[PATCH] media: rkisp1: rksip1-capture.c: Improve comments and fix typos

Dafna Hirschfeld dafna.hirschfeld at collabora.com
Tue Apr 20 10:24:07 BST 2021


Hi,
Thanks for the fixes

On 18.04.21 12:34, Sebastian Fricke wrote:
> Improve the wording of the function description to increase readability.
> 
> Fix three typos:
> s/during processing a frame/while processing a frame/
> s/it also update/it also updates/
> s/there's not buf in shadow/there's no buffer in a shadow register/
> 
> Replace the abbreviation 'buf' with the full word buffer, the
> abbreviation 'config' with the verb configure, and 'regs' with registers.
> The goal of this change is to ease the reading flow of the comment.
> 
> Signed-off-by: Sebastian Fricke <sebastian.fricke at posteo.net>
> ---
> Side-note:
> I also believe there should be a description for the abbreviation FE, as
> it is not described anywhere. I think it means frame end, right?.

yes, this is Frame End. You can send a patch replacing "FE" with frame end if you like.

Reviewed-by: Dafna Hirschfeld <dafna.hirschfeld at collabora.com>

> ---
>   .../platform/rockchip/rkisp1/rkisp1-capture.c    | 16 ++++++++--------
>   1 file changed, 8 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/media/platform/rockchip/rkisp1/rkisp1-capture.c b/drivers/media/platform/rockchip/rkisp1/rkisp1-capture.c
> index 5f6c9d1623e4..9643bdd05b7b 100644
> --- a/drivers/media/platform/rockchip/rkisp1/rkisp1-capture.c
> +++ b/drivers/media/platform/rockchip/rkisp1/rkisp1-capture.c
> @@ -830,8 +830,8 @@ static void rkisp1_return_all_buffers(struct rkisp1_capture *cap,
>   }
>   
>   /*
> - * Most of registers inside rockchip ISP1 have shadow register since
> - * they must be not be changed during processing a frame.
> + * Most registers inside the rockchip ISP1 have shadow register since
> + * they must not be changed while processing a frame.
>    * Usually, each sub-module updates its shadow register after
>    * processing the last pixel of a frame.
>    */
> @@ -847,14 +847,14 @@ static void rkisp1_cap_stream_enable(struct rkisp1_capture *cap)
>   	spin_lock_irq(&cap->buf.lock);
>   	rkisp1_set_next_buf(cap);
>   	cap->ops->enable(cap);
> -	/* It's safe to config ACTIVE and SHADOW regs for the
> +	/* It's safe to configure ACTIVE and SHADOW registers for the
>   	 * first stream. While when the second is starting, do NOT
> -	 * force update because it also update the first one.
> +	 * force update because it also updates the first one.
>   	 *
> -	 * The latter case would drop one more buf(that is 2) since
> -	 * there's not buf in shadow when the second FE received. This's
> -	 * also required because the second FE maybe corrupt especially
> -	 * when run at 120fps.
> +	 * The latter case would drop one more buffer(that is 2) since
> +	 * there's no buffer in a shadow register when the second FE received.
> +	 * This's also required because the second FE maybe corrupt
> +	 * especially when run at 120fps.
>   	 */
>   	if (!other->is_streaming) {
>   		/* force cfg update */
> 



More information about the linux-arm-kernel mailing list