[RFC v03 00/15] dmaengine: New 'universal' API for requesting channel

[RFC v03 00/15] dmaengine: New 'universal' API for requesting channel

[Peter Ujfalusi]

Hi,

I keep this still as RFC.

Changes since RFC v02:
- Using has_acpi_companion() instead ACPI_HANDLE()
- mask matching change within private_candidate()
- Fallback in dma_request_chan() when DT/ACPI lookup fails.
- Rename dma_get_channel() -> find_candidate()
- Arch code changes as suggested by Arnd
- Some documentation updated, more need to be done.

Changes since RFC v01:
- dma_request_chan(); lost the mask parameter
- The new API does not rely on RESOURCE_DMA, instead the dma_filter_map table
  will be used to provide the needed information to the filter function in
  legacy mode
- Extended the example patches to convert most of daVinci to use the new API to
  request the DMA channels.

TODO: Documentation update ;)

As it has been discussed in the following thread:
http://www.gossamer-threads.com/lists/linux/kernel/2181487#2181487

Andy: I did looked at the unified device properties, but I decided to not to use
it as I don't see it to fit well and most of the legacy board files are using
resources to specify at least their memory regions so adding the DMA resource
to them would be more inline with the rest of the code.

The ARM, mmc and spi patches are converting daVinci drivers board files to use
the new method of requesting DMA channel.

With this series I have taken a path which would result two new API, which can
be used to convert most of the current users already and with some work all
users might be able to move to this set.
With this set the filter_fn used for legacy (non DT/ACPI) channel request is no
longer needed to be exported to client drivers since the selection of the
correct filter_fn will be done in the core.

So, the first proposal is to have:

struct dma_chan *dma_request_chan(struct device *dev, const char *name);
struct dma_chan *dma_request_chan_by_mask(const dma_cap_mask_t *mask);

The dma_request_chan_by_mask() is to request any channel matching with the
requested capabilities, can be used to request channel for memcpy, memset, xor,
etc where no hardware synchronization is needed.

The dma_request_chan() is to request a slave channel. The dma_request_chan() will try to find the
channel via DT, ACPI or in case if the kernel booted in non DT/ACPI mode
it will use a filter lookup table and retrieves the needed information from
the dma_filter_map provided by the DMA drivers.
This legacy mode needs changes in platform code, in dmaengine drivers and
finally the dmaengine user drivers can be converted:

For each dmaengine driver an array of DMA device, slave and the parameter
for the filter function needs to be added:

static const struct dma_filter_map da830_edma_map[] = {
	{ "davinci-mcasp.0", "rx", EDMA_FILTER_PARAM(0, 0) },
	{ "davinci-mcasp.0", "tx", EDMA_FILTER_PARAM(0, 1) },
	{ "davinci-mcasp.1", "rx", EDMA_FILTER_PARAM(0, 2) },
	{ "davinci-mcasp.1", "tx", EDMA_FILTER_PARAM(0, 3) },
	{ "davinci-mcasp.2", "rx", EDMA_FILTER_PARAM(0, 4) },
	{ "davinci-mcasp.2", "tx", EDMA_FILTER_PARAM(0, 5) },
	{ "spi_davinci.0", "rx", EDMA_FILTER_PARAM(0, 14) },
	{ "spi_davinci.0", "tx", EDMA_FILTER_PARAM(0, 15) },
	{ "da830-mmc.0", "rx", EDMA_FILTER_PARAM(0, 16) },
	{ "da830-mmc.0", "tx", EDMA_FILTER_PARAM(0, 17) },
	{ "spi_davinci.1", "rx", EDMA_FILTER_PARAM(0, 18) },
	{ "spi_davinci.1", "tx", EDMA_FILTER_PARAM(0, 19) },
};

This information is going to be needed by the dmaengine driver, so
modification to the platform_data is needed, and the driver map should be
added to the pdata of the DMA driver:

da8xx_edma0_pdata.slave_map = da830_edma_map;
da8xx_edma0_pdata.slavecnt = ARRAY_SIZE(da830_edma_map);

The DMA driver then needs to configure the needed device -> filter_fn
mapping before it registers with dma_async_device_register() :

if (info->slave_map) {
	ecc->dma_slave.filter_map.map = info->slave_map;
	ecc->dma_slave.filter_map.mapcnt = info->slavecnt;
	ecc->dma_slave.filter_map.filter_fn = edma_filter_fn;
}

When neither DT or ACPI lookup is available the dma_request_chan() will
try to match the requester's device name with the filter_map's list of
device names, when a match found it will use the information from the
dma_filter_map to get the channel with the dma_get_channel() internal
function.

Regards,
Peter
---
Peter Ujfalusi (15):
  dmaengine: core: Skip mask matching when it is not provided to
    private_candidate
  dmaengine: core: Move and merge the code paths using private_candidate
  dmaengine: core: Introduce new, universal API to request a channel
  dmaengine: edma: Add support for DMA filter mapping to slave devices
  ARM: davinci: devices-da8xx: Add dma_filter_map to edma
  ARM: davinci: dm355: Add dma_filter_map to edma
  ARM: davinci: dm365: Add dma_filter_map to edma
  ARM: davinci: dm644x: Add dma_filter_map to edma
  ARM: davinci: dm646x: Add dma_filter_map to edma
  mmc: davinci_mmc: Use dma_request_chan() to requesting DMA channel
  spi: davinci: Use dma_request_chan() to requesting DMA channel
  ARM: davinci: devices-da8xx: Remove DMA resources for MMC and SPI
  ARM: davinci: devices: Remove DMA resources for MMC
  ARM: davinci: dm355: Remove DMA resources for SPI
  ARM: davinci: dm365: Remove DMA resources for SPI

 Documentation/dmaengine/client.txt    |  23 ++---
 arch/arm/mach-davinci/devices-da8xx.c |  95 +++++++++---------
 arch/arm/mach-davinci/devices.c       |  19 ----
 arch/arm/mach-davinci/dm355.c         |  28 ++++--
 arch/arm/mach-davinci/dm365.c         |  30 ++++--
 arch/arm/mach-davinci/dm644x.c        |  12 +++
 arch/arm/mach-davinci/dm646x.c        |  11 +++
 drivers/dma/dmaengine.c               | 181 ++++++++++++++++++++++++++--------
 drivers/dma/edma.c                    |   6 ++
 drivers/mmc/host/davinci_mmc.c        |  52 +++-------
 drivers/spi/spi-davinci.c             |  76 +++++---------
 include/linux/dmaengine.h             |  27 +++++
 include/linux/platform_data/edma.h    |   7 ++
 13 files changed, 336 insertions(+), 231 deletions(-)

-- 
2.6.3

[RFC v03 01/15] dmaengine: core: Skip mask matching when it is not provided to private_candidate

[Peter Ujfalusi]

If mask is NULL skip the mask matching against the DMA device capabilities.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 drivers/dma/dmaengine.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index daf54a39bcc7..6311e1fc80be 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -515,7 +515,7 @@ static struct dma_chan *private_candidate(const dma_cap_mask_t *mask,
 {
 	struct dma_chan *chan;
 
-	if (!__dma_device_satisfies_mask(dev, mask)) {
+	if (mask && !__dma_device_satisfies_mask(dev, mask)) {
 		pr_debug("%s: wrong capabilities\n", __func__);
 		return NULL;
 	}
-- 
2.6.3

[RFC v03 02/15] dmaengine: core: Move and merge the code paths using private_candidate

[Peter Ujfalusi]

Channel matching with private_candidate() is used in two paths, the error
checking is slightly different in them and they are duplicating code also.
Move the code under find_candidate() to provide consistent execution and
going to allow us to reuse this mode of channel lookup later.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 drivers/dma/dmaengine.c | 81 +++++++++++++++++++++++++------------------------
 1 file changed, 42 insertions(+), 39 deletions(-)

diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index 6311e1fc80be..ea9d66982d40 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -546,6 +546,42 @@ static struct dma_chan *private_candidate(const dma_cap_mask_t *mask,
 	return NULL;
 }
 
+static struct dma_chan *find_candidate(struct dma_device *device,
+				       const dma_cap_mask_t *mask,
+				       dma_filter_fn fn, void *fn_param)
+{
+	struct dma_chan *chan = private_candidate(mask, device, fn, fn_param);
+	int err;
+
+	if (chan) {
+		/* Found a suitable channel, try to grab, prep, and return it.
+		 * We first set DMA_PRIVATE to disable balance_ref_count as this
+		 * channel will not be published in the general-purpose
+		 * allocator
+		 */
+		dma_cap_set(DMA_PRIVATE, device->cap_mask);
+		device->privatecnt++;
+		err = dma_chan_get(chan);
+
+		if (err) {
+			if (err == -ENODEV) {
+				pr_debug("%s: %s module removed\n", __func__,
+					 dma_chan_name(chan));
+				list_del_rcu(&device->global_node);
+			} else
+				pr_debug("%s: failed to get %s: (%d)\n",
+					 __func__, dma_chan_name(chan), err);
+
+			if (--device->privatecnt == 0)
+				dma_cap_clear(DMA_PRIVATE, device->cap_mask);
+
+			chan = ERR_PTR(err);
+		}
+	}
+
+	return chan ? chan : ERR_PTR(-EPROBE_DEFER);
+}
+
 /**
  * dma_get_slave_channel - try to get specific channel exclusively
  * @chan: target channel
@@ -584,7 +620,6 @@ struct dma_chan *dma_get_any_slave_channel(struct dma_device *device)
 {
 	dma_cap_mask_t mask;
 	struct dma_chan *chan;
-	int err;
 
 	dma_cap_zero(mask);
 	dma_cap_set(DMA_SLAVE, mask);
@@ -592,23 +627,11 @@ struct dma_chan *dma_get_any_slave_channel(struct dma_device *device)
 	/* lock against __dma_request_channel */
 	mutex_lock(&dma_list_mutex);
 
-	chan = private_candidate(&mask, device, NULL, NULL);
-	if (chan) {
-		dma_cap_set(DMA_PRIVATE, device->cap_mask);
-		device->privatecnt++;
-		err = dma_chan_get(chan);
-		if (err) {
-			pr_debug("%s: failed to get %s: (%d)\n",
-				__func__, dma_chan_name(chan), err);
-			chan = NULL;
-			if (--device->privatecnt == 0)
-				dma_cap_clear(DMA_PRIVATE, device->cap_mask);
-		}
-	}
+	chan = find_candidate(device, &mask, NULL, NULL);
 
 	mutex_unlock(&dma_list_mutex);
 
