[Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.9.2a1pre) Gecko/20090318 Minefield/3.6a1pre] (home, optim default) (W2Ksp4) (http://hg.mozilla.org/mozilla-central/rev/6f21cfb3a12d) No leak. [Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.9.2a1pre) Gecko/20090322 Minefield/3.6a1pre] (home, optim default) (W2Ksp4) (http://hg.mozilla.org/mozilla-central/rev/8d2c566f3256) Leaks. Regression timeframe: "Should" be one of bug 483626, bug 453185 or bug 421611.

Confirmed by { http://tinderbox.mozilla.org/showlog.cgi?log=Thunderbird/1238183877.1238186339.30647.gz&fulltext=1 Linux comm-central check on 2009/03/27 12:57:57 }

Confirmed by { http://tinderbox.mozilla.org/showlog.cgi?log=Thunderbird3.0/1238257155.1238260151.7894.gz&fulltext=1 MacOSX 10.4 comm-1.9.1 check on 2009/03/28 09:19:15 }

Really needs an owner if we're going to consider it wanted... If you get a patch, please nominate it for approval.

I came across this leak working on bug 411511 and thought at first it was my fault :) It turns out to be caused by set_environment added in bug 421611. See comment at http://mxr.mozilla.org/mozilla-central/source/xpcom/threads/nsEnvironment.cpp#135: * Environment strings must have static duration; We're gonna leak all of this * at shutdown: this is by design, caused how Unix/Linux implement environment * vars.

Thanks for the analysis! R.WontFix, like bug 203764.