+++ This bug was initially created as a clone of Bug #454412 +++ Started with: http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1221025516.1221032273.8436.gz Linux comm-central dep unit test on 2008/09/09 22:45:16

Looks like you're right, I don't see a metacity process on that machine any more - looks like some patch managed to crash the window manager on tests. In the future, could you please give me some pointers in comment #0 of the bug what exactly is "bug xxx symptoms"? And a unit test machine specific thing doesn't block a release.

I started metacity again. As mochichrome was running right at that moment, I saw some messages spit out by metacity when our tests are running: Window manager warning: Buggy client sent a _NET_ACTIVE_WINDOW message with a timestamp of 0 for 0x200081 (Chrome Tes) Window manager warning: meta_window_activate called by a pager with a 0 timestamp; the pager needs to be fixed. Window manager warning: Buggy client sent a _NET_ACTIVE_WINDOW message with a timestamp of 0 for 0x200081 (SeaMonkey) Window manager warning: meta_window_activate called by a pager with a 0 timestamp; the pager needs to be fixed. (lines like these repeat over and over) There might be some bug in Mozilla code causing this, might be worth to look into.

marking FIXED as mochichrome succeeded (I started metacity somewhere at the beginning of that, so mochitest still has those failures in that cycle, should be good next cycle unless we have the other issue of flaky tests again).

(In reply to comment #2) > There might be some bug in Mozilla code causing this, might be worth to look into. Bug 449723 ;-) (In reply to comment #3) > marking FIXED as mochichrome succeeded It took 2-3 more cycles to fully disappear :-/ But it's gone eventually :-) V.Fixed