Bug 283203 – REGRESSION (283213@main): [ macOS Debug ] fast/canvas/imageData-consistency.html is a consistent timeout

• Component: New Bugs
• Status: RESOLVED FIXED

[Karl Rackler]

Created attachment 473237 [details]
layout-tests stdio

Description:
fast/canvas/imageData-consistency.html

I cannot reproduce down to the commit that introduced the change due to a lack of roots in the 283216@main...283212@main range. 

Looking at commits in range https://commits.webkit.org/compare/283216@main...283212@main, the only relevant change is 283213@main, which disabled partial layout when replaced bot with relative dimension(s) is present.

This issue can be bisected to 283213@main using the command: 
run-webkit-tests --debug --iterations 2000 --exit-after-n-failures 1 --exit-after-n-crashes-or-timeouts 1 
--no-retry --clobber-old-results --force -f fast/canvas/imageData-consistency.html

History:
https://results.webkit.org/?suite=layout-tests&test=fast%2Fcanvas%2FimageData-consistency.html&platform=mac&style=debug

Diff:
There is no diff since the test timeout.

Here is the relevant section from the bot layout-test run:
06:10:25.963 79034 worker/6 fast/inline/incorrect-content-clip-when-list-marker-is-present.html passed
06:10:25.969 79034 worker/3 "/usr/bin/sample 79140 10 10 -file /var/folders/h3/f42c_9t53gg3ms5thjdfbd480000gn/T/tmpidzu9900/DumpRenderTree-79140-sample.txt" took 10.65s
06:10:25.985 79034 worker/9 fast/history/visited-link-background-color.html passed
06:10:26.020 79034 worker/5 fast/dom/rangeContainsRange.html passed
06:10:26.020 79034 worker/11 fast/invalid/015.html passed
06:10:26.150 79034 worker/11 fast/invalid/016.html passed
06:10:26.202 79034 worker/3 Couldn't quit 79140 with SIGTERM, sending SIGKILL.
06:10:26.205 79034 worker/6 fast/inline/incorrect-content-height-with-empty-generated-content.html passed
06:10:26.209 79034 worker/5 fast/dom/rangeIntersectsNode.html passed
06:10:26.252 79034 worker/11 fast/invalid/017.html passed
06:10:26.256 79034 worker/3 killed pid 79140
06:10:26.260 79034 [15309/72044] fast/canvas/imageData-consistency.html failed unexpectedly (test timed out, no output from test)
06:10:26.260 79034 worker/3 killing driver
06:10:26.264 79034 worker/3 fast/canvas/imageData-consistency.html failed:
06:10:26.264 79034 worker/3 test timed out
06:10:26.264 79034 worker/3 no output from test

[Radar WebKit Bug Importer]

<rdar://problem/140003573>

[zalan]

It passes locally on my trunk build but the test is indeed slow. However it is not necessarily an indication of a bug but it could simply be that we are now running _full_ layout with all the extra debug only codepath (e.g. slow path vs. fast path verifications etc).

[Karl Rackler]

(In reply to zalan from comment #2)
> It passes locally on my trunk build but the test is indeed slow. However it
> is not necessarily an indication of a bug but it could simply be that we are
> now running _full_ layout with all the extra debug only codepath (e.g. slow
> path vs. fast path verifications etc).

You may be right.  There is a definite pass before the change and timeout after the 283213@main change.

[zalan]

Created attachment 473924 [details]
Patch

[zalan]

Created attachment 473925 [details]
[fast-cq]Patch

[EWS]

Committed 289024@main (bc2afade7b0d): <https://commits.webkit.org/289024@main>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 473925 [details].