Skip to contentSkip to conversation

patches against slave dma subsystem

PATCH Applied Andy Shevchenko 3 participants 15 messages Jan 10, 2013 – Jan 12, 2013

Subsystems (3)
dmaengine Arm/nomadik/Ux500 architectures Microchip AT91 DMA drivers
lore
Share & tools
lore

Andy Shevchenko

author
#0 lore raw

v2 cover patches against slave dma subsystem

Recend to Vinod for review.

Since v1:
- is_slave_xfer renamed to is_slave_direction
- changelog fixed for "dma: dw_dmac: clear suspend bit during termination"
- typo fixed in the changelog of "dw_dmac: backlink to dw_dma in dw_dma_chan is superfluous"

Andy Shevchenko (10):
dmaengine: introduce is_slave_direction function
dma: at_hdmac: check direction properly for cyclic transfers
dma: dw_dmac: check direction properly in dw_dma_cyclic_prep
dma: ep93xx_dma: reuse is_slave_direction helper
dma: ipu_idmac: reuse is_slave_direction helper
dma: ste_dma40: reuse is_slave_direction helper
dw_dmac: call .probe after we have a device in place
dw_dmac: store direction in the custom channel structure
dw_dmac: make usage of dw_dma_slave optional
dw_dmac: backlink to dw_dma in dw_dma_chan is superfluous

Heikki Krogerus (1):
dma: dw_dmac: clear suspend bit during termination

 drivers/dma/at_hdmac.c      |   10 +++----
 drivers/dma/dw_dmac.c       |   62 ++++++++++++++++++++++++++++++-------------
 drivers/dma/dw_dmac_regs.h  |   17 ++++++------
 drivers/dma/ep93xx_dma.c    |    3 +--
 drivers/dma/ipu/ipu_idmac.c |    2 +-
 drivers/dma/ste_dma40.c     |    2 +-
 include/linux/dmaengine.h   |    5 ++++
 7 files changed, 64 insertions(+), 37 deletions(-)
Signature · 1 line
1.7.10.4
author
#1 lore raw

PATCH v2 1/11 dmaengine: introduce is_slave_direction function

This function helps to distinguish the slave type of transfer by checking the direction parameter.

Signed-off-by Andy Shevchenko <[email protected]>
Reviewed-by Viresh Kumar <[email protected]>
Reviewed-by Mika Westerberg <[email protected]>
Reviewed-by Linus Walleij <[email protected]>
Cc Nicolas Ferre <[email protected]>
Cc Guennadi Liakhovetski <[email protected]>
---
 include/linux/dmaengine.h |    5 +++++
 1 file changed, 5 insertions(+)
include/linux/dmaengine.h +5 -0
index 4ca9cf7..bfcdecb 100644
@@ -621,6 +621,11 @@ static inline int dmaengine_slave_config(struct dma_chan *chan,
 			(unsigned long)config);
 }
 
+static inline bool is_slave_direction(enum dma_transfer_direction direction)
+{
+	return (direction == DMA_MEM_TO_DEV) || (direction == DMA_DEV_TO_MEM);
+}
+
 static inline struct dma_async_tx_descriptor *dmaengine_prep_slave_single(
 	struct dma_chan *chan, dma_addr_t buf, size_t len,
 	enum dma_transfer_direction dir, unsigned long flags)
Signature · 1 line
1.7.10.4
author
#2 lore raw

PATCH v2 2/11 dma: at_hdmac: check direction properly for cyclic transfers

In reply to Andy Shevchenko

dma_transfer_direction is a normal enum. It means we can't usually use the values as bit fields. Let's adjust this check and move it above the usage of the direction parameter, due to the nature of the following usage of it.

