9:34 a.m.
1. Read my mail using VM (6.62).
2. Mail message has an HTML attachment, so I click on it with button
2 and attachment is sent to a running netscape.
3. Hit 'q' in the VM window --> crash
Lisp backtrace follows:
set-window-buffer(#<window on "INBOX" 0x8a31> #<buffer
"INBOX">)
# bind (delete-me done all-frames start w obj-2 obj-1 action)
vm-window-loop(replace #<buffer "INBOX Presentation"> #<buffer
"INBOX">)
# bind (new old)
vm-replace-buffer-in-windows(#<buffer "INBOX Presentation"> #<buffer
"INBOX">)
# (unwind-protect ...)
# bind (vm-sbe-buffer)
vm-preview-current-message()
# bind (buffer)
#<compiled-function (buffer) "...(47)" [symbol-name buffer vm-system-state
vm-...
And here's the core dump:
XEmacs details:
uname -a: SunOS picard 5.5 Generic_103093-23 sun4m sparc SUNW,SPARCstation-10
./configure '--site-includes=/opt/local/include /usr/demo/SOUND/include'
'--site-libraries=/opt/local/lib /usr/demo/SOUND/lib' '--cflags=-g'
'--with-site-lisp=yes' '--with-dialogs=athena' '--debug'
'--error-checking=all' '--verbose'
XEmacs 21.0 "Poitou" configured for `sparc-sun-solaris2.5'.
Where should the build process find the source code? /disk1/xemacs-21.0
What installation prefix should install use? /usr/local
What operating system and machine description files should XEmacs use?
`s/sol2.h' and `m/sparc.h'
What compiler should XEmacs be built with? cc -g
Should XEmacs use the GNU version of malloc? yes
Should XEmacs use the relocating allocator for buffers? yes
What window system should XEmacs use? x11
Where do we find X Windows header files? /usr/dt/include
/usr/openwin/include
Where do we find X Windows libraries? /usr/dt/lib /usr/openwin/lib
Additional header files: /opt/local/include
/usr/demo/SOUND/include
Additional libraries: /opt/local/lib
/usr/demo/SOUND/lib
Runtime library search path:
/usr/dt/lib:/usr/openwin/lib:/opt/SUNWdt/lib
Compiling in support for XAUTH.
Compiling in support for XPM images.
Compiling in support for PNG image handling.
Compiling in support for (builtin) GIF image handling.
Compiling in support for JPEG image handling.
Compiling in support for X-Face message headers.
Compiling in native sound support.
Compiling in support for GNU DBM.
Compiling in support for CDE.
Compiling in support for ToolTalk.
Compiling in EXPERIMENTAL support for Drag'n'Drop ( CDE ).
Compiling in support for proper session-management.
Using Lucid menubars.
Using Lucid scrollbars.
Using Athena dialog boxes.
Compiling in DLL support.
movemail will use "dot-locking" for locking mail spool files.
Compiling in extra code for debugging.
Compiling in code for checking XEmacs memory usage.
WARNING: ---------------------------------------------------------
WARNING: Compiling in support for runtime error checking.
WARNING: XEmacs will run noticeably more slowly as a result.
WARNING: Error checking is on by default for XEmacs beta releases.
WARNING: ---------------------------------------------------------
--
8:15 a.m.
Amir J. Katz writes:
set-window-buffer(#<window on "INBOX" 0x8a31>
#<buffer "INBOX">)
This seems to be a recurring theme. I'm also seeing at least one
other weird recentering display behavior that makes me think
something bad has happened to the redisplay. If anyone knows of
recent redisplay code changes that might be responsible, please
speak up.
8:59 a.m.
Kyle Jones writes:
| Amir J. Katz writes:
| > set-window-buffer(#<window on "INBOX" 0x8a31> #<buffer
"INBOX">)
|
| This seems to be a recurring theme. I'm also seeing at least one
| other weird recentering display behavior that makes me think
| something bad has happened to the redisplay. If anyone knows of
| recent redisplay code changes that might be responsible, please
| speak up.
