Bug 240551 – REGRESSION(r294317): [iOS] 3* editing* Layout tests are constant timeouts

• Status: RESOLVED
• Resolution: FIXED
• Component: New Bugs

[Dawn Morningstar]

editing/input/cocoa/autocorrect-off.html 
editing/input/cocoa/autocorrect-on.html 
editing/undo-manager/undo-manager-item-labels.html 

HISTORY:
https://results.webkit.org/?suite=layout-tests&suite=layout-tests&suite=layout-tests&test=editing%2Fundo-manager%2Fundo-manager-item-labels.html&test=editing%2Finput%2Fcocoa%2Fautocorrect-off.html&test=editing%2Finput%2Fcocoa%2Fautocorrect-on.html

DIFF:
--- /Volumes/Data/worker/Apple-iPadOS-15-Simulator-Debug-WK2-Tests/build/layout-test-results/editing/undo-manager/undo-manager-item-labels-expected.txt
+++ /Volumes/Data/worker/Apple-iPadOS-15-Simulator-Debug-WK2-Tests/build/layout-test-results/editing/undo-manager/undo-manager-item-labels-actual.txt
@@ -1,3 +1,5 @@
+FAIL: Timed out waiting for notifyDone to be called
+
 
 Verifies that setting the label attribute of UndoItem affects the undo and redo action names in the platform undo manager.
 
@@ -28,9 +30,4 @@
 PASS undoName is "Paste"
 PASS redoName is ""
 After undoing the paste command:
-PASS undoName is "First 🥇"
-PASS redoName is "Paste"
-PASS successfullyParsed is true
 
-TEST COMPLETE
-

This appears to be a regression due to changes in r294317.

[Radar WebKit Bug Importer]

<rdar://problem/93461381>

[Dawn Morningstar]

fast/forms/ios/autocapitalize-words.html
fast/forms/ios/form-control-refresh/select/select-multiple-picker.html
fast/forms/ios/inputmode-none.html
fast/forms/ios/no-scrolling-when-moving-focus-between-adjacent-fields.html
fast/forms/validation-messages.html
fast/forms/datalist/datalist.html

These tests all appear to have a similar regression point.

https://results.webkit.org/?suite=layout-tests&suite=layout-tests&suite=layout-tests&suite=layout-tests&suite=layout-tests&suite=layout-tests&test=fast%2Fforms%2Fdatalist%2Fdatalist.html&test=fast%2Fforms%2Fios%2Fautocapitalize-words.html&test=fast%2Fforms%2Fios%2Fform-control-refresh%2Fselect%2Fselect-multiple-picker.html&test=fast%2Fforms%2Fios%2Finputmode-none.html&test=fast%2Fforms%2Fios%2Fno-scrolling-when-moving-focus-between-adjacent-fields.html&test=fast%2Fforms%2Fvalidation-messages.html

[Dawn Morningstar]

(In reply to Matteo Flores from comment #2)
> fast/forms/ios/autocapitalize-words.html
> fast/forms/ios/form-control-refresh/select/select-multiple-picker.html
> fast/forms/ios/inputmode-none.html
> fast/forms/ios/no-scrolling-when-moving-focus-between-adjacent-fields.html
> fast/forms/validation-messages.html
> fast/forms/datalist/datalist.html
> 
> These tests all appear to have a similar regression point.
> 
> https://results.webkit.org/?suite=layout-tests&suite=layout-
> tests&suite=layout-tests&suite=layout-tests&suite=layout-tests&suite=layout-
> tests&test=fast%2Fforms%2Fdatalist%2Fdatalist.
> html&test=fast%2Fforms%2Fios%2Fautocapitalize-words.
> html&test=fast%2Fforms%2Fios%2Fform-control-refresh%2Fselect%2Fselect-
> multiple-picker.html&test=fast%2Fforms%2Fios%2Finputmode-none.
> html&test=fast%2Fforms%2Fios%2Fno-scrolling-when-moving-focus-between-
> adjacent-fields.html&test=fast%2Fforms%2Fvalidation-messages.html

Forgot this one fast/forms/ios/inputmode-removing-none.html

[Dawn Morningstar]

These also appear to have the same regression point:
fast/events/ios/do-not-show-keyboard-when-focusing-after-blur.html 
fast/events/ios/keyup.html 
fast/events/ios/unscalable-viewport-clicks-on-doubletap.html 
fast/shadow-dom/ios/accessory-bar-work-on-input-with-tabindex-in-shadow-tree.html 
http/wpt/webauthn/public-key-credential-unfocused-document.https.html 
mathml/fractions-invalid.html

[Dawn Morningstar]

Adding two more:

TestWebKitAPI.DragAndDropTests.CanDragImageWhenNotFirstResponder 
TestWebKitAPI.DragAndDropTests.InjectedBundleImageElementData

[Robert Jenner]

I was able to reproduce all of these timeouts using the following tests:

For Layout tests:

Run-WebKit-Tests --iOS-simulator <test goes here>

For API-Tests:

run-api-tests --iOS-simulator <test goes here>

I have also confirmed that r294317 is our regression point, and did break all of these tests. For all of the tests mentioned above I was able to reproduce the timeouts at r294317, but they all passed when I ran the testing at r294316.

We should revert r294317 since it broke 16 layout-tests, and 2 api-tests.

[Dawn Morningstar]

fixed with a revert here: https://bugs.webkit.org/show_bug.cgi?id=240613