diff mbox

[1/3] dmaengine: of: Allow #dma-cells to be zero

Message ID 1443193000-457-2-git-send-email-jonathanh@nvidia.com
State New, archived
Headers show

Commit Message

Jon Hunter Sept. 25, 2015, 2:56 p.m. UTC
Currently, the #dma-cells property must be 1 or more. However, for some
DMA controllers, where DMA clients may use any DMA channel and it is not
necessary to specify any other DMA information in the device-tree DMA
binding, setting #dma-cells to 0 is desirable. The Tegra210 ADMA
controller is an example of a DMA controller where neither the DMA
channel or the hardware request signal need to be encoded in the
device-tree binding.

By allowing DMA controllers to set #dma-cells to 0, means that the
"dma-names" property for these controllers is not needed. Therefore,
update the of_dma_request_slave_channel() and of_dma_match_channel()
functions to ignore this property if no name is provided.

Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
---
 Documentation/devicetree/bindings/dma/dma.txt | 12 ++++++++----
 drivers/dma/of-dma.c                          | 23 +++++++++++++++--------
 2 files changed, 23 insertions(+), 12 deletions(-)

Comments

Arnd Bergmann Sept. 25, 2015, 3:05 p.m. UTC | #1
On Friday 25 September 2015 15:56:38 Jon Hunter wrote:
> Currently, the #dma-cells property must be 1 or more. However, for some
> DMA controllers, where DMA clients may use any DMA channel and it is not
> necessary to specify any other DMA information in the device-tree DMA
> binding, setting #dma-cells to 0 is desirable. The Tegra210 ADMA
> controller is an example of a DMA controller where neither the DMA
> channel or the hardware request signal need to be encoded in the
> device-tree binding.
> 
> By allowing DMA controllers to set #dma-cells to 0, means that the
> "dma-names" property for these controllers is not needed. Therefore,
> update the of_dma_request_slave_channel() and of_dma_match_channel()
> functions to ignore this property if no name is provided.
> 
> Signed-off-by: Jon Hunter <jonathanh@nvidia.com>

Sorry, but this makes no sense. #dma-cells=0 would imply that there
is only one slave per DMA controller, but that is not the case here.
As commented in patch 3, you actually support multiple slaves, you
just use the wrong interface.

	Arnd

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/Documentation/devicetree/bindings/dma/dma.txt b/Documentation/devicetree/bindings/dma/dma.txt
index 6312fb00ce8d..4d73916fc879 100644
--- a/Documentation/devicetree/bindings/dma/dma.txt
+++ b/Documentation/devicetree/bindings/dma/dma.txt
@@ -8,9 +8,12 @@  controller.
 * DMA controller
 
 Required property:
-- #dma-cells: 		Must be at least 1. Used to provide DMA controller
-			specific information. See DMA client binding below for
-			more details.
+- #dma-cells: 		Used to provide DMA controller specific information.
+			Can be 0 for DMA controllers where clients may use
+			any channel and any DMA specific information, such
+			as a hardware request signal, does not need to be
+			encoded in the binding. See DMA client binding below
+			for more details.
 
 Optional properties:
 - dma-channels: 	Number of DMA channels supported by the controller.
@@ -79,7 +82,8 @@  Required property:
 			are defined in the binding of the DMA client device.
 			Multiple DMA specifiers can be used to represent
 			alternatives and in this case the dma-names for those
-			DMA specifiers must be identical (see examples).
+			DMA specifiers must be identical (see examples). This
+			is optional for DMA controllers where #dma-cells is 0.
 
 Examples:
 
diff --git a/drivers/dma/of-dma.c b/drivers/dma/of-dma.c
index 1e1f2986eba8..ef3d260610ea 100644
--- a/drivers/dma/of-dma.c
+++ b/drivers/dma/of-dma.c
@@ -214,11 +214,13 @@  static int of_dma_match_channel(struct device_node *np, const char *name,
 {
 	const char *s;
 
-	if (of_property_read_string_index(np, "dma-names", index, &s))
-		return -ENODEV;
+	if (name) {
+		if (of_property_read_string_index(np, "dma-names", index, &s))
+			return -ENODEV;
 
-	if (strcmp(name, s))
-		return -ENODEV;
+		if (strcmp(name, s))
+			return -ENODEV;
+	}
 
 	if (of_parse_phandle_with_args(np, "dmas", "#dma-cells", index,
 				       dma_spec))
@@ -230,7 +232,7 @@  static int of_dma_match_channel(struct device_node *np, const char *name,
 /**
  * of_dma_request_slave_channel - Get the DMA slave channel
  * @np:		device node to get DMA request from
- * @name:	name of desired channel
+ * @name:	name of desired channel (optional)
  *
  * Returns pointer to appropriate DMA channel on success or an error pointer.
  */
@@ -243,8 +245,8 @@  struct dma_chan *of_dma_request_slave_channel(struct device_node *np,
 	int			count, i;
 	int			ret_no_channel = -ENODEV;
 
-	if (!np || !name) {
-		pr_err("%s: not enough information provided\n", __func__);
+	if (!np) {
+		pr_err("%s: device node is not valid\n", __func__);
 		return ERR_PTR(-ENODEV);
 	}
 
@@ -252,7 +254,12 @@  struct dma_chan *of_dma_request_slave_channel(struct device_node *np,
 	if (!of_find_property(np, "dmas", NULL))
 		return ERR_PTR(-ENODEV);
 
-	count = of_property_count_strings(np, "dma-names");
+	/*
+	 * If a name is not provided, then assume that there is one
+	 * DMA specifier in the list for the client and so set the
+	 * count to 1 and see if this matches.
+	 */
+	count = name ? of_property_count_strings(np, "dma-names") : 1;
 	if (count < 0) {
 		pr_err("%s: dma-names property of node '%s' missing or empty\n",
 			__func__, np->full_name);