Gunnar Evermann already started speculation about that one since he
sent a redisplay and a Motif patch. He sent a fix which Georg Nikodym
and I tested:
--- window.c.orig-vmcrash Fri Oct 2 22:01:32 1998
+++ window.c Sat Oct 3 10:50:04 1998
＠＠ -1635,7 +1635,8 ＠＠
/* this is not right, but much easier than doing what is right. */
/* w->start_at_line_beg = 0; */
/* WTF is this supposed to mean? GE */
- w->start_at_line_beg = beginning_of_line_p (XBUFFER (w->buffer), XINT (pos));
+ w->start_at_line_beg = beginning_of_line_p (XBUFFER (w->buffer),
+ marker_position (w->start[CURRENT_DISP]));
if (NILP (noforce))
w->force_start = 1;
w->redo_modeline = 1;
＠＠ -3167,7 +3168,7 ＠＠
Fset_marker (w->sb_point, w->start[CURRENT_DISP], buffer);
/* set start_at_line_beg correctly. GE */
w->start_at_line_beg = beginning_of_line_p (XBUFFER (buffer),
- XBUFFER (buffer)->last_window_start);
+ marker_position (w->start[CURRENT_DISP]));
w->force_start = 0; /* Lucid fix */
SET_LAST_MODIFIED (w, 1);
SET_LAST_FACECHANGE (w);
I think you have to apply this patch manually but it fixed the VM
crash problem reported before which looked pretty much the same!
--Marcus
2:30 p.m.
On Wed, 28 Oct 1998, Marcus Thiessel wrote:
Kyle Jones writes:
| Amir J. Katz writes:
| > set-window-buffer(#<window on "INBOX" 0x8a31> #<buffer
"INBOX">)
|
| This seems to be a recurring theme. I'm also seeing at least one
| other weird recentering display behavior that makes me think
| something bad has happened to the redisplay. If anyone knows of
| recent redisplay code changes that might be responsible, please
| speak up.
Actually I think something good has happened to redisplay but then others
might disagree...
Gunnar Evermann already started speculation about that one since he
sent a redisplay and a Motif patch. He sent a fix which Georg Nikodym
and I tested:
[patch snipped]
Uh, it was rather a bit more than speculation! I produced a fix for the
problem of XEmacs crahsing in redisplay when displaying a window that had
just been widened. I rushed this fix a bit and it contained a dumb bug
(which now causes the crashes in VM) which is fixed by the patch Marcus
resent (it's also in CVS already).
Concerning the 'weird recentering' : What my patch does is to look at the
value of w->start_at_line_beg and if it is true, BUT 'start' is not really
at the beginnig of a line, it triggers a 'regenerate_window_point_center'.
Actually the code looked like this was the way someone had intended it to
work, but it didn't.
I had to modify some places (set_window_buffer and set_window_start or
something similar) to make sure that w->start_at_line_beg is actually set
to the correct value. There might still be other places that need
fixing...
I think in VM you use set-window-start for scrolling or something like
that. If you set start to some value in the middle of a line this might
cause some recentering. if you can describe more precisely when this weird
stuff happens I can look into it when I get home (I only have XEmacs 19.13
here at the moment).
Sorry for the confusion.
Gunnar
3:53 p.m.
"G" == G Evermann <ge204(a)cus.cam.ac.uk> writes:
G> I think in VM you use set-window-start for scrolling or something
G> like that. If you set start to some value in the middle of a line
G> this might cause some recentering.
I haven't tried XEmacs 21.* at all, but FWIW, the `scroll-in-place' package
uses `set-window-start' to effect scrolling, and it may set the window start to
a non-line-start position.
`scroll-in-place' might be useful in reproducing and diagnosing the undesirable
recentering behavior. Since it's a lot smaller than VM, it might be easier for
one to figure out what is going on.
Additionally, it's easy to scroll at a fine grain with `scroll-in-place': try
`C-u 1 C-v' or `C-u 1 M-v'. The prefix argument ``sticks'' for
immediately
subsequent invocations of `C-v' and `M-v'.
Eric.
--
-------------------------------------------------------------------------------
Eric Eide <eeide(a)cs.utah.edu> . University of Utah Dept. of Computer Science
http://www.cs.utah.edu/~eeide . +1 (801) 585-5512 voice, +1 (801) 581-5843 FAX
9524
days inactive
9524
days old
4 comments
5 participants
participants (5)
-
Amir J. Katz -
Eric Eide -
Gunnar Evermann -
Kyle Jones -
Marcus Thiessel