This series is the culmination of quite a bit of poking at OpenCL kernels, exposing a bunch of details about integer and reduced-precision modes that don't normally come up in OpenGL (but come up in enough truly bizarre corners, i.e. blend shaders, that without this knowledge, debugging can be herculean).
Alyssa Rosenzweig (6): panfrost/midgard: Add rounding mode specific opcodes panfrost/midgard: Note floating compares type convert panfrost/midgard: Add integer outmods panfrost/midgard: Add shifting int modifiers panfrost/midgard: Expose vec8/vec16 modes panfrost/midgard: Disambiguate register mode .../drivers/panfrost/midgard/disassemble.c | 559 ++++++++++-------- .../drivers/panfrost/midgard/midgard.h | 46 +- .../panfrost/midgard/midgard_compile.c | 64 +- .../drivers/panfrost/midgard/midgard_ops.c | 32 +- 4 files changed, 404 insertions(+), 297 deletions(-) -- 2.20.1 _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev
