Re: emacs-22.1 with GTK dumps core when Gnome wigets clicked

[Gary Mills]

On Sun, Jun 17, 2007 at 03:41:26PM +0900, YAMAMOTO Mitsuharu wrote:
> 
> Gosh!  I noticed that the mutex code in gmalloc.c did not take effect
> at all because `HAVE_GTK_AND_PTHREAD' is checked before including
> config.h.  Could you try moving the following part at the beginning of
> gmalloc.c to just before the first occurrence of '#ifdef USE_PTHREAD'?
> 
> #ifdef HAVE_GTK_AND_PTHREAD
> #define USE_PTHREAD
> #endif

Well, I got a different core dump this time...

$ dbx src/emacs core
...
t@2 (l@2) terminated by signal KILL (Killed)
0xfd3416b8: __pollsys+0x0008:   blu      _cerror        ! 0xfd2a2600
(dbx) lwps
  l@1 LWP suspended in _write()
 >l@2 LWP suspended in __pollsys()
o l@3 signal SIGSEGV in 0x0()
(dbx) lwp l@3
Current function is malloc (optimized)
  967     return (__malloc_hook != NULL ? *__malloc_hook : _malloc_internal) 
(size);
t@3 (l@3) stopped in (unknown) at 0x0
0x00000000:     <bad address 0x0>
(dbx) regs
current thread: t@3
current frame:  [2]
g0-g1    0x00000000 0x00000000 0x00000000 0x00678000
g2-g3    0x00000000 0x0000006d 0x00000000 0x0084eec0
g4-g5    0x00000000 0x003b2000 0x00000000 0x003a8800
g6-g7    0x00000000 0x00000000 0x00000000 0xfa0f0000
o0-o1    0x00000000 0x00000010 0x00000000 0x00001000
o2-o3    0x00000000 0xfd36ec80 0x00000000 0xfa0f0000
o4-o5    0x00000000 0x00368ae0 0x00000000 0x00000000
o6-o7    0x00000000 0xfa14f5e0 0x00000000 0x00221cc0
l0-l1    0x00000000 0x00000056 0x00000000 0x00000000
l2-l3    0x00000000 0xfd7f2000 0x00000000 0xff39c000
l4-l5    0x00000000 0x00000004 0x00000000 0x00000000
l6-l7    0x00000000 0x00000001 0x00000000 0x001aeba8
i0-i1    0x00000000 0x00000010 0x00000000 0x003ad400
i2-i3    0x00000000 0x00000000 0x00000000 0x003ad400
i4-i5    0x00000000 0x00000000 0x00000000 0x00000001
i6-i7    0x00000000 0xfa14f640 0x00000000 0x0022286c
y        0x00000000 0x00000000
ccr      0x00000000 0xfe001000
pc       0x00000000 0x00221cc0:malloc+0x58      call     %i2
npc      0x00000000 0x00000004:0x00000004       <bad address 0x4>

-- 
-Gary Mills-    -Unix Support-    -U of M Academic Computing and Networking-