On Thu, Sep 01, 2022 at 11:15:57AM +0200, Frederic Weisbecker wrote:
> On Wed, Aug 31, 2022 at 11:10:44AM -0700, Paul E. McKenney wrote:
> > From: Zhen Lei <thunder.leizhen@xxxxxxxxxx>
> >
> > In kernels built with either CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y or
> > CONFIG_NO_HZ_FULL=y, additional CPUs must be added to rcu_nocb_mask.
> > Except that kernels booted without the rcu_nocbs= will not have
> > allocated rcu_nocb_mask. And the current rcu_init_nohz() function uses
> > its need_rcu_nocb_mask and offload_all local variables to track the
> > rcu_nocb and nohz_full state.
> >
> > But there is a much simpler approach, namely creating a cpumask pointer
> > to track the default and then using cpumask_available() to check the
> > rcu_nocb_mask state. This commit takes this approach, thereby simplifying
> > and shortening the rcu_init_nohz() function.
> >
> > Signed-off-by: Zhen Lei <thunder.leizhen@xxxxxxxxxx>
> > Signed-off-by: Paul E. McKenney <paulmck@xxxxxxxxxx>
> > Reviewed-by: Joel Fernandes (Google) <joel@xxxxxxxxxxxxxxxxx>
> > ---
> > kernel/rcu/tree_nocb.h | 32 +++++++++-----------------------
> > 1 file changed, 9 insertions(+), 23 deletions(-)
> >
> > diff --git a/kernel/rcu/tree_nocb.h b/kernel/rcu/tree_nocb.h
> > index 0a5f0ef414845..c8167be2288fa 100644
> > --- a/kernel/rcu/tree_nocb.h
> > +++ b/kernel/rcu/tree_nocb.h
> > @@ -1210,45 +1210,31 @@ EXPORT_SYMBOL_GPL(rcu_nocb_cpu_offload);
> > void __init rcu_init_nohz(void)
> > {
> > int cpu;
> > - bool need_rcu_nocb_mask = false;
> > - bool offload_all = false;
> > struct rcu_data *rdp;
> > + const struct cpumask *cpumask = NULL;
> >
> > #if defined(CONFIG_RCU_NOCB_CPU_DEFAULT_ALL)
> > - if (!rcu_state.nocb_is_setup) {
> > - need_rcu_nocb_mask = true;
> > - offload_all = true;
> > - }
> > -#endif /* #if defined(CONFIG_RCU_NOCB_CPU_DEFAULT_ALL) */
> > -
> > -#if defined(CONFIG_NO_HZ_FULL)
> > - if (tick_nohz_full_running && !cpumask_empty(tick_nohz_full_mask)) {
> > - need_rcu_nocb_mask = true;
> > - offload_all = false; /* NO_HZ_FULL has its own mask. */
> > - }
> > -#endif /* #if defined(CONFIG_NO_HZ_FULL) */
> > + cpumask = cpu_possible_mask;
>
> You're missing the rcu_state.nocb_is_setup check, so
> CONFIG_RCU_NOCB_CPU_DEFAULT_ALL will now always override the rcu_nocbs=
> parameter (should be the other way around).
This is indeed a problem, so especially thank you for looking this over!
I will revert this patch.
> > +#elif defined(CONFIG_NO_HZ_FULL)
> > + if (tick_nohz_full_running && !cpumask_empty(tick_nohz_full_mask))
> > + cpumask = tick_nohz_full_mask;
> > +#endif
>
> A subtle behaviour difference here too: CONFIG_RCU_NOCB_CPU_DEFAULT_ALL will
> now override nohz_full=
>
> I don't mind, it's probably what we want in the end, but the changelog should
> tell about it, or even better, this should be a separate change.
Good point. Perhaps the key point is that if there is nohz_full=,
rcu_nocbs=, and CONFIG_RCU_NOCB_CPU_DEFAULT_ALL, we still need rcu_nocbs=
to include at least those bits set by nohz_full=.
Are there any other constraints?
Zhen Lei, please feel free to submit an updated patch.
Thanx, Paul
> Thanks.
>
> >
> > - if (need_rcu_nocb_mask) {
> > + if (cpumask) {
> > if (!cpumask_available(rcu_nocb_mask)) {
> > if (!zalloc_cpumask_var(&rcu_nocb_mask, GFP_KERNEL)) {
> > pr_info("rcu_nocb_mask allocation failed, callback offloading disabled.\n");
> > return;
> > }
> > }
> > +
> > + cpumask_or(rcu_nocb_mask, rcu_nocb_mask, cpumask);
> > rcu_state.nocb_is_setup = true;
> > }
> >
> > if (!rcu_state.nocb_is_setup)
> > return;
> >
> > -#if defined(CONFIG_NO_HZ_FULL)
> > - if (tick_nohz_full_running)
> > - cpumask_or(rcu_nocb_mask, rcu_nocb_mask, tick_nohz_full_mask);
> > -#endif /* #if defined(CONFIG_NO_HZ_FULL) */
> > -
> > - if (offload_all)
> > - cpumask_setall(rcu_nocb_mask);
> > -
> > if (!cpumask_subset(rcu_nocb_mask, cpu_possible_mask)) {
> > pr_info("\tNote: kernel parameter 'rcu_nocbs=', 'nohz_full', or 'isolcpus=' contains nonexistent CPUs.\n");
> > cpumask_and(rcu_nocb_mask, cpu_possible_mask,
> > --
> > 2.31.1.189.g2e36527f23
> >
