https://treeherder.mozilla.org/logviewer.html#?job_id=114602723&repo=autoland Tons of this for almost 2 hours: 11:18:17 INFO - --DOMWINDOW == 70 (0000002D0F0C6000) [pid = 4260] [serial = 52015] [outer = 0000000000000000] [url = file:///C:/slave/test/build/tests/reftest/tests/gfx/tests/crashtests/783041-2.html] 11:18:17 INFO - --DOMWINDOW == 69 (0000002D0F0CB800) [pid = 4260] [serial = 52016] [outer = 0000000000000000] [url = file:///C:/slave/test/build/tests/reftest/tests/gfx/tests/crashtests/783041-2.html] 11:18:17 INFO - --DOMWINDOW == 68 (0000002D0F0D6800) [pid = 4260] [serial = 52017] [outer = 0000000000000000] [url = file:///C:/slave/test/build/tests/reftest/tests/gfx/tests/crashtests/783041-2.html] 11:18:17 INFO - ++DOMWINDOW == 69 (0000002D06D1B800) [pid = 4260] [serial = 52064] [outer = 0000002D0A21F000] 11:18:17 INFO - ++DOMWINDOW == 70 (0000002D07924000) [pid = 4260] [serial = 52065] [outer = 0000002D0A21F000]

Two more runs with the issue: https://treeherder.mozilla.org/logviewer.html#?job_id=114595171&repo=mozilla-central https://treeherder.mozilla.org/logviewer.html#?job_id=114617440&repo=mozilla-central Milan, please take a look at this or redirect. Thank you.

Comment on attachment 8887207 [details] Bug 1381283: The original test is meant to simulate reloading until the crash. Now that we're not crashing, we limit it to 10 reloads, to avoid any kind of a race condition and not terminating the test. https://reviewboard.mozilla.org/r/157982/#review163466

fwiw, lots more failure reports in bug 1204281. The earliest I have found is https://treeherder.mozilla.org/#/jobs?repo=mozilla-inbound&revision=7745e88c749719bc64077ebb7e940ffe25d17b88, but it was so infrequent at that time it is hard to get an accurate regression range.

Pushed by msreckovic@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/7a8c2db428e0 The original test is meant to simulate reloading until the crash. Now that we're not crashing, we limit it to 10 reloads, to avoid any kind of a race condition and not terminating the test. r=aosmond

(In reply to Geoff Brown [:gbrown] from comment #5) > fwiw, lots more failure reports in bug 1204281. The earliest I have found is > https://treeherder.mozilla.org/#/jobs?repo=mozilla- > inbound&revision=7745e88c749719bc64077ebb7e940ffe25d17b88, but it was so > infrequent at that time it is hard to get an accurate regression range. Indeed, it looks like that is when this started. See https://bugzilla.mozilla.org/show_bug.cgi?id=1362903#c20 and https://bugzilla.mozilla.org/show_bug.cgi?id=1381839#c4.

(In reply to Pulsebot from comment #6) > Pushed by msreckovic@mozilla.com: > https://hg.mozilla.org/integration/autoland/rev/7a8c2db428e0 > The original test is meant to simulate reloading until the crash. Now that > we're not crashing, we limit it to 10 reloads, to avoid any kind of a race > condition and not terminating the test. r=aosmond :milan -- I don't see any calls to reload() in your patch. Shouldn't MaybeReload() sometimes call reload()?

(In reply to Geoff Brown [:gbrown] (pto July 21, July 24) from comment #8) > ... > > :milan -- I don't see any calls to reload() in your patch. Shouldn't > MaybeReload() sometimes call reload()? Right - changing the value of location.search reloads. It appears to be what the standard allows ("Can be set, to navigate to the same URL with a changed query...") and seems to happen: ... [task 2017-07-18T17:46:37.167591Z] 17:46:37 INFO - --DOMWINDOW == 45 (0x7fd9d1ab7800) [pid = 1013] [serial = 2500] [outer = (nil)] [url = file:///home/worker/workspace/build/tests/reftest/tests/gfx/tests/crashtests/783041-2.html] [task 2017-07-18T17:46:37.168232Z] 17:46:37 INFO - --DOMWINDOW == 44 (0x7fd9995e9000) [pid = 1013] [serial = 2501] [outer = (nil)] [url = file:///home/worker/workspace/build/tests/reftest/tests/gfx/tests/crashtests/783041-2.html?10] [task 2017-07-18T17:46:37.168406Z] 17:46:37 INFO - --DOMWINDOW == 43 (0x7fd9ce460800) [pid = 1013] [serial = 2502] [outer = (nil)] [url = file:///home/worker/workspace/build/tests/reftest/tests/gfx/tests/crashtests/783041-2.html?9] ... although we still actually have a hang (see bug 1381933.)

Conversation continued in bug 1381933