On Tue, May 26, 2020 at 10:14:56AM +0200, Ingo Molnar wrote:
>
> * Paul E. McKenney wrote:
>
> > > + if (!tick_nohz_full_cpu(rdp->cpu) ||
> > > + !READ_ONCE(rdp->rcu_urgent_qs) ||
> > > + READ_ONCE(rdp->rcu_forced_tick)) {
> > > + // RCU doesn't need nohz_full help from this
* Paul E. McKenney wrote:
> > + if (!tick_nohz_full_cpu(rdp->cpu) ||
> > + !READ_ONCE(rdp->rcu_urgent_qs) ||
> > + READ_ONCE(rdp->rcu_forced_tick)) {
> > + // RCU doesn't need nohz_full help from this CPU, or it is
> > + // already getting that help.
> > +
"Paul E. McKenney" writes:
> On Thu, May 21, 2020 at 10:05:15PM +0200, Thomas Gleixner wrote:
>> +void __rcu_irq_enter_check_tick(void);
>> +
>> +static __always_inline void rcu_irq_enter_check_tick(void)
>> +{
>> +if (context_tracking_enabled())
>> +__rcu_irq_enter_check_tick();
On Thu, May 21, 2020 at 10:05:15PM +0200, Thomas Gleixner wrote:
> From: Paul E. McKenney
>
> There will likely be exception handlers that can sleep, which rules
> out the usual approach of invoking rcu_nmi_enter() on entry and also
> rcu_nmi_exit() on all exit paths. However, the alternative
From: Paul E. McKenney
There will likely be exception handlers that can sleep, which rules
out the usual approach of invoking rcu_nmi_enter() on entry and also
rcu_nmi_exit() on all exit paths. However, the alternative approach of
just not calling anything can prevent RCU from coaxing quiescent
5 matches
Mail list logo