Re: Crashes of GNU Emacs 23

[Peter Dyballa]

Am 07.07.2005 um 13:54 schrieb Jan D.:

> (gdb) n
> (gdb) n
> (gdb) p buf1
>
> You forgot the stack trace.

So hopefully this is now right:

(gdb) where
#0  x_error_quitter (display=0x4907400, error=0xbfffba00) at  
xterm.c:7795
#1  0x000f18a0 in x_error_handler (display=0x4907400, error=0xbfffba00)  
at xterm.c:7760
#2  0x8acb614c in _XError ()
#3  0x8acb4884 in _XReply ()
#4  0x8acafaa8 in XSync ()
#5  0x8acafb3c in _XSyncFunction ()
#6  0x8ac95f84 in XCreateGC ()
#7  0x00100828 in x_make_gc (f=0xdbce60) at xfns.c:2828
#8  0x001055b4 in x_create_tip_frame (dpyinfo=0xd1acd0, parms=10561309,  
text=77542483) at xfns.c:4805
#9  0x00106258 in Fx_show_tip (string=77542483, frame=13704404,  
parms=10561309, timeout=80, dx=40, dy=-\
80) at xfns.c:5040
#10 0x001edd8c in Ffuncall (nargs=7, args=0xbfffbfc0) at eval.c:2872
#11 0x00242098 in Fbyte_code (bytestr=3758955, vector=3759044,  
maxdepth=56) at bytecode.c:689
#12 0x001ec530 in Feval (form=3758941) at eval.c:2200
#13 0x001ea420 in Fcondition_case (args=10561237) at eval.c:1390
#14 0x0024301c in Fbyte_code (bytestr=3758843, vector=3758876,  
maxdepth=24) at bytecode.c:867
#15 0x001ee770 in funcall_lambda (fun=3758788, nargs=2,  
arg_vector=0xbfffc954) at eval.c:3040
#16 0x001edeac in Ffuncall (nargs=3, args=0xbfffc950) at eval.c:2899
#17 0x00242098 in Fbyte_code (bytestr=3761099, vector=3761132,  
maxdepth=24) at bytecode.c:689
#18 0x001ee770 in funcall_lambda (fun=3761060, nargs=1,  
arg_vector=0xbfffcd78) at eval.c:3040
#19 0x001edeac in Ffuncall (nargs=2, args=0xbfffcd74) at eval.c:2899
#20 0x001ed0f4 in run_hook_with_args (nargs=2, args=0xbfffcd74,  
cond=until_success) at eval.c:2517
#21 0x001ecdd0 in Frun_hook_with_args_until_success (nargs=2,  
args=0xbfffcd74) at eval.c:2427
#22 0x001eda8c in Ffuncall (nargs=3, args=0xbfffcd70) at eval.c:2832
#23 0x00242098 in Fbyte_code (bytestr=3758507, vector=3758532,  
maxdepth=24) at bytecode.c:689
#24 0x001ee770 in funcall_lambda (fun=3758468, nargs=1,  
arg_vector=0xbfffd138) at eval.c:3040
#25 0x001edeac in Ffuncall (nargs=2, args=0xbfffd134) at eval.c:2899
#26 0x001ec9ec in Fapply (nargs=2, args=0xbfffd134) at eval.c:2294
#27 0x001eda8c in Ffuncall (nargs=3, args=0xbfffd130) at eval.c:2832
#28 0x00242098 in Fbyte_code (bytestr=3551435, vector=3551476,  
maxdepth=32) at bytecode.c:689
#29 0x001ec530 in Feval (form=3551421) at eval.c:2200
#30 0x001ea420 in Fcondition_case (args=10561221) at eval.c:1390
#31 0x0024301c in Fbyte_code (bytestr=3551171, vector=3551316,  
maxdepth=40) at bytecode.c:867
#32 0x001ee770 in funcall_lambda (fun=3551132, nargs=1,  
arg_vector=0xbfffdad4) at eval.c:3040
#33 0x001edeac in Ffuncall (nargs=2, args=0xbfffdad0) at eval.c:2899
#34 0x001ed414 in call1 (fn=41971049, arg1=14405076) at eval.c:2639
#35 0x0013609c in timer_check (do_it_now=1) at keyboard.c:4470
#36 0x00133c8c in readable_events (flags=1) at keyboard.c:3509
#37 0x0013a314 in get_input_pending (addr=0x3b1b10, flags=1) at  
keyboard.c:6552
#38 0x00142c6c in detect_input_pending_run_timers (do_display=1) at  
keyboard.c:9910
#39 0x0024e35c in wait_reading_process_output (time_limit=30,  
microsecs=0, read_kbd=-1, do_display=1, w\
ait_for_cell=41944073, wait_proc=0x0, just_wait_proc=0) at  
process.c:4460
#40 0x000118a4 in sit_for (sec=30, usec=0, reading=1, display=1,  
initial_display=0) at dispnew.c:6386
#41 0x00131d44 in read_char (commandflag=1, nmaps=2, maps=0xbfffe930,  
prev_event=41944073, used_mouse_m\
enu=0xbfffe9e0) at keyboard.c:2768
#42 0x0013f948 in read_key_sequence (keybuf=0xbfffec10, bufsize=30,  
prompt=41944073, dont_downcase_last\
=0, can_return_switch_frame=1, fix_current_buffer=1) at keyboard.c:8821
#43 0x0012d2a4 in command_loop_1 () at keyboard.c:1528
#44 0x001ea5f4 in internal_condition_case (bfun=0x12cd80  
<command_loop_1>, handlers=41978953, hfun=0x12\
c440 <cmd_error>) at eval.c:1443
#45 0x0012ca2c in command_loop_2 () at keyboard.c:1318
#46 0x001e9de0 in internal_catch (tag=41974217, func=0x12c9ec  
<command_loop_2>, arg=41944073) at eval.c\
:1202
#47 0x0012c994 in command_loop () at keyboard.c:1297
#48 0x0012bed8 in recursive_edit_1 () at keyboard.c:990
#49 0x0012c164 in Frecursive_edit () at keyboard.c:1051
#49 0x0012c164 in Frecursive_edit () at keyboard.c:1051
#50 0x00129e58 in main (argc=3, argv=0xbffff8e4) at emacs.c:1784
(gdb) n
7796      sprintf (buf1, "X protocol error: %s on protocol request %d",
(gdb) n
7798      x_connection_closed (display, buf1);
(gdb) p buf1
$1 = "X protocol error: BadFont (invalid Font parameter) on protocol  
request 55\0000\313\000\000\000\02\
0\001\0010\314\000\000\000\000\001\200\000\032\000\034\037$\000\035\317\ 
300\000\000\000\000\001\n\224\3\
04\002\a\260\000\000\000\000\000\277\377\272\240\004\220t\000\000\017\26 
2P\277\377\270\300\000\000\000\\
000\212\312\373<\000\017\262P\277\377\270\360\204\000\002(\212\312\372\2 
50\277\377\271\200\277\377\271\\
320\200\000\000\000\000\323S\000\212\314Y0\000\000\000\000\000\000\000\0 
00\277\377\370\340\000\000\000\\
020\000\000\000 \277\377\272\240"...

So it comes probably to the question: should I throw away my X11  
fontsets for GNU Emacs 23? They are based on the function  
create-fontset-from-fontset-spec, two of them are a bit complicated  
since they have additions with decode-char à la:

    (create-fontset-from-fontset-spec  
"-adobe-courier-medium-r-*-*-9-*-*-*-*-*-fontset-09pt_adobe_courier" t  
'noerror)
	(set-fontset-font "fontset-09pt_adobe_courier"       'latin-iso8859-1   
'("adobe-courier" . "iso8859-1"))
	(set-fontset-font "fontset-09pt_adobe_courier"       'latin-iso8859-2   
'("adobe-courier" . "iso8859-2"))
... ; more Latin scripts
	(set-fontset-font "fontset-09pt_adobe_courier"       'latin-iso8859-15  
'("adobe-courier" . "iso8859-15"))
	(set-fontset-font "fontset-09pt_adobe_courier" 'mule-unicode-0100-24ff  
'("adobe-courier" . "iso10646-1"))
	(set-fontset-font "fontset-09pt_adobe_courier" 'mule-unicode-2500-33ff  
'("adobe-courier" . "iso10646-1"))
	(set-fontset-font "fontset-09pt_adobe_courier" 'mule-unicode-e000-ffff  
'("adobe-courier" . "iso10646-1"))
	(set-fontset-font "fontset-09pt_adobe_courier" (cons (decode-char 'ucs  
#x0370) (decode-char 'ucs #x03cf)) '("courier new" . "iso10646-1"))	;  
Greek
... ; more non-Latin scripts

But they aren't the culprits ... I reduced the number of fontsets until  
the initial one only was left. To do so, to get its name, I opened  
.emacs and saw this (here already corrected):

        (setq default-frame-alist '(
          (background-color . "LightCyan1")
;         (background-color . "ivory")
;         (background-color . "ghost white")
          (cursor-color . "purple")
          (cursor-type . box)
          (foreground-color . "grey10")
          (vertical-scroll-bars . left)
          (font . "fontset-11pt_adobe_courier")
;         (font . "-adobe-courier-medium-r-*-11-*-iso10646-1")
          (top . 25) (left . 50) (width . 89) (height . 50))
        )

The reason for the crashes obviously is the use of:

          (font . "-adobe-courier-medium-r-*-11-*-iso10646-1")

Having made it a comment GNU Emacs 23 now behaves fine again.


I have a backup file from July 4 -- I either made this change then or  
the updates I fetched this weekend made this statement incompatible --  
I can't tell ... GNU Emacs 22.0.50 still uses that setting and works  
stable.


Thank you very much for giving me the right clues!

--
Mit friedvollen Grüßen

  Pete

For some reason, this fortune reminds everyone of Marvin Zelkowitz.