Merge tag 'spi-v4.5' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi

Pull spi updates from Mark Brown:
 "A quiet release for SPI, not even many driver updates:

   - Add a dummy loopback driver for use in exercising framework
     features during development.

   - Move the test utilities to tools/ and add support for transferring
     data to and from a file instead of stdin and stdout to spidev_test.

   - Support for Mediatek MT2701 and Renesas AG5 deices"

* tag 'spi-v4.5' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi: (69 commits)
  spi: loopback: fix typo in MODULE_PARM_DESC
  spi: sun4i: Prevent chip-select from being activated twice before a transfer
  spi: loopback-test: spi_check_rx_ranges can get always done
  spi: loopback-test: rename method spi_test_fill_tx to spi_test_fill_pattern
  spi: loopback-test: write rx pattern also when running without tx_buf
  spi: fsl-espi: expose maximum transfer size limit
  spi: expose master transfer size limitation.
  spi: zynq: use to_platform_device()
  spi: cadence: use to_platform_device()
  spi: mediatek: Add spi support for mt2701 IC
  spi: mediatek: merge all identical compat to mtk_common_compat
  spi: mtk: Add bindings for mediatek MT2701 soc platform
  spi: mediatek: Prevent overflows in FIFO transfers
  spi: s3c64xx: Remove unused platform_device_id entries
  spi: use to_spi_device
  spi: dw: Use SPI_TMOD_TR rather than magic const 0 to set tmode
  spi: imx: defer spi initialization, if DMA engine is
  spi: imx: return error from dma channel request
  spi: imx: enable loopback only for ECSPI controller family
  spi: imx: fix loopback mode setup after controller reset
  ...

36 files changed, 1536 insertions, 232 deletions

diff --git a/Documentation/Makefile b/Documentation/Makefile
index bc0548201755..1207d7907650 100644
--- a/Documentation/Makefile
+++ b/Documentation/Makefile
@@ -1,4 +1,4 @@
 subdir-y := accounting auxdisplay blackfin connector \
 	filesystems filesystems ia64 laptops mic misc-devices \
-	networking pcmcia prctl ptp spi timers vDSO video4linux \
+	networking pcmcia prctl ptp timers vDSO video4linux \
 	watchdog
diff --git a/Documentation/devicetree/bindings/spi/sh-msiof.txt b/Documentation/devicetree/bindings/spi/sh-msiof.txt
index 705075da2f10..aa005c1d10d9 100644
--- a/Documentation/devicetree/bindings/spi/sh-msiof.txt
+++ b/Documentation/devicetree/bindings/spi/sh-msiof.txt
@@ -10,6 +10,7 @@ Required properties:
 			 "renesas,msiof-r8a7792" (R-Car V2H)
 			 "renesas,msiof-r8a7793" (R-Car M2-N)
 			 "renesas,msiof-r8a7794" (R-Car E2)
+			 "renesas,msiof-sh73a0" (SH-Mobile AG5)
 - reg                  : A list of offsets and lengths of the register sets for
 			 the device.
 			 If only one register set is present, it is to be used
diff --git a/Documentation/devicetree/bindings/spi/spi-mt65xx.txt b/Documentation/devicetree/bindings/spi/spi-mt65xx.txt
index ce363c923f44..e43f4cf4cf35 100644
--- a/Documentation/devicetree/bindings/spi/spi-mt65xx.txt
+++ b/Documentation/devicetree/bindings/spi/spi-mt65xx.txt
@@ -2,9 +2,10 @@ Binding for MTK SPI controller
 
 Required properties:
 - compatible: should be one of the following.
-    - mediatek,mt8173-spi: for mt8173 platforms
-    - mediatek,mt8135-spi: for mt8135 platforms
+    - mediatek,mt2701-spi: for mt2701 platforms
     - mediatek,mt6589-spi: for mt6589 platforms
+    - mediatek,mt8135-spi: for mt8135 platforms
+    - mediatek,mt8173-spi: for mt8173 platforms
 
 - #address-cells: should be 1.
 
@@ -29,10 +30,10 @@ Required properties:
   muxes clock, and "spi-clk" for the clock gate.
 
 Optional properties:
--cs-gpios: see spi-bus.txt, only required for MT8173.
+-cs-gpios: see spi-bus.txt.
 
 - mediatek,pad-select: specify which pins group(ck/mi/mo/cs) spi
-  controller used. This is a array, the element value should be 0~3,
+  controller used. This is an array, the element value should be 0~3,
   only required for MT8173.
     0: specify GPIO69,70,71,72 for spi pins.
     1: specify GPIO102,103,104,105 for spi pins.
diff --git a/Documentation/spi/00-INDEX b/Documentation/spi/00-INDEX
index a128fa835512..4644bf0d9832 100644
--- a/Documentation/spi/00-INDEX
+++ b/Documentation/spi/00-INDEX
@@ -10,13 +10,9 @@ pxa2xx
 	- PXA2xx SPI master controller build by spi_message fifo wq
 spidev
 	- Intro to the userspace API for spi devices
-spidev_fdx.c
-	- spidev example file
 spi-lm70llp
 	- Connecting an LM70-LLP sensor to the kernel via the SPI subsys.
 spi-sc18is602
 	- NXP SC18IS602/603 I2C-bus to SPI bridge
 spi-summary
 	- (Linux) SPI overview. If unsure about SPI or SPI in Linux, start here.
-spidev_test.c
-	- SPI testing utility.
diff --git a/Documentation/spi/Makefile b/Documentation/spi/Makefile
deleted file mode 100644
index efa255813e9d..000000000000
--- a/Documentation/spi/Makefile
+++ /dev/null
@@ -1,8 +0,0 @@
-# List of programs to build
-hostprogs-y := spidev_test spidev_fdx
-
-# Tell kbuild to always build the programs
-always := $(hostprogs-y)
-
-HOSTCFLAGS_spidev_test.o += -I$(objtree)/usr/include
-HOSTCFLAGS_spidev_fdx.o += -I$(objtree)/usr/include
diff --git a/arch/arm/plat-samsung/devs.c b/arch/arm/plat-samsung/devs.c
index 74ef8891254e..42f6f9cd844c 100644
--- a/arch/arm/plat-samsung/devs.c
+++ b/arch/arm/plat-samsung/devs.c
@@ -1100,9 +1100,7 @@ struct platform_device s3c_device_wdt = {
 #ifdef CONFIG_S3C64XX_DEV_SPI0
 static struct resource s3c64xx_spi0_resource[] = {
 	[0] = DEFINE_RES_MEM(S3C_PA_SPI0, SZ_256),
-	[1] = DEFINE_RES_DMA(DMACH_SPI0_TX),
-	[2] = DEFINE_RES_DMA(DMACH_SPI0_RX),
-	[3] = DEFINE_RES_IRQ(IRQ_SPI0),
+	[1] = DEFINE_RES_IRQ(IRQ_SPI0),
 };
 
 struct platform_device s3c64xx_device_spi0 = {
@@ -1130,6 +1128,8 @@ void __init s3c64xx_spi0_set_platdata(int (*cfg_gpio)(void), int src_clk_nr,
 	pd.num_cs = num_cs;
 	pd.src_clk_nr = src_clk_nr;
 	pd.cfg_gpio = (cfg_gpio) ? cfg_gpio : s3c64xx_spi0_cfg_gpio;
+	pd.dma_tx = (void *)DMACH_SPI0_TX;
+	pd.dma_rx = (void *)DMACH_SPI0_RX;
 #if defined(CONFIG_PL330_DMA)
 	pd.filter = pl330_filter;
 #elif defined(CONFIG_S3C64XX_PL080)
@@ -1145,9 +1145,7 @@ void __init s3c64xx_spi0_set_platdata(int (*cfg_gpio)(void), int src_clk_nr,
 #ifdef CONFIG_S3C64XX_DEV_SPI1
 static struct resource s3c64xx_spi1_resource[] = {
 	[0] = DEFINE_RES_MEM(S3C_PA_SPI1, SZ_256),
-	[1] = DEFINE_RES_DMA(DMACH_SPI1_TX),
-	[2] = DEFINE_RES_DMA(DMACH_SPI1_RX),
-	[3] = DEFINE_RES_IRQ(IRQ_SPI1),
+	[1] = DEFINE_RES_IRQ(IRQ_SPI1),
 };
 
 struct platform_device s3c64xx_device_spi1 = {
@@ -1175,12 +1173,15 @@ void __init s3c64xx_spi1_set_platdata(int (*cfg_gpio)(void), int src_clk_nr,
 	pd.num_cs = num_cs;
 	pd.src_clk_nr = src_clk_nr;
 	pd.cfg_gpio = (cfg_gpio) ? cfg_gpio : s3c64xx_spi1_cfg_gpio;
+	pd.dma_tx = (void *)DMACH_SPI1_TX;
+	pd.dma_rx = (void *)DMACH_SPI1_RX;
 #if defined(CONFIG_PL330_DMA)
 	pd.filter = pl330_filter;
 #elif defined(CONFIG_S3C64XX_PL080)
 	pd.filter = pl08x_filter_id;
 #endif
 
+
 	s3c_set_platdata(&pd, sizeof(pd), &s3c64xx_device_spi1);
 }
 #endif /* CONFIG_S3C64XX_DEV_SPI1 */
@@ -1188,9 +1189,7 @@ void __init s3c64xx_spi1_set_platdata(int (*cfg_gpio)(void), int src_clk_nr,
 #ifdef CONFIG_S3C64XX_DEV_SPI2
 static struct resource s3c64xx_spi2_resource[] = {
 	[0] = DEFINE_RES_MEM(S3C_PA_SPI2, SZ_256),
-	[1] = DEFINE_RES_DMA(DMACH_SPI2_TX),
-	[2] = DEFINE_RES_DMA(DMACH_SPI2_RX),
-	[3] = DEFINE_RES_IRQ(IRQ_SPI2),
+	[1] = DEFINE_RES_IRQ(IRQ_SPI2),
 };
 
 struct platform_device s3c64xx_device_spi2 = {
@@ -1218,6 +1217,8 @@ void __init s3c64xx_spi2_set_platdata(int (*cfg_gpio)(void), int src_clk_nr,
 	pd.num_cs = num_cs;
 	pd.src_clk_nr = src_clk_nr;
 	pd.cfg_gpio = (cfg_gpio) ? cfg_gpio : s3c64xx_spi2_cfg_gpio;
+	pd.dma_tx = (void *)DMACH_SPI2_TX;
+	pd.dma_rx = (void *)DMACH_SPI2_RX;
 #if defined(CONFIG_PL330_DMA)
 	pd.filter = pl330_filter;
 #elif defined(CONFIG_S3C64XX_PL080)
diff --git a/drivers/spi/Kconfig b/drivers/spi/Kconfig
index 8b9c2a38d1cc..77064160dd76 100644
--- a/drivers/spi/Kconfig
+++ b/drivers/spi/Kconfig
@@ -585,7 +585,7 @@ config SPI_TEGRA20_SLINK
 
 config SPI_TOPCLIFF_PCH
 	tristate "Intel EG20T PCH/LAPIS Semicon IOH(ML7213/ML7223/ML7831) SPI"
-	depends on PCI && (X86_32 || COMPILE_TEST)
+	depends on PCI && (X86_32 || MIPS || COMPILE_TEST)
 	help
 	  SPI driver for the Topcliff PCH (Platform Controller Hub) SPI bus
 	  used in some x86 embedded processors.
@@ -689,6 +689,15 @@ config SPI_SPIDEV
 	  Note that this application programming interface is EXPERIMENTAL
 	  and hence SUBJECT TO CHANGE WITHOUT NOTICE while it stabilizes.
 
+config SPI_LOOPBACK_TEST
+	tristate "spi loopback test framework support"
+	depends on m
+	help
+	  This enables the SPI loopback testing framework driver
+
+	  primarily used for development of spi_master drivers
+	  and to detect regressions
+
 config SPI_TLE62X0
 	tristate "Infineon TLE62X0 (for power switching)"
 	depends on SYSFS
diff --git a/drivers/spi/Makefile b/drivers/spi/Makefile
index 31fb7fb2a0b6..8991ffce6e12 100644
--- a/drivers/spi/Makefile
+++ b/drivers/spi/Makefile
@@ -8,6 +8,7 @@ ccflags-$(CONFIG_SPI_DEBUG) := -DDEBUG
 # config declarations into driver model code
 obj-$(CONFIG_SPI_MASTER)		+= spi.o
 obj-$(CONFIG_SPI_SPIDEV)		+= spidev.o
+obj-$(CONFIG_SPI_LOOPBACK_TEST)		+= spi-loopback-test.o
 
 # SPI master controller drivers (bus)
 obj-$(CONFIG_SPI_ALTERA)		+= spi-altera.o
diff --git a/drivers/spi/spi-bcm63xx.c b/drivers/spi/spi-bcm63xx.c
index bf9a610e5b89..fee747030ee6 100644
--- a/drivers/spi/spi-bcm63xx.c
+++ b/drivers/spi/spi-bcm63xx.c
@@ -207,6 +207,9 @@ static void bcm63xx_spi_setup_transfer(struct spi_device *spi,
 	u8 clk_cfg, reg;
 	int i;
 
+	/* Default to lowest clock configuration */
+	clk_cfg = SPI_CLK_0_391MHZ;
+
 	/* Find the closest clock configuration */
 	for (i = 0; i < SPI_CLK_MASK; i++) {
 		if (t->speed_hz >= bcm63xx_spi_freq_table[i][0]) {
@@ -215,10 +218,6 @@ static void bcm63xx_spi_setup_transfer(struct spi_device *spi,
 		}
 	}
 
-	/* No matching configuration found, default to lowest */
-	if (i == SPI_CLK_MASK)
-		clk_cfg = SPI_CLK_0_391MHZ;
-
 	/* clear existing clock configuration bits of the register */
 	reg = bcm_spi_readb(bs, SPI_CLK_CFG);
 	reg &= ~SPI_CLK_MASK;
diff --git a/drivers/spi/spi-butterfly.c b/drivers/spi/spi-butterfly.c
index 9a95862986c8..22a31e4a1a11 100644
--- a/drivers/spi/spi-butterfly.c
+++ b/drivers/spi/spi-butterfly.c
@@ -27,7 +27,6 @@
 
 #include <linux/mtd/partitions.h>
 
-
 /*
  * This uses SPI to talk with an "AVR Butterfly", which is a $US20 card
  * with a battery powered AVR microcontroller and lots of goodies.  You
@@ -37,7 +36,6 @@
  * and use this custom parallel port cable.
  */
 
