[Webkit-unassigned] [Bug 252905] [GTK] Lots of WebGL2 tests are flaky

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=252905

--- Comment #2 from Carlos Alberto Lopez Perez <clopez at igalia.com> ---
(In reply to Carlos Alberto Lopez Perez from comment #0)
> 
> Something weird is going on with this tests.
> It is not normal that they can give 11 times a failure on a row and then 11
> times a pass on another row.
> It looks like if the flakiness was decided at build-time rather than a
> run-time (either you got a build that is going to fail always or always
> going to pass). So the result doesn't change until you rebuild or something
> like that. Just speculation on my part here.
> 

Could be something also related to clean build issues like: the build-only bot is creating builds that fail because a bug that triggers due to a continous-build issue but the test-only bot creates build that work because it is doing continous builds from a further point.

In any case the gardening done here is just based on the info from the post-commit bots (not the EWS) ones so I believe there is something else going here than just a 'clean build' issue. I also checked the info from other post-commit bots like the skip-failing tests one and they report same issue (lots of flakies).

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20230224/83e44230/attachment-0001.htm>