Merge branch spi-next from git://git.kernel.org/pub/scm/linux/kernel/git/broonie/misc.git

Signed-off-by: Grant Likely <grant.likely@secretlab.ca>

6 files changed, 1451 insertions, 578 deletions

diff --git a/drivers/spi/Kconfig b/drivers/spi/Kconfig
index a90393d..e79884e9 100644
--- a/drivers/spi/Kconfig
+++ b/drivers/spi/Kconfig
@@ -297,9 +297,20 @@ config SPI_PPC4xx
 	help
 	  This selects a driver for the PPC4xx SPI Controller.
 
+config SPI_PXA2XX_PXADMA
+	bool "PXA2xx SSP legacy PXA DMA API support"
+	depends on SPI_PXA2XX && ARCH_PXA
+	help
+	  Enable PXA private legacy DMA API support. Note that this is
+	  deprecated in favor of generic DMA engine API.
+
+config SPI_PXA2XX_DMA
+	def_bool y
+	depends on SPI_PXA2XX && !SPI_PXA2XX_PXADMA
+
 config SPI_PXA2XX
 	tristate "PXA2xx SSP SPI master"
-	depends on ARCH_PXA || PCI
+	depends on ARCH_PXA || PCI || ACPI
 	select PXA_SSP if ARCH_PXA
 	help
 	  This enables using a PXA2xx or Sodaville SSP port as a SPI master
diff --git a/drivers/spi/Makefile b/drivers/spi/Makefile
index 64e970b..e53c309 100644
--- a/drivers/spi/Makefile
+++ b/drivers/spi/Makefile
@@ -47,7 +47,10 @@ obj-$(CONFIG_SPI_OMAP24XX)		+= spi-omap2-mcspi.o
 obj-$(CONFIG_SPI_ORION)			+= spi-orion.o
 obj-$(CONFIG_SPI_PL022)			+= spi-pl022.o
 obj-$(CONFIG_SPI_PPC4xx)		+= spi-ppc4xx.o
-obj-$(CONFIG_SPI_PXA2XX)		+= spi-pxa2xx.o
+spi-pxa2xx-platform-objs		:= spi-pxa2xx.o
+spi-pxa2xx-platform-$(CONFIG_SPI_PXA2XX_PXADMA)	+= spi-pxa2xx-pxadma.o
+spi-pxa2xx-platform-$(CONFIG_SPI_PXA2XX_DMA)	+= spi-pxa2xx-dma.o
+obj-$(CONFIG_SPI_PXA2XX)		+= spi-pxa2xx-platform.o
 obj-$(CONFIG_SPI_PXA2XX_PCI)		+= spi-pxa2xx-pci.o
 obj-$(CONFIG_SPI_RSPI)			+= spi-rspi.o
 obj-$(CONFIG_SPI_S3C24XX)		+= spi-s3c24xx-hw.o
diff --git a/drivers/spi/spi-pxa2xx-dma.c b/drivers/spi/spi-pxa2xx-dma.c
new file mode 100644
index 0000000..c735c5a
--- /dev/null
+++ b/drivers/spi/spi-pxa2xx-dma.c
@@ -0,0 +1,392 @@
+/*
+ * PXA2xx SPI DMA engine support.
+ *
+ * Copyright (C) 2013, Intel Corporation
+ * Author: Mika Westerberg <mika.westerberg@linux.intel.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/init.h>
+#include <linux/device.h>
+#include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
+#include <linux/pxa2xx_ssp.h>
+#include <linux/scatterlist.h>
+#include <linux/sizes.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/pxa2xx_spi.h>
+
+#include "spi-pxa2xx.h"
+
+static int pxa2xx_spi_map_dma_buffer(struct driver_data *drv_data,
+				     enum dma_data_direction dir)
+{
+	int i, nents, len = drv_data->len;
+	struct scatterlist *sg;
+	struct device *dmadev;
+	struct sg_table *sgt;
+	void *buf, *pbuf;
+
+	/*
+	 * Some DMA controllers have problems transferring buffers that are
+	 * not multiple of 4 bytes. So we truncate the transfer so that it
+	 * is suitable for such controllers, and handle the trailing bytes
+	 * manually after the DMA completes.
+	 *
+	 * REVISIT: It would be better if this information could be
+	 * retrieved directly from the DMA device in a similar way than
+	 * ->copy_align etc. is done.
+	 */
+	len = ALIGN(drv_data->len, 4);
+
+	if (dir == DMA_TO_DEVICE) {
+		dmadev = drv_data->tx_chan->device->dev;
+		sgt = &drv_data->tx_sgt;
+		buf = drv_data->tx;
+		drv_data->tx_map_len = len;
+	} else {
+		dmadev = drv_data->rx_chan->device->dev;
+		sgt = &drv_data->rx_sgt;
+		buf = drv_data->rx;
+		drv_data->rx_map_len = len;
+	}
+
+	nents = DIV_ROUND_UP(len, SZ_2K);
+	if (nents != sgt->nents) {
+		int ret;
+
+		sg_free_table(sgt);
+		ret = sg_alloc_table(sgt, nents, GFP_KERNEL);
+		if (ret)
+			return ret;
+	}
+
+	pbuf = buf;
+	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
+		size_t bytes = min_t(size_t, len, SZ_2K);
+
+		if (buf)
+			sg_set_buf(sg, pbuf, bytes);
+		else
+			sg_set_buf(sg, drv_data->dummy, bytes);
+
+		pbuf += bytes;
+		len -= bytes;
+	}
+
+	nents = dma_map_sg(dmadev, sgt->sgl, sgt->nents, dir);
+	if (!nents)
+		return -ENOMEM;
+
+	return nents;
+}
+
+static void pxa2xx_spi_unmap_dma_buffer(struct driver_data *drv_data,
+					enum dma_data_direction dir)
+{
+	struct device *dmadev;
+	struct sg_table *sgt;
+
+	if (dir == DMA_TO_DEVICE) {
+		dmadev = drv_data->tx_chan->device->dev;
+		sgt = &drv_data->tx_sgt;
+	} else {
+		dmadev = drv_data->rx_chan->device->dev;
+		sgt = &drv_data->rx_sgt;
+	}
+
+	dma_unmap_sg(dmadev, sgt->sgl, sgt->nents, dir);
+}
+
+static void pxa2xx_spi_unmap_dma_buffers(struct driver_data *drv_data)
+{
+	if (!drv_data->dma_mapped)
+		return;
+
+	pxa2xx_spi_unmap_dma_buffer(drv_data, DMA_FROM_DEVICE);
+	pxa2xx_spi_unmap_dma_buffer(drv_data, DMA_TO_DEVICE);
+
+	drv_data->dma_mapped = 0;
+}
+
+static void pxa2xx_spi_dma_transfer_complete(struct driver_data *drv_data,
+					     bool error)
+{
+	struct spi_message *msg = drv_data->cur_msg;
+
+	/*
+	 * It is possible that one CPU is handling ROR interrupt and other
+	 * just gets DMA completion. Calling pump_transfers() twice for the
+	 * same transfer leads to problems thus we prevent concurrent calls
+	 * by using ->dma_running.
+	 */
+	if (atomic_dec_and_test(&drv_data->dma_running)) {
+		void __iomem *reg = drv_data->ioaddr;
+
+		/*
+		 * If the other CPU is still handling the ROR interrupt we
+		 * might not know about the error yet. So we re-check the
+		 * ROR bit here before we clear the status register.
+		 */
+		if (!error) {
+			u32 status = read_SSSR(reg) & drv_data->mask_sr;
+			error = status & SSSR_ROR;
+		}
+
+		/* Clear status & disable interrupts */
+		write_SSCR1(read_SSCR1(reg) & ~drv_data->dma_cr1, reg);
+		write_SSSR_CS(drv_data, drv_data->clear_sr);
+		if (!pxa25x_ssp_comp(drv_data))
+			write_SSTO(0, reg);
+
+		if (!error) {
+			pxa2xx_spi_unmap_dma_buffers(drv_data);
+
+			/* Handle the last bytes of unaligned transfer */
+			drv_data->tx += drv_data->tx_map_len;
+			drv_data->write(drv_data);
+
+			drv_data->rx += drv_data->rx_map_len;
+			drv_data->read(drv_data);
+
+			msg->actual_length += drv_data->len;
+			msg->state = pxa2xx_spi_next_transfer(drv_data);
+		} else {
+			/* In case we got an error we disable the SSP now */
+			write_SSCR0(read_SSCR0(reg) & ~SSCR0_SSE, reg);
+
+			msg->state = ERROR_STATE;
+		}
+
+		tasklet_schedule(&drv_data->pump_transfers);
+	}
+}
+
+static void pxa2xx_spi_dma_callback(void *data)
+{
+	pxa2xx_spi_dma_transfer_complete(data, false);
+}
+
+static struct dma_async_tx_descriptor *
+pxa2xx_spi_dma_prepare_one(struct driver_data *drv_data,
+			   enum dma_transfer_direction dir)
+{
+	struct pxa2xx_spi_master *pdata = drv_data->master_info;
+	struct chip_data *chip = drv_data->cur_chip;
+	enum dma_slave_buswidth width;
+	struct dma_slave_config cfg;
+	struct dma_chan *chan;
+	struct sg_table *sgt;
+	int nents, ret;
+
+	switch (drv_data->n_bytes) {
+	case 1:
+		width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		break;
+	case 2:
+		width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		break;
+	default:
+		width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		break;
+	}
+
+	memset(&cfg, 0, sizeof(cfg));
+	cfg.direction = dir;
+
+	if (dir == DMA_MEM_TO_DEV) {
+		cfg.dst_addr = drv_data->ssdr_physical;
+		cfg.dst_addr_width = width;
+		cfg.dst_maxburst = chip->dma_burst_size;
+		cfg.slave_id = pdata->tx_slave_id;
+
+		sgt = &drv_data->tx_sgt;
+		nents = drv_data->tx_nents;
+		chan = drv_data->tx_chan;
+	} else {
+		cfg.src_addr = drv_data->ssdr_physical;
+		cfg.src_addr_width = width;
+		cfg.src_maxburst = chip->dma_burst_size;
+		cfg.slave_id = pdata->rx_slave_id;
+
+		sgt = &drv_data->rx_sgt;
+		nents = drv_data->rx_nents;
+		chan = drv_data->rx_chan;
+	}
+
+	ret = dmaengine_slave_config(chan, &cfg);
+	if (ret) {
+		dev_warn(&drv_data->pdev->dev, "DMA slave config failed\n");
+		return NULL;
+	}
+
+	return dmaengine_prep_slave_sg(chan, sgt->sgl, nents, dir,
+				       DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+}
+
+static bool pxa2xx_spi_dma_filter(struct dma_chan *chan, void *param)
+{
+	const struct pxa2xx_spi_master *pdata = param;
+
+	return chan->chan_id == pdata->tx_chan_id ||
+	       chan->chan_id == pdata->rx_chan_id;
+}
+
+bool pxa2xx_spi_dma_is_possible(size_t len)
+{
+	return len <= MAX_DMA_LEN;
+}
+
+int pxa2xx_spi_map_dma_buffers(struct driver_data *drv_data)
+{
+	const struct chip_data *chip = drv_data->cur_chip;
+	int ret;
+
+	if (!chip->enable_dma)
+		return 0;
+
+	/* Don't bother with DMA if we can't do even a single burst */
+	if (drv_data->len < chip->dma_burst_size)
+		return 0;
+
+	ret = pxa2xx_spi_map_dma_buffer(drv_data, DMA_TO_DEVICE);
+	if (ret <= 0) {
+		dev_warn(&drv_data->pdev->dev, "failed to DMA map TX\n");
+		return 0;
+	}
+
+	drv_data->tx_nents = ret;
+
+	ret = pxa2xx_spi_map_dma_buffer(drv_data, DMA_FROM_DEVICE);
+	if (ret <= 0) {
+		pxa2xx_spi_unmap_dma_buffer(drv_data, DMA_TO_DEVICE);
+		dev_warn(&drv_data->pdev->dev, "failed to DMA map RX\n");
+		return 0;
+	}
+
+	drv_data->rx_nents = ret;
+	return 1;
+}
+
+irqreturn_t pxa2xx_spi_dma_transfer(struct driver_data *drv_data)
+{
+	u32 status;
+
+	status = read_SSSR(drv_data->ioaddr) & drv_data->mask_sr;
+	if (status & SSSR_ROR) {
+		dev_err(&drv_data->pdev->dev, "FIFO overrun\n");
+
+		dmaengine_terminate_all(drv_data->rx_chan);
+		dmaengine_terminate_all(drv_data->tx_chan);
+
+		pxa2xx_spi_dma_transfer_complete(drv_data, true);
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_NONE;
+}
+
+int pxa2xx_spi_dma_prepare(struct driver_data *drv_data, u32 dma_burst)
+{
+	struct dma_async_tx_descriptor *tx_desc, *rx_desc;
+
+	tx_desc = pxa2xx_spi_dma_prepare_one(drv_data, DMA_MEM_TO_DEV);
+	if (!tx_desc) {
+		dev_err(&drv_data->pdev->dev,
+			"failed to get DMA TX descriptor\n");
+		return -EBUSY;
+	}
+
+	rx_desc = pxa2xx_spi_dma_prepare_one(drv_data, DMA_DEV_TO_MEM);
+	if (!rx_desc) {
+		dev_err(&drv_data->pdev->dev,
+			"failed to get DMA RX descriptor\n");
+		return -EBUSY;
+	}
+
+	/* We are ready when RX completes */
+	rx_desc->callback = pxa2xx_spi_dma_callback;
+	rx_desc->callback_param = drv_data;
+
+	dmaengine_submit(rx_desc);
+	dmaengine_submit(tx_desc);
+	return 0;
+}
+
+void pxa2xx_spi_dma_start(struct driver_data *drv_data)
+{
+	dma_async_issue_pending(drv_data->rx_chan);
+	dma_async_issue_pending(drv_data->tx_chan);
+
+	atomic_set(&drv_data->dma_running, 1);
+}
+
+int pxa2xx_spi_dma_setup(struct driver_data *drv_data)
+{
+	struct pxa2xx_spi_master *pdata = drv_data->master_info;
+	dma_cap_mask_t mask;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+
+	drv_data->dummy = devm_kzalloc(&drv_data->pdev->dev, SZ_2K, GFP_KERNEL);
+	if (!drv_data->dummy)
+		return -ENOMEM;
+
+	drv_data->tx_chan = dma_request_channel(mask, pxa2xx_spi_dma_filter,
+						pdata);
+	if (!drv_data->tx_chan)
+		return -ENODEV;
+
+	drv_data->rx_chan = dma_request_channel(mask, pxa2xx_spi_dma_filter,
+						pdata);
+	if (!drv_data->rx_chan) {
+		dma_release_channel(drv_data->tx_chan);
+		drv_data->tx_chan = NULL;
+		return -ENODEV;
+	}
+
+	return 0;
+}
+
+void pxa2xx_spi_dma_release(struct driver_data *drv_data)
+{
+	if (drv_data->rx_chan) {
+		dmaengine_terminate_all(drv_data->rx_chan);
+		dma_release_channel(drv_data->rx_chan);
+		sg_free_table(&drv_data->rx_sgt);
+		drv_data->rx_chan = NULL;
+	}
+	if (drv_data->tx_chan) {
+		dmaengine_terminate_all(drv_data->tx_chan);
+		dma_release_channel(drv_data->tx_chan);
+		sg_free_table(&drv_data->tx_sgt);
+		drv_data->tx_chan = NULL;
+	}
+}
+
+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,
+					   u32 *threshold)
+{
+	struct pxa2xx_spi_chip *chip_info = spi->controller_data;
+
+	/*
+	 * If the DMA burst size is given in chip_info we use that,
+	 * otherwise we use the default. Also we use the default FIFO
+	 * thresholds for now.
+	 */
+	*burst_code = chip_info ? chip_info->dma_burst_size : 16;
+	*threshold = SSCR1_RxTresh(RX_THRESH_DFLT)
+		   | SSCR1_TxTresh(TX_THRESH_DFLT);
+
+	return 0;
+}
diff --git a/drivers/spi/spi-pxa2xx-pxadma.c b/drivers/spi/spi-pxa2xx-pxadma.c
new file mode 100644
index 0000000..2916efc
--- /dev/null
+++ b/drivers/spi/spi-pxa2xx-pxadma.c
@@ -0,0 +1,490 @@
+/*
+ * PXA2xx SPI private DMA support.
+ *
+ * Copyright (C) 2005 Stephen Street / StreetFire Sound Labs
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/dma-mapping.h>
+#include <linux/pxa2xx_ssp.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/pxa2xx_spi.h>
+
+#include "spi-pxa2xx.h"
+
+#define DMA_INT_MASK		(DCSR_ENDINTR | DCSR_STARTINTR | DCSR_BUSERR)
+#define RESET_DMA_CHANNEL	(DCSR_NODESC | DMA_INT_MASK)
+
+bool pxa2xx_spi_dma_is_possible(size_t len)
+{
+	/* Try to map dma buffer and do a dma transfer if successful, but
+	 * only if the length is non-zero and less than MAX_DMA_LEN.
+	 *
+	 * Zero-length non-descriptor DMA is illegal on PXA2xx; force use
+	 * of PIO instead.  Care is needed above because the transfer may
+	 * have have been passed with buffers that are already dma mapped.
+	 * A zero-length transfer in PIO mode will not try to write/read
+	 * to/from the buffers
+	 *
+	 * REVISIT large transfers are exactly where we most want to be
+	 * using DMA.  If this happens much, split those transfers into
+	 * multiple DMA segments rather than forcing PIO.
+	 */
+	return len > 0 && len <= MAX_DMA_LEN;
+}
+
+int pxa2xx_spi_map_dma_buffers(struct driver_data *drv_data)
+{
+	struct spi_message *msg = drv_data->cur_msg;
+	struct device *dev = &msg->spi->dev;
+
+	if (!drv_data->cur_chip->enable_dma)
+		return 0;
+
+	if (msg->is_dma_mapped)
+		return  drv_data->rx_dma && drv_data->tx_dma;
+
+	if (!IS_DMA_ALIGNED(drv_data->rx) || !IS_DMA_ALIGNED(drv_data->tx))
+		return 0;
+
+	/* Modify setup if rx buffer is null */
+	if (drv_data->rx == NULL) {
+		*drv_data->null_dma_buf = 0;
+		drv_data->rx = drv_data->null_dma_buf;
+		drv_data->rx_map_len = 4;
+	} else
+		drv_data->rx_map_len = drv_data->len;
+
+
+	/* Modify setup if tx buffer is null */
+	if (drv_data->tx == NULL) {
+		*drv_data->null_dma_buf = 0;
+		drv_data->tx = drv_data->null_dma_buf;
+		drv_data->tx_map_len = 4;
+	} else
+		drv_data->tx_map_len = drv_data->len;
+
+	/* Stream map the tx buffer. Always do DMA_TO_DEVICE first
+	 * so we flush the cache *before* invalidating it, in case
+	 * the tx and rx buffers overlap.
+	 */
+	drv_data->tx_dma = dma_map_single(dev, drv_data->tx,
+					drv_data->tx_map_len, DMA_TO_DEVICE);
+	if (dma_mapping_error(dev, drv_data->tx_dma))
+		return 0;
+
+	/* Stream map the rx buffer */
+	drv_data->rx_dma = dma_map_single(dev, drv_data->rx,
+					drv_data->rx_map_len, DMA_FROM_DEVICE);
+	if (dma_mapping_error(dev, drv_data->rx_dma)) {
+		dma_unmap_single(dev, drv_data->tx_dma,
+					drv_data->tx_map_len, DMA_TO_DEVICE);
+		return 0;
+	}
+
+	return 1;
+}
+
+static void pxa2xx_spi_unmap_dma_buffers(struct driver_data *drv_data)
+{
+	struct device *dev;
+
+	if (!drv_data->dma_mapped)
+		return;
+
+	if (!drv_data->cur_msg->is_dma_mapped) {
+		dev = &drv_data->cur_msg->spi->dev;
+		dma_unmap_single(dev, drv_data->rx_dma,
+					drv_data->rx_map_len, DMA_FROM_DEVICE);
+		dma_unmap_single(dev, drv_data->tx_dma,
+					drv_data->tx_map_len, DMA_TO_DEVICE);
+	}
+
+	drv_data->dma_mapped = 0;
+}
+
+static int wait_ssp_rx_stall(void const __iomem *ioaddr)
+{
+	unsigned long limit = loops_per_jiffy << 1;
+
+	while ((read_SSSR(ioaddr) & SSSR_BSY) && --limit)
+		cpu_relax();
+
+	return limit;
+}
+
+static int wait_dma_channel_stop(int channel)
+{
+	unsigned long limit = loops_per_jiffy << 1;
+
+	while (!(DCSR(channel) & DCSR_STOPSTATE) && --limit)
+		cpu_relax();
+
+	return limit;
+}
+
+static void pxa2xx_spi_dma_error_stop(struct driver_data *drv_data,
+				      const char *msg)
+{
+	void __iomem *reg = drv_data->ioaddr;
+
+	/* Stop and reset */
+	DCSR(drv_data->rx_channel) = RESET_DMA_CHANNEL;
+	DCSR(drv_data->tx_channel) = RESET_DMA_CHANNEL;
+	write_SSSR_CS(drv_data, drv_data->clear_sr);
+	write_SSCR1(read_SSCR1(reg) & ~drv_data->dma_cr1, reg);
+	if (!pxa25x_ssp_comp(drv_data))
+		write_SSTO(0, reg);
+	pxa2xx_spi_flush(drv_data);
+	write_SSCR0(read_SSCR0(reg) & ~SSCR0_SSE, reg);
+
+	pxa2xx_spi_unmap_dma_buffers(drv_data);
+
+	dev_err(&drv_data->pdev->dev, "%s\n", msg);
+
+	drv_data->cur_msg->state = ERROR_STATE;
+	tasklet_schedule(&drv_data->pump_transfers);
+}
+
+static void pxa2xx_spi_dma_transfer_complete(struct driver_data *drv_data)
+{
+	void __iomem *reg = drv_data->ioaddr;
+	struct spi_message *msg = drv_data->cur_msg;
+
+	/* Clear and disable interrupts on SSP and DMA channels*/
+	write_SSCR1(read_SSCR1(reg) & ~drv_data->dma_cr1, reg);
+	write_SSSR_CS(drv_data, drv_data->clear_sr);
+	DCSR(drv_data->tx_channel) = RESET_DMA_CHANNEL;
+	DCSR(drv_data->rx_channel) = RESET_DMA_CHANNEL;
+
+	if (wait_dma_channel_stop(drv_data->rx_channel) == 0)
+		dev_err(&drv_data->pdev->dev,
+			"dma_handler: dma rx channel stop failed\n");
+
+	if (wait_ssp_rx_stall(drv_data->ioaddr) == 0)
+		dev_err(&drv_data->pdev->dev,
+			"dma_transfer: ssp rx stall failed\n");
+
+	pxa2xx_spi_unmap_dma_buffers(drv_data);
+
+	/* update the buffer pointer for the amount completed in dma */
+	drv_data->rx += drv_data->len -
+			(DCMD(drv_data->rx_channel) & DCMD_LENGTH);
+
+	/* read trailing data from fifo, it does not matter how many
+	 * bytes are in the fifo just read until buffer is full
+	 * or fifo is empty, which ever occurs first */
+	drv_data->read(drv_data);
+
+	/* return count of what was actually read */
+	msg->actual_length += drv_data->len -
+				(drv_data->rx_end - drv_data->rx);
+
+	/* Transfer delays and chip select release are
+	 * handled in pump_transfers or giveback
+	 */
+
+	/* Move to next transfer */
+	msg->state = pxa2xx_spi_next_transfer(drv_data);
+
+	/* Schedule transfer tasklet */
+	tasklet_schedule(&drv_data->pump_transfers);
+}
+
+void pxa2xx_spi_dma_handler(int channel, void *data)
+{
+	struct driver_data *drv_data = data;
+	u32 irq_status = DCSR(channel) & DMA_INT_MASK;
+
+	if (irq_status & DCSR_BUSERR) {
+
+		if (channel == drv_data->tx_channel)
+			pxa2xx_spi_dma_error_stop(drv_data,
+				"dma_handler: bad bus address on tx channel");
+		else
+			pxa2xx_spi_dma_error_stop(drv_data,
+				"dma_handler: bad bus address on rx channel");
+		return;
+	}
+
+	/* PXA255x_SSP has no timeout interrupt, wait for tailing bytes */
+	if ((channel == drv_data->tx_channel)
+		&& (irq_status & DCSR_ENDINTR)
+		&& (drv_data->ssp_type == PXA25x_SSP)) {
+
+		/* Wait for rx to stall */
+		if (wait_ssp_rx_stall(drv_data->ioaddr) == 0)
+			dev_err(&drv_data->pdev->dev,
+				"dma_handler: ssp rx stall failed\n");
+
+		/* finish this transfer, start the next */
+		pxa2xx_spi_dma_transfer_complete(drv_data);
+	}
+}
+
+irqreturn_t pxa2xx_spi_dma_transfer(struct driver_data *drv_data)
+{
+	u32 irq_status;
+	void __iomem *reg = drv_data->ioaddr;
+
+	irq_status = read_SSSR(reg) & drv_data->mask_sr;
+	if (irq_status & SSSR_ROR) {
+		pxa2xx_spi_dma_error_stop(drv_data,
+					  "dma_transfer: fifo overrun");
+		return IRQ_HANDLED;
+	}
+
+	/* Check for false positive timeout */
+	if ((irq_status & SSSR_TINT)
+		&& (DCSR(drv_data->tx_channel) & DCSR_RUN)) {
+		write_SSSR(SSSR_TINT, reg);
+		return IRQ_HANDLED;
+	}
+
+	if (irq_status & SSSR_TINT || drv_data->rx == drv_data->rx_end) {
+
+		/* Clear and disable timeout interrupt, do the rest in
+		 * dma_transfer_complete */
+		if (!pxa25x_ssp_comp(drv_data))
+			write_SSTO(0, reg);
+
+		/* finish this transfer, start the next */
+		pxa2xx_spi_dma_transfer_complete(drv_data);
+
+		return IRQ_HANDLED;
+	}
+
+	/* Opps problem detected */
+	return IRQ_NONE;
+}
+
+int pxa2xx_spi_dma_prepare(struct driver_data *drv_data, u32 dma_burst)
+{
+	u32 dma_width;
+
+	switch (drv_data->n_bytes) {
+	case 1:
+		dma_width = DCMD_WIDTH1;
+		break;
+	case 2:
+		dma_width = DCMD_WIDTH2;
+		break;
+	default:
+		dma_width = DCMD_WIDTH4;
+		break;
+	}
+
+	/* Setup rx DMA Channel */
+	DCSR(drv_data->rx_channel) = RESET_DMA_CHANNEL;
+	DSADR(drv_data->rx_channel) = drv_data->ssdr_physical;
+	DTADR(drv_data->rx_channel) = drv_data->rx_dma;
+	if (drv_data->rx == drv_data->null_dma_buf)
+		/* No target address increment */
+		DCMD(drv_data->rx_channel) = DCMD_FLOWSRC
+						| dma_width
+						| dma_burst
+						| drv_data->len;
+	else
+		DCMD(drv_data->rx_channel) = DCMD_INCTRGADDR
+						| DCMD_FLOWSRC
+						| dma_width
+						| dma_burst
+						| drv_data->len;
+
+	/* Setup tx DMA Channel */
+	DCSR(drv_data->tx_channel) = RESET_DMA_CHANNEL;
+	DSADR(drv_data->tx_channel) = drv_data->tx_dma;
+	DTADR(drv_data->tx_channel) = drv_data->ssdr_physical;
+	if (drv_data->tx == drv_data->null_dma_buf)
+		/* No source address increment */
+		DCMD(drv_data->tx_channel) = DCMD_FLOWTRG
+						| dma_width
+						| dma_burst
+						| drv_data->len;
+	else
+		DCMD(drv_data->tx_channel) = DCMD_INCSRCADDR
+						| DCMD_FLOWTRG
+						| dma_width
+						| dma_burst
+						| drv_data->len;
+
+	/* Enable dma end irqs on SSP to detect end of transfer */
+	if (drv_data->ssp_type == PXA25x_SSP)
+		DCMD(drv_data->tx_channel) |= DCMD_ENDIRQEN;
+
+	return 0;
+}
+
+void pxa2xx_spi_dma_start(struct driver_data *drv_data)
+{
+	DCSR(drv_data->rx_channel) |= DCSR_RUN;
+	DCSR(drv_data->tx_channel) |= DCSR_RUN;
+}
+
+int pxa2xx_spi_dma_setup(struct driver_data *drv_data)
+{
+	struct device *dev = &drv_data->pdev->dev;
+	struct ssp_device *ssp = drv_data->ssp;
+
+	/* Get two DMA channels	(rx and tx) */
+	drv_data->rx_channel = pxa_request_dma("pxa2xx_spi_ssp_rx",
+						DMA_PRIO_HIGH,
+						pxa2xx_spi_dma_handler,
+						drv_data);
+	if (drv_data->rx_channel < 0) {
+		dev_err(dev, "problem (%d) requesting rx channel\n",
+			drv_data->rx_channel);
+		return -ENODEV;
+	}
+	drv_data->tx_channel = pxa_request_dma("pxa2xx_spi_ssp_tx",
+						DMA_PRIO_MEDIUM,
+						pxa2xx_spi_dma_handler,
+						drv_data);
+	if (drv_data->tx_channel < 0) {
+		dev_err(dev, "problem (%d) requesting tx channel\n",
+			drv_data->tx_channel);
+		pxa_free_dma(drv_data->rx_channel);
+		return -ENODEV;
+	}
+
+	DRCMR(ssp->drcmr_rx) = DRCMR_MAPVLD | drv_data->rx_channel;
+	DRCMR(ssp->drcmr_tx) = DRCMR_MAPVLD | drv_data->tx_channel;
+
+	return 0;
+}
+
+void pxa2xx_spi_dma_release(struct driver_data *drv_data)
+{
+	struct ssp_device *ssp = drv_data->ssp;
+
+	DRCMR(ssp->drcmr_rx) = 0;
+	DRCMR(ssp->drcmr_tx) = 0;
+
+	if (drv_data->tx_channel != 0)
+		pxa_free_dma(drv_data->tx_channel);
+	if (drv_data->rx_channel != 0)
+		pxa_free_dma(drv_data->rx_channel);
+}
+
+void pxa2xx_spi_dma_resume(struct driver_data *drv_data)
+{
+	if (drv_data->rx_channel != -1)
+		DRCMR(drv_data->ssp->drcmr_rx) =
+			DRCMR_MAPVLD | drv_data->rx_channel;
+	if (drv_data->tx_channel != -1)
+		DRCMR(drv_data->ssp->drcmr_tx) =
+			DRCMR_MAPVLD | drv_data->tx_channel;
+}
+
+int pxa2xx_spi_set_dma_burst_and_threshold(struct chip_data *chip,
+					   struct spi_device *spi,
+					   u8 bits_per_word, u32 *burst_code,
+					   u32 *threshold)
+{
+	struct pxa2xx_spi_chip *chip_info =
+			(struct pxa2xx_spi_chip *)spi->controller_data;
+	int bytes_per_word;
+	int burst_bytes;
+	int thresh_words;
+	int req_burst_size;
+	int retval = 0;
+
+	/* Set the threshold (in registers) to equal the same amount of data
+	 * as represented by burst size (in bytes).  The computation below
+	 * is (burst_size rounded up to nearest 8 byte, word or long word)
+	 * divided by (bytes/register); the tx threshold is the inverse of
+	 * the rx, so that there will always be enough data in the rx fifo
+	 * to satisfy a burst, and there will always be enough space in the
+	 * tx fifo to accept a burst (a tx burst will overwrite the fifo if
+	 * there is not enough space), there must always remain enough empty
+	 * space in the rx fifo for any data loaded to the tx fifo.
+	 * Whenever burst_size (in bytes) equals bits/word, the fifo threshold
+	 * will be 8, or half the fifo;
+	 * The threshold can only be set to 2, 4 or 8, but not 16, because
+	 * to burst 16 to the tx fifo, the fifo would have to be empty;
+	 * however, the minimum fifo trigger level is 1, and the tx will
+	 * request service when the fifo is at this level, with only 15 spaces.
+	 */
+
+	/* find bytes/word */
+	if (bits_per_word <= 8)
+		bytes_per_word = 1;
+	else if (bits_per_word <= 16)
+		bytes_per_word = 2;
+	else
+		bytes_per_word = 4;
+
+	/* use struct pxa2xx_spi_chip->dma_burst_size if available */
+	if (chip_info)
+		req_burst_size = chip_info->dma_burst_size;
+	else {
+		switch (chip->dma_burst_size) {
+		default:
+			/* if the default burst size is not set,
+			 * do it now */
+			chip->dma_burst_size = DCMD_BURST8;
+		case DCMD_BURST8:
+			req_burst_size = 8;
+			break;
+		case DCMD_BURST16:
+			req_burst_size = 16;
+			break;
+		case DCMD_BURST32:
+			req_burst_size = 32;
+			break;
+		}
+	}
+	if (req_burst_size <= 8) {
+		*burst_code = DCMD_BURST8;
+		burst_bytes = 8;
+	} else if (req_burst_size <= 16) {
+		if (bytes_per_word == 1) {
+			/* don't burst more than 1/2 the fifo */
+			*burst_code = DCMD_BURST8;
+			burst_bytes = 8;
+			retval = 1;
+		} else {
+			*burst_code = DCMD_BURST16;
+			burst_bytes = 16;
+		}
+	} else {
+		if (bytes_per_word == 1) {
+			/* don't burst more than 1/2 the fifo */
+			*burst_code = DCMD_BURST8;
+			burst_bytes = 8;
+			retval = 1;
+		} else if (bytes_per_word == 2) {
+			/* don't burst more than 1/2 the fifo */
+			*burst_code = DCMD_BURST16;
+			burst_bytes = 16;
+			retval = 1;
+		} else {
+			*burst_code = DCMD_BURST32;
+			burst_bytes = 32;
+		}
+	}
+
+	thresh_words = burst_bytes / bytes_per_word;
+
+	/* thresh_words will be between 2 and 8 */
+	*threshold = (SSCR1_RxTresh(thresh_words) & SSCR1_RFT)
+			| (SSCR1_TxTresh(16-thresh_words) & SSCR1_TFT);
+
+	return retval;
+}
diff --git a/drivers/spi/spi-pxa2xx.c b/drivers/spi/spi-pxa2xx.c
index 304cf6e..90b27a3 100644
--- a/drivers/spi/spi-pxa2xx.c
+++ b/drivers/spi/spi-pxa2xx.c
@@ -1,5 +1,6 @@
 /*
  * Copyright (C) 2005 Stephen Street / StreetFire Sound Labs
+ * Copyright (C) 2013, Intel Corporation
  *
  * 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
@@ -24,18 +25,20 @@
 #include <linux/interrupt.h>
 #include <linux/platform_device.h>
 #include <linux/spi/pxa2xx_spi.h>
-#include <linux/dma-mapping.h>
 #include <linux/spi/spi.h>
 #include <linux/workqueue.h>
 #include <linux/delay.h>
 #include <linux/gpio.h>
 #include <linux/slab.h>
 #include <linux/clk.h>
+#include <linux/pm_runtime.h>
+#include <linux/acpi.h>
 
 #include <asm/io.h>
 #include <asm/irq.h>
 #include <asm/delay.h>
 
+#include "spi-pxa2xx.h"
 
 MODULE_AUTHOR("Stephen Street");
 MODULE_DESCRIPTION("PXA2xx SSP SPI Controller");
@@ -46,12 +49,6 @@ MODULE_ALIAS("platform:pxa2xx-spi");
 
 #define TIMOUT_DFLT		1000
 
-#define DMA_INT_MASK		(DCSR_ENDINTR | DCSR_STARTINTR | DCSR_BUSERR)
-#define RESET_DMA_CHANNEL	(DCSR_NODESC | DMA_INT_MASK)
-#define IS_DMA_ALIGNED(x)	IS_ALIGNED((unsigned long)(x), DMA_ALIGNMENT)
-#define MAX_DMA_LEN		8191
-#define DMA_ALIGNMENT		8
-
 /*
  * for testing SSCR1 changes that require SSP restart, basically
  * everything except the service and interrupt enables, the pxa270 developer
@@ -66,105 +63,97 @@ MODULE_ALIAS("platform:pxa2xx-spi");
 				| SSCR1_RFT | SSCR1_TFT | SSCR1_MWDS \
 				| SSCR1_SPH | SSCR1_SPO | SSCR1_LBM)
 
-#define DEFINE_SSP_REG(reg, off) \
-static inline u32 read_##reg(void const __iomem *p) \
-{ return __raw_readl(p + (off)); } \
-\
-static inline void write_##reg(u32 v, void __iomem *p) \
-{ __raw_writel(v, p + (off)); }
-
-DEFINE_SSP_REG(SSCR0, 0x00)
-DEFINE_SSP_REG(SSCR1, 0x04)
-DEFINE_SSP_REG(SSSR, 0x08)
-DEFINE_SSP_REG(SSITR, 0x0c)
-DEFINE_SSP_REG(SSDR, 0x10)
-DEFINE_SSP_REG(SSTO, 0x28)
-DEFINE_SSP_REG(SSPSP, 0x2c)
-
-#define START_STATE ((void*)0)
-#define RUNNING_STATE ((void*)1)
-#define DONE_STATE ((void*)2)
-#define ERROR_STATE ((void*)-1)
-
-struct driver_data {
-	/* Driver model hookup */
-	struct platform_device *pdev;
-
-	/* SSP Info */
-	struct ssp_device *ssp;
+#define LPSS_RX_THRESH_DFLT	64
+#define LPSS_TX_LOTHRESH_DFLT	160
+#define LPSS_TX_HITHRESH_DFLT	224
 
-	/* SPI framework hookup */
-	enum pxa_ssp_type ssp_type;
-	struct spi_master *master;
+/* Offset from drv_data->lpss_base */
+#define SPI_CS_CONTROL		0x18
+#define SPI_CS_CONTROL_SW_MODE	BIT(0)
+#define SPI_CS_CONTROL_CS_HIGH	BIT(1)
 
-	/* PXA hookup */
-	struct pxa2xx_spi_master *master_info;
-
-	/* DMA setup stuff */
-	int rx_channel;
-	int tx_channel;
-	u32 *null_dma_buf;
-
-	/* SSP register addresses */
-	void __iomem *ioaddr;
-	u32 ssdr_physical;
-
-	/* SSP masks*/
-	u32 dma_cr1;
-	u32 int_cr1;
-	u32 clear_sr;
-	u32 mask_sr;
-
-	/* Maximun clock rate */
-	unsigned long max_clk_rate;
-
-	/* Message Transfer pump */
-	struct tasklet_struct pump_transfers;
-
-	/* Current message transfer state info */
-	struct spi_message* cur_msg;
-	struct spi_transfer* cur_transfer;
-	struct chip_data *cur_chip;
-	size_t len;
-	void *tx;
-	void *tx_end;
-	void *rx;
-	void *rx_end;
-	int dma_mapped;
-	dma_addr_t rx_dma;
-	dma_addr_t tx_dma;
-	size_t rx_map_len;
-	size_t tx_map_len;
-	u8 n_bytes;
-	u32 dma_width;
-	int (*write)(struct driver_data *drv_data);
-	int (*read)(struct driver_data *drv_data);
-	irqreturn_t (*transfer_handler)(struct driver_data *drv_data);
-	void (*cs_control)(u32 command);
-};
+static bool is_lpss_ssp(const struct driver_data *drv_data)
+{
+	return drv_data->ssp_type == LPSS_SSP;
+}
 
-struct chip_data {
-	u32 cr0;
-	u32 cr1;
-	u32 psp;
-	u32 timeout;
-	u8 n_bytes;
-	u32 dma_width;
-	u32 dma_burst_size;
-	u32 threshold;
-	u32 dma_threshold;
-	u8 enable_dma;
-	u8 bits_per_word;
-	u32 speed_hz;
-	union {
-		int gpio_cs;
-		unsigned int frm;
-	};
-	int gpio_cs_inverted;
-	int (*write)(struct driver_data *drv_data);
-	int (*read)(struct driver_data *drv_data);
-	void (*cs_control)(u32 command);
-};
+/*
+ * Read and write LPSS SSP private registers. Caller must first check that
+ * is_lpss_ssp() returns true before these can be called.
+ */
+static u32 __lpss_ssp_read_priv(struct driver_data *drv_data, unsigned offset)
+{
+	WARN_ON(!drv_data->lpss_base);
+	return readl(drv_data->lpss_base + offset);
+}
+
+static void __lpss_ssp_write_priv(struct driver_data *drv_data,
+				  unsigned offset, u32 value)
+{
+	WARN_ON(!drv_data->lpss_base);
+	writel(value, drv_data->lpss_base + offset);
+}
+
+/*
+ * lpss_ssp_setup - perform LPSS SSP specific setup
+ * @drv_data: pointer to the driver private data
+ *
+ * Perform LPSS SSP specific setup. This function must be called first if
+ * one is going to use LPSS SSP private registers.
+ */
+static void lpss_ssp_setup(struct driver_data *drv_data)
+{
+	unsigned offset = 0x400;
+	u32 value, orig;
+
+	if (!is_lpss_ssp(drv_data))
+		return;
+
+	/*
+	 * Perform auto-detection of the LPSS SSP private registers. They
+	 * can be either at 1k or 2k offset from the base address.
+	 */
+	orig = readl(drv_data->ioaddr + offset + SPI_CS_CONTROL);
+
+	value = orig | SPI_CS_CONTROL_SW_MODE;
+	writel(value, drv_data->ioaddr + offset + SPI_CS_CONTROL);
+	value = readl(drv_data->ioaddr + offset + SPI_CS_CONTROL);
+	if (value != (orig | SPI_CS_CONTROL_SW_MODE)) {
+		offset = 0x800;
+		goto detection_done;
+	}
+
+	value &= ~SPI_CS_CONTROL_SW_MODE;
+	writel(value, drv_data->ioaddr + offset + SPI_CS_CONTROL);
+	value = readl(drv_data->ioaddr + offset + SPI_CS_CONTROL);
+	if (value != orig) {
+		offset = 0x800;
+		goto detection_done;
+	}
+
+detection_done:
+	/* Now set the LPSS base */
+	drv_data->lpss_base = drv_data->ioaddr + offset;
+
+	/* Enable software chip select control */
+	value = SPI_CS_CONTROL_SW_MODE | SPI_CS_CONTROL_CS_HIGH;
+	__lpss_ssp_write_priv(drv_data, SPI_CS_CONTROL, value);
+}
+
+static void lpss_ssp_cs_control(struct driver_data *drv_data, bool enable)
+{
+	u32 value;
+
+	if (!is_lpss_ssp(drv_data))
+		return;
+
+	value = __lpss_ssp_read_priv(drv_data, SPI_CS_CONTROL);
+	if (enable)
+		value &= ~SPI_CS_CONTROL_CS_HIGH;
+	else
+		value |= SPI_CS_CONTROL_CS_HIGH;
+	__lpss_ssp_write_priv(drv_data, SPI_CS_CONTROL, value);
+}
 
 static void cs_assert(struct driver_data *drv_data)
 {
@@ -180,8 +169,12 @@ static void cs_assert(struct driver_data *drv_data)
 		return;
 	}
 
-	if (gpio_is_valid(chip->gpio_cs))
+	if (gpio_is_valid(chip->gpio_cs)) {
 		gpio_set_value(chip->gpio_cs, chip->gpio_cs_inverted);
+		return;
+	}
+
+	lpss_ssp_cs_control(drv_data, true);
 }
 
 static void cs_deassert(struct driver_data *drv_data)
@@ -196,30 +189,15 @@ static void cs_deassert(struct driver_data *drv_data)
 		return;
 	}
 
-	if (gpio_is_valid(chip->gpio_cs))
+	if (gpio_is_valid(chip->gpio_cs)) {
 		gpio_set_value(chip->gpio_cs, !chip->gpio_cs_inverted);
-}
-
-static void write_SSSR_CS(struct driver_data *drv_data, u32 val)
-{
-	void __iomem *reg = drv_data->ioaddr;
-
-	if (drv_data->ssp_type == CE4100_SSP)
-		val |= read_SSSR(reg) & SSSR_ALT_FRM_MASK;
-
-	write_SSSR(val, reg);
-}
+		return;
+	}
 
-static int pxa25x_ssp_comp(struct driver_data *drv_data)
-{
-	if (drv_data->ssp_type == PXA25x_SSP)
-		return 1;
-	if (drv_data->ssp_type == CE4100_SSP)
-		return 1;
-	return 0;
+	lpss_ssp_cs_control(drv_data, false);
 }
 
-static int flush(struct driver_data *drv_data)
+int pxa2xx_spi_flush(struct driver_data *drv_data)
 {
 	unsigned long limit = loops_per_jiffy << 1;
 
@@ -345,7 +323,7 @@ static int u32_reader(struct driver_data *drv_data)
 	return drv_data->rx == drv_data->rx_end;
 }
 
-static void *next_transfer(struct driver_data *drv_data)
+void *pxa2xx_spi_next_transfer(struct driver_data *drv_data)
 {
 	struct spi_message *msg = drv_data->cur_msg;
 	struct spi_transfer *trans = drv_data->cur_transfer;
@@ -361,76 +339,6 @@ static void *next_transfer(struct driver_data *drv_data)
 		return DONE_STATE;
 }
 
-static int map_dma_buffers(struct driver_data *drv_data)
-{
-	struct spi_message *msg = drv_data->cur_msg;
-	struct device *dev = &msg->spi->dev;
-
-	if (!drv_data->cur_chip->enable_dma)
-		return 0;
-
-	if (msg->is_dma_mapped)
-		return  drv_data->rx_dma && drv_data->tx_dma;
-
-	if (!IS_DMA_ALIGNED(drv_data->rx) || !IS_DMA_ALIGNED(drv_data->tx))
-		return 0;
-
-	/* Modify setup if rx buffer is null */
-	if (drv_data->rx == NULL) {
-		*drv_data->null_dma_buf = 0;
-		drv_data->rx = drv_data->null_dma_buf;
-		drv_data->rx_map_len = 4;
-	} else
-		drv_data->rx_map_len = drv_data->len;
-
-
-	/* Modify setup if tx buffer is null */
-	if (drv_data->tx == NULL) {
-		*drv_data->null_dma_buf = 0;
-		drv_data->tx = drv_data->null_dma_buf;
-		drv_data->tx_map_len = 4;
-	} else
-		drv_data->tx_map_len = drv_data->len;
-
-	/* Stream map the tx buffer. Always do DMA_TO_DEVICE first
-	 * so we flush the cache *before* invalidating it, in case
-	 * the tx and rx buffers overlap.
-	 */
-	drv_data->tx_dma = dma_map_single(dev, drv_data->tx,
-					drv_data->tx_map_len, DMA_TO_DEVICE);
-	if (dma_mapping_error(dev, drv_data->tx_dma))
-		return 0;
-
-	/* Stream map the rx buffer */
-	drv_data->rx_dma = dma_map_single(dev, drv_data->rx,
-					drv_data->rx_map_len, DMA_FROM_DEVICE);
-	if (dma_mapping_error(dev, drv_data->rx_dma)) {
-		dma_unmap_single(dev, drv_data->tx_dma,
-					drv_data->tx_map_len, DMA_TO_DEVICE);
-		return 0;
-	}
-
-	return 1;
-}
-
-static void unmap_dma_buffers(struct driver_data *drv_data)
-{
-	struct device *dev;
-
-	if (!drv_data->dma_mapped)
-		return;
-
-	if (!drv_data->cur_msg->is_dma_mapped) {
-		dev = &drv_data->cur_msg->spi->dev;
-		dma_unmap_single(dev, drv_data->rx_dma,
-					drv_data->rx_map_len, DMA_FROM_DEVICE);
-		dma_unmap_single(dev, drv_data->tx_dma,
-					drv_data->tx_map_len, DMA_TO_DEVICE);
-	}
-
-	drv_data->dma_mapped = 0;
-}
-
 /* caller already set message->status; dma and pio irqs are blocked */
 static void giveback(struct driver_data *drv_data)
 {
@@ -483,161 +391,6 @@ static void giveback(struct driver_data *drv_data)
 	drv_data->cur_chip = NULL;
 }
 
-static int wait_ssp_rx_stall(void const __iomem *ioaddr)
-{
-	unsigned long limit = loops_per_jiffy << 1;
-
-	while ((read_SSSR(ioaddr) & SSSR_BSY) && --limit)
-		cpu_relax();
-
-	return limit;
-}
-
-static int wait_dma_channel_stop(int channel)
-{
-	unsigned long limit = loops_per_jiffy << 1;
-
-	while (!(DCSR(channel) & DCSR_STOPSTATE) && --limit)
-		cpu_relax();
-
-	return limit;
-}
-
-static void dma_error_stop(struct driver_data *drv_data, const char *msg)
-{
-	void __iomem *reg = drv_data->ioaddr;
-
-	/* Stop and reset */
-	DCSR(drv_data->rx_channel) = RESET_DMA_CHANNEL;
-	DCSR(drv_data->tx_channel) = RESET_DMA_CHANNEL;
-	write_SSSR_CS(drv_data, drv_data->clear_sr);
-	write_SSCR1(read_SSCR1(reg) & ~drv_data->dma_cr1, reg);
-	if (!pxa25x_ssp_comp(drv_data))
-		write_SSTO(0, reg);
-	flush(drv_data);
-	write_SSCR0(read_SSCR0(reg) & ~SSCR0_SSE, reg);
-
-	unmap_dma_buffers(drv_data);
-
-	dev_err(&drv_data->pdev->dev, "%s\n", msg);
-
-	drv_data->cur_msg->state = ERROR_STATE;
-	tasklet_schedule(&drv_data->pump_transfers);
-}
-
-static void dma_transfer_complete(struct driver_data *drv_data)
-{
-	void __iomem *reg = drv_data->ioaddr;
-	struct spi_message *msg = drv_data->cur_msg;
-
-	/* Clear and disable interrupts on SSP and DMA channels*/
-	write_SSCR1(read_SSCR1(reg) & ~drv_data->dma_cr1, reg);
-	write_SSSR_CS(drv_data, drv_data->clear_sr);
-	DCSR(drv_data->tx_channel) = RESET_DMA_CHANNEL;
-	DCSR(drv_data->rx_channel) = RESET_DMA_CHANNEL;
-
-	if (wait_dma_channel_stop(drv_data->rx_channel) == 0)
-		dev_err(&drv_data->pdev->dev,
-			"dma_handler: dma rx channel stop failed\n");
-
-	if (wait_ssp_rx_stall(drv_data->ioaddr) == 0)
-		dev_err(&drv_data->pdev->dev,
-			"dma_transfer: ssp rx stall failed\n");
-
-	unmap_dma_buffers(drv_data);
-
-	/* update the buffer pointer for the amount completed in dma */
-	drv_data->rx += drv_data->len -
-			(DCMD(drv_data->rx_channel) & DCMD_LENGTH);
-
-	/* read trailing data from fifo, it does not matter how many
-	 * bytes are in the fifo just read until buffer is full
-	 * or fifo is empty, which ever occurs first */
-	drv_data->read(drv_data);
-
-	/* return count of what was actually read */
-	msg->actual_length += drv_data->len -
-				(drv_data->rx_end - drv_data->rx);
-
-	/* Transfer delays and chip select release are
-	 * handled in pump_transfers or giveback
-	 */
-
-	/* Move to next transfer */
-	msg->state = next_transfer(drv_data);
-
-	/* Schedule transfer tasklet */
-	tasklet_schedule(&drv_data->pump_transfers);
-}
-
-static void dma_handler(int channel, void *data)
-{
-	struct driver_data *drv_data = data;
-	u32 irq_status = DCSR(channel) & DMA_INT_MASK;
-
-	if (irq_status & DCSR_BUSERR) {
-
-		if (channel == drv_data->tx_channel)
-			dma_error_stop(drv_data,
-					"dma_handler: "
-					"bad bus address on tx channel");
-		else
-			dma_error_stop(drv_data,
-					"dma_handler: "
-					"bad bus address on rx channel");
-		return;
-	}
-
-	/* PXA255x_SSP has no timeout interrupt, wait for tailing bytes */
-	if ((channel == drv_data->tx_channel)
-		&& (irq_status & DCSR_ENDINTR)
-		&& (drv_data->ssp_type == PXA25x_SSP)) {
-
-		/* Wait for rx to stall */
-		if (wait_ssp_rx_stall(drv_data->ioaddr) == 0)
-			dev_err(&drv_data->pdev->dev,
-				"dma_handler: ssp rx stall failed\n");
-
-		/* finish this transfer, start the next */
-		dma_transfer_complete(drv_data);
-	}
-}
-
-static irqreturn_t dma_transfer(struct driver_data *drv_data)
-{
-	u32 irq_status;
-	void __iomem *reg = drv_data->ioaddr;
-
-	irq_status = read_SSSR(reg) & drv_data->mask_sr;
-	if (irq_status & SSSR_ROR) {
-		dma_error_stop(drv_data, "dma_transfer: fifo overrun");
-		return IRQ_HANDLED;
-	}
-
-	/* Check for false positive timeout */
-	if ((irq_status & SSSR_TINT)
-		&& (DCSR(drv_data->tx_channel) & DCSR_RUN)) {
-		write_SSSR(SSSR_TINT, reg);
-		return IRQ_HANDLED;
-	}
-
-	if (irq_status & SSSR_TINT || drv_data->rx == drv_data->rx_end) {
-
-		/* Clear and disable timeout interrupt, do the rest in
-		 * dma_transfer_complete */
-		if (!pxa25x_ssp_comp(drv_data))
-			write_SSTO(0, reg);
-
-		/* finish this transfer, start the next */
-		dma_transfer_complete(drv_data);
-
-		return IRQ_HANDLED;
-	}
-
-	/* Opps problem detected */
-	return IRQ_NONE;
-}
-
 static void reset_sccr1(struct driver_data *drv_data)
 {
 	void __iomem *reg = drv_data->ioaddr;
@@ -659,7 +412,7 @@ static void int_error_stop(struct driver_data *drv_data, const char* msg)
 	reset_sccr1(drv_data);
 	if (!pxa25x_ssp_comp(drv_data))
 		write_SSTO(0, reg);
-	flush(drv_data);
+	pxa2xx_spi_flush(drv_data);
 	write_SSCR0(read_SSCR0(reg) & ~SSCR0_SSE, reg);
 
 	dev_err(&drv_data->pdev->dev, "%s\n", msg);
@@ -687,7 +440,7 @@ static void int_transfer_complete(struct driver_data *drv_data)
 	 */
 
 	/* Move to next transfer */
-	drv_data->cur_msg->state = next_transfer(drv_data);
+	drv_data->cur_msg->state = pxa2xx_spi_next_transfer(drv_data);
 
 	/* Schedule transfer tasklet */
 	tasklet_schedule(&drv_data->pump_transfers);
@@ -767,10 +520,20 @@ static irqreturn_t ssp_int(int irq, void *dev_id)
 {
 	struct driver_data *drv_data = dev_id;
 	void __iomem *reg = drv_data->ioaddr;
-	u32 sccr1_reg = read_SSCR1(reg);
+	u32 sccr1_reg;
 	u32 mask = drv_data->mask_sr;
 	u32 status;
 
+	/*
+	 * The IRQ might be shared with other peripherals so we must first
+	 * check that are we RPM suspended or not. If we are we assume that
+	 * the IRQ was not for us (we shouldn't be RPM suspended when the
+	 * interrupt is enabled).
+	 */
+	if (pm_runtime_suspended(&drv_data->pdev->dev))
+		return IRQ_NONE;
+
+	sccr1_reg = read_SSCR1(reg);
 	status = read_SSSR(reg);
 
 	/* Ignore possible writes if we don't need to write */
@@ -798,103 +561,6 @@ static irqreturn_t ssp_int(int irq, void *dev_id)
 	return drv_data->transfer_handler(drv_data);
 }
 
-static int set_dma_burst_and_threshold(struct chip_data *chip,
-				struct spi_device *spi,
-				u8 bits_per_word, u32 *burst_code,
-				u32 *threshold)
-{
-	struct pxa2xx_spi_chip *chip_info =
-			(struct pxa2xx_spi_chip *)spi->controller_data;
-	int bytes_per_word;
-	int burst_bytes;
-	int thresh_words;
-	int req_burst_size;
-	int retval = 0;
-
-	/* Set the threshold (in registers) to equal the same amount of data
-	 * as represented by burst size (in bytes).  The computation below
-	 * is (burst_size rounded up to nearest 8 byte, word or long word)
-	 * divided by (bytes/register); the tx threshold is the inverse of
-	 * the rx, so that there will always be enough data in the rx fifo
-	 * to satisfy a burst, and there will always be enough space in the
-	 * tx fifo to accept a burst (a tx burst will overwrite the fifo if
-	 * there is not enough space), there must always remain enough empty
-	 * space in the rx fifo for any data loaded to the tx fifo.
-	 * Whenever burst_size (in bytes) equals bits/word, the fifo threshold
-	 * will be 8, or half the fifo;
-	 * The threshold can only be set to 2, 4 or 8, but not 16, because
-	 * to burst 16 to the tx fifo, the fifo would have to be empty;
-	 * however, the minimum fifo trigger level is 1, and the tx will
-	 * request service when the fifo is at this level, with only 15 spaces.
-	 */
-
-	/* find bytes/word */
-	if (bits_per_word <= 8)
-		bytes_per_word = 1;
-	else if (bits_per_word <= 16)
-		bytes_per_word = 2;
-	else
-		bytes_per_word = 4;
-
-	/* use struct pxa2xx_spi_chip->dma_burst_size if available */
-	if (chip_info)
-		req_burst_size = chip_info->dma_burst_size;
-	else {
-		switch (chip->dma_burst_size) {
-		default:
-			/* if the default burst size is not set,
-			 * do it now */
-			chip->dma_burst_size = DCMD_BURST8;
-		case DCMD_BURST8:
-			req_burst_size = 8;
-			break;
-		case DCMD_BURST16:
-			req_burst_size = 16;
-			break;
-		case DCMD_BURST32:
-			req_burst_size = 32;
-			break;
-		}
-	}
-	if (req_burst_size <= 8) {
-		*burst_code = DCMD_BURST8;
-		burst_bytes = 8;
-	} else if (req_burst_size <= 16) {
-		if (bytes_per_word == 1) {
-			/* don't burst more than 1/2 the fifo */
-			*burst_code = DCMD_BURST8;
-			burst_bytes = 8;
-			retval = 1;
-		} else {
-			*burst_code = DCMD_BURST16;
-			burst_bytes = 16;
-		}
-	} else {
-		if (bytes_per_word == 1) {
-			/* don't burst more than 1/2 the fifo */
-			*burst_code = DCMD_BURST8;
-			burst_bytes = 8;
-			retval = 1;
-		} else if (bytes_per_word == 2) {
-			/* don't burst more than 1/2 the fifo */
-			*burst_code = DCMD_BURST16;
-			burst_bytes = 16;
-			retval = 1;
-		} else {
-			*burst_code = DCMD_BURST32;
-			burst_bytes = 32;
-		}
-	}
-
-	thresh_words = burst_bytes / bytes_per_word;
-
-	/* thresh_words will be between 2 and 8 */
-	*threshold = (SSCR1_RxTresh(thresh_words) & SSCR1_RFT)
-			| (SSCR1_TxTresh(16-thresh_words) & SSCR1_TFT);
-
-	return retval;
-}
-
 static unsigned int ssp_get_clk_div(struct driver_data *drv_data, int rate)
 {
 	unsigned long ssp_clk = drv_data->max_clk_rate;
@@ -956,8 +622,8 @@ static void pump_transfers(unsigned long data)
 			cs_deassert(drv_data);
 	}
 
-	/* Check for transfers that need multiple DMA segments */
-	if (transfer->len > MAX_DMA_LEN && chip->enable_dma) {
+	/* Check if we can DMA this transfer */
+	if (!pxa2xx_spi_dma_is_possible(transfer->len) && chip->enable_dma) {
 
 		/* reject already-mapped transfers; PIO won't always work */
 		if (message->is_dma_mapped
@@ -980,21 +646,20 @@ static void pump_transfers(unsigned long data)
 	}
 
 	/* Setup the transfer state based on the type of transfer */
-	if (flush(drv_data) == 0) {
+	if (pxa2xx_spi_flush(drv_data) == 0) {
 		dev_err(&drv_data->pdev->dev, "pump_transfers: flush failed\n");
 		message->status = -EIO;
 		giveback(drv_data);
 		return;
 	}
 	drv_data->n_bytes = chip->n_bytes;
-	drv_data->dma_width = chip->dma_width;
 	drv_data->tx = (void *)transfer->tx_buf;
 	drv_data->tx_end = drv_data->tx + transfer->len;
 	drv_data->rx = transfer->rx_buf;
 	drv_data->rx_end = drv_data->rx + transfer->len;
 	drv_data->rx_dma = transfer->rx_dma;
 	drv_data->tx_dma = transfer->tx_dma;
-	drv_data->len = transfer->len & DCMD_LENGTH;
+	drv_data->len = transfer->len;
 	drv_data->write = drv_data->tx ? chip->write : null_writer;
 	drv_data->read = drv_data->rx ? chip->read : null_reader;
 
@@ -1015,21 +680,18 @@ static void pump_transfers(unsigned long data)
 
 		if (bits <= 8) {
 			drv_data->n_bytes = 1;
-			drv_data->dma_width = DCMD_WIDTH1;
 			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->dma_width = DCMD_WIDTH2;
 			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->dma_width = DCMD_WIDTH4;
 			drv_data->read = drv_data->read != null_reader ?
 						u32_reader : null_reader;
 			drv_data->write = drv_data->write != null_writer ?
@@ -1038,7 +700,8 @@ static void pump_transfers(unsigned long data)
 		/* if bits/word is changed in dma mode, then must check the
 		 * thresholds and burst also */
 		if (chip->enable_dma) {
-			if (set_dma_burst_and_threshold(chip, message->spi,
+			if (pxa2xx_spi_set_dma_burst_and_threshold(chip,
+							message->spi,
 							bits, &dma_burst,
 							&dma_thresh))
 				if (printk_ratelimit())
@@ -1057,70 +720,21 @@ static void pump_transfers(unsigned long data)
 
 	message->state = RUNNING_STATE;
 
-	/* Try to map dma buffer and do a dma transfer if successful, but
-	 * only if the length is non-zero and less than MAX_DMA_LEN.
-	 *
-	 * Zero-length non-descriptor DMA is illegal on PXA2xx; force use
-	 * of PIO instead.  Care is needed above because the transfer may
-	 * have have been passed with buffers that are already dma mapped.
-	 * A zero-length transfer in PIO mode will not try to write/read
-	 * to/from the buffers
-	 *
-	 * REVISIT large transfers are exactly where we most want to be
-	 * using DMA.  If this happens much, split those transfers into
-	 * multiple DMA segments rather than forcing PIO.
-	 */
 	drv_data->dma_mapped = 0;
-	if (drv_data->len > 0 && drv_data->len <= MAX_DMA_LEN)
-		drv_data->dma_mapped = map_dma_buffers(drv_data);
+	if (pxa2xx_spi_dma_is_possible(drv_data->len))
+		drv_data->dma_mapped = pxa2xx_spi_map_dma_buffers(drv_data);
 	if (drv_data->dma_mapped) {
 
 		/* Ensure we have the correct interrupt handler */
-		drv_data->transfer_handler = dma_transfer;
-
-		/* Setup rx DMA Channel */
-		DCSR(drv_data->rx_channel) = RESET_DMA_CHANNEL;
-		DSADR(drv_data->rx_channel) = drv_data->ssdr_physical;
-		DTADR(drv_data->rx_channel) = drv_data->rx_dma;
-		if (drv_data->rx == drv_data->null_dma_buf)
-			/* No target address increment */
-			DCMD(drv_data->rx_channel) = DCMD_FLOWSRC
-							| drv_data->dma_width
-							| dma_burst
-							| drv_data->len;
-		else
-			DCMD(drv_data->rx_channel) = DCMD_INCTRGADDR
-							| DCMD_FLOWSRC
-							| drv_data->dma_width
-							| dma_burst
-							| drv_data->len;
-
-		/* Setup tx DMA Channel */
-		DCSR(drv_data->tx_channel) = RESET_DMA_CHANNEL;
-		DSADR(drv_data->tx_channel) = drv_data->tx_dma;
-		DTADR(drv_data->tx_channel) = drv_data->ssdr_physical;
-		if (drv_data->tx == drv_data->null_dma_buf)
-			/* No source address increment */
-			DCMD(drv_data->tx_channel) = DCMD_FLOWTRG
-							| drv_data->dma_width
-							| dma_burst
-							| drv_data->len;
-		else
-			DCMD(drv_data->tx_channel) = DCMD_INCSRCADDR
-							| DCMD_FLOWTRG
-							| drv_data->dma_width
-							| dma_burst
-							| drv_data->len;
-
-		/* Enable dma end irqs on SSP to detect end of transfer */
-		if (drv_data->ssp_type == PXA25x_SSP)
-			DCMD(drv_data->tx_channel) |= DCMD_ENDIRQEN;
+		drv_data->transfer_handler = pxa2xx_spi_dma_transfer;
+
+		pxa2xx_spi_dma_prepare(drv_data, dma_burst);
 
 		/* Clear status and start DMA engine */
 		cr1 = chip->cr1 | dma_thresh | drv_data->dma_cr1;
 		write_SSSR(drv_data->clear_sr, reg);
-		DCSR(drv_data->rx_channel) |= DCSR_RUN;
-		DCSR(drv_data->tx_channel) |= DCSR_RUN;
+
+		pxa2xx_spi_dma_start(drv_data);
 	} else {
 		/* Ensure we have the correct interrupt handler	*/
 		drv_data->transfer_handler = interrupt_transfer;
@@ -1130,6 +744,13 @@ static void pump_transfers(unsigned long data)
 		write_SSSR_CS(drv_data, drv_data->clear_sr);
 	}
 
+	if (is_lpss_ssp(drv_data)) {
+		if ((read_SSIRF(reg) & 0xff) != chip->lpss_rx_threshold)
+			write_SSIRF(chip->lpss_rx_threshold, reg);
+		if ((read_SSITF(reg) & 0xffff) != chip->lpss_tx_threshold)
+			write_SSITF(chip->lpss_tx_threshold, reg);
+	}
+
 	/* see if we need to reload the config registers */
 	if ((read_SSCR0(reg) != cr0)
 		|| (read_SSCR1(reg) & SSCR1_CHANGE_MASK) !=
@@ -1177,6 +798,27 @@ static int pxa2xx_spi_transfer_one_message(struct spi_master *master,
 	return 0;
 }
 
+static int pxa2xx_spi_prepare_transfer(struct spi_master *master)
+{
+	struct driver_data *drv_data = spi_master_get_devdata(master);
+
+	pm_runtime_get_sync(&drv_data->pdev->dev);
+	return 0;
+}
+
+static int pxa2xx_spi_unprepare_transfer(struct spi_master *master)
+{
+	struct driver_data *drv_data = spi_master_get_devdata(master);
+
+	/* Disable the SSP now */
+	write_SSCR0(read_SSCR0(drv_data->ioaddr) & ~SSCR0_SSE,
+		    drv_data->ioaddr);
+
+	pm_runtime_mark_last_busy(&drv_data->pdev->dev);
+	pm_runtime_put_autosuspend(&drv_data->pdev->dev);
+	return 0;
+}
+
 static int setup_cs(struct spi_device *spi, struct chip_data *chip,
 		    struct pxa2xx_spi_chip *chip_info)
 {
@@ -1221,8 +863,17 @@ static int setup(struct spi_device *spi)
 	struct chip_data *chip;
 	struct driver_data *drv_data = spi_master_get_devdata(spi->master);
 	unsigned int clk_div;
-	uint tx_thres = TX_THRESH_DFLT;
-	uint rx_thres = RX_THRESH_DFLT;
+	uint tx_thres, tx_hi_thres, rx_thres;
+
+	if (is_lpss_ssp(drv_data)) {
+		tx_thres = LPSS_TX_LOTHRESH_DFLT;
+		tx_hi_thres = LPSS_TX_HITHRESH_DFLT;
+		rx_thres = LPSS_RX_THRESH_DFLT;
+	} else {
+		tx_thres = TX_THRESH_DFLT;
+		tx_hi_thres = 0;
+		rx_thres = RX_THRESH_DFLT;
+	}
 
 	if (!pxa25x_ssp_comp(drv_data)
 		&& (spi->bits_per_word < 4 || spi->bits_per_word > 32)) {
@@ -1262,8 +913,6 @@ static int setup(struct spi_device *spi)
 			chip->gpio_cs = -1;
 		chip->enable_dma = 0;
 		chip->timeout = TIMOUT_DFLT;
-		chip->dma_burst_size = drv_data->master_info->enable_dma ?
-					DCMD_BURST8 : 0;
 	}
 
 	/* protocol drivers may change the chip settings, so...
@@ -1277,23 +926,37 @@ static int setup(struct spi_device *spi)
 			chip->timeout = chip_info->timeout;
 		if (chip_info->tx_threshold)
 			tx_thres = chip_info->tx_threshold;
+		if (chip_info->tx_hi_threshold)
+			tx_hi_thres = chip_info->tx_hi_threshold;
 		if (chip_info->rx_threshold)
 			rx_thres = chip_info->rx_threshold;
 		chip->enable_dma = drv_data->master_info->enable_dma;
 		chip->dma_threshold = 0;
 		if (chip_info->enable_loopback)
 			chip->cr1 = SSCR1_LBM;
+	} else if (ACPI_HANDLE(&spi->dev)) {
+		/*
+		 * Slave devices enumerated from ACPI namespace don't
+		 * usually have chip_info but we still might want to use
+		 * DMA with them.
+		 */
+		chip->enable_dma = drv_data->master_info->enable_dma;
 	}
 
 	chip->threshold = (SSCR1_RxTresh(rx_thres) & SSCR1_RFT) |
 			(SSCR1_TxTresh(tx_thres) & SSCR1_TFT);
 
+	chip->lpss_rx_threshold = SSIRF_RxThresh(rx_thres);
+	chip->lpss_tx_threshold = SSITF_TxLoThresh(tx_thres)
+				| SSITF_TxHiThresh(tx_hi_thres);
+
 	/* set dma burst and threshold outside of chip_info path so that if
 	 * chip_info goes away after setting chip->enable_dma, the
 	 * burst and threshold can still respond to changes in bits_per_word */
 	if (chip->enable_dma) {
 		/* set up legal burst and threshold for dma */
-		if (set_dma_burst_and_threshold(chip, spi, spi->bits_per_word,
+		if (pxa2xx_spi_set_dma_burst_and_threshold(chip, spi,
+						spi->bits_per_word,
 						&chip->dma_burst_size,
 						&chip->dma_threshold)) {
 			dev_warn(&spi->dev, "in setup: DMA burst size reduced "
@@ -1314,6 +977,9 @@ static int setup(struct spi_device *spi)
 	chip->cr1 |= (((spi->mode & SPI_CPHA) != 0) ? SSCR1_SPH : 0)
 			| (((spi->mode & SPI_CPOL) != 0) ? SSCR1_SPO : 0);
 
+	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",
@@ -1328,18 +994,15 @@ static int setup(struct spi_device *spi)
 
 	if (spi->bits_per_word <= 8) {
 		chip->n_bytes = 1;
-		chip->dma_width = DCMD_WIDTH1;
 		chip->read = u8_reader;
 		chip->write = u8_writer;
 	} else if (spi->bits_per_word <= 16) {
 		chip->n_bytes = 2;
-		chip->dma_width = DCMD_WIDTH2;
 		chip->read = u16_reader;
 		chip->write = u16_writer;
 	} else if (spi->bits_per_word <= 32) {
 		chip->cr0 |= SSCR0_EDSS;
 		chip->n_bytes = 4;
-		chip->dma_width = DCMD_WIDTH4;
 		chip->read = u32_reader;
 		chip->write = u32_writer;
 	} else {
@@ -1370,6 +1033,99 @@ static void cleanup(struct spi_device *spi)
 	kfree(chip);
 }
 
+#ifdef CONFIG_ACPI
+static int pxa2xx_spi_acpi_add_dma(struct acpi_resource *res, void *data)
+{
+	struct pxa2xx_spi_master *pdata = data;
+
+	if (res->type == ACPI_RESOURCE_TYPE_FIXED_DMA) {
+		const struct acpi_resource_fixed_dma *dma;
+
+		dma = &res->data.fixed_dma;
+		if (pdata->tx_slave_id < 0) {
+			pdata->tx_slave_id = dma->request_lines;
+			pdata->tx_chan_id = dma->channels;
+		} else if (pdata->rx_slave_id < 0) {
+			pdata->rx_slave_id = dma->request_lines;
+			pdata->rx_chan_id = dma->channels;
+		}
+	}
+
+	/* Tell the ACPI core to skip this resource */
+	return 1;
+}
+
+static struct pxa2xx_spi_master *
+pxa2xx_spi_acpi_get_pdata(struct platform_device *pdev)
+{
+	struct pxa2xx_spi_master *pdata;
+	struct list_head resource_list;
+	struct acpi_device *adev;
+	struct ssp_device *ssp;
+	struct resource *res;
+	int devid;
+
+	if (!ACPI_HANDLE(&pdev->dev) ||
+	    acpi_bus_get_device(ACPI_HANDLE(&pdev->dev), &adev))
+		return NULL;
+
+	pdata = devm_kzalloc(&pdev->dev, sizeof(*ssp), GFP_KERNEL);
+	if (!pdata) {
+		dev_err(&pdev->dev,
+			"failed to allocate memory for platform data\n");
+		return NULL;
+	}
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res)
+		return NULL;
+
+	ssp = &pdata->ssp;
+
+	ssp->phys_base = res->start;
+	ssp->mmio_base = devm_request_and_ioremap(&pdev->dev, res);
+	if (!ssp->mmio_base) {
+		dev_err(&pdev->dev, "failed to ioremap mmio_base\n");
+		return NULL;
+	}
+
+	ssp->clk = devm_clk_get(&pdev->dev, NULL);
+	ssp->irq = platform_get_irq(pdev, 0);
+	ssp->type = LPSS_SSP;
+	ssp->pdev = pdev;
+
+	ssp->port_id = -1;
+	if (adev->pnp.unique_id && !kstrtoint(adev->pnp.unique_id, 0, &devid))
+		ssp->port_id = devid;
+
+	pdata->num_chipselect = 1;
+	pdata->rx_slave_id = -1;
+	pdata->tx_slave_id = -1;
+
+	INIT_LIST_HEAD(&resource_list);
+	acpi_dev_get_resources(adev, &resource_list, pxa2xx_spi_acpi_add_dma,
+			       pdata);
+	acpi_dev_free_resource_list(&resource_list);
+
+	pdata->enable_dma = pdata->rx_slave_id >= 0 && pdata->tx_slave_id >= 0;
+
+	return pdata;
+}
+
+static struct acpi_device_id pxa2xx_spi_acpi_match[] = {
+	{ "INT33C0", 0 },
+	{ "INT33C1", 0 },
+	{ },
+};
+MODULE_DEVICE_TABLE(acpi, pxa2xx_spi_acpi_match);
+#else
+static inline struct pxa2xx_spi_master *
+pxa2xx_spi_acpi_get_pdata(struct platform_device *pdev)
+{
+	return NULL;
+}
+#endif
+
 static int pxa2xx_spi_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
@@ -1381,8 +1137,11 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 
 	platform_info = dev_get_platdata(dev);
 	if (!platform_info) {
-		dev_err(&pdev->dev, "missing platform data\n");
-		return -ENODEV;
+		platform_info = pxa2xx_spi_acpi_get_pdata(pdev);
+		if (!platform_info) {
+			dev_err(&pdev->dev, "missing platform data\n");
+			return -ENODEV;
+		}
 	}
 
 	ssp = pxa_ssp_request(pdev->id, pdev->name);
@@ -1409,8 +1168,9 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 
 	master->dev.parent = &pdev->dev;
 	master->dev.of_node = pdev->dev.of_node;
+	ACPI_HANDLE_SET(&master->dev, ACPI_HANDLE(&pdev->dev));
 	/* the spi->mode bits understood by this driver: */
-	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	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;
@@ -1418,6 +1178,8 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 	master->cleanup = cleanup;
 	master->setup = setup;
 	master->transfer_one_message = pxa2xx_spi_transfer_one_message;
+	master->prepare_transfer_hardware = pxa2xx_spi_prepare_transfer;
+	master->unprepare_transfer_hardware = pxa2xx_spi_unprepare_transfer;
 
 	drv_data->ssp_type = ssp->type;
 	drv_data->null_dma_buf = (u32 *)PTR_ALIGN(&drv_data[1], DMA_ALIGNMENT);
@@ -1431,7 +1193,7 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 		drv_data->mask_sr = SSSR_RFS | SSSR_TFS | SSSR_ROR;
 	} else {
 		drv_data->int_cr1 = SSCR1_TIE | SSCR1_RIE | SSCR1_TINTE;
-		drv_data->dma_cr1 = SSCR1_TSRE | SSCR1_RSRE | SSCR1_TINTE;
+		drv_data->dma_cr1 = DEFAULT_DMA_CR1;
 		drv_data->clear_sr = SSSR_ROR | SSSR_TINT;
 		drv_data->mask_sr = SSSR_TINT | SSSR_RFS | SSSR_TFS | SSSR_ROR;
 	}
@@ -1447,31 +1209,11 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 	drv_data->tx_channel = -1;
 	drv_data->rx_channel = -1;
 	if (platform_info->enable_dma) {
-
-		/* Get two DMA channels	(rx and tx) */
-		drv_data->rx_channel = pxa_request_dma("pxa2xx_spi_ssp_rx",
-							DMA_PRIO_HIGH,
-							dma_handler,
-							drv_data);
-		if (drv_data->rx_channel < 0) {
-			dev_err(dev, "problem (%d) requesting rx channel\n",
-				drv_data->rx_channel);
-			status = -ENODEV;
-			goto out_error_irq_alloc;
-		}
-		drv_data->tx_channel = pxa_request_dma("pxa2xx_spi_ssp_tx",
-							DMA_PRIO_MEDIUM,
-							dma_handler,
-							drv_data);
-		if (drv_data->tx_channel < 0) {
-			dev_err(dev, "problem (%d) requesting tx channel\n",
-				drv_data->tx_channel);
-			status = -ENODEV;
-			goto out_error_dma_alloc;
+		status = pxa2xx_spi_dma_setup(drv_data);
+		if (status) {
+			dev_warn(dev, "failed to setup DMA, using PIO\n");
+			platform_info->enable_dma = false;
 		}
-
-		DRCMR(ssp->drcmr_rx) = DRCMR_MAPVLD | drv_data->rx_channel;
-		DRCMR(ssp->drcmr_tx) = DRCMR_MAPVLD | drv_data->tx_channel;
 	}
 
 	/* Enable SOC clock */
@@ -1492,6 +1234,8 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 		write_SSTO(0, drv_data->ioaddr);
 	write_SSPSP(0, drv_data->ioaddr);
 
+	lpss_ssp_setup(drv_data);
+
 	tasklet_init(&drv_data->pump_transfers, pump_transfers,
 		     (unsigned long)drv_data);
 
@@ -1503,18 +1247,16 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 		goto out_error_clock_enabled;
 	}
 
+	pm_runtime_set_autosuspend_delay(&pdev->dev, 50);
+	pm_runtime_use_autosuspend(&pdev->dev);
+	pm_runtime_set_active(&pdev->dev);
+	pm_runtime_enable(&pdev->dev);
+
 	return status;
 
 out_error_clock_enabled:
 	clk_disable_unprepare(ssp->clk);
-
-out_error_dma_alloc:
-	if (drv_data->tx_channel != -1)
-		pxa_free_dma(drv_data->tx_channel);
-	if (drv_data->rx_channel != -1)
-		pxa_free_dma(drv_data->rx_channel);
-
-out_error_irq_alloc:
+	pxa2xx_spi_dma_release(drv_data);
 	free_irq(ssp->irq, drv_data);
 
 out_error_master_alloc:
@@ -1532,17 +1274,18 @@ static int pxa2xx_spi_remove(struct platform_device *pdev)
 		return 0;
 	ssp = drv_data->ssp;
 
+	pm_runtime_get_sync(&pdev->dev);
+
 	/* Disable the SSP at the peripheral and SOC level */
 	write_SSCR0(0, drv_data->ioaddr);
 	clk_disable_unprepare(ssp->clk);
 
 	/* Release DMA */
-	if (drv_data->master_info->enable_dma) {
-		DRCMR(ssp->drcmr_rx) = 0;
-		DRCMR(ssp->drcmr_tx) = 0;
-		pxa_free_dma(drv_data->tx_channel);
-		pxa_free_dma(drv_data->rx_channel);
-	}
+	if (drv_data->master_info->enable_dma)
+		pxa2xx_spi_dma_release(drv_data);
+
+	pm_runtime_put_noidle(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
 
 	/* Release IRQ */
 	free_irq(ssp->irq, drv_data);
@@ -1589,12 +1332,7 @@ static int pxa2xx_spi_resume(struct device *dev)
 	struct ssp_device *ssp = drv_data->ssp;
 	int status = 0;
 
-	if (drv_data->rx_channel != -1)
-		DRCMR(drv_data->ssp->drcmr_rx) =
-			DRCMR_MAPVLD | drv_data->rx_channel;
-	if (drv_data->tx_channel != -1)
-		DRCMR(drv_data->ssp->drcmr_tx) =
-			DRCMR_MAPVLD | drv_data->tx_channel;
+	pxa2xx_spi_dma_resume(drv_data);
 
 	/* Enable the SSP clock */
 	clk_prepare_enable(ssp->clk);
@@ -1608,20 +1346,38 @@ static int pxa2xx_spi_resume(struct device *dev)
 
 	return 0;
 }
+#endif
+
+#ifdef CONFIG_PM_RUNTIME
+static int pxa2xx_spi_runtime_suspend(struct device *dev)
+{
+	struct driver_data *drv_data = dev_get_drvdata(dev);
+
+	clk_disable_unprepare(drv_data->ssp->clk);
+	return 0;
+}
+
+static int pxa2xx_spi_runtime_resume(struct device *dev)
+{
+	struct driver_data *drv_data = dev_get_drvdata(dev);
+
+	clk_prepare_enable(drv_data->ssp->clk);
+	return 0;
+}
+#endif
 
 static const struct dev_pm_ops pxa2xx_spi_pm_ops = {
-	.suspend	= pxa2xx_spi_suspend,
-	.resume		= pxa2xx_spi_resume,
+	SET_SYSTEM_SLEEP_PM_OPS(pxa2xx_spi_suspend, pxa2xx_spi_resume)
+	SET_RUNTIME_PM_OPS(pxa2xx_spi_runtime_suspend,
+			   pxa2xx_spi_runtime_resume, NULL)
 };
-#endif
 
 static struct platform_driver driver = {
 	.driver = {
 		.name	= "pxa2xx-spi",
 		.owner	= THIS_MODULE,
-#ifdef CONFIG_PM
 		.pm	= &pxa2xx_spi_pm_ops,
-#endif
+		.acpi_match_table = ACPI_PTR(pxa2xx_spi_acpi_match),
 	},
 	.probe = pxa2xx_spi_probe,
 	.remove = pxa2xx_spi_remove,
diff --git a/drivers/spi/spi-pxa2xx.h b/drivers/spi/spi-pxa2xx.h
new file mode 100644
index 0000000..5adc2a1
--- /dev/null
+++ b/drivers/spi/spi-pxa2xx.h
@@ -0,0 +1,221 @@
+/*
+ * Copyright (C) 2005 Stephen Street / StreetFire Sound Labs
+ * Copyright (C) 2013, Intel Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef SPI_PXA2XX_H
+#define SPI_PXA2XX_H
+
+#include <linux/atomic.h>
+#include <linux/dmaengine.h>
+#include <linux/errno.h>
+#include <linux/io.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/pxa2xx_ssp.h>
+#include <linux/scatterlist.h>
+#include <linux/sizes.h>
+#include <linux/spi/spi.h>
+#include <linux/spi/pxa2xx_spi.h>
+
+struct driver_data {
+	/* Driver model hookup */
+	struct platform_device *pdev;
+
+	/* SSP Info */
+	struct ssp_device *ssp;
+
+	/* SPI framework hookup */
+	enum pxa_ssp_type ssp_type;
+	struct spi_master *master;
+
+	/* PXA hookup */
+	struct pxa2xx_spi_master *master_info;
+
+	/* PXA private DMA setup stuff */
+	int rx_channel;
+	int tx_channel;
+	u32 *null_dma_buf;
+
+	/* SSP register addresses */
+	void __iomem *ioaddr;
+	u32 ssdr_physical;
+
+	/* SSP masks*/
+	u32 dma_cr1;
+	u32 int_cr1;
+	u32 clear_sr;
+	u32 mask_sr;
+
+	/* Maximun clock rate */
+	unsigned long max_clk_rate;
+
+	/* Message Transfer pump */
+	struct tasklet_struct pump_transfers;
+
+	/* DMA engine support */
+	struct dma_chan *rx_chan;
+	struct dma_chan *tx_chan;
+	struct sg_table rx_sgt;
+	struct sg_table tx_sgt;
+	int rx_nents;
+	int tx_nents;
+	void *dummy;
+	atomic_t dma_running;
+
+	/* Current message transfer state info */
+	struct spi_message *cur_msg;
+	struct spi_transfer *cur_transfer;
+	struct chip_data *cur_chip;
+	size_t len;
+	void *tx;
+	void *tx_end;
+	void *rx;
+	void *rx_end;
+	int dma_mapped;
+	dma_addr_t rx_dma;
+	dma_addr_t tx_dma;
+	size_t rx_map_len;
+	size_t tx_map_len;
+	u8 n_bytes;
+	int (*write)(struct driver_data *drv_data);
+	int (*read)(struct driver_data *drv_data);
+	irqreturn_t (*transfer_handler)(struct driver_data *drv_data);
+	void (*cs_control)(u32 command);
+
+	void __iomem *lpss_base;
+};
+
+struct chip_data {
+	u32 cr0;
+	u32 cr1;
+	u32 psp;
+	u32 timeout;
+	u8 n_bytes;
+	u32 dma_burst_size;
+	u32 threshold;
+	u32 dma_threshold;
+	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;
+	};
+	int gpio_cs_inverted;
+	int (*write)(struct driver_data *drv_data);
+	int (*read)(struct driver_data *drv_data);
+	void (*cs_control)(u32 command);
+};
+
+#define DEFINE_SSP_REG(reg, off) \
+static inline u32 read_##reg(void const __iomem *p) \
+{ return __raw_readl(p + (off)); } \
+\
+static inline void write_##reg(u32 v, void __iomem *p) \
+{ __raw_writel(v, p + (off)); }
+
+DEFINE_SSP_REG(SSCR0, 0x00)
+DEFINE_SSP_REG(SSCR1, 0x04)
+DEFINE_SSP_REG(SSSR, 0x08)
+DEFINE_SSP_REG(SSITR, 0x0c)
+DEFINE_SSP_REG(SSDR, 0x10)
+DEFINE_SSP_REG(SSTO, 0x28)
+DEFINE_SSP_REG(SSPSP, 0x2c)
+DEFINE_SSP_REG(SSITF, SSITF)
+DEFINE_SSP_REG(SSIRF, SSIRF)
+
+#define START_STATE ((void *)0)
+#define RUNNING_STATE ((void *)1)
+#define DONE_STATE ((void *)2)
+#define ERROR_STATE ((void *)-1)
+
+#define IS_DMA_ALIGNED(x)	IS_ALIGNED((unsigned long)(x), DMA_ALIGNMENT)
+#define DMA_ALIGNMENT		8
+
+static inline int pxa25x_ssp_comp(struct driver_data *drv_data)
+{
+	if (drv_data->ssp_type == PXA25x_SSP)
+		return 1;
+	if (drv_data->ssp_type == CE4100_SSP)
+		return 1;
+	return 0;
+}
+
+static inline void write_SSSR_CS(struct driver_data *drv_data, u32 val)
+{
+	void __iomem *reg = drv_data->ioaddr;
+
+	if (drv_data->ssp_type == CE4100_SSP)
+		val |= read_SSSR(reg) & SSSR_ALT_FRM_MASK;
+
+	write_SSSR(val, reg);
+}
+
+extern int pxa2xx_spi_flush(struct driver_data *drv_data);
+extern void *pxa2xx_spi_next_transfer(struct driver_data *drv_data);
+
+/*
+ * Select the right DMA implementation.
+ */
+#if defined(CONFIG_SPI_PXA2XX_PXADMA)
+#define SPI_PXA2XX_USE_DMA	1
+#define MAX_DMA_LEN		8191
+#define DEFAULT_DMA_CR1		(SSCR1_TSRE | SSCR1_RSRE | SSCR1_TINTE)
+#elif defined(CONFIG_SPI_PXA2XX_DMA)
+#define SPI_PXA2XX_USE_DMA	1
+#define MAX_DMA_LEN		SZ_64K
+#define DEFAULT_DMA_CR1		(SSCR1_TSRE | SSCR1_RSRE | SSCR1_TRAIL)
+#else
+#undef SPI_PXA2XX_USE_DMA
+#define MAX_DMA_LEN		0
+#define DEFAULT_DMA_CR1		0
+#endif
+
+#ifdef SPI_PXA2XX_USE_DMA
+extern bool pxa2xx_spi_dma_is_possible(size_t len);
+extern int pxa2xx_spi_map_dma_buffers(struct driver_data *drv_data);
+extern irqreturn_t pxa2xx_spi_dma_transfer(struct driver_data *drv_data);
+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,
+						  u32 *burst_code,
+						  u32 *threshold);
+#else
+static inline bool pxa2xx_spi_dma_is_possible(size_t len) { return false; }
+static inline int pxa2xx_spi_map_dma_buffers(struct driver_data *drv_data)
+{
+	return 0;
+}
+#define pxa2xx_spi_dma_transfer NULL
+static inline void pxa2xx_spi_dma_prepare(struct driver_data *drv_data,
+					  u32 dma_burst) {}
+static inline void pxa2xx_spi_dma_start(struct driver_data *drv_data) {}
+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,
+							 u32 *burst_code,
+							 u32 *threshold)
+{
+	return -ENODEV;
+}
+#endif
+
+#endif /* SPI_PXA2XX_H */