diff --git a/Documentation/powerpc/dts-bindings/fsl/spi.txt b/Documentation/powerpc/dts-bindings/fsl/spi.txt
index 80510c0..777abd7 100644
--- a/Documentation/powerpc/dts-bindings/fsl/spi.txt
+++ b/Documentation/powerpc/dts-bindings/fsl/spi.txt
@@ -1,7 +1,9 @@
 * SPI (Serial Peripheral Interface)
 
 Required properties:
-- cell-index : SPI controller index.
+- cell-index : QE SPI subblock index.
+		0: QE subblock SPI1
+		1: QE subblock SPI2
 - compatible : should be "fsl,spi".
 - mode : the SPI operation mode, it can be "cpu" or "cpu-qe".
 - reg : Offset and length of the register set for the device
@@ -29,3 +31,23 @@
 		gpios = <&gpio 18 1	// device reg=<0>
 			 &gpio 19 1>;	// device reg=<1>
 	};
+
+
+* eSPI (Enhanced Serial Peripheral Interface)
+
+Required properties:
+- compatible : should be "fsl,mpc8536-espi".
+- reg : Offset and length of the register set for the device.
+- interrupts : should contain eSPI interrupt, the device has one interrupt.
+- fsl,espi-num-chipselects : the number of the chipselect signals.
+
+Example:
+	spi@110000 {
+		#address-cells = <1>;
+		#size-cells = <0>;
+		compatible = "fsl,mpc8536-espi";
+		reg = <0x110000 0x1000>;
+		interrupts = <53 0x2>;
+		interrupt-parent = <&mpic>;
+		fsl,espi-num-chipselects = <4>;
+	};
diff --git a/arch/arm/mach-lpc32xx/phy3250.c b/arch/arm/mach-lpc32xx/phy3250.c
index bc9a42d..0c936cf 100644
--- a/arch/arm/mach-lpc32xx/phy3250.c
+++ b/arch/arm/mach-lpc32xx/phy3250.c
@@ -172,18 +172,12 @@
 }
 
 static struct pl022_config_chip spi0_chip_info = {
-	.lbm			= LOOPBACK_DISABLED,
 	.com_mode		= INTERRUPT_TRANSFER,
 	.iface			= SSP_INTERFACE_MOTOROLA_SPI,
 	.hierarchy		= SSP_MASTER,
 	.slave_tx_disable	= 0,
-	.endian_tx		= SSP_TX_LSB,
-	.endian_rx		= SSP_RX_LSB,
-	.data_size		= SSP_DATA_BITS_8,
 	.rx_lev_trig		= SSP_RX_4_OR_MORE_ELEM,
 	.tx_lev_trig		= SSP_TX_4_OR_MORE_EMPTY_LOC,
-	.clk_phase		= SSP_CLK_FIRST_EDGE,
-	.clk_pol		= SSP_CLK_POL_IDLE_LOW,
 	.ctrl_len		= SSP_BITS_8,
 	.wait_state		= SSP_MWIRE_WAIT_ZERO,
 	.duplex			= SSP_MICROWIRE_CHANNEL_FULL_DUPLEX,
@@ -239,6 +233,7 @@
 			.max_speed_hz = 5000000,
 			.bus_num = 0,
 			.chip_select = 0,
+			.mode = SPI_MODE_0,
 			.platform_data = &eeprom,
 			.controller_data = &spi0_chip_info,
 		},
diff --git a/arch/arm/mach-u300/dummyspichip.c b/arch/arm/mach-u300/dummyspichip.c
index 5f55012..03f7936 100644
--- a/arch/arm/mach-u300/dummyspichip.c
+++ b/arch/arm/mach-u300/dummyspichip.c
@@ -46,7 +46,6 @@
 	 * struct, this is just used here to alter the behaviour of the chip
 	 * in order to perform tests.
 	 */
-	struct pl022_config_chip *chip_info = spi->controller_data;
 	int status;
 	u8 txbuf[14] = {0xDE, 0xAD, 0xBE, 0xEF, 0x2B, 0xAD,
 			0xCA, 0xFE, 0xBA, 0xBE, 0xB1, 0x05,
@@ -72,7 +71,7 @@
 	 * Force chip to 8 bit mode
 	 * WARNING: NEVER DO THIS IN REAL DRIVER CODE, THIS SHOULD BE STATIC!
 	 */
-	chip_info->data_size = SSP_DATA_BITS_8;
+	spi->bits_per_word = 8;
 	/* You should NOT DO THIS EITHER */
 	spi->master->setup(spi);
 
@@ -159,7 +158,7 @@
 	 * Force chip to 16 bit mode
 	 * WARNING: NEVER DO THIS IN REAL DRIVER CODE, THIS SHOULD BE STATIC!
 	 */
-	chip_info->data_size = SSP_DATA_BITS_16;
+	spi->bits_per_word = 16;
 	/* You should NOT DO THIS EITHER */
 	spi->master->setup(spi);
 
diff --git a/arch/arm/mach-u300/spi.c b/arch/arm/mach-u300/spi.c
index f0e887b..edb2c0d 100644
--- a/arch/arm/mach-u300/spi.c
+++ b/arch/arm/mach-u300/spi.c
@@ -30,8 +30,6 @@
 }
 
 struct pl022_config_chip dummy_chip_info = {
-	/* Nominally this is LOOPBACK_DISABLED, but this is our dummy chip! */
-	.lbm = LOOPBACK_ENABLED,
 	/*
 	 * available POLLING_TRANSFER and INTERRUPT_TRANSFER,
 	 * DMA_TRANSFER does not work
@@ -42,14 +40,8 @@
 	.hierarchy = SSP_MASTER,
 	/* 0 = drive TX even as slave, 1 = do not drive TX as slave */
 	.slave_tx_disable = 0,
-	/* LSB first */
-	.endian_tx = SSP_TX_LSB,
-	.endian_rx = SSP_RX_LSB,
-	.data_size = SSP_DATA_BITS_8, /* used to be 12 in some default */
 	.rx_lev_trig = SSP_RX_1_OR_MORE_ELEM,
 	.tx_lev_trig = SSP_TX_1_OR_MORE_EMPTY_LOC,
-	.clk_phase = SSP_CLK_SECOND_EDGE,
-	.clk_pol = SSP_CLK_POL_IDLE_LOW,
 	.ctrl_len = SSP_BITS_12,
 	.wait_state = SSP_MWIRE_WAIT_ZERO,
 	.duplex = SSP_MICROWIRE_CHANNEL_FULL_DUPLEX,
@@ -75,7 +67,7 @@
 		.bus_num        = 0, /* Only one bus on this chip */
 		.chip_select    = 0,
 		/* Means SPI_CS_HIGH, change if e.g low CS */
-		.mode           = 0,
+		.mode           = SPI_MODE_1 | SPI_LSB_FIRST | SPI_LOOP,
 	},
 #endif
 };
diff --git a/arch/arm/mach-ux500/board-mop500.c b/arch/arm/mach-ux500/board-mop500.c
index 0e8fd13..219ae0c 100644
--- a/arch/arm/mach-ux500/board-mop500.c
+++ b/arch/arm/mach-ux500/board-mop500.c
@@ -55,19 +55,13 @@
 }
 
 struct pl022_config_chip ab4500_chip_info = {
-	.lbm = LOOPBACK_DISABLED,
 	.com_mode = INTERRUPT_TRANSFER,
 	.iface = SSP_INTERFACE_MOTOROLA_SPI,
 	/* we can act as master only */
 	.hierarchy = SSP_MASTER,
 	.slave_tx_disable = 0,
-	.endian_rx = SSP_RX_MSB,
-	.endian_tx = SSP_TX_MSB,
-	.data_size = SSP_DATA_BITS_24,
 	.rx_lev_trig = SSP_RX_1_OR_MORE_ELEM,
 	.tx_lev_trig = SSP_TX_1_OR_MORE_EMPTY_LOC,
-	.clk_phase = SSP_CLK_SECOND_EDGE,
-	.clk_pol = SSP_CLK_POL_IDLE_HIGH,
 	.cs_control = ab4500_spi_cs_control,
 };
 
@@ -83,7 +77,7 @@
 		.max_speed_hz = 12000000,
 		.bus_num = 0,
 		.chip_select = 0,
-		.mode = SPI_MODE_0,
+		.mode = SPI_MODE_3,
 		.irq = IRQ_DB8500_AB8500,
 	},
 };
diff --git a/arch/arm/plat-samsung/include/plat/s3c64xx-spi.h b/arch/arm/plat-samsung/include/plat/s3c64xx-spi.h
index e5aba8f..b226f740 100644
--- a/arch/arm/plat-samsung/include/plat/s3c64xx-spi.h
+++ b/arch/arm/plat-samsung/include/plat/s3c64xx-spi.h
@@ -32,6 +32,8 @@
  * struct s3c64xx_spi_info - SPI Controller defining structure
  * @src_clk_nr: Clock source index for the CLK_CFG[SPI_CLKSEL] field.
  * @src_clk_name: Platform name of the corresponding clock.
+ * @clk_from_cmu: If the SPI clock/prescalar control block is present
+ *     by the platform's clock-management-unit and not in SPI controller.
  * @num_cs: Number of CS this controller emulates.
  * @cfg_gpio: Configure pins for this SPI controller.
  * @fifo_lvl_mask: All tx fifo_lvl fields start at offset-6
@@ -41,6 +43,7 @@
 struct s3c64xx_spi_info {
 	int src_clk_nr;
 	char *src_clk_name;
+	bool clk_from_cmu;
 
 	int num_cs;
 
diff --git a/arch/blackfin/include/asm/bfin5xx_spi.h b/arch/blackfin/include/asm/bfin5xx_spi.h
index ed4f8c6..4223cf0 100644
--- a/arch/blackfin/include/asm/bfin5xx_spi.h
+++ b/arch/blackfin/include/asm/bfin5xx_spi.h
@@ -11,26 +11,17 @@
 
 #define MIN_SPI_BAUD_VAL	2
 
-#define SPI_READ              0
-#define SPI_WRITE             1
-
-#define SPI_CTRL_OFF            0x0
-#define SPI_FLAG_OFF            0x4
-#define SPI_STAT_OFF            0x8
-#define SPI_TXBUFF_OFF          0xc
-#define SPI_RXBUFF_OFF          0x10
-#define SPI_BAUD_OFF            0x14
-#define SPI_SHAW_OFF            0x18
-
-
 #define BIT_CTL_ENABLE      0x4000
 #define BIT_CTL_OPENDRAIN   0x2000
 #define BIT_CTL_MASTER      0x1000
-#define BIT_CTL_POLAR       0x0800
-#define BIT_CTL_PHASE       0x0400
-#define BIT_CTL_BITORDER    0x0200
+#define BIT_CTL_CPOL        0x0800
+#define BIT_CTL_CPHA        0x0400
+#define BIT_CTL_LSBF        0x0200
 #define BIT_CTL_WORDSIZE    0x0100
-#define BIT_CTL_MISOENABLE  0x0020
+#define BIT_CTL_EMISO       0x0020
+#define BIT_CTL_PSSE        0x0010
+#define BIT_CTL_GM          0x0008
+#define BIT_CTL_SZ          0x0004
 #define BIT_CTL_RXMOD       0x0000
 #define BIT_CTL_TXMOD       0x0001
 #define BIT_CTL_TIMOD_DMA_TX 0x0003
@@ -50,61 +41,7 @@
 #define BIT_STU_SENDOVER    0x0001
 #define BIT_STU_RECVFULL    0x0020
 
-#define CFG_SPI_ENABLE      1
-#define CFG_SPI_DISABLE     0
-
-#define CFG_SPI_OUTENABLE   1
-#define CFG_SPI_OUTDISABLE  0
-
-#define CFG_SPI_ACTLOW      1
-#define CFG_SPI_ACTHIGH     0
-
-#define CFG_SPI_PHASESTART  1
-#define CFG_SPI_PHASEMID    0
-
-#define CFG_SPI_MASTER      1
-#define CFG_SPI_SLAVE       0
-
-#define CFG_SPI_SENELAST    0
-#define CFG_SPI_SENDZERO    1
-
-#define CFG_SPI_RCVFLUSH    1
-#define CFG_SPI_RCVDISCARD  0
-
-#define CFG_SPI_LSBFIRST    1
-#define CFG_SPI_MSBFIRST    0
-
-#define CFG_SPI_WORDSIZE16  1
-#define CFG_SPI_WORDSIZE8   0
-
-#define CFG_SPI_MISOENABLE   1
-#define CFG_SPI_MISODISABLE  0
-
-#define CFG_SPI_READ      0x00
-#define CFG_SPI_WRITE     0x01
-#define CFG_SPI_DMAREAD   0x02
-#define CFG_SPI_DMAWRITE  0x03
-
-#define CFG_SPI_CSCLEARALL  0
-#define CFG_SPI_CHIPSEL1    1
-#define CFG_SPI_CHIPSEL2    2
-#define CFG_SPI_CHIPSEL3    3
-#define CFG_SPI_CHIPSEL4    4
-#define CFG_SPI_CHIPSEL5    5
-#define CFG_SPI_CHIPSEL6    6
-#define CFG_SPI_CHIPSEL7    7
-
-#define CFG_SPI_CS1VALUE    1
-#define CFG_SPI_CS2VALUE    2
-#define CFG_SPI_CS3VALUE    3
-#define CFG_SPI_CS4VALUE    4
-#define CFG_SPI_CS5VALUE    5
-#define CFG_SPI_CS6VALUE    6
-#define CFG_SPI_CS7VALUE    7
-
-#define CMD_SPI_SET_BAUDRATE  2
-#define CMD_SPI_GET_SYSTEMCLOCK   25
-#define CMD_SPI_SET_WRITECONTINUOUS     26
+#define MAX_CTRL_CS          8  /* cs in spi controller */
 
 /* device.platform_data for SSP controller devices */
 struct bfin5xx_spi_master {
@@ -120,9 +57,7 @@
 	u16 ctl_reg;
 	u8 enable_dma;
 	u8 bits_per_word;
-	u8 cs_change_per_word;
 	u16 cs_chg_udelay; /* Some devices require 16-bit delays */
-	u32 cs_gpio;
 	/* Value to send if no TX value is supplied, usually 0x0 or 0xFFFF */
 	u16 idle_tx_val;
 	u8 pio_interrupt; /* Enable spi data irq */
diff --git a/arch/powerpc/boot/dts/mpc8536ds.dts b/arch/powerpc/boot/dts/mpc8536ds.dts
index 815cebb..a75c10e 100644
--- a/arch/powerpc/boot/dts/mpc8536ds.dts
+++ b/arch/powerpc/boot/dts/mpc8536ds.dts
@@ -108,6 +108,58 @@
 			};
 		};
 
+		spi@7000 {
+			#address-cells = <1>;
+			#size-cells = <0>;
+			compatible = "fsl,mpc8536-espi";
+			reg = <0x7000 0x1000>;
+			interrupts = <59 0x2>;
+			interrupt-parent = <&mpic>;
+			fsl,espi-num-chipselects = <4>;
+
+			flash@0 {
+				#address-cells = <1>;
+				#size-cells = <1>;
+				compatible = "spansion,s25sl12801";
+				reg = <0>;
+				spi-max-frequency = <40000000>;
+				partition@u-boot {
+					label = "u-boot";
+					reg = <0x00000000 0x00100000>;
+					read-only;
+				};
+				partition@kernel {
+					label = "kernel";
+					reg = <0x00100000 0x00500000>;
+					read-only;
+				};
+				partition@dtb {
+					label = "dtb";
+					reg = <0x00600000 0x00100000>;
+					read-only;
+				};
+				partition@fs {
+					label = "file system";
+					reg = <0x00700000 0x00900000>;
+				};
+			};
+			flash@1 {
+				compatible = "spansion,s25sl12801";
+				reg = <1>;
+				spi-max-frequency = <40000000>;
+			};
+			flash@2 {
+				compatible = "spansion,s25sl12801";
+				reg = <2>;
+				spi-max-frequency = <40000000>;
+			};
+			flash@3 {
+				compatible = "spansion,s25sl12801";
+				reg = <3>;
+				spi-max-frequency = <40000000>;
+			};
+		};
+
 		dma@21300 {
 			#address-cells = <1>;
 			#size-cells = <1>;
diff --git a/arch/powerpc/boot/dts/p4080ds.dts b/arch/powerpc/boot/dts/p4080ds.dts
index 2f0de24..5b7fc29 100644
--- a/arch/powerpc/boot/dts/p4080ds.dts
+++ b/arch/powerpc/boot/dts/p4080ds.dts
@@ -236,22 +236,19 @@
 		};
 
 		spi@110000 {
-			cell-index = <0>;
 			#address-cells = <1>;
 			#size-cells = <0>;
-			compatible = "fsl,espi";
+			compatible = "fsl,p4080-espi", "fsl,mpc8536-espi";
 			reg = <0x110000 0x1000>;
 			interrupts = <53 0x2>;
 			interrupt-parent = <&mpic>;
-			espi,num-ss-bits = <4>;
-			mode = "cpu";
+			fsl,espi-num-chipselects = <4>;
 
-			fsl_m25p80@0 {
+			flash@0 {
 				#address-cells = <1>;
 				#size-cells = <1>;
-				compatible = "fsl,espi-flash";
+				compatible = "spansion,s25sl12801";
 				reg = <0>;
-				linux,modalias = "fsl_m25p80";
 				spi-max-frequency = <40000000>; /* input clock */
 				partition@u-boot {
 					label = "u-boot";
diff --git a/drivers/mfd/ab8500-spi.c b/drivers/mfd/ab8500-spi.c
index e1c8b62..01b6d58 100644
--- a/drivers/mfd/ab8500-spi.c
+++ b/drivers/mfd/ab8500-spi.c
@@ -83,6 +83,11 @@
 	struct ab8500 *ab8500;
 	int ret;
 
+	spi->bits_per_word = 24;
+	ret = spi_setup(spi);
+	if (ret < 0)
+		return ret;
+
 	ab8500 = kzalloc(sizeof *ab8500, GFP_KERNEL);
 	if (!ab8500)
 		return -ENOMEM;
diff --git a/drivers/spi/Kconfig b/drivers/spi/Kconfig
index 91c2f4f..9949c25 100644
--- a/drivers/spi/Kconfig
+++ b/drivers/spi/Kconfig
@@ -182,12 +182,27 @@
 	  This enables using the Freescale MPC5121 Programmable Serial
 	  Controller in SPI master mode.
 
-config SPI_MPC8xxx
-	tristate "Freescale MPC8xxx SPI controller"
+config SPI_FSL_LIB
+	tristate
 	depends on FSL_SOC
+
+config SPI_FSL_SPI
+	tristate "Freescale SPI controller"
+	depends on FSL_SOC
+	select SPI_FSL_LIB
 	help
-	  This enables using the Freescale MPC8xxx SPI controllers in master
-	  mode.
+	  This enables using the Freescale SPI controllers in master mode.
+	  MPC83xx platform uses the controller in cpu mode or CPM/QE mode.
+	  MPC8569 uses the controller in QE mode, MPC8610 in cpu mode.
+
+config SPI_FSL_ESPI
+	tristate "Freescale eSPI controller"
+	depends on FSL_SOC
+	select SPI_FSL_LIB
+	help
+	  This enables using the Freescale eSPI controllers in master mode.
+	  From MPC8536, 85xx platform uses the controller, and all P10xx,
+	  P20xx, P30xx,P40xx, P50xx uses this controller.
 
 config SPI_OMAP_UWIRE
 	tristate "OMAP1 MicroWire"
@@ -298,6 +313,13 @@
 	help
 	  SPI driver for Freescale STMP37xx/378x SoC SSP interface
 
+config SPI_TOPCLIFF_PCH
+	tristate "Topcliff PCH SPI Controller"
+	depends on PCI
+	help
+	  SPI driver for the Topcliff PCH (Platform Controller Hub) SPI bus
+	  used in some x86 embedded processors.
+
 config SPI_TXX9
 	tristate "Toshiba TXx9 SPI controller"
 	depends on GENERIC_GPIO && CPU_TX49XX
diff --git a/drivers/spi/Makefile b/drivers/spi/Makefile
index e9cbd18..557aaad 100644
--- a/drivers/spi/Makefile
+++ b/drivers/spi/Makefile
@@ -2,9 +2,7 @@
 # Makefile for kernel SPI drivers.
 #
 
-ifeq ($(CONFIG_SPI_DEBUG),y)
-EXTRA_CFLAGS += -DDEBUG
-endif
+ccflags-$(CONFIG_SPI_DEBUG) := -DDEBUG
 
 # small core, mostly translating board-specific
 # config declarations into driver model code
@@ -34,11 +32,14 @@
 obj-$(CONFIG_SPI_MPC512x_PSC)		+= mpc512x_psc_spi.o
 obj-$(CONFIG_SPI_MPC52xx_PSC)		+= mpc52xx_psc_spi.o
 obj-$(CONFIG_SPI_MPC52xx)		+= mpc52xx_spi.o
-obj-$(CONFIG_SPI_MPC8xxx)		+= spi_mpc8xxx.o
+obj-$(CONFIG_SPI_FSL_LIB)		+= spi_fsl_lib.o
+obj-$(CONFIG_SPI_FSL_ESPI)		+= spi_fsl_espi.o
+obj-$(CONFIG_SPI_FSL_SPI)		+= spi_fsl_spi.o
 obj-$(CONFIG_SPI_PPC4xx)		+= spi_ppc4xx.o
 obj-$(CONFIG_SPI_S3C24XX_GPIO)		+= spi_s3c24xx_gpio.o
 obj-$(CONFIG_SPI_S3C24XX)		+= spi_s3c24xx_hw.o
 obj-$(CONFIG_SPI_S3C64XX)		+= spi_s3c64xx.o
+obj-$(CONFIG_SPI_TOPCLIFF_PCH)		+= spi_topcliff_pch.o
 obj-$(CONFIG_SPI_TXX9)			+= spi_txx9.o
 obj-$(CONFIG_SPI_XILINX)		+= xilinx_spi.o
 obj-$(CONFIG_SPI_XILINX_OF)		+= xilinx_spi_of.o
diff --git a/drivers/spi/amba-pl022.c b/drivers/spi/amba-pl022.c
index 4c37c4e2..fb3d1b3 100644
--- a/drivers/spi/amba-pl022.c
+++ b/drivers/spi/amba-pl022.c
@@ -27,7 +27,6 @@
 /*
  * TODO:
  * - add timeout on polled transfers
- * - add generic DMA framework support
  */
 
 #include <linux/init.h>
@@ -45,6 +44,9 @@
 #include <linux/amba/pl022.h>
 #include <linux/io.h>
 #include <linux/slab.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/scatterlist.h>
 
 /*
  * This macro is used to define some register default values.
@@ -381,6 +383,14 @@
 	enum ssp_reading		read;
 	enum ssp_writing		write;
 	u32				exp_fifo_level;
+	/* DMA settings */
+#ifdef CONFIG_DMA_ENGINE
+	struct dma_chan			*dma_rx_channel;
+	struct dma_chan			*dma_tx_channel;
+	struct sg_table			sgt_rx;
+	struct sg_table			sgt_tx;
+	char				*dummypage;
+#endif
 };
 
 /**
@@ -406,7 +416,7 @@
 	u16 dmacr;
 	u16 cpsr;
 	u8 n_bytes;
-	u8 enable_dma:1;
+	bool enable_dma;
 	enum ssp_reading read;
 	enum ssp_writing write;
 	void (*cs_control) (u32 command);
@@ -763,6 +773,371 @@
 	}
 	return STATE_DONE;
 }
+
+/*
+ * This DMA functionality is only compiled in if we have
+ * access to the generic DMA devices/DMA engine.
+ */
+#ifdef CONFIG_DMA_ENGINE
+static void unmap_free_dma_scatter(struct pl022 *pl022)
+{
+	/* Unmap and free the SG tables */
+	dma_unmap_sg(&pl022->adev->dev, pl022->sgt_tx.sgl,
+		     pl022->sgt_tx.nents, DMA_TO_DEVICE);
+	dma_unmap_sg(&pl022->adev->dev, pl022->sgt_rx.sgl,
+		     pl022->sgt_rx.nents, DMA_FROM_DEVICE);
+	sg_free_table(&pl022->sgt_rx);
+	sg_free_table(&pl022->sgt_tx);
+}
+
+static void dma_callback(void *data)
+{
+	struct pl022 *pl022 = data;
+	struct spi_message *msg = pl022->cur_msg;
+
+	BUG_ON(!pl022->sgt_rx.sgl);
+
+#ifdef VERBOSE_DEBUG
+	/*
+	 * Optionally dump out buffers to inspect contents, this is
+	 * good if you want to convince yourself that the loopback
+	 * read/write contents are the same, when adopting to a new
+	 * DMA engine.
+	 */
+	{
+		struct scatterlist *sg;
+		unsigned int i;
+
+		dma_sync_sg_for_cpu(&pl022->adev->dev,
+				    pl022->sgt_rx.sgl,
+				    pl022->sgt_rx.nents,
+				    DMA_FROM_DEVICE);
+
+		for_each_sg(pl022->sgt_rx.sgl, sg, pl022->sgt_rx.nents, i) {
+			dev_dbg(&pl022->adev->dev, "SPI RX SG ENTRY: %d", i);
+			print_hex_dump(KERN_ERR, "SPI RX: ",
+				       DUMP_PREFIX_OFFSET,
+				       16,
+				       1,
+				       sg_virt(sg),
+				       sg_dma_len(sg),
+				       1);
+		}
+		for_each_sg(pl022->sgt_tx.sgl, sg, pl022->sgt_tx.nents, i) {
+			dev_dbg(&pl022->adev->dev, "SPI TX SG ENTRY: %d", i);
+			print_hex_dump(KERN_ERR, "SPI TX: ",
+				       DUMP_PREFIX_OFFSET,
+				       16,
+				       1,
+				       sg_virt(sg),
+				       sg_dma_len(sg),
+				       1);
+		}
+	}
+#endif
+
+	unmap_free_dma_scatter(pl022);
+
+	/* Update total bytes transfered */
+	msg->actual_length += pl022->cur_transfer->len;
+	if (pl022->cur_transfer->cs_change)
+		pl022->cur_chip->
+			cs_control(SSP_CHIP_DESELECT);
+
+	/* Move to next transfer */
+	msg->state = next_transfer(pl022);
+	tasklet_schedule(&pl022->pump_transfers);
+}
+
+static void setup_dma_scatter(struct pl022 *pl022,
+			      void *buffer,
+			      unsigned int length,
+			      struct sg_table *sgtab)
+{
+	struct scatterlist *sg;
+	int bytesleft = length;
+	void *bufp = buffer;
+	int mapbytes;
+	int i;
+
+	if (buffer) {
+		for_each_sg(sgtab->sgl, sg, sgtab->nents, i) {
+			/*
+			 * If there are less bytes left than what fits
+			 * in the current page (plus page alignment offset)
+			 * we just feed in this, else we stuff in as much
+			 * as we can.
+			 */
+			if (bytesleft < (PAGE_SIZE - offset_in_page(bufp)))
+				mapbytes = bytesleft;
+			else
+				mapbytes = PAGE_SIZE - offset_in_page(bufp);
+			sg_set_page(sg, virt_to_page(bufp),
+				    mapbytes, offset_in_page(bufp));
+			bufp += mapbytes;
+			bytesleft -= mapbytes;
+			dev_dbg(&pl022->adev->dev,
+				"set RX/TX target page @ %p, %d bytes, %d left\n",
+				bufp, mapbytes, bytesleft);
+		}
+	} else {
+		/* Map the dummy buffer on every page */
+		for_each_sg(sgtab->sgl, sg, sgtab->nents, i) {
+			if (bytesleft < PAGE_SIZE)
+				mapbytes = bytesleft;
+			else
+				mapbytes = PAGE_SIZE;
+			sg_set_page(sg, virt_to_page(pl022->dummypage),
+				    mapbytes, 0);
+			bytesleft -= mapbytes;
+			dev_dbg(&pl022->adev->dev,
+				"set RX/TX to dummy page %d bytes, %d left\n",
+				mapbytes, bytesleft);
+
+		}
+	}
+	BUG_ON(bytesleft);
+}
+
+/**
+ * configure_dma - configures the channels for the next transfer
+ * @pl022: SSP driver's private data structure
+ */
+static int configure_dma(struct pl022 *pl022)
+{
+	struct dma_slave_config rx_conf = {
+		.src_addr = SSP_DR(pl022->phybase),
+		.direction = DMA_FROM_DEVICE,
+		.src_maxburst = pl022->vendor->fifodepth >> 1,
+	};
+	struct dma_slave_config tx_conf = {
+		.dst_addr = SSP_DR(pl022->phybase),
+		.direction = DMA_TO_DEVICE,
+		.dst_maxburst = pl022->vendor->fifodepth >> 1,
+	};
+	unsigned int pages;
+	int ret;
+	int sglen;
+	struct dma_chan *rxchan = pl022->dma_rx_channel;
+	struct dma_chan *txchan = pl022->dma_tx_channel;
+	struct dma_async_tx_descriptor *rxdesc;
+	struct dma_async_tx_descriptor *txdesc;
+	dma_cookie_t cookie;
+
+	/* Check that the channels are available */
+	if (!rxchan || !txchan)
+		return -ENODEV;
+
+	switch (pl022->read) {
+	case READING_NULL:
+		/* Use the same as for writing */
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_UNDEFINED;
+		break;
+	case READING_U8:
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		break;
+	case READING_U16:
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		break;
+	case READING_U32:
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		break;
+	}
+
+	switch (pl022->write) {
+	case WRITING_NULL:
+		/* Use the same as for reading */
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_UNDEFINED;
+		break;
+	case WRITING_U8:
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		break;
+	case WRITING_U16:
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		break;
+	case WRITING_U32:
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;;
+		break;
+	}
+
+	/* SPI pecularity: we need to read and write the same width */
+	if (rx_conf.src_addr_width == DMA_SLAVE_BUSWIDTH_UNDEFINED)
+		rx_conf.src_addr_width = tx_conf.dst_addr_width;
+	if (tx_conf.dst_addr_width == DMA_SLAVE_BUSWIDTH_UNDEFINED)
+		tx_conf.dst_addr_width = rx_conf.src_addr_width;
+	BUG_ON(rx_conf.src_addr_width != tx_conf.dst_addr_width);
+
+	rxchan->device->device_control(rxchan, DMA_SLAVE_CONFIG,
+				       (unsigned long) &rx_conf);
+	txchan->device->device_control(txchan, DMA_SLAVE_CONFIG,
+				       (unsigned long) &tx_conf);
+
+	/* Create sglists for the transfers */
+	pages = (pl022->cur_transfer->len >> PAGE_SHIFT) + 1;
+	dev_dbg(&pl022->adev->dev, "using %d pages for transfer\n", pages);
+
+	ret = sg_alloc_table(&pl022->sgt_rx, pages, GFP_KERNEL);
+	if (ret)
+		goto err_alloc_rx_sg;
+
+	ret = sg_alloc_table(&pl022->sgt_tx, pages, GFP_KERNEL);
+	if (ret)
+		goto err_alloc_tx_sg;
+
+	/* Fill in the scatterlists for the RX+TX buffers */
+	setup_dma_scatter(pl022, pl022->rx,
+			  pl022->cur_transfer->len, &pl022->sgt_rx);
+	setup_dma_scatter(pl022, pl022->tx,
+			  pl022->cur_transfer->len, &pl022->sgt_tx);
+
+	/* Map DMA buffers */
+	sglen = dma_map_sg(&pl022->adev->dev, pl022->sgt_rx.sgl,
+			   pl022->sgt_rx.nents, DMA_FROM_DEVICE);
+	if (!sglen)
+		goto err_rx_sgmap;
+
+	sglen = dma_map_sg(&pl022->adev->dev, pl022->sgt_tx.sgl,
+			   pl022->sgt_tx.nents, DMA_TO_DEVICE);
+	if (!sglen)
+		goto err_tx_sgmap;
+
+	/* Send both scatterlists */
+	rxdesc = rxchan->device->device_prep_slave_sg(rxchan,
+				      pl022->sgt_rx.sgl,
+				      pl022->sgt_rx.nents,
+				      DMA_FROM_DEVICE,
+				      DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!rxdesc)
+		goto err_rxdesc;
+
+	txdesc = txchan->device->device_prep_slave_sg(txchan,
+				      pl022->sgt_tx.sgl,
+				      pl022->sgt_tx.nents,
+				      DMA_TO_DEVICE,
+				      DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!txdesc)
+		goto err_txdesc;
+
+	/* Put the callback on the RX transfer only, that should finish last */
+	rxdesc->callback = dma_callback;
+	rxdesc->callback_param = pl022;
+
+	/* Submit and fire RX and TX with TX last so we're ready to read! */
+	cookie = rxdesc->tx_submit(rxdesc);
+	if (dma_submit_error(cookie))
+		goto err_submit_rx;
+	cookie = txdesc->tx_submit(txdesc);
+	if (dma_submit_error(cookie))
+		goto err_submit_tx;
+	rxchan->device->device_issue_pending(rxchan);
+	txchan->device->device_issue_pending(txchan);
+
+	return 0;
+
+err_submit_tx:
+err_submit_rx:
+err_txdesc:
+	txchan->device->device_control(txchan, DMA_TERMINATE_ALL, 0);
+err_rxdesc:
+	rxchan->device->device_control(rxchan, DMA_TERMINATE_ALL, 0);
+	dma_unmap_sg(&pl022->adev->dev, pl022->sgt_tx.sgl,
+		     pl022->sgt_tx.nents, DMA_TO_DEVICE);
+err_tx_sgmap:
+	dma_unmap_sg(&pl022->adev->dev, pl022->sgt_rx.sgl,
+		     pl022->sgt_tx.nents, DMA_FROM_DEVICE);
+err_rx_sgmap:
+	sg_free_table(&pl022->sgt_tx);
+err_alloc_tx_sg:
+	sg_free_table(&pl022->sgt_rx);
+err_alloc_rx_sg:
+	return -ENOMEM;
+}
+
+static int __init pl022_dma_probe(struct pl022 *pl022)
+{
+	dma_cap_mask_t mask;
+
+	/* Try to acquire a generic DMA engine slave channel */
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+	/*
+	 * We need both RX and TX channels to do DMA, else do none
+	 * of them.
+	 */
+	pl022->dma_rx_channel = dma_request_channel(mask,
+					    pl022->master_info->dma_filter,
+					    pl022->master_info->dma_rx_param);
+	if (!pl022->dma_rx_channel) {
+		dev_err(&pl022->adev->dev, "no RX DMA channel!\n");
+		goto err_no_rxchan;
+	}
+
+	pl022->dma_tx_channel = dma_request_channel(mask,
+					    pl022->master_info->dma_filter,
+					    pl022->master_info->dma_tx_param);
+	if (!pl022->dma_tx_channel) {
+		dev_err(&pl022->adev->dev, "no TX DMA channel!\n");
+		goto err_no_txchan;
+	}
+
+	pl022->dummypage = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!pl022->dummypage) {
+		dev_err(&pl022->adev->dev, "no DMA dummypage!\n");
+		goto err_no_dummypage;
+	}
+
+	dev_info(&pl022->adev->dev, "setup for DMA on RX %s, TX %s\n",
+		 dma_chan_name(pl022->dma_rx_channel),
+		 dma_chan_name(pl022->dma_tx_channel));
+
+	return 0;
+
+err_no_dummypage:
+	dma_release_channel(pl022->dma_tx_channel);
+err_no_txchan:
+	dma_release_channel(pl022->dma_rx_channel);
+	pl022->dma_rx_channel = NULL;
+err_no_rxchan:
+	return -ENODEV;
+}
+
+static void terminate_dma(struct pl022 *pl022)
+{
+	struct dma_chan *rxchan = pl022->dma_rx_channel;
+	struct dma_chan *txchan = pl022->dma_tx_channel;
+
+	rxchan->device->device_control(rxchan, DMA_TERMINATE_ALL, 0);
+	txchan->device->device_control(txchan, DMA_TERMINATE_ALL, 0);
+	unmap_free_dma_scatter(pl022);
+}
+
+static void pl022_dma_remove(struct pl022 *pl022)
+{
+	if (pl022->busy)
+		terminate_dma(pl022);
+	if (pl022->dma_tx_channel)
+		dma_release_channel(pl022->dma_tx_channel);
+	if (pl022->dma_rx_channel)
+		dma_release_channel(pl022->dma_rx_channel);
+	kfree(pl022->dummypage);
+}
+
+#else
+static inline int configure_dma(struct pl022 *pl022)
+{
+	return -ENODEV;
+}
+
+static inline int pl022_dma_probe(struct pl022 *pl022)
+{
+	return 0;
+}
+
+static inline void pl022_dma_remove(struct pl022 *pl022)
+{
+}
+#endif
+
 /**
  * pl022_interrupt_handler - Interrupt handler for SSP controller
  *
@@ -794,14 +1169,17 @@
 	if (unlikely(!irq_status))
 		return IRQ_NONE;
 
