On Tue, Nov 14, 2017 at 04:58:34PM +0200, Juha-Pekka Heikkila wrote:
> DRM_FORMAT_C8 is needed for kms_rotation_crc test
>
> Signed-off-by: Juha-Pekka Heikkila <juhapekka.heikk...@gmail.com>
> ---
> lib/igt_fb.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/lib/igt_fb.c b/lib/igt_fb.c
> index d4eaed7..00d52f7 100644
> --- a/lib/igt_fb.c
> +++ b/lib/igt_fb.c
> @@ -67,6 +67,7 @@ static struct format_desc_struct {
> DF(XRGB8888, RGB24, 32, 24),
> DF(XRGB2101010, RGB30, 32, 30),
> DF(ARGB8888, ARGB32, 32, 32),
> + DF(C8, A8, 8, 8),
What I've been thinking for pixel formats not supported by cairo is that
we'd add software routines to convert between them and XRGB8888 and hook
that up so that we do the conversion to/from XRGB888 in the cairo
surface create/destroy hooks. And if we go nuts we could even expand
rendercopy to do some of the format conversions for us on the GPU.
For the time being, I think it's probably better to just check for C8
in the test and skip the cairo stuff entirely. Since this is just for
the bad-format test we don't actually need the fb to have any real
content.
> };
> #undef DF
>
> --
> 2.7.4
>
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
--
Ville Syrjälä
Intel OTC
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
