[PATCH v6 0/2] spi: Add DMA support for ti-qspi

[PATCH v6 0/2] spi: Add DMA support for ti-qspi

[Vignesh R]

This series adds support for DMA during QSPI flash read using memory
mapped mode.

Tested on DRA74 EVM, DRA72 EVM and AM437x SK on linux-next.

Vignesh R (2):
  spi: Add support to handle kmap'd buffers in spi_map_buf()
  spi: spi-ti-qspi: Add DMA support for QSPI mmap read

 drivers/spi/spi-ti-qspi.c | 139 ++++++++++++++++++++++++++++++++++++++++------
 drivers/spi/spi.c         |  17 +++++-
 2 files changed, 136 insertions(+), 20 deletions(-)

-- 
2.9.2

[PATCH v6 1/2] spi: Add support to handle kmap'd buffers in spi_map_buf()

[Vignesh R]

JFFS2 FS might sometime provide kmap'd buffers as destination
buffers to read data from flash. Update spi_map_buf() function to
generate sg_list for such buffers, so that SPI controllers drivers can
use DMA to read data into such buffers.

Signed-off-by: Vignesh R <vigneshr@ti.com>
---

v6: New patch. Extends spi_map_buf() to handle kmap buffers

 drivers/spi/spi.c | 17 ++++++++++++++---
 1 file changed, 14 insertions(+), 3 deletions(-)

diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
index b4f6951256b1..a58e102326a1 100644
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -37,6 +37,7 @@
 #include <linux/kthread.h>
 #include <linux/ioport.h>
 #include <linux/acpi.h>
+#include <linux/highmem.h>
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/spi.h>
@@ -709,6 +710,13 @@ static int spi_map_buf(struct spi_master *master, struct device *dev,
 {
 	const bool vmalloced_buf = is_vmalloc_addr(buf);
 	unsigned int max_seg_size = dma_get_max_seg_size(dev);
+#ifdef CONFIG_HIGHMEM
+	const bool kmap_buf = ((unsigned long)buf >= PKMAP_BASE &&
+				(unsigned long)buf < (PKMAP_BASE +
+					(LAST_PKMAP * PAGE_SIZE)));
+#else
+	const bool kmap_buf = false;
+#endif
 	int desc_len;
 	int sgs;
 	struct page *vm_page;
@@ -716,7 +724,7 @@ static int spi_map_buf(struct spi_master *master, struct device *dev,
 	size_t min;
 	int i, ret;
 
-	if (vmalloced_buf) {
+	if (vmalloced_buf || kmap_buf) {
 		desc_len = min_t(int, max_seg_size, PAGE_SIZE);
 		sgs = DIV_ROUND_UP(len + offset_in_page(buf), desc_len);
 	} else if (virt_addr_valid(buf)) {
@@ -732,10 +740,13 @@ static int spi_map_buf(struct spi_master *master, struct device *dev,
 
 	for (i = 0; i < sgs; i++) {
 
-		if (vmalloced_buf) {
+		if (vmalloced_buf || kmap_buf) {
 			min = min_t(size_t,
 				    len, desc_len - offset_in_page(buf));
-			vm_page = vmalloc_to_page(buf);
+			if (vmalloced_buf)
+				vm_page = vmalloc_to_page(buf);
+			else
+				vm_page = kmap_to_page(buf);
 			if (!vm_page) {
 				sg_free_table(sgt);
 				return -ENOMEM;
-- 
2.9.2

Applied "spi: Add support to handle kmap'd buffers in spi_map_buf()" to the spi tree

[Mark Brown]

The patch

   spi: Add support to handle kmap'd buffers in spi_map_buf()

has been applied to the spi tree at

   git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi.git 

All being well this means that it will be integrated into the linux-next
tree (usually sometime in the next 24 hours) and sent to Linus during
the next merge window (or sooner if it is a bug fix), however if
problems are discovered then the patch may be dropped or reverted.  

You may get further e-mails resulting from automated or manual testing
and review of the tree, please engage with people reporting problems and
send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they
should be sent as incremental updates against current git, existing
patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying
to this mail.

Thanks,
Mark

>From b1b8153cf0aeeb7ae6d4f012b8beb2dcfc92c68a Mon Sep 17 00:00:00 2001
From: Vignesh R <vigneshr@ti.com>
Date: Wed, 17 Aug 2016 15:22:36 +0530
Subject: [PATCH] spi: Add support to handle kmap'd buffers in spi_map_buf()

JFFS2 FS might sometime provide kmap'd buffers as destination
buffers to read data from flash. Update spi_map_buf() function to
generate sg_list for such buffers, so that SPI controllers drivers can
use DMA to read data into such buffers.

Signed-off-by: Vignesh R <vigneshr@ti.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
---
 drivers/spi/spi.c | 17 ++++++++++++++---
 1 file changed, 14 insertions(+), 3 deletions(-)

diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
index 51ad42fad567..2a5dd22efa34 100644
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -37,6 +37,7 @@
 #include <linux/kthread.h>
 #include <linux/ioport.h>
 #include <linux/acpi.h>
+#include <linux/highmem.h>
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/spi.h>
@@ -709,6 +710,13 @@ static int spi_map_buf(struct spi_master *master, struct device *dev,
 {
 	const bool vmalloced_buf = is_vmalloc_addr(buf);
 	unsigned int max_seg_size = dma_get_max_seg_size(dev);
+#ifdef CONFIG_HIGHMEM
+	const bool kmap_buf = ((unsigned long)buf >= PKMAP_BASE &&
+				(unsigned long)buf < (PKMAP_BASE +
+					(LAST_PKMAP * PAGE_SIZE)));
+#else
+	const bool kmap_buf = false;
+#endif
 	int desc_len;
 	int sgs;
 	struct page *vm_page;
@@ -716,7 +724,7 @@ static int spi_map_buf(struct spi_master *master, struct device *dev,
 	size_t min;
 	int i, ret;
 
-	if (vmalloced_buf) {
+	if (vmalloced_buf || kmap_buf) {
 		desc_len = min_t(int, max_seg_size, PAGE_SIZE);
 		sgs = DIV_ROUND_UP(len + offset_in_page(buf), desc_len);
 	} else if (virt_addr_valid(buf)) {
@@ -732,10 +740,13 @@ static int spi_map_buf(struct spi_master *master, struct device *dev,
 
 	for (i = 0; i < sgs; i++) {
 
-		if (vmalloced_buf) {
+		if (vmalloced_buf || kmap_buf) {
 			min = min_t(size_t,
 				    len, desc_len - offset_in_page(buf));
-			vm_page = vmalloc_to_page(buf);
+			if (vmalloced_buf)
+				vm_page = vmalloc_to_page(buf);
+			else
+				vm_page = kmap_to_page(buf);
 			if (!vm_page) {
 				sg_free_table(sgt);
 				return -ENOMEM;
-- 
2.8.1

[PATCH v6 2/2] spi: spi-ti-qspi: Add DMA support for QSPI mmap read

[Vignesh R]

Use mem-to-mem DMA to read from flash when reading in mmap mode. This
gives improved read performance and reduces CPU load.

With this patch the raw-read throughput is ~16MB/s on DRA74 EVM. And CPU
load is <20%. UBIFS read throughput ~13 MB/s.

Signed-off-by: Vignesh R <vigneshr@ti.com>
---

v6: Move handling of kmap'd buffers to spi core code.

 drivers/spi/spi-ti-qspi.c | 139 ++++++++++++++++++++++++++++++++++++++++------
 1 file changed, 122 insertions(+), 17 deletions(-)

diff --git a/drivers/spi/spi-ti-qspi.c b/drivers/spi/spi-ti-qspi.c
index ac0b072815a3..caeac66a3977 100644
--- a/drivers/spi/spi-ti-qspi.c
+++ b/drivers/spi/spi-ti-qspi.c
@@ -41,6 +41,8 @@ struct ti_qspi_regs {
 };
 
 struct ti_qspi {
+	struct completion	transfer_complete;
+
 	/* list synchronization */
 	struct mutex            list_lock;
 
@@ -54,6 +56,9 @@ struct ti_qspi {
 
 	struct ti_qspi_regs     ctx_reg;
 
+	dma_addr_t		mmap_phys_base;
+	struct dma_chan		*rx_chan;
+
 	u32 spi_max_frequency;
 	u32 cmd;
 	u32 dc;
@@ -379,6 +384,72 @@ static int qspi_transfer_msg(struct ti_qspi *qspi, struct spi_transfer *t,
 	return 0;
 }
 
