Merge tag 'spi-v4.4' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi

Pull spi updates from Mark Brown:
 "Quite a lot of activity in SPI this cycle, almost all of it in drivers
  with a few minor improvements and tweaks in the core.

   - Updates to pxa2xx to support Intel Broxton and multiple chip selects.
   - Support for big endian in the bcm63xx driver.
   - Multiple slave support for the mt8173
   - New driver for the auxiliary SPI controller in bcm2835 SoCs.
   - Support for Layerscale SoCs in the Freescale DSPI driver"

* tag 'spi-v4.4' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi: (87 commits)
  spi: pxa2xx: Rework self-initiated platform data creation for non-ACPI
  spi: pxa2xx: Add support for Intel Broxton
  spi: pxa2xx: Detect number of enabled Intel LPSS SPI chip select signals
  spi: pxa2xx: Add output control for multiple Intel LPSS chip selects
  spi: pxa2xx: Use LPSS prefix for defines that are Intel LPSS specific
  spi: Add DSPI support for layerscape family
  spi: ti-qspi: improve ->remove() callback
  spi/spi-xilinx: Fix race condition on last word read
  spi: Drop owner assignment from spi_drivers
  spi: Add THIS_MODULE to spi_driver in SPI core
  spi: Setup the master controller driver before setting the chipselect
  spi: dw: replace magic constant by DW_SPI_DR
  spi: mediatek: mt8173 spi multiple devices support
  spi: mediatek: handle controller_data in mtk_spi_setup
  spi: mediatek: remove mtk_spi_config
  spi: mediatek: Update document devicetree bindings to support multiple devices
  spi: fix kernel-doc warnings about missing return desc in spi.c
  spi: fix kernel-doc warnings about missing return desc in spi.h
  spi: pxa2xx: Align a few defines
  spi: pxa2xx: Save other reg_cs_ctrl bits when configuring chip select
  ...

39 files changed, 1331 insertions, 625 deletions

diff --git a/drivers/spi/Kconfig b/drivers/spi/Kconfig
index 4887f317ea58..8b9c2a38d1cc 100644
--- a/drivers/spi/Kconfig
+++ b/drivers/spi/Kconfig
@@ -88,6 +88,17 @@ config SPI_BCM2835
 	  is for the regular SPI controller. Slave mode operation is not also
 	  not supported.
 
+config SPI_BCM2835AUX
+	tristate "BCM2835 SPI auxiliary controller"
+	depends on ARCH_BCM2835 || COMPILE_TEST
+	depends on GPIOLIB
+	help
+	  This selects a driver for the Broadcom BCM2835 SPI aux master.
+
+	  The BCM2835 contains two types of SPI master controller; the
+	  "universal SPI master", and the regular SPI controller.
+	  This driver is for the universal/auxiliary SPI controller.
+
 config SPI_BFIN5XX
 	tristate "SPI controller driver for ADI Blackfin5xx"
 	depends on BLACKFIN && !BF60x
@@ -125,7 +136,7 @@ config SPI_BCM53XX
 
 config SPI_BCM63XX
 	tristate "Broadcom BCM63xx SPI controller"
-	depends on BCM63XX
+	depends on BCM63XX || COMPILE_TEST
 	help
           Enable support for the SPI controller on the Broadcom BCM63xx SoCs.
 
@@ -304,7 +315,7 @@ config SPI_FSL_SPI
 config SPI_FSL_DSPI
 	tristate "Freescale DSPI controller"
 	select REGMAP_MMIO
-	depends on SOC_VF610 || SOC_LS1021A || COMPILE_TEST
+	depends on SOC_VF610 || SOC_LS1021A || ARCH_LAYERSCAPE || COMPILE_TEST
 	help
 	  This enables support for the Freescale DSPI controller in master
 	  mode. VF610 platform uses the controller.
diff --git a/drivers/spi/Makefile b/drivers/spi/Makefile
index 6a7f6f9d0d1c..31fb7fb2a0b6 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-$(CONFIG_SPI_BCM2835AUX)		+= spi-bcm2835aux.o
 obj-$(CONFIG_SPI_BCM53XX)		+= spi-bcm53xx.o
 obj-$(CONFIG_SPI_BCM63XX)		+= spi-bcm63xx.o
 obj-$(CONFIG_SPI_BCM63XX_HSSPI)		+= spi-bcm63xx-hsspi.o
diff --git a/drivers/spi/spi-ath79.c b/drivers/spi/spi-ath79.c
index bf1f9b32c597..6165bf21d427 100644
--- a/drivers/spi/spi-ath79.c
+++ b/drivers/spi/spi-ath79.c
@@ -240,14 +240,9 @@ static int ath79_spi_probe(struct platform_device *pdev)
 	sp->bitbang.flags = SPI_CS_HIGH;
 
 	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (r == NULL) {
-		ret = -ENOENT;
-		goto err_put_master;
-	}
-
-	sp->base = devm_ioremap(&pdev->dev, r->start, resource_size(r));
-	if (!sp->base) {
-		ret = -ENXIO;
+	sp->base = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(sp->base)) {
+		ret = PTR_ERR(sp->base);
 		goto err_put_master;
 	}
 
diff --git a/drivers/spi/spi-atmel.c b/drivers/spi/spi-atmel.c
index 63318e2afba1..aebad36391c9 100644
--- a/drivers/spi/spi-atmel.c
+++ b/drivers/spi/spi-atmel.c
@@ -773,7 +773,8 @@ static int atmel_spi_next_xfer_dma_submit(struct spi_master *master,
 
 	*plen = len;
 
-	if (atmel_spi_dma_slave_config(as, &slave_config, 8))
+	if (atmel_spi_dma_slave_config(as, &slave_config,
+				       xfer->bits_per_word))
 		goto err_exit;
 
 	/* Send both scatterlists */
@@ -871,14 +872,7 @@ static int atmel_spi_set_xfer_speed(struct atmel_spi *as,
 	 * Calculate the lowest divider that satisfies the
 	 * constraint, assuming div32/fdiv/mbz == 0.
 	 */
-	if (xfer->speed_hz)
-		scbr = DIV_ROUND_UP(bus_hz, xfer->speed_hz);
-	else
-		/*
-		 * This can happend if max_speed is null.
-		 * In this case, we set the lowest possible speed
-		 */
-		scbr = 0xff;
+	scbr = DIV_ROUND_UP(bus_hz, xfer->speed_hz);
 
 	/*
 	 * If the resulting divider doesn't fit into the
@@ -1300,14 +1294,12 @@ static int atmel_spi_one_transfer(struct spi_master *master,
 		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,
+	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;
-		}
+		return -ENOPROTOOPT;
 	}
 
 	/*
diff --git a/drivers/spi/spi-au1550.c b/drivers/spi/spi-au1550.c
index f45e085c01a6..afd239d6dec1 100644
--- a/drivers/spi/spi-au1550.c
+++ b/drivers/spi/spi-au1550.c
@@ -233,13 +233,12 @@ static int au1550_spi_setupxfer(struct spi_device *spi, struct spi_transfer *t)
 	unsigned bpw, hz;
 	u32 cfg, stat;
 
-	bpw = spi->bits_per_word;
-	hz = spi->max_speed_hz;
 	if (t) {
-		if (t->bits_per_word)
-			bpw = t->bits_per_word;
-		if (t->speed_hz)
-			hz = t->speed_hz;
+		bpw = t->bits_per_word;
+		hz = t->speed_hz;
+	} else {
+		bpw = spi->bits_per_word;
+		hz = spi->max_speed_hz;
 	}
 
 	if (!hz)
diff --git a/drivers/spi/spi-bcm2835.c b/drivers/spi/spi-bcm2835.c
index 3e8eeb23d4e9..cf04960cc3e6 100644
--- a/drivers/spi/spi-bcm2835.c
+++ b/drivers/spi/spi-bcm2835.c
@@ -777,7 +777,7 @@ static int bcm2835_spi_probe(struct platform_device *pdev)
 		goto out_master_put;
 	}
 
-	bs->irq = irq_of_parse_and_map(pdev->dev.of_node, 0);
+	bs->irq = platform_get_irq(pdev, 0);
 	if (bs->irq <= 0) {
 		dev_err(&pdev->dev, "could not get IRQ: %d\n", bs->irq);
 		err = bs->irq ? bs->irq : -ENODEV;
@@ -786,6 +786,12 @@ static int bcm2835_spi_probe(struct platform_device *pdev)
 
 	clk_prepare_enable(bs->clk);
 
+	bcm2835_dma_init(master, &pdev->dev);
+
+	/* initialise the hardware with the default polarities */
+	bcm2835_wr(bs, BCM2835_SPI_CS,
+		   BCM2835_SPI_CS_CLEAR_RX | BCM2835_SPI_CS_CLEAR_TX);
+
 	err = devm_request_irq(&pdev->dev, bs->irq, bcm2835_spi_interrupt, 0,
 			       dev_name(&pdev->dev), master);
 	if (err) {
@@ -793,12 +799,6 @@ static int bcm2835_spi_probe(struct platform_device *pdev)
 		goto out_clk_disable;
 	}
 
-	bcm2835_dma_init(master, &pdev->dev);
-
-	/* initialise the hardware with the default polarities */
-	bcm2835_wr(bs, BCM2835_SPI_CS,
-		   BCM2835_SPI_CS_CLEAR_RX | BCM2835_SPI_CS_CLEAR_TX);
-
 	err = devm_spi_register_master(&pdev->dev, master);
 	if (err) {
 		dev_err(&pdev->dev, "could not register SPI master: %d\n", err);
diff --git a/drivers/spi/spi-bcm2835aux.c b/drivers/spi/spi-bcm2835aux.c
new file mode 100644
index 000000000000..7de6f8472a81
--- /dev/null
+++ b/drivers/spi/spi-bcm2835aux.c
@@ -0,0 +1,512 @@
+/*
+ * Driver for Broadcom BCM2835 auxiliary SPI Controllers
+ *
+ * the driver does not rely on the native chipselects at all
+ * but only uses the gpio type chipselects
+ *
+ * Based on: spi-bcm2835.c
+ *
+ * Copyright (C) 2015 Martin Sperl
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/clk.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_device.h>
+#include <linux/of_gpio.h>
+#include <linux/of_irq.h>
+#include <linux/regmap.h>
+#include <linux/spi/spi.h>
+#include <linux/spinlock.h>
+
+/*
+ * spi register defines
+ *
+ * note there is garbage in the "official" documentation,
+ * so some data is taken from the file:
+ *   brcm_usrlib/dag/vmcsx/vcinclude/bcm2708_chip/aux_io.h
+ * inside of:
+ *   http://www.broadcom.com/docs/support/videocore/Brcm_Android_ICS_Graphics_Stack.tar.gz
+ */
+
+/* SPI register offsets */
+#define BCM2835_AUX_SPI_CNTL0	0x00
+#define BCM2835_AUX_SPI_CNTL1	0x04
+#define BCM2835_AUX_SPI_STAT	0x08
+#define BCM2835_AUX_SPI_PEEK	0x0C
+#define BCM2835_AUX_SPI_IO	0x20
+#define BCM2835_AUX_SPI_TXHOLD	0x30
+
+/* Bitfields in CNTL0 */
+#define BCM2835_AUX_SPI_CNTL0_SPEED	0xFFF00000
+#define BCM2835_AUX_SPI_CNTL0_SPEED_MAX	0xFFF
+#define BCM2835_AUX_SPI_CNTL0_SPEED_SHIFT	20
+#define BCM2835_AUX_SPI_CNTL0_CS	0x000E0000
+#define BCM2835_AUX_SPI_CNTL0_POSTINPUT	0x00010000
+#define BCM2835_AUX_SPI_CNTL0_VAR_CS	0x00008000
+#define BCM2835_AUX_SPI_CNTL0_VAR_WIDTH	0x00004000
+#define BCM2835_AUX_SPI_CNTL0_DOUTHOLD	0x00003000
+#define BCM2835_AUX_SPI_CNTL0_ENABLE	0x00000800
+#define BCM2835_AUX_SPI_CNTL0_CPHA_IN	0x00000400
+#define BCM2835_AUX_SPI_CNTL0_CLEARFIFO	0x00000200
+#define BCM2835_AUX_SPI_CNTL0_CPHA_OUT	0x00000100
+#define BCM2835_AUX_SPI_CNTL0_CPOL	0x00000080
+#define BCM2835_AUX_SPI_CNTL0_MSBF_OUT	0x00000040
+#define BCM2835_AUX_SPI_CNTL0_SHIFTLEN	0x0000003F
+
+/* Bitfields in CNTL1 */
+#define BCM2835_AUX_SPI_CNTL1_CSHIGH	0x00000700
+#define BCM2835_AUX_SPI_CNTL1_IDLE	0x00000080
+#define BCM2835_AUX_SPI_CNTL1_TXEMPTY	0x00000040
+#define BCM2835_AUX_SPI_CNTL1_MSBF_IN	0x00000002
+#define BCM2835_AUX_SPI_CNTL1_KEEP_IN	0x00000001
+
+/* Bitfields in STAT */
+#define BCM2835_AUX_SPI_STAT_TX_LVL	0xFF000000
+#define BCM2835_AUX_SPI_STAT_RX_LVL	0x00FF0000
+#define BCM2835_AUX_SPI_STAT_TX_FULL	0x00000400
+#define BCM2835_AUX_SPI_STAT_TX_EMPTY	0x00000200
+#define BCM2835_AUX_SPI_STAT_RX_FULL	0x00000100
+#define BCM2835_AUX_SPI_STAT_RX_EMPTY	0x00000080
+#define BCM2835_AUX_SPI_STAT_BUSY	0x00000040
+#define BCM2835_AUX_SPI_STAT_BITCOUNT	0x0000003F
+
+/* timeout values */
+#define BCM2835_AUX_SPI_POLLING_LIMIT_US	30
+#define BCM2835_AUX_SPI_POLLING_JIFFIES		2
+
+#define BCM2835_AUX_SPI_MODE_BITS (SPI_CPOL | SPI_CPHA | SPI_CS_HIGH \
+				  | SPI_NO_CS)
+
+struct bcm2835aux_spi {
+	void __iomem *regs;
+	struct clk *clk;
+	int irq;
+	u32 cntl[2];
+	const u8 *tx_buf;
+	u8 *rx_buf;
+	int tx_len;
+	int rx_len;
+	int pending;
+};
+
+static inline u32 bcm2835aux_rd(struct bcm2835aux_spi *bs, unsigned reg)
+{
+	return readl(bs->regs + reg);
+}
+
+static inline void bcm2835aux_wr(struct bcm2835aux_spi *bs, unsigned reg,
+				 u32 val)
+{
+	writel(val, bs->regs + reg);
+}
+
+static inline void bcm2835aux_rd_fifo(struct bcm2835aux_spi *bs)
+{
+	u32 data;
+	int count = min(bs->rx_len, 3);
+
+	data = bcm2835aux_rd(bs, BCM2835_AUX_SPI_IO);
+	if (bs->rx_buf) {
+		switch (count) {
+		case 4:
+			*bs->rx_buf++ = (data >> 24) & 0xff;
+			/* fallthrough */
+		case 3:
+			*bs->rx_buf++ = (data >> 16) & 0xff;
+			/* fallthrough */
+		case 2:
+			*bs->rx_buf++ = (data >> 8) & 0xff;
+			/* fallthrough */
+		case 1:
+			*bs->rx_buf++ = (data >> 0) & 0xff;
+			/* fallthrough - no default */
+		}
+	}
+	bs->rx_len -= count;
+	bs->pending -= count;
+}
+
+static inline void bcm2835aux_wr_fifo(struct bcm2835aux_spi *bs)
+{
+	u32 data;
+	u8 byte;
+	int count;
+	int i;
+
+	/* gather up to 3 bytes to write to the FIFO */
+	count = min(bs->tx_len, 3);
+	data = 0;
+	for (i = 0; i < count; i++) {
+		byte = bs->tx_buf ? *bs->tx_buf++ : 0;
+		data |= byte << (8 * (2 - i));
+	}
+
+	/* and set the variable bit-length */
+	data |= (count * 8) << 24;
+
+	/* and decrement length */
+	bs->tx_len -= count;
+	bs->pending += count;
+
+	/* write to the correct TX-register */
+	if (bs->tx_len)
+		bcm2835aux_wr(bs, BCM2835_AUX_SPI_TXHOLD, data);
+	else
+		bcm2835aux_wr(bs, BCM2835_AUX_SPI_IO, data);
+}
+
+static void bcm2835aux_spi_reset_hw(struct bcm2835aux_spi *bs)
+{
+	/* disable spi clearing fifo and interrupts */
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL1, 0);
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL0,
+		      BCM2835_AUX_SPI_CNTL0_CLEARFIFO);
+}
+
+static irqreturn_t bcm2835aux_spi_interrupt(int irq, void *dev_id)
+{
+	struct spi_master *master = dev_id;
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+	irqreturn_t ret = IRQ_NONE;
+
+	/* check if we have data to read */
+	while (bs->rx_len &&
+	       (!(bcm2835aux_rd(bs, BCM2835_AUX_SPI_STAT) &
+		  BCM2835_AUX_SPI_STAT_RX_EMPTY))) {
+		bcm2835aux_rd_fifo(bs);
+		ret = IRQ_HANDLED;
+	}
+
+	/* check if we have data to write */
+	while (bs->tx_len &&
+	       (bs->pending < 12) &&
+	       (!(bcm2835aux_rd(bs, BCM2835_AUX_SPI_STAT) &
+		  BCM2835_AUX_SPI_STAT_TX_FULL))) {
+		bcm2835aux_wr_fifo(bs);
+		ret = IRQ_HANDLED;
+	}
+
+	/* and check if we have reached "done" */
+	while (bs->rx_len &&
+	       (!(bcm2835aux_rd(bs, BCM2835_AUX_SPI_STAT) &
+		  BCM2835_AUX_SPI_STAT_BUSY))) {
+		bcm2835aux_rd_fifo(bs);
+		ret = IRQ_HANDLED;
+	}
+
+	/* and if rx_len is 0 then wake up completion and disable spi */
+	if (!bs->rx_len) {
+		bcm2835aux_spi_reset_hw(bs);
+		complete(&master->xfer_completion);
+	}
+
+	/* and return */
+	return ret;
+}
+
+static int __bcm2835aux_spi_transfer_one_irq(struct spi_master *master,
+					     struct spi_device *spi,
+					     struct spi_transfer *tfr)
+{
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+
+	/* enable interrupts */
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL1, bs->cntl[1] |
+		BCM2835_AUX_SPI_CNTL1_TXEMPTY |
+		BCM2835_AUX_SPI_CNTL1_IDLE);
+
+	/* and wait for finish... */
+	return 1;
+}
+
+static int bcm2835aux_spi_transfer_one_irq(struct spi_master *master,
+					   struct spi_device *spi,
+					   struct spi_transfer *tfr)
+{
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+
+	/* fill in registers and fifos before enabling interrupts */
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL1, bs->cntl[1]);
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL0, bs->cntl[0]);
+
+	/* fill in tx fifo with data before enabling interrupts */
+	while ((bs->tx_len) &&
+	       (bs->pending < 12) &&
+	       (!(bcm2835aux_rd(bs, BCM2835_AUX_SPI_STAT) &
+		  BCM2835_AUX_SPI_STAT_TX_FULL))) {
+		bcm2835aux_wr_fifo(bs);
+	}
+
+	/* now run the interrupt mode */
+	return __bcm2835aux_spi_transfer_one_irq(master, spi, tfr);
+}
+
+static int bcm2835aux_spi_transfer_one_poll(struct spi_master *master,
+					    struct spi_device *spi,
+					struct spi_transfer *tfr)
+{
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+	unsigned long timeout;
+	u32 stat;
+
+	/* configure spi */
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL1, bs->cntl[1]);
+	bcm2835aux_wr(bs, BCM2835_AUX_SPI_CNTL0, bs->cntl[0]);
+
+	/* set the timeout */
+	timeout = jiffies + BCM2835_AUX_SPI_POLLING_JIFFIES;
+
+	/* loop until finished the transfer */
+	while (bs->rx_len) {
+		/* read status */
+		stat = bcm2835aux_rd(bs, BCM2835_AUX_SPI_STAT);
+
+		/* fill in tx fifo with remaining data */
+		if ((bs->tx_len) && (!(stat & BCM2835_AUX_SPI_STAT_TX_FULL))) {
+			bcm2835aux_wr_fifo(bs);
+			continue;
+		}
+
+		/* read data from fifo for both cases */
+		if (!(stat & BCM2835_AUX_SPI_STAT_RX_EMPTY)) {
+			bcm2835aux_rd_fifo(bs);
+			continue;
+		}
+		if (!(stat & BCM2835_AUX_SPI_STAT_BUSY)) {
+			bcm2835aux_rd_fifo(bs);
+			continue;
+		}
+
+		/* there is still data pending to read check the timeout */
+		if (bs->rx_len && time_after(jiffies, timeout)) {
+			dev_dbg_ratelimited(&spi->dev,
+					    "timeout period reached: jiffies: %lu remaining tx/rx: %d/%d - falling back to interrupt mode\n",
+					    jiffies - timeout,
+					    bs->tx_len, bs->rx_len);
+			/* forward to interrupt handler */
+			return __bcm2835aux_spi_transfer_one_irq(master,
+							       spi, tfr);
+		}
+	}
+
+	/* Transfer complete - reset SPI HW */
+	bcm2835aux_spi_reset_hw(bs);
+
+	/* and return without waiting for completion */
+	return 0;
+}
+
+static int bcm2835aux_spi_transfer_one(struct spi_master *master,
+				       struct spi_device *spi,
+				       struct spi_transfer *tfr)
+{
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+	unsigned long spi_hz, clk_hz, speed;
+	unsigned long spi_used_hz;
+	unsigned long long xfer_time_us;
+
+	/* calculate the registers to handle
+	 *
+	 * note that we use the variable data mode, which
+	 * is not optimal for longer transfers as we waste registers
+	 * resulting (potentially) in more interrupts when transferring
+	 * more than 12 bytes
+	 */
+	bs->cntl[0] = BCM2835_AUX_SPI_CNTL0_ENABLE |
+		      BCM2835_AUX_SPI_CNTL0_VAR_WIDTH |
+		      BCM2835_AUX_SPI_CNTL0_MSBF_OUT;
+	bs->cntl[1] = BCM2835_AUX_SPI_CNTL1_MSBF_IN;
+
+	/* set clock */
+	spi_hz = tfr->speed_hz;
+	clk_hz = clk_get_rate(bs->clk);
+
+	if (spi_hz >= clk_hz / 2) {
+		speed = 0;
+	} else if (spi_hz) {
+		speed = DIV_ROUND_UP(clk_hz, 2 * spi_hz) - 1;
+		if (speed >  BCM2835_AUX_SPI_CNTL0_SPEED_MAX)
+			speed = BCM2835_AUX_SPI_CNTL0_SPEED_MAX;
+	} else { /* the slowest we can go */
+		speed = BCM2835_AUX_SPI_CNTL0_SPEED_MAX;
+	}
+	bs->cntl[0] |= speed << BCM2835_AUX_SPI_CNTL0_SPEED_SHIFT;
+
+	spi_used_hz = clk_hz / (2 * (speed + 1));
+
+	/* handle all the modes */
+	if (spi->mode & SPI_CPOL)
+		bs->cntl[0] |= BCM2835_AUX_SPI_CNTL0_CPOL;
+	if (spi->mode & SPI_CPHA)
+		bs->cntl[0] |= BCM2835_AUX_SPI_CNTL0_CPHA_OUT |
+			       BCM2835_AUX_SPI_CNTL0_CPHA_IN;
+
+	/* set transmit buffers and length */
+	bs->tx_buf = tfr->tx_buf;
+	bs->rx_buf = tfr->rx_buf;
+	bs->tx_len = tfr->len;
+	bs->rx_len = tfr->len;
+	bs->pending = 0;
+
+	/* calculate the estimated time in us the transfer runs
+	 * note that there are are 2 idle clocks after each
+	 * chunk getting transferred - in our case the chunk size
+	 * is 3 bytes, so we approximate this by 9 bits/byte
+	 */
+	xfer_time_us = tfr->len * 9 * 1000000;
+	do_div(xfer_time_us, spi_used_hz);
+
+	/* run in polling mode for short transfers */
+	if (xfer_time_us < BCM2835_AUX_SPI_POLLING_LIMIT_US)
+		return bcm2835aux_spi_transfer_one_poll(master, spi, tfr);
+
+	/* run in interrupt mode for all others */
+	return bcm2835aux_spi_transfer_one_irq(master, spi, tfr);
+}
+
+static void bcm2835aux_spi_handle_err(struct spi_master *master,
+				      struct spi_message *msg)
+{
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+
+	bcm2835aux_spi_reset_hw(bs);
+}
+
+static int bcm2835aux_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct bcm2835aux_spi *bs;
+	struct resource *res;
+	unsigned long clk_hz;
+	int err;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*bs));
+	if (!master) {
+		dev_err(&pdev->dev, "spi_alloc_master() failed\n");
+		return -ENOMEM;
+	}
+
+	platform_set_drvdata(pdev, master);
+	master->mode_bits = BCM2835_AUX_SPI_MODE_BITS;
+	master->bits_per_word_mask = SPI_BPW_MASK(8);
+	master->num_chipselect = -1;
+	master->transfer_one = bcm2835aux_spi_transfer_one;
+	master->handle_err = bcm2835aux_spi_handle_err;
+	master->dev.of_node = pdev->dev.of_node;
+
+	bs = spi_master_get_devdata(master);
+
+	/* the main area */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	bs->regs = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(bs->regs)) {
+		err = PTR_ERR(bs->regs);
+		goto out_master_put;
+	}
+
+	bs->clk = devm_clk_get(&pdev->dev, NULL);
+	if ((!bs->clk) || (IS_ERR(bs->clk))) {
+		err = PTR_ERR(bs->clk);
+		dev_err(&pdev->dev, "could not get clk: %d\n", err);
+		goto out_master_put;
+	}
+
+	bs->irq = platform_get_irq(pdev, 0);
+	if (bs->irq <= 0) {
+		dev_err(&pdev->dev, "could not get IRQ: %d\n", bs->irq);
+		err = bs->irq ? bs->irq : -ENODEV;
+		goto out_master_put;
+	}
+
+	/* this also enables the HW block */
+	err = clk_prepare_enable(bs->clk);
+	if (err) {
+		dev_err(&pdev->dev, "could not prepare clock: %d\n", err);
+		goto out_master_put;
+	}
+
+	/* just checking if the clock returns a sane value */
+	clk_hz = clk_get_rate(bs->clk);
+	if (!clk_hz) {
+		dev_err(&pdev->dev, "clock returns 0 Hz\n");
+		err = -ENODEV;
+		goto out_clk_disable;
+	}
+
+	/* reset SPI-HW block */
+	bcm2835aux_spi_reset_hw(bs);
+
+	err = devm_request_irq(&pdev->dev, bs->irq,
+			       bcm2835aux_spi_interrupt,
+			       IRQF_SHARED,
+			       dev_name(&pdev->dev), master);
+	if (err) {
+		dev_err(&pdev->dev, "could not request IRQ: %d\n", err);
+		goto out_clk_disable;
+	}
+
+	err = devm_spi_register_master(&pdev->dev, master);
+	if (err) {
+		dev_err(&pdev->dev, "could not register SPI master: %d\n", err);
+		goto out_clk_disable;
+	}
+
+	return 0;
+
+out_clk_disable:
+	clk_disable_unprepare(bs->clk);
+out_master_put:
+	spi_master_put(master);
+	return err;
+}
+
+static int bcm2835aux_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = platform_get_drvdata(pdev);
+	struct bcm2835aux_spi *bs = spi_master_get_devdata(master);
+
+	bcm2835aux_spi_reset_hw(bs);
+
+	/* disable the HW block by releasing the clock */
+	clk_disable_unprepare(bs->clk);
+
+	return 0;
+}
+
+static const struct of_device_id bcm2835aux_spi_match[] = {
+	{ .compatible = "brcm,bcm2835-aux-spi", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, bcm2835aux_spi_match);
+
+static struct platform_driver bcm2835aux_spi_driver = {
+	.driver		= {
+		.name		= "spi-bcm2835aux",
+		.of_match_table	= bcm2835aux_spi_match,
+	},
+	.probe		= bcm2835aux_spi_probe,
+	.remove		= bcm2835aux_spi_remove,
+};
+module_platform_driver(bcm2835aux_spi_driver);
+
+MODULE_DESCRIPTION("SPI controller driver for Broadcom BCM2835 aux");
+MODULE_AUTHOR("Martin Sperl <kernel@martin.sperl.org>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/spi-bcm53xx.c b/drivers/spi/spi-bcm53xx.c
index 1520554978a3..cc3f938f0a6b 100644
--- a/drivers/spi/spi-bcm53xx.c
+++ b/drivers/spi/spi-bcm53xx.c
@@ -247,28 +247,19 @@ static int bcm53xxspi_bcma_probe(struct bcma_device *core)
 	if (err) {
 		spi_master_put(master);
 		bcma_set_drvdata(core, NULL);
-		goto out;
+		return err;
 	}
 
 	/* 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);
+	return 0;
 }
 
 static struct bcma_driver bcm53xxspi_bcma_driver = {
 	.name		= KBUILD_MODNAME,
 	.id_table	= bcm53xxspi_bcma_tbl,
 	.probe		= bcm53xxspi_bcma_probe,
-	.remove		= bcm53xxspi_bcma_remove,
 };
 
 /**************************************************
diff --git a/drivers/spi/spi-bcm63xx.c b/drivers/spi/spi-bcm63xx.c
index e73e2b052c9c..06858e04ec59 100644
--- a/drivers/spi/spi-bcm63xx.c
+++ b/drivers/spi/spi-bcm63xx.c
@@ -27,10 +27,117 @@
 #include <linux/err.h>
 #include <linux/pm_runtime.h>
 
-#include <bcm63xx_dev_spi.h>
+/* BCM 6338/6348 SPI core */
+#define SPI_6348_RSET_SIZE		64
+#define SPI_6348_CMD			0x00	/* 16-bits register */
+#define SPI_6348_INT_STATUS		0x02
+#define SPI_6348_INT_MASK_ST		0x03
+#define SPI_6348_INT_MASK		0x04
+#define SPI_6348_ST			0x05
+#define SPI_6348_CLK_CFG		0x06
+#define SPI_6348_FILL_BYTE		0x07
+#define SPI_6348_MSG_TAIL		0x09
+#define SPI_6348_RX_TAIL		0x0b
+#define SPI_6348_MSG_CTL		0x40	/* 8-bits register */
+#define SPI_6348_MSG_CTL_WIDTH		8
+#define SPI_6348_MSG_DATA		0x41
+#define SPI_6348_MSG_DATA_SIZE		0x3f
+#define SPI_6348_RX_DATA		0x80
+#define SPI_6348_RX_DATA_SIZE		0x3f
+
+/* BCM 3368/6358/6262/6368 SPI core */
+#define SPI_6358_RSET_SIZE		1804
+#define SPI_6358_MSG_CTL		0x00	/* 16-bits register */
+#define SPI_6358_MSG_CTL_WIDTH		16
+#define SPI_6358_MSG_DATA		0x02
+#define SPI_6358_MSG_DATA_SIZE		0x21e
+#define SPI_6358_RX_DATA		0x400
+#define SPI_6358_RX_DATA_SIZE		0x220
+#define SPI_6358_CMD			0x700	/* 16-bits register */
+#define SPI_6358_INT_STATUS		0x702
+#define SPI_6358_INT_MASK_ST		0x703
+#define SPI_6358_INT_MASK		0x704
+#define SPI_6358_ST			0x705
+#define SPI_6358_CLK_CFG		0x706
+#define SPI_6358_FILL_BYTE		0x707
+#define SPI_6358_MSG_TAIL		0x709
+#define SPI_6358_RX_TAIL		0x70B
+
+/* Shared SPI definitions */
+
+/* Message configuration */
+#define SPI_FD_RW			0x00
+#define SPI_HD_W			0x01
+#define SPI_HD_R			0x02
+#define SPI_BYTE_CNT_SHIFT		0
+#define SPI_6348_MSG_TYPE_SHIFT		6
+#define SPI_6358_MSG_TYPE_SHIFT		14
+
+/* Command */
+#define SPI_CMD_NOOP			0x00
+#define SPI_CMD_SOFT_RESET		0x01
+#define SPI_CMD_HARD_RESET		0x02
+#define SPI_CMD_START_IMMEDIATE		0x03
+#define SPI_CMD_COMMAND_SHIFT		0
+#define SPI_CMD_COMMAND_MASK		0x000f
+#define SPI_CMD_DEVICE_ID_SHIFT		4
+#define SPI_CMD_PREPEND_BYTE_CNT_SHIFT	8
+#define SPI_CMD_ONE_BYTE_SHIFT		11
+#define SPI_CMD_ONE_WIRE_SHIFT		12
+#define SPI_DEV_ID_0			0
+#define SPI_DEV_ID_1			1
+#define SPI_DEV_ID_2			2
+#define SPI_DEV_ID_3			3
+
+/* Interrupt mask */
+#define SPI_INTR_CMD_DONE		0x01
+#define SPI_INTR_RX_OVERFLOW		0x02
+#define SPI_INTR_TX_UNDERFLOW		0x04
+#define SPI_INTR_TX_OVERFLOW		0x08
+#define SPI_INTR_RX_UNDERFLOW		0x10
+#define SPI_INTR_CLEAR_ALL		0x1f
+
+/* Status */
+#define SPI_RX_EMPTY			0x02
+#define SPI_CMD_BUSY			0x04
+#define SPI_SERIAL_BUSY			0x08
+
+/* Clock configuration */
+#define SPI_CLK_20MHZ			0x00
+#define SPI_CLK_0_391MHZ		0x01
+#define SPI_CLK_0_781MHZ		0x02	/* default */
+#define SPI_CLK_1_563MHZ		0x03
+#define SPI_CLK_3_125MHZ		0x04
+#define SPI_CLK_6_250MHZ		0x05
+#define SPI_CLK_12_50MHZ		0x06
+#define SPI_CLK_MASK			0x07
+#define SPI_SSOFFTIME_MASK		0x38
+#define SPI_SSOFFTIME_SHIFT		3
+#define SPI_BYTE_SWAP			0x80
+
+enum bcm63xx_regs_spi {
+	SPI_CMD,
+	SPI_INT_STATUS,
+	SPI_INT_MASK_ST,
+	SPI_INT_MASK,
+	SPI_ST,
+	SPI_CLK_CFG,
+	SPI_FILL_BYTE,
+	SPI_MSG_TAIL,
+	SPI_RX_TAIL,
+	SPI_MSG_CTL,
+	SPI_MSG_DATA,
+	SPI_RX_DATA,
+	SPI_MSG_TYPE_SHIFT,
+	SPI_MSG_CTL_WIDTH,
+	SPI_MSG_DATA_SIZE,
+};
 
 #define BCM63XX_SPI_MAX_PREPEND		15
 
+#define BCM63XX_SPI_MAX_CS		8
+#define BCM63XX_SPI_BUS_NUM		0
+
 struct bcm63xx_spi {
 	struct completion	done;
 
@@ -38,6 +145,7 @@ struct bcm63xx_spi {
 	int			irq;
 
 	/* Platform data */
+	const unsigned long	*reg_offsets;
 	unsigned		fifo_size;
 	unsigned int		msg_type_shift;
 	unsigned int		msg_ctl_width;
@@ -51,27 +159,35 @@ struct bcm63xx_spi {
 };
 
 static inline u8 bcm_spi_readb(struct bcm63xx_spi *bs,
-				unsigned int offset)
+			       unsigned int offset)
 {
-	return bcm_readb(bs->regs + bcm63xx_spireg(offset));
+	return readb(bs->regs + bs->reg_offsets[offset]);
 }
 
 static inline u16 bcm_spi_readw(struct bcm63xx_spi *bs,
 				unsigned int offset)
 {
-	return bcm_readw(bs->regs + bcm63xx_spireg(offset));
+#ifdef CONFIG_CPU_BIG_ENDIAN
+	return ioread16be(bs->regs + bs->reg_offsets[offset]);
+#else
+	return readw(bs->regs + bs->reg_offsets[offset]);
+#endif
 }
 
 static inline void bcm_spi_writeb(struct bcm63xx_spi *bs,
 				  u8 value, unsigned int offset)
 {
-	bcm_writeb(value, bs->regs + bcm63xx_spireg(offset));
+	writeb(value, bs->regs + bs->reg_offsets[offset]);
 }
 
 static inline void bcm_spi_writew(struct bcm63xx_spi *bs,
 				  u16 value, unsigned int offset)
 {
-	bcm_writew(value, bs->regs + bcm63xx_spireg(offset));
+#ifdef CONFIG_CPU_BIG_ENDIAN
+	iowrite16be(value, bs->regs + bs->reg_offsets[offset]);
+#else
+	writew(value, bs->regs + bs->reg_offsets[offset]);
+#endif
 }
 
 static const unsigned bcm63xx_spi_freq_table[SPI_CLK_MASK][2] = {
@@ -122,7 +238,6 @@ static int bcm63xx_txrx_bufs(struct spi_device *spi, struct spi_transfer *first,
 	struct bcm63xx_spi *bs = spi_master_get_devdata(spi->master);
 	u16 msg_ctl;
 	u16 cmd;
-	u8 rx_tail;
 	unsigned int i, timeout = 0, prepend_len = 0, len = 0;
 	struct spi_transfer *t = first;
 	bool do_rx = false;
@@ -314,18 +429,71 @@ static irqreturn_t bcm63xx_spi_interrupt(int irq, void *dev_id)
 	return IRQ_HANDLED;
 }
 
+static const unsigned long bcm6348_spi_reg_offsets[] = {
+	[SPI_CMD]		= SPI_6348_CMD,
+	[SPI_INT_STATUS]	= SPI_6348_INT_STATUS,
+	[SPI_INT_MASK_ST]	= SPI_6348_INT_MASK_ST,
+	[SPI_INT_MASK]		= SPI_6348_INT_MASK,
+	[SPI_ST]		= SPI_6348_ST,
+	[SPI_CLK_CFG]		= SPI_6348_CLK_CFG,
+	[SPI_FILL_BYTE]		= SPI_6348_FILL_BYTE,
+	[SPI_MSG_TAIL]		= SPI_6348_MSG_TAIL,
+	[SPI_RX_TAIL]		= SPI_6348_RX_TAIL,
+	[SPI_MSG_CTL]		= SPI_6348_MSG_CTL,
+	[SPI_MSG_DATA]		= SPI_6348_MSG_DATA,
+	[SPI_RX_DATA]		= SPI_6348_RX_DATA,
+	[SPI_MSG_TYPE_SHIFT]	= SPI_6348_MSG_TYPE_SHIFT,
+	[SPI_MSG_CTL_WIDTH]	= SPI_6348_MSG_CTL_WIDTH,
+	[SPI_MSG_DATA_SIZE]	= SPI_6348_MSG_DATA_SIZE,
+};
+
+static const unsigned long bcm6358_spi_reg_offsets[] = {
+	[SPI_CMD]		= SPI_6358_CMD,
+	[SPI_INT_STATUS]	= SPI_6358_INT_STATUS,
+	[SPI_INT_MASK_ST]	= SPI_6358_INT_MASK_ST,
+	[SPI_INT_MASK]		= SPI_6358_INT_MASK,
+	[SPI_ST]		= SPI_6358_ST,
+	[SPI_CLK_CFG]		= SPI_6358_CLK_CFG,
+	[SPI_FILL_BYTE]		= SPI_6358_FILL_BYTE,
+	[SPI_MSG_TAIL]		= SPI_6358_MSG_TAIL,
+	[SPI_RX_TAIL]		= SPI_6358_RX_TAIL,
+	[SPI_MSG_CTL]		= SPI_6358_MSG_CTL,
+	[SPI_MSG_DATA]		= SPI_6358_MSG_DATA,
+	[SPI_RX_DATA]		= SPI_6358_RX_DATA,
+	[SPI_MSG_TYPE_SHIFT]	= SPI_6358_MSG_TYPE_SHIFT,
+	[SPI_MSG_CTL_WIDTH]	= SPI_6358_MSG_CTL_WIDTH,
+	[SPI_MSG_DATA_SIZE]	= SPI_6358_MSG_DATA_SIZE,
+};
+
+static const struct platform_device_id bcm63xx_spi_dev_match[] = {
+	{
+		.name = "bcm6348-spi",
+		.driver_data = (unsigned long)bcm6348_spi_reg_offsets,
+	},
+	{
+		.name = "bcm6358-spi",
+		.driver_data = (unsigned long)bcm6358_spi_reg_offsets,
+	},
+	{
+	},
+};
 
 static int bcm63xx_spi_probe(struct platform_device *pdev)
 {
 	struct resource *r;
+	const unsigned long *bcm63xx_spireg;
 	struct device *dev = &pdev->dev;
-	struct bcm63xx_spi_pdata *pdata = dev_get_platdata(&pdev->dev);
 	int irq;
 	struct spi_master *master;
 	struct clk *clk;
 	struct bcm63xx_spi *bs;
 	int ret;
 
+	if (!pdev->id_entry->driver_data)
+		return -EINVAL;
+
+	bcm63xx_spireg = (const unsigned long *)pdev->id_entry->driver_data;
+
 	irq = platform_get_irq(pdev, 0);
 	if (irq < 0) {
 		dev_err(dev, "no irq\n");
@@ -359,7 +527,8 @@ static int bcm63xx_spi_probe(struct platform_device *pdev)
 
 	bs->irq = irq;
 	bs->clk = clk;
-	bs->fifo_size = pdata->fifo_size;
+	bs->reg_offsets = bcm63xx_spireg;
+	bs->fifo_size = bs->reg_offsets[SPI_MSG_DATA_SIZE];
 
 	ret = devm_request_irq(&pdev->dev, irq, bcm63xx_spi_interrupt, 0,
 							pdev->name, master);
@@ -368,26 +537,16 @@ static int bcm63xx_spi_probe(struct platform_device *pdev)
 		goto out_err;
 	}
 
-	master->bus_num = pdata->bus_num;
-	master->num_chipselect = pdata->num_chipselect;
+	master->bus_num = BCM63XX_SPI_BUS_NUM;
+	master->num_chipselect = BCM63XX_SPI_MAX_CS;
 	master->transfer_one_message = bcm63xx_spi_transfer_one;
 	master->mode_bits = MODEBITS;
 	master->bits_per_word_mask = SPI_BPW_MASK(8);
 	master->auto_runtime_pm = true;
-	bs->msg_type_shift = pdata->msg_type_shift;
-	bs->msg_ctl_width = pdata->msg_ctl_width;
-	bs->tx_io = (u8 *)(bs->regs + bcm63xx_spireg(SPI_MSG_DATA));
-	bs->rx_io = (const u8 *)(bs->regs + bcm63xx_spireg(SPI_RX_DATA));
-
-	switch (bs->msg_ctl_width) {
-	case 8:
-	case 16:
-		break;
-	default:
-		dev_err(dev, "unsupported MSG_CTL width: %d\n",
-			 bs->msg_ctl_width);
-		goto out_err;
-	}
+	bs->msg_type_shift = bs->reg_offsets[SPI_MSG_TYPE_SHIFT];
+	bs->msg_ctl_width = bs->reg_offsets[SPI_MSG_CTL_WIDTH];
+	bs->tx_io = (u8 *)(bs->regs + bs->reg_offsets[SPI_MSG_DATA]);
+	bs->rx_io = (const u8 *)(bs->regs + bs->reg_offsets[SPI_RX_DATA]);
 
 	/* Initialize hardware */
 	ret = clk_prepare_enable(bs->clk);
@@ -467,6 +626,7 @@ static struct platform_driver bcm63xx_spi_driver = {
 		.name	= "bcm63xx-spi",
 		.pm	= &bcm63xx_spi_pm_ops,
 	},
+	.id_table	= bcm63xx_spi_dev_match,
 	.probe		= bcm63xx_spi_probe,
 	.remove		= bcm63xx_spi_remove,
 };
diff --git a/drivers/spi/spi-bfin-sport.c b/drivers/spi/spi-bfin-sport.c
index a78693189f45..6c967555a56a 100644
--- a/drivers/spi/spi-bfin-sport.c
+++ b/drivers/spi/spi-bfin-sport.c
@@ -352,10 +352,7 @@ bfin_sport_spi_pump_transfers(unsigned long data)
 	transfer = drv_data->cur_transfer;
 	chip = drv_data->cur_chip;
 
-	if (transfer->speed_hz)
-		transfer_speed = bfin_sport_hz_to_spi_baud(transfer->speed_hz);
-	else
-		transfer_speed = chip->baud;
+	transfer_speed = bfin_sport_hz_to_spi_baud(transfer->speed_hz);
 	bfin_write(&drv_data->regs->tclkdiv, transfer_speed);
 	SSYNC();
 
diff --git a/drivers/spi/spi-bfin5xx.c b/drivers/spi/spi-bfin5xx.c
index a3d65b4f4944..1e91325bf39c 100644
--- a/drivers/spi/spi-bfin5xx.c
+++ b/drivers/spi/spi-bfin5xx.c
@@ -661,11 +661,7 @@ static void bfin_spi_pump_transfers(unsigned long data)
 	message->state = RUNNING_STATE;
 	dma_config = 0;
 
-	/* Speed setup (surely valid because already checked) */
-	if (transfer->speed_hz)
-		bfin_write(&drv_data->regs->baud, hz_to_spi_baud(transfer->speed_hz));
-	else
-		bfin_write(&drv_data->regs->baud, chip->baud);
+	bfin_write(&drv_data->regs->baud, hz_to_spi_baud(transfer->speed_hz));
 
 	bfin_write(&drv_data->regs->stat, BIT_STAT_CLR);
 	bfin_spi_cs_active(drv_data, chip);
diff --git a/drivers/spi/spi-bitbang.c b/drivers/spi/spi-bitbang.c
index 840a4984d365..3aa9e6e3dac8 100644
--- a/drivers/spi/spi-bitbang.c
+++ b/drivers/spi/spi-bitbang.c
@@ -24,6 +24,8 @@
 #include <linux/spi/spi.h>
 #include <linux/spi/spi_bitbang.h>
 
+#define SPI_BITBANG_CS_DELAY	100
+
 
 /*----------------------------------------------------------------------*/
 
@@ -180,7 +182,6 @@ int spi_bitbang_setup(struct spi_device *spi)
 {
 	struct spi_bitbang_cs	*cs = spi->controller_state;
 	struct spi_bitbang	*bitbang;
-	unsigned long		flags;
 
 	bitbang = spi_master_get_devdata(spi->master);
 
@@ -210,12 +211,12 @@ int spi_bitbang_setup(struct spi_device *spi)
 	 */
 
 	/* deselect chip (low or high) */
-	spin_lock_irqsave(&bitbang->lock, flags);
+	mutex_lock(&bitbang->lock);
 	if (!bitbang->busy) {
 		bitbang->chipselect(spi, BITBANG_CS_INACTIVE);
 		ndelay(cs->nsecs);
 	}
-	spin_unlock_irqrestore(&bitbang->lock, flags);
+	mutex_unlock(&bitbang->lock);
 
 	return 0;
 }
@@ -255,122 +256,39 @@ static int spi_bitbang_bufs(struct spi_device *spi, struct spi_transfer *t)
 static int spi_bitbang_prepare_hardware(struct spi_master *spi)
 {
 	struct spi_bitbang	*bitbang;
-	unsigned long		flags;
 
 	bitbang = spi_master_get_devdata(spi);
 
-	spin_lock_irqsave(&bitbang->lock, flags);
+	mutex_lock(&bitbang->lock);
 	bitbang->busy = 1;
-	spin_unlock_irqrestore(&bitbang->lock, flags);
+	mutex_unlock(&bitbang->lock);
 
 	return 0;
 }
 
 static int spi_bitbang_transfer_one(struct spi_master *master,
-				    struct spi_message *m)
+				    struct spi_device *spi,
+				    struct spi_transfer *transfer)
 {
-	struct spi_bitbang	*bitbang;
-	unsigned		nsecs;
-	struct spi_transfer	*t = NULL;
-	unsigned		cs_change;
-	int			status;
-	int			do_setup = -1;
-	struct spi_device	*spi = m->spi;
-
-	bitbang = spi_master_get_devdata(master);
-
-	/* FIXME this is made-up ... the correct value is known to
-	 * word-at-a-time bitbang code, and presumably chipselect()
-	 * should enforce these requirements too?
-	 */
-	nsecs = 100;
-
-	cs_change = 1;
-	status = 0;
-
-	list_for_each_entry(t, &m->transfers, transfer_list) {
-
-		/* override speed or wordsize? */
-		if (t->speed_hz || t->bits_per_word)
-			do_setup = 1;
-
-		/* init (-1) or override (1) transfer params */
-		if (do_setup != 0) {
-			if (bitbang->setup_transfer) {
-				status = bitbang->setup_transfer(spi, t);
-				if (status < 0)
-					break;
-			}
-			if (do_setup == -1)
-				do_setup = 0;
-		}
-
-		/* set up default clock polarity, and activate chip;
-		 * this implicitly updates clock and spi modes as
-		 * previously recorded for this device via setup().
-		 * (and also deselects any other chip that might be
-		 * selected ...)
-		 */
-		if (cs_change) {
-			bitbang->chipselect(spi, BITBANG_CS_ACTIVE);
-			ndelay(nsecs);
-		}
-		cs_change = t->cs_change;
-		if (!t->tx_buf && !t->rx_buf && t->len) {
-			status = -EINVAL;
-			break;
-		}
-
-		/* transfer data.  the lower level code handles any
-		 * new dma mappings it needs. our caller always gave
-		 * us dma-safe buffers.
-		 */
-		if (t->len) {
-			/* REVISIT dma API still needs a designated
-			 * DMA_ADDR_INVALID; ~0 might be better.
-			 */
-			if (!m->is_dma_mapped)
-				t->rx_dma = t->tx_dma = 0;
-			status = bitbang->txrx_bufs(spi, t);
-		}
-		if (status > 0)
-			m->actual_length += status;
-		if (status != t->len) {
-			/* always report some kind of error */
-			if (status >= 0)
-				status = -EREMOTEIO;
-			break;
-		}
-		status = 0;
+	struct spi_bitbang *bitbang = spi_master_get_devdata(master);
+	int status = 0;
 
-		/* protocol tweaks before next transfer */
-		if (t->delay_usecs)
-			udelay(t->delay_usecs);
-
-		if (cs_change &&
-		    !list_is_last(&t->transfer_list, &m->transfers)) {
-			/* sometimes a short mid-message deselect of the chip
-			 * may be needed to terminate a mode or command
-			 */
-			ndelay(nsecs);
-			bitbang->chipselect(spi, BITBANG_CS_INACTIVE);
-			ndelay(nsecs);
-		}
+	if (bitbang->setup_transfer) {
+		status = bitbang->setup_transfer(spi, transfer);
+		if (status < 0)
+			goto out;
 	}
 
-	m->status = status;
+	if (transfer->len)
+		status = bitbang->txrx_bufs(spi, transfer);
 
-	/* normally deactivate chipselect ... unless no error and
-	 * cs_change has hinted that the next message will probably
-	 * be for this chip too.
-	 */
-	if (!(status == 0 && cs_change)) {
-		ndelay(nsecs);
-		bitbang->chipselect(spi, BITBANG_CS_INACTIVE);
-		ndelay(nsecs);
-	}
+	if (status == transfer->len)
+		status = 0;
+	else if (status >= 0)
+		status = -EREMOTEIO;
 
-	spi_finalize_current_message(master);
+out:
+	spi_finalize_current_transfer(master);
 
 	return status;
 }
@@ -378,17 +296,32 @@ static int spi_bitbang_transfer_one(struct spi_master *master,
 static int spi_bitbang_unprepare_hardware(struct spi_master *spi)
 {
 	struct spi_bitbang	*bitbang;
-	unsigned long		flags;
 
 	bitbang = spi_master_get_devdata(spi);
 
-	spin_lock_irqsave(&bitbang->lock, flags);
+	mutex_lock(&bitbang->lock);
 	bitbang->busy = 0;
-	spin_unlock_irqrestore(&bitbang->lock, flags);
+	mutex_unlock(&bitbang->lock);
 
 	return 0;
 }
 
+static void spi_bitbang_set_cs(struct spi_device *spi, bool enable)
+{
+	struct spi_bitbang *bitbang = spi_master_get_devdata(spi->master);
+
+	/* SPI core provides CS high / low, but bitbang driver
+	 * expects CS active
+	 * spi device driver takes care of handling SPI_CS_HIGH
+	 */
+	enable = (!!(spi->mode & SPI_CS_HIGH) == enable);
+
+	ndelay(SPI_BITBANG_CS_DELAY);
+	bitbang->chipselect(spi, enable ? BITBANG_CS_ACTIVE :
+			    BITBANG_CS_INACTIVE);
+	ndelay(SPI_BITBANG_CS_DELAY);
+}
+
 /*----------------------------------------------------------------------*/
 
 /**
@@ -427,7 +360,7 @@ int spi_bitbang_start(struct spi_bitbang *bitbang)
 	if (!master || !bitbang->chipselect)
 		return -EINVAL;
 
-	spin_lock_init(&bitbang->lock);
+	mutex_init(&bitbang->lock);
 
 	if (!master->mode_bits)
 		master->mode_bits = SPI_CPOL | SPI_CPHA | bitbang->flags;
@@ -437,7 +370,8 @@ int spi_bitbang_start(struct spi_bitbang *bitbang)
 
 	master->prepare_transfer_hardware = spi_bitbang_prepare_hardware;
 	master->unprepare_transfer_hardware = spi_bitbang_unprepare_hardware;
-	master->transfer_one_message = spi_bitbang_transfer_one;
+	master->transfer_one = spi_bitbang_transfer_one;
+	master->set_cs = spi_bitbang_set_cs;
 
 	if (!bitbang->txrx_bufs) {
 		bitbang->use_dma = 0;
diff --git a/drivers/spi/spi-coldfire-qspi.c b/drivers/spi/spi-coldfire-qspi.c
index 688956ff5095..23f6fffd75e1 100644
--- a/drivers/spi/spi-coldfire-qspi.c
+++ b/drivers/spi/spi-coldfire-qspi.c
@@ -420,19 +420,20 @@ static int mcfqspi_probe(struct platform_device *pdev)
 	master->auto_runtime_pm = true;
 
 	platform_set_drvdata(pdev, master);
+	pm_runtime_enable(&pdev->dev);
 
 	status = devm_spi_register_master(&pdev->dev, master);
 	if (status) {
 		dev_dbg(&pdev->dev, "spi_register_master failed\n");
 		goto fail2;
 	}
-	pm_runtime_enable(&pdev->dev);
 
 	dev_info(&pdev->dev, "Coldfire QSPI bus driver\n");
 
 	return 0;
 
 fail2:
+	pm_runtime_disable(&pdev->dev);
 	mcfqspi_cs_teardown(mcfqspi);
 fail1:
 	clk_disable(mcfqspi->clk);
diff --git a/drivers/spi/spi-davinci.c b/drivers/spi/spi-davinci.c
index a85d863d4a44..7d3af3eacf57 100644
--- a/drivers/spi/spi-davinci.c
+++ b/drivers/spi/spi-davinci.c
@@ -215,18 +215,10 @@ static void davinci_spi_chipselect(struct spi_device *spi, int value)
 	struct davinci_spi_config *spicfg = spi->controller_data;
 	u8 chip_sel = spi->chip_select;
 	u16 spidat1 = CS_DEFAULT;
-	bool gpio_chipsel = false;
-	int gpio;
 
 	dspi = spi_master_get_devdata(spi->master);
 	pdata = &dspi->pdata;
 
-	if (spi->cs_gpio >= 0) {
-		/* SPI core parse and update master->cs_gpio */
-		gpio_chipsel = true;
-		gpio = spi->cs_gpio;
-	}
-
 	/* program delay transfers if tx_delay is non zero */
 	if (spicfg->wdelay)
 		spidat1 |= SPIDAT1_WDEL;
@@ -235,11 +227,12 @@ static void davinci_spi_chipselect(struct spi_device *spi, int value)
 	 * Board specific chip select logic decides the polarity and cs
 	 * line for the controller
 	 */
-	if (gpio_chipsel) {
+	if (spi->cs_gpio >= 0) {
 		if (value == BITBANG_CS_ACTIVE)
-			gpio_set_value(gpio, spi->mode & SPI_CS_HIGH);
+			gpio_set_value(spi->cs_gpio, spi->mode & SPI_CS_HIGH);
 		else
-			gpio_set_value(gpio, !(spi->mode & SPI_CS_HIGH));
+			gpio_set_value(spi->cs_gpio,
+				!(spi->mode & SPI_CS_HIGH));
 	} else {
 		if (value == BITBANG_CS_ACTIVE) {
 			spidat1 |= SPIDAT1_CSHOLD_MASK;
diff --git a/drivers/spi/spi-dw-mmio.c b/drivers/spi/spi-dw-mmio.c
index 7edede6e024b..a6d7029a85ac 100644
--- a/drivers/spi/spi-dw-mmio.c
+++ b/drivers/spi/spi-dw-mmio.c
@@ -19,6 +19,7 @@
 #include <linux/of.h>
 #include <linux/of_gpio.h>
 #include <linux/of_platform.h>
+#include <linux/property.h>
 
 #include "spi-dw.h"
 
@@ -74,13 +75,11 @@ static int dw_spi_mmio_probe(struct platform_device *pdev)
 
 	dws->max_freq = clk_get_rate(dwsmmio->clk);
 
-	of_property_read_u32(pdev->dev.of_node, "reg-io-width",
-			     &dws->reg_io_width);
+	device_property_read_u32(&pdev->dev, "reg-io-width", &dws->reg_io_width);
 
 	num_cs = 4;
 
-	if (pdev->dev.of_node)
-		of_property_read_u32(pdev->dev.of_node, "num-cs", &num_cs);
+	device_property_read_u32(&pdev->dev, "num-cs", &num_cs);
 
 	dws->num_cs = num_cs;
 
diff --git a/drivers/spi/spi-dw-pci.c b/drivers/spi/spi-dw-pci.c
index 6d331e0db331..332ccb0539a7 100644
--- a/drivers/spi/spi-dw-pci.c
+++ b/drivers/spi/spi-dw-pci.c
@@ -23,11 +23,6 @@
 
 #define DRIVER_NAME "dw_spi_pci"
 
-struct dw_spi_pci {
-	struct pci_dev	*pdev;
-	struct dw_spi	dws;
-};
-
 struct spi_pci_desc {
 	int	(*setup)(struct dw_spi *);
 	u16	num_cs;
@@ -48,7 +43,6 @@ static struct spi_pci_desc spi_pci_mid_desc_2 = {
 
 static int spi_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
-	struct dw_spi_pci *dwpci;
 	struct dw_spi *dws;
 	struct spi_pci_desc *desc = (struct spi_pci_desc *)ent->driver_data;
 	int pci_bar = 0;
@@ -58,14 +52,10 @@ static int spi_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (ret)
 		return ret;
 
-	dwpci = devm_kzalloc(&pdev->dev, sizeof(struct dw_spi_pci),
-			GFP_KERNEL);
-	if (!dwpci)
+	dws = devm_kzalloc(&pdev->dev, sizeof(*dws), GFP_KERNEL);
+	if (!dws)
 		return -ENOMEM;
 
-	dwpci->pdev = pdev;
-	dws = &dwpci->dws;
-
 	/* Get basic io resource and map it */
 	dws->paddr = pci_resource_start(pdev, pci_bar);
 
@@ -74,7 +64,6 @@ static int spi_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		return ret;
 
 	dws->regs = pcim_iomap_table(pdev)[pci_bar];
-
 	dws->irq = pdev->irq;
 
 	/*
@@ -99,7 +88,7 @@ static int spi_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		return ret;
 
 	/* PCI hook and SPI hook use the same drv data */
-	pci_set_drvdata(pdev, dwpci);
+	pci_set_drvdata(pdev, dws);
 
 	dev_info(&pdev->dev, "found PCI SPI controller(ID: %04x:%04x)\n",
 		pdev->vendor, pdev->device);
@@ -109,26 +98,26 @@ static int spi_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 static void spi_pci_remove(struct pci_dev *pdev)
 {
-	struct dw_spi_pci *dwpci = pci_get_drvdata(pdev);
+	struct dw_spi *dws = pci_get_drvdata(pdev);
 
-	dw_spi_remove_host(&dwpci->dws);
+	dw_spi_remove_host(dws);
 }
 
 #ifdef CONFIG_PM_SLEEP
 static int spi_suspend(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
-	struct dw_spi_pci *dwpci = pci_get_drvdata(pdev);
+	struct dw_spi *dws = pci_get_drvdata(pdev);
 
-	return dw_spi_suspend_host(&dwpci->dws);
+	return dw_spi_suspend_host(dws);
 }
 
 static int spi_resume(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
-	struct dw_spi_pci *dwpci = pci_get_drvdata(pdev);
+	struct dw_spi *dws = pci_get_drvdata(pdev);
 
-	return dw_spi_resume_host(&dwpci->dws);
+	return dw_spi_resume_host(dws);
 }
 #endif
 
diff --git a/drivers/spi/spi-dw.c b/drivers/spi/spi-dw.c
index 4fbfcdc5cb24..882cd6618cd5 100644
--- a/drivers/spi/spi-dw.c
+++ b/drivers/spi/spi-dw.c
@@ -30,19 +30,13 @@
 
 /* Slave spi_dev related */
 struct chip_data {
-	u16 cr0;
 	u8 cs;			/* chip select pin */
-	u8 n_bytes;		/* current is a 1/2/4 byte op */
 	u8 tmode;		/* TR/TO/RO/EEPROM */
 	u8 type;		/* SPI/SSP/MicroWire */
 
 	u8 poll_mode;		/* 1 means use poll mode */
 
-	u32 dma_width;
-	u32 rx_threshold;
-	u32 tx_threshold;
 	u8 enable_dma;
-	u8 bits_per_word;
 	u16 clk_div;		/* baud rate divider */
 	u32 speed_hz;		/* baud rate */
 	void (*cs_control)(u32 command);
@@ -289,14 +283,11 @@ static int dw_spi_transfer_one(struct spi_master *master,
 	struct chip_data *chip = spi_get_ctldata(spi);
 	u8 imask = 0;
 	u16 txlevel = 0;
-	u16 clk_div = 0;
-	u32 speed = 0;
-	u32 cr0 = 0;
+	u16 clk_div;
+	u32 cr0;
 	int ret;
 
 	dws->dma_mapped = 0;
-	dws->n_bytes = chip->n_bytes;
-	dws->dma_width = chip->dma_width;
 
 	dws->tx = (void *)transfer->tx_buf;
 	dws->tx_end = dws->tx + transfer->len;
@@ -306,37 +297,30 @@ static int dw_spi_transfer_one(struct spi_master *master,
 
 	spi_enable_chip(dws, 0);
 
-	cr0 = chip->cr0;
-
 	/* Handle per transfer options for bpw and speed */
-	if (transfer->speed_hz) {
-		speed = chip->speed_hz;
-
-		if ((transfer->speed_hz != speed) || !chip->clk_div) {
-			speed = transfer->speed_hz;
-
-			/* clk_div doesn't support odd number */
-			clk_div = (dws->max_freq / speed + 1) & 0xfffe;
+	if (transfer->speed_hz != chip->speed_hz) {
+		/* clk_div doesn't support odd number */
+		clk_div = (dws->max_freq / transfer->speed_hz + 1) & 0xfffe;
 
-			chip->speed_hz = speed;
-			chip->clk_div = clk_div;
+		chip->speed_hz = transfer->speed_hz;
+		chip->clk_div = clk_div;
 
-			spi_set_clk(dws, chip->clk_div);
-		}
+		spi_set_clk(dws, chip->clk_div);
 	}
-	if (transfer->bits_per_word) {
-		if (transfer->bits_per_word == 8) {
-			dws->n_bytes = 1;
-			dws->dma_width = 1;
-		} else if (transfer->bits_per_word == 16) {
-			dws->n_bytes = 2;
-			dws->dma_width = 2;
-		}
-		cr0 = (transfer->bits_per_word - 1)
-			| (chip->type << SPI_FRF_OFFSET)
-			| (spi->mode << SPI_MODE_OFFSET)
-			| (chip->tmode << SPI_TMOD_OFFSET);
+	if (transfer->bits_per_word == 8) {
+		dws->n_bytes = 1;
+		dws->dma_width = 1;
+	} else if (transfer->bits_per_word == 16) {
+		dws->n_bytes = 2;
+		dws->dma_width = 2;
+	} else {
+		return -EINVAL;
 	}
+	/* Default SPI mode is SCPOL = 0, SCPH = 0 */
+	cr0 = (transfer->bits_per_word - 1)
+		| (chip->type << SPI_FRF_OFFSET)
+		| (spi->mode << SPI_MODE_OFFSET)
+		| (chip->tmode << SPI_TMOD_OFFSET);
 
 	/*
 	 * Adjust transfer mode if necessary. Requires platform dependent
@@ -439,34 +423,9 @@ static int dw_spi_setup(struct spi_device *spi)
 
 		chip->poll_mode = chip_info->poll_mode;
 		chip->type = chip_info->type;
-
-		chip->rx_threshold = 0;
-		chip->tx_threshold = 0;
-	}
-
-	if (spi->bits_per_word == 8) {
-		chip->n_bytes = 1;
-		chip->dma_width = 1;
-	} else if (spi->bits_per_word == 16) {
-		chip->n_bytes = 2;
-		chip->dma_width = 2;
-	}
-	chip->bits_per_word = spi->bits_per_word;
-
-	if (!spi->max_speed_hz) {
-		dev_err(&spi->dev, "No max speed HZ parameter\n");
-		return -EINVAL;
 	}
 
 	chip->tmode = 0; /* Tx & Rx */
-	/* Default SPI mode is SCPOL = 0, SCPH = 0 */
-	chip->cr0 = (chip->bits_per_word - 1)
-			| (chip->type << SPI_FRF_OFFSET)
-			| (spi->mode  << SPI_MODE_OFFSET)
-			| (chip->tmode << SPI_TMOD_OFFSET);
-
-	if (spi->mode & SPI_LOOP)
-		chip->cr0 |= 1 << SPI_SRL_OFFSET;
 
 	if (gpio_is_valid(spi->cs_gpio)) {
 		ret = gpio_direction_output(spi->cs_gpio,
@@ -524,13 +483,12 @@ int dw_spi_add_host(struct device *dev, struct dw_spi *dws)
 	dws->master = master;
 	dws->type = SSI_MOTO_SPI;
 	dws->dma_inited = 0;
-	dws->dma_addr = (dma_addr_t)(dws->paddr + 0x60);
+	dws->dma_addr = (dma_addr_t)(dws->paddr + DW_SPI_DR);
 	snprintf(dws->name, sizeof(dws->name), "dw_spi%d", dws->bus_num);
 
-	ret = devm_request_irq(dev, dws->irq, dw_spi_irq, IRQF_SHARED,
-			dws->name, master);
+	ret = request_irq(dws->irq, dw_spi_irq, IRQF_SHARED, dws->name, master);
 	if (ret < 0) {
-		dev_err(&master->dev, "can not get IRQ\n");
+		dev_err(dev, "can not get IRQ\n");
 		goto err_free_master;
 	}
 
@@ -573,6 +531,7 @@ err_dma_exit:
 	if (dws->dma_ops && dws->dma_ops->dma_exit)
 		dws->dma_ops->dma_exit(dws);
 	spi_enable_chip(dws, 0);
+	free_irq(dws->irq, master);
 err_free_master:
 	spi_master_put(master);
 	return ret;
@@ -581,28 +540,27 @@ EXPORT_SYMBOL_GPL(dw_spi_add_host);
 
 void dw_spi_remove_host(struct dw_spi *dws)
 {
-	if (!dws)
-		return;
 	dw_spi_debugfs_remove(dws);
 
 	if (dws->dma_ops && dws->dma_ops->dma_exit)
 		dws->dma_ops->dma_exit(dws);
-	spi_enable_chip(dws, 0);
-	/* Disable clk */
-	spi_set_clk(dws, 0);
+
+	spi_shutdown_chip(dws);
+
+	free_irq(dws->irq, dws->master);
 }
 EXPORT_SYMBOL_GPL(dw_spi_remove_host);
 
 int dw_spi_suspend_host(struct dw_spi *dws)
 {
-	int ret = 0;
+	int ret;
 
 	ret = spi_master_suspend(dws->master);
 	if (ret)
 		return ret;
-	spi_enable_chip(dws, 0);
-	spi_set_clk(dws, 0);
-	return ret;
+
+	spi_shutdown_chip(dws);
+	return 0;
 }
 EXPORT_SYMBOL_GPL(dw_spi_suspend_host);
 
diff --git a/drivers/spi/spi-dw.h b/drivers/spi/spi-dw.h
index b75ed327d5a2..35589a270468 100644
--- a/drivers/spi/spi-dw.h
+++ b/drivers/spi/spi-dw.h
@@ -225,6 +225,12 @@ static inline void spi_reset_chip(struct dw_spi *dws)
 	spi_enable_chip(dws, 1);
 }
 
+static inline void spi_shutdown_chip(struct dw_spi *dws)
+{
+	spi_enable_chip(dws, 0);
+	spi_set_clk(dws, 0);
+}
+
 /*
  * Each SPI slave device to work with dw_api controller should
  * has such a structure claiming its working mode (poll or PIO/DMA),
diff --git a/drivers/spi/spi-fsl-dspi.c b/drivers/spi/spi-fsl-dspi.c
index 86bcdd68c1fe..59a11437db70 100644
--- a/drivers/spi/spi-fsl-dspi.c
+++ b/drivers/spi/spi-fsl-dspi.c
@@ -409,9 +409,6 @@ static int dspi_transfer_one_message(struct spi_master *master,
 				SPI_MCR_CLR_TXF | SPI_MCR_CLR_RXF);
 		regmap_write(dspi->regmap, SPI_CTAR(dspi->cs),
 				dspi->cur_chip->ctar_val);
-		if (transfer->speed_hz)
-			regmap_write(dspi->regmap, SPI_CTAR(dspi->cs),
-					dspi->cur_chip->ctar_val);
 
 		trans_mode = dspi->devtype_data->trans_mode;
 		switch (trans_mode) {
diff --git a/drivers/spi/spi-imx.c b/drivers/spi/spi-imx.c
index f9deb84e4e55..0e5723ab47f0 100644
--- a/drivers/spi/spi-imx.c
+++ b/drivers/spi/spi-imx.c
@@ -336,13 +336,20 @@ static int __maybe_unused mx51_ecspi_config(struct spi_imx_data *spi_imx,
 
 	if (config->mode & SPI_CPHA)
 		cfg |= MX51_ECSPI_CONFIG_SCLKPHA(config->cs);
+	else
+		cfg &= ~MX51_ECSPI_CONFIG_SCLKPHA(config->cs);
 
 	if (config->mode & SPI_CPOL) {
 		cfg |= MX51_ECSPI_CONFIG_SCLKPOL(config->cs);
 		cfg |= MX51_ECSPI_CONFIG_SCLKCTL(config->cs);
+	} else {
+		cfg &= ~MX51_ECSPI_CONFIG_SCLKPOL(config->cs);
+		cfg &= ~MX51_ECSPI_CONFIG_SCLKCTL(config->cs);
 	}
 	if (config->mode & SPI_CS_HIGH)
 		cfg |= MX51_ECSPI_CONFIG_SSBPOL(config->cs);
+	else
+		cfg &= ~MX51_ECSPI_CONFIG_SSBPOL(config->cs);
 
 	writel(ctrl, spi_imx->base + MX51_ECSPI_CTRL);
 	writel(cfg, spi_imx->base + MX51_ECSPI_CONFIG);
diff --git a/drivers/spi/spi-mpc512x-psc.c b/drivers/spi/spi-mpc512x-psc.c
index 1e75341689a6..c3ec46cd9f91 100644
--- a/drivers/spi/spi-mpc512x-psc.c
+++ b/drivers/spi/spi-mpc512x-psc.c
@@ -302,11 +302,9 @@ static int mpc512x_psc_spi_msg_xfer(struct spi_master *master,
 	cs_change = 1;
 	status = 0;
 	list_for_each_entry(t, &m->transfers, transfer_list) {
-		if (t->bits_per_word || t->speed_hz) {
-			status = mpc512x_psc_spi_transfer_setup(spi, t);
-			if (status < 0)
-				break;
-		}
+		status = mpc512x_psc_spi_transfer_setup(spi, t);
+		if (status < 0)
+			break;
 
 		if (cs_change)
 			mpc512x_psc_spi_activate_cs(spi);
diff --git a/drivers/spi/spi-mt65xx.c b/drivers/spi/spi-mt65xx.c
index ecb6c58238c4..563954a61424 100644
--- a/drivers/spi/spi-mt65xx.c
+++ b/drivers/spi/spi-mt65xx.c
@@ -20,6 +20,7 @@
 #include <linux/ioport.h>
 #include <linux/module.h>
 #include <linux/of.h>
+#include <linux/of_gpio.h>
 #include <linux/platform_device.h>
 #include <linux/platform_data/spi-mt65xx.h>
 #include <linux/pm_runtime.h>
@@ -84,7 +85,8 @@ struct mtk_spi_compatible {
 struct mtk_spi {
 	void __iomem *base;
 	u32 state;
-	u32 pad_sel;
+	int pad_num;
+	u32 *pad_sel;
 	struct clk *parent_clk, *sel_clk, *spi_clk;
 	struct spi_transfer *cur_transfer;
 	u32 xfer_len;
@@ -131,10 +133,28 @@ static void mtk_spi_reset(struct mtk_spi *mdata)
 	writel(reg_val, mdata->base + SPI_CMD_REG);
 }
 
-static void mtk_spi_config(struct mtk_spi *mdata,
-			   struct mtk_chip_config *chip_config)
+static int mtk_spi_prepare_message(struct spi_master *master,
+				   struct spi_message *msg)
 {
+	u16 cpha, cpol;
 	u32 reg_val;
+	struct spi_device *spi = msg->spi;
+	struct mtk_chip_config *chip_config = spi->controller_data;
+	struct mtk_spi *mdata = spi_master_get_devdata(master);
+
+	cpha = spi->mode & SPI_CPHA ? 1 : 0;
+	cpol = spi->mode & SPI_CPOL ? 1 : 0;
+
+	reg_val = readl(mdata->base + SPI_CMD_REG);
+	if (cpha)
+		reg_val |= SPI_CMD_CPHA;
+	else
+		reg_val &= ~SPI_CMD_CPHA;
+	if (cpol)
+		reg_val |= SPI_CMD_CPOL;
+	else
+		reg_val &= ~SPI_CMD_CPOL;
+	writel(reg_val, mdata->base + SPI_CMD_REG);
 
 	reg_val = readl(mdata->base + SPI_CMD_REG);
 
@@ -170,38 +190,8 @@ static void mtk_spi_config(struct mtk_spi *mdata,
 
 	/* pad select */
 	if (mdata->dev_comp->need_pad_sel)
-		writel(mdata->pad_sel, mdata->base + SPI_PAD_SEL_REG);
-}
-
-static int mtk_spi_prepare_message(struct spi_master *master,
-				   struct spi_message *msg)
-{
-	u32 reg_val;
-	u8 cpha, cpol;
-	struct mtk_chip_config *chip_config;
-	struct spi_device *spi = msg->spi;
-	struct mtk_spi *mdata = spi_master_get_devdata(master);
-
-	cpha = spi->mode & SPI_CPHA ? 1 : 0;
-	cpol = spi->mode & SPI_CPOL ? 1 : 0;
-
-	reg_val = readl(mdata->base + SPI_CMD_REG);
-	if (cpha)
-		reg_val |= SPI_CMD_CPHA;
-	else
-		reg_val &= ~SPI_CMD_CPHA;
-	if (cpol)
-		reg_val |= SPI_CMD_CPOL;
-	else
-		reg_val &= ~SPI_CMD_CPOL;
-	writel(reg_val, mdata->base + SPI_CMD_REG);
-
-	chip_config = spi->controller_data;
-	if (!chip_config) {
-		chip_config = (void *)&mtk_default_chip_info;
-		spi->controller_data = chip_config;
-	}
-	mtk_spi_config(mdata, chip_config);
+		writel(mdata->pad_sel[spi->chip_select],
+		       mdata->base + SPI_PAD_SEL_REG);
 
 	return 0;
 }
@@ -413,6 +403,19 @@ static bool mtk_spi_can_dma(struct spi_master *master,
 	return xfer->len > MTK_SPI_MAX_FIFO_SIZE;
 }
 
+static int mtk_spi_setup(struct spi_device *spi)
+{
+	struct mtk_spi *mdata = spi_master_get_devdata(spi->master);
+
+	if (!spi->controller_data)
+		spi->controller_data = (void *)&mtk_default_chip_info;
+
+	if (mdata->dev_comp->need_pad_sel)
+		gpio_direction_output(spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
+
+	return 0;
+}
+
 static irqreturn_t mtk_spi_interrupt(int irq, void *dev_id)
 {
 	u32 cmd, reg_val, cnt;
@@ -484,7 +487,7 @@ static int mtk_spi_probe(struct platform_device *pdev)
 	struct mtk_spi *mdata;
 	const struct of_device_id *of_id;
 	struct resource *res;
-	int irq, ret;
+	int i, irq, ret;
 
 	master = spi_alloc_master(&pdev->dev, sizeof(*mdata));
 	if (!master) {
@@ -500,6 +503,7 @@ static int mtk_spi_probe(struct platform_device *pdev)
 	master->prepare_message = mtk_spi_prepare_message;
 	master->transfer_one = mtk_spi_transfer_one;
 	master->can_dma = mtk_spi_can_dma;
+	master->setup = mtk_spi_setup;
 
 	of_id = of_match_node(mtk_spi_of_match, pdev->dev.of_node);
 	if (!of_id) {
@@ -514,21 +518,34 @@ static int mtk_spi_probe(struct platform_device *pdev)
 		master->flags = SPI_MASTER_MUST_TX;
 
 	if (mdata->dev_comp->need_pad_sel) {
-		ret = of_property_read_u32(pdev->dev.of_node,
-					   "mediatek,pad-select",
-					   &mdata->pad_sel);
-		if (ret) {
-			dev_err(&pdev->dev, "failed to read pad select: %d\n",
-				ret);
+		mdata->pad_num = of_property_count_u32_elems(
+			pdev->dev.of_node,
+			"mediatek,pad-select");
+		if (mdata->pad_num < 0) {
+			dev_err(&pdev->dev,
+				"No 'mediatek,pad-select' property\n");
+			ret = -EINVAL;
 			goto err_put_master;
 		}
 
-		if (mdata->pad_sel > MT8173_SPI_MAX_PAD_SEL) {
-			dev_err(&pdev->dev, "wrong pad-select: %u\n",
-				mdata->pad_sel);
-			ret = -EINVAL;
+		mdata->pad_sel = devm_kmalloc_array(&pdev->dev, mdata->pad_num,
+						    sizeof(u32), GFP_KERNEL);
+		if (!mdata->pad_sel) {
+			ret = -ENOMEM;
 			goto err_put_master;
 		}
+
+		for (i = 0; i < mdata->pad_num; i++) {
+			of_property_read_u32_index(pdev->dev.of_node,
+						   "mediatek,pad-select",
+						   i, &mdata->pad_sel[i]);
+			if (mdata->pad_sel[i] > MT8173_SPI_MAX_PAD_SEL) {
+				dev_err(&pdev->dev, "wrong pad-sel[%d]: %u\n",
+					i, mdata->pad_sel[i]);
+				ret = -EINVAL;
+				goto err_put_master;
+			}
+		}
 	}
 
 	platform_set_drvdata(pdev, master);
@@ -606,6 +623,26 @@ static int mtk_spi_probe(struct platform_device *pdev)
 		goto err_put_master;
 	}
 
+	if (mdata->dev_comp->need_pad_sel) {
+		if (mdata->pad_num != master->num_chipselect) {
+			dev_err(&pdev->dev,
+				"pad_num does not match num_chipselect(%d != %d)\n",
+				mdata->pad_num, master->num_chipselect);
+			ret = -EINVAL;
+			goto err_put_master;
+		}
+
+		for (i = 0; i < master->num_chipselect; i++) {
+			ret = devm_gpio_request(&pdev->dev, master->cs_gpios[i],
+						dev_name(&pdev->dev));
+			if (ret) {
+				dev_err(&pdev->dev,
+					"can't get CS GPIO %i\n", i);
+				goto err_put_master;
+			}
+		}
+	}
+
 	return 0;
 
 err_disable_clk:
diff --git a/drivers/spi/spi-oc-tiny.c b/drivers/spi/spi-oc-tiny.c
index 76656a77ec12..b5911282a611 100644
--- a/drivers/spi/spi-oc-tiny.c
+++ b/drivers/spi/spi-oc-tiny.c
@@ -207,8 +207,7 @@ static int tiny_spi_of_probe(struct platform_device *pdev)
 	struct tiny_spi *hw = platform_get_drvdata(pdev);
 	struct device_node *np = pdev->dev.of_node;
 	unsigned int i;
-	const __be32 *val;
-	int len;
+	u32 val;
 
 	if (!np)
 		return 0;
@@ -226,13 +225,10 @@ static int tiny_spi_of_probe(struct platform_device *pdev)
 			return -ENODEV;
 	}
 	hw->bitbang.master->dev.of_node = pdev->dev.of_node;
-	val = of_get_property(pdev->dev.of_node,
-			      "clock-frequency", &len);
-	if (val && len >= sizeof(__be32))
-		hw->freq = be32_to_cpup(val);
-	val = of_get_property(pdev->dev.of_node, "baud-width", &len);
-	if (val && len >= sizeof(__be32))
-		hw->baudwidth = be32_to_cpup(val);
+	if (!of_property_read_u32(np, "clock-frequency", &val))
+		hw->freq = val;
+	if (!of_property_read_u32(np, "baud-width", &val))
+		hw->baudwidth = val;
 	return 0;
 }
 #else /* !CONFIG_OF */
diff --git a/drivers/spi/spi-octeon.c b/drivers/spi/spi-octeon.c
index e99d6a93d394..07e4ce8273df 100644
--- a/drivers/spi/spi-octeon.c
+++ b/drivers/spi/spi-octeon.c
@@ -65,7 +65,7 @@ static int octeon_spi_do_transfer(struct octeon_spi *p,
 	cpha = mode & SPI_CPHA;
 	cpol = mode & SPI_CPOL;
 
-	speed_hz = xfer->speed_hz ? : spi->max_speed_hz;
+	speed_hz = xfer->speed_hz;
 
 	clkdiv = octeon_get_io_clock_rate() / (2 * speed_hz);
 
diff --git a/drivers/spi/spi-omap-100k.c b/drivers/spi/spi-omap-100k.c
index 35b332dacb13..76a8425be227 100644
--- a/drivers/spi/spi-omap-100k.c
+++ b/drivers/spi/spi-omap-100k.c
@@ -244,12 +244,12 @@ static int omap1_spi100k_setup_transfer(struct spi_device *spi,
 {
 	struct omap1_spi100k *spi100k = spi_master_get_devdata(spi->master);
 	struct omap1_spi100k_cs *cs = spi->controller_state;
-	u8 word_len = spi->bits_per_word;
+	u8 word_len;
 
-	if (t != NULL && t->bits_per_word)
+	if (t != NULL)
 		word_len = t->bits_per_word;
-	if (!word_len)
-		word_len = 8;
+	else
+		word_len = spi->bits_per_word;
 
 	if (spi->bits_per_word > 32)
 		return -EINVAL;
@@ -302,7 +302,6 @@ static int omap1_spi100k_transfer_one_message(struct spi_master *master,
 	struct spi_device *spi = m->spi;
 	struct spi_transfer *t = NULL;
 	int cs_active = 0;
-	int par_override = 0;
 	int status = 0;
 
 	list_for_each_entry(t, &m->transfers, transfer_list) {
@@ -310,14 +309,9 @@ static int omap1_spi100k_transfer_one_message(struct spi_master *master,
 			status = -EINVAL;
 			break;
 		}
-		if (par_override || t->speed_hz || t->bits_per_word) {
-			par_override = 1;
-			status = omap1_spi100k_setup_transfer(spi, t);
-			if (status < 0)
-				break;
-			if (!t->speed_hz && !t->bits_per_word)
-				par_override = 0;
-		}
+		status = omap1_spi100k_setup_transfer(spi, t);
+		if (status < 0)
+			break;
 
 		if (!cs_active) {
 			omap1_spi100k_force_cs(spi100k, 1);
@@ -347,11 +341,7 @@ static int omap1_spi100k_transfer_one_message(struct spi_master *master,
 		}
 	}
 
-	/* Restore defaults if they were overriden */
-	if (par_override) {
-		par_override = 0;
-		status = omap1_spi100k_setup_transfer(spi, NULL);
-	}
+	status = omap1_spi100k_setup_transfer(spi, NULL);
 
 	if (cs_active)
 		omap1_spi100k_force_cs(spi100k, 0);
diff --git a/drivers/spi/spi-omap-uwire.c b/drivers/spi/spi-omap-uwire.c
index 55576db31549..ce8dbdbce312 100644
--- a/drivers/spi/spi-omap-uwire.c
+++ b/drivers/spi/spi-omap-uwire.c
@@ -205,7 +205,7 @@ static void uwire_chipselect(struct spi_device *spi, int value)
 static int uwire_txrx(struct spi_device *spi, struct spi_transfer *t)
 {
 	unsigned	len = t->len;
-	unsigned	bits = t->bits_per_word ? : spi->bits_per_word;
+	unsigned	bits = t->bits_per_word;
 	unsigned	bytes;
 	u16		val, w;
 	int		status = 0;
@@ -344,9 +344,10 @@ static int uwire_setup_transfer(struct spi_device *spi, struct spi_transfer *t)
 	/* assume it's already enabled */
 	rate = clk_get_rate(uwire->ck);
 
-	hz = spi->max_speed_hz;
-	if (t != NULL && t->speed_hz)
+	if (t != NULL)
 		hz = t->speed_hz;
+	else
+		hz = spi->max_speed_hz;
 
 	if (!hz) {
 		pr_debug("%s: zero speed?\n", dev_name(&spi->dev));
diff --git a/drivers/spi/spi-omap2-mcspi.c b/drivers/spi/spi-omap2-mcspi.c
index 3d09e0b69b73..1f8903d356e5 100644
--- a/drivers/spi/spi-omap2-mcspi.c
+++ b/drivers/spi/spi-omap2-mcspi.c
@@ -1217,6 +1217,33 @@ out:
 	return status;
 }
 
+static int omap2_mcspi_prepare_message(struct spi_master *master,
+				       struct spi_message *msg)
+{
+	struct omap2_mcspi	*mcspi = spi_master_get_devdata(master);
+	struct omap2_mcspi_regs	*ctx = &mcspi->ctx;
+	struct omap2_mcspi_cs	*cs;
+
+	/* Only a single channel can have the FORCE bit enabled
+	 * in its chconf0 register.
+	 * Scan all channels and disable them except the current one.
+	 * A FORCE can remain from a last transfer having cs_change enabled
+	 */
+	list_for_each_entry(cs, &ctx->cs, node) {
+		if (msg->spi->controller_state == cs)
+			continue;
+
+		if ((cs->chconf0 & OMAP2_MCSPI_CHCONF_FORCE)) {
+			cs->chconf0 &= ~OMAP2_MCSPI_CHCONF_FORCE;
+			writel_relaxed(cs->chconf0,
+					cs->base + OMAP2_MCSPI_CHCONF0);
+			readl_relaxed(cs->base + OMAP2_MCSPI_CHCONF0);
+		}
+	}
+
+	return 0;
+}
+
 static int omap2_mcspi_transfer_one(struct spi_master *master,
 		struct spi_device *spi, struct spi_transfer *t)
 {
@@ -1344,6 +1371,7 @@ static int omap2_mcspi_probe(struct platform_device *pdev)
 	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 32);
 	master->setup = omap2_mcspi_setup;
 	master->auto_runtime_pm = true;
+	master->prepare_message = omap2_mcspi_prepare_message;
 	master->transfer_one = omap2_mcspi_transfer_one;
 	master->set_cs = omap2_mcspi_set_cs;
 	master->cleanup = omap2_mcspi_cleanup;
diff --git a/drivers/spi/spi-ppc4xx.c b/drivers/spi/spi-ppc4xx.c
index 54fb984a3e17..dd3d0a218d8b 100644
--- a/drivers/spi/spi-ppc4xx.c
+++ b/drivers/spi/spi-ppc4xx.c
@@ -210,12 +210,12 @@ static int spi_ppc4xx_setupxfer(struct spi_device *spi, struct spi_transfer *t)
 	if (in_8(&hw->regs->cdm) != cdm)
 		out_8(&hw->regs->cdm, cdm);
 
-	spin_lock(&hw->bitbang.lock);
+	mutex_lock(&hw->bitbang.lock);
 	if (!hw->bitbang.busy) {
 		hw->bitbang.chipselect(spi, BITBANG_CS_INACTIVE);
 		/* Need to ndelay here? */
 	}
-	spin_unlock(&hw->bitbang.lock);
+	mutex_unlock(&hw->bitbang.lock);
 
 	return 0;
 }
diff --git a/drivers/spi/spi-pxa2xx-dma.c b/drivers/spi/spi-pxa2xx-dma.c
index 66a173939be8..bd8b369a343c 100644
--- a/drivers/spi/spi-pxa2xx-dma.c
+++ b/drivers/spi/spi-pxa2xx-dma.c
@@ -344,10 +344,6 @@ void pxa2xx_spi_dma_release(struct driver_data *drv_data)
 	}
 }
 
-void pxa2xx_spi_dma_resume(struct driver_data *drv_data)
-{
-}
-
 int pxa2xx_spi_set_dma_burst_and_threshold(struct chip_data *chip,
 					   struct spi_device *spi,
 					   u8 bits_per_word, u32 *burst_code,
diff --git a/drivers/spi/spi-pxa2xx.c b/drivers/spi/spi-pxa2xx.c
index a8ef38ebb9c9..b25dc71b0ea9 100644
--- a/drivers/spi/spi-pxa2xx.c
+++ b/drivers/spi/spi-pxa2xx.c
@@ -13,6 +13,7 @@
  * GNU General Public License for more details.
  */
 
+#include <linux/bitops.h>
 #include <linux/init.h>
 #include <linux/module.h>
 #include <linux/device.h>
@@ -61,9 +62,13 @@ MODULE_ALIAS("platform:pxa2xx-spi");
 				| QUARK_X1000_SSCR1_TFT		\
 				| SSCR1_SPH | SSCR1_SPO | SSCR1_LBM)
 
-#define GENERAL_REG_RXTO_HOLDOFF_DISABLE BIT(24)
-#define SPI_CS_CONTROL_SW_MODE	BIT(0)
-#define SPI_CS_CONTROL_CS_HIGH	BIT(1)
+#define LPSS_GENERAL_REG_RXTO_HOLDOFF_DISABLE	BIT(24)
+#define LPSS_CS_CONTROL_SW_MODE			BIT(0)
+#define LPSS_CS_CONTROL_CS_HIGH			BIT(1)
+#define LPSS_CS_CONTROL_CS_SEL_SHIFT		8
+#define LPSS_CS_CONTROL_CS_SEL_MASK		(3 << LPSS_CS_CONTROL_CS_SEL_SHIFT)
+#define LPSS_CAPS_CS_EN_SHIFT			9
+#define LPSS_CAPS_CS_EN_MASK			(0xf << LPSS_CAPS_CS_EN_SHIFT)
 
 struct lpss_config {
 	/* LPSS offset from drv_data->ioaddr */
@@ -72,6 +77,7 @@ struct lpss_config {
 	int reg_general;
 	int reg_ssp;
 	int reg_cs_ctrl;
+	int reg_capabilities;
 	/* FIFO thresholds */
 	u32 rx_threshold;
 	u32 tx_threshold_lo;
@@ -85,6 +91,7 @@ static const struct lpss_config lpss_platforms[] = {
 		.reg_general = 0x08,
 		.reg_ssp = 0x0c,
 		.reg_cs_ctrl = 0x18,
+		.reg_capabilities = -1,
 		.rx_threshold = 64,
 		.tx_threshold_lo = 160,
 		.tx_threshold_hi = 224,
@@ -94,6 +101,7 @@ static const struct lpss_config lpss_platforms[] = {
 		.reg_general = 0x08,
 		.reg_ssp = 0x0c,
 		.reg_cs_ctrl = 0x18,
+		.reg_capabilities = -1,
 		.rx_threshold = 64,
 		.tx_threshold_lo = 160,
 		.tx_threshold_hi = 224,
@@ -103,10 +111,21 @@ static const struct lpss_config lpss_platforms[] = {
 		.reg_general = -1,
 		.reg_ssp = 0x20,
 		.reg_cs_ctrl = 0x24,
+		.reg_capabilities = 0xfc,
 		.rx_threshold = 1,
 		.tx_threshold_lo = 32,
 		.tx_threshold_hi = 56,
 	},
+	{	/* LPSS_BXT_SSP */
+		.offset = 0x200,
+		.reg_general = -1,
+		.reg_ssp = 0x20,
+		.reg_cs_ctrl = 0x24,
+		.reg_capabilities = 0xfc,
+		.rx_threshold = 1,
+		.tx_threshold_lo = 16,
+		.tx_threshold_hi = 48,
+	},
 };
 
 static inline const struct lpss_config
@@ -121,6 +140,7 @@ static bool is_lpss_ssp(const struct driver_data *drv_data)
 	case LPSS_LPT_SSP:
 	case LPSS_BYT_SSP:
 	case LPSS_SPT_SSP:
+	case LPSS_BXT_SSP:
 		return true;
 	default:
 		return false;
@@ -249,7 +269,9 @@ static void lpss_ssp_setup(struct driver_data *drv_data)
 	drv_data->lpss_base = drv_data->ioaddr + config->offset;
 
 	/* Enable software chip select control */
-	value = SPI_CS_CONTROL_SW_MODE | SPI_CS_CONTROL_CS_HIGH;
+	value = __lpss_ssp_read_priv(drv_data, config->reg_cs_ctrl);
+	value &= ~(LPSS_CS_CONTROL_SW_MODE | LPSS_CS_CONTROL_CS_HIGH);
+	value |= LPSS_CS_CONTROL_SW_MODE | LPSS_CS_CONTROL_CS_HIGH;
 	__lpss_ssp_write_priv(drv_data, config->reg_cs_ctrl, value);
 
 	/* Enable multiblock DMA transfers */
@@ -259,7 +281,7 @@ static void lpss_ssp_setup(struct driver_data *drv_data)
 		if (config->reg_general >= 0) {
 			value = __lpss_ssp_read_priv(drv_data,
 						     config->reg_general);
-			value |= GENERAL_REG_RXTO_HOLDOFF_DISABLE;
+			value |= LPSS_GENERAL_REG_RXTO_HOLDOFF_DISABLE;
 			__lpss_ssp_write_priv(drv_data,
 					      config->reg_general, value);
 		}
@@ -269,15 +291,34 @@ static void lpss_ssp_setup(struct driver_data *drv_data)
 static void lpss_ssp_cs_control(struct driver_data *drv_data, bool enable)
 {
 	const struct lpss_config *config;
-	u32 value;
+	u32 value, cs;
 
 	config = lpss_get_config(drv_data);
 
 	value = __lpss_ssp_read_priv(drv_data, config->reg_cs_ctrl);
-	if (enable)
-		value &= ~SPI_CS_CONTROL_CS_HIGH;
-	else
-		value |= SPI_CS_CONTROL_CS_HIGH;
+	if (enable) {
+		cs = drv_data->cur_msg->spi->chip_select;
+		cs <<= LPSS_CS_CONTROL_CS_SEL_SHIFT;
+		if (cs != (value & LPSS_CS_CONTROL_CS_SEL_MASK)) {
+			/*
+			 * When switching another chip select output active
+			 * the output must be selected first and wait 2 ssp_clk
+			 * cycles before changing state to active. Otherwise
+			 * a short glitch will occur on the previous chip
+			 * select since output select is latched but state
+			 * control is not.
+			 */
+			value &= ~LPSS_CS_CONTROL_CS_SEL_MASK;
+			value |= cs;
+			__lpss_ssp_write_priv(drv_data,
+					      config->reg_cs_ctrl, value);
+			ndelay(1000000000 /
+			       (drv_data->master->max_speed_hz / 2));
+		}
+		value &= ~LPSS_CS_CONTROL_CS_HIGH;
+	} else {
+		value |= LPSS_CS_CONTROL_CS_HIGH;
+	}
 	__lpss_ssp_write_priv(drv_data, config->reg_cs_ctrl, value);
 }
 
@@ -734,7 +775,7 @@ static unsigned int quark_x1000_get_clk_div(int rate, u32 *dds)
 	mul = (1 << 24) >> 1;
 
 	/* Calculate initial quot */
-	q1 = DIV_ROUND_CLOSEST(fref1, rate);
+	q1 = DIV_ROUND_UP(fref1, rate);
 
 	/* Scale q1 if it's too big */
 	if (q1 > 256) {
@@ -759,7 +800,7 @@ static unsigned int quark_x1000_get_clk_div(int rate, u32 *dds)
 
 	/* Case 2 */
 
-	q2 = DIV_ROUND_CLOSEST(fref2, rate);
+	q2 = DIV_ROUND_UP(fref2, rate);
 	r2 = abs(fref2 / q2 - rate);
 
 	/*
@@ -778,13 +819,13 @@ static unsigned int quark_x1000_get_clk_div(int rate, u32 *dds)
 		mul = (1 << 24) * 2 / 5;
 	}
 
-	/* Check case 3 only If the divisor is big enough */
+	/* Check case 3 only if the divisor is big enough */
 	if (fref / rate >= 80) {
 		u64 fssp;
 		u32 m;
 
 		/* Calculate initial quot */
-		q1 = DIV_ROUND_CLOSEST(fref, rate);
+		q1 = DIV_ROUND_UP(fref, rate);
 		m = (1 << 24) / q1;
 
 		/* Get the remainder */
@@ -806,7 +847,7 @@ static unsigned int quark_x1000_get_clk_div(int rate, u32 *dds)
 
 static unsigned int ssp_get_clk_div(struct driver_data *drv_data, int rate)
 {
-	unsigned long ssp_clk = drv_data->max_clk_rate;
+	unsigned long ssp_clk = drv_data->master->max_speed_hz;
 	const struct ssp_device *ssp = drv_data->ssp;
 
 	rate = min_t(int, ssp_clk, rate);
@@ -818,8 +859,9 @@ static unsigned int ssp_get_clk_div(struct driver_data *drv_data, int rate)
 }
 
 static unsigned int pxa2xx_ssp_get_clk_div(struct driver_data *drv_data,
-					   struct chip_data *chip, int rate)
+					   int rate)
 {
+	struct chip_data *chip = drv_data->cur_chip;
 	unsigned int clk_div;
 
 	switch (drv_data->ssp_type) {
@@ -922,52 +964,55 @@ static void pump_transfers(unsigned long data)
 	drv_data->read = drv_data->rx ? chip->read : null_reader;
 
 	/* Change speed and bit per word on a per transfer */
-	cr0 = chip->cr0;
-	if (transfer->speed_hz || transfer->bits_per_word) {
-
-		bits = chip->bits_per_word;
-		speed = chip->speed_hz;
-
-		if (transfer->speed_hz)
-			speed = transfer->speed_hz;
-
-		if (transfer->bits_per_word)
-			bits = transfer->bits_per_word;
-
-		clk_div = pxa2xx_ssp_get_clk_div(drv_data, chip, speed);
-
-		if (bits <= 8) {
-			drv_data->n_bytes = 1;
-			drv_data->read = drv_data->read != null_reader ?
-						u8_reader : null_reader;
-			drv_data->write = drv_data->write != null_writer ?
-						u8_writer : null_writer;
-		} else if (bits <= 16) {
-			drv_data->n_bytes = 2;
-			drv_data->read = drv_data->read != null_reader ?
-						u16_reader : null_reader;
-			drv_data->write = drv_data->write != null_writer ?
-						u16_writer : null_writer;
-		} else if (bits <= 32) {
-			drv_data->n_bytes = 4;
-			drv_data->read = drv_data->read != null_reader ?
-						u32_reader : null_reader;
-			drv_data->write = drv_data->write != null_writer ?
-						u32_writer : null_writer;
-		}
-		/* if bits/word is changed in dma mode, then must check the
-		 * thresholds and burst also */
-		if (chip->enable_dma) {
-			if (pxa2xx_spi_set_dma_burst_and_threshold(chip,
-							message->spi,
-							bits, &dma_burst,
-							&dma_thresh))
-				dev_warn_ratelimited(&message->spi->dev,
-						     "pump_transfers: DMA burst size reduced to match bits_per_word\n");
-		}
-
-		cr0 = pxa2xx_configure_sscr0(drv_data, clk_div, bits);
+	bits = transfer->bits_per_word;
+	speed = transfer->speed_hz;
+
+	clk_div = pxa2xx_ssp_get_clk_div(drv_data, speed);
+
+	if (bits <= 8) {
+		drv_data->n_bytes = 1;
+		drv_data->read = drv_data->read != null_reader ?
+					u8_reader : null_reader;
+		drv_data->write = drv_data->write != null_writer ?
+					u8_writer : null_writer;
+	} else if (bits <= 16) {
+		drv_data->n_bytes = 2;
+		drv_data->read = drv_data->read != null_reader ?
+					u16_reader : null_reader;
+		drv_data->write = drv_data->write != null_writer ?
+					u16_writer : null_writer;
+	} else if (bits <= 32) {
+		drv_data->n_bytes = 4;
+		drv_data->read = drv_data->read != null_reader ?
+					u32_reader : null_reader;
+		drv_data->write = drv_data->write != null_writer ?
+					u32_writer : null_writer;
 	}
+	/*
+	 * if bits/word is changed in dma mode, then must check the
+	 * thresholds and burst also
+	 */
+	if (chip->enable_dma) {
+		if (pxa2xx_spi_set_dma_burst_and_threshold(chip,
+						message->spi,
+						bits, &dma_burst,
+						&dma_thresh))
+			dev_warn_ratelimited(&message->spi->dev,
+					     "pump_transfers: DMA burst size reduced to match bits_per_word\n");
+	}
+
+	/* NOTE:  PXA25x_SSP _could_ use external clocking ... */
+	cr0 = pxa2xx_configure_sscr0(drv_data, clk_div, bits);
+	if (!pxa25x_ssp_comp(drv_data))
+		dev_dbg(&message->spi->dev, "%u Hz actual, %s\n",
+			drv_data->master->max_speed_hz
+				/ (1 + ((cr0 & SSCR0_SCR(0xfff)) >> 8)),
+			chip->enable_dma ? "DMA" : "PIO");
+	else
+		dev_dbg(&message->spi->dev, "%u Hz actual, %s\n",
+			drv_data->master->max_speed_hz / 2
+				/ (1 + ((cr0 & SSCR0_SCR(0x0ff)) >> 8)),
+			chip->enable_dma ? "DMA" : "PIO");
 
 	message->state = RUNNING_STATE;
 
@@ -1111,7 +1156,6 @@ static int setup(struct spi_device *spi)
 	struct chip_data *chip;
 	const struct lpss_config *config;
 	struct driver_data *drv_data = spi_master_get_devdata(spi->master);
-	unsigned int clk_div;
 	uint tx_thres, tx_hi_thres, rx_thres;
 
 	switch (drv_data->ssp_type) {
@@ -1123,6 +1167,7 @@ static int setup(struct spi_device *spi)
 	case LPSS_LPT_SSP:
 	case LPSS_BYT_SSP:
 	case LPSS_SPT_SSP:
+	case LPSS_BXT_SSP:
 		config = lpss_get_config(drv_data);
 		tx_thres = config->tx_threshold_lo;
 		tx_hi_thres = config->tx_threshold_hi;
@@ -1203,11 +1248,6 @@ static int setup(struct spi_device *spi)
 		}
 	}
 
-	clk_div = pxa2xx_ssp_get_clk_div(drv_data, chip, spi->max_speed_hz);
-	chip->speed_hz = spi->max_speed_hz;
-
-	chip->cr0 = pxa2xx_configure_sscr0(drv_data, clk_div,
-					   spi->bits_per_word);
 	switch (drv_data->ssp_type) {
 	case QUARK_X1000_SSP:
 		chip->threshold = (QUARK_X1000_SSCR1_RxTresh(rx_thres)
@@ -1228,18 +1268,6 @@ static int setup(struct spi_device *spi)
 	if (spi->mode & SPI_LOOP)
 		chip->cr1 |= SSCR1_LBM;
 
-	/* NOTE:  PXA25x_SSP _could_ use external clocking ... */
-	if (!pxa25x_ssp_comp(drv_data))
-		dev_dbg(&spi->dev, "%ld Hz actual, %s\n",
-			drv_data->max_clk_rate
-				/ (1 + ((chip->cr0 & SSCR0_SCR(0xfff)) >> 8)),
-			chip->enable_dma ? "DMA" : "PIO");
-	else
-		dev_dbg(&spi->dev, "%ld Hz actual, %s\n",
-			drv_data->max_clk_rate / 2
-				/ (1 + ((chip->cr0 & SSCR0_SCR(0x0ff)) >> 8)),
-			chip->enable_dma ? "DMA" : "PIO");
-
 	if (spi->bits_per_word <= 8) {
 		chip->n_bytes = 1;
 		chip->read = u8_reader;
@@ -1249,13 +1277,10 @@ static int setup(struct spi_device *spi)
 		chip->read = u16_reader;
 		chip->write = u16_writer;
 	} else if (spi->bits_per_word <= 32) {
-		if (!is_quark_x1000_ssp(drv_data))
-			chip->cr0 |= SSCR0_EDSS;
 		chip->n_bytes = 4;
 		chip->read = u32_reader;
 		chip->write = u32_writer;
 	}
-	chip->bits_per_word = spi->bits_per_word;
 
 	spi_set_ctldata(spi, chip);
 
@@ -1279,6 +1304,7 @@ static void cleanup(struct spi_device *spi)
 	kfree(chip);
 }
 
+#ifdef CONFIG_PCI
 #ifdef CONFIG_ACPI
 
 static const struct acpi_device_id pxa2xx_spi_acpi_match[] = {
@@ -1292,6 +1318,23 @@ static const struct acpi_device_id pxa2xx_spi_acpi_match[] = {
 };
 MODULE_DEVICE_TABLE(acpi, pxa2xx_spi_acpi_match);
 
+static int pxa2xx_spi_get_port_id(struct acpi_device *adev)
+{
+	unsigned int devid;
+	int port_id = -1;
+
+	if (adev && adev->pnp.unique_id &&
+	    !kstrtouint(adev->pnp.unique_id, 0, &devid))
+		port_id = devid;
+	return port_id;
+}
+#else /* !CONFIG_ACPI */
+static int pxa2xx_spi_get_port_id(struct acpi_device *adev)
+{
+	return -1;
+}
+#endif
+
 /*
  * PCI IDs of compound devices that integrate both host controller and private
  * integrated DMA engine. Please note these are not used in module
@@ -1304,6 +1347,14 @@ static const struct pci_device_id pxa2xx_spi_pci_compound_match[] = {
 	/* SPT-H */
 	{ PCI_VDEVICE(INTEL, 0xa129), LPSS_SPT_SSP },
 	{ PCI_VDEVICE(INTEL, 0xa12a), LPSS_SPT_SSP },
+	/* BXT */
+	{ PCI_VDEVICE(INTEL, 0x0ac2), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x0ac4), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x0ac6), LPSS_BXT_SSP },
+	/* APL */
+	{ PCI_VDEVICE(INTEL, 0x5ac2), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x5ac4), LPSS_BXT_SSP },
+	{ PCI_VDEVICE(INTEL, 0x5ac6), LPSS_BXT_SSP },
 	{ },
 };
 
@@ -1318,7 +1369,7 @@ static bool pxa2xx_spi_idma_filter(struct dma_chan *chan, void *param)
 }
 
 static struct pxa2xx_spi_master *
-pxa2xx_spi_acpi_get_pdata(struct platform_device *pdev)
+pxa2xx_spi_init_pdata(struct platform_device *pdev)
 {
 	struct pxa2xx_spi_master *pdata;
 	struct acpi_device *adev;
@@ -1326,18 +1377,18 @@ pxa2xx_spi_acpi_get_pdata(struct platform_device *pdev)
 	struct resource *res;
 	const struct acpi_device_id *adev_id = NULL;
 	const struct pci_device_id *pcidev_id = NULL;
-	int devid, type;
+	int type;
 
-	if (!ACPI_HANDLE(&pdev->dev) ||
-	    acpi_bus_get_device(ACPI_HANDLE(&pdev->dev), &adev))
-		return NULL;
+	adev = ACPI_COMPANION(&pdev->dev);
 
 	if (dev_is_pci(pdev->dev.parent))
 		pcidev_id = pci_match_id(pxa2xx_spi_pci_compound_match,
 					 to_pci_dev(pdev->dev.parent));
-	else
+	else if (adev)
 		adev_id = acpi_match_device(pdev->dev.driver->acpi_match_table,
 					    &pdev->dev);
+	else
+		return NULL;
 
 	if (adev_id)
 		type = (int)adev_id->driver_data;
@@ -1371,10 +1422,7 @@ pxa2xx_spi_acpi_get_pdata(struct platform_device *pdev)
 	ssp->irq = platform_get_irq(pdev, 0);
 	ssp->type = type;
 	ssp->pdev = pdev;
-
-	ssp->port_id = -1;
-	if (adev->pnp.unique_id && !kstrtoint(adev->pnp.unique_id, 0, &devid))
-		ssp->port_id = devid;
+	ssp->port_id = pxa2xx_spi_get_port_id(adev);
 
 	pdata->num_chipselect = 1;
 	pdata->enable_dma = true;
@@ -1382,9 +1430,9 @@ pxa2xx_spi_acpi_get_pdata(struct platform_device *pdev)
 	return pdata;
 }
 
-#else
+#else /* !CONFIG_PCI */
 static inline struct pxa2xx_spi_master *
-pxa2xx_spi_acpi_get_pdata(struct platform_device *pdev)
+pxa2xx_spi_init_pdata(struct platform_device *pdev)
 {
 	return NULL;
 }
@@ -1397,12 +1445,13 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 	struct spi_master *master;
 	struct driver_data *drv_data;
 	struct ssp_device *ssp;
+	const struct lpss_config *config;
 	int status;
 	u32 tmp;
 
 	platform_info = dev_get_platdata(dev);
 	if (!platform_info) {
-		platform_info = pxa2xx_spi_acpi_get_pdata(pdev);
+		platform_info = pxa2xx_spi_init_pdata(pdev);
 		if (!platform_info) {
 			dev_err(&pdev->dev, "missing platform data\n");
 			return -ENODEV;
@@ -1436,7 +1485,6 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LOOP;
 
 	master->bus_num = ssp->port_id;
-	master->num_chipselect = platform_info->num_chipselect;
 	master->dma_alignment = DMA_ALIGNMENT;
 	master->cleanup = cleanup;
 	master->setup = setup;
@@ -1489,7 +1537,7 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 	/* Enable SOC clock */
 	clk_prepare_enable(ssp->clk);
 
-	drv_data->max_clk_rate = clk_get_rate(ssp->clk);
+	master->max_speed_hz = clk_get_rate(ssp->clk);
 
 	/* Load default SSP configuration */
 	pxa2xx_spi_write(drv_data, SSCR0, 0);
@@ -1522,6 +1570,19 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 	if (is_lpss_ssp(drv_data))
 		lpss_ssp_setup(drv_data);
 
+	if (is_lpss_ssp(drv_data)) {
+		lpss_ssp_setup(drv_data);
+		config = lpss_get_config(drv_data);
+		if (config->reg_capabilities >= 0) {
+			tmp = __lpss_ssp_read_priv(drv_data,
+						   config->reg_capabilities);
+			tmp &= LPSS_CAPS_CS_EN_MASK;
+			tmp >>= LPSS_CAPS_CS_EN_SHIFT;
+			platform_info->num_chipselect = ffz(tmp);
+		}
+	}
+	master->num_chipselect = platform_info->num_chipselect;
+
 	tasklet_init(&drv_data->pump_transfers, pump_transfers,
 		     (unsigned long)drv_data);
 
@@ -1614,8 +1675,6 @@ static int pxa2xx_spi_resume(struct device *dev)
 	struct ssp_device *ssp = drv_data->ssp;
 	int status = 0;
 
-	pxa2xx_spi_dma_resume(drv_data);
-
 	/* Enable the SSP clock */
 	if (!pm_runtime_suspended(dev))
 		clk_prepare_enable(ssp->clk);
diff --git a/drivers/spi/spi-pxa2xx.h b/drivers/spi/spi-pxa2xx.h
index 0a9b6390a817..58efa98313aa 100644
--- a/drivers/spi/spi-pxa2xx.h
+++ b/drivers/spi/spi-pxa2xx.h
@@ -46,9 +46,6 @@ struct driver_data {
 	u32 clear_sr;
 	u32 mask_sr;
 
-	/* Maximun clock rate */
-	unsigned long max_clk_rate;
-
 	/* Message Transfer pump */
 	struct tasklet_struct pump_transfers;
 
@@ -86,10 +83,8 @@ struct driver_data {
 };
 
 struct chip_data {
-	u32 cr0;
 	u32 cr1;
 	u32 dds_rate;
-	u32 psp;
 	u32 timeout;
 	u8 n_bytes;
 	u32 dma_burst_size;
@@ -98,8 +93,6 @@ struct chip_data {
 	u16 lpss_rx_threshold;
 	u16 lpss_tx_threshold;
 	u8 enable_dma;
-	u8 bits_per_word;
-	u32 speed_hz;
 	union {
 		int gpio_cs;
 		unsigned int frm;
@@ -175,7 +168,6 @@ extern int pxa2xx_spi_dma_prepare(struct driver_data *drv_data, u32 dma_burst);
 extern void pxa2xx_spi_dma_start(struct driver_data *drv_data);
 extern int pxa2xx_spi_dma_setup(struct driver_data *drv_data);
 extern void pxa2xx_spi_dma_release(struct driver_data *drv_data);
-extern void pxa2xx_spi_dma_resume(struct driver_data *drv_data);
 extern int pxa2xx_spi_set_dma_burst_and_threshold(struct chip_data *chip,
 						  struct spi_device *spi,
 						  u8 bits_per_word,
@@ -196,7 +188,6 @@ static inline int pxa2xx_spi_dma_setup(struct driver_data *drv_data)
 	return 0;
 }
 static inline void pxa2xx_spi_dma_release(struct driver_data *drv_data) {}
-static inline void pxa2xx_spi_dma_resume(struct driver_data *drv_data) {}
 static inline int pxa2xx_spi_set_dma_burst_and_threshold(struct chip_data *chip,
 							 struct spi_device *spi,
 							 u8 bits_per_word,
diff --git a/drivers/spi/spi-s3c24xx.c b/drivers/spi/spi-s3c24xx.c
index f36bc320a807..4e7d1bfed7e6 100644
--- a/drivers/spi/spi-s3c24xx.c
+++ b/drivers/spi/spi-s3c24xx.c
@@ -198,12 +198,12 @@ static int s3c24xx_spi_setup(struct spi_device *spi)
 	if (ret)
 		return ret;
 
-	spin_lock(&hw->bitbang.lock);
+	mutex_lock(&hw->bitbang.lock);
 	if (!hw->bitbang.busy) {
 		hw->bitbang.chipselect(spi, BITBANG_CS_INACTIVE);
 		/* need to ndelay for 0.5 clocktick ? */
 	}
-	spin_unlock(&hw->bitbang.lock);
+	mutex_unlock(&hw->bitbang.lock);
 
 	return 0;
 }
diff --git a/drivers/spi/spi-s3c64xx.c b/drivers/spi/spi-s3c64xx.c
index cd1cfac0447f..8e86e7f6663a 100644
--- a/drivers/spi/spi-s3c64xx.c
+++ b/drivers/spi/spi-s3c64xx.c
@@ -32,6 +32,7 @@
 #define MAX_SPI_PORTS		6
 #define S3C64XX_SPI_QUIRK_POLL		(1 << 0)
 #define S3C64XX_SPI_QUIRK_CS_AUTO	(1 << 1)
+#define AUTOSUSPEND_TIMEOUT	2000
 
 /* Registers and bit-fields */
 
@@ -682,7 +683,7 @@ static int s3c64xx_spi_transfer_one(struct spi_master *master,
 
 	/* Only BPW and Speed may change across transfers */
 	bpw = xfer->bits_per_word;
-	speed = xfer->speed_hz ? : spi->max_speed_hz;
+	speed = xfer->speed_hz;
 
 	if (bpw != sdd->cur_bpw || speed != sdd->cur_speed) {
 		sdd->cur_bpw = bpw;
@@ -859,13 +860,15 @@ static int s3c64xx_spi_setup(struct spi_device *spi)
 		}
 	}
 
-	pm_runtime_put(&sdd->pdev->dev);
+	pm_runtime_mark_last_busy(&sdd->pdev->dev);
+	pm_runtime_put_autosuspend(&sdd->pdev->dev);
 	if (!(sdd->port_conf->quirks & S3C64XX_SPI_QUIRK_CS_AUTO))
 		writel(S3C64XX_SPI_SLAVE_SIG_INACT, sdd->regs + S3C64XX_SPI_SLAVE_SEL);
 	return 0;
 
 setup_exit:
-	pm_runtime_put(&sdd->pdev->dev);
+	pm_runtime_mark_last_busy(&sdd->pdev->dev);
+	pm_runtime_put_autosuspend(&sdd->pdev->dev);
 	/* setup() returns with device de-selected */
 	if (!(sdd->port_conf->quirks & S3C64XX_SPI_QUIRK_CS_AUTO))
 		writel(S3C64XX_SPI_SLAVE_SIG_INACT, sdd->regs + S3C64XX_SPI_SLAVE_SEL);
@@ -1162,6 +1165,12 @@ static int s3c64xx_spi_probe(struct platform_device *pdev)
 		goto err2;
 	}
 
+	pm_runtime_set_autosuspend_delay(&pdev->dev, AUTOSUSPEND_TIMEOUT);
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+	pm_runtime_get_sync(&pdev->dev);
+
 	/* Setup Deufult Mode */
 	s3c64xx_spi_hwinit(sdd, sdd->port_id);
 
@@ -1180,9 +1189,6 @@ static int s3c64xx_spi_probe(struct platform_device *pdev)
 	       S3C64XX_SPI_INT_TX_OVERRUN_EN | S3C64XX_SPI_INT_TX_UNDERRUN_EN,
 	       sdd->regs + S3C64XX_SPI_INT_EN);
 
-	pm_runtime_set_active(&pdev->dev);
-	pm_runtime_enable(&pdev->dev);
-
 	ret = devm_spi_register_master(&pdev->dev, master);
 	if (ret != 0) {
 		dev_err(&pdev->dev, "cannot register SPI master: %d\n", ret);
@@ -1195,9 +1201,16 @@ static int s3c64xx_spi_probe(struct platform_device *pdev)
 					mem_res, (FIFO_LVL_MASK(sdd) >> 1) + 1,
 					sdd->rx_dma.dmach, sdd->tx_dma.dmach);
 
+	pm_runtime_mark_last_busy(&pdev->dev);
+	pm_runtime_put_autosuspend(&pdev->dev);
+
 	return 0;
 
 err3:
+	pm_runtime_put_noidle(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+	pm_runtime_set_suspended(&pdev->dev);
+
 	clk_disable_unprepare(sdd->src_clk);
 err2:
 	clk_disable_unprepare(sdd->clk);
@@ -1212,7 +1225,7 @@ static int s3c64xx_spi_remove(struct platform_device *pdev)
 	struct spi_master *master = spi_master_get(platform_get_drvdata(pdev));
 	struct s3c64xx_spi_driver_data *sdd = spi_master_get_devdata(master);
 
-	pm_runtime_disable(&pdev->dev);
+	pm_runtime_get_sync(&pdev->dev);
 
 	writel(0, sdd->regs + S3C64XX_SPI_INT_EN);
 
@@ -1220,6 +1233,10 @@ static int s3c64xx_spi_remove(struct platform_device *pdev)
 
 	clk_disable_unprepare(sdd->clk);
 
+	pm_runtime_put_noidle(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+	pm_runtime_set_suspended(&pdev->dev);
+
 	return 0;
 }
 
@@ -1233,10 +1250,9 @@ static int s3c64xx_spi_suspend(struct device *dev)
 	if (ret)
 		return ret;
 
-	if (!pm_runtime_suspended(dev)) {
-		clk_disable_unprepare(sdd->clk);
-		clk_disable_unprepare(sdd->src_clk);
-	}
+	ret = pm_runtime_force_suspend(dev);
+	if (ret < 0)
+		return ret;
 
 	sdd->cur_speed = 0; /* Output Clock is stopped */
 
@@ -1248,14 +1264,14 @@ static int s3c64xx_spi_resume(struct device *dev)
 	struct spi_master *master = dev_get_drvdata(dev);
 	struct s3c64xx_spi_driver_data *sdd = spi_master_get_devdata(master);
 	struct s3c64xx_spi_info *sci = sdd->cntrlr_info;
+	int ret;
 
 	if (sci->cfg_gpio)
 		sci->cfg_gpio();
 
-	if (!pm_runtime_suspended(dev)) {
-		clk_prepare_enable(sdd->src_clk);
-		clk_prepare_enable(sdd->clk);
-	}
+	ret = pm_runtime_force_resume(dev);
+	if (ret < 0)
+		return ret;
 
 	s3c64xx_spi_hwinit(sdd, sdd->port_id);
 
diff --git a/drivers/spi/spi-ti-qspi.c b/drivers/spi/spi-ti-qspi.c
index aa6d284131e0..64318fcfacf2 100644
--- a/drivers/spi/spi-ti-qspi.c
+++ b/drivers/spi/spi-ti-qspi.c
@@ -39,8 +39,6 @@ struct ti_qspi_regs {
 };
 
 struct ti_qspi {
-	struct completion       transfer_complete;
-
 	/* list synchronization */
 	struct mutex            list_lock;
 
@@ -62,10 +60,6 @@ struct ti_qspi {
 
 #define QSPI_PID			(0x0)
 #define QSPI_SYSCONFIG			(0x10)
-#define QSPI_INTR_STATUS_RAW_SET	(0x20)
-#define QSPI_INTR_STATUS_ENABLED_CLEAR	(0x24)
-#define QSPI_INTR_ENABLE_SET_REG	(0x28)
-#define QSPI_INTR_ENABLE_CLEAR_REG	(0x2c)
 #define QSPI_SPI_CLOCK_CNTRL_REG	(0x40)
 #define QSPI_SPI_DC_REG			(0x44)
 #define QSPI_SPI_CMD_REG		(0x48)
@@ -97,7 +91,6 @@ struct ti_qspi {
 #define QSPI_RD_DUAL			(3 << 16)
 #define QSPI_RD_QUAD			(7 << 16)
 #define QSPI_INVAL			(4 << 16)
-#define QSPI_WC_CMD_INT_EN			(1 << 14)
 #define QSPI_FLEN(n)			((n - 1) << 0)
 #define QSPI_WLEN_MAX_BITS		128
 #define QSPI_WLEN_MAX_BYTES		16
@@ -106,10 +99,6 @@ struct ti_qspi {
 #define BUSY				0x01
 #define WC				0x02
 
-/* INTERRUPT REGISTER */
-#define QSPI_WC_INT_EN				(1 << 1)
-#define QSPI_WC_INT_DISABLE			(1 << 1)
-
 /* Device Control */
 #define QSPI_DD(m, n)			(m << (3 + n * 8))
 #define QSPI_CKPHA(n)			(1 << (2 + n * 8))
@@ -217,6 +206,24 @@ static inline u32 qspi_is_busy(struct ti_qspi *qspi)
 	return stat & BUSY;
 }
 
+static inline int ti_qspi_poll_wc(struct ti_qspi *qspi)
+{
+	u32 stat;
+	unsigned long timeout = jiffies + QSPI_COMPLETION_TIMEOUT;
+
+	do {
+		stat = ti_qspi_read(qspi, QSPI_SPI_STATUS_REG);
+		if (stat & WC)
+			return 0;
+		cpu_relax();
+	} while (time_after(timeout, jiffies));
+
+	stat = ti_qspi_read(qspi, QSPI_SPI_STATUS_REG);
+	if (stat & WC)
+		return 0;
+	return  -ETIMEDOUT;
+}
+
 static int qspi_write_msg(struct ti_qspi *qspi, struct spi_transfer *t)
 {
 	int wlen, count, xfer_len;
@@ -275,8 +282,7 @@ static int qspi_write_msg(struct ti_qspi *qspi, struct spi_transfer *t)
 		}
 
 		ti_qspi_write(qspi, cmd, QSPI_SPI_CMD_REG);
-		if (!wait_for_completion_timeout(&qspi->transfer_complete,
-						 QSPI_COMPLETION_TIMEOUT)) {
+		if (ti_qspi_poll_wc(qspi)) {
 			dev_err(qspi->dev, "write timed out\n");
 			return -ETIMEDOUT;
 		}
@@ -315,8 +321,7 @@ static int qspi_read_msg(struct ti_qspi *qspi, struct spi_transfer *t)
 			return -EBUSY;
 
 		ti_qspi_write(qspi, cmd, QSPI_SPI_CMD_REG);
-		if (!wait_for_completion_timeout(&qspi->transfer_complete,
-						 QSPI_COMPLETION_TIMEOUT)) {
+		if (ti_qspi_poll_wc(qspi)) {
 			dev_err(qspi->dev, "read timed out\n");
 			return -ETIMEDOUT;
 		}
@@ -388,9 +393,7 @@ static int ti_qspi_start_transfer_one(struct spi_master *master,
 	qspi->cmd = 0;
 	qspi->cmd |= QSPI_EN_CS(spi->chip_select);
 	qspi->cmd |= QSPI_FLEN(frame_length);
-	qspi->cmd |= QSPI_WC_CMD_INT_EN;
 
-	ti_qspi_write(qspi, QSPI_WC_INT_EN, QSPI_INTR_ENABLE_SET_REG);
 	ti_qspi_write(qspi, qspi->dc, QSPI_SPI_DC_REG);
 
 	mutex_lock(&qspi->list_lock);
@@ -410,39 +413,13 @@ static int ti_qspi_start_transfer_one(struct spi_master *master,
 
 	mutex_unlock(&qspi->list_lock);
 
+	ti_qspi_write(qspi, qspi->cmd | QSPI_INVAL, QSPI_SPI_CMD_REG);
 	m->status = status;
 	spi_finalize_current_message(master);
 
-	ti_qspi_write(qspi, qspi->cmd | QSPI_INVAL, QSPI_SPI_CMD_REG);
-
 	return status;
 }
 
-static irqreturn_t ti_qspi_isr(int irq, void *dev_id)
-{
-	struct ti_qspi *qspi = dev_id;
-	u16 int_stat;
-	u32 stat;
-
-	irqreturn_t ret = IRQ_HANDLED;
-
-	int_stat = ti_qspi_read(qspi, QSPI_INTR_STATUS_ENABLED_CLEAR);
-	stat = ti_qspi_read(qspi, QSPI_SPI_STATUS_REG);
-
-	if (!int_stat) {
-		dev_dbg(qspi->dev, "No IRQ triggered\n");
-		ret = IRQ_NONE;
-		goto out;
-	}
-
-	ti_qspi_write(qspi, QSPI_WC_INT_DISABLE,
-				QSPI_INTR_STATUS_ENABLED_CLEAR);
-	if (stat & WC)
-		complete(&qspi->transfer_complete);
-out:
-	return ret;
-}
-
 static int ti_qspi_runtime_resume(struct device *dev)
 {
 	struct ti_qspi      *qspi;
@@ -551,22 +528,12 @@ static int ti_qspi_probe(struct platform_device *pdev)
 		}
 	}
 
-	ret = devm_request_irq(&pdev->dev, irq, ti_qspi_isr, 0,
-			dev_name(&pdev->dev), qspi);
-	if (ret < 0) {
-		dev_err(&pdev->dev, "Failed to register ISR for IRQ %d\n",
-				irq);
-		goto free_master;
-	}
-
 	qspi->fclk = devm_clk_get(&pdev->dev, "fck");
 	if (IS_ERR(qspi->fclk)) {
 		ret = PTR_ERR(qspi->fclk);
 		dev_err(&pdev->dev, "could not get clk: %d\n", ret);
 	}
 
-	init_completion(&qspi->transfer_complete);
-
 	pm_runtime_use_autosuspend(&pdev->dev);
 	pm_runtime_set_autosuspend_delay(&pdev->dev, QSPI_AUTOSUSPEND_TIMEOUT);
 	pm_runtime_enable(&pdev->dev);
@@ -587,18 +554,7 @@ free_master:
 
 static int ti_qspi_remove(struct platform_device *pdev)
 {
-	struct ti_qspi *qspi = platform_get_drvdata(pdev);
-	int ret;
-
-	ret = pm_runtime_get_sync(qspi->dev);
-	if (ret < 0) {
-		dev_err(qspi->dev, "pm_runtime_get_sync() failed\n");
-		return ret;
-	}
-
-	ti_qspi_write(qspi, QSPI_WC_INT_DISABLE, QSPI_INTR_ENABLE_CLEAR_REG);
-
-	pm_runtime_put(qspi->dev);
+	pm_runtime_put_sync(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
 
 	return 0;
diff --git a/drivers/spi/spi-tle62x0.c b/drivers/spi/spi-tle62x0.c
index daf5aa1c24c3..c6ae775289e5 100644
--- a/drivers/spi/spi-tle62x0.c
+++ b/drivers/spi/spi-tle62x0.c
@@ -307,7 +307,6 @@ static int tle62x0_remove(struct spi_device *spi)
 static struct spi_driver tle62x0_driver = {
 	.driver = {
 		.name	= "tle62x0",
-		.owner	= THIS_MODULE,
 	},
 	.probe		= tle62x0_probe,
 	.remove		= tle62x0_remove,
diff --git a/drivers/spi/spi-txx9.c b/drivers/spi/spi-txx9.c
index 9190124b6d90..d69f8f8f3fa6 100644
--- a/drivers/spi/spi-txx9.c
+++ b/drivers/spi/spi-txx9.c
@@ -181,7 +181,7 @@ static void txx9spi_work_one(struct txx9spi *c, struct spi_message *m)
 		u32 data;
 		unsigned int len = t->len;
 		unsigned int wsize;
-		u32 speed_hz = t->speed_hz ? : spi->max_speed_hz;
+		u32 speed_hz = t->speed_hz;
 		u8 bits_per_word = t->bits_per_word;
 
 		wsize = bits_per_word >> 3; /* in bytes */
diff --git a/drivers/spi/spi-xilinx.c b/drivers/spi/spi-xilinx.c
index a339c1e9997a..3009121173cd 100644
--- a/drivers/spi/spi-xilinx.c
+++ b/drivers/spi/spi-xilinx.c
@@ -270,6 +270,7 @@ static int xilinx_spi_txrx_bufs(struct spi_device *spi, struct spi_transfer *t)
 
 	while (remaining_words) {
 		int n_words, tx_words, rx_words;
+		u32 sr;
 
 		n_words = min(remaining_words, xspi->buffer_size);
 
@@ -284,24 +285,33 @@ static int xilinx_spi_txrx_bufs(struct spi_device *spi, struct spi_transfer *t)
 		if (use_irq) {
 			xspi->write_fn(cr, xspi->regs + XSPI_CR_OFFSET);
 			wait_for_completion(&xspi->done);
-		} else
-			while (!(xspi->read_fn(xspi->regs + XSPI_SR_OFFSET) &
-						XSPI_SR_TX_EMPTY_MASK))
-				;
-
-		/* A transmit has just completed. Process received data and
-		 * check for more data to transmit. Always inhibit the
-		 * transmitter while the Isr refills the transmit register/FIFO,
-		 * or make sure it is stopped if we're done.
-		 */
-		if (use_irq)
+			/* A transmit has just completed. Process received data
+			 * and check for more data to transmit. Always inhibit
+			 * the transmitter while the Isr refills the transmit
+			 * register/FIFO, or make sure it is stopped if we're
+			 * done.
+			 */
 			xspi->write_fn(cr | XSPI_CR_TRANS_INHIBIT,
-			       xspi->regs + XSPI_CR_OFFSET);
+				       xspi->regs + XSPI_CR_OFFSET);
+			sr = XSPI_SR_TX_EMPTY_MASK;
+		} else
+			sr = xspi->read_fn(xspi->regs + XSPI_SR_OFFSET);
 
 		/* Read out all the data from the Rx FIFO */
 		rx_words = n_words;
-		while (rx_words--)
-			xilinx_spi_rx(xspi);
+		while (rx_words) {
+			if ((sr & XSPI_SR_TX_EMPTY_MASK) && (rx_words > 1)) {
+				xilinx_spi_rx(xspi);
+				rx_words--;
+				continue;
+			}
+
+			sr = xspi->read_fn(xspi->regs + XSPI_SR_OFFSET);
+			if (!(sr & XSPI_SR_RX_EMPTY_MASK)) {
+				xilinx_spi_rx(xspi);
+				rx_words--;
+			}
+		}
 
 		remaining_words -= n_words;
 	}
diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
index a5f53de813d3..e2415be209d5 100644
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -123,6 +123,28 @@ SPI_STATISTICS_SHOW(bytes, "%llu");
 SPI_STATISTICS_SHOW(bytes_rx, "%llu");
 SPI_STATISTICS_SHOW(bytes_tx, "%llu");
 
+#define SPI_STATISTICS_TRANSFER_BYTES_HISTO(index, number)		\
+	SPI_STATISTICS_SHOW_NAME(transfer_bytes_histo##index,		\
+				 "transfer_bytes_histo_" number,	\
+				 transfer_bytes_histo[index],  "%lu")
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(0,  "0-1");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(1,  "2-3");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(2,  "4-7");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(3,  "8-15");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(4,  "16-31");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(5,  "32-63");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(6,  "64-127");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(7,  "128-255");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(8,  "256-511");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(9,  "512-1023");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(10, "1024-2047");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(11, "2048-4095");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(12, "4096-8191");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(13, "8192-16383");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(14, "16384-32767");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(15, "32768-65535");
+SPI_STATISTICS_TRANSFER_BYTES_HISTO(16, "65536+");
+
 static struct attribute *spi_dev_attrs[] = {
 	&dev_attr_modalias.attr,
 	NULL,
@@ -143,6 +165,23 @@ static struct attribute *spi_device_statistics_attrs[] = {
 	&dev_attr_spi_device_bytes.attr,
 	&dev_attr_spi_device_bytes_rx.attr,
 	&dev_attr_spi_device_bytes_tx.attr,
+	&dev_attr_spi_device_transfer_bytes_histo0.attr,
+	&dev_attr_spi_device_transfer_bytes_histo1.attr,
+	&dev_attr_spi_device_transfer_bytes_histo2.attr,
+	&dev_attr_spi_device_transfer_bytes_histo3.attr,
+	&dev_attr_spi_device_transfer_bytes_histo4.attr,
+	&dev_attr_spi_device_transfer_bytes_histo5.attr,
+	&dev_attr_spi_device_transfer_bytes_histo6.attr,
+	&dev_attr_spi_device_transfer_bytes_histo7.attr,
+	&dev_attr_spi_device_transfer_bytes_histo8.attr,
+	&dev_attr_spi_device_transfer_bytes_histo9.attr,
+	&dev_attr_spi_device_transfer_bytes_histo10.attr,
+	&dev_attr_spi_device_transfer_bytes_histo11.attr,
+	&dev_attr_spi_device_transfer_bytes_histo12.attr,
+	&dev_attr_spi_device_transfer_bytes_histo13.attr,
+	&dev_attr_spi_device_transfer_bytes_histo14.attr,
+	&dev_attr_spi_device_transfer_bytes_histo15.attr,
+	&dev_attr_spi_device_transfer_bytes_histo16.attr,
 	NULL,
 };
 
@@ -168,6 +207,23 @@ static struct attribute *spi_master_statistics_attrs[] = {
 	&dev_attr_spi_master_bytes.attr,
 	&dev_attr_spi_master_bytes_rx.attr,
 	&dev_attr_spi_master_bytes_tx.attr,
+	&dev_attr_spi_master_transfer_bytes_histo0.attr,
+	&dev_attr_spi_master_transfer_bytes_histo1.attr,
+	&dev_attr_spi_master_transfer_bytes_histo2.attr,
+	&dev_attr_spi_master_transfer_bytes_histo3.attr,
+	&dev_attr_spi_master_transfer_bytes_histo4.attr,
+	&dev_attr_spi_master_transfer_bytes_histo5.attr,
+	&dev_attr_spi_master_transfer_bytes_histo6.attr,
+	&dev_attr_spi_master_transfer_bytes_histo7.attr,
+	&dev_attr_spi_master_transfer_bytes_histo8.attr,
+	&dev_attr_spi_master_transfer_bytes_histo9.attr,
+	&dev_attr_spi_master_transfer_bytes_histo10.attr,
+	&dev_attr_spi_master_transfer_bytes_histo11.attr,
+	&dev_attr_spi_master_transfer_bytes_histo12.attr,
+	&dev_attr_spi_master_transfer_bytes_histo13.attr,
+	&dev_attr_spi_master_transfer_bytes_histo14.attr,
+	&dev_attr_spi_master_transfer_bytes_histo15.attr,
+	&dev_attr_spi_master_transfer_bytes_histo16.attr,
 	NULL,
 };
 
@@ -186,10 +242,15 @@ void spi_statistics_add_transfer_stats(struct spi_statistics *stats,
 				       struct spi_master *master)
 {
 	unsigned long flags;
+	int l2len = min(fls(xfer->len), SPI_STATISTICS_HISTO_SIZE) - 1;
+
+	if (l2len < 0)
+		l2len = 0;
 
 	spin_lock_irqsave(&stats->lock, flags);
 
 	stats->transfers++;
+	stats->transfer_bytes_histo[l2len]++;
 
 	stats->bytes += xfer->len;
 	if ((xfer->tx_buf) &&
@@ -270,15 +331,24 @@ EXPORT_SYMBOL_GPL(spi_bus_type);
 static int spi_drv_probe(struct device *dev)
 {
 	const struct spi_driver		*sdrv = to_spi_driver(dev->driver);
+	struct spi_device		*spi = to_spi_device(dev);
 	int ret;
 
 	ret = of_clk_set_defaults(dev->of_node, false);
 	if (ret)
 		return ret;
 
+	if (dev->of_node) {
+		spi->irq = of_irq_get(dev->of_node, 0);
+		if (spi->irq == -EPROBE_DEFER)
+			return -EPROBE_DEFER;
+		if (spi->irq < 0)
+			spi->irq = 0;
+	}
+
 	ret = dev_pm_domain_attach(dev, true);
 	if (ret != -EPROBE_DEFER) {
-		ret = sdrv->probe(to_spi_device(dev));
+		ret = sdrv->probe(spi);
 		if (ret)
 			dev_pm_domain_detach(dev, true);
 	}
@@ -305,12 +375,15 @@ static void spi_drv_shutdown(struct device *dev)
 }
 
 /**
- * spi_register_driver - register a SPI driver
+ * __spi_register_driver - register a SPI driver
  * @sdrv: the driver to register
  * Context: can sleep
+ *
+ * Return: zero on success, else a negative error code.
  */
-int spi_register_driver(struct spi_driver *sdrv)
+int __spi_register_driver(struct module *owner, struct spi_driver *sdrv)
 {
+	sdrv->driver.owner = owner;
 	sdrv->driver.bus = &spi_bus_type;
 	if (sdrv->probe)
 		sdrv->driver.probe = spi_drv_probe;
@@ -320,7 +393,7 @@ int spi_register_driver(struct spi_driver *sdrv)
 		sdrv->driver.shutdown = spi_drv_shutdown;
 	return driver_register(&sdrv->driver);
 }
-EXPORT_SYMBOL_GPL(spi_register_driver);
+EXPORT_SYMBOL_GPL(__spi_register_driver);
 
 /*-------------------------------------------------------------------------*/
 
@@ -359,7 +432,7 @@ static DEFINE_MUTEX(board_lock);
  * needs to discard the spi_device without adding it, then it should
  * call spi_dev_put() on it.
  *
- * Returns a pointer to the new device, or NULL.
+ * Return: a pointer to the new device, or NULL.
  */
 struct spi_device *spi_alloc_device(struct spi_master *master)
 {
@@ -418,7 +491,7 @@ static int spi_dev_check(struct device *dev, void *data)
  * Companion function to spi_alloc_device.  Devices allocated with
  * spi_alloc_device can be added onto the spi bus with this function.
  *
- * Returns 0 on success; negative errno on failure
+ * Return: 0 on success; negative errno on failure
  */
 int spi_add_device(struct spi_device *spi)
 {
@@ -491,7 +564,7 @@ EXPORT_SYMBOL_GPL(spi_add_device);
  * this is exported so that for example a USB or parport based adapter
  * driver could add devices (which it would learn about out-of-band).
  *
- * Returns the new device, or NULL.
+ * Return: the new device, or NULL.
  */
 struct spi_device *spi_new_device(struct spi_master *master,
 				  struct spi_board_info *chip)
@@ -563,6 +636,8 @@ static void spi_match_master_to_boardinfo(struct spi_master *master,
  *
  * The board info passed can safely be __initdata ... but be careful of
  * any embedded pointers (platform_data, etc), they're copied as-is.
+ *
+ * Return: zero on success, else a negative error code.
  */
 int spi_register_board_info(struct spi_board_info const *info, unsigned n)
 {
@@ -597,7 +672,7 @@ static void spi_set_cs(struct spi_device *spi, bool enable)
 	if (spi->mode & SPI_CS_HIGH)
 		enable = !enable;
 
-	if (spi->cs_gpio >= 0)
+	if (gpio_is_valid(spi->cs_gpio))
 		gpio_set_value(spi->cs_gpio, !enable);
 	else if (spi->master->set_cs)
 		spi->master->set_cs(spi, !enable);
@@ -1140,6 +1215,8 @@ static int spi_init_queue(struct spi_master *master)
  *
  * If there are more messages in the queue, the next message is returned from
  * this call.
+ *
+ * Return: the next message in the queue, else NULL if the queue is empty.
  */
 struct spi_message *spi_get_next_queued_message(struct spi_master *master)
 {
@@ -1303,6 +1380,8 @@ static int __spi_queued_transfer(struct spi_device *spi,
  * spi_queued_transfer - transfer function for queued transfers
  * @spi: spi device which is requesting transfer
  * @msg: spi message which is to handled is queued to driver queue
+ *
+ * Return: zero on success, else a negative error code.
  */
 static int spi_queued_transfer(struct spi_device *spi, struct spi_message *msg)
 {
@@ -1433,9 +1512,6 @@ of_register_spi_device(struct spi_master *master, struct device_node *nc)
 	}
 	spi->max_speed_hz = value;
 
-	/* IRQ */
-	spi->irq = irq_of_parse_and_map(nc, 0);
-
 	/* Store a pointer to the node in the device structure */
 	of_node_get(nc);
 	spi->dev.of_node = nc;
@@ -1605,12 +1681,13 @@ static struct class spi_master_class = {
  * only ones directly touching chip registers.  It's how they allocate
  * an spi_master structure, prior to calling spi_register_master().
  *
- * This must be called from context that can sleep.  It returns the SPI
- * master structure on success, else NULL.
+ * This must be called from context that can sleep.
  *
  * The caller is responsible for assigning the bus number and initializing
  * the master's methods before calling spi_register_master(); and (after errors
  * adding the device) calling spi_master_put() to prevent a memory leak.
+ *
+ * Return: the SPI master structure on success, else NULL.
  */
 struct spi_master *spi_alloc_master(struct device *dev, unsigned size)
 {
@@ -1694,6 +1771,8 @@ static int of_spi_register_master(struct spi_master *master)
  * success, else a negative error code (dropping the master's refcount).
  * After a successful return, the caller is responsible for calling
  * spi_unregister_master().
+ *
+ * Return: zero on success, else a negative error code.
  */
 int spi_register_master(struct spi_master *master)
 {
@@ -1787,6 +1866,8 @@ static void devm_spi_unregister(struct device *dev, void *res)
  *
  * Register a SPI device as with spi_register_master() which will
  * automatically be unregister
+ *
+ * Return: zero on success, else a negative error code.
  */
 int devm_spi_register_master(struct device *dev, struct spi_master *master)
 {
@@ -1892,6 +1973,8 @@ static int __spi_master_match(struct device *dev, const void *data)
  * arch init time.  It returns a refcounted pointer to the relevant
  * spi_master (which the caller must release), or NULL if there is
  * no such master registered.
+ *
+ * Return: the SPI master structure on success, else NULL.
  */
 struct spi_master *spi_busnum_to_master(u16 bus_num)
 {
@@ -1945,11 +2028,13 @@ static int __spi_validate_bits_per_word(struct spi_master *master, u8 bits_per_w
  * that the underlying controller or its driver does not support.  For
  * example, not all hardware supports wire transfers using nine bit words,
  * LSB-first wire encoding, or active-high chipselects.
+ *
+ * Return: zero on success, else a negative error code.
  */
 int spi_setup(struct spi_device *spi)
 {
 	unsigned	bad_bits, ugly_bits;
-	int		status = 0;
+	int		status;
 
 	/* check mode to prevent that DUAL and QUAD set at the same time
 	 */
@@ -1986,17 +2071,18 @@ int spi_setup(struct spi_device *spi)
 	if (!spi->bits_per_word)
 		spi->bits_per_word = 8;
 
-	if (__spi_validate_bits_per_word(spi->master, spi->bits_per_word))
-		return -EINVAL;
+	status = __spi_validate_bits_per_word(spi->master, spi->bits_per_word);
+	if (status)
+		return status;
 
 	if (!spi->max_speed_hz)
 		spi->max_speed_hz = spi->master->max_speed_hz;
 
-	spi_set_cs(spi, false);
-
 	if (spi->master->setup)
 		status = spi->master->setup(spi);
 
+	spi_set_cs(spi, false);
+
 	dev_dbg(&spi->dev, "setup mode %d, %s%s%s%s%u bits/w, %u Hz max --> %d\n",
 			(int) (spi->mode & (SPI_CPOL | SPI_CPHA)),
 			(spi->mode & SPI_CS_HIGH) ? "cs_high, " : "",
@@ -2162,6 +2248,8 @@ static int __spi_async(struct spi_device *spi, struct spi_message *message)
  * no other spi_message queued to that device will be processed.
  * (This rule applies equally to all the synchronous transfer calls,
  * which are wrappers around this core asynchronous primitive.)
+ *
+ * Return: zero on success, else a negative error code.
  */
 int spi_async(struct spi_device *spi, struct spi_message *message)
 {
@@ -2214,6 +2302,8 @@ EXPORT_SYMBOL_GPL(spi_async);
  * no other spi_message queued to that device will be processed.
  * (This rule applies equally to all the synchronous transfer calls,
  * which are wrappers around this core asynchronous primitive.)
+ *
+ * Return: zero on success, else a negative error code.
  */
 int spi_async_locked(struct spi_device *spi, struct spi_message *message)
 {
@@ -2329,7 +2419,7 @@ static int __spi_sync(struct spi_device *spi, struct spi_message *message,
  * Also, the caller is guaranteeing that the memory associated with the
  * message will not be freed before this call returns.
  *
- * It returns zero on success, else a negative error code.
+ * Return: zero on success, else a negative error code.
  */
 int spi_sync(struct spi_device *spi, struct spi_message *message)
 {
@@ -2351,7 +2441,7 @@ EXPORT_SYMBOL_GPL(spi_sync);
  * SPI bus. It has to be preceded by a spi_bus_lock call. The SPI bus must
  * be released by a spi_bus_unlock call when the exclusive access is over.
  *
- * It returns zero on success, else a negative error code.
+ * Return: zero on success, else a negative error code.
  */
 int spi_sync_locked(struct spi_device *spi, struct spi_message *message)
 {
@@ -2372,7 +2462,7 @@ EXPORT_SYMBOL_GPL(spi_sync_locked);
  * exclusive access is over. Data transfer must be done by spi_sync_locked
  * and spi_async_locked calls when the SPI bus lock is held.
  *
- * It returns zero on success, else a negative error code.
+ * Return: always zero.
  */
 int spi_bus_lock(struct spi_master *master)
 {
@@ -2401,7 +2491,7 @@ EXPORT_SYMBOL_GPL(spi_bus_lock);
  * This call releases an SPI bus lock previously obtained by an spi_bus_lock
  * call.
  *
- * It returns zero on success, else a negative error code.
+ * Return: always zero.
  */
 int spi_bus_unlock(struct spi_master *master)
 {
@@ -2436,6 +2526,8 @@ static u8	*buf;
  * portable code should never use this for more than 32 bytes.
  * Performance-sensitive or bulk transfer code should instead use
  * spi_{async,sync}() calls with dma-safe buffers.
+ *
+ * Return: zero on success, else a negative error code.
  */
 int spi_write_then_read(struct spi_device *spi,
 		const void *txbuf, unsigned n_tx,
diff --git a/drivers/spi/spidev.c b/drivers/spi/spidev.c
index ef008e52f953..91a0fcd72423 100644
--- a/drivers/spi/spidev.c
+++ b/drivers/spi/spidev.c
@@ -788,7 +788,6 @@ static int spidev_remove(struct spi_device *spi)
 static struct spi_driver spidev_spi_driver = {
 	.driver = {
 		.name =		"spidev",
-		.owner =	THIS_MODULE,
 		.of_match_table = of_match_ptr(spidev_dt_ids),
 	},
 	.probe =	spidev_probe,