270757 2024-03-09 16:21:31 -0800 Do not run the references when a test crashes or times out 2024-04-25 17:13:04 -0700 1 1 1 Unclassified WebKit Tools / Tests WebKit Nightly Build Unspecified Unspecified RESOLVED DUPLICATE 273212 InRadar P2 Normal --- 1 ahmad.saleem792 webkit-unassigned ap gsnedders jbedard ryanhaddad webkit-bug-importer oldest_to_newest 2019878 0 ahmad.saleem792 2024-03-09 16:21:31 -0800 Hi Team, While going through Blink's commit, I came across this optimization for test suite but don't know whether we have done similar in one way or shape but just wanted to raise to get input. Blink Commit - https://chromium.googlesource.com/chromium/blink/+/a7ba83f0e6c7c9c1ebcab9ea4f53e899b26c397e WebKit Source: https://searchfox.org/wubkat/rev/93c3eb8ac99a6b6e3894deaf14e5d38304898b48/Tools/Scripts/webkitpy/layout_tests/controllers/single_test_runner.py#354 and https://searchfox.org/wubkat/rev/93c3eb8ac99a6b6e3894deaf14e5d38304898b48/Tools/Scripts/webkitpy/layout_tests/controllers/single_test_runner.py#378 __ Blink Commit Details: If the actual test in a reftest crashes or times out, we probably shouldn't bother executing the reference as well, since we have nothing to compare it to, and the test failed regardless. This can save a bunch of execution time and avoid some confusion. __ Just raising to get input. Thanks! 2021616 1 webkit-bug-importer 2024-03-16 17:22:13 -0700 <rdar://problem/124790780> 2028838 2 jbedard 2024-04-16 09:37:41 -0700 This doesn't seem like a bad idea, although in practice, it seems unlikely to save us much time because we tend to not leave CRASH and TIMEOUT tests running, because such expectations are of very low value. 2031376 3 jbedard 2024-04-25 17:13:04 -0700 I'm resolving this as a duplicate of https://bugs.webkit.org/show_bug.cgi?id=273212. The rationale in that bug is quite different than in this one, but since crashes and timeouts also don't have test output, https://bugs.webkit.org/show_bug.cgi?id=273212 will end up with the same effect as the one requested here. *** This bug has been marked as a duplicate of bug 273212 ***