Re: [GIT PULL rcu/next] fixes and breakup of memory-barrier-decrease patch

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

On Sun, May 22, 2011 at 11:04:40AM +0200, Ingo Molnar wrote:
> 
> * Paul E. McKenney <paulmck@linux.vnet.ibm.com> wrote:
> 
> > > I mean, without Frederic's patch we are getting very long hangs due to the 
> > > barrier patch, right?
> > 
> > Yes.  The reason we are seeing these hangs is that HARDIRQ_ENTER() invoked 
> > irq_enter(), which calls rcu_irq_enter() but that the matching HARDIRQ_EXIT() 
> > invoked __irq_exit(), which does not call rcu_irq_exit(). This resulted in 
> > calls to rcu_irq_enter() that were not balanced by matching calls to 
> > rcu_irq_exit().  Therefore, after these tests completed, RCU's dyntick-idle 
> > nesting count was a large number, which caused RCU to conclude that the 
> > affected CPU was not in dyntick-idle mode when in fact it was.
> > 
> > RCU would therefore incorrectly wait for this dyntick-idle CPU.
> > 
> > With Frederic's patch, these tests don't ever call either rcu_irq_enter() or 
> > rcu_irq_exit(), which works because the CPU running the test is already 
> > marked as not being in dyntick-idle mode.
> > 
> > So, with Frederic's patch, the rcu_irq_enter() and rcu_irq_exit() calls are 
> > balanced and things work.
> > 
> > The reason that the imbalance was not noticed before the barrier patch was 
> > applied is that the old implementation of rcu_enter_nohz() ignored the 
> > nesting depth.  This could still result in delays, but much shorter ones.  
> > Whenever there was a delay, RCU would IPI the CPU with the unbalanced nesting 
> > level, which would eventually result in rcu_enter_nohz() being called, which 
> > in turn would force RCU to see that the CPU was in dyntick-idle mode.
> > 
> > Hmmm...  I should add this line of reasoning to one of the commit logs, 
> > shouldn't I?  (Added it.  Which of course invalidates my pull request.)
> 
> Well, the thing i was missing from the tree was Frederic's fix patch. Or was 
> that included in one of the commits?

Ah!  I don't see any evidence of anyone else having taken it, so I just
now queued it.

> I mean, if we just revert the revert, we reintroduce the delay, no matter who 
> is to blame - not good! :-)

Good point!  ;-)

> > > Even if the barrier patch is not to blame - somehow it still managed to 
> > > produce these hangs - and we do not understand it yet.
> > 
> > >From Yinghai's message https://lkml.org/lkml/2011/5/12/465, I believe
> > that the residual delay he is seeing is not due to the barrier patch,
> > but rather due to a26ac2455 (move TREE_RCU from softirq to kthrea).
> > 
> > More on this below.
> 
> Ok - we can treat that regression differently. Also, that seems like a much 
> shorter delay, correct? The delays fixed by Frederic's patch were huge (i think 
> i saw a 1 hour delay once) - they were essentially not delays but hangs.

Yes, the delays fixed by Frederic's patch were hours in length, while
the remaining delays measure in seconds.  And I am looking at the code
and at how grace-period duration has varied, so hope to get to the
bottom of it in a few days.  Hopefully sooner.  ;-)

							Thanx, Paul