We are not locking down our hashtables. This needs to be fixed and an total analysis done on the thread saftey of xpcom/proxy.
marking as dogfood to get attention. This will bit us soon. We should fix this asap.
Without a reproducible testcase, cannot put on PDT+. Putting on PDT- for now. But yes, you have our attention :-)
will never make it for m12.
*** Bug 22798 has been marked as a duplicate of this bug. ***
there is a test case for this bug attached to bug 22798 as well as lots of commentary that i'm too lazy to copy.
oops, i meant bug 21556. The word is that its a dup of this one but no one has actually marked it as a dup. Its a dependency.
Moving out of m13 to m14.
Hmm...I noticed you've moved it out the M14. While I may be restating the obvious, this bug completely wastes mozilla on an SMP machine. Thus I feel this is a very important bug and the number of votes for it seem to reflect that. It should be quite easy to debug (I know, I should do it myself, rather than whine) on a SMP machine. Personally I think this is required for beta and/or dogfood.
niles, bug #21556, depending on 18110, is marked PDT+. Clearing Status Whiteboard.
*** Bug 21556 has been marked as a duplicate of this bug. ***
I believe that xpcom/proxy is now relatively threadsafe. The stack crawls that I saw today on the smp machine are related to a necko bug.
ok to mark Verified per dp.
Do you know the bug number for the necko issue? I still seem to be getting crashes, although they are now MUCH less frequent, so I'd like to follow up on that one. Thanx!