-
 /* DATA output bits (pins 2..9 == D0..D7) */
 #define	butterfly_nreset (1 << 1)		/* pin 3 */
 
@@ -52,14 +50,11 @@
 /* CONTROL output bits */
 #define	spi_cs_bit	PARPORT_CONTROL_SELECT	/* pin 17 */
 
-
-
 static inline struct butterfly *spidev_to_pp(struct spi_device *spi)
 {
 	return spi->controller_data;
 }
 
-
 struct butterfly {
 	/* REVISIT ... for now, this must be first */
 	struct spi_bitbang	bitbang;
@@ -140,7 +135,6 @@ static void butterfly_chipselect(struct spi_device *spi, int value)
 	parport_frob_control(pp->port, spi_cs_bit, value ? spi_cs_bit : 0);
 }
 
-
 /* we only needed to implement one mode here, and choose SPI_MODE_0 */
 
 #define spidelay(X)	do { } while (0)
@@ -149,9 +143,8 @@ static void butterfly_chipselect(struct spi_device *spi, int value)
 #include "spi-bitbang-txrx.h"
 
 static u32
-butterfly_txrx_word_mode0(struct spi_device *spi,
-		unsigned nsecs,
-		u32 word, u8 bits)
+butterfly_txrx_word_mode0(struct spi_device *spi, unsigned nsecs, u32 word,
+			  u8 bits)
 {
 	return bitbang_txrx_be_cpha0(spi, nsecs, 0, 0, word, bits);
 }
@@ -186,7 +179,6 @@ static struct flash_platform_data flash = {
 	.nr_parts	= ARRAY_SIZE(partitions),
 };
 
-
 /* REVISIT remove this ugly global and its "only one" limitation */
 static struct butterfly *butterfly;
 
@@ -197,6 +189,7 @@ static void butterfly_attach(struct parport *p)
 	struct butterfly	*pp;
 	struct spi_master	*master;
 	struct device		*dev = p->physport->dev;
+	struct pardev_cb	butterfly_cb;
 
 	if (butterfly || !dev)
 		return;
@@ -229,9 +222,9 @@ static void butterfly_attach(struct parport *p)
 	 * parport hookup
 	 */
 	pp->port = p;
-	pd = parport_register_device(p, "spi_butterfly",
-			NULL, NULL, NULL,
-			0 /* FLAGS */, pp);
+	memset(&butterfly_cb, 0, sizeof(butterfly_cb));
+	butterfly_cb.private = pp;
+	pd = parport_register_dev_model(p, "spi_butterfly", &butterfly_cb, 0);
 	if (!pd) {
 		status = -ENOMEM;
 		goto clean0;
@@ -262,7 +255,6 @@ static void butterfly_attach(struct parport *p)
 	parport_write_data(pp->port, pp->lastbyte);
 	msleep(100);
 
-
 	/*
 	 * Start SPI ... for now, hide that we're two physical busses.
 	 */
@@ -283,7 +275,7 @@ static void butterfly_attach(struct parport *p)
 	pp->dataflash = spi_new_device(pp->bitbang.master, &pp->info[0]);
 	if (pp->dataflash)
 		pr_debug("%s: dataflash at %s\n", p->name,
-				dev_name(&pp->dataflash->dev));
+			 dev_name(&pp->dataflash->dev));
 
 	pr_info("%s: AVR Butterfly\n", p->name);
 	butterfly = pp;
@@ -297,7 +289,7 @@ clean2:
 clean1:
 	parport_unregister_device(pd);
 clean0:
-	(void) spi_master_put(pp->bitbang.master);
+	spi_master_put(pp->bitbang.master);
 done:
 	pr_debug("%s: butterfly probe, fail %d\n", p->name, status);
 }
@@ -325,16 +317,16 @@ static void butterfly_detach(struct parport *p)
 	parport_release(pp->pd);
 	parport_unregister_device(pp->pd);
 
-	(void) spi_master_put(pp->bitbang.master);
+	spi_master_put(pp->bitbang.master);
 }
 
 static struct parport_driver butterfly_driver = {
 	.name =		"spi_butterfly",
-	.attach =	butterfly_attach,
+	.match_port =	butterfly_attach,
 	.detach =	butterfly_detach,
+	.devmodel = true,
 };
 
-
 static int __init butterfly_init(void)
 {
 	return parport_register_driver(&butterfly_driver);
diff --git a/drivers/spi/spi-cadence.c b/drivers/spi/spi-cadence.c
index 5a6749881ff9..121a4135b540 100644
--- a/drivers/spi/spi-cadence.c
+++ b/drivers/spi/spi-cadence.c
@@ -617,8 +617,7 @@ static int cdns_spi_remove(struct platform_device *pdev)
  */
 static int __maybe_unused cdns_spi_suspend(struct device *dev)
 {
-	struct platform_device *pdev = container_of(dev,
-			struct platform_device, dev);
+	struct platform_device *pdev = to_platform_device(dev);
 	struct spi_master *master = platform_get_drvdata(pdev);
 	struct cdns_spi *xspi = spi_master_get_devdata(master);
 
@@ -641,8 +640,7 @@ static int __maybe_unused cdns_spi_suspend(struct device *dev)
  */
 static int __maybe_unused cdns_spi_resume(struct device *dev)
 {
-	struct platform_device *pdev = container_of(dev,
-			struct platform_device, dev);
+	struct platform_device *pdev = to_platform_device(dev);
 	struct spi_master *master = platform_get_drvdata(pdev);
 	struct cdns_spi *xspi = spi_master_get_devdata(master);
 	int ret = 0;
diff --git a/drivers/spi/spi-davinci.c b/drivers/spi/spi-davinci.c
index 7d3af3eacf57..fddb7a3be322 100644
--- a/drivers/spi/spi-davinci.c
+++ b/drivers/spi/spi-davinci.c
@@ -477,33 +477,33 @@ static int davinci_spi_check_error(struct davinci_spi *dspi, int int_status)
 	struct device *sdev = dspi->bitbang.master->dev.parent;
 
 	if (int_status & SPIFLG_TIMEOUT_MASK) {
-		dev_dbg(sdev, "SPI Time-out Error\n");
+		dev_err(sdev, "SPI Time-out Error\n");
 		return -ETIMEDOUT;
 	}
 	if (int_status & SPIFLG_DESYNC_MASK) {
-		dev_dbg(sdev, "SPI Desynchronization Error\n");
+		dev_err(sdev, "SPI Desynchronization Error\n");
 		return -EIO;
 	}
 	if (int_status & SPIFLG_BITERR_MASK) {
-		dev_dbg(sdev, "SPI Bit error\n");
+		dev_err(sdev, "SPI Bit error\n");
 		return -EIO;
 	}
 
 	if (dspi->version == SPI_VERSION_2) {
 		if (int_status & SPIFLG_DLEN_ERR_MASK) {
-			dev_dbg(sdev, "SPI Data Length Error\n");
+			dev_err(sdev, "SPI Data Length Error\n");
 			return -EIO;
 		}
 		if (int_status & SPIFLG_PARERR_MASK) {
-			dev_dbg(sdev, "SPI Parity Error\n");
+			dev_err(sdev, "SPI Parity Error\n");
 			return -EIO;
 		}
 		if (int_status & SPIFLG_OVRRUN_MASK) {
-			dev_dbg(sdev, "SPI Data Overrun error\n");
+			dev_err(sdev, "SPI Data Overrun error\n");
 			return -EIO;
 		}
 		if (int_status & SPIFLG_BUF_INIT_ACTIVE_MASK) {
-			dev_dbg(sdev, "SPI Buffer Init Active\n");
+			dev_err(sdev, "SPI Buffer Init Active\n");
 			return -EBUSY;
 		}
 	}
@@ -703,7 +703,8 @@ static int davinci_spi_bufs(struct spi_device *spi, struct spi_transfer *t)
 
 	/* Wait for the transfer to complete */
 	if (spicfg->io_type != SPI_IO_TYPE_POLL) {
-		wait_for_completion_interruptible(&(dspi->done));
+		if (wait_for_completion_timeout(&dspi->done, HZ) == 0)
+			errors = SPIFLG_TIMEOUT_MASK;
 	} else {
 		while (dspi->rcount > 0 || dspi->wcount > 0) {
 			errors = davinci_spi_process_events(dspi);
diff --git a/drivers/spi/spi-dw-mid.c b/drivers/spi/spi-dw-mid.c
index bb1052e748f2..9185f6c08459 100644
--- a/drivers/spi/spi-dw-mid.c
+++ b/drivers/spi/spi-dw-mid.c
@@ -283,7 +283,7 @@ static void mid_spi_dma_stop(struct dw_spi *dws)
 	}
 }
 
-static struct dw_spi_dma_ops mid_dma_ops = {
+static const struct dw_spi_dma_ops mid_dma_ops = {
 	.dma_init	= mid_spi_dma_init,
 	.dma_exit	= mid_spi_dma_exit,
 	.dma_setup	= mid_spi_dma_setup,
diff --git a/drivers/spi/spi-dw.c b/drivers/spi/spi-dw.c
index 882cd6618cd5..c09bb745693a 100644
--- a/drivers/spi/spi-dw.c
+++ b/drivers/spi/spi-dw.c
@@ -425,7 +425,7 @@ static int dw_spi_setup(struct spi_device *spi)
 		chip->type = chip_info->type;
 	}
 
-	chip->tmode = 0; /* Tx & Rx */
+	chip->tmode = SPI_TMOD_TR;
 
 	if (gpio_is_valid(spi->cs_gpio)) {
 		ret = gpio_direction_output(spi->cs_gpio,
diff --git a/drivers/spi/spi-dw.h b/drivers/spi/spi-dw.h
index 35589a270468..61bc3cbab38d 100644
--- a/drivers/spi/spi-dw.h
+++ b/drivers/spi/spi-dw.h
@@ -130,7 +130,7 @@ struct dw_spi {
 	struct dma_chan		*rxchan;
 	unsigned long		dma_chan_busy;
 	dma_addr_t		dma_addr; /* phy address of the Data register */
-	struct dw_spi_dma_ops	*dma_ops;
+	const struct dw_spi_dma_ops *dma_ops;
 	void			*dma_tx;
 	void			*dma_rx;
 
diff --git a/drivers/spi/spi-fsl-espi.c b/drivers/spi/spi-fsl-espi.c
index c27124a5ec8e..7fd6a4c009d2 100644
--- a/drivers/spi/spi-fsl-espi.c
+++ b/drivers/spi/spi-fsl-espi.c
@@ -643,6 +643,11 @@ static int fsl_espi_runtime_resume(struct device *dev)
 }
 #endif
 
+static size_t fsl_espi_max_transfer_size(struct spi_device *spi)
+{
+	return SPCOM_TRANLEN_MAX;
+}
+
 static struct spi_master * fsl_espi_probe(struct device *dev,
 		struct resource *mem, unsigned int irq)
 {
@@ -670,6 +675,7 @@ static struct spi_master * fsl_espi_probe(struct device *dev,
 	master->cleanup = fsl_espi_cleanup;
 	master->transfer_one_message = fsl_espi_do_one_msg;
 	master->auto_runtime_pm = true;
+	master->max_transfer_size = fsl_espi_max_transfer_size;
 
 	mpc8xxx_spi = spi_master_get_devdata(master);
 
diff --git a/drivers/spi/spi-imx.c b/drivers/spi/spi-imx.c
index 0e5723ab47f0..d98c33cb64f9 100644
--- a/drivers/spi/spi-imx.c
+++ b/drivers/spi/spi-imx.c
@@ -104,9 +104,7 @@ struct spi_imx_data {
 	unsigned int dma_is_inited;
 	unsigned int dma_finished;
 	bool usedma;
-	u32 rx_wml;
-	u32 tx_wml;
-	u32 rxt_wml;
+	u32 wml;
 	struct completion dma_rx_completion;
 	struct completion dma_tx_completion;
 
@@ -124,9 +122,14 @@ static inline int is_imx35_cspi(struct spi_imx_data *d)
 	return d->devtype_data->devtype == IMX35_CSPI;
 }
 
+static inline int is_imx51_ecspi(struct spi_imx_data *d)
+{
+	return d->devtype_data->devtype == IMX51_ECSPI;
+}
+
 static inline unsigned spi_imx_get_fifosize(struct spi_imx_data *d)
 {
-	return (d->devtype_data->devtype == IMX51_ECSPI) ? 64 : 8;
+	return is_imx51_ecspi(d) ? 64 : 8;
 }
 
 #define MXC_SPI_BUF_RX(type)						\
@@ -201,9 +204,8 @@ static bool spi_imx_can_dma(struct spi_master *master, struct spi_device *spi,
 {
 	struct spi_imx_data *spi_imx = spi_master_get_devdata(master);
 
-	if (spi_imx->dma_is_inited
-	    && transfer->len > spi_imx->rx_wml * sizeof(u32)
-	    && transfer->len > spi_imx->tx_wml * sizeof(u32))
+	if (spi_imx->dma_is_inited &&
+	    transfer->len > spi_imx->wml * sizeof(u32))
 		return true;
 	return false;
 }
@@ -244,6 +246,9 @@ static bool spi_imx_can_dma(struct spi_master *master, struct spi_device *spi,
 #define MX51_ECSPI_STAT		0x18
 #define MX51_ECSPI_STAT_RR		(1 <<  3)
 
+#define MX51_ECSPI_TESTREG	0x20
+#define MX51_ECSPI_TESTREG_LBC	BIT(31)
+
 /* MX51 eCSPI */
 static unsigned int mx51_ecspi_clkdiv(unsigned int fin, unsigned int fspi,
 				      unsigned int *fres)
@@ -313,7 +318,7 @@ static int __maybe_unused mx51_ecspi_config(struct spi_imx_data *spi_imx,
 {
 	u32 ctrl = MX51_ECSPI_CTRL_ENABLE, cfg = 0, dma = 0;
 	u32 tx_wml_cfg, rx_wml_cfg, rxt_wml_cfg;
-	u32 clk = config->speed_hz, delay;
+	u32 clk = config->speed_hz, delay, reg;
 
 	/*
 	 * The hardware seems to have a race condition when changing modes. The
@@ -351,7 +356,16 @@ static int __maybe_unused mx51_ecspi_config(struct spi_imx_data *spi_imx,
 	else
 		cfg &= ~MX51_ECSPI_CONFIG_SSBPOL(config->cs);
 
+	/* CTRL register always go first to bring out controller from reset */
 	writel(ctrl, spi_imx->base + MX51_ECSPI_CTRL);
+
+	reg = readl(spi_imx->base + MX51_ECSPI_TESTREG);
+	if (config->mode & SPI_LOOP)
+		reg |= MX51_ECSPI_TESTREG_LBC;
+	else
+		reg &= ~MX51_ECSPI_TESTREG_LBC;
+	writel(reg, spi_imx->base + MX51_ECSPI_TESTREG);
+
 	writel(cfg, spi_imx->base + MX51_ECSPI_CONFIG);
 
 	/*
@@ -378,10 +392,9 @@ static int __maybe_unused mx51_ecspi_config(struct spi_imx_data *spi_imx,
 	if (spi_imx->dma_is_inited) {
 		dma = readl(spi_imx->base + MX51_ECSPI_DMA);
 
-		spi_imx->rxt_wml = spi_imx_get_fifosize(spi_imx) / 2;
-		rx_wml_cfg = spi_imx->rx_wml << MX51_ECSPI_DMA_RX_WML_OFFSET;
-		tx_wml_cfg = spi_imx->tx_wml << MX51_ECSPI_DMA_TX_WML_OFFSET;
-		rxt_wml_cfg = spi_imx->rxt_wml << MX51_ECSPI_DMA_RXT_WML_OFFSET;
+		rx_wml_cfg = spi_imx->wml << MX51_ECSPI_DMA_RX_WML_OFFSET;
+		tx_wml_cfg = spi_imx->wml << MX51_ECSPI_DMA_TX_WML_OFFSET;
+		rxt_wml_cfg = spi_imx->wml << MX51_ECSPI_DMA_RXT_WML_OFFSET;
 		dma = (dma & ~MX51_ECSPI_DMA_TX_WML_MASK
 			   & ~MX51_ECSPI_DMA_RX_WML_MASK
 			   & ~MX51_ECSPI_DMA_RXT_WML_MASK)
@@ -832,18 +845,21 @@ static int spi_imx_sdma_init(struct device *dev, struct spi_imx_data *spi_imx,
 	if (of_machine_is_compatible("fsl,imx6dl"))
 		return 0;
 
+	spi_imx->wml = spi_imx_get_fifosize(spi_imx) / 2;
+
 	/* Prepare for TX DMA: */
-	master->dma_tx = dma_request_slave_channel(dev, "tx");
-	if (!master->dma_tx) {
-		dev_err(dev, "cannot get the TX DMA channel!\n");
-		ret = -EINVAL;
+	master->dma_tx = dma_request_slave_channel_reason(dev, "tx");
+	if (IS_ERR(master->dma_tx)) {
+		ret = PTR_ERR(master->dma_tx);
+		dev_dbg(dev, "can't get the TX DMA channel, error %d!\n", ret);
+		master->dma_tx = NULL;
 		goto err;
 	}
 
 	slave_config.direction = DMA_MEM_TO_DEV;
 	slave_config.dst_addr = res->start + MXC_CSPITXDATA;
 	slave_config.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
-	slave_config.dst_maxburst = spi_imx_get_fifosize(spi_imx) / 2;
+	slave_config.dst_maxburst = spi_imx->wml;
 	ret = dmaengine_slave_config(master->dma_tx, &slave_config);
 	if (ret) {
 		dev_err(dev, "error in TX dma configuration.\n");
@@ -851,17 +867,18 @@ static int spi_imx_sdma_init(struct device *dev, struct spi_imx_data *spi_imx,
 	}
 
 	/* Prepare for RX : */
-	master->dma_rx = dma_request_slave_channel(dev, "rx");
-	if (!master->dma_rx) {
-		dev_dbg(dev, "cannot get the DMA channel.\n");
-		ret = -EINVAL;
+	master->dma_rx = dma_request_slave_channel_reason(dev, "rx");
+	if (IS_ERR(master->dma_rx)) {
+		ret = PTR_ERR(master->dma_rx);
+		dev_dbg(dev, "can't get the RX DMA channel, error %d\n", ret);
+		master->dma_rx = NULL;
 		goto err;
 	}
 
 	slave_config.direction = DMA_DEV_TO_MEM;
 	slave_config.src_addr = res->start + MXC_CSPIRXDATA;
 	slave_config.src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
-	slave_config.src_maxburst = spi_imx_get_fifosize(spi_imx) / 2;
+	slave_config.src_maxburst = spi_imx->wml;
 	ret = dmaengine_slave_config(master->dma_rx, &slave_config);
 	if (ret) {
 		dev_err(dev, "error in RX dma configuration.\n");
@@ -874,8 +891,6 @@ static int spi_imx_sdma_init(struct device *dev, struct spi_imx_data *spi_imx,
 	master->max_dma_len = MAX_SDMA_BD_BYTES;
 	spi_imx->bitbang.master->flags = SPI_MASTER_MUST_RX |
 					 SPI_MASTER_MUST_TX;
-	spi_imx->tx_wml = spi_imx_get_fifosize(spi_imx) / 2;
-	spi_imx->rx_wml = spi_imx_get_fifosize(spi_imx) / 2;
 	spi_imx->dma_is_inited = 1;
 
 	return 0;
@@ -942,14 +957,22 @@ static int spi_imx_dma_transfer(struct spi_imx_data *spi_imx,
 	dma = readl(spi_imx->base + MX51_ECSPI_DMA);
 	dma = dma & (~MX51_ECSPI_DMA_RXT_WML_MASK);
 	/* Change RX_DMA_LENGTH trigger dma fetch tail data */
-	left = transfer->len % spi_imx->rxt_wml;
+	left = transfer->len % spi_imx->wml;
 	if (left)
 		writel(dma | (left << MX51_ECSPI_DMA_RXT_WML_OFFSET),
 				spi_imx->base + MX51_ECSPI_DMA);
+	/*
+	 * Set these order to avoid potential RX overflow. The overflow may
+	 * happen if we enable SPI HW before starting RX DMA due to rescheduling
+	 * for another task and/or interrupt.
+	 * So RX DMA enabled first to make sure data would be read out from FIFO
+	 * ASAP. TX DMA enabled next to start filling TX FIFO with new data.
+	 * And finaly SPI HW enabled to start actual data transfer.
+	 */
+	dma_async_issue_pending(master->dma_rx);
+	dma_async_issue_pending(master->dma_tx);
 	spi_imx->devtype_data->trigger(spi_imx);
 
-	dma_async_issue_pending(master->dma_tx);
-	dma_async_issue_pending(master->dma_rx);
 	/* Wait SDMA to finish the data transfer.*/
 	timeout = wait_for_completion_timeout(&spi_imx->dma_tx_completion,
 						IMX_DMA_TIMEOUT);
@@ -958,6 +981,7 @@ static int spi_imx_dma_transfer(struct spi_imx_data *spi_imx,
 			dev_driver_string(&master->dev),
 			dev_name(&master->dev));
 		dmaengine_terminate_all(master->dma_tx);
+		dmaengine_terminate_all(master->dma_rx);
 	} else {
 		timeout = wait_for_completion_timeout(
 				&spi_imx->dma_rx_completion, IMX_DMA_TIMEOUT);
@@ -968,8 +992,9 @@ static int spi_imx_dma_transfer(struct spi_imx_data *spi_imx,
 			spi_imx->devtype_data->reset(spi_imx);
 			dmaengine_terminate_all(master->dma_rx);
 		}
+		dma &= ~MX51_ECSPI_DMA_RXT_WML_MASK;
 		writel(dma |
-		       spi_imx->rxt_wml << MX51_ECSPI_DMA_RXT_WML_OFFSET,
+		       spi_imx->wml << MX51_ECSPI_DMA_RXT_WML_OFFSET,
 		       spi_imx->base + MX51_ECSPI_DMA);
 	}
 
@@ -1117,6 +1142,9 @@ static int spi_imx_probe(struct platform_device *pdev)
 	spi_imx = spi_master_get_devdata(master);
 	spi_imx->bitbang.master = master;
 
+	spi_imx->devtype_data = of_id ? of_id->data :
+		(struct spi_imx_devtype_data *)pdev->id_entry->driver_data;
+
 	for (i = 0; i < master->num_chipselect; i++) {
 		int cs_gpio = of_get_named_gpio(np, "cs-gpios", i);
 		if (!gpio_is_valid(cs_gpio) && mxc_platform_info)
@@ -1142,12 +1170,11 @@ static int spi_imx_probe(struct platform_device *pdev)
 	spi_imx->bitbang.master->prepare_message = spi_imx_prepare_message;
 	spi_imx->bitbang.master->unprepare_message = spi_imx_unprepare_message;
 	spi_imx->bitbang.master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	if (is_imx51_ecspi(spi_imx))
+		spi_imx->bitbang.master->mode_bits |= SPI_LOOP;
 
 	init_completion(&spi_imx->xfer_done);
 
-	spi_imx->devtype_data = of_id ? of_id->data :
-		(struct spi_imx_devtype_data *) pdev->id_entry->driver_data;
-
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	spi_imx->base = devm_ioremap_resource(&pdev->dev, res);
 	if (IS_ERR(spi_imx->base)) {
@@ -1193,9 +1220,15 @@ static int spi_imx_probe(struct platform_device *pdev)
 	 * Only validated on i.mx6 now, can remove the constrain if validated on
 	 * other chips.
 	 */
-	if (spi_imx->devtype_data == &imx51_ecspi_devtype_data
-	    && spi_imx_sdma_init(&pdev->dev, spi_imx, master, res))
-		dev_err(&pdev->dev, "dma setup error,use pio instead\n");
+	if (is_imx51_ecspi(spi_imx)) {
+		ret = spi_imx_sdma_init(&pdev->dev, spi_imx, master, res);
+		if (ret == -EPROBE_DEFER)
+			goto out_clk_put;
+
+		if (ret < 0)
+			dev_err(&pdev->dev, "dma setup error %d, use pio\n",
+				ret);
+	}
 
 	spi_imx->devtype_data->reset(spi_imx);
 
diff --git a/drivers/spi/spi-lm70llp.c b/drivers/spi/spi-lm70llp.c
index ba72347cb99d..61ee0f4269ae 100644
--- a/drivers/spi/spi-lm70llp.c
+++ b/drivers/spi/spi-lm70llp.c
@@ -14,6 +14,8 @@
  * GNU General Public License for more details.
  */
 
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
 #include <linux/init.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
@@ -23,11 +25,9 @@
 #include <linux/sysfs.h>
 #include <linux/workqueue.h>
 
-
 #include <linux/spi/spi.h>
 #include <linux/spi/spi_bitbang.h>
 
-
 /*
  * The LM70 communicates with a host processor using a 3-wire variant of
  * the SPI/Microwire bus interface. This driver specifically supports an
@@ -88,7 +88,6 @@ struct spi_lm70llp {
 /* REVISIT : ugly global ; provides "exclusive open" facility */
 static struct spi_lm70llp *lm70llp;
 
-
 /*-------------------------------------------------------------------*/
 
 static inline struct spi_lm70llp *spidev_to_pp(struct spi_device *spi)
@@ -122,12 +121,14 @@ static inline void assertCS(struct spi_lm70llp *pp)
 static inline void clkHigh(struct spi_lm70llp *pp)
 {
 	u8 data = parport_read_data(pp->port);
+
 	parport_write_data(pp->port, data | SCLK);
 }
 
 static inline void clkLow(struct spi_lm70llp *pp)
 {
 	u8 data = parport_read_data(pp->port);
+
 	parport_write_data(pp->port, data & ~SCLK);
 }
 
@@ -166,8 +167,10 @@ static inline void setmosi(struct spi_device *s, int is_on)
 static inline int getmiso(struct spi_device *s)
 {
 	struct spi_lm70llp *pp = spidev_to_pp(s);
-	return ((SIO == (parport_read_status(pp->port) & SIO)) ? 0 : 1 );
+
+	return ((SIO == (parport_read_status(pp->port) & SIO)) ? 0 : 1);
 }
+
 /*--------------------------------------------------------------------*/
 
 #include "spi-bitbang-txrx.h"
@@ -196,11 +199,10 @@ static void spi_lm70llp_attach(struct parport *p)
 	struct spi_lm70llp	*pp;
 	struct spi_master	*master;
 	int			status;
+	struct pardev_cb	lm70llp_cb;
 
 	if (lm70llp) {
-		printk(KERN_WARNING
-			"%s: spi_lm70llp instance already loaded. Aborting.\n",
-			DRVNAME);
+		pr_warn("spi_lm70llp instance already loaded. Aborting.\n");
 		return;
 	}
 
@@ -227,9 +229,11 @@ static void spi_lm70llp_attach(struct parport *p)
 	 * Parport hookup
 	 */
 	pp->port = p;
-	pd = parport_register_device(p, DRVNAME,
-			NULL, NULL, NULL,
-			PARPORT_FLAG_EXCL, pp);
+	memset(&lm70llp_cb, 0, sizeof(lm70llp_cb));
+	lm70llp_cb.private = pp;
+	lm70llp_cb.flags = PARPORT_FLAG_EXCL;
+	pd = parport_register_dev_model(p, DRVNAME, &lm70llp_cb, 0);
+
 	if (!pd) {
 		status = -ENOMEM;
 		goto out_free_master;
@@ -245,9 +249,8 @@ static void spi_lm70llp_attach(struct parport *p)
 	 */
 	status = spi_bitbang_start(&pp->bitbang);
 	if (status < 0) {
-		printk(KERN_WARNING
-			"%s: spi_bitbang_start failed with status %d\n",
-			DRVNAME, status);
+		dev_warn(&pd->dev, "spi_bitbang_start failed with status %d\n",
+			 status);
 		goto out_off_and_release;
 	}
 
@@ -272,9 +275,9 @@ static void spi_lm70llp_attach(struct parport *p)
 	pp->spidev_lm70 = spi_new_device(pp->bitbang.master, &pp->info);
 	if (pp->spidev_lm70)
 		dev_dbg(&pp->spidev_lm70->dev, "spidev_lm70 at %s\n",
-				dev_name(&pp->spidev_lm70->dev));
+			dev_name(&pp->spidev_lm70->dev));
 	else {
-		printk(KERN_WARNING "%s: spi_new_device failed\n", DRVNAME);
+		dev_warn(&pd->dev, "spi_new_device failed\n");
 		status = -ENODEV;
 		goto out_bitbang_stop;
 	}
@@ -293,9 +296,9 @@ out_off_and_release:
 out_parport_unreg:
 	parport_unregister_device(pd);
 out_free_master:
-	(void) spi_master_put(master);
+	spi_master_put(master);
 out_fail:
-	pr_info("%s: spi_lm70llp probe fail, status %d\n", DRVNAME, status);
+	pr_info("spi_lm70llp probe fail, status %d\n", status);
 }
 
 static void spi_lm70llp_detach(struct parport *p)
@@ -314,16 +317,16 @@ static void spi_lm70llp_detach(struct parport *p)
 	parport_release(pp->pd);
 	parport_unregister_device(pp->pd);
 
-	(void) spi_master_put(pp->bitbang.master);
+	spi_master_put(pp->bitbang.master);
 
 	lm70llp = NULL;
 }
 
-
 static struct parport_driver spi_lm70llp_drv = {
 	.name =		DRVNAME,
-	.attach =	spi_lm70llp_attach,
+	.match_port =	spi_lm70llp_attach,
 	.detach =	spi_lm70llp_detach,
+	.devmodel =	true,
 };
 
 static int __init init_spi_lm70llp(void)
diff --git a/drivers/spi/spi-loopback-test.c b/drivers/spi/spi-loopback-test.c
new file mode 100644
index 000000000000..894616f687b0
--- /dev/null
+++ b/drivers/spi/spi-loopback-test.c
@@ -0,0 +1,1005 @@
+/*
+ *  linux/drivers/spi/spi-loopback-test.c
+ *
+ *  (c) Martin Sperl <kernel@martin.sperl.org>
+ *
+ *  Loopback test driver to test several typical spi_message conditions
+ *  that a spi_master driver may encounter
+ *  this can also get used for regression testing
+ *
+ *  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/delay.h>
+#include <linux/kernel.h>
+#include <linux/list.h>
+#include <linux/list_sort.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/printk.h>
+#include <linux/spi/spi.h>
+
+#include "spi-test.h"
+
+/* flag to only simulate transfers */
+int simulate_only;
+module_param(simulate_only, int, 0);
+MODULE_PARM_DESC(simulate_only, "if not 0 do not execute the spi message");
+
+/* dump spi messages */
+int dump_messages;
+module_param(dump_messages, int, 0);
+MODULE_PARM_DESC(dump_messages,
+		 "=1 dump the basic spi_message_structure, " \
+		 "=2 dump the spi_message_structure including data, " \
+		 "=3 dump the spi_message structure before and after execution");
+/* the device is jumpered for loopback - enabling some rx_buf tests */
+int loopback;
+module_param(loopback, int, 0);
+MODULE_PARM_DESC(loopback,
+		 "if set enable loopback mode, where the rx_buf "	\
+		 "is checked to match tx_buf after the spi_message "	\
+		 "is executed");
+
+/* run only a specific test */
+int run_only_test = -1;
+module_param(run_only_test, int, 0);
+MODULE_PARM_DESC(run_only_test,
+		 "only run the test with this number (0-based !)");
+
+/* the actual tests to execute */
+static struct spi_test spi_tests[] = {
+	{
+		.description	= "tx/rx-transfer - start of page",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_rx_align = ITERATE_ALIGN,
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "tx/rx-transfer - crossing PAGE_SIZE",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_rx_align = ITERATE_ALIGN,
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(PAGE_SIZE - 4),
+				.rx_buf = RX(PAGE_SIZE - 4),
+			},
+		},
+	},
+	{
+		.description	= "tx-transfer - only",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+			},
+		},
+	},
+	{
+		.description	= "rx-transfer - only",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_rx_align = ITERATE_ALIGN,
+		.transfers		= {
+			{
+				.len = 1,
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx-transfers - alter both",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0) | BIT(1),
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+			},
+			{
+				.len = 1,
+				/* this is why we cant use ITERATE_MAX_LEN */
+				.tx_buf = TX(SPI_TEST_MAX_SIZE_HALF),
+			},
+		},
+	},
+	{
+		.description	= "two tx-transfers - alter first",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(1),
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(64),
+			},
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx-transfers - alter second",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0),
+		.transfers		= {
+			{
+				.len = 16,
+				.tx_buf = TX(0),
+			},
+			{
+				.len = 1,
+				.tx_buf = TX(64),
+			},
+		},
+	},
+	{
+		.description	= "two transfers tx then rx - alter both",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0) | BIT(1),
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+			},
+			{
+				.len = 1,
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two transfers tx then rx - alter tx",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0),
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+			},
+			{
+				.len = 1,
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two transfers tx then rx - alter rx",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(1),
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+			},
+			{
+				.len = 1,
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx+rx transfers - alter both",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0) | BIT(1),
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+			{
+				.len = 1,
+				/* making sure we align without overwrite
+				 * the reason we can not use ITERATE_MAX_LEN
+				 */
+				.tx_buf = TX(SPI_TEST_MAX_SIZE_HALF),
+				.rx_buf = RX(SPI_TEST_MAX_SIZE_HALF),
+			},
+		},
+	},
+	{
+		.description	= "two tx+rx transfers - alter first",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0),
+		.transfers		= {
+			{
+				.len = 1,
+				/* making sure we align without overwrite */
+				.tx_buf = TX(1024),
+				.rx_buf = RX(1024),
+			},
+			{
+				.len = 1,
+				/* making sure we align without overwrite */
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx+rx transfers - alter second",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(1),
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+			{
+				.len = 1,
+				/* making sure we align without overwrite */
+				.tx_buf = TX(1024),
+				.rx_buf = RX(1024),
+			},
+		},
+	},
+
+	{ /* end of tests sequence */ }
+};
+
+static int spi_loopback_test_probe(struct spi_device *spi)
+{
+	int ret;
+
+	dev_info(&spi->dev, "Executing spi-loopback-tests\n");
+
+	ret = spi_test_run_tests(spi, spi_tests);
+
+	dev_info(&spi->dev, "Finished spi-loopback-tests with return: %i\n",
+		 ret);
+
+	return ret;
+}
+
+/* non const match table to permit to change via a module parameter */
+static struct of_device_id spi_loopback_test_of_match[] = {
+	{ .compatible	= "linux,spi-loopback-test", },
+	{ }
+};
+
+/* allow to override the compatible string via a module_parameter */
+module_param_string(compatible, spi_loopback_test_of_match[0].compatible,
+		    sizeof(spi_loopback_test_of_match[0].compatible),
+		    0000);
+
+MODULE_DEVICE_TABLE(of, spi_loopback_test_of_match);
+
+static struct spi_driver spi_loopback_test_driver = {
+	.driver = {
+		.name = "spi-loopback-test",
+		.owner = THIS_MODULE,
+		.of_match_table = spi_loopback_test_of_match,
+	},
+	.probe = spi_loopback_test_probe,
+};
+
+module_spi_driver(spi_loopback_test_driver);
+
+MODULE_AUTHOR("Martin Sperl <kernel@martin.sperl.org>");
+MODULE_DESCRIPTION("test spi_driver to check core functionality");
+MODULE_LICENSE("GPL");
+
+/*-------------------------------------------------------------------------*/
+
+/* spi_test implementation */
+
+#define RANGE_CHECK(ptr, plen, start, slen) \
+	((ptr >= start) && (ptr + plen <= start + slen))
+
+/* we allocate one page more, to allow for offsets */
+#define SPI_TEST_MAX_SIZE_PLUS (SPI_TEST_MAX_SIZE + PAGE_SIZE)
+
+static void spi_test_print_hex_dump(char *pre, const void *ptr, size_t len)
+{
+	/* limit the hex_dump */
+	if (len < 1024) {
+		print_hex_dump(KERN_INFO, pre,
+			       DUMP_PREFIX_OFFSET, 16, 1,
+			       ptr, len, 0);
+		return;
+	}
+	/* print head */
+	print_hex_dump(KERN_INFO, pre,
+		       DUMP_PREFIX_OFFSET, 16, 1,
+		       ptr, 512, 0);
+	/* print tail */
+	pr_info("%s truncated - continuing at offset %04zx\n",
+		pre, len - 512);
+	print_hex_dump(KERN_INFO, pre,
+		       DUMP_PREFIX_OFFSET, 16, 1,
+		       ptr + (len - 512), 512, 0);
+}
+
+static void spi_test_dump_message(struct spi_device *spi,
+				  struct spi_message *msg,
+				  bool dump_data)
+{
+	struct spi_transfer *xfer;
+	int i;
+	u8 b;
+
+	dev_info(&spi->dev, "  spi_msg@%pK\n", msg);
+	if (msg->status)
+		dev_info(&spi->dev, "    status:        %i\n",
+			 msg->status);
+	dev_info(&spi->dev, "    frame_length:  %i\n",
+		 msg->frame_length);
+	dev_info(&spi->dev, "    actual_length: %i\n",
+		 msg->actual_length);
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		dev_info(&spi->dev, "    spi_transfer@%pK\n", xfer);
+		dev_info(&spi->dev, "      len:    %i\n", xfer->len);
+		dev_info(&spi->dev, "      tx_buf: %pK\n", xfer->tx_buf);
+		if (dump_data && xfer->tx_buf)
+			spi_test_print_hex_dump("          TX: ",
+						xfer->tx_buf,
+						xfer->len);
+
+		dev_info(&spi->dev, "      rx_buf: %pK\n", xfer->rx_buf);
+		if (dump_data && xfer->rx_buf)
+			spi_test_print_hex_dump("          RX: ",
+						xfer->rx_buf,
+						xfer->len);
+		/* check for unwritten test pattern on rx_buf */
+		if (xfer->rx_buf) {
+			for (i = 0 ; i < xfer->len ; i++) {
+				b = ((u8 *)xfer->rx_buf)[xfer->len - 1 - i];
+				if (b != SPI_TEST_PATTERN_UNWRITTEN)
+					break;
+			}
+			if (i)
+				dev_info(&spi->dev,
+					 "      rx_buf filled with %02x starts at offset: %i\n",
+					 SPI_TEST_PATTERN_UNWRITTEN,
+					 xfer->len - i);
+		}
+	}
+}
+
+struct rx_ranges {
+	struct list_head list;
+	u8 *start;
+	u8 *end;
+};
+
+int rx_ranges_cmp(void *priv, struct list_head *a, struct list_head *b)
+{
+	struct rx_ranges *rx_a = list_entry(a, struct rx_ranges, list);
+	struct rx_ranges *rx_b = list_entry(b, struct rx_ranges, list);
+
+	if (rx_a->start > rx_b->start)
+		return 1;
+	if (rx_a->start < rx_b->start)
+		return -1;
+	return 0;
+}
+
+static int spi_check_rx_ranges(struct spi_device *spi,
+			       struct spi_message *msg,
+			       void *rx)
+{
+	struct spi_transfer *xfer;
+	struct rx_ranges ranges[SPI_TEST_MAX_TRANSFERS], *r;
+	int i = 0;
+	LIST_HEAD(ranges_list);
+	u8 *addr;
+	int ret = 0;
+
+	/* loop over all transfers to fill in the rx_ranges */
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		/* if there is no rx, then no check is needed */
+		if (!xfer->rx_buf)
+			continue;
+		/* fill in the rx_range */
+		if (RANGE_CHECK(xfer->rx_buf, xfer->len,
+				rx, SPI_TEST_MAX_SIZE_PLUS)) {
+			ranges[i].start = xfer->rx_buf;
+			ranges[i].end = xfer->rx_buf + xfer->len;
+			list_add(&ranges[i].list, &ranges_list);
+			i++;
+		}
+	}
+
+	/* if no ranges, then we can return and avoid the checks...*/
+	if (!i)
+		return 0;
+
+	/* sort the list */
+	list_sort(NULL, &ranges_list, rx_ranges_cmp);
+
+	/* and iterate over all the rx addresses */
+	for (addr = rx; addr < (u8 *)rx + SPI_TEST_MAX_SIZE_PLUS; addr++) {
+		/* if we are the DO not write pattern,
+		 * then continue with the loop...
+		 */
+		if (*addr == SPI_TEST_PATTERN_DO_NOT_WRITE)
+			continue;
+
+		/* check if we are inside a range */
+		list_for_each_entry(r, &ranges_list, list) {
+			/* if so then set to end... */
+			if ((addr >= r->start) && (addr < r->end))
+				addr = r->end;
+		}
+		/* second test after a (hopefull) translation */
+		if (*addr == SPI_TEST_PATTERN_DO_NOT_WRITE)
+			continue;
+
+		/* if still not found then something has modified too much */
+		/* we could list the "closest" transfer here... */
+		dev_err(&spi->dev,
+			"loopback strangeness - rx changed outside of allowed range at: %pK\n",
+			addr);
+		/* do not return, only set ret,
+		 * so that we list all addresses
+		 */
+		ret = -ERANGE;
+	}
+
+	return ret;
+}
+
+static int spi_test_check_loopback_result(struct spi_device *spi,
+					  struct spi_message *msg,
+					  void *tx, void *rx)
+{
+	struct spi_transfer *xfer;
+	u8 rxb, txb;
+	size_t i;
+	int ret;
+
+	/* checks rx_buffer pattern are valid with loopback or without */
+	ret = spi_check_rx_ranges(spi, msg, rx);
+	if (ret)
+		return ret;
+
+	/* if we run without loopback, then return now */
+	if (!loopback)
+		return 0;
+
+	/* if applicable to transfer check that rx_buf is equal to tx_buf */
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		/* if there is no rx, then no check is needed */
+		if (!xfer->rx_buf)
+			continue;
+		/* so depending on tx_buf we need to handle things */
+		if (xfer->tx_buf) {
+			for (i = 1; i < xfer->len; i++) {
+				txb = ((u8 *)xfer->tx_buf)[i];
+				rxb = ((u8 *)xfer->rx_buf)[i];
+				if (txb != rxb)
+					goto mismatch_error;
+			}
+		} else {
+			/* first byte received */
+			txb = ((u8 *)xfer->rx_buf)[0];
+			/* first byte may be 0 or xff */
+			if (!((txb == 0) || (txb == 0xff))) {
+				dev_err(&spi->dev,
+					"loopback strangeness - we expect 0x00 or 0xff, but not 0x%02x\n",
+					txb);
+				return -EINVAL;
+			}
+			/* check that all bytes are identical */
+			for (i = 1; i < xfer->len; i++) {
+				rxb = ((u8 *)xfer->rx_buf)[i];
+				if (rxb != txb)
+					goto mismatch_error;
+			}
+		}
+	}
+
+	return 0;
+
+mismatch_error:
+	dev_err(&spi->dev,
+		"loopback strangeness - transfer missmatch on byte %04zx - expected 0x%02x, but got 0x%02x\n",
+		i, txb, rxb);
+
+	return -EINVAL;
+}
+
+static int spi_test_translate(struct spi_device *spi,
+			      void **ptr, size_t len,
+			      void *tx, void *rx)
+{
+	size_t off;
+
+	/* return on null */
+	if (!*ptr)
+		return 0;
+
+	/* in the MAX_SIZE_HALF case modify the pointer */
+	if (((size_t)*ptr) & SPI_TEST_MAX_SIZE_HALF)
+		/* move the pointer to the correct range */
+		*ptr += (SPI_TEST_MAX_SIZE_PLUS / 2) -
+			SPI_TEST_MAX_SIZE_HALF;
+
+	/* RX range
+	 * - we check against MAX_SIZE_PLUS to allow for automated alignment
+	 */
+	if (RANGE_CHECK(*ptr, len,  RX(0), SPI_TEST_MAX_SIZE_PLUS)) {
+		off = *ptr - RX(0);
+		*ptr = rx + off;
+
+		return 0;
+	}
+
+	/* TX range */
+	if (RANGE_CHECK(*ptr, len,  TX(0), SPI_TEST_MAX_SIZE_PLUS)) {
+		off = *ptr - TX(0);
+		*ptr = tx + off;
+
+		return 0;
+	}
+
+	dev_err(&spi->dev,
+		"PointerRange [%pK:%pK[ not in range [%pK:%pK[ or [%pK:%pK[\n",
+		*ptr, *ptr + len,
+		RX(0), RX(SPI_TEST_MAX_SIZE),
+		TX(0), TX(SPI_TEST_MAX_SIZE));
+
+	return -EINVAL;
+}
+
+static int spi_test_fill_pattern(struct spi_device *spi,
+				 struct spi_test *test)
+{
+	struct spi_transfer *xfers = test->transfers;
+	u8 *tx_buf;
+	size_t count = 0;
+	int i, j;
+
+#ifdef __BIG_ENDIAN
+#define GET_VALUE_BYTE(value, index, bytes) \
+	(value >> (8 * (bytes - 1 - count % bytes)))
+#else
+#define GET_VALUE_BYTE(value, index, bytes) \
+	(value >> (8 * (count % bytes)))
+#endif
+
+	/* fill all transfers with the pattern requested */
+	for (i = 0; i < test->transfer_count; i++) {
+		/* fill rx_buf with SPI_TEST_PATTERN_UNWRITTEN */
+		if (xfers[i].rx_buf)
+			memset(xfers[i].rx_buf, SPI_TEST_PATTERN_UNWRITTEN,
+			       xfers[i].len);
+		/* if tx_buf is NULL then skip */
+		tx_buf = (u8 *)xfers[i].tx_buf;
+		if (!tx_buf)
+			continue;
+		/* modify all the transfers */
+		for (j = 0; j < xfers[i].len; j++, tx_buf++, count++) {
+			/* fill tx */
+			switch (test->fill_option) {
+			case FILL_MEMSET_8:
+				*tx_buf = test->fill_pattern;
+				break;
+			case FILL_MEMSET_16:
+				*tx_buf = GET_VALUE_BYTE(test->fill_pattern,
+							 count, 2);
+				break;
+			case FILL_MEMSET_24:
+				*tx_buf = GET_VALUE_BYTE(test->fill_pattern,
+							 count, 3);
+				break;
+			case FILL_MEMSET_32:
+				*tx_buf = GET_VALUE_BYTE(test->fill_pattern,
+							 count, 4);
+				break;
+			case FILL_COUNT_8:
+				*tx_buf = count;
+				break;
+			case FILL_COUNT_16:
+				*tx_buf = GET_VALUE_BYTE(count, count, 2);
+				break;
+			case FILL_COUNT_24:
+				*tx_buf = GET_VALUE_BYTE(count, count, 3);
+				break;
+			case FILL_COUNT_32:
+				*tx_buf = GET_VALUE_BYTE(count, count, 4);
+				break;
+			case FILL_TRANSFER_BYTE_8:
+				*tx_buf = j;
+				break;
+			case FILL_TRANSFER_BYTE_16:
+				*tx_buf = GET_VALUE_BYTE(j, j, 2);
+				break;
+			case FILL_TRANSFER_BYTE_24:
+				*tx_buf = GET_VALUE_BYTE(j, j, 3);
+				break;
+			case FILL_TRANSFER_BYTE_32:
+				*tx_buf = GET_VALUE_BYTE(j, j, 4);
+				break;
+			case FILL_TRANSFER_NUM:
+				*tx_buf = i;
+				break;
+			default:
+				dev_err(&spi->dev,
+					"unsupported fill_option: %i\n",
+					test->fill_option);
+				return -EINVAL;
+			}
+		}
+	}
+
+	return 0;
+}
+
+static int _spi_test_run_iter(struct spi_device *spi,
+			      struct spi_test *test,
+			      void *tx, void *rx)
+{
+	struct spi_message *msg = &test->msg;
+	struct spi_transfer *x;
+	int i, ret;
+
+	/* initialize message - zero-filled via static initialization */
+	spi_message_init_no_memset(msg);
+
+	/* fill rx with the DO_NOT_WRITE pattern */
+	memset(rx, SPI_TEST_PATTERN_DO_NOT_WRITE, SPI_TEST_MAX_SIZE_PLUS);
+
+	/* add the individual transfers */
+	for (i = 0; i < test->transfer_count; i++) {
+		x = &test->transfers[i];
+
+		/* patch the values of tx_buf */
+		ret = spi_test_translate(spi, (void **)&x->tx_buf, x->len,
+					 (void *)tx, rx);
+		if (ret)
+			return ret;
+
+		/* patch the values of rx_buf */
+		ret = spi_test_translate(spi, &x->rx_buf, x->len,
+					 (void *)tx, rx);
+		if (ret)
+			return ret;
+
+		/* and add it to the list */
+		spi_message_add_tail(x, msg);
+	}
+
+	/* fill in the transfer buffers with pattern */
+	ret = spi_test_fill_pattern(spi, test);
+	if (ret)
+		return ret;
+
+	/* and execute */
+	if (test->execute_msg)
+		ret = test->execute_msg(spi, test, tx, rx);
+	else
+		ret = spi_test_execute_msg(spi, test, tx, rx);
+
+	/* handle result */
+	if (ret == test->expected_return)
+		return 0;
+
+	dev_err(&spi->dev,
+		"test failed - test returned %i, but we expect %i\n",
+		ret, test->expected_return);
+
+	if (ret)
+		return ret;
+
+	/* if it is 0, as we expected something else,
+	 * then return something special
+	 */
+	return -EFAULT;
+}
+
+static int spi_test_run_iter(struct spi_device *spi,
+			     const struct spi_test *testtemplate,
+			     void *tx, void *rx,
+			     size_t len,
+			     size_t tx_off,
+			     size_t rx_off
+	)
+{
+	struct spi_test test;
+	int i, tx_count, rx_count;
+
+	/* copy the test template to test */
+	memcpy(&test, testtemplate, sizeof(test));
+
+	/* set up test->transfers to the correct count */
+	if (!test.transfer_count) {
+		for (i = 0;
+		    (i < SPI_TEST_MAX_TRANSFERS) && test.transfers[i].len;
+		    i++) {
+			test.transfer_count++;
+		}
+	}
+
+	/* if iterate_transfer_mask is not set,
+	 * then set it to first transfer only
+	 */
+	if (!(test.iterate_transfer_mask & (BIT(test.transfer_count) - 1)))
+		test.iterate_transfer_mask = 1;
+
+	/* count number of transfers with tx/rx_buf != NULL */
+	for (i = 0; i < test.transfer_count; i++) {
+		if (test.transfers[i].tx_buf)
+			tx_count++;
+		if (test.transfers[i].rx_buf)
+			rx_count++;
+	}
+
+	/* in some iteration cases warn and exit early,
+	 * as there is nothing to do, that has not been tested already...
+	 */
+	if (tx_off && (!tx_count)) {
+		dev_warn_once(&spi->dev,
+			      "%s: iterate_tx_off configured with tx_buf==NULL - ignoring\n",
+			      test.description);
+		return 0;
+	}
+	if (rx_off && (!rx_count)) {
+		dev_warn_once(&spi->dev,
+			      "%s: iterate_rx_off configured with rx_buf==NULL - ignoring\n",
+			      test.description);
+		return 0;
+	}
+
+	/* write out info */
+	if (!(len || tx_off || rx_off)) {
+		dev_info(&spi->dev, "Running test %s\n", test.description);
+	} else {
+		dev_info(&spi->dev,
+			 "  with iteration values: len = %zu, tx_off = %zu, rx_off = %zu\n",
+			 len, tx_off, rx_off);
+	}
+
+	/* update in the values from iteration values */
+	for (i = 0; i < test.transfer_count; i++) {
+		/* only when bit in transfer mask is set */
+		if (!(test.iterate_transfer_mask & BIT(i)))
+			continue;
+		if (len)
+			test.transfers[i].len = len;
+		if (test.transfers[i].tx_buf)
+			test.transfers[i].tx_buf += tx_off;
+		if (test.transfers[i].tx_buf)
+			test.transfers[i].rx_buf += rx_off;
+	}
+
+	/* and execute */
+	return _spi_test_run_iter(spi, &test, tx, rx);
+}
+
+/**
+ * spi_test_execute_msg - default implementation to run a test
+ *
+ * spi: @spi_device on which to run the @spi_message
+ * test: the test to execute, which already contains @msg
+ * tx:   the tx buffer allocated for the test sequence
+ * rx:   the rx buffer allocated for the test sequence
+ *
+ * Returns: error code of spi_sync as well as basic error checking
+ */
+int spi_test_execute_msg(struct spi_device *spi, struct spi_test *test,
+			 void *tx, void *rx)
+{
+	struct spi_message *msg = &test->msg;
+	int ret = 0;
+	int i;
+
+	/* only if we do not simulate */
+	if (!simulate_only) {
+		/* dump the complete message before and after the transfer */
+		if (dump_messages == 3)
+			spi_test_dump_message(spi, msg, true);
+
+		/* run spi message */
+		ret = spi_sync(spi, msg);
+		if (ret == -ETIMEDOUT) {
+			dev_info(&spi->dev,
+				 "spi-message timed out - reruning...\n");
+			/* rerun after a few explicit schedules */
+			for (i = 0; i < 16; i++)
+				schedule();
+			ret = spi_sync(spi, msg);
+		}
+		if (ret) {
+			dev_err(&spi->dev,
+				"Failed to execute spi_message: %i\n",
+				ret);
+			goto exit;
+		}
+
+		/* do some extra error checks */
+		if (msg->frame_length != msg->actual_length) {
+			dev_err(&spi->dev,
+				"actual length differs from expected\n");
+			ret = -EIO;
+			goto exit;
+		}
+
+		/* run rx-buffer tests */
+		ret = spi_test_check_loopback_result(spi, msg, tx, rx);
+	}
+
+	/* if requested or on error dump message (including data) */
+exit:
+	if (dump_messages || ret)
+		spi_test_dump_message(spi, msg,
+				      (dump_messages >= 2) || (ret));
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(spi_test_execute_msg);
+
+/**
+ * spi_test_run_test - run an individual spi_test
+ *                     including all the relevant iterations on:
+ *                     length and buffer alignment
+ *
+ * spi:  the spi_device to send the messages to
+ * test: the test which we need to execute
+ * tx:   the tx buffer allocated for the test sequence
+ * rx:   the rx buffer allocated for the test sequence
+ *
+ * Returns: status code of spi_sync or other failures
+ */
+
+int spi_test_run_test(struct spi_device *spi, const struct spi_test *test,
+		      void *tx, void *rx)
+{
+	int idx_len;
+	size_t len;
+	size_t tx_align, rx_align;
+	int ret;
+
+	/* test for transfer limits */
+	if (test->transfer_count >= SPI_TEST_MAX_TRANSFERS) {
+		dev_err(&spi->dev,
+			"%s: Exceeded max number of transfers with %i\n",
+			test->description, test->transfer_count);
+		return -E2BIG;
+	}
+
+	/* setting up some values in spi_message
+	 * based on some settings in spi_master
+	 * some of this can also get done in the run() method
+	 */
+
+	/* iterate over all the iterable values using macros
+	 * (to make it a bit more readable...
+	 */
+#define FOR_EACH_ITERATE(var, defaultvalue)				\
+	for (idx_##var = -1, var = defaultvalue;			\
+	     ((idx_##var < 0) ||					\
+		     (							\
+			     (idx_##var < SPI_TEST_MAX_ITERATE) &&	\
+			     (var = test->iterate_##var[idx_##var])	\
+		     )							\
+	     );								\
+	     idx_##var++)
+#define FOR_EACH_ALIGNMENT(var)						\
+	for (var = 0;							\
+	    var < (test->iterate_##var ?				\
+			(spi->master->dma_alignment ?			\
+			 spi->master->dma_alignment :			\
+			 test->iterate_##var) :				\
+			1);						\
+	    var++)
+
+	FOR_EACH_ITERATE(len, 0) {
+		FOR_EACH_ALIGNMENT(tx_align) {
+			FOR_EACH_ALIGNMENT(rx_align) {
+				/* and run the iteration */
+				ret = spi_test_run_iter(spi, test,
+							tx, rx,
+							len,
+							tx_align,
+							rx_align);
+				if (ret)
+					return ret;
+			}
+		}
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(spi_test_run_test);
+
+/**
+ * spi_test_run_tests - run an array of spi_messages tests
+ * @spi: the spi device on which to run the tests
+ * @tests: NULL-terminated array of @spi_test
+ *
+ * Returns: status errors as per @spi_test_run_test()
+ */
+
+int spi_test_run_tests(struct spi_device *spi,
+		       struct spi_test *tests)
+{
+	char *rx = NULL, *tx = NULL;
+	int ret = 0, count = 0;
+	struct spi_test *test;
+
+	/* allocate rx/tx buffers of 128kB size without devm
+	 * in the hope that is on a page boundary
+	 */
+	rx = kzalloc(SPI_TEST_MAX_SIZE_PLUS, GFP_KERNEL);
+	if (!rx) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	tx = kzalloc(SPI_TEST_MAX_SIZE_PLUS, GFP_KERNEL);
+	if (!tx) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	/* now run the individual tests in the table */
+	for (test = tests, count = 0; test->description[0];
+	     test++, count++) {
+		/* only run test if requested */
+		if ((run_only_test > -1) && (count != run_only_test))
+			continue;
+		/* run custom implementation */
+		if (test->run_test)
+			ret = test->run_test(spi, test, tx, rx);
+		else
+			ret = spi_test_run_test(spi, test, tx, rx);
+		if (ret)
+			goto out;
+		/* add some delays so that we can easily
+		 * detect the individual tests when using a logic analyzer
+		 * we also add scheduling to avoid potential spi_timeouts...
+		 */
+		mdelay(100);
+		schedule();
+	}
+
+out:
+	kfree(rx);
+	kfree(tx);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(spi_test_run_tests);
diff --git a/drivers/spi/spi-mt65xx.c b/drivers/spi/spi-mt65xx.c
index 7840067062a8..0be89e052428 100644
--- a/drivers/spi/spi-mt65xx.c
+++ b/drivers/spi/spi-mt65xx.c
@@ -95,8 +95,7 @@ struct mtk_spi {
 	const struct mtk_spi_compatible *dev_comp;
 };
 
-static const struct mtk_spi_compatible mt6589_compat;
-static const struct mtk_spi_compatible mt8135_compat;
+static const struct mtk_spi_compatible mtk_common_compat;
 static const struct mtk_spi_compatible mt8173_compat = {
 	.need_pad_sel = true,
 	.must_tx = true,
@@ -112,9 +111,18 @@ static const struct mtk_chip_config mtk_default_chip_info = {
 };
 
 static const struct of_device_id mtk_spi_of_match[] = {
-	{ .compatible = "mediatek,mt6589-spi", .data = (void *)&mt6589_compat },
-	{ .compatible = "mediatek,mt8135-spi", .data = (void *)&mt8135_compat },
-	{ .compatible = "mediatek,mt8173-spi", .data = (void *)&mt8173_compat },
+	{ .compatible = "mediatek,mt2701-spi",
+		.data = (void *)&mtk_common_compat,
+	},
+	{ .compatible = "mediatek,mt6589-spi",
+		.data = (void *)&mtk_common_compat,
+	},
+	{ .compatible = "mediatek,mt8135-spi",
+		.data = (void *)&mtk_common_compat,
+	},
+	{ .compatible = "mediatek,mt8173-spi",
+		.data = (void *)&mt8173_compat,
+	},
 	{}
 };
 MODULE_DEVICE_TABLE(of, mtk_spi_of_match);
@@ -154,9 +162,6 @@ static int mtk_spi_prepare_message(struct spi_master *master,
 		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);
 
 	/* set the mlsbx and mlsbtx */
 	if (chip_config->tx_mlsb)
@@ -323,7 +328,8 @@ static int mtk_spi_fifo_transfer(struct spi_master *master,
 				 struct spi_device *spi,
 				 struct spi_transfer *xfer)
 {
-	int cnt;
+	int cnt, remainder;
+	u32 reg_val;
 	struct mtk_spi *mdata = spi_master_get_devdata(master);
 
 	mdata->cur_transfer = xfer;
@@ -331,12 +337,16 @@ static int mtk_spi_fifo_transfer(struct spi_master *master,
 	mtk_spi_prepare_transfer(master, xfer);
 	mtk_spi_setup_packet(master);
 
-	if (xfer->len % 4)
-		cnt = xfer->len / 4 + 1;
-	else
-		cnt = xfer->len / 4;
+	cnt = xfer->len / 4;
 	iowrite32_rep(mdata->base + SPI_TX_DATA_REG, xfer->tx_buf, cnt);
 
+	remainder = xfer->len % 4;
+	if (remainder > 0) {
+		reg_val = 0;
+		memcpy(&reg_val, xfer->tx_buf + (cnt * 4), remainder);
+		writel(reg_val, mdata->base + SPI_TX_DATA_REG);
+	}
+
 	mtk_spi_enable_transfer(master);
 
 	return 1;
@@ -418,7 +428,7 @@ static int mtk_spi_setup(struct spi_device *spi)
 
 static irqreturn_t mtk_spi_interrupt(int irq, void *dev_id)
 {
-	u32 cmd, reg_val, cnt;
+	u32 cmd, reg_val, cnt, remainder;
 	struct spi_master *master = dev_id;
 	struct mtk_spi *mdata = spi_master_get_devdata(master);
 	struct spi_transfer *trans = mdata->cur_transfer;
@@ -431,12 +441,15 @@ static irqreturn_t mtk_spi_interrupt(int irq, void *dev_id)
 
 	if (!master->can_dma(master, master->cur_msg->spi, trans)) {
 		if (trans->rx_buf) {
-			if (mdata->xfer_len % 4)
-				cnt = mdata->xfer_len / 4 + 1;
-			else
-				cnt = mdata->xfer_len / 4;
+			cnt = mdata->xfer_len / 4;
 			ioread32_rep(mdata->base + SPI_RX_DATA_REG,
 				     trans->rx_buf, cnt);
+			remainder = mdata->xfer_len % 4;
+			if (remainder > 0) {
+				reg_val = readl(mdata->base + SPI_RX_DATA_REG);
+				memcpy(trans->rx_buf + (cnt * 4),
+					&reg_val, remainder);
+			}
 		}
 		spi_finalize_current_transfer(master);
 		return IRQ_HANDLED;
@@ -610,7 +623,8 @@ static int mtk_spi_probe(struct platform_device *pdev)
 	ret = clk_set_parent(mdata->sel_clk, mdata->parent_clk);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "failed to clk_set_parent (%d)\n", ret);
-		goto err_disable_clk;
+		clk_disable_unprepare(mdata->spi_clk);
+		goto err_put_master;
 	}
 
 	clk_disable_unprepare(mdata->spi_clk);
@@ -620,7 +634,7 @@ static int mtk_spi_probe(struct platform_device *pdev)
 	ret = devm_spi_register_master(&pdev->dev, master);
 	if (ret) {
 		dev_err(&pdev->dev, "failed to register master (%d)\n", ret);
-		goto err_put_master;
+		goto err_disable_runtime_pm;
 	}
 
 	if (mdata->dev_comp->need_pad_sel) {
@@ -629,14 +643,14 @@ static int mtk_spi_probe(struct platform_device *pdev)
 				"pad_num does not match num_chipselect(%d != %d)\n",
 				mdata->pad_num, master->num_chipselect);
 			ret = -EINVAL;
-			goto err_put_master;
+			goto err_disable_runtime_pm;
 		}
 
 		if (!master->cs_gpios && master->num_chipselect > 1) {
 			dev_err(&pdev->dev,
 				"cs_gpios not specified and num_chipselect > 1\n");
 			ret = -EINVAL;
-			goto err_put_master;
+			goto err_disable_runtime_pm;
 		}
 
 		if (master->cs_gpios) {
@@ -647,7 +661,7 @@ static int mtk_spi_probe(struct platform_device *pdev)
 				if (ret) {
 					dev_err(&pdev->dev,
 						"can't get CS GPIO %i\n", i);
-					goto err_put_master;
+					goto err_disable_runtime_pm;
 				}
 			}
 		}
@@ -655,8 +669,8 @@ static int mtk_spi_probe(struct platform_device *pdev)
 
 	return 0;
 
-err_disable_clk:
-	clk_disable_unprepare(mdata->spi_clk);
+err_disable_runtime_pm:
+	pm_runtime_disable(&pdev->dev);
 err_put_master:
 	spi_master_put(master);
 
diff --git a/drivers/spi/spi-omap2-mcspi.c b/drivers/spi/spi-omap2-mcspi.c
index 1f8903d356e5..7273820275e9 100644
--- a/drivers/spi/spi-omap2-mcspi.c
+++ b/drivers/spi/spi-omap2-mcspi.c
@@ -24,6 +24,7 @@
 #include <linux/dma-mapping.h>
 #include <linux/dmaengine.h>
 #include <linux/omap-dma.h>
+#include <linux/pinctrl/consumer.h>
 #include <linux/platform_device.h>
 #include <linux/err.h>
 #include <linux/clk.h>
@@ -1024,6 +1025,16 @@ static int omap2_mcspi_setup(struct spi_device *spi)
 		spi->controller_state = cs;
 		/* Link this to context save list */
 		list_add_tail(&cs->node, &ctx->cs);
+
+		if (gpio_is_valid(spi->cs_gpio)) {
+			ret = gpio_request(spi->cs_gpio, dev_name(&spi->dev));
+			if (ret) {
+				dev_err(&spi->dev, "failed to request gpio\n");
+				return ret;
+			}
+			gpio_direction_output(spi->cs_gpio,
+					 !(spi->mode & SPI_CS_HIGH));
+		}
 	}
 
 	if (!mcspi_dma->dma_rx || !mcspi_dma->dma_tx) {
@@ -1032,15 +1043,6 @@ static int omap2_mcspi_setup(struct spi_device *spi)
 			return ret;
 	}
 
-	if (gpio_is_valid(spi->cs_gpio)) {
-		ret = gpio_request(spi->cs_gpio, dev_name(&spi->dev));
-		if (ret) {
-			dev_err(&spi->dev, "failed to request gpio\n");
-			return ret;
-		}
-		gpio_direction_output(spi->cs_gpio, !(spi->mode & SPI_CS_HIGH));
-	}
-
 	ret = pm_runtime_get_sync(mcspi->dev);
 	if (ret < 0)
 		return ret;
@@ -1536,14 +1538,23 @@ static int omap2_mcspi_resume(struct device *dev)
 	}
 	pm_runtime_mark_last_busy(mcspi->dev);
 	pm_runtime_put_autosuspend(mcspi->dev);
-	return 0;
+
+	return pinctrl_pm_select_default_state(dev);
+}
+
+static int omap2_mcspi_suspend(struct device *dev)
+{
+	return pinctrl_pm_select_sleep_state(dev);
 }
+
 #else
+#define omap2_mcspi_suspend	NULL
 #define	omap2_mcspi_resume	NULL
 #endif
 
 static const struct dev_pm_ops omap2_mcspi_pm_ops = {
 	.resume = omap2_mcspi_resume,
+	.suspend = omap2_mcspi_suspend,
 	.runtime_resume	= omap_mcspi_runtime_resume,
 };
 
diff --git a/drivers/spi/spi-pxa2xx.c b/drivers/spi/spi-pxa2xx.c
index b25dc71b0ea9..ab9914ad8365 100644
--- a/drivers/spi/spi-pxa2xx.c
+++ b/drivers/spi/spi-pxa2xx.c
@@ -1567,9 +1567,6 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 	if (!is_quark_x1000_ssp(drv_data))
 		pxa2xx_spi_write(drv_data, SSPSP, 0);
 
-	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);
diff --git a/drivers/spi/spi-s3c64xx.c b/drivers/spi/spi-s3c64xx.c
index 8e86e7f6663a..5a76a50063b5 100644
--- a/drivers/spi/spi-s3c64xx.c
+++ b/drivers/spi/spi-s3c64xx.c
@@ -133,7 +133,6 @@
 struct s3c64xx_spi_dma_data {
 	struct dma_chan *ch;
 	enum dma_transfer_direction direction;
-	unsigned int dmach;
 };
 
 /**
@@ -325,7 +324,7 @@ static int s3c64xx_spi_prepare_transfer(struct spi_master *spi)
 
 		/* Acquire DMA channels */
 		sdd->rx_dma.ch = dma_request_slave_channel_compat(mask, filter,
-				   (void *)(long)sdd->rx_dma.dmach, dev, "rx");
+				   sdd->cntrlr_info->dma_rx, dev, "rx");
 		if (!sdd->rx_dma.ch) {
 			dev_err(dev, "Failed to get RX DMA channel\n");
 			ret = -EBUSY;
@@ -334,7 +333,7 @@ static int s3c64xx_spi_prepare_transfer(struct spi_master *spi)
 		spi->dma_rx = sdd->rx_dma.ch;
 
 		sdd->tx_dma.ch = dma_request_slave_channel_compat(mask, filter,
-				   (void *)(long)sdd->tx_dma.dmach, dev, "tx");
+				   sdd->cntrlr_info->dma_tx, dev, "tx");
 		if (!sdd->tx_dma.ch) {
 			dev_err(dev, "Failed to get TX DMA channel\n");
 			ret = -EBUSY;
@@ -1028,7 +1027,6 @@ static inline struct s3c64xx_spi_port_config *s3c64xx_spi_get_port_config(
 static int s3c64xx_spi_probe(struct platform_device *pdev)
 {
 	struct resource	*mem_res;
-	struct resource	*res;
 	struct s3c64xx_spi_driver_data *sdd;
 	struct s3c64xx_spi_info *sci = dev_get_platdata(&pdev->dev);
 	struct spi_master *master;
@@ -1087,20 +1085,9 @@ static int s3c64xx_spi_probe(struct platform_device *pdev)
 
 	sdd->cur_bpw = 8;
 
-	if (!sdd->pdev->dev.of_node) {
-		res = platform_get_resource(pdev, IORESOURCE_DMA,  0);
-		if (!res) {
-			dev_warn(&pdev->dev, "Unable to get SPI tx dma resource. Switching to poll mode\n");
-			sdd->port_conf->quirks = S3C64XX_SPI_QUIRK_POLL;
-		} else
-			sdd->tx_dma.dmach = res->start;
-
-		res = platform_get_resource(pdev, IORESOURCE_DMA,  1);
-		if (!res) {
-			dev_warn(&pdev->dev, "Unable to get SPI rx dma resource. Switching to poll mode\n");
-			sdd->port_conf->quirks = S3C64XX_SPI_QUIRK_POLL;
-		} else
-			sdd->rx_dma.dmach = res->start;
+	if (!sdd->pdev->dev.of_node && (!sci->dma_tx || !sci->dma_rx)) {
+		dev_warn(&pdev->dev, "Unable to get SPI tx/rx DMA data. Switching to poll mode\n");
+		sdd->port_conf->quirks = S3C64XX_SPI_QUIRK_POLL;
 	}
 
 	sdd->tx_dma.direction = DMA_MEM_TO_DEV;
@@ -1197,9 +1184,9 @@ static int s3c64xx_spi_probe(struct platform_device *pdev)
 
 	dev_dbg(&pdev->dev, "Samsung SoC SPI Driver loaded for Bus SPI-%d with %d Slaves attached\n",
 					sdd->port_id, master->num_chipselect);
-	dev_dbg(&pdev->dev, "\tIOmem=[%pR]\tFIFO %dbytes\tDMA=[Rx-%d, Tx-%d]\n",
+	dev_dbg(&pdev->dev, "\tIOmem=[%pR]\tFIFO %dbytes\tDMA=[Rx-%p, Tx-%p]\n",
 					mem_res, (FIFO_LVL_MASK(sdd) >> 1) + 1,
-					sdd->rx_dma.dmach, sdd->tx_dma.dmach);
+					sci->dma_rx, sci->dma_tx);
 
 	pm_runtime_mark_last_busy(&pdev->dev);
 	pm_runtime_put_autosuspend(&pdev->dev);
@@ -1370,12 +1357,6 @@ static const struct platform_device_id s3c64xx_spi_driver_ids[] = {
 	}, {
 		.name		= "s3c6410-spi",
 		.driver_data	= (kernel_ulong_t)&s3c6410_spi_port_config,
-	}, {
-		.name		= "s5pv210-spi",
-		.driver_data	= (kernel_ulong_t)&s5pv210_spi_port_config,
-	}, {
-		.name		= "exynos4210-spi",
-		.driver_data	= (kernel_ulong_t)&exynos4_spi_port_config,
 	},
 	{ },
 };
diff --git a/drivers/spi/spi-sun4i.c b/drivers/spi/spi-sun4i.c
index fbb0a4d74e91..1ddd9e2309b6 100644
--- a/drivers/spi/spi-sun4i.c
+++ b/drivers/spi/spi-sun4i.c
@@ -140,6 +140,9 @@ static void sun4i_spi_set_cs(struct spi_device *spi, bool enable)
 	reg &= ~SUN4I_CTL_CS_MASK;
 	reg |= SUN4I_CTL_CS(spi->chip_select);
 
+	/* We want to control the chip select manually */
+	reg |= SUN4I_CTL_CS_MANUAL;
+
 	if (enable)
 		reg |= SUN4I_CTL_CS_LEVEL;
 	else
@@ -222,15 +225,12 @@ static int sun4i_spi_transfer_one(struct spi_master *master,
 	else
 		reg |= SUN4I_CTL_DHB;
 
-	/* We want to control the chip select manually */
-	reg |= SUN4I_CTL_CS_MANUAL;
-
 	sun4i_spi_write(sspi, SUN4I_CTL_REG, reg);
 
 	/* Ensure that we have a parent clock fast enough */
 	mclk_rate = clk_get_rate(sspi->mclk);
-	if (mclk_rate < (2 * spi->max_speed_hz)) {
-		clk_set_rate(sspi->mclk, 2 * spi->max_speed_hz);
+	if (mclk_rate < (2 * tfr->speed_hz)) {
+		clk_set_rate(sspi->mclk, 2 * tfr->speed_hz);
 		mclk_rate = clk_get_rate(sspi->mclk);
 	}
 
@@ -248,14 +248,14 @@ static int sun4i_spi_transfer_one(struct spi_master *master,
 	 * First try CDR2, and if we can't reach the expected
 	 * frequency, fall back to CDR1.
 	 */
-	div = mclk_rate / (2 * spi->max_speed_hz);
+	div = mclk_rate / (2 * tfr->speed_hz);
 	if (div <= (SUN4I_CLK_CTL_CDR2_MASK + 1)) {
 		if (div > 0)
 			div--;
 
 		reg = SUN4I_CLK_CTL_CDR2(div) | SUN4I_CLK_CTL_DRS;
 	} else {
-		div = ilog2(mclk_rate) - ilog2(spi->max_speed_hz);
+		div = ilog2(mclk_rate) - ilog2(tfr->speed_hz);
 		reg = SUN4I_CLK_CTL_CDR1(div);
 	}
 
diff --git a/drivers/spi/spi-sun6i.c b/drivers/spi/spi-sun6i.c
index ac48f59705a8..42e2c4bd690a 100644
--- a/drivers/spi/spi-sun6i.c
+++ b/drivers/spi/spi-sun6i.c
@@ -217,8 +217,8 @@ static int sun6i_spi_transfer_one(struct spi_master *master,
 
 	/* Ensure that we have a parent clock fast enough */
 	mclk_rate = clk_get_rate(sspi->mclk);
-	if (mclk_rate < (2 * spi->max_speed_hz)) {
-		clk_set_rate(sspi->mclk, 2 * spi->max_speed_hz);
+	if (mclk_rate < (2 * tfr->speed_hz)) {
+		clk_set_rate(sspi->mclk, 2 * tfr->speed_hz);
 		mclk_rate = clk_get_rate(sspi->mclk);
 	}
 
@@ -236,14 +236,14 @@ static int sun6i_spi_transfer_one(struct spi_master *master,
 	 * First try CDR2, and if we can't reach the expected
 	 * frequency, fall back to CDR1.
 	 */
-	div = mclk_rate / (2 * spi->max_speed_hz);
+	div = mclk_rate / (2 * tfr->speed_hz);
 	if (div <= (SUN6I_CLK_CTL_CDR2_MASK + 1)) {
 		if (div > 0)
 			div--;
 
 		reg = SUN6I_CLK_CTL_CDR2(div) | SUN6I_CLK_CTL_DRS;
 	} else {
-		div = ilog2(mclk_rate) - ilog2(spi->max_speed_hz);
+		div = ilog2(mclk_rate) - ilog2(tfr->speed_hz);
 		reg = SUN6I_CLK_CTL_CDR1(div);
 	}
 
diff --git a/drivers/spi/spi-test.h b/drivers/spi/spi-test.h
new file mode 100644
index 000000000000..922c52833239
--- /dev/null
+++ b/drivers/spi/spi-test.h
@@ -0,0 +1,136 @@
+/*
+ *  linux/drivers/spi/spi-test.h
+ *
+ *  (c) Martin Sperl <kernel@martin.sperl.org>
+ *
+ *  spi_test definitions
+ *
+ *  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/spi/spi.h>
+
+#define SPI_TEST_MAX_TRANSFERS 4
+#define SPI_TEST_MAX_SIZE (32 * PAGE_SIZE)
+#define SPI_TEST_MAX_ITERATE 32
+
+/* the "dummy" start addresses used in spi_test
+ * these addresses get translated at a later stage
+ */
+#define RX_START	BIT(30)
+#define TX_START	BIT(31)
+#define RX(off)		((void *)(RX_START + off))
+#define TX(off)		((void *)(TX_START + off))
+
+/* some special defines for offsets */
+#define SPI_TEST_MAX_SIZE_HALF	BIT(29)
+
+/* detection pattern for unfinished reads...
+ * - 0x00 or 0xff could be valid levels for tx_buf = NULL,
+ * so we do not use either of them
+ */
+#define SPI_TEST_PATTERN_UNWRITTEN 0xAA
+#define SPI_TEST_PATTERN_DO_NOT_WRITE 0x55
+#define SPI_TEST_CHECK_DO_NOT_WRITE 64
+
+/**
+ * struct spi_test - describes a specific (set of) tests to execute
+ *
+ * @description:      description of the test
+ *
+ * @msg:              a template @spi_message usedfor the default settings
+ * @transfers:        array of @spi_transfers that are part of the
+ *                    resulting spi_message. The first transfer with len == 0
+ *                    signifies the end of the list
+ * @transfer_count:   normally computed number of transfers with len > 0
+ *
+ * @run_test:         run a specific spi_test - this allows to override
+ *                    the default implementation of @spi_test_run_transfer
+ *                    either to add some custom filters for a specific test
+ *                    or to effectively run some very custom tests...
+ * @execute_msg:      run the spi_message for real - this allows to override
+ *                    @spi_test_execute_msg to apply final modifications
+ *                    on the spi_message
+ * @expected_return:  the expected return code - in some cases we want to
+ *                    test also for error conditions
+ *
+ * @iterate_len:      list of length to iterate on (in addition to the
+ *                    explicitly set @spi_transfer.len)
+ * @iterate_tx_align: change the alignment of @spi_transfer.tx_buf
+ *                    for all values in the below range if set.
+ *                    the ranges are:
+ *                    [0 : @spi_master.dma_alignment[ if set
+ *                    [0 : iterate_tx_align[ if unset
+ * @iterate_rx_align: change the alignment of @spi_transfer.rx_buf
+ *                    see @iterate_tx_align for details
+ * @iterate_transfer_mask: the bitmask of transfers to which the iterations
+ *                         apply - if 0, then it applies to all transfer
+ *
+ * @fill_option:      define the way how tx_buf is filled
+ * @fill_pattern:     fill pattern to apply to the tx_buf
+ *                    (used in some of the @fill_options)
+ */
+
+struct spi_test {
+	char description[64];
+	struct spi_message msg;
+	struct spi_transfer transfers[SPI_TEST_MAX_TRANSFERS];
+	unsigned int transfer_count;
+	int (*run_test)(struct spi_device *spi, struct spi_test *test,
+			void *tx, void *rx);
+	int (*execute_msg)(struct spi_device *spi, struct spi_test *test,
+			   void *tx, void *rx);
+	int expected_return;
+	/* iterate over all the non-zero values */
+	int iterate_len[SPI_TEST_MAX_ITERATE];
+	int iterate_tx_align;
+	int iterate_rx_align;
+	u32 iterate_transfer_mask;
+	/* the tx-fill operation */
+	u32 fill_option;
+#define FILL_MEMSET_8	0	/* just memset with 8 bit */
+#define FILL_MEMSET_16	1	/* just memset with 16 bit */
+#define FILL_MEMSET_24	2	/* just memset with 24 bit */
+#define FILL_MEMSET_32	3	/* just memset with 32 bit */
+#define FILL_COUNT_8	4	/* fill with a 8 byte counter */
+#define FILL_COUNT_16	5	/* fill with a 16 bit counter */
+#define FILL_COUNT_24	6	/* fill with a 24 bit counter */
+#define FILL_COUNT_32	7	/* fill with a 32 bit counter */
+#define FILL_TRANSFER_BYTE_8  8	/* fill with the transfer byte - 8 bit */
+#define FILL_TRANSFER_BYTE_16 9	/* fill with the transfer byte - 16 bit */
+#define FILL_TRANSFER_BYTE_24 10 /* fill with the transfer byte - 24 bit */
+#define FILL_TRANSFER_BYTE_32 11 /* fill with the transfer byte - 32 bit */
+#define FILL_TRANSFER_NUM     16 /* fill with the transfer number */
+	u32 fill_pattern;
+};
+
+/* default implementation for @spi_test.run_test */
+int spi_test_run_test(struct spi_device *spi,
+		      const struct spi_test *test,
+		      void *tx, void *rx);
+
+/* default implementation for @spi_test.execute_msg */
+int spi_test_execute_msg(struct spi_device *spi,
+			 struct spi_test *test,
+			 void *tx, void *rx);
+
+/* function to execute a set of tests */
+int spi_test_run_tests(struct spi_device *spi,
+		       struct spi_test *tests);
+
+/* some of the default @spi_transfer.len to test */
+#define ITERATE_LEN 2, 3, 7, 11, 16, 31, 32, 64, 97, 128, 251, 256, \
+		1021, 1024, 1031, 4093, PAGE_SIZE, 4099, 65536, 65537
+
+#define ITERATE_MAX_LEN ITERATE_LEN, SPI_TEST_MAX_SIZE - 1, SPI_TEST_MAX_SIZE
+
+/* the default alignment to test */
+#define ITERATE_ALIGN sizeof(int)
diff --git a/drivers/spi/spi-zynqmp-gqspi.c b/drivers/spi/spi-zynqmp-gqspi.c
index f23f36ebaf3d..aab9b492c627 100644
--- a/drivers/spi/spi-zynqmp-gqspi.c
+++ b/drivers/spi/spi-zynqmp-gqspi.c
@@ -917,9 +917,7 @@ static int zynqmp_qspi_start_transfer(struct spi_master *master,
  */
 static int __maybe_unused zynqmp_qspi_suspend(struct device *dev)
 {
-	struct platform_device *pdev = container_of(dev,
-						    struct platform_device,
-						    dev);
+	struct platform_device *pdev = to_platform_device(dev);
 	struct spi_master *master = platform_get_drvdata(pdev);
 
 	spi_master_suspend(master);
@@ -940,9 +938,7 @@ static int __maybe_unused zynqmp_qspi_suspend(struct device *dev)
  */
 static int __maybe_unused zynqmp_qspi_resume(struct device *dev)
 {
-	struct platform_device *pdev = container_of(dev,
-						    struct platform_device,
-						    dev);
+	struct platform_device *pdev = to_platform_device(dev);
 	struct spi_master *master = platform_get_drvdata(pdev);
 	struct zynqmp_qspi *xqspi = spi_master_get_devdata(master);
 	int ret = 0;
diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
index 151b01c25b40..47eff8012a77 100644
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -84,8 +84,7 @@ static ssize_t spi_device_##field##_show(struct device *dev,		\
 					 struct device_attribute *attr,	\
 					char *buf)			\
 {									\
-	struct spi_device *spi = container_of(dev,			\
-					      struct spi_device, dev);	\
+	struct spi_device *spi = to_spi_device(dev);			\
 	return spi_statistics_##field##_show(&spi->statistics, buf);	\
 }									\
 static struct device_attribute dev_attr_spi_device_##field = {		\
@@ -605,6 +604,24 @@ struct spi_device *spi_new_device(struct spi_master *master,
 }
 EXPORT_SYMBOL_GPL(spi_new_device);
 
+/**
+ * spi_unregister_device - unregister a single SPI device
+ * @spi: spi_device to unregister
+ *
+ * Start making the passed SPI device vanish. Normally this would be handled
+ * by spi_unregister_master().
+ */
+void spi_unregister_device(struct spi_device *spi)
+{
+	if (!spi)
+		return;
+
+	if (spi->dev.of_node)
+		of_node_clear_flag(spi->dev.of_node, OF_POPULATED);
+	device_unregister(&spi->dev);
+}
+EXPORT_SYMBOL_GPL(spi_unregister_device);
+
 static void spi_match_master_to_boardinfo(struct spi_master *master,
 				struct spi_board_info *bi)
 {
@@ -1548,6 +1565,8 @@ static void of_register_spi_devices(struct spi_master *master)
 		return;
 
 	for_each_available_child_of_node(master->dev.of_node, nc) {
+		if (of_node_test_and_set_flag(nc, OF_POPULATED))
+			continue;
 		spi = of_register_spi_device(master, nc);
 		if (IS_ERR(spi))
 			dev_warn(&master->dev, "Failed to create SPI device for %s\n",
@@ -2636,6 +2655,11 @@ static int of_spi_notify(struct notifier_block *nb, unsigned long action,
 		if (master == NULL)
 			return NOTIFY_OK;	/* not for us */
 