-	return chan;
+	return IS_ERR(chan) ? NULL : chan;
 }
 EXPORT_SYMBOL_GPL(dma_get_any_slave_channel);
 
@@ -625,35 +648,15 @@ struct dma_chan *__dma_request_channel(const dma_cap_mask_t *mask,
 {
 	struct dma_device *device, *_d;
 	struct dma_chan *chan = NULL;
-	int err;
 
 	/* Find a channel */
 	mutex_lock(&dma_list_mutex);
 	list_for_each_entry_safe(device, _d, &dma_device_list, global_node) {
-		chan = private_candidate(mask, device, fn, fn_param);
-		if (chan) {
-			/* Found a suitable channel, try to grab, prep, and
-			 * return it.  We first set DMA_PRIVATE to disable
-			 * balance_ref_count as this channel will not be
-			 * published in the general-purpose allocator
-			 */
-			dma_cap_set(DMA_PRIVATE, device->cap_mask);
-			device->privatecnt++;
-			err = dma_chan_get(chan);
+		chan = find_candidate(device, mask, fn, fn_param);
+		if (!IS_ERR(chan))
+			break;
 
-			if (err == -ENODEV) {
-				pr_debug("%s: %s module removed\n",
-					 __func__, dma_chan_name(chan));
-				list_del_rcu(&device->global_node);
-			} else if (err)
-				pr_debug("%s: failed to get %s: (%d)\n",
-					 __func__, dma_chan_name(chan), err);
-			else
-				break;
-			if (--device->privatecnt == 0)
-				dma_cap_clear(DMA_PRIVATE, device->cap_mask);
-			chan = NULL;
-		}
+		chan = NULL;
 	}
 	mutex_unlock(&dma_list_mutex);
 
-- 
2.6.3

[RFC v03 03/15] dmaengine: core: Introduce new, universal API to request a channel

[Peter Ujfalusi]

The two API function can cover most, if not all current APIs used to
request a channel. With minimal effort dmaengine drivers, platforms and
dmaengine user drivers can be converted to use the two function.

struct dma_chan *dma_request_chan_by_mask(const dma_cap_mask_t *mask);

To request any channel matching with the requested capabilities, can be
used to request channel for memcpy, memset, xor, etc where no hardware
synchronization is needed.

struct dma_chan *dma_request_chan(struct device *dev, const char *name);
To request a slave channel. The dma_request_chan() will try to find the
channel via DT, ACPI or in case if the kernel booted in non DT/ACPI mode
it will use a filter lookup table and retrieves the needed information from
the dma_filter_map provided by the DMA drivers.
This legacy mode needs changes in platform code, in dmaengine drivers and
finally the dmaengine user drivers can be converted:

For each dmaengine driver an array of DMA device, slave and the parameter
for the filter function needs to be added:

static const struct dma_filter_map da830_edma_map[] = {
	{ "davinci-mcasp.0", "rx", EDMA_FILTER_PARAM(0, 0) },
	{ "davinci-mcasp.0", "tx", EDMA_FILTER_PARAM(0, 1) },
	{ "davinci-mcasp.1", "rx", EDMA_FILTER_PARAM(0, 2) },
	{ "davinci-mcasp.1", "tx", EDMA_FILTER_PARAM(0, 3) },
	{ "davinci-mcasp.2", "rx", EDMA_FILTER_PARAM(0, 4) },
	{ "davinci-mcasp.2", "tx", EDMA_FILTER_PARAM(0, 5) },
	{ "spi_davinci.0", "rx", EDMA_FILTER_PARAM(0, 14) },
	{ "spi_davinci.0", "tx", EDMA_FILTER_PARAM(0, 15) },
	{ "da830-mmc.0", "rx", EDMA_FILTER_PARAM(0, 16) },
	{ "da830-mmc.0", "tx", EDMA_FILTER_PARAM(0, 17) },
	{ "spi_davinci.1", "rx", EDMA_FILTER_PARAM(0, 18) },
	{ "spi_davinci.1", "tx", EDMA_FILTER_PARAM(0, 19) },
};

This information is going to be needed by the dmaengine driver, so
modification to the platform_data is needed, and the driver map should be
added to the pdata of the DMA driver:

da8xx_edma0_pdata.slave_map = da830_edma_map;
da8xx_edma0_pdata.slavecnt = ARRAY_SIZE(da830_edma_map);

The DMA driver then needs to configure the needed device -> filter_fn
mapping before it registers with dma_async_device_register() :

if (info->slave_map) {
	ecc->dma_slave.filter_map.map = info->slave_map;
	ecc->dma_slave.filter_map.mapcnt = info->slavecnt;
	ecc->dma_slave.filter_map.filter_fn = edma_filter_fn;
}

When neither DT or ACPI lookup is available the dma_request_chan() will
try to match the requester's device name with the filter_map's list of
device names, when a match found it will use the information from the
dma_filter_map to get the channel with the dma_get_channel() internal
function.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 Documentation/dmaengine/client.txt | 23 +++------
 drivers/dma/dmaengine.c            | 98 ++++++++++++++++++++++++++++++++++++++
 include/linux/dmaengine.h          | 27 +++++++++++
 3 files changed, 131 insertions(+), 17 deletions(-)

diff --git a/Documentation/dmaengine/client.txt b/Documentation/dmaengine/client.txt
index d9f9f461102a..6c72a06eb1a5 100644
--- a/Documentation/dmaengine/client.txt
+++ b/Documentation/dmaengine/client.txt
@@ -22,25 +22,14 @@ The slave DMA usage consists of following steps:
    Channel allocation is slightly different in the slave DMA context,
    client drivers typically need a channel from a particular DMA
    controller only and even in some cases a specific channel is desired.
-   To request a channel dma_request_channel() API is used.
+   To request a channel dma_request_chan() API is used.
 
    Interface:
-	struct dma_chan *dma_request_channel(dma_cap_mask_t mask,
-			dma_filter_fn filter_fn,
-			void *filter_param);
-   where dma_filter_fn is defined as:
-	typedef bool (*dma_filter_fn)(struct dma_chan *chan, void *filter_param);
-
-   The 'filter_fn' parameter is optional, but highly recommended for
-   slave and cyclic channels as they typically need to obtain a specific
-   DMA channel.
-
-   When the optional 'filter_fn' parameter is NULL, dma_request_channel()
-   simply returns the first channel that satisfies the capability mask.
-
-   Otherwise, the 'filter_fn' routine will be called once for each free
-   channel which has a capability in 'mask'.  'filter_fn' is expected to
-   return 'true' when the desired DMA channel is found.
+	struct dma_chan *dma_request_chan(struct device *dev, const char *name);
+
+   Which will find and return the 'name' DMA channel associated with the 'dev'
+   device. The association is done via DT, ACPI or board file based
+   dma_map_filter matching table.
 
    A channel allocated via this interface is exclusive to the caller,
    until dma_release_channel() is called.
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index ea9d66982d40..b9dc1512c4aa 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -43,6 +43,7 @@
 
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
+#include <linux/platform_device.h>
 #include <linux/dma-mapping.h>
 #include <linux/init.h>
 #include <linux/module.h>
@@ -712,6 +713,103 @@ struct dma_chan *dma_request_slave_channel(struct device *dev,
 }
 EXPORT_SYMBOL_GPL(dma_request_slave_channel);
 
+const static struct dma_filter_map *dma_filter_match(struct dma_device *device,
+					       const char *name,
+					       struct device *dev)
+{
+	const struct dma_filter_map *map_found = NULL;
+	int i;
+
+	if (!device->filter_map.mapcnt)
+		return NULL;
+
+	for (i = 0; i < device->filter_map.mapcnt; i++) {
+		const struct dma_filter_map *map = &device->filter_map.map[i];
+
+		if (!strcmp(map->devname, dev_name(dev)) &&
+		    !strcmp(map->slave, name)) {
+			map_found = map;
+			break;
+		}
+	}
+
+	return map_found;
+}
+
+/**
+ * dma_request_chan - try to allocate an exclusive slave channel
+ * @dev:	pointer to client device structure
+ * @name:	slave channel name
+ *
+ * Returns pointer to appropriate DMA channel on success or an error pointer.
+ */
+struct dma_chan *dma_request_chan(struct device *dev, const char *name)
+{
+	struct dma_device *device, *_d;
+	struct dma_chan *chan = NULL;
+
+	/* If device-tree is present get slave info from here */
+	if (dev->of_node)
+		chan = of_dma_request_slave_channel(dev->of_node, name);
+
+	/* If device was enumerated by ACPI get slave info from here */
+	if (has_acpi_companion(dev) && !chan)
+		chan = acpi_dma_request_slave_chan_by_name(dev, name);
+
+	if (chan) {
+		/* Valid channel found */
+		if (!IS_ERR(chan))
+			return chan;
+
+		pr_warn("%s: %s DMA request failed, falling back to legacy\n",
+			__func__, dev->of_node ? "OF" : "ACPI");
+	}
+
+	/* Try to find the channel via the DMA filter map(s) */
+	mutex_lock(&dma_list_mutex);
+	list_for_each_entry_safe(device, _d, &dma_device_list, global_node) {
+		dma_cap_mask_t mask;
+		const struct dma_filter_map *map = dma_filter_match(device,
+								    name, dev);
+
+		if (!map)
+			continue;
+
+		dma_cap_zero(mask);
+		dma_cap_set(DMA_SLAVE, mask);
+
+		chan = find_candidate(device, &mask,
+				      device->filter_map.filter_fn, map->param);
+		if (!IS_ERR(chan))
+			break;
+	}
+	mutex_unlock(&dma_list_mutex);
+
+	return chan ? chan : ERR_PTR(-EPROBE_DEFER);
+}
+EXPORT_SYMBOL_GPL(dma_request_chan);
+
+/**
+ * dma_request_chan_by_mask - allocate a channel satisfying certain capabilities
+ * @mask: capabilities that the channel must satisfy
+ *
+ * Returns pointer to appropriate DMA channel on success or an error pointer.
+ */
+struct dma_chan *dma_request_chan_by_mask(const dma_cap_mask_t *mask)
+{
+	struct dma_chan *chan;
+
+	if (!mask)
+		return ERR_PTR(-ENODEV);
+
+	chan = __dma_request_channel(mask, NULL, NULL);
+	if (!chan)
+		chan = ERR_PTR(-ENODEV);
+
+	return chan;
+}
+EXPORT_SYMBOL_GPL(dma_request_chan_by_mask);
+
 void dma_release_channel(struct dma_chan *chan)
 {
 	mutex_lock(&dma_list_mutex);
diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index a2b7c2071cf4..49d48e69c179 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -606,6 +606,18 @@ enum dmaengine_alignment {
 	DMAENGINE_ALIGN_64_BYTES = 6,
 };
 
