Re: [PATCH] a local-timer-free version of RCU

[Mathieu Desnoyers <mathieu.desnoyers@efficios.com>]

* Mathieu Desnoyers (mathieu.desnoyers@efficios.com) wrote:
> > > +/**
> > > + * rcu_read_unlock - marks the end of an RCU read-side critical section.
> > > + * Check if a RCU batch was started while we were in the critical
> > > + * section.  If so, call rcu_quiescent() join the rendezvous.
> > > + *
> > > + * See rcu_read_lock() for more information.
> > > + */
> > > +void __rcu_read_unlock(void)
> > > +{
> > > +	struct rcu_data *r;
> > > +	int	cpu, flags;
> > > +
> 
> Another memory barrier would be needed here to ensure that the memory accesses
> performed within the C.S. are not reordered wrt nest_count decrement.

Nevermind. xchg() acts as a memory barrier, and nest_count is only ever touched
by the local CPU. No memory barrier needed here.

Thanks,

Mathieu

> 
> > > +	cpu = smp_processor_id();
> > > +	r = &per_cpu(rcu_data, cpu);
> > > +	if (--r->nest_count == 0) {
> > > +		flags = xchg(&r->flags, 0);
> > > +		if (flags & DO_RCU_COMPLETION)
> > > +			rcu_quiescent(cpu);
> > > +	}
> > > +}
> > > +EXPORT_SYMBOL(__rcu_read_unlock);
> 
> Thanks,
> 
> Mathieu
> 
> -- 
> Mathieu Desnoyers
> Operating System Efficiency R&D Consultant
> EfficiOS Inc.
> http://www.efficios.com

-- 
Mathieu Desnoyers
Operating System Efficiency R&D Consultant
EfficiOS Inc.
http://www.efficios.com