On Wed, Apr 3, 2024 at 3:43 AM Pekka Paalanen <pekka.paala...@collabora.com> wrote: > > On Tue, 2 Apr 2024 19:28:13 -0400 > Zack Rusin <zack.ru...@broadcom.com> wrote: > > > The table of primary plane formats wasn't sorted at all, leading to > > applications picking our least desirable formats by defaults. > > > > Sort the primary plane formats according to our order of preference. > > This is good. > > > Fixes IGT's kms_atomic plane-invalid-params which assumes that the > > preferred format is a 32bpp format. > > That sounds strange, why would IGT depend on preferred format being > 32bpp? > > That must be an oversight. IGT cannot dictate the format that hardware > must prefer. XRGB8888 is strongly suggested to be supported in general, > but why also preferred?
I think it's just a side-effect of the pixman's assert that's failing: https://cgit.freedesktop.org/drm/igt-gpu-tools/tree/lib/igt_fb.c#n4190 i.e. pixman assumes everything is 4 byte aligned. I should have rephrased the message as "IGT assumes that the preferred fb format is 4 byte aligned because our 16bpp formats are packed and pixman can't convert them". z