Hi Gregory, My apologies I only just found this series, when you said your Mesa patchs I thought you meant the ones in the bug comments.
It seems we have taken very different appoaches to the problem with inputs, I do think my change here is simpler and my series also address outputs being removed too. It would be great if you could test my series with PCSX2. I will review and test out your patch 3 tomorrow, thanks for this and sorry for the confusion on my side. Tim On Sun, 2015-10-25 at 15:01 +0100, Gregory Hainaut wrote: > v4: > Rebase against lastest master and fix bad variable name in assert. > > A new test was developed to properly check commit 4 behavior. > I ran most of the piglit test without regression. > > v3: > Squash old commit 1&2 > * Use a better name for the new attribute: always_active_io > * Use ir_variable directly instead of ir_variable_refcount_visitor > * Put related code in linker.cpp > > Add 2 new commits to fix wrong interface matching in more complex case. > Commit 3: avoid collision between user and linker slot assignment > Commit 4: avoid unpredictable sorting of varying > > Commit 1/2/3 fix the piglit test: arb_separate_shader_object > -rendezvous_by_name posted on piglit ML > Commit 4 was tested on the PCSX2 application. > > Gregory Hainaut (4): > glsl IR: add always_active_io attribute to ir_variable > glsl IR: only allow optimization of interstage variable > glsl: avoid linker and user varying location to overlap > glsl: don't sort varying in separate shader mode > > src/glsl/ir.cpp | 1 + > src/glsl/ir.h | 7 +++++ > src/glsl/link_varyings.cpp | 76 ++++++++++++++++++++++++++++++++++++++++++- > --- > src/glsl/linker.cpp | 73 > ++++++++++++++++++++++++++++++++++++++++++++ > src/glsl/opt_dead_code.cpp | 18 +++++++++++ > 5 files changed, 169 insertions(+), 6 deletions(-) > _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev