Re: emacs from head segfaults when run with -nw

[Eli Zaretskii <eliz@gnu.org>]

> Date: Mon, 05 Apr 2010 17:06:41 +0300
> From: Eli Zaretskii <eliz@gnu.org>
> Cc: emacs-devel@gnu.org
> 
> In my case, I found so far that one of the submaps of
> Buffer-menu-mode-map is corrupted (a NULL pointer):
> 
>   #2  0x01068682 in mark_char_table (ptr=0x2fee200) at alloc.c:5393
>   5393                mark_char_table (XVECTOR (val));
>   (gdb) p *ptr
>   $33 = {
>     size = 3222274066,
>     next = 0x2ff3000,
>     contents = {4}
>   }
>   (gdb) p ptr->contents[0]
>   $34 = 4
>   (gdb) p size
>   $35 = 18
>   (gdb) p ptr->contents[1]
>   $36 = 0
>   (gdb) p ptr->contents[2]
>   $37 = 50252549
>   (gdb) xtype
>   Lisp_Vectorlike
>   PVEC_SUB_CHAR_TABLE
>   (gdb) xvector
>   $38 = (struct Lisp_Vector *) 0x0       <<<<<<<<<<<<<<<<
>   Cannot access memory at address 0x0    <<<<<<<<<<<<<<<<
>   (gdb)

That was incorrect, sorry.  Here is a more accurate account:

  #2  0x01068682 in mark_char_table (ptr=0x2fee200) at alloc.c:5393
  5393                mark_char_table (XVECTOR (val));
  (gdb) p val
  $29 = 0
  (gdb) down
  #1  0x01068649 in mark_char_table (ptr=0x2fecb00) at alloc.c:5396
  5396            mark_object (val);
  (gdb) p size
  $30 = 34
  (gdb) p i
  $31 = 8
  (gdb) p *ptr->contents@34
  $32 = {8, 0, 50277381, 46032898, 46032898, 46032898, 46032898, 46024706,
    46032898 <repeats 26 times>}
  (gdb) p ptr->contents[7]
  $33 = 46024706
  (gdb) p *$33
  $34 = 702
  (gdb) xtype
  Lisp_Cons
  (gdb) xcons
  $35 = (struct Lisp_Cons *) 0x2b8       <<<<<<<<<<<<<<<<
  Cannot access memory at address 0x2b8  <<<<<<<<<<<<<<<<
  (gdb) p ptr->contents[6]
  $36 = 46032898
  (gdb) xtype
  Lisp_Symbol
  (gdb) xsymbol
  $37 = (struct Lisp_Symbol *) 0x2be6800
  "nil"

So 46032898 is nil, but what is the mysterious 46024706?  Observe:

  (gdb) p/x 46032898
  $41 = 0x2be6802
  (gdb) p/x 46024706
  $42 = 0x2be4802

So 46024706 is nil with one of its bits reset!  What does it mean?  A
bad memory chip in my computer?

In the "good" session (unoptimized build), btw, the contents of the
same char-table is dandy:

  #0  mark_char_table (ptr=0x3011b00) at alloc.c:5388
  5388          if (INTEGERP (val) || SYMBOLP (val) && XSYMBOL(val)->gcmarkbit)
  (gdb) p val
  $56 = 46221314
  (gdb) xtype
  Lisp_Symbol
  (gdb) xsymbol
  $57 = (struct Lisp_Symbol *) 0x2c14800
  "nil"
  (gdb) p size
  $58 = 34
  (gdb) p *ptr->contents@34
  $59 = {8, 0, 50483205, 46221314 <repeats 31 times>}

46221314 is nil, as you see above.

Ideas?