+struct dma_filter_map {
+        char *devname;
+        char *slave;
+        void *param;
+};
+
+struct dma_filter {
+        dma_filter_fn filter_fn;
+        int mapcnt;
+        const struct dma_filter_map *map;
+};
+
 /**
  * struct dma_device - info on the entity supplying DMA services
  * @chancnt: how many DMA channels are supported
@@ -669,6 +681,7 @@ struct dma_device {
 	unsigned int privatecnt;
 	struct list_head channels;
 	struct list_head global_node;
+	struct dma_filter filter_map;
 	dma_cap_mask_t  cap_mask;
 	unsigned short max_xor;
 	unsigned short max_pq;
@@ -1235,6 +1248,10 @@ struct dma_chan *__dma_request_channel(const dma_cap_mask_t *mask,
 struct dma_chan *dma_request_slave_channel_reason(struct device *dev,
 						  const char *name);
 struct dma_chan *dma_request_slave_channel(struct device *dev, const char *name);
+
+struct dma_chan *dma_request_chan(struct device *dev, const char *name);
+struct dma_chan *dma_request_chan_by_mask(const dma_cap_mask_t *mask);
+
 void dma_release_channel(struct dma_chan *chan);
 int dma_get_slave_caps(struct dma_chan *chan, struct dma_slave_caps *caps);
 #else
@@ -1268,6 +1285,16 @@ static inline struct dma_chan *dma_request_slave_channel(struct device *dev,
 {
 	return NULL;
 }
+static inline struct dma_chan *dma_request_chan(struct device *dev,
+						const char *name)
+{
+	return ERR_PTR(-ENODEV);
+}
+static inline struct dma_chan *dma_request_chan_by_mask(
+						const dma_cap_mask_t *mask)
+{
+	return ERR_PTR(-ENODEV);
+}
 static inline void dma_release_channel(struct dma_chan *chan)
 {
 }
-- 
2.6.3

[RFC v03 04/15] dmaengine: edma: Add support for DMA filter mapping to slave devices

[Peter Ujfalusi]

Add support for providing device to filter_fn mapping so client drivers
can switch to use the dma_request_chan() API.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 drivers/dma/edma.c                 | 6 ++++++
 include/linux/platform_data/edma.h | 7 +++++++
 2 files changed, 13 insertions(+)

diff --git a/drivers/dma/edma.c b/drivers/dma/edma.c
index 0675e268d577..46b305ea0d21 100644
--- a/drivers/dma/edma.c
+++ b/drivers/dma/edma.c
@@ -2297,6 +2297,12 @@ static int edma_probe(struct platform_device *pdev)
 		edma_set_chmap(&ecc->slave_chans[i], ecc->dummy_slot);
 	}
 
+	if (info->slave_map) {
+		ecc->dma_slave.filter_map.map = info->slave_map;
+		ecc->dma_slave.filter_map.mapcnt = info->slavecnt;
+		ecc->dma_slave.filter_map.filter_fn = edma_filter_fn;
+	}
+
 	ret = dma_async_device_register(&ecc->dma_slave);
 	if (ret) {
 		dev_err(dev, "slave ddev registration failed (%d)\n", ret);
diff --git a/include/linux/platform_data/edma.h b/include/linux/platform_data/edma.h
index e2878baeb90e..4e78a946e0c1 100644
--- a/include/linux/platform_data/edma.h
+++ b/include/linux/platform_data/edma.h
@@ -53,12 +53,16 @@ enum dma_event_q {
 #define EDMA_CTLR(i)			((i) >> 16)
 #define EDMA_CHAN_SLOT(i)		((i) & 0xffff)
 
+#define EDMA_FILTER_PARAM(ctlr, chan)	((int[]) { EDMA_CTLR_CHAN(ctlr, chan) })
+
 struct edma_rsv_info {
 
 	const s16	(*rsv_chans)[2];
 	const s16	(*rsv_slots)[2];
 };
 
+struct dma_filter_map;
+
 /* platform_data for EDMA driver */
 struct edma_soc_info {
 	/*
@@ -76,6 +80,9 @@ struct edma_soc_info {
 
 	s8	(*queue_priority_mapping)[2];
 	const s16	(*xbar_chans)[2];
+
+	const struct dma_filter_map *slave_map;
+	int slavecnt;
 };
 
 #endif
-- 
2.6.3

[RFC v03 05/15] ARM: davinci: devices-da8xx: Add dma_filter_map to edma

[Peter Ujfalusi]

Provide the dma_filter_map to edma which will allow us to move the drivers
to the new, simpler dmaengine API and we can remove the DMA resources also
for the devices.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 arch/arm/mach-davinci/devices-da8xx.c | 46 +++++++++++++++++++++++++++++++++++
 1 file changed, 46 insertions(+)

diff --git a/arch/arm/mach-davinci/devices-da8xx.c b/arch/arm/mach-davinci/devices-da8xx.c
index 28c90bc372bd..9ffdcd5de0f8 100644
--- a/arch/arm/mach-davinci/devices-da8xx.c
+++ b/arch/arm/mach-davinci/devices-da8xx.c
@@ -17,6 +17,7 @@
 #include <linux/ahci_platform.h>
 #include <linux/clk.h>
 #include <linux/reboot.h>
+#include <linux/dmaengine.h>
 
 #include <mach/cputype.h>
 #include <mach/common.h>
@@ -233,16 +234,54 @@ static const struct platform_device_info da850_edma1_device __initconst = {
 	.size_data	= sizeof(da850_edma1_pdata),
 };
 
+static const struct dma_filter_map da830_edma_map[] = {
+	{ "davinci-mcasp.0", "rx", EDMA_FILTER_PARAM(0, 0) },
+	{ "davinci-mcasp.0", "tx", EDMA_FILTER_PARAM(0, 1) },
+	{ "davinci-mcasp.1", "rx", EDMA_FILTER_PARAM(0, 2) },
+	{ "davinci-mcasp.1", "tx", EDMA_FILTER_PARAM(0, 3) },
+	{ "davinci-mcasp.2", "rx", EDMA_FILTER_PARAM(0, 4) },
+	{ "davinci-mcasp.2", "tx", EDMA_FILTER_PARAM(0, 5) },
+	{ "spi_davinci.0", "rx", EDMA_FILTER_PARAM(0, 14) },
+	{ "spi_davinci.0", "tx", EDMA_FILTER_PARAM(0, 15) },
+	{ "da830-mmc.0", "rx", EDMA_FILTER_PARAM(0, 16) },
+	{ "da830-mmc.0", "tx", EDMA_FILTER_PARAM(0, 17) },
+	{ "spi_davinci.1", "rx", EDMA_FILTER_PARAM(0, 18) },
+	{ "spi_davinci.1", "tx", EDMA_FILTER_PARAM(0, 19) },
+};
+
 int __init da830_register_edma(struct edma_rsv_info *rsv)
 {
 	struct platform_device *edma_pdev;
 
 	da8xx_edma0_pdata.rsv = rsv;
 
+	da8xx_edma0_pdata.slave_map = da830_edma_map;
+	da8xx_edma0_pdata.slavecnt = ARRAY_SIZE(da830_edma_map);
+
 	edma_pdev = platform_device_register_full(&da8xx_edma0_device);
 	return IS_ERR(edma_pdev) ? PTR_ERR(edma_pdev) : 0;
 }
 
+static const struct dma_filter_map da850_edma0_map[] = {
+	{ "davinci-mcasp.0", "rx", EDMA_FILTER_PARAM(0, 0) },
+	{ "davinci-mcasp.0", "tx", EDMA_FILTER_PARAM(0, 1) },
+	{ "davinci-mcbsp.0", "rx", EDMA_FILTER_PARAM(0, 2) },
+	{ "davinci-mcbsp.0", "tx", EDMA_FILTER_PARAM(0, 3) },
+	{ "davinci-mcbsp.1", "rx", EDMA_FILTER_PARAM(0, 4) },
+	{ "davinci-mcbsp.1", "tx", EDMA_FILTER_PARAM(0, 5) },
+	{ "spi_davinci.0", "rx", EDMA_FILTER_PARAM(0, 14) },
+	{ "spi_davinci.0", "tx", EDMA_FILTER_PARAM(0, 15) },
+	{ "da830-mmc.0", "rx", EDMA_FILTER_PARAM(0, 16) },
+	{ "da830-mmc.0", "tx", EDMA_FILTER_PARAM(0, 17) },
+	{ "spi_davinci.1", "rx", EDMA_FILTER_PARAM(0, 18) },
+	{ "spi_davinci.1", "tx", EDMA_FILTER_PARAM(0, 19) },
+};
+
+static const struct dma_filter_map da850_edma1_map[] = {
+	{ "da830-mmc.1", "tx", EDMA_FILTER_PARAM(0, 28) },
+	{ "da830-mmc.1", "rx", EDMA_FILTER_PARAM(0, 29) },
+};
+
 int __init da850_register_edma(struct edma_rsv_info *rsv[2])
 {
 	struct platform_device *edma_pdev;
@@ -252,11 +291,18 @@ int __init da850_register_edma(struct edma_rsv_info *rsv[2])
 		da850_edma1_pdata.rsv = rsv[1];
 	}
 
+	da8xx_edma0_pdata.slave_map = da850_edma0_map;
+	da8xx_edma0_pdata.slavecnt = ARRAY_SIZE(da850_edma0_map);
+
 	edma_pdev = platform_device_register_full(&da8xx_edma0_device);
 	if (IS_ERR(edma_pdev)) {
 		pr_warn("%s: Failed to register eDMA0\n", __func__);
 		return PTR_ERR(edma_pdev);
 	}
+
+	da850_edma1_pdata.slave_map = da850_edma1_map;
+	da850_edma1_pdata.slavecnt = ARRAY_SIZE(da850_edma1_map);
+
 	edma_pdev = platform_device_register_full(&da850_edma1_device);
 	return IS_ERR(edma_pdev) ? PTR_ERR(edma_pdev) : 0;
 }
-- 
2.6.3

[RFC v03 06/15] ARM: davinci: dm355: Add dma_filter_map to edma

[Peter Ujfalusi]

