[PATCH v3 0/2] spi: Add DMA support for ti-qspi

[PATCH v3 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.

v2: https://lkml.org/lkml/2016/4/25/187
v1: https://lkml.org/lkml/2016/4/4/855


Vignesh R (2):
  spi: Add DMA support for spi_flash_read()
  spi: spi-ti-qspi: Add DMA support for QSPI mmap read

 drivers/spi/spi-ti-qspi.c | 189 ++++++++++++++++++++++++++++++++++++++++++----
 drivers/spi/spi.c         |  14 ++++
 include/linux/spi/spi.h   |   4 +
 3 files changed, 194 insertions(+), 13 deletions(-)

-- 
2.8.3

[PATCH v3 1/2] spi: Add DMA support for spi_flash_read()

[Vignesh R]

Few SPI devices provide accelerated read interfaces to read from
SPI-NOR flash devices. These hardwares also support DMA to transfer data
from flash to memory either via mem-to-mem DMA or dedicated slave DMA
channels. Hence, add support for DMA in order to improve throughput and
reduce CPU load.
Use spi_map_buf() to get sg table for the buffer and pass it to SPI
driver.

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

v3: No changes.
v2: use cur_msg_mapped flag to indicate success/failure of spi_map_buf()


 drivers/spi/spi.c       | 14 ++++++++++++++
 include/linux/spi/spi.h |  4 ++++
 2 files changed, 18 insertions(+)

diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
index 77e6e45951f4..da2cbe205654 100644
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -2725,6 +2725,7 @@ int spi_flash_read(struct spi_device *spi,
 
 {
 	struct spi_master *master = spi->master;
+	struct device *rx_dev = NULL;
 	int ret;
 
 	if ((msg->opcode_nbits == SPI_NBITS_DUAL ||
@@ -2750,9 +2751,22 @@ int spi_flash_read(struct spi_device *spi,
 			return ret;
 		}
 	}
+
 	mutex_lock(&master->bus_lock_mutex);
+	if (master->dma_rx) {
+		rx_dev = master->dma_rx->device->dev;
+		ret = spi_map_buf(master, rx_dev, &msg->rx_sg,
+				  msg->buf, msg->len,
+				  DMA_FROM_DEVICE);
+		if (!ret)
+			msg->cur_msg_mapped = true;
+	}
 	ret = master->spi_flash_read(spi, msg);
+	if (msg->cur_msg_mapped)
+		spi_unmap_buf(master, rx_dev, &msg->rx_sg,
+			      DMA_FROM_DEVICE);
 	mutex_unlock(&master->bus_lock_mutex);
+
 	if (master->auto_runtime_pm)
 		pm_runtime_put(master->dev.parent);
 
diff --git a/include/linux/spi/spi.h b/include/linux/spi/spi.h
index 1f03483f61e5..7b53af4ba5f8 100644
--- a/include/linux/spi/spi.h
+++ b/include/linux/spi/spi.h
@@ -1143,6 +1143,8 @@ static inline ssize_t spi_w8r16be(struct spi_device *spi, u8 cmd)
  * @opcode_nbits: number of lines to send opcode
  * @addr_nbits: number of lines to send address
  * @data_nbits: number of lines for data
+ * @rx_sg: Scatterlist for receive data read from flash
+ * @cur_msg_mapped: message has been mapped for DMA
  */
 struct spi_flash_read_message {
 	void *buf;
@@ -1155,6 +1157,8 @@ struct spi_flash_read_message {
 	u8 opcode_nbits;
 	u8 addr_nbits;
 	u8 data_nbits;
+	struct sg_table rx_sg;
+	bool cur_msg_mapped;
 };
 
 /* SPI core interface for flash read support */
-- 
2.8.3

Re: [PATCH v3 1/2] spi: Add DMA support for spi_flash_read()

[kbuild test robot]

[-- Attachment #1: Type: text/plain, Size: 1982 bytes --]

