| Message ID | 20220831181044.2694488-3-paulmck@kernel.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Callback-offload (nocb) updates for v6.1 | expand |
On Wed, Aug 31, 2022 at 11:10:44AM -0700, Paul E. McKenney wrote: > From: Zhen Lei <thunder.leizhen@huawei.com> > > 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@huawei.com> > Signed-off-by: Paul E. McKenney <paulmck@kernel.org> > Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org> > --- > 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). > +#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. 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 >
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@huawei.com> > > > > 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@huawei.com> > > Signed-off-by: Paul E. McKenney <paulmck@kernel.org> > > Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org> > > --- > > 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 > >
On Thu, Sep 01, 2022 at 03:25:20AM -0700, Paul E. McKenney wrote: > On Thu, Sep 01, 2022 at 11:15:57AM +0200, Frederic Weisbecker wrote: > > > +#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=. Not sure I get what you mean. nohz_full= should in any case always force rcu_nocbs at least on the nohz_full CPUs. For example assuming the following combination: rcu_nocbs=6, nohz_full=7 AND CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y, then the result should be: NOCB CPUs = 6,7 NOHZ_FULL CPUs = 7 (CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y is overriden by rcu_nocbs=6). Now if we have nohz_full=7 AND CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y, then the result is expected to be either: NOCB CPUs = 7 (upstream behaviour) NOHZ_FULL CPUs = 7 or NOCB CPUs = all NOHZ_FULL CPUs = 7 The second makes more sense IMHO but that should be in a separate change. Thanks.
On 2022/9/1 19:11, Frederic Weisbecker wrote: > On Thu, Sep 01, 2022 at 03:25:20AM -0700, Paul E. McKenney wrote: >> On Thu, Sep 01, 2022 at 11:15:57AM +0200, Frederic Weisbecker wrote: >>>> +#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=. > > Not sure I get what you mean. nohz_full= should in any case always force > rcu_nocbs at least on the nohz_full CPUs. > > For example assuming the following combination: rcu_nocbs=6, nohz_full=7 AND > CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y, then the result should be: > > NOCB CPUs = 6,7 > NOHZ_FULL CPUs = 7 > > (CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y is overriden by rcu_nocbs=6). > > Now if we have nohz_full=7 AND CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y, then the > result is expected to be either: > > NOCB CPUs = 7 (upstream behaviour) > NOHZ_FULL CPUs = 7 OK, I got it. CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y takes effect only when there are no 'rcu_nocbs=' and 'nohz_full='. > > or > > NOCB CPUs = all > NOHZ_FULL CPUs = 7 > > The second makes more sense IMHO but that should be in a separate change. OK, in this case, CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y takes effect only when there is no 'rcu_nocbs='. > > Thanks. > . >
On Thu, Sep 01, 2022 at 01:11:14PM +0200, Frederic Weisbecker wrote: > On Thu, Sep 01, 2022 at 03:25:20AM -0700, Paul E. McKenney wrote: > > On Thu, Sep 01, 2022 at 11:15:57AM +0200, Frederic Weisbecker wrote: > > > > +#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=. > > Not sure I get what you mean. nohz_full= should in any case always force > rcu_nocbs at least on the nohz_full CPUs. > > For example assuming the following combination: rcu_nocbs=6, nohz_full=7 AND > CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y, then the result should be: > > NOCB CPUs = 6,7 > NOHZ_FULL CPUs = 7 > > (CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y is overriden by rcu_nocbs=6). > > Now if we have nohz_full=7 AND CONFIG_RCU_NOCB_CPU_DEFAULT_ALL=y, then the > result is expected to be either: > > NOCB CPUs = 7 (upstream behaviour) > NOHZ_FULL CPUs = 7 > > or > > NOCB CPUs = all > NOHZ_FULL CPUs = 7 > > The second makes more sense IMHO but that should be in a separate change. Your examples are consistent with what I was trying to say. ;-) Thanx, Paul
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; +#elif defined(CONFIG_NO_HZ_FULL) + if (tick_nohz_full_running && !cpumask_empty(tick_nohz_full_mask)) + cpumask = tick_nohz_full_mask; +#endif - 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,