Scheduler maintains a per LLC info which tells if there is any idle core
in that LLC. However this information doesn't provide which core is idle.
So when iterating for idle-cores, if select_idle_core() finds an
idle-core, then it doesn't try to reset this information.
So if there was only one idle core in the LLC and select_idle_core()
selected the idle-core, the LLC will maintain that it still has a
idle-core.
On the converse, if a task is pinned, and has a restricted
cpus_allowed_list and LLC has multiple idle-cores, but select_idle_core
cannot find a idle-core, LLC will no more maintain that it has an
idle-core.
As a first step to solve this problem, LLC will maintain the identity of
the idle core instead of just the information that LLC has an idle core
Along with maintaining, this change will solve both the problems listed
above. However there are other problems that exist with the current
infrastructure and those will continue to exist with this change and
would be handled in subsequent patches.
Cc: LKML <[email protected]>
Cc: Gautham R Shenoy <[email protected]>
Cc: Parth Shah <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Valentin Schneider <[email protected]>
Cc: Dietmar Eggemann <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Vincent Guittot <[email protected]>
Cc: Rik van Riel <[email protected]>
Signed-off-by: Srikar Dronamraju <[email protected]>
---
include/linux/sched/topology.h | 2 +-
kernel/sched/fair.c | 43 +++++++++++++++++++---------------
kernel/sched/sched.h | 3 +++
kernel/sched/topology.c | 7 ++++++
4 files changed, 35 insertions(+), 20 deletions(-)
diff --git a/include/linux/sched/topology.h b/include/linux/sched/topology.h
index 8f0f778b7c91..285165a35f21 100644
--- a/include/linux/sched/topology.h
+++ b/include/linux/sched/topology.h
@@ -73,7 +73,7 @@ struct sched_group;
struct sched_domain_shared {
atomic_t ref;
atomic_t nr_busy_cpus;
- int has_idle_cores;
+ int idle_core;
};
struct sched_domain {
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index a258a84cfdfd..03083eacdaf0 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -1563,11 +1563,11 @@ numa_type numa_classify(unsigned int imbalance_pct,
#ifdef CONFIG_SCHED_SMT
/* Forward declarations of select_idle_sibling helpers */
-static inline bool test_idle_cores(int cpu, bool def);
+static inline int get_idle_core(int cpu, int def);
static inline int numa_idle_core(int idle_core, int cpu)
{
if (!static_branch_likely(&sched_smt_present) ||
- idle_core >= 0 || !test_idle_cores(cpu, false))
+ idle_core >= 0 || get_idle_core(cpu, -1) == -1)
return idle_core;
/*
@@ -6015,29 +6015,31 @@ static inline int __select_idle_cpu(int cpu)
DEFINE_STATIC_KEY_FALSE(sched_smt_present);
EXPORT_SYMBOL_GPL(sched_smt_present);
-static inline void set_idle_cores(int cpu, int val)
+static inline void set_idle_core(int cpu, int val)
{
struct sched_domain_shared *sds;
sds = rcu_dereference(per_cpu(sd_llc_shared, cpu));
if (sds)
- WRITE_ONCE(sds->has_idle_cores, val);
+ WRITE_ONCE(sds->idle_core, val);
}
-static inline bool test_idle_cores(int cpu, bool def)
+static inline int get_idle_core(int cpu, int def)
{
struct sched_domain_shared *sds;
- sds = rcu_dereference(per_cpu(sd_llc_shared, cpu));
- if (sds)
- return READ_ONCE(sds->has_idle_cores);
+ if (static_branch_likely(&sched_smt_present)) {
+ sds = rcu_dereference(per_cpu(sd_llc_shared, cpu));
+ if (sds)
+ return READ_ONCE(sds->idle_core);
+ }
return def;
}
/*
* Scans the local SMT mask to see if the entire core is idle, and records this
- * information in sd_llc_shared->has_idle_cores.
+ * information in sd_llc_shared->idle_core.
*
* Since SMT siblings share all cache levels, inspecting this limited remote
* state should be fairly cheap.
@@ -6048,7 +6050,7 @@ void __update_idle_core(struct rq *rq)
int cpu;
rcu_read_lock();
- if (test_idle_cores(core, true))
+ if (get_idle_core(core, 0) != -1)
goto unlock;
for_each_cpu(cpu, cpu_smt_mask(core)) {
@@ -6059,7 +6061,7 @@ void __update_idle_core(struct rq *rq)
goto unlock;
}
- set_idle_cores(core, 1);
+ set_idle_core(core, per_cpu(smt_id, core));
unlock:
rcu_read_unlock();
}
@@ -6067,7 +6069,7 @@ void __update_idle_core(struct rq *rq)
/*
* Scan the entire LLC domain for idle cores; this dynamically switches off if
* there are no idle cores left in the system; tracked through
- * sd_llc->shared->has_idle_cores and enabled through update_idle_core() above.
+ * sd_llc->shared->idle_core and enabled through update_idle_core() above.
*/
static int select_idle_core(struct task_struct *p, int core, struct cpumask *cpus, int *idle_cpu)
{
@@ -6102,11 +6104,11 @@ static int select_idle_core(struct task_struct *p, int core, struct cpumask *cpu
#else /* CONFIG_SCHED_SMT */
-static inline void set_idle_cores(int cpu, int val)
+static inline void set_idle_core(int cpu, int val)
{
}
-static inline bool test_idle_cores(int cpu, bool def)
+static inline bool get_idle_core(int cpu, int def)
{
return def;
}
@@ -6127,7 +6129,8 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t
{
struct cpumask *cpus = this_cpu_cpumask_var_ptr(select_idle_mask);
int i, cpu, idle_cpu = -1, nr = INT_MAX;
- bool smt = test_idle_cores(target, false);
+ int idle_core = get_idle_core(target, -1);
+ bool smt = (idle_core != -1);
int this = smp_processor_id();
struct sched_domain *this_sd;
u64 time;
@@ -6160,8 +6163,13 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t
for_each_cpu_wrap(cpu, cpus, target) {
if (smt) {
i = select_idle_core(p, cpu, cpus, &idle_cpu);
- if ((unsigned int)i < nr_cpumask_bits)
+ if ((unsigned int)i < nr_cpumask_bits) {
+#ifdef CONFIG_SCHED_SMT
+ if ((per_cpu(smt_id, i)) == idle_core)
+ set_idle_core(i, -1);
+#endif
return i;
+ }
} else {
if (!--nr)
@@ -6172,9 +6180,6 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t
}
}
- if (smt)
- set_idle_cores(this, false);
-
if (sched_feat(SIS_PROP) && !smt) {
time = cpu_clock(this) - time;
update_avg(&this_sd->avg_scan_cost, time);
diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index 10a1522b1e30..46d40a281724 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -1478,6 +1478,9 @@ static inline struct sched_domain *lowest_flag_domain(int cpu, int flag)
DECLARE_PER_CPU(struct sched_domain __rcu *, sd_llc);
DECLARE_PER_CPU(int, sd_llc_size);
DECLARE_PER_CPU(int, sd_llc_id);
+#ifdef CONFIG_SCHED_SMT
+DECLARE_PER_CPU(int, smt_id);
+#endif
DECLARE_PER_CPU(struct sched_domain_shared __rcu *, sd_llc_shared);
DECLARE_PER_CPU(struct sched_domain __rcu *, sd_numa);
DECLARE_PER_CPU(struct sched_domain __rcu *, sd_asym_packing);
diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c
index 09d35044bd88..8db40c8a6ad0 100644
--- a/kernel/sched/topology.c
+++ b/kernel/sched/topology.c
@@ -644,6 +644,9 @@ static void destroy_sched_domains(struct sched_domain *sd)
DEFINE_PER_CPU(struct sched_domain __rcu *, sd_llc);
DEFINE_PER_CPU(int, sd_llc_size);
DEFINE_PER_CPU(int, sd_llc_id);
+#ifdef CONFIG_SCHED_SMT
+DEFINE_PER_CPU(int, smt_id);
+#endif
DEFINE_PER_CPU(struct sched_domain_shared __rcu *, sd_llc_shared);
DEFINE_PER_CPU(struct sched_domain __rcu *, sd_numa);
DEFINE_PER_CPU(struct sched_domain __rcu *, sd_asym_packing);
@@ -667,6 +670,9 @@ static void update_top_cache_domain(int cpu)
rcu_assign_pointer(per_cpu(sd_llc, cpu), sd);
per_cpu(sd_llc_size, cpu) = size;
per_cpu(sd_llc_id, cpu) = id;
+#ifdef CONFIG_SCHED_SMT
+ per_cpu(smt_id, cpu) = cpumask_first(cpu_smt_mask(cpu));
+#endif
rcu_assign_pointer(per_cpu(sd_llc_shared, cpu), sds);
sd = lowest_flag_domain(cpu, SD_NUMA);
@@ -1466,6 +1472,7 @@ sd_init(struct sched_domain_topology_level *tl,
sd->shared = *per_cpu_ptr(sdd->sds, sd_id);
atomic_inc(&sd->shared->ref);
atomic_set(&sd->shared->nr_busy_cpus, sd_weight);
+ sd->shared->idle_core = -1;
}
sd->private = sdd;
--
2.18.2