Signed-off-by Andy Shevchenko <[email protected]>
Reviewed-by Viresh Kumar <[email protected]>
Acked-by Nicolas Ferre <[email protected]>
---
 drivers/dma/at_hdmac.c |   10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
drivers/dma/at_hdmac.c +5 -5
index 13a02f4..6e13f26 100644
@@ -778,7 +778,7 @@ err:
  */
 static int
 atc_dma_cyclic_check_values(unsigned int reg_width, dma_addr_t buf_addr,
-		size_t period_len, enum dma_transfer_direction direction)
+		size_t period_len)
 {
 	if (period_len > (ATC_BTSIZE_MAX << reg_width))
 		goto err_out;

@@ -786,8 +786,6 @@ atc_dma_cyclic_check_values(unsigned int reg_width, dma_addr_t buf_addr,
 		goto err_out;
 	if (unlikely(buf_addr & ((1 << reg_width) - 1)))
 		goto err_out;
-	if (unlikely(!(direction & (DMA_DEV_TO_MEM | DMA_MEM_TO_DEV))))
-		goto err_out;
 
 	return 0;
 

@@ -886,14 +884,16 @@ atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
 		return NULL;
 	}
 
+	if (unlikely(!is_slave_direction(direction)))
+		goto err_out;
+
 	if (sconfig->direction == DMA_MEM_TO_DEV)
 		reg_width = convert_buswidth(sconfig->dst_addr_width);
 	else
 		reg_width = convert_buswidth(sconfig->src_addr_width);
 
 	/* Check for too big/unaligned periods and unaligned DMA buffer */
-	if (atc_dma_cyclic_check_values(reg_width, buf_addr,
-					period_len, direction))
+	if (atc_dma_cyclic_check_values(reg_width, buf_addr, period_len))
 		goto err_out;
 
 	/* build cyclic linked list */
Signature · 1 line
1.7.10.4
author
#3 lore raw

PATCH v2 3/11 dma: dw_dmac: check direction properly in dw_dma_cyclic_prep

In reply to Andy Shevchenko

dma_transfer_direction is a normal enum. It means we can't usually use the values as bit fields. Let's adjust this check and move it above the usage of the direction parameter, due to the nature of the following usage of it.

Signed-off-by Andy Shevchenko <[email protected]>
Acked-by Viresh Kumar <[email protected]>
---
 drivers/dma/dw_dmac.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/dma/dw_dmac.c +3 -2
index 6c9e20a..ca996bc 100644
@@ -1355,6 +1355,9 @@ struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan,
 
 	retval = ERR_PTR(-EINVAL);
 
+	if (unlikely(!is_slave_direction(direction)))
+		goto out_err;
+
 	if (direction == DMA_MEM_TO_DEV)
 		reg_width = __ffs(sconfig->dst_addr_width);
 	else

@@ -1369,8 +1372,6 @@ struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan,
 		goto out_err;
 	if (unlikely(buf_addr & ((1 << reg_width) - 1)))
 		goto out_err;
-	if (unlikely(!(direction & (DMA_MEM_TO_DEV | DMA_DEV_TO_MEM))))
-		goto out_err;
 
 	retval = ERR_PTR(-ENOMEM);
Signature · 1 line
1.7.10.4
author
#4 lore raw

PATCH v2 4/11 dma: ep93xx_dma: reuse is_slave_direction helper

In reply to Andy Shevchenko

The is_slave_direction helps to check if the transfer type is slave.

Signed-off-by Andy Shevchenko <[email protected]>
Reviewed-by Viresh Kumar <[email protected]>
Acked-by Mika Westerberg <[email protected]>
---
 drivers/dma/ep93xx_dma.c |    3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/dma/ep93xx_dma.c +1 -2
index bcfde40..f2bf8c0 100644
@@ -903,8 +903,7 @@ static int ep93xx_dma_alloc_chan_resources(struct dma_chan *chan)
 			switch (data->port) {
 			case EP93XX_DMA_SSP:
 			case EP93XX_DMA_IDE:
-				if (data->direction != DMA_MEM_TO_DEV &&
-				    data->direction != DMA_DEV_TO_MEM)
+				if (!is_slave_direction(data->direction))
 					return -EINVAL;
 				break;
 			default:
Signature · 1 line
1.7.10.4
author
#5 lore raw

PATCH v2 5/11 dma: ipu_idmac: reuse is_slave_direction helper

In reply to Andy Shevchenko

The is_slave_direction helps to check if the transfer type is slave.

