Re: crash: divide by 0

[B. Anyos]

It seems that
        FRAME_SPACE_WIDTH (it->f)

is zero (i.e. it->f->space_width is zero)
While debugging and changing this to anything but zero
the crash goes away.

This space_width - according to the Changelog - is a
relatively new code (added by Kenichi Handa on 2004-12-30)

Could it be the problem that it is not initialized ?

Bela


B. Anyos said the following on 1/3/2005 17:28 PM:
> I can confirm this, too.
> I had a CVS update and a 'make bootstrap' today.
> To reproduce the problem just try to load a file.
> It crashes every time.
> I also have a Windows client. Does this happen on other platform(s) ?
>
> Regards,
>  Bela
>
> Here's the stack dump:
>
> --------------------------
> x_produce_glyphs(it * 0x0009f010) line 19222 + 4 bytes
> display_line(it * 0x00000060) line 14775 + 13 bytes
> try_window(int -2127184384, text_pos {...}) line 12387 + 12 bytes
> redisplay_window(int -2127184384, int 0) line 12027 + 10 bytes
> redisplay_window_0(int -2127184384) line 10698 + 10 bytes
> internal_condition_case_1(int (void)* 0x01050f61 
> redisplay_window_0(int), int -2127184384, int -1590823544, int (void)* 
> 0x01050f4a redisplay_window_error(void)) line 1426 + 53 bytes
> redisplay_windows(int -2127184384) line 10679 + 29 bytes
> redisplay_internal(int 0) line 10259 + 8 bytes
> redisplay() line 9489 + 7 bytes
> read_char(int 1, int 9, int * 0x0082fc7c, int 556606464, int * 
> 0x0082fce0) line 2539
> read_key_sequence(int * 0x0082fd7c, int 30, int 556606464, int 0, int 1, 
> int 1) line 8820 + 36 bytes
> command_loop_1() line 1532 + 40 bytes
> internal_condition_case(int (void)* 0x0100e792 command_loop_1(void), int 
> 556692080, int (void)* 0x0100e3e9 cmd_error(void)) line 1386
> command_loop_2() line 1308 + 21 bytes
> internal_catch(int 556688344, int (void)* 0x0100e65d 
> command_loop_2(void), int 556606464) line 1144 + 6 bytes
> command_loop() line 1292
> recursive_edit_1() line 984 + 5 bytes
> Frecursive_edit() line 1046
> main() line 1763 + 5 bytes
> EMACS! mainCRTStartup + 180 bytes
> KERNEL32! 77e8141a()
> --------------------------
>
> Sam Steingold said the following on 1/3/2005 16:58 PM:
>
> > GNU Emacs 21.3.50.1 (i386-msvc-nt5.0.2195)
> >  of 2005-01-03 on WINSTEINGOLDLAP
> > --with-msvc (12.00)
> >
> > with the current CVS HEAD I am getting consistent crashes (division by
> > 0) in xdisp.c:19222 (apparently "tab_width" is 0)
>
>
> _______________________________________________
> Emacs-devel mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/emacs-devel