GNU bug report logs - #18516
24.4.50; EMACS crashes when I try to open Elisp manual

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 18516 in the body.
You can then email your comments to 18516 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
To: bug-gnu-emacs <at> gnu.org
Cc: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Subject: 24.4.50; EMACS crashes when I try to open Elisp manual
Date: Sun, 21 Sep 2014 10:55:48 +0200

The crash does not happen with the -Q option, only with my init
file. The recipe to make the crash is

C-h
i
m
Elisp
RET

Actually, it is not the Elisp manual as such that is problematic, but
any manual bundled with Emacs causes the crash, while other manuals are
OK. I suspect that the difference is because of the gzipping of these
bundled manual.

I tried to start Emacs from gdb (the latter started from an MSYS
console) but then I got that xbacktrace is not supported. And the crash
happened earlier, just with `C-h i'.

I think I should try with a simpler init file --- to find the smallest
thing causing the crash, but I have not yet done this investigation.






In GNU Emacs 24.4.50.1 (i686-pc-mingw32)
 of 2014-09-19 on CHOUNEK
Repository revision: 117905 dgutov <at> yandex.ru-20140919034142-2etpyoq7g9fq3yfd
Windowing system distributor `Microsoft Corp.', version 5.1.2600
Configured using:
 `configure --prefix=c:/Programme/GNU/Emacs --without-jpeg
 --without-tiff --without-gif --without-png 'CPPFLAGS= -DFOR_MSW=1 -I
 C:/Programme/GNU/installation/emacs-install/libXpm-3.5.8/include -I
 C:/Programme/GNU/installation/emacs-install/libXpm-3.5.8/src -L
 c:/Programme/GNU/installation/emacs-install/libXpm-3.5.8/src'
 CPP=/mingw/bin/cpp.exe
 PKG_CONFIG_PATH=/mingw/lib/pkgconfig:/usr/local/lib/pkgconfig'

Configured features:
XPM SOUND NOTIFY ACL ZLIB

Important settings:
  value of $EMACSPATH: c:\Programme\NGNU\CVS;C:\Programme\GNU\GnuPG;c:\Programme\apache-ant-1.8.0\bin;c:\msys\1.0\bin;c:\msys\1.0\mingw\bin;
  value of $LANG: FRA
  locale-coding-system: cp1252

Major mode: Dired by name

Minor modes in effect:
  recentf-mode: t
  mail-abbrevs-mode: t
  iswitchb-mode: t
  tooltip-mode: t
  electric-indent-mode: t
  mouse-wheel-mode: t
  menu-bar-mode: t
  file-name-shadow-mode: t
  global-font-lock-mode: t
  font-lock-mode: t
  blink-cursor-mode: t
  auto-composition-mode: t
  auto-encryption-mode: t
  auto-compression-mode: t
  buffer-read-only: t
  line-number-mode: t

Recent input:
e f v a r SPC c o l ) , SPC e t SPC d e SPC p a s s 
<backspace> <backspace> <backspace> <backspace> <backspace> 
<backspace> ' a s s o c i e r SPC d y n a m i q u e 
m e n t r <backspace> SPC l e s SPC s y m b o l SPC 
r o w SPC e t SPC c o l SPC à SPC l e u r SPC a v e 
c <backspace> <backspace> <backspace> <backspace> v 
a l e u r SPC p o u r SPC l ' a f f i c h i e r <backspace> 
<backspace> <backspace> e r SPC ( <left> <left> <backspace> 
<backspace> a g f <backspace> e <end> s e s - p r i 
n c <backspace> t - c e l l ) <left> . . . <end> . 
SPC E s t r - <backspace> <backspace> - c e SPC q u 
e SPC t u SPC e r e <backspace> <backspace> <backspace> 
t e r o u <backspace> <backspace> <backspace> <backspace> 
r o u v e SPC <backspace> s SPC q u e SPC c ' e s t 
SPC u n e SPC b o n n e SPC i d é e . <return> <return> 
SPC SPC SPC V i n c e n t . C-c C-s y C-x d <M-backspace> 
<M-backspace> <M-backspace> <M-backspace> P r o <tab> 
m <tab> g n <tab> e m <tab> <return> <down> <down> 
i <down> <down> <down> <down> i <down> <down> <down> 
<down> <down> <down> <next> C-s e l i s <down> <up> 
<return> C-x k <return> M-x r e p o r t - b u <tab> 
<tab> e m <tab> <tab> <return>

Recent messages:
Sending email 
Sending email done
250 2.0.0 tkfG1o00K2maGi203kfHFS mail accepted for delivery
221 2.0.0 mwinf5d32 ME closing connection
Sending...done
Deleting article d:/msys/1.0/home/Vincent/News/drafts/drafts/175 in drafts...
Mark set [2 times]
Mark saved where search started
uncompressing elisp.info.gz...done
Making completion list...

