[PATCH] dmaengine: axi-dmac: add reset deassert logic

[PATCH] dmaengine: axi-dmac: add reset deassert logic

[Sergey Matsievskiy]

From: Sergey Matsievskiy <matsievskiysv@gmail.com>

Add optional shared reset deassert logic, sometimes required for
communication with the device.

Signed-off-by: Sergey Matsievskiy <matsievskiysv@gmail.com>
---
 drivers/dma/dma-axi-dmac.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/dma/dma-axi-dmac.c b/drivers/dma/dma-axi-dmac.c
index d47ff27e1408..98aaccd26a26 100644
--- a/drivers/dma/dma-axi-dmac.c
+++ b/drivers/dma/dma-axi-dmac.c
@@ -24,6 +24,7 @@
 #include <linux/of_address.h>
 #include <linux/platform_device.h>
 #include <linux/regmap.h>
+#include <linux/reset.h>
 #include <linux/slab.h>
 
 #include <dt-bindings/dma/axi-dmac.h>
@@ -1211,10 +1212,16 @@ static int axi_dmac_probe(struct platform_device *pdev)
 	struct axi_dmac *dmac;
 	struct clk *clk;
 	struct regmap *regmap;
+	struct reset_control *rstc;
 	unsigned int version;
 	u32 irq_mask = 0;
 	int ret;
 
+	rstc = devm_reset_control_get_optional_shared_deasserted(&pdev->dev, NULL);
+	if (IS_ERR(rstc))
+		return dev_err_probe(&pdev->dev, PTR_ERR(rstc),
+				     "failed to deassert reset\n");
+
 	dmac = devm_kzalloc(&pdev->dev, sizeof(*dmac), GFP_KERNEL);
 	if (!dmac)
 		return -ENOMEM;
-- 
2.47.3

Re: [PATCH] dmaengine: axi-dmac: add reset deassert logic

[Frank Li]

On Thu, Nov 05, 2026 at 04:40:40PM +0300, Sergey Matsievskiy wrote:
> [You don't often get email from matsievskiysv@gmail.com. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]
>
> From: Sergey Matsievskiy <matsievskiysv@gmail.com>
>
> Add optional shared reset deassert logic, sometimes required for
> communication with the device.

" sometimes required for ..." is it correct? most likely some SoC require
a addtional reset for this dma controller.

Frank

Re: [PATCH] dmaengine: axi-dmac: add reset deassert logic

[Sergey Matsievskiy]

> " sometimes required for ..." is it correct? most likely some SoC require
> a addtional reset for this dma controller.

I'd say, when this controller is used in FPGA, it could be used without the
external reset. In SoC, it almost certainly would be behind the reset
controller.