Jason Thorpe wrote:
> On Thu, 12 Mar 1998 16:59:37 +0100 (MET) 
>  Jaromir Dolecek <dolecek@ics.muni.cz> wrote:
> 
>  > Well, it's vanilla 1.3 kernel, not current.
>  > Unfortunately, I can't get the arguments of vm_fault() from crash dump.
> 
> Try:
> 
> 	(gdb) proc curproc
> 	(gdb) bt
> 
> ...that might help a little.

You are right, it helps just a little ;)

#0  0xf020e548 in sysent ()
#1  0xf4881f40 in ?? ()
#2  0xf012220d in tsleep (ident=0xf0123aad, priority=-259349696, 
    wmesg=0xf092be00 "\220h\"ð", timo=-192405624)
    at ../../../../kern/kern_synch.c:359

Note I'm using 1.3 gdb - 1.2.1 ``gdb -k netbsd.gdb netbsd.0.core '' crashes.

>  > I'try to write down the arguments too next time this happens.
>  > Should I find out some more info ?
> 
> Well, you got the PC from the DDB traceback, so try using gdb's "list"
> command like:
> 
> 	(gdb) list *(vnode_pager_unmount+0x<whatever>)

(gdb) list *(vnode_pager_umount+0x1d)
0xf01c6ea1 is in vnode_pager_umount (../../../../vm/vnode_pager.c:465).
460                     /*
461                      * Save the next pointer now since uncaching may
462                      * terminate the object and render pager invalid
463                      */
464                     npager = pager->pg_list.tqe_next;
465                     vp = ((vn_pager_t)pager->pg_data)->vnp_vp;
466                     if (mp == (struct mount *)0 || vp->v_mount == mp) {
467                             VOP_LOCK(vp);
468                             (void) vnode_pager_uncache(vp);
469                             VOP_UNLOCK(vp);

Is there any way I can get the contents the structure used
on line 465 in gdb ? I'm unable to find it in help and
I'm not very familiar with this post-death debugging ;).

-- 
Jaromir Dolecek <dolecek@ics.muni.cz>	http://www.ics.muni.cz/~dolecek/
-------------------------------------------------------------------------
It is better never to have been born.  But who among us has such luck?
One in a million, perhaps.