GitLab

Add functions:
    piglit_probe_pixel_stencil
    piglit_probe_rect_stencil
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>

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.

This was caused by bead3e50

The proper macro for Windows operating systems is _WIN32 and not WIN32.
Thanks to Vinson Lee to remind me of that.

which don't print anything, but can return the probed pixel color.

Replace the two identical definitions in piglit-util-gl.c and in
piglit-util-gles2.c with a single definition in piglit-util.c.
Signed-off-by: Chad Versace <chad@chad-versace.us>

API-specific files are now named as COMMON_NAME-API.EXT, and
API-independent files are now named COMMON_NAME.EXT. For example,
	piglit-util.h
	piglit-util.c
	piglit-util-gl.c
	piglit-util-gles2.c
Signed-off-by: Chad Versace <chad.versace@intel.com>

And replace all occurences of glutExtensionSupported() in piglit-util.c.

Note: This belongs to a series that adds GLES2 support to Piglit.
Signed-off-by: Chad Versace <chad.versace@intel.com>

This reverts commit b51e9ef1.  While a
good idea, it causes errors with Mesa debug builds due to glGetError()
being called after the context is destroyed in automatic mode
framework tests.

Silences this GCC warning.
piglit-util.c: In function 'strchrnul':
piglit-util.c:1060: warning: return discards qualifiers from pointer target type

This alpha modification is optional.

It will be needed for testing fbo-generatemipmap-formats/GL_ALPHA,
which currently always passes.

This commit does not change behavior of any test (note the GL_FALSE
parameter).

I don't have an implementation with some of these extensions handy, so
they may fail due to testcase bugs, but it's already showing off
failure in EXT_texture_sRGB like I hoped, and there's bonus fail I
wasn't expecting in GL_ARB_depth_texture and friends.

Fixes glsl-vs-normalscale result.

This also manages to hang 965 on 7.8 and master.

This exercises some obviously never-tested code paths, because all drivers
here fail. (swrast, r300, all gallium ones)

This significantly clarifies the problem in glsl-routing in the 965
driver, since we don't have to worry about issues with sparse varying
usage.  We just fail at handling more than 7 vec4 varyings (on my
Ironlake -- the number is likely a bit higher pre-Ironlake).

I've copied that function around a few texturing/ files, and wanted to
use it in shader_runner now.

This allows to decouple building piglit from the testing with piglit.