Load-path shadows:
d:/msys/1.0/home/Vincent/.emacs.d/etc/custom hides c:/Programme/GNU/Emacs/share/emacs/24.4.50/lisp/custom
c:/Programme/GNU/Emacs/share/emacs/24.4.50/lisp/loaddefs hides c:/Programme/GNU/installation/cedet-install/cedet/lisp/cedet/loaddefs

Features:
(shadow warnings emacsbug jka-compr misearch multi-isearch dired-aux
mailalias smtpmail qp mail-extr sort gnus-bcklg gnus-cite bbdb-message
sendmail nnir gnus-draft gnus-ml nndraft nnmh nndoc nnfolder bbdb-gnus
bbdb-mua bbdb-com crm network-stream auth-source starttls gnus-agent
gnus-srvr gnus-score score-mode nnvirtual gnus-msg gnus-art mm-uu
mml2015 epg-config mm-view mml-smime smime password-cache dig mailcap
nntp gnus-cache gnus-sum gnus-group gnus-undo gnus-start gnus-cloud
nnimap nnmail mail-source tls utf7 netrc nnoo parse-time gnus-spec
gnus-int gnus-range message dired-x dired format-spec rfc822 mml mml-sec
mm-decode mm-bodies mm-encode mail-parse rfc2231 rfc2047 rfc2045
ietf-drums gmm-utils mailheader gnus-win gnus gnus-ems nnheader
gnus-util mail-utils mm-util help-fns mail-prsvr hl-line accents-ascii
eieio byte-opt bytecomp byte-compile cconv eieio-core tex-mik
preview-latex tex-site auto-loads calc-mathfloat calc-math edmacro
kmacro w32utils java-init cl-macs cl gv bsh-init recentf tree-widget
wid-edit cl-loaddefs cl-lib generic-x cc-mode cc-fonts cc-guess cc-menus
cc-cmds cc-styles cc-align cc-engine cc-vars cc-defs template mailabbrev
iswitchb cus-start cus-load bbdb easymenu bbdb-site timezone
bbdb-loaddefs calc-misc calc-arith calc-ext calc calc-loaddefs calc-macs
skeleton load-path-to-cedet-svn time-date tooltip electric uniquify
ediff-hook vc-hooks lisp-float-type mwheel dos-w32 ls-lisp
w32-common-fns disp-table w32-win w32-vars tool-bar dnd fontset image
regexp-opt fringe tabulated-list newcomment lisp-mode prog-mode register
page menu-bar rfn-eshadow timer select scroll-bar mouse jit-lock
font-lock syntax facemenu font-core frame cham georgian utf-8-lang
misc-lang vietnamese tibetan thai tai-viet lao korean japanese hebrew
greek romanian slovak czech european ethiopic indian cyrillic chinese
case-table epa-hook jka-cmpr-hook help simple abbrev minibuffer nadvice
loaddefs button faces cus-face macroexp files text-properties overlay
sha1 md5 base64 format env code-pages mule custom widget
hashtable-print-readable backquote make-network-process w32notify w32
multi-tty emacs)

Memory information:
((conses 8 307965 27511)
 (symbols 24 42644 0)
 (miscs 20 1204 784)
 (strings 16 69259 8672)
 (string-bytes 1 2999325)
 (vectors 8 33295)
 (vector-slots 4 1369977 56986)
 (floats 8 261 542)
 (intervals 28 2243 1407)
 (buffers 516 28))

Message #8 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
To: 18516 <at> debbugs.gnu.org
Subject: RE: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Sun, 21 Sep 2014 13:50:50 +0200

[Message part 1 (text/plain, inline)]

Hello,

I have made a little more investigation, trying to boil down the dozen
of lines in my init file to what ultimately causes the crash.

It occurs that this is caused by a tiny missing slash in some info path.

I made an MSDos batch file that start emacs with -Q and a few other
command line arguments ( --eval and -f ) to do the minial inits and
actions to cause the crash. So on my machine when I open a DOS console,
and I type

call bug18516_1.bat

The it lauchnes and immediately after crashes.

Hopefully that additional investigation can help you locate what is
happening.

Please make me know if you have any instruction how to get gdb working,
FYI I have this one on my machine:

-----------------------------------------------------------------------
GNU gdb (GDB) 7.4
Copyright (C) 2012 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "i686-pc-mingw32".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
-----------------------------------------------------------------------

Please believe that I remain at your disposal if you need me to do any
kind of experiments that could help investigating the problem.

VBR,
  Vincent.

[bug18516_1.bat.gz (application/octet-stream, attachment)]

Message #11 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Cc: 18516 <at> debbugs.gnu.org
Subject: Re: bug#18516: 24.4.50; EMACS crashes when I try to open Elisp manual
Date: Sun, 21 Sep 2014 18:59:25 +0300

> From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
> Date: Sun, 21 Sep 2014 13:50:50 +0200
> 
> I have made a little more investigation, trying to boil down the dozen
> of lines in my init file to what ultimately causes the crash.
> 
> It occurs that this is caused by a tiny missing slash in some info path.
> 
> I made an MSDos batch file that start emacs with -Q and a few other
> command line arguments ( --eval and -f ) to do the minial inits and
> actions to cause the crash. So on my machine when I open a DOS console,
> and I type
> 
> call bug18516_1.bat
> 
> The it lauchnes and immediately after crashes.

It doesn't crash for me.

The comments in your batch file say "it is on purpose that there is no
'/' between 'c:' and 'Invalid'", but actually there is a slash there.
Anyway, I tried both ways, and it didn't crash for me.

I don't understand why you use forward slashes in the "start" command;
did you invoke this batch file from MSYS Bash or something?

Message #14 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Cc: 18516 <at> debbugs.gnu.org
Subject: Re: bug#18516: 24.4.50; EMACS crashes when I try to open Elisp manual
Date: Sun, 21 Sep 2014 19:00:11 +0300

> From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
> Date: Sun, 21 Sep 2014 10:55:48 +0200
> Cc: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
> 
> I tried to start Emacs from gdb (the latter started from an MSYS
> console) but then I got that xbacktrace is not supported.

You need to start GDB from the src directory of the Emacs tree.

Message #17 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
To: 18516 <at> debbugs.gnu.org, Eli Zaretskii <eliz <at> gnu.org>
Subject: RE: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Sun, 21 Sep 2014 19:09:01 +0200

Hello Eli,

Answers inserted.

----------------------------------------
> Date: Sun, 21 Sep 2014 18:59:25 +0300
> From: eliz <at> gnu.org
> Subject: Re: bug#18516: 24.4.50; EMACS crashes when I try to open Elisp manual
> To: vincent.b.1 <at> hotmail.fr
> CC: 18516 <at> debbugs.gnu.org
>
> > From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
> > Date: Sun, 21 Sep 2014 13:50:50 +0200
> >
> > I have made a little more investigation, trying to boil down the dozen
> > of lines in my init file to what ultimately causes the crash.
> >
> > It occurs that this is caused by a tiny missing slash in some info path.
> >
> > I made an MSDos batch file that start emacs with -Q and a few other
> > command line arguments ( --eval and -f ) to do the minial inits and
> > actions to cause the crash. So on my machine when I open a DOS console,
> > and I type
> >
> > call bug18516_1.bat
> >
> > The it lauchnes and immediately after crashes.
>
> It doesn't crash for me.
>
> The comments in your batch file say "it is on purpose that there is no
> '/' between 'c:' and 'Invalid'", but actually there is a slash there.

Sorry, it is my mistake, this is because I tried both with and without
the forward slash to check that the missing forward slash is the
problem. BTW, c:/Invalid/path/info does not need to be an invalid path
when the forward slash is present. It just becomes invalid when you
remove this slash and just write c:Invalid/path/info.

> Anyway, I tried both ways, and it didn't crash for me.
>
> I don't understand why you use forward slashes in the "start" command;
> did you invoke this batch file from MSYS Bash or something?

No, I did call bug18516_1.bat from an MSDos console, which I started by
typing "cmd" in the start/execute menu.

I wrote forward slash just because I copied and paste the full path from
a Dired view. It does not prevent the crash to use backslashes anyway.

   Vincent.

Message #20 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Cc: 18516 <at> debbugs.gnu.org
Subject: Re: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Sun, 21 Sep 2014 20:28:54 +0300

> From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
> Date: Sun, 21 Sep 2014 19:09:01 +0200
> 
> > I don't understand why you use forward slashes in the "start" command;
> > did you invoke this batch file from MSYS Bash or something?
> 
> No, I did call bug18516_1.bat from an MSDos console, which I started by
> typing "cmd" in the start/execute menu.
> 
> I wrote forward slash just because I copied and paste the full path from
> a Dired view. It does not prevent the crash to use backslashes anyway.

Then I guess I'll have to say "not reproducible".

Perhaps if you run under GDB and show where it crashes and why, we
could understand the reasons.

Message #23 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
To: 18516 <at> debbugs.gnu.org, Eli Zaretskii <eliz <at> gnu.org>
Subject: RE: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Sun, 21 Sep 2014 20:00:35 +0200

Dear Eli,

Here is what I get with gdb after going to the src directory:

-----------------------------------------------------------------------
(gdb) target exec ../nt/runemacs.exe
(gdb) run -Q "--eval=(setq source-directory \"c:/Programme/GNU/Emacs/\")" "--eval=(eval-after-load \"Info\"  '(dolist (i '(\"c:/Programme/GNU/emacs/share/info/\" \"c:/Invalid/path/info\" )) (add-to-list 'Info-directory-list i)))" -f info "--eval=(Info-menu \"Elisp\")"
Starting program: c:\Programme\GNU\installation\emacs-install\emacs\trunk\nt\runemacs.exe -Q "--eval=(setq source-directory \"c:/Programme/GNU/Emacs/\")" "--eval=(eval-after-load \"Info\"  '(dolist (i '(\"c:/Programme/GNU/emacs/share/info/\" \"c:/Invalid/path/info\" )) (add-to-list 'Info-directory-list i)))" -f info "--eval=(Info-menu \"Elisp\")"
[New Thread 85856.0x14dc0]
[Inferior 1 (process 85856) exited normally]
-----------------------------------------------------------------------

It did not crash when started from the debugger --- it got open on the Elisp manual top page, and then
I typed 'C-x C-c' to exit. The `[Inferior 1 (process 85856) exited
normally]' output was made before I exited.

FYI, I simplified the batch causing the crash, the minimal content is now:
-----------------------------------------------------------------------
start c:\Programme\GNU\Emacs\bin\runemacs.exe -Q  "--eval=(eval-after-load \"Info\"  '(dolist (i '(\"c:/Programme/GNU/emacs/share/info/\" \"c:Invalid/path/info\" )) (add-to-list 'Info-directory-list i)))" -f info "--eval=(Info-menu \"Elisp\")"
-----------------------------------------------------------------------

I just removed from it the setting of source-directory, and used
backslashes for the program path.

FYI, it did not launch addpm.exe, and I installed on the same location
as a previous installation.

The previous installation was also a build from sources, here is its
`M-x version' output:

GNU Emacs 24.3.50.2 (i686-pc-mingw32) of 2013-08-25 on CHOUNEK

This 24.3.50.2 Emacs does *NOT* crash when I do the same experiment.

My current Emacs is

GNU Emacs 24.4.50.1 (i686-pc-mingw32) of 2014-09-19 on CHOUNEK

So, it seems that the problem is fairly recent --- well on my time
scale, which is rather slow compared to you who are active developpers
--- as it happened over the last year or so (the interval between these
versions is 390 days).

   Vincent.

----------------------------------------
> Date: Sun, 21 Sep 2014 19:00:11 +0300
> From: eliz <at> gnu.org
> Subject: Re: bug#18516: 24.4.50; EMACS crashes when I try to open Elisp manual
> To: vincent.b.1 <at> hotmail.fr
> CC: 18516 <at> debbugs.gnu.org
>
> > From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
> > Date: Sun, 21 Sep 2014 10:55:48 +0200
> > Cc: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
> >
> > I tried to start Emacs from gdb (the latter started from an MSYS
> > console) but then I got that xbacktrace is not supported.
>
> You need to start GDB from the src directory of the Emacs tree.

Message #26 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Cc: 18516 <at> debbugs.gnu.org
Subject: Re: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Sun, 21 Sep 2014 21:09:38 +0300

> From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
> Date: Sun, 21 Sep 2014 20:00:35 +0200
> 
> It did not crash when started from the debugger --- it got open on the Elisp manual top page, and then
> I typed 'C-x C-c' to exit. The `[Inferior 1 (process 85856) exited
> normally]' output was made before I exited.

So you cannot reproduce it, too.

Can you describe what happens when it crashes?  What do you see,
exactly?

Message #29 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
To: 18516 <at> debbugs.gnu.org, Eli Zaretskii <eliz <at> gnu.org>
Cc: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Subject: RE: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Sun, 21 Sep 2014 23:13:33 +0200

[Message part 1 (text/plain, inline)]

Dear Eli et alii,

Sorry for my previous email, actually I had done a mistake when trying
to launch from gdb as I had provided to `run' the slash the absence of
which is needed for crashing, that is why there was no crash with gdb!
Sorry for loosing your precious time... :-/

So, actually, I *can* make it crash under gdb.

The problem is that there is not much information provided by gdb --- no
more information than if there is no crash --- if I type:

(gdb) bt full

I get:

(gdb) No stack.

whether it is before run, after it (crash done), or before or after
clicking on the "Ne pas envoyer" (=Don't send) button of the MSW crash
popup windown.

and `bt full' does not seem to have any other effect.

and if I type

(gdb) xbacktrace

Then I get:

(gdb) Undefined command: "xbacktrace".  Try "help".

I have gdb 7.4 (maybe there are more recent ones for MSWindows XP)

What I see when the crash occurs is the usual MSWindows crash popup that
propose you to send a file that contains a dump of your memory ---
possibly containing confidential data from other applications, this is
why I never accept to do such a thing and use as much as possible free
S/W.

  Vincent.

PS: attached is a screenshot of after the crash (I only moved the
3 windows (1= gdb in console, 2=Emacs, 3=MSW crash popup) for all of
them to be visible.

[Sans titre.PNG (image/png, attachment)]

Message #32 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Cc: 18516 <at> debbugs.gnu.org
Subject: Re: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Mon, 22 Sep 2014 05:49:22 +0300

> From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
> Cc: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
> Date: Sun, 21 Sep 2014 23:13:33 +0200
> 
> So, actually, I *can* make it crash under gdb.
> 
> The problem is that there is not much information provided by gdb --- no
> more information than if there is no crash --- if I type:
> 
> (gdb) bt full
> 
> I get:
> 
> (gdb) No stack.
> 
> whether it is before run, after it (crash done), or before or after
> clicking on the "Ne pas envoyer" (=Don't send) button of the MSW crash
> popup windown.
> 
> and `bt full' does not seem to have any other effect.

That's because Emacs already exited by that time.  You need to catch
the moment when it crashes.

The problem here is that you run the wrong program under the debugger:
runemacs.exe.  You need to invoke emacs.exe instead.

Please invoke Emacs under GDB like this:

 gdb c:/Programme/GNU/Emacs/bin/emacs.exe
 (gdb) run -Q --eval=... <put here the rest of the command line>

Message #35 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
To: 18516 <at> debbugs.gnu.org, Eli Zaretskii <eliz <at> gnu.org>
Cc: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Subject: RE: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Mon, 22 Sep 2014 06:21:45 +0200

[Message part 1 (text/plain, inline)]

Dear Eli,

[...]

>
> That's because Emacs already exited by that time. You need to catch
> the moment when it crashes.
>
> The problem here is that you run the wrong program under the debugger:
> runemacs.exe. You need to invoke emacs.exe instead.
>
> Please invoke Emacs under GDB like this:
>
> gdb c:/Programme/GNU/Emacs/bin/emacs.exe
> (gdb) run -Q --eval=... <put here the rest of the command line>

Ok...

Here is attached the log of the gdb sessione encoded in raw-text coding
system, please feel free to ask me for any further step --- like setting
break points, dumping some variable contents, and so on --- if it can
help locate the problem.

BR,
  Vincent.

[bug18516_2.log (application/octet-stream, attachment)]

Message #38 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
To: 18516 <at> debbugs.gnu.org, Eli Zaretskii <eliz <at> gnu.org>
Cc: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Subject: RE: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Mon, 22 Sep 2014 06:26:17 +0200

>Dear Eli,
>

[...]

>
>Here is attached the log of the gdb sessione encoded in raw-text coding
>system, 
[...]

Just to clarify that to make this file, I did `M-x shell', then all of
the gdb stuff, and then saved the whole Emacs buffer to a log file ---
at that time I was asked by Emacs whether I want to encode some stange
char as utf-8, and I replied raw-text rather than utf-8.

  Vincent.

Message #41 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Dmitry Antipov <dmantipov <at> yandex.ru>
To: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Cc: Eli Zaretskii <eliz <at> gnu.org>, 18516 <at> debbugs.gnu.org
Subject: Re: bug#18516: Acknowledgement (24.4.50; EMACS crashes when I try
 to open Elisp manual)
Date: Mon, 22 Sep 2014 10:21:57 +0400

On 09/22/2014 08:21 AM, Vincent Belaïche wrote:

> Here is attached the log of the gdb sessione encoded in raw-text coding
> system, please feel free to ask me for any further step --- like setting
> break points, dumping some variable contents, and so on --- if it can
> help locate the problem.

Crash in Fexpand_file_name?  Hm...there was some changes in r11780[123]
and r117806.  Can you try to revert to r117800 and reproduce this crash?

Dmitry

Message #46 received at 18516-done <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Dmitry Antipov <dmantipov <at> yandex.ru>
Cc: vincent.b.1 <at> hotmail.fr, 18516-done <at> debbugs.gnu.org
Subject: Re: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Mon, 22 Sep 2014 17:39:36 +0300

> Date: Mon, 22 Sep 2014 10:21:57 +0400
> From: Dmitry Antipov <dmantipov <at> yandex.ru>
> CC: 18516 <at> debbugs.gnu.org, Eli Zaretskii <eliz <at> gnu.org>
> 
> On 09/22/2014 08:21 AM, Vincent Belaïche wrote:
> 
> > Here is attached the log of the gdb sessione encoded in raw-text coding
> > system, please feel free to ask me for any further step --- like setting
> > break points, dumping some variable contents, and so on --- if it can
> > help locate the problem.
> 
> Crash in Fexpand_file_name?  Hm...there was some changes in r11780[123]
> and r117806.  Can you try to revert to r117800 and reproduce this crash?

This was introduced in r117806 (alloca doesn't like negative arguments).

Fixed in trunk revision 117918.

Message #49 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
To: 18516 <at> debbugs.gnu.org, Eli Zaretskii <eliz <at> gnu.org>
Cc: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Subject: RE: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Mon, 22 Sep 2014 19:13:54 +0200

[Message part 1 (text/plain, inline)]

Дорогие Дмитрий, Илья, и другие,

----------------------------------------
> Date: Mon, 22 Sep 2014 17:39:36 +0300
> From: eliz <at> gnu.org
> Subject: Re: bug#18516: Acknowledgement (24.4.50; EMACS crashes when I try to open Elisp manual)
> To: dmantipov <at> yandex.ru
> CC: vincent.b.1 <at> hotmail.fr; 18516-done <at> debbugs.gnu.org
>
> > Date: Mon, 22 Sep 2014 10:21:57 +0400
> > From: Dmitry Antipov <dmantipov <at> yandex.ru>
> > CC: 18516 <at> debbugs.gnu.org, Eli Zaretskii <eliz <at> gnu.org>
> >
> > On 09/22/2014 08:21 AM, Vincent Belaïche wrote:
> >
> > > Here is attached the log of the gdb sessione encoded in raw-text coding
> > > system, please feel free to ask me for any further step --- like setting
> > > break points, dumping some variable contents, and so on --- if it can
> > > help locate the problem.
> >
> > Crash in Fexpand_file_name? Hm...there was some changes in r11780[123]
> > and r117806. Can you try to revert to r117800 and reproduce this crash?
>
> This was introduced in r117806 (alloca doesn't like negative arguments).
>
> Fixed in trunk revision 117918.

Could you please send to me the exact bzr command you would like me to
do in order to revert to whatever revision you think is of interest, and
then I can rebuild Emacs and try again to crash.

Another option would be to instrument Fexpand_file_name in order to find
when it crashes in more details, just send me a fileio.c as attachment
if you want me to do that.

After reading of Eli that the crash is within Fexpand_file_name, I have
done some more investigation how to provoke the crash with a simpler
command line, and I have found that the following arguments suffice:

-Q "--eval=(expand-file-name \"c:Invalid/path/info\" \"C:\\Programme\\GNU\\installation\\emacs-install\\emacs\\trunk\\src/\")"


Please note that:

1) C:\Programme\GNU\installation\emacs-install\emacs\trunk\src/ is the
   value of variable default-directory, and I need to pass it explicitly
   to expand-file-name for the crash to happen

2) It is on purpose that
   C:\Programme\GNU\installation\emacs-install\emacs\trunk\src/ mixes
   backward and forward slashes, it is the way it is when I launch Emacs
   with -Q and
   current directory=C:\Programme\GNU\installation\emacs-install\emacs\trunk\src

3) C:\Programme\GNU\installation\emacs-install\emacs\trunk\src is the
   directory where the C source code is found, and in which I lauch gdb
   according to Eli's recommendation.

4) c:Invalid/path/info is an invalid path, and it is on purpose that
   first slash after c: is missing.

Hopefully, this shorter settup will allow you to reproduce the
crash. Could you please try that before I have to do any bzr reverting.


VBR,
  Vincent.

PS: Attached is a gdb session log with the crash causing simpler settup.

[bug18516_3.log (application/octet-stream, attachment)]

Message #52 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Cc: 18516 <at> debbugs.gnu.org
Subject: Re: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Mon, 22 Sep 2014 20:53:16 +0300

> From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
> Cc: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
> Date: Mon, 22 Sep 2014 19:13:54 +0200
> 
> > > Crash in Fexpand_file_name? Hm...there was some changes in r11780[123]
> > > and r117806. Can you try to revert to r117800 and reproduce this crash?
> >
> > This was introduced in r117806 (alloca doesn't like negative arguments).
> >
> > Fixed in trunk revision 117918.
> 
> Could you please send to me the exact bzr command you would like me to
> do in order to revert to whatever revision you think is of interest, and
> then I can rebuild Emacs and try again to crash.

You don't need to revert anything, just resync with trunk and rebuild.
The problem should be gone.

> After reading of Eli that the crash is within Fexpand_file_name, I have
> done some more investigation how to provoke the crash with a simpler
> command line, and I have found that the following arguments suffice:
> 
> -Q "--eval=(expand-file-name \"c:Invalid/path/info\" \"C:\\Programme\\GNU\\installation\\emacs-install\\emacs\\trunk\\src/\")"

The easiest recipe is just this:

 emacs -Q
 M-: (let (default-directory) (expand-file-name "foo" "c:bar/baz")) RET

> Hopefully, this shorter settup will allow you to reproduce the
> crash.

I already did reproduce it, and I already fixed the crash in the
repository.  That's why I closed the bug.

> Could you please try that before I have to do any bzr reverting.

No need to revert anything.

Message #55 received at 18516 <at> debbugs.gnu.org (full text, mbox):

From: Vincent Belaïche <vincent.b.1 <at> hotmail.fr> 
To: Eli Zaretskii <eliz <at> gnu.org> , 18516 <at> debbugs.gnu.org
Cc: Vincent Belaïche <vincent.b.1 <at> hotmail.fr>
Subject: RE: bug#18516: Acknowledgement (24.4.50;
 EMACS crashes when I try to open Elisp manual)
Date: Tue, 23 Sep 2014 11:51:10 +0200

----------------------------------------
> Date: Mon, 22 Sep 2014 20:53:16 +0300
> From: eliz <at> gnu.org
> Subject: Re: bug#18516: Acknowledgement (24.4.50; EMACS crashes when I try to open Elisp manual)
> To: vincent.b.1 <at> hotmail.fr
> CC: 18516 <at> debbugs.gnu.org
>

[...]

>
> You don't need to revert anything, just resync with trunk and rebuild.
> The problem should be gone.
>

[...]

I did it, and the bug is gone for me also.

Thank you very much Eli for your patience & kind help.

  Vincent.

Previous Next

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