Provide the dma_filter_map to edma which will allow us to move the drivers
to the new, simpler dmaengine API and we can remove the DMA resources also
for the devices.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 arch/arm/mach-davinci/dm355.c | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)

diff --git a/arch/arm/mach-davinci/dm355.c b/arch/arm/mach-davinci/dm355.c
index 609950b8c191..ee7656fa0c52 100644
--- a/arch/arm/mach-davinci/dm355.c
+++ b/arch/arm/mach-davinci/dm355.c
@@ -13,6 +13,7 @@
 #include <linux/serial_8250.h>
 #include <linux/platform_device.h>
 #include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
 #include <linux/spi/spi.h>
 #include <linux/platform_data/edma.h>
 #include <linux/platform_data/gpio-davinci.h>
@@ -576,9 +577,28 @@ static s8 queue_priority_mapping[][2] = {
 	{-1, -1},
 };
 
+static const struct dma_filter_map da355_edma_map[] = {
+	{ "davinci-mcbsp.0", "tx", EDMA_FILTER_PARAM(0, 2) },
+	{ "davinci-mcbsp.0", "rx", EDMA_FILTER_PARAM(0, 3) },
+	{ "davinci-mcbsp.1", "tx", EDMA_FILTER_PARAM(0, 8) },
+	{ "davinci-mcbsp.1", "rx", EDMA_FILTER_PARAM(0, 9) },
+	{ "spi_davinci.2", "tx", EDMA_FILTER_PARAM(0, 10) },
+	{ "spi_davinci.2", "rx", EDMA_FILTER_PARAM(0, 11) },
+	{ "spi_davinci.1", "tx", EDMA_FILTER_PARAM(0, 14) },
+	{ "spi_davinci.1", "rx", EDMA_FILTER_PARAM(0, 15) },
+	{ "spi_davinci.0", "tx", EDMA_FILTER_PARAM(0, 16) },
+	{ "spi_davinci.0", "rx", EDMA_FILTER_PARAM(0, 17) },
+	{ "dm6441-mmc.0", "rx", EDMA_FILTER_PARAM(0, 26) },
+	{ "dm6441-mmc.0", "tx", EDMA_FILTER_PARAM(0, 27) },
+	{ "dm6441-mmc.1", "rx", EDMA_FILTER_PARAM(0, 30) },
+	{ "dm6441-mmc.1", "tx", EDMA_FILTER_PARAM(0, 31) },
+};
+
 static struct edma_soc_info dm355_edma_pdata = {
 	.queue_priority_mapping	= queue_priority_mapping,
 	.default_queue		= EVENTQ_1,
+	.slave_map		= da355_edma_map,
+	.slavecnt		= ARRAY_SIZE(da355_edma_map),
 };
 
 static struct resource edma_resources[] = {
-- 
2.6.3

[RFC v03 07/15] ARM: davinci: dm365: Add dma_filter_map to edma

[Peter Ujfalusi]

Provide the dma_filter_map to edma which will allow us to move the drivers
to the new, simpler dmaengine API and we can remove the DMA resources also
for the devices.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 arch/arm/mach-davinci/dm365.c | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)

diff --git a/arch/arm/mach-davinci/dm365.c b/arch/arm/mach-davinci/dm365.c
index 2068cbeaeb03..e794bff7d589 100644
--- a/arch/arm/mach-davinci/dm365.c
+++ b/arch/arm/mach-davinci/dm365.c
@@ -17,6 +17,7 @@
 #include <linux/serial_8250.h>
 #include <linux/platform_device.h>
 #include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
 #include <linux/spi/spi.h>
 #include <linux/platform_data/edma.h>
 #include <linux/platform_data/gpio-davinci.h>
@@ -862,9 +863,30 @@ static s8 dm365_queue_priority_mapping[][2] = {
 	{-1, -1},
 };
 
+static const struct dma_filter_map da365_edma_map[] = {
+	{ "davinci-mcbsp.0", "tx", EDMA_FILTER_PARAM(0, 2) },
+	{ "davinci-mcbsp.0", "rx", EDMA_FILTER_PARAM(0, 3) },
+	{ "davinci_voicecodec", "tx", EDMA_FILTER_PARAM(0, 2) },
+	{ "davinci_voicecodec", "rx", EDMA_FILTER_PARAM(0, 3) },
+	{ "spi_davinci.2", "tx", EDMA_FILTER_PARAM(0, 10) },
+	{ "spi_davinci.2", "rx", EDMA_FILTER_PARAM(0, 11) },
+	{ "spi_davinci.1", "tx", EDMA_FILTER_PARAM(0, 14) },
+	{ "spi_davinci.1", "rx", EDMA_FILTER_PARAM(0, 15) },
+	{ "spi_davinci.0", "tx", EDMA_FILTER_PARAM(0, 16) },
+	{ "spi_davinci.0", "rx", EDMA_FILTER_PARAM(0, 17) },
+	{ "spi_davinci.3", "tx", EDMA_FILTER_PARAM(0, 18) },
+	{ "spi_davinci.3", "rx", EDMA_FILTER_PARAM(0, 19) },
+	{ "dm6441-mmc.0", "rx", EDMA_FILTER_PARAM(0, 26) },
+	{ "dm6441-mmc.0", "tx", EDMA_FILTER_PARAM(0, 27) },
+	{ "dm6441-mmc.1", "rx", EDMA_FILTER_PARAM(0, 30) },
+	{ "dm6441-mmc.1", "tx", EDMA_FILTER_PARAM(0, 31) },
+};
+
 static struct edma_soc_info dm365_edma_pdata = {
 	.queue_priority_mapping	= dm365_queue_priority_mapping,
 	.default_queue		= EVENTQ_3,
+	.slave_map		= da365_edma_map,
+	.slavecnt		= ARRAY_SIZE(da365_edma_map),
 };
 
 static struct resource edma_resources[] = {
-- 
2.6.3

[RFC v03 08/15] ARM: davinci: dm644x: Add dma_filter_map to edma

[Peter Ujfalusi]

Provide the dma_filter_map to edma which will allow us to move the drivers
to the new, simpler dmaengine API and we can remove the DMA resources also
for the devices.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 arch/arm/mach-davinci/dm644x.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/arch/arm/mach-davinci/dm644x.c b/arch/arm/mach-davinci/dm644x.c
index d38f5049d56e..59fae2c3b8f4 100644
--- a/arch/arm/mach-davinci/dm644x.c
+++ b/arch/arm/mach-davinci/dm644x.c
@@ -11,6 +11,7 @@
 #include <linux/init.h>
 #include <linux/clk.h>
 #include <linux/serial_8250.h>
+#include <linux/dmaengine.h>
 #include <linux/platform_device.h>
 #include <linux/platform_data/edma.h>
 #include <linux/platform_data/gpio-davinci.h>
@@ -505,9 +506,20 @@ static s8 queue_priority_mapping[][2] = {
 	{-1, -1},
 };
 
+static const struct dma_filter_map da644x_edma_map[] = {
+	{ "davinci-mcbsp", "tx", EDMA_FILTER_PARAM(0, 2) },
+	{ "davinci-mcbsp", "rx", EDMA_FILTER_PARAM(0, 3) },
+	{ "spi_davinci", "tx", EDMA_FILTER_PARAM(0, 16) },
+	{ "spi_davinci", "rx", EDMA_FILTER_PARAM(0, 17) },
+	{ "dm6441-mmc.0", "rx", EDMA_FILTER_PARAM(0, 26) },
+	{ "dm6441-mmc.0", "tx", EDMA_FILTER_PARAM(0, 27) },
+};
+
 static struct edma_soc_info dm644x_edma_pdata = {
 	.queue_priority_mapping	= queue_priority_mapping,
 	.default_queue		= EVENTQ_1,
+	.slave_map		= da644x_edma_map,
+	.slavecnt		= ARRAY_SIZE(da644x_edma_map),
 };
 
 static struct resource edma_resources[] = {
-- 
2.6.3

[RFC v03 09/15] ARM: davinci: dm646x: Add dma_filter_map to edma

[Peter Ujfalusi]

Provide the dma_filter_map to edma which will allow us to move the drivers
to the new, simpler dmaengine API and we can remove the DMA resources also
for the devices.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 arch/arm/mach-davinci/dm646x.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/arch/arm/mach-davinci/dm646x.c b/arch/arm/mach-davinci/dm646x.c
index 70eb42725eec..0f2dada231a1 100644
--- a/arch/arm/mach-davinci/dm646x.c
+++ b/arch/arm/mach-davinci/dm646x.c
@@ -9,6 +9,7 @@
  * or implied.
  */
 #include <linux/dma-mapping.h>
+#include <linux/dmaengine.h>
 #include <linux/init.h>
 #include <linux/clk.h>
 #include <linux/serial_8250.h>
@@ -540,9 +541,19 @@ static s8 dm646x_queue_priority_mapping[][2] = {
 	{-1, -1},
 };
 
+static const struct dma_filter_map da646x_edma_map[] = {
+	{ "davinci-mcasp.0", "tx", EDMA_FILTER_PARAM(0, 6) },
+	{ "davinci-mcasp.0", "rx", EDMA_FILTER_PARAM(0, 9) },
+	{ "davinci-mcasp.1", "tx", EDMA_FILTER_PARAM(0, 12) },
+	{ "spi_davinci", "tx", EDMA_FILTER_PARAM(0, 16) },
+	{ "spi_davinci", "rx", EDMA_FILTER_PARAM(0, 17) },
+};
+
 static struct edma_soc_info dm646x_edma_pdata = {
 	.queue_priority_mapping	= dm646x_queue_priority_mapping,
 	.default_queue		= EVENTQ_1,
+	.slave_map		= da646x_edma_map,
+	.slavecnt		= ARRAY_SIZE(da646x_edma_map),
 };
 
 static struct resource edma_resources[] = {
-- 
2.6.3

[RFC v03 10/15] mmc: davinci_mmc: Use dma_request_chan() to requesting DMA channel

[Peter Ujfalusi]

With the new dma_request_chan() the clinet driver does not need to look for
the DMA resource and it does not need to pass filter_fn anymore.
By switching to the new API the davinci_mmc driver can now support deferred
probing against DMA.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 drivers/mmc/host/davinci_mmc.c | 52 ++++++++++++------------------------------
 1 file changed, 14 insertions(+), 38 deletions(-)

diff --git a/drivers/mmc/host/davinci_mmc.c b/drivers/mmc/host/davinci_mmc.c
index ea2a2ebc6b91..8a0a2c291e0c 100644
--- a/drivers/mmc/host/davinci_mmc.c
+++ b/drivers/mmc/host/davinci_mmc.c
@@ -32,12 +32,10 @@
 #include <linux/delay.h>
 #include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
-#include <linux/edma.h>
 #include <linux/mmc/mmc.h>
 #include <linux/of.h>
 #include <linux/of_device.h>
 
-#include <linux/platform_data/edma.h>
 #include <linux/platform_data/mmc-davinci.h>
 
 /*
@@ -517,35 +515,20 @@ davinci_release_dma_channels(struct mmc_davinci_host *host)
 
 static int __init davinci_acquire_dma_channels(struct mmc_davinci_host *host)
 {
-	int r;
-	dma_cap_mask_t mask;
-
-	dma_cap_zero(mask);
-	dma_cap_set(DMA_SLAVE, mask);
-
-	host->dma_tx =
-		dma_request_slave_channel_compat(mask, edma_filter_fn,
-				&host->txdma, mmc_dev(host->mmc), "tx");
-	if (!host->dma_tx) {
+	host->dma_tx = dma_request_chan(mmc_dev(host->mmc), "tx");
+	if (IS_ERR(host->dma_tx)) {
 		dev_err(mmc_dev(host->mmc), "Can't get dma_tx channel\n");
-		return -ENODEV;
+		return PTR_ERR(host->dma_tx);
 	}
 
-	host->dma_rx =
-		dma_request_slave_channel_compat(mask, edma_filter_fn,
-				&host->rxdma, mmc_dev(host->mmc), "rx");
-	if (!host->dma_rx) {
+	host->dma_rx = dma_request_chan(mmc_dev(host->mmc), "rx");
+	if (IS_ERR(host->dma_rx)) {
 		dev_err(mmc_dev(host->mmc), "Can't get dma_rx channel\n");
-		r = -ENODEV;
-		goto free_master_write;
+		dma_release_channel(host->dma_tx);
+		return PTR_ERR(host->dma_rx);
 	}
 
 	return 0;
-
-free_master_write:
-	dma_release_channel(host->dma_tx);
-
-	return r;
 }
 
 /*----------------------------------------------------------------------*/
@@ -1262,18 +1245,6 @@ static int __init davinci_mmcsd_probe(struct platform_device *pdev)
 	host = mmc_priv(mmc);
 	host->mmc = mmc;	/* Important */
 
-	r = platform_get_resource(pdev, IORESOURCE_DMA, 0);
-	if (!r)
-		dev_warn(&pdev->dev, "RX DMA resource not specified\n");
-	else
-		host->rxdma = r->start;
-
-	r = platform_get_resource(pdev, IORESOURCE_DMA, 1);
-	if (!r)
-		dev_warn(&pdev->dev, "TX DMA resource not specified\n");
-	else
-		host->txdma = r->start;
-
 	host->mem_res = mem;
 	host->base = ioremap(mem->start, mem_size);
 	if (!host->base)
@@ -1300,8 +1271,13 @@ static int __init davinci_mmcsd_probe(struct platform_device *pdev)
 	host->mmc_irq = irq;
 	host->sdio_irq = platform_get_irq(pdev, 1);
 
-	if (host->use_dma && davinci_acquire_dma_channels(host) != 0)
-		host->use_dma = 0;
+	if (host->use_dma) {
+		ret = davinci_acquire_dma_channels(host);
+		if (ret == -EPROBE_DEFER)
+			goto out;
+		else if (ret)
+			host->use_dma = 0;
+	}
 
 	/* REVISIT:  someday, support IRQ-driven card detection.  */
 	mmc->caps |= MMC_CAP_NEEDS_POLL;
-- 
2.6.3

[RFC v03 11/15] spi: davinci: Use dma_request_chan() to requesting DMA channel

[Peter Ujfalusi]

With the new dma_request_chan() the clinet driver does not need to look for
the DMA resource and it does not need to pass filter_fn anymore.
By switching to the new API the davinci_mmc driver can now support deferred
probing against DMA.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 drivers/spi/spi-davinci.c | 76 +++++++++++++++--------------------------------
 1 file changed, 24 insertions(+), 52 deletions(-)

diff --git a/drivers/spi/spi-davinci.c b/drivers/spi/spi-davinci.c
index 7d3af3eacf57..540fc8754085 100644
--- a/drivers/spi/spi-davinci.c
+++ b/drivers/spi/spi-davinci.c
@@ -23,7 +23,6 @@
 #include <linux/clk.h>
 #include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
-#include <linux/edma.h>
 #include <linux/of.h>
 #include <linux/of_device.h>
 #include <linux/of_gpio.h>
@@ -33,8 +32,6 @@
 
 #include <linux/platform_data/spi-davinci.h>
 
-#define SPI_NO_RESOURCE		((resource_size_t)-1)
-
 #define CS_DEFAULT	0xFF
 
 #define SPIFMT_PHASE_MASK	BIT(16)
@@ -130,8 +127,6 @@ struct davinci_spi {
 
 	struct dma_chan		*dma_rx;
 	struct dma_chan		*dma_tx;
-	int			dma_rx_chnum;
-	int			dma_tx_chnum;
 
 	struct davinci_spi_platform_data pdata;
 
@@ -796,35 +791,19 @@ static irqreturn_t davinci_spi_irq(s32 irq, void *data)
 
 static int davinci_spi_request_dma(struct davinci_spi *dspi)
 {
-	dma_cap_mask_t mask;
 	struct device *sdev = dspi->bitbang.master->dev.parent;
-	int r;
-
-	dma_cap_zero(mask);
-	dma_cap_set(DMA_SLAVE, mask);
 
-	dspi->dma_rx = dma_request_channel(mask, edma_filter_fn,
-					   &dspi->dma_rx_chnum);
-	if (!dspi->dma_rx) {
-		dev_err(sdev, "request RX DMA channel failed\n");
-		r = -ENODEV;
-		goto rx_dma_failed;
-	}
+	dspi->dma_rx = dma_request_chan(sdev, "rx");
+	if (IS_ERR(dspi->dma_rx))
+		return PTR_ERR(dspi->dma_rx);
 
-	dspi->dma_tx = dma_request_channel(mask, edma_filter_fn,
-					   &dspi->dma_tx_chnum);
-	if (!dspi->dma_tx) {
-		dev_err(sdev, "request TX DMA channel failed\n");
-		r = -ENODEV;
-		goto tx_dma_failed;
+	dspi->dma_tx = dma_request_chan(sdev, "tx");
+	if (IS_ERR(dspi->dma_tx)) {
+		dma_release_channel(dspi->dma_rx);
+		return PTR_ERR(dspi->dma_tx);
 	}
 
 	return 0;
-
-tx_dma_failed:
-	dma_release_channel(dspi->dma_rx);
-rx_dma_failed:
-	return r;
 }
 
 #if defined(CONFIG_OF)
@@ -935,8 +914,6 @@ static int davinci_spi_probe(struct platform_device *pdev)
 	struct davinci_spi *dspi;
 	struct davinci_spi_platform_data *pdata;
 	struct resource *r;
-	resource_size_t dma_rx_chan = SPI_NO_RESOURCE;
-	resource_size_t	dma_tx_chan = SPI_NO_RESOURCE;
 	int ret = 0;
 	u32 spipc0;
 
@@ -1043,27 +1020,15 @@ static int davinci_spi_probe(struct platform_device *pdev)
 		}
 	}
 
-	r = platform_get_resource(pdev, IORESOURCE_DMA, 0);
-	if (r)
-		dma_rx_chan = r->start;
-	r = platform_get_resource(pdev, IORESOURCE_DMA, 1);
-	if (r)
-		dma_tx_chan = r->start;
-
 	dspi->bitbang.txrx_bufs = davinci_spi_bufs;
-	if (dma_rx_chan != SPI_NO_RESOURCE &&
-	    dma_tx_chan != SPI_NO_RESOURCE) {
-		dspi->dma_rx_chnum = dma_rx_chan;
-		dspi->dma_tx_chnum = dma_tx_chan;
-
-		ret = davinci_spi_request_dma(dspi);
-		if (ret)
-			goto free_clk;
-
-		dev_info(&pdev->dev, "DMA: supported\n");
-		dev_info(&pdev->dev, "DMA: RX channel: %pa, TX channel: %pa, event queue: %d\n",
-				&dma_rx_chan, &dma_tx_chan,
-				pdata->dma_event_q);
+
+	ret = davinci_spi_request_dma(dspi);
+	if (ret == -EPROBE_DEFER) {
+		goto free_clk;
+	} else if (ret) {
+		dev_info(&pdev->dev, "DMA is not supported (%d)\n", ret);
+		dspi->dma_rx = NULL;
+		dspi->dma_tx = NULL;
 	}
 
 	dspi->get_rx = davinci_spi_rx_buf_u8;
@@ -1101,8 +1066,10 @@ static int davinci_spi_probe(struct platform_device *pdev)
 	return ret;
 
 free_dma:
-	dma_release_channel(dspi->dma_rx);
-	dma_release_channel(dspi->dma_tx);
+	if (dspi->dma_rx) {
+		dma_release_channel(dspi->dma_rx);
+		dma_release_channel(dspi->dma_tx);
+	}
 free_clk:
 	clk_disable_unprepare(dspi->clk);
 free_master:
@@ -1133,6 +1100,11 @@ static int davinci_spi_remove(struct platform_device *pdev)
 	clk_disable_unprepare(dspi->clk);
 	spi_master_put(master);
 
+	if (dspi->dma_rx) {
+		dma_release_channel(dspi->dma_rx);
+		dma_release_channel(dspi->dma_tx);
+	}
+
 	return 0;
 }
 
-- 
2.6.3

[RFC v03 12/15] ARM: davinci: devices-da8xx: Remove DMA resources for MMC and SPI

[Peter Ujfalusi]

The drivers are now converted to not use the DMA resource.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 arch/arm/mach-davinci/devices-da8xx.c | 49 -----------------------------------
 1 file changed, 49 deletions(-)

diff --git a/arch/arm/mach-davinci/devices-da8xx.c b/arch/arm/mach-davinci/devices-da8xx.c
index 9ffdcd5de0f8..242fb48371ae 100644
--- a/arch/arm/mach-davinci/devices-da8xx.c
+++ b/arch/arm/mach-davinci/devices-da8xx.c
@@ -57,15 +57,6 @@
 #define DA8XX_EMAC_RAM_OFFSET		0x0000
 #define DA8XX_EMAC_CTRL_RAM_SIZE	SZ_8K
 
