Mutt Segfaults in 14.0
 

New install of 14.0 with mutt-1.5.21-i486-2; mutt segfaults. Recompiled from Slack source with SlackBuild, no joy.

Anyone else seen this? Ideas? Fix?

Run Whatever that mentions is missing. It's i486. Have you a 32 bit cpu, or a multilib 64 bit system?

That was a good idea; I'm embarrassed not to have thought of it. Alas, no libs are missing. BTW, it's a 32-bit Intel Core 2 Duo E6300 with Intel D975XBX2KR board. For now I'm happily using an older version of mutt; not sure what the new version in 14.0 actually buys me.

Maybe a few more/less fleas? :-). From 'man mutt'

did it work with 13.37? this might just be mutt's reaction to an option in the .muttrc that is not compiled in. you could compare `mutt -v` to your config file.

Hello HaroldWho,

Sometimes when I have misconfigured my muttrc, mutt does not start at all.
Have you been able to reproduce the segfaults?

On 1.5.21-x86_64 mutt has behaved well, although a while ago I had
a problem with mutt freezing. It seemed to happen randomly, or at times when
I was doing something fancy such as reading a message while logged into a
remote mail server, then following a url in the message, then closing the
browser and resuming my session with mutt. These were not segfaults, however.
Mutt would not exit and dump its memory. I would have to chvt to restart.

I haven't used mutt-1.5.21 before. as I had Slack 13.1. As for the .muttrc config file, mutt only complains about the "alternates" variable. Still segfaults without any further messages if the "alternates" var is commented out.

I d/l'd the source from mutt.org and configured it w/o POP, NLS, or IMAP. Compiles but still segfaults. I'm gonna try recompiling with S-lang instead of ncurses.

Since I have a working version of mutt I'm not going down this road too far though.

Using Mutt on 14.0 32 Bit and 64 Bit (64 Bit now updated to -current) without any problems.

Nor any problems here.

HaroldWho, in the last few days I have changed muttrc and have managed to produce some segfaults.
What I do not understand is why you must recompile. For me, segfault ends when I
revert whatever change I've made that produces a segfault.

From what I've seen here so far I need to pay closer attention to my ~/.muttrc file. BTW, mutt seems to ignore the -F field to specify another config file. Only seems to look for ~/.muttrc or /etc/mutt/Muttrc.

I recompiled because I don't need POP, NLS, or IMAP in Mutt. BTW, segfaults whether compiled against ncurses or S-lang.

Here's what the syslog shows after a segfault Means nothing to me:

I still have a few ideas to try; then it stick with 1.4.2.3i

Thanks for all the ideas.

Just tried that option, and mutt did not ignore it. After mutt started it was configured exactly as specified.

Thanks for keeping us posted.

What does the +nls add?

OK, the segfault problem is due to some unrecognized/screwed-up variable in my old ~/.muttrc file. Mutt complains about the "alternates" variable, but something is still there that just causes a segfault. I built a quick-and-dirty Muttrc using http://muttrcbuilder.org/page (a very interesting site BTW) and no segfault. Downside: appearance not at all what I want, and I doubt I'll spend the time on a new Muttrc when mu old Mutt works fine.

BTW, all NLS gives you is support for a lot of non-English languages.

Again, thank you all for the ideas and suggestions.

Nice site. The muttrcbuilder is very interesting.
(edit link?)

Muttrc does do more than set variables. For example, read the mutt documentation about Alternative Addresses.
Confusingly, "alternates" isn't a variable.

You can also read the same documentation when using mutt by pressing F1 -if that key is not already mapped.