User-Agent: Mozilla/5.0 (X11; U; Linux i686; fr-FR; rv:1.7b) Gecko/20040421 Build Identifier: Mozilla/5.0 (X11; U; Linux i686; fr-FR; rv:1.7b) Gecko/20040421 If I launch mozilla and enter www.toolinux.com, it receives data from www.toolinux.com and then crash with a seg fault when trying to display. If I try to access www.toolinux.com with a hidden tab (with open in a new tab), it freezes for 4 secs and crash. Reproducible: Always Steps to Reproduce: 1.launch mozilla 2.trying to display www.toolinux.com 3. Actual Results: segmentation fault
It doesn't crash on my Linux (2004050105) although the web is very slow and complex.
can you post a Talkback ID (using Firefox latest nightly build or Mozilla 1.7RC) ? If not, please run mozilla in a a debugger: "mozilla -g" Once in GDB, start Mozilla with "GDB> run" When Mozilla crashes, back in GDB, type "GDB> stacktrace" and attach the output here (via "create a new attachment").
Created attachment 147501 [details] Output when using gdb I am using Mozilla 1.7b Mozilla/5.0 (X11; U; Linux i686; fr-FR; rv:1.7b) Gecko/20040421
That stack is obviously bogus, since UseComputedHeight does not call any functions that could possibly call UseComputedHeight (and furthermore, because UseComputedHeight could not possibly crash unless getenv() crashes, in which case the crash wouldn't be in UseComputedHeight). Olivier, you should know by now that stacks gotten from nigtlies via gdb are utterly useless, since the binaries are stripped... The only reasonable way to get a stack out of a nightly is talkback -- that's why we have it.
> Olivier, you should know by now that stacks gotten from nigtlies via gdb are > utterly useless, since the binaries are stripped... Boris, I know that, that's why I ask for a Talkback ID first, then again, there are a few cases where the crash is located outside Mozilla and it's often a quick way to distinguish a Flash, mplayerplug-in crasher for e.g., thus avoid wasting your time trying to reproduce the crash. :) Now, if the stacks can even be bogus (I didn't know that could happen) then I'll stop asking for GDB stack traces.
WFM: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7) Gecko/20040601 WFM: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8a2) Gecko/20040601