Re: [PATCH 01/11] drm/vkms: rename VKMS_LUT_SIZE to VKMS_GAMMA_LUT_SIZE

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] Incorrect size calculation for gamma_lut validation causes atomic commits to be rejected on 32-bit systems.
--

--- Patch [1]: [PATCH 01/11] drm/vkms: rename VKMS_LUT_SIZE to VKMS_GAMMA_LUT_SIZE ---
commit 2362b93c64416bccb5039e7ee169799f46cd254c
Author: Leandro Ribeiro <[email protected]>

drm/vkms: rename VKMS_LUT_SIZE to VKMS_GAMMA_LUT_SIZE

This commit renames VKMS_LUT_SIZE to VKMS_GAMMA_LUT_SIZE to clarify that it
represents the size of the CRTC gamma LUT.

> diff --git a/drivers/gpu/drm/vkms/vkms_drv.c b/drivers/gpu/drm/vkms/vkms_drv.c
> index 5a640b531d88f..86a8a95694fd7 100644
> --- a/drivers/gpu/drm/vkms/vkms_drv.c
> +++ b/drivers/gpu/drm/vkms/vkms_drv.c
> @@ -113,7 +113,7 @@ static int vkms_atomic_check(struct drm_device *dev, struct drm_atomic_commit *s
>  			continue;
>  
>  		if (new_crtc_state->gamma_lut->length / sizeof(struct drm_color_lut *)
> -		    > VKMS_LUT_SIZE)
> +		    > VKMS_GAMMA_LUT_SIZE)

[Severity: High]
This is a pre-existing issue, but does this calculation incorrectly reject
atomic commits with valid gamma LUTs on 32-bit systems?

Looking at vkms_atomic_check() in drivers/gpu/drm/vkms/vkms_drv.c, it
divides the length by the size of a pointer (sizeof(struct drm_color_lut *))
rather than the actual struct size (sizeof(struct drm_color_lut)).

On 32-bit architectures, the pointer size is 4 bytes, while the struct itself
is 8 bytes. When userspace provides a valid 256-element LUT (2048 bytes), this
division calculates the length as 512 elements. This would exceed the 
VKMS_GAMMA_LUT_SIZE of 256, causing the atomic check to incorrectly fail 
with -EINVAL.

>  			return -EINVAL;
>  	}
>

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1