On Sun, Jul 06, 2014 at 01:49:04PM -0700, Neel Natu wrote:
> Hi Steve,
>
> On Sun, Jul 6, 2014 at 8:46 AM, Steve Wills <swi...@freebsd.org> wrote:
> > I should have noted this system is running in bhyve. Also I'm told this
> > panic
> > may be related to the fact that the system is running in bhyve.
> >
> > Looking at it a little more closely:
> >
> > (kgdb) list *__mtx_lock_sleep+0xb1
> > 0xffffffff809638d1 is in __mtx_lock_sleep
> > (/usr/src/sys/kern/kern_mutex.c:431).
> > 426 * owner stops running or the state of the lock
> > changes.
> > 427 */
> > 428 v = m->mtx_lock;
> > 429 if (v != MTX_UNOWNED) {
> > 430 owner = (struct thread *)(v &
> > ~MTX_FLAGMASK);
> > 431 if (TD_IS_RUNNING(owner)) {
> > 432 if (LOCK_LOG_TEST(&m->lock_object,
> > 0))
> > 433 CTR3(KTR_LOCK,
> > 434 "%s: spinning on %p
> > held by %p",
> > 435 __func__, m, owner);
> > (kgdb)
> >
> > I'm told that MTX_CONTESTED was set on the unlocked mtx and that
> > MTX_CONTENDED
> > is spuriously left behind, and to ask how lock prefix is handled in bhyve.
> > Any
> > of that make sense to anyone?
> >
>
> Regarding the lock prefix: since bhyve only supports hardware that has
> nested paging, the hypervisor doesn't get in the way of instructions
> that access memory. This includes instructions with lock prefixes or
> any other prefixes for that matter. If there is a VM exit due to a
> nested page fault then the faulting instruction is restarted after
> resolving the fault.
>
> Having said that, there are more plausible explanations that might
> implicate bhyve: incorrect translations in the nested page tables,
> stale translations in the TLB etc.
>
> Do you have a core file for the panic? It would be very useful to
> debug this further.No, unfortunately I did not have swap or dumpdev setup at the time so I was unable to get a core dump from the crashed kernel. (Bhyve did not crash.) I've setup swap in the VM and set the dumpdev as well, so if it happens again I should get a core. Steve
pgpqBGlv4h2w9.pgp
Description: PGP signature
