[PATCH v11 1/2] sched/cache: Reduce the overhead of task_cache_work by only scan the visisted cpus

From: Luo Gengkun

Date: Wed Nov 11 2026 - 05:34:17 EST


The overhead of task_cache_work() is high, especially in multi-NUMA systems.
Currently, task_cache_work() tries to find the pref_llc by scanning all CPUs
in the system. However, most of these scans are meaningless, such as those
for CPUs that have never been visited or were accessed a long time ago.

To address this problem, introduce visited_cpus to track the visited CPUs
and evict them once they have not been accessed for a duration exceeding
llc_epoch_affinity_timeout.

Now that we know exactly which CPUs to scan from visited_cpus, we can remove
get_scan_cpumasks().

Below are the Redis test results conducted on AMD server.

When NUMA balancing is disabled:

valkey-benchmark rps | baseline | schedcache | schedcache_visit
| (no cache aware) | (without this series) | (with this series)
----------------------+---------------------+---------------------+-----------+-------------------+--------
| p99 latency(ms) | p99 latency(ms) | DIFF(%) | p99 latency(ms) | DIFF(%)
400000 | 0.436 | 0.554 | -25.68% | 0.441 | -1.14%

Additionally, the output of perf top -e cycles:k highlights the overhead
incurred by task_cache_work():

valkey-benchmark rps | schedcache | schedcache_visit
| (without this series) | (with this series)
----------------------+-------------------------------------------+---------------------------------------
400000 | 0.81% [kernel] [k] task_cache_work | 0.02% [kernel] [k] task_cache_work

The trace data is shown below:

echo NO_SC_VISIT > /sys/kernel/debug/sched/features
redis-server-24660 [195] ..... 2916.005989: sched_cache_scan: comm=redis-server pid=24660 scan=384
valkey-benchmar-48836 [317] ..... 2916.005993: sched_cache_scan: comm=valkey-benchmar pid=48836 scan=384
valkey-benchmar-48487 [310] ..... 2916.006002: sched_cache_scan: comm=valkey-benchmar pid=48487 scan=384

echo SC_VISIT > /sys/kernel/debug/sched/features
redis-server-29080 [225] ..... 2876.257909: sched_cache_scan: comm=redis-server pid=29080 scan=14
redis-server-26694 [018] ..... 2876.257909: sched_cache_scan: comm=redis-server pid=26694 scan=13
valkey-benchmar-48198 [103] ..... 2876.257909: sched_cache_scan: comm=valkey-benchmar pid=48198 scan=16

The above results show that this series significantly reduces the overhead
of task_cache_work() by decreasing the number of CPUs that need to be
scanned, from 384 to 16, when NUMA balancing is disabled.

Even with NUMA balancing enabled, this patch provides better performance
benefits. The following tests were conducted with NUMA balancing enabled:

valkey-benchmark rps | baseline | schedcache | schedcache_visit
| (no cache aware) | (without this series) | (with this series)
----------------------+---------------------+---------------------+-----------+-------------------+--------
| p99 latency(ms) | p99 latency(ms) | DIFF(%) | p99 latency(ms) | DIFF(%)
400000 | 0.437 | 0.454 | -3.89% | 0.442 | -1.14%

The overhead of task_cache_work() is as follows:

valkey-benchmark rps | schedcache | schedcache_visit
| (without this series) | (with this series)
----------------------+-------------------------------------------+---------------------------------------
400000 | 0.13% [kernel] [k] task_cache_work | 0.03% [kernel] [k] task_cache_work

Hackbench testing was also conducted. The results below show that this
series does not affect the accuracy of cache-aware scheduling.

=========================================
Hackbench Comparison: baseline vs schedcache
=========================================
MODE GROUPS FDS | baseline(std) | schedcache(std) | DIFF(%) | VERDICT
---------- ------ -----+--------------------+--------------------+------------+-----------
threads 1 10 | 113.200 (4.22%) | 67.300 (1.32%) | 40.55% | IMPROVED
threads 1 2 | 16.555 (4.11%) | 11.020 (1.66%) | 33.43% | IMPROVED
threads 1 20 | 250.774 (1.26%) | 265.026 (5.44%) | -5.68% | REGRESSED
threads 1 4 | 42.117 (1.44%) | 27.758 (1.64%) | 34.09% | IMPROVED
threads 1 6 | 65.140 (4.31%) | 39.182 (1.38%) | 39.85% | IMPROVED
threads 1 8 | 84.286 (1.29%) | 53.721 (1.58%) | 36.26% | IMPROVED
threads 2 10 | 122.592 (0.44%) | 113.365 (4.93%) | 7.53% | IMPROVED
threads 2 2 | 17.702 (4.09%) | 10.473 (0.42%) | 40.84% | IMPROVED
threads 2 20 | 336.457 (1.77%) | 314.108 (1.51%) | 6.64% | IMPROVED
threads 2 4 | 43.989 (0.88%) | 27.067 (3.38%) | 38.47% | IMPROVED
threads 2 6 | 69.322 (0.85%) | 41.707 (4.19%) | 39.84% | IMPROVED
threads 2 8 | 103.767 (1.81%) | 58.518 (3.00%) | 43.61% | IMPROVED
threads 4 10 | 148.882 (3.56%) | 149.449 (1.06%) | -0.38% | REGRESSED
threads 4 2 | 18.909 (2.96%) | 11.063 (2.08%) | 41.49% | IMPROVED
threads 4 20 | 724.943 (2.14%) | 631.222 (3.92%) | 12.93% | IMPROVED
threads 4 4 | 48.191 (1.91%) | 27.352 (5.35%) | 43.24% | IMPROVED
threads 4 6 | 79.725 (3.84%) | 78.732 (4.10%) | 1.25% | IMPROVED
threads 4 8 | 108.768 (1.36%) | 105.928 (1.65%) | 2.61% | IMPROVED

Hackbench Comparison: schedcache vs schedcache_visit
=========================================
MODE GROUPS FDS | schedcache(std) |schedcache_visit(std) | DIFF(%) | VERDICT
---------- ------ -----+--------------------+----------------------+---------+-----------
threads 1 10 | 67.300 (1.32%) | 67.014 (0.96%) | 0.42% | IMPROVED
threads 1 2 | 11.020 (1.66%) | 10.557 (1.46%) | 4.20% | IMPROVED
threads 1 20 | 265.026 (5.44%) | 212.366 (16.32%) | 19.87% | IMPROVED
threads 1 4 | 27.758 (1.64%) | 25.711 (1.32%) | 7.37% | IMPROVED
threads 1 6 | 39.182 (1.38%) | 38.914 (0.34%) | 0.68% | IMPROVED
threads 1 8 | 53.721 (1.58%) | 52.889 (0.27%) | 1.55% | IMPROVED
threads 2 10 | 121.203 (6.99%) | 124.254 (1.38%) | -2.52% | REGRESSED
threads 2 2 | 10.473 (0.42%) | 11.206 (5.91%) | -7.00% | REGRESSED
threads 2 20 | 314.108 (1.51%) | 301.754 (1.95%) | 3.93% | IMPROVED
threads 2 4 | 27.067 (3.38%) | 28.028 (2.01%) | -3.55% | REGRESSED
threads 2 6 | 41.707 (4.19%) | 42.149 (3.35%) | -1.06% | REGRESSED
threads 2 8 | 58.518 (3.00%) | 57.133 (4.39%) | 2.37% | IMPROVED
threads 4 10 | 149.449 (1.06%) | 141.407 (0.08%) | 5.38% | IMPROVED
threads 4 2 | 11.063 (2.08%) | 11.360 (5.85%) | -2.68% | REGRESSED
threads 4 20 | 631.222 (3.92%) | 622.780 (2.49%) | 1.34% | IMPROVED
threads 4 4 | 27.352 (5.35%) | 27.947 (5.37%) | -2.18% | REGRESSED
threads 4 6 | 78.732 (4.10%) | 73.911 (0.70%) | 6.12% | IMPROVED
threads 4 8 | 105.928 (1.65%) | 107.535 (3.29%) | -1.52% | REGRESSED

Tested-by: K Prateek Nayak <kprateek.nayak@xxxxxxx>
Tested-by: Chen Yu <yu.c.chen@xxxxxxxxx>
Reviewed-by: Chen Yu <yu.c.chen@xxxxxxxxx>
Signed-off-by: Luo Gengkun <luogengkun@xxxxxxxxxxxxxxx>
---
include/linux/sched.h | 2 +
kernel/sched/fair.c | 99 ++++++++++++++++++-------------------------
2 files changed, 43 insertions(+), 58 deletions(-)

diff --git a/include/linux/sched.h b/include/linux/sched.h
index e4f6323d7855..1b08925c6358 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -2385,6 +2385,7 @@ static __always_inline int task_mm_cid(struct task_struct *t)
struct sched_cache_time {
u64 runtime;
unsigned long epoch;
+ unsigned long epoch_last_visit;
};

struct sched_cache_group {
@@ -2397,6 +2398,7 @@ struct sched_cache_group {
int cpu;
refcount_t refcnt;
struct rcu_head rcu;
+ cpumask_var_t visited_cpus;
} ____cacheline_aligned_in_smp;

struct sched_cache_group *sched_cache_group_get(struct sched_cache_group *grp);
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 79f6b3a4fd7f..c97f08eaa9e5 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -1599,11 +1599,11 @@ int mm_init_sched(struct mm_struct *mm,
int i;

grp = kzalloc_obj(*grp);
- if (!grp) {
- free_percpu(_pcpu_sched);
- mm->sched_cache_grp = NULL;
- return -ENOMEM;
- }
+ if (!grp)
+ goto err_free_pcpu;
+
+ if (!zalloc_cpumask_var(&grp->visited_cpus, GFP_KERNEL))
+ goto err_free_grp;

for_each_possible_cpu(i) {
struct sched_cache_time *pcpu_sched = per_cpu_ptr(_pcpu_sched, i);
@@ -1612,6 +1612,7 @@ int mm_init_sched(struct mm_struct *mm,
pcpu_sched->runtime = 0;
/* a slightly stale cpu epoch is acceptible */
pcpu_sched->epoch = rq->cpu_epoch;
+ pcpu_sched->epoch_last_visit = rq->cpu_epoch;
epoch = rq->cpu_epoch;
}

@@ -1637,6 +1638,14 @@ int mm_init_sched(struct mm_struct *mm,
*/
smp_store_release(&mm->sched_cache_grp, grp);
return 0;
+
+err_free_grp:
+ kfree(grp);
+
+err_free_pcpu:
+ free_percpu(_pcpu_sched);
+ mm->sched_cache_grp = NULL;
+ return -ENOMEM;
}

static void sched_cache_group_free_rcu(struct rcu_head *rcu)
@@ -1645,6 +1654,7 @@ static void sched_cache_group_free_rcu(struct rcu_head *rcu)
container_of(rcu, struct sched_cache_group, rcu);

free_percpu(grp->pcpu_sched);
+ free_cpumask_var(grp->visited_cpus);
kfree(grp);
}

@@ -1785,13 +1795,24 @@ static inline void __update_mm_sched(struct rq *rq,
}
}

-static unsigned long fraction_mm_sched(struct rq *rq,
- struct sched_cache_time *pcpu_sched)
+static unsigned long fraction_mm_sched(int cpu,
+ struct sched_cache_group *grp)
{
+ struct sched_cache_time *pcpu_sched =
+ per_cpu_ptr(grp->pcpu_sched, cpu);
+ struct rq *rq = cpu_rq(cpu);
+
guard(raw_spinlock_irqsave)(&rq->cpu_epoch_lock);

__update_mm_sched(rq, pcpu_sched);

+ /* Evict the CPU if this sched group has not run on it for a while */
+ if ((rq->cpu_epoch - pcpu_sched->epoch_last_visit) >
+ llc_epoch_affinity_timeout) {
+ cpumask_clear_cpu(cpu, grp->visited_cpus);
+ return 0;
+ }
+
/*
* Runtime is a geometric series (r=0.5) and as such will sum to twice
* the accumulation period, this means the multiplcation here should
@@ -1867,6 +1888,9 @@ void account_mm_sched(struct rq *rq, struct task_struct *p, s64 delta_exec)
pcpu_sched->runtime += delta_exec;
rq->cpu_runtime += delta_exec;
epoch = rq->cpu_epoch;
+ pcpu_sched->epoch_last_visit = epoch;
+ if (!cpumask_test_cpu(cpu_of(rq), grp->visited_cpus))
+ cpumask_set_cpu(cpu_of(rq), grp->visited_cpus);
}

/*
@@ -1917,52 +1941,6 @@ static void task_tick_cache(struct rq *rq, struct task_struct *p)
}
}

-static void get_scan_cpumasks(cpumask_var_t cpus, struct task_struct *p,
- struct sched_cache_group *grp)
-{
-#ifdef CONFIG_NUMA_BALANCING
- int cpu, curr_cpu, nid, pref_nid;
-
- if (!static_branch_likely(&sched_numa_balancing))
- goto out;
-
- cpu = READ_ONCE(grp->cpu);
- if (cpu != -1)
- nid = cpu_to_node(cpu);
- curr_cpu = task_cpu(p);
-
- /*
- * Scanning in the preferred NUMA node is ideal. However, the NUMA
- * preferred node is per-task rather than per-process. It is possible
- * for different threads of the process to have distinct preferred
- * nodes; consequently, the process-wide preferred LLC may bounce
- * between different nodes. As a workaround, maintain the scan
- * CPU mask to also cover the process's current preferred LLC and the
- * current running node to mitigate the bouncing risk.
- * TBD: numa_group should be considered during task aggregation.
- */
- pref_nid = p->numa_preferred_nid;
- /* honor the task's preferred node */
- if (pref_nid == NUMA_NO_NODE)
- goto out;
-
- cpumask_or(cpus, cpus, cpumask_of_node(pref_nid));
-
- /* honor the task's preferred LLC CPU */
- if (cpu != -1 && !cpumask_test_cpu(cpu, cpus) && nid != NUMA_NO_NODE)
- cpumask_or(cpus, cpus, cpumask_of_node(nid));
-
- /* make sure the task's current running node is included */
- if (!cpumask_test_cpu(curr_cpu, cpus))
- cpumask_or(cpus, cpus, cpumask_of_node(cpu_to_node(curr_cpu)));
-
- return;
-
-out:
-#endif
- cpumask_copy(cpus, cpu_online_mask);
-}
-
static inline void update_avg_scale(u64 *avg, u64 sample)
{
int factor = per_cpu(sd_llc_size, raw_smp_processor_id());
@@ -2011,7 +1989,7 @@ static void task_cache_work(struct callback_head *work)
if (time_before(now, next_scan))
return;

- /* only 1 thread is allowed to scan */
+ /* elect a single scanner per epoch */
if (!try_cmpxchg(&grp->next_scan, &next_scan,
now + max_t(unsigned long,
READ_ONCE(llc_epoch_period), 1)))
@@ -2032,7 +2010,13 @@ static void task_cache_work(struct callback_head *work)
scoped_guard (cpus_read_lock) {
guard(rcu)();

- get_scan_cpumasks(cpus, p, grp);
+ /*
+ * account_mm_sched() can set a bit in visited_cpus while
+ * it is read here locklessly for efficiency. A CPU missed
+ * this way has only just been used by this sched group,
+ * so its share small, and the next scan will include it.
+ */
+ cpumask_and(cpus, cpu_online_mask, grp->visited_cpus);

for_each_cpu(cpu, cpus) {
/* XXX sched_cluster_active */
@@ -2043,9 +2027,8 @@ static void task_cache_work(struct callback_head *work)
if (!sd)
continue;

- for_each_cpu(i, sched_domain_span(sd)) {
- occ = fraction_mm_sched(cpu_rq(i),
- per_cpu_ptr(grp->pcpu_sched, i));
+ for_each_cpu_and(i, sched_domain_span(sd), cpus) {
+ occ = fraction_mm_sched(i, grp);
a_occ += occ;
if (occ > m_occ) {
m_occ = occ;
--
2.34.1