| Message ID | 20220825181210.284283-8-vschneid@redhat.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | sched, net: NUMA-aware CPU spreading interface | expand |
On 2022/8/26 2:12, Valentin Schneider wrote: > Tariq has pointed out that drivers allocating IRQ vectors would benefit > from having smarter NUMA-awareness - cpumask_local_spread() only knows > about the local node and everything outside is in the same bucket. > > sched_domains_numa_masks is pretty much what we want to hand out (a cpumask > of CPUs reachable within a given distance budget), introduce > sched_numa_hop_mask() to export those cpumasks. > > Link: http://lore.kernel.org/r/20220728191203.4055-1-tariqt@nvidia.com > Signed-off-by: Valentin Schneider <vschneid@redhat.com> > --- > include/linux/topology.h | 9 +++++++++ > kernel/sched/topology.c | 28 ++++++++++++++++++++++++++++ > 2 files changed, 37 insertions(+) > > diff --git a/include/linux/topology.h b/include/linux/topology.h > index 4564faafd0e1..13b82b83e547 100644 > --- a/include/linux/topology.h > +++ b/include/linux/topology.h > @@ -245,5 +245,14 @@ static inline const struct cpumask *cpu_cpu_mask(int cpu) > return cpumask_of_node(cpu_to_node(cpu)); > } > > +#ifdef CONFIG_NUMA > +extern const struct cpumask *sched_numa_hop_mask(int node, int hops); > +#else > +static inline const struct cpumask *sched_numa_hop_mask(int node, int hops) > +{ > + return ERR_PTR(-EOPNOTSUPP); > +} > +#endif /* CONFIG_NUMA */ > + > I think it should be better to return cpu_online_mask() if CONFIG_NUMA=n and hop is 0. Then we can keep the behaviour consistent with cpumask_local_spread() which for_each_numa_hop_cpu is going to replace. The macro checking maybe unnecessary, check whether node is NUMA_NO_NODE will handle the case where NUMA is not configured. Thanks. > #endif /* _LINUX_TOPOLOGY_H */ > diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c > index 8739c2a5a54e..f0236a0ae65c 100644 > --- a/kernel/sched/topology.c > +++ b/kernel/sched/topology.c > @@ -2067,6 +2067,34 @@ int sched_numa_find_closest(const struct cpumask *cpus, int cpu) > return found; > } > > +/** > + * sched_numa_hop_mask() - Get the cpumask of CPUs at most @hops hops away. > + * @node: The node to count hops from. > + * @hops: Include CPUs up to that many hops away. 0 means local node. > + * > + * Requires rcu_lock to be held. Returned cpumask is only valid within that > + * read-side section, copy it if required beyond that. > + * > + * Note that not all hops are equal in size; see sched_init_numa() for how > + * distances and masks are handled. > + * > + * Also note that this is a reflection of sched_domains_numa_masks, which may change > + * during the lifetime of the system (offline nodes are taken out of the masks). > + */ > +const struct cpumask *sched_numa_hop_mask(int node, int hops) > +{ > + struct cpumask ***masks = rcu_dereference(sched_domains_numa_masks); > + > + if (node >= nr_node_ids || hops >= sched_domains_numa_levels) > + return ERR_PTR(-EINVAL); > + > + if (!masks) > + return NULL; > + > + return masks[hops][node]; > +} > +EXPORT_SYMBOL_GPL(sched_numa_hop_mask); > + > #endif /* CONFIG_NUMA */ > > static int __sdt_alloc(const struct cpumask *cpu_map) >
