GitLab

Designated initializers are not supported by MSVC.

This patch also fixes a MSVC build error. getopt.h is not available on
MSVC.

stdbool.h is not available with MSVC. piglit-util.h includes stdbool.h
and defines the equivalent types for MSVC.

Fixes Linux build.

hiz_check_fbo_depth_test() has return type bool and should have returned
true (1) on success and false (0) on failure. Instead, it returned
PIGLIT_PASS (0) on success and PIGLIT_FAIL (1) on failure.

Change extension requirement from GL_ARB_fbo to GL_ARB_framebuffer_object.
Signed-off-by: Chad Versace <chad.versace@intel.com>

Remove duplicate definitions of piglit_require_extension() from
piglit-util-{gl,gles2}.c and define it just once in piglit-util.c. Do the
same for piglit_require_not_extension().
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Chad Versace <chad.versace@intel.com>

It is created by the CMake and contains just build artifacts.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Chad Versace <chad.versace@intel.com>

Add the following tests:
    - hiz-depth-test-fbo-d24
    - hiz-depth-test-fbo-d24s8
    - hiz-depth-test-fbo-d24-s8
    - hiz-depth-test-window-stencil0
    - hiz-depth-test-window-stencil1
Signed-off-by: Chad Versace <chad.versace@intel.com>

Add three tests, in which the FBO has the following attachments:
    1. GL_DEPTH_ATTACHMENT: GL_DEPTH_COMPONENT24
    2. GL_DEPTH_ATTACHMENT: GL_DEPTH_COMPONENT24
       GL_STENCIL_COMPONENT: GL_STENCIL_INDEX8
    3. GL_DEPTH_STENCIL_ATTACHMENT: GL_DEPTH24_STENCIL8
Reviewed-By: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Chad Versace <chad.versace@intel.com>

Add tests for two cases: when the stencil buffer is, and is not, present.
Reviewed-By: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Chad Versace <chad.versace@intel.com>

Eventually there will be over a dozen HiZ tests, so placing common
functionality in a library will reduce a good hunk of code.
Reviewed-By: Ian Romanick <ian.d.romanick@intel.com>
Signed-off-by: Chad Versace <chad.versace@intel.com>

Verifies that at least one fbconfig exists that complies with the
minimums described by the spec.  Similar to glx-fbconfig-sanity but
different enough to be its own test.

Reviewed-by: Chad Versace <chad.versace@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>

On current Mesa, some of these cases just fail.  Some fail with
assertion failures.
Reported-by: Bryan Cain <bryancain3@gmail.com>
Reviewed-by: Chad Versace <chad.versace@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>

Reviewed-by: Eric Anholt <eric@anholt.net>

One of these is at least one of the bugs that Trine is hitting --
fragment.color was overriding other fragment.color[n] when present.

No tests successfully use rgbw and RGTC yet, since we can't render to
rgtc.  But when we go write some, this should reduce surprises.

These test that several operations correctly are passed/not passed by
conditional rendering.

The rest of piglit assumes that forward slash is the separator for the
group name. This patch fixes the summary HTML results on Windows. Tests
are now properly displayed in subgroups instead of the entire group name
listed as part of the test name.

These two tests were trying to check what buffer 0's type was, not
whether the pipeline as a whole was "fixed" for the purpose of
clamping.

This seems to have been removed by the commit below.  This is probably
the result of a merge gone wrong.

    commit e466e012
    Author: Brian Paul <brianp@vmware.com>
    Date:   Sun Apr 10 13:26:07 2011 -0600

        sampler-objects: test for GL_ARB_sampler_objects

The paths were previously generated incorrectly on Windows hosts and
resulted in shader tests not being found.

Julian Adams authored Mar 31, 2011 and Brian Paul committed Apr 13, 2011

Test that both vertex and fragment shader texture sampling samples from the
correct textures.
Signed-off-by: Brian Paul <brianp@vmware.com>

The point here is to try to catch the cases where the compiler turns a
min/max of a variable float with constant vectors into a saturate of a
float.

We never actually failed, becase we aren't writemasking in ir_to_mesa
(so the full temporary vector would be filled anyway), and because in
brw_fs.cpp we had split up the expression to scalars already.