Hi,

[auto build test ERROR on spi/for-next]
[also build test ERROR on v4.7-rc2 next-20160607]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url:    https://github.com/0day-ci/linux/commits/Vignesh-R/spi-Add-DMA-support-for-ti-qspi/20160607-162134
base:   https://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi for-next
config: m32r-allyesconfig (attached as .config)
compiler: m32r-linux-gcc (GCC) 4.9.0
reproduce:
        wget https://git.kernel.org/cgit/linux/kernel/git/wfg/lkp-tests.git/plain/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # save the attached .config to linux build tree
        make.cross ARCH=m32r 

All errors (new ones prefixed by >>):

   drivers/spi/spi.c: In function 'spi_flash_read':
>> drivers/spi/spi.c:2758:3: error: implicit declaration of function 'spi_map_buf' [-Werror=implicit-function-declaration]
      ret = spi_map_buf(master, rx_dev, &msg->rx_sg,
      ^
>> drivers/spi/spi.c:2766:3: error: implicit declaration of function 'spi_unmap_buf' [-Werror=implicit-function-declaration]
      spi_unmap_buf(master, rx_dev, &msg->rx_sg,
      ^
   cc1: some warnings being treated as errors

vim +/spi_map_buf +2758 drivers/spi/spi.c

  2752			}
  2753		}
  2754	
  2755		mutex_lock(&master->bus_lock_mutex);
  2756		if (master->dma_rx) {
  2757			rx_dev = master->dma_rx->device->dev;
> 2758			ret = spi_map_buf(master, rx_dev, &msg->rx_sg,
  2759					  msg->buf, msg->len,
  2760					  DMA_FROM_DEVICE);
  2761			if (!ret)
  2762				msg->cur_msg_mapped = true;
  2763		}
  2764		ret = master->spi_flash_read(spi, msg);
  2765		if (msg->cur_msg_mapped)
> 2766			spi_unmap_buf(master, rx_dev, &msg->rx_sg,
  2767				      DMA_FROM_DEVICE);
  2768		mutex_unlock(&master->bus_lock_mutex);
  2769	

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

[-- Attachment #2: .config.gz --]
[-- Type: application/octet-stream, Size: 35831 bytes --]

Re: [PATCH v3 1/2] spi: Add DMA support for spi_flash_read()

[Vignesh R]

On Wednesday 08 June 2016 03:59 AM, kbuild test robot wrote:
> Hi,
> 
> [auto build test ERROR on spi/for-next]
> [also build test ERROR on v4.7-rc2 next-20160607]
> [if your patch is applied to the wrong git tree, please drop us a note to help improve the system]
> 
> url:    https://github.com/0day-ci/linux/commits/Vignesh-R/spi-Add-DMA-support-for-ti-qspi/20160607-162134
> base:   https://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi for-next
> config: m32r-allyesconfig (attached as .config)
> compiler: m32r-linux-gcc (GCC) 4.9.0
> reproduce:
>         wget https://git.kernel.org/cgit/linux/kernel/git/wfg/lkp-tests.git/plain/sbin/make.cross -O ~/bin/make.cross
>         chmod +x ~/bin/make.cross
>         # save the attached .config to linux build tree
>         make.cross ARCH=m32r 
> 
> All errors (new ones prefixed by >>):
> 
>    drivers/spi/spi.c: In function 'spi_flash_read':
>>> drivers/spi/spi.c:2758:3: error: implicit declaration of function 'spi_map_buf' [-Werror=implicit-function-declaration]
>       ret = spi_map_buf(master, rx_dev, &msg->rx_sg,
>       ^
>>> drivers/spi/spi.c:2766:3: error: implicit declaration of function 'spi_unmap_buf' [-Werror=implicit-function-declaration]
>       spi_unmap_buf(master, rx_dev, &msg->rx_sg,
>       ^
>    cc1: some warnings being treated as errors
> 


Oops, posted v4 fixing these errors.


-- 
Regards
Vignesh

[PATCH v3 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 ~13 MB/s.

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

v3: Cleanup code based on review comments for v2.
v2: Handle kmap'd buffers of JFFS2 FS.

 drivers/spi/spi-ti-qspi.c | 189 ++++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 176 insertions(+), 13 deletions(-)

diff --git a/drivers/spi/spi-ti-qspi.c b/drivers/spi/spi-ti-qspi.c
index 29ea8d2f9824..1f6c59c29157 100644
--- a/drivers/spi/spi-ti-qspi.c
+++ b/drivers/spi/spi-ti-qspi.c
@@ -33,6 +33,7 @@
 #include <linux/pinctrl/consumer.h>
 #include <linux/mfd/syscon.h>
 #include <linux/regmap.h>
+#include <linux/highmem.h>
 
 #include <linux/spi/spi.h>
 
@@ -41,6 +42,8 @@ struct ti_qspi_regs {
 };
 
 struct ti_qspi {
+	struct completion	transfer_complete;
+
 	/* list synchronization */
 	struct mutex            list_lock;
 
@@ -54,6 +57,9 @@ 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 +385,72 @@ 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 +498,40 @@ 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,
+#ifdef CONFIG_HIGHMEM
+static int ti_qspi_map_buf(struct ti_qspi *qspi, void *buf,
+			   unsigned int len, struct sg_table *sgt)
+{
+	unsigned int max_seg_size =
+		dma_get_max_seg_size(qspi->rx_chan->device->dev);
+	unsigned int desc_len = min_t(int, max_seg_size, PAGE_SIZE);
+	int sgs = DIV_ROUND_UP(len + offset_in_page(buf), desc_len);
+	struct page *vm_page;
+	size_t min;
+	int i, ret;
+
+	ret = sg_alloc_table(sgt, sgs, GFP_KERNEL);
+	if (ret)
+		return ret;
+
+	for (i = 0; i < sgs; i++) {
+		min = min_t(size_t, len, desc_len -
+			    offset_in_page(buf));
+		vm_page = kmap_to_page(buf);
+		if (!vm_page) {
+			sg_free_table(sgt);
+			return -ENOMEM;
+		}
+		sg_set_page(&sgt->sgl[i], vm_page, min,
+			    offset_in_page(buf));
+		buf += min;
+		len -= min;
+	}
+	return 0;
+}
+#endif
+
+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 +542,46 @@ 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) {
+		struct device *dev = qspi->rx_chan->device->dev;
+		void *buf = msg->buf;
+		struct sg_table sgt;
+
+		if (msg->cur_msg_mapped) {
+			ret = ti_qspi_dma_xfer_sg(qspi, msg->rx_sg, msg->from);
+			if (ret)
+				goto err_unlock;
+#ifdef CONFIG_HIGHMEM
+		} else if ((unsigned long)buf >= PKMAP_BASE &&
+				(unsigned long)buf < (PKMAP_BASE +
+					(LAST_PKMAP * PAGE_SIZE))) {
+			/* Generate sg_table for kmap buffers */
+			ret = ti_qspi_map_buf(qspi, buf, msg->len, &sgt);
+			if (ret)
+				goto err_unlock;
+			ret = dma_map_sg(dev, sgt.sgl, sgt.nents,
+					 DMA_FROM_DEVICE);
+			if (!ret) {
+				ret = -ENOMEM;
+				goto err_unlock;
+			}
+			ret = ti_qspi_dma_xfer_sg(qspi, sgt, msg->from);
+			dma_unmap_sg(dev, sgt.sgl, sgt.orig_nents,
+				     DMA_FROM_DEVICE);
+			sg_free_table(&sgt);
+#endif
+		} 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 +678,7 @@ 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 +693,7 @@ 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 +736,6 @@ 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 =
@@ -637,6 +770,33 @@ static int ti_qspi_probe(struct platform_device *pdev)
 	if (ret)
 		goto free_master;
 
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_MEMCPY, mask);
+
+	qspi->rx_chan = dma_request_channel(mask, NULL, NULL);
+	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;
 	return 0;
 
 free_master:
@@ -656,6 +816,9 @@ 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;
 }
 
