On 30-Dec-00 Szilveszter Adam wrote:
> On Sat, Dec 30, 2000 at 12:08:48PM -0800, John Baldwin wrote:
>> 
>> On 30-Dec-00 Michael Harnois wrote:
>> > panic: lockable mtx_enter() of lockmgr interlock when not legal @
>> > ../../kern/kern_lock.c: 247
>> > 
>> > which is
>> > 
>> > mtx_enter(lkp->lk_interlock, MTX_DEF);
>> 
>> We need to know where interrupts were disabled (since that is what makes the
>> blockable mtx_enter() not legal).  The most likely reason is that sched_lock
>> is
>> held.  Take a crash dump if you can, and then examine the
>> '__mtx_debug_sched_lock' variable.  Esp. the mtxd_line and mtd_file members
>> which tell us where sched_lock was last acquired.
> 
> John, 
> 
> I can already supply the needed information.
> 
> (kgdb) print __mtx_debug_sched_lock
> $1 = {mtxd_witness = 0xc03d98f0, mtxd_held = {le_next = 0x0, le_prev =
> 0x0},
>   mtxd_file = 0xc033c27c "../../kern/kern_sig.c", mtxd_line = 1111,
>   mtxd_description = 0xc0368cbf "sched lock"}

That's line 1111 of sys/kern/kern_sig.c:

        /*
         * Defer further processing for signals which are held,
         * except that stopped processes must be continued by SIGCONT.
         */
        mtx_enter(&sched_lock, MTX_SPIN);
        if (action == SIG_HOLD && (!(prop & SA_CONT) || p->p_stat != SSTOP)) {
                mtx_exit(&sched_lock, MTX_SPIN);
                return;
        }


Hmm, ok, which signal is this?

-- 

John Baldwin <[EMAIL PROTECTED]> -- http://www.FreeBSD.org/~jhb/
PGP Key: http://www.Baldwin.cx/~john/pgpkey.asc
"Power Users Use the Power to Serve!"  -  http://www.FreeBSD.org/


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message

Reply via email to