1:50 a.m.
Here is what happens with with xemacs version r21-5-latest-beta. This
happens when starting xemacs built on either of two different computers.
It also happens when built with gnome+gtk, but it takes a little longer,
that is the window comes up and lets you do a bit before crashing.
Lisp backtrace follows:
# bind (frame-being-created)
make-frame(nil #<x-device on ":0.0" 0x2865>)
frame-initialize()
# bind (debugger debug-on-error command-line-args-left)
command-line()
# (unwind-protect ...)
normal-top-level()
# (condition-case ... . error)
# (catch top-level ...)
zsh: 14675 segmentation fault (core dumped) xemacs
#0 0x404d6a41 in kill () from /lib/libc.so.6
#1 0x080ba9c3 in fatal_error_signal ()
#2 0x404d6898 in killpg () from /lib/libc.so.6
#3 0x40268d32 in XtInitializeWidgetClass () from
/usr/X11R6/lib/libXt.so.6
#4 0x40269bd9 in XtCreatePopupShell () from /usr/X11R6/lib/libXt.so.6
#5 0x081c8c80 in x_create_widgets ()
#6 0x081ca351 in x_init_frame_1 ()
#7 0x08130673 in Fmake_frame ()
#8 0x080beaab in Ffuncall ()
#9 0x08093bff in execute_optimized_program ()
#10 0x080937ba in funcall_compiled_function ()
#11 0x080bea03 in Ffuncall ()
#12 0x08093bff in execute_optimized_program ()
#13 0x080937ba in funcall_compiled_function ()
#14 0x080bea03 in Ffuncall ()
#15 0x08093bff in execute_optimized_program ()
#16 0x080937ba in funcall_compiled_function ()
#17 0x080be01f in Feval ()
#18 0x080bbc7c in condition_case_1 ()
#19 0x0809fc0e in top_level_1 ()
#20 0x080bbb3d in internal_catch ()
#21 0x0809f3a2 in initial_command_loop ()
#22 0x080b8330 in xemacs_21_5_b3_i686_pc_linux ()
#23 0x080b9235 in main ()
#24 0x404c46a0 in __libc_start_main () from /lib/libc.so.6
(gdb)
2:51 a.m.
>>>> "Richard" == Richard Ketchersid
<ketchers(a)math.ucla.edu> writes:
Richard> Here is what happens with with xemacs version r21-5-latest-beta.
If you have a working version of XEmacs, please use M-x
report-xemacs-bug so that we get your Installation file and other
important configuration information. Also tell us the versions of the
OS, any packager, and of X11. But let me guess: recently updated
Debian sid (how apt), XFree86 4.1.0?
The Debian X11 package maintainer thinks XFree86 is broken, FWIW. I
have two otherwise identical builds of XEmacs, one built before my
upgrade of Debian, one after. The latter crashes as you describe, I'm
using the former to write this post. Hmmm....
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
7 a.m.
Below is the installation file. The distribution is Mandrake Cooker (8.2+)
uname -a: Linux lsanca1-ar12-170-121 2.4.13-7mdk #1 Wed Nov 14 18:55:29 CET 2001 i686
unknown
/usr/src/CVS/xemacs-21.5/configure '--prefix=/usr' '--libs=-L/usr/X11R6/lib
-lX11 -lXext -lXp' '--extra-verbose' '--cflags=-O3'
'--dynamic=yes' '--with-mule'
'--infopath=/usr/info:/usr/local/info:/usr/share/info:/usr/local/share/info:/usr/lib/xemacs/site-packages/info/:/usr/lib/xemacs/mule-packages/info/:/usr/lib/xemacs/xemacs-packages/info/:/usr/lib/xemacs/info/:.'
'--infodir=/usr/lib/xemacs/info/' '--with-site-lisp=yes'
'--error-checking=none'
XEmacs 21.5-b3 "asparagus" configured for `i686-pc-linux'.
Compilation / Installation:
Source code location: /usr/src/CVS/xemacs-21.5
Installation prefix: /usr
Operating system description file: `s/linux.h'
Machine description file: `m/intel386.h'
Compiler: gcc -O3
Relocating allocator for buffers: no
GNU version of malloc: yes
- Using Doug Lea's new malloc from the GNU C Library.
Window System:
Compiling in support for the X window system:
- X Windows headers location: /usr/X11R6/include
- X Windows libraries location: /usr/X11R6/lib
- Handling WM_COMMAND properly.
Using Lucid menubars.
Using Lucid scrollbars.
Using Motif dialog boxes.
Using Motif native widgets.
TTY:
Compiling in support for ncurses.
Images:
Compiling in support for GIF images (builtin).
