bug#9081: 24.0.50; crash in X

bug#9081: 24.0.50; crash in X

[Sam Steingold]

In GNU Emacs 24.0.50.2 (x86_64-unknown-linux-gnu, X toolkit)
 of 2011-07-07 on tbox
Windowing system distributor `Colin Harrison', version 11.0.60900031
configured using `configure  '--with-x-toolkit=lucid' '--without-gconf''

crash:

Breakpoint 1, abort () at /home/sds/src/emacs/trunk/src/emacs.c:374
374     {
(gdb) where
#0  abort () at /home/sds/src/emacs/trunk/src/emacs.c:374
#1  0x000000000055f0ff in internal_condition_case (
    bfun=0x4eed80 <safe_run_hooks_1>, handlers=11796306, 
    hfun=0x4f9350 <safe_run_hooks_error>)
    at /home/sds/src/emacs/trunk/src/eval.c:1468
#2  0x00000000004f92fa in safe_run_hook_funcall (nargs=<value optimized out>, 
    args=<value optimized out>)
    at /home/sds/src/emacs/trunk/src/keyboard.c:1923
#3  0x000000000055af85 in run_hook_with_args (nargs=1, args=0x7fff276f5ec8, 
    funcall=0x4f92c0 <safe_run_hook_funcall>)
    at /home/sds/src/emacs/trunk/src/eval.c:2715
#4  0x00000000004f685f in safe_run_hooks (hook=497989842)
    at /home/sds/src/emacs/trunk/src/keyboard.c:1940
#5  0x000000000042c141 in update_menu_bar (f=0x2099da50, save_match_data=0, 
    hooks_run=1) at /home/sds/src/emacs/trunk/src/xdisp.c:9930
#6  0x000000000043b716 in prepare_menu_bars ()
    at /home/sds/src/emacs/trunk/src/xdisp.c:9826
#7  0x00000000004cb187 in x_handle_selection_clear (
    event=<value optimized out>) at /home/sds/src/emacs/trunk/src/xselect.c:987
#8  0x00000000004cc0ae in x_handle_selection_event (event=0x4eed80)
    at /home/sds/src/emacs/trunk/src/xselect.c:996
#9  0x00000000004f456c in swallow_events (do_display=0)
    at /home/sds/src/emacs/trunk/src/keyboard.c:4181
#10 0x000000000059821d in wait_reading_process_output (time_limit=20, 
    microsecs=0, read_kbd=0, do_display=0, wait_for_cell=12087398, 
    wait_proc=0x0, just_wait_proc=0)
    at /home/sds/src/emacs/trunk/src/process.c:4658
#11 0x00000000004ca419 in wait_for_property_change (location=0x1f4207f0)
    at /home/sds/src/emacs/trunk/src/xselect.c:1141
#12 0x00000000004cc59c in x_handle_selection_request (
    event=<value optimized out>) at /home/sds/src/emacs/trunk/src/xselect.c:682
#13 x_handle_selection_event (event=<value optimized out>)
    at /home/sds/src/emacs/trunk/src/xselect.c:1000
#14 0x00000000004f456c in swallow_events (do_display=1)
    at /home/sds/src/emacs/trunk/src/keyboard.c:4181
#15 0x0000000000598c83 in wait_reading_process_output (time_limit=30, 
    microsecs=0, read_kbd=-1, do_display=1, wait_for_cell=11796258, 
    wait_proc=0x0, just_wait_proc=0)
    at /home/sds/src/emacs/trunk/src/process.c:4627
#16 0x00000000004124a8 in sit_for (timeout=120, reading=1, do_display=1)
    at /home/sds/src/emacs/trunk/src/dispnew.c:5988
#17 0x00000000004f7d89 in read_char (commandflag=1, nmaps=3, 
    maps=0x7fff276f7c90, prev_event=11796258, used_mouse_menu=0x7fff276f7e3c, 
    end_time=0x0) at /home/sds/src/emacs/trunk/src/keyboard.c:2688
#18 0x00000000004f9f8a in read_key_sequence (keybuf=0x7fff276f7ea0, 
    bufsize=30, prompt=11796258, dont_downcase_last=0, 
    can_return_switch_frame=1, fix_current_buffer=1)
    at /home/sds/src/emacs/trunk/src/keyboard.c:9283
#19 0x00000000004fc1ef in command_loop_1 ()
    at /home/sds/src/emacs/trunk/src/keyboard.c:1445
#20 0x000000000055f0d1 in internal_condition_case (
    bfun=0x4fc030 <command_loop_1>, handlers=11848338, 
    hfun=0x4f9710 <cmd_error>) at /home/sds/src/emacs/trunk/src/eval.c:1493
#21 0x00000000004f933a in command_loop_2 (ignore=<value optimized out>)
    at /home/sds/src/emacs/trunk/src/keyboard.c:1156
#22 0x000000000055f1c8 in internal_catch (tag=<value optimized out>, 
    func=0x4f9320 <command_loop_2>, arg=11796258)
    at /home/sds/src/emacs/trunk/src/eval.c:1247
#23 0x00000000004f999c in command_loop ()
    at /home/sds/src/emacs/trunk/src/keyboard.c:1135
#24 recursive_edit_1 () at /home/sds/src/emacs/trunk/src/keyboard.c:756
#25 0x00000000004f9ae2 in Frecursive_edit ()
    at /home/sds/src/emacs/trunk/src/keyboard.c:820
#26 0x00000000004e9817 in main (argc=2, argv=0x7fff276f8768)
    at /home/sds/src/emacs/trunk/src/emacs.c:1698
(gdb) up
#1  0x000000000055f0ff in internal_condition_case (
    bfun=0x4eed80 <safe_run_hooks_1>, handlers=11796306, 
    hfun=0x4f9350 <safe_run_hooks_error>)
    at /home/sds/src/emacs/trunk/src/eval.c:1468
1468        abort ();

no lisp backtrace

-- 
Sam Steingold (http://sds.podval.org/) on CentOS release 5.6 (Final) X 11.0.60900031
http://thereligionofpeace.com http://www.PetitionOnline.com/tap12009/
http://jihadwatch.org http://memri.org http://www.memritv.org
A computer scientist is someone who fixes things that aren't broken.

bug#9081: 24.0.50; crash in X

[Chong Yidong]

Thanks for the bug report.

This crash occurs when Emacs is waiting for a reply to a selection
event.  During this waiting period, X errors are being caught.  One of
the events that come in during this period caused Lisp code to run via
safe_run_hooks, which triggers the following in internal_condition_case:

  /* Since Fsignal will close off all calls to x_catch_errors,
     we will get the wrong results if some are not closed now.  */
#if HAVE_X_WINDOWS
  if (x_catching_errors ())
    abort ();
#endif

Anyone have a clue about the rationale of the above code?  It was
inserted by RMS on 2006-03-11, but I don't see anything in the
emacs-devel or bug-gnu-emacs archives discussing this.  Note that there
is no corresponding check in internal_lisp_condition_case.  Note also
that x_reply_selection_request has a record_unwind_protect to uncatch X
errors.

If no one has any idea, I propose to remove those lines.

bug#9081: 24.0.50; crash in X

[Chong Yidong]

Chong Yidong <cyd@stupidchicken.com> writes:

>   /* Since Fsignal will close off all calls to x_catch_errors,
>      we will get the wrong results if some are not closed now.  */
> #if HAVE_X_WINDOWS
>   if (x_catching_errors ())
>     abort ();
> #endif
>
> Anyone have a clue about the rationale of the above code?  It was
> inserted by RMS on 2006-03-11, but I don't see anything in the
> emacs-devel or bug-gnu-emacs archives discussing this.  Note that there
> is no corresponding check in internal_lisp_condition_case.  Note also
> that x_reply_selection_request has a record_unwind_protect to uncatch X
> errors.

I removed those lines; let us see if that causes any trouble.