Re: [V2 PATCH] kthread_work: Make lockdep happy

[Andy Walls <awalls@md.metrocast.net>]

On Tue, 2010-12-21 at 13:59 +0100, Tejun Heo wrote:
> Hello,
> 
> On Tue, Dec 21, 2010 at 12:40:50PM +0800, Yong Zhang wrote:
> > From: Yong Zhang <yong.zhang0@gmail.com>
> > Subject: [V2 PATCH] kthread_work: Make lockdep happy
> > 
> > spinlock in kthread_worker and wait_queue_head in kthread_work
> > both should be lockdep sensible.
> > So change the interface to make it suiltable for CONFIG_LOCKDEP.
> > 
> > Reported-by: Nicolas <nicolas.mailhot@laposte.net>
> > Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
> > Cc: Tejun Heo <tj@kernel.org>
> > Cc: Andrew Morton <akpm@linux-foundation.org>
> > Cc: Andy Walls <awalls@md.metrocast.net>
> > ---
> > Changes from V1:
> >  *According to Tejun, kthread_worker could be defined on stack,
> >   So introduce DEFINE_KTHREAD_WORKER_ONSTACK.
> >  *Change wrong setting to kthread_work->task. Thanks Adny for
> >   pointing it.
> >  *including some minor issue.
> > 
> > BTW, only passed build.
> 
> If somebody can confirm this makes lockdep behave correctly, I'll
> route it through the wq tree.

I will attempt to test later tonight with at least one ivtv (PVR-350)
card installed.  If I have time, I'll test with two cards (PVR-350 and
PVR-150) installed in the machine, since that will create two different
kthread_workers, each with their own lock.

How can I dump information on lockdep spinlock tracking to verify that
the two distinct locks are tracked separately by lockdep?

Regards,
Andy