2021-10-20 14:38:32

by Dongliang Mu

[permalink] [raw]
Subject: [PATCH] dmaengine: rcar-dmac: refactor the error handling code of rcar_dmac_probe

In rcar_dmac_probe, if pm_runtime_resume_and_get fails, it forgets to
disable runtime PM. And of_dma_controller_free should only be invoked
after the success of of_dma_controller_register.

Fix this by refactoring the error handling code.

Signed-off-by: Dongliang Mu <[email protected]>
---
drivers/dma/sh/rcar-dmac.c | 13 +++++++------
1 file changed, 7 insertions(+), 6 deletions(-)

diff --git a/drivers/dma/sh/rcar-dmac.c b/drivers/dma/sh/rcar-dmac.c
index 6885b3dcd7a9..5c7716fd6bc5 100644
--- a/drivers/dma/sh/rcar-dmac.c
+++ b/drivers/dma/sh/rcar-dmac.c
@@ -1916,7 +1916,7 @@ static int rcar_dmac_probe(struct platform_device *pdev)
ret = pm_runtime_resume_and_get(&pdev->dev);
if (ret < 0) {
dev_err(&pdev->dev, "runtime PM get sync failed (%d)\n", ret);
- return ret;
+ goto err_pm_disable;
}

ret = rcar_dmac_init(dmac);
@@ -1924,7 +1924,7 @@ static int rcar_dmac_probe(struct platform_device *pdev)

if (ret) {
dev_err(&pdev->dev, "failed to reset device\n");
- goto error;
+ goto err_pm_disable;
}

/* Initialize engine */
@@ -1958,14 +1958,14 @@ static int rcar_dmac_probe(struct platform_device *pdev)
for_each_rcar_dmac_chan(i, dmac, chan) {
ret = rcar_dmac_chan_probe(dmac, chan);
if (ret < 0)
- goto error;
+ goto err_pm_disable;
}

/* Register the DMAC as a DMA provider for DT. */
ret = of_dma_controller_register(pdev->dev.of_node, rcar_dmac_of_xlate,
NULL);
if (ret < 0)
- goto error;
+ goto err_pm_disable;

/*
* Register the DMA engine device.
@@ -1974,12 +1974,13 @@ static int rcar_dmac_probe(struct platform_device *pdev)
*/
ret = dma_async_device_register(engine);
if (ret < 0)
- goto error;
+ goto err_dma_free;

return 0;

-error:
+err_dma_free:
of_dma_controller_free(pdev->dev.of_node);
+err_pm_disable:
pm_runtime_disable(&pdev->dev);
return ret;
}
--
2.25.1


2021-10-21 09:02:13

by Geert Uytterhoeven

[permalink] [raw]
Subject: Re: [PATCH] dmaengine: rcar-dmac: refactor the error handling code of rcar_dmac_probe

Hi Dongliang,

Thanks for your patch!

On Wed, Oct 20, 2021 at 4:36 PM Dongliang Mu <[email protected]> wrote:
> In rcar_dmac_probe, if pm_runtime_resume_and_get fails, it forgets to
> disable runtime PM. And of_dma_controller_free should only be invoked
> after the success of of_dma_controller_register.

The second issue is actually harmless, as of_dma_controller_free()
is a no-op if the DMA controller was never registered.
Of course it doesn't hurt to improve symmetry.

> Fix this by refactoring the error handling code.
>
> Signed-off-by: Dongliang Mu <[email protected]>

Reviewed-by: Geert Uytterhoeven <[email protected]>

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds

2021-10-22 06:38:25

by Laurent Pinchart

[permalink] [raw]
Subject: Re: [PATCH] dmaengine: rcar-dmac: refactor the error handling code of rcar_dmac_probe

Hi Dongliang,

Thank you for the patch.

On Wed, Oct 20, 2021 at 10:35:33PM +0800, Dongliang Mu wrote:
> In rcar_dmac_probe, if pm_runtime_resume_and_get fails, it forgets to
> disable runtime PM. And of_dma_controller_free should only be invoked
> after the success of of_dma_controller_register.
>
> Fix this by refactoring the error handling code.
>
> Signed-off-by: Dongliang Mu <[email protected]>

Reviewed-by: Laurent Pinchart <[email protected]>

> ---
> drivers/dma/sh/rcar-dmac.c | 13 +++++++------
> 1 file changed, 7 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/dma/sh/rcar-dmac.c b/drivers/dma/sh/rcar-dmac.c
> index 6885b3dcd7a9..5c7716fd6bc5 100644
> --- a/drivers/dma/sh/rcar-dmac.c
> +++ b/drivers/dma/sh/rcar-dmac.c
> @@ -1916,7 +1916,7 @@ static int rcar_dmac_probe(struct platform_device *pdev)
> ret = pm_runtime_resume_and_get(&pdev->dev);
> if (ret < 0) {
> dev_err(&pdev->dev, "runtime PM get sync failed (%d)\n", ret);
> - return ret;
> + goto err_pm_disable;
> }
>
> ret = rcar_dmac_init(dmac);
> @@ -1924,7 +1924,7 @@ static int rcar_dmac_probe(struct platform_device *pdev)
>
> if (ret) {
> dev_err(&pdev->dev, "failed to reset device\n");
> - goto error;
> + goto err_pm_disable;
> }
>
> /* Initialize engine */
> @@ -1958,14 +1958,14 @@ static int rcar_dmac_probe(struct platform_device *pdev)
> for_each_rcar_dmac_chan(i, dmac, chan) {
> ret = rcar_dmac_chan_probe(dmac, chan);
> if (ret < 0)
> - goto error;
> + goto err_pm_disable;
> }
>
> /* Register the DMAC as a DMA provider for DT. */
> ret = of_dma_controller_register(pdev->dev.of_node, rcar_dmac_of_xlate,
> NULL);
> if (ret < 0)
> - goto error;
> + goto err_pm_disable;
>
> /*
> * Register the DMA engine device.
> @@ -1974,12 +1974,13 @@ static int rcar_dmac_probe(struct platform_device *pdev)
> */
> ret = dma_async_device_register(engine);
> if (ret < 0)
> - goto error;
> + goto err_dma_free;
>
> return 0;
>
> -error:
> +err_dma_free:
> of_dma_controller_free(pdev->dev.of_node);
> +err_pm_disable:
> pm_runtime_disable(&pdev->dev);
> return ret;
> }

--
Regards,

Laurent Pinchart

2021-10-25 04:26:41

by Vinod Koul

[permalink] [raw]
Subject: Re: [PATCH] dmaengine: rcar-dmac: refactor the error handling code of rcar_dmac_probe

On 20-10-21, 22:35, Dongliang Mu wrote:
> In rcar_dmac_probe, if pm_runtime_resume_and_get fails, it forgets to
> disable runtime PM. And of_dma_controller_free should only be invoked
> after the success of of_dma_controller_register.
>
> Fix this by refactoring the error handling code.

Applied, thanks

--
~Vinod