WebGL 2 texture specification tests failing on Windows Intel |
||||||||
Issue descriptionLooks like Intel Windows is failing a few of these tests consistently. This is not on the CQ but we should add suppressions because it fails on the FYI waterfall. First failing build: https://build.chromium.org/p/chromium.gpu.fyi/builders/Win7%20Release%20%28New%20Intel%29/builds/708 Certainly because of https://chromium.googlesource.com/chromium/src/+/fb04cbe9cd74bedddbb9f3e9ae888b4a2bc73dd4 We should suppress the following tests on Wintel: WebglConformance.deqp_functional_gles3_texturespecification_teximage3d_depth WebglConformance.deqp_functional_gles3_texturespecification_texsubimage3d_depth WebglConformance.deqp_functional_gles3_texturespecification_texstorage3d_format_depth_stencil
,
May 24 2016
Can someone from ANGLE team take this bug? I don't have a local Windows machine with Intel GPU, nor do I have the bandwidth at this point.
,
May 24 2016
Sure, that's fine. I think in this case we might want to deprioritize the fix or workaround if it's Intel-only. It's not clear that we'll be able to support fully conformant WebGL 2 on Intel Windows immediately.
,
Jan 6 2017
Looking at conformance failures on Win / Intel HD 620.
We have the following expectations pointing at this bug (indirectly):
self.Fail('deqp/functional/gles3/textureformat/sized_color_3d_pot_00.html',
['win', 'intel'], bug= 614418 )
self.Fail('deqp/functional/gles3/textureformat/sized_color_3d_pot_01.html',
['win', 'intel'], bug= 614418 )
self.Fail('deqp/functional/gles3/textureformat/sized_color_3d_pot_02.html',
['win', 'intel'], bug= 614418 )
self.Fail('deqp/functional/gles3/textureformat/sized_color_3d_pot_03.html',
['win', 'intel'], bug= 614418 )
These still fail but so do many others in this category, so either this bug is hitting more tests or another bug is hitting all of them.
self.Fail('deqp/functional/gles3/textureformat/sized_depth_stencil.html',
['win', 'intel'], bug= 614418 )
Still failing.
self.Flaky('deqp/functional/gles3/textureformat/unsized_3d.html',
['win', 'intel'], bug= 614418 )
This one is NOT failing anymore.
Here are all of the textureformat/ tests failing on this cfg:
deqp/functional/gles3/textureformat/sized_color_2d_pot_00.html: 6 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_pot_01.html: 6 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_pot_02.html: 6 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_pot_03.html: 7 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_npot_00.html: 6 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_npot_01.html: 6 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_npot_02.html: 6 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_npot_03.html: 7 tests failed
deqp/functional/gles3/textureformat/sized_color_cube_pot_00.html: 28 tests failed
deqp/functional/gles3/textureformat/sized_color_cube_pot_01.html: 32 tests failed
deqp/functional/gles3/textureformat/sized_color_cube_pot_02.html: 32 tests failed
deqp/functional/gles3/textureformat/sized_color_cube_pot_03.html: 34 tests failed
deqp/functional/gles3/textureformat/sized_color_cube_npot_00.html: 28 tests failed
deqp/functional/gles3/textureformat/sized_color_cube_npot_01.html: 32 tests failed
deqp/functional/gles3/textureformat/sized_color_cube_npot_02.html: 32 tests failed
deqp/functional/gles3/textureformat/sized_color_cube_npot_03.html: 34 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_array_pot_00.html: 48 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_array_pot_01.html: 48 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_array_pot_02.html: 48 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_array_pot_03.html: 56 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_array_npot_00.html: 42 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_array_npot_01.html: 42 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_array_npot_02.html: 42 tests failed
deqp/functional/gles3/textureformat/sized_color_2d_array_npot_03.html: 49 tests failed
deqp/functional/gles3/textureformat/sized_color_3d_pot_00.html: 96 tests failed
deqp/functional/gles3/textureformat/sized_color_3d_pot_01.html: 96 tests failed
deqp/functional/gles3/textureformat/sized_color_3d_pot_02.html: 96 tests failed
deqp/functional/gles3/textureformat/sized_color_3d_pot_03.html: 112 tests failed
deqp/functional/gles3/textureformat/sized_color_3d_npot_00.html: 42 tests failed
deqp/functional/gles3/textureformat/sized_color_3d_npot_01.html: 42 tests failed
deqp/functional/gles3/textureformat/sized_color_3d_npot_02.html: 42 tests failed
deqp/functional/gles3/textureformat/sized_color_3d_npot_03.html: 49 tests failed
deqp/functional/gles3/textureformat/sized_depth_stencil.html: 3 tests failed
,
Jan 6 2017
CC'ing Yunchao, Qiankun and Yang from Intel's Web GPU team. I wonder whether we could collaborate to get these issues fixed, and WebGL 2.0 conformance runs clean on Windows with Intel GPUs?
,
Jan 6 2017
FYI: I emailed the full results of my conformance run to the Intel Web GPU team for them to take a look (including these and a number of other failures).
,
Jan 9 2017
Can you cc me on that email Kai? What GPU family were you using?
,
Jan 9 2017
FYI. I tried to roll-in webgl CTS and run the tests on all bots. It seems to me that there were many failures for multiple GPU vendors. see https://codereview.chromium.org/2614313002/
,
Feb 20 2017
Hi Kai: Could you verify if these failures still occur on HD620 with the newly released driver 4590?
,
Jun 20 2017
,
Jun 21 2018
This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue. Sorry for the inconvenience if the bug really should have been left as Available. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Jun 21 2018
Intel team: could you please send a tryjob to win_optional_gpu_tests_rel which un-suppresses these tests and see whether they are now passing on the new Win/Intel bots? Thanks.
,
Jun 27 2018
This failures are not existed in the webgl expectation file now. The failures were removed at this issue long ago: https://bugs.chromium.org/p/chromium/issues/detail?id=614418. Mark this issue as fixed. |
||||||||
►
Sign in to add a comment |
||||||||
Comment 1 by zmo@chromium.org
, May 24 2016