Bug 112603
| Summary: | [chromium] fast/innerHTML/innerHTML-iframe.html is flaky | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Kenneth Russell <kbr> |
| Component: | Tools / Tests | Assignee: | Kenneth Russell <kbr> |
| Status: | RESOLVED FIXED | ||
| Severity: | Normal | CC: | abarth, eric, jamesr |
| Priority: | P2 | ||
| Version: | 528+ (Nightly build) | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
Kenneth Russell
Layout test fast/innerHTML/innerHTML-iframe.html is flaky on all platforms. See:
http://test-results.appspot.com/dashboards/flakiness_dashboard.html#group=%40ToT%20-%20chromium.org&showAllRuns=true&showExpectations=true&tests=fast%2FinnerHTML%2FinnerHTML-iframe.html
Seems strange that the onload handler wouldn't be able to reference document.body. Could the new multithreaded parser have something to do with these failures?
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
James Robinson
Hmm, I thought I fixed the document.body flake in r145857 but it looks like the setTimeout is still racing with onload. The failing diff now is:
TypeError: Cannot set property 'innerHTML' of null
PASS: body and iframe cleared without crashing.
which means the setTimeout() blew away <div id="x"> before the onload fired.
James Robinson
Perhaps the right fix is to set the timeout from inside the onload handler so it always happens after the .innerHTML.
Kenneth Russell
Is the existing pattern in the test supposed to be guaranteed to work though?
James Robinson
No, it's racy.
Kenneth Russell
OK. I'll update the test.
Kenneth Russell
Fixed the test in http://trac.webkit.org/changeset/146126 .