5:03 a.m.
this is with andy's gutter/gc patch from this past weekend.
i was composing an email from vm, when I saw the tabs disappear. i.e,
not tabs, just a solid motif-blue background where the tabs would have
been. XEmacs hung, thrashed awhile, then died.
no lisp backtrace, i think i had killed off the xterm I started that
session from.
here is the gdb backtrace
#0 0x403cc4e1 in __kill () from /lib/libc.so.6
#1 0x8080991 in fatal_error_signal (sig=11) at emacs.c:272
#2 0x403cc408 in __restore ()
at ../sysdeps/unix/sysv/linux/i386/sigaction.c:127
#3 0x809c3be in x_update_scrollbar_instance_status (w=0x88d5b10, active=1,
size=0, instance=0x8abcf40) at scrollbar-x.c:250
#4 0x8095d0b in update_window_scrollbars (w=0x88d5b10, mirror=0x8b53dc0,
active=1, horiz_only=0) at scrollbar.c:314
#5 0x813300d in redisplay_output_window (w=0x88d5b10)
at redisplay-output.c:2251
#6 0x812833a in redisplay_window (window=143481616, skip_selected=0)
at redisplay.c:6196
#7 0x812875c in redisplay_frame (f=0x8b5d2b0, preemption_check=0)
at redisplay.c:6412
#8 0x8128938 in redisplay_device (d=0x84e4200, automatic=1)
at redisplay.c:6490
#9 0x8128c79 in redisplay_without_hooks () at redisplay.c:6579
#10 0x812eddc in redisplay () at redisplay.c:6636
#11 0x80b1bb4 in Fsit_for (seconds=1, nodisplay=136572516)
at event-stream.c:2780
#12 0x8085020 in Ffuncall (nargs=2, args=0xbfffeac0) at eval.c:3196
#13 0x8060544 in execute_optimized_program (
program=0x89656f0
"\b«\035\t«\032\tÂa«\004ê\023ÄÅÆ\t\"\tÇa«\004Ȫ\002ÉQª\002Ê\026\013\b«'\016\f¬\bÊËD\026\fª\034\016\f;«\t\016\fËD\026\fª\017Ë\016\fs¬\tÍ\016\fËC\"\026\f\212Î
q\210)ÏÐ !\210ÑÂ!\207", stack_depth=4,
constants_data=0x8552338) at bytecode.c:747
#14 0x80601a9 in funcall_compiled_function (fun=140000228, nargs=1,
args=0xbfffebc0) at bytecode.c:519
#15 0x808506f in Ffuncall (nargs=2, args=0xbfffebbc) at eval.c:3210
#16 0x8060544 in execute_optimized_program (
program=0x89655d0 "\b\205\005\001Á\032Ã
\034Å\f8\036\006\fA＠\036\a\016\006È_\016\a\\\211\036\t\016\nU\036\013\016\n\016\tV\036\f\016\t\026\n\016\f«\005ÍÎ!\210\b\203Ë",
stack_depth=5, constants_data=0x8552c80) at bytecode.c:747
#17 0x80601a9 in funcall_compiled_function (fun=140000340, nargs=0,
3:21 p.m.
>>>> "JM" == Jeff Miller
<jmiller(a)smart.net> writes:
JM> this is with andy's gutter/gc patch from this past weekend.
JM>
JM> i was composing an email from vm, when I saw the tabs disappear. i.e,
JM> not tabs, just a solid motif-blue background where the tabs would have
JM> been. XEmacs hung, thrashed awhile, then died.
JM>
JM> no lisp backtrace, i think i had killed off the xterm I started that
JM> session from.
I've seen one of these since the patches as well. Exactly the same
description - composing in vm, tabs become blue, thrash, kaboom.
It was running so well before that that I had switched to running the
optimized version, though, so I didn't get anything useful.
After rebuilding debug with full error-checking I haven't had the
crash.
>>>> "JV" == Jan Vroonhof
<vroonhof(a)math.ethz.ch> writes:
JV> Does applying
JV>
JV> http://www.xemacs.org/list-archives/xemacs-patches/200003/msg00011.html
JV>
JV> help?
Maybe that's why I haven't seen the crash, since I updated and rebuilt
on 3/8. I'll keep running vm with the debug version since that seems
to be the harshest test ("VM - MUA or Emacs torture-test? you decide...").
--
John A. Turner, Senior Research Associate
Blue Sky Studios, One South Rd, Harrison, NY 10528
http://www.blueskystudios.com/ (914) 825-8319
5:26 p.m.
>>>> "JAT" == John A Turner
<turner(a)blueskystudios.com> writes:
JAT> I've seen one of these since the patches as well. Exactly the same
JAT> description - composing in vm, tabs become blue, thrash, kaboom.
JAT>
JAT> It was running so well before that that I had switched to running the
JAT> optimized version, though, so I didn't get anything useful.
JAT>
JAT> After rebuilding debug with full error-checking I haven't had the
JAT> crash.
of course as soon as I said that, boom... nice C stack trace, but once
again no lisp backtrace... same env and build as reported in:
http://www.xemacs.org/list-archives/xemacs-beta/200003/msg00021.html
(gdb) where
#0 0x102a86c0 in lw_add_widget_value_arg (wv=0x0, name=0xf510d0e "foreground",
value=10936319) at lwlib.c:1368
#1 0x1026d05c in update_widget_face (wv=0x0, ii=0x10d7ee80, domain=281544192) at
glyphs-x.c:2356
#2 0x1026c7e0 in x_update_widget (p=0x10d7ee80) at glyphs-x.c:2211
#3 0x10177db8 in update_widget (widget=282586752) at glyphs-widget.c:352
#4 0x1016d2c8 in update_subwindow (subwindow=282586752) at glyphs.c:4312
#5 0x1016d538 in update_frame_subwindows (f=0x10c80600) at glyphs.c:4364
#6 0x101f0664 in redisplay_frame (f=0x10c80600, preemption_check=0) at redisplay.c:6341
#7 0x101f0da0 in redisplay_device (d=0x10642000, automatic=1) at redisplay.c:6490
#8 0x101f1720 in redisplay_without_hooks () at redisplay.c:6579
#9 0x101f1a34 in redisplay () at redisplay.c:6636
#10 0x10105d08 in Fnext_event (event=279769556, prompt=272336900) at event-stream.c:2236
#11 0x10076568 in Fcommand_loop_1 () at cmdloop.c:566
#12 0x100761ec in command_loop_1 (dummy=272336900) at cmdloop.c:490
#13 0x100acd34 in condition_case_1 (handlers=272336996, bfun=0x1007619c
<command_loop_1>, barg=272336900, hfun=0x10075560 <cmd_error>, harg=272336900)
at eval.c:1629
#14 0x10075788 in command_loop_3 () at cmdloop.c:252
#15 0x100757f8 in command_loop_2 (dummy=272336900) at cmdloop.c:263
#16 0x100ac628 in internal_catch (tag=272412644, func=0x100757c4 <command_loop_2>,
arg=272336900, threw=0x0) at eval.c:1304
#17 0x10075b30 in initial_command_loop (load_me=272336900) at cmdloop.c:301
#18 0x100a515c in xemacs_21_2_b31_mips_sgi_irix6_5 (argc=1, argv=0x7fff2e34,
envp=0x7fff2e3c, restart=0) at emacs.c:1879
#19 0x100a6490 in main (argc=1, argv=0x7fff2e34, envp=0x7fff2e3c) at emacs.c:2305
7:18 a.m.
At 12:26 PM 3/9/00 -0500, John A. Turner wrote:
#0 0x102a86c0 in lw_add_widget_value_arg (wv=0x0, name=0xf510d0e
"foreground", value=10936319) at lwlib.c:1368
#1 0x1026d05c in update_widget_face (wv=0x0, ii=0x10d7ee80,
domain=281544192) at glyphs-x.c:2356
Someone else has reported this. I do not understand why wv could be 0, it
seems to indicate that the widget has been deleted. Try the attached patch
and see if you can go for any longer. Its not really the right fix but it
might help a bit.
andy
--------------------------------------------------------------
Dr Andy Piper
Senior Consultant Architect, BEA Systems Ltd
8:31 p.m.
>>>> "AP" == Andy Piper <andy(a)xemacs.org>
writes:
AP> Someone else has reported this. I do not understand why wv could
AP> be 0, it seems to indicate that the widget has been deleted. Try
AP> the attached patch and see if you can go for any longer. Its not
AP> really the right fix but it might help a bit.
that may have helped a little, but now I'm crashing in a slightly
different place
as before, this is only happening in The XEmacs Torture Test
(a.k.a. VM) - I happily do all my real work in a separate fully-
optimized build of the same source
this is after updating this AM, so it has other mods in addition to
this glyphs-x.c hackery
C stack follows, but I'm still not able to get a lisp backtrace (see
end of C stack for what happens when I try to cont)
Program received signal SIGSEGV, Segmentation fault.
0x1026e170 in update_tab_widget_face (wv=0x0, ii=0x10b3f400,
domain=279985152) at glyphs-x.c:2388
2388 if (wv->contents)
(gdb) where
#0 0x1026e170 in update_tab_widget_face (wv=0x0, ii=0x10b3f400,
domain=279985152) at glyphs-x.c:2388
#1 0x1026f7bc in x_tab_control_update (image_instance=280228864)
at glyphs-x.c:2698
#2 0x101783dc in update_widget (widget=280228864)
at glyphs-widget.c:358
#3 0x1016d878 in update_subwindow (subwindow=280228864)
at glyphs.c:4312
#4 0x10201d74 in redisplay_output_subwindow (w=0x10ab5c00,
image_instance=280228864, db=0x7fff1f10, dga=0x7fff1f20,
findex=7, cursor_start=0, cursor_width=0, cursor_height=0)
at redisplay-output.c:1181
#5 0x102776f4 in x_output_display_block (w=0x10ab5c00,
dl=0x10700000, block=0, start=0, end=1, start_pixpos=0,
cursor_start=0, cursor_width=0, cursor_height=0)
at redisplay-x.c:482
#6 0x10201a38 in redisplay_output_display_block (w=0x10ab5c00,
dl=0x10700000, block=0, start=0, end=1, start_pixpos=0,
cursor_start=0, cursor_width=0, cursor_height=0)
at redisplay-output.c:1099
#7 0x1020052c in output_display_line (w=0x10ab5c00, cdla=0x10b69dc0,
ddla=0x10b69ea0, line=0, force_start=-1, force_end=-1)
at redisplay-output.c:759
#8 0x1018d638 in output_gutter (f=0x10b03c00, pos=TOP_GUTTER,
force=0) at gutter.c:314
#9 0x1018ed90 in update_frame_gutters (f=0x10b03c00) at gutter.c:524
#10 0x101f1138 in redisplay_frame (f=0x10b03c00, preemption_check=0)
at redisplay.c:6406
#11 0x101f1a7c in redisplay_device (d=0x10612c00, automatic=1)
at redisplay.c:6523
#12 0x101f2148 in redisplay_without_hooks () at redisplay.c:6582
#13 0x101f245c in redisplay () at redisplay.c:6639
#14 0x1010795c in Fnext_event (event=275646160, prompt=272336900)
at event-stream.c:2173
#15 0x100767b8 in Fcommand_loop_1 () at cmdloop.c:566
#16 0x1007643c in command_loop_1 (dummy=272336900) at cmdloop.c:490
#17 0x100acdd4 in condition_case_1 (handlers=272336996,
bfun=0x100763ec <command_loop_1>, barg=272336900,
hfun=0x100757b0 <cmd_error>, harg=272336900) at eval.c:1644
#18 0x100759d8 in command_loop_3 () at cmdloop.c:252
#19 0x10075a48 in command_loop_2 (dummy=272336900) at cmdloop.c:263
#20 0x100ac658 in internal_catch (tag=272412644,
func=0x10075a14 <command_loop_2>, arg=272336900, threw=0x0)
at eval.c:1310
#21 0x10075d80 in initial_command_loop (load_me=272336900)
at cmdloop.c:301
#22 0x100a5190 in xemacs_21_2_b31_mips_sgi_irix6_5 (argc=1,
argv=0x7fff2e34, envp=0x7fff2e3c, restart=0) at emacs.c:1828
#23 0x100a64c4 in main (argc=1, argv=0x7fff2e34, envp=0x7fff2e3c)
at emacs.c:2254
(gdb) cont
Continuing.
Program received signal SIGSEGV, Segmentation fault.
0x1026e170 in update_tab_widget_face (wv=0x0, ii=0x10b3f400,
domain=279985152) at glyphs-x.c:2388
2388 if (wv->contents)
(gdb) cont
Continuing.
Program received signal SIGSEGV, Segmentation fault.
kill () at engine.c:648
648 engine.c: No such file or directory.
(gdb) cont
Continuing.
Program terminated with signal SIGSEGV, Segmentation fault.
The program no longer exists.
-JT
11:23 a.m.
At 03:31 PM 3/15/00 -0500, John A. Turner wrote:
that may have helped a little, but now I'm crashing in a slightly
different place
Ok, I forgot to apply the same kludge to tab_controls. The patch going into
CVS RSN should fix this.
andy
--------------------------------------------------------------
Dr Andy Piper
Senior Consultant Architect, BEA Systems Ltd
9072
days inactive
9080
days old
5 comments
3 participants
participants (3)
-
Andy Piper -
Jeff Miller -
John A. Turner