On Thu, Mar 10, 2016 at 09:03:37AM +0000, Matthew Auld wrote: > > Y tiled fb will be rejected for gen < 9, should use X/linear for those. > > Are you referring to the __gem_set_tiling call? Since this didn't fail > on BDW and I don't see any logic in i915_gem_set_tiling which > implements this. No, I mean the addfb. Note that the current logic in addfb is sort of silly, and I tried to fix it in https://lists.freedesktop.org/archives/intel-gfx/2016-February/087670.html but no review so far. > > > Actually using X/linear for gen >= 9 would also be interesting to > make sure we reject 90/270 > > Yup, good point > > > I would suggest looking at the supported rotations for each plane > and picking some invalid ones based on that. > > Sorry, I must be missing something, isn't that what it already does? > I've just picked the invalid rotations for the primary, cursor and sprite > planes. I mean that instead of hardcoding the assumption which rotation are unsupported, you could actually look at what the driver reports as supported vs. not. -- Ville Syrjälä Intel OTC _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
