On Wed, 2016-10-26 at 11:47 +1100, Timothy Arceri wrote: > Did you test early_fragment_tests specifically? This is one that I > checked is only in merge_in_qualifier() and not merge_qualifier()
AFAIK, early_fragment_tests is the only default input layout-qualifier- name for a fragment shader (see the table in the section 4.4 (Layout Qualifiers) of the GLSL 4.5 spec), so I don't know how it could happen something like: ~~~~~~~~~~~~~~~~~~~~~~~~~ layout(...) layout(early_fragment_tests) in; ~~~~~~~~~~~~~~~~~~~~~~~~~ or viceversa. The only possibility would be: ~~~~~~~~~~~~~~~~~~~~~~~~~ layout(early_fragment_tests) layout(early_fragment_tests) in; ~~~~~~~~~~~~~~~~~~~~~~~~~ Therefore, I could add management for that layout-qualifier-name in merge_qualifier(), but that won't make a difference. In addition, the generated tests for early_fragment_tests are passing, the same than tests/spec/arb_shader_image_load_store/early-z.c FTR, I did a quick modification of tests/spec/arb_shader_image_load_store/early-z.c to check that the only possible combination works. It does. I don't think a test for that is valuable, though, so I won't send a patch for piglit. -- Br, Andres _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev