618365 - 3 "WARNING: Failed to get XPConnect?!: 'rts', file ../../../xpcom/base/nsCycleCollector.cpp, line 3415"

Status: VERIFIED FIXED

(Core :: XPCOM, defect)

Description

[Serge Gautherie (:sgautherie)]

(First noticed in bug 617967.)

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1292004418.1292005837.8018.gz&fulltext=1
Linux mozilla-central leak test build on 2010/12/10 10:06:58
http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1291989887.1291993512.15084.gz&fulltext=1
OS X 10.5 comm-central-trunk leak test build on 2010/12/10 06:04:47
{
WARNING: Failed to get XPConnect?!: 'rts', file ../../../xpcom/base/nsCycleCollector.cpp, line 3415
WARNING: Failed to get XPConnect?!: 'rts', file ../../../xpcom/base/nsCycleCollector.cpp, line 3415
WARNING: Failed to get XPConnect?!: 'rts', file ../../../xpcom/base/nsCycleCollector.cpp, line 3415
}

This warning was added in bug 580096.

Comment 1

[Ben Turner (not reading bugmail, use the needinfo flag!)]

Hm. So the only way this happens is if the process shuts down before our first cycle collection, looks like on XPCOM registration only (passing --register on the command line) and on some of the C++ unit tests. It's not a problem in that case (who cares if we don't CC - we're shutting down), but it would be a problem if it ever happened on a real browser run. I don't know if I want to make that warning go away or not.

Not a major problem in any case, I certainly wouldn't block on it.

Comment 2

[Andrew McCreight [:mccr8]]

The cycle collector no longer produces a warning if it fails to get the JS runtime when calling the GC.

Comment 3

[Serge Gautherie (:sgautherie)]

Right: bug 626768 replaced "Failed to get XPConnect?!".


https://tbpl.mozilla.org/php/getParsedLog.php?id=11974527&tree=Firefox&full=1
Linux mozilla-central leak test build on 2012-05-22 21:26:29 PDT for push 36e938e51481

No "Failed to get JS runtime!" nor "Cannot cycle collect if GC has not run first!" :-)

V.Fixed