Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uwm.edu!ogicse!schaefer
From: schaefer@ogicse.ogi.edu (Barton E. Schaefer)
Newsgroups: comp.mail.mush
Subject: Re: 'longjmp botch' in 7.0.4 on 4.3Tahoe
Message-ID: <7843@ogicse.ogi.edu>
Date: 9 Mar 90 02:14:52 GMT
References: <253@elrond.locus.com>
Organization: Oregon Graduate Institute (formerly OGC), Beaverton, OR
Lines: 29

In article <253@elrond.locus.com> ghb@elrond.locus.com (George Bray) writes:
} 
} If I start to reply to a message, and decide I don't really want to,
} and hit ^C, I sometimes get "longjmp botch" and a core dump.

I've just applied a patch that should considerably narrow the hole in
which this can happen.  An *UN*official patch you can make to mail.c,
line 1733:

  1732      /* now check whether we should abort the letter */
  1733      if (sig > 0 && !killme) {

becomes:

  1733      if (sig > 0 && !killme && ison(glob_flags, IS_GETTING)) {

This will be official whenever all the other changes are ...

} I have also seen a longjmp botch if I ^Z (suspend) mush when it isn't
} at a prompt (like while it's updating a folder) and then resume it.

I have no idea how this could happen, as none of the stop/start code
does any longjmping whatsoever.  If you see it again, please try to
tell us more specifically what the circumstances are.
-- 
Bart Schaefer          "EARTH: Surrender IMMEDIATELY or we PICKLE DAN QUAYLE"

                                                                    "THPPFT!"
schaefer@cse.ogi.edu (used to be cse.ogc.edu)