This is the mail archive of the cygwin mailing list for the Cygwin project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [ANNOUNCEMENT] Updated: emacs-24.3.93-1 [TEST]

On 8/21/2014 8:12 AM, Achim Gratz wrote:
Achim Gratz <Stromeko <at>> writes:
There are still crashes of emacs-w32 that take down Emacs completely
(sometimes with but mostly without it asking for the debugger to be
attached).  So far these have all occured directly in response to
keyboard input (as in the window disappears while the key is still
pressed down).  Let me know if the stackdump would be useful or if I
need to wait out an occasion for when it asks for the debugger to attach
and what to do then.

Just had another one w/ the message and managed to attach the debugger.  I
don't know if the back trace makes any sense.  When entering the debugger I
think the crash was indicated to be inside the cyggobject library (sorry,
that line scrolled off the screen), but that may just have been the message

Thread 1 (Thread 10724.0x3340):
#0  0x000007fefdba3ca3 in KERNELBASE!DebugBreak () from
#1  0x0000000100577794 in emacs_abort () at
#2  0x00000001004ca456 in overlays_at (pos=pos@entry=23459,
extend=extend@entry=false, vec_ptr=0x42a2d0, vec_ptr@entry=0x100927f60 <Qnil>,
     len_ptr=len_ptr@entry=0x42a2d8, next_ptr=next_ptr@entry=0x0,
prev_ptr=prev_ptr@entry=0x0, change_req=change_req@entry=false)
     at /usr/src/debug/emacs-24.3.93-1/src/buffer.c:2854
#3  0x000000010056968f in get_char_property_and_overlay
(position=position@entry=93836, prop=prop@entry=4305239426,
     overlay=overlay@entry=0x0) at
#4  0x00000001005697b8 in Fget_char_property (position=position@entry=93836,
prop=prop@entry=4305239426, object=object@entry=25772655621)
     at /usr/src/debug/emacs-24.3.93-1/src/textprop.c:703
#5  0x000000010056622a in get_local_map (position=<optimized out>,
position@entry=23459, buffer=0x6002b8400, type=4305239426)
     at /usr/src/debug/emacs-24.3.93-1/src/intervals.c:2234
#6  0x00000001004c0688 in Fcurrent_active_maps (olp=4304926818,
position=position@entry=4304926770) at
#7  0x00000001004c09f6 in Fkey_binding (key=key@entry=4304949141,
accept_default=accept_default@entry=4304926770, no_remap=4304926818,
     position=4304926770) at /usr/src/debug/emacs-24.3.93-1/src/keymap.c:1721
#8  0x00000001004c0ba8 in Fcommand_remapping (command=<optimized out>,
position=<optimized out>, keymaps=<optimized out>)
     at /usr/src/debug/emacs-24.3.93-1/src/keymap.c:1248
#9  0x00000001004c0bca in Fcommand_remapping (command=<optimized out>,
position=<optimized out>, keymaps=<optimized out>)
     at /usr/src/debug/emacs-24.3.93-1/src/keymap.c:1253
#10 0x00000001004b8cfa in read_key_sequence (keybuf=keybuf@entry=0x42a710,
prompt=<optimized out>, dont_downcase_last=dont_downcase_last@entry=false,
     prevent_redisplay=prevent_redisplay@entry=false, bufsize=30) at
#11 0x00000001004b9c54 in command_loop_1 () at
#12 0x000000010051a0dd in internal_condition_case
(bfun=bfun@entry=0x1004b9a30 <command_loop_1>, handlers=<optimized out>,
     hfun=hfun@entry=0x1004b0560 <cmd_error>) at
#13 0x00000001004ab9ea in command_loop_2 (ignore=ignore@entry=4304926770) at
#14 0x0000000100519fdc in internal_catch (tag=4304985026,
func=func@entry=0x1004ab9c0 <command_loop_2>, arg=4304926770)
     at /usr/src/debug/emacs-24.3.93-1/src/eval.c:1118
#15 0x00000001004b0154 in command_loop () at
#16 recursive_edit_1 () at /usr/src/debug/emacs-24.3.93-1/src/keyboard.c:777
#17 0x00000001004b0476 in Frecursive_edit () at
#18 0x00000001005b5953 in main (argc=<optimized out>, argv=<optimized out>)
at /usr/src/debug/emacs-24.3.93-1/src/emacs.c:1646

Thanks for the report.  The abort comes from here (in buffer.c):

  for (tail = current_buffer->overlays_before; tail; tail = tail->next)
      ptrdiff_t startpos, endpos;

      XSETMISC (overlay, tail);

      start = OVERLAY_START (overlay);
      end = OVERLAY_END (overlay);
      endpos = OVERLAY_POSITION (end);   <<<<<<<<<<<<<<<<<<<<<

where OVERLAY_POSITION is defined as follows in buffer.h:

 (MARKERP (P) ? marker_position (P) : (emacs_abort (), 0))

I'm not at all familiar with the code in this area, but I'm copying Eli in the hopes that he can see something in this backtrace. I don't know if this will turn out to be another one of those backtraces that don't make sense.

Do you remember what you were doing at the time of the crash?


Problem reports:
Unsubscribe info:

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]