Re: segfault causes system freeze

[steve]

Le 23-11-2018, à 21:31:25 +0100, Jörg Sommer a écrit :


steve hat am Do 22. Nov, 16:19 (+0100) geschrieben:

Le 22-11-2018, à 16:09:50 +0100, Jörg Sommer a écrit :

> > #0  0x5592c9a59c68 in index_make_entry (s=0x7fff20afb090 "", l=1024, 
menu=, num=) at ../../curs_main.c:300
> > h = 0x5592cbef1670
> > flag = (MUTT_FORMAT_TREE | MUTT_FORMAT_MAKEPRINT | 
MUTT_FORMAT_ARROWCURSOR | MUTT_FORMAT_INDEX)
> > edgemsgno = 40
> > reverse = 
> > tmp = 
>
> This is here
> https://sources.debian.org/src/mutt/1.7.2-1+deb9u1/curs_main.c/#L300 and
> in the current version (of neomutt) here
> https://salsa.debian.org/mutt-team/neomutt/blob/master/curs_main.c#L550
>
> The best would be to create a bug report and if you don't care too much, I
> can send the developer the coredump in a *private* mail. He might dig out
> something more.

That would be really nice of you. But does it mean I also have to open a
bug report?


I think that's the best to do. Do you have reportbug installed? This eases
the creation of bug reports.


Yes I have. Problem, it asks me if I really want to report a bug since
my version is said to be outdated. Moreover, looking at the already
reported bug, some might be the same, or very close, as mine. And other
problem, I can not reproduce it at will. So not enough information yet.
And as Kevin said, it might be related to offlineimap. 


With therefore wait a bit and try to collect more info before opening a
bug.

Re: segfault causes system freeze

[steve]

Le 23-11-2018, à 09:52:59 -0800, Kevin J. McCarthy a écrit :


On Fri, Nov 23, 2018 at 07:30:00AM +0100, steve wrote:

I think it's when I just go back from editing with vim to the pager.


Also, are you using IMAP?


Yes I do via offlineimap.


It sounds like this may be triggered by offlineimap updating while you 
are in the middle of composing the message.


NeoMutt checks for new mail in the pager too, but it looks like in 
1.7.2 they botched the redraw-data updates, setting "max" before the 
index data structures were updated.


It looks like this is fixed in the latest version link posted by Jörg. 
However, I think it would be tricky to get Debian/Ubuntu to patch this 
for a non-security issue.  Your best bet would be to update.


You mean compiling myself the last git version? I'd like to stick to
Debian's version if possible.


Another workaround might be turning off $pager_index_lines.


Currently it's set to
pager_index_lines=9
which I find useful. 


Since the problem hasn't occurred for the 4 last days, I'll wait a bit
before your suggested workaround.

Thanks a lot for your help.

Have a nice weekend.

Steve