-	/* This handles the error code interrupts */
+	/*
+	 * This handles the FIFO interrupts, the timeout
+	 * interrupts are flatly ignored, they cannot be
+	 * trusted.
+	 */
 	if (unlikely(irq_status & SSP_MIS_MASK_RORMIS)) {
 		/*
 		 * Overrun interrupt - bail out since our Data has been
 		 * corrupted
 		 */
-		dev_err(&pl022->adev->dev,
-			"FIFO overrun\n");
+		dev_err(&pl022->adev->dev, "FIFO overrun\n");
 		if (readw(SSP_SR(pl022->virtbase)) & SSP_SR_MASK_RFF)
 			dev_err(&pl022->adev->dev,
 				"RXFIFO is full\n");
@@ -896,8 +1274,8 @@
 }
 
 /**
- * pump_transfers - Tasklet function which schedules next interrupt transfer
- * when running in interrupt transfer mode.
+ * pump_transfers - Tasklet function which schedules next transfer
+ * when running in interrupt or DMA transfer mode.
  * @data: SSP driver private data structure
  *
  */
@@ -954,65 +1332,23 @@
 	}
 	/* Flush the FIFOs and let's go! */
 	flush(pl022);
+
+	if (pl022->cur_chip->enable_dma) {
+		if (configure_dma(pl022)) {
+			dev_dbg(&pl022->adev->dev,
+				"configuration of DMA failed, fall back to interrupt mode\n");
+			goto err_config_dma;
+		}
+		return;
+	}
+
+err_config_dma:
 	writew(ENABLE_ALL_INTERRUPTS, SSP_IMSC(pl022->virtbase));
 }
 
-/**
- * NOT IMPLEMENTED
- * configure_dma - It configures the DMA pipes for DMA transfers
- * @data: SSP driver's private data structure
- *
- */
-static int configure_dma(void *data)
+static void do_interrupt_dma_transfer(struct pl022 *pl022)
 {
-	struct pl022 *pl022 = data;
-	dev_dbg(&pl022->adev->dev, "configure DMA\n");
-	return -ENOTSUPP;
-}
-
-/**
- * do_dma_transfer - It handles transfers of the current message
- * if it is DMA xfer.
- * NOT FULLY IMPLEMENTED
- * @data: SSP driver's private data structure
- */
-static void do_dma_transfer(void *data)
-{
-	struct pl022 *pl022 = data;
-
-	if (configure_dma(data)) {
-		dev_dbg(&pl022->adev->dev, "configuration of DMA Failed!\n");
-		goto err_config_dma;
-	}
-
-	/* TODO: Implememt DMA setup of pipes here */
-
-	/* Enable target chip, set up transfer */
-	pl022->cur_chip->cs_control(SSP_CHIP_SELECT);
-	if (set_up_next_transfer(pl022, pl022->cur_transfer)) {
-		/* Error path */
-		pl022->cur_msg->state = STATE_ERROR;
-		pl022->cur_msg->status = -EIO;
-		giveback(pl022);
-		return;
-	}
-	/* Enable SSP */
-	writew((readw(SSP_CR1(pl022->virtbase)) | SSP_CR1_MASK_SSE),
-	       SSP_CR1(pl022->virtbase));
-
-	/* TODO: Enable the DMA transfer here */
-	return;
-
- err_config_dma:
-	pl022->cur_msg->state = STATE_ERROR;
-	pl022->cur_msg->status = -EIO;
-	giveback(pl022);
-	return;
-}
-
-static void do_interrupt_transfer(void *data)
-{
-	struct pl022 *pl022 = data;
+	u32 irqflags = ENABLE_ALL_INTERRUPTS;
 
 	/* Enable target chip */
 	pl022->cur_chip->cs_control(SSP_CHIP_SELECT);
@@ -1023,15 +1359,26 @@
 		giveback(pl022);
 		return;
 	}
+	/* If we're using DMA, set up DMA here */
+	if (pl022->cur_chip->enable_dma) {
+		/* Configure DMA transfer */
+		if (configure_dma(pl022)) {
+			dev_dbg(&pl022->adev->dev,
+				"configuration of DMA failed, fall back to interrupt mode\n");
+			goto err_config_dma;
+		}
+		/* Disable interrupts in DMA mode, IRQ from DMA controller */
+		irqflags = DISABLE_ALL_INTERRUPTS;
+	}
+err_config_dma:
 	/* Enable SSP, turn on interrupts */
 	writew((readw(SSP_CR1(pl022->virtbase)) | SSP_CR1_MASK_SSE),
 	       SSP_CR1(pl022->virtbase));
-	writew(ENABLE_ALL_INTERRUPTS, SSP_IMSC(pl022->virtbase));
+	writew(irqflags, SSP_IMSC(pl022->virtbase));
 }
 
-static void do_polling_transfer(void *data)
+static void do_polling_transfer(struct pl022 *pl022)
 {
-	struct pl022 *pl022 = data;
 	struct spi_message *message = NULL;
 	struct spi_transfer *transfer = NULL;
 	struct spi_transfer *previous = NULL;
@@ -1101,7 +1448,7 @@
  *
  * This function checks if there is any spi message in the queue that
  * needs processing and delegate control to appropriate function
- * do_polling_transfer()/do_interrupt_transfer()/do_dma_transfer()
+ * do_polling_transfer()/do_interrupt_dma_transfer()
  * based on the kind of the transfer
  *
  */
@@ -1150,10 +1497,8 @@
 
 	if (pl022->cur_chip->xfer_type == POLLING_TRANSFER)
 		do_polling_transfer(pl022);
-	else if (pl022->cur_chip->xfer_type == INTERRUPT_TRANSFER)
-		do_interrupt_transfer(pl022);
 	else
-		do_dma_transfer(pl022);
+		do_interrupt_dma_transfer(pl022);
 }
 
 
@@ -1248,100 +1593,56 @@
 }
 
 static int verify_controller_parameters(struct pl022 *pl022,
-					struct pl022_config_chip *chip_info)
+				struct pl022_config_chip const *chip_info)
 {
-	if ((chip_info->lbm != LOOPBACK_ENABLED)
-	    && (chip_info->lbm != LOOPBACK_DISABLED)) {
-		dev_err(chip_info->dev,
-			"loopback Mode is configured incorrectly\n");
-		return -EINVAL;
-	}
 	if ((chip_info->iface < SSP_INTERFACE_MOTOROLA_SPI)
 	    || (chip_info->iface > SSP_INTERFACE_UNIDIRECTIONAL)) {
-		dev_err(chip_info->dev,
+		dev_err(&pl022->adev->dev,
 			"interface is configured incorrectly\n");
 		return -EINVAL;
 	}
 	if ((chip_info->iface == SSP_INTERFACE_UNIDIRECTIONAL) &&
 	    (!pl022->vendor->unidir)) {
-		dev_err(chip_info->dev,
+		dev_err(&pl022->adev->dev,
 			"unidirectional mode not supported in this "
 			"hardware version\n");
 		return -EINVAL;
 	}
 	if ((chip_info->hierarchy != SSP_MASTER)
 	    && (chip_info->hierarchy != SSP_SLAVE)) {
-		dev_err(chip_info->dev,
+		dev_err(&pl022->adev->dev,
 			"hierarchy is configured incorrectly\n");
 		return -EINVAL;
 	}
-	if (((chip_info->clk_freq).cpsdvsr < CPSDVR_MIN)
-	    || ((chip_info->clk_freq).cpsdvsr > CPSDVR_MAX)) {
-		dev_err(chip_info->dev,
-			"cpsdvsr is configured incorrectly\n");
-		return -EINVAL;
-	}
-	if ((chip_info->endian_rx != SSP_RX_MSB)
-	    && (chip_info->endian_rx != SSP_RX_LSB)) {
-		dev_err(chip_info->dev,
-			"RX FIFO endianess is configured incorrectly\n");
-		return -EINVAL;
-	}
-	if ((chip_info->endian_tx != SSP_TX_MSB)
-	    && (chip_info->endian_tx != SSP_TX_LSB)) {
-		dev_err(chip_info->dev,
-			"TX FIFO endianess is configured incorrectly\n");
-		return -EINVAL;
-	}
-	if ((chip_info->data_size < SSP_DATA_BITS_4)
-	    || (chip_info->data_size > SSP_DATA_BITS_32)) {
-		dev_err(chip_info->dev,
-			"DATA Size is configured incorrectly\n");
-		return -EINVAL;
-	}
 	if ((chip_info->com_mode != INTERRUPT_TRANSFER)
 	    && (chip_info->com_mode != DMA_TRANSFER)
 	    && (chip_info->com_mode != POLLING_TRANSFER)) {
-		dev_err(chip_info->dev,
+		dev_err(&pl022->adev->dev,
 			"Communication mode is configured incorrectly\n");
 		return -EINVAL;
 	}
 	if ((chip_info->rx_lev_trig < SSP_RX_1_OR_MORE_ELEM)
 	    || (chip_info->rx_lev_trig > SSP_RX_32_OR_MORE_ELEM)) {
-		dev_err(chip_info->dev,
+		dev_err(&pl022->adev->dev,
 			"RX FIFO Trigger Level is configured incorrectly\n");
 		return -EINVAL;
 	}
 	if ((chip_info->tx_lev_trig < SSP_TX_1_OR_MORE_EMPTY_LOC)
 	    || (chip_info->tx_lev_trig > SSP_TX_32_OR_MORE_EMPTY_LOC)) {
-		dev_err(chip_info->dev,
+		dev_err(&pl022->adev->dev,
 			"TX FIFO Trigger Level is configured incorrectly\n");
 		return -EINVAL;
 	}
-	if (chip_info->iface == SSP_INTERFACE_MOTOROLA_SPI) {
-		if ((chip_info->clk_phase != SSP_CLK_FIRST_EDGE)
-		    && (chip_info->clk_phase != SSP_CLK_SECOND_EDGE)) {
-			dev_err(chip_info->dev,
-				"Clock Phase is configured incorrectly\n");
-			return -EINVAL;
-		}
-		if ((chip_info->clk_pol != SSP_CLK_POL_IDLE_LOW)
-		    && (chip_info->clk_pol != SSP_CLK_POL_IDLE_HIGH)) {
-			dev_err(chip_info->dev,
-				"Clock Polarity is configured incorrectly\n");
-			return -EINVAL;
-		}
-	}
 	if (chip_info->iface == SSP_INTERFACE_NATIONAL_MICROWIRE) {
 		if ((chip_info->ctrl_len < SSP_BITS_4)
 		    || (chip_info->ctrl_len > SSP_BITS_32)) {
-			dev_err(chip_info->dev,
+			dev_err(&pl022->adev->dev,
 				"CTRL LEN is configured incorrectly\n");
 			return -EINVAL;
 		}
 		if ((chip_info->wait_state != SSP_MWIRE_WAIT_ZERO)
 		    && (chip_info->wait_state != SSP_MWIRE_WAIT_ONE)) {
-			dev_err(chip_info->dev,
+			dev_err(&pl022->adev->dev,
 				"Wait State is configured incorrectly\n");
 			return -EINVAL;
 		}
@@ -1350,24 +1651,20 @@
 			if ((chip_info->duplex !=
 			     SSP_MICROWIRE_CHANNEL_FULL_DUPLEX)
 			    && (chip_info->duplex !=
-				SSP_MICROWIRE_CHANNEL_HALF_DUPLEX))
-				dev_err(chip_info->dev,
+				SSP_MICROWIRE_CHANNEL_HALF_DUPLEX)) {
+				dev_err(&pl022->adev->dev,
 					"Microwire duplex mode is configured incorrectly\n");
 				return -EINVAL;
+			}
 		} else {
 			if (chip_info->duplex != SSP_MICROWIRE_CHANNEL_FULL_DUPLEX)
-				dev_err(chip_info->dev,
+				dev_err(&pl022->adev->dev,
 					"Microwire half duplex mode requested,"
 					" but this is only available in the"
 					" ST version of PL022\n");
 			return -EINVAL;
 		}
 	}
-	if (chip_info->cs_control == NULL) {
-		dev_warn(chip_info->dev,
-			"Chip Select Function is NULL for this chip\n");
-		chip_info->cs_control = null_cs_control;
-	}
 	return 0;
 }
 
@@ -1467,22 +1764,24 @@
 	return 0;
 }
 
-/**
- * NOT IMPLEMENTED
- * process_dma_info - Processes the DMA info provided by client drivers
- * @chip_info: chip info provided by client device
- * @chip: Runtime state maintained by the SSP controller for each spi device
- *
- * This function processes and stores DMA config provided by client driver
- * into the runtime state maintained by the SSP controller driver
+
+/*
+ * A piece of default chip info unless the platform
+ * supplies it.
  */
-static int process_dma_info(struct pl022_config_chip *chip_info,
-			    struct chip_data *chip)
-{
-	dev_err(chip_info->dev,
-		"cannot process DMA info, DMA not implemented!\n");
-	return -ENOTSUPP;
-}
+static const struct pl022_config_chip pl022_default_chip_info = {
+	.com_mode = POLLING_TRANSFER,
+	.iface = SSP_INTERFACE_MOTOROLA_SPI,
+	.hierarchy = SSP_SLAVE,
+	.slave_tx_disable = DO_NOT_DRIVE_TX,
+	.rx_lev_trig = SSP_RX_1_OR_MORE_ELEM,
+	.tx_lev_trig = SSP_TX_1_OR_MORE_EMPTY_LOC,
+	.ctrl_len = SSP_BITS_8,
+	.wait_state = SSP_MWIRE_WAIT_ZERO,
+	.duplex = SSP_MICROWIRE_CHANNEL_FULL_DUPLEX,
+	.cs_control = null_cs_control,
+};
+
 
 /**
  * pl022_setup - setup function registered to SPI master framework
@@ -1496,23 +1795,15 @@
  * controller hardware here, that is not done until the actual transfer
  * commence.
  */
-
-/* FIXME: JUST GUESSING the spi->mode bits understood by this driver */
-#define MODEBITS	(SPI_CPOL | SPI_CPHA | SPI_CS_HIGH \
-			| SPI_LSB_FIRST | SPI_LOOP)
-
 static int pl022_setup(struct spi_device *spi)
 {
-	struct pl022_config_chip *chip_info;
+	struct pl022_config_chip const *chip_info;
 	struct chip_data *chip;
+	struct ssp_clock_params clk_freq;
 	int status = 0;
 	struct pl022 *pl022 = spi_master_get_devdata(spi->master);
-
-	if (spi->mode & ~MODEBITS) {
-		dev_dbg(&spi->dev, "unsupported mode bits %x\n",
-			spi->mode & ~MODEBITS);
-		return -EINVAL;
-	}
+	unsigned int bits = spi->bits_per_word;
+	u32 tmp;
 
 	if (!spi->max_speed_hz)
 		return -EINVAL;
@@ -1535,48 +1826,13 @@
 	chip_info = spi->controller_data;
 
 	if (chip_info == NULL) {
+		chip_info = &pl022_default_chip_info;
 		/* spi_board_info.controller_data not is supplied */
 		dev_dbg(&spi->dev,
 			"using default controller_data settings\n");
-
-		chip_info =
-			kzalloc(sizeof(struct pl022_config_chip), GFP_KERNEL);
-
-		if (!chip_info) {
-			dev_err(&spi->dev,
-				"cannot allocate controller data\n");
-			status = -ENOMEM;
-			goto err_first_setup;
-		}
-
-		dev_dbg(&spi->dev, "allocated memory for controller data\n");
-
-		/* Pointer back to the SPI device */
-		chip_info->dev = &spi->dev;
-		/*
-		 * Set controller data default values:
-		 * Polling is supported by default
-		 */
-		chip_info->lbm = LOOPBACK_DISABLED;
-		chip_info->com_mode = POLLING_TRANSFER;
-		chip_info->iface = SSP_INTERFACE_MOTOROLA_SPI;
-		chip_info->hierarchy = SSP_SLAVE;
-		chip_info->slave_tx_disable = DO_NOT_DRIVE_TX;
-		chip_info->endian_tx = SSP_TX_LSB;
-		chip_info->endian_rx = SSP_RX_LSB;
-		chip_info->data_size = SSP_DATA_BITS_12;
-		chip_info->rx_lev_trig = SSP_RX_1_OR_MORE_ELEM;
-		chip_info->tx_lev_trig = SSP_TX_1_OR_MORE_EMPTY_LOC;
-		chip_info->clk_phase = SSP_CLK_SECOND_EDGE;
-		chip_info->clk_pol = SSP_CLK_POL_IDLE_LOW;
-		chip_info->ctrl_len = SSP_BITS_8;
-		chip_info->wait_state = SSP_MWIRE_WAIT_ZERO;
-		chip_info->duplex = SSP_MICROWIRE_CHANNEL_FULL_DUPLEX;
-		chip_info->cs_control = null_cs_control;
-	} else {
+	} else
 		dev_dbg(&spi->dev,
 			"using user supplied controller_data settings\n");
-	}
 
 	/*
 	 * We can override with custom divisors, else we use the board
@@ -1586,29 +1842,48 @@
 	    && (0 == chip_info->clk_freq.scr)) {
 		status = calculate_effective_freq(pl022,
 						  spi->max_speed_hz,
-						  &chip_info->clk_freq);
+						  &clk_freq);
 		if (status < 0)
 			goto err_config_params;
 	} else {
-		if ((chip_info->clk_freq.cpsdvsr % 2) != 0)
-			chip_info->clk_freq.cpsdvsr =
-				chip_info->clk_freq.cpsdvsr - 1;
+		memcpy(&clk_freq, &chip_info->clk_freq, sizeof(clk_freq));
+		if ((clk_freq.cpsdvsr % 2) != 0)
+			clk_freq.cpsdvsr =
+				clk_freq.cpsdvsr - 1;
 	}
+	if ((clk_freq.cpsdvsr < CPSDVR_MIN)
+	    || (clk_freq.cpsdvsr > CPSDVR_MAX)) {
+		dev_err(&spi->dev,
+			"cpsdvsr is configured incorrectly\n");
+		goto err_config_params;
+	}
+
+
 	status = verify_controller_parameters(pl022, chip_info);
 	if (status) {
 		dev_err(&spi->dev, "controller data is incorrect");
 		goto err_config_params;
 	}
+
 	/* Now set controller state based on controller data */
 	chip->xfer_type = chip_info->com_mode;
-	chip->cs_control = chip_info->cs_control;
+	if (!chip_info->cs_control) {
+		chip->cs_control = null_cs_control;
+		dev_warn(&spi->dev,
+			 "chip select function is NULL for this chip\n");
+	} else
+		chip->cs_control = chip_info->cs_control;
 
-	if (chip_info->data_size <= 8) {
-		dev_dbg(&spi->dev, "1 <= n <=8 bits per word\n");
+	if (bits <= 3) {
+		/* PL022 doesn't support less than 4-bits */
+		status = -ENOTSUPP;
+		goto err_config_params;
+	} else if (bits <= 8) {
+		dev_dbg(&spi->dev, "4 <= n <=8 bits per word\n");
 		chip->n_bytes = 1;
 		chip->read = READING_U8;
 		chip->write = WRITING_U8;
-	} else if (chip_info->data_size <= 16) {
+	} else if (bits <= 16) {
 		dev_dbg(&spi->dev, "9 <= n <= 16 bits per word\n");
 		chip->n_bytes = 2;
 		chip->read = READING_U16;
@@ -1625,6 +1900,7 @@
 			dev_err(&spi->dev,
 				"a standard pl022 can only handle "
 				"1 <= n <= 16 bit words\n");
+			status = -ENOTSUPP;
 			goto err_config_params;
 		}
 	}
@@ -1636,9 +1912,8 @@
 	chip->cpsr = 0;
 	if ((chip_info->com_mode == DMA_TRANSFER)
 	    && ((pl022->master_info)->enable_dma)) {
-		chip->enable_dma = 1;
+		chip->enable_dma = true;
 		dev_dbg(&spi->dev, "DMA mode set in controller state\n");
-		status = process_dma_info(chip_info, chip);
 		if (status < 0)
 			goto err_config_params;
 		SSP_WRITE_BITS(chip->dmacr, SSP_DMA_ENABLED,
@@ -1646,7 +1921,7 @@
 		SSP_WRITE_BITS(chip->dmacr, SSP_DMA_ENABLED,
 			       SSP_DMACR_MASK_TXDMAE, 1);
 	} else {
-		chip->enable_dma = 0;
+		chip->enable_dma = false;
 		dev_dbg(&spi->dev, "DMA mode NOT set in controller state\n");
 		SSP_WRITE_BITS(chip->dmacr, SSP_DMA_DISABLED,
 			       SSP_DMACR_MASK_RXDMAE, 0);
@@ -1654,10 +1929,12 @@
 			       SSP_DMACR_MASK_TXDMAE, 1);
 	}
 
-	chip->cpsr = chip_info->clk_freq.cpsdvsr;
+	chip->cpsr = clk_freq.cpsdvsr;
 
 	/* Special setup for the ST micro extended control registers */
 	if (pl022->vendor->extended_cr) {
+		u32 etx;
+
 		if (pl022->vendor->pl023) {
 			/* These bits are only in the PL023 */
 			SSP_WRITE_BITS(chip->cr1, chip_info->clkdelay,
@@ -1673,29 +1950,51 @@
 			SSP_WRITE_BITS(chip->cr1, chip_info->wait_state,
 				       SSP_CR1_MASK_MWAIT_ST, 6);
 		}
-		SSP_WRITE_BITS(chip->cr0, chip_info->data_size,
+		SSP_WRITE_BITS(chip->cr0, bits - 1,
 			       SSP_CR0_MASK_DSS_ST, 0);
-		SSP_WRITE_BITS(chip->cr1, chip_info->endian_rx,
-			       SSP_CR1_MASK_RENDN_ST, 4);
-		SSP_WRITE_BITS(chip->cr1, chip_info->endian_tx,
-			       SSP_CR1_MASK_TENDN_ST, 5);
+
+		if (spi->mode & SPI_LSB_FIRST) {
+			tmp = SSP_RX_LSB;
+			etx = SSP_TX_LSB;
+		} else {
+			tmp = SSP_RX_MSB;
+			etx = SSP_TX_MSB;
+		}
+		SSP_WRITE_BITS(chip->cr1, tmp, SSP_CR1_MASK_RENDN_ST, 4);
+		SSP_WRITE_BITS(chip->cr1, etx, SSP_CR1_MASK_TENDN_ST, 5);
 		SSP_WRITE_BITS(chip->cr1, chip_info->rx_lev_trig,
 			       SSP_CR1_MASK_RXIFLSEL_ST, 7);
 		SSP_WRITE_BITS(chip->cr1, chip_info->tx_lev_trig,
 			       SSP_CR1_MASK_TXIFLSEL_ST, 10);
 	} else {
-		SSP_WRITE_BITS(chip->cr0, chip_info->data_size,
+		SSP_WRITE_BITS(chip->cr0, bits - 1,
 			       SSP_CR0_MASK_DSS, 0);
 		SSP_WRITE_BITS(chip->cr0, chip_info->iface,
 			       SSP_CR0_MASK_FRF, 4);
 	}
+
 	/* Stuff that is common for all versions */
-	SSP_WRITE_BITS(chip->cr0, chip_info->clk_pol, SSP_CR0_MASK_SPO, 6);
-	SSP_WRITE_BITS(chip->cr0, chip_info->clk_phase, SSP_CR0_MASK_SPH, 7);
-	SSP_WRITE_BITS(chip->cr0, chip_info->clk_freq.scr, SSP_CR0_MASK_SCR, 8);
+	if (spi->mode & SPI_CPOL)
+		tmp = SSP_CLK_POL_IDLE_HIGH;
+	else
+		tmp = SSP_CLK_POL_IDLE_LOW;
+	SSP_WRITE_BITS(chip->cr0, tmp, SSP_CR0_MASK_SPO, 6);
+
+	if (spi->mode & SPI_CPHA)
+		tmp = SSP_CLK_SECOND_EDGE;
+	else
+		tmp = SSP_CLK_FIRST_EDGE;
+	SSP_WRITE_BITS(chip->cr0, tmp, SSP_CR0_MASK_SPH, 7);
+
+	SSP_WRITE_BITS(chip->cr0, clk_freq.scr, SSP_CR0_MASK_SCR, 8);
 	/* Loopback is available on all versions except PL023 */
-	if (!pl022->vendor->pl023)
-		SSP_WRITE_BITS(chip->cr1, chip_info->lbm, SSP_CR1_MASK_LBM, 0);
+	if (!pl022->vendor->pl023) {
+		if (spi->mode & SPI_LOOP)
+			tmp = LOOPBACK_ENABLED;
+		else
+			tmp = LOOPBACK_DISABLED;
+		SSP_WRITE_BITS(chip->cr1, tmp, SSP_CR1_MASK_LBM, 0);
+	}
 	SSP_WRITE_BITS(chip->cr1, SSP_DISABLED, SSP_CR1_MASK_SSE, 1);
 	SSP_WRITE_BITS(chip->cr1, chip_info->hierarchy, SSP_CR1_MASK_MS, 2);
 	SSP_WRITE_BITS(chip->cr1, chip_info->slave_tx_disable, SSP_CR1_MASK_SOD, 3);
@@ -1704,7 +2003,7 @@
 	spi_set_ctldata(spi, chip);
 	return status;
  err_config_params:
- err_first_setup:
+	spi_set_ctldata(spi, NULL);
 	kfree(chip);
 	return status;
 }
@@ -1766,12 +2065,21 @@
 	master->setup = pl022_setup;
 	master->transfer = pl022_transfer;
 
+	/*
+	 * Supports mode 0-3, loopback, and active low CS. Transfers are
+	 * always MS bit first on the original pl022.
+	 */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LOOP;
+	if (pl022->vendor->extended_cr)
+		master->mode_bits |= SPI_LSB_FIRST;
+
 	dev_dbg(&adev->dev, "BUSNO: %d\n", master->bus_num);
 
 	status = amba_request_regions(adev, NULL);
 	if (status)
 		goto err_no_ioregion;
 
+	pl022->phybase = adev->res.start;
 	pl022->virtbase = ioremap(adev->res.start, resource_size(&adev->res));
 	if (pl022->virtbase == NULL) {
 		status = -ENOMEM;
@@ -1798,6 +2106,14 @@
 		dev_err(&adev->dev, "probe - cannot get IRQ (%d)\n", status);
 		goto err_no_irq;
 	}
+
+	/* Get DMA channels */
+	if (platform_info->enable_dma) {
+		status = pl022_dma_probe(pl022);
+		if (status != 0)
+			goto err_no_dma;
+	}
+
 	/* Initialize and start queue */
 	status = init_queue(pl022);
 	if (status != 0) {
@@ -1826,6 +2142,8 @@
  err_start_queue:
  err_init_queue:
 	destroy_queue(pl022);
+	pl022_dma_remove(pl022);
+ err_no_dma:
 	free_irq(adev->irq[0], pl022);
  err_no_irq:
 	clk_put(pl022->clk);
@@ -1856,6 +2174,7 @@
 		return status;
 	}
 	load_ssp_default_config(pl022);
+	pl022_dma_remove(pl022);
 	free_irq(adev->irq[0], pl022);
 	clk_disable(pl022->clk);
 	clk_put(pl022->clk);
diff --git a/drivers/spi/atmel_spi.c b/drivers/spi/atmel_spi.c
index c4e0442..154529a 100644
--- a/drivers/spi/atmel_spi.c
+++ b/drivers/spi/atmel_spi.c
@@ -654,6 +654,8 @@
 	struct spi_transfer	*xfer;
 	unsigned long		flags;
 	struct device		*controller = spi->master->dev.parent;
+	u8			bits;
+	struct atmel_spi_device	*asd;
 
 	as = spi_master_get_devdata(spi->master);
 
@@ -672,8 +674,18 @@
 			return -EINVAL;
 		}
 
+		if (xfer->bits_per_word) {
+			asd = spi->controller_state;
+			bits = (asd->csr >> 4) & 0xf;
+			if (bits != xfer->bits_per_word - 8) {
+				dev_dbg(&spi->dev, "you can't yet change "
+					 "bits_per_word in transfers\n");
+				return -ENOPROTOOPT;
+			}
+		}
+
 		/* FIXME implement these protocol options!! */
-		if (xfer->bits_per_word || xfer->speed_hz) {
+		if (xfer->speed_hz) {
 			dev_dbg(&spi->dev, "no protocol options yet\n");
 			return -ENOPROTOOPT;
 		}
diff --git a/drivers/spi/omap2_mcspi.c b/drivers/spi/omap2_mcspi.c
index b3a94ca..2a651e6 100644
--- a/drivers/spi/omap2_mcspi.c
+++ b/drivers/spi/omap2_mcspi.c
@@ -296,6 +296,19 @@
 	return 0;
 }
 
+static int mcspi_wait_for_reg_bit(void __iomem *reg, unsigned long bit)
+{
+	unsigned long timeout;
+
+	timeout = jiffies + msecs_to_jiffies(1000);
+	while (!(__raw_readl(reg) & bit)) {
+		if (time_after(jiffies, timeout))
+			return -1;
+		cpu_relax();
+	}
+	return 0;
+}
+
 static unsigned
 omap2_mcspi_txrx_dma(struct spi_device *spi, struct spi_transfer *xfer)
 {
@@ -309,11 +322,14 @@
 	u32			l;
 	u8			* rx;
 	const u8		* tx;
+	void __iomem		*chstat_reg;
 
 	mcspi = spi_master_get_devdata(spi->master);
 	mcspi_dma = &mcspi->dma_channels[spi->chip_select];
 	l = mcspi_cached_chconf0(spi);
 
+	chstat_reg = cs->base + OMAP2_MCSPI_CHSTAT0;
+
 	count = xfer->len;
 	c = count;
 	word_len = cs->word_len;
@@ -382,6 +398,16 @@
 	if (tx != NULL) {
 		wait_for_completion(&mcspi_dma->dma_tx_completion);
 		dma_unmap_single(NULL, xfer->tx_dma, count, DMA_TO_DEVICE);
+
+		/* for TX_ONLY mode, be sure all words have shifted out */
+		if (rx == NULL) {
+			if (mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_TXS) < 0)
+				dev_err(&spi->dev, "TXS timed out\n");
+			else if (mcspi_wait_for_reg_bit(chstat_reg,
+						OMAP2_MCSPI_CHSTAT_EOT) < 0)
+				dev_err(&spi->dev, "EOT timed out\n");
+		}
 	}
 
 	if (rx != NULL) {
@@ -435,19 +461,6 @@
 	return count;
 }
 
-static int mcspi_wait_for_reg_bit(void __iomem *reg, unsigned long bit)
-{
-	unsigned long timeout;
-
-	timeout = jiffies + msecs_to_jiffies(1000);
-	while (!(__raw_readl(reg) & bit)) {
-		if (time_after(jiffies, timeout))
-			return -1;
-		cpu_relax();
-	}
-	return 0;
-}
-
 static unsigned
 omap2_mcspi_txrx_pio(struct spi_device *spi, struct spi_transfer *xfer)
 {
@@ -489,10 +502,8 @@
 					dev_err(&spi->dev, "TXS timed out\n");
 					goto out;
 				}
-#ifdef VERBOSE
-				dev_dbg(&spi->dev, "write-%d %02x\n",
+				dev_vdbg(&spi->dev, "write-%d %02x\n",
 						word_len, *tx);
-#endif
 				__raw_writel(*tx++, tx_reg);
 			}
 			if (rx != NULL) {
@@ -506,10 +517,8 @@
 				    (l & OMAP2_MCSPI_CHCONF_TURBO)) {
 					omap2_mcspi_set_enable(spi, 0);
 					*rx++ = __raw_readl(rx_reg);
-#ifdef VERBOSE
-					dev_dbg(&spi->dev, "read-%d %02x\n",
+					dev_vdbg(&spi->dev, "read-%d %02x\n",
 						    word_len, *(rx - 1));
-#endif
 					if (mcspi_wait_for_reg_bit(chstat_reg,
 						OMAP2_MCSPI_CHSTAT_RXS) < 0) {
 						dev_err(&spi->dev,
@@ -522,10 +531,8 @@
 				}
 
 				*rx++ = __raw_readl(rx_reg);
-#ifdef VERBOSE
-				dev_dbg(&spi->dev, "read-%d %02x\n",
+				dev_vdbg(&spi->dev, "read-%d %02x\n",
 						word_len, *(rx - 1));
-#endif
 			}
 		} while (c);
 	} else if (word_len <= 16) {
@@ -542,10 +549,8 @@
 					dev_err(&spi->dev, "TXS timed out\n");
 					goto out;
 				}
-#ifdef VERBOSE
-				dev_dbg(&spi->dev, "write-%d %04x\n",
+				dev_vdbg(&spi->dev, "write-%d %04x\n",
 						word_len, *tx);
-#endif
 				__raw_writel(*tx++, tx_reg);
 			}
 			if (rx != NULL) {
@@ -559,10 +564,8 @@
 				    (l & OMAP2_MCSPI_CHCONF_TURBO)) {
 					omap2_mcspi_set_enable(spi, 0);
 					*rx++ = __raw_readl(rx_reg);
-#ifdef VERBOSE
-					dev_dbg(&spi->dev, "read-%d %04x\n",
+					dev_vdbg(&spi->dev, "read-%d %04x\n",
 						    word_len, *(rx - 1));
-#endif
 					if (mcspi_wait_for_reg_bit(chstat_reg,
 						OMAP2_MCSPI_CHSTAT_RXS) < 0) {
 						dev_err(&spi->dev,
@@ -575,10 +578,8 @@
 				}
 
 				*rx++ = __raw_readl(rx_reg);
-#ifdef VERBOSE
-				dev_dbg(&spi->dev, "read-%d %04x\n",
+				dev_vdbg(&spi->dev, "read-%d %04x\n",
 						word_len, *(rx - 1));
-#endif
 			}
 		} while (c);
 	} else if (word_len <= 32) {
@@ -595,10 +596,8 @@
 					dev_err(&spi->dev, "TXS timed out\n");
 					goto out;
 				}
-#ifdef VERBOSE
-				dev_dbg(&spi->dev, "write-%d %08x\n",
+				dev_vdbg(&spi->dev, "write-%d %08x\n",
 						word_len, *tx);
-#endif
 				__raw_writel(*tx++, tx_reg);
 			}
 			if (rx != NULL) {
@@ -612,10 +611,8 @@
 				    (l & OMAP2_MCSPI_CHCONF_TURBO)) {
 					omap2_mcspi_set_enable(spi, 0);
 					*rx++ = __raw_readl(rx_reg);
-#ifdef VERBOSE
-					dev_dbg(&spi->dev, "read-%d %08x\n",
+					dev_vdbg(&spi->dev, "read-%d %08x\n",
 						    word_len, *(rx - 1));
-#endif
 					if (mcspi_wait_for_reg_bit(chstat_reg,
 						OMAP2_MCSPI_CHSTAT_RXS) < 0) {
 						dev_err(&spi->dev,
@@ -628,10 +625,8 @@
 				}
 
 				*rx++ = __raw_readl(rx_reg);
-#ifdef VERBOSE
-				dev_dbg(&spi->dev, "read-%d %08x\n",
+				dev_vdbg(&spi->dev, "read-%d %08x\n",
 						word_len, *(rx - 1));
-#endif
 			}
 		} while (c);
 	}
@@ -644,6 +639,12 @@
 		} else if (mcspi_wait_for_reg_bit(chstat_reg,
 				OMAP2_MCSPI_CHSTAT_EOT) < 0)
 			dev_err(&spi->dev, "EOT timed out\n");
+
+		/* disable chan to purge rx datas received in TX_ONLY transfer,
+		 * otherwise these rx datas will affect the direct following
+		 * RX_ONLY transfer.
+		 */
+		omap2_mcspi_set_enable(spi, 0);
 	}
 out:
 	omap2_mcspi_set_enable(spi, 1);
diff --git a/drivers/spi/orion_spi.c b/drivers/spi/orion_spi.c
index 3aea50d..0b677dc 100644
--- a/drivers/spi/orion_spi.c
+++ b/drivers/spi/orion_spi.c
@@ -404,7 +404,7 @@
 			goto msg_rejected;
 		}
 
-		if ((t != NULL) && t->bits_per_word)
+		if (t->bits_per_word)
 			bits_per_word = t->bits_per_word;
 
 		if ((bits_per_word != 8) && (bits_per_word != 16)) {
@@ -415,7 +415,7 @@
 			goto msg_rejected;
 		}
 		/*make sure buffer length is even when working in 16 bit mode*/
-		if ((t != NULL) && (t->bits_per_word == 16) && (t->len & 1)) {
+		if ((t->bits_per_word == 16) && (t->len & 1)) {
 			dev_err(&spi->dev,
 				"message rejected : "
 				"odd data length (%d) while in 16 bit mode\n",
diff --git a/drivers/spi/spi_bfin5xx.c b/drivers/spi/spi_bfin5xx.c
index 10a6dc3..ab483a0 100644
--- a/drivers/spi/spi_bfin5xx.c
+++ b/drivers/spi/spi_bfin5xx.c
@@ -1,7 +1,7 @@
 /*
  * Blackfin On-Chip SPI Driver
  *
- * Copyright 2004-2007 Analog Devices Inc.
+ * Copyright 2004-2010 Analog Devices Inc.
  *
  * Enter bugs at http://blackfin.uclinux.org/
  *
@@ -41,13 +41,16 @@
 #define RUNNING_STATE	((void *)1)
 #define DONE_STATE	((void *)2)
 #define ERROR_STATE	((void *)-1)
-#define QUEUE_RUNNING	0
-#define QUEUE_STOPPED	1
 
-/* Value to send if no TX value is supplied */
-#define SPI_IDLE_TXVAL 0x0000
+struct bfin_spi_master_data;
 
-struct driver_data {
+struct bfin_spi_transfer_ops {
+	void (*write) (struct bfin_spi_master_data *);
+	void (*read) (struct bfin_spi_master_data *);
+	void (*duplex) (struct bfin_spi_master_data *);
+};
+
+struct bfin_spi_master_data {
 	/* Driver model hookup */
 	struct platform_device *pdev;
 
@@ -69,7 +72,7 @@
 	spinlock_t lock;
 	struct list_head queue;
 	int busy;
-	int run;
+	bool running;
 
 	/* Message Transfer pump */
 	struct tasklet_struct pump_transfers;
@@ -77,7 +80,7 @@
 	/* Current message transfer state info */
 	struct spi_message *cur_msg;
 	struct spi_transfer *cur_transfer;
-	struct chip_data *cur_chip;
+	struct bfin_spi_slave_data *cur_chip;
 	size_t len_in_bytes;
 	size_t len;
 	void *tx;
@@ -92,38 +95,37 @@
 	dma_addr_t rx_dma;
 	dma_addr_t tx_dma;
 
+	int irq_requested;
+	int spi_irq;
+
 	size_t rx_map_len;
 	size_t tx_map_len;
 	u8 n_bytes;
+	u16 ctrl_reg;
+	u16 flag_reg;
+
 	int cs_change;
-	void (*write) (struct driver_data *);
-	void (*read) (struct driver_data *);
-	void (*duplex) (struct driver_data *);
+	const struct bfin_spi_transfer_ops *ops;
 };
 
-struct chip_data {
+struct bfin_spi_slave_data {
 	u16 ctl_reg;
 	u16 baud;
 	u16 flag;
 
 	u8 chip_select_num;
-	u8 n_bytes;
-	u8 width;		/* 0 or 1 */
 	u8 enable_dma;
-	u8 bits_per_word;	/* 8 or 16 */
-	u8 cs_change_per_word;
 	u16 cs_chg_udelay;	/* Some devices require > 255usec delay */
 	u32 cs_gpio;
 	u16 idle_tx_val;
-	void (*write) (struct driver_data *);
-	void (*read) (struct driver_data *);
-	void (*duplex) (struct driver_data *);
+	u8 pio_interrupt;	/* use spi data irq */
+	const struct bfin_spi_transfer_ops *ops;
 };
 
 #define DEFINE_SPI_REG(reg, off) \
-static inline u16 read_##reg(struct driver_data *drv_data) \
+static inline u16 read_##reg(struct bfin_spi_master_data *drv_data) \
 	{ return bfin_read16(drv_data->regs_base + off); } \
-static inline void write_##reg(struct driver_data *drv_data, u16 v) \
+static inline void write_##reg(struct bfin_spi_master_data *drv_data, u16 v) \
 	{ bfin_write16(drv_data->regs_base + off, v); }
 
 DEFINE_SPI_REG(CTRL, 0x00)
@@ -134,7 +136,7 @@
 DEFINE_SPI_REG(BAUD, 0x14)
 DEFINE_SPI_REG(SHAW, 0x18)
 
-static void bfin_spi_enable(struct driver_data *drv_data)
+static void bfin_spi_enable(struct bfin_spi_master_data *drv_data)
 {
 	u16 cr;
 
@@ -142,7 +144,7 @@
 	write_CTRL(drv_data, (cr | BIT_CTL_ENABLE));
 }
 
-static void bfin_spi_disable(struct driver_data *drv_data)
+static void bfin_spi_disable(struct bfin_spi_master_data *drv_data)
 {
 	u16 cr;
 
@@ -165,7 +167,7 @@
 	return spi_baud;
 }
 
-static int bfin_spi_flush(struct driver_data *drv_data)
+static int bfin_spi_flush(struct bfin_spi_master_data *drv_data)
 {
 	unsigned long limit = loops_per_jiffy << 1;
 
@@ -179,13 +181,12 @@
 }
 
 /* Chip select operation functions for cs_change flag */
-static void bfin_spi_cs_active(struct driver_data *drv_data, struct chip_data *chip)
+static void bfin_spi_cs_active(struct bfin_spi_master_data *drv_data, struct bfin_spi_slave_data *chip)
 {
-	if (likely(chip->chip_select_num)) {
+	if (likely(chip->chip_select_num < MAX_CTRL_CS)) {
 		u16 flag = read_FLAG(drv_data);
 
-		flag |= chip->flag;
-		flag &= ~(chip->flag << 8);
+		flag &= ~chip->flag;
 
 		write_FLAG(drv_data, flag);
 	} else {
@@ -193,13 +194,13 @@
 	}
 }
 
-static void bfin_spi_cs_deactive(struct driver_data *drv_data, struct chip_data *chip)
+static void bfin_spi_cs_deactive(struct bfin_spi_master_data *drv_data,
+                                 struct bfin_spi_slave_data *chip)
 {
-	if (likely(chip->chip_select_num)) {
+	if (likely(chip->chip_select_num < MAX_CTRL_CS)) {
 		u16 flag = read_FLAG(drv_data);
 
-		flag &= ~chip->flag;
-		flag |= (chip->flag << 8);
+		flag |= chip->flag;
 
 		write_FLAG(drv_data, flag);
 	} else {
@@ -211,16 +212,43 @@
 		udelay(chip->cs_chg_udelay);
 }
 
-/* stop controller and re-config current chip*/
-static void bfin_spi_restore_state(struct driver_data *drv_data)
+/* enable or disable the pin muxed by GPIO and SPI CS to work as SPI CS */
+static inline void bfin_spi_cs_enable(struct bfin_spi_master_data *drv_data,
+                                      struct bfin_spi_slave_data *chip)
 {
-	struct chip_data *chip = drv_data->cur_chip;
+	if (chip->chip_select_num < MAX_CTRL_CS) {
+		u16 flag = read_FLAG(drv_data);
+
+		flag |= (chip->flag >> 8);
+
+		write_FLAG(drv_data, flag);
+	}
+}
+
+static inline void bfin_spi_cs_disable(struct bfin_spi_master_data *drv_data,
+                                       struct bfin_spi_slave_data *chip)
+{
+	if (chip->chip_select_num < MAX_CTRL_CS) {
+		u16 flag = read_FLAG(drv_data);
+
+		flag &= ~(chip->flag >> 8);
+
+		write_FLAG(drv_data, flag);
+	}
+}
+
+/* stop controller and re-config current chip*/
+static void bfin_spi_restore_state(struct bfin_spi_master_data *drv_data)
+{
+	struct bfin_spi_slave_data *chip = drv_data->cur_chip;
 
 	/* Clear status and disable clock */
 	write_STAT(drv_data, BIT_STAT_CLR);
 	bfin_spi_disable(drv_data);
 	dev_dbg(&drv_data->pdev->dev, "restoring spi ctl state\n");
 
+	SSYNC();
+
 	/* Load the registers */
 	write_CTRL(drv_data, chip->ctl_reg);
 	write_BAUD(drv_data, chip->baud);
@@ -230,49 +258,12 @@
 }
 
 /* used to kick off transfer in rx mode and read unwanted RX data */
-static inline void bfin_spi_dummy_read(struct driver_data *drv_data)
+static inline void bfin_spi_dummy_read(struct bfin_spi_master_data *drv_data)
 {
 	(void) read_RDBR(drv_data);
 }
 
-static void bfin_spi_null_writer(struct driver_data *drv_data)
-{
-	u8 n_bytes = drv_data->n_bytes;
-	u16 tx_val = drv_data->cur_chip->idle_tx_val;
-
-	/* clear RXS (we check for RXS inside the loop) */
-	bfin_spi_dummy_read(drv_data);
-
-	while (drv_data->tx < drv_data->tx_end) {
-		write_TDBR(drv_data, tx_val);
-		drv_data->tx += n_bytes;
-		/* wait until transfer finished.
-		   checking SPIF or TXS may not guarantee transfer completion */
-		while (!(read_STAT(drv_data) & BIT_STAT_RXS))
-			cpu_relax();
-		/* discard RX data and clear RXS */
-		bfin_spi_dummy_read(drv_data);
-	}
-}
-
-static void bfin_spi_null_reader(struct driver_data *drv_data)
-{
-	u8 n_bytes = drv_data->n_bytes;
-	u16 tx_val = drv_data->cur_chip->idle_tx_val;
-
-	/* discard old RX data and clear RXS */
-	bfin_spi_dummy_read(drv_data);
-
-	while (drv_data->rx < drv_data->rx_end) {
-		write_TDBR(drv_data, tx_val);
-		drv_data->rx += n_bytes;
-		while (!(read_STAT(drv_data) & BIT_STAT_RXS))
-			cpu_relax();
-		bfin_spi_dummy_read(drv_data);
-	}
-}
-
-static void bfin_spi_u8_writer(struct driver_data *drv_data)
+static void bfin_spi_u8_writer(struct bfin_spi_master_data *drv_data)
 {
 	/* clear RXS (we check for RXS inside the loop) */
 	bfin_spi_dummy_read(drv_data);
@@ -288,25 +279,7 @@
 	}
 }
 
-static void bfin_spi_u8_cs_chg_writer(struct driver_data *drv_data)
-{
-	struct chip_data *chip = drv_data->cur_chip;
-
-	/* clear RXS (we check for RXS inside the loop) */
-	bfin_spi_dummy_read(drv_data);
-
-	while (drv_data->tx < drv_data->tx_end) {
-		bfin_spi_cs_active(drv_data, chip);
-		write_TDBR(drv_data, (*(u8 *) (drv_data->tx++)));
-		/* make sure transfer finished before deactiving CS */
-		while (!(read_STAT(drv_data) & BIT_STAT_RXS))
-			cpu_relax();
-		bfin_spi_dummy_read(drv_data);
-		bfin_spi_cs_deactive(drv_data, chip);
-	}
-}
-
-static void bfin_spi_u8_reader(struct driver_data *drv_data)
+static void bfin_spi_u8_reader(struct bfin_spi_master_data *drv_data)
 {
 	u16 tx_val = drv_data->cur_chip->idle_tx_val;
 
@@ -321,25 +294,7 @@
 	}
 }
 
-static void bfin_spi_u8_cs_chg_reader(struct driver_data *drv_data)
-{
-	struct chip_data *chip = drv_data->cur_chip;
-	u16 tx_val = chip->idle_tx_val;
-
-	/* discard old RX data and clear RXS */
-	bfin_spi_dummy_read(drv_data);
-
-	while (drv_data->rx < drv_data->rx_end) {
-		bfin_spi_cs_active(drv_data, chip);
-		write_TDBR(drv_data, tx_val);
-		while (!(read_STAT(drv_data) & BIT_STAT_RXS))
-			cpu_relax();
-		*(u8 *) (drv_data->rx++) = read_RDBR(drv_data);
-		bfin_spi_cs_deactive(drv_data, chip);
-	}
-}
-
-static void bfin_spi_u8_duplex(struct driver_data *drv_data)
+static void bfin_spi_u8_duplex(struct bfin_spi_master_data *drv_data)
 {
 	/* discard old RX data and clear RXS */
 	bfin_spi_dummy_read(drv_data);
@@ -352,24 +307,13 @@
 	}
 }
 
-static void bfin_spi_u8_cs_chg_duplex(struct driver_data *drv_data)
-{
-	struct chip_data *chip = drv_data->cur_chip;
+static const struct bfin_spi_transfer_ops bfin_bfin_spi_transfer_ops_u8 = {
+	.write  = bfin_spi_u8_writer,
+	.read   = bfin_spi_u8_reader,
+	.duplex = bfin_spi_u8_duplex,
+};
 
-	/* discard old RX data and clear RXS */
-	bfin_spi_dummy_read(drv_data);
-
-	while (drv_data->rx < drv_data->rx_end) {
-		bfin_spi_cs_active(drv_data, chip);
-		write_TDBR(drv_data, (*(u8 *) (drv_data->tx++)));
-		while (!(read_STAT(drv_data) & BIT_STAT_RXS))
-			cpu_relax();
-		*(u8 *) (drv_data->rx++) = read_RDBR(drv_data);
-		bfin_spi_cs_deactive(drv_data, chip);
-	}
-}
-
-static void bfin_spi_u16_writer(struct driver_data *drv_data)
+static void bfin_spi_u16_writer(struct bfin_spi_master_data *drv_data)
 {
 	/* clear RXS (we check for RXS inside the loop) */
 	bfin_spi_dummy_read(drv_data);
@@ -386,26 +330,7 @@
 	}
 }
 
-static void bfin_spi_u16_cs_chg_writer(struct driver_data *drv_data)
-{
-	struct chip_data *chip = drv_data->cur_chip;
-
-	/* clear RXS (we check for RXS inside the loop) */
-	bfin_spi_dummy_read(drv_data);
-
-	while (drv_data->tx < drv_data->tx_end) {
-		bfin_spi_cs_active(drv_data, chip);
-		write_TDBR(drv_data, (*(u16 *) (drv_data->tx)));
-		drv_data->tx += 2;
-		/* make sure transfer finished before deactiving CS */
-		while (!(read_STAT(drv_data) & BIT_STAT_RXS))
-			cpu_relax();
-		bfin_spi_dummy_read(drv_data);
-		bfin_spi_cs_deactive(drv_data, chip);
-	}
-}
-
-static void bfin_spi_u16_reader(struct driver_data *drv_data)
+static void bfin_spi_u16_reader(struct bfin_spi_master_data *drv_data)
 {
 	u16 tx_val = drv_data->cur_chip->idle_tx_val;
 
@@ -421,26 +346,7 @@
 	}
 }
 
-static void bfin_spi_u16_cs_chg_reader(struct driver_data *drv_data)
-{
-	struct chip_data *chip = drv_data->cur_chip;
-	u16 tx_val = chip->idle_tx_val;
-
-	/* discard old RX data and clear RXS */
-	bfin_spi_dummy_read(drv_data);
-
-	while (drv_data->rx < drv_data->rx_end) {
-		bfin_spi_cs_active(drv_data, chip);
-		write_TDBR(drv_data, tx_val);
-		while (!(read_STAT(drv_data) & BIT_STAT_RXS))
-			cpu_relax();
-		*(u16 *) (drv_data->rx) = read_RDBR(drv_data);
-		drv_data->rx += 2;
-		bfin_spi_cs_deactive(drv_data, chip);
-	}
-}
-
-static void bfin_spi_u16_duplex(struct driver_data *drv_data)
+static void bfin_spi_u16_duplex(struct bfin_spi_master_data *drv_data)
 {
 	/* discard old RX data and clear RXS */
 	bfin_spi_dummy_read(drv_data);
@@ -455,27 +361,14 @@
 	}
 }
 
-static void bfin_spi_u16_cs_chg_duplex(struct driver_data *drv_data)
-{
-	struct chip_data *chip = drv_data->cur_chip;
+static const struct bfin_spi_transfer_ops bfin_bfin_spi_transfer_ops_u16 = {
+	.write  = bfin_spi_u16_writer,
+	.read   = bfin_spi_u16_reader,
+	.duplex = bfin_spi_u16_duplex,
+};
 
-	/* discard old RX data and clear RXS */
-	bfin_spi_dummy_read(drv_data);
-
-	while (drv_data->rx < drv_data->rx_end) {
-		bfin_spi_cs_active(drv_data, chip);
-		write_TDBR(drv_data, (*(u16 *) (drv_data->tx)));
-		drv_data->tx += 2;
-		while (!(read_STAT(drv_data) & BIT_STAT_RXS))
-			cpu_relax();
-		*(u16 *) (drv_data->rx) = read_RDBR(drv_data);
-		drv_data->rx += 2;
-		bfin_spi_cs_deactive(drv_data, chip);
-	}
-}
-
-/* test if ther is more transfer to be done */
-static void *bfin_spi_next_transfer(struct driver_data *drv_data)
+/* test if there is more transfer to be done */
+static void *bfin_spi_next_transfer(struct bfin_spi_master_data *drv_data)
 {
 	struct spi_message *msg = drv_data->cur_msg;
 	struct spi_transfer *trans = drv_data->cur_transfer;
@@ -494,9 +387,9 @@
  * caller already set message->status;
  * dma and pio irqs are blocked give finished message back
  */
-static void bfin_spi_giveback(struct driver_data *drv_data)
+static void bfin_spi_giveback(struct bfin_spi_master_data *drv_data)
 {
-	struct chip_data *chip = drv_data->cur_chip;
+	struct bfin_spi_slave_data *chip = drv_data->cur_chip;
 	struct spi_transfer *last_transfer;
 	unsigned long flags;
 	struct spi_message *msg;
@@ -525,10 +418,83 @@
 		msg->complete(msg->context);
 }
 
+/* spi data irq handler */
+static irqreturn_t bfin_spi_pio_irq_handler(int irq, void *dev_id)
+{
+	struct bfin_spi_master_data *drv_data = dev_id;
+	struct bfin_spi_slave_data *chip = drv_data->cur_chip;
+	struct spi_message *msg = drv_data->cur_msg;
+	int n_bytes = drv_data->n_bytes;
+
+	/* wait until transfer finished. */
+	while (!(read_STAT(drv_data) & BIT_STAT_RXS))
+		cpu_relax();
+
+	if ((drv_data->tx && drv_data->tx >= drv_data->tx_end) ||
+		(drv_data->rx && drv_data->rx >= (drv_data->rx_end - n_bytes))) {
+		/* last read */
+		if (drv_data->rx) {
+			dev_dbg(&drv_data->pdev->dev, "last read\n");
+			if (n_bytes == 2)
+				*(u16 *) (drv_data->rx) = read_RDBR(drv_data);
+			else if (n_bytes == 1)
+				*(u8 *) (drv_data->rx) = read_RDBR(drv_data);
+			drv_data->rx += n_bytes;
+		}
+
+		msg->actual_length += drv_data->len_in_bytes;
+		if (drv_data->cs_change)
+			bfin_spi_cs_deactive(drv_data, chip);
+		/* Move to next transfer */
+		msg->state = bfin_spi_next_transfer(drv_data);
+
+		disable_irq_nosync(drv_data->spi_irq);
+
+		/* Schedule transfer tasklet */
+		tasklet_schedule(&drv_data->pump_transfers);
+		return IRQ_HANDLED;
+	}
+
+	if (drv_data->rx && drv_data->tx) {
+		/* duplex */
+		dev_dbg(&drv_data->pdev->dev, "duplex: write_TDBR\n");
+		if (drv_data->n_bytes == 2) {
+			*(u16 *) (drv_data->rx) = read_RDBR(drv_data);
+			write_TDBR(drv_data, (*(u16 *) (drv_data->tx)));
+		} else if (drv_data->n_bytes == 1) {
+			*(u8 *) (drv_data->rx) = read_RDBR(drv_data);
+			write_TDBR(drv_data, (*(u8 *) (drv_data->tx)));
+		}
+	} else if (drv_data->rx) {
+		/* read */
+		dev_dbg(&drv_data->pdev->dev, "read: write_TDBR\n");
+		if (drv_data->n_bytes == 2)
+			*(u16 *) (drv_data->rx) = read_RDBR(drv_data);
+		else if (drv_data->n_bytes == 1)
+			*(u8 *) (drv_data->rx) = read_RDBR(drv_data);
+		write_TDBR(drv_data, chip->idle_tx_val);
+	} else if (drv_data->tx) {
+		/* write */
+		dev_dbg(&drv_data->pdev->dev, "write: write_TDBR\n");
+		bfin_spi_dummy_read(drv_data);
+		if (drv_data->n_bytes == 2)
+			write_TDBR(drv_data, (*(u16 *) (drv_data->tx)));
+		else if (drv_data->n_bytes == 1)
+			write_TDBR(drv_data, (*(u8 *) (drv_data->tx)));
+	}
+
+	if (drv_data->tx)
+		drv_data->tx += n_bytes;
+	if (drv_data->rx)
+		drv_data->rx += n_bytes;
+
+	return IRQ_HANDLED;
+}
+
 static irqreturn_t bfin_spi_dma_irq_handler(int irq, void *dev_id)
 {
-	struct driver_data *drv_data = dev_id;
-	struct chip_data *chip = drv_data->cur_chip;
+	struct bfin_spi_master_data *drv_data = dev_id;
+	struct bfin_spi_slave_data *chip = drv_data->cur_chip;
 	struct spi_message *msg = drv_data->cur_msg;
 	unsigned long timeout;
 	unsigned short dmastat = get_dma_curr_irqstat(drv_data->dma_channel);
@@ -540,10 +506,6 @@
 
 	clear_dma_irqstat(drv_data->dma_channel);
 
-	/* Wait for DMA to complete */
-	while (get_dma_curr_irqstat(drv_data->dma_channel) & DMA_RUN)
-		cpu_relax();
-
 	/*
 	 * wait for the last transaction shifted out.  HRM states:
 	 * at this point there may still be data in the SPI DMA FIFO waiting
@@ -551,8 +513,8 @@
 	 * register until it goes low for 2 successive reads
 	 */
 	if (drv_data->tx != NULL) {
-		while ((read_STAT(drv_data) & TXS) ||
-		       (read_STAT(drv_data) & TXS))
+		while ((read_STAT(drv_data) & BIT_STAT_TXS) ||
+		       (read_STAT(drv_data) & BIT_STAT_TXS))
 			cpu_relax();
 	}
 
@@ -561,14 +523,14 @@
 		dmastat, read_STAT(drv_data));
 
 	timeout = jiffies + HZ;
-	while (!(read_STAT(drv_data) & SPIF))
+	while (!(read_STAT(drv_data) & BIT_STAT_SPIF))
 		if (!time_before(jiffies, timeout)) {
 			dev_warn(&drv_data->pdev->dev, "timeout waiting for SPIF");
 			break;
 		} else
 			cpu_relax();
 
-	if ((dmastat & DMA_ERR) && (spistat & RBSY)) {
+	if ((dmastat & DMA_ERR) && (spistat & BIT_STAT_RBSY)) {
 		msg->state = ERROR_STATE;
 		dev_err(&drv_data->pdev->dev, "dma receive: fifo/buffer overflow\n");
 	} else {
@@ -588,20 +550,20 @@
 	dev_dbg(&drv_data->pdev->dev,
 		"disable dma channel irq%d\n",
 		drv_data->dma_channel);
-	dma_disable_irq(drv_data->dma_channel);
+	dma_disable_irq_nosync(drv_data->dma_channel);
 
 	return IRQ_HANDLED;
 }
 
 static void bfin_spi_pump_transfers(unsigned long data)
 {
-	struct driver_data *drv_data = (struct driver_data *)data;
+	struct bfin_spi_master_data *drv_data = (struct bfin_spi_master_data *)data;
 	struct spi_message *message = NULL;
 	struct spi_transfer *transfer = NULL;
 	struct spi_transfer *previous = NULL;
-	struct chip_data *chip = NULL;
-	u8 width;
-	u16 cr, dma_width, dma_config;
+	struct bfin_spi_slave_data *chip = NULL;
+	unsigned int bits_per_word;
+	u16 cr, cr_width, dma_width, dma_config;
 	u32 tranf_success = 1;
 	u8 full_duplex = 0;
 
@@ -639,7 +601,7 @@
 			udelay(previous->delay_usecs);
 	}
 
-	/* Setup the transfer state based on the type of transfer */
+	/* Flush any existing transfers that may be sitting in the hardware */
 	if (bfin_spi_flush(drv_data) == 0) {
 		dev_err(&drv_data->pdev->dev, "pump_transfers: flush failed\n");
 		message->status = -EIO;
@@ -679,52 +641,31 @@
 	drv_data->cs_change = transfer->cs_change;
 
 	/* Bits per word setup */
-	switch (transfer->bits_per_word) {
-	case 8:
+	bits_per_word = transfer->bits_per_word ? : message->spi->bits_per_word;
+	if (bits_per_word == 8) {
 		drv_data->n_bytes = 1;
-		width = CFG_SPI_WORDSIZE8;
-		drv_data->read = chip->cs_change_per_word ?
-			bfin_spi_u8_cs_chg_reader : bfin_spi_u8_reader;
-		drv_data->write = chip->cs_change_per_word ?
-			bfin_spi_u8_cs_chg_writer : bfin_spi_u8_writer;
-		drv_data->duplex = chip->cs_change_per_word ?
-			bfin_spi_u8_cs_chg_duplex : bfin_spi_u8_duplex;
-		break;
-
-	case 16:
+		drv_data->len = transfer->len;
+		cr_width = 0;
+		drv_data->ops = &bfin_bfin_spi_transfer_ops_u8;
+	} else if (bits_per_word == 16) {
 		drv_data->n_bytes = 2;
-		width = CFG_SPI_WORDSIZE16;
-		drv_data->read = chip->cs_change_per_word ?
-			bfin_spi_u16_cs_chg_reader : bfin_spi_u16_reader;
-		drv_data->write = chip->cs_change_per_word ?
-			bfin_spi_u16_cs_chg_writer : bfin_spi_u16_writer;
-		drv_data->duplex = chip->cs_change_per_word ?
-			bfin_spi_u16_cs_chg_duplex : bfin_spi_u16_duplex;
-		break;
-
-	default:
-		/* No change, the same as default setting */
-		drv_data->n_bytes = chip->n_bytes;
-		width = chip->width;
-		drv_data->write = drv_data->tx ? chip->write : bfin_spi_null_writer;
-		drv_data->read = drv_data->rx ? chip->read : bfin_spi_null_reader;
-		drv_data->duplex = chip->duplex ? chip->duplex : bfin_spi_null_writer;
-		break;
+		drv_data->len = (transfer->len) >> 1;
+		cr_width = BIT_CTL_WORDSIZE;
+		drv_data->ops = &bfin_bfin_spi_transfer_ops_u16;
+	} else {
+		dev_err(&drv_data->pdev->dev, "transfer: unsupported bits_per_word\n");
+		message->status = -EINVAL;
+		bfin_spi_giveback(drv_data);
+		return;
 	}
-	cr = (read_CTRL(drv_data) & (~BIT_CTL_TIMOD));
-	cr |= (width << 8);
+	cr = read_CTRL(drv_data) & ~(BIT_CTL_TIMOD | BIT_CTL_WORDSIZE);
+	cr |= cr_width;
 	write_CTRL(drv_data, cr);
 
-	if (width == CFG_SPI_WORDSIZE16) {
-		drv_data->len = (transfer->len) >> 1;
-	} else {
-		drv_data->len = transfer->len;
-	}
 	dev_dbg(&drv_data->pdev->dev,
-		"transfer: drv_data->write is %p, chip->write is %p, null_wr is %p\n",
-		drv_data->write, chip->write, bfin_spi_null_writer);
+		"transfer: drv_data->ops is %p, chip->ops is %p, u8_ops is %p\n",
+		drv_data->ops, chip->ops, &bfin_bfin_spi_transfer_ops_u8);
 
-	/* speed and width has been set on per message */
 	message->state = RUNNING_STATE;
 	dma_config = 0;
 
@@ -735,13 +676,11 @@
 		write_BAUD(drv_data, chip->baud);
 
 	write_STAT(drv_data, BIT_STAT_CLR);
-	cr = (read_CTRL(drv_data) & (~BIT_CTL_TIMOD));
-	if (drv_data->cs_change)
-		bfin_spi_cs_active(drv_data, chip);
+	bfin_spi_cs_active(drv_data, chip);
 
 	dev_dbg(&drv_data->pdev->dev,
 		"now pumping a transfer: width is %d, len is %d\n",
-		width, transfer->len);
+		cr_width, transfer->len);
 
 	/*
 	 * Try to map dma buffer and do a dma transfer.  If successful use,
@@ -760,7 +699,7 @@
 		/* config dma channel */
 		dev_dbg(&drv_data->pdev->dev, "doing dma transfer\n");
 		set_dma_x_count(drv_data->dma_channel, drv_data->len);
-		if (width == CFG_SPI_WORDSIZE16) {
+		if (cr_width == BIT_CTL_WORDSIZE) {
 			set_dma_x_modify(drv_data->dma_channel, 2);
 			dma_width = WDSIZE_16;
 		} else {
@@ -846,73 +785,100 @@
 		dma_enable_irq(drv_data->dma_channel);
 		local_irq_restore(flags);
 