Signed-off-by Andy Shevchenko <[email protected]>
Reviewed-by Viresh Kumar <[email protected]>
Cc Guennadi Liakhovetski <[email protected]>
---
 drivers/dma/ipu/ipu_idmac.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/dma/ipu/ipu_idmac.c +1 -1
index 6585537..8c61d17 100644
@@ -1347,7 +1347,7 @@ static struct dma_async_tx_descriptor *idmac_prep_slave_sg(struct dma_chan *chan
 	    chan->chan_id != IDMAC_IC_7)
 		return NULL;
 
-	if (direction != DMA_DEV_TO_MEM && direction != DMA_MEM_TO_DEV) {
+	if (!is_slave_direction(direction)) {
 		dev_err(chan->device->dev, "Invalid DMA direction %d!\n", direction);
 		return NULL;
 	}
Signature · 1 line
1.7.10.4
author
#6 lore raw

PATCH v2 6/11 dma: ste_dma40: reuse is_slave_direction helper

In reply to Andy Shevchenko

The is_slave_direction helps to check if the transfer type is slave.

Signed-off-by Andy Shevchenko <[email protected]>
Reviewed-by Viresh Kumar <[email protected]>
Reviewed-by Linus Walleij <[email protected]>
---
 drivers/dma/ste_dma40.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/dma/ste_dma40.c +1 -1
index 23c5573..e5e60bb 100644
@@ -2337,7 +2337,7 @@ static struct dma_async_tx_descriptor *d40_prep_slave_sg(struct dma_chan *chan,
 							 unsigned long dma_flags,
 							 void *context)
 {
-	if (direction != DMA_DEV_TO_MEM && direction != DMA_MEM_TO_DEV)
+	if (!is_slave_direction(direction))
 		return NULL;
 
 	return d40_prep_sg(chan, sgl, sgl, sg_len, direction, dma_flags);
Signature · 1 line
1.7.10.4
author
#7 lore raw

PATCH v2 7/11 dw_dmac: call .probe after we have a device in place

In reply to Andy Shevchenko

If we don't yet have the platform device for the driver when it is being loaded we fail to probe the driver. So instead of calling probe() directly we call platform_driver_register(). It will call the probe() immediately if we have the device but also makes the driver to work on platforms where the platform device is created later.

Signed-off-by Andy Shevchenko <[email protected]>
Acked-by Viresh Kumar <[email protected]>
---
 drivers/dma/dw_dmac.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/dma/dw_dmac.c +2 -1
index ca996bc..e5540278 100644
@@ -1858,6 +1858,7 @@ MODULE_DEVICE_TABLE(of, dw_dma_id_table);
 #endif
 
 static struct platform_driver dw_driver = {
+	.probe		= dw_probe,
 	.remove		= dw_remove,
 	.shutdown	= dw_shutdown,
 	.driver = {

@@ -1869,7 +1870,7 @@ static struct platform_driver dw_driver = {
 
 static int __init dw_init(void)
 {
-	return platform_driver_probe(&dw_driver, dw_probe);
+	return platform_driver_register(&dw_driver);
 }
 subsys_initcall(dw_init);
Signature · 1 line
1.7.10.4
author
#8 lore raw

PATCH v2 8/11 dw_dmac: store direction in the custom channel structure

In reply to Andy Shevchenko

Currently the direction value comes from the generic slave configuration structure and explicitly as a preparation function parameter. The first one is kinda obsoleted. Thus, we have to store the value passed to the preparation function somewhere in our structures to be able to use it later. The best candidate to provide the storage is a custom channel structure. Until now we still keep and check the direction field of the slave config structure as well.

