4BSD Scheduler Problem on 5.3

[John Baldwin]

On Thursday 31 March 2005 08:03 pm, Robert Watson wrote:
> On Thu, 31 Mar 2005, John Baldwin wrote:
> > On Thursday 31 March 2005 03:38 pm, William Michael Grim wrote:
> > > Hello.
> > >
> > > I keep having kernel panics every couple weeks on my system.  It occurs
> > > in the sched_switch() function.  There are several other statements in
> > > the backtrace involving "??"; what are those?
> > >
> > > I have attached the dump output and system info to this email.  Any
> > > feedback would be helpful.
> > >
> > > Thanks so much for your help.
> >
> > The real trace ends with Xint0x80_syscall().  The rest after that is
> > garbage memory.  Your real problem is in exit1() or ttywakeup().  Since
> > ttywakeup() doesn't call exit1() (AFAIK), the exit1() frame is probably
> > bogus (gdb doesn't grok trapframes maybe?) and the real bug is a NULL
> > pointer deref in ttywakeup().  Perhaps it's a bug in the ptc driver?
> > (ptcopen is in the trace).  What is the ptc driver anyway?
>
> I think we have a race in -STABLE relating to tty wakeups and
> open/close/device teardown.  I've seen a panic relating to sio during a
> tty close on RELENG_5 about 5-6 months ago, but was unable to get a dump.
> Scott has since fixed dumps with twe, but I've not yet been able to get
> the bug to recur.  I'll give it another try.

Sounds very plausible.  Does Poul-Henning have any ideas?

-- 
John Baldwin <jhb at FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve"  =  http://www.FreeBSD.org