I was running a build from yesterday on Window with the mail smoketest enabled. When I get the imap password dialog, I just dismiss it and then the Application crash into void nsImapProtocol::SendSetBiffIndicatorEvent(nsMsgBiffState newState) because m_imapMiscellaneousSink was null. here is the stack trace: nsImapProtocol::SendSetBiffIndicatorEvent(int 0x00000002) line 3408 + 28 bytes nsImapProtocol::TryToLogon() line 6820 nsImapProtocol::ProcessCurrentURL() line 1155 + 8 bytes nsImapProtocol::ImapThreadMainLoop() line 981 + 14 bytes nsImapProtocol::Run(nsImapProtocol * const 0x05673e84) line 778 nsThread::Main(void * 0x043fed48) line 120 + 26 bytes _PR_NativeRunThread(void * 0x051206f8) line 433 + 13 bytes MSVCRTD! 0019c323() KERNEL32! 77e96523() my guess is we should just test m_imapMiscellaneousSink before acceessing it.

nominating nsbeta1

This crash has only shown up in talkback reports 3 times this year. So, I'm not convinced we need to pursue this yet. However, if it's just a null ptr check, then I'd take it if we thought that wouldn't cause other problems. I'm putting an nsbeta1- on this, but, would consider a fix.

By the definitions on <http://bugzilla.mozilla.org/bug_status.html#severity> and <http://bugzilla.mozilla.org/enter_bug.cgi?format=guided>, crashing and dataloss bugs are of critical or possibly higher severity. Only changing open bugs to minimize unnecessary spam. Keywords to trigger this would be crash, topcrash, topcrash+, zt4newcrash, dataloss.

Unable to reproduce this crash with Mozilla 1.7 beta. I see a similar stack trace in the long-resolved bug 103337. Not seeing any recent reports of this in our talkback or in bugzilla so resolving wfm.