WebKit Bugzilla
New
Browse
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED FIXED
171784
fast/shadow-dom/link-element-in-shadow-tree.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=171784
Summary
fast/shadow-dom/link-element-in-shadow-tree.html is flaky
Alexey Proskuryakov
Reported
2017-05-06 20:01:59 PDT
@@ -1,5 +1,5 @@ -PASS Link elements should load stylesheets inside a connected shadow tree with closed mode +FAIL Link elements should load stylesheets inside a connected shadow tree with closed mode promise_test: Unhandled rejection with value: "timeout" PASS Link elements should load stylesheets inside a connected shadow tree with open mode PASS Link elements must not load stylesheets inside a disconnceted shadow tree with closed mode PASS Link elements must not load stylesheets inside a disconnceted shadow tree with open mode
Attachments
Fixes the test
(2.26 KB, patch)
2020-01-13 23:45 PST
,
Ryosuke Niwa
koivisto
: review+
Details
Formatted Diff
Diff
View All
Add attachment
proposed patch, testcase, etc.
Truitt Savell
Comment 1
2019-12-20 13:22:45 PST
This same failure still appears to happen. History:
https://results.webkit.org/?suite=layout-tests&test=fast%2Fshadow-dom%2Flink-element-in-shadow-tree.html
The diff is the exact same.
Radar WebKit Bug Importer
Comment 2
2019-12-20 13:23:11 PST
<
rdar://problem/58123584
>
Truitt Savell
Comment 3
2019-12-20 13:27:48 PST
History shows this flakiness became a lot worse recently around
r253500
Truitt Savell
Comment 4
2019-12-20 13:45:51 PST
marked test as failing in
https://trac.webkit.org/changeset/253843/webkit
Ryosuke Niwa
Comment 5
2020-01-13 23:10:23 PST
Hm... I can't reproduce this with: ./Tools/Scripts/run-webkit-tests --debug fast/shadow-dom/link-element-in-shadow-tree.html --iterations 100 --exit-after-n-failures=1 --no-show-results
Ryosuke Niwa
Comment 6
2020-01-13 23:45:37 PST
Created
attachment 387626
[details]
Fixes the test
Antti Koivisto
Comment 7
2020-01-13 23:51:11 PST
Comment on
attachment 387626
[details]
Fixes the test View in context:
https://bugs.webkit.org/attachment.cgi?id=387626&action=review
> LayoutTests/ChangeLog:10 > + Load and cache the styleshe before any test cases run to avoid this timeout.
stylesheet
Ryosuke Niwa
Comment 8
2020-01-14 17:05:51 PST
(In reply to Antti Koivisto from
comment #7
)
> Comment on
attachment 387626
[details]
> Fixes the test > > View in context: >
https://bugs.webkit.org/attachment.cgi?id=387626&action=review
> > > LayoutTests/ChangeLog:10 > > + Load and cache the styleshe before any test cases run to avoid this timeout. > > stylesheet
Oops, fixed.
Ryosuke Niwa
Comment 9
2020-01-14 17:09:39 PST
Oops, also forgot to remove the flaky test expectation. Doing that in the same patch.
Ryosuke Niwa
Comment 10
2020-01-14 17:11:11 PST
This isn't really a regression. As far as I can tell, this test had already been flaky.
Ryosuke Niwa
Comment 11
2020-01-14 17:11:15 PST
Committed
r254546
: <
https://trac.webkit.org/changeset/254546
>
Ryosuke Niwa
Comment 12
2020-01-17 22:10:03 PST
The test appears to be no longer flaky.
Alexey Proskuryakov
Comment 13
2020-01-17 22:40:07 PST
I do not concur. On open source bots, the test stopped failing even before this fix. On internal ones, I see failures even after the fix. PASS Link elements must not load stylesheets inside a disconnceted shadow tree with open mode PASS Style rules loaded inside a shadow tree with closed mode must apply to a div in the shadow tree PASS Style rules loaded inside a shadow tree with open mode must apply to a div in the shadow tree -PASS :host style rules loaded inside a shadow tree with closed mode must apply to the host element +FAIL :host style rules loaded inside a shadow tree with closed mode must apply to the host element promise_test: Unhandled rejection with value: "timeout" PASS :host style rules loaded inside a shadow tree with open mode must apply to the host element
Alexey Proskuryakov
Comment 14
2020-01-17 22:43:31 PST
Not sure if it's best to reopen or to have a new bug for a new fix in this case. And thank you for following up after landing the fix!
Ryosuke Niwa
Comment 15
2020-01-17 22:57:31 PST
(In reply to Alexey Proskuryakov from
comment #14
)
> Not sure if it's best to reopen or to have a new bug for a new fix in this > case. > > And thank you for following up after landing the fix!
Oh, interesting. Maybe you can file a radar for the internal failure. I can follow up on that there.
Alexey Proskuryakov
Comment 16
2020-01-18 13:17:25 PST
Filed
rdar://problem/58713068
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug