Bug 253252

Summary: REBASELINE REGRESSION(iPad 9th gen): [ iPad Simulator ] 2x editing/selection tests are constant failures.
Product: WebKit Reporter: Bri Harris <bharris9>
Component: New BugsAssignee: Bri Harris <bharris9>
Status: NEW ---    
Severity: Normal CC: ap, ntim, thorton, webkit-bot-watchers-bugzilla, webkit-bug-importer, wenson_hsieh
Priority: P2 Keywords: InRadar
Version: Other   
Hardware: Unspecified   
OS: Unspecified   
Bug Depends on: 253490    
Bug Blocks:    

Description Bri Harris 2023-03-02 09:53:23 PST 
The following 2 editing/selection layout-tests are flaky test failures on iPad Simulators:

editing/selection/ios/show-selection-in-transformed-container-2.html
editing/selection/ios/show-selection-in-transformed-container.html


HISTORY:
https://results.webkit.org/?suite=layout-tests&suite=layout-tests&test=editing%2Fselection%2Fios%2Fshow-selection-in-transformed-container-2.html&test=editing%2Fselection%2Fios%2Fshow-selection-in-transformed-container.html

DIFF TEXT: 

 @@ -4,11 +4,12 @@
 On success, you will see a series of "PASS" messages, followed by "TEST COMPLETE".
 
 
-PASS caretRect.left is 10
+FAIL caretRect.left should be 10. Was 9.
 PASS caretRect.top is 4
-PASS caretRect.width is 2
+FAIL caretRect.width should be 2. Was 3.
 PASS caretRect.height is 24
 PASS successfullyParsed is true
+Some tests failed.
 

These test seemed to start failing after the iPad generation was changed on simulators. It appears these will need to be re-baselined.
Comment 1 Radar WebKit Bug Importer 2023-03-02 09:53:43 PST 
<rdar://problem/106150327>
Comment 2 Alexey Proskuryakov 2023-03-05 11:47:41 PST 
A rebaseline cannot address a flaky failure, because by definition the results are unstable. But looking at the history, I don't see it being flaky?
Comment 3 Wenson Hsieh 2023-03-05 11:50:31 PST 
(In reply to Alexey Proskuryakov from comment #2)
> A rebaseline cannot address a flaky failure, because by definition the
> results are unstable. But looking at the history, I don't see it being flaky?

Indeed — these tests have been failing consistently since they've started running on iPad for the first time, fairly recently:

https://results.webkit.org/?suite=layout-tests&suite=layout-tests&test=editing%2Fselection%2Fios%2Fshow-selection-in-transformed-container-2.html&test=editing%2Fselection%2Fios%2Fshow-selection-in-transformed-container.html&model=iPad%20(9th%20generation)
Comment 4 EWS 2023-03-06 14:24:42 PST 
Test gardening commit 261292@main (c1f7133c1b39): <https://commits.webkit.org/261292@main>

Reviewed commits have been landed. Closing PR #11133 and removing active labels.
Comment 5 Tim Nguyen (:ntim) 2023-03-07 00:11:15 PST 
(In reply to EWS from comment #4)
> Test gardening commit 261292@main (c1f7133c1b39):
> <https://commits.webkit.org/261292@main>
> 
> Reviewed commits have been landed. Closing PR #11133 and removing active
> labels.

I reverted this in bug 253490, since it was slowing down EWS queues. Will see if I can reland a correct patch.
Comment 6 Bri Harris 2023-03-07 16:02:45 PST 
Pull request: https://github.com/WebKit/WebKit/pull/11200
Comment 7 EWS 2023-03-07 16:07:28 PST 
Test gardening commit 261347@main (a9d2ee0df5a8): <https://commits.webkit.org/261347@main>

Reviewed commits have been landed. Closing PR #11200 and removing active labels.