Merge branch 'fix/dw' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi into spi-dw

Conflicts:
	drivers/spi/spi-dw-mid.c

31 files changed, 1257 insertions, 358 deletions

diff --git a/drivers/spi/Kconfig b/drivers/spi/Kconfig
index 6343df8..84e7c9e 100644
--- a/drivers/spi/Kconfig
+++ b/drivers/spi/Kconfig
@@ -69,6 +69,7 @@ config SPI_ATH79
 
 config SPI_ATMEL
 	tristate "Atmel SPI Controller"
+	depends on HAS_DMA
 	depends on (ARCH_AT91 || AVR32 || COMPILE_TEST)
 	help
 	  This selects a driver for the Atmel SPI Controller, present on
@@ -112,6 +113,14 @@ config SPI_AU1550
 	  If you say yes to this option, support will be included for the
 	  PSC SPI controller found on Au1550, Au1200 and Au1300 series.
 
+config SPI_BCM53XX
+	tristate "Broadcom BCM53xx SPI controller"
+	depends on ARCH_BCM_5301X
+	depends on BCMA_POSSIBLE
+	select BCMA
+	help
+          Enable support for the SPI controller on Broadcom BCM53xx ARM SoCs.
+
 config SPI_BCM63XX
 	tristate "Broadcom BCM63xx SPI controller"
 	depends on BCM63XX
@@ -185,6 +194,7 @@ config SPI_EFM32
 
 config SPI_EP93XX
 	tristate "Cirrus Logic EP93xx SPI controller"
+	depends on HAS_DMA
 	depends on ARCH_EP93XX || COMPILE_TEST
 	help
 	  This enables using the Cirrus EP93xx SPI controller in master
@@ -314,6 +324,7 @@ config SPI_OMAP_UWIRE
 
 config SPI_OMAP24XX
 	tristate "McSPI driver for OMAP"
+	depends on HAS_DMA
 	depends on ARM || ARM64 || AVR32 || HEXAGON || MIPS || SUPERH
 	depends on ARCH_OMAP2PLUS || COMPILE_TEST
 	help
@@ -380,7 +391,7 @@ config SPI_PXA2XX
 	  additional documentation can be found a Documentation/spi/pxa2xx.
 
 config SPI_PXA2XX_PCI
-	def_tristate SPI_PXA2XX && PCI
+	def_tristate SPI_PXA2XX && PCI && COMMON_CLK
 
 config SPI_ROCKCHIP
 	tristate "Rockchip SPI controller driver"
@@ -500,7 +511,7 @@ config SPI_MXS
 config SPI_TEGRA114
 	tristate "NVIDIA Tegra114 SPI Controller"
 	depends on (ARCH_TEGRA && TEGRA20_APB_DMA) || COMPILE_TEST
-	depends on RESET_CONTROLLER
+	depends on RESET_CONTROLLER && HAS_DMA
 	help
 	  SPI driver for NVIDIA Tegra114 SPI Controller interface. This controller
 	  is different than the older SoCs SPI controller and also register interface
@@ -518,7 +529,7 @@ config SPI_TEGRA20_SFLASH
 config SPI_TEGRA20_SLINK
 	tristate "Nvidia Tegra20/Tegra30 SLINK Controller"
 	depends on (ARCH_TEGRA && TEGRA20_APB_DMA) || COMPILE_TEST
-	depends on RESET_CONTROLLER
+	depends on RESET_CONTROLLER && HAS_DMA
 	help
 	  SPI driver for Nvidia Tegra20/Tegra30 SLINK Controller interface.
 
diff --git a/drivers/spi/Makefile b/drivers/spi/Makefile
index 762da07..78f24ca 100644
--- a/drivers/spi/Makefile
+++ b/drivers/spi/Makefile
@@ -15,6 +15,7 @@ obj-$(CONFIG_SPI_ATMEL)			+= spi-atmel.o
 obj-$(CONFIG_SPI_ATH79)			+= spi-ath79.o
 obj-$(CONFIG_SPI_AU1550)		+= spi-au1550.o
 obj-$(CONFIG_SPI_BCM2835)		+= spi-bcm2835.o
+obj-$(CONFIG_SPI_BCM53XX)		+= spi-bcm53xx.o
 obj-$(CONFIG_SPI_BCM63XX)		+= spi-bcm63xx.o
 obj-$(CONFIG_SPI_BCM63XX_HSSPI)		+= spi-bcm63xx-hsspi.o
 obj-$(CONFIG_SPI_BFIN5XX)		+= spi-bfin5xx.o
diff --git a/drivers/spi/spi-au1550.c b/drivers/spi/spi-au1550.c
index 40c3d43..f40b34c 100644
--- a/drivers/spi/spi-au1550.c
+++ b/drivers/spi/spi-au1550.c
@@ -945,7 +945,7 @@ static int au1550_spi_remove(struct platform_device *pdev)
 	spi_bitbang_stop(&hw->bitbang);
 	free_irq(hw->irq, hw);
 	iounmap((void __iomem *)hw->regs);
-	release_mem_region(r->start, sizeof(psc_spi_t));
+	release_mem_region(hw->ioarea->start, sizeof(psc_spi_t));
 
 	if (hw->usedma) {
 		au1550_spi_dma_rxtmp_free(hw);
diff --git a/drivers/spi/spi-bcm53xx.c b/drivers/spi/spi-bcm53xx.c
new file mode 100644
index 0000000..17b34cb
--- /dev/null
+++ b/drivers/spi/spi-bcm53xx.c
@@ -0,0 +1,299 @@
+#define pr_fmt(fmt)		KBUILD_MODNAME ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/delay.h>
+#include <linux/bcma/bcma.h>
+#include <linux/spi/spi.h>
+
+#include "spi-bcm53xx.h"
+
+#define BCM53XXSPI_MAX_SPI_BAUD	13500000	/* 216 MHz? */
+
+/* The longest observed required wait was 19 ms */
+#define BCM53XXSPI_SPE_TIMEOUT_MS	80
+
+struct bcm53xxspi {
+	struct bcma_device *core;
+	struct spi_master *master;
+
+	size_t read_offset;
+};
+
+static inline u32 bcm53xxspi_read(struct bcm53xxspi *b53spi, u16 offset)
+{
+	return bcma_read32(b53spi->core, offset);
+}
+
+static inline void bcm53xxspi_write(struct bcm53xxspi *b53spi, u16 offset,
+				    u32 value)
+{
+	bcma_write32(b53spi->core, offset, value);
+}
+
+static inline unsigned int bcm53xxspi_calc_timeout(size_t len)
+{
+	/* Do some magic calculation based on length and buad. Add 10% and 1. */
+	return (len * 9000 / BCM53XXSPI_MAX_SPI_BAUD * 110 / 100) + 1;
+}
+
+static int bcm53xxspi_wait(struct bcm53xxspi *b53spi, unsigned int timeout_ms)
+{
+	unsigned long deadline;
+	u32 tmp;
+
+	/* SPE bit has to be 0 before we read MSPI STATUS */
+	deadline = jiffies + BCM53XXSPI_SPE_TIMEOUT_MS * HZ / 1000;
+	do {
+		tmp = bcm53xxspi_read(b53spi, B53SPI_MSPI_SPCR2);
+		if (!(tmp & B53SPI_MSPI_SPCR2_SPE))
+			break;
+		udelay(5);
+	} while (!time_after_eq(jiffies, deadline));
+
+	if (tmp & B53SPI_MSPI_SPCR2_SPE)
+		goto spi_timeout;
+
+	/* Check status */
+	deadline = jiffies + timeout_ms * HZ / 1000;
+	do {
+		tmp = bcm53xxspi_read(b53spi, B53SPI_MSPI_MSPI_STATUS);
+		if (tmp & B53SPI_MSPI_MSPI_STATUS_SPIF) {
+			bcm53xxspi_write(b53spi, B53SPI_MSPI_MSPI_STATUS, 0);
+			return 0;
+		}
+
+		cpu_relax();
+		udelay(100);
+	} while (!time_after_eq(jiffies, deadline));
+
+spi_timeout:
+	bcm53xxspi_write(b53spi, B53SPI_MSPI_MSPI_STATUS, 0);
+
+	pr_err("Timeout waiting for SPI to be ready!\n");
+
+	return -EBUSY;
+}
+
+static void bcm53xxspi_buf_write(struct bcm53xxspi *b53spi, u8 *w_buf,
+				 size_t len, bool cont)
+{
+	u32 tmp;
+	int i;
+
+	for (i = 0; i < len; i++) {
+		/* Transmit Register File MSB */
+		bcm53xxspi_write(b53spi, B53SPI_MSPI_TXRAM + 4 * (i * 2),
+				 (unsigned int)w_buf[i]);
+	}
+
+	for (i = 0; i < len; i++) {
+		tmp = B53SPI_CDRAM_CONT | B53SPI_CDRAM_PCS_DISABLE_ALL |
+		      B53SPI_CDRAM_PCS_DSCK;
+		if (!cont && i == len - 1)
+			tmp &= ~B53SPI_CDRAM_CONT;
+		tmp &= ~0x1;
+		/* Command Register File */
+		bcm53xxspi_write(b53spi, B53SPI_MSPI_CDRAM + 4 * i, tmp);
+	}
+
+	/* Set queue pointers */
+	bcm53xxspi_write(b53spi, B53SPI_MSPI_NEWQP, 0);
+	bcm53xxspi_write(b53spi, B53SPI_MSPI_ENDQP, len - 1);
+
+	if (cont)
+		bcm53xxspi_write(b53spi, B53SPI_MSPI_WRITE_LOCK, 1);
+
+	/* Start SPI transfer */
+	tmp = bcm53xxspi_read(b53spi, B53SPI_MSPI_SPCR2);
+	tmp |= B53SPI_MSPI_SPCR2_SPE;
+	if (cont)
+		tmp |= B53SPI_MSPI_SPCR2_CONT_AFTER_CMD;
+	bcm53xxspi_write(b53spi, B53SPI_MSPI_SPCR2, tmp);
+
+	/* Wait for SPI to finish */
+	bcm53xxspi_wait(b53spi, bcm53xxspi_calc_timeout(len));
+
+	if (!cont)
+		bcm53xxspi_write(b53spi, B53SPI_MSPI_WRITE_LOCK, 0);
+
+	b53spi->read_offset = len;
+}
+
+static void bcm53xxspi_buf_read(struct bcm53xxspi *b53spi, u8 *r_buf,
+				size_t len, bool cont)
+{
+	u32 tmp;
+	int i;
+
+	for (i = 0; i < b53spi->read_offset + len; i++) {
+		tmp = B53SPI_CDRAM_CONT | B53SPI_CDRAM_PCS_DISABLE_ALL |
+		      B53SPI_CDRAM_PCS_DSCK;
+		if (!cont && i == b53spi->read_offset + len - 1)
+			tmp &= ~B53SPI_CDRAM_CONT;
+		tmp &= ~0x1;
+		/* Command Register File */
+		bcm53xxspi_write(b53spi, B53SPI_MSPI_CDRAM + 4 * i, tmp);
+	}
+
+	/* Set queue pointers */
+	bcm53xxspi_write(b53spi, B53SPI_MSPI_NEWQP, 0);
+	bcm53xxspi_write(b53spi, B53SPI_MSPI_ENDQP,
+			 b53spi->read_offset + len - 1);
+
+	if (cont)
+		bcm53xxspi_write(b53spi, B53SPI_MSPI_WRITE_LOCK, 1);
+
+	/* Start SPI transfer */
+	tmp = bcm53xxspi_read(b53spi, B53SPI_MSPI_SPCR2);
+	tmp |= B53SPI_MSPI_SPCR2_SPE;
+	if (cont)
+		tmp |= B53SPI_MSPI_SPCR2_CONT_AFTER_CMD;
+	bcm53xxspi_write(b53spi, B53SPI_MSPI_SPCR2, tmp);
+
+	/* Wait for SPI to finish */
+	bcm53xxspi_wait(b53spi, bcm53xxspi_calc_timeout(len));
+
+	if (!cont)
+		bcm53xxspi_write(b53spi, B53SPI_MSPI_WRITE_LOCK, 0);
+
+	for (i = 0; i < len; ++i) {
+		int offset = b53spi->read_offset + i;
+
+		/* Data stored in the transmit register file LSB */
+		r_buf[i] = (u8)bcm53xxspi_read(b53spi, B53SPI_MSPI_RXRAM + 4 * (1 + offset * 2));
+	}
+
+	b53spi->read_offset = 0;
+}
+
+static int bcm53xxspi_transfer_one(struct spi_master *master,
+				   struct spi_device *spi,
+				   struct spi_transfer *t)
+{
+	struct bcm53xxspi *b53spi = spi_master_get_devdata(master);
+	u8 *buf;
+	size_t left;
+
+	if (t->tx_buf) {
+		buf = (u8 *)t->tx_buf;
+		left = t->len;
+		while (left) {
+			size_t to_write = min_t(size_t, 16, left);
+			bool cont = left - to_write > 0;
+
+			bcm53xxspi_buf_write(b53spi, buf, to_write, cont);
+			left -= to_write;
+			buf += to_write;
+		}
+	}
+
+	if (t->rx_buf) {
+		buf = (u8 *)t->rx_buf;
+		left = t->len;
+		while (left) {
+			size_t to_read = min_t(size_t, 16 - b53spi->read_offset,
+					       left);
+			bool cont = left - to_read > 0;
+
+			bcm53xxspi_buf_read(b53spi, buf, to_read, cont);
+			left -= to_read;
+			buf += to_read;
+		}
+	}
+
+	return 0;
+}
+
+/**************************************************
+ * BCMA
+ **************************************************/
+
+static struct spi_board_info bcm53xx_info = {
+	.modalias	= "bcm53xxspiflash",
+};
+
+static const struct bcma_device_id bcm53xxspi_bcma_tbl[] = {
+	BCMA_CORE(BCMA_MANUF_BCM, BCMA_CORE_NS_QSPI, BCMA_ANY_REV, BCMA_ANY_CLASS),
+	BCMA_CORETABLE_END
+};
+MODULE_DEVICE_TABLE(bcma, bcm53xxspi_bcma_tbl);
+
+static int bcm53xxspi_bcma_probe(struct bcma_device *core)
+{
+	struct bcm53xxspi *b53spi;
+	struct spi_master *master;
+	int err;
+
+	if (core->bus->drv_cc.core->id.rev != 42) {
+		pr_err("SPI on SoC with unsupported ChipCommon rev\n");
+		return -ENOTSUPP;
+	}
+
+	master = spi_alloc_master(&core->dev, sizeof(*b53spi));
+	if (!master)
+		return -ENOMEM;
+
+	b53spi = spi_master_get_devdata(master);
+	b53spi->master = master;
+	b53spi->core = core;
+
+	master->transfer_one = bcm53xxspi_transfer_one;
+
+	bcma_set_drvdata(core, b53spi);
+
+	err = devm_spi_register_master(&core->dev, master);
+	if (err) {
+		spi_master_put(master);
+		bcma_set_drvdata(core, NULL);
+		goto out;
+	}
+
+	/* Broadcom SoCs (at least with the CC rev 42) use SPI for flash only */
+	spi_new_device(master, &bcm53xx_info);
+
+out:
+	return err;
+}
+
+static void bcm53xxspi_bcma_remove(struct bcma_device *core)
+{
+	struct bcm53xxspi *b53spi = bcma_get_drvdata(core);
+
+	spi_unregister_master(b53spi->master);
+}
+
+static struct bcma_driver bcm53xxspi_bcma_driver = {
+	.name		= KBUILD_MODNAME,
+	.id_table	= bcm53xxspi_bcma_tbl,
+	.probe		= bcm53xxspi_bcma_probe,
+	.remove		= bcm53xxspi_bcma_remove,
+};
+
+/**************************************************
+ * Init & exit
+ **************************************************/
+
+static int __init bcm53xxspi_module_init(void)
+{
+	int err = 0;
+
+	err = bcma_driver_register(&bcm53xxspi_bcma_driver);
+	if (err)
+		pr_err("Failed to register bcma driver: %d\n", err);
+
+	return err;
+}
+
+static void __exit bcm53xxspi_module_exit(void)
+{
+	bcma_driver_unregister(&bcm53xxspi_bcma_driver);
+}
+
+module_init(bcm53xxspi_module_init);
+module_exit(bcm53xxspi_module_exit);
+
+MODULE_DESCRIPTION("Broadcom BCM53xx SPI Controller driver");
+MODULE_AUTHOR("Rafał Miłecki <zajec5@gmail.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/spi/spi-bcm53xx.h b/drivers/spi/spi-bcm53xx.h
new file mode 100644
index 0000000..73575df
--- /dev/null
+++ b/drivers/spi/spi-bcm53xx.h
@@ -0,0 +1,72 @@
+#ifndef SPI_BCM53XX_H
+#define SPI_BCM53XX_H
+
+#define B53SPI_BSPI_REVISION_ID			0x000
+#define B53SPI_BSPI_SCRATCH			0x004
+#define B53SPI_BSPI_MAST_N_BOOT_CTRL		0x008
+#define B53SPI_BSPI_BUSY_STATUS			0x00c
+#define B53SPI_BSPI_INTR_STATUS			0x010
+#define B53SPI_BSPI_B0_STATUS			0x014
+#define B53SPI_BSPI_B0_CTRL			0x018
+#define B53SPI_BSPI_B1_STATUS			0x01c
+#define B53SPI_BSPI_B1_CTRL			0x020
+#define B53SPI_BSPI_STRAP_OVERRIDE_CTRL		0x024
+#define B53SPI_BSPI_FLEX_MODE_ENABLE		0x028
+#define B53SPI_BSPI_BITS_PER_CYCLE		0x02c
+#define B53SPI_BSPI_BITS_PER_PHASE		0x030
+#define B53SPI_BSPI_CMD_AND_MODE_BYTE		0x034
+#define B53SPI_BSPI_BSPI_FLASH_UPPER_ADDR_BYTE	0x038
+#define B53SPI_BSPI_BSPI_XOR_VALUE		0x03c
+#define B53SPI_BSPI_BSPI_XOR_ENABLE		0x040
+#define B53SPI_BSPI_BSPI_PIO_MODE_ENABLE	0x044
+#define B53SPI_BSPI_BSPI_PIO_IODIR		0x048
+#define B53SPI_BSPI_BSPI_PIO_DATA		0x04c
+
+/* RAF */
+#define B53SPI_RAF_START_ADDR			0x100
+#define B53SPI_RAF_NUM_WORDS			0x104
+#define B53SPI_RAF_CTRL				0x108
+#define B53SPI_RAF_FULLNESS			0x10c
+#define B53SPI_RAF_WATERMARK			0x110
+#define B53SPI_RAF_STATUS			0x114
+#define B53SPI_RAF_READ_DATA			0x118
+#define B53SPI_RAF_WORD_CNT			0x11c
+#define B53SPI_RAF_CURR_ADDR			0x120
+
+/* MSPI */
+#define B53SPI_MSPI_SPCR0_LSB			0x200
+#define B53SPI_MSPI_SPCR0_MSB			0x204
+#define B53SPI_MSPI_SPCR1_LSB			0x208
+#define B53SPI_MSPI_SPCR1_MSB			0x20c
+#define B53SPI_MSPI_NEWQP			0x210
+#define B53SPI_MSPI_ENDQP			0x214
+#define B53SPI_MSPI_SPCR2			0x218
+#define  B53SPI_MSPI_SPCR2_SPE			0x00000040
+#define  B53SPI_MSPI_SPCR2_CONT_AFTER_CMD	0x00000080
+#define B53SPI_MSPI_MSPI_STATUS			0x220
+#define  B53SPI_MSPI_MSPI_STATUS_SPIF		0x00000001
+#define B53SPI_MSPI_CPTQP			0x224
+#define B53SPI_MSPI_TXRAM			0x240 /* 32 registers, up to 0x2b8 */
+#define B53SPI_MSPI_RXRAM			0x2c0 /* 32 registers, up to 0x33c */
+#define B53SPI_MSPI_CDRAM			0x340 /* 16 registers, up to 0x37c */
+#define  B53SPI_CDRAM_PCS_PCS0			0x00000001
+#define  B53SPI_CDRAM_PCS_PCS1			0x00000002
+#define  B53SPI_CDRAM_PCS_PCS2			0x00000004
+#define  B53SPI_CDRAM_PCS_PCS3			0x00000008
+#define  B53SPI_CDRAM_PCS_DISABLE_ALL		0x0000000f
+#define  B53SPI_CDRAM_PCS_DSCK			0x00000010
+#define  B53SPI_CDRAM_BITSE			0x00000040
+#define  B53SPI_CDRAM_CONT			0x00000080
+#define B53SPI_MSPI_WRITE_LOCK			0x380
+#define B53SPI_MSPI_DISABLE_FLUSH_GEN		0x384
+
+/* Interrupt */
+#define B53SPI_INTR_RAF_LR_FULLNESS_REACHED	0x3a0
+#define B53SPI_INTR_RAF_LR_TRUNCATED		0x3a4
+#define B53SPI_INTR_RAF_LR_IMPATIENT		0x3a8
+#define B53SPI_INTR_RAF_LR_SESSION_DONE		0x3ac
+#define B53SPI_INTR_RAF_LR_OVERREAD		0x3b0
+#define B53SPI_INTR_MSPI_DONE			0x3b4
+#define B53SPI_INTR_MSPI_HALT_SET_TRANSACTION_DONE	0x3b8
+
+#endif /* SPI_BCM53XX_H */
diff --git a/drivers/spi/spi-cadence.c b/drivers/spi/spi-cadence.c
index 562ff83..7b811e3 100644
--- a/drivers/spi/spi-cadence.c
+++ b/drivers/spi/spi-cadence.c
@@ -677,7 +677,6 @@ static struct platform_driver cdns_spi_driver = {
 	.remove	= cdns_spi_remove,
 	.driver = {
 		.name = CDNS_SPI_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table = cdns_spi_of_match,
 		.pm = &cdns_spi_dev_pm_ops,
 	},
diff --git a/drivers/spi/spi-clps711x.c b/drivers/spi/spi-clps711x.c
index ce538da..181cf22 100644
--- a/drivers/spi/spi-clps711x.c
+++ b/drivers/spi/spi-clps711x.c
@@ -30,7 +30,6 @@
 struct spi_clps711x_data {
 	void __iomem		*syncio;
 	struct regmap		*syscon;
-	struct regmap		*syscon1;
 	struct clk		*spi_clk;
 
 	u8			*tx_buf;
@@ -47,27 +46,6 @@ static int spi_clps711x_setup(struct spi_device *spi)
 	return 0;
 }
 
-static void spi_clps711x_setup_xfer(struct spi_device *spi,
-				    struct spi_transfer *xfer)
-{
-	struct spi_master *master = spi->master;
-	struct spi_clps711x_data *hw = spi_master_get_devdata(master);
-
-	/* Setup SPI frequency divider */
-	if (xfer->speed_hz >= master->max_speed_hz)
-		regmap_update_bits(hw->syscon1, SYSCON_OFFSET,
-				   SYSCON1_ADCKSEL_MASK, SYSCON1_ADCKSEL(3));
-	else if (xfer->speed_hz >= (master->max_speed_hz / 2))
-		regmap_update_bits(hw->syscon1, SYSCON_OFFSET,
-				   SYSCON1_ADCKSEL_MASK, SYSCON1_ADCKSEL(2));
-	else if (xfer->speed_hz >= (master->max_speed_hz / 8))
-		regmap_update_bits(hw->syscon1, SYSCON_OFFSET,
-				   SYSCON1_ADCKSEL_MASK, SYSCON1_ADCKSEL(1));
-	else
-		regmap_update_bits(hw->syscon1, SYSCON_OFFSET,
-				   SYSCON1_ADCKSEL_MASK, SYSCON1_ADCKSEL(0));
-}
-
 static int spi_clps711x_prepare_message(struct spi_master *master,
 					struct spi_message *msg)
 {
@@ -87,7 +65,7 @@ static int spi_clps711x_transfer_one(struct spi_master *master,
 	struct spi_clps711x_data *hw = spi_master_get_devdata(master);
 	u8 data;
 
-	spi_clps711x_setup_xfer(spi, xfer);
+	clk_set_rate(hw->spi_clk, xfer->speed_hz ? : spi->max_speed_hz);
 
 	hw->len = xfer->len;
 	hw->bpw = xfer->bits_per_word;
@@ -176,13 +154,11 @@ static int spi_clps711x_probe(struct platform_device *pdev)
 		}
 	}
 
-	hw->spi_clk = devm_clk_get(&pdev->dev, "spi");
+	hw->spi_clk = devm_clk_get(&pdev->dev, NULL);
 	if (IS_ERR(hw->spi_clk)) {
-		dev_err(&pdev->dev, "Can't get clocks\n");
 		ret = PTR_ERR(hw->spi_clk);
 		goto err_out;
 	}
-	master->max_speed_hz = clk_get_rate(hw->spi_clk);
 
 	hw->syscon = syscon_regmap_lookup_by_pdevname("syscon.3");
 	if (IS_ERR(hw->syscon)) {
@@ -190,12 +166,6 @@ static int spi_clps711x_probe(struct platform_device *pdev)
 		goto err_out;
 	}
 
-	hw->syscon1 = syscon_regmap_lookup_by_pdevname("syscon.1");
-	if (IS_ERR(hw->syscon1)) {
-		ret = PTR_ERR(hw->syscon1);
-		goto err_out;
-	}
-
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	hw->syncio = devm_ioremap_resource(&pdev->dev, res);
 	if (IS_ERR(hw->syncio)) {
diff --git a/drivers/spi/spi-davinci.c b/drivers/spi/spi-davinci.c
index 514852c..63700ab 100644
--- a/drivers/spi/spi-davinci.c
+++ b/drivers/spi/spi-davinci.c
@@ -65,6 +65,7 @@
 
 /* SPIDAT1 (upper 16 bit defines) */
 #define SPIDAT1_CSHOLD_MASK	BIT(12)
+#define SPIDAT1_WDEL		BIT(10)
 
 /* SPIGCR1 */
 #define SPIGCR1_CLKMOD_MASK	BIT(1)
@@ -213,6 +214,7 @@ static void davinci_spi_chipselect(struct spi_device *spi, int value)
 {
 	struct davinci_spi *dspi;
 	struct davinci_spi_platform_data *pdata;
+	struct davinci_spi_config *spicfg = spi->controller_data;
 	u8 chip_sel = spi->chip_select;
 	u16 spidat1 = CS_DEFAULT;
 	bool gpio_chipsel = false;
@@ -227,6 +229,10 @@ static void davinci_spi_chipselect(struct spi_device *spi, int value)
 		gpio = spi->cs_gpio;
 	}
 
+	/* program delay transfers if tx_delay is non zero */
+	if (spicfg->wdelay)
+		spidat1 |= SPIDAT1_WDEL;
+
 	/*
 	 * Board specific chip select logic decides the polarity and cs
 	 * line for the controller
@@ -241,9 +247,9 @@ static void davinci_spi_chipselect(struct spi_device *spi, int value)
 			spidat1 |= SPIDAT1_CSHOLD_MASK;
 			spidat1 &= ~(0x1 << chip_sel);
 		}
-
-		iowrite16(spidat1, dspi->base + SPIDAT1 + 2);
 	}
+
+	iowrite16(spidat1, dspi->base + SPIDAT1 + 2);
 }
 
 /**
@@ -289,7 +295,7 @@ static int davinci_spi_setup_transfer(struct spi_device *spi,
 	int prescale;
 
 	dspi = spi_master_get_devdata(spi->master);
-	spicfg = (struct davinci_spi_config *)spi->controller_data;
+	spicfg = spi->controller_data;
 	if (!spicfg)
 		spicfg = &davinci_spi_default_cfg;
 
@@ -337,6 +343,14 @@ static int davinci_spi_setup_transfer(struct spi_device *spi,
 		spifmt |= SPIFMT_PHASE_MASK;
 
 	/*
+	* Assume wdelay is used only on SPI peripherals that has this field
+	* in SPIFMTn register and when it's configured from board file or DT.
+	*/
+	if (spicfg->wdelay)
+		spifmt |= ((spicfg->wdelay << SPIFMT_WDELAY_SHIFT)
+				& SPIFMT_WDELAY_MASK);
+
+	/*
 	 * Version 1 hardware supports two basic SPI modes:
 	 *  - Standard SPI mode uses 4 pins, with chipselect
 	 *  - 3 pin SPI is a 4 pin variant without CS (SPI_NO_CS)
@@ -353,9 +367,6 @@ static int davinci_spi_setup_transfer(struct spi_device *spi,
 
 		u32 delay = 0;
 
-		spifmt |= ((spicfg->wdelay << SPIFMT_WDELAY_SHIFT)
-							& SPIFMT_WDELAY_MASK);
-
 		if (spicfg->odd_parity)
 			spifmt |= SPIFMT_ODD_PARITY_MASK;
 
@@ -387,6 +398,26 @@ static int davinci_spi_setup_transfer(struct spi_device *spi,
 	return 0;
 }
 
+static int davinci_spi_of_setup(struct spi_device *spi)
+{
+	struct davinci_spi_config *spicfg = spi->controller_data;
+	struct device_node *np = spi->dev.of_node;
+	u32 prop;
+
+	if (spicfg == NULL && np) {
+		spicfg = kzalloc(sizeof(*spicfg), GFP_KERNEL);
+		if (!spicfg)
+			return -ENOMEM;
+		*spicfg = davinci_spi_default_cfg;
+		/* override with dt configured values */
+		if (!of_property_read_u32(np, "ti,spi-wdelay", &prop))
+			spicfg->wdelay = (u8)prop;
+		spi->controller_data = spicfg;
+	}
+
+	return 0;
+}
+
 /**
  * davinci_spi_setup - This functions will set default transfer method
  * @spi: spi device on which data transfer to be done
@@ -401,36 +432,33 @@ static int davinci_spi_setup(struct spi_device *spi)
 	struct spi_master *master = spi->master;
 	struct device_node *np = spi->dev.of_node;
 	bool internal_cs = true;
-	unsigned long flags = GPIOF_DIR_OUT;
 
 	dspi = spi_master_get_devdata(spi->master);
 	pdata = &dspi->pdata;
 
-	flags |= (spi->mode & SPI_CS_HIGH) ? GPIOF_INIT_LOW : GPIOF_INIT_HIGH;
-
 	if (!(spi->mode & SPI_NO_CS)) {
 		if (np && (master->cs_gpios != NULL) && (spi->cs_gpio >= 0)) {
-			retval = gpio_request_one(spi->cs_gpio,
-						  flags, dev_name(&spi->dev));
+			retval = gpio_direction_output(
+				      spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
 			internal_cs = false;
 		} else if (pdata->chip_sel &&
 			   spi->chip_select < pdata->num_chipselect &&
 			   pdata->chip_sel[spi->chip_select] != SPI_INTERN_CS) {
 			spi->cs_gpio = pdata->chip_sel[spi->chip_select];
-			retval = gpio_request_one(spi->cs_gpio,
-						  flags, dev_name(&spi->dev));
+			retval = gpio_direction_output(
+				      spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
 			internal_cs = false;
 		}
-	}
 
-	if (retval) {
-		dev_err(&spi->dev, "GPIO %d setup failed (%d)\n",
-			spi->cs_gpio, retval);
-		return retval;
-	}
+		if (retval) {
+			dev_err(&spi->dev, "GPIO %d setup failed (%d)\n",
+				spi->cs_gpio, retval);
+			return retval;
+		}
 
-	if (internal_cs)
-		set_io_bits(dspi->base + SPIPC0, 1 << spi->chip_select);
+		if (internal_cs)
+			set_io_bits(dspi->base + SPIPC0, 1 << spi->chip_select);
+	}
 
 	if (spi->mode & SPI_READY)
 		set_io_bits(dspi->base + SPIPC0, SPIPC0_SPIENA_MASK);
@@ -440,13 +468,16 @@ static int davinci_spi_setup(struct spi_device *spi)
 	else
 		clear_io_bits(dspi->base + SPIGCR1, SPIGCR1_LOOPBACK_MASK);
 
-	return retval;
+	return davinci_spi_of_setup(spi);
 }
 
 static void davinci_spi_cleanup(struct spi_device *spi)
 {
-	if (spi->cs_gpio >= 0)
-		gpio_free(spi->cs_gpio);
+	struct davinci_spi_config *spicfg = spi->controller_data;
+
+	spi->controller_data = NULL;
+	if (spi->dev.of_node)
+		kfree(spicfg);
 }
 
 static int davinci_spi_check_error(struct davinci_spi *dspi, int int_status)
@@ -971,6 +1002,27 @@ static int davinci_spi_probe(struct platform_device *pdev)
 	if (dspi->version == SPI_VERSION_2)
 		dspi->bitbang.flags |= SPI_READY;
 
+	if (pdev->dev.of_node) {
+		int i;
+
+		for (i = 0; i < pdata->num_chipselect; i++) {
+			int cs_gpio = of_get_named_gpio(pdev->dev.of_node,
+							"cs-gpios", i);
+
+			if (cs_gpio == -EPROBE_DEFER) {
+				ret = cs_gpio;
+				goto free_clk;
+			}
+
+			if (gpio_is_valid(cs_gpio)) {
+				ret = devm_gpio_request(&pdev->dev, cs_gpio,
+							dev_name(&pdev->dev));
+				if (ret)
+					goto free_clk;
+			}
+		}
+	}
+
 	r = platform_get_resource(pdev, IORESOURCE_DMA, 0);
 	if (r)
 		dma_rx_chan = r->start;
diff --git a/drivers/spi/spi-dw-mid.c b/drivers/spi/spi-dw-mid.c
index 1ca8f66..46c6d58 100644
--- a/drivers/spi/spi-dw-mid.c
+++ b/drivers/spi/spi-dw-mid.c
@@ -113,8 +113,7 @@ static void dw_spi_dma_done(void *arg)
 
 static int mid_spi_dma_transfer(struct dw_spi *dws, int cs_change)
 {
-	struct dma_async_tx_descriptor *txdesc = NULL, *rxdesc = NULL;
-	struct dma_chan *txchan, *rxchan;
+	struct dma_async_tx_descriptor *txdesc, *rxdesc;
 	struct dma_slave_config txconf, rxconf;
 	u16 dma_ctrl = 0;
 
@@ -124,16 +123,14 @@ static int mid_spi_dma_transfer(struct dw_spi *dws, int cs_change)
 		dw_writew(dws, DW_SPI_DMARDLR, 0xf);
 		dw_writew(dws, DW_SPI_DMATDLR, 0x10);
 		if (dws->tx_dma)
-			dma_ctrl |= 0x2;
+			dma_ctrl |= SPI_DMA_TDMAE;
 		if (dws->rx_dma)
-			dma_ctrl |= 0x1;
+			dma_ctrl |= SPI_DMA_RDMAE;
 		dw_writew(dws, DW_SPI_DMACR, dma_ctrl);
 		spi_enable_chip(dws, 1);
 	}
 
 	dws->dma_chan_done = 0;
-	txchan = dws->txchan;
-	rxchan = dws->rxchan;
 
 	/* 2. Prepare the TX dma transfer */
 	txconf.direction = DMA_MEM_TO_DEV;
@@ -143,17 +140,17 @@ static int mid_spi_dma_transfer(struct dw_spi *dws, int cs_change)
 	txconf.dst_addr_width = dws->dma_width;
 	txconf.device_fc = false;
 
-	dmaengine_slave_config(txchan, &txconf);
+	dmaengine_slave_config(dws->txchan, &txconf);
 
 	memset(&dws->tx_sgl, 0, sizeof(dws->tx_sgl));
 	dws->tx_sgl.dma_address = dws->tx_dma;
 	dws->tx_sgl.length = dws->len;
 
-	txdesc = dmaengine_prep_slave_sg(txchan,
+	txdesc = dmaengine_prep_slave_sg(dws->txchan,
 				&dws->tx_sgl,
 				1,
 				DMA_MEM_TO_DEV,
-				DMA_PREP_INTERRUPT);
+				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	txdesc->callback = dw_spi_dma_done;
 	txdesc->callback_param = dws;
 
@@ -165,23 +162,27 @@ static int mid_spi_dma_transfer(struct dw_spi *dws, int cs_change)
 	rxconf.src_addr_width = dws->dma_width;
 	rxconf.device_fc = false;
 
-	dmaengine_slave_config(txchan, &rxconf);
+	dmaengine_slave_config(dws->rxchan, &rxconf);
 
 	memset(&dws->rx_sgl, 0, sizeof(dws->rx_sgl));
 	dws->rx_sgl.dma_address = dws->rx_dma;
 	dws->rx_sgl.length = dws->len;
 
-	rxdesc = dmaengine_prep_slave_sg(rxchan,
+	rxdesc = dmaengine_prep_slave_sg(dws->rxchan,
 				&dws->rx_sgl,
 				1,
 				DMA_DEV_TO_MEM,
-				DMA_PREP_INTERRUPT);
+				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	rxdesc->callback = dw_spi_dma_done;
 	rxdesc->callback_param = dws;
 
 	/* rx must be started before tx due to spi instinct */
-	rxdesc->tx_submit(rxdesc);
-	txdesc->tx_submit(txdesc);
+	dmaengine_submit(rxdesc);
+	dma_async_issue_pending(dws->rxchan);
+
+	dmaengine_submit(txdesc);
+	dma_async_issue_pending(dws->txchan);
+
 	return 0;
 }
 
diff --git a/drivers/spi/spi-dw-pci.c b/drivers/spi/spi-dw-pci.c
index c90a450..ba68da1 100644
--- a/drivers/spi/spi-dw-pci.c
+++ b/drivers/spi/spi-dw-pci.c
@@ -63,6 +63,8 @@ static int spi_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (ret)
 		return ret;
 
+	dws->regs = pcim_iomap_table(pdev)[pci_bar];
+
 	dws->bus_num = 0;
 	dws->num_cs = 4;
 	dws->irq = pdev->irq;
diff --git a/drivers/spi/spi-dw.c b/drivers/spi/spi-dw.c
index 6bb4849..72e12ba 100644
--- a/drivers/spi/spi-dw.c
+++ b/drivers/spi/spi-dw.c
@@ -265,7 +265,7 @@ static void giveback(struct dw_spi *dws)
 					transfer_list);
 
 	if (!last_transfer->cs_change)
-		spi_chip_sel(dws, dws->cur_msg->spi, 0);
+		spi_chip_sel(dws, msg->spi, 0);
 
 	spi_finalize_current_message(dws->master);
 }
@@ -542,8 +542,7 @@ static int dw_spi_setup(struct spi_device *spi)
 	/* Only alloc on first setup */
 	chip = spi_get_ctldata(spi);
 	if (!chip) {
-		chip = devm_kzalloc(&spi->dev, sizeof(struct chip_data),
-				GFP_KERNEL);
+		chip = kzalloc(sizeof(struct chip_data), GFP_KERNEL);
 		if (!chip)
 			return -ENOMEM;
 		spi_set_ctldata(spi, chip);
@@ -604,6 +603,14 @@ static int dw_spi_setup(struct spi_device *spi)
 	return 0;
 }
 
+static void dw_spi_cleanup(struct spi_device *spi)
+{
+	struct chip_data *chip = spi_get_ctldata(spi);
+
+	kfree(chip);
+	spi_set_ctldata(spi, NULL);
+}
+
 /* Restart the controller, disable all interrupts, clean rx fifo */
 static void spi_hw_init(struct dw_spi *dws)
 {
@@ -659,8 +666,10 @@ int dw_spi_add_host(struct device *dev, struct dw_spi *dws)
 	master->bus_num = dws->bus_num;
 	master->num_chipselect = dws->num_cs;
 	master->setup = dw_spi_setup;
+	master->cleanup = dw_spi_cleanup;
 	master->transfer_one_message = dw_spi_transfer_one_message;
 	master->max_speed_hz = dws->max_freq;
+	master->dev.of_node = dev->of_node;
 
 	/* Basic HW init */
 	spi_hw_init(dws);
diff --git a/drivers/spi/spi-dw.h b/drivers/spi/spi-dw.h
index 089fc4b..83a103a 100644
--- a/drivers/spi/spi-dw.h
+++ b/drivers/spi/spi-dw.h
@@ -74,6 +74,10 @@
 #define SPI_INT_RXFI			(1 << 4)
 #define SPI_INT_MSTI			(1 << 5)
 
+/* Bit fields in DMACR */
+#define SPI_DMA_RDMAE			(1 << 0)
+#define SPI_DMA_TDMAE			(1 << 1)
+
 /* TX RX interrupt level threshold, max can be 256 */
 #define SPI_INT_THRESHOLD		32
 
diff --git a/drivers/spi/spi-fsl-cpm.c b/drivers/spi/spi-fsl-cpm.c
index 54b0637..c5dd20b 100644
--- a/drivers/spi/spi-fsl-cpm.c
+++ b/drivers/spi/spi-fsl-cpm.c
@@ -15,17 +15,17 @@
  * Free Software Foundation;  either version 2 of the  License, or (at your
  * option) any later version.
  */
-#include <linux/types.h>
-#include <linux/kernel.h>
-#include <linux/spi/spi.h>
-#include <linux/fsl_devices.h>
-#include <linux/dma-mapping.h>
-#include <linux/of_address.h>
 #include <asm/cpm.h>
 #include <asm/qe.h>
+#include <linux/dma-mapping.h>
+#include <linux/fsl_devices.h>
+#include <linux/kernel.h>
+#include <linux/of_address.h>
+#include <linux/spi/spi.h>
+#include <linux/types.h>
 
-#include "spi-fsl-lib.h"
 #include "spi-fsl-cpm.h"
+#include "spi-fsl-lib.h"
 #include "spi-fsl-spi.h"
 
 /* CPM1 and CPM2 are mutually exclusive. */
diff --git a/drivers/spi/spi-fsl-dspi.c b/drivers/spi/spi-fsl-dspi.c
index 5021ddf..4482160 100644
--- a/drivers/spi/spi-fsl-dspi.c
+++ b/drivers/spi/spi-fsl-dspi.c
@@ -13,22 +13,22 @@
  *
  */
 
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
-#include <linux/interrupt.h>
-#include <linux/errno.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
 #include <linux/regmap.h>
 #include <linux/sched.h>
-#include <linux/delay.h>
-#include <linux/io.h>
-#include <linux/clk.h>
-#include <linux/err.h>
 #include <linux/spi/spi.h>
 #include <linux/spi/spi_bitbang.h>
-#include <linux/pm_runtime.h>
-#include <linux/of.h>
-#include <linux/of_device.h>
 
 #define DRIVER_NAME "fsl-dspi"
 
@@ -493,9 +493,6 @@ static int dspi_probe(struct platform_device *pdev)
 	}
 
 	dspi_regmap_config.lock_arg = dspi;
-	dspi_regmap_config.val_format_endian =
-		of_property_read_bool(np, "big-endian")
-			? REGMAP_ENDIAN_BIG : REGMAP_ENDIAN_DEFAULT;
 	dspi->regmap = devm_regmap_init_mmio_clk(&pdev->dev, "dspi", base,
 						&dspi_regmap_config);
 	if (IS_ERR(dspi->regmap)) {
@@ -535,7 +532,6 @@ static int dspi_probe(struct platform_device *pdev)
 		goto out_clk_put;
 	}
 
-	pr_info(KERN_INFO "Freescale DSPI master initialized\n");
 	return ret;
 
 out_clk_put:
diff --git a/drivers/spi/spi-fsl-espi.c b/drivers/spi/spi-fsl-espi.c
index 8ebd724..a7f94b6 100644
--- a/drivers/spi/spi-fsl-espi.c
+++ b/drivers/spi/spi-fsl-espi.c
@@ -8,19 +8,19 @@
  * 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/err.h>
 #include <linux/fsl_devices.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/module.h>
 #include <linux/mm.h>
 #include <linux/of.h>
 #include <linux/of_address.h>
 #include <linux/of_irq.h>
 #include <linux/of_platform.h>
-#include <linux/interrupt.h>
-#include <linux/err.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
 #include <sysdev/fsl_soc.h>
 
 #include "spi-fsl-lib.h"
@@ -452,16 +452,16 @@ static int fsl_espi_setup(struct spi_device *spi)
 	int retval;
 	u32 hw_mode;
 	u32 loop_mode;
-	struct spi_mpc8xxx_cs *cs = spi->controller_state;
+	struct spi_mpc8xxx_cs *cs = spi_get_ctldata(spi);
 
 	if (!spi->max_speed_hz)
 		return -EINVAL;
 
 	if (!cs) {
-		cs = devm_kzalloc(&spi->dev, sizeof(*cs), GFP_KERNEL);
+		cs = kzalloc(sizeof(*cs), GFP_KERNEL);
 		if (!cs)
 			return -ENOMEM;
-		spi->controller_state = cs;
+		spi_set_ctldata(spi, cs);
 	}
 
 	mpc8xxx_spi = spi_master_get_devdata(spi->master);
@@ -496,6 +496,14 @@ static int fsl_espi_setup(struct spi_device *spi)
 	return 0;
 }
 
+static void fsl_espi_cleanup(struct spi_device *spi)
+{
+	struct spi_mpc8xxx_cs *cs = spi_get_ctldata(spi);
+
+	kfree(cs);
+	spi_set_ctldata(spi, NULL);
+}
+
 void fsl_espi_cpu_irq(struct mpc8xxx_spi *mspi, u32 events)
 {
 	struct fsl_espi_reg *reg_base = mspi->reg_base;
@@ -605,6 +613,7 @@ static struct spi_master * fsl_espi_probe(struct device *dev,
 
 	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 16);
 	master->setup = fsl_espi_setup;
+	master->cleanup = fsl_espi_cleanup;
 
 	mpc8xxx_spi = spi_master_get_devdata(master);
 	mpc8xxx_spi->spi_do_one_msg = fsl_espi_do_one_msg;
diff --git a/drivers/spi/spi-fsl-lib.c b/drivers/spi/spi-fsl-lib.c
index e0b773f..5ddb5b0 100644
--- a/drivers/spi/spi-fsl-lib.c
+++ b/drivers/spi/spi-fsl-lib.c
@@ -16,10 +16,10 @@
  * 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/fsl_devices.h>
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
 #include <linux/mm.h>
 #include <linux/of_platform.h>
 #include <linux/spi/spi.h>
diff --git a/drivers/spi/spi-fsl-spi.c b/drivers/spi/spi-fsl-spi.c
index 9452f674..ed79288 100644
--- a/drivers/spi/spi-fsl-spi.c
+++ b/drivers/spi/spi-fsl-spi.c
@@ -19,25 +19,25 @@
  * Free Software Foundation;  either version 2 of the  License, or (at your
  * option) any later version.
  */
-#include <linux/module.h>
-#include <linux/types.h>
-#include <linux/kernel.h>
-#include <linux/interrupt.h>
 #include <linux/delay.h>
-#include <linux/irq.h>
-#include <linux/spi/spi.h>
-#include <linux/spi/spi_bitbang.h>
-#include <linux/platform_device.h>
-#include <linux/fsl_devices.h>
 #include <linux/dma-mapping.h>
+#include <linux/fsl_devices.h>
+#include <linux/gpio.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/kernel.h>
 #include <linux/mm.h>
+#include <linux/module.h>
 #include <linux/mutex.h>
 #include <linux/of.h>
-#include <linux/of_platform.h>
 #include <linux/of_address.h>
 #include <linux/of_irq.h>
-#include <linux/gpio.h>
 #include <linux/of_gpio.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/spi_bitbang.h>
+#include <linux/types.h>
 
 #include "spi-fsl-lib.h"
 #include "spi-fsl-cpm.h"
@@ -425,16 +425,16 @@ static int fsl_spi_setup(struct spi_device *spi)
 	struct fsl_spi_reg *reg_base;
 	int retval;
 	u32 hw_mode;
-	struct spi_mpc8xxx_cs	*cs = spi->controller_state;
+	struct spi_mpc8xxx_cs *cs = spi_get_ctldata(spi);
 
 	if (!spi->max_speed_hz)
 		return -EINVAL;
 
 	if (!cs) {
-		cs = devm_kzalloc(&spi->dev, sizeof(*cs), GFP_KERNEL);
+		cs = kzalloc(sizeof(*cs), GFP_KERNEL);
 		if (!cs)
 			return -ENOMEM;
-		spi->controller_state = cs;
+		spi_set_ctldata(spi, cs);
 	}
 	mpc8xxx_spi = spi_master_get_devdata(spi->master);
 
@@ -496,9 +496,13 @@ static int fsl_spi_setup(struct spi_device *spi)
 static void fsl_spi_cleanup(struct spi_device *spi)
 {
 	struct mpc8xxx_spi *mpc8xxx_spi = spi_master_get_devdata(spi->master);
+	struct spi_mpc8xxx_cs *cs = spi_get_ctldata(spi);
 
 	if (mpc8xxx_spi->type == TYPE_GRLIB && gpio_is_valid(spi->cs_gpio))
 		gpio_free(spi->cs_gpio);
+
+	kfree(cs);
+	spi_set_ctldata(spi, NULL);
 }
 
 static void fsl_spi_cpu_irq(struct mpc8xxx_spi *mspi, u32 events)
diff --git a/drivers/spi/spi-imx.c b/drivers/spi/spi-imx.c
index 5daff20..3637847 100644
--- a/drivers/spi/spi-imx.c
+++ b/drivers/spi/spi-imx.c
@@ -21,6 +21,8 @@
 #include <linux/clk.h>
 #include <linux/completion.h>
 #include <linux/delay.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
 #include <linux/err.h>
 #include <linux/gpio.h>
 #include <linux/interrupt.h>
@@ -37,6 +39,7 @@
 #include <linux/of_device.h>
 #include <linux/of_gpio.h>
 
+#include <linux/platform_data/dma-imx.h>
 #include <linux/platform_data/spi-imx.h>
 
 #define DRIVER_NAME "spi_imx"
@@ -51,6 +54,9 @@
 #define MXC_INT_RR	(1 << 0) /* Receive data ready interrupt */
 #define MXC_INT_TE	(1 << 1) /* Transmit FIFO empty interrupt */
 
+/* The maximum  bytes that a sdma BD can transfer.*/
+#define MAX_SDMA_BD_BYTES  (1 << 15)
+#define IMX_DMA_TIMEOUT (msecs_to_jiffies(3000))
 struct spi_imx_config {
 	unsigned int speed_hz;
 	unsigned int bpw;
@@ -95,6 +101,16 @@ struct spi_imx_data {
 	const void *tx_buf;
 	unsigned int txfifo; /* number of words pushed in tx FIFO */
 
+	/* DMA */
+	unsigned int dma_is_inited;
+	unsigned int dma_finished;
+	bool usedma;
+	u32 rx_wml;
+	u32 tx_wml;
+	u32 rxt_wml;
+	struct completion dma_rx_completion;
+	struct completion dma_tx_completion;
+
 	const struct spi_imx_devtype_data *devtype_data;
 	int chipselect[0];
 };
@@ -181,9 +197,21 @@ static unsigned int spi_imx_clkdiv_2(unsigned int fin,
 	return 7;
 }
 
+static bool spi_imx_can_dma(struct spi_master *master, struct spi_device *spi,
+			 struct spi_transfer *transfer)
+{
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(master);
+
+	if (spi_imx->dma_is_inited && (transfer->len > spi_imx->rx_wml)
+	    && (transfer->len > spi_imx->tx_wml))
+		return true;
+	return false;
+}
+
 #define MX51_ECSPI_CTRL		0x08
 #define MX51_ECSPI_CTRL_ENABLE		(1 <<  0)
 #define MX51_ECSPI_CTRL_XCH		(1 <<  2)
+#define MX51_ECSPI_CTRL_SMC		(1 << 3)
 #define MX51_ECSPI_CTRL_MODE_MASK	(0xf << 4)
 #define MX51_ECSPI_CTRL_POSTDIV_OFFSET	8
 #define MX51_ECSPI_CTRL_PREDIV_OFFSET	12
@@ -201,6 +229,18 @@ static unsigned int spi_imx_clkdiv_2(unsigned int fin,
 #define MX51_ECSPI_INT_TEEN		(1 <<  0)
 #define MX51_ECSPI_INT_RREN		(1 <<  3)
 
+#define MX51_ECSPI_DMA      0x14
+#define MX51_ECSPI_DMA_TX_WML_OFFSET	0
+#define MX51_ECSPI_DMA_TX_WML_MASK	0x3F
+#define MX51_ECSPI_DMA_RX_WML_OFFSET	16
+#define MX51_ECSPI_DMA_RX_WML_MASK	(0x3F << 16)
+#define MX51_ECSPI_DMA_RXT_WML_OFFSET	24
+#define MX51_ECSPI_DMA_RXT_WML_MASK	(0x3F << 24)
+
+#define MX51_ECSPI_DMA_TEDEN_OFFSET	7
+#define MX51_ECSPI_DMA_RXDEN_OFFSET	23
+#define MX51_ECSPI_DMA_RXTDEN_OFFSET	31
+
 #define MX51_ECSPI_STAT		0x18
 #define MX51_ECSPI_STAT_RR		(1 <<  3)
 
@@ -257,17 +297,22 @@ static void __maybe_unused mx51_ecspi_intctrl(struct spi_imx_data *spi_imx, int
 
 static void __maybe_unused mx51_ecspi_trigger(struct spi_imx_data *spi_imx)
 {
-	u32 reg;
-
-	reg = readl(spi_imx->base + MX51_ECSPI_CTRL);
-	reg |= MX51_ECSPI_CTRL_XCH;
+	u32 reg = readl(spi_imx->base + MX51_ECSPI_CTRL);
+
+	if (!spi_imx->usedma)
+		reg |= MX51_ECSPI_CTRL_XCH;
+	else if (!spi_imx->dma_finished)
+		reg |= MX51_ECSPI_CTRL_SMC;
+	else
+		reg &= ~MX51_ECSPI_CTRL_SMC;
 	writel(reg, spi_imx->base + MX51_ECSPI_CTRL);
 }
 
 static int __maybe_unused mx51_ecspi_config(struct spi_imx_data *spi_imx,
 		struct spi_imx_config *config)
 {
-	u32 ctrl = MX51_ECSPI_CTRL_ENABLE, cfg = 0;
+	u32 ctrl = MX51_ECSPI_CTRL_ENABLE, cfg = 0, dma = 0;
+	u32 tx_wml_cfg, rx_wml_cfg, rxt_wml_cfg;
 	u32 clk = config->speed_hz, delay;
 
 	/*
@@ -319,6 +364,30 @@ static int __maybe_unused mx51_ecspi_config(struct spi_imx_data *spi_imx,
 	else			/* SCLK is _very_ slow */
 		usleep_range(delay, delay + 10);
 
+	/*
+	 * Configure the DMA register: setup the watermark
+	 * and enable DMA request.
+	 */
+	if (spi_imx->dma_is_inited) {
+		dma = readl(spi_imx->base + MX51_ECSPI_DMA);
+
+		spi_imx->tx_wml = spi_imx_get_fifosize(spi_imx) / 2;
+		spi_imx->rx_wml = spi_imx_get_fifosize(spi_imx) / 2;
+		spi_imx->rxt_wml = spi_imx_get_fifosize(spi_imx) / 2;
+		rx_wml_cfg = spi_imx->rx_wml << MX51_ECSPI_DMA_RX_WML_OFFSET;
+		tx_wml_cfg = spi_imx->tx_wml << MX51_ECSPI_DMA_TX_WML_OFFSET;
+		rxt_wml_cfg = spi_imx->rxt_wml << MX51_ECSPI_DMA_RXT_WML_OFFSET;
+		dma = (dma & ~MX51_ECSPI_DMA_TX_WML_MASK
+			   & ~MX51_ECSPI_DMA_RX_WML_MASK
+			   & ~MX51_ECSPI_DMA_RXT_WML_MASK)
+			   | rx_wml_cfg | tx_wml_cfg | rxt_wml_cfg
+			   |(1 << MX51_ECSPI_DMA_TEDEN_OFFSET)
+			   |(1 << MX51_ECSPI_DMA_RXDEN_OFFSET)
+			   |(1 << MX51_ECSPI_DMA_RXTDEN_OFFSET);
+
+		writel(dma, spi_imx->base + MX51_ECSPI_DMA);
+	}
+
 	return 0;
 }
 
@@ -730,7 +799,186 @@ static int spi_imx_setupxfer(struct spi_device *spi,
 	return 0;
 }
 
-static int spi_imx_transfer(struct spi_device *spi,
+static void spi_imx_sdma_exit(struct spi_imx_data *spi_imx)
+{
+	struct spi_master *master = spi_imx->bitbang.master;
+
+	if (master->dma_rx) {
+		dma_release_channel(master->dma_rx);
+		master->dma_rx = NULL;
+	}
+
+	if (master->dma_tx) {
+		dma_release_channel(master->dma_tx);
+		master->dma_tx = NULL;
+	}
+
+	spi_imx->dma_is_inited = 0;
+}
+
+static int spi_imx_sdma_init(struct device *dev, struct spi_imx_data *spi_imx,
+			     struct spi_master *master,
+			     const struct resource *res)
+{
+	struct dma_slave_config slave_config = {};
+	int ret;
+
+	/* Prepare for TX DMA: */
+	master->dma_tx = dma_request_slave_channel(dev, "tx");
+	if (!master->dma_tx) {
+		dev_err(dev, "cannot get the TX DMA channel!\n");
+		ret = -EINVAL;
+		goto err;
+	}
+
+	slave_config.direction = DMA_MEM_TO_DEV;
+	slave_config.dst_addr = res->start + MXC_CSPITXDATA;
+	slave_config.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	slave_config.dst_maxburst = spi_imx_get_fifosize(spi_imx) / 2;
+	ret = dmaengine_slave_config(master->dma_tx, &slave_config);
+	if (ret) {
+		dev_err(dev, "error in TX dma configuration.\n");
+		goto err;
+	}
+
+	/* Prepare for RX : */
+	master->dma_rx = dma_request_slave_channel(dev, "rx");
+	if (!master->dma_rx) {
+		dev_dbg(dev, "cannot get the DMA channel.\n");
+		ret = -EINVAL;
+		goto err;
+	}
+
+	slave_config.direction = DMA_DEV_TO_MEM;
+	slave_config.src_addr = res->start + MXC_CSPIRXDATA;
+	slave_config.src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	slave_config.src_maxburst = spi_imx_get_fifosize(spi_imx) / 2;
+	ret = dmaengine_slave_config(master->dma_rx, &slave_config);
+	if (ret) {
+		dev_err(dev, "error in RX dma configuration.\n");
+		goto err;
+	}
+
+	init_completion(&spi_imx->dma_rx_completion);
+	init_completion(&spi_imx->dma_tx_completion);
+	master->can_dma = spi_imx_can_dma;
+	master->max_dma_len = MAX_SDMA_BD_BYTES;
+	spi_imx->bitbang.master->flags = SPI_MASTER_MUST_RX |
+					 SPI_MASTER_MUST_TX;
+	spi_imx->dma_is_inited = 1;
+
+	return 0;
+err:
+	spi_imx_sdma_exit(spi_imx);
+	return ret;
+}
+
+static void spi_imx_dma_rx_callback(void *cookie)
+{
+	struct spi_imx_data *spi_imx = (struct spi_imx_data *)cookie;
+
+	complete(&spi_imx->dma_rx_completion);
+}
+
+static void spi_imx_dma_tx_callback(void *cookie)
+{
+	struct spi_imx_data *spi_imx = (struct spi_imx_data *)cookie;
+
+	complete(&spi_imx->dma_tx_completion);
+}
+
+static int spi_imx_dma_transfer(struct spi_imx_data *spi_imx,
+				struct spi_transfer *transfer)
+{
+	struct dma_async_tx_descriptor *desc_tx = NULL, *desc_rx = NULL;
+	int ret;
+	u32 dma;
+	int left;
+	struct spi_master *master = spi_imx->bitbang.master;
+	struct sg_table *tx = &transfer->tx_sg, *rx = &transfer->rx_sg;
+
+	if (tx) {
+		desc_tx = dmaengine_prep_slave_sg(master->dma_tx,
+					tx->sgl, tx->nents, DMA_TO_DEVICE,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!desc_tx)
+			goto no_dma;
+
+		desc_tx->callback = spi_imx_dma_tx_callback;
+		desc_tx->callback_param = (void *)spi_imx;
+		dmaengine_submit(desc_tx);
+	}
+
+	if (rx) {
+		desc_rx = dmaengine_prep_slave_sg(master->dma_rx,
+					rx->sgl, rx->nents, DMA_FROM_DEVICE,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!desc_rx)
+			goto no_dma;
+
+		desc_rx->callback = spi_imx_dma_rx_callback;
+		desc_rx->callback_param = (void *)spi_imx;
+		dmaengine_submit(desc_rx);
+	}
+
+	reinit_completion(&spi_imx->dma_rx_completion);
+	reinit_completion(&spi_imx->dma_tx_completion);
+
+	/* Trigger the cspi module. */
+	spi_imx->dma_finished = 0;
+
+	dma = readl(spi_imx->base + MX51_ECSPI_DMA);
+	dma = dma & (~MX51_ECSPI_DMA_RXT_WML_MASK);
+	/* Change RX_DMA_LENGTH trigger dma fetch tail data */
+	left = transfer->len % spi_imx->rxt_wml;
+	if (left)
+		writel(dma | (left << MX51_ECSPI_DMA_RXT_WML_OFFSET),
+				spi_imx->base + MX51_ECSPI_DMA);
+	spi_imx->devtype_data->trigger(spi_imx);
+
+	dma_async_issue_pending(master->dma_tx);
+	dma_async_issue_pending(master->dma_rx);
+	/* Wait SDMA to finish the data transfer.*/
+	ret = wait_for_completion_timeout(&spi_imx->dma_tx_completion,
+						IMX_DMA_TIMEOUT);
+	if (!ret) {
+		pr_warn("%s %s: I/O Error in DMA TX\n",
+			dev_driver_string(&master->dev),
+			dev_name(&master->dev));
+		dmaengine_terminate_all(master->dma_tx);
+	} else {
+		ret = wait_for_completion_timeout(&spi_imx->dma_rx_completion,
+				IMX_DMA_TIMEOUT);
+		if (!ret) {
+			pr_warn("%s %s: I/O Error in DMA RX\n",
+				dev_driver_string(&master->dev),
+				dev_name(&master->dev));
+			spi_imx->devtype_data->reset(spi_imx);
+			dmaengine_terminate_all(master->dma_rx);
+		}
+		writel(dma |
+		       spi_imx->rxt_wml << MX51_ECSPI_DMA_RXT_WML_OFFSET,
+		       spi_imx->base + MX51_ECSPI_DMA);
+	}
+
+	spi_imx->dma_finished = 1;
+	spi_imx->devtype_data->trigger(spi_imx);
+
+	if (!ret)
+		ret = -ETIMEDOUT;
+	else if (ret > 0)
+		ret = transfer->len;
+
+	return ret;
+
+no_dma:
+	pr_warn_once("%s %s: DMA not available, falling back to PIO\n",
+		     dev_driver_string(&master->dev),
+		     dev_name(&master->dev));
+	return -EAGAIN;
+}
+
+static int spi_imx_pio_transfer(struct spi_device *spi,
 				struct spi_transfer *transfer)
 {
 	struct spi_imx_data *spi_imx = spi_master_get_devdata(spi->master);
@@ -751,6 +999,24 @@ static int spi_imx_transfer(struct spi_device *spi,
 	return transfer->len;
 }
 
+static int spi_imx_transfer(struct spi_device *spi,
+				struct spi_transfer *transfer)
+{
+	int ret;
+	struct spi_imx_data *spi_imx = spi_master_get_devdata(spi->master);
+
+	if (spi_imx->bitbang.master->can_dma &&
+	    spi_imx_can_dma(spi_imx->bitbang.master, spi, transfer)) {
+		spi_imx->usedma = true;
+		ret = spi_imx_dma_transfer(spi_imx, transfer);
+		if (ret != -EAGAIN)
+			return ret;
+	}
+	spi_imx->usedma = false;
+
+	return spi_imx_pio_transfer(spi, transfer);
+}
+
 static int spi_imx_setup(struct spi_device *spi)
 {
 	struct spi_imx_data *spi_imx = spi_master_get_devdata(spi->master);
@@ -911,6 +1177,13 @@ static int spi_imx_probe(struct platform_device *pdev)
 		goto out_put_per;
 
 	spi_imx->spi_clk = clk_get_rate(spi_imx->clk_per);
+	/*
+	 * Only validated on i.mx6 now, can remove the constrain if validated on
+	 * other chips.
+	 */
+	if (spi_imx->devtype_data == &imx51_ecspi_devtype_data
+	    && spi_imx_sdma_init(&pdev->dev, spi_imx, master, res))
+		dev_err(&pdev->dev, "dma setup error,use pio instead\n");
 
 	spi_imx->devtype_data->reset(spi_imx);
 
@@ -949,6 +1222,7 @@ static int spi_imx_remove(struct platform_device *pdev)
 	writel(0, spi_imx->base + MXC_CSPICTRL);
 	clk_unprepare(spi_imx->clk_ipg);
 	clk_unprepare(spi_imx->clk_per);
+	spi_imx_sdma_exit(spi_imx);
 	spi_master_put(master);
 
 	return 0;
diff --git a/drivers/spi/spi-mxs.c b/drivers/spi/spi-mxs.c
index c3f8d3a..5146087 100644
--- a/drivers/spi/spi-mxs.c
+++ b/drivers/spi/spi-mxs.c
@@ -85,7 +85,7 @@ static int mxs_spi_setup_transfer(struct spi_device *dev,
 		mxs_ssp_set_clk_rate(ssp, hz);
 		/*
 		 * Save requested rate, hz, rather than the actual rate,
-		 * ssp->clk_rate.  Otherwise we would set the rate every trasfer
+		 * ssp->clk_rate.  Otherwise we would set the rate every transfer
 		 * when the actual rate is not quite the same as requested rate.
 		 */
 		spi->sck = hz;
diff --git a/drivers/spi/spi-omap-100k.c b/drivers/spi/spi-omap-100k.c
index 5e91858..fb52276 100644
--- a/drivers/spi/spi-omap-100k.c
+++ b/drivers/spi/spi-omap-100k.c
@@ -70,10 +70,6 @@
 #define SPI_STATUS_WE                   (1UL << 1)
 #define SPI_STATUS_RD                   (1UL << 0)
 
-#define WRITE 0
-#define READ  1
-
-
 /* use PIO for small transfers, avoiding DMA setup/teardown overhead and
  * cache operations; better heuristics consider wordsize and bitrate.
  */
diff --git a/drivers/spi/spi-omap2-mcspi.c b/drivers/spi/spi-omap2-mcspi.c
index 68441fa..352eed7 100644
--- a/drivers/spi/spi-omap2-mcspi.c
+++ b/drivers/spi/spi-omap2-mcspi.c
@@ -329,7 +329,8 @@ static void omap2_mcspi_set_fifo(const struct spi_device *spi,
 disable_fifo:
 	if (t->rx_buf != NULL)
 		chconf &= ~OMAP2_MCSPI_CHCONF_FFER;
-	else
+
+	if (t->tx_buf != NULL)
 		chconf &= ~OMAP2_MCSPI_CHCONF_FFET;
 
 	mcspi_write_chconf0(spi, chconf);
diff --git a/drivers/spi/spi-orion.c b/drivers/spi/spi-orion.c
index 345e7d6..835cdda 100644
--- a/drivers/spi/spi-orion.c
+++ b/drivers/spi/spi-orion.c
@@ -18,6 +18,7 @@
 #include <linux/module.h>
 #include <linux/pm_runtime.h>
 #include <linux/of.h>
+#include <linux/of_device.h>
 #include <linux/clk.h>
 #include <linux/sizes.h>
 #include <asm/unaligned.h>
@@ -40,13 +41,27 @@
 #define ORION_SPI_MODE_CPHA		(1 << 12)
 #define ORION_SPI_IF_8_16_BIT_MODE	(1 << 5)
 #define ORION_SPI_CLK_PRESCALE_MASK	0x1F
+#define ARMADA_SPI_CLK_PRESCALE_MASK	0xDF
 #define ORION_SPI_MODE_MASK		(ORION_SPI_MODE_CPOL | \
 					 ORION_SPI_MODE_CPHA)
 
+enum orion_spi_type {
+	ORION_SPI,
+	ARMADA_SPI,
+};
+
+struct orion_spi_dev {
+	enum orion_spi_type	typ;
+	unsigned int		min_divisor;
+	unsigned int		max_divisor;
+	u32			prescale_mask;
+};
+
 struct orion_spi {
 	struct spi_master	*master;
 	void __iomem		*base;
 	struct clk              *clk;
+	const struct orion_spi_dev *devdata;
 };
 
 static inline void __iomem *spi_reg(struct orion_spi *orion_spi, u32 reg)
@@ -83,30 +98,66 @@ static int orion_spi_baudrate_set(struct spi_device *spi, unsigned int speed)
 	u32 prescale;
 	u32 reg;
 	struct orion_spi *orion_spi;
+	const struct orion_spi_dev *devdata;
 
 	orion_spi = spi_master_get_devdata(spi->master);
+	devdata = orion_spi->devdata;
 
 	tclk_hz = clk_get_rate(orion_spi->clk);
 
-	/*
-	 * the supported rates are: 4,6,8...30
-	 * round up as we look for equal or less speed
-	 */
-	rate = DIV_ROUND_UP(tclk_hz, speed);
-	rate = roundup(rate, 2);
+	if (devdata->typ == ARMADA_SPI) {
+		unsigned int clk, spr, sppr, sppr2, err;
+		unsigned int best_spr, best_sppr, best_err;
+
+		best_err = speed;
+		best_spr = 0;
+		best_sppr = 0;
+
+		/* Iterate over the valid range looking for best fit */
+		for (sppr = 0; sppr < 8; sppr++) {
+			sppr2 = 0x1 << sppr;
+
+			spr = tclk_hz / sppr2;
+			spr = DIV_ROUND_UP(spr, speed);
+			if ((spr == 0) || (spr > 15))
+				continue;
 
-	/* check if requested speed is too small */
-	if (rate > 30)
-		return -EINVAL;
+			clk = tclk_hz / (spr * sppr2);
+			err = speed - clk;
 
-	if (rate < 4)
-		rate = 4;
+			if (err < best_err) {
+				best_spr = spr;
+				best_sppr = sppr;
+				best_err = err;
+			}
+		}
+
+		if ((best_sppr == 0) && (best_spr == 0))
+			return -EINVAL;
+
+		prescale = ((best_sppr & 0x6) << 5) |
+			((best_sppr & 0x1) << 4) | best_spr;
+	} else {
+		/*
+		 * the supported rates are: 4,6,8...30
+		 * round up as we look for equal or less speed
+		 */
+		rate = DIV_ROUND_UP(tclk_hz, speed);
+		rate = roundup(rate, 2);
+
+		/* check if requested speed is too small */
+		if (rate > 30)
+			return -EINVAL;
 
-	/* Convert the rate to SPI clock divisor value.	*/
-	prescale = 0x10 + rate/2;
+		if (rate < 4)
+			rate = 4;
+
+		/* Convert the rate to SPI clock divisor value.	*/
+		prescale = 0x10 + rate/2;
+	}
 
 	reg = readl(spi_reg(orion_spi, ORION_SPI_IF_CONFIG_REG));
-	reg = ((reg & ~ORION_SPI_CLK_PRESCALE_MASK) | prescale);
+	reg = ((reg & ~devdata->prescale_mask) | prescale);
 	writel(reg, spi_reg(orion_spi, ORION_SPI_IF_CONFIG_REG));
 
 	return 0;
@@ -342,8 +393,31 @@ static int orion_spi_reset(struct orion_spi *orion_spi)
 	return 0;
 }
 
+static const struct orion_spi_dev orion_spi_dev_data = {
+	.typ = ORION_SPI,
+	.min_divisor = 4,
+	.max_divisor = 30,
+	.prescale_mask = ORION_SPI_CLK_PRESCALE_MASK,
+};
+
+static const struct orion_spi_dev armada_spi_dev_data = {
+	.typ = ARMADA_SPI,
+	.min_divisor = 1,
+	.max_divisor = 1920,
+	.prescale_mask = ARMADA_SPI_CLK_PRESCALE_MASK,
+};
+
+static const struct of_device_id orion_spi_of_match_table[] = {
+	{ .compatible = "marvell,orion-spi", .data = &orion_spi_dev_data, },
+	{ .compatible = "marvell,armada-370-spi", .data = &armada_spi_dev_data, },
+	{}
+};
+MODULE_DEVICE_TABLE(of, orion_spi_of_match_table);
+
 static int orion_spi_probe(struct platform_device *pdev)
 {
+	const struct of_device_id *of_id;
+	const struct orion_spi_dev *devdata;
 	struct spi_master *master;
 	struct orion_spi *spi;
 	struct resource *r;
@@ -379,6 +453,10 @@ static int orion_spi_probe(struct platform_device *pdev)
 	spi = spi_master_get_devdata(master);
 	spi->master = master;
 
+	of_id = of_match_device(orion_spi_of_match_table, &pdev->dev);
+	devdata = of_id->data;
+	spi->devdata = devdata;
+
 	spi->clk = devm_clk_get(&pdev->dev, NULL);
 	if (IS_ERR(spi->clk)) {
 		status = PTR_ERR(spi->clk);
@@ -390,8 +468,8 @@ static int orion_spi_probe(struct platform_device *pdev)
 		goto out;
 
 	tclk_hz = clk_get_rate(spi->clk);
-	master->max_speed_hz = DIV_ROUND_UP(tclk_hz, 4);
-	master->min_speed_hz = DIV_ROUND_UP(tclk_hz, 30);
+	master->max_speed_hz = DIV_ROUND_UP(tclk_hz, devdata->min_divisor);
+	master->min_speed_hz = DIV_ROUND_UP(tclk_hz, devdata->max_divisor);
 
 	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	spi->base = devm_ioremap_resource(&pdev->dev, r);
@@ -470,12 +548,6 @@ static const struct dev_pm_ops orion_spi_pm_ops = {
 			   NULL)
 };
 
-static const struct of_device_id orion_spi_of_match_table[] = {
-	{ .compatible = "marvell,orion-spi", },
-	{}
-};
-MODULE_DEVICE_TABLE(of, orion_spi_of_match_table);
-
 static struct platform_driver orion_spi_driver = {
 	.driver = {
 		.name	= DRIVER_NAME,
diff --git a/drivers/spi/spi-pl022.c b/drivers/spi/spi-pl022.c
index 1189cfd..f35f723 100644
--- a/drivers/spi/spi-pl022.c
+++ b/drivers/spi/spi-pl022.c
@@ -82,6 +82,7 @@
 #define SSP_MIS(r)	(r + 0x01C)
 #define SSP_ICR(r)	(r + 0x020)
 #define SSP_DMACR(r)	(r + 0x024)
+#define SSP_CSR(r)	(r + 0x030) /* vendor extension */
 #define SSP_ITCR(r)	(r + 0x080)
 #define SSP_ITIP(r)	(r + 0x084)
 #define SSP_ITOP(r)	(r + 0x088)
@@ -198,6 +199,12 @@
 #define SSP_DMACR_MASK_TXDMAE		(0x1UL << 1)
 
 /*
+ * SSP Chip Select Control Register - SSP_CSR
+ * (vendor extension)
+ */
+#define SSP_CSR_CSVALUE_MASK		(0x1FUL << 0)
+
+/*
  * SSP Integration Test control Register - SSP_ITCR
  */
 #define SSP_ITCR_MASK_ITEN		(0x1UL << 0)
@@ -313,6 +320,7 @@ enum ssp_writing {
  * @extended_cr: 32 bit wide control register 0 with extra
  * features and extra features in CR1 as found in the ST variants
  * @pl023: supports a subset of the ST extensions called "PL023"
+ * @internal_cs_ctrl: supports chip select control register
  */
 struct vendor_data {
 	int fifodepth;
@@ -321,6 +329,7 @@ struct vendor_data {
 	bool extended_cr;
 	bool pl023;
 	bool loopback;
+	bool internal_cs_ctrl;
 };
 
