Hi Rob, On 22 July 2016 at 17:22, Rob Herring <r...@kernel.org> wrote: > Another version of common pipe_screen reference counting. Please help > test on AMD and Nouveau as those are the more complicated ones and I > don't have h/w to test. > > Changes in v4: > - Move fd dup() back into driver winsys create screen functions which > sometimes need the dup'ed fd before the pipe_screen is created. > - Update vmwgfx driver which I missed updating in v3 > - Update vc4 commit msg to reflect this is a new feature. > > Changes in v3: > - dup() fd and store in pipe_screen as the lifetime of the > pipe_loader_drm_device and pipe_screen are different. > - Fix leaking of pipe_loader_drm_device. Only the last one closed was > getting freed. > - Move mutex for fd hash table into u_screen.c > > Rob > > > Rob Herring (11): > gallium: move pipe_screen destroy into pipe-loader > pipe-loader-drm: protect create_screen() calls with a mutex > gallium: add common pipe_screen reference counting functions > pipe-loader-drm: use pipe_screen_unreference to destroy screen > nouveau: use common screen ref counting > freedreno: use common screen ref counting > amdgpu: use common screen ref counting > radeon: use common screen ref counting > vmwgfx: use common screen ref counting > virgl: use common screen ref counting > vc4: use common screen ref counting > Thank you very much for working on these and pardon for dropping the ball.
Barring the AMDGPU and close(fd) + ::destroy(foo) things look pretty good. Will you have the change to respin the series one final time so that we can land the lot ? Thanks Emil _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev
