1. Go to a page requiring http authentication such as the one cited above 2. Fill in any username and password (valid or not, doesn't matter) 3. Click on OK Get the following crash. _free_dbg_lk(void * 0x034d2150, int 1) line 1033 + 60 bytes _free_dbg(void * 0x034d2150, int 1) line 970 + 13 bytes free(void * 0x034d2150) line 926 + 11 bytes PR_Free(void * 0x034d2150) line 66 + 10 bytes nsAllocatorImpl::Free(nsAllocatorImpl * const 0x00c64b50, void * 0x034d2150) line 94 + 10 bytes nsAllocator::Free(void * 0x034d2150) line 161 nsBasicAuth::Authenticate(nsBasicAuth * const 0x02d53100, nsIURI * 0x02e66340, const char * 0x014dda20, const char * 0x02d53180, const unsigned short * 0x034d2030, const unsigned short * 0x034d3cf0, nsIPrompt * 0x02e27be0, char * * 0x0012f794) line 84 + 9 bytes nsHTTPChannel::Authenticate(const char * 0x02d53180, int 0) line 1907 + 151 bytes nsHTTPChannel::ProcessAuthentication(int 401) line 2283 + 31 bytes nsHTTPChannel::ProcessStatusCode() line 2105 + 12 bytes nsHTTPChannel::FinishedResponseHeaders() line 1996 + 8 bytes nsHTTPServerListener::FinishedResponseHeaders() line 1063 + 11 bytes nsHTTPServerListener::OnDataAvailable(nsHTTPServerListener * const 0x03115090, nsIChannel * 0x02e35ce4, nsISupports * 0x02e624c0, nsIInputStream * 0x030b124c, unsigned int 0, unsigned int 0) line 405 + 8 bytes nsOnDataAvailableEvent::HandleEvent(nsOnDataAvailableEvent * const 0x03122cd0) line 406 + 47 bytes nsStreamListenerEvent::HandlePLEvent(PLEvent * 0x03122e70) line 97 + 12 bytes PL_HandleEvent(PLEvent * 0x03122e70) line 575 + 10 bytes PL_ProcessPendingEvents(PLEventQueue * 0x010f4160) line 520 + 9 bytes _md_EventReceiverProc(HWND__ * 0x00bc05d2, unsigned int 49420, unsigned int 0, long 17776992) line 1032 + 9 bytes USER32! 77e71268() 010f4
Nominating for dogfood since this prevents visiting any sites that require http authentication.
Probably related to bug 41404. cc shaver.
Agreed. (And I have a fix for 41404 now.) *** This bug has been marked as a duplicate of 41404 ***
No, it's not a dup. This crash occurs when clicking OK and 41404 talks about a crash when clicking cancel. And the stack traces of the two crashes are different. See bug 41466 for the cancel stack trace. So keep them as two separate bugs and if the same code-change fixes both, so much the better. But if we dup them and the code change for the cancel case doesn't fix this case, then we have lost a bug report. And this one is definitely dogfood (I hope pdt agrees) whereas the cancel crash is not.
Reassigning to shaver because of the similarity to bug 41404
18 years ago
Touche'. I don't think the fix for 41404 will get this one right now, but I'll get a patch together that fixes both. (cvs.mozilla.org isn't playing nice right now, for some reason.)
*** Bug 41517 has been marked as a duplicate of this bug. ***
verified: winNT 2000070508 Linux 2000070420 Mac OS8.6 2000070511