Re: [PATCH RFC v1 1/2] rcu/tree: Add basic support for kfree_rcu batching

[Joel Fernandes]

On Fri, Aug 09, 2019 at 08:38:14PM -0700, Paul E. McKenney wrote:
> On Fri, Aug 09, 2019 at 10:42:32PM -0400, Joel Fernandes wrote:
> > On Wed, Aug 07, 2019 at 10:52:15AM -0700, Paul E. McKenney wrote:
> > [snip] 
> > > > > > @@ -3459,6 +3645,8 @@ void __init rcu_init(void)
> > > > > >  {
> > > > > >  	int cpu;
> > > > > >  
> > > > > > +	kfree_rcu_batch_init();
> > > > > 
> > > > > What happens if someone does a kfree_rcu() before this point?  It looks
> > > > > like it should work, but have you tested it?
> > > > > 
> > > > > >  	rcu_early_boot_tests();
> > > > > 
> > > > > For example, by testing it in rcu_early_boot_tests() and moving the
> > > > > call to kfree_rcu_batch_init() here.
> > > > 
> > > > I have not tried to do the kfree_rcu() this early. I will try it out.
> > > 
> > > Yeah, well, call_rcu() this early came as a surprise to me back in the
> > > day, so...  ;-)
> > 
> > I actually did get surprised as well!
> > 
> > It appears the timers are not fully initialized so the really early
> > kfree_rcu() call from rcu_init() does cause a splat about an initialized
> > timer spinlock (even though future kfree_rcu()s and the system are working
> > fine all the way into the torture tests).
> > 
> > I think to resolve this, we can just not do batching until early_initcall,
> > during which I have an initialization function which switches batching on.
> > >From that point it is safe.
> 
> Just go ahead and batch, but don't bother with the timer until
> after single-threaded boot is done.  For example, you could check
> rcu_scheduler_active similar to how sync_rcu_exp_select_cpus() does.
> (See kernel/rcu/tree_exp.h.)

Cool, that works nicely and I tested it. Actually I made it such that we
don't need to batch even, before the scheduler is up. I don't see any benefit
of that unless we can see a kfree_rcu() flood happening that early at boot
which seems highly doubtful as a real world case.

> If needed, use an early_initcall() to handle the case where early boot
> kfree_rcu() calls needed to set the timer but could not.

And it would also need this complexity of early_initcall.

> > Below is the diff on top of this patch, I think this should be good but let
> > me know if anything looks odd to you. I tested it and it works.
> 
> Keep in mind that a call_rcu() callback can't possibly be invoked until
> quite some time after the scheduler is up and running.  So it will be
> a lot simpler to just skip setting the timer during early boot.

Sure. Skipping batching would skip the timer too :-D

If in the future, batching is needed this early, then I am happy to add an
early_initcall to setup the timer for any batched calls that could not setup
the timer. Hope that is ok with you?

thanks,

 - Joel

[snip]