Issue metadata
Sign in to add a comment
|
WebglConformance_conformance_textures_misc_tex_video_using_tex_unit_non_zero flaky on Win10 NV Debug |
||||||||||||||||||||||||
Issue descriptionThis bot: https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Win10%20FYI%20Debug%20%28NVIDIA%29?limit=200 Starting from this build: https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Win10%20FYI%20Debug%20%28NVIDIA%29/1164 Flakes on WebglConformance_conformance_textures_misc_tex_video_using_tex_unit_non_zero by having the pixel compare of that test fail. Looking in the changelogs for the 4 last builds, I found one slightly related commit if the failing video is h264, maybe: https://chromium.googlesource.com/chromium/src/+/22c1a0ffad44a60ab1ca4f18c1d414b311dc08aa Unfortunately I don't have a Windows machine to try and reproduce this.
,
Jul 18
Surfaces just got enabled it looks like too.
,
Jul 18
See b/111593092
,
Jul 18
i don't think that this bot runs with the new hw video decoder enabled, so it's likely not https://chromium.googlesource.com/chromium/src/+/22c1a0ffad44a60ab1ca4f18c1d414b311dc08aa . i wouldn't be surprised if it were related to surfaces in some way. we use the media gl context there, and maybe we're not resetting something properly. i'm wfh, and can't repro either.
,
Jul 18
It turns out this was reported earlier on Android, and is now happening on macOS as well. Duplicating this into the original bug report and raising the priority. The first flakiness was seen on Android on April 9; http://crrev.com/572793 didn't land until July 5. The current flakiness seems to have been introduced today. |
|||||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||||
Comment 1 by kbr@chromium.org
, Jul 18Cc: -vmi...@chromium.org jrumm...@chromium.org rkuroiwa@chromium.org liber...@chromium.org
Components: Internals>Media>Video Blink>WebGL
Labels: -Type-Bug -Pri-3 Pri-1 Type-Bug-Regression