-#define DA8XX_DMA_SPI0_RX	EDMA_CTLR_CHAN(0, 14)
-#define DA8XX_DMA_SPI0_TX	EDMA_CTLR_CHAN(0, 15)
-#define DA8XX_DMA_MMCSD0_RX	EDMA_CTLR_CHAN(0, 16)
-#define DA8XX_DMA_MMCSD0_TX	EDMA_CTLR_CHAN(0, 17)
-#define DA8XX_DMA_SPI1_RX	EDMA_CTLR_CHAN(0, 18)
-#define DA8XX_DMA_SPI1_TX	EDMA_CTLR_CHAN(0, 19)
-#define DA850_DMA_MMCSD1_RX	EDMA_CTLR_CHAN(1, 28)
-#define DA850_DMA_MMCSD1_TX	EDMA_CTLR_CHAN(1, 29)
-
 void __iomem *da8xx_syscfg0_base;
 void __iomem *da8xx_syscfg1_base;
 
@@ -751,16 +742,6 @@ static struct resource da8xx_mmcsd0_resources[] = {
 		.end	= IRQ_DA8XX_MMCSDINT0,
 		.flags	= IORESOURCE_IRQ,
 	},
-	{		/* DMA RX */
-		.start	= DA8XX_DMA_MMCSD0_RX,
-		.end	= DA8XX_DMA_MMCSD0_RX,
-		.flags	= IORESOURCE_DMA,
-	},
-	{		/* DMA TX */
-		.start	= DA8XX_DMA_MMCSD0_TX,
-		.end	= DA8XX_DMA_MMCSD0_TX,
-		.flags	= IORESOURCE_DMA,
-	},
 };
 
 static struct platform_device da8xx_mmcsd0_device = {
@@ -788,16 +769,6 @@ static struct resource da850_mmcsd1_resources[] = {
 		.end	= IRQ_DA850_MMCSDINT0_1,
 		.flags	= IORESOURCE_IRQ,
 	},
-	{		/* DMA RX */
-		.start	= DA850_DMA_MMCSD1_RX,
-		.end	= DA850_DMA_MMCSD1_RX,
-		.flags	= IORESOURCE_DMA,
-	},
-	{		/* DMA TX */
-		.start	= DA850_DMA_MMCSD1_TX,
-		.end	= DA850_DMA_MMCSD1_TX,
-		.flags	= IORESOURCE_DMA,
-	},
 };
 
 static struct platform_device da850_mmcsd1_device = {
@@ -984,16 +955,6 @@ static struct resource da8xx_spi0_resources[] = {
 		.end	= IRQ_DA8XX_SPINT0,
 		.flags	= IORESOURCE_IRQ,
 	},
-	[2] = {
-		.start	= DA8XX_DMA_SPI0_RX,
-		.end	= DA8XX_DMA_SPI0_RX,
-		.flags	= IORESOURCE_DMA,
-	},
-	[3] = {
-		.start	= DA8XX_DMA_SPI0_TX,
-		.end	= DA8XX_DMA_SPI0_TX,
-		.flags	= IORESOURCE_DMA,
-	},
 };
 
 static struct resource da8xx_spi1_resources[] = {
@@ -1007,16 +968,6 @@ static struct resource da8xx_spi1_resources[] = {
 		.end	= IRQ_DA8XX_SPINT1,
 		.flags	= IORESOURCE_IRQ,
 	},
-	[2] = {
-		.start	= DA8XX_DMA_SPI1_RX,
-		.end	= DA8XX_DMA_SPI1_RX,
-		.flags	= IORESOURCE_DMA,
-	},
-	[3] = {
-		.start	= DA8XX_DMA_SPI1_TX,
-		.end	= DA8XX_DMA_SPI1_TX,
-		.flags	= IORESOURCE_DMA,
-	},
 };
 
 static struct davinci_spi_platform_data da8xx_spi_pdata[] = {
-- 
2.6.3

[RFC v03 13/15] ARM: davinci: devices: Remove DMA resources for MMC

[Peter Ujfalusi]

The driver is converted to not use the DMA resource.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 arch/arm/mach-davinci/devices.c | 19 -------------------
 1 file changed, 19 deletions(-)

diff --git a/arch/arm/mach-davinci/devices.c b/arch/arm/mach-davinci/devices.c
index 6257aa452568..3ae70f2909b0 100644
--- a/arch/arm/mach-davinci/devices.c
+++ b/arch/arm/mach-davinci/devices.c
@@ -36,9 +36,6 @@
 #define DM365_MMCSD0_BASE	     0x01D11000
 #define DM365_MMCSD1_BASE	     0x01D00000
 
-#define DAVINCI_DMA_MMCRXEVT	26
-#define DAVINCI_DMA_MMCTXEVT	27
-
 void __iomem  *davinci_sysmod_base;
 
 void davinci_map_sysmod(void)
@@ -144,14 +141,6 @@ static struct resource mmcsd0_resources[] = {
 		.start = IRQ_SDIOINT,
 		.flags = IORESOURCE_IRQ,
 	},
-	/* DMA channels: RX, then TX */
-	{
-		.start = EDMA_CTLR_CHAN(0, DAVINCI_DMA_MMCRXEVT),
-		.flags = IORESOURCE_DMA,
-	}, {
-		.start = EDMA_CTLR_CHAN(0, DAVINCI_DMA_MMCTXEVT),
-		.flags = IORESOURCE_DMA,
-	},
 };
 
 static struct platform_device davinci_mmcsd0_device = {
@@ -181,14 +170,6 @@ static struct resource mmcsd1_resources[] = {
 		.start = IRQ_DM355_SDIOINT1,
 		.flags = IORESOURCE_IRQ,
 	},
-	/* DMA channels: RX, then TX */
-	{
-		.start = EDMA_CTLR_CHAN(0, 30),	/* rx */
-		.flags = IORESOURCE_DMA,
-	}, {
-		.start = EDMA_CTLR_CHAN(0, 31),	/* tx */
-		.flags = IORESOURCE_DMA,
-	},
 };
 
 static struct platform_device davinci_mmcsd1_device = {
-- 
2.6.3

[RFC v03 14/15] ARM: davinci: dm355: Remove DMA resources for SPI

[Peter Ujfalusi]

The driver is converted to not use the DMA resource.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 arch/arm/mach-davinci/dm355.c | 8 --------
 1 file changed, 8 deletions(-)

diff --git a/arch/arm/mach-davinci/dm355.c b/arch/arm/mach-davinci/dm355.c
index ee7656fa0c52..bed8f49eb60c 100644
--- a/arch/arm/mach-davinci/dm355.c
+++ b/arch/arm/mach-davinci/dm355.c
@@ -397,14 +397,6 @@ static struct resource dm355_spi0_resources[] = {
 		.start = IRQ_DM355_SPINT0_0,
 		.flags = IORESOURCE_IRQ,
 	},
-	{
-		.start = 17,
-		.flags = IORESOURCE_DMA,
-	},
-	{
-		.start = 16,
-		.flags = IORESOURCE_DMA,
-	},
 };
 
 static struct davinci_spi_platform_data dm355_spi0_pdata = {
-- 
2.6.3

[RFC v03 15/15] ARM: davinci: dm365: Remove DMA resources for SPI

[Peter Ujfalusi]

The driver is converted to not use the DMA resource.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---
 arch/arm/mach-davinci/dm365.c | 8 --------
 1 file changed, 8 deletions(-)

diff --git a/arch/arm/mach-davinci/dm365.c b/arch/arm/mach-davinci/dm365.c
index e794bff7d589..664ee6bbef22 100644
--- a/arch/arm/mach-davinci/dm365.c
+++ b/arch/arm/mach-davinci/dm365.c
@@ -660,14 +660,6 @@ static struct resource dm365_spi0_resources[] = {
 		.start = IRQ_DM365_SPIINT0_0,
 		.flags = IORESOURCE_IRQ,
 	},
-	{
-		.start = 17,
-		.flags = IORESOURCE_DMA,
-	},
-	{
-		.start = 16,
-		.flags = IORESOURCE_DMA,
-	},
 };
 
 static struct platform_device dm365_spi0_device = {
-- 
2.6.3

[RFC v03 03/15] dmaengine: core: Introduce new, universal API to request a channel

[Andy Shevchenko]

On Wed, Dec 2, 2015 at 3:59 PM, Peter Ujfalusi <peter.ujfalusi@ti.com> wrote:
> The two API function can cover most, if not all current APIs used to
> request a channel. With minimal effort dmaengine drivers, platforms and
> dmaengine user drivers can be converted to use the two function.
>
> struct dma_chan *dma_request_chan_by_mask(const dma_cap_mask_t *mask);
>
> To request any channel matching with the requested capabilities, can be
> used to request channel for memcpy, memset, xor, etc where no hardware
> synchronization is needed.
>
> struct dma_chan *dma_request_chan(struct device *dev, const char *name);
> To request a slave channel. The dma_request_chan() will try to find the
> channel via DT, ACPI or in case if the kernel booted in non DT/ACPI mode
> it will use a filter lookup table and retrieves the needed information from
> the dma_filter_map provided by the DMA drivers.
> This legacy mode needs changes in platform code, in dmaengine drivers and
> finally the dmaengine user drivers can be converted:
>
> For each dmaengine driver an array of DMA device, slave and the parameter
> for the filter function needs to be added:
>
> static const struct dma_filter_map da830_edma_map[] = {
>         { "davinci-mcasp.0", "rx", EDMA_FILTER_PARAM(0, 0) },
>         { "davinci-mcasp.0", "tx", EDMA_FILTER_PARAM(0, 1) },
>         { "davinci-mcasp.1", "rx", EDMA_FILTER_PARAM(0, 2) },
>         { "davinci-mcasp.1", "tx", EDMA_FILTER_PARAM(0, 3) },
>         { "davinci-mcasp.2", "rx", EDMA_FILTER_PARAM(0, 4) },
>         { "davinci-mcasp.2", "tx", EDMA_FILTER_PARAM(0, 5) },
>         { "spi_davinci.0", "rx", EDMA_FILTER_PARAM(0, 14) },
>         { "spi_davinci.0", "tx", EDMA_FILTER_PARAM(0, 15) },
>         { "da830-mmc.0", "rx", EDMA_FILTER_PARAM(0, 16) },
>         { "da830-mmc.0", "tx", EDMA_FILTER_PARAM(0, 17) },
>         { "spi_davinci.1", "rx", EDMA_FILTER_PARAM(0, 18) },
>         { "spi_davinci.1", "tx", EDMA_FILTER_PARAM(0, 19) },
> };
>
> This information is going to be needed by the dmaengine driver, so
> modification to the platform_data is needed, and the driver map should be
> added to the pdata of the DMA driver:
>
> da8xx_edma0_pdata.slave_map = da830_edma_map;
> da8xx_edma0_pdata.slavecnt = ARRAY_SIZE(da830_edma_map);
>
> The DMA driver then needs to configure the needed device -> filter_fn
> mapping before it registers with dma_async_device_register() :
>
> if (info->slave_map) {
>         ecc->dma_slave.filter_map.map = info->slave_map;
>         ecc->dma_slave.filter_map.mapcnt = info->slavecnt;
>         ecc->dma_slave.filter_map.filter_fn = edma_filter_fn;
> }
>
> When neither DT or ACPI lookup is available the dma_request_chan() will
> try to match the requester's device name with the filter_map's list of
> device names, when a match found it will use the information from the
> dma_filter_map to get the channel with the dma_get_channel() internal
> function.
>

Few nitpicks below.

> Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
> ---
>  Documentation/dmaengine/client.txt | 23 +++------
>  drivers/dma/dmaengine.c            | 98 ++++++++++++++++++++++++++++++++++++++
>  include/linux/dmaengine.h          | 27 +++++++++++
>  3 files changed, 131 insertions(+), 17 deletions(-)
>
> diff --git a/Documentation/dmaengine/client.txt b/Documentation/dmaengine/client.txt
> index d9f9f461102a..6c72a06eb1a5 100644
> --- a/Documentation/dmaengine/client.txt
> +++ b/Documentation/dmaengine/client.txt
> @@ -22,25 +22,14 @@ The slave DMA usage consists of following steps:
>     Channel allocation is slightly different in the slave DMA context,
>     client drivers typically need a channel from a particular DMA
>     controller only and even in some cases a specific channel is desired.
> -   To request a channel dma_request_channel() API is used.
> +   To request a channel dma_request_chan() API is used.
>
>     Interface:
> -       struct dma_chan *dma_request_channel(dma_cap_mask_t mask,
> -                       dma_filter_fn filter_fn,
> -                       void *filter_param);
> -   where dma_filter_fn is defined as:
> -       typedef bool (*dma_filter_fn)(struct dma_chan *chan, void *filter_param);
> -
> -   The 'filter_fn' parameter is optional, but highly recommended for
> -   slave and cyclic channels as they typically need to obtain a specific
> -   DMA channel.
> -
> -   When the optional 'filter_fn' parameter is NULL, dma_request_channel()
> -   simply returns the first channel that satisfies the capability mask.
> -
> -   Otherwise, the 'filter_fn' routine will be called once for each free
> -   channel which has a capability in 'mask'.  'filter_fn' is expected to
> -   return 'true' when the desired DMA channel is found.
> +       struct dma_chan *dma_request_chan(struct device *dev, const char *name);
> +
> +   Which will find and return the 'name' DMA channel associated with the 'dev'
> +   device. The association is done via DT, ACPI or board file based
> +   dma_map_filter matching table.
>
>     A channel allocated via this interface is exclusive to the caller,
>     until dma_release_channel() is called.
> diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
> index ea9d66982d40..b9dc1512c4aa 100644
> --- a/drivers/dma/dmaengine.c
> +++ b/drivers/dma/dmaengine.c
> @@ -43,6 +43,7 @@
>
>  #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
>
> +#include <linux/platform_device.h>
>  #include <linux/dma-mapping.h>
>  #include <linux/init.h>
>  #include <linux/module.h>
> @@ -712,6 +713,103 @@ struct dma_chan *dma_request_slave_channel(struct device *dev,
>  }
>  EXPORT_SYMBOL_GPL(dma_request_slave_channel);
>
> +const static struct dma_filter_map *dma_filter_match(struct dma_device *device,
> +                                              const char *name,
> +                                              struct device *dev)
> +{
> +       const struct dma_filter_map *map_found = NULL;
> +       int i;
> +
> +       if (!device->filter_map.mapcnt)
> +               return NULL;
> +
> +       for (i = 0; i < device->filter_map.mapcnt; i++) {
> +               const struct dma_filter_map *map = &device->filter_map.map[i];
> +
> +               if (!strcmp(map->devname, dev_name(dev)) &&
> +                   !strcmp(map->slave, name)) {
> +                       map_found = map;
> +                       break;

return map?

> +               }
> +       }
> +

return NULL;

> +       return map_found;
> +}
> +
> +/**
> + * dma_request_chan - try to allocate an exclusive slave channel
> + * @dev:       pointer to client device structure
> + * @name:      slave channel name
> + *
> + * Returns pointer to appropriate DMA channel on success or an error pointer.
> + */
> +struct dma_chan *dma_request_chan(struct device *dev, const char *name)
> +{
> +       struct dma_device *device, *_d;

If you name *d, *_d; it would?

> +       struct dma_chan *chan = NULL;
> +

> +       /* If device-tree is present get slave info from here */
> +       if (dev->of_node)
> +               chan = of_dma_request_slave_channel(dev->of_node, name);
> +
> +       /* If device was enumerated by ACPI get slave info from here */
> +       if (has_acpi_companion(dev) && !chan)
> +               chan = acpi_dma_request_slave_chan_by_name(dev, name);

This part might be a good candidate to be moved to
drivers/base/property.c as
struct dma_chan *device_property_dma_request_chan(?) or alike.


> +
> +       if (chan) {
> +               /* Valid channel found */
> +               if (!IS_ERR(chan))
> +                       return chan;
> +

They might return EPROBE_DEFER. Is any specific handling needed here?

> +               pr_warn("%s: %s DMA request failed, falling back to legacy\n",
> +                       __func__, dev->of_node ? "OF" : "ACPI");
> +       }
> +
> +       /* Try to find the channel via the DMA filter map(s) */
> +       mutex_lock(&dma_list_mutex);
> +       list_for_each_entry_safe(device, _d, &dma_device_list, global_node) {
> +               dma_cap_mask_t mask;
> +               const struct dma_filter_map *map = dma_filter_match(device,
> +                                                                   name, dev);
> +
> +               if (!map)
> +                       continue;
> +
> +               dma_cap_zero(mask);
> +               dma_cap_set(DMA_SLAVE, mask);
> +
> +               chan = find_candidate(device, &mask,
> +                                     device->filter_map.filter_fn, map->param);

?allow to put this on single line I believe.

> +               if (!IS_ERR(chan))
> +                       break;
> +       }
> +       mutex_unlock(&dma_list_mutex);
> +
> +       return chan ? chan : ERR_PTR(-EPROBE_DEFER);
> +}
> +EXPORT_SYMBOL_GPL(dma_request_chan);
> +
> +/**
> + * dma_request_chan_by_mask - allocate a channel satisfying certain capabilities
> + * @mask: capabilities that the channel must satisfy
> + *
> + * Returns pointer to appropriate DMA channel on success or an error pointer.
> + */
> +struct dma_chan *dma_request_chan_by_mask(const dma_cap_mask_t *mask)
> +{
> +       struct dma_chan *chan;
> +
> +       if (!mask)
> +               return ERR_PTR(-ENODEV);
> +
> +       chan = __dma_request_channel(mask, NULL, NULL);
> +       if (!chan)
> +               chan = ERR_PTR(-ENODEV);
> +
> +       return chan;
> +}
> +EXPORT_SYMBOL_GPL(dma_request_chan_by_mask);
> +
>  void dma_release_channel(struct dma_chan *chan)
>  {
>         mutex_lock(&dma_list_mutex);
> diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
> index a2b7c2071cf4..49d48e69c179 100644
> --- a/include/linux/dmaengine.h
> +++ b/include/linux/dmaengine.h
> @@ -606,6 +606,18 @@ enum dmaengine_alignment {
>         DMAENGINE_ALIGN_64_BYTES = 6,
>  };
>
> +struct dma_filter_map {
> +        char *devname;

const ?

> +        char *slave;

Ditto.

> +        void *param;
> +};
> +
> +struct dma_filter {
> +        dma_filter_fn filter_fn;
> +        int mapcnt;
> +        const struct dma_filter_map *map;
> +};


-- 
With Best Regards,
Andy Shevchenko

[RFC v03 00/15] dmaengine: New 'universal' API for requesting channel

[Andy Shevchenko]

On Wed, Dec 2, 2015 at 3:59 PM, Peter Ujfalusi <peter.ujfalusi@ti.com> wrote:
> Hi,
>
> I keep this still as RFC.
>
> Changes since RFC v02:
> - Using has_acpi_companion() instead ACPI_HANDLE()
> - mask matching change within private_candidate()
> - Fallback in dma_request_chan() when DT/ACPI lookup fails.
> - Rename dma_get_channel() -> find_candidate()
> - Arch code changes as suggested by Arnd
> - Some documentation updated, more need to be done.
>
> Changes since RFC v01:
> - dma_request_chan(); lost the mask parameter
> - The new API does not rely on RESOURCE_DMA, instead the dma_filter_map table
>   will be used to provide the needed information to the filter function in
>   legacy mode
> - Extended the example patches to convert most of daVinci to use the new API to
>   request the DMA channels.
>
> TODO: Documentation update ;)
>
> As it has been discussed in the following thread:
> http://www.gossamer-threads.com/lists/linux/kernel/2181487#2181487
>
> Andy: I did looked at the unified device properties, but I decided to not to use
> it as I don't see it to fit well and most of the legacy board files are using
> resources to specify at least their memory regions so adding the DMA resource
> to them would be more inline with the rest of the code.
>
> The ARM, mmc and spi patches are converting daVinci drivers board files to use
> the new method of requesting DMA channel.
>
> With this series I have taken a path which would result two new API, which can
> be used to convert most of the current users already and with some work all
> users might be able to move to this set.
> With this set the filter_fn used for legacy (non DT/ACPI) channel request is no
> longer needed to be exported to client drivers since the selection of the
> correct filter_fn will be done in the core.
>
> So, the first proposal is to have:
>
> struct dma_chan *dma_request_chan(struct device *dev, const char *name);
> struct dma_chan *dma_request_chan_by_mask(const dma_cap_mask_t *mask);
>
> The dma_request_chan_by_mask() is to request any channel matching with the
> requested capabilities, can be used to request channel for memcpy, memset, xor,
> etc where no hardware synchronization is needed.
>
> The dma_request_chan() is to request a slave channel. The dma_request_chan() will try to find the
> channel via DT, ACPI or in case if the kernel booted in non DT/ACPI mode
> it will use a filter lookup table and retrieves the needed information from
> the dma_filter_map provided by the DMA drivers.
> This legacy mode needs changes in platform code, in dmaengine drivers and
> finally the dmaengine user drivers can be converted:
>
> For each dmaengine driver an array of DMA device, slave and the parameter
> for the filter function needs to be added:
>
> static const struct dma_filter_map da830_edma_map[] = {
>         { "davinci-mcasp.0", "rx", EDMA_FILTER_PARAM(0, 0) },
>         { "davinci-mcasp.0", "tx", EDMA_FILTER_PARAM(0, 1) },
>         { "davinci-mcasp.1", "rx", EDMA_FILTER_PARAM(0, 2) },
>         { "davinci-mcasp.1", "tx", EDMA_FILTER_PARAM(0, 3) },
>         { "davinci-mcasp.2", "rx", EDMA_FILTER_PARAM(0, 4) },
>         { "davinci-mcasp.2", "tx", EDMA_FILTER_PARAM(0, 5) },
>         { "spi_davinci.0", "rx", EDMA_FILTER_PARAM(0, 14) },
>         { "spi_davinci.0", "tx", EDMA_FILTER_PARAM(0, 15) },
>         { "da830-mmc.0", "rx", EDMA_FILTER_PARAM(0, 16) },
>         { "da830-mmc.0", "tx", EDMA_FILTER_PARAM(0, 17) },
>         { "spi_davinci.1", "rx", EDMA_FILTER_PARAM(0, 18) },
>         { "spi_davinci.1", "tx", EDMA_FILTER_PARAM(0, 19) },
> };
>
> This information is going to be needed by the dmaengine driver, so
> modification to the platform_data is needed, and the driver map should be
> added to the pdata of the DMA driver:
>
> da8xx_edma0_pdata.slave_map = da830_edma_map;
> da8xx_edma0_pdata.slavecnt = ARRAY_SIZE(da830_edma_map);
>
> The DMA driver then needs to configure the needed device -> filter_fn
> mapping before it registers with dma_async_device_register() :
>
> if (info->slave_map) {
>         ecc->dma_slave.filter_map.map = info->slave_map;
>         ecc->dma_slave.filter_map.mapcnt = info->slavecnt;
>         ecc->dma_slave.filter_map.filter_fn = edma_filter_fn;
> }
>
> When neither DT or ACPI lookup is available the dma_request_chan() will
> try to match the requester's device name with the filter_map's list of
> device names, when a match found it will use the information from the
> dma_filter_map to get the channel with the dma_get_channel() internal
> function.
>

