[PATCH v1] media:Fix repeated initialization

[PATCH v1] media:Fix repeated initialization

[Wang Ming]

If the first report returns NULL,
there is no need to execute the
fimc_capture_try_format() function
again.

Signed-off-by: Wang Ming <machel@vivo.com>
---
 .../media/platform/samsung/exynos4-is/fimc-capture.c   | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/media/platform/samsung/exynos4-is/fimc-capture.c b/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
index a0d43bf892e6..5ce2f04afa1e 100644
--- a/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
+++ b/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
@@ -825,10 +825,12 @@ static int fimc_pipeline_try_format(struct fimc_ctx *ctx,
 		tfmt->height = mf->height;
 		ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
 					NULL, &fcc, FIMC_SD_PAD_SINK_CAM);
-		ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
-					NULL, &fcc, FIMC_SD_PAD_SOURCE);
-		if (ffmt && ffmt->mbus_code)
-			mf->code = ffmt->mbus_code;
+		if (ffmt) {
+			ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
+				NULL, &fcc, FIMC_SD_PAD_SOURCE);
+			if (ffmt && ffmt->mbus_code)
+				mf->code = ffmt->mbus_code;
+		}
 		if (mf->width != tfmt->width || mf->height != tfmt->height)
 			continue;
 		tfmt->code = mf->code;
-- 
2.25.1

Re: [PATCH v1] media:Fix repeated initialization

[Krzysztof Kozlowski]

On 05/07/2023 16:21, Wang Ming wrote:
> If the first report returns NULL,
> there is no need to execute the
> fimc_capture_try_format() function
> again.

Few nitpicks only, as I don't know the code unfortunately:

Please wrap commit message according to Linux coding style / submission
process (neither too early nor over the limit):
https://elixir.bootlin.com/linux/v6.4-rc1/source/Documentation/process/submitting-patches.rst#L597

Please use subject prefixes matching the subsystem. You can get them for
example with `git log --oneline -- DIRECTORY_OR_FILE` on the directory
your patch is touching.

> 
> Signed-off-by: Wang Ming <machel@vivo.com>
> ---
>  .../media/platform/samsung/exynos4-is/fimc-capture.c   | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/media/platform/samsung/exynos4-is/fimc-capture.c b/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
> index a0d43bf892e6..5ce2f04afa1e 100644
> --- a/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
> +++ b/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
> @@ -825,10 +825,12 @@ static int fimc_pipeline_try_format(struct fimc_ctx *ctx,
>  		tfmt->height = mf->height;
>  		ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
>  					NULL, &fcc, FIMC_SD_PAD_SINK_CAM);
> -		ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
> -					NULL, &fcc, FIMC_SD_PAD_SOURCE);
> -		if (ffmt && ffmt->mbus_code)
> -			mf->code = ffmt->mbus_code;
> +		if (ffmt) {
> +			ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
> +				NULL, &fcc, FIMC_SD_PAD_SOURCE);

This does not look aligned.



Best regards,
Krzysztof

回复: [PATCH v1] media:Fix repeated initialization

[王明-软件底层技术部]

Hi Krzysztof Kozlowski.
Look closely, there seems to be a misalignment problem here. In addition, what modifications do you need me to make, I will submit them repeatedly according to your requirements.
: )
-----邮件原件-----
发件人: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> 
发送时间: 2023年7月5日 23:10
收件人: 王明-软件底层技术部 <machel@vivo.com>; Sylwester Nawrocki <s.nawrocki@samsung.com>; Mauro Carvalho Chehab <mchehab@kernel.org>; Alim Akhtar <alim.akhtar@samsung.com>; linux-media@vger.kernel.org; linux-arm-kernel@lists.infradead.org; linux-samsung-soc@vger.kernel.org; linux-kernel@vger.kernel.org
抄送: opensource.kernel <opensource.kernel@vivo.com>
主题: Re: [PATCH v1] media:Fix repeated initialization

On 05/07/2023 16:21, Wang Ming wrote:
> If the first report returns NULL,
> there is no need to execute the
> fimc_capture_try_format() function
> again.

Few nitpicks only, as I don't know the code unfortunately:

Please wrap commit message according to Linux coding style / submission process (neither too early nor over the limit):
https://elixir.bootlin.com/linux/v6.4-rc1/source/Documentation/process/submitting-patches.rst#L597

Please use subject prefixes matching the subsystem. You can get them for example with `git log --oneline -- DIRECTORY_OR_FILE` on the directory your patch is touching.

> 
> Signed-off-by: Wang Ming <machel@vivo.com>
> ---
>  .../media/platform/samsung/exynos4-is/fimc-capture.c   | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/media/platform/samsung/exynos4-is/fimc-capture.c 
> b/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
> index a0d43bf892e6..5ce2f04afa1e 100644
> --- a/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
> +++ b/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
> @@ -825,10 +825,12 @@ static int fimc_pipeline_try_format(struct fimc_ctx *ctx,
>  		tfmt->height = mf->height;
>  		ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
>  					NULL, &fcc, FIMC_SD_PAD_SINK_CAM);
> -		ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
> -					NULL, &fcc, FIMC_SD_PAD_SOURCE);
> -		if (ffmt && ffmt->mbus_code)
> -			mf->code = ffmt->mbus_code;
> +		if (ffmt) {
> +			ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
> +				NULL, &fcc, FIMC_SD_PAD_SOURCE);

This does not look aligned.



Best regards,
Krzysztof

Re: [PATCH v1] media:Fix repeated initialization

[Hans Verkuil]

On 05/07/2023 16:21, Wang Ming wrote:
> If the first report returns NULL,
> there is no need to execute the
> fimc_capture_try_format() function
> again.
> 
> Signed-off-by: Wang Ming <machel@vivo.com>
> ---
>  .../media/platform/samsung/exynos4-is/fimc-capture.c   | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/media/platform/samsung/exynos4-is/fimc-capture.c b/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
> index a0d43bf892e6..5ce2f04afa1e 100644
> --- a/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
> +++ b/drivers/media/platform/samsung/exynos4-is/fimc-capture.c
> @@ -825,10 +825,12 @@ static int fimc_pipeline_try_format(struct fimc_ctx *ctx,
>  		tfmt->height = mf->height;
>  		ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
>  					NULL, &fcc, FIMC_SD_PAD_SINK_CAM);
> -		ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
> -					NULL, &fcc, FIMC_SD_PAD_SOURCE);
> -		if (ffmt && ffmt->mbus_code)
> -			mf->code = ffmt->mbus_code;
> +		if (ffmt) {
> +			ffmt = fimc_capture_try_format(ctx, &tfmt->width, &tfmt->height,
> +				NULL, &fcc, FIMC_SD_PAD_SOURCE);
> +			if (ffmt && ffmt->mbus_code)
> +				mf->code = ffmt->mbus_code;
> +		}

I'm not so sure this is a bug at all. I think (not quite certain) that the
first fimc_capture_try_format() call constrains the width and height to
whatever is the SINK_CAM pad, then further constrains it to that of the
SOURCE pad.

But really, someone from Samsung should take a look as well. Adding Marek and
Sylwester.

Regards,

	Hans

>  		if (mf->width != tfmt->width || mf->height != tfmt->height)
>  			continue;
>  		tfmt->code = mf->code;