#[markdown(off)]
Filed by: nbeleuzu [at] mozilla.com

https://treeherder.mozilla.org/logviewer.html#?job_id=224772272&repo=autoland

https://queue.taskcluster.net/v1/task/Vw35gwpmTi6TvvQImUyC4w/runs/0/artifacts/public/logs/live_backing.log

10:13:36 INFO - GECKO(1947) | --DOCSHELL 0x11c786800 == 3 [pid = 1947] [id = {f0eaec0e-3f6a-fe40-97b4-70253766d734}]
10:13:36 INFO - GECKO(1947) | --DOCSHELL 0x123faf800 == 2 [pid = 1947] [id = {38be86b8-ef19-ec4f-a280-5b4a359d81a5}]
10:13:36 INFO - GECKO(1947) | --DOMWINDOW == 7 (0x12907c000) [pid = 1947] [serial = 11] [outer = 0x0] [url = about:blank]
10:13:36 INFO - GECKO(1947) | --DOMWINDOW == 6 (0x1190ddc00) [pid = 1947] [serial = 1] [outer = 0x0] [url = chrome://browser/content/hiddenWindow.xul]
10:13:36 INFO - GECKO(1947) | --DOCSHELL 0x11b8cb000 == 1 [pid = 1947] [id = {0de519a1-7de3-ce43-98c6-926cce2ccfde}]
10:13:36 INFO - GECKO(1947) | --DOMWINDOW == 5 (0x124f34000) [pid = 1947] [serial = 6] [outer = 0x0] [url = chrome://extensions/content/dummy.xul]
10:13:36 INFO - GECKO(1947) | --DOCSHELL 0x124f70800 == 0 [pid = 1947] [id = {39d32784-d7e2-3245-9225-452a499b4c64}]
10:13:36 INFO - GECKO(1947) | --DOMWINDOW == 4 (0x125532400) [pid = 1947] [serial = 8] [outer = 0x0] [url = chrome://extensions/content/dummy.xul]
10:13:36 INFO - GECKO(1947) | --DOMWINDOW == 3 (0x1190de800) [pid = 1947] [serial = 2] [outer = 0x0] [url = about:blank]
10:13:36 INFO - GECKO(1947) | --DOMWINDOW == 2 (0x124731400) [pid = 1947] [serial = 5] [outer = 0x0] [url = about:blank]
10:13:36 INFO - GECKO(1947) | --DOMWINDOW == 1 (0x13471e800) [pid = 1947] [serial = 33] [outer = 0x0] [url = about:blank]
10:13:36 INFO - GECKO(1947) | --DOMWINDOW == 0 (0x11b459c00) [pid = 1947] [serial = 4] [outer = 0x0] [url = about:blank]
10:13:36 INFO - GECKO(1947) | nsStringStats
10:13:36 INFO - GECKO(1947) | => mAllocCount: 98444
10:13:36 INFO - GECKO(1947) | => mReallocCount: 0
10:13:36 INFO - GECKO(1947) | => mFreeCount: 98444
10:13:36 INFO - GECKO(1947) | => mShareCount: 121552
10:13:36 INFO - GECKO(1947) | => mAdoptCount: 2662
10:13:36 INFO - GECKO(1947) | => mAdoptFreeCount: 2702
10:13:36 INFO - GECKO(1947) | => Process ID: 1947, Thread ID: 140735267107584
10:13:36 INFO - TEST-INFO | Main app process: exit 0
10:13:36 ERROR - TEST-UNEXPECTED-FAIL | browser/components/urlbar/tests/browser/browser_UrlbarInput_unit.js | leaked 1 window(s) until shutdown [url = about:blank]
10:13:36 ERROR - TEST-UNEXPECTED-FAIL | browser/components/urlbar/tests/browser/browser_UrlbarInput_unit.js | leaked 1 window(s) until shutdown [url = chrome://mochitests/content/browser/browser/components/urlbar/tests/browser/empty.xul]
10:13:36 INFO - TEST-INFO | browser/components/urlbar/tests/browser/browser_UrlbarInput_unit.js | windows(s) leaked: [pid = 1947] [serial = 31], [pid = 1947] [serial = 30]
10:13:36 ERROR - TEST-UNEXPECTED-FAIL | browser/components/urlbar/tests/browser/browser_UrlbarInput_unit.js | leaked 1 docShell(s) until shutdown
10:13:36 INFO - TEST-INFO | browser/components/urlbar/tests/browser/browser_UrlbarInput_unit.js | docShell(s) leaked: [pid = 1947] [id = {a176af5c-4f3b-8f43-b874-8c54a705563c}]

This seems to have started here:
https://treeherder.mozilla.org/#/jobs?repo=autoland&searchStr=test-verify&fromchange=22ec021f1f92a342023765975e0ab9cffa85baef&tochange=905f08f4eab0a91d06092dcf4fea9cb69f090c23&selectedJob=224765474

And eventually was fixed with a backout:
https://hg.mozilla.org/integration/autoland/rev/9865a7358969d86d1b657ba2b9890e4ea5189f11
+
https://hg.mozilla.org/integration/autoland/rev/2b73321e629d

could be due to the test itself failing. Especially if the relanding didn't trigger the same failure. Drew?

I never relanded the input tests after they were backed out, so that's why this bug isn't happening any more. The failure though was in the controller test, not this input test, so I still need to fix whatever's leaking.

I'll mark this bug as fixed by the backout and see if I can fix the leak before relanding.

This test is leaking for me locally on a debug build using a fresh m-c tree even without my patch. It leaks every time I run it. I wonder why this bug was only filed now, i.e., why it didn't show up on our infrastructure until now?

Reopening because there seems to be an actual leak in this test.

browser_UrlbarInput_unit.js leaks the empty.xul window every time I run it on a debug build. I narrowed down the leak to the objects/properties that this patch nulls out after closing the window. Simply cleaning up these properties in a register-cleanup function isn't enough to stop the leak. It happens after the test creates either one of its two UrlbarInputs. Cleaning up the properties after each UrlbarInput is created doesn't stop the leak either. Cleaning up the properties plus opening a new window for each UrlbarInput does stop the leak though, so that's what this patch does.

However, the leakcheck test still logs UNEXPECTED-FAIL leakcheck: default 1392 bytes leaked at the end, which it does currently without this patch. That doesn't cause the test to fail though, unlike the leaks reported in this bug.

https://hg.mozilla.org/mozilla-central/rev/e688af125f98