I like patches 1 & 2 in current shape.
Though few comments to patch 3.

> Regards,
> Peter
> ---
> Peter Ujfalusi (15):
>   dmaengine: core: Skip mask matching when it is not provided to
>     private_candidate
>   dmaengine: core: Move and merge the code paths using private_candidate
>   dmaengine: core: Introduce new, universal API to request a channel
>   dmaengine: edma: Add support for DMA filter mapping to slave devices
>   ARM: davinci: devices-da8xx: Add dma_filter_map to edma
>   ARM: davinci: dm355: Add dma_filter_map to edma
>   ARM: davinci: dm365: Add dma_filter_map to edma
>   ARM: davinci: dm644x: Add dma_filter_map to edma
>   ARM: davinci: dm646x: Add dma_filter_map to edma
>   mmc: davinci_mmc: Use dma_request_chan() to requesting DMA channel
>   spi: davinci: Use dma_request_chan() to requesting DMA channel
>   ARM: davinci: devices-da8xx: Remove DMA resources for MMC and SPI
>   ARM: davinci: devices: Remove DMA resources for MMC
>   ARM: davinci: dm355: Remove DMA resources for SPI
>   ARM: davinci: dm365: Remove DMA resources for SPI
>
>  Documentation/dmaengine/client.txt    |  23 ++---
>  arch/arm/mach-davinci/devices-da8xx.c |  95 +++++++++---------
>  arch/arm/mach-davinci/devices.c       |  19 ----
>  arch/arm/mach-davinci/dm355.c         |  28 ++++--
>  arch/arm/mach-davinci/dm365.c         |  30 ++++--
>  arch/arm/mach-davinci/dm644x.c        |  12 +++
>  arch/arm/mach-davinci/dm646x.c        |  11 +++
>  drivers/dma/dmaengine.c               | 181 ++++++++++++++++++++++++++--------
>  drivers/dma/edma.c                    |   6 ++
>  drivers/mmc/host/davinci_mmc.c        |  52 +++-------
>  drivers/spi/spi-davinci.c             |  76 +++++---------
>  include/linux/dmaengine.h             |  27 +++++
>  include/linux/platform_data/edma.h    |   7 ++
>  13 files changed, 336 insertions(+), 231 deletions(-)
>
> --
> 2.6.3
>



-- 
With Best Regards,
Andy Shevchenko

[RFC v03 03/15] dmaengine: core: Introduce new, universal API to request a channel

[Peter Ujfalusi]

On 12/02/2015 04:35 PM, Andy Shevchenko wrote:
>> +const static struct dma_filter_map *dma_filter_match(struct dma_device *device,
>> +                                              const char *name,
>> +                                              struct device *dev)
>> +{
>> +       const struct dma_filter_map *map_found = NULL;
>> +       int i;
>> +
>> +       if (!device->filter_map.mapcnt)
>> +               return NULL;
>> +
>> +       for (i = 0; i < device->filter_map.mapcnt; i++) {
>> +               const struct dma_filter_map *map = &device->filter_map.map[i];
>> +
>> +               if (!strcmp(map->devname, dev_name(dev)) &&
>> +                   !strcmp(map->slave, name)) {
>> +                       map_found = map;
>> +                       break;
> 
> return map?
> 
>> +               }
>> +       }
>> +
> 
> return NULL;

OK.

> 
>> +       return map_found;
>> +}
>> +
>> +/**
>> + * dma_request_chan - try to allocate an exclusive slave channel
>> + * @dev:       pointer to client device structure
>> + * @name:      slave channel name
>> + *
>> + * Returns pointer to appropriate DMA channel on success or an error pointer.
>> + */
>> +struct dma_chan *dma_request_chan(struct device *dev, const char *name)
>> +{
>> +       struct dma_device *device, *_d;
> 
> If you name *d, *_d; it would?
> 
>> +       struct dma_chan *chan = NULL;
>> +
> 
>> +       /* If device-tree is present get slave info from here */
>> +       if (dev->of_node)
>> +               chan = of_dma_request_slave_channel(dev->of_node, name);
>> +
>> +       /* If device was enumerated by ACPI get slave info from here */
>> +       if (has_acpi_companion(dev) && !chan)
>> +               chan = acpi_dma_request_slave_chan_by_name(dev, name);
> 
> This part might be a good candidate to be moved to
> drivers/base/property.c as
> struct dma_chan *device_property_dma_request_chan(?) or alike.

Not sure if it is a good idea. We want users to use the dmaengine API for
requesting DMA channels, but if we just add renamed
dma_request_slave_channel_reason() - essentially the
device_property_dma_request_chan() would be the same - what users will stop to
use some random API to request the channel?
I'm not really sure if something which is returning 'struct dma_chan *' can be
considered as property to anything. The DMA request number can be seen as a
property for a given device, but a dmaengine related pointer?

Actually I was thinking to move the declaration for these from
include/linux/of_dma.h/acpi_dma.h to a header under drivers/dma/

Also move as much local to dmaengine as it is possible to have cleaner
interface towards the client drivers.

> 
>> +
>> +       if (chan) {
>> +               /* Valid channel found */
>> +               if (!IS_ERR(chan))
>> +                       return chan;
>> +
> 
> They might return EPROBE_DEFER. Is any specific handling needed here?

-EPROBE_DEFER means that the DMA driver is not yet loaded and in this case the
lookup for the filter function would also fail, but we have additional and
needless warning printed out. It is better to return with the deferred code also.

> 
>> +               pr_warn("%s: %s DMA request failed, falling back to legacy\n",
>> +                       __func__, dev->of_node ? "OF" : "ACPI");
>> +       }
>> +
>> +       /* Try to find the channel via the DMA filter map(s) */
>> +       mutex_lock(&dma_list_mutex);
>> +       list_for_each_entry_safe(device, _d, &dma_device_list, global_node) {
>> +               dma_cap_mask_t mask;
>> +               const struct dma_filter_map *map = dma_filter_match(device,
>> +                                                                   name, dev);
>> +
>> +               if (!map)
>> +                       continue;
>> +
>> +               dma_cap_zero(mask);
>> +               dma_cap_set(DMA_SLAVE, mask);
>> +
>> +               chan = find_candidate(device, &mask,
>> +                                     device->filter_map.filter_fn, map->param);
> 
> ?allow to put this on single line I believe.

if not in one line, but will look much better.

> 
>> +               if (!IS_ERR(chan))
>> +                       break;
>> +       }
>> +       mutex_unlock(&dma_list_mutex);
>> +
>> +       return chan ? chan : ERR_PTR(-EPROBE_DEFER);
>> +}
>> +EXPORT_SYMBOL_GPL(dma_request_chan);
>> +
>> +/**
>> + * dma_request_chan_by_mask - allocate a channel satisfying certain capabilities
>> + * @mask: capabilities that the channel must satisfy
>> + *
>> + * Returns pointer to appropriate DMA channel on success or an error pointer.
>> + */
>> +struct dma_chan *dma_request_chan_by_mask(const dma_cap_mask_t *mask)
>> +{
>> +       struct dma_chan *chan;
>> +
>> +       if (!mask)
>> +               return ERR_PTR(-ENODEV);
>> +
>> +       chan = __dma_request_channel(mask, NULL, NULL);
>> +       if (!chan)
>> +               chan = ERR_PTR(-ENODEV);
>> +
>> +       return chan;
>> +}
>> +EXPORT_SYMBOL_GPL(dma_request_chan_by_mask);
>> +
>>  void dma_release_channel(struct dma_chan *chan)
>>  {
>>         mutex_lock(&dma_list_mutex);
>> diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
>> index a2b7c2071cf4..49d48e69c179 100644
>> --- a/include/linux/dmaengine.h
>> +++ b/include/linux/dmaengine.h
>> @@ -606,6 +606,18 @@ enum dmaengine_alignment {
>>         DMAENGINE_ALIGN_64_BYTES = 6,
>>  };
>>
>> +struct dma_filter_map {
>> +        char *devname;
> 
> const ?
> 
>> +        char *slave;
> 
> Ditto.

Sure.
I'm also going to change the name of the struct to dma_slave_map

> 
>> +        void *param;
>> +};
>> +
>> +struct dma_filter {
>> +        dma_filter_fn filter_fn;
>> +        int mapcnt;
>> +        const struct dma_filter_map *map;
>> +};
> 
> 


-- 
P?ter