sched/uclamp: Optimize sched_uclamp_used static key enabling

Repeat calls of static_branch_enable() to an already enabled
static key introduce overhead, because it calls cpus_read_lock().

Users may frequently set the uclamp value of tasks, triggering
the repeat enabling of the sched_uclamp_used static key.

Optimize this and avoid repeat calls to static_branch_enable()
by checking whether it's enabled already.

[ mingo: Rewrote the changelog for legibility ]

Signed-off-by: Xuewen Yan <xuewen.yan@unisoc.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Christian Loehle <christian.loehle@arm.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20250219093747.2612-2-xuewen.yan@unisoc.com

Author: Xuewen Yan

kernel/sched/core.c

@@ -1941,12 +1941,12 @@ static int sysctl_sched_uclamp_handler(const struct ctl_table *table, int write,
 	}
 
 	if (update_root_tg) {
-		static_branch_enable(&sched_uclamp_used);
+		sched_uclamp_enable();
 		uclamp_update_root_tg();
 	}
 
 	if (old_min_rt != sysctl_sched_uclamp_util_min_rt_