On Mon, 06 Jul 1998, "Lance A. Brown" wrote:
I am *still* getting segmentation faults from running
flist +inbox -recurse -noshow
As I was pulling 0.27 from the FTP site, 0.26 started doing this to me
again.
in 0.27e, occasionally. The core dumps show a stack trace in
__libc_free but no useful stack data from any nmh code.
Same thing I got. I recompiled 0.27 with -ggdb and ran the 0.27
flist. It worked. I ran with 0.26. It worked. The problem fixed
itself! Ack! I've got a version of flist now that I can run through
gdb the next time this happens.
As far as I can tell the seg fault is happening while reading a
context file or .mh_profile.
My strace run still shows that mine bailed out on a directory with no
messages in it, and .mh_sequences had "cur: 1".
Linux 2.1.105, glibc-2.0.6 system.
--
Brandon M. Browning, Benevolent Ruler Mail brandon+pgp for PGP key
-*- Making the Internet safer for humans, one spammer at a time. -*-
It is impossible to experience one's death objectively and still carry
a tune.
-- Woody Allen