Yeah, after looking at one more instance of this in bug 1787569, I don't think `<meta name="timeout" content="long">` spot-treatment would be worthwhile here. On top of what Henrik noted, I think would have to deploy it too broadly. In bug 1787569, the "slow" test (`outline-valid-mandatory.html`) **literally just calls `test_valid_value` 20 times**, and that's it. There are no special timeouts or looping/async fanciness that might be a spot where you could imagine chaos mode surfacing some real potential for edge-casey slowness. Under the normal-configuration part of that bug's test-verify run, that test usually completed in 70-90ms; but with chaos mode, the slowest testrun was 100x slower and took `9294ms`, which is close enough to the 10s threshold that the test-verify run was considered a failure (hence the bug having been filed).
Bug 1734020 Comment 6 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
Yeah, after looking at one more instance of this in bug 1787569, I don't think `<meta name="timeout" content="long">` spot-treatment would be worthwhile here. On top of what Henrik noted, I think would have to deploy it too broadly. In bug 1787569, the "slow" test (`outline-valid-mandatory.html`) **literally just calls `test_valid_value` 20 times**, and that's it. There are no special timeouts or looping/async fanciness that might be a spot where you could imagine chaos mode surfacing some real potential for edge-casey slowness. Under the "regular-configuration" (non-chaos-mode) part of that bug's test-verify run, the test completes in 70-90ms in nearly all of the runs (the exception being a single 133ms run). Whereas with chaos mode, the slowest testrun was about 100x slower, with duration `9294ms`, which is close enough to the 10s timeout-threshold that the test-verify run was considered a failure (hence the bug having been filed).