[Piglit] [PATCH] arb_clear_texture: add a trivial test

[Ilia Mirkin]

On Fri, Mar 7, 2014 at 11:26 PM, Ilia Mirkin <imirkin at alum.mit.edu> wrote:
> On Fri, Mar 7, 2014 at 8:02 PM, Eric Anholt <eric at anholt.net> wrote:
>> Ilia Mirkin <imirkin at alum.mit.edu> writes:
>>
>>> On Fri, Mar 7, 2014 at 2:38 PM, Eric Anholt <eric at anholt.net> wrote:
>>>> Ilia Mirkin <imirkin at alum.mit.edu> writes:
>>
>>>>
>>>> Other than that, this is a nice simple test.  We're going to want an
>>>> all-the-formats testcase before we land clear_texture in Mesa, but this
>>>> is a good start.
>>>
>>> Absolutely. And testing glClearTexSubImage too (where the visual
>>> representation may become more interesting). And an api-errors style
>>> test. And I want to test various widths/heights, in case e.g. POT
>>> works but NPOT doesn't. I'm thinking that 32 and 33 actually cover
>>> most cases.
>>
>> This all sounds good!
>>
>>> Should I make these into subtests, or is it fine to just have the one
>>> test? Long-term, should this 'simple' test go away entirely in favor
>>> of a all-formats test that does it all?
>>
>> I think there's a lot to be said for having a dead simple sanity test
>> test, along with all the complicated stuff.  When you're first lighting
>> up the extension, having all the other variants in the way obscures the
>> initial debugging.
>>
>> For similar reasons, I lean toward separate test files instead of
>> subtests unless the code sharing is massive.  I'm trying to optimize for
>> me debugging the driver in the future, not me writing the test now.  The
>> bodies of tests don't change much over time, so there's less maintenance
>> burden from code duplication than in most codebases we work on.
>
> I'm trying to implement the piglit_draw_rect_tex suggestion (based on
> ext_unpack_subimage.c), and am having a lot of trouble with the GL
> basics. Here's the code that I have now:
>
> program = piglit_build_simple_program(vertex_shader, fragment_shader);
> glUseProgram(program);
>
> ... do the clears on texture[0] and texture[1] ...
>
> /* Render both textures to the screen */
> glBindFramebuffer(GL_FRAMEBUFFER, piglit_winsys_fbo);
> glBindTexture(GL_TEXTURE_2D, texture[0]);
> piglit_draw_rect_tex(0, 0, 32, 32, 0, 0, 32, 32);
>
> glBindTexture(GL_TEXTURE_2D, texture[1]);
> piglit_draw_rect_tex(32, 32, 32, 32, 0, 0, 32, 32);
>
> This does not work. I'm sure the solution to this problem is quite
> trivial, but I have no idea what to look for. Semi-relatedly, what
> should the split between piglit_init and piglit_display be?

Nevermind, got it all figured out. Turns out that
piglit_present_results() is actually quite important :) I've ended up
with a solution copied from fbo-drawbuffers.c which does not rely on
shaders either. I guess I'll eventually have to go that route (to get
at MS samples), but this is all much simpler for now. v3 should be
coming up soon.

  -ilia