Compiling in support for XPM images.
Compiling in support for PNG images.
Compiling in support for JPEG images.
Compiling in support for TIFF images.
Sound:
Compiling in support for sound (native).
Compiling in support for ESD (Enlightened Sound Daemon).
Databases:
Compiling in support for Berkeley database.
Compiling in support for LDAP.
Compiling in support for PostgreSQL.
- Using PostgreSQL header file: pgsql/libpq-fe.h
- Using PostgreSQL V7 bindings.
Internationalization:
Compiling in support for Mule (multi-lingual Emacs).
Compiling in support for XIM (X11R5+ I18N input method).
- Using raw Xlib to provide XIM support.
Mail:
Compiling in support for "dot-locking" mail spool file locking method.
Other Features:
Compiling in support for dynamic shared object modules.
Compiling in support for extra debugging code.
On 15 Nov 2001, Stephen J. Turnbull wrote:
>>>>> "Richard" == Richard Ketchersid
<ketchers(a)math.ucla.edu> writes:
Richard> Here is what happens with with xemacs version r21-5-latest-beta.
If you have a working version of XEmacs, please use M-x
report-xemacs-bug so that we get your Installation file and other
important configuration information. Also tell us the versions of the
OS, any packager, and of X11. But let me guess: recently updated
Debian sid (how apt), XFree86 4.1.0?
The Debian X11 package maintainer thinks XFree86 is broken, FWIW. I
have two otherwise identical builds of XEmacs, one built before my
upgrade of Debian, one after. The latter crashes as you describe, I'm
using the former to write this post. Hmmm....
11:48 a.m.
New subject: Segfaults in recent XEmacsen trying to open first frame, solved
Actually about two weeks ago, information from Mike Fabian＠SuSE (and
Jens Petersen＠Red Hat also seems aware of it), but it took _me_ a
while to catch on.
>>>> "Richard" == Richard Ketchersid
<ketchers(a)math.ucla.edu> writes:
Richard> Below is the installation file. The distribution is
Richard> Mandrake Cooker (8.2+)
Ah, YALD with the problem ... let's see ... looks familiar ... yup,
XEmacs maintainer senility. We been there, done that, but I forgot.
See http://list-archive.xemacs.org/xemacs-beta/200110/msg00294.html
for details.
"./configure --ldflags='-s -znocombreloc" ..." should fix. (Ie,
it does for me.)
Alternatively, './configure --pdump ...' should fix. (Ie, it does for
me.)
There should be a fix (to configure) in 21.4.6, which likely will use
the conservative `nocombreloc' approach. I propose to default pdump
on in 21.5 and we'll see how it goes.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
4:50 p.m.
New subject: Segfaults in recent XEmacsen trying to open first frame, solved
>>>> "ST" == Stephen J Turnbull
<stephen(a)xemacs.org> writes:
ST> Ah, YALD with the problem ... let's see ... looks familiar
ST> ... yup, XEmacs maintainer senility. We been there, done that,
ST> but I forgot.
ST>
ST> See
ST> http://list-archive.xemacs.org/xemacs-beta/200110/msg00294.html
ST> for details.
ST>
ST> "./configure --ldflags='-s -znocombreloc" ..." should fix.
(Ie,
ST> it does for me.)
ST>
ST> Alternatively, './configure --pdump ...' should fix. (Ie, it
ST> does for me.)
ST>
ST> There should be a fix (to configure) in 21.4.6, which likely
ST> will use the conservative `nocombreloc' approach. I propose to
ST> default pdump on in 21.5 and we'll see how it goes.
ST>
Stephen,
Both of the above configure lines work fine on my Debian "sid"/unstable
system. Thanks for the heads-up.
BTW, I am still not clear what the source of this problem was. Was this a
bug in XEmacs or in other libraries ?
Thanks.
--
Salman Ahmed
ssahmed AT pathcom DOT com
7:12 p.m.
New subject: Segfaults in recent XEmacsen trying to open first frame, solved
>>>> "S" == S Salman Ahmed
<ssahmed(a)pathcom.com> writes:
S> BTW, I am still not clear what the source of this problem
S> was. Was this a bug in XEmacs or in other libraries ?
Actually, it's a conflict with an optimization in GNU ld. XEmacs
assumes a certain layout for the binary when it "unexec"s the
preloaded memory image into an executable file. ld's -z comboreloc
option violates that by combining and sorting the various ELF sections
for more efficient loading. Evidently the optimization makes a
perceptible difference in loading monstrous apps like KDE, but it also
confuses XEmacs at run time.
This option was made the default in a very recent release of GNU
binutils. It only affects apps the unexec, like XEmacs and GNU Emacs.
I don't know if the GNU developers are even aware of it (I didn't see
anything last I looked at emacs-devel), but the Linux distros all are
dealing with it (at least Red Hat and SuSE have already fixed their
packages).
I don't know if you can call the unexec design a bug, although it is
acknowledged to be fragile to such changes in the toolchain. There
really wasn't much other choice in the early days of Emacs.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
12:03 p.m.
Do we still have problems with ESD? I guess I would recommend
configuring the package '--with-sound=native,noesd'.
Richard Ketchersid <ketchers(a)math.ucla.edu> writes:
Below is the installation file. The distribution is Mandrake Cooker
(8.2+)
uname -a: Linux lsanca1-ar12-170-121 2.4.13-7mdk #1 Wed Nov 14 18:55:29 CET 2001 i686
unknown
/usr/src/CVS/xemacs-21.5/configure '--prefix=/usr' '--libs=-L/usr/X11R6/lib
-lX11 -lXext -lXp' '--extra-verbose' '--cflags=-O3'
'--dynamic=yes' '--with-mule'
'--infopath=/usr/info:/usr/local/info:/usr/share/info:/usr/local/share/info:/usr/lib/xemacs/site-packages/info/:/usr/lib/xemacs/mule-packages/info/:/usr/lib/xemacs/xemacs-packages/info/:/usr/lib/xemacs/info/:.'
'--infodir=/usr/lib/xemacs/info/' '--with-site-lisp=yes'
'--error-checking=none'
XEmacs 21.5-b3 "asparagus" configured for `i686-pc-linux'.
Compilation / Installation:
Source code location: /usr/src/CVS/xemacs-21.5
Installation prefix: /usr
Operating system description file: `s/linux.h'
Machine description file: `m/intel386.h'
Compiler: gcc -O3
Relocating allocator for buffers: no
GNU version of malloc: yes
- Using Doug Lea's new malloc from the GNU C Library.
Window System:
Compiling in support for the X window system:
- X Windows headers location: /usr/X11R6/include
- X Windows libraries location: /usr/X11R6/lib
- Handling WM_COMMAND properly.
Using Lucid menubars.
Using Lucid scrollbars.
Using Motif dialog boxes.
Using Motif native widgets.
TTY:
Compiling in support for ncurses.
Images:
Compiling in support for GIF images (builtin).
Compiling in support for XPM images.
Compiling in support for PNG images.
Compiling in support for JPEG images.
Compiling in support for TIFF images.
Sound:
Compiling in support for sound (native).
Compiling in support for ESD (Enlightened Sound Daemon).
Databases:
Compiling in support for Berkeley database.
Compiling in support for LDAP.
Compiling in support for PostgreSQL.
- Using PostgreSQL header file: pgsql/libpq-fe.h
- Using PostgreSQL V7 bindings.
Internationalization:
Compiling in support for Mule (multi-lingual Emacs).
Compiling in support for XIM (X11R5+ I18N input method).
- Using raw Xlib to provide XIM support.
Mail:
Compiling in support for "dot-locking" mail spool file locking method.
Other Features:
Compiling in support for dynamic shared object modules.
Compiling in support for extra debugging code.
On 15 Nov 2001, Stephen J. Turnbull wrote:
> >>>>> "Richard" == Richard Ketchersid
<ketchers(a)math.ucla.edu> writes:
>
> Richard> Here is what happens with with xemacs version r21-5-latest-beta.
>
> If you have a working version of XEmacs, please use M-x
> report-xemacs-bug so that we get your Installation file and other
> important configuration information. Also tell us the versions of the
> OS, any packager, and of X11. But let me guess: recently updated
> Debian sid (how apt), XFree86 4.1.0?
>
> The Debian X11 package maintainer thinks XFree86 is broken, FWIW. I
> have two otherwise identical builds of XEmacs, one built before my
> upgrade of Debian, one after. The latter crashes as you describe, I'm
> using the former to write this post. Hmmm....
>
>
2:51 p.m.
New subject: ESD issues [was: Segfault in XEmacs-21.5-b3]
>>>> "vin" == Vin Shelton <acs(a)xemacs.org>
writes:
vin> Do we still have problems with ESD? I guess I would
vin> recommend configuring the package
vin> '--with-sound=native,noesd'.
Yes, ESD is still a problem AFAIK.
I got sick of seeing the reports, since there's nothing we can do
about it, and 21.4 now defaults to noesd. I wonder if I should port
the patch to 21.5?
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
8205
days inactive
8205
days old
7 comments
4 participants
participants (4)
-
Richard Ketchersid -
S. Salman Ahmed -
Stephen J. Turnbull -
Vin Shelton