[PATCH v2] sched/core: Adapt WARN_DOUBLE_CLOCK machinery for core-sched

From: Hao Jia
Date: Wed Feb 15 2023 - 02:39:49 EST


When sched_core_enabled(), we sometimes need to call update_rq_clock()
to update the rq clock of sibling CPUs on the same core, before that we
need to clear RQCF_UPDATED of rq->clock_update_flags to avoid the
WARN_DOUBLE_CLOCK warning. Because at this time the rq->clock_update_flags
of sibling CPUs may be RQCF_UPDATED. If sched_core_enabled(), we will get
a core wide rq->lock, so at this point we can safely clear RQCF_UPDATED of
rq->clock_update_flags of all CPUs on this core to avoid the
WARN_DOUBLE_CLOCK warning.

We cannot clear rq->clock_update_flags of other cpus on the same core in
rq_pin_lock(). Because in some functions, we will temporarily give up
core wide rq->lock, and then use raw_spin_rq_lock() to obtain core wide
rq->lock, such as newidle_balance() and _double_lock_balance().

Steps to reproduce:
1. Enable CONFIG_SCHED_DEBUG and CONFIG_SCHED_CORE when compiling
the kernel
2. echo 1 > /sys/kernel/debug/clear_warn_once
echo "WARN_DOUBLE_CLOCK" > /sys/kernel/debug/sched/features
3. Run the linux/tools/testing/selftests/sched/cs_prctl_test test

Signed-off-by: Hao Jia <jiahao.os@xxxxxxxxxxxxx>
---
- Adapt WARN_DOUBLE_CLOCK machinery for core-sched instead of clearing
WARN_DOUBLE_CLOCK warning one by one.
- Modify commit information
[v1] https://lore.kernel.org/all/20221206070550.31763-1-jiahao.os@xxxxxxxxxxxxx

kernel/sched/core.c | 17 +++++++++++++++++
1 file changed, 17 insertions(+)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index e838feb6adc5..16a33e5adb77 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -427,11 +427,27 @@ void sched_core_put(void)
schedule_work(&_work);
}

+static inline void sched_core_rq_clock_clear_update(struct rq *rq)
+{
+#ifdef CONFIG_SCHED_DEBUG
+ const struct cpumask *smt_mask;
+ int i;
+
+ if (rq->core_enabled) {
+ smt_mask = cpu_smt_mask(rq->cpu);
+ for_each_cpu(i, smt_mask) {
+ if (rq->cpu != i)
+ cpu_rq(i)->clock_update_flags &= (RQCF_REQ_SKIP|RQCF_ACT_SKIP);
+ }
+ }
+#endif
+}
#else /* !CONFIG_SCHED_CORE */

static inline void sched_core_enqueue(struct rq *rq, struct task_struct *p) { }
static inline void
sched_core_dequeue(struct rq *rq, struct task_struct *p, int flags) { }
+static inline void sched_core_rq_clock_clear_update(struct rq *rq) { }

#endif /* CONFIG_SCHED_CORE */

@@ -546,6 +562,7 @@ void raw_spin_rq_lock_nested(struct rq *rq, int subclass)
if (likely(lock == __rq_lockp(rq))) {
/* preempt_count *MUST* be > 1 */
preempt_enable_no_resched();
+ sched_core_rq_clock_clear_update(rq);
return;
}
raw_spin_unlock(lock);
--
2.37.0 (Apple Git-136)