 /**
@@ -440,9 +449,32 @@ static void null_cs_control(u32 command)
 	pr_debug("pl022: dummy chip select control, CS=0x%x\n", command);
 }
 
+/**
+ * internal_cs_control - Control chip select signals via SSP_CSR.
+ * @pl022: SSP driver private data structure
+ * @command: select/delect the chip
+ *
+ * Used on controller with internal chip select control via SSP_CSR register
+ * (vendor extension). Each of the 5 LSB in the register controls one chip
+ * select signal.
+ */
+static void internal_cs_control(struct pl022 *pl022, u32 command)
+{
+	u32 tmp;
+
+	tmp = readw(SSP_CSR(pl022->virtbase));
+	if (command == SSP_CHIP_SELECT)
+		tmp &= ~BIT(pl022->cur_cs);
+	else
+		tmp |= BIT(pl022->cur_cs);
+	writew(tmp, SSP_CSR(pl022->virtbase));
+}
+
 static void pl022_cs_control(struct pl022 *pl022, u32 command)
 {
-	if (gpio_is_valid(pl022->cur_cs))
+	if (pl022->vendor->internal_cs_ctrl)
+		internal_cs_control(pl022, command);
+	else if (gpio_is_valid(pl022->cur_cs))
 		gpio_set_value(pl022->cur_cs, command);
 	else
 		pl022->cur_chip->cs_control(command);
@@ -2100,6 +2132,10 @@ static int pl022_probe(struct amba_device *adev, const struct amba_id *id)
 	pl022->vendor = id->data;
 	pl022->chipselects = devm_kzalloc(dev, num_cs * sizeof(int),
 					  GFP_KERNEL);
+	if (!pl022->chipselects) {
+		status = -ENOMEM;
+		goto err_no_mem;
+	}
 
 	/*
 	 * Bus Number Which has been Assigned to this SSP controller
@@ -2118,6 +2154,9 @@ static int pl022_probe(struct amba_device *adev, const struct amba_id *id)
 	if (platform_info->num_chipselect && platform_info->chipselects) {
 		for (i = 0; i < num_cs; i++)
 			pl022->chipselects[i] = platform_info->chipselects[i];
+	} else if (pl022->vendor->internal_cs_ctrl) {
+		for (i = 0; i < num_cs; i++)
+			pl022->chipselects[i] = i;
 	} else if (IS_ENABLED(CONFIG_OF)) {
 		for (i = 0; i < num_cs; i++) {
 			int cs_gpio = of_get_named_gpio(np, "cs-gpios", i);
@@ -2136,7 +2175,7 @@ static int pl022_probe(struct amba_device *adev, const struct amba_id *id)
 						cs_gpio);
 				else if (gpio_direction_output(cs_gpio, 1))
 					dev_err(&adev->dev,
-						"could set gpio %d as output\n",
+						"could not set gpio %d as output\n",
 						cs_gpio);
 			}
 		}
@@ -2241,6 +2280,7 @@ static int pl022_probe(struct amba_device *adev, const struct amba_id *id)
 	amba_release_regions(adev);
  err_no_ioregion:
  err_no_gpio:
+ err_no_mem:
 	spi_master_put(master);
 	return status;
 }
@@ -2347,6 +2387,7 @@ static struct vendor_data vendor_arm = {
 	.extended_cr = false,
 	.pl023 = false,
 	.loopback = true,
+	.internal_cs_ctrl = false,
 };
 
 static struct vendor_data vendor_st = {
@@ -2356,6 +2397,7 @@ static struct vendor_data vendor_st = {
 	.extended_cr = true,
 	.pl023 = false,
 	.loopback = true,
+	.internal_cs_ctrl = false,
 };
 
 static struct vendor_data vendor_st_pl023 = {
@@ -2365,6 +2407,17 @@ static struct vendor_data vendor_st_pl023 = {
 	.extended_cr = true,
 	.pl023 = true,
 	.loopback = false,
+	.internal_cs_ctrl = false,
+};
+
+static struct vendor_data vendor_lsi = {
+	.fifodepth = 8,
+	.max_bpw = 16,
+	.unidir = false,
+	.extended_cr = false,
+	.pl023 = false,
+	.loopback = true,
+	.internal_cs_ctrl = true,
 };
 
 static struct amba_id pl022_ids[] = {
@@ -2398,6 +2451,15 @@ static struct amba_id pl022_ids[] = {
 		.mask	= 0xffffffff,
 		.data	= &vendor_st_pl023,
 	},
+	{
+		/*
+		 * PL022 variant that has a chip select control register whih
+		 * allows control of 5 output signals nCS[0:4].
+		 */
+		.id	= 0x000b6022,
+		.mask	= 0x000fffff,
+		.data	= &vendor_lsi,
+	},
 	{ 0, 0 },
 };
 
diff --git a/drivers/spi/spi-pxa2xx-pci.c b/drivers/spi/spi-pxa2xx-pci.c
index c1865c9..536c863 100644
--- a/drivers/spi/spi-pxa2xx-pci.c
+++ b/drivers/spi/spi-pxa2xx-pci.c
@@ -7,6 +7,8 @@
 #include <linux/of_device.h>
 #include <linux/module.h>
 #include <linux/spi/pxa2xx_spi.h>
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
 
 enum {
 	PORT_CE4100,
@@ -21,6 +23,7 @@ struct pxa_spi_info {
 	int tx_chan_id;
 	int rx_slave_id;
 	int rx_chan_id;
+	unsigned long max_clk_rate;
 };
 
 static struct pxa_spi_info spi_info_configs[] = {
@@ -32,6 +35,7 @@ static struct pxa_spi_info spi_info_configs[] = {
 		.tx_chan_id = -1,
 		.rx_slave_id = -1,
 		.rx_chan_id = -1,
+		.max_clk_rate = 3686400,
 	},
 	[PORT_BYT] = {
 		.type = LPSS_SSP,
@@ -41,6 +45,7 @@ static struct pxa_spi_info spi_info_configs[] = {
 		.tx_chan_id = 0,
 		.rx_slave_id = 1,
 		.rx_chan_id = 1,
+		.max_clk_rate = 50000000,
 	},
 };
 
@@ -53,6 +58,7 @@ static int pxa2xx_spi_pci_probe(struct pci_dev *dev,
 	struct pxa2xx_spi_master spi_pdata;
 	struct ssp_device *ssp;
 	struct pxa_spi_info *c;
+	char buf[40];
 
 	ret = pcim_enable_device(dev);
 	if (ret)
@@ -84,6 +90,12 @@ static int pxa2xx_spi_pci_probe(struct pci_dev *dev,
 	ssp->port_id = (c->port_id >= 0) ? c->port_id : dev->devfn;
 	ssp->type = c->type;
 
+	snprintf(buf, sizeof(buf), "pxa2xx-spi.%d", ssp->port_id);
+	ssp->clk = clk_register_fixed_rate(&dev->dev, buf , NULL,
+					CLK_IS_ROOT, c->max_clk_rate);
+	 if (IS_ERR(ssp->clk))
+		return PTR_ERR(ssp->clk);
+
 	memset(&pi, 0, sizeof(pi));
 	pi.parent = &dev->dev;
 	pi.name = "pxa2xx-spi";
@@ -92,8 +104,10 @@ static int pxa2xx_spi_pci_probe(struct pci_dev *dev,
 	pi.size_data = sizeof(spi_pdata);
 
 	pdev = platform_device_register_full(&pi);
-	if (IS_ERR(pdev))
+	if (IS_ERR(pdev)) {
+		clk_unregister(ssp->clk);
 		return PTR_ERR(pdev);
+	}
 
 	pci_set_drvdata(dev, pdev);
 
@@ -103,8 +117,12 @@ static int pxa2xx_spi_pci_probe(struct pci_dev *dev,
 static void pxa2xx_spi_pci_remove(struct pci_dev *dev)
 {
 	struct platform_device *pdev = pci_get_drvdata(dev);
+	struct pxa2xx_spi_master *spi_pdata;
+
+	spi_pdata = dev_get_platdata(&pdev->dev);
 
 	platform_device_unregister(pdev);
+	clk_unregister(spi_pdata->ssp.clk);
 }
 
 static const struct pci_device_id pxa2xx_spi_pci_devices[] = {
diff --git a/drivers/spi/spi-pxa2xx.c b/drivers/spi/spi-pxa2xx.c
index fe79210..46f45ca 100644
--- a/drivers/spi/spi-pxa2xx.c
+++ b/drivers/spi/spi-pxa2xx.c
@@ -1074,6 +1074,7 @@ static struct acpi_device_id pxa2xx_spi_acpi_match[] = {
 	{ "INT3430", 0 },
 	{ "INT3431", 0 },
 	{ "80860F0E", 0 },
+	{ "8086228E", 0 },
 	{ },
 };
 MODULE_DEVICE_TABLE(acpi, pxa2xx_spi_acpi_match);
diff --git a/drivers/spi/spi-rockchip.c b/drivers/spi/spi-rockchip.c
index c074360..f96ea8a 100644
--- a/drivers/spi/spi-rockchip.c
+++ b/drivers/spi/spi-rockchip.c
@@ -220,7 +220,7 @@ static inline void wait_for_idle(struct rockchip_spi *rs)
 	do {
 		if (!(readl_relaxed(rs->regs + ROCKCHIP_SPI_SR) & SR_BUSY))
 			return;
-	} while (time_before(jiffies, timeout));
+	} while (!time_after(jiffies, timeout));
 
 	dev_warn(rs->dev, "spi controller is in busy state!\n");
 }
@@ -415,7 +415,7 @@ static void rockchip_spi_dma_txcb(void *data)
 	spin_unlock_irqrestore(&rs->lock, flags);
 }
 
-static int rockchip_spi_dma_transfer(struct rockchip_spi *rs)
+static void rockchip_spi_prepare_dma(struct rockchip_spi *rs)
 {
 	unsigned long flags;
 	struct dma_slave_config rxconf, txconf;
@@ -474,8 +474,6 @@ static int rockchip_spi_dma_transfer(struct rockchip_spi *rs)
 		dmaengine_submit(txdesc);
 		dma_async_issue_pending(rs->dma_tx.ch);
 	}
-
-	return 1;
 }
 
 static void rockchip_spi_config(struct rockchip_spi *rs)
@@ -499,7 +497,7 @@ static void rockchip_spi_config(struct rockchip_spi *rs)
 	}
 
 	/* div doesn't support odd number */
-	div = rs->max_freq / rs->speed;
+	div = max_t(u32, rs->max_freq / rs->speed, 1);
 	div = (div + 1) & 0xfffe;
 
 	spi_enable_chip(rs, 0);
@@ -529,7 +527,8 @@ static int rockchip_spi_transfer_one(
 	int ret = 0;
 	struct rockchip_spi *rs = spi_master_get_devdata(master);
 
-	WARN_ON((readl_relaxed(rs->regs + ROCKCHIP_SPI_SR) & SR_BUSY));
+	WARN_ON(readl_relaxed(rs->regs + ROCKCHIP_SPI_SSIENR) &&
+		(readl_relaxed(rs->regs + ROCKCHIP_SPI_SR) & SR_BUSY));
 
 	if (!xfer->tx_buf && !xfer->rx_buf) {
 		dev_err(rs->dev, "No buffer for transfer\n");
@@ -556,16 +555,17 @@ static int rockchip_spi_transfer_one(
 	else if (rs->rx)
 		rs->tmode = CR0_XFM_RO;
 
-	if (master->can_dma && master->can_dma(master, spi, xfer))
+	/* we need prepare dma before spi was enabled */
+	if (master->can_dma && master->can_dma(master, spi, xfer)) {
 		rs->use_dma = 1;
-	else
+		rockchip_spi_prepare_dma(rs);
+	} else {
 		rs->use_dma = 0;
+	}
 
 	rockchip_spi_config(rs);
 
-	if (rs->use_dma)
-		ret = rockchip_spi_dma_transfer(rs);
-	else
+	if (!rs->use_dma)
 		ret = rockchip_spi_pio_transfer(rs);
 
 	return ret;
@@ -678,7 +678,7 @@ static int rockchip_spi_probe(struct platform_device *pdev)
 		rs->dma_tx.addr = (dma_addr_t)(mem->start + ROCKCHIP_SPI_TXDR);
 		rs->dma_rx.addr = (dma_addr_t)(mem->start + ROCKCHIP_SPI_RXDR);
 		rs->dma_tx.direction = DMA_MEM_TO_DEV;
-		rs->dma_tx.direction = DMA_DEV_TO_MEM;
+		rs->dma_rx.direction = DMA_DEV_TO_MEM;
 
 		master->can_dma = rockchip_spi_can_dma;
 		master->dma_tx = rs->dma_tx.ch;
diff --git a/drivers/spi/spi-rspi.c b/drivers/spi/spi-rspi.c
index c850dfd..54bb0fa 100644
--- a/drivers/spi/spi-rspi.c
+++ b/drivers/spi/spi-rspi.c
@@ -87,7 +87,7 @@
 /* RSPI on SH only */
 #define SPCR_TXMD		0x02	/* TX Only Mode (vs. Full Duplex) */
 #define SPCR_SPMS		0x01	/* 3-wire Mode (vs. 4-wire) */
-/* QSPI on R-Car M2 only */
+/* QSPI on R-Car Gen2 only */
 #define SPCR_WSWAP		0x02	/* Word Swap of read-data for DMAC */
 #define SPCR_BSWAP		0x01	/* Byte Swap of read-data for DMAC */
 
@@ -472,25 +472,52 @@ static int rspi_dma_transfer(struct rspi_data *rspi, struct sg_table *tx,
 	dma_cookie_t cookie;
 	int ret;
 
-	if (tx) {
-		desc_tx = dmaengine_prep_slave_sg(rspi->master->dma_tx,
-					tx->sgl, tx->nents, DMA_TO_DEVICE,
-					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
-		if (!desc_tx)
-			goto no_dma;
-
-		irq_mask |= SPCR_SPTIE;
-	}
+	/* First prepare and submit the DMA request(s), as this may fail */
 	if (rx) {
 		desc_rx = dmaengine_prep_slave_sg(rspi->master->dma_rx,
 					rx->sgl, rx->nents, DMA_FROM_DEVICE,
 					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
-		if (!desc_rx)
-			goto no_dma;
+		if (!desc_rx) {
+			ret = -EAGAIN;
+			goto no_dma_rx;
+		}
+
+		desc_rx->callback = rspi_dma_complete;
+		desc_rx->callback_param = rspi;
+		cookie = dmaengine_submit(desc_rx);
+		if (dma_submit_error(cookie)) {
+			ret = cookie;
+			goto no_dma_rx;
+		}
 
 		irq_mask |= SPCR_SPRIE;
 	}
 
+	if (tx) {
+		desc_tx = dmaengine_prep_slave_sg(rspi->master->dma_tx,
+					tx->sgl, tx->nents, DMA_TO_DEVICE,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!desc_tx) {
+			ret = -EAGAIN;
+			goto no_dma_tx;
+		}
+
+		if (rx) {
+			/* No callback */
+			desc_tx->callback = NULL;
+		} else {
+			desc_tx->callback = rspi_dma_complete;
+			desc_tx->callback_param = rspi;
+		}
+		cookie = dmaengine_submit(desc_tx);
+		if (dma_submit_error(cookie)) {
+			ret = cookie;
+			goto no_dma_tx;
+		}
+
+		irq_mask |= SPCR_SPTIE;
+	}
+
 	/*
 	 * DMAC needs SPxIE, but if SPxIE is set, the IRQ routine will be
 	 * called. So, this driver disables the IRQ while DMA transfer.
@@ -503,34 +530,24 @@ static int rspi_dma_transfer(struct rspi_data *rspi, struct sg_table *tx,
 	rspi_enable_irq(rspi, irq_mask);
 	rspi->dma_callbacked = 0;
 
-	if (rx) {
-		desc_rx->callback = rspi_dma_complete;
-		desc_rx->callback_param = rspi;
-		cookie = dmaengine_submit(desc_rx);
-		if (dma_submit_error(cookie))
-			return cookie;
+	/* Now start DMA */
+	if (rx)
 		dma_async_issue_pending(rspi->master->dma_rx);
-	}
-	if (tx) {
-		if (rx) {
-			/* No callback */
-			desc_tx->callback = NULL;
-		} else {
-			desc_tx->callback = rspi_dma_complete;
-			desc_tx->callback_param = rspi;
-		}
-		cookie = dmaengine_submit(desc_tx);
-		if (dma_submit_error(cookie))
-			return cookie;
+	if (tx)
 		dma_async_issue_pending(rspi->master->dma_tx);
-	}
 
 	ret = wait_event_interruptible_timeout(rspi->wait,
 					       rspi->dma_callbacked, HZ);
 	if (ret > 0 && rspi->dma_callbacked)
 		ret = 0;
-	else if (!ret)
+	else if (!ret) {
+		dev_err(&rspi->master->dev, "DMA timeout\n");
 		ret = -ETIMEDOUT;
+		if (tx)
+			dmaengine_terminate_all(rspi->master->dma_tx);
+		if (rx)
+			dmaengine_terminate_all(rspi->master->dma_rx);
+	}
 
 	rspi_disable_irq(rspi, irq_mask);
 
@@ -541,11 +558,16 @@ static int rspi_dma_transfer(struct rspi_data *rspi, struct sg_table *tx,
 
 	return ret;
 
-no_dma:
-	pr_warn_once("%s %s: DMA not available, falling back to PIO\n",
-		     dev_driver_string(&rspi->master->dev),
-		     dev_name(&rspi->master->dev));
-	return -EAGAIN;
+no_dma_tx:
+	if (rx)
+		dmaengine_terminate_all(rspi->master->dma_rx);
+no_dma_rx:
+	if (ret == -EAGAIN) {
+		pr_warn_once("%s %s: DMA not available, falling back to PIO\n",
+			     dev_driver_string(&rspi->master->dev),
+			     dev_name(&rspi->master->dev));
+	}
+	return ret;
 }
 
 static void rspi_receive_init(const struct rspi_data *rspi)
@@ -887,20 +909,24 @@ static struct dma_chan *rspi_request_dma_chan(struct device *dev,
 	dma_cap_zero(mask);
 	dma_cap_set(DMA_SLAVE, mask);
 
-	chan = dma_request_channel(mask, shdma_chan_filter,
-				   (void *)(unsigned long)id);
+	chan = dma_request_slave_channel_compat(mask, shdma_chan_filter,
+				(void *)(unsigned long)id, dev,
+				dir == DMA_MEM_TO_DEV ? "tx" : "rx");
 	if (!chan) {
-		dev_warn(dev, "dma_request_channel failed\n");
+		dev_warn(dev, "dma_request_slave_channel_compat failed\n");
 		return NULL;
 	}
 
 	memset(&cfg, 0, sizeof(cfg));
 	cfg.slave_id = id;
 	cfg.direction = dir;
-	if (dir == DMA_MEM_TO_DEV)
+	if (dir == DMA_MEM_TO_DEV) {
 		cfg.dst_addr = port_addr;
-	else
+		cfg.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	} else {
 		cfg.src_addr = port_addr;
+		cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+	}
 
 	ret = dmaengine_slave_config(chan, &cfg);
 	if (ret) {
@@ -916,22 +942,30 @@ static int rspi_request_dma(struct device *dev, struct spi_master *master,
 			    const struct resource *res)
 {
 	const struct rspi_plat_data *rspi_pd = dev_get_platdata(dev);
+	unsigned int dma_tx_id, dma_rx_id;
+
+	if (dev->of_node) {
+		/* In the OF case we will get the slave IDs from the DT */
+		dma_tx_id = 0;
+		dma_rx_id = 0;
+	} else if (rspi_pd && rspi_pd->dma_tx_id && rspi_pd->dma_rx_id) {
+		dma_tx_id = rspi_pd->dma_tx_id;
+		dma_rx_id = rspi_pd->dma_rx_id;
+	} else {
+		/* The driver assumes no error. */
+		return 0;
+	}
 
-	if (!rspi_pd || !rspi_pd->dma_rx_id || !rspi_pd->dma_tx_id)
-		return 0;	/* The driver assumes no error. */
-
-	master->dma_rx = rspi_request_dma_chan(dev, DMA_DEV_TO_MEM,
-					       rspi_pd->dma_rx_id,
+	master->dma_tx = rspi_request_dma_chan(dev, DMA_MEM_TO_DEV, dma_tx_id,
 					       res->start + RSPI_SPDR);
-	if (!master->dma_rx)
+	if (!master->dma_tx)
 		return -ENODEV;
 
-	master->dma_tx = rspi_request_dma_chan(dev, DMA_MEM_TO_DEV,
-					       rspi_pd->dma_tx_id,
+	master->dma_rx = rspi_request_dma_chan(dev, DMA_DEV_TO_MEM, dma_rx_id,
 					       res->start + RSPI_SPDR);
-	if (!master->dma_tx) {
-		dma_release_channel(master->dma_rx);
-		master->dma_rx = NULL;
+	if (!master->dma_rx) {
+		dma_release_channel(master->dma_tx);
+		master->dma_tx = NULL;
 		return -ENODEV;
 	}
 
@@ -1024,12 +1058,11 @@ static int rspi_request_irq(struct device *dev, unsigned int irq,
 			    irq_handler_t handler, const char *suffix,
 			    void *dev_id)
 {
-	const char *base = dev_name(dev);
-	size_t len = strlen(base) + strlen(suffix) + 2;
-	char *name = devm_kzalloc(dev, len, GFP_KERNEL);
+	const char *name = devm_kasprintf(dev, GFP_KERNEL, "%s:%s",
+					  dev_name(dev), suffix);
 	if (!name)
 		return -ENOMEM;
-	snprintf(name, len, "%s:%s", base, suffix);
+
 	return devm_request_irq(dev, irq, handler, 0, name, dev_id);
 }
 
@@ -1062,7 +1095,7 @@ static int rspi_probe(struct platform_device *pdev)
 			master->num_chipselect = rspi_pd->num_chipselect;
 		else
 			master->num_chipselect = 2; /* default */
-	};
+	}
 
 	/* ops parameter check */
 	if (!ops->set_config_register) {
diff --git a/drivers/spi/spi-sh-msiof.c b/drivers/spi/spi-sh-msiof.c
index 2a4354d..3f36540 100644
--- a/drivers/spi/spi-sh-msiof.c
+++ b/drivers/spi/spi-sh-msiof.c
@@ -636,17 +636,7 @@ static int sh_msiof_dma_once(struct sh_msiof_spi_priv *p, const void *tx,
 	dma_cookie_t cookie;
 	int ret;
 
-	if (tx) {
-		ier_bits |= IER_TDREQE | IER_TDMAE;
-		dma_sync_single_for_device(p->master->dma_tx->device->dev,
-					   p->tx_dma_addr, len, DMA_TO_DEVICE);
-		desc_tx = dmaengine_prep_slave_single(p->master->dma_tx,
-					p->tx_dma_addr, len, DMA_TO_DEVICE,
-					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
-		if (!desc_tx)
-			return -EAGAIN;
-	}
-
+	/* First prepare and submit the DMA request(s), as this may fail */
 	if (rx) {
 		ier_bits |= IER_RDREQE | IER_RDMAE;
 		desc_rx = dmaengine_prep_slave_single(p->master->dma_rx,
@@ -654,30 +644,26 @@ static int sh_msiof_dma_once(struct sh_msiof_spi_priv *p, const void *tx,
 					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 		if (!desc_rx)
 			return -EAGAIN;
-	}
-
-	/* 1 stage FIFO watermarks for DMA */
-	sh_msiof_write(p, FCTR, FCTR_TFWM_1 | FCTR_RFWM_1);
 
-	/* setup msiof transfer mode registers (32-bit words) */
-	sh_msiof_spi_set_mode_regs(p, tx, rx, 32, len / 4);
-
-	sh_msiof_write(p, IER, ier_bits);
-
-	reinit_completion(&p->done);
-
-	if (rx) {
 		desc_rx->callback = sh_msiof_dma_complete;
 		desc_rx->callback_param = p;
 		cookie = dmaengine_submit(desc_rx);
-		if (dma_submit_error(cookie)) {
-			ret = cookie;
-			goto stop_ier;
-		}
-		dma_async_issue_pending(p->master->dma_rx);
+		if (dma_submit_error(cookie))
+			return cookie;
 	}
 
 	if (tx) {
+		ier_bits |= IER_TDREQE | IER_TDMAE;
+		dma_sync_single_for_device(p->master->dma_tx->device->dev,
+					   p->tx_dma_addr, len, DMA_TO_DEVICE);
+		desc_tx = dmaengine_prep_slave_single(p->master->dma_tx,
+					p->tx_dma_addr, len, DMA_TO_DEVICE,
+					DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+		if (!desc_tx) {
+			ret = -EAGAIN;
+			goto no_dma_tx;
+		}
+
 		if (rx) {
 			/* No callback */
 			desc_tx->callback = NULL;
@@ -688,15 +674,30 @@ static int sh_msiof_dma_once(struct sh_msiof_spi_priv *p, const void *tx,
 		cookie = dmaengine_submit(desc_tx);
 		if (dma_submit_error(cookie)) {
 			ret = cookie;
-			goto stop_rx;
+			goto no_dma_tx;
 		}
-		dma_async_issue_pending(p->master->dma_tx);
 	}
 
+	/* 1 stage FIFO watermarks for DMA */
+	sh_msiof_write(p, FCTR, FCTR_TFWM_1 | FCTR_RFWM_1);
+
+	/* setup msiof transfer mode registers (32-bit words) */
+	sh_msiof_spi_set_mode_regs(p, tx, rx, 32, len / 4);
+
+	sh_msiof_write(p, IER, ier_bits);
+
+	reinit_completion(&p->done);
+
+	/* Now start DMA */
+	if (rx)
+		dma_async_issue_pending(p->master->dma_rx);
+	if (tx)
+		dma_async_issue_pending(p->master->dma_tx);
+
 	ret = sh_msiof_spi_start(p, rx);
 	if (ret) {
 		dev_err(&p->pdev->dev, "failed to start hardware\n");
-		goto stop_tx;
+		goto stop_dma;
 	}
 
 	/* wait for tx fifo to be emptied / rx fifo to be filled */
@@ -726,13 +727,12 @@ static int sh_msiof_dma_once(struct sh_msiof_spi_priv *p, const void *tx,
 stop_reset:
 	sh_msiof_reset_str(p);
 	sh_msiof_spi_stop(p, rx);
-stop_tx:
+stop_dma:
 	if (tx)
 		dmaengine_terminate_all(p->master->dma_tx);
-stop_rx:
+no_dma_tx:
 	if (rx)
 		dmaengine_terminate_all(p->master->dma_rx);
-stop_ier:
 	sh_msiof_write(p, IER, 0);
 	return ret;
 }
@@ -928,6 +928,9 @@ static const struct of_device_id sh_msiof_match[] = {
 	{ .compatible = "renesas,sh-mobile-msiof", .data = &sh_data },
 	{ .compatible = "renesas,msiof-r8a7790",   .data = &r8a779x_data },
 	{ .compatible = "renesas,msiof-r8a7791",   .data = &r8a779x_data },
+	{ .compatible = "renesas,msiof-r8a7792",   .data = &r8a779x_data },
+	{ .compatible = "renesas,msiof-r8a7793",   .data = &r8a779x_data },
+	{ .compatible = "renesas,msiof-r8a7794",   .data = &r8a779x_data },
 	{},
 };
 MODULE_DEVICE_TABLE(of, sh_msiof_match);
@@ -972,20 +975,24 @@ static struct dma_chan *sh_msiof_request_dma_chan(struct device *dev,
 	dma_cap_zero(mask);
 	dma_cap_set(DMA_SLAVE, mask);
 
-	chan = dma_request_channel(mask, shdma_chan_filter,
-				  (void *)(unsigned long)id);
+	chan = dma_request_slave_channel_compat(mask, shdma_chan_filter,
+				(void *)(unsigned long)id, dev,
+				dir == DMA_MEM_TO_DEV ? "tx" : "rx");
 	if (!chan) {
-		dev_warn(dev, "dma_request_channel failed\n");
+		dev_warn(dev, "dma_request_slave_channel_compat failed\n");
 		return NULL;
 	}
 
 	memset(&cfg, 0, sizeof(cfg));
 	cfg.slave_id = id;
 	cfg.direction = dir;
-	if (dir == DMA_MEM_TO_DEV)
+	if (dir == DMA_MEM_TO_DEV) {
 		cfg.dst_addr = port_addr;
-	else
+		cfg.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+	} else {
 		cfg.src_addr = port_addr;
+		cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+	}
 
 	ret = dmaengine_slave_config(chan, &cfg);
 	if (ret) {
@@ -1002,12 +1009,22 @@ static int sh_msiof_request_dma(struct sh_msiof_spi_priv *p)
 	struct platform_device *pdev = p->pdev;
 	struct device *dev = &pdev->dev;
 	const struct sh_msiof_spi_info *info = dev_get_platdata(dev);
+	unsigned int dma_tx_id, dma_rx_id;
 	const struct resource *res;
 	struct spi_master *master;
 	struct device *tx_dev, *rx_dev;
 
-	if (!info || !info->dma_tx_id || !info->dma_rx_id)
-		return 0;	/* The driver assumes no error */
+	if (dev->of_node) {
+		/* In the OF case we will get the slave IDs from the DT */
+		dma_tx_id = 0;
+		dma_rx_id = 0;
+	} else if (info && info->dma_tx_id && info->dma_rx_id) {
+		dma_tx_id = info->dma_tx_id;
+		dma_rx_id = info->dma_rx_id;
+	} else {
+		/* The driver assumes no error */
+		return 0;
+	}
 
 	/* The DMA engine uses the second register set, if present */
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
@@ -1016,13 +1033,13 @@ static int sh_msiof_request_dma(struct sh_msiof_spi_priv *p)
 
 	master = p->master;
 	master->dma_tx = sh_msiof_request_dma_chan(dev, DMA_MEM_TO_DEV,
-						   info->dma_tx_id,
+						   dma_tx_id,
 						   res->start + TFDR);
 	if (!master->dma_tx)
 		return -ENODEV;
 
 	master->dma_rx = sh_msiof_request_dma_chan(dev, DMA_DEV_TO_MEM,
-						   info->dma_rx_id,
+						   dma_rx_id,
 						   res->start + RFDR);
 	if (!master->dma_rx)
 		goto free_tx_chan;
@@ -1205,6 +1222,9 @@ static struct platform_device_id spi_driver_ids[] = {
 	{ "spi_sh_msiof",	(kernel_ulong_t)&sh_data },
 	{ "spi_r8a7790_msiof",	(kernel_ulong_t)&r8a779x_data },
 	{ "spi_r8a7791_msiof",	(kernel_ulong_t)&r8a779x_data },
+	{ "spi_r8a7792_msiof",	(kernel_ulong_t)&r8a779x_data },
+	{ "spi_r8a7793_msiof",	(kernel_ulong_t)&r8a779x_data },
+	{ "spi_r8a7794_msiof",	(kernel_ulong_t)&r8a779x_data },
 	{},
 };
 MODULE_DEVICE_TABLE(platform, spi_driver_ids);
diff --git a/drivers/spi/spi-sirf.c b/drivers/spi/spi-sirf.c
index 95ac276..39e2c0a 100644
--- a/drivers/spi/spi-sirf.c
+++ b/drivers/spi/spi-sirf.c
@@ -62,15 +62,15 @@
 #define SIRFSOC_SPI_TRAN_DAT_FORMAT_12	(1 << 26)
 #define SIRFSOC_SPI_TRAN_DAT_FORMAT_16	(2 << 26)
 #define SIRFSOC_SPI_TRAN_DAT_FORMAT_32	(3 << 26)
-#define SIRFSOC_SPI_CMD_BYTE_NUM(x)		((x & 3) << 28)
-#define SIRFSOC_SPI_ENA_AUTO_CLR		BIT(30)
-#define SIRFSOC_SPI_MUL_DAT_MODE		BIT(31)
+#define SIRFSOC_SPI_CMD_BYTE_NUM(x)	((x & 3) << 28)
+#define SIRFSOC_SPI_ENA_AUTO_CLR	BIT(30)
+#define SIRFSOC_SPI_MUL_DAT_MODE	BIT(31)
 
 /* Interrupt Enable */
-#define SIRFSOC_SPI_RX_DONE_INT_EN		BIT(0)
-#define SIRFSOC_SPI_TX_DONE_INT_EN		BIT(1)
-#define SIRFSOC_SPI_RX_OFLOW_INT_EN		BIT(2)
-#define SIRFSOC_SPI_TX_UFLOW_INT_EN		BIT(3)
+#define SIRFSOC_SPI_RX_DONE_INT_EN	BIT(0)
+#define SIRFSOC_SPI_TX_DONE_INT_EN	BIT(1)
+#define SIRFSOC_SPI_RX_OFLOW_INT_EN	BIT(2)
+#define SIRFSOC_SPI_TX_UFLOW_INT_EN	BIT(3)
 #define SIRFSOC_SPI_RX_IO_DMA_INT_EN	BIT(4)
 #define SIRFSOC_SPI_TX_IO_DMA_INT_EN	BIT(5)
 #define SIRFSOC_SPI_RXFIFO_FULL_INT_EN	BIT(6)
@@ -79,7 +79,7 @@
 #define SIRFSOC_SPI_TXFIFO_THD_INT_EN	BIT(9)
 #define SIRFSOC_SPI_FRM_END_INT_EN	BIT(10)
 
-#define SIRFSOC_SPI_INT_MASK_ALL		0x1FFF
+#define SIRFSOC_SPI_INT_MASK_ALL	0x1FFF
 
 /* Interrupt status */
 #define SIRFSOC_SPI_RX_DONE		BIT(0)
@@ -170,8 +170,7 @@ struct sirfsoc_spi {
 	 * command model
 	 */
 	bool	tx_by_cmd;
-
-	int chipselect[0];
+	bool	hw_cs;
 };
 
 static void spi_sirfsoc_rx_word_u8(struct sirfsoc_spi *sspi)
@@ -304,7 +303,7 @@ static void spi_sirfsoc_dma_fini_callback(void *data)
 	complete(dma_complete);
 }
 
-static int spi_sirfsoc_cmd_transfer(struct spi_device *spi,
+static void spi_sirfsoc_cmd_transfer(struct spi_device *spi,
 	struct spi_transfer *t)
 {
 	struct sirfsoc_spi *sspi;
@@ -312,6 +311,8 @@ static int spi_sirfsoc_cmd_transfer(struct spi_device *spi,
 	u32 cmd;
 
 	sspi = spi_master_get_devdata(spi->master);
+	writel(SIRFSOC_SPI_FIFO_RESET, sspi->base + SIRFSOC_SPI_TXFIFO_OP);
+	writel(SIRFSOC_SPI_FIFO_START, sspi->base + SIRFSOC_SPI_TXFIFO_OP);
 	memcpy(&cmd, sspi->tx, t->len);
 	if (sspi->word_width == 1 && !(spi->mode & SPI_LSB_FIRST))
 		cmd = cpu_to_be32(cmd) >>
@@ -326,10 +327,9 @@ static int spi_sirfsoc_cmd_transfer(struct spi_device *spi,
 		sspi->base + SIRFSOC_SPI_TX_RX_EN);
 	if (wait_for_completion_timeout(&sspi->tx_done, timeout) == 0) {
 		dev_err(&spi->dev, "cmd transfer timeout\n");
-		return 0;
+		return;
 	}
-
-	return t->len;
+	sspi->left_rx_word -= t->len;
 }
 
 static void spi_sirfsoc_dma_transfer(struct spi_device *spi,
@@ -438,7 +438,8 @@ static void spi_sirfsoc_pio_transfer(struct spi_device *spi,
 			sspi->tx_word(sspi);
 		writel(SIRFSOC_SPI_TXFIFO_EMPTY_INT_EN |
 			SIRFSOC_SPI_TX_UFLOW_INT_EN |
-			SIRFSOC_SPI_RX_OFLOW_INT_EN,
+			SIRFSOC_SPI_RX_OFLOW_INT_EN |
+			SIRFSOC_SPI_RX_IO_DMA_INT_EN,
 			sspi->base + SIRFSOC_SPI_INT_EN);
 		writel(SIRFSOC_SPI_RX_EN | SIRFSOC_SPI_TX_EN,
 			sspi->base + SIRFSOC_SPI_TX_RX_EN);
@@ -484,7 +485,7 @@ static void spi_sirfsoc_chipselect(struct spi_device *spi, int value)
 {
 	struct sirfsoc_spi *sspi = spi_master_get_devdata(spi->master);
 
-	if (sspi->chipselect[spi->chip_select] == 0) {
+	if (sspi->hw_cs) {
 		u32 regval = readl(sspi->base + SIRFSOC_SPI_CTRL);
 		switch (value) {
 		case BITBANG_CS_ACTIVE:
@@ -502,14 +503,13 @@ static void spi_sirfsoc_chipselect(struct spi_device *spi, int value)
 		}
 		writel(regval, sspi->base + SIRFSOC_SPI_CTRL);
 	} else {
-		int gpio = sspi->chipselect[spi->chip_select];
 		switch (value) {
 		case BITBANG_CS_ACTIVE:
-			gpio_direction_output(gpio,
+			gpio_direction_output(spi->cs_gpio,
 					spi->mode & SPI_CS_HIGH ? 1 : 0);
 			break;
 		case BITBANG_CS_INACTIVE:
-			gpio_direction_output(gpio,
+			gpio_direction_output(spi->cs_gpio,
 					spi->mode & SPI_CS_HIGH ? 0 : 1);
 			break;
 		}
@@ -603,8 +603,8 @@ spi_sirfsoc_setup_transfer(struct spi_device *spi, struct spi_transfer *t)
 		sspi->tx_by_cmd = false;
 	}
 	/*
-	 * set spi controller in RISC chipselect mode, we are controlling CS by
-	 * software BITBANG_CS_ACTIVE and BITBANG_CS_INACTIVE.
+	 * it should never set to hardware cs mode because in hardware cs mode,
+	 * cs signal can't controlled by driver.
 	 */
 	regval |= SIRFSOC_SPI_CS_IO_MODE;
 	writel(regval, sspi->base + SIRFSOC_SPI_CTRL);
@@ -627,9 +627,17 @@ spi_sirfsoc_setup_transfer(struct spi_device *spi, struct spi_transfer *t)
 
 static int spi_sirfsoc_setup(struct spi_device *spi)
 {
+	struct sirfsoc_spi *sspi;
+
 	if (!spi->max_speed_hz)
 		return -EINVAL;
 
+	sspi = spi_master_get_devdata(spi->master);
+
+	if (spi->cs_gpio == -ENOENT)
+		sspi->hw_cs = true;
+	else
+		sspi->hw_cs = false;
 	return spi_sirfsoc_setup_transfer(spi, NULL);
 }
 
@@ -638,19 +646,10 @@ static int spi_sirfsoc_probe(struct platform_device *pdev)
 	struct sirfsoc_spi *sspi;
 	struct spi_master *master;
 	struct resource *mem_res;
-	int num_cs, cs_gpio, irq;
-	int i;
-	int ret;
+	int irq;
+	int i, ret;
 
-	ret = of_property_read_u32(pdev->dev.of_node,
-			"sirf,spi-num-chipselects", &num_cs);
-	if (ret < 0) {
-		dev_err(&pdev->dev, "Unable to get chip select number\n");
-		goto err_cs;
-	}
-
-	master = spi_alloc_master(&pdev->dev,
-			sizeof(*sspi) + sizeof(int) * num_cs);
+	master = spi_alloc_master(&pdev->dev, sizeof(*sspi));
 	if (!master) {
 		dev_err(&pdev->dev, "Unable to allocate SPI master\n");
 		return -ENOMEM;
@@ -658,32 +657,6 @@ static int spi_sirfsoc_probe(struct platform_device *pdev)
 	platform_set_drvdata(pdev, master);
 	sspi = spi_master_get_devdata(master);
 
-	master->num_chipselect = num_cs;
-
-	for (i = 0; i < master->num_chipselect; i++) {
-		cs_gpio = of_get_named_gpio(pdev->dev.of_node, "cs-gpios", i);
-		if (cs_gpio < 0) {
-			dev_err(&pdev->dev, "can't get cs gpio from DT\n");
-			ret = -ENODEV;
-			goto free_master;
-		}
-
-		sspi->chipselect[i] = cs_gpio;
-		if (cs_gpio == 0)
-			continue; /* use cs from spi controller */
-
-		ret = gpio_request(cs_gpio, DRIVER_NAME);
-		if (ret) {
-			while (i > 0) {
-				i--;
-				if (sspi->chipselect[i] > 0)
-					gpio_free(sspi->chipselect[i]);
-			}
-			dev_err(&pdev->dev, "fail to request cs gpios\n");
-			goto free_master;
-		}
-	}
-
 	mem_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	sspi->base = devm_ioremap_resource(&pdev->dev, mem_res);
 	if (IS_ERR(sspi->base)) {
@@ -753,7 +726,21 @@ static int spi_sirfsoc_probe(struct platform_device *pdev)
 	ret = spi_bitbang_start(&sspi->bitbang);
 	if (ret)
 		goto free_dummypage;
-
+	for (i = 0; master->cs_gpios && i < master->num_chipselect; i++) {
+		if (master->cs_gpios[i] == -ENOENT)
+			continue;
+		if (!gpio_is_valid(master->cs_gpios[i])) {
+			dev_err(&pdev->dev, "no valid gpio\n");
+			ret = -EINVAL;
+			goto free_dummypage;
+		}
+		ret = devm_gpio_request(&pdev->dev,
+				master->cs_gpios[i], DRIVER_NAME);
+		if (ret) {
+			dev_err(&pdev->dev, "failed to request gpio\n");
+			goto free_dummypage;
+		}
+	}
 	dev_info(&pdev->dev, "registerred, bus number = %d\n", master->bus_num);
 
 	return 0;
@@ -768,7 +755,7 @@ free_rx_dma:
 	dma_release_channel(sspi->rx_chan);
 free_master:
 	spi_master_put(master);
-err_cs:
+
 	return ret;
 }
 
@@ -776,16 +763,11 @@ static int  spi_sirfsoc_remove(struct platform_device *pdev)
 {
 	struct spi_master *master;
 	struct sirfsoc_spi *sspi;
-	int i;
 
 	master = platform_get_drvdata(pdev);
 	sspi = spi_master_get_devdata(master);
 
 	spi_bitbang_stop(&sspi->bitbang);
-	for (i = 0; i < master->num_chipselect; i++) {
-		if (sspi->chipselect[i] > 0)
-			gpio_free(sspi->chipselect[i]);
-	}
 	kfree(sspi->dummypage);
 	clk_disable_unprepare(sspi->clk);
 	clk_put(sspi->clk);
diff --git a/drivers/spi/spi-xilinx.c b/drivers/spi/spi-xilinx.c
index 4d8efb1..79bd84f 100644
--- a/drivers/spi/spi-xilinx.c
+++ b/drivers/spi/spi-xilinx.c
@@ -471,7 +471,6 @@ static struct platform_driver xilinx_spi_driver = {
 	.remove = xilinx_spi_remove,
 	.driver = {
 		.name = XILINX_SPI_NAME,
-		.owner = THIS_MODULE,
 		.of_match_table = xilinx_spi_of_match,
 	},
 };
diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
index e0531ba..e19512f 100644
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -552,6 +552,9 @@ int spi_register_board_info(struct spi_board_info const *info, unsigned n)
 	struct boardinfo *bi;
 	int i;
 
+	if (!n)
+		return -EINVAL;
+
 	bi = kzalloc(n * sizeof(*bi), GFP_KERNEL);
 	if (!bi)
 		return -ENOMEM;
@@ -789,27 +792,35 @@ static int spi_transfer_one_message(struct spi_master *master,
 	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
 		trace_spi_transfer_start(msg, xfer);
 
-		reinit_completion(&master->xfer_completion);
+		if (xfer->tx_buf || xfer->rx_buf) {
+			reinit_completion(&master->xfer_completion);
 
-		ret = master->transfer_one(master, msg->spi, xfer);
-		if (ret < 0) {
-			dev_err(&msg->spi->dev,
-				"SPI transfer failed: %d\n", ret);
-			goto out;
-		}
+			ret = master->transfer_one(master, msg->spi, xfer);
+			if (ret < 0) {
+				dev_err(&msg->spi->dev,
+					"SPI transfer failed: %d\n", ret);
+				goto out;
+			}
 
-		if (ret > 0) {
-			ret = 0;
-			ms = xfer->len * 8 * 1000 / xfer->speed_hz;
-			ms += ms + 100; /* some tolerance */
+			if (ret > 0) {
+				ret = 0;
+				ms = xfer->len * 8 * 1000 / xfer->speed_hz;
+				ms += ms + 100; /* some tolerance */
 
-			ms = wait_for_completion_timeout(&master->xfer_completion,
-							 msecs_to_jiffies(ms));
-		}
+				ms = wait_for_completion_timeout(&master->xfer_completion,
+								 msecs_to_jiffies(ms));
+			}
 
-		if (ms == 0) {
-			dev_err(&msg->spi->dev, "SPI transfer timed out\n");
-			msg->status = -ETIMEDOUT;
+			if (ms == 0) {
+				dev_err(&msg->spi->dev,
+					"SPI transfer timed out\n");
+				msg->status = -ETIMEDOUT;
+			}
+		} else {
+			if (xfer->len)
+				dev_err(&msg->spi->dev,
+					"Bufferless transfer has length %u\n",
+					xfer->len);
 		}
 
 		trace_spi_transfer_stop(msg, xfer);
@@ -848,6 +859,7 @@ out:
 
 /**
  * spi_finalize_current_transfer - report completion of a transfer
+ * @master: the master reporting completion
  *
  * Called by SPI drivers using the core transfer_one_message()
  * implementation to notify it that the current interrupt driven