[PATCH 3/3] mm/mmu_gather: send tlb_remove_table_smp_sync IPI only to CPUs in kernel mode

[Peter Zijlstra]

On Wed, Apr 05, 2023 at 01:45:02PM +0100, Valentin Schneider wrote:
> On 05/04/23 14:05, Frederic Weisbecker wrote:
> >  static void smp_call_function_many_cond(const struct cpumask *mask,
> >                                       smp_call_func_t func, void *info,
> > @@ -946,10 +948,13 @@ static void smp_call_function_many_cond(const struct cpumask *mask,
> >  #endif
> >                       cfd_seq_store(pcpu->seq_queue, this_cpu, cpu, CFD_SEQ_QUEUE);
> >                       if (llist_add(&csd->node.llist, &per_cpu(call_single_queue, cpu))) {
> > -				__cpumask_set_cpu(cpu, cfd->cpumask_ipi);
> > -				nr_cpus++;
> > -				last_cpu = cpu;
> > -
> > +				if (!(scf_flags & SCF_NO_USER) ||
> > +				    !IS_ENABLED(CONFIG_GENERIC_ENTRY) ||
> > +				     ct_state_cpu(cpu) != CONTEXT_USER) {
> > +					__cpumask_set_cpu(cpu, cfd->cpumask_ipi);
> > +					nr_cpus++;
> > +					last_cpu = cpu;
> > +				}
> 
> I've been hacking on something like this (CSD deferral for NOHZ-full),
> and unfortunately this uses the CPU-local cfd_data storage thing, which
> means any further smp_call_function() from the same CPU to the same
> destination will spin on csd_lock_wait(), waiting for the target CPU to
> come out of userspace and flush the queue - and we've just spent extra
> effort into *not* disturbing it, so that'll take a while :(

I'm not sure I buy into deferring stuff.. a NOHZ_FULL cpu might 'never'
come back. Queueing data just in case it does seems wasteful.