1395785 - browser_586068-apptabs_ondemand.js, browser_unrestored_crashedTabs.js, and browser_bug1303838.js are going to permafail on Linux when Gecko 57 merges to Beta on 2017-09-20

Status: VERIFIED WORKSFORME

(Firefox :: Session Restore, defect)

Description

[Ryan VanderMeulen [:RyanVM]]

[Tracking Requested - why for this release]: Permafailing Linux tests on the next merge day.

Linux-only from what I can tell. Bug 1373814 looks suspicious by virtue of it being the only thing to touch SessionStore.jsm recently, but I'll try to bisect it to a real culprit before setting deps and needinfoing people.

Interestingly, the browser_bug1303838.js failures are also consistently showing up in all the failing logs.

https://treeherder.mozilla.org/logviewer.html#?job_id=127569911&repo=try

TEST-UNEXPECTED-FAIL | browser/components/sessionstore/test/browser_586068-apptabs_ondemand.js | uncaught exception - TypeError: url is undefined at prepareConnectionToHost@resource:///modules/sessionstore/SessionStore.jsm:3443:9
speculativeConnectOnTabHover@resource:///modules/sessionstore/SessionStore.jsm:3463:22
speculativeConnectOnTabHover@resource:///modules/sessionstore/SessionStore.jsm:346:5
_mouseenter@chrome://browser/content/tabbrowser.xml:7689:11
onxblmouseover@chrome://browser/content/tabbrowser.xml:7840:9

TEST-PASS | dom/base/test/browser_bug1303838.js | check tabs.length - 
TEST-PASS | dom/base/test/browser_bug1303838.js | check selectedTab - 
Waiting for 1 LocationChange
Buffered messages logged at 23:10:15
Clicking #frame-link-3
LocationChange: data:text/html;charset=utf-8,ibaz
TEST-PASS | dom/base/test/browser_bug1303838.js | check tabs.length - 
TEST-PASS | dom/base/test/browser_bug1303838.js | check selectedTab - 
Leaving test bound 
Buffered messages finished
TEST-UNEXPECTED-FAIL | dom/base/test/browser_bug1303838.js | This test exceeded the timeout threshold. It should be rewritten or split up. If that's not possible, use requestLongerTimeout(N), but only as a last resort.

Comment 1

[Ryan VanderMeulen [:RyanVM]]

Ugh, bisection on Try is pointing at bug 1393900 as the culprit (at least that seems more likely than some jemalloc changes). Mark, I'm wondering if there's similar issues at play to what you ran into with the Places test reorderings?

https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=96bea603414c&tochange=d87bd54dc409

Comment 2

[Mark Banner (:standard8)]

Sorry, I've not really got any ideas here. I don't think I was hitting timeout limits.

Comment 3

[Ryan VanderMeulen [:RyanVM]]

Mike, can you maybe help me find an owner for this? I'd prefer to not resort to disabling if at all possible.

Comment 4

[Ryan VanderMeulen [:RyanVM]]

Also affects browser_unrestored_crashedTabs.js.

Comment 5

[Ryan VanderMeulen [:RyanVM]]

Re-assuringly, this appears to have gone away on its own. I'll reopen the bug if it comes back.