Re: crashes on x86_64

[Kim F. Storm]

Reiner Steib <address@hidden> writes:

> Hi,
>
> during the last few days (I didn't build for about one month, so the
> problem might have started earlier), Emacs built on x86_64 (AMD 64
> bit) crashes quite often (several times per day).

In lisp.h, can you try to enable extended GC checking by
changing the 0 to 1 in the following line in lisp.h:

#if 0
/* Define this temporarily to hunt a bug.  If defined, the size of

Then do:
cd src
rm *.o
make emacs

and try again.

Thanks.


>
> Here's some gdb output:
>
> Program received signal SIGSEGV, Segmentation fault.
> [Switching to Thread 182940272320 (LWP 31751)]
> make_uninit_multibyte_string (nchars=10, nbytes=10) at 
> [...]/emacs/src/alloc.c:1884
> 1884      string_free_list = NEXT_FREE_LISP_STRING (s);
>
> (gdb) bt full
> #0  make_uninit_multibyte_string (nchars=10, nbytes=10) at 
> [...]/emacs/src/alloc.c:1884
>         string = Variable "string" is not available.

Can you investigate why there is no backtrace.

Perhaps you need to compile with -g option (in src/Makefile, add -g to the 
CFLAGS= line).

> (gdb) xbacktrace 
> (gdb) p string_free_list
> $1 = (struct Lisp_String *) 0xc8

When I got crashes myself (a long time ago), the bogus pointer value
often was equal to 0xc8.   But I never managed to find out where
it came from.

> (gdb) pr string_free_list

Doing pr on a process which crashed in GC often fails...

-- 
Kim F. Storm <address@hidden> http://www.cua.dk