On Mon, Nov 12, 2001 at 06:03:25PM -0000, eeh@netbsd.org wrote:
> 
> | On Mon, Nov 12, 2001 at 05:52:34PM -0000, eeh@netbsd.org wrote:
> | > It might be usefult to find out what instructions you have at 0100.8958 and 
> | > 0115.5ebc.  From the looks of things something nasty has probably happened
> | > to your stack.
> |
> | How can I do this ?
> 
> You can run gdb on the kernel and do:
> 
> gdb> x/i <addr>

OK, no problems:
(gdb) x/i 0x01008958
0x1008958 <datafault+92>:       stx  %g1, [ %g6 + 0xc8 ]
(gdb) x/i 0x01155ebc
0x1155ebc <uvm_page_init_buckets+56>:   clr  [ %g2 + %o0 ]
(gdb) l *(datafault+92)
No source file for address 0x1008958.
(gdb) l *(uvm_page_init_buckets+56)
0x1155ebc is in uvm_page_init_buckets (/home/cvs.netbsd.org/src/sys/arch/sparc64/compile/SWING32/../../../../uvm/uvm_page.c:217).
warning: Source file is more recent than executable.
 XXX hum, I just ran a CVS update, I hope this didn't change this file
212     
213     static void
214     uvm_page_init_buckets(struct pgfreelist *pgfl)
215     {
216             int color, i;
217     
218             for (color = 0; color < uvmexp.ncolors; color++) {
219                     for (i = 0; i < PGFL_NQUEUES; i++) {
220                             TAILQ_INIT(&pgfl->pgfl_buckets[
221                                 color].pgfl_queues[i]);


--
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
--