+static void ti_qspi_dma_callback(void *param)
+{
+	struct ti_qspi *qspi = param;
+
+	complete(&qspi->transfer_complete);
+}
+
+static int ti_qspi_dma_xfer(struct ti_qspi *qspi, dma_addr_t dma_dst,
+			    dma_addr_t dma_src, size_t len)
+{
+	struct dma_chan *chan = qspi->rx_chan;
+	struct dma_device *dma_dev = chan->device;
+	dma_cookie_t cookie;
+	enum dma_ctrl_flags flags = DMA_CTRL_ACK | DMA_PREP_INTERRUPT;
+	struct dma_async_tx_descriptor *tx;
+	int ret;
+
+	tx = dma_dev->device_prep_dma_memcpy(chan, dma_dst, dma_src,
+					     len, flags);
+	if (!tx) {
+		dev_err(qspi->dev, "device_prep_dma_memcpy error\n");
+		return -EIO;
+	}
+
+	tx->callback = ti_qspi_dma_callback;
+	tx->callback_param = qspi;
+	cookie = tx->tx_submit(tx);
+
+	ret = dma_submit_error(cookie);
+	if (ret) {
+		dev_err(qspi->dev, "dma_submit_error %d\n", cookie);
+		return -EIO;
+	}
+
+	dma_async_issue_pending(chan);
+	ret = wait_for_completion_timeout(&qspi->transfer_complete,
+					  msecs_to_jiffies(len));
+	if (ret <= 0) {
+		dmaengine_terminate_sync(chan);
+		dev_err(qspi->dev, "DMA wait_for_completion_timeout\n");
+		return -ETIMEDOUT;
+	}
+
+	return 0;
+}
+
+static int ti_qspi_dma_xfer_sg(struct ti_qspi *qspi, struct sg_table rx_sg,
+			       loff_t from)
+{
+	struct scatterlist *sg;
+	dma_addr_t dma_src = qspi->mmap_phys_base + from;
+	dma_addr_t dma_dst;
+	int i, len, ret;
+
+	for_each_sg(rx_sg.sgl, sg, rx_sg.nents, i) {
+		dma_dst = sg_dma_address(sg);
+		len = sg_dma_len(sg);
+		ret = ti_qspi_dma_xfer(qspi, dma_dst, dma_src, len);
+		if (ret)
+			return ret;
+		dma_src += len;
+	}
+
+	return 0;
+}
+
 static void ti_qspi_enable_memory_map(struct spi_device *spi)
 {
 	struct ti_qspi  *qspi = spi_master_get_devdata(spi->master);
@@ -426,7 +497,7 @@ static void ti_qspi_setup_mmap_read(struct spi_device *spi,
 		      QSPI_SPI_SETUP_REG(spi->chip_select));
 }
 
-static int ti_qspi_spi_flash_read(struct  spi_device *spi,
+static int ti_qspi_spi_flash_read(struct spi_device *spi,
 				  struct spi_flash_read_message *msg)
 {
 	struct ti_qspi *qspi = spi_master_get_devdata(spi->master);
@@ -437,9 +508,23 @@ static int ti_qspi_spi_flash_read(struct  spi_device *spi,
 	if (!qspi->mmap_enabled)
 		ti_qspi_enable_memory_map(spi);
 	ti_qspi_setup_mmap_read(spi, msg);
-	memcpy_fromio(msg->buf, qspi->mmap_base + msg->from, msg->len);
+
+	if (qspi->rx_chan) {
+		if (msg->cur_msg_mapped) {
+			ret = ti_qspi_dma_xfer_sg(qspi, msg->rx_sg, msg->from);
+			if (ret)
+				goto err_unlock;
+		} else {
+			dev_err(qspi->dev, "Invalid address for DMA\n");
+			ret = -EIO;
+			goto err_unlock;
+		}
+	} else {
+		memcpy_fromio(msg->buf, qspi->mmap_base + msg->from, msg->len);
+	}
 	msg->retlen = msg->len;
 
+err_unlock:
 	mutex_unlock(&qspi->list_lock);
 
 	return ret;
@@ -536,6 +621,7 @@ static int ti_qspi_probe(struct platform_device *pdev)
 	struct device_node *np = pdev->dev.of_node;
 	u32 max_freq;
 	int ret = 0, num_cs, irq;
+	dma_cap_mask_t mask;
 
 	master = spi_alloc_master(&pdev->dev, sizeof(*qspi));
 	if (!master)
@@ -550,6 +636,7 @@ static int ti_qspi_probe(struct platform_device *pdev)
 	master->dev.of_node = pdev->dev.of_node;
 	master->bits_per_word_mask = SPI_BPW_MASK(32) | SPI_BPW_MASK(16) |
 				     SPI_BPW_MASK(8);
+	master->spi_flash_read = ti_qspi_spi_flash_read;
 
 	if (!of_property_read_u32(np, "num-cs", &num_cs))
 		master->num_chipselect = num_cs;
@@ -592,17 +679,6 @@ static int ti_qspi_probe(struct platform_device *pdev)
 		goto free_master;
 	}
 
-	if (res_mmap) {
-		qspi->mmap_base = devm_ioremap_resource(&pdev->dev,
-							res_mmap);
-		master->spi_flash_read = ti_qspi_spi_flash_read;
-		if (IS_ERR(qspi->mmap_base)) {
-			dev_err(&pdev->dev,
-				"falling back to PIO mode\n");
-			master->spi_flash_read = NULL;
-		}
-	}
-	qspi->mmap_enabled = false;
 
 	if (of_property_read_bool(np, "syscon-chipselects")) {
 		qspi->ctrl_base =
@@ -633,11 +709,37 @@ static int ti_qspi_probe(struct platform_device *pdev)
 	if (!of_property_read_u32(np, "spi-max-frequency", &max_freq))
 		qspi->spi_max_frequency = max_freq;
 
-	ret = devm_spi_register_master(&pdev->dev, master);
-	if (ret)
-		goto free_master;
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_MEMCPY, mask);
 
-	return 0;
+	qspi->rx_chan = dma_request_chan_by_mask(&mask);
+	if (!qspi->rx_chan) {
+		dev_err(qspi->dev,
+			"No Rx DMA available, trying mmap mode\n");
+		ret = 0;
+		goto no_dma;
+	}
+	master->dma_rx = qspi->rx_chan;
+	init_completion(&qspi->transfer_complete);
+	if (res_mmap)
+		qspi->mmap_phys_base = (dma_addr_t)res_mmap->start;
+
+no_dma:
+	if (!qspi->rx_chan && res_mmap) {
+		qspi->mmap_base = devm_ioremap_resource(&pdev->dev, res_mmap);
+		if (IS_ERR(qspi->mmap_base)) {
+			dev_info(&pdev->dev,
+				 "mmap failed with error %ld using PIO mode\n",
+				 PTR_ERR(qspi->mmap_base));
+			qspi->mmap_base = NULL;
+			master->spi_flash_read = NULL;
+		}
+	}
+	qspi->mmap_enabled = false;
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (!ret)
+		return 0;
 
 free_master:
 	spi_master_put(master);
@@ -656,6 +758,9 @@ static int ti_qspi_remove(struct platform_device *pdev)
 	pm_runtime_put_sync(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
 
+	if (qspi->rx_chan)
+		dma_release_channel(qspi->rx_chan);
+
 	return 0;
 }
 
-- 
2.9.2

Applied "spi: spi-ti-qspi: Add DMA support for QSPI mmap read" to the spi tree

[Mark Brown]

The patch

   spi: spi-ti-qspi: Add DMA support for QSPI mmap read

has been applied to the spi tree at

   git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi.git 

All being well this means that it will be integrated into the linux-next
tree (usually sometime in the next 24 hours) and sent to Linus during
the next merge window (or sooner if it is a bug fix), however if
problems are discovered then the patch may be dropped or reverted.  

You may get further e-mails resulting from automated or manual testing
and review of the tree, please engage with people reporting problems and
send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they
should be sent as incremental updates against current git, existing
patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying
to this mail.

Thanks,
Mark

>From 5720ec0a6d2605930934f3f154b048e8be3d8a40 Mon Sep 17 00:00:00 2001
From: Vignesh R <vigneshr@ti.com>
Date: Wed, 17 Aug 2016 15:22:37 +0530
Subject: [PATCH] spi: spi-ti-qspi: Add DMA support for QSPI mmap read

Use mem-to-mem DMA to read from flash when reading in mmap mode. This
gives improved read performance and reduces CPU load.

With this patch the raw-read throughput is ~16MB/s on DRA74 EVM. And CPU
load is <20%. UBIFS read throughput ~13 MB/s.

Signed-off-by: Vignesh R <vigneshr@ti.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
---
 drivers/spi/spi-ti-qspi.c | 139 ++++++++++++++++++++++++++++++++++++++++------
 1 file changed, 122 insertions(+), 17 deletions(-)

diff --git a/drivers/spi/spi-ti-qspi.c b/drivers/spi/spi-ti-qspi.c
index ac0b072815a3..caeac66a3977 100644
--- a/drivers/spi/spi-ti-qspi.c
+++ b/drivers/spi/spi-ti-qspi.c
@@ -41,6 +41,8 @@ struct ti_qspi_regs {
 };
 
 struct ti_qspi {
+	struct completion	transfer_complete;
+
 	/* list synchronization */
 	struct mutex            list_lock;
 
@@ -54,6 +56,9 @@ struct ti_qspi {
 
 	struct ti_qspi_regs     ctx_reg;
 
+	dma_addr_t		mmap_phys_base;
+	struct dma_chan		*rx_chan;
+
 	u32 spi_max_frequency;
 	u32 cmd;
 	u32 dc;
@@ -379,6 +384,72 @@ static int qspi_transfer_msg(struct ti_qspi *qspi, struct spi_transfer *t,
 	return 0;
 }
 
+static void ti_qspi_dma_callback(void *param)
+{
+	struct ti_qspi *qspi = param;
+
+	complete(&qspi->transfer_complete);
+}
+
+static int ti_qspi_dma_xfer(struct ti_qspi *qspi, dma_addr_t dma_dst,
+			    dma_addr_t dma_src, size_t len)
+{
+	struct dma_chan *chan = qspi->rx_chan;
+	struct dma_device *dma_dev = chan->device;
+	dma_cookie_t cookie;
+	enum dma_ctrl_flags flags = DMA_CTRL_ACK | DMA_PREP_INTERRUPT;
+	struct dma_async_tx_descriptor *tx;
+	int ret;
+
+	tx = dma_dev->device_prep_dma_memcpy(chan, dma_dst, dma_src,
+					     len, flags);
+	if (!tx) {
+		dev_err(qspi->dev, "device_prep_dma_memcpy error\n");
+		return -EIO;
+	}
+
+	tx->callback = ti_qspi_dma_callback;
+	tx->callback_param = qspi;
+	cookie = tx->tx_submit(tx);
+
+	ret = dma_submit_error(cookie);
+	if (ret) {
+		dev_err(qspi->dev, "dma_submit_error %d\n", cookie);
+		return -EIO;
+	}
+
+	dma_async_issue_pending(chan);
+	ret = wait_for_completion_timeout(&qspi->transfer_complete,
+					  msecs_to_jiffies(len));
+	if (ret <= 0) {
+		dmaengine_terminate_sync(chan);
+		dev_err(qspi->dev, "DMA wait_for_completion_timeout\n");
+		return -ETIMEDOUT;
+	}
+
+	return 0;
+}
+
+static int ti_qspi_dma_xfer_sg(struct ti_qspi *qspi, struct sg_table rx_sg,
+			       loff_t from)
+{
+	struct scatterlist *sg;
+	dma_addr_t dma_src = qspi->mmap_phys_base + from;
+	dma_addr_t dma_dst;
+	int i, len, ret;
+
+	for_each_sg(rx_sg.sgl, sg, rx_sg.nents, i) {
+		dma_dst = sg_dma_address(sg);
+		len = sg_dma_len(sg);
+		ret = ti_qspi_dma_xfer(qspi, dma_dst, dma_src, len);
+		if (ret)
+			return ret;
+		dma_src += len;
+	}
+
+	return 0;
+}
+
 static void ti_qspi_enable_memory_map(struct spi_device *spi)
 {
 	struct ti_qspi  *qspi = spi_master_get_devdata(spi->master);
@@ -426,7 +497,7 @@ static void ti_qspi_setup_mmap_read(struct spi_device *spi,
 		      QSPI_SPI_SETUP_REG(spi->chip_select));
 }
 
-static int ti_qspi_spi_flash_read(struct  spi_device *spi,
+static int ti_qspi_spi_flash_read(struct spi_device *spi,
 				  struct spi_flash_read_message *msg)
 {
 	struct ti_qspi *qspi = spi_master_get_devdata(spi->master);
@@ -437,9 +508,23 @@ static int ti_qspi_spi_flash_read(struct  spi_device *spi,
 	if (!qspi->mmap_enabled)
 		ti_qspi_enable_memory_map(spi);
 	ti_qspi_setup_mmap_read(spi, msg);
-	memcpy_fromio(msg->buf, qspi->mmap_base + msg->from, msg->len);
+
+	if (qspi->rx_chan) {
+		if (msg->cur_msg_mapped) {
+			ret = ti_qspi_dma_xfer_sg(qspi, msg->rx_sg, msg->from);
+			if (ret)
+				goto err_unlock;
+		} else {
+			dev_err(qspi->dev, "Invalid address for DMA\n");
+			ret = -EIO;
+			goto err_unlock;
+		}
+	} else {
+		memcpy_fromio(msg->buf, qspi->mmap_base + msg->from, msg->len);
+	}
 	msg->retlen = msg->len;
 
+err_unlock:
 	mutex_unlock(&qspi->list_lock);
 
 	return ret;
@@ -536,6 +621,7 @@ static int ti_qspi_probe(struct platform_device *pdev)
 	struct device_node *np = pdev->dev.of_node;
 	u32 max_freq;
 	int ret = 0, num_cs, irq;
+	dma_cap_mask_t mask;
 
 	master = spi_alloc_master(&pdev->dev, sizeof(*qspi));
 	if (!master)
@@ -550,6 +636,7 @@ static int ti_qspi_probe(struct platform_device *pdev)
 	master->dev.of_node = pdev->dev.of_node;
 	master->bits_per_word_mask = SPI_BPW_MASK(32) | SPI_BPW_MASK(16) |
 				     SPI_BPW_MASK(8);
+	master->spi_flash_read = ti_qspi_spi_flash_read;
 
 	if (!of_property_read_u32(np, "num-cs", &num_cs))
 		master->num_chipselect = num_cs;
@@ -592,17 +679,6 @@ static int ti_qspi_probe(struct platform_device *pdev)
 		goto free_master;
 	}
 
-	if (res_mmap) {
-		qspi->mmap_base = devm_ioremap_resource(&pdev->dev,
-							res_mmap);
-		master->spi_flash_read = ti_qspi_spi_flash_read;
-		if (IS_ERR(qspi->mmap_base)) {
-			dev_err(&pdev->dev,
-				"falling back to PIO mode\n");
-			master->spi_flash_read = NULL;
-		}
-	}
-	qspi->mmap_enabled = false;
 
 	if (of_property_read_bool(np, "syscon-chipselects")) {
 		qspi->ctrl_base =
@@ -633,11 +709,37 @@ static int ti_qspi_probe(struct platform_device *pdev)
 	if (!of_property_read_u32(np, "spi-max-frequency", &max_freq))
 		qspi->spi_max_frequency = max_freq;
 
-	ret = devm_spi_register_master(&pdev->dev, master);
-	if (ret)
-		goto free_master;
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_MEMCPY, mask);
 
-	return 0;
+	qspi->rx_chan = dma_request_chan_by_mask(&mask);
+	if (!qspi->rx_chan) {
+		dev_err(qspi->dev,
+			"No Rx DMA available, trying mmap mode\n");
+		ret = 0;
+		goto no_dma;
+	}
+	master->dma_rx = qspi->rx_chan;
+	init_completion(&qspi->transfer_complete);
+	if (res_mmap)
+		qspi->mmap_phys_base = (dma_addr_t)res_mmap->start;
+
+no_dma:
+	if (!qspi->rx_chan && res_mmap) {
+		qspi->mmap_base = devm_ioremap_resource(&pdev->dev, res_mmap);
+		if (IS_ERR(qspi->mmap_base)) {
+			dev_info(&pdev->dev,
+				 "mmap failed with error %ld using PIO mode\n",
+				 PTR_ERR(qspi->mmap_base));
+			qspi->mmap_base = NULL;
+			master->spi_flash_read = NULL;
+		}
+	}
+	qspi->mmap_enabled = false;
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (!ret)
+		return 0;
 
 free_master:
 	spi_master_put(master);
@@ -656,6 +758,9 @@ static int ti_qspi_remove(struct platform_device *pdev)
 	pm_runtime_put_sync(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
 
+	if (qspi->rx_chan)
+		dma_release_channel(qspi->rx_chan);
+
 	return 0;
 }
 
-- 
2.8.1