GNU bug report logs - #716
23.0.60; opening tgz file causes emacs crash

Previous Next

Full log

Message #12 received at 716 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Jason Rumney <jasonrumney <at> gmail.com>
To: robert marshall <robert.marshall <at> tnei.co.uk>,
        716 <at> debbugs.gnu.org
Subject: Re: bug#716: 23.0.60; opening tgz file causes emacs crash
Date: Fri, 22 Aug 2008 20:10:44 +0800

robert marshall wrote:
> If I open a tgz (tar and gzipped file) in emacs it immediately crashes 
> (giving the windows emacs abort dialog)

In trying to debug this, I think I have found where it is going wrong, 
but I have no idea why and how to debug the relevant bytecode. It could 
be a sign of the byte compilation going wrong on Windows (line-end or 
other coding problems?), or a bug in Fbytecode somewhere (which is only 
affecting Windows for some reason). The stack trace when debugging 
includes the following:

#8  0x010a5e65 in Finsert (nargs=2, args=0x0) at editfns.c:2224
#9  0x0115795b in Fbyte_code (bytestr=48986435, vector=49414916, 
maxdepth=48)
   at bytecode.c:1265
#10 0x01023232 in funcall_lambda (fun=49839780, nargs=0, 
arg_vector=0x82d854)
   at eval.c:3229
#11 0x01022d11 in Ffuncall (nargs=1, args=0x82d850) at eval.c:3088

In frame #11, args[0] is tar-summarize-buffer, so at that point all 
appears normal.
By frame #8 things have clearly gone wrong. How did args become a NULL 
pointer, yet there are 2 args?

Currently I don't have access to a GNU/Linux machine to try a copy of 
tar-mode.elc compiled there. Can someone else with access to both 
platforms try that? To reproduce the bug, you need to open a tar file in 
Emacs (trunk) maybe a few times (the most I have had to open one before 
triggering the bug is 3 times, but often it happens first time).

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.