-- 
2.8.3

Re: [PATCH v3 2/2] spi: spi-ti-qspi: Add DMA support for QSPI mmap read

[Peter Ujfalusi]

On 06/07/16 11:18, Vignesh R wrote:
> 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 ~13 MB/s.
> 
> Signed-off-by: Vignesh R <vigneshr@ti.com>
> ---
> 
> v3: Cleanup code based on review comments for v2.
> v2: Handle kmap'd buffers of JFFS2 FS.
> 
>  drivers/spi/spi-ti-qspi.c | 189 ++++++++++++++++++++++++++++++++++++++++++----
>  1 file changed, 176 insertions(+), 13 deletions(-)
> 
> diff --git a/drivers/spi/spi-ti-qspi.c b/drivers/spi/spi-ti-qspi.c
> index 29ea8d2f9824..1f6c59c29157 100644
> --- a/drivers/spi/spi-ti-qspi.c
> +++ b/drivers/spi/spi-ti-qspi.c
> @@ -33,6 +33,7 @@
>  #include <linux/pinctrl/consumer.h>
>  #include <linux/mfd/syscon.h>
>  #include <linux/regmap.h>
> +#include <linux/highmem.h>
>  
>  #include <linux/spi/spi.h>
>  
> @@ -41,6 +42,8 @@ struct ti_qspi_regs {
>  };
>  
>  struct ti_qspi {
> +	struct completion	transfer_complete;
> +
>  	/* list synchronization */
>  	struct mutex            list_lock;
>  
> @@ -54,6 +57,9 @@ 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 +385,72 @@ 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 +498,40 @@ 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,
> +#ifdef CONFIG_HIGHMEM
> +static int ti_qspi_map_buf(struct ti_qspi *qspi, void *buf,
> +			   unsigned int len, struct sg_table *sgt)
> +{
> +	unsigned int max_seg_size =
> +		dma_get_max_seg_size(qspi->rx_chan->device->dev);
> +	unsigned int desc_len = min_t(int, max_seg_size, PAGE_SIZE);
> +	int sgs = DIV_ROUND_UP(len + offset_in_page(buf), desc_len);
> +	struct page *vm_page;
> +	size_t min;
> +	int i, ret;
> +
> +	ret = sg_alloc_table(sgt, sgs, GFP_KERNEL);
> +	if (ret)
> +		return ret;
> +
> +	for (i = 0; i < sgs; i++) {
> +		min = min_t(size_t, len, desc_len -
> +			    offset_in_page(buf));
> +		vm_page = kmap_to_page(buf);
> +		if (!vm_page) {
> +			sg_free_table(sgt);
> +			return -ENOMEM;
> +		}
> +		sg_set_page(&sgt->sgl[i], vm_page, min,
> +			    offset_in_page(buf));
> +		buf += min;
> +		len -= min;
> +	}
> +	return 0;
> +}
> +#endif
> +
> +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 +542,46 @@ 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) {
> +		struct device *dev = qspi->rx_chan->device->dev;
> +		void *buf = msg->buf;
> +		struct sg_table sgt;
> +
> +		if (msg->cur_msg_mapped) {
> +			ret = ti_qspi_dma_xfer_sg(qspi, msg->rx_sg, msg->from);
> +			if (ret)
> +				goto err_unlock;
> +#ifdef CONFIG_HIGHMEM
> +		} else if ((unsigned long)buf >= PKMAP_BASE &&
> +				(unsigned long)buf < (PKMAP_BASE +
> +					(LAST_PKMAP * PAGE_SIZE))) {
> +			/* Generate sg_table for kmap buffers */
> +			ret = ti_qspi_map_buf(qspi, buf, msg->len, &sgt);
> +			if (ret)
> +				goto err_unlock;
> +			ret = dma_map_sg(dev, sgt.sgl, sgt.nents,
> +					 DMA_FROM_DEVICE);
> +			if (!ret) {
> +				ret = -ENOMEM;
> +				goto err_unlock;
> +			}
> +			ret = ti_qspi_dma_xfer_sg(qspi, sgt, msg->from);
> +			dma_unmap_sg(dev, sgt.sgl, sgt.orig_nents,
> +				     DMA_FROM_DEVICE);
> +			sg_free_table(&sgt);
> +#endif
> +		} 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 +678,7 @@ 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 +693,7 @@ 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 +736,6 @@ 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 =
> @@ -637,6 +770,33 @@ static int ti_qspi_probe(struct platform_device *pdev)
>  	if (ret)
>  		goto free_master;
>  
> +	dma_cap_zero(mask);
> +	dma_cap_set(DMA_MEMCPY, mask);
> +
> +	qspi->rx_chan = dma_request_channel(mask, NULL, NULL);

dma_request_channel is deprecated, please use the:
dma_request_chan_by_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;
>  	return 0;
>  
>  free_master:
> @@ -656,6 +816,9 @@ 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;
>  }
>  
> 


-- 
Péter

Re: [PATCH v3 2/2] spi: spi-ti-qspi: Add DMA support for QSPI mmap read

[Vignesh R]

On Tuesday 07 June 2016 02:47 PM, Peter Ujfalusi wrote:
[...]
>> @@ -637,6 +770,33 @@ static int ti_qspi_probe(struct platform_device *pdev)
>>  	if (ret)
>>  		goto free_master;
>>  
>> +	dma_cap_zero(mask);
>> +	dma_cap_set(DMA_MEMCPY, mask);
>> +
>> +	qspi->rx_chan = dma_request_channel(mask, NULL, NULL);
> 
> dma_request_channel is deprecated, please use the:
> dma_request_chan_by_mask()
> 

Updated to use dma_request_chan_by_mask() in v4, Thanks.


-- 
Regards
Vignesh

Re: [PATCH v3 2/2] spi: spi-ti-qspi: Add DMA support for QSPI mmap read

[kbuild test robot]

[-- Attachment #1: Type: text/plain, Size: 1866 bytes --]

Hi,

[auto build test WARNING on spi/for-next]
[also build test WARNING on v4.7-rc2 next-20160606]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url:    https://github.com/0day-ci/linux/commits/Vignesh-R/spi-Add-DMA-support-for-ti-qspi/20160607-162134
base:   https://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi for-next
config: mn10300-allmodconfig (attached as .config)
compiler: am33_2.0-linux-gcc (GCC) 4.9.0
reproduce:
        wget https://git.kernel.org/cgit/linux/kernel/git/wfg/lkp-tests.git/plain/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # save the attached .config to linux build tree
        make.cross ARCH=mn10300 

All warnings (new ones prefixed by >>):

   drivers/spi/spi-ti-qspi.c: In function 'ti_qspi_spi_flash_read':
>> drivers/spi/spi-ti-qspi.c:549:19: warning: unused variable 'sgt' [-Wunused-variable]
      struct sg_table sgt;
                      ^
>> drivers/spi/spi-ti-qspi.c:548:9: warning: unused variable 'buf' [-Wunused-variable]
      void *buf = msg->buf;
            ^
>> drivers/spi/spi-ti-qspi.c:547:18: warning: unused variable 'dev' [-Wunused-variable]
      struct device *dev = qspi->rx_chan->device->dev;
                     ^

vim +/sgt +549 drivers/spi/spi-ti-qspi.c

   541	
   542		if (!qspi->mmap_enabled)
   543			ti_qspi_enable_memory_map(spi);
   544		ti_qspi_setup_mmap_read(spi, msg);
   545	
   546		if (qspi->rx_chan) {
 > 547			struct device *dev = qspi->rx_chan->device->dev;
 > 548			void *buf = msg->buf;
 > 549			struct sg_table sgt;
   550	
   551			if (msg->cur_msg_mapped) {
   552				ret = ti_qspi_dma_xfer_sg(qspi, msg->rx_sg, msg->from);

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

[-- Attachment #2: .config.gz --]
[-- Type: application/octet-stream, Size: 39042 bytes --]