[Intel-gfx] [igt-dev] [PATCH i-g-t 4/6] lib: Don't call igt_require_fb_modifiers() when no modifier

[Ville Syrjälä]

On Wed, Oct 10, 2018 at 05:21:02PM -0700, Deepak Rawat wrote:
> vmwgfx doesn't support fb modifier so skip igt_require_fb_modifiers()
> when modifier are not passed.
> 
> Signed-off-by: Deepak Rawat <drawat at vmware.com>
> ---
>  lib/ioctl_wrappers.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/lib/ioctl_wrappers.c b/lib/ioctl_wrappers.c
> index 0929c43f..3a11cb6e 100644
> --- a/lib/ioctl_wrappers.c
> +++ b/lib/ioctl_wrappers.c
> @@ -1678,7 +1678,10 @@ int __kms_addfb(int fd, uint32_t handle,
>  	struct drm_mode_fb_cmd2 f;
>  	int ret, i;
>  
> -	igt_require_fb_modifiers(fd);
> +	if (modifier)
> +		igt_require_fb_modifiers(fd);
> +	else
> +		flags &= ~DRM_MODE_FB_MODIFIERS;

This would theoretically change the behviour for i915 at least. Without
the modifiers flag the kernel will pick the modifier for us based on 
the bo tiling, which in theory might not be what we wanted. But at least
igt_fb should be fine with that.

Maybe it would be better to just not pass the flags from the caller at
all, and instead have __kms_addfb() check if the driver has modifiers
or not and set the flag based on that?

>  
>  	memset(&f, 0, sizeof(f));
>  
> -- 
> 2.17.1
> 
> _______________________________________________
> igt-dev mailing list
> igt-dev at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/igt-dev

-- 
Ville Syrjälä
Intel