-	} else {
-		/* IO mode write then read */
-		dev_dbg(&drv_data->pdev->dev, "doing IO transfer\n");
-
-		/* we always use SPI_WRITE mode. SPI_READ mode
-		   seems to have problems with setting up the
-		   output value in TDBR prior to the transfer. */
-		write_CTRL(drv_data, (cr | CFG_SPI_WRITE));
-
-		if (full_duplex) {
-			/* full duplex mode */
-			BUG_ON((drv_data->tx_end - drv_data->tx) !=
-			       (drv_data->rx_end - drv_data->rx));
-			dev_dbg(&drv_data->pdev->dev,
-				"IO duplex: cr is 0x%x\n", cr);
-
-			drv_data->duplex(drv_data);
-
-			if (drv_data->tx != drv_data->tx_end)
-				tranf_success = 0;
-		} else if (drv_data->tx != NULL) {
-			/* write only half duplex */
-			dev_dbg(&drv_data->pdev->dev,
-				"IO write: cr is 0x%x\n", cr);
-
-			drv_data->write(drv_data);
-
-			if (drv_data->tx != drv_data->tx_end)
-				tranf_success = 0;
-		} else if (drv_data->rx != NULL) {
-			/* read only half duplex */
-			dev_dbg(&drv_data->pdev->dev,
-				"IO read: cr is 0x%x\n", cr);
-
-			drv_data->read(drv_data);
-			if (drv_data->rx != drv_data->rx_end)
-				tranf_success = 0;
-		}
-
-		if (!tranf_success) {
-			dev_dbg(&drv_data->pdev->dev,
-				"IO write error!\n");
-			message->state = ERROR_STATE;
-		} else {
-			/* Update total byte transfered */
-			message->actual_length += drv_data->len_in_bytes;
-			/* Move to next transfer of this msg */
-			message->state = bfin_spi_next_transfer(drv_data);
-			if (drv_data->cs_change)
-				bfin_spi_cs_deactive(drv_data, chip);
-		}
-		/* Schedule next transfer tasklet */
-		tasklet_schedule(&drv_data->pump_transfers);
+		return;
 	}
+
+	/*
+	 * We always use SPI_WRITE mode (transfer starts with TDBR write).
+	 * SPI_READ mode (transfer starts with RDBR read) seems to have
+	 * problems with setting up the output value in TDBR prior to the
+	 * start of the transfer.
+	 */
+	write_CTRL(drv_data, cr | BIT_CTL_TXMOD);
+
+	if (chip->pio_interrupt) {
+		/* SPI irq should have been disabled by now */
+
+		/* discard old RX data and clear RXS */
+		bfin_spi_dummy_read(drv_data);
+
+		/* start transfer */
+		if (drv_data->tx == NULL)
+			write_TDBR(drv_data, chip->idle_tx_val);
+		else {
+			if (bits_per_word == 8)
+				write_TDBR(drv_data, (*(u8 *) (drv_data->tx)));
+			else
+				write_TDBR(drv_data, (*(u16 *) (drv_data->tx)));
+			drv_data->tx += drv_data->n_bytes;
+		}
+
+		/* once TDBR is empty, interrupt is triggered */
+		enable_irq(drv_data->spi_irq);
+		return;
+	}
+
+	/* IO mode */
+	dev_dbg(&drv_data->pdev->dev, "doing IO transfer\n");
+
+	if (full_duplex) {
+		/* full duplex mode */
+		BUG_ON((drv_data->tx_end - drv_data->tx) !=
+		       (drv_data->rx_end - drv_data->rx));
+		dev_dbg(&drv_data->pdev->dev,
+			"IO duplex: cr is 0x%x\n", cr);
+
+		drv_data->ops->duplex(drv_data);
+
+		if (drv_data->tx != drv_data->tx_end)
+			tranf_success = 0;
+	} else if (drv_data->tx != NULL) {
+		/* write only half duplex */
+		dev_dbg(&drv_data->pdev->dev,
+			"IO write: cr is 0x%x\n", cr);
+
+		drv_data->ops->write(drv_data);
+
+		if (drv_data->tx != drv_data->tx_end)
+			tranf_success = 0;
+	} else if (drv_data->rx != NULL) {
+		/* read only half duplex */
+		dev_dbg(&drv_data->pdev->dev,
+			"IO read: cr is 0x%x\n", cr);
+
+		drv_data->ops->read(drv_data);
+		if (drv_data->rx != drv_data->rx_end)
+			tranf_success = 0;
+	}
+
+	if (!tranf_success) {
+		dev_dbg(&drv_data->pdev->dev,
+			"IO write error!\n");
+		message->state = ERROR_STATE;
+	} else {
+		/* Update total byte transfered */
+		message->actual_length += drv_data->len_in_bytes;
+		/* Move to next transfer of this msg */
+		message->state = bfin_spi_next_transfer(drv_data);
+		if (drv_data->cs_change)
+			bfin_spi_cs_deactive(drv_data, chip);
+	}
+
+	/* Schedule next transfer tasklet */
+	tasklet_schedule(&drv_data->pump_transfers);
 }
 
 /* pop a msg from queue and kick off real transfer */
 static void bfin_spi_pump_messages(struct work_struct *work)
 {
-	struct driver_data *drv_data;
+	struct bfin_spi_master_data *drv_data;
 	unsigned long flags;
 
-	drv_data = container_of(work, struct driver_data, pump_messages);
+	drv_data = container_of(work, struct bfin_spi_master_data, pump_messages);
 
 	/* Lock queue and check for queue work */
 	spin_lock_irqsave(&drv_data->lock, flags);
-	if (list_empty(&drv_data->queue) || drv_data->run == QUEUE_STOPPED) {
+	if (list_empty(&drv_data->queue) || !drv_data->running) {
 		/* pumper kicked off but no work to do */
 		drv_data->busy = 0;
 		spin_unlock_irqrestore(&drv_data->lock, flags);
@@ -962,12 +928,12 @@
  */
 static int bfin_spi_transfer(struct spi_device *spi, struct spi_message *msg)
 {
-	struct driver_data *drv_data = spi_master_get_devdata(spi->master);
+	struct bfin_spi_master_data *drv_data = spi_master_get_devdata(spi->master);
 	unsigned long flags;
 
 	spin_lock_irqsave(&drv_data->lock, flags);
 
-	if (drv_data->run == QUEUE_STOPPED) {
+	if (!drv_data->running) {
 		spin_unlock_irqrestore(&drv_data->lock, flags);
 		return -ESHUTDOWN;
 	}
@@ -979,7 +945,7 @@
 	dev_dbg(&spi->dev, "adding an msg in transfer() \n");
 	list_add_tail(&msg->queue, &drv_data->queue);
 
-	if (drv_data->run == QUEUE_RUNNING && !drv_data->busy)
+	if (drv_data->running && !drv_data->busy)
 		queue_work(drv_data->workqueue, &drv_data->pump_messages);
 
 	spin_unlock_irqrestore(&drv_data->lock, flags);
@@ -1003,147 +969,184 @@
 	P_SPI2_SSEL6, P_SPI2_SSEL7},
 };
 
-/* first setup for new devices */
+/* setup for devices (may be called multiple times -- not just first setup) */
 static int bfin_spi_setup(struct spi_device *spi)
 {
-	struct bfin5xx_spi_chip *chip_info = NULL;
-	struct chip_data *chip;
-	struct driver_data *drv_data = spi_master_get_devdata(spi->master);
-	int ret;
-
-	if (spi->bits_per_word != 8 && spi->bits_per_word != 16)
-		return -EINVAL;
+	struct bfin5xx_spi_chip *chip_info;
+	struct bfin_spi_slave_data *chip = NULL;
+	struct bfin_spi_master_data *drv_data = spi_master_get_devdata(spi->master);
+	u16 bfin_ctl_reg;
+	int ret = -EINVAL;
 
 	/* Only alloc (or use chip_info) on first setup */
+	chip_info = NULL;
 	chip = spi_get_ctldata(spi);
 	if (chip == NULL) {
-		chip = kzalloc(sizeof(struct chip_data), GFP_KERNEL);
-		if (!chip)
-			return -ENOMEM;
+		chip = kzalloc(sizeof(*chip), GFP_KERNEL);
+		if (!chip) {
+			dev_err(&spi->dev, "cannot allocate chip data\n");
+			ret = -ENOMEM;
+			goto error;
+		}
 
 		chip->enable_dma = 0;
 		chip_info = spi->controller_data;
 	}
 
+	/* Let people set non-standard bits directly */
+	bfin_ctl_reg = BIT_CTL_OPENDRAIN | BIT_CTL_EMISO |
+		BIT_CTL_PSSE | BIT_CTL_GM | BIT_CTL_SZ;
+
 	/* chip_info isn't always needed */
 	if (chip_info) {
 		/* Make sure people stop trying to set fields via ctl_reg
 		 * when they should actually be using common SPI framework.
-		 * Currently we let through: WOM EMISO PSSE GM SZ TIMOD.
+		 * Currently we let through: WOM EMISO PSSE GM SZ.
 		 * Not sure if a user actually needs/uses any of these,
 		 * but let's assume (for now) they do.
 		 */
-		if (chip_info->ctl_reg & (SPE|MSTR|CPOL|CPHA|LSBF|SIZE)) {
+		if (chip_info->ctl_reg & ~bfin_ctl_reg) {
 			dev_err(&spi->dev, "do not set bits in ctl_reg "
 				"that the SPI framework manages\n");
-			return -EINVAL;
+			goto error;
 		}
-
 		chip->enable_dma = chip_info->enable_dma != 0
 		    && drv_data->master_info->enable_dma;
 		chip->ctl_reg = chip_info->ctl_reg;
-		chip->bits_per_word = chip_info->bits_per_word;
-		chip->cs_change_per_word = chip_info->cs_change_per_word;
 		chip->cs_chg_udelay = chip_info->cs_chg_udelay;
-		chip->cs_gpio = chip_info->cs_gpio;
 		chip->idle_tx_val = chip_info->idle_tx_val;
+		chip->pio_interrupt = chip_info->pio_interrupt;
+		spi->bits_per_word = chip_info->bits_per_word;
+	} else {
+		/* force a default base state */
+		chip->ctl_reg &= bfin_ctl_reg;
+	}
+
+	if (spi->bits_per_word != 8 && spi->bits_per_word != 16) {
+		dev_err(&spi->dev, "%d bits_per_word is not supported\n",
+				spi->bits_per_word);
+		goto error;
 	}
 
 	/* translate common spi framework into our register */
-	if (spi->mode & SPI_CPOL)
-		chip->ctl_reg |= CPOL;
-	if (spi->mode & SPI_CPHA)
-		chip->ctl_reg |= CPHA;
-	if (spi->mode & SPI_LSB_FIRST)
-		chip->ctl_reg |= LSBF;
-	/* we dont support running in slave mode (yet?) */
-	chip->ctl_reg |= MSTR;
-
-	/*
-	 * if any one SPI chip is registered and wants DMA, request the
-	 * DMA channel for it
-	 */
-	if (chip->enable_dma && !drv_data->dma_requested) {
-		/* register dma irq handler */
-		if (request_dma(drv_data->dma_channel, "BFIN_SPI_DMA") < 0) {
-			dev_dbg(&spi->dev,
-				"Unable to request BlackFin SPI DMA channel\n");
-			return -ENODEV;
-		}
-		if (set_dma_callback(drv_data->dma_channel,
-		    bfin_spi_dma_irq_handler, drv_data) < 0) {
-			dev_dbg(&spi->dev, "Unable to set dma callback\n");
-			return -EPERM;
-		}
-		dma_disable_irq(drv_data->dma_channel);
-		drv_data->dma_requested = 1;
+	if (spi->mode & ~(SPI_CPOL | SPI_CPHA | SPI_LSB_FIRST)) {
+		dev_err(&spi->dev, "unsupported spi modes detected\n");
+		goto error;
 	}
+	if (spi->mode & SPI_CPOL)
+		chip->ctl_reg |= BIT_CTL_CPOL;
+	if (spi->mode & SPI_CPHA)
+		chip->ctl_reg |= BIT_CTL_CPHA;
+	if (spi->mode & SPI_LSB_FIRST)
+		chip->ctl_reg |= BIT_CTL_LSBF;
+	/* we dont support running in slave mode (yet?) */
+	chip->ctl_reg |= BIT_CTL_MASTER;
 
 	/*
 	 * Notice: for blackfin, the speed_hz is the value of register
 	 * SPI_BAUD, not the real baudrate
 	 */
 	chip->baud = hz_to_spi_baud(spi->max_speed_hz);
-	chip->flag = 1 << (spi->chip_select);
 	chip->chip_select_num = spi->chip_select;
+	if (chip->chip_select_num < MAX_CTRL_CS) {
+		if (!(spi->mode & SPI_CPHA))
+			dev_warn(&spi->dev, "Warning: SPI CPHA not set:"
+				" Slave Select not under software control!\n"
+				" See Documentation/blackfin/bfin-spi-notes.txt");
 
-	if (chip->chip_select_num == 0) {
+		chip->flag = (1 << spi->chip_select) << 8;
+	} else
+		chip->cs_gpio = chip->chip_select_num - MAX_CTRL_CS;
+
+	if (chip->enable_dma && chip->pio_interrupt) {
+		dev_err(&spi->dev, "enable_dma is set, "
+				"do not set pio_interrupt\n");
+		goto error;
+	}
+	/*
+	 * if any one SPI chip is registered and wants DMA, request the
+	 * DMA channel for it
+	 */
+	if (chip->enable_dma && !drv_data->dma_requested) {
+		/* register dma irq handler */
+		ret = request_dma(drv_data->dma_channel, "BFIN_SPI_DMA");
+		if (ret) {
+			dev_err(&spi->dev,
+				"Unable to request BlackFin SPI DMA channel\n");
+			goto error;
+		}
+		drv_data->dma_requested = 1;
+
+		ret = set_dma_callback(drv_data->dma_channel,
+			bfin_spi_dma_irq_handler, drv_data);
+		if (ret) {
+			dev_err(&spi->dev, "Unable to set dma callback\n");
+			goto error;
+		}
+		dma_disable_irq(drv_data->dma_channel);
+	}
+
+	if (chip->pio_interrupt && !drv_data->irq_requested) {
+		ret = request_irq(drv_data->spi_irq, bfin_spi_pio_irq_handler,
+			IRQF_DISABLED, "BFIN_SPI", drv_data);
+		if (ret) {
+			dev_err(&spi->dev, "Unable to register spi IRQ\n");
+			goto error;
+		}
+		drv_data->irq_requested = 1;
+		/* we use write mode, spi irq has to be disabled here */
+		disable_irq(drv_data->spi_irq);
+	}
+
+	if (chip->chip_select_num >= MAX_CTRL_CS) {
 		ret = gpio_request(chip->cs_gpio, spi->modalias);
 		if (ret) {
-			if (drv_data->dma_requested)
-				free_dma(drv_data->dma_channel);
-			return ret;
+			dev_err(&spi->dev, "gpio_request() error\n");
+			goto pin_error;
 		}
 		gpio_direction_output(chip->cs_gpio, 1);
 	}
 
-	switch (chip->bits_per_word) {
-	case 8:
-		chip->n_bytes = 1;
-		chip->width = CFG_SPI_WORDSIZE8;
-		chip->read = chip->cs_change_per_word ?
-			bfin_spi_u8_cs_chg_reader : bfin_spi_u8_reader;
-		chip->write = chip->cs_change_per_word ?
-			bfin_spi_u8_cs_chg_writer : bfin_spi_u8_writer;
-		chip->duplex = chip->cs_change_per_word ?
-			bfin_spi_u8_cs_chg_duplex : bfin_spi_u8_duplex;
-		break;
-
-	case 16:
-		chip->n_bytes = 2;
-		chip->width = CFG_SPI_WORDSIZE16;
-		chip->read = chip->cs_change_per_word ?
-			bfin_spi_u16_cs_chg_reader : bfin_spi_u16_reader;
-		chip->write = chip->cs_change_per_word ?
-			bfin_spi_u16_cs_chg_writer : bfin_spi_u16_writer;
-		chip->duplex = chip->cs_change_per_word ?
-			bfin_spi_u16_cs_chg_duplex : bfin_spi_u16_duplex;
-		break;
-
-	default:
-		dev_err(&spi->dev, "%d bits_per_word is not supported\n",
-				chip->bits_per_word);
-		if (chip_info)
-			kfree(chip);
-		return -ENODEV;
-	}
-
 	dev_dbg(&spi->dev, "setup spi chip %s, width is %d, dma is %d\n",
-			spi->modalias, chip->width, chip->enable_dma);
+			spi->modalias, spi->bits_per_word, chip->enable_dma);
 	dev_dbg(&spi->dev, "ctl_reg is 0x%x, flag_reg is 0x%x\n",
 			chip->ctl_reg, chip->flag);
 
 	spi_set_ctldata(spi, chip);
 
 	dev_dbg(&spi->dev, "chip select number is %d\n", chip->chip_select_num);
-	if ((chip->chip_select_num > 0)
-		&& (chip->chip_select_num <= spi->master->num_chipselect))
-		peripheral_request(ssel[spi->master->bus_num]
-			[chip->chip_select_num-1], spi->modalias);
+	if (chip->chip_select_num < MAX_CTRL_CS) {
+		ret = peripheral_request(ssel[spi->master->bus_num]
+		                         [chip->chip_select_num-1], spi->modalias);
+		if (ret) {
+			dev_err(&spi->dev, "peripheral_request() error\n");
+			goto pin_error;
+		}
+	}
 
+	bfin_spi_cs_enable(drv_data, chip);
 	bfin_spi_cs_deactive(drv_data, chip);
 
 	return 0;
+
+ pin_error:
+	if (chip->chip_select_num >= MAX_CTRL_CS)
+		gpio_free(chip->cs_gpio);
+	else
+		peripheral_free(ssel[spi->master->bus_num]
+			[chip->chip_select_num - 1]);
+ error:
+	if (chip) {
+		if (drv_data->dma_requested)
+			free_dma(drv_data->dma_channel);
+		drv_data->dma_requested = 0;
+
+		kfree(chip);
+		/* prevent free 'chip' twice */
+		spi_set_ctldata(spi, NULL);
+	}
+
+	return ret;
 }
 
 /*
@@ -1152,28 +1155,30 @@
  */
 static void bfin_spi_cleanup(struct spi_device *spi)
 {
-	struct chip_data *chip = spi_get_ctldata(spi);
+	struct bfin_spi_slave_data *chip = spi_get_ctldata(spi);
+	struct bfin_spi_master_data *drv_data = spi_master_get_devdata(spi->master);
 
 	if (!chip)
 		return;
 
-	if ((chip->chip_select_num > 0)
-		&& (chip->chip_select_num <= spi->master->num_chipselect))
+	if (chip->chip_select_num < MAX_CTRL_CS) {
 		peripheral_free(ssel[spi->master->bus_num]
 					[chip->chip_select_num-1]);
-
-	if (chip->chip_select_num == 0)
+		bfin_spi_cs_disable(drv_data, chip);
+	} else
 		gpio_free(chip->cs_gpio);
 
 	kfree(chip);
+	/* prevent free 'chip' twice */
+	spi_set_ctldata(spi, NULL);
 }
 
-static inline int bfin_spi_init_queue(struct driver_data *drv_data)
+static inline int bfin_spi_init_queue(struct bfin_spi_master_data *drv_data)
 {
 	INIT_LIST_HEAD(&drv_data->queue);
 	spin_lock_init(&drv_data->lock);
 
-	drv_data->run = QUEUE_STOPPED;
+	drv_data->running = false;
 	drv_data->busy = 0;
 
 	/* init transfer tasklet */
@@ -1190,18 +1195,18 @@
 	return 0;
 }
 
-static inline int bfin_spi_start_queue(struct driver_data *drv_data)
+static inline int bfin_spi_start_queue(struct bfin_spi_master_data *drv_data)
 {
 	unsigned long flags;
 
 	spin_lock_irqsave(&drv_data->lock, flags);
 
-	if (drv_data->run == QUEUE_RUNNING || drv_data->busy) {
+	if (drv_data->running || drv_data->busy) {
 		spin_unlock_irqrestore(&drv_data->lock, flags);
 		return -EBUSY;
 	}
 
-	drv_data->run = QUEUE_RUNNING;
+	drv_data->running = true;
 	drv_data->cur_msg = NULL;
 	drv_data->cur_transfer = NULL;
 	drv_data->cur_chip = NULL;
@@ -1212,7 +1217,7 @@
 	return 0;
 }
 
-static inline int bfin_spi_stop_queue(struct driver_data *drv_data)
+static inline int bfin_spi_stop_queue(struct bfin_spi_master_data *drv_data)
 {
 	unsigned long flags;
 	unsigned limit = 500;
@@ -1226,7 +1231,7 @@
 	 * execution path (pump_messages) would be required to call wake_up or
 	 * friends on every SPI message. Do this instead
 	 */
-	drv_data->run = QUEUE_STOPPED;
+	drv_data->running = false;
 	while (!list_empty(&drv_data->queue) && drv_data->busy && limit--) {
 		spin_unlock_irqrestore(&drv_data->lock, flags);
 		msleep(10);
@@ -1241,7 +1246,7 @@
 	return status;
 }
 
-static inline int bfin_spi_destroy_queue(struct driver_data *drv_data)
+static inline int bfin_spi_destroy_queue(struct bfin_spi_master_data *drv_data)
 {
 	int status;
 
@@ -1259,14 +1264,14 @@
 	struct device *dev = &pdev->dev;
 	struct bfin5xx_spi_master *platform_info;
 	struct spi_master *master;
-	struct driver_data *drv_data = 0;
+	struct bfin_spi_master_data *drv_data;
 	struct resource *res;
 	int status = 0;
 
 	platform_info = dev->platform_data;
 
 	/* Allocate master with space for drv_data */
-	master = spi_alloc_master(dev, sizeof(struct driver_data) + 16);
+	master = spi_alloc_master(dev, sizeof(*drv_data));
 	if (!master) {
 		dev_err(&pdev->dev, "can not alloc spi_master\n");
 		return -ENOMEM;
@@ -1302,11 +1307,19 @@
 		goto out_error_ioremap;
 	}
 
-	drv_data->dma_channel = platform_get_irq(pdev, 0);
-	if (drv_data->dma_channel < 0) {
+	res = platform_get_resource(pdev, IORESOURCE_DMA, 0);
+	if (res == NULL) {
 		dev_err(dev, "No DMA channel specified\n");
 		status = -ENOENT;
-		goto out_error_no_dma_ch;
+		goto out_error_free_io;
+	}
+	drv_data->dma_channel = res->start;
+
+	drv_data->spi_irq = platform_get_irq(pdev, 0);
+	if (drv_data->spi_irq < 0) {
+		dev_err(dev, "No spi pio irq specified\n");
+		status = -ENOENT;
+		goto out_error_free_io;
 	}
 
 	/* Initial and start queue */
@@ -1328,6 +1341,12 @@
 		goto out_error_queue_alloc;
 	}
 
+	/* Reset SPI registers. If these registers were used by the boot loader,
+	 * the sky may fall on your head if you enable the dma controller.
+	 */
+	write_CTRL(drv_data, BIT_CTL_CPHA | BIT_CTL_MASTER);
+	write_FLAG(drv_data, 0xFF00);
+
 	/* Register with the SPI framework */
 	platform_set_drvdata(pdev, drv_data);
 	status = spi_register_master(master);
@@ -1343,7 +1362,7 @@
 
 out_error_queue_alloc:
 	bfin_spi_destroy_queue(drv_data);
-out_error_no_dma_ch:
+out_error_free_io:
 	iounmap((void *) drv_data->regs_base);
 out_error_ioremap:
 out_error_get_res:
@@ -1355,7 +1374,7 @@
 /* stop hardware and remove the driver */
 static int __devexit bfin_spi_remove(struct platform_device *pdev)
 {
-	struct driver_data *drv_data = platform_get_drvdata(pdev);
+	struct bfin_spi_master_data *drv_data = platform_get_drvdata(pdev);
 	int status = 0;
 
 	if (!drv_data)
@@ -1375,6 +1394,11 @@
 			free_dma(drv_data->dma_channel);
 	}
 
+	if (drv_data->irq_requested) {
+		free_irq(drv_data->spi_irq, drv_data);
+		drv_data->irq_requested = 0;
+	}
+
 	/* Disconnect from the SPI framework */
 	spi_unregister_master(drv_data->master);
 
@@ -1389,26 +1413,32 @@
 #ifdef CONFIG_PM
 static int bfin_spi_suspend(struct platform_device *pdev, pm_message_t state)
 {
-	struct driver_data *drv_data = platform_get_drvdata(pdev);
+	struct bfin_spi_master_data *drv_data = platform_get_drvdata(pdev);
 	int status = 0;
 
 	status = bfin_spi_stop_queue(drv_data);
 	if (status != 0)
 		return status;
 
-	/* stop hardware */
-	bfin_spi_disable(drv_data);
+	drv_data->ctrl_reg = read_CTRL(drv_data);
+	drv_data->flag_reg = read_FLAG(drv_data);
+
+	/*
+	 * reset SPI_CTL and SPI_FLG registers
+	 */
+	write_CTRL(drv_data, BIT_CTL_CPHA | BIT_CTL_MASTER);
+	write_FLAG(drv_data, 0xFF00);
 
 	return 0;
 }
 
 static int bfin_spi_resume(struct platform_device *pdev)
 {
-	struct driver_data *drv_data = platform_get_drvdata(pdev);
+	struct bfin_spi_master_data *drv_data = platform_get_drvdata(pdev);
 	int status = 0;
 
-	/* Enable the SPI interface */
-	bfin_spi_enable(drv_data);
+	write_CTRL(drv_data, drv_data->ctrl_reg);
+	write_FLAG(drv_data, drv_data->flag_reg);
 
 	/* Start the queue running */
 	status = bfin_spi_start_queue(drv_data);
@@ -1439,7 +1469,7 @@
 {
 	return platform_driver_probe(&bfin_spi_driver, bfin_spi_probe);
 }
-module_init(bfin_spi_init);
+subsys_initcall(bfin_spi_init);
 
 static void __exit bfin_spi_exit(void)
 {
diff --git a/drivers/spi/spi_fsl_espi.c b/drivers/spi/spi_fsl_espi.c
new file mode 100644
index 0000000..e3b4f645
--- /dev/null
+++ b/drivers/spi/spi_fsl_espi.c
@@ -0,0 +1,748 @@
+/*
+ * Freescale eSPI controller driver.
+ *
+ * Copyright 2010 Freescale Semiconductor, Inc.
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+#include <linux/module.h>
+#include <linux/delay.h>
+#include <linux/irq.h>
+#include <linux/spi/spi.h>
+#include <linux/platform_device.h>
+#include <linux/fsl_devices.h>
+#include <linux/mm.h>
+#include <linux/of.h>
+#include <linux/of_platform.h>
+#include <linux/of_spi.h>
+#include <linux/interrupt.h>
+#include <linux/err.h>
+#include <sysdev/fsl_soc.h>
+
+#include "spi_fsl_lib.h"
+
+/* eSPI Controller registers */
+struct fsl_espi_reg {
+	__be32 mode;		/* 0x000 - eSPI mode register */
+	__be32 event;		/* 0x004 - eSPI event register */
+	__be32 mask;		/* 0x008 - eSPI mask register */
+	__be32 command;		/* 0x00c - eSPI command register */
+	__be32 transmit;	/* 0x010 - eSPI transmit FIFO access register*/
+	__be32 receive;		/* 0x014 - eSPI receive FIFO access register*/
+	u8 res[8];		/* 0x018 - 0x01c reserved */
+	__be32 csmode[4];	/* 0x020 - 0x02c eSPI cs mode register */
+};
+
+struct fsl_espi_transfer {
+	const void *tx_buf;
+	void *rx_buf;
+	unsigned len;
+	unsigned n_tx;
+	unsigned n_rx;
+	unsigned actual_length;
+	int status;
+};
+
+/* eSPI Controller mode register definitions */
+#define SPMODE_ENABLE		(1 << 31)
+#define SPMODE_LOOP		(1 << 30)
+#define SPMODE_TXTHR(x)		((x) << 8)
+#define SPMODE_RXTHR(x)		((x) << 0)
+
+/* eSPI Controller CS mode register definitions */
+#define CSMODE_CI_INACTIVEHIGH	(1 << 31)
+#define CSMODE_CP_BEGIN_EDGECLK	(1 << 30)
+#define CSMODE_REV		(1 << 29)
+#define CSMODE_DIV16		(1 << 28)
+#define CSMODE_PM(x)		((x) << 24)
+#define CSMODE_POL_1		(1 << 20)
+#define CSMODE_LEN(x)		((x) << 16)
+#define CSMODE_BEF(x)		((x) << 12)
+#define CSMODE_AFT(x)		((x) << 8)
+#define CSMODE_CG(x)		((x) << 3)
+
+/* Default mode/csmode for eSPI controller */
+#define SPMODE_INIT_VAL (SPMODE_TXTHR(4) | SPMODE_RXTHR(3))
+#define CSMODE_INIT_VAL (CSMODE_POL_1 | CSMODE_BEF(0) \
+		| CSMODE_AFT(0) | CSMODE_CG(1))
+
+/* SPIE register values */
+#define	SPIE_NE		0x00000200	/* Not empty */
+#define	SPIE_NF		0x00000100	/* Not full */
+
+/* SPIM register values */
+#define	SPIM_NE		0x00000200	/* Not empty */
+#define	SPIM_NF		0x00000100	/* Not full */
+#define SPIE_RXCNT(reg)     ((reg >> 24) & 0x3F)
+#define SPIE_TXCNT(reg)     ((reg >> 16) & 0x3F)
+
+/* SPCOM register values */
+#define SPCOM_CS(x)		((x) << 30)
+#define SPCOM_TRANLEN(x)	((x) << 0)
+#define	SPCOM_TRANLEN_MAX	0xFFFF	/* Max transaction length */
+
+static void fsl_espi_change_mode(struct spi_device *spi)
+{
+	struct mpc8xxx_spi *mspi = spi_master_get_devdata(spi->master);
+	struct spi_mpc8xxx_cs *cs = spi->controller_state;
+	struct fsl_espi_reg *reg_base = mspi->reg_base;
+	__be32 __iomem *mode = &reg_base->csmode[spi->chip_select];
+	__be32 __iomem *espi_mode = &reg_base->mode;
+	u32 tmp;
+	unsigned long flags;
+
+	/* Turn off IRQs locally to minimize time that SPI is disabled. */
+	local_irq_save(flags);
+
+	/* Turn off SPI unit prior changing mode */
+	tmp = mpc8xxx_spi_read_reg(espi_mode);
+	mpc8xxx_spi_write_reg(espi_mode, tmp & ~SPMODE_ENABLE);
+	mpc8xxx_spi_write_reg(mode, cs->hw_mode);
+	mpc8xxx_spi_write_reg(espi_mode, tmp);
+
+	local_irq_restore(flags);
+}
+
+static u32 fsl_espi_tx_buf_lsb(struct mpc8xxx_spi *mpc8xxx_spi)
+{
+	u32 data;
+	u16 data_h;
+	u16 data_l;
+	const u32 *tx = mpc8xxx_spi->tx;
+
+	if (!tx)
+		return 0;
+
+	data = *tx++ << mpc8xxx_spi->tx_shift;
+	data_l = data & 0xffff;
+	data_h = (data >> 16) & 0xffff;
+	swab16s(&data_l);
+	swab16s(&data_h);
+	data = data_h | data_l;
+
+	mpc8xxx_spi->tx = tx;
+	return data;
+}
+
+static int fsl_espi_setup_transfer(struct spi_device *spi,
+					struct spi_transfer *t)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(spi->master);
+	int bits_per_word = 0;
+	u8 pm;
+	u32 hz = 0;
+	struct spi_mpc8xxx_cs *cs = spi->controller_state;
+
+	if (t) {
+		bits_per_word = t->bits_per_word;
+		hz = t->speed_hz;
+	}
+
+	/* spi_transfer level calls that work per-word */
+	if (!bits_per_word)
+		bits_per_word = spi->bits_per_word;
+
+	/* Make sure its a bit width we support [4..16] */
+	if ((bits_per_word < 4) || (bits_per_word > 16))
+		return -EINVAL;
+
+	if (!hz)
+		hz = spi->max_speed_hz;
+
+	cs->rx_shift = 0;
+	cs->tx_shift = 0;
+	cs->get_rx = mpc8xxx_spi_rx_buf_u32;
+	cs->get_tx = mpc8xxx_spi_tx_buf_u32;
+	if (bits_per_word <= 8) {
+		cs->rx_shift = 8 - bits_per_word;
+	} else if (bits_per_word <= 16) {
+		cs->rx_shift = 16 - bits_per_word;
+		if (spi->mode & SPI_LSB_FIRST)
+			cs->get_tx = fsl_espi_tx_buf_lsb;
+	} else {
+		return -EINVAL;
+	}
+
+	mpc8xxx_spi->rx_shift = cs->rx_shift;
+	mpc8xxx_spi->tx_shift = cs->tx_shift;
+	mpc8xxx_spi->get_rx = cs->get_rx;
+	mpc8xxx_spi->get_tx = cs->get_tx;
+
+	bits_per_word = bits_per_word - 1;
+
+	/* mask out bits we are going to set */
+	cs->hw_mode &= ~(CSMODE_LEN(0xF) | CSMODE_DIV16 | CSMODE_PM(0xF));
+
+	cs->hw_mode |= CSMODE_LEN(bits_per_word);
+
+	if ((mpc8xxx_spi->spibrg / hz) > 64) {
+		cs->hw_mode |= CSMODE_DIV16;
+		pm = (mpc8xxx_spi->spibrg - 1) / (hz * 64) + 1;
+
+		WARN_ONCE(pm > 16, "%s: Requested speed is too low: %d Hz. "
+			  "Will use %d Hz instead.\n", dev_name(&spi->dev),
+			  hz, mpc8xxx_spi->spibrg / 1024);
+		if (pm > 16)
+			pm = 16;
+	} else {
+		pm = (mpc8xxx_spi->spibrg - 1) / (hz * 4) + 1;
+	}
+	if (pm)
+		pm--;
+
+	cs->hw_mode |= CSMODE_PM(pm);
+
+	fsl_espi_change_mode(spi);
+	return 0;
+}
+
+static int fsl_espi_cpu_bufs(struct mpc8xxx_spi *mspi, struct spi_transfer *t,
+		unsigned int len)
+{
+	u32 word;
+	struct fsl_espi_reg *reg_base = mspi->reg_base;
+
+	mspi->count = len;
+
+	/* enable rx ints */
+	mpc8xxx_spi_write_reg(&reg_base->mask, SPIM_NE);
+
+	/* transmit word */
+	word = mspi->get_tx(mspi);
+	mpc8xxx_spi_write_reg(&reg_base->transmit, word);
+
+	return 0;
+}
+
+static int fsl_espi_bufs(struct spi_device *spi, struct spi_transfer *t)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(spi->master);
+	struct fsl_espi_reg *reg_base = mpc8xxx_spi->reg_base;
+	unsigned int len = t->len;
+	u8 bits_per_word;
+	int ret;
+
+	bits_per_word = spi->bits_per_word;
+	if (t->bits_per_word)
+		bits_per_word = t->bits_per_word;
+
+	mpc8xxx_spi->len = t->len;
+	len = roundup(len, 4) / 4;
+
+	mpc8xxx_spi->tx = t->tx_buf;
+	mpc8xxx_spi->rx = t->rx_buf;
+
+	INIT_COMPLETION(mpc8xxx_spi->done);
+
+	/* Set SPCOM[CS] and SPCOM[TRANLEN] field */
+	if ((t->len - 1) > SPCOM_TRANLEN_MAX) {
+		dev_err(mpc8xxx_spi->dev, "Transaction length (%d)"
+				" beyond the SPCOM[TRANLEN] field\n", t->len);
+		return -EINVAL;
+	}
+	mpc8xxx_spi_write_reg(&reg_base->command,
+		(SPCOM_CS(spi->chip_select) | SPCOM_TRANLEN(t->len - 1)));
+
+	ret = fsl_espi_cpu_bufs(mpc8xxx_spi, t, len);
+	if (ret)
+		return ret;
+
+	wait_for_completion(&mpc8xxx_spi->done);
+
+	/* disable rx ints */
+	mpc8xxx_spi_write_reg(&reg_base->mask, 0);
+
+	return mpc8xxx_spi->count;
+}
+
+static void fsl_espi_addr2cmd(unsigned int addr, u8 *cmd)
+{
+	if (cmd[1] && cmd[2] && cmd[3]) {
+		cmd[1] = (u8)(addr >> 16);
+		cmd[2] = (u8)(addr >> 8);
+		cmd[3] = (u8)(addr >> 0);
+	}
+}
+
+static unsigned int fsl_espi_cmd2addr(u8 *cmd)
+{
+	if (cmd[1] && cmd[2] && cmd[3])
+		return cmd[1] << 16 | cmd[2] << 8 | cmd[3] << 0;
+
+	return 0;
+}
+
+static void fsl_espi_do_trans(struct spi_message *m,
+				struct fsl_espi_transfer *tr)
+{
+	struct spi_device *spi = m->spi;
+	struct mpc8xxx_spi *mspi = spi_master_get_devdata(spi->master);
+	struct fsl_espi_transfer *espi_trans = tr;
+	struct spi_message message;
+	struct spi_transfer *t, *first, trans;
+	int status = 0;
+
+	spi_message_init(&message);
+	memset(&trans, 0, sizeof(trans));
+
+	first = list_first_entry(&m->transfers, struct spi_transfer,
+			transfer_list);
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if ((first->bits_per_word != t->bits_per_word) ||
+			(first->speed_hz != t->speed_hz)) {
+			espi_trans->status = -EINVAL;
+			dev_err(mspi->dev, "bits_per_word/speed_hz should be"
+					" same for the same SPI transfer\n");
+			return;
+		}
+
+		trans.speed_hz = t->speed_hz;
+		trans.bits_per_word = t->bits_per_word;
+		trans.delay_usecs = max(first->delay_usecs, t->delay_usecs);
+	}
+
+	trans.len = espi_trans->len;
+	trans.tx_buf = espi_trans->tx_buf;
+	trans.rx_buf = espi_trans->rx_buf;
+	spi_message_add_tail(&trans, &message);
+
+	list_for_each_entry(t, &message.transfers, transfer_list) {
+		if (t->bits_per_word || t->speed_hz) {
+			status = -EINVAL;
+
+			status = fsl_espi_setup_transfer(spi, t);
+			if (status < 0)
+				break;
+		}
+
+		if (t->len)
+			status = fsl_espi_bufs(spi, t);
+
+		if (status) {
+			status = -EMSGSIZE;
+			break;
+		}
+
+		if (t->delay_usecs)
+			udelay(t->delay_usecs);
+	}
+
+	espi_trans->status = status;
+	fsl_espi_setup_transfer(spi, NULL);
+}
+
+static void fsl_espi_cmd_trans(struct spi_message *m,
+				struct fsl_espi_transfer *trans, u8 *rx_buff)
+{
+	struct spi_transfer *t;
+	u8 *local_buf;
+	int i = 0;
+	struct fsl_espi_transfer *espi_trans = trans;
+
+	local_buf = kzalloc(SPCOM_TRANLEN_MAX, GFP_KERNEL);
+	if (!local_buf) {
+		espi_trans->status = -ENOMEM;
+		return;
+	}
+
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if (t->tx_buf) {
+			memcpy(local_buf + i, t->tx_buf, t->len);
+			i += t->len;
+		}
+	}
+
+	espi_trans->tx_buf = local_buf;
+	espi_trans->rx_buf = local_buf + espi_trans->n_tx;
+	fsl_espi_do_trans(m, espi_trans);
+
+	espi_trans->actual_length = espi_trans->len;
+	kfree(local_buf);
+}
+
+static void fsl_espi_rw_trans(struct spi_message *m,
+				struct fsl_espi_transfer *trans, u8 *rx_buff)
+{
+	struct fsl_espi_transfer *espi_trans = trans;
+	unsigned int n_tx = espi_trans->n_tx;
+	unsigned int n_rx = espi_trans->n_rx;
+	struct spi_transfer *t;
+	u8 *local_buf;
+	u8 *rx_buf = rx_buff;
+	unsigned int trans_len;
+	unsigned int addr;
+	int i, pos, loop;
+
+	local_buf = kzalloc(SPCOM_TRANLEN_MAX, GFP_KERNEL);
+	if (!local_buf) {
+		espi_trans->status = -ENOMEM;
+		return;
+	}
+
+	for (pos = 0, loop = 0; pos < n_rx; pos += trans_len, loop++) {
+		trans_len = n_rx - pos;
+		if (trans_len > SPCOM_TRANLEN_MAX - n_tx)
+			trans_len = SPCOM_TRANLEN_MAX - n_tx;
+
+		i = 0;
+		list_for_each_entry(t, &m->transfers, transfer_list) {
+			if (t->tx_buf) {
+				memcpy(local_buf + i, t->tx_buf, t->len);
+				i += t->len;
+			}
+		}
+
+		addr = fsl_espi_cmd2addr(local_buf);
+		addr += pos;
+		fsl_espi_addr2cmd(addr, local_buf);
+
+		espi_trans->n_tx = n_tx;
+		espi_trans->n_rx = trans_len;
+		espi_trans->len = trans_len + n_tx;
+		espi_trans->tx_buf = local_buf;
+		espi_trans->rx_buf = local_buf + n_tx;
+		fsl_espi_do_trans(m, espi_trans);
+
+		memcpy(rx_buf + pos, espi_trans->rx_buf + n_tx, trans_len);
+
+		if (loop > 0)
+			espi_trans->actual_length += espi_trans->len - n_tx;
+		else
+			espi_trans->actual_length += espi_trans->len;
+	}
+
+	kfree(local_buf);
+}
+
+static void fsl_espi_do_one_msg(struct spi_message *m)
+{
+	struct spi_transfer *t;
+	u8 *rx_buf = NULL;
+	unsigned int n_tx = 0;
+	unsigned int n_rx = 0;
+	struct fsl_espi_transfer espi_trans;
+
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if (t->tx_buf)
+			n_tx += t->len;
+		if (t->rx_buf) {
+			n_rx += t->len;
+			rx_buf = t->rx_buf;
+		}
+	}
+
+	espi_trans.n_tx = n_tx;
+	espi_trans.n_rx = n_rx;
+	espi_trans.len = n_tx + n_rx;
+	espi_trans.actual_length = 0;
+	espi_trans.status = 0;
+
+	if (!rx_buf)
+		fsl_espi_cmd_trans(m, &espi_trans, NULL);
+	else
+		fsl_espi_rw_trans(m, &espi_trans, rx_buf);
+
+	m->actual_length = espi_trans.actual_length;
+	m->status = espi_trans.status;
+	m->complete(m->context);
+}
+
+static int fsl_espi_setup(struct spi_device *spi)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi;
+	struct fsl_espi_reg *reg_base;
+	int retval;
+	u32 hw_mode;
+	u32 loop_mode;
+	struct spi_mpc8xxx_cs *cs = spi->controller_state;
+
+	if (!spi->max_speed_hz)
+		return -EINVAL;
+
+	if (!cs) {
+		cs = kzalloc(sizeof *cs, GFP_KERNEL);
+		if (!cs)
+			return -ENOMEM;
+		spi->controller_state = cs;
+	}
+
+	mpc8xxx_spi = spi_master_get_devdata(spi->master);
+	reg_base = mpc8xxx_spi->reg_base;
+
+	hw_mode = cs->hw_mode; /* Save orginal settings */
+	cs->hw_mode = mpc8xxx_spi_read_reg(
+			&reg_base->csmode[spi->chip_select]);
+	/* mask out bits we are going to set */
+	cs->hw_mode &= ~(CSMODE_CP_BEGIN_EDGECLK | CSMODE_CI_INACTIVEHIGH
+			 | CSMODE_REV);
+
+	if (spi->mode & SPI_CPHA)
+		cs->hw_mode |= CSMODE_CP_BEGIN_EDGECLK;
+	if (spi->mode & SPI_CPOL)
+		cs->hw_mode |= CSMODE_CI_INACTIVEHIGH;
+	if (!(spi->mode & SPI_LSB_FIRST))
+		cs->hw_mode |= CSMODE_REV;
+
+	/* Handle the loop mode */
+	loop_mode = mpc8xxx_spi_read_reg(&reg_base->mode);
+	loop_mode &= ~SPMODE_LOOP;
+	if (spi->mode & SPI_LOOP)
+		loop_mode |= SPMODE_LOOP;
+	mpc8xxx_spi_write_reg(&reg_base->mode, loop_mode);
+
+	retval = fsl_espi_setup_transfer(spi, NULL);
+	if (retval < 0) {
+		cs->hw_mode = hw_mode; /* Restore settings */
+		return retval;
+	}
+	return 0;
+}
+
+void fsl_espi_cpu_irq(struct mpc8xxx_spi *mspi, u32 events)
+{
+	struct fsl_espi_reg *reg_base = mspi->reg_base;
+
+	/* We need handle RX first */
+	if (events & SPIE_NE) {
+		u32 rx_data;
+
+		/* Spin until RX is done */
+		while (SPIE_RXCNT(events) < min(4, mspi->len)) {
+			cpu_relax();
+			events = mpc8xxx_spi_read_reg(&reg_base->event);
+		}
+		mspi->len -= 4;
+
+		rx_data = mpc8xxx_spi_read_reg(&reg_base->receive);
+
+		if (mspi->rx)
+			mspi->get_rx(rx_data, mspi);
+	}
+
+	if (!(events & SPIE_NF)) {
+		int ret;
+
+		/* spin until TX is done */
+		ret = spin_event_timeout(((events = mpc8xxx_spi_read_reg(
+				&reg_base->event)) & SPIE_NF) == 0, 1000, 0);
+		if (!ret) {
+			dev_err(mspi->dev, "tired waiting for SPIE_NF\n");
+			return;
+		}
+	}
+
+	/* Clear the events */
+	mpc8xxx_spi_write_reg(&reg_base->event, events);
+
+	mspi->count -= 1;
+	if (mspi->count) {
+		u32 word = mspi->get_tx(mspi);
+
+		mpc8xxx_spi_write_reg(&reg_base->transmit, word);
+	} else {
+		complete(&mspi->done);
+	}
+}
+
+static irqreturn_t fsl_espi_irq(s32 irq, void *context_data)
+{
+	struct mpc8xxx_spi *mspi = context_data;
+	struct fsl_espi_reg *reg_base = mspi->reg_base;
+	irqreturn_t ret = IRQ_NONE;
+	u32 events;
+
+	/* Get interrupt events(tx/rx) */
+	events = mpc8xxx_spi_read_reg(&reg_base->event);
+	if (events)
+		ret = IRQ_HANDLED;
+
+	dev_vdbg(mspi->dev, "%s: events %x\n", __func__, events);
+
+	fsl_espi_cpu_irq(mspi, events);
+
+	return ret;
+}
+
+static void fsl_espi_remove(struct mpc8xxx_spi *mspi)
+{
+	iounmap(mspi->reg_base);
+}
+
+static struct spi_master * __devinit fsl_espi_probe(struct device *dev,
+		struct resource *mem, unsigned int irq)
+{
+	struct fsl_spi_platform_data *pdata = dev->platform_data;
+	struct spi_master *master;
+	struct mpc8xxx_spi *mpc8xxx_spi;
+	struct fsl_espi_reg *reg_base;
+	u32 regval;
+	int i, ret = 0;
+
+	master = spi_alloc_master(dev, sizeof(struct mpc8xxx_spi));
+	if (!master) {
+		ret = -ENOMEM;
+		goto err;
+	}
+
+	dev_set_drvdata(dev, master);
+
+	ret = mpc8xxx_spi_probe(dev, mem, irq);
+	if (ret)
+		goto err_probe;
+
+	master->setup = fsl_espi_setup;
+
+	mpc8xxx_spi = spi_master_get_devdata(master);
+	mpc8xxx_spi->spi_do_one_msg = fsl_espi_do_one_msg;
+	mpc8xxx_spi->spi_remove = fsl_espi_remove;
+
+	mpc8xxx_spi->reg_base = ioremap(mem->start, resource_size(mem));
+	if (!mpc8xxx_spi->reg_base) {
+		ret = -ENOMEM;
+		goto err_probe;
+	}
+
+	reg_base = mpc8xxx_spi->reg_base;
+
+	/* Register for SPI Interrupt */
+	ret = request_irq(mpc8xxx_spi->irq, fsl_espi_irq,
+			  0, "fsl_espi", mpc8xxx_spi);
+	if (ret)
+		goto free_irq;
+
+	if (mpc8xxx_spi->flags & SPI_QE_CPU_MODE) {
+		mpc8xxx_spi->rx_shift = 16;
+		mpc8xxx_spi->tx_shift = 24;
+	}
+
+	/* SPI controller initializations */
+	mpc8xxx_spi_write_reg(&reg_base->mode, 0);
+	mpc8xxx_spi_write_reg(&reg_base->mask, 0);
+	mpc8xxx_spi_write_reg(&reg_base->command, 0);
+	mpc8xxx_spi_write_reg(&reg_base->event, 0xffffffff);
+
+	/* Init eSPI CS mode register */
+	for (i = 0; i < pdata->max_chipselect; i++)
+		mpc8xxx_spi_write_reg(&reg_base->csmode[i], CSMODE_INIT_VAL);
+
+	/* Enable SPI interface */
+	regval = pdata->initial_spmode | SPMODE_INIT_VAL | SPMODE_ENABLE;
+
+	mpc8xxx_spi_write_reg(&reg_base->mode, regval);
+
+	ret = spi_register_master(master);
+	if (ret < 0)
+		goto unreg_master;
+
+	dev_info(dev, "at 0x%p (irq = %d)\n", reg_base, mpc8xxx_spi->irq);
+
+	return master;
+
+unreg_master:
+	free_irq(mpc8xxx_spi->irq, mpc8xxx_spi);
+free_irq:
+	iounmap(mpc8xxx_spi->reg_base);
+err_probe:
+	spi_master_put(master);
+err:
+	return ERR_PTR(ret);
+}
+
+static int of_fsl_espi_get_chipselects(struct device *dev)
+{
+	struct device_node *np = dev->of_node;
+	struct fsl_spi_platform_data *pdata = dev->platform_data;
+	const u32 *prop;
+	int len;
+
+	prop = of_get_property(np, "fsl,espi-num-chipselects", &len);
+	if (!prop || len < sizeof(*prop)) {
+		dev_err(dev, "No 'fsl,espi-num-chipselects' property\n");
+		return -EINVAL;
+	}
+
+	pdata->max_chipselect = *prop;
+	pdata->cs_control = NULL;
+
+	return 0;
+}
+
+static int __devinit of_fsl_espi_probe(struct platform_device *ofdev,
+					const struct of_device_id *ofid)
+{
+	struct device *dev = &ofdev->dev;
+	struct device_node *np = ofdev->dev.of_node;
+	struct spi_master *master;
+	struct resource mem;
+	struct resource irq;
+	int ret = -ENOMEM;
+
+	ret = of_mpc8xxx_spi_probe(ofdev, ofid);
+	if (ret)
+		return ret;
+
+	ret = of_fsl_espi_get_chipselects(dev);
+	if (ret)
+		goto err;
+
+	ret = of_address_to_resource(np, 0, &mem);
+	if (ret)
+		goto err;
+
+	ret = of_irq_to_resource(np, 0, &irq);
+	if (!ret) {
+		ret = -EINVAL;
+		goto err;
+	}
+
+	master = fsl_espi_probe(dev, &mem, irq.start);
+	if (IS_ERR(master)) {
+		ret = PTR_ERR(master);
+		goto err;
+	}
+
+	return 0;
+
+err:
+	return ret;
+}
+
+static int __devexit of_fsl_espi_remove(struct platform_device *dev)
+{
+	return mpc8xxx_spi_remove(&dev->dev);
+}
+
+static const struct of_device_id of_fsl_espi_match[] = {
+	{ .compatible = "fsl,mpc8536-espi" },
+	{}
+};
+MODULE_DEVICE_TABLE(of, of_fsl_espi_match);
+
+static struct of_platform_driver fsl_espi_driver = {
+	.driver = {
+		.name = "fsl_espi",
+		.owner = THIS_MODULE,
+		.of_match_table = of_fsl_espi_match,
+	},
+	.probe		= of_fsl_espi_probe,
+	.remove		= __devexit_p(of_fsl_espi_remove),
+};
+
+static int __init fsl_espi_init(void)
+{
+	return of_register_platform_driver(&fsl_espi_driver);
+}
+module_init(fsl_espi_init);
+
+static void __exit fsl_espi_exit(void)
+{
+	of_unregister_platform_driver(&fsl_espi_driver);
+}
+module_exit(fsl_espi_exit);
+
+MODULE_AUTHOR("Mingkai Hu");
+MODULE_DESCRIPTION("Enhanced Freescale SPI Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi_fsl_lib.c b/drivers/spi/spi_fsl_lib.c
new file mode 100644
index 0000000..5cd741f
--- /dev/null
+++ b/drivers/spi/spi_fsl_lib.c
@@ -0,0 +1,237 @@
+/*
+ * Freescale SPI/eSPI controller driver library.
+ *
+ * Maintainer: Kumar Gala
+ *
+ * Copyright (C) 2006 Polycom, Inc.
+ *
+ * CPM SPI and QE buffer descriptors mode support:
+ * Copyright (c) 2009  MontaVista Software, Inc.
+ * Author: Anton Vorontsov <avorontsov@ru.mvista.com>
+ *
+ * Copyright 2010 Freescale Semiconductor, Inc.
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+#include <linux/kernel.h>
+#include <linux/interrupt.h>
+#include <linux/fsl_devices.h>
+#include <linux/dma-mapping.h>
+#include <linux/mm.h>
+#include <linux/of_platform.h>
+#include <linux/of_spi.h>
+#include <sysdev/fsl_soc.h>
+
+#include "spi_fsl_lib.h"
+
+#define MPC8XXX_SPI_RX_BUF(type) 					  \
+void mpc8xxx_spi_rx_buf_##type(u32 data, struct mpc8xxx_spi *mpc8xxx_spi) \
+{									  \
+	type *rx = mpc8xxx_spi->rx;					  \
+	*rx++ = (type)(data >> mpc8xxx_spi->rx_shift);			  \
+	mpc8xxx_spi->rx = rx;						  \
+}
+
+#define MPC8XXX_SPI_TX_BUF(type)				\
+u32 mpc8xxx_spi_tx_buf_##type(struct mpc8xxx_spi *mpc8xxx_spi)	\
+{								\
+	u32 data;						\
+	const type *tx = mpc8xxx_spi->tx;			\
+	if (!tx)						\
+		return 0;					\
+	data = *tx++ << mpc8xxx_spi->tx_shift;			\
+	mpc8xxx_spi->tx = tx;					\
+	return data;						\
+}
+
+MPC8XXX_SPI_RX_BUF(u8)
+MPC8XXX_SPI_RX_BUF(u16)
+MPC8XXX_SPI_RX_BUF(u32)
+MPC8XXX_SPI_TX_BUF(u8)
+MPC8XXX_SPI_TX_BUF(u16)
+MPC8XXX_SPI_TX_BUF(u32)
+
+struct mpc8xxx_spi_probe_info *to_of_pinfo(struct fsl_spi_platform_data *pdata)
+{
+	return container_of(pdata, struct mpc8xxx_spi_probe_info, pdata);
+}
+
+void mpc8xxx_spi_work(struct work_struct *work)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi = container_of(work, struct mpc8xxx_spi,
+						       work);
+
+	spin_lock_irq(&mpc8xxx_spi->lock);
+	while (!list_empty(&mpc8xxx_spi->queue)) {
+		struct spi_message *m = container_of(mpc8xxx_spi->queue.next,
+						   struct spi_message, queue);
+
+		list_del_init(&m->queue);
+		spin_unlock_irq(&mpc8xxx_spi->lock);
+
+		if (mpc8xxx_spi->spi_do_one_msg)
+			mpc8xxx_spi->spi_do_one_msg(m);
+
+		spin_lock_irq(&mpc8xxx_spi->lock);
+	}
+	spin_unlock_irq(&mpc8xxx_spi->lock);
+}
+
+int mpc8xxx_spi_transfer(struct spi_device *spi,
+				struct spi_message *m)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(spi->master);
+	unsigned long flags;
+
+	m->actual_length = 0;
+	m->status = -EINPROGRESS;
+
+	spin_lock_irqsave(&mpc8xxx_spi->lock, flags);
+	list_add_tail(&m->queue, &mpc8xxx_spi->queue);
+	queue_work(mpc8xxx_spi->workqueue, &mpc8xxx_spi->work);
+	spin_unlock_irqrestore(&mpc8xxx_spi->lock, flags);
+
+	return 0;
+}
+
+void mpc8xxx_spi_cleanup(struct spi_device *spi)
+{
+	kfree(spi->controller_state);
+}
+
+const char *mpc8xxx_spi_strmode(unsigned int flags)
+{
+	if (flags & SPI_QE_CPU_MODE) {
+		return "QE CPU";
+	} else if (flags & SPI_CPM_MODE) {
+		if (flags & SPI_QE)
+			return "QE";
+		else if (flags & SPI_CPM2)
+			return "CPM2";
+		else
+			return "CPM1";
+	}
+	return "CPU";
+}
+
+int mpc8xxx_spi_probe(struct device *dev, struct resource *mem,
+			unsigned int irq)
+{
+	struct fsl_spi_platform_data *pdata = dev->platform_data;
+	struct spi_master *master;
+	struct mpc8xxx_spi *mpc8xxx_spi;
+	int ret = 0;
+
+	master = dev_get_drvdata(dev);
+
+	/* the spi->mode bits understood by this driver: */
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH
+			| SPI_LSB_FIRST | SPI_LOOP;
+
+	master->transfer = mpc8xxx_spi_transfer;
+	master->cleanup = mpc8xxx_spi_cleanup;
+	master->dev.of_node = dev->of_node;
+
+	mpc8xxx_spi = spi_master_get_devdata(master);
+	mpc8xxx_spi->dev = dev;
+	mpc8xxx_spi->get_rx = mpc8xxx_spi_rx_buf_u8;
+	mpc8xxx_spi->get_tx = mpc8xxx_spi_tx_buf_u8;
+	mpc8xxx_spi->flags = pdata->flags;
+	mpc8xxx_spi->spibrg = pdata->sysclk;
+	mpc8xxx_spi->irq = irq;
+
+	mpc8xxx_spi->rx_shift = 0;
+	mpc8xxx_spi->tx_shift = 0;
+
+	init_completion(&mpc8xxx_spi->done);
+
+	master->bus_num = pdata->bus_num;
+	master->num_chipselect = pdata->max_chipselect;
+
+	spin_lock_init(&mpc8xxx_spi->lock);
+	init_completion(&mpc8xxx_spi->done);
+	INIT_WORK(&mpc8xxx_spi->work, mpc8xxx_spi_work);
+	INIT_LIST_HEAD(&mpc8xxx_spi->queue);
+
+	mpc8xxx_spi->workqueue = create_singlethread_workqueue(
+		dev_name(master->dev.parent));
+	if (mpc8xxx_spi->workqueue == NULL) {
+		ret = -EBUSY;
+		goto err;
+	}
+
+	return 0;
+
+err:
+	return ret;
+}
+
+int __devexit mpc8xxx_spi_remove(struct device *dev)
+{
+	struct mpc8xxx_spi *mpc8xxx_spi;
+	struct spi_master *master;
+
+	master = dev_get_drvdata(dev);
+	mpc8xxx_spi = spi_master_get_devdata(master);
+
+	flush_workqueue(mpc8xxx_spi->workqueue);
+	destroy_workqueue(mpc8xxx_spi->workqueue);
+	spi_unregister_master(master);
+
+	free_irq(mpc8xxx_spi->irq, mpc8xxx_spi);
+
+	if (mpc8xxx_spi->spi_remove)
+		mpc8xxx_spi->spi_remove(mpc8xxx_spi);
+
+	return 0;
+}
+
+int __devinit of_mpc8xxx_spi_probe(struct platform_device *ofdev,
+					const struct of_device_id *ofid)
+{
+	struct device *dev = &ofdev->dev;
+	struct device_node *np = ofdev->dev.of_node;
+	struct mpc8xxx_spi_probe_info *pinfo;
+	struct fsl_spi_platform_data *pdata;
+	const void *prop;
+	int ret = -ENOMEM;
+
+	pinfo = kzalloc(sizeof(*pinfo), GFP_KERNEL);
+	if (!pinfo)
+		return -ENOMEM;
+
+	pdata = &pinfo->pdata;
+	dev->platform_data = pdata;
+
+	/* Allocate bus num dynamically. */
+	pdata->bus_num = -1;
+
+	/* SPI controller is either clocked from QE or SoC clock. */
+	pdata->sysclk = get_brgfreq();
+	if (pdata->sysclk == -1) {
+		pdata->sysclk = fsl_get_sys_freq();
+		if (pdata->sysclk == -1) {
+			ret = -ENODEV;
+			goto err;
+		}
+	}
+
+	prop = of_get_property(np, "mode", NULL);
+	if (prop && !strcmp(prop, "cpu-qe"))
+		pdata->flags = SPI_QE_CPU_MODE;
+	else if (prop && !strcmp(prop, "qe"))
+		pdata->flags = SPI_CPM_MODE | SPI_QE;
+	else if (of_device_is_compatible(np, "fsl,cpm2-spi"))
+		pdata->flags = SPI_CPM_MODE | SPI_CPM2;
+	else if (of_device_is_compatible(np, "fsl,cpm1-spi"))
+		pdata->flags = SPI_CPM_MODE | SPI_CPM1;
+
+	return 0;
+
+err:
+	kfree(pinfo);
+	return ret;
+}
diff --git a/drivers/spi/spi_fsl_lib.h b/drivers/spi/spi_fsl_lib.h
new file mode 100644
index 0000000..281e060
--- /dev/null
+++ b/drivers/spi/spi_fsl_lib.h
@@ -0,0 +1,124 @@
+/*
+ * Freescale SPI/eSPI controller driver library.
+ *
+ * Maintainer: Kumar Gala
+ *
+ * Copyright 2010 Freescale Semiconductor, Inc.
+ * Copyright (C) 2006 Polycom, Inc.
+ *
+ * CPM SPI and QE buffer descriptors mode support:
+ * Copyright (c) 2009  MontaVista Software, Inc.
+ * Author: Anton Vorontsov <avorontsov@ru.mvista.com>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+#ifndef __SPI_FSL_LIB_H__
+#define __SPI_FSL_LIB_H__
+
+#include <asm/io.h>
+
+/* SPI/eSPI Controller driver's private data. */
+struct mpc8xxx_spi {
+	struct device *dev;
+	void *reg_base;
+
+	/* rx & tx bufs from the spi_transfer */
+	const void *tx;
+	void *rx;
+#ifdef CONFIG_SPI_FSL_ESPI
+	int len;
+#endif
+
+	int subblock;
+	struct spi_pram __iomem *pram;
+	struct cpm_buf_desc __iomem *tx_bd;
+	struct cpm_buf_desc __iomem *rx_bd;
+
+	struct spi_transfer *xfer_in_progress;
+
+	/* dma addresses for CPM transfers */
+	dma_addr_t tx_dma;
+	dma_addr_t rx_dma;
+	bool map_tx_dma;
+	bool map_rx_dma;
+
+	dma_addr_t dma_dummy_tx;
+	dma_addr_t dma_dummy_rx;
+
+	/* functions to deal with different sized buffers */
+	void (*get_rx) (u32 rx_data, struct mpc8xxx_spi *);
+	u32(*get_tx) (struct mpc8xxx_spi *);
+
+	/* hooks for different controller driver */
+	void (*spi_do_one_msg) (struct spi_message *m);
+	void (*spi_remove) (struct mpc8xxx_spi *mspi);
+
+	unsigned int count;
+	unsigned int irq;
+
+	unsigned nsecs;		/* (clock cycle time)/2 */
+
+	u32 spibrg;		/* SPIBRG input clock */
+	u32 rx_shift;		/* RX data reg shift when in qe mode */
+	u32 tx_shift;		/* TX data reg shift when in qe mode */
+
+	unsigned int flags;
+
+	struct workqueue_struct *workqueue;
+	struct work_struct work;
+
+	struct list_head queue;
+	spinlock_t lock;
+
+	struct completion done;
+};
+
+struct spi_mpc8xxx_cs {
+	/* functions to deal with different sized buffers */
+	void (*get_rx) (u32 rx_data, struct mpc8xxx_spi *);
+	u32 (*get_tx) (struct mpc8xxx_spi *);
+	u32 rx_shift;		/* RX data reg shift when in qe mode */
+	u32 tx_shift;		/* TX data reg shift when in qe mode */
+	u32 hw_mode;		/* Holds HW mode register settings */
+};
+
+static inline void mpc8xxx_spi_write_reg(__be32 __iomem *reg, u32 val)
+{
+	out_be32(reg, val);
+}
+
+static inline u32 mpc8xxx_spi_read_reg(__be32 __iomem *reg)
+{
+	return in_be32(reg);
+}
+
+struct mpc8xxx_spi_probe_info {
+	struct fsl_spi_platform_data pdata;
+	int *gpios;
+	bool *alow_flags;
+};
+
+extern u32 mpc8xxx_spi_tx_buf_u8(struct mpc8xxx_spi *mpc8xxx_spi);
+extern u32 mpc8xxx_spi_tx_buf_u16(struct mpc8xxx_spi *mpc8xxx_spi);
+extern u32 mpc8xxx_spi_tx_buf_u32(struct mpc8xxx_spi *mpc8xxx_spi);
+extern void mpc8xxx_spi_rx_buf_u8(u32 data, struct mpc8xxx_spi *mpc8xxx_spi);
+extern void mpc8xxx_spi_rx_buf_u16(u32 data, struct mpc8xxx_spi *mpc8xxx_spi);
+extern void mpc8xxx_spi_rx_buf_u32(u32 data, struct mpc8xxx_spi *mpc8xxx_spi);
+
+extern struct mpc8xxx_spi_probe_info *to_of_pinfo(
+		struct fsl_spi_platform_data *pdata);
+extern int mpc8xxx_spi_bufs(struct mpc8xxx_spi *mspi,
+		struct spi_transfer *t, unsigned int len);
+extern int mpc8xxx_spi_transfer(struct spi_device *spi, struct spi_message *m);
+extern void mpc8xxx_spi_cleanup(struct spi_device *spi);
+extern const char *mpc8xxx_spi_strmode(unsigned int flags);
+extern int mpc8xxx_spi_probe(struct device *dev, struct resource *mem,
+		unsigned int irq);
+extern int mpc8xxx_spi_remove(struct device *dev);
+extern int of_mpc8xxx_spi_probe(struct platform_device *ofdev,
+				const struct of_device_id *ofid);
+
+#endif /* __SPI_FSL_LIB_H__ */
diff --git a/drivers/spi/spi_mpc8xxx.c b/drivers/spi/spi_fsl_spi.c
similarity index 64%
rename from drivers/spi/spi_mpc8xxx.c
rename to drivers/spi/spi_fsl_spi.c
index 1dd86b8..7ca52d3 100644
--- a/drivers/spi/spi_mpc8xxx.c
+++ b/drivers/spi/spi_fsl_spi.c
@@ -1,9 +1,10 @@
 /*
- * MPC8xxx SPI controller driver.
+ * Freescale SPI controller driver.
  *
  * Maintainer: Kumar Gala
  *
  * Copyright (C) 2006 Polycom, Inc.
+ * Copyright 2010 Freescale Semiconductor, Inc.
  *
  * CPM SPI and QE buffer descriptors mode support:
  * Copyright (c) 2009  MontaVista Software, Inc.
@@ -15,18 +16,11 @@
  * option) any later version.
  */
 #include <linux/module.h>
-#include <linux/init.h>
 #include <linux/types.h>
 #include <linux/kernel.h>
-#include <linux/bug.h>
-#include <linux/errno.h>
-#include <linux/err.h>
-#include <linux/io.h>
-#include <linux/completion.h>
 #include <linux/interrupt.h>
 #include <linux/delay.h>
 #include <linux/irq.h>
-#include <linux/device.h>
 #include <linux/spi/spi.h>
 #include <linux/spi/spi_bitbang.h>
 #include <linux/platform_device.h>
@@ -38,12 +32,12 @@
 #include <linux/of_platform.h>
 #include <linux/gpio.h>
 #include <linux/of_gpio.h>
-#include <linux/slab.h>
 
 #include <sysdev/fsl_soc.h>
 #include <asm/cpm.h>
 #include <asm/qe.h>
-#include <asm/irq.h>
+
+#include "spi_fsl_lib.h"
 
 /* CPM1 and CPM2 are mutually exclusive. */
 #ifdef CONFIG_CPM1
@@ -55,7 +49,7 @@
 #endif
 
 /* SPI Controller registers */
-struct mpc8xxx_spi_reg {
+struct fsl_spi_reg {
 	u8 res1[0x20];
 	__be32 mode;
 	__be32 event;
@@ -80,7 +74,7 @@
 
 /*
  * Default for SPI Mode:
- * 	SPI MODE 0 (inactive low, phase middle, MSB, 8-bit length, slow clk
+ *	SPI MODE 0 (inactive low, phase middle, MSB, 8-bit length, slow clk
  */
 #define	SPMODE_INIT_VAL (SPMODE_CI_INACTIVEHIGH | SPMODE_DIV16 | SPMODE_REV | \
 			 SPMODE_MS | SPMODE_LEN(7) | SPMODE_PM(0xf))
@@ -102,112 +96,16 @@
 #define	SPI_PRAM_SIZE	0x100
 #define	SPI_MRBLR	((unsigned int)PAGE_SIZE)
 
-/* SPI Controller driver's private data. */
-struct mpc8xxx_spi {
-	struct device *dev;
-	struct mpc8xxx_spi_reg __iomem *base;
+static void *fsl_dummy_rx;
+static DEFINE_MUTEX(fsl_dummy_rx_lock);
+static int fsl_dummy_rx_refcnt;
 
-	/* rx & tx bufs from the spi_transfer */
-	const void *tx;
-	void *rx;
-
-	int subblock;
-	struct spi_pram __iomem *pram;
-	struct cpm_buf_desc __iomem *tx_bd;
-	struct cpm_buf_desc __iomem *rx_bd;
-
-	struct spi_transfer *xfer_in_progress;
-
-	/* dma addresses for CPM transfers */
-	dma_addr_t tx_dma;
-	dma_addr_t rx_dma;
-	bool map_tx_dma;
-	bool map_rx_dma;
-
-	dma_addr_t dma_dummy_tx;
-	dma_addr_t dma_dummy_rx;
-
-	/* functions to deal with different sized buffers */
-	void (*get_rx) (u32 rx_data, struct mpc8xxx_spi *);
-	u32(*get_tx) (struct mpc8xxx_spi *);
-
-	unsigned int count;
-	unsigned int irq;
-
-	unsigned nsecs;		/* (clock cycle time)/2 */
-
-	u32 spibrg;		/* SPIBRG input clock */
-	u32 rx_shift;		/* RX data reg shift when in qe mode */
-	u32 tx_shift;		/* TX data reg shift when in qe mode */
-
-	unsigned int flags;
-
-	struct workqueue_struct *workqueue;
-	struct work_struct work;
-
-	struct list_head queue;
-	spinlock_t lock;
-
-	struct completion done;
-};
-
-static void *mpc8xxx_dummy_rx;
-static DEFINE_MUTEX(mpc8xxx_dummy_rx_lock);
-static int mpc8xxx_dummy_rx_refcnt;
-
-struct spi_mpc8xxx_cs {
-	/* functions to deal with different sized buffers */
-	void (*get_rx) (u32 rx_data, struct mpc8xxx_spi *);
-	u32 (*get_tx) (struct mpc8xxx_spi *);
-	u32 rx_shift;		/* RX data reg shift when in qe mode */
-	u32 tx_shift;		/* TX data reg shift when in qe mode */
-	u32 hw_mode;		/* Holds HW mode register settings */
-};
-
-static inline void mpc8xxx_spi_write_reg(__be32 __iomem *reg, u32 val)
-{
-	out_be32(reg, val);
-}
-
-static inline u32 mpc8xxx_spi_read_reg(__be32 __iomem *reg)
-{
-	return in_be32(reg);
-}
-
-#define MPC83XX_SPI_RX_BUF(type) 					  \
-static									  \
-void mpc8xxx_spi_rx_buf_##type(u32 data, struct mpc8xxx_spi *mpc8xxx_spi) \
-{									  \
-	type *rx = mpc8xxx_spi->rx;					  \
-	*rx++ = (type)(data >> mpc8xxx_spi->rx_shift);			  \
-	mpc8xxx_spi->rx = rx;						  \
-}
-
-#define MPC83XX_SPI_TX_BUF(type)				\
-static								\
-u32 mpc8xxx_spi_tx_buf_##type(struct mpc8xxx_spi *mpc8xxx_spi)	\
-{								\
-	u32 data;						\
-	const type *tx = mpc8xxx_spi->tx;			\
-	if (!tx)						\
-		return 0;					\
-	data = *tx++ << mpc8xxx_spi->tx_shift;			\
-	mpc8xxx_spi->tx = tx;					\
-	return data;						\
-}
-
-MPC83XX_SPI_RX_BUF(u8)
-MPC83XX_SPI_RX_BUF(u16)
-MPC83XX_SPI_RX_BUF(u32)
-MPC83XX_SPI_TX_BUF(u8)
-MPC83XX_SPI_TX_BUF(u16)
-MPC83XX_SPI_TX_BUF(u32)
-
-static void mpc8xxx_spi_change_mode(struct spi_device *spi)
+static void fsl_spi_change_mode(struct spi_device *spi)
 {
 	struct mpc8xxx_spi *mspi = spi_master_get_devdata(spi->master);
 	struct spi_mpc8xxx_cs *cs = spi->controller_state;
-	__be32 __iomem *mode = &mspi->base->mode;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
+	__be32 __iomem *mode = &reg_base->mode;
 	unsigned long flags;
 
 	if (cs->hw_mode == mpc8xxx_spi_read_reg(mode))
@@ -238,7 +136,7 @@
 	local_irq_restore(flags);
 }
 
-static void mpc8xxx_spi_chipselect(struct spi_device *spi, int value)
+static void fsl_spi_chipselect(struct spi_device *spi, int value)
 {
 	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(spi->master);
 	struct fsl_spi_platform_data *pdata = spi->dev.parent->platform_data;
@@ -256,18 +154,17 @@
 		mpc8xxx_spi->get_rx = cs->get_rx;
 		mpc8xxx_spi->get_tx = cs->get_tx;
 
-		mpc8xxx_spi_change_mode(spi);
+		fsl_spi_change_mode(spi);
 
 		if (pdata->cs_control)
 			pdata->cs_control(spi, pol);
 	}
 }
 
-static int
-mspi_apply_cpu_mode_quirks(struct spi_mpc8xxx_cs *cs,
-			   struct spi_device *spi,
-			   struct mpc8xxx_spi *mpc8xxx_spi,
-			   int bits_per_word)
+static int mspi_apply_cpu_mode_quirks(struct spi_mpc8xxx_cs *cs,
+				struct spi_device *spi,
+				struct mpc8xxx_spi *mpc8xxx_spi,
+				int bits_per_word)
 {
 	cs->rx_shift = 0;
 	cs->tx_shift = 0;
@@ -307,10 +204,9 @@
 	return bits_per_word;
 }
 
-static int
-mspi_apply_qe_mode_quirks(struct spi_mpc8xxx_cs *cs,
-			  struct spi_device *spi,
-			  int bits_per_word)
+static int mspi_apply_qe_mode_quirks(struct spi_mpc8xxx_cs *cs,
+				struct spi_device *spi,
+				int bits_per_word)
 {
 	/* QE uses Little Endian for words > 8
 	 * so transform all words > 8 into 8 bits
@@ -326,13 +222,13 @@
 	return bits_per_word;
 }
 
-static
-int mpc8xxx_spi_setup_transfer(struct spi_device *spi, struct spi_transfer *t)
+static int fsl_spi_setup_transfer(struct spi_device *spi,
+					struct spi_transfer *t)
 {
 	struct mpc8xxx_spi *mpc8xxx_spi;
-	int bits_per_word;
+	int bits_per_word = 0;
 	u8 pm;
-	u32 hz;
+	u32 hz = 0;
 	struct spi_mpc8xxx_cs	*cs = spi->controller_state;
 
 	mpc8xxx_spi = spi_master_get_devdata(spi->master);
@@ -340,9 +236,6 @@
 	if (t) {
 		bits_per_word = t->bits_per_word;
 		hz = t->speed_hz;
-	} else {
-		bits_per_word = 0;
-		hz = 0;
 	}
 
 	/* spi_transfer level calls that work per-word */
@@ -388,23 +281,25 @@
 			  hz, mpc8xxx_spi->spibrg / 1024);
 		if (pm > 16)
 			pm = 16;
-	} else
+	} else {
 		pm = (mpc8xxx_spi->spibrg - 1) / (hz * 4) + 1;
+	}
 	if (pm)
 		pm--;
 
 	cs->hw_mode |= SPMODE_PM(pm);
 
-	mpc8xxx_spi_change_mode(spi);
+	fsl_spi_change_mode(spi);
 	return 0;
 }
 
-static void mpc8xxx_spi_cpm_bufs_start(struct mpc8xxx_spi *mspi)
+static void fsl_spi_cpm_bufs_start(struct mpc8xxx_spi *mspi)
 {
 	struct cpm_buf_desc __iomem *tx_bd = mspi->tx_bd;
 	struct cpm_buf_desc __iomem *rx_bd = mspi->rx_bd;
 	unsigned int xfer_len = min(mspi->count, SPI_MRBLR);
 	unsigned int xfer_ofs;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
 
 	xfer_ofs = mspi->xfer_in_progress->len - mspi->count;
 
@@ -424,13 +319,14 @@
 				 BD_SC_LAST);
 
 	/* start transfer */
-	mpc8xxx_spi_write_reg(&mspi->base->command, SPCOM_STR);
+	mpc8xxx_spi_write_reg(&reg_base->command, SPCOM_STR);
 }
 
-static int mpc8xxx_spi_cpm_bufs(struct mpc8xxx_spi *mspi,
+static int fsl_spi_cpm_bufs(struct mpc8xxx_spi *mspi,
 				struct spi_transfer *t, bool is_dma_mapped)
 {
 	struct device *dev = mspi->dev;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
 
 	if (is_dma_mapped) {
 		mspi->map_tx_dma = 0;
@@ -475,13 +371,13 @@
 	}
 
 	/* enable rx ints */
-	mpc8xxx_spi_write_reg(&mspi->base->mask, SPIE_RXB);
+	mpc8xxx_spi_write_reg(&reg_base->mask, SPIE_RXB);
 
 	mspi->xfer_in_progress = t;
 	mspi->count = t->len;
 
 	/* start CPM transfers */
-	mpc8xxx_spi_cpm_bufs_start(mspi);
+	fsl_spi_cpm_bufs_start(mspi);
 
 	return 0;
 
@@ -491,7 +387,7 @@
 	return -ENOMEM;
 }
 
-static void mpc8xxx_spi_cpm_bufs_complete(struct mpc8xxx_spi *mspi)
+static void fsl_spi_cpm_bufs_complete(struct mpc8xxx_spi *mspi)
 {
 	struct device *dev = mspi->dev;
 	struct spi_transfer *t = mspi->xfer_in_progress;
@@ -503,31 +399,34 @@
 	mspi->xfer_in_progress = NULL;
 }
 
-static int mpc8xxx_spi_cpu_bufs(struct mpc8xxx_spi *mspi,
+static int fsl_spi_cpu_bufs(struct mpc8xxx_spi *mspi,
 				struct spi_transfer *t, unsigned int len)
 {
 	u32 word;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
 
 	mspi->count = len;
 
 	/* enable rx ints */
-	mpc8xxx_spi_write_reg(&mspi->base->mask, SPIM_NE);
+	mpc8xxx_spi_write_reg(&reg_base->mask, SPIM_NE);
 
 	/* transmit word */
 	word = mspi->get_tx(mspi);
-	mpc8xxx_spi_write_reg(&mspi->base->transmit, word);
+	mpc8xxx_spi_write_reg(&reg_base->transmit, word);
 
 	return 0;
 }
 
-static int mpc8xxx_spi_bufs(struct spi_device *spi, struct spi_transfer *t,
+static int fsl_spi_bufs(struct spi_device *spi, struct spi_transfer *t,
 			    bool is_dma_mapped)
 {
 	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(spi->master);
+	struct fsl_spi_reg *reg_base;
 	unsigned int len = t->len;
 	u8 bits_per_word;
 	int ret;
 
+	reg_base = mpc8xxx_spi->reg_base;
 	bits_per_word = spi->bits_per_word;
 	if (t->bits_per_word)
 		bits_per_word = t->bits_per_word;
@@ -551,24 +450,24 @@
 	INIT_COMPLETION(mpc8xxx_spi->done);
 
 	if (mpc8xxx_spi->flags & SPI_CPM_MODE)
-		ret = mpc8xxx_spi_cpm_bufs(mpc8xxx_spi, t, is_dma_mapped);
+		ret = fsl_spi_cpm_bufs(mpc8xxx_spi, t, is_dma_mapped);
 	else
-		ret = mpc8xxx_spi_cpu_bufs(mpc8xxx_spi, t, len);
+		ret = fsl_spi_cpu_bufs(mpc8xxx_spi, t, len);
 	if (ret)
 		return ret;
 
 	wait_for_completion(&mpc8xxx_spi->done);
 
 	/* disable rx ints */
-	mpc8xxx_spi_write_reg(&mpc8xxx_spi->base->mask, 0);
+	mpc8xxx_spi_write_reg(&reg_base->mask, 0);
 
 	if (mpc8xxx_spi->flags & SPI_CPM_MODE)
-		mpc8xxx_spi_cpm_bufs_complete(mpc8xxx_spi);
+		fsl_spi_cpm_bufs_complete(mpc8xxx_spi);
 
 	return mpc8xxx_spi->count;
 }
 
-static void mpc8xxx_spi_do_one_msg(struct spi_message *m)
+static void fsl_spi_do_one_msg(struct spi_message *m)
 {
 	struct spi_device *spi = m->spi;
 	struct spi_transfer *t;
@@ -584,18 +483,18 @@
 			status = -EINVAL;
 
 			if (cs_change)
-				status = mpc8xxx_spi_setup_transfer(spi, t);
+				status = fsl_spi_setup_transfer(spi, t);
 			if (status < 0)
 				break;
 		}
 
 		if (cs_change) {
-			mpc8xxx_spi_chipselect(spi, BITBANG_CS_ACTIVE);
+			fsl_spi_chipselect(spi, BITBANG_CS_ACTIVE);
 			ndelay(nsecs);
 		}
 		cs_change = t->cs_change;
 		if (t->len)
-			status = mpc8xxx_spi_bufs(spi, t, m->is_dma_mapped);
+			status = fsl_spi_bufs(spi, t, m->is_dma_mapped);
 		if (status) {
 			status = -EMSGSIZE;
 			break;
@@ -607,7 +506,7 @@
 
 		if (cs_change) {
 			ndelay(nsecs);
-			mpc8xxx_spi_chipselect(spi, BITBANG_CS_INACTIVE);
+			fsl_spi_chipselect(spi, BITBANG_CS_INACTIVE);
 			ndelay(nsecs);
 		}
 	}
@@ -617,35 +516,16 @@
 
 	if (status || !cs_change) {
 		ndelay(nsecs);
-		mpc8xxx_spi_chipselect(spi, BITBANG_CS_INACTIVE);
+		fsl_spi_chipselect(spi, BITBANG_CS_INACTIVE);
 	}
 
-	mpc8xxx_spi_setup_transfer(spi, NULL);
+	fsl_spi_setup_transfer(spi, NULL);
 }
 
-static void mpc8xxx_spi_work(struct work_struct *work)
-{
-	struct mpc8xxx_spi *mpc8xxx_spi = container_of(work, struct mpc8xxx_spi,
-						       work);
-
-	spin_lock_irq(&mpc8xxx_spi->lock);
-	while (!list_empty(&mpc8xxx_spi->queue)) {
-		struct spi_message *m = container_of(mpc8xxx_spi->queue.next,
-						   struct spi_message, queue);
-
-		list_del_init(&m->queue);
-		spin_unlock_irq(&mpc8xxx_spi->lock);
-
-		mpc8xxx_spi_do_one_msg(m);
-
-		spin_lock_irq(&mpc8xxx_spi->lock);
-	}
-	spin_unlock_irq(&mpc8xxx_spi->lock);
-}
-
-static int mpc8xxx_spi_setup(struct spi_device *spi)
+static int fsl_spi_setup(struct spi_device *spi)
 {
 	struct mpc8xxx_spi *mpc8xxx_spi;
+	struct fsl_spi_reg *reg_base;
 	int retval;
 	u32 hw_mode;
 	struct spi_mpc8xxx_cs	*cs = spi->controller_state;
@@ -661,8 +541,10 @@
 	}
 	mpc8xxx_spi = spi_master_get_devdata(spi->master);
 
+	reg_base = mpc8xxx_spi->reg_base;
+
 	hw_mode = cs->hw_mode; /* Save original settings */
-	cs->hw_mode = mpc8xxx_spi_read_reg(&mpc8xxx_spi->base->mode);
+	cs->hw_mode = mpc8xxx_spi_read_reg(&reg_base->mode);
 	/* mask out bits we are going to set */
 	cs->hw_mode &= ~(SPMODE_CP_BEGIN_EDGECLK | SPMODE_CI_INACTIVEHIGH
 			 | SPMODE_REV | SPMODE_LOOP);
@@ -676,7 +558,7 @@
 	if (spi->mode & SPI_LOOP)
 		cs->hw_mode |= SPMODE_LOOP;
 
-	retval = mpc8xxx_spi_setup_transfer(spi, NULL);
+	retval = fsl_spi_setup_transfer(spi, NULL);
 	if (retval < 0) {
 		cs->hw_mode = hw_mode; /* Restore settings */
 		return retval;
@@ -684,9 +566,10 @@
 	return 0;
 }
 
-static void mpc8xxx_spi_cpm_irq(struct mpc8xxx_spi *mspi, u32 events)
+static void fsl_spi_cpm_irq(struct mpc8xxx_spi *mspi, u32 events)
 {
 	u16 len;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
 
 	dev_dbg(mspi->dev, "%s: bd datlen %d, count %d\n", __func__,
 		in_be16(&mspi->rx_bd->cbd_datlen), mspi->count);
@@ -698,20 +581,22 @@
 	}
 
 	/* Clear the events */
-	mpc8xxx_spi_write_reg(&mspi->base->event, events);
+	mpc8xxx_spi_write_reg(&reg_base->event, events);
 
 	mspi->count -= len;
 	if (mspi->count)
-		mpc8xxx_spi_cpm_bufs_start(mspi);
+		fsl_spi_cpm_bufs_start(mspi);
 	else
 		complete(&mspi->done);
 }
 
-static void mpc8xxx_spi_cpu_irq(struct mpc8xxx_spi *mspi, u32 events)
+static void fsl_spi_cpu_irq(struct mpc8xxx_spi *mspi, u32 events)
 {
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
+
 	/* We need handle RX first */
 	if (events & SPIE_NE) {
-		u32 rx_data = mpc8xxx_spi_read_reg(&mspi->base->receive);
+		u32 rx_data = mpc8xxx_spi_read_reg(&reg_base->receive);
 
 		if (mspi->rx)
 			mspi->get_rx(rx_data, mspi);
@@ -720,102 +605,80 @@
 	if ((events & SPIE_NF) == 0)
 		/* spin until TX is done */
 		while (((events =
-			mpc8xxx_spi_read_reg(&mspi->base->event)) &
+			mpc8xxx_spi_read_reg(&reg_base->event)) &
 						SPIE_NF) == 0)
 			cpu_relax();
 
 	/* Clear the events */
-	mpc8xxx_spi_write_reg(&mspi->base->event, events);
+	mpc8xxx_spi_write_reg(&reg_base->event, events);
 
 	mspi->count -= 1;
 	if (mspi->count) {
 		u32 word = mspi->get_tx(mspi);
 
-		mpc8xxx_spi_write_reg(&mspi->base->transmit, word);
+		mpc8xxx_spi_write_reg(&reg_base->transmit, word);
 	} else {
 		complete(&mspi->done);
 	}
 }
 
-static irqreturn_t mpc8xxx_spi_irq(s32 irq, void *context_data)
+static irqreturn_t fsl_spi_irq(s32 irq, void *context_data)
 {
 	struct mpc8xxx_spi *mspi = context_data;
 	irqreturn_t ret = IRQ_NONE;
 	u32 events;
+	struct fsl_spi_reg *reg_base = mspi->reg_base;
 
 	/* Get interrupt events(tx/rx) */
-	events = mpc8xxx_spi_read_reg(&mspi->base->event);
+	events = mpc8xxx_spi_read_reg(&reg_base->event);
 	if (events)
 		ret = IRQ_HANDLED;
 
 	dev_dbg(mspi->dev, "%s: events %x\n", __func__, events);
 
 	if (mspi->flags & SPI_CPM_MODE)
-		mpc8xxx_spi_cpm_irq(mspi, events);
+		fsl_spi_cpm_irq(mspi, events);
 	else
-		mpc8xxx_spi_cpu_irq(mspi, events);
+		fsl_spi_cpu_irq(mspi, events);
 
 	return ret;
 }
 
-static int mpc8xxx_spi_transfer(struct spi_device *spi,
-				struct spi_message *m)
+static void *fsl_spi_alloc_dummy_rx(void)
 {
-	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(spi->master);
-	unsigned long flags;
+	mutex_lock(&fsl_dummy_rx_lock);
 
-	m->actual_length = 0;
-	m->status = -EINPROGRESS;
+	if (!fsl_dummy_rx)
+		fsl_dummy_rx = kmalloc(SPI_MRBLR, GFP_KERNEL);
+	if (fsl_dummy_rx)
+		fsl_dummy_rx_refcnt++;
 
-	spin_lock_irqsave(&mpc8xxx_spi->lock, flags);
-	list_add_tail(&m->queue, &mpc8xxx_spi->queue);
-	queue_work(mpc8xxx_spi->workqueue, &mpc8xxx_spi->work);
-	spin_unlock_irqrestore(&mpc8xxx_spi->lock, flags);
+	mutex_unlock(&fsl_dummy_rx_lock);
 
-	return 0;
+	return fsl_dummy_rx;
 }
 
-
-static void mpc8xxx_spi_cleanup(struct spi_device *spi)
+static void fsl_spi_free_dummy_rx(void)
 {
-	kfree(spi->controller_state);
-}
+	mutex_lock(&fsl_dummy_rx_lock);
 
-static void *mpc8xxx_spi_alloc_dummy_rx(void)
-{
-	mutex_lock(&mpc8xxx_dummy_rx_lock);
-
-	if (!mpc8xxx_dummy_rx)
-		mpc8xxx_dummy_rx = kmalloc(SPI_MRBLR, GFP_KERNEL);
-	if (mpc8xxx_dummy_rx)
-		mpc8xxx_dummy_rx_refcnt++;
-
-	mutex_unlock(&mpc8xxx_dummy_rx_lock);
-
-	return mpc8xxx_dummy_rx;
-}
-
-static void mpc8xxx_spi_free_dummy_rx(void)
-{
-	mutex_lock(&mpc8xxx_dummy_rx_lock);
-
-	switch (mpc8xxx_dummy_rx_refcnt) {
+	switch (fsl_dummy_rx_refcnt) {
 	case 0:
 		WARN_ON(1);
 		break;
 	case 1:
-		kfree(mpc8xxx_dummy_rx);
-		mpc8xxx_dummy_rx = NULL;
+		kfree(fsl_dummy_rx);
+		fsl_dummy_rx = NULL;
 		/* fall through */
 	default:
-		mpc8xxx_dummy_rx_refcnt--;
+		fsl_dummy_rx_refcnt--;
 		break;
 	}
 
-	mutex_unlock(&mpc8xxx_dummy_rx_lock);
+	mutex_unlock(&fsl_dummy_rx_lock);
 }
 
-static unsigned long mpc8xxx_spi_cpm_get_pram(struct mpc8xxx_spi *mspi)
+static unsigned long fsl_spi_cpm_get_pram(struct mpc8xxx_spi *mspi)
 {
 	struct device *dev = mspi->dev;
 	struct device_node *np = dev->of_node;
@@ -869,7 +732,7 @@
 	return pram_ofs;
 }
 
-static int mpc8xxx_spi_cpm_init(struct mpc8xxx_spi *mspi)
+static int fsl_spi_cpm_init(struct mpc8xxx_spi *mspi)
 {
 	struct device *dev = mspi->dev;
 	struct device_node *np = dev->of_node;
@@ -881,7 +744,7 @@
 	if (!(mspi->flags & SPI_CPM_MODE))
 		return 0;
 
-	if (!mpc8xxx_spi_alloc_dummy_rx())
+	if (!fsl_spi_alloc_dummy_rx())
 		return -ENOMEM;
 
 	if (mspi->flags & SPI_QE) {
@@ -902,7 +765,7 @@
 		}
 	}
 
-	pram_ofs = mpc8xxx_spi_cpm_get_pram(mspi);
+	pram_ofs = fsl_spi_cpm_get_pram(mspi);
 	if (IS_ERR_VALUE(pram_ofs)) {
 		dev_err(dev, "can't allocate spi parameter ram\n");
 		goto err_pram;
@@ -922,7 +785,7 @@
 		goto err_dummy_tx;
 	}
 
-	mspi->dma_dummy_rx = dma_map_single(dev, mpc8xxx_dummy_rx, SPI_MRBLR,
+	mspi->dma_dummy_rx = dma_map_single(dev, fsl_dummy_rx, SPI_MRBLR,
 					    DMA_FROM_DEVICE);
 	if (dma_mapping_error(dev, mspi->dma_dummy_rx)) {
 		dev_err(dev, "unable to map dummy rx buffer\n");
@@ -960,11 +823,11 @@
 err_bds:
 	cpm_muram_free(pram_ofs);
 err_pram:
-	mpc8xxx_spi_free_dummy_rx();
+	fsl_spi_free_dummy_rx();
 	return -ENOMEM;
 }
 
-static void mpc8xxx_spi_cpm_free(struct mpc8xxx_spi *mspi)
+static void fsl_spi_cpm_free(struct mpc8xxx_spi *mspi)
 {
 	struct device *dev = mspi->dev;
 
@@ -972,30 +835,22 @@
 	dma_unmap_single(dev, mspi->dma_dummy_tx, PAGE_SIZE, DMA_TO_DEVICE);
 	cpm_muram_free(cpm_muram_offset(mspi->tx_bd));
 	cpm_muram_free(cpm_muram_offset(mspi->pram));
-	mpc8xxx_spi_free_dummy_rx();
+	fsl_spi_free_dummy_rx();
 }
 
-static const char *mpc8xxx_spi_strmode(unsigned int flags)
+static void fsl_spi_remove(struct mpc8xxx_spi *mspi)
 {
-	if (flags & SPI_QE_CPU_MODE) {
-		return "QE CPU";
-	} else if (flags & SPI_CPM_MODE) {
-		if (flags & SPI_QE)
-			return "QE";
-		else if (flags & SPI_CPM2)
-			return "CPM2";
-		else
-			return "CPM1";
-	}
-	return "CPU";
+	iounmap(mspi->reg_base);
+	fsl_spi_cpm_free(mspi);
 }
 
-static struct spi_master * __devinit
-mpc8xxx_spi_probe(struct device *dev, struct resource *mem, unsigned int irq)
+static struct spi_master * __devinit fsl_spi_probe(struct device *dev,
+		struct resource *mem, unsigned int irq)
 {
 	struct fsl_spi_platform_data *pdata = dev->platform_data;
 	struct spi_master *master;
 	struct mpc8xxx_spi *mpc8xxx_spi;
+	struct fsl_spi_reg *reg_base;
 	u32 regval;
 	int ret = 0;
 
@@ -1007,132 +862,77 @@
 
 	dev_set_drvdata(dev, master);
 
-	/* the spi->mode bits understood by this driver: */
-	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH
-			| SPI_LSB_FIRST | SPI_LOOP;
+	ret = mpc8xxx_spi_probe(dev, mem, irq);
+	if (ret)
+		goto err_probe;
 
-	master->setup = mpc8xxx_spi_setup;
-	master->transfer = mpc8xxx_spi_transfer;
-	master->cleanup = mpc8xxx_spi_cleanup;
-	master->dev.of_node = dev->of_node;
+	master->setup = fsl_spi_setup;
 
 	mpc8xxx_spi = spi_master_get_devdata(master);
-	mpc8xxx_spi->dev = dev;
-	mpc8xxx_spi->get_rx = mpc8xxx_spi_rx_buf_u8;
-	mpc8xxx_spi->get_tx = mpc8xxx_spi_tx_buf_u8;
-	mpc8xxx_spi->flags = pdata->flags;
-	mpc8xxx_spi->spibrg = pdata->sysclk;
+	mpc8xxx_spi->spi_do_one_msg = fsl_spi_do_one_msg;
+	mpc8xxx_spi->spi_remove = fsl_spi_remove;
 
-	ret = mpc8xxx_spi_cpm_init(mpc8xxx_spi);
+
+	ret = fsl_spi_cpm_init(mpc8xxx_spi);
 	if (ret)
 		goto err_cpm_init;
 
-	mpc8xxx_spi->rx_shift = 0;
-	mpc8xxx_spi->tx_shift = 0;
 	if (mpc8xxx_spi->flags & SPI_QE_CPU_MODE) {
 		mpc8xxx_spi->rx_shift = 16;
 		mpc8xxx_spi->tx_shift = 24;
 	}
 
-	init_completion(&mpc8xxx_spi->done);
-
-	mpc8xxx_spi->base = ioremap(mem->start, resource_size(mem));
-	if (mpc8xxx_spi->base == NULL) {
+	mpc8xxx_spi->reg_base = ioremap(mem->start, resource_size(mem));
+	if (mpc8xxx_spi->reg_base == NULL) {
 		ret = -ENOMEM;
 		goto err_ioremap;
 	}
 
-	mpc8xxx_spi->irq = irq;
-
 	/* Register for SPI Interrupt */
-	ret = request_irq(mpc8xxx_spi->irq, mpc8xxx_spi_irq,
-			  0, "mpc8xxx_spi", mpc8xxx_spi);
+	ret = request_irq(mpc8xxx_spi->irq, fsl_spi_irq,
+			  0, "fsl_spi", mpc8xxx_spi);
 
 	if (ret != 0)
-		goto unmap_io;
+		goto free_irq;
 
-	master->bus_num = pdata->bus_num;
-	master->num_chipselect = pdata->max_chipselect;
+	reg_base = mpc8xxx_spi->reg_base;
 
 	/* SPI controller initializations */
-	mpc8xxx_spi_write_reg(&mpc8xxx_spi->base->mode, 0);
-	mpc8xxx_spi_write_reg(&mpc8xxx_spi->base->mask, 0);
-	mpc8xxx_spi_write_reg(&mpc8xxx_spi->base->command, 0);
-	mpc8xxx_spi_write_reg(&mpc8xxx_spi->base->event, 0xffffffff);
+	mpc8xxx_spi_write_reg(&reg_base->mode, 0);
+	mpc8xxx_spi_write_reg(&reg_base->mask, 0);
+	mpc8xxx_spi_write_reg(&reg_base->command, 0);
+	mpc8xxx_spi_write_reg(&reg_base->event, 0xffffffff);
 
 	/* Enable SPI interface */
 	regval = pdata->initial_spmode | SPMODE_INIT_VAL | SPMODE_ENABLE;
 	if (mpc8xxx_spi->flags & SPI_QE_CPU_MODE)
 		regval |= SPMODE_OP;
 
-	mpc8xxx_spi_write_reg(&mpc8xxx_spi->base->mode, regval);
-	spin_lock_init(&mpc8xxx_spi->lock);
-	init_completion(&mpc8xxx_spi->done);
-	INIT_WORK(&mpc8xxx_spi->work, mpc8xxx_spi_work);
-	INIT_LIST_HEAD(&mpc8xxx_spi->queue);
-
-	mpc8xxx_spi->workqueue = create_singlethread_workqueue(
-		dev_name(master->dev.parent));
-	if (mpc8xxx_spi->workqueue == NULL) {
-		ret = -EBUSY;
-		goto free_irq;
-	}
+	mpc8xxx_spi_write_reg(&reg_base->mode, regval);
 
 	ret = spi_register_master(master);
 	if (ret < 0)
 		goto unreg_master;
 
-	dev_info(dev, "at 0x%p (irq = %d), %s mode\n", mpc8xxx_spi->base,
+	dev_info(dev, "at 0x%p (irq = %d), %s mode\n", reg_base,
 		 mpc8xxx_spi->irq, mpc8xxx_spi_strmode(mpc8xxx_spi->flags));
 
 	return master;
 
 unreg_master:
-	destroy_workqueue(mpc8xxx_spi->workqueue);
-free_irq:
 	free_irq(mpc8xxx_spi->irq, mpc8xxx_spi);
-unmap_io:
-	iounmap(mpc8xxx_spi->base);
+free_irq:
+	iounmap(mpc8xxx_spi->reg_base);
 err_ioremap:
-	mpc8xxx_spi_cpm_free(mpc8xxx_spi);
+	fsl_spi_cpm_free(mpc8xxx_spi);
 err_cpm_init:
+err_probe:
 	spi_master_put(master);
 err:
 	return ERR_PTR(ret);
 }
 
-static int __devexit mpc8xxx_spi_remove(struct device *dev)
-{
-	struct mpc8xxx_spi *mpc8xxx_spi;
-	struct spi_master *master;
-
-	master = dev_get_drvdata(dev);
-	mpc8xxx_spi = spi_master_get_devdata(master);
-
-	flush_workqueue(mpc8xxx_spi->workqueue);
-	destroy_workqueue(mpc8xxx_spi->workqueue);
-	spi_unregister_master(master);
-
-	free_irq(mpc8xxx_spi->irq, mpc8xxx_spi);
-	iounmap(mpc8xxx_spi->base);
-	mpc8xxx_spi_cpm_free(mpc8xxx_spi);
-
-	return 0;
-}
-
-struct mpc8xxx_spi_probe_info {
-	struct fsl_spi_platform_data pdata;
-	int *gpios;
-	bool *alow_flags;
-};
-
-static struct mpc8xxx_spi_probe_info *
-to_of_pinfo(struct fsl_spi_platform_data *pdata)
-{
-	return container_of(pdata, struct mpc8xxx_spi_probe_info, pdata);
-}
-
-static void mpc8xxx_spi_cs_control(struct spi_device *spi, bool on)
+static void fsl_spi_cs_control(struct spi_device *spi, bool on)
 {
 	struct device *dev = spi->dev.parent;
 	struct mpc8xxx_spi_probe_info *pinfo = to_of_pinfo(dev->platform_data);
@@ -1143,7 +943,7 @@
 	gpio_set_value(gpio, on ^ alow);
 }
 
-static int of_mpc8xxx_spi_get_chipselects(struct device *dev)
+static int of_fsl_spi_get_chipselects(struct device *dev)
 {
 	struct device_node *np = dev->of_node;
 	struct fsl_spi_platform_data *pdata = dev->platform_data;
@@ -1204,7 +1004,7 @@
 	}
 
 	pdata->max_chipselect = ngpios;
-	pdata->cs_control = mpc8xxx_spi_cs_control;
+	pdata->cs_control = fsl_spi_cs_control;
 
 	return 0;
 
@@ -1223,7 +1023,7 @@
 	return ret;
 }
 
-static int of_mpc8xxx_spi_free_chipselects(struct device *dev)
+static int of_fsl_spi_free_chipselects(struct device *dev)
 {
 	struct fsl_spi_platform_data *pdata = dev->platform_data;
 	struct mpc8xxx_spi_probe_info *pinfo = to_of_pinfo(pdata);
@@ -1242,50 +1042,21 @@
 	return 0;
 }
 
-static int __devinit of_mpc8xxx_spi_probe(struct platform_device *ofdev,
-					  const struct of_device_id *ofid)
+static int __devinit of_fsl_spi_probe(struct platform_device *ofdev,
+					const struct of_device_id *ofid)
 {
 	struct device *dev = &ofdev->dev;
 	struct device_node *np = ofdev->dev.of_node;
-	struct mpc8xxx_spi_probe_info *pinfo;
-	struct fsl_spi_platform_data *pdata;
 	struct spi_master *master;
 	struct resource mem;
 	struct resource irq;
-	const void *prop;
 	int ret = -ENOMEM;
 
-	pinfo = kzalloc(sizeof(*pinfo), GFP_KERNEL);
-	if (!pinfo)
-		return -ENOMEM;
+	ret = of_mpc8xxx_spi_probe(ofdev, ofid);
+	if (ret)
+		return ret;
 
-	pdata = &pinfo->pdata;
-	dev->platform_data = pdata;
-
-	/* Allocate bus num dynamically. */
-	pdata->bus_num = -1;
-
-	/* SPI controller is either clocked from QE or SoC clock. */
-	pdata->sysclk = get_brgfreq();
-	if (pdata->sysclk == -1) {
-		pdata->sysclk = fsl_get_sys_freq();
-		if (pdata->sysclk == -1) {
-			ret = -ENODEV;
-			goto err_clk;
-		}
-	}
-
-	prop = of_get_property(np, "mode", NULL);
-	if (prop && !strcmp(prop, "cpu-qe"))
-		pdata->flags = SPI_QE_CPU_MODE;
-	else if (prop && !strcmp(prop, "qe"))
-		pdata->flags = SPI_CPM_MODE | SPI_QE;
-	else if (of_device_is_compatible(np, "fsl,cpm2-spi"))
-		pdata->flags = SPI_CPM_MODE | SPI_CPM2;
-	else if (of_device_is_compatible(np, "fsl,cpm1-spi"))
-		pdata->flags = SPI_CPM_MODE | SPI_CPM1;
-
-	ret = of_mpc8xxx_spi_get_chipselects(dev);
+	ret = of_fsl_spi_get_chipselects(dev);
 	if (ret)
 		goto err;
 
@@ -1299,7 +1070,7 @@
 		goto err;
 	}
 
-	master = mpc8xxx_spi_probe(dev, &mem, irq.start);
+	master = fsl_spi_probe(dev, &mem, irq.start);
 	if (IS_ERR(master)) {
 		ret = PTR_ERR(master);
 		goto err;
@@ -1308,42 +1079,40 @@
 	return 0;
 
 err:
-	of_mpc8xxx_spi_free_chipselects(dev);
-err_clk:
-	kfree(pinfo);
+	of_fsl_spi_free_chipselects(dev);
 	return ret;
 }
 
-static int __devexit of_mpc8xxx_spi_remove(struct platform_device *ofdev)
+static int __devexit of_fsl_spi_remove(struct platform_device *ofdev)
 {
 	int ret;
 
 	ret = mpc8xxx_spi_remove(&ofdev->dev);
 	if (ret)
 		return ret;
-	of_mpc8xxx_spi_free_chipselects(&ofdev->dev);
+	of_fsl_spi_free_chipselects(&ofdev->dev);
 	return 0;
 }
 
-static const struct of_device_id of_mpc8xxx_spi_match[] = {
+static const struct of_device_id of_fsl_spi_match[] = {
 	{ .compatible = "fsl,spi" },
-	{},
+	{}
 };
-MODULE_DEVICE_TABLE(of, of_mpc8xxx_spi_match);
+MODULE_DEVICE_TABLE(of, of_fsl_spi_match);
 
-static struct of_platform_driver of_mpc8xxx_spi_driver = {
+static struct of_platform_driver of_fsl_spi_driver = {
 	.driver = {
-		.name = "mpc8xxx_spi",
+		.name = "fsl_spi",
 		.owner = THIS_MODULE,
-		.of_match_table = of_mpc8xxx_spi_match,
+		.of_match_table = of_fsl_spi_match,
 	},
-	.probe		= of_mpc8xxx_spi_probe,
-	.remove		= __devexit_p(of_mpc8xxx_spi_remove),
+	.probe		= of_fsl_spi_probe,
+	.remove		= __devexit_p(of_fsl_spi_remove),
 };
 
 #ifdef CONFIG_MPC832x_RDB
 /*
- * 				XXX XXX XXX
+ * XXX XXX XXX
  * This is "legacy" platform driver, was used by the MPC8323E-RDB boards
  * only. The driver should go away soon, since newer MPC8323E-RDB's device
  * tree can work with OpenFirmware driver. But for now we support old trees
@@ -1366,7 +1135,7 @@
 	if (irq <= 0)
 		return -EINVAL;
 
-	master = mpc8xxx_spi_probe(&pdev->dev, mem, irq);
+	master = fsl_spi_probe(&pdev->dev, mem, irq);
 	if (IS_ERR(master))
 		return PTR_ERR(master);
 	return 0;
@@ -1405,21 +1174,20 @@
 static void __exit legacy_driver_unregister(void) {}
 #endif /* CONFIG_MPC832x_RDB */
 
-static int __init mpc8xxx_spi_init(void)
+static int __init fsl_spi_init(void)
 {
 	legacy_driver_register();
-	return of_register_platform_driver(&of_mpc8xxx_spi_driver);
+	return of_register_platform_driver(&of_fsl_spi_driver);
 }
+module_init(fsl_spi_init);
 
-static void __exit mpc8xxx_spi_exit(void)
+static void __exit fsl_spi_exit(void)
 {
-	of_unregister_platform_driver(&of_mpc8xxx_spi_driver);
+	of_unregister_platform_driver(&of_fsl_spi_driver);
 	legacy_driver_unregister();
 }
-
-module_init(mpc8xxx_spi_init);
-module_exit(mpc8xxx_spi_exit);
+module_exit(fsl_spi_exit);
 
 MODULE_AUTHOR("Kumar Gala");
-MODULE_DESCRIPTION("Simple MPC8xxx SPI Driver");
+MODULE_DESCRIPTION("Simple Freescale SPI Driver");
 MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi_s3c64xx.c b/drivers/spi/spi_s3c64xx.c
index c3038da..795828b 100644
--- a/drivers/spi/spi_s3c64xx.c
+++ b/drivers/spi/spi_s3c64xx.c
@@ -261,15 +261,25 @@
 		chcfg |= S3C64XX_SPI_CH_TXCH_ON;
 		if (dma_mode) {
 			modecfg |= S3C64XX_SPI_MODE_TXDMA_ON;
-			s3c2410_dma_config(sdd->tx_dmach, 1);
+			s3c2410_dma_config(sdd->tx_dmach, sdd->cur_bpw / 8);
 			s3c2410_dma_enqueue(sdd->tx_dmach, (void *)sdd,
 						xfer->tx_dma, xfer->len);
 			s3c2410_dma_ctrl(sdd->tx_dmach, S3C2410_DMAOP_START);
 		} else {
-			unsigned char *buf = (unsigned char *) xfer->tx_buf;
-			int i = 0;
-			while (i < xfer->len)
-				writeb(buf[i++], regs + S3C64XX_SPI_TX_DATA);
+			switch (sdd->cur_bpw) {
+			case 32:
+				iowrite32_rep(regs + S3C64XX_SPI_TX_DATA,
+					xfer->tx_buf, xfer->len / 4);
+				break;
+			case 16:
+				iowrite16_rep(regs + S3C64XX_SPI_TX_DATA,
+					xfer->tx_buf, xfer->len / 2);
+				break;
+			default:
+				iowrite8_rep(regs + S3C64XX_SPI_TX_DATA,
+					xfer->tx_buf, xfer->len);
+				break;
+			}
 		}
 	}
 
@@ -286,7 +296,7 @@
 			writel(((xfer->len * 8 / sdd->cur_bpw) & 0xffff)
 					| S3C64XX_SPI_PACKET_CNT_EN,
 					regs + S3C64XX_SPI_PACKET_CNT);
-			s3c2410_dma_config(sdd->rx_dmach, 1);
+			s3c2410_dma_config(sdd->rx_dmach, sdd->cur_bpw / 8);
 			s3c2410_dma_enqueue(sdd->rx_dmach, (void *)sdd,
 						xfer->rx_dma, xfer->len);
 			s3c2410_dma_ctrl(sdd->rx_dmach, S3C2410_DMAOP_START);
@@ -366,20 +376,26 @@
 				return -EIO;
 		}
 	} else {
-		unsigned char *buf;
-		int i;
-
 		/* If it was only Tx */
 		if (xfer->rx_buf == NULL) {
 			sdd->state &= ~TXBUSY;
 			return 0;
 		}
 
-		i = 0;
-		buf = xfer->rx_buf;
-		while (i < xfer->len)
-			buf[i++] = readb(regs + S3C64XX_SPI_RX_DATA);
-
+		switch (sdd->cur_bpw) {
+		case 32:
+			ioread32_rep(regs + S3C64XX_SPI_RX_DATA,
+				xfer->rx_buf, xfer->len / 4);
+			break;
+		case 16:
+			ioread16_rep(regs + S3C64XX_SPI_RX_DATA,
+				xfer->rx_buf, xfer->len / 2);
+			break;
+		default:
+			ioread8_rep(regs + S3C64XX_SPI_RX_DATA,
+				xfer->rx_buf, xfer->len);
+			break;
+		}
 		sdd->state &= ~RXBUSY;
 	}
 
@@ -399,13 +415,18 @@
 
 static void s3c64xx_spi_config(struct s3c64xx_spi_driver_data *sdd)
 {
+	struct s3c64xx_spi_info *sci = sdd->cntrlr_info;
 	void __iomem *regs = sdd->regs;
 	u32 val;
 
 	/* Disable Clock */
-	val = readl(regs + S3C64XX_SPI_CLK_CFG);
-	val &= ~S3C64XX_SPI_ENCLK_ENABLE;
-	writel(val, regs + S3C64XX_SPI_CLK_CFG);
+	if (sci->clk_from_cmu) {
+		clk_disable(sdd->src_clk);
+	} else {
+		val = readl(regs + S3C64XX_SPI_CLK_CFG);
+		val &= ~S3C64XX_SPI_ENCLK_ENABLE;
+		writel(val, regs + S3C64XX_SPI_CLK_CFG);
+	}
 
 	/* Set Polarity and Phase */
 	val = readl(regs + S3C64XX_SPI_CH_CFG);
@@ -429,29 +450,39 @@
 	switch (sdd->cur_bpw) {
 	case 32:
 		val |= S3C64XX_SPI_MODE_BUS_TSZ_WORD;
+		val |= S3C64XX_SPI_MODE_CH_TSZ_WORD;
 		break;
 	case 16:
 		val |= S3C64XX_SPI_MODE_BUS_TSZ_HALFWORD;
+		val |= S3C64XX_SPI_MODE_CH_TSZ_HALFWORD;
 		break;
 	default:
 		val |= S3C64XX_SPI_MODE_BUS_TSZ_BYTE;
+		val |= S3C64XX_SPI_MODE_CH_TSZ_BYTE;
 		break;
 	}
-	val |= S3C64XX_SPI_MODE_CH_TSZ_BYTE; /* Always 8bits wide */
 
 	writel(val, regs + S3C64XX_SPI_MODE_CFG);
 
-	/* Configure Clock */
-	val = readl(regs + S3C64XX_SPI_CLK_CFG);
-	val &= ~S3C64XX_SPI_PSR_MASK;
-	val |= ((clk_get_rate(sdd->src_clk) / sdd->cur_speed / 2 - 1)
-			& S3C64XX_SPI_PSR_MASK);
-	writel(val, regs + S3C64XX_SPI_CLK_CFG);
+	if (sci->clk_from_cmu) {
+		/* Configure Clock */
+		/* There is half-multiplier before the SPI */
+		clk_set_rate(sdd->src_clk, sdd->cur_speed * 2);
+		/* Enable Clock */
+		clk_enable(sdd->src_clk);
+	} else {
+		/* Configure Clock */
+		val = readl(regs + S3C64XX_SPI_CLK_CFG);
+		val &= ~S3C64XX_SPI_PSR_MASK;
+		val |= ((clk_get_rate(sdd->src_clk) / sdd->cur_speed / 2 - 1)
+				& S3C64XX_SPI_PSR_MASK);
+		writel(val, regs + S3C64XX_SPI_CLK_CFG);
 
-	/* Enable Clock */
-	val = readl(regs + S3C64XX_SPI_CLK_CFG);
-	val |= S3C64XX_SPI_ENCLK_ENABLE;
-	writel(val, regs + S3C64XX_SPI_CLK_CFG);
+		/* Enable Clock */
+		val = readl(regs + S3C64XX_SPI_CLK_CFG);
+		val |= S3C64XX_SPI_ENCLK_ENABLE;
+		writel(val, regs + S3C64XX_SPI_CLK_CFG);
+	}
 }
 
 static void s3c64xx_spi_dma_rxcb(struct s3c2410_dma_chan *chan, void *buf_id,
@@ -499,6 +530,7 @@
 static int s3c64xx_spi_map_mssg(struct s3c64xx_spi_driver_data *sdd,
 						struct spi_message *msg)
 {
+	struct s3c64xx_spi_info *sci = sdd->cntrlr_info;
 	struct device *dev = &sdd->pdev->dev;
 	struct spi_transfer *xfer;
 
@@ -514,6 +546,9 @@
 	/* Map until end or first fail */
 	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
 
+		if (xfer->len <= ((sci->fifo_lvl_mask >> 1) + 1))
+			continue;
+
 		if (xfer->tx_buf != NULL) {
 			xfer->tx_dma = dma_map_single(dev,
 					(void *)xfer->tx_buf, xfer->len,
@@ -545,6 +580,7 @@
 static void s3c64xx_spi_unmap_mssg(struct s3c64xx_spi_driver_data *sdd,
 						struct spi_message *msg)
 {
+	struct s3c64xx_spi_info *sci = sdd->cntrlr_info;
 	struct device *dev = &sdd->pdev->dev;
 	struct spi_transfer *xfer;
 
@@ -553,6 +589,9 @@
 
 	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
 
+		if (xfer->len <= ((sci->fifo_lvl_mask >> 1) + 1))
+			continue;
+
 		if (xfer->rx_buf != NULL
 				&& xfer->rx_dma != XFER_DMAADDR_INVALID)
 			dma_unmap_single(dev, xfer->rx_dma,
@@ -608,6 +647,14 @@
 		bpw = xfer->bits_per_word ? : spi->bits_per_word;
 		speed = xfer->speed_hz ? : spi->max_speed_hz;
 
+		if (xfer->len % (bpw / 8)) {
+			dev_err(&spi->dev,
+				"Xfer length(%u) not a multiple of word size(%u)\n",
+				xfer->len, bpw / 8);
+			status = -EIO;
+			goto out;
+		}
+
 		if (bpw != sdd->cur_bpw || speed != sdd->cur_speed) {
 			sdd->cur_bpw = bpw;
 			sdd->cur_speed = speed;
@@ -798,7 +845,6 @@
 	struct s3c64xx_spi_driver_data *sdd;
 	struct s3c64xx_spi_info *sci;
 	struct spi_message *msg;
-	u32 psr, speed;
 	unsigned long flags;
 	int err = 0;
 
@@ -841,31 +887,36 @@
 	}
 
 	/* Check if we can provide the requested rate */
-	speed = clk_get_rate(sdd->src_clk) / 2 / (0 + 1); /* Max possible */
+	if (!sci->clk_from_cmu) {
+		u32 psr, speed;
 
-	if (spi->max_speed_hz > speed)
-		spi->max_speed_hz = speed;
+		/* Max possible */
+		speed = clk_get_rate(sdd->src_clk) / 2 / (0 + 1);
 
-	psr = clk_get_rate(sdd->src_clk) / 2 / spi->max_speed_hz - 1;
-	psr &= S3C64XX_SPI_PSR_MASK;
-	if (psr == S3C64XX_SPI_PSR_MASK)
-		psr--;
+		if (spi->max_speed_hz > speed)
+			spi->max_speed_hz = speed;
 
-	speed = clk_get_rate(sdd->src_clk) / 2 / (psr + 1);
-	if (spi->max_speed_hz < speed) {
-		if (psr+1 < S3C64XX_SPI_PSR_MASK) {
-			psr++;
-		} else {
-			err = -EINVAL;
-			goto setup_exit;
+		psr = clk_get_rate(sdd->src_clk) / 2 / spi->max_speed_hz - 1;
+		psr &= S3C64XX_SPI_PSR_MASK;
+		if (psr == S3C64XX_SPI_PSR_MASK)
+			psr--;
+
+		speed = clk_get_rate(sdd->src_clk) / 2 / (psr + 1);
+		if (spi->max_speed_hz < speed) {
+			if (psr+1 < S3C64XX_SPI_PSR_MASK) {
+				psr++;
+			} else {
+				err = -EINVAL;
+				goto setup_exit;
+			}
 		}
-	}
 
-	speed = clk_get_rate(sdd->src_clk) / 2 / (psr + 1);
-	if (spi->max_speed_hz >= speed)
-		spi->max_speed_hz = speed;
-	else
-		err = -EINVAL;
+		speed = clk_get_rate(sdd->src_clk) / 2 / (psr + 1);
+		if (spi->max_speed_hz >= speed)
+			spi->max_speed_hz = speed;
+		else
+			err = -EINVAL;
+	}
 
 setup_exit:
 
@@ -888,7 +939,8 @@
 	/* Disable Interrupts - we use Polling if not DMA mode */
 	writel(0, regs + S3C64XX_SPI_INT_EN);
 
-	writel(sci->src_clk_nr << S3C64XX_SPI_CLKSEL_SRCSHFT,
+	if (!sci->clk_from_cmu)
+		writel(sci->src_clk_nr << S3C64XX_SPI_CLKSEL_SRCSHFT,
 				regs + S3C64XX_SPI_CLK_CFG);
 	writel(0, regs + S3C64XX_SPI_MODE_CFG);
 	writel(0, regs + S3C64XX_SPI_PACKET_CNT);
diff --git a/drivers/spi/spi_topcliff_pch.c b/drivers/spi/spi_topcliff_pch.c
new file mode 100644
index 0000000..58e187f
--- /dev/null
+++ b/drivers/spi/spi_topcliff_pch.c
@@ -0,0 +1,1303 @@
+/*
+ * SPI bus driver for the Topcliff PCH used by Intel SoCs
+ *
+ * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
+ *
+ * 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; version 2 of the License.
+ *
+ * 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.
+ */
+
+#include <linux/delay.h>
+#include <linux/pci.h>
+#include <linux/wait.h>
+#include <linux/spi/spi.h>
+#include <linux/interrupt.h>
+#include <linux/sched.h>
+#include <linux/spi/spidev.h>
+#include <linux/module.h>
+#include <linux/device.h>
+
+/* Register offsets */
+#define PCH_SPCR		0x00	/* SPI control register */
+#define PCH_SPBRR		0x04	/* SPI baud rate register */
+#define PCH_SPSR		0x08	/* SPI status register */
+#define PCH_SPDWR		0x0C	/* SPI write data register */
+#define PCH_SPDRR		0x10	/* SPI read data register */
+#define PCH_SSNXCR		0x18	/* SSN Expand Control Register */
+#define PCH_SRST		0x1C	/* SPI reset register */
+
+#define PCH_SPSR_TFD		0x000007C0
+#define PCH_SPSR_RFD		0x0000F800
+
+#define PCH_READABLE(x)		(((x) & PCH_SPSR_RFD)>>11)
+#define PCH_WRITABLE(x)		(((x) & PCH_SPSR_TFD)>>6)
+
+#define PCH_RX_THOLD		7
+#define PCH_RX_THOLD_MAX	15
+
+#define PCH_MAX_BAUDRATE	5000000
+#define PCH_MAX_FIFO_DEPTH	16
+
+#define STATUS_RUNNING		1
+#define STATUS_EXITING		2
+#define PCH_SLEEP_TIME		10
+
+#define PCH_ADDRESS_SIZE	0x20
+
+#define SSN_LOW			0x02U
+#define SSN_NO_CONTROL		0x00U
+#define PCH_MAX_CS		0xFF
+#define PCI_DEVICE_ID_GE_SPI	0x8816
+
+#define SPCR_SPE_BIT		(1 << 0)
+#define SPCR_MSTR_BIT		(1 << 1)
+#define SPCR_LSBF_BIT		(1 << 4)
+#define SPCR_CPHA_BIT		(1 << 5)
+#define SPCR_CPOL_BIT		(1 << 6)
+#define SPCR_TFIE_BIT		(1 << 8)
+#define SPCR_RFIE_BIT		(1 << 9)
+#define SPCR_FIE_BIT		(1 << 10)
+#define SPCR_ORIE_BIT		(1 << 11)
+#define SPCR_MDFIE_BIT		(1 << 12)
+#define SPCR_FICLR_BIT		(1 << 24)
+#define SPSR_TFI_BIT		(1 << 0)
+#define SPSR_RFI_BIT		(1 << 1)
+#define SPSR_FI_BIT		(1 << 2)
+#define SPBRR_SIZE_BIT		(1 << 10)
+
+#define PCH_ALL			(SPCR_TFIE_BIT|SPCR_RFIE_BIT|SPCR_FIE_BIT|SPCR_ORIE_BIT|SPCR_MDFIE_BIT)
+
+#define SPCR_RFIC_FIELD		20
+#define SPCR_TFIC_FIELD		16
+
+#define SPSR_INT_BITS		0x1F
+#define MASK_SPBRR_SPBR_BITS	(~((1 << 10) - 1))
+#define MASK_RFIC_SPCR_BITS	(~(0xf << 20))
+#define MASK_TFIC_SPCR_BITS	(~(0xf000f << 12))
+
+#define PCH_CLOCK_HZ		50000000
+#define PCH_MAX_SPBR		1023
+
+
+/**
+ * struct pch_spi_data - Holds the SPI channel specific details
+ * @io_remap_addr:		The remapped PCI base address
+ * @master:			Pointer to the SPI master structure
+ * @work:			Reference to work queue handler
+ * @wk:				Workqueue for carrying out execution of the
+ *				requests
+ * @wait:			Wait queue for waking up upon receiving an
+ *				interrupt.
+ * @transfer_complete:		Status of SPI Transfer
+ * @bcurrent_msg_processing:	Status flag for message processing
+ * @lock:			Lock for protecting this structure
+ * @queue:			SPI Message queue
+ * @status:			Status of the SPI driver
+ * @bpw_len:			Length of data to be transferred in bits per
+ *				word
+ * @transfer_active:		Flag showing active transfer
+ * @tx_index:			Transmit data count; for bookkeeping during
+ *				transfer
+ * @rx_index:			Receive data count; for bookkeeping during
+ *				transfer
+ * @tx_buff:			Buffer for data to be transmitted
+ * @rx_index:			Buffer for Received data
+ * @n_curnt_chip:		The chip number that this SPI driver currently
+ *				operates on
+ * @current_chip:		Reference to the current chip that this SPI
+ *				driver currently operates on
+ * @current_msg:		The current message that this SPI driver is
+ *				handling
+ * @cur_trans:			The current transfer that this SPI driver is
+ *				handling
+ * @board_dat:			Reference to the SPI device data structure
+ */
+struct pch_spi_data {
+	void __iomem *io_remap_addr;
+	struct spi_master *master;
+	struct work_struct work;
+	struct workqueue_struct *wk;
+	wait_queue_head_t wait;
+	u8 transfer_complete;
+	u8 bcurrent_msg_processing;
+	spinlock_t lock;
+	struct list_head queue;
+	u8 status;
+	u32 bpw_len;
+	u8 transfer_active;
+	u32 tx_index;
+	u32 rx_index;
+	u16 *pkt_tx_buff;
+	u16 *pkt_rx_buff;
+	u8 n_curnt_chip;
+	struct spi_device *current_chip;
+	struct spi_message *current_msg;
+	struct spi_transfer *cur_trans;
+	struct pch_spi_board_data *board_dat;
+};
+
+/**
+ * struct pch_spi_board_data - Holds the SPI device specific details
+ * @pdev:		Pointer to the PCI device
+ * @irq_reg_sts:	Status of IRQ registration
+ * @pci_req_sts:	Status of pci_request_regions
+ * @suspend_sts:	Status of suspend
+ * @data:		Pointer to SPI channel data structure
+ */
+struct pch_spi_board_data {
+	struct pci_dev *pdev;
+	u8 irq_reg_sts;
+	u8 pci_req_sts;
+	u8 suspend_sts;
+	struct pch_spi_data *data;
+};
+
+static struct pci_device_id pch_spi_pcidev_id[] = {
+	{PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_GE_SPI)},
+	{0,}
+};
+
+/**
+ * pch_spi_writereg() - Performs  register writes
+ * @master:	Pointer to struct spi_master.
+ * @idx:	Register offset.
+ * @val:	Value to be written to register.
+ */
+static inline void pch_spi_writereg(struct spi_master *master, int idx, u32 val)
+{
+	struct pch_spi_data *data = spi_master_get_devdata(master);
+	iowrite32(val, (data->io_remap_addr + idx));
+}
+
+/**
+ * pch_spi_readreg() - Performs register reads
+ * @master:	Pointer to struct spi_master.
+ * @idx:	Register offset.
+ */
+static inline u32 pch_spi_readreg(struct spi_master *master, int idx)
+{
+	struct pch_spi_data *data = spi_master_get_devdata(master);
+	return ioread32(data->io_remap_addr + idx);
+}
+
+static inline void pch_spi_setclr_reg(struct spi_master *master, int idx,
+				      u32 set, u32 clr)
+{
+	u32 tmp = pch_spi_readreg(master, idx);
+	tmp = (tmp & ~clr) | set;
+	pch_spi_writereg(master, idx, tmp);
+}
+
+static void pch_spi_set_master_mode(struct spi_master *master)
+{
+	pch_spi_setclr_reg(master, PCH_SPCR, SPCR_MSTR_BIT, 0);
+}
+
+/**
+ * pch_spi_clear_fifo() - Clears the Transmit and Receive FIFOs
+ * @master:	Pointer to struct spi_master.
+ */
+static void pch_spi_clear_fifo(struct spi_master *master)
+{
+	pch_spi_setclr_reg(master, PCH_SPCR, SPCR_FICLR_BIT, 0);
+	pch_spi_setclr_reg(master, PCH_SPCR, 0, SPCR_FICLR_BIT);
+}
+
+static void pch_spi_handler_sub(struct pch_spi_data *data, u32 reg_spsr_val,
+				void __iomem *io_remap_addr)
+{
+	u32 n_read, tx_index, rx_index, bpw_len;
+	u16 *pkt_rx_buffer, *pkt_tx_buff;
+	int read_cnt;
+	u32 reg_spcr_val;
+	void __iomem *spsr;
+	void __iomem *spdrr;
+	void __iomem *spdwr;
+
+	spsr = io_remap_addr + PCH_SPSR;
+	iowrite32(reg_spsr_val, spsr);
+
+	if (data->transfer_active) {
+		rx_index = data->rx_index;
+		tx_index = data->tx_index;
+		bpw_len = data->bpw_len;
+		pkt_rx_buffer = data->pkt_rx_buff;
+		pkt_tx_buff = data->pkt_tx_buff;
+
+		spdrr = io_remap_addr + PCH_SPDRR;
+		spdwr = io_remap_addr + PCH_SPDWR;
+
+		n_read = PCH_READABLE(reg_spsr_val);
+
+		for (read_cnt = 0; (read_cnt < n_read); read_cnt++) {
+			pkt_rx_buffer[rx_index++] = ioread32(spdrr);
+			if (tx_index < bpw_len)
+				iowrite32(pkt_tx_buff[tx_index++], spdwr);
+		}
+
+		/* disable RFI if not needed */
+		if ((bpw_len - rx_index) <= PCH_MAX_FIFO_DEPTH) {
+			reg_spcr_val = ioread32(io_remap_addr + PCH_SPCR);
+			reg_spcr_val &= ~SPCR_RFIE_BIT; /* disable RFI */
+
+			/* reset rx threshold */
+			reg_spcr_val &= MASK_RFIC_SPCR_BITS;
+			reg_spcr_val |= (PCH_RX_THOLD_MAX << SPCR_RFIC_FIELD);
+			iowrite32(((reg_spcr_val) &= (~(SPCR_RFIE_BIT))),
+				 (io_remap_addr + PCH_SPCR));
+		}
+
+		/* update counts */
+		data->tx_index = tx_index;
+		data->rx_index = rx_index;
+
+	}
+
+	/* if transfer complete interrupt */
+	if (reg_spsr_val & SPSR_FI_BIT) {
+		/* disable FI & RFI interrupts */
+		pch_spi_setclr_reg(data->master, PCH_SPCR, 0,
+				   SPCR_FIE_BIT | SPCR_TFIE_BIT);
+
+		/* transfer is completed;inform pch_spi_process_messages */
+		data->transfer_complete = true;
+		wake_up(&data->wait);
+	}
+}
+
+/**
+ * pch_spi_handler() - Interrupt handler
+ * @irq:	The interrupt number.
+ * @dev_id:	Pointer to struct pch_spi_board_data.
+ */
+static irqreturn_t pch_spi_handler(int irq, void *dev_id)
+{
+	u32 reg_spsr_val;
+	struct pch_spi_data *data;
+	void __iomem *spsr;
+	void __iomem *io_remap_addr;
+	irqreturn_t ret = IRQ_NONE;
+	struct pch_spi_board_data *board_dat = dev_id;
+
+	if (board_dat->suspend_sts) {
+		dev_dbg(&board_dat->pdev->dev,
+			"%s returning due to suspend\n", __func__);
+		return IRQ_NONE;
+	}
+
+	data = board_dat->data;
+	io_remap_addr = data->io_remap_addr;
+	spsr = io_remap_addr + PCH_SPSR;
+
+	reg_spsr_val = ioread32(spsr);
+
+	/* Check if the interrupt is for SPI device */
+	if (reg_spsr_val & (SPSR_FI_BIT | SPSR_RFI_BIT)) {
+		pch_spi_handler_sub(data, reg_spsr_val, io_remap_addr);
+		ret = IRQ_HANDLED;
+	}
+
+	dev_dbg(&board_dat->pdev->dev, "%s EXIT return value=%d\n",
+		__func__, ret);
+
+	return ret;
+}
+
+/**
+ * pch_spi_set_baud_rate() - Sets SPBR field in SPBRR
+ * @master:	Pointer to struct spi_master.
+ * @speed_hz:	Baud rate.
+ */
+static void pch_spi_set_baud_rate(struct spi_master *master, u32 speed_hz)
+{
+	u32 n_spbr = PCH_CLOCK_HZ / (speed_hz * 2);
+
+	/* if baud rate is less than we can support limit it */
+	if (n_spbr > PCH_MAX_SPBR)
+		n_spbr = PCH_MAX_SPBR;
+
+	pch_spi_setclr_reg(master, PCH_SPBRR, n_spbr, ~MASK_SPBRR_SPBR_BITS);
+}
+
+/**
+ * pch_spi_set_bits_per_word() - Sets SIZE field in SPBRR
+ * @master:		Pointer to struct spi_master.
+ * @bits_per_word:	Bits per word for SPI transfer.
+ */
+static void pch_spi_set_bits_per_word(struct spi_master *master,
+				      u8 bits_per_word)
+{
+	if (bits_per_word == 8)
+		pch_spi_setclr_reg(master, PCH_SPBRR, 0, SPBRR_SIZE_BIT);
+	else
+		pch_spi_setclr_reg(master, PCH_SPBRR, SPBRR_SIZE_BIT, 0);
+}
+
+/**
+ * pch_spi_setup_transfer() - Configures the PCH SPI hardware for transfer
+ * @spi:	Pointer to struct spi_device.
+ */
+static void pch_spi_setup_transfer(struct spi_device *spi)
+{
+	u32 flags = 0;
+
+	dev_dbg(&spi->dev, "%s SPBRR content =%x setting baud rate=%d\n",
+		__func__, pch_spi_readreg(spi->master, PCH_SPBRR),
+		spi->max_speed_hz);
+	pch_spi_set_baud_rate(spi->master, spi->max_speed_hz);
+
+	/* set bits per word */
+	pch_spi_set_bits_per_word(spi->master, spi->bits_per_word);
+
+	if (!(spi->mode & SPI_LSB_FIRST))
+		flags |= SPCR_LSBF_BIT;
+	if (spi->mode & SPI_CPOL)
+		flags |= SPCR_CPOL_BIT;
+	if (spi->mode & SPI_CPHA)
+		flags |= SPCR_CPHA_BIT;
+	pch_spi_setclr_reg(spi->master, PCH_SPCR, flags,
+			   (SPCR_LSBF_BIT | SPCR_CPOL_BIT | SPCR_CPHA_BIT));
+
+	/* Clear the FIFO by toggling  FICLR to 1 and back to 0 */
+	pch_spi_clear_fifo(spi->master);
+}
+
+/**
+ * pch_spi_reset() - Clears SPI registers
+ * @master:	Pointer to struct spi_master.
+ */
+static void pch_spi_reset(struct spi_master *master)
+{
+	/* write 1 to reset SPI */
+	pch_spi_writereg(master, PCH_SRST, 0x1);
+
+	/* clear reset */
+	pch_spi_writereg(master, PCH_SRST, 0x0);
+}
+
+static int pch_spi_setup(struct spi_device *pspi)
+{
+	/* check bits per word */
+	if (pspi->bits_per_word == 0) {
+		pspi->bits_per_word = 8;
+		dev_dbg(&pspi->dev, "%s 8 bits per word\n", __func__);
+	}
+
+	if ((pspi->bits_per_word != 8) && (pspi->bits_per_word != 16)) {
+		dev_err(&pspi->dev, "%s Invalid bits per word\n", __func__);
+		return -EINVAL;
+	}
+
+	/* Check baud rate setting */
+	/* if baud rate of chip is greater than
+	   max we can support,return error */
+	if ((pspi->max_speed_hz) > PCH_MAX_BAUDRATE)
+		pspi->max_speed_hz = PCH_MAX_BAUDRATE;
+
+	dev_dbg(&pspi->dev, "%s MODE = %x\n", __func__,
+		(pspi->mode) & (SPI_CPOL | SPI_CPHA));
+
+	return 0;
+}
+
+static int pch_spi_transfer(struct spi_device *pspi, struct spi_message *pmsg)
+{
+
+	struct spi_transfer *transfer;
+	struct pch_spi_data *data = spi_master_get_devdata(pspi->master);
+	int retval;
+	unsigned long flags;
+
+	/* validate spi message and baud rate */
+	if (unlikely(list_empty(&pmsg->transfers) == 1)) {
+		dev_err(&pspi->dev, "%s list empty\n", __func__);
+		retval = -EINVAL;
+		goto err_out;
+	}
+
+	if (unlikely(pspi->max_speed_hz == 0)) {
+		dev_err(&pspi->dev, "%s pch_spi_tranfer maxspeed=%d\n",
+			__func__, pspi->max_speed_hz);
+		retval = -EINVAL;
+		goto err_out;
+	}
+
+	dev_dbg(&pspi->dev, "%s Transfer List not empty. "
+		"Transfer Speed is set.\n", __func__);
+
+	/* validate Tx/Rx buffers and Transfer length */
+	list_for_each_entry(transfer, &pmsg->transfers, transfer_list) {
+		if (!transfer->tx_buf && !transfer->rx_buf) {
+			dev_err(&pspi->dev,
+				"%s Tx and Rx buffer NULL\n", __func__);
+			retval = -EINVAL;
+			goto err_out;
+		}
+
+		if (!transfer->len) {
+			dev_err(&pspi->dev, "%s Transfer length invalid\n",
+				__func__);
+			retval = -EINVAL;
+			goto err_out;
+		}
+
+		dev_dbg(&pspi->dev, "%s Tx/Rx buffer valid. Transfer length"
+			" valid\n", __func__);
+
+		/* if baud rate hs been specified validate the same */
+		if (transfer->speed_hz > PCH_MAX_BAUDRATE)
+			transfer->speed_hz = PCH_MAX_BAUDRATE;
+
+		/* if bits per word has been specified validate the same */
+		if (transfer->bits_per_word) {
+			if ((transfer->bits_per_word != 8)
+			    && (transfer->bits_per_word != 16)) {
+				retval = -EINVAL;
+				dev_err(&pspi->dev,
+					"%s Invalid bits per word\n", __func__);
+				goto err_out;
+			}
+		}
+	}
+
+	spin_lock_irqsave(&data->lock, flags);
+
+	/* We won't process any messages if we have been asked to terminate */
+	if (data->status == STATUS_EXITING) {
+		dev_err(&pspi->dev, "%s status = STATUS_EXITING.\n", __func__);
+		retval = -ESHUTDOWN;
+		goto err_return_spinlock;
+	}
+
+	/* If suspended ,return -EINVAL */
+	if (data->board_dat->suspend_sts) {
+		dev_err(&pspi->dev, "%s suspend; returning EINVAL\n", __func__);
+		retval = -EINVAL;
+		goto err_return_spinlock;
+	}
+
+	/* set status of message */
+	pmsg->actual_length = 0;
+	dev_dbg(&pspi->dev, "%s - pmsg->status =%d\n", __func__, pmsg->status);
+
+	pmsg->status = -EINPROGRESS;
+
+	/* add message to queue */
+	list_add_tail(&pmsg->queue, &data->queue);
+	dev_dbg(&pspi->dev, "%s - Invoked list_add_tail\n", __func__);
+
+	/* schedule work queue to run */
+	queue_work(data->wk, &data->work);
+	dev_dbg(&pspi->dev, "%s - Invoked queue work\n", __func__);
+
+	retval = 0;
+
+err_return_spinlock:
+	spin_unlock_irqrestore(&data->lock, flags);
+err_out:
+	dev_dbg(&pspi->dev, "%s RETURN=%d\n", __func__, retval);
+	return retval;
+}
+
+static inline void pch_spi_select_chip(struct pch_spi_data *data,
+				       struct spi_device *pspi)
+{
+	if (data->current_chip != NULL) {
+		if (pspi->chip_select != data->n_curnt_chip) {
+			dev_dbg(&pspi->dev, "%s : different slave\n", __func__);
+			data->current_chip = NULL;
+		}
+	}
+
+	data->current_chip = pspi;
+
+	data->n_curnt_chip = data->current_chip->chip_select;
+
+	dev_dbg(&pspi->dev, "%s :Invoking pch_spi_setup_transfer\n", __func__);
+	pch_spi_setup_transfer(pspi);
+}
+
+static void pch_spi_set_tx(struct pch_spi_data *data, int *bpw,
+			   struct spi_message **ppmsg)
+{
+	int size;
+	u32 n_writes;
+	int j;
+	struct spi_message *pmsg;
+	const u8 *tx_buf;
+	const u16 *tx_sbuf;
+
+	pmsg = *ppmsg;
+
+	/* set baud rate if needed */
+	if (data->cur_trans->speed_hz) {
+		dev_dbg(&data->master->dev, "%s:setting baud rate\n", __func__);
+		pch_spi_set_baud_rate(data->master, data->cur_trans->speed_hz);
+	}
+
+	/* set bits per word if needed */
+	if (data->cur_trans->bits_per_word &&
+	    (data->current_msg->spi->bits_per_word != data->cur_trans->bits_per_word)) {
+		dev_dbg(&data->master->dev, "%s:set bits per word\n", __func__);
+		pch_spi_set_bits_per_word(data->master,
+					  data->cur_trans->bits_per_word);
+		*bpw = data->cur_trans->bits_per_word;
+	} else {
+		*bpw = data->current_msg->spi->bits_per_word;
+	}
+
+	/* reset Tx/Rx index */
+	data->tx_index = 0;
+	data->rx_index = 0;
+
+	data->bpw_len = data->cur_trans->len / (*bpw / 8);
+
+	/* find alloc size */
+	size = data->cur_trans->len * sizeof(*data->pkt_tx_buff);
+
+	/* allocate memory for pkt_tx_buff & pkt_rx_buffer */
+	data->pkt_tx_buff = kzalloc(size, GFP_KERNEL);
+	if (data->pkt_tx_buff != NULL) {
+		data->pkt_rx_buff = kzalloc(size, GFP_KERNEL);
+		if (!data->pkt_rx_buff)
+			kfree(data->pkt_tx_buff);
+	}
+
+	if (!data->pkt_rx_buff) {
+		/* flush queue and set status of all transfers to -ENOMEM */
+		dev_err(&data->master->dev, "%s :kzalloc failed\n", __func__);
+		list_for_each_entry(pmsg, data->queue.next, queue) {
+			pmsg->status = -ENOMEM;
+
+			if (pmsg->complete != 0)
+				pmsg->complete(pmsg->context);
+
+			/* delete from queue */
+			list_del_init(&pmsg->queue);
+		}
+		return;
+	}
+
+	/* copy Tx Data */
+	if (data->cur_trans->tx_buf != NULL) {
+		if (*bpw == 8) {
+			tx_buf = data->cur_trans->tx_buf;
+			for (j = 0; j < data->bpw_len; j++)
+				data->pkt_tx_buff[j] = *tx_buf++;
+		} else {
+			tx_sbuf = data->cur_trans->tx_buf;
+			for (j = 0; j < data->bpw_len; j++)
+				data->pkt_tx_buff[j] = *tx_sbuf++;
+		}
+	}
+
+	/* if len greater than PCH_MAX_FIFO_DEPTH, write 16,else len bytes */
+	n_writes = data->bpw_len;
+	if (n_writes > PCH_MAX_FIFO_DEPTH)
+		n_writes = PCH_MAX_FIFO_DEPTH;
+
+	dev_dbg(&data->master->dev, "\n%s:Pulling down SSN low - writing "
+		"0x2 to SSNXCR\n", __func__);
+	pch_spi_writereg(data->master, PCH_SSNXCR, SSN_LOW);
+
+	for (j = 0; j < n_writes; j++)
+		pch_spi_writereg(data->master, PCH_SPDWR, data->pkt_tx_buff[j]);
+
+	/* update tx_index */
+	data->tx_index = j;
+
+	/* reset transfer complete flag */
+	data->transfer_complete = false;
+	data->transfer_active = true;
+}
+
+
+static void pch_spi_nomore_transfer(struct pch_spi_data *data,
+						struct spi_message *pmsg)
+{
+	dev_dbg(&data->master->dev, "%s called\n", __func__);
+	/* Invoke complete callback
+	 * [To the spi core..indicating end of transfer] */
+	data->current_msg->status = 0;
+
+	if (data->current_msg->complete != 0) {
+		dev_dbg(&data->master->dev,
+			"%s:Invoking callback of SPI core\n", __func__);
+		data->current_msg->complete(data->current_msg->context);
+	}
+
+	/* update status in global variable */
+	data->bcurrent_msg_processing = false;
+
+	dev_dbg(&data->master->dev,
+		"%s:data->bcurrent_msg_processing = false\n", __func__);
+
+	data->current_msg = NULL;
+	data->cur_trans = NULL;
+
+	/* check if we have items in list and not suspending
+	 * return 1 if list empty */
+	if ((list_empty(&data->queue) == 0) &&
+	    (!data->board_dat->suspend_sts) &&
+	    (data->status != STATUS_EXITING)) {
+		/* We have some more work to do (either there is more tranint
+		 * bpw;sfer requests in the current message or there are
+		 *more messages)
+		 */
+		dev_dbg(&data->master->dev, "%s:Invoke queue_work\n", __func__);
+		queue_work(data->wk, &data->work);
+	} else if (data->board_dat->suspend_sts ||
+		   data->status == STATUS_EXITING) {
+		dev_dbg(&data->master->dev,
+			"%s suspend/remove initiated, flushing queue\n",
+			__func__);
+		list_for_each_entry(pmsg, data->queue.next, queue) {
+			pmsg->status = -EIO;
+
+			if (pmsg->complete)
+				pmsg->complete(pmsg->context);
+
+			/* delete from queue */
+			list_del_init(&pmsg->queue);
+		}
+	}
+}
+
+static void pch_spi_set_ir(struct pch_spi_data *data)
+{
+	/* enable interrupts */
+	if ((data->bpw_len) > PCH_MAX_FIFO_DEPTH) {
+		/* set receive threhold to PCH_RX_THOLD */
+		pch_spi_setclr_reg(data->master, PCH_SPCR,
+				   PCH_RX_THOLD << SPCR_TFIC_FIELD,
+				   ~MASK_TFIC_SPCR_BITS);
+		/* enable FI and RFI interrupts */
+		pch_spi_setclr_reg(data->master, PCH_SPCR,
+				   SPCR_RFIE_BIT | SPCR_TFIE_BIT, 0);
+	} else {
+		/* set receive threhold to maximum */
+		pch_spi_setclr_reg(data->master, PCH_SPCR,
+				   PCH_RX_THOLD_MAX << SPCR_TFIC_FIELD,
+				   ~MASK_TFIC_SPCR_BITS);
+		/* enable FI interrupt */
+		pch_spi_setclr_reg(data->master, PCH_SPCR, SPCR_FIE_BIT, 0);
+	}
+
+	dev_dbg(&data->master->dev,
+		"%s:invoking pch_spi_set_enable to enable SPI\n", __func__);
+
+	/* SPI set enable */
+	pch_spi_setclr_reg(data->current_chip->master, PCH_SPCR, SPCR_SPE_BIT, 0);
+
+	/* Wait until the transfer completes; go to sleep after
+				 initiating the transfer. */
+	dev_dbg(&data->master->dev,
+		"%s:waiting for transfer to get over\n", __func__);
+
+	wait_event_interruptible(data->wait, data->transfer_complete);
+
+	pch_spi_writereg(data->master, PCH_SSNXCR, SSN_NO_CONTROL);
+	dev_dbg(&data->master->dev,
+		"%s:no more control over SSN-writing 0 to SSNXCR.", __func__);
+
+	data->transfer_active = false;
+	dev_dbg(&data->master->dev,
+		"%s set data->transfer_active = false\n", __func__);
+
+	/* clear all interrupts */
+	pch_spi_writereg(data->master, PCH_SPSR,
+			 pch_spi_readreg(data->master, PCH_SPSR));
+	/* disable interrupts */
+	pch_spi_setclr_reg(data->master, PCH_SPCR, 0, PCH_ALL);
+}
+
+static void pch_spi_copy_rx_data(struct pch_spi_data *data, int bpw)
+{
+	int j;
+	u8 *rx_buf;
+	u16 *rx_sbuf;
+
+	/* copy Rx Data */
+	if (!data->cur_trans->rx_buf)
+		return;
+
+	if (bpw == 8) {
+		rx_buf = data->cur_trans->rx_buf;
+		for (j = 0; j < data->bpw_len; j++)
+			*rx_buf++ = data->pkt_rx_buff[j] & 0xFF;
+	} else {
+		rx_sbuf = data->cur_trans->rx_buf;
+		for (j = 0; j < data->bpw_len; j++)
+			*rx_sbuf++ = data->pkt_rx_buff[j];
+	}
+}
+
+
+static void pch_spi_process_messages(struct work_struct *pwork)
+{
+	struct spi_message *pmsg;
+	struct pch_spi_data *data;
+	int bpw;
+
+	data = container_of(pwork, struct pch_spi_data, work);
+	dev_dbg(&data->master->dev, "%s data initialized\n", __func__);
+
+	spin_lock(&data->lock);
+
+	/* check if suspend has been initiated;if yes flush queue */
+	if (data->board_dat->suspend_sts || (data->status == STATUS_EXITING)) {
+		dev_dbg(&data->master->dev,
+			"%s suspend/remove initiated,flushing queue\n",
+			__func__);
+
+		list_for_each_entry(pmsg, data->queue.next, queue) {
+			pmsg->status = -EIO;
+
+			if (pmsg->complete != 0) {
+				spin_unlock(&data->lock);
+				pmsg->complete(pmsg->context);
+				spin_lock(&data->lock);
+			}
+
+			/* delete from queue */
+			list_del_init(&pmsg->queue);
+		}
+
+		spin_unlock(&data->lock);
+		return;
+	}
+
+	data->bcurrent_msg_processing = true;
+	dev_dbg(&data->master->dev,
+		"%s Set data->bcurrent_msg_processing= true\n", __func__);
+
+	/* Get the message from the queue and delete it from there. */
+	data->current_msg = list_entry(data->queue.next, struct spi_message,
+					queue);
+
+	list_del_init(&data->current_msg->queue);
+
+	data->current_msg->status = 0;
+
+	pch_spi_select_chip(data, data->current_msg->spi);
+
+	spin_unlock(&data->lock);
+
+	do {
+		/* If we are already processing a message get the next
+		transfer structure from the message otherwise retrieve
+		the 1st transfer request from the message. */
+		spin_lock(&data->lock);
+
+		if (data->cur_trans == NULL) {
+			data->cur_trans =
+			    list_entry(data->current_msg->transfers.
+				       next, struct spi_transfer,
+				       transfer_list);
+			dev_dbg(&data->master->dev,
+				"%s :Getting 1st transfer message\n", __func__);
+		} else {
+			data->cur_trans =
+			    list_entry(data->cur_trans->transfer_list.next,
+				       struct spi_transfer,
+				       transfer_list);
+			dev_dbg(&data->master->dev,
+				"%s :Getting next transfer message\n",
+				__func__);
+		}
+
+		spin_unlock(&data->lock);
+
+		pch_spi_set_tx(data, &bpw, &pmsg);
+
+		/* Control interrupt*/
+		pch_spi_set_ir(data);
+
+		/* Disable SPI transfer */
+		pch_spi_setclr_reg(data->current_chip->master, PCH_SPCR, 0,
+				   SPCR_SPE_BIT);
+
+		/* clear FIFO */
+		pch_spi_clear_fifo(data->master);
+
+		/* copy Rx Data */
+		pch_spi_copy_rx_data(data, bpw);
+
+		/* free memory */
+		kfree(data->pkt_rx_buff);
+		data->pkt_rx_buff = NULL;
+
+		kfree(data->pkt_tx_buff);
+		data->pkt_tx_buff = NULL;
+
+		/* increment message count */
+		data->current_msg->actual_length += data->cur_trans->len;
+
+		dev_dbg(&data->master->dev,
+			"%s:data->current_msg->actual_length=%d\n",
+			__func__, data->current_msg->actual_length);
+
+		/* check for delay */
+		if (data->cur_trans->delay_usecs) {
+			dev_dbg(&data->master->dev, "%s:"
+				"delay in usec=%d\n", __func__,
+				data->cur_trans->delay_usecs);
+			udelay(data->cur_trans->delay_usecs);
+		}
+
+		spin_lock(&data->lock);
+
+		/* No more transfer in this message. */
+		if ((data->cur_trans->transfer_list.next) ==
+		    &(data->current_msg->transfers)) {
+			pch_spi_nomore_transfer(data, pmsg);
+		}
+
+		spin_unlock(&data->lock);
+
+	} while (data->cur_trans != NULL);
+}
+
+static void pch_spi_free_resources(struct pch_spi_board_data *board_dat)
+{
+	dev_dbg(&board_dat->pdev->dev, "%s ENTRY\n", __func__);
+
+	/* free workqueue */
+	if (board_dat->data->wk != NULL) {
+		destroy_workqueue(board_dat->data->wk);
+		board_dat->data->wk = NULL;
+		dev_dbg(&board_dat->pdev->dev,
+			"%s destroy_workqueue invoked successfully\n",
+			__func__);
+	}
+
+	/* disable interrupts & free IRQ */
+	if (board_dat->irq_reg_sts) {
+		/* disable interrupts */
+		pch_spi_setclr_reg(board_dat->data->master, PCH_SPCR, 0,
+				   PCH_ALL);
+
+		/* free IRQ */
+		free_irq(board_dat->pdev->irq, board_dat);
+
+		dev_dbg(&board_dat->pdev->dev,
+			"%s free_irq invoked successfully\n", __func__);
+
+		board_dat->irq_reg_sts = false;
+	}
+
+	/* unmap PCI base address */
+	if (board_dat->data->io_remap_addr != 0) {
+		pci_iounmap(board_dat->pdev, board_dat->data->io_remap_addr);
+
+		board_dat->data->io_remap_addr = 0;
+
+		dev_dbg(&board_dat->pdev->dev,
+			"%s pci_iounmap invoked successfully\n", __func__);
+	}
+
+	/* release PCI region */
+	if (board_dat->pci_req_sts) {
+		pci_release_regions(board_dat->pdev);
+		dev_dbg(&board_dat->pdev->dev,
+			"%s pci_release_regions invoked successfully\n",
+			__func__);
+		board_dat->pci_req_sts = false;
+	}
+}
+
+static int pch_spi_get_resources(struct pch_spi_board_data *board_dat)
+{
+	void __iomem *io_remap_addr;
+	int retval;
+	dev_dbg(&board_dat->pdev->dev, "%s ENTRY\n", __func__);
+
+	/* create workqueue */
+	board_dat->data->wk = create_singlethread_workqueue(KBUILD_MODNAME);
+	if (!board_dat->data->wk) {
+		dev_err(&board_dat->pdev->dev,
+			"%s create_singlet hread_workqueue failed\n", __func__);
+		retval = -EBUSY;
+		goto err_return;
+	}
+
+	dev_dbg(&board_dat->pdev->dev,
+		"%s create_singlethread_workqueue success\n", __func__);
+
+	retval = pci_request_regions(board_dat->pdev, KBUILD_MODNAME);
+	if (retval != 0) {
+		dev_err(&board_dat->pdev->dev,
+			"%s request_region failed\n", __func__);
+		goto err_return;
+	}
+
+	board_dat->pci_req_sts = true;
+
+	io_remap_addr = pci_iomap(board_dat->pdev, 1, 0);
+	if (io_remap_addr == 0) {
+		dev_err(&board_dat->pdev->dev,
+			"%s pci_iomap failed\n", __func__);
+		retval = -ENOMEM;
+		goto err_return;
+	}
+
+	/* calculate base address for all channels */
+	board_dat->data->io_remap_addr = io_remap_addr;
+
+	/* reset PCH SPI h/w */
+	pch_spi_reset(board_dat->data->master);
+	dev_dbg(&board_dat->pdev->dev,
+		"%s pch_spi_reset invoked successfully\n", __func__);
+
+	/* register IRQ */
+	retval = request_irq(board_dat->pdev->irq, pch_spi_handler,
+			     IRQF_SHARED, KBUILD_MODNAME, board_dat);
+	if (retval != 0) {
+		dev_err(&board_dat->pdev->dev,
+			"%s request_irq failed\n", __func__);
+		goto err_return;
+	}
+
+	dev_dbg(&board_dat->pdev->dev, "%s request_irq returned=%d\n",
+		__func__, retval);
+
+	board_dat->irq_reg_sts = true;
+	dev_dbg(&board_dat->pdev->dev, "%s data->irq_reg_sts=true\n", __func__);
+
+err_return:
+	if (retval != 0) {
+		dev_err(&board_dat->pdev->dev,
+			"%s FAIL:invoking pch_spi_free_resources\n", __func__);
+		pch_spi_free_resources(board_dat);
+	}
+
+	dev_dbg(&board_dat->pdev->dev, "%s Return=%d\n", __func__, retval);
+
+	return retval;
+}
+
+static int pch_spi_probe(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+
+	struct spi_master *master;
+
+	struct pch_spi_board_data *board_dat;
+	int retval;
+
+	dev_dbg(&pdev->dev, "%s ENTRY\n", __func__);
+
+	/* allocate memory for private data */
+	board_dat = kzalloc(sizeof(struct pch_spi_board_data), GFP_KERNEL);
+	if (board_dat == NULL) {
+		dev_err(&pdev->dev,
+			" %s memory allocation for private data failed\n",
+			__func__);
+		retval = -ENOMEM;
+		goto err_kmalloc;
+	}
+
+	dev_dbg(&pdev->dev,
+		"%s memory allocation for private data success\n", __func__);
+
+	/* enable PCI device */
+	retval = pci_enable_device(pdev);
+	if (retval != 0) {
+		dev_err(&pdev->dev, "%s pci_enable_device FAILED\n", __func__);
+
+		goto err_pci_en_device;
+	}
+
+	dev_dbg(&pdev->dev, "%s pci_enable_device returned=%d\n",
+		__func__, retval);
+
+	board_dat->pdev = pdev;
+
+	/* alllocate memory for SPI master */
+	master = spi_alloc_master(&pdev->dev, sizeof(struct pch_spi_data));
+	if (master == NULL) {
+		retval = -ENOMEM;
+		dev_err(&pdev->dev, "%s Fail.\n", __func__);
+		goto err_spi_alloc_master;
+	}
+
+	dev_dbg(&pdev->dev,
+		"%s spi_alloc_master returned non NULL\n", __func__);
+
+	/* initialize members of SPI master */
+	master->bus_num = -1;
+	master->num_chipselect = PCH_MAX_CS;
+	master->setup = pch_spi_setup;
+	master->transfer = pch_spi_transfer;
+	dev_dbg(&pdev->dev,
+		"%s transfer member of SPI master initialized\n", __func__);
+
+	board_dat->data = spi_master_get_devdata(master);
+
+	board_dat->data->master = master;
+	board_dat->data->n_curnt_chip = 255;
+	board_dat->data->board_dat = board_dat;
+	board_dat->data->status = STATUS_RUNNING;
+
+	INIT_LIST_HEAD(&board_dat->data->queue);
+	spin_lock_init(&board_dat->data->lock);
+	INIT_WORK(&board_dat->data->work, pch_spi_process_messages);
+	init_waitqueue_head(&board_dat->data->wait);
+
+	/* allocate resources for PCH SPI */
+	retval = pch_spi_get_resources(board_dat);
+	if (retval) {
+		dev_err(&pdev->dev, "%s fail(retval=%d)\n", __func__, retval);
+		goto err_spi_get_resources;
+	}
+
+	dev_dbg(&pdev->dev, "%s pch_spi_get_resources returned=%d\n",
+		__func__, retval);
+
+	/* save private data in dev */
+	pci_set_drvdata(pdev, board_dat);
+	dev_dbg(&pdev->dev, "%s invoked pci_set_drvdata\n", __func__);
+
+	/* set master mode */
+	pch_spi_set_master_mode(master);
+	dev_dbg(&pdev->dev,
+		"%s invoked pch_spi_set_master_mode\n", __func__);
+
+	/* Register the controller with the SPI core. */
+	retval = spi_register_master(master);
+	if (retval != 0) {
+		dev_err(&pdev->dev,
+			"%s spi_register_master FAILED\n", __func__);
+		goto err_spi_reg_master;
+	}
+
+	dev_dbg(&pdev->dev, "%s spi_register_master returned=%d\n",
+		__func__, retval);
+
+
+	return 0;
+
+err_spi_reg_master:
+	spi_unregister_master(master);
+err_spi_get_resources:
+err_spi_alloc_master:
+	spi_master_put(master);
+	pci_disable_device(pdev);
+err_pci_en_device:
+	kfree(board_dat);
+err_kmalloc:
+	return retval;
+}
+
+static void pch_spi_remove(struct pci_dev *pdev)
+{
+	struct pch_spi_board_data *board_dat = pci_get_drvdata(pdev);
+	int count;
+
+	dev_dbg(&pdev->dev, "%s ENTRY\n", __func__);
+
+	if (!board_dat) {
+		dev_err(&pdev->dev,
+			"%s pci_get_drvdata returned NULL\n", __func__);
+		return;
+	}
+
+	/* check for any pending messages; no action is taken if the queue
+	 * is still full; but at least we tried.  Unload anyway */
+	count = 500;
+	spin_lock(&board_dat->data->lock);
+	board_dat->data->status = STATUS_EXITING;
+	while ((list_empty(&board_dat->data->queue) == 0) && --count) {
+		dev_dbg(&board_dat->pdev->dev, "%s :queue not empty\n",
+			__func__);
+		spin_unlock(&board_dat->data->lock);
+		msleep(PCH_SLEEP_TIME);
+		spin_lock(&board_dat->data->lock);
+	}
+	spin_unlock(&board_dat->data->lock);
+
+	/* Free resources allocated for PCH SPI */
+	pch_spi_free_resources(board_dat);
+
+	spi_unregister_master(board_dat->data->master);
+
+	/* free memory for private data */
+	kfree(board_dat);
+
+	pci_set_drvdata(pdev, NULL);
+
+	/* disable PCI device */
+	pci_disable_device(pdev);
+
+	dev_dbg(&pdev->dev, "%s invoked pci_disable_device\n", __func__);
+}
+
+#ifdef CONFIG_PM
+static int pch_spi_suspend(struct pci_dev *pdev, pm_message_t state)
+{
+	u8 count;
+	int retval;
+
+	struct pch_spi_board_data *board_dat = pci_get_drvdata(pdev);
+
+	dev_dbg(&pdev->dev, "%s ENTRY\n", __func__);
+
+	if (!board_dat) {
+		dev_err(&pdev->dev,
+			"%s pci_get_drvdata returned NULL\n", __func__);
+		return -EFAULT;
+	}
+
+	retval = 0;
+	board_dat->suspend_sts = true;
+
+	/* check if the current message is processed:
+	   Only after thats done the transfer will be suspended */
+	count = 255;
+	while ((--count) > 0) {
+		if (!(board_dat->data->bcurrent_msg_processing)) {
+			dev_dbg(&pdev->dev, "%s board_dat->data->bCurrent_"
+				"msg_processing = false\n", __func__);
+			break;
+		} else {
+			dev_dbg(&pdev->dev, "%s board_dat->data->bCurrent_msg_"
+				"processing = true\n", __func__);
+		}
+		msleep(PCH_SLEEP_TIME);
+	}
+
+	/* Free IRQ */
+	if (board_dat->irq_reg_sts) {
+		/* disable all interrupts */
+		pch_spi_setclr_reg(board_dat->data->master, PCH_SPCR, 0,
+				   PCH_ALL);
+		pch_spi_reset(board_dat->data->master);
+
+		free_irq(board_dat->pdev->irq, board_dat);
+
+		board_dat->irq_reg_sts = false;
+		dev_dbg(&pdev->dev,
+			"%s free_irq invoked successfully.\n", __func__);
+	}
+
+	/* save config space */
+	retval = pci_save_state(pdev);
+
+	if (retval == 0) {
+		dev_dbg(&pdev->dev, "%s pci_save_state returned=%d\n",
+			__func__, retval);
+		/* disable PM notifications */
+		pci_enable_wake(pdev, PCI_D3hot, 0);
+		dev_dbg(&pdev->dev,
+			"%s pci_enable_wake invoked successfully\n", __func__);
+		/* disable PCI device */
+		pci_disable_device(pdev);
+		dev_dbg(&pdev->dev,
+			"%s pci_disable_device invoked successfully\n",
+			__func__);
+		/* move device to D3hot  state */
+		pci_set_power_state(pdev, PCI_D3hot);
+		dev_dbg(&pdev->dev,
+			"%s pci_set_power_state invoked successfully\n",
+			__func__);
+	} else {
+		dev_err(&pdev->dev, "%s pci_save_state failed\n", __func__);
+	}
+
+	dev_dbg(&pdev->dev, "%s return=%d\n", __func__, retval);
+
+	return retval;
+}
+
+static int pch_spi_resume(struct pci_dev *pdev)
+{
+	int retval;
+
+	struct pch_spi_board_data *board = pci_get_drvdata(pdev);
+	dev_dbg(&pdev->dev, "%s ENTRY\n", __func__);
+
+	if (!board) {
+		dev_err(&pdev->dev,
+			"%s pci_get_drvdata returned NULL\n", __func__);
+		return -EFAULT;
+	}
+
+	/* move device to DO power state */
+	pci_set_power_state(pdev, PCI_D0);
+
+	/* restore state */
+	pci_restore_state(pdev);
+
+	retval = pci_enable_device(pdev);
+	if (retval < 0) {
+		dev_err(&pdev->dev,
+			"%s pci_enable_device failed\n", __func__);
+	} else {
+		/* disable PM notifications */
+		pci_enable_wake(pdev, PCI_D3hot, 0);
+
+		/* register IRQ handler */
+		if (!board->irq_reg_sts) {
+			/* register IRQ */
+			retval = request_irq(board->pdev->irq, pch_spi_handler,
+					     IRQF_SHARED, KBUILD_MODNAME,
+					     board);
+			if (retval < 0) {
+				dev_err(&pdev->dev,
+					"%s request_irq failed\n", __func__);
+				return retval;
+			}
+			board->irq_reg_sts = true;
+
+			/* reset PCH SPI h/w */
+			pch_spi_reset(board->data->master);
+			pch_spi_set_master_mode(board->data->master);
+
+			/* set suspend status to false */
+			board->suspend_sts = false;
+
+		}
+	}
+
+	dev_dbg(&pdev->dev, "%s returning=%d\n", __func__, retval);
+
+	return retval;
+}
+#else
+#define pch_spi_suspend NULL
+#define pch_spi_resume NULL
+
+#endif
+
+static struct pci_driver pch_spi_pcidev = {
+	.name = "pch_spi",
+	.id_table = pch_spi_pcidev_id,
+	.probe = pch_spi_probe,
+	.remove = pch_spi_remove,
+	.suspend = pch_spi_suspend,
+	.resume = pch_spi_resume,
+};
+
+static int __init pch_spi_init(void)
+{
+	return pci_register_driver(&pch_spi_pcidev);
+}
+module_init(pch_spi_init);
+
+static void __exit pch_spi_exit(void)
+{
+	pci_unregister_driver(&pch_spi_pcidev);
+}
+module_exit(pch_spi_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Topcliff PCH SPI PCI Driver");
diff --git a/include/linux/amba/pl022.h b/include/linux/amba/pl022.h
index abf26cc..4ce98f5 100644
--- a/include/linux/amba/pl022.h
+++ b/include/linux/amba/pl022.h
@@ -228,6 +228,7 @@
 };
 
 
+struct dma_chan;
 /**
  * struct pl022_ssp_master - device.platform_data for SPI controller devices.
  * @num_chipselect: chipselects are used to distinguish individual
@@ -235,11 +236,16 @@
  *     each slave has a chipselect signal, but it's common that not
  *     every chipselect is connected to a slave.
  * @enable_dma: if true enables DMA driven transfers.
+ * @dma_rx_param: parameter to locate an RX DMA channel.
+ * @dma_tx_param: parameter to locate a TX DMA channel.
  */
 struct pl022_ssp_controller {
 	u16 bus_id;
 	u8 num_chipselect;
 	u8 enable_dma:1;
+	bool (*dma_filter)(struct dma_chan *chan, void *filter_param);
+	void *dma_rx_param;
+	void *dma_tx_param;
 };
 
 /**
@@ -270,20 +276,13 @@
  * @dma_config: DMA configuration for SSP controller and peripheral
  */
 struct pl022_config_chip {
-	struct device *dev;
-	enum ssp_loopback lbm;
 	enum ssp_interface iface;
 	enum ssp_hierarchy hierarchy;
 	bool slave_tx_disable;
 	struct ssp_clock_params clk_freq;
-	enum ssp_rx_endian endian_rx;
-	enum ssp_tx_endian endian_tx;
-	enum ssp_data_size data_size;
 	enum ssp_mode com_mode;
 	enum ssp_rx_level_trig rx_lev_trig;
 	enum ssp_tx_level_trig tx_lev_trig;
-	enum ssp_spi_clk_phase clk_phase;
-	enum ssp_spi_clk_pol clk_pol;
 	enum ssp_microwire_ctrl_len ctrl_len;
 	enum ssp_microwire_wait_state wait_state;
 	enum ssp_duplex duplex;