On 26/08/22 16:14, Yicong Yang wrote: > On 2022/8/26 2:12, Valentin Schneider wrote: >> Tariq has pointed out that drivers allocating IRQ vectors would benefit >> from having smarter NUMA-awareness - cpumask_local_spread() only knows >> about the local node and everything outside is in the same bucket. >> >> sched_domains_numa_masks is pretty much what we want to hand out (a cpumask >> of CPUs reachable within a given distance budget), introduce >> sched_numa_hop_mask() to export those cpumasks. >> >> Link: http://lore.kernel.org/r/20220728191203.4055-1-tariqt@nvidia.com >> Signed-off-by: Valentin Schneider <vschneid@redhat.com> >> --- >> include/linux/topology.h | 9 +++++++++ >> kernel/sched/topology.c | 28 ++++++++++++++++++++++++++++ >> 2 files changed, 37 insertions(+) >> >> diff --git a/include/linux/topology.h b/include/linux/topology.h >> index 4564faafd0e1..13b82b83e547 100644 >> --- a/include/linux/topology.h >> +++ b/include/linux/topology.h >> @@ -245,5 +245,14 @@ static inline const struct cpumask *cpu_cpu_mask(int cpu) >> return cpumask_of_node(cpu_to_node(cpu)); >> } >> >> +#ifdef CONFIG_NUMA >> +extern const struct cpumask *sched_numa_hop_mask(int node, int hops); >> +#else >> +static inline const struct cpumask *sched_numa_hop_mask(int node, int hops) >> +{ >> + return ERR_PTR(-EOPNOTSUPP); >> +} >> +#endif /* CONFIG_NUMA */ >> + >> > > I think it should be better to return cpu_online_mask() if CONFIG_NUMA=n and hop is 0. Then we > can keep the behaviour consistent with cpumask_local_spread() which for_each_numa_hop_cpu is > going to replace. > That's a good point, thanks. > The macro checking maybe unnecessary, check whether node is NUMA_NO_NODE will handle the case > where NUMA is not configured. > > Thanks.
diff --git a/include/linux/topology.h b/include/linux/topology.h index 4564faafd0e1..13b82b83e547 100644 --- a/include/linux/topology.h +++ b/include/linux/topology.h @@ -245,5 +245,14 @@ static inline const struct cpumask *cpu_cpu_mask(int cpu) return cpumask_of_node(cpu_to_node(cpu)); } +#ifdef CONFIG_NUMA +extern const struct cpumask *sched_numa_hop_mask(int node, int hops); +#else +static inline const struct cpumask *sched_numa_hop_mask(int node, int hops) +{ + return ERR_PTR(-EOPNOTSUPP); +} +#endif /* CONFIG_NUMA */ + #endif /* _LINUX_TOPOLOGY_H */ diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c index 8739c2a5a54e..f0236a0ae65c 100644 --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -2067,6 +2067,34 @@ int sched_numa_find_closest(const struct cpumask *cpus, int cpu) return found; } +/** + * sched_numa_hop_mask() - Get the cpumask of CPUs at most @hops hops away. + * @node: The node to count hops from. + * @hops: Include CPUs up to that many hops away. 0 means local node. + * + * Requires rcu_lock to be held. Returned cpumask is only valid within that + * read-side section, copy it if required beyond that. + * + * Note that not all hops are equal in size; see sched_init_numa() for how + * distances and masks are handled. + * + * Also note that this is a reflection of sched_domains_numa_masks, which may change + * during the lifetime of the system (offline nodes are taken out of the masks). + */ +const struct cpumask *sched_numa_hop_mask(int node, int hops) +{ + struct cpumask ***masks = rcu_dereference(sched_domains_numa_masks); + + if (node >= nr_node_ids || hops >= sched_domains_numa_levels) + return ERR_PTR(-EINVAL); + + if (!masks) + return NULL; + + return masks[hops][node]; +} +EXPORT_SYMBOL_GPL(sched_numa_hop_mask); + #endif /* CONFIG_NUMA */ static int __sdt_alloc(const struct cpumask *cpu_map)
Tariq has pointed out that drivers allocating IRQ vectors would benefit from having smarter NUMA-awareness - cpumask_local_spread() only knows about the local node and everything outside is in the same bucket. sched_domains_numa_masks is pretty much what we want to hand out (a cpumask of CPUs reachable within a given distance budget), introduce sched_numa_hop_mask() to export those cpumasks. Link: http://lore.kernel.org/r/20220728191203.4055-1-tariqt@nvidia.com Signed-off-by: Valentin Schneider <vschneid@redhat.com> --- include/linux/topology.h | 9 +++++++++ kernel/sched/topology.c | 28 ++++++++++++++++++++++++++++ 2 files changed, 37 insertions(+)