Re: Segmentation fault
Hello Alex,
> Yesterday it happened again (see the attached log file) and now
> I'm running two instances of the patched Mutt, one in gdb and the
> other in valgrind (though the two are connected to different IMAP
> servers).
I think the imap server doesn't matter, it is more a question of your
config (sloppy threads) and the way you are using mutt to trigger the
segfault that fast. Maybe also the machine you're running on. It pops up
faster if you use a busy machine for example.
> ==31957== Invalid read of size 1
> ==31957== at 0x1B901A63: strcmp (mac_replace_strmem.c:250)
> ==31957== by 0x807153D: hash_find_hash (hash.c:108)
> ==31957== by 0x80A24B1: mutt_sort_threads (thread.c:921)
> ==31957== by 0x80A0519: mutt_sort_headers (sort.c:291)
> ==31957== by 0x805F5F4: update_index (curs_main.c:317)
> ==31957== by 0x8064BC3: mutt_index_menu (curs_main.c:492)
> ==31957== by 0x8077A85: main (main.c:934)
is this the only invalid read? Because if it is on a long term valgrind
usage scenario than we have one of the segfault reasons eliminated.
Btw. at the moment only stuff with 'invalid read of size bla' is
interesting to find the remaining bug(s). - But I think that I already
mentioned that.
Thomas