Hey Emil, On 3 May 2017 at 15:32, Emil Velikov <emil.l.veli...@gmail.com> wrote: > On 3 May 2017 at 09:43, Daniel Stone <dani...@collabora.com> wrote: >> @@ -805,6 +807,7 @@ intel_create_image_from_fds(__DRIscreen *dri_screen, >> if (image == NULL) >> return NULL; >> >> + image->modifier = DRM_FORMAT_MOD_INVALID; > I think we want to fix intel_from_planar, > intel_create_image_from_renderbuffer and > intel_create_image_from_texture as well?
I think intel_from_planar should copy the modifier from its parent BO; not sure how I missed that, since I did specifically look at it! For renderbuffer, I _think_ we can derive from the existing renderbuffer's miptree tiling flag; hopefully the same as texture as well. Let's see what happens. Do you know of any good testcases which exercise these? > With that > Fixes: d78a36ea624 ("i965/dri: Handle the linear fb modifier") > Reviewed-by: Emil Velikov <emil.veli...@collabora.com> Thanks! I'll send a v2. > As a follow-up we might consider factoring and sharing the common bits > for all the "create image" functions. > Just an idea anyway. I agree, but figured I'd leave that one to the Intel guys. Or maybe for a very, very, very rainy day. ;) Cheers, Dan _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev