GitLab

Consequence of previous commit:
Revert "glsl_parser_test: Add config opt 'override_extensions'"

This reverts commit 3fd587bc.

The email that precipitated the revert:
On 01/27/2011 11:52 AM, Ian Romanick wrote:
    I don't understand why it is useful / necessary to have this in the
    test.  If you're trying to test driver features while they're being
    implemented, you can / should do this by hand.  If you're not trying to
    test driver features while they're being implemented, you don't any
    this to happen behind your back.

This is likely related to Mesa bugzilla #33197 and piglit test
glsl-vs-post-increment-01.  Christoph's original test was a
stand-alone executable, but it was converted to use shader_runner by
Ian Romanick.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>

vs-redeclaration-01.frag -> vs-redeclaration-01.vert

Reproduces Mesa FDO bug #33565.

It's just been one of those days...

Also add some quotations from the spec to justify the behavior tested.

Reproduces Mesa bug #33507.

These are related to Mesa bugzilla #33197.

This test triggers a crash in the AMD Catalyst 10.12 GLSL compiler.

The option manually enables or disables extensions via the environment
variable MESA_EXTENSION_OVERRIDE.

... and update all test file using the option.

The new name makes clear what the option does.

Replace functional, stateless style with more traditional stateful style.
This does not alter the public behavior of GLSLParserTest nor Piglit test
results.

Storytime: When I initially wrote GLSLParserTest, I attempted to use
a functional style in which the class maintained a minimal amount of
state. Now that I have a need to modify the class, I realize how unsuited
the stateless style is to Python. I atone for my function sins :)

These reproduce several bugs related to Mesa bugzilla #33219.

Catalyst 10.12 crashes on 4-dimensional arrays.

Array sizes greater than 256 trigger an assertion in the Mesa 7.8 GLSL
compiler.

In case glx.h / glxext.h aren't new enough.

Interestingly, only invalid-logic-not-06.vert reproduces Mesa bug #33314.

This reproduces Mesa bug #33308.  The crazy thing is that this is
*NOT* illegal.  The return type is void, and the return "value" has
the type void.  Madness.

invalid-equality-04.vert reproduces Mesa bugzilla #33303.

In GLSL 1.00, 1.10, and 1.20, check that vertex shader attributes and
fragment shader varyings cannot be written to.

In GLSL 1.30, check that global variables declared 'in' cannot be written
to.

forgot to disable texture after drawing the FBO.

made TestResults::write() thread-safe by writing results in one
chunk to the file so that when threaded tests are implemented
there will be no interleaving.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>

These tests originally wanted the FBO to be incomplete.  However, this
just verified incorrect behavior in another vendor's driver.  The spec
says that all of these cases should produce a complete FBO.

The format of the texture was also change from RGB to RGBA.  This
should prevent FRAMEBUFFER_UNSUPPORTED on some drivers.

In test precision-global-01.frag, a global variable was declared with the
'in' qualifier, which caused the test to unintentionally fail. Fix the
test by changing the qualifier to 'varying'.