+		if (of_node_test_and_set_flag(rd->dn, OF_POPULATED)) {
+			put_device(&master->dev);
+			return NOTIFY_OK;
+		}
+
 		spi = of_register_spi_device(master, rd->dn);
 		put_device(&master->dev);
 
@@ -2647,6 +2671,10 @@ static int of_spi_notify(struct notifier_block *nb, unsigned long action,
 		break;
 
 	case OF_RECONFIG_CHANGE_REMOVE:
+		/* already depopulated? */
+		if (!of_node_check_flag(rd->dn, OF_POPULATED))
+			return NOTIFY_OK;
+
 		/* find our device by node */
 		spi = of_find_spi_device_by_node(rd->dn);
 		if (spi == NULL)
diff --git a/drivers/spi/spidev.c b/drivers/spi/spidev.c
index d0e7dfc647cf..e3c19f30f591 100644
--- a/drivers/spi/spidev.c
+++ b/drivers/spi/spidev.c
@@ -284,7 +284,7 @@ static int spidev_message(struct spidev_data *spidev,
 			k_tmp->speed_hz = spidev->speed_hz;
 #ifdef VERBOSE
 		dev_dbg(&spidev->spi->dev,
-			"  xfer len %zd %s%s%s%dbits %u usec %uHz\n",
+			"  xfer len %u %s%s%s%dbits %u usec %uHz\n",
 			u_tmp->len,
 			u_tmp->rx_buf ? "rx " : "",
 			u_tmp->tx_buf ? "tx " : "",
diff --git a/include/linux/platform_data/spi-s3c64xx.h b/include/linux/platform_data/spi-s3c64xx.h
index d3889b98a1a1..fb5625bcca9a 100644
--- a/include/linux/platform_data/spi-s3c64xx.h
+++ b/include/linux/platform_data/spi-s3c64xx.h
@@ -40,6 +40,8 @@ struct s3c64xx_spi_info {
 	int num_cs;
 	int (*cfg_gpio)(void);
 	dma_filter_fn filter;
+	void *dma_tx;
+	void *dma_rx;
 };
 
 /**
diff --git a/include/linux/spi/spi.h b/include/linux/spi/spi.h
index cce80e6dc7d1..53be3a4c60cb 100644
--- a/include/linux/spi/spi.h
+++ b/include/linux/spi/spi.h
@@ -425,6 +425,12 @@ struct spi_master {
 #define SPI_MASTER_MUST_RX      BIT(3)		/* requires rx */
 #define SPI_MASTER_MUST_TX      BIT(4)		/* requires tx */
 
+	/*
+	 * on some hardware transfer size may be constrained
+	 * the limit may depend on device transfer settings
+	 */
+	size_t (*max_transfer_size)(struct spi_device *spi);
+
 	/* lock and mutex for SPI bus locking */
 	spinlock_t		bus_lock_spinlock;
 	struct mutex		bus_lock_mutex;
@@ -762,10 +768,15 @@ struct spi_message {
 	void			*state;
 };
 
+static inline void spi_message_init_no_memset(struct spi_message *m)
+{
+	INIT_LIST_HEAD(&m->transfers);
+}
+
 static inline void spi_message_init(struct spi_message *m)
 {
 	memset(m, 0, sizeof *m);
-	INIT_LIST_HEAD(&m->transfers);
+	spi_message_init_no_memset(m);
 }
 
 static inline void
@@ -832,6 +843,15 @@ extern int spi_async(struct spi_device *spi, struct spi_message *message);
 extern int spi_async_locked(struct spi_device *spi,
 			    struct spi_message *message);
 
+static inline size_t
+spi_max_transfer_size(struct spi_device *spi)
+{
+	struct spi_master *master = spi->master;
+	if (!master->max_transfer_size)
+		return SIZE_MAX;
+	return master->max_transfer_size(spi);
+}
+
 /*---------------------------------------------------------------------------*/
 
 /* All these synchronous SPI transfer routines are utilities layered
@@ -1115,12 +1135,7 @@ spi_add_device(struct spi_device *spi);
 extern struct spi_device *
 spi_new_device(struct spi_master *, struct spi_board_info *);
 
-static inline void
-spi_unregister_device(struct spi_device *spi)
-{
-	if (spi)
-		device_unregister(&spi->dev);
-}
+extern void spi_unregister_device(struct spi_device *spi);
 
 extern const struct spi_device_id *
 spi_get_device_id(const struct spi_device *sdev);
diff --git a/tools/Makefile b/tools/Makefile
index 0ba0df3b516f..1dc6900468ed 100644
--- a/tools/Makefile
+++ b/tools/Makefile
@@ -17,6 +17,7 @@ help:
 	@echo '  lguest     - a minimal 32-bit x86 hypervisor'
 	@echo '  perf       - Linux performance measurement and analysis tool'
 	@echo '  selftests  - various kernel selftests'
+	@echo '  spi        - spi tools'
 	@echo '  turbostat  - Intel CPU idle stats and freq reporting tool'
 	@echo '  usb        - USB testing tools'
 	@echo '  virtio     - vhost test module'
@@ -52,7 +53,7 @@ acpi: FORCE
 cpupower: FORCE
 	$(call descend,power/$@)
 
-cgroup firewire hv guest usb virtio vm net iio: FORCE
+cgroup firewire hv guest spi usb virtio vm net iio: FORCE
 	$(call descend,$@)
 
 liblockdep: FORCE
@@ -118,7 +119,7 @@ acpi_clean:
 cpupower_clean:
 	$(call descend,power/cpupower,clean)
 
-cgroup_clean hv_clean firewire_clean lguest_clean usb_clean virtio_clean vm_clean net_clean iio_clean:
+cgroup_clean hv_clean firewire_clean lguest_clean spi_clean usb_clean virtio_clean vm_clean net_clean iio_clean:
 	$(call descend,$(@:_clean=),clean)
 
 liblockdep_clean:
@@ -143,7 +144,7 @@ freefall_clean:
 	$(call descend,laptop/freefall,clean)
 
 clean: acpi_clean cgroup_clean cpupower_clean hv_clean firewire_clean lguest_clean \
-		perf_clean selftests_clean turbostat_clean usb_clean virtio_clean \
+		perf_clean selftests_clean turbostat_clean spi_clean usb_clean virtio_clean \
 		vm_clean net_clean iio_clean x86_energy_perf_policy_clean tmon_clean \
 		freefall_clean
 
diff --git a/Documentation/spi/.gitignore b/tools/spi/.gitignore
index 4280576397e8..4280576397e8 100644
--- a/Documentation/spi/.gitignore
+++ b/tools/spi/.gitignore
diff --git a/tools/spi/Makefile b/tools/spi/Makefile
new file mode 100644
index 000000000000..cd0db62e4d9d
--- /dev/null
+++ b/tools/spi/Makefile
@@ -0,0 +1,4 @@
+all: spidev_test spidev_fdx
+
+clean:
+	$(RM) spidev_test spidev_fdx
diff --git a/Documentation/spi/spidev_fdx.c b/tools/spi/spidev_fdx.c
index 0ea3e51292fc..0ea3e51292fc 100644
--- a/Documentation/spi/spidev_fdx.c
+++ b/tools/spi/spidev_fdx.c
diff --git a/Documentation/spi/spidev_test.c b/tools/spi/spidev_test.c
index 135b3f592b83..8a73d8185316 100644
--- a/Documentation/spi/spidev_test.c
+++ b/tools/spi/spidev_test.c
@@ -19,6 +19,7 @@
 #include <getopt.h>
 #include <fcntl.h>
 #include <sys/ioctl.h>
+#include <sys/stat.h>
 #include <linux/types.h>
 #include <linux/spi/spidev.h>
 
@@ -33,6 +34,8 @@ static void pabort(const char *s)
 static const char *device = "/dev/spidev1.1";
 static uint32_t mode;
 static uint8_t bits = 8;
+static char *input_file;
+static char *output_file;
 static uint32_t speed = 500000;
 static uint16_t delay;
 static int verbose;
@@ -49,7 +52,8 @@ uint8_t default_tx[] = {
 uint8_t default_rx[ARRAY_SIZE(default_tx)] = {0, };
 char *input_tx;
 
-static void hex_dump(const void *src, size_t length, size_t line_size, char *prefix)
+static void hex_dump(const void *src, size_t length, size_t line_size,
+		     char *prefix)
 {
 	int i = 0;
 	const unsigned char *address = src;
@@ -83,13 +87,17 @@ static void hex_dump(const void *src, size_t length, size_t line_size, char *pre
 static int unescape(char *_dst, char *_src, size_t len)
 {
 	int ret = 0;
+	int match;
 	char *src = _src;
 	char *dst = _dst;
 	unsigned int ch;
 
 	while (*src) {
 		if (*src == '\\' && *(src+1) == 'x') {
-			sscanf(src + 2, "%2x", &ch);
+			match = sscanf(src + 2, "%2x", &ch);
+			if (!match)
+				pabort("malformed input string");
+
 			src += 4;
 			*dst++ = (unsigned char)ch;
 		} else {
@@ -103,7 +111,7 @@ static int unescape(char *_dst, char *_src, size_t len)
 static void transfer(int fd, uint8_t const *tx, uint8_t const *rx, size_t len)
 {
 	int ret;
-
+	int out_fd;
 	struct spi_ioc_transfer tr = {
 		.tx_buf = (unsigned long)tx,
 		.rx_buf = (unsigned long)rx,
@@ -134,7 +142,21 @@ static void transfer(int fd, uint8_t const *tx, uint8_t const *rx, size_t len)
 
 	if (verbose)
 		hex_dump(tx, len, 32, "TX");
-	hex_dump(rx, len, 32, "RX");
+
+	if (output_file) {
+		out_fd = open(output_file, O_WRONLY | O_CREAT | O_TRUNC, 0666);
+		if (out_fd < 0)
+			pabort("could not open output file");
+
+		ret = write(out_fd, rx, len);
+		if (ret != len)
+			pabort("not all bytes written to output file");
+
+		close(out_fd);
+	}
+
+	if (verbose || !output_file)
+		hex_dump(rx, len, 32, "RX");
 }
 
 static void print_usage(const char *prog)
@@ -143,7 +165,9 @@ static void print_usage(const char *prog)
 	puts("  -D --device   device to use (default /dev/spidev1.1)\n"
 	     "  -s --speed    max speed (Hz)\n"
 	     "  -d --delay    delay (usec)\n"
-	     "  -b --bpw      bits per word \n"
+	     "  -b --bpw      bits per word\n"
+	     "  -i --input    input data from a file (e.g. \"test.bin\")\n"
+	     "  -o --output   output data to a file (e.g. \"results.bin\")\n"
 	     "  -l --loop     loopback\n"
 	     "  -H --cpha     clock phase\n"
 	     "  -O --cpol     clock polarity\n"
@@ -167,6 +191,8 @@ static void parse_opts(int argc, char *argv[])
 			{ "speed",   1, 0, 's' },
 			{ "delay",   1, 0, 'd' },
 			{ "bpw",     1, 0, 'b' },
+			{ "input",   1, 0, 'i' },
+			{ "output",  1, 0, 'o' },
 			{ "loop",    0, 0, 'l' },
 			{ "cpha",    0, 0, 'H' },
 			{ "cpol",    0, 0, 'O' },
@@ -182,7 +208,8 @@ static void parse_opts(int argc, char *argv[])
 		};
 		int c;
 
-		c = getopt_long(argc, argv, "D:s:d:b:lHOLC3NR24p:v", lopts, NULL);
+		c = getopt_long(argc, argv, "D:s:d:b:i:o:lHOLC3NR24p:v",
+				lopts, NULL);
 
 		if (c == -1)
 			break;
@@ -200,6 +227,12 @@ static void parse_opts(int argc, char *argv[])
 		case 'b':
 			bits = atoi(optarg);
 			break;
+		case 'i':
+			input_file = optarg;
+			break;
+		case 'o':
+			output_file = optarg;
+			break;
 		case 'l':
 			mode |= SPI_LOOP;
 			break;
@@ -249,13 +282,63 @@ static void parse_opts(int argc, char *argv[])
 	}
 }
 
+static void transfer_escaped_string(int fd, char *str)
+{
+	size_t size = strlen(str + 1);
+	uint8_t *tx;
+	uint8_t *rx;
+
+	tx = malloc(size);
+	if (!tx)
+		pabort("can't allocate tx buffer");
+
+	rx = malloc(size);
+	if (!rx)
+		pabort("can't allocate rx buffer");
+
+	size = unescape((char *)tx, str, size);
+	transfer(fd, tx, rx, size);
+	free(rx);
+	free(tx);
+}
+
+static void transfer_file(int fd, char *filename)
+{
+	ssize_t bytes;
+	struct stat sb;
+	int tx_fd;
+	uint8_t *tx;
+	uint8_t *rx;
+
+	if (stat(filename, &sb) == -1)
+		pabort("can't stat input file");
+
+	tx_fd = open(filename, O_RDONLY);
+	if (fd < 0)
+		pabort("can't open input file");
+
+	tx = malloc(sb.st_size);
+	if (!tx)
+		pabort("can't allocate tx buffer");
+
+	rx = malloc(sb.st_size);
+	if (!rx)
+		pabort("can't allocate rx buffer");
+
+	bytes = read(tx_fd, tx, sb.st_size);
+	if (bytes != sb.st_size)
+		pabort("failed to read input file");
+
+	transfer(fd, tx, rx, sb.st_size);
+	free(rx);
+	free(tx);
+	close(tx_fd);
+}
+
 int main(int argc, char *argv[])
 {
 	int ret = 0;
 	int fd;
-	uint8_t *tx;
-	uint8_t *rx;
-	int size;
 
 	parse_opts(argc, argv);
 
@@ -300,17 +383,15 @@ int main(int argc, char *argv[])
 	printf("bits per word: %d\n", bits);
 	printf("max speed: %d Hz (%d KHz)\n", speed, speed/1000);
 
-	if (input_tx) {
-		size = strlen(input_tx+1);
-		tx = malloc(size);
-		rx = malloc(size);
-		size = unescape((char *)tx, input_tx, size);
-		transfer(fd, tx, rx, size);
-		free(rx);
-		free(tx);
-	} else {
+	if (input_tx && input_file)
+		pabort("only one of -p and --input may be selected");
+
+	if (input_tx)
+		transfer_escaped_string(fd, input_tx);
+	else if (input_file)
+		transfer_file(fd, input_file);
+	else
 		transfer(fd, default_tx, default_rx, sizeof(default_tx));
-	}
 
 	close(fd);