[PATCH 1/2] dumb_buffer: Remove invalid-bpp subtest

[Kamil Konieczny]

Hi Rob,
On 2024-08-14 at 11:54:25 -0700, Rob Clark wrote:
> From: Rob Clark <robdclark at chromium.org>
> 
> These allocations are not necessarily invalid.

Please write a little more why. Also look at gitlab log,
you need to delete comments with this subtest name and description
as our CI is creating testlist at build time and that info
is there.

Add also tests/ prefix in subject:

[PATCH 1/2] tests/dumb_buffer: Remove invalid-bpp subtest

> 
> Signed-off-by: Rob Clark <robdclark at chromium.org>
> ---
>  tests/dumb_buffer.c | 22 ----------------------
>  1 file changed, 22 deletions(-)
> 
> diff --git a/tests/dumb_buffer.c b/tests/dumb_buffer.c
> index 1d4bf296c53e..579d2f906b6d 100644
> --- a/tests/dumb_buffer.c
> +++ b/tests/dumb_buffer.c
> @@ -136,25 +136,6 @@ static void dumb_destroy(int fd, uint32_t handle)
>  	igt_assert_eq(__dumb_destroy(fd, handle), 0);
>  }
>  
> -static void invalid_dimensions_test(int fd)
> -{
> -	struct drm_mode_create_dumb create;
> -
> -	memset(&create, 0, sizeof(create));
> -	create.width = 4032;
> -	create.height = 2016;
> -	create.bpp = 24;
> -	igt_assert_eq(__dumb_create(fd, &create), -EINVAL);

These use 3-Bytes bpp with divisible by 4 screen W x H, so here
you could be right with removing this.

> -
> -	create.bpp = 32;
> -	create.width = 0;
> -	igt_assert_eq(__dumb_create(fd, &create), -EINVAL);

Why 0-width is valid?

> -
> -	create.width = 4032;
> -	create.height = 0;
> -	igt_assert_eq(__dumb_create(fd, &create), -EINVAL);

Same here, imho looks invalid. Btw what about other settings?
Like bpp not divisible by 8 bits?

Regards,
Kamil

> -}
> -
>  static void valid_dumb_creation_test(int fd)
>  {
>  	struct drm_mode_create_dumb create = {
> @@ -397,9 +378,6 @@ igt_main
>  		fd = drm_open_driver(DRIVER_ANY);
>  	}
>  
> -	igt_subtest("invalid-bpp")
> -		invalid_dimensions_test(fd);
> -
>  	igt_subtest("create-valid-dumb")
>  		valid_dumb_creation_test(fd);
>  
> -- 
> 2.46.0
>