LayoutTest workers/sab/postMessage-clones.html is a flaky failure https://build.webkit.org/results/Apple%20El%20Capitan%20Debug%20WK2%20(Tests)/r212114%20(10957)/results.html https://webkit-test-results.webkit.org/dashboards/flakiness_dashboard.html#showAllRuns=true&tests=workers%2Fsab%2FpostMessage-clones.html --- /Volumes/Data/slave/elcapitan-debug-tests-wk2/build/layout-test-results/workers/sab/postMessage-clones-expected.txt +++ /Volumes/Data/slave/elcapitan-debug-tests-wk2/build/layout-test-results/workers/sab/postMessage-clones-actual.txt @@ -3,11 +3,11 @@ On success, you will see a series of "PASS" messages, followed by "TEST COMPLETE". -PASS successfullyParsed is true - -TEST COMPLETE PASS memory[0] is 42 PASS otherMemory[0] is 0 PASS memory[0] is 42 PASS otherMemory[0] is 43 +PASS successfullyParsed is true +TEST COMPLETE +
This is a new test added with https://trac.webkit.org/changeset/212035
Darn, that means I did a bad job writing the test.
I think I fixed it in https://trac.webkit.org/changeset/212132.
It appears that this test is still a flaky failure. https://build.webkit.org/builders/Apple%20El%20Capitan%20Release%20WK2%20(Tests)/builds/13234
Created attachment 302918 [details] proposed fix This is actually straightforward. Two issues: 1. waitUntilDone cannot be used in js-tests. This is very clear if you look at js-test-post.js. This is why expected test results look wrong, you don't want TEST COMPLETE to be printed out before the test is complete 2. Even with setTimeout, the test races with HTML parser. To avoid the race, the test could use onload instead of a zero delay timer. But that's not necessary once #1 is fixed. While at it, also updated doctype to HTML5, and migrated to js-test.js, which are both inconsequential changes.
Filed bug 168961 about many more tests that mouse waiUntilDone.
Comment on attachment 302918 [details] proposed fix Clearing flags on attachment: 302918 Committed r213174: <http://trac.webkit.org/changeset/213174>
All reviewed patches have been landed. Closing bug.