+
+	/* I2C */
+	DM644X_I2C,
+
+	/* ASP function */
+	DM644X_MCBSP,
+
+	/* UART1 */
+	DM644X_UART1,
+
+	/* UART2 */
+	DM644X_UART2,
+
+	/* PWM0 */
+	DM644X_PWM0,
+
+	/* PWM1 */
+	DM644X_PWM1,
+
+	/* PWM2 */
+	DM644X_PWM2,
+
+	/* VLYNQ function */
+	DM644X_VLYNQEN,
+	DM644X_VLSCREN,
+	DM644X_VLYNQWD,
+
+	/* EMAC and MDIO function */
+	DM644X_EMACEN,
+
+	/* GPIO3V[0:16] pins */
+	DM644X_GPIO3V,
+
+	/* GPIO pins */
+	DM644X_GPIO0,
+	DM644X_GPIO3,
+	DM644X_GPIO43_44,
+	DM644X_GPIO46_47,
+
+	/* VPBE */
+	DM644X_RGB666,
+
+	/* LCD */
+	DM644X_LOEEN,
+	DM644X_LFLDEN,
+};
+
+enum davinci_dm646x_index {
+	/* ATA function */
+	DM646X_ATAEN,
+
+	/* AUDIO Clock */
+	DM646X_AUDCK1,
+	DM646X_AUDCK0,
+
+	/* CRGEN Control */
+	DM646X_CRGMUX,
+
+	/* VPIF Control */
+	DM646X_STSOMUX_DISABLE,
+	DM646X_STSIMUX_DISABLE,
+	DM646X_PTSOMUX_DISABLE,
+	DM646X_PTSIMUX_DISABLE,
+
+	/* TSIF Control */
+	DM646X_STSOMUX,
+	DM646X_STSIMUX,
+	DM646X_PTSOMUX_PARALLEL,
+	DM646X_PTSIMUX_PARALLEL,
+	DM646X_PTSOMUX_SERIAL,
+	DM646X_PTSIMUX_SERIAL,
+};
+
+enum davinci_dm355_index {
+	/* MMC/SD 0 */
+	DM355_MMCSD0,
+
+	/* MMC/SD 1 */
+	DM355_SD1_CLK,
+	DM355_SD1_CMD,
+	DM355_SD1_DATA3,
+	DM355_SD1_DATA2,
+	DM355_SD1_DATA1,
+	DM355_SD1_DATA0,
+
+	/* I2C */
+	DM355_I2C_SDA,
+	DM355_I2C_SCL,
+
+	/* ASP0 function */
+	DM355_MCBSP0_BDX,
+	DM355_MCBSP0_X,
+	DM355_MCBSP0_BFSX,
+	DM355_MCBSP0_BDR,
+	DM355_MCBSP0_R,
+	DM355_MCBSP0_BFSR,
+
+	/* SPI0 */
+	DM355_SPI0_SDI,
+	DM355_SPI0_SDENA0,
+	DM355_SPI0_SDENA1,
+
+	/* IRQ muxing */
+	DM355_INT_EDMA_CC,
+	DM355_INT_EDMA_TC0_ERR,
+	DM355_INT_EDMA_TC1_ERR,
+
+	/* EDMA event muxing */
+	DM355_EVT8_ASP1_TX,
+	DM355_EVT9_ASP1_RX,
+	DM355_EVT26_MMC0_RX,
+};
+
+#ifdef CONFIG_DAVINCI_MUX
+/* setup pin muxing */
+extern void davinci_mux_init(void);
+extern int davinci_mux_register(const struct mux_config *pins,
+				unsigned long size);
+extern int davinci_cfg_reg(unsigned long reg_cfg);
+#else
+/* boot loader does it all (no warnings from CONFIG_DAVINCI_MUX_WARNINGS) */
+static inline void davinci_mux_init(void) {}
+static inline int davinci_mux_register(const struct mux_config *pins,
+				       unsigned long size) { return 0; }
+static inline int davinci_cfg_reg(unsigned long reg_cfg) { return 0; }
+#endif
+
+#endif /* __INC_MACH_MUX_H */
diff --git a/arch/arm/mach-davinci/include/mach/psc.h b/arch/arm/mach-davinci/include/mach/psc.h
index 4977aa071e1..55a90d419fa 100644
--- a/arch/arm/mach-davinci/include/mach/psc.h
+++ b/arch/arm/mach-davinci/include/mach/psc.h
@@ -38,8 +38,6 @@
 #define DAVINCI_LPSC_TPTC1          4
 #define DAVINCI_LPSC_EMAC           5
 #define DAVINCI_LPSC_EMAC_WRAPPER   6
-#define DAVINCI_LPSC_MDIO           7
-#define DAVINCI_LPSC_IEEE1394       8
 #define DAVINCI_LPSC_USB            9
 #define DAVINCI_LPSC_ATA            10
 #define DAVINCI_LPSC_VLYNQ          11
@@ -47,7 +45,6 @@
 #define DAVINCI_LPSC_DDR_EMIF       13
 #define DAVINCI_LPSC_AEMIF          14
 #define DAVINCI_LPSC_MMC_SD         15
-#define DAVINCI_LPSC_MEMSTICK       16
 #define DAVINCI_LPSC_McBSP          17
 #define DAVINCI_LPSC_I2C            18
 #define DAVINCI_LPSC_UART0          19
@@ -73,4 +70,54 @@
 #define DAVINCI_LPSC_GEM            39
 #define DAVINCI_LPSC_IMCOP          40
 
+#define DM355_LPSC_TIMER3		5
+#define DM355_LPSC_SPI1			6
+#define DM355_LPSC_MMC_SD1		7
+#define DM355_LPSC_McBSP1		8
+#define DM355_LPSC_PWM3			10
+#define DM355_LPSC_SPI2			11
+#define DM355_LPSC_RTO			12
+#define DM355_LPSC_VPSS_DAC		41
+
+/*
+ * LPSC Assignments
+ */
+#define DM646X_LPSC_ARM            0
+#define DM646X_LPSC_C64X_CPU       1
+#define DM646X_LPSC_HDVICP0        2
+#define DM646X_LPSC_HDVICP1        3
+#define DM646X_LPSC_TPCC           4
+#define DM646X_LPSC_TPTC0          5
+#define DM646X_LPSC_TPTC1          6
+#define DM646X_LPSC_TPTC2          7
+#define DM646X_LPSC_TPTC3          8
+#define DM646X_LPSC_PCI            13
+#define DM646X_LPSC_EMAC           14
+#define DM646X_LPSC_VDCE           15
+#define DM646X_LPSC_VPSSMSTR       16
+#define DM646X_LPSC_VPSSSLV        17
+#define DM646X_LPSC_TSIF0          18
+#define DM646X_LPSC_TSIF1          19
+#define DM646X_LPSC_DDR_EMIF       20
+#define DM646X_LPSC_AEMIF          21
+#define DM646X_LPSC_McASP0         22
+#define DM646X_LPSC_McASP1         23
+#define DM646X_LPSC_CRGEN0         24
+#define DM646X_LPSC_CRGEN1         25
+#define DM646X_LPSC_UART0          26
+#define DM646X_LPSC_UART1          27
+#define DM646X_LPSC_UART2          28
+#define DM646X_LPSC_PWM0           29
+#define DM646X_LPSC_PWM1           30
+#define DM646X_LPSC_I2C            31
+#define DM646X_LPSC_SPI            32
+#define DM646X_LPSC_GPIO           33
+#define DM646X_LPSC_TIMER0         34
+#define DM646X_LPSC_TIMER1         35
+#define DM646X_LPSC_ARM_INTC       45
+
+extern int davinci_psc_is_clk_active(unsigned int id);
+extern void davinci_psc_config(unsigned int domain, unsigned int id,
+			       char enable);
+
 #endif /* __ASM_ARCH_PSC_H */
diff --git a/arch/arm/mach-davinci/include/mach/serial.h b/arch/arm/mach-davinci/include/mach/serial.h
index fb8cb229bfd..632847d74a1 100644
--- a/arch/arm/mach-davinci/include/mach/serial.h
+++ b/arch/arm/mach-davinci/include/mach/serial.h
@@ -13,8 +13,23 @@
 
 #include <mach/io.h>
 
-#define DAVINCI_UART0_BASE   (IO_PHYS + 0x20000)
-#define DAVINCI_UART1_BASE   (IO_PHYS + 0x20400)
-#define DAVINCI_UART2_BASE   (IO_PHYS + 0x20800)
+#define DAVINCI_MAX_NR_UARTS	3
+#define DAVINCI_UART0_BASE	(IO_PHYS + 0x20000)
+#define DAVINCI_UART1_BASE	(IO_PHYS + 0x20400)
+#define DAVINCI_UART2_BASE	(IO_PHYS + 0x20800)
+
+#define DM355_UART2_BASE	(IO_PHYS + 0x206000)
+
+/* DaVinci UART register offsets */
+#define UART_DAVINCI_PWREMU		0x0c
+#define UART_DM646X_SCR			0x10
+#define UART_DM646X_SCR_TX_WATERMARK	0x08
+
+struct davinci_uart_config {
+	/* Bit field of UARTs present; bit 0 --> UART1 */
+	unsigned int enabled_uarts;
+};
+
+extern void davinci_serial_init(struct davinci_uart_config *);
 
 #endif /* __ASM_ARCH_SERIAL_H */
diff --git a/arch/arm/mach-davinci/io.c b/arch/arm/mach-davinci/io.c
index 299515f70b8..a548abb513e 100644
--- a/arch/arm/mach-davinci/io.c
+++ b/arch/arm/mach-davinci/io.c
@@ -51,7 +51,26 @@ void __init davinci_map_common_io(void)
 	davinci_check_revision();
 }
 
-void __init davinci_init_common_hw(void)
+#define BETWEEN(p, st, sz)	((p) >= (st) && (p) < ((st) + (sz)))
+#define XLATE(p, pst, vst)	((void __iomem *)((p) - (pst) + (vst)))
+
+/*
+ * Intercept ioremap() requests for addresses in our fixed mapping regions.
+ */
+void __iomem *davinci_ioremap(unsigned long p, size_t size, unsigned int type)
+{
+	if (BETWEEN(p, IO_PHYS, IO_SIZE))
+		return XLATE(p, IO_PHYS, IO_VIRT);
+
+	return __arm_ioremap(p, size, type);
+}
+EXPORT_SYMBOL(davinci_ioremap);
+
+void davinci_iounmap(volatile void __iomem *addr)
 {
-	davinci_clk_init();
+	unsigned long virt = (unsigned long)addr;
+
+	if (virt >= VMALLOC_START && virt < VMALLOC_END)
+		__iounmap(addr);
 }
+EXPORT_SYMBOL(davinci_iounmap);
diff --git a/arch/arm/mach-davinci/irq.c b/arch/arm/mach-davinci/irq.c
index 38021af8359..5a324c90e29 100644
--- a/arch/arm/mach-davinci/irq.c
+++ b/arch/arm/mach-davinci/irq.c
@@ -25,6 +25,7 @@
 #include <linux/io.h>
 
 #include <mach/hardware.h>
+#include <mach/cputype.h>
 #include <asm/mach/irq.h>
 
 #define IRQ_BIT(irq)		((irq) & 0x1f)
@@ -40,14 +41,18 @@
 #define IRQ_INTPRI0_REG_OFFSET	0x0030
 #define IRQ_INTPRI7_REG_OFFSET	0x004C
 
+const u8 *davinci_def_priorities;
+
+#define INTC_BASE IO_ADDRESS(DAVINCI_ARM_INTC_BASE)
+
 static inline unsigned int davinci_irq_readl(int offset)
 {
-	return davinci_readl(DAVINCI_ARM_INTC_BASE + offset);
+	return __raw_readl(INTC_BASE + offset);
 }
 
 static inline void davinci_irq_writel(unsigned long value, int offset)
 {
-	davinci_writel(value, DAVINCI_ARM_INTC_BASE + offset);
+	__raw_writel(value, INTC_BASE + offset);
 }
 
 /* Disable interrupt */
@@ -108,9 +113,8 @@ static struct irq_chip davinci_irq_chip_0 = {
 	.unmask = davinci_unmask_irq,
 };
 
-
 /* FIQ are pri 0-1; otherwise 2-7, with 7 lowest priority */
-static const u8 default_priorities[DAVINCI_N_AINTC_IRQ] __initdata = {
+static const u8 dm644x_default_priorities[DAVINCI_N_AINTC_IRQ] __initdata = {
 	[IRQ_VDINT0]		= 2,
 	[IRQ_VDINT1]		= 6,
 	[IRQ_VDINT2]		= 6,
@@ -177,11 +181,149 @@ static const u8 default_priorities[DAVINCI_N_AINTC_IRQ] __initdata = {
 	[IRQ_EMUINT]		= 7,
 };
 
+static const u8 dm646x_default_priorities[DAVINCI_N_AINTC_IRQ] = {
+	[IRQ_DM646X_VP_VERTINT0]        = 7,
+	[IRQ_DM646X_VP_VERTINT1]        = 7,
+	[IRQ_DM646X_VP_VERTINT2]        = 7,
+	[IRQ_DM646X_VP_VERTINT3]        = 7,
+	[IRQ_DM646X_VP_ERRINT]          = 7,
+	[IRQ_DM646X_RESERVED_1]         = 7,
+	[IRQ_DM646X_RESERVED_2]         = 7,
+	[IRQ_DM646X_WDINT]              = 7,
+	[IRQ_DM646X_CRGENINT0]          = 7,
+	[IRQ_DM646X_CRGENINT1]          = 7,
+	[IRQ_DM646X_TSIFINT0]           = 7,
+	[IRQ_DM646X_TSIFINT1]           = 7,
+	[IRQ_DM646X_VDCEINT]            = 7,
+	[IRQ_DM646X_USBINT]             = 7,
+	[IRQ_DM646X_USBDMAINT]          = 7,
+	[IRQ_DM646X_PCIINT]             = 7,
+	[IRQ_CCINT0]                    = 7,    /* dma */
+	[IRQ_CCERRINT]                  = 7,    /* dma */
+	[IRQ_TCERRINT0]                 = 7,    /* dma */
+	[IRQ_TCERRINT]                  = 7,    /* dma */
+	[IRQ_DM646X_TCERRINT2]          = 7,
+	[IRQ_DM646X_TCERRINT3]          = 7,
+	[IRQ_DM646X_IDE]                = 7,
+	[IRQ_DM646X_HPIINT]             = 7,
+	[IRQ_DM646X_EMACRXTHINT]        = 7,
+	[IRQ_DM646X_EMACRXINT]          = 7,
+	[IRQ_DM646X_EMACTXINT]          = 7,
+	[IRQ_DM646X_EMACMISCINT]        = 7,
+	[IRQ_DM646X_MCASP0TXINT]        = 7,
+	[IRQ_DM646X_MCASP0RXINT]        = 7,
+	[IRQ_AEMIFINT]                  = 7,
+	[IRQ_DM646X_RESERVED_3]         = 7,
+	[IRQ_DM646X_MCASP1TXINT]        = 7,    /* clockevent */
+	[IRQ_TINT0_TINT34]              = 7,    /* clocksource */
+	[IRQ_TINT1_TINT12]              = 7,    /* DSP timer */
+	[IRQ_TINT1_TINT34]              = 7,    /* system tick */
+	[IRQ_PWMINT0]                   = 7,
+	[IRQ_PWMINT1]                   = 7,
+	[IRQ_DM646X_VLQINT]             = 7,
+	[IRQ_I2C]                       = 7,
+	[IRQ_UARTINT0]                  = 7,
+	[IRQ_UARTINT1]                  = 7,
+	[IRQ_DM646X_UARTINT2]           = 7,
+	[IRQ_DM646X_SPINT0]             = 7,
+	[IRQ_DM646X_SPINT1]             = 7,
+	[IRQ_DM646X_DSP2ARMINT]         = 7,
+	[IRQ_DM646X_RESERVED_4]         = 7,
+	[IRQ_DM646X_PSCINT]             = 7,
+	[IRQ_DM646X_GPIO0]              = 7,
+	[IRQ_DM646X_GPIO1]              = 7,
+	[IRQ_DM646X_GPIO2]              = 7,
+	[IRQ_DM646X_GPIO3]              = 7,
+	[IRQ_DM646X_GPIO4]              = 7,
+	[IRQ_DM646X_GPIO5]              = 7,
+	[IRQ_DM646X_GPIO6]              = 7,
+	[IRQ_DM646X_GPIO7]              = 7,
+	[IRQ_DM646X_GPIOBNK0]           = 7,
+	[IRQ_DM646X_GPIOBNK1]           = 7,
+	[IRQ_DM646X_GPIOBNK2]           = 7,
+	[IRQ_DM646X_DDRINT]             = 7,
+	[IRQ_DM646X_AEMIFINT]           = 7,
+	[IRQ_COMMTX]                    = 7,
+	[IRQ_COMMRX]                    = 7,
+	[IRQ_EMUINT]                    = 7,
+};
+
+static const u8 dm355_default_priorities[DAVINCI_N_AINTC_IRQ] = {
+	[IRQ_DM355_CCDC_VDINT0]		= 2,
+	[IRQ_DM355_CCDC_VDINT1]		= 6,
+	[IRQ_DM355_CCDC_VDINT2]		= 6,
+	[IRQ_DM355_IPIPE_HST]		= 6,
+	[IRQ_DM355_H3AINT]		= 6,
+	[IRQ_DM355_IPIPE_SDR]		= 6,
+	[IRQ_DM355_IPIPEIFINT]		= 6,
+	[IRQ_DM355_OSDINT]		= 7,
+	[IRQ_DM355_VENCINT]		= 6,
+	[IRQ_ASQINT]			= 6,
+	[IRQ_IMXINT]			= 6,
+	[IRQ_USBINT]			= 4,
+	[IRQ_DM355_RTOINT]		= 4,
+	[IRQ_DM355_UARTINT2]		= 7,
+	[IRQ_DM355_TINT6]		= 7,
+	[IRQ_CCINT0]			= 5,	/* dma */
+	[IRQ_CCERRINT]			= 5,	/* dma */
+	[IRQ_TCERRINT0]			= 5,	/* dma */
+	[IRQ_TCERRINT]			= 5,	/* dma */
+	[IRQ_DM355_SPINT2_1]		= 7,
+	[IRQ_DM355_TINT7]		= 4,
+	[IRQ_DM355_SDIOINT0]		= 7,
+	[IRQ_MBXINT]			= 7,
+	[IRQ_MBRINT]			= 7,
+	[IRQ_MMCINT]			= 7,
+	[IRQ_DM355_MMCINT1]		= 7,
+	[IRQ_DM355_PWMINT3]		= 7,
+	[IRQ_DDRINT]			= 7,
+	[IRQ_AEMIFINT]			= 7,
+	[IRQ_DM355_SDIOINT1]		= 4,
+	[IRQ_TINT0_TINT12]		= 2,	/* clockevent */
+	[IRQ_TINT0_TINT34]		= 2,	/* clocksource */
+	[IRQ_TINT1_TINT12]		= 7,	/* DSP timer */
+	[IRQ_TINT1_TINT34]		= 7,	/* system tick */
+	[IRQ_PWMINT0]			= 7,
+	[IRQ_PWMINT1]			= 7,
+	[IRQ_PWMINT2]			= 7,
+	[IRQ_I2C]			= 3,
+	[IRQ_UARTINT0]			= 3,
+	[IRQ_UARTINT1]			= 3,
+	[IRQ_DM355_SPINT0_0]		= 3,
+	[IRQ_DM355_SPINT0_1]		= 3,
+	[IRQ_DM355_GPIO0]		= 3,
+	[IRQ_DM355_GPIO1]		= 7,
+	[IRQ_DM355_GPIO2]		= 4,
+	[IRQ_DM355_GPIO3]		= 4,
+	[IRQ_DM355_GPIO4]		= 7,
+	[IRQ_DM355_GPIO5]		= 7,
+	[IRQ_DM355_GPIO6]		= 7,
+	[IRQ_DM355_GPIO7]		= 7,
+	[IRQ_DM355_GPIO8]		= 7,
+	[IRQ_DM355_GPIO9]		= 7,
+	[IRQ_DM355_GPIOBNK0]		= 7,
+	[IRQ_DM355_GPIOBNK1]		= 7,
+	[IRQ_DM355_GPIOBNK2]		= 7,
+	[IRQ_DM355_GPIOBNK3]		= 7,
+	[IRQ_DM355_GPIOBNK4]		= 7,
+	[IRQ_DM355_GPIOBNK5]		= 7,
+	[IRQ_DM355_GPIOBNK6]		= 7,
+	[IRQ_COMMTX]			= 7,
+	[IRQ_COMMRX]			= 7,
+	[IRQ_EMUINT]			= 7,
+};
+
 /* ARM Interrupt Controller Initialization */
 void __init davinci_irq_init(void)
 {
 	unsigned i;
-	const u8 *priority = default_priorities;
+
+	if (cpu_is_davinci_dm644x())
+		davinci_def_priorities = dm644x_default_priorities;
+	else if (cpu_is_davinci_dm646x())
+		davinci_def_priorities = dm646x_default_priorities;
+	else if (cpu_is_davinci_dm355())
+		davinci_def_priorities = dm355_default_priorities;
 
 	/* Clear all interrupt requests */
 	davinci_irq_writel(~0x0, FIQ_REG0_OFFSET);
@@ -209,8 +351,8 @@ void __init davinci_irq_init(void)
 		unsigned	j;
 		u32		pri;
 
-		for (j = 0, pri = 0; j < 32; j += 4, priority++)
-			pri |= (*priority & 0x07) << j;
+		for (j = 0, pri = 0; j < 32; j += 4, davinci_def_priorities++)
+			pri |= (*davinci_def_priorities & 0x07) << j;
 		davinci_irq_writel(pri, i);
 	}
 
diff --git a/arch/arm/mach-davinci/mux.c b/arch/arm/mach-davinci/mux.c
index 8ff9d8aca60..bbba0b247a4 100644
--- a/arch/arm/mach-davinci/mux.c
+++ b/arch/arm/mach-davinci/mux.c
@@ -1,41 +1,103 @@
 /*
- * DaVinci pin multiplexing configurations
+ * Utility to set the DAVINCI MUX register from a table in mux.h
  *
  * Author: Vladimir Barinov, MontaVista Software, Inc. <source@mvista.com>
  *
+ * Based on linux/arch/arm/plat-omap/mux.c:
+ * Copyright (C) 2003 - 2005 Nokia Corporation
+ *
+ * Written by Tony Lindgren
+ *
  * 2007 (c) MontaVista Software, Inc. This file is licensed under
  * the terms of the GNU General Public License version 2. This program
  * is licensed "as is" without any warranty of any kind, whether express
  * or implied.
+ *
+ * Copyright (C) 2008 Texas Instruments.
  */
 #include <linux/io.h>
+#include <linux/module.h>
 #include <linux/spinlock.h>
 
 #include <mach/hardware.h>
-
 #include <mach/mux.h>
 
-/* System control register offsets */
-#define PINMUX0         0x00
-#define PINMUX1         0x04
+static const struct mux_config *mux_table;
+static unsigned long pin_table_sz;
+
+int __init davinci_mux_register(const struct mux_config *pins,
+				unsigned long size)
+{
+	mux_table = pins;
+	pin_table_sz = size;
 
-static DEFINE_SPINLOCK(mux_lock);
+	return 0;
+}
 
-void davinci_mux_peripheral(unsigned int mux, unsigned int enable)
+/*
+ * Sets the DAVINCI MUX register based on the table
+ */
+int __init_or_module davinci_cfg_reg(const unsigned long index)
 {
-	u32 pinmux, muxreg = PINMUX0;
+	static DEFINE_SPINLOCK(mux_spin_lock);
+	void __iomem *base = IO_ADDRESS(DAVINCI_SYSTEM_MODULE_BASE);
+	unsigned long flags;
+	const struct mux_config *cfg;
+	unsigned int reg_orig = 0, reg = 0;
+	unsigned int mask, warn = 0;
+
+	if (!mux_table)
+		BUG();
+
+	if (index >= pin_table_sz) {
+		printk(KERN_ERR "Invalid pin mux index: %lu (%lu)\n",
+		       index, pin_table_sz);
+		dump_stack();
+		return -ENODEV;
+	}
+
+	cfg = &mux_table[index];
+
+	if (cfg->name == NULL) {
+		printk(KERN_ERR "No entry for the specified index\n");
+		return -ENODEV;
+	}
+
+	/* Update the mux register in question */
+	if (cfg->mask) {
+		unsigned	tmp1, tmp2;
+
+		spin_lock_irqsave(&mux_spin_lock, flags);
+		reg_orig = __raw_readl(base + cfg->mux_reg);
+
+		mask = (cfg->mask << cfg->mask_offset);
+		tmp1 = reg_orig & mask;
+		reg = reg_orig & ~mask;
+
+		tmp2 = (cfg->mode << cfg->mask_offset);
+		reg |= tmp2;
+
+		if (tmp1 != tmp2)
+			warn = 1;
+
+		__raw_writel(reg, base + cfg->mux_reg);
+		spin_unlock_irqrestore(&mux_spin_lock, flags);
+	}
+
+	if (warn) {
+#ifdef CONFIG_DAVINCI_MUX_WARNINGS
+		printk(KERN_WARNING "MUX: initialized %s\n", cfg->name);
+#endif
+	}
 
-	if (mux >= DAVINCI_MUX_LEVEL2) {
-		muxreg = PINMUX1;
-		mux -= DAVINCI_MUX_LEVEL2;
+#ifdef CONFIG_DAVINCI_MUX_DEBUG
+	if (cfg->debug || warn) {
+		printk(KERN_WARNING "MUX: Setting register %s\n", cfg->name);
+		printk(KERN_WARNING "	   %s (0x%08x) = 0x%08x -> 0x%08x\n",
+		       cfg->mux_reg_name, cfg->mux_reg, reg_orig, reg);
 	}
+#endif
 
-	spin_lock(&mux_lock);
-	pinmux = davinci_readl(DAVINCI_SYSTEM_MODULE_BASE + muxreg);
-	if (enable)
-		pinmux |= (1 << mux);
-	else
-		pinmux &= ~(1 << mux);
-	davinci_writel(pinmux, DAVINCI_SYSTEM_MODULE_BASE + muxreg);
-	spin_unlock(&mux_lock);
+	return 0;
 }
+EXPORT_SYMBOL(davinci_cfg_reg);
diff --git a/arch/arm/mach-davinci/mux.h b/arch/arm/mach-davinci/mux.h
new file mode 100644
index 00000000000..adc86941337
--- /dev/null
+++ b/arch/arm/mach-davinci/mux.h
@@ -0,0 +1,51 @@
+/*
+ * Pin-multiplex helper macros for TI DaVinci family devices
+ *
+ * Author: Vladimir Barinov, MontaVista Software, Inc. <source@mvista.com>
+ *
+ * 2007 (c) MontaVista Software, Inc. This file is licensed under
+ * the terms of the GNU General Public License version 2. This program
+ * is licensed "as is" without any warranty of any kind, whether express
+ * or implied.
+ *
+ * Copyright (C) 2008 Texas Instruments.
+ */
+#ifndef _MACH_DAVINCI_MUX_H_
+#define _MACH_DAVINCI_MUX_H_
+
+#include <mach/mux.h>
+
+#define MUX_CFG(soc, desc, muxreg, mode_offset, mode_mask, mux_mode, dbg)\
+[soc##_##desc] = {							\
+			.name =  #desc,					\
+			.debug = dbg,					\
+			.mux_reg_name = "PINMUX"#muxreg,		\
+			.mux_reg = PINMUX##muxreg,			\
+			.mask_offset = mode_offset,			\
+			.mask = mode_mask,				\
+			.mode = mux_mode,				\
+		},
+
+#define INT_CFG(soc, desc, mode_offset, mode_mask, mux_mode, dbg)	\
+[soc##_##desc] = {							\
+			.name =  #desc,					\
+			.debug = dbg,					\
+			.mux_reg_name = "INTMUX",			\
+			.mux_reg = INTMUX,				\
+			.mask_offset = mode_offset,			\
+			.mask = mode_mask,				\
+			.mode = mux_mode,				\
+		},
+
+#define EVT_CFG(soc, desc, mode_offset, mode_mask, mux_mode, dbg)	\
+[soc##_##desc] = {							\
+			.name =  #desc,					\
+			.debug = dbg,					\
+			.mux_reg_name = "EVTMUX",			\
+			.mux_reg = EVTMUX,				\
+			.mask_offset = mode_offset,			\
+			.mask = mode_mask,				\
+			.mode = mux_mode,				\
+		},
+
+#endif /* _MACH_DAVINCI_MUX_H */
diff --git a/arch/arm/mach-davinci/psc.c b/arch/arm/mach-davinci/psc.c
index 58754f066d5..84171abf5f7 100644
--- a/arch/arm/mach-davinci/psc.c
+++ b/arch/arm/mach-davinci/psc.c
@@ -23,10 +23,13 @@
 #include <linux/init.h>
 #include <linux/io.h>
 
+#include <mach/cputype.h>
 #include <mach/hardware.h>
 #include <mach/psc.h>
 #include <mach/mux.h>
 
+#define DAVINCI_PWR_SLEEP_CNTRL_BASE 0x01C41000
+
 /* PSC register offsets */
 #define EPCPR		0x070
 #define PTCMD		0x120
@@ -36,102 +39,61 @@
 #define MDSTAT		0x800
 #define MDCTL		0xA00
 
-/* System control register offsets */
-#define VDD3P3V_PWDN	0x48
+#define MDSTAT_STATE_MASK 0x1f
 
-static void davinci_psc_mux(unsigned int id)
+/* Return nonzero iff the domain's clock is active */
+int __init davinci_psc_is_clk_active(unsigned int id)
 {
-	switch (id) {
-	case DAVINCI_LPSC_ATA:
-		davinci_mux_peripheral(DAVINCI_MUX_HDIREN, 1);
-		davinci_mux_peripheral(DAVINCI_MUX_ATAEN, 1);
-		break;
-	case DAVINCI_LPSC_MMC_SD:
-		/* VDD power manupulations are done in U-Boot for CPMAC
-		 * so applies to MMC as well
-		 */
-		/*Set up the pull regiter for MMC */
-		davinci_writel(0, DAVINCI_SYSTEM_MODULE_BASE + VDD3P3V_PWDN);
-		davinci_mux_peripheral(DAVINCI_MUX_MSTK, 0);
-		break;
-	case DAVINCI_LPSC_I2C:
-		davinci_mux_peripheral(DAVINCI_MUX_I2C, 1);
-		break;
-	case DAVINCI_LPSC_McBSP:
-		davinci_mux_peripheral(DAVINCI_MUX_ASP, 1);
-		break;
-	default:
-		break;
-	}
+	void __iomem *psc_base = IO_ADDRESS(DAVINCI_PWR_SLEEP_CNTRL_BASE);
+	u32 mdstat = __raw_readl(psc_base + MDSTAT + 4 * id);
+
+	/* if clocked, state can be "Enable" or "SyncReset" */
+	return mdstat & BIT(12);
 }
 
 /* Enable or disable a PSC domain */
 void davinci_psc_config(unsigned int domain, unsigned int id, char enable)
 {
-	u32 epcpr, ptcmd, ptstat, pdstat, pdctl1, mdstat, mdctl, mdstat_mask;
+	u32 epcpr, ptcmd, ptstat, pdstat, pdctl1, mdstat, mdctl;
+	void __iomem *psc_base = IO_ADDRESS(DAVINCI_PWR_SLEEP_CNTRL_BASE);
+	u32 next_state = enable ? 0x3 : 0x2; /* 0x3 enables, 0x2 disables */
 
-	mdctl = davinci_readl(DAVINCI_PWR_SLEEP_CNTRL_BASE + MDCTL + 4 * id);
-	if (enable)
-		mdctl |= 0x00000003;	/* Enable Module */
-	else
-		mdctl &= 0xFFFFFFF2;	/* Disable Module */
-	davinci_writel(mdctl, DAVINCI_PWR_SLEEP_CNTRL_BASE + MDCTL + 4 * id);
+	mdctl = __raw_readl(psc_base + MDCTL + 4 * id);
+	mdctl &= ~MDSTAT_STATE_MASK;
+	mdctl |= next_state;
+	__raw_writel(mdctl, psc_base + MDCTL + 4 * id);
 
-	pdstat = davinci_readl(DAVINCI_PWR_SLEEP_CNTRL_BASE + PDSTAT);
+	pdstat = __raw_readl(psc_base + PDSTAT);
 	if ((pdstat & 0x00000001) == 0) {
-		pdctl1 = davinci_readl(DAVINCI_PWR_SLEEP_CNTRL_BASE + PDCTL1);
+		pdctl1 = __raw_readl(psc_base + PDCTL1);
 		pdctl1 |= 0x1;
-		davinci_writel(pdctl1, DAVINCI_PWR_SLEEP_CNTRL_BASE + PDCTL1);
+		__raw_writel(pdctl1, psc_base + PDCTL1);
 
 		ptcmd = 1 << domain;
-		davinci_writel(ptcmd, DAVINCI_PWR_SLEEP_CNTRL_BASE + PTCMD);
+		__raw_writel(ptcmd, psc_base + PTCMD);
 
 		do {
-			epcpr = davinci_readl(DAVINCI_PWR_SLEEP_CNTRL_BASE +
-					      EPCPR);
+			epcpr = __raw_readl(psc_base + EPCPR);
 		} while ((((epcpr >> domain) & 1) == 0));
 
-		pdctl1 = davinci_readl(DAVINCI_PWR_SLEEP_CNTRL_BASE + PDCTL1);
+		pdctl1 = __raw_readl(psc_base + PDCTL1);
 		pdctl1 |= 0x100;
-		davinci_writel(pdctl1, DAVINCI_PWR_SLEEP_CNTRL_BASE + PDCTL1);
+		__raw_writel(pdctl1, psc_base + PDCTL1);
 
 		do {
-			ptstat = davinci_readl(DAVINCI_PWR_SLEEP_CNTRL_BASE +
+			ptstat = __raw_readl(psc_base +
 					       PTSTAT);
 		} while (!(((ptstat >> domain) & 1) == 0));
 	} else {
 		ptcmd = 1 << domain;
-		davinci_writel(ptcmd, DAVINCI_PWR_SLEEP_CNTRL_BASE + PTCMD);
+		__raw_writel(ptcmd, psc_base + PTCMD);
 
 		do {
-			ptstat = davinci_readl(DAVINCI_PWR_SLEEP_CNTRL_BASE +
-					       PTSTAT);
+			ptstat = __raw_readl(psc_base + PTSTAT);
 		} while (!(((ptstat >> domain) & 1) == 0));
 	}
 
-	if (enable)
-		mdstat_mask = 0x3;
-	else
-		mdstat_mask = 0x2;
-
 	do {
-		mdstat = davinci_readl(DAVINCI_PWR_SLEEP_CNTRL_BASE +
-				       MDSTAT + 4 * id);
-	} while (!((mdstat & 0x0000001F) == mdstat_mask));
-
-	if (enable)
-		davinci_psc_mux(id);
-}
-
-void __init davinci_psc_init(void)
-{
-	davinci_psc_config(DAVINCI_GPSC_ARMDOMAIN, DAVINCI_LPSC_VPSSMSTR, 1);
-	davinci_psc_config(DAVINCI_GPSC_ARMDOMAIN, DAVINCI_LPSC_VPSSSLV, 1);
-	davinci_psc_config(DAVINCI_GPSC_ARMDOMAIN, DAVINCI_LPSC_TPCC, 1);
-	davinci_psc_config(DAVINCI_GPSC_ARMDOMAIN, DAVINCI_LPSC_TPTC0, 1);
-	davinci_psc_config(DAVINCI_GPSC_ARMDOMAIN, DAVINCI_LPSC_TPTC1, 1);
-	davinci_psc_config(DAVINCI_GPSC_ARMDOMAIN, DAVINCI_LPSC_GPIO, 1);
-
-	/* Turn on WatchDog timer LPSC.	 Needed for RESET to work */
-	davinci_psc_config(DAVINCI_GPSC_ARMDOMAIN, DAVINCI_LPSC_TIMER2, 1);
+		mdstat = __raw_readl(psc_base + MDSTAT + 4 * id);
+	} while (!((mdstat & MDSTAT_STATE_MASK) == next_state));
 }
diff --git a/arch/arm/mach-davinci/serial.c b/arch/arm/mach-davinci/serial.c
index 3010f997125..69507579652 100644
--- a/arch/arm/mach-davinci/serial.c
+++ b/arch/arm/mach-davinci/serial.c
@@ -32,32 +32,47 @@
 #include <mach/hardware.h>
 #include <mach/serial.h>
 #include <mach/irqs.h>
+#include <mach/cputype.h>
+#include "clock.h"
 
-#define UART_DAVINCI_PWREMU 0x0c
-
-static inline unsigned int davinci_serial_in(struct plat_serial8250_port *up,
-					  int offset)
+static inline unsigned int serial_read_reg(struct plat_serial8250_port *up,
+					   int offset)
 {
 	offset <<= up->regshift;
-	return (unsigned int)__raw_readb(up->membase + offset);
+	return (unsigned int)__raw_readl(IO_ADDRESS(up->mapbase) + offset);
 }
 
-static inline void davinci_serial_outp(struct plat_serial8250_port *p,
-				       int offset, int value)
+static inline void serial_write_reg(struct plat_serial8250_port *p, int offset,
+				    int value)
 {
 	offset <<= p->regshift;
-	__raw_writeb(value, p->membase + offset);
+	__raw_writel(value, IO_ADDRESS(p->mapbase) + offset);
 }
 
 static struct plat_serial8250_port serial_platform_data[] = {
 	{
-		.membase	= (char *)IO_ADDRESS(DAVINCI_UART0_BASE),
-		.mapbase	= (unsigned long)DAVINCI_UART0_BASE,
+		.mapbase	= DAVINCI_UART0_BASE,
 		.irq		= IRQ_UARTINT0,
-		.flags		= UPF_BOOT_AUTOCONF | UPF_SKIP_TEST,
+		.flags		= UPF_BOOT_AUTOCONF | UPF_SKIP_TEST |
+				  UPF_IOREMAP,
+		.iotype		= UPIO_MEM,
+		.regshift	= 2,
+	},
+	{
+		.mapbase	= DAVINCI_UART1_BASE,
+		.irq		= IRQ_UARTINT1,
+		.flags		= UPF_BOOT_AUTOCONF | UPF_SKIP_TEST |
+				  UPF_IOREMAP,
+		.iotype		= UPIO_MEM,
+		.regshift	= 2,
+	},
+	{
+		.mapbase	= DAVINCI_UART2_BASE,
+		.irq		= IRQ_UARTINT2,
+		.flags		= UPF_BOOT_AUTOCONF | UPF_SKIP_TEST |
+				  UPF_IOREMAP,
 		.iotype		= UPIO_MEM,
 		.regshift	= 2,
-		.uartclk	= 27000000,
 	},
 	{
 		.flags		= 0
@@ -74,22 +89,68 @@ static struct platform_device serial_device = {
 
 static void __init davinci_serial_reset(struct plat_serial8250_port *p)
 {
-	/* reset both transmitter and receiver: bits 14,13 = UTRST, URRST */
 	unsigned int pwremu = 0;
 
-	davinci_serial_outp(p, UART_IER, 0);  /* disable all interrupts */
+	serial_write_reg(p, UART_IER, 0);  /* disable all interrupts */
 
-	davinci_serial_outp(p, UART_DAVINCI_PWREMU, pwremu);
+	/* reset both transmitter and receiver: bits 14,13 = UTRST, URRST */
+	serial_write_reg(p, UART_DAVINCI_PWREMU, pwremu);
 	mdelay(10);
 
 	pwremu |= (0x3 << 13);
 	pwremu |= 0x1;
-	davinci_serial_outp(p, UART_DAVINCI_PWREMU, pwremu);
+	serial_write_reg(p, UART_DAVINCI_PWREMU, pwremu);
+
+	if (cpu_is_davinci_dm646x())
+		serial_write_reg(p, UART_DM646X_SCR,
+				 UART_DM646X_SCR_TX_WATERMARK);
+}
+
+void __init davinci_serial_init(struct davinci_uart_config *info)
+{
+	int i;
+	char name[16];
+	struct clk *uart_clk;
+	struct device *dev = &serial_device.dev;
+
+	/*
+	 * Make sure the serial ports are muxed on at this point.
+	 * You have to mux them off in device drivers later on
+	 * if not needed.
+	 */
+	for (i = 0; i < DAVINCI_MAX_NR_UARTS; i++) {
+		struct plat_serial8250_port *p = serial_platform_data + i;
+
+		if (!(info->enabled_uarts & (1 << i))) {
+			p->flags = 0;
+			continue;
+		}
+
+		if (cpu_is_davinci_dm646x())
+			p->iotype = UPIO_MEM32;
+
+		if (cpu_is_davinci_dm355()) {
+			if (i == 2) {
+				p->mapbase = (unsigned long)DM355_UART2_BASE;
+				p->irq = IRQ_DM355_UARTINT2;
+			}
+		}
+
+		sprintf(name, "uart%d", i);
+		uart_clk = clk_get(dev, name);
+		if (IS_ERR(uart_clk))
+			printk(KERN_ERR "%s:%d: failed to get UART%d clock\n",
+					__func__, __LINE__, i);
+		else {
+			clk_enable(uart_clk);
+			p->uartclk = clk_get_rate(uart_clk);
+			davinci_serial_reset(p);
+		}
+	}
 }
 
 static int __init davinci_init(void)
 {
-	davinci_serial_reset(&serial_platform_data[0]);
 	return platform_device_register(&serial_device);
 }
 
diff --git a/arch/arm/mach-davinci/time.c b/arch/arm/mach-davinci/time.c
index f8bcd29d17a..494e01bff5c 100644
--- a/arch/arm/mach-davinci/time.c
+++ b/arch/arm/mach-davinci/time.c
@@ -16,6 +16,9 @@
 #include <linux/clockchips.h>
 #include <linux/spinlock.h>
 #include <linux/io.h>
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/device.h>
 
 #include <mach/hardware.h>
 #include <asm/system.h>
@@ -24,8 +27,11 @@
 #include <asm/mach/time.h>
 #include <asm/errno.h>
 #include <mach/io.h>
+#include <mach/cputype.h>
+#include "clock.h"
 
 static struct clock_event_device clockevent_davinci;
+static unsigned int davinci_clock_tick_rate;
 
 #define DAVINCI_TIMER0_BASE (IO_PHYS + 0x21400)
 #define DAVINCI_TIMER1_BASE (IO_PHYS + 0x21800)
@@ -99,9 +105,9 @@ struct timer_s {
 	unsigned int id;
 	unsigned long period;
 	unsigned long opts;
-	unsigned long reg_base;
-	unsigned long tim_reg;
-	unsigned long prd_reg;
+	void __iomem *base;
+	unsigned long tim_off;
+	unsigned long prd_off;
 	unsigned long enamode_shift;
 	struct irqaction irqaction;
 };
@@ -114,15 +120,15 @@ static struct timer_s timers[];
 
 static int timer32_config(struct timer_s *t)
 {
-	u32 tcr = davinci_readl(t->reg_base + TCR);
+	u32 tcr = __raw_readl(t->base + TCR);
 
 	/* disable timer */
 	tcr &= ~(TCR_ENAMODE_MASK << t->enamode_shift);
-	davinci_writel(tcr, t->reg_base + TCR);
+	__raw_writel(tcr, t->base + TCR);
 
 	/* reset counter to zero, set new period */
-	davinci_writel(0, t->tim_reg);
-	davinci_writel(t->period, t->prd_reg);
+	__raw_writel(0, t->base + t->tim_off);
+	__raw_writel(t->period, t->base + t->prd_off);
 
 	/* Set enable mode */
 	if (t->opts & TIMER_OPTS_ONESHOT) {
@@ -131,13 +137,13 @@ static int timer32_config(struct timer_s *t)
 		tcr |= TCR_ENAMODE_PERIODIC << t->enamode_shift;
 	}
 
-	davinci_writel(tcr, t->reg_base + TCR);
+	__raw_writel(tcr, t->base + TCR);
 	return 0;
 }
 
 static inline u32 timer32_read(struct timer_s *t)
 {
-	return davinci_readl(t->tim_reg);
+	return __raw_readl(t->base + t->tim_off);
 }
 
 static irqreturn_t timer_interrupt(int irq, void *dev_id)
@@ -176,51 +182,54 @@ static struct timer_s timers[] = {
 
 static void __init timer_init(void)
 {
-	u32 bases[] = {DAVINCI_TIMER0_BASE, DAVINCI_TIMER1_BASE};
+	u32 phys_bases[] = {DAVINCI_TIMER0_BASE, DAVINCI_TIMER1_BASE};
 	int i;
 
 	/* Global init of each 64-bit timer as a whole */
 	for(i=0; i<2; i++) {
-		u32 tgcr, base = bases[i];
+		u32 tgcr;
+		void __iomem *base = IO_ADDRESS(phys_bases[i]);
 
 		/* Disabled, Internal clock source */
-		davinci_writel(0, base + TCR);
+		__raw_writel(0, base + TCR);
 
 		/* reset both timers, no pre-scaler for timer34 */
 		tgcr = 0;
-		davinci_writel(tgcr, base + TGCR);
+		__raw_writel(tgcr, base + TGCR);
 
 		/* Set both timers to unchained 32-bit */
 		tgcr = TGCR_TIMMODE_32BIT_UNCHAINED << TGCR_TIMMODE_SHIFT;
-		davinci_writel(tgcr, base + TGCR);
+		__raw_writel(tgcr, base + TGCR);
 
 		/* Unreset timers */
 		tgcr |= (TGCR_UNRESET << TGCR_TIM12RS_SHIFT) |
 			(TGCR_UNRESET << TGCR_TIM34RS_SHIFT);
-		davinci_writel(tgcr, base + TGCR);
+		__raw_writel(tgcr, base + TGCR);
 
 		/* Init both counters to zero */
-		davinci_writel(0, base + TIM12);
-		davinci_writel(0, base + TIM34);
+		__raw_writel(0, base + TIM12);
+		__raw_writel(0, base + TIM34);
 	}
 
 	/* Init of each timer as a 32-bit timer */
 	for (i=0; i< ARRAY_SIZE(timers); i++) {
 		struct timer_s *t = &timers[i];
+		u32 phys_base;
 
 		if (t->name) {
 			t->id = i;
-			t->reg_base = (IS_TIMER1(t->id) ?
+			phys_base = (IS_TIMER1(t->id) ?
 			       DAVINCI_TIMER1_BASE : DAVINCI_TIMER0_BASE);
+			t->base = IO_ADDRESS(phys_base);
 
 			if (IS_TIMER_BOT(t->id)) {
 				t->enamode_shift = 6;
-				t->tim_reg = t->reg_base + TIM12;
-				t->prd_reg = t->reg_base + PRD12;
+				t->tim_off = TIM12;
+				t->prd_off = PRD12;
 			} else {
 				t->enamode_shift = 22;
-				t->tim_reg = t->reg_base + TIM34;
-				t->prd_reg = t->reg_base + PRD34;
+				t->tim_off = TIM34;
+				t->prd_off = PRD34;
 			}
 
 			/* Register interrupt */
@@ -238,7 +247,7 @@ static void __init timer_init(void)
 /*
  * clocksource
  */
-static cycle_t read_cycles(void)
+static cycle_t read_cycles(struct clocksource *cs)
 {
 	struct timer_s *t = &timers[TID_CLOCKSOURCE];
 
@@ -274,7 +283,7 @@ static void davinci_set_mode(enum clock_event_mode mode,
 
 	switch (mode) {
 	case CLOCK_EVT_MODE_PERIODIC:
-		t->period = CLOCK_TICK_RATE / (HZ);
+		t->period = davinci_clock_tick_rate / (HZ);
 		t->opts = TIMER_OPTS_PERIODIC;
 		timer32_config(t);
 		break;
@@ -301,21 +310,29 @@ static struct clock_event_device clockevent_davinci = {
 
 static void __init davinci_timer_init(void)
 {
+	struct clk *timer_clk;
+
 	static char err[] __initdata = KERN_ERR
 		"%s: can't register clocksource!\n";
 
 	/* init timer hw */
 	timer_init();
 
+	timer_clk = clk_get(NULL, "timer0");
+	BUG_ON(IS_ERR(timer_clk));
+	clk_enable(timer_clk);
+
+	davinci_clock_tick_rate = clk_get_rate(timer_clk);
+
 	/* setup clocksource */
 	clocksource_davinci.mult =
-		clocksource_khz2mult(CLOCK_TICK_RATE/1000,
+		clocksource_khz2mult(davinci_clock_tick_rate/1000,
 				     clocksource_davinci.shift);
 	if (clocksource_register(&clocksource_davinci))
 		printk(err, clocksource_davinci.name);
 
 	/* setup clockevent */
-	clockevent_davinci.mult = div_sc(CLOCK_TICK_RATE, NSEC_PER_SEC,
+	clockevent_davinci.mult = div_sc(davinci_clock_tick_rate, NSEC_PER_SEC,
 					 clockevent_davinci.shift);
 	clockevent_davinci.max_delta_ns =
 		clockevent_delta2ns(0xfffffffe, &clockevent_davinci);
@@ -333,42 +350,52 @@ struct sys_timer davinci_timer = {
 
 /* reset board using watchdog timer */
 void davinci_watchdog_reset(void) {
-	u32 tgcr, wdtcr, base = DAVINCI_WDOG_BASE;
+	u32 tgcr, wdtcr;
+	void __iomem *base = IO_ADDRESS(DAVINCI_WDOG_BASE);
+	struct device dev;
+	struct clk *wd_clk;
+	char *name = "watchdog";
+
+	dev_set_name(&dev, name);
+	wd_clk = clk_get(&dev, NULL);
+	if (WARN_ON(IS_ERR(wd_clk)))
+		return;
+	clk_enable(wd_clk);
 
 	/* disable, internal clock source */
-	davinci_writel(0, base + TCR);
+	__raw_writel(0, base + TCR);
 
 	/* reset timer, set mode to 64-bit watchdog, and unreset */
 	tgcr = 0;
-	davinci_writel(tgcr, base + TCR);
+	__raw_writel(tgcr, base + TCR);
 	tgcr = TGCR_TIMMODE_64BIT_WDOG << TGCR_TIMMODE_SHIFT;
 	tgcr |= (TGCR_UNRESET << TGCR_TIM12RS_SHIFT) |
 		(TGCR_UNRESET << TGCR_TIM34RS_SHIFT);
-	davinci_writel(tgcr, base + TCR);
+	__raw_writel(tgcr, base + TCR);
 
 	/* clear counter and period regs */
-	davinci_writel(0, base + TIM12);
-	davinci_writel(0, base + TIM34);
-	davinci_writel(0, base + PRD12);
-	davinci_writel(0, base + PRD34);
+	__raw_writel(0, base + TIM12);
+	__raw_writel(0, base + TIM34);
+	__raw_writel(0, base + PRD12);
+	__raw_writel(0, base + PRD34);
 
 	/* enable */
-	wdtcr = davinci_readl(base + WDTCR);
+	wdtcr = __raw_readl(base + WDTCR);
 	wdtcr |= WDTCR_WDEN_ENABLE << WDTCR_WDEN_SHIFT;
-	davinci_writel(wdtcr, base + WDTCR);
+	__raw_writel(wdtcr, base + WDTCR);
 
 	/* put watchdog in pre-active state */
 	wdtcr = (WDTCR_WDKEY_SEQ0 << WDTCR_WDKEY_SHIFT) |
 		(WDTCR_WDEN_ENABLE << WDTCR_WDEN_SHIFT);
-	davinci_writel(wdtcr, base + WDTCR);
+	__raw_writel(wdtcr, base + WDTCR);
 
 	/* put watchdog in active state */
 	wdtcr = (WDTCR_WDKEY_SEQ1 << WDTCR_WDKEY_SHIFT) |
 		(WDTCR_WDEN_ENABLE << WDTCR_WDEN_SHIFT);
-	davinci_writel(wdtcr, base + WDTCR);
+	__raw_writel(wdtcr, base + WDTCR);
 
 	/* write an invalid value to the WDKEY field to trigger
 	 * a watchdog reset */
 	wdtcr = 0x00004000;
-	davinci_writel(wdtcr, base + WDTCR);
+	__raw_writel(wdtcr, base + WDTCR);
 }
diff --git a/arch/arm/mach-davinci/usb.c b/arch/arm/mach-davinci/usb.c
index 2429b79f6da..abedb633718 100644
--- a/arch/arm/mach-davinci/usb.c
+++ b/arch/arm/mach-davinci/usb.c
@@ -14,6 +14,8 @@
 #include <mach/hardware.h>
 #include <mach/irqs.h>
 
+#define DAVINCI_USB_OTG_BASE 0x01C64000
+
 #if defined(CONFIG_USB_MUSB_HDRC) || defined(CONFIG_USB_MUSB_HDRC_MODULE)
 static struct musb_hdrc_eps_bits musb_eps[] = {
 	{ "ep1_tx", 8, },