[PATCH v2] media: rppx1: describe the MAIN_POST white balance gains block

[PATCH v2] media: rppx1: describe the MAIN_POST white balance gains block

[Linmao Li]

rppx1_ext_params_blocks_info[] is indexed by block type and is built
with designated initialisers, so a block type without an entry is left
as a zero-sized hole. RPPX1_PARAMS_BLOCK_TYPE_AWBG_POST is the only
such hole: a block reporting that type and a size of zero passes
v4l2_isp_params_validate_buffer(), which then has nothing to advance
the walk with.

Describe the block, so the array covers all block types the uAPI
defines. The MAIN_POST white balance gains module is probed and started
with the rest of the POST pipe, and RPPX1_PARAMS_MAX_SIZE already
reserves room for all three white balance gains blocks.

The block is not dispatched by rppx1_params(), in line with the other
described blocks that have no users yet.

Signed-off-by: Linmao Li <lilinmao@kylinos.cn>
---
Changes since v1 [1]:
- dropped the rppx1_params() dispatch hunk, the block stays undispatched
  like the other described blocks that have no users yet (Niklas);
- retitled, and the commit message now leads with the hole in the array;
- dropped the Fixes: tag, the fix is expected to be collected together
  with the driver for v7.3 (Jacopo).

The second patch of v1, "media: rppx1: bls: read the raw pattern from
the PRE2 acquisition module", is dropped for now and will be resent when
PRE2 support lands.

The zero-sized block problem in the common validator is addressed by
David Carlier's "media: v4l2-isp: reject zero-sized parameter blocks"
[2], so it is not touched here.

[1] https://lore.kernel.org/linux-media/20260817104529.3132552-1-lilinmao@kylinos.cn/
[2] https://lore.kernel.org/linux-media/20260815193839.141406-1-devnexen@gmail.com/

 drivers/media/platform/dreamchip/rppx1/rpp_params.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/media/platform/dreamchip/rppx1/rpp_params.c b/drivers/media/platform/dreamchip/rppx1/rpp_params.c
index a75a27a8afd09..5e6727d589464 100644
--- a/drivers/media/platform/dreamchip/rppx1/rpp_params.c
+++ b/drivers/media/platform/dreamchip/rppx1/rpp_params.c
@@ -25,6 +25,7 @@ rppx1_ext_params_blocks_info[] = {
 	RPPX1_PARAMS_BLOCK_INFO(LSC_PRE2, lsc),
 	RPPX1_PARAMS_BLOCK_INFO(AWBG_PRE1, awbg),
 	RPPX1_PARAMS_BLOCK_INFO(AWBG_PRE2, awbg),
+	RPPX1_PARAMS_BLOCK_INFO(AWBG_POST, awbg),
 	RPPX1_PARAMS_BLOCK_INFO(CCOR_POST, ccor),
 	RPPX1_PARAMS_BLOCK_INFO(HIST_PRE1, hist),
 	RPPX1_PARAMS_BLOCK_INFO(HIST_PRE2, hist),
-- 
2.25.1

Re: [PATCH v2] media: rppx1: describe the MAIN_POST white balance gains block

[Niklas Söderlund]

Hi Linmao,

Thanks for your work.

On 2026-08-18 10:09:33 +0800, Linmao Li wrote:
> rppx1_ext_params_blocks_info[] is indexed by block type and is built
> with designated initialisers, so a block type without an entry is left
> as a zero-sized hole. RPPX1_PARAMS_BLOCK_TYPE_AWBG_POST is the only
> such hole: a block reporting that type and a size of zero passes
> v4l2_isp_params_validate_buffer(), which then has nothing to advance
> the walk with.
> 
> Describe the block, so the array covers all block types the uAPI
> defines. The MAIN_POST white balance gains module is probed and started
> with the rest of the POST pipe, and RPPX1_PARAMS_MAX_SIZE already
> reserves room for all three white balance gains blocks.
> 
> The block is not dispatched by rppx1_params(), in line with the other
> described blocks that have no users yet.
> 
> Signed-off-by: Linmao Li <lilinmao@kylinos.cn>

Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>

> ---
> Changes since v1 [1]:
> - dropped the rppx1_params() dispatch hunk, the block stays undispatched
>   like the other described blocks that have no users yet (Niklas);
> - retitled, and the commit message now leads with the hole in the array;
> - dropped the Fixes: tag, the fix is expected to be collected together
>   with the driver for v7.3 (Jacopo).
> 
> The second patch of v1, "media: rppx1: bls: read the raw pattern from
> the PRE2 acquisition module", is dropped for now and will be resent when
> PRE2 support lands.
> 
> The zero-sized block problem in the common validator is addressed by
> David Carlier's "media: v4l2-isp: reject zero-sized parameter blocks"
> [2], so it is not touched here.
> 
> [1] https://lore.kernel.org/linux-media/20260817104529.3132552-1-lilinmao@kylinos.cn/
> [2] https://lore.kernel.org/linux-media/20260815193839.141406-1-devnexen@gmail.com/
> 
>  drivers/media/platform/dreamchip/rppx1/rpp_params.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/media/platform/dreamchip/rppx1/rpp_params.c b/drivers/media/platform/dreamchip/rppx1/rpp_params.c
> index a75a27a8afd09..5e6727d589464 100644
> --- a/drivers/media/platform/dreamchip/rppx1/rpp_params.c
> +++ b/drivers/media/platform/dreamchip/rppx1/rpp_params.c
> @@ -25,6 +25,7 @@ rppx1_ext_params_blocks_info[] = {
>  	RPPX1_PARAMS_BLOCK_INFO(LSC_PRE2, lsc),
>  	RPPX1_PARAMS_BLOCK_INFO(AWBG_PRE1, awbg),
>  	RPPX1_PARAMS_BLOCK_INFO(AWBG_PRE2, awbg),
> +	RPPX1_PARAMS_BLOCK_INFO(AWBG_POST, awbg),
>  	RPPX1_PARAMS_BLOCK_INFO(CCOR_POST, ccor),
>  	RPPX1_PARAMS_BLOCK_INFO(HIST_PRE1, hist),
>  	RPPX1_PARAMS_BLOCK_INFO(HIST_PRE2, hist),
> -- 
> 2.25.1
> 

-- 
Kind Regards,
Niklas Söderlund

Re: [PATCH v2] media: rppx1: describe the MAIN_POST white balance gains block

[Jacopo Mondi]

Hello Linmao

On Tue, Aug 18, 2026 at 10:09:33AM +0800, Linmao Li wrote:
> rppx1_ext_params_blocks_info[] is indexed by block type and is built
> with designated initialisers, so a block type without an entry is left
> as a zero-sized hole. RPPX1_PARAMS_BLOCK_TYPE_AWBG_POST is the only
> such hole: a block reporting that type and a size of zero passes
> v4l2_isp_params_validate_buffer(), which then has nothing to advance
> the walk with.
>
> Describe the block, so the array covers all block types the uAPI
> defines. The MAIN_POST white balance gains module is probed and started
> with the rest of the POST pipe, and RPPX1_PARAMS_MAX_SIZE already
> reserves room for all three white balance gains blocks.
>
> The block is not dispatched by rppx1_params(), in line with the other
> described blocks that have no users yet.
>
> Signed-off-by: Linmao Li <lilinmao@kylinos.cn>

Thanks

Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>

> ---
> Changes since v1 [1]:
> - dropped the rppx1_params() dispatch hunk, the block stays undispatched
>   like the other described blocks that have no users yet (Niklas);
> - retitled, and the commit message now leads with the hole in the array;
> - dropped the Fixes: tag, the fix is expected to be collected together
>   with the driver for v7.3 (Jacopo).

I'll sync with Sakari/Hans to see if this can collected as soon as
-rc1 is out

>
> The second patch of v1, "media: rppx1: bls: read the raw pattern from
> the PRE2 acquisition module", is dropped for now and will be resent when
> PRE2 support lands.

If you wish to resend, we could collect that one as well!

Thanks
  j

>
> The zero-sized block problem in the common validator is addressed by
> David Carlier's "media: v4l2-isp: reject zero-sized parameter blocks"
> [2], so it is not touched here.
>
> [1] https://lore.kernel.org/linux-media/20260817104529.3132552-1-lilinmao@kylinos.cn/
> [2] https://lore.kernel.org/linux-media/20260815193839.141406-1-devnexen@gmail.com/
>
>  drivers/media/platform/dreamchip/rppx1/rpp_params.c | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/drivers/media/platform/dreamchip/rppx1/rpp_params.c b/drivers/media/platform/dreamchip/rppx1/rpp_params.c
> index a75a27a8afd09..5e6727d589464 100644
> --- a/drivers/media/platform/dreamchip/rppx1/rpp_params.c
> +++ b/drivers/media/platform/dreamchip/rppx1/rpp_params.c
> @@ -25,6 +25,7 @@ rppx1_ext_params_blocks_info[] = {
>  	RPPX1_PARAMS_BLOCK_INFO(LSC_PRE2, lsc),
>  	RPPX1_PARAMS_BLOCK_INFO(AWBG_PRE1, awbg),
>  	RPPX1_PARAMS_BLOCK_INFO(AWBG_PRE2, awbg),
> +	RPPX1_PARAMS_BLOCK_INFO(AWBG_POST, awbg),
>  	RPPX1_PARAMS_BLOCK_INFO(CCOR_POST, ccor),
>  	RPPX1_PARAMS_BLOCK_INFO(HIST_PRE1, hist),
>  	RPPX1_PARAMS_BLOCK_INFO(HIST_PRE2, hist),
> --
> 2.25.1
>
>