[Piglit] [PATCH 1/3] Test that gl_ClipVertex respects clip plane enable flags.

[Paul Berry]

On 6 October 2011 14:20, Ian Romanick <idr at freedesktop.org> wrote:

> On 10/03/2011 04:19 PM, Paul Berry wrote:
>
>> This test is similar to the existing
>> "vs-clip-distance-enables.**shader_test", except that it tests the
>> behavior of gl_ClipVertex instead of gl_ClipDistance.  Because of
>> differences in the way gl_ClipVertex and gl_ClipDistance are
>> implemented in Mesa, it's possible that one of these tests might pass
>> while the other fails.
>>
>
> There is one thing that I don't think I noticed when
> vs-clip-distance-enables.**shader_test went around for review.  If the
> test fails in, say, the first draw will be difficult to debug because it's
> overdrawn by later operations.  Other than breaking this into several
> separate tests, I don't see a good way to fix that.
>

Yeah, good point.  It would be cool if we could add a "pause" command to the
shader_test that causes the window to be displayed and then waits for a
keypress (except in "-auto" mode, of course, when it would just merrily
continue along).  Then we could sprinkle "pause" commands into tests like
this for easier debugging.  Hmm, that might not be that hard to implement.
I'll add it to my "back burner" list.

Another shader_runner feature that would be cool would be a "-step"
command-line option that causes it to execute the commands one at a time,
printing each command as it does so, and waiting for a keypress before it
moves on to the next command.


>
> So, other than just wanting to complain, the series is
>
> Reviewed-by: Ian Romanick <ian.d.romanick at intel.com>


Thanks!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/piglit/attachments/20111006/ad4f945d/attachment.htm>