Re: [PATCH RFC tip/core/rcu 7/7] rcu: Quiet RCU-lockdep warnings involving interrupt disabling

["Paul E. McKenney" <paulmck@linux.vnet.ibm.com>]

On Tue, Dec 06, 2011 at 08:04:55AM -0800, Paul E. McKenney wrote:
> On Tue, Dec 06, 2011 at 07:26:45AM -0500, Steven Rostedt wrote:
> > On Tue, 2011-12-06 at 11:32 +0100, Peter Zijlstra wrote:
> > 
> > > > > 
> > > > > > Maybe we could teach might_sleep() about this special case?
> > > > > 
> > > > > Sounds horrid.
> > > > 
> > > > Maybe, any alternative?
> > > 
> > > Maybe someone explain this mess first?
> > 
> > Me too, because this looks like a hack that's just like a lie. The first
> > lie to be said gets you out of a bit of trouble, but then something else
> > happens based on that lie, in which you need to make another bigger lie.
> > This new lie affects more people and requires new more ingenious lies to
> > control the chaos. But eventually the lies required to keep everything
> > going become so overwhelming that it all blows up in your face and you
> > end up looking like a jackass.
> > 
> > Why is rcu using rt_mutex_lock() in strange ways? It's lying about its
> > use. And now this patch is the bigger lie to get around the issues of
> > the first lie. Eventually this code will continue to expand largely
> > based on these lies and will explode in our faces, and I feel sorry for
> > the poor jackass that needs to fix it.
> > 
> > Perhaps the real answer is that we need to create an API for priority
> > inheritance, that things like RCU could use. Attach a task that another
> > task requires to finish something and boost the priority of that task.
> > Maybe even completions could use such a thing?
> 
> I would be OK with that -- that was in fact the approach I was taking
> when I was advised to use mutexes instead.  ;-)

But in the spirit of full disclosure -- moving from explicit priority
boosting to mutexes simplified the code greatly.  It eliminated a bunch
of races between one task boosting and the boostee exiting its RCU
read-side critical section.  The difficulty is (or maybe was) the need
to block when boosting priority, which meant that there were races
where the booster marked the task, dropped the rcu_node ->lock,
the boostee exited its RCU read-side critical section and uselessly
deboosted itself, then the booster incorrectly did the boost.

But perhaps I was just suffering a failure of imagination.

							Thanx, Paul