Signed-off-by Andy Shevchenko <[email protected]>
Acked-by Viresh Kumar <[email protected]>
---
 drivers/dma/dw_dmac.c      |   12 ++++++++++--
 drivers/dma/dw_dmac_regs.h |   14 ++++++++------
 2 files changed, 18 insertions(+), 8 deletions(-)
drivers/dma/dw_dmac.c +10 -2
index e5540278..a2c5a60 100644
@@ -178,9 +178,9 @@ static void dwc_initialize(struct dw_dma_chan *dwc)
 		cfghi = dws->cfg_hi;
 		cfglo |= dws->cfg_lo & ~DWC_CFGL_CH_PRIOR_MASK;
 	} else {
-		if (dwc->dma_sconfig.direction == DMA_MEM_TO_DEV)
+		if (dwc->direction == DMA_MEM_TO_DEV)
 			cfghi = DWC_CFGH_DST_PER(dwc->dma_sconfig.slave_id);
-		else if (dwc->dma_sconfig.direction == DMA_DEV_TO_MEM)
+		else if (dwc->direction == DMA_DEV_TO_MEM)
 			cfghi = DWC_CFGH_SRC_PER(dwc->dma_sconfig.slave_id);
 	}
 

@@ -721,6 +721,8 @@ dwc_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 		return NULL;
 	}
 
+	dwc->direction = DMA_MEM_TO_MEM;
+
 	data_width = min_t(unsigned int, dwc->dw->data_width[dwc_get_sms(dws)],
 					 dwc->dw->data_width[dwc_get_dms(dws)]);
 

@@ -805,6 +807,8 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 	if (unlikely(!dws || !sg_len))
 		return NULL;
 
+	dwc->direction = direction;
+
 	prev = first = NULL;
 
 	switch (direction) {

@@ -981,6 +985,7 @@ set_runtime_config(struct dma_chan *chan, struct dma_slave_config *sconfig)
 		return -EINVAL;
 
 	memcpy(&dwc->dma_sconfig, sconfig, sizeof(*sconfig));
+	dwc->direction = sconfig->direction;
 
 	convert_burst(&dwc->dma_sconfig.src_maxburst);
 	convert_burst(&dwc->dma_sconfig.dst_maxburst);

@@ -1358,6 +1363,8 @@ struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan,
 	if (unlikely(!is_slave_direction(direction)))
 		goto out_err;
 
+	dwc->direction = direction;
+
 	if (direction == DMA_MEM_TO_DEV)
 		reg_width = __ffs(sconfig->dst_addr_width);
 	else

@@ -1732,6 +1739,7 @@ static int dw_probe(struct platform_device *pdev)
 		channel_clear_bit(dw, CH_EN, dwc->mask);
 
 		dwc->dw = dw;
+		dwc->direction = DMA_TRANS_NONE;
 
 		/* hardware configuration */
 		if (autocfg) {
drivers/dma/dw_dmac_regs.h +8 -6
index 8881e9b..f9532c2 100644
@@ -9,6 +9,7 @@
  * published by the Free Software Foundation.
  */
 
+#include <linux/dmaengine.h>
 #include <linux/dw_dmac.h>
 
 #define DW_DMA_MAX_NR_CHANNELS	8

@@ -184,12 +185,13 @@ enum dw_dmac_flags {
 };
 
 struct dw_dma_chan {
-	struct dma_chan		chan;
-	void __iomem		*ch_regs;
-	u8			mask;
-	u8			priority;
-	bool			paused;
-	bool			initialized;
+	struct dma_chan			chan;
+	void __iomem			*ch_regs;
+	u8				mask;
+	u8				priority;
+	enum dma_transfer_direction	direction;
+	bool				paused;
+	bool				initialized;
 
 	/* software emulation of the LLP transfers */
 	struct list_head	*tx_list;
Signature · 1 line
1.7.10.4
author
#9 lore raw

PATCH v2 9/11 dw_dmac: make usage of dw_dma_slave optional

In reply to Andy Shevchenko

The driver requires a custom slave configuration to be present to be able to make the slave transfers. Nevertheless, in some cases we need only the request line as an additional information to the generic slave configuration. The request line is provided by slave_id parameter of the dma_slave_config structure. That's why the custom slave configuration could be optional for such cases.

Signed-off-by Andy Shevchenko <[email protected]>
Acked-by Viresh Kumar <[email protected]>
---
 drivers/dma/dw_dmac.c |   13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)
drivers/dma/dw_dmac.c +7 -6
index a2c5a60..c87d8ab 100644
@@ -50,11 +50,12 @@ static inline unsigned int dwc_get_sms(struct dw_dma_slave *slave)
 		struct dw_dma_slave *__slave = (_chan->private);	\
 		struct dw_dma_chan *_dwc = to_dw_dma_chan(_chan);	\
 		struct dma_slave_config	*_sconfig = &_dwc->dma_sconfig;	\
+		bool _is_slave = is_slave_xfer(_dwc->direction);	\
 		int _dms = dwc_get_dms(__slave);		\
 		int _sms = dwc_get_sms(__slave);		\
-		u8 _smsize = __slave ? _sconfig->src_maxburst :	\
+		u8 _smsize = _is_slave ? _sconfig->src_maxburst :	\
 			DW_DMA_MSIZE_16;			\
-		u8 _dmsize = __slave ? _sconfig->dst_maxburst :	\
+		u8 _dmsize = _is_slave ? _sconfig->dst_maxburst :	\
 			DW_DMA_MSIZE_16;			\
 								\
 		(DWC_CTLL_DST_MSIZE(_dmsize)			\

@@ -328,7 +329,7 @@ dwc_descriptor_complete(struct dw_dma_chan *dwc, struct dw_desc *desc,
 	list_splice_init(&desc->tx_list, &dwc->free_list);
 	list_move(&desc->desc_node, &dwc->free_list);
 
-	if (!dwc->chan.private) {
+	if (!is_slave_xfer(dwc->direction)) {
 		struct device *parent = chan2parent(&dwc->chan);
 		if (!(txd->flags & DMA_COMPL_SKIP_DEST_UNMAP)) {
 			if (txd->flags & DMA_COMPL_DEST_UNMAP_SINGLE)

@@ -804,7 +805,7 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 
 	dev_vdbg(chan2dev(chan), "%s\n", __func__);
 
-	if (unlikely(!dws || !sg_len))
+	if (unlikely(!is_slave_xfer(direction) || !sg_len))
 		return NULL;
 
 	dwc->direction = direction;

@@ -980,8 +981,8 @@ set_runtime_config(struct dma_chan *chan, struct dma_slave_config *sconfig)
 {
 	struct dw_dma_chan *dwc = to_dw_dma_chan(chan);
 
-	/* Check if it is chan is configured for slave transfers */
-	if (!chan->private)
+	/* Check if chan will be configured for slave transfers */
+	if (!is_slave_xfer(sconfig->direction))
 		return -EINVAL;
 
 	memcpy(&dwc->dma_sconfig, sconfig, sizeof(*sconfig));
Signature · 1 line
1.7.10.4
#10 lore raw

Re: [PATCHv2 09/11] dw_dmac: make usage of dw_dma_slave optional

Please, don't push this one yet. I forgot to update the name of is_slave_xfer.

Quoted 69 lines · Andy Shevchenko

The driver requires a custom slave configuration to be present to be able to make the slave transfers. Nevertheless, in some cases we need only the request line as an additional information to the generic slave configuration. The request line is provided by slave_id parameter of the dma_slave_config structure. That's why the custom slave configuration could be optional for such cases.

Signed-off-by Andy Shevchenko <[email protected]>
Acked-by Viresh Kumar <[email protected]>
---
 drivers/dma/dw_dmac.c |   13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)
drivers/dma/dw_dmac.c +7 -7
index a2c5a60..c87d8ab 100644
@@ -50,11 +50,12 @@ static inline unsigned int dwc_get_sms(struct dw_dma_slave *slave)
                struct dw_dma_slave *__slave = (_chan->private);        \
                struct dw_dma_chan *_dwc = to_dw_dma_chan(_chan);       \
                struct dma_slave_config *_sconfig = &_dwc->dma_sconfig; \
+               bool _is_slave = is_slave_xfer(_dwc->direction);        \
                int _dms = dwc_get_dms(__slave);                \
                int _sms = dwc_get_sms(__slave);                \
-               u8 _smsize = __slave ? _sconfig->src_maxburst : \
+               u8 _smsize = _is_slave ? _sconfig->src_maxburst :       \
                        DW_DMA_MSIZE_16;                        \
-               u8 _dmsize = __slave ? _sconfig->dst_maxburst : \
+               u8 _dmsize = _is_slave ? _sconfig->dst_maxburst :       \
                        DW_DMA_MSIZE_16;                        \
                                                                \
                (DWC_CTLL_DST_MSIZE(_dmsize)                    \

@@ -328,7 +329,7 @@ dwc_descriptor_complete(struct dw_dma_chan *dwc, struct dw_desc *desc,
        list_splice_init(&desc->tx_list, &dwc->free_list);
        list_move(&desc->desc_node, &dwc->free_list);
 
-       if (!dwc->chan.private) {
+       if (!is_slave_xfer(dwc->direction)) {
                struct device *parent = chan2parent(&dwc->chan);
                if (!(txd->flags & DMA_COMPL_SKIP_DEST_UNMAP)) {
                        if (txd->flags & DMA_COMPL_DEST_UNMAP_SINGLE)

@@ -804,7 +805,7 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 
        dev_vdbg(chan2dev(chan), "%s\n", __func__);
 
-       if (unlikely(!dws || !sg_len))
+       if (unlikely(!is_slave_xfer(direction) || !sg_len))
                return NULL;
 
        dwc->direction = direction;

@@ -980,8 +981,8 @@ set_runtime_config(struct dma_chan *chan, struct dma_slave_config *sconfig)
 {
        struct dw_dma_chan *dwc = to_dw_dma_chan(chan);
 
-       /* Check if it is chan is configured for slave transfers */
-       if (!chan->private)
+       /* Check if chan will be configured for slave transfers */
+       if (!is_slave_xfer(sconfig->direction))
                return -EINVAL;
 
        memcpy(&dwc->dma_sconfig, sconfig, sizeof(*sconfig));
--

1.7.10.4

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Signature · 2 lines
With Best Regards,
Andy Shevchenko
author
#11 lore raw

[PATCHv2.1] dw_dmac: make usage of dw_dma_slave optional

The driver requires a custom slave configuration to be present to be able to make the slave transfers. Nevertheless, in some cases we need only the request line as an additional information to the generic slave configuration. The request line is provided by slave_id parameter of the dma_slave_config structure. That's why the custom slave configuration could be optional for such cases.

Signed-off-by Andy Shevchenko <[email protected]>
Acked-by Viresh Kumar <[email protected]>
---
 drivers/dma/dw_dmac.c |   13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)
drivers/dma/dw_dmac.c +7 -6
index a2c5a60..e5378b5 100644
@@ -50,11 +50,12 @@ static inline unsigned int dwc_get_sms(struct dw_dma_slave *slave)
 		struct dw_dma_slave *__slave = (_chan->private);	\
 		struct dw_dma_chan *_dwc = to_dw_dma_chan(_chan);	\
 		struct dma_slave_config	*_sconfig = &_dwc->dma_sconfig;	\
+		bool _is_slave = is_slave_direction(_dwc->direction);	\
 		int _dms = dwc_get_dms(__slave);		\
 		int _sms = dwc_get_sms(__slave);		\
-		u8 _smsize = __slave ? _sconfig->src_maxburst :	\
+		u8 _smsize = _is_slave ? _sconfig->src_maxburst :	\
 			DW_DMA_MSIZE_16;			\
-		u8 _dmsize = __slave ? _sconfig->dst_maxburst :	\
+		u8 _dmsize = _is_slave ? _sconfig->dst_maxburst :	\
 			DW_DMA_MSIZE_16;			\
 								\
 		(DWC_CTLL_DST_MSIZE(_dmsize)			\

@@ -328,7 +329,7 @@ dwc_descriptor_complete(struct dw_dma_chan *dwc, struct dw_desc *desc,
 	list_splice_init(&desc->tx_list, &dwc->free_list);
 	list_move(&desc->desc_node, &dwc->free_list);
 
-	if (!dwc->chan.private) {
+	if (!is_slave_direction(dwc->direction)) {
 		struct device *parent = chan2parent(&dwc->chan);
 		if (!(txd->flags & DMA_COMPL_SKIP_DEST_UNMAP)) {
 			if (txd->flags & DMA_COMPL_DEST_UNMAP_SINGLE)

@@ -804,7 +805,7 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 
 	dev_vdbg(chan2dev(chan), "%s\n", __func__);
 
-	if (unlikely(!dws || !sg_len))
+	if (unlikely(!is_slave_direction(direction) || !sg_len))
 		return NULL;
 
 	dwc->direction = direction;

@@ -980,8 +981,8 @@ set_runtime_config(struct dma_chan *chan, struct dma_slave_config *sconfig)
 {
 	struct dw_dma_chan *dwc = to_dw_dma_chan(chan);
 
-	/* Check if it is chan is configured for slave transfers */
-	if (!chan->private)
+	/* Check if chan will be configured for slave transfers */
+	if (!is_slave_direction(sconfig->direction))
 		return -EINVAL;
 
 	memcpy(&dwc->dma_sconfig, sconfig, sizeof(*sconfig));
Signature · 1 line
1.7.10.4
author
#12 lore raw

PATCH v2 10/11 dw_dmac: backlink to dw_dma in dw_dma_chan is superfluous

In reply to Andy Shevchenko

The same information could be extracted from the struct dma_chan. The patch introduces helper function dwc_get_data_width() as well.

Signed-off-by Andy Shevchenko <[email protected]>
Acked-by Viresh Kumar <[email protected]>
---
 drivers/dma/dw_dmac.c      |   27 ++++++++++++++++++++-------
 drivers/dma/dw_dmac_regs.h |    3 ---
 2 files changed, 20 insertions(+), 10 deletions(-)
drivers/dma/dw_dmac.c +20 -7
index c87d8ab..f9d37ec 100644
@@ -73,6 +73,22 @@ static inline unsigned int dwc_get_sms(struct dw_dma_slave *slave)
  */
 #define NR_DESCS_PER_CHANNEL	64
 
+#define SRC_MASTER	0
+#define DST_MASTER	1
+
+static inline unsigned int dwc_get_data_width(struct dma_chan *chan, int master)
+{
+	struct dw_dma *dw = to_dw_dma(chan->device);
+	struct dw_dma_slave *dws = chan->private;
+
+	if (master == SRC_MASTER)
+		return dw->data_width[dwc_get_sms(dws)];
+	else if (master == DST_MASTER)
+		return dw->data_width[dwc_get_dms(dws)];
+
+	return 0;
+}
+
 /*----------------------------------------------------------------------*/
 
 /*

@@ -701,7 +717,6 @@ dwc_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 		size_t len, unsigned long flags)
 {
 	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
-	struct dw_dma_slave	*dws = chan->private;
 	struct dw_desc		*desc;
 	struct dw_desc		*first;
 	struct dw_desc		*prev;

@@ -724,8 +739,8 @@ dwc_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
 
 	dwc->direction = DMA_MEM_TO_MEM;
 
-	data_width = min_t(unsigned int, dwc->dw->data_width[dwc_get_sms(dws)],
-					 dwc->dw->data_width[dwc_get_dms(dws)]);
+	data_width = min_t(unsigned int, dwc_get_data_width(chan, SRC_MASTER),
+			   dwc_get_data_width(chan, DST_MASTER));
 
 	src_width = dst_width = min_t(unsigned int, data_width,
 				      dwc_fast_fls(src | dest | len));

@@ -790,7 +805,6 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		unsigned long flags, void *context)
 {
 	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
-	struct dw_dma_slave	*dws = chan->private;
 	struct dma_slave_config	*sconfig = &dwc->dma_sconfig;
 	struct dw_desc		*prev;
 	struct dw_desc		*first;

@@ -824,7 +838,7 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 		ctllo |= sconfig->device_fc ? DWC_CTLL_FC(DW_DMA_FC_P_M2P) :
 			DWC_CTLL_FC(DW_DMA_FC_D_M2P);
 
-		data_width = dwc->dw->data_width[dwc_get_sms(dws)];
+		data_width = dwc_get_data_width(chan, SRC_MASTER);
 
 		for_each_sg(sgl, sg, sg_len, i) {
 			struct dw_desc	*desc;

@@ -887,7 +901,7 @@ slave_sg_todev_fill_desc:
 		ctllo |= sconfig->device_fc ? DWC_CTLL_FC(DW_DMA_FC_P_P2M) :
 			DWC_CTLL_FC(DW_DMA_FC_D_P2M);
 
-		data_width = dwc->dw->data_width[dwc_get_dms(dws)];
+		data_width = dwc_get_data_width(chan, DST_MASTER);
 
 		for_each_sg(sgl, sg, sg_len, i) {
 			struct dw_desc	*desc;

@@ -1739,7 +1753,6 @@ static int dw_probe(struct platform_device *pdev)
 
 		channel_clear_bit(dw, CH_EN, dwc->mask);
 
-		dwc->dw = dw;
 		dwc->direction = DMA_TRANS_NONE;
 
 		/* hardware configuration */
drivers/dma/dw_dmac_regs.h +0 -3
index f9532c2..577f2dd 100644
@@ -214,9 +214,6 @@ struct dw_dma_chan {
 
 	/* configuration passed via DMA_SLAVE_CONFIG */
 	struct dma_slave_config dma_sconfig;
-
-	/* backlink to dw_dma */
-	struct dw_dma		*dw;
 };
 
 static inline struct dw_dma_chan_regs __iomem *
Signature · 1 line
1.7.10.4
author
#13 lore raw

PATCH v2 11/11 dma: dw_dmac: clear suspend bit during termination

In reply to Andy Shevchenko

From: Heikki Krogerus <[email protected]>

The DMA transfer could not be established if previously it was paused and terminated. In that case the channel's suspend bit remains set that prevents to transfer anything until channel is resumed.

The patch adds the dwc_chan_resume() call instead of a plain flag assignment. That clears the DWC_CFGL_CH_SUSP bit as well during termination.

Signed-off-by Heikki Krogerus <[email protected]>
Signed-off-by Andy Shevchenko <[email protected]>
Acked-by Viresh Kumar <[email protected]>
---
 drivers/dma/dw_dmac.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/dma/dw_dmac.c +1 -1
index f9d37ec..684a9e1e 100644
@@ -1059,7 +1059,7 @@ static int dwc_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd,
 
 		dwc_chan_disable(dw, dwc);
 
-		dwc->paused = false;
+		dwc_chan_resume(dwc);
 
 		/* active_list entries will end up before queued entries */
 		list_splice_init(&dwc->queue, &list);
Signature · 1 line
1.7.10.4
Applied
#14 lore raw
In reply to Andy Shevchenko
Quoted 7 lines · Andy Shevchenko

Recend to Vinod for review.

Since v1:
- is_slave_xfer renamed to is_slave_direction
- changelog fixed for "dma: dw_dmac: clear suspend bit during termination"
- typo fixed in the changelog of "dw_dmac: backlink to dw_dma in dw_dma_chan is superfluous"

Applied all, Thanks

--
~Vinod

Quoted 28 lines

Andy Shevchenko (10):
dmaengine: introduce is_slave_direction function
dma: at_hdmac: check direction properly for cyclic transfers
dma: dw_dmac: check direction properly in dw_dma_cyclic_prep
dma: ep93xx_dma: reuse is_slave_direction helper
dma: ipu_idmac: reuse is_slave_direction helper
dma: ste_dma40: reuse is_slave_direction helper
dw_dmac: call .probe after we have a device in place
dw_dmac: store direction in the custom channel structure
dw_dmac: make usage of dw_dma_slave optional
dw_dmac: backlink to dw_dma in dw_dma_chan is superfluous

Heikki Krogerus (1):
dma: dw_dmac: clear suspend bit during termination

 drivers/dma/at_hdmac.c      |   10 +++----
 drivers/dma/dw_dmac.c       |   62 ++++++++++++++++++++++++++++++-------------
 drivers/dma/dw_dmac_regs.h  |   17 ++++++------
 drivers/dma/ep93xx_dma.c    |    3 +--
 drivers/dma/ipu/ipu_idmac.c |    2 +-
 drivers/dma/ste_dma40.c     |    2 +-
 include/linux/dmaengine.h   |    5 ++++
 7 files changed, 64 insertions(+), 37 deletions(-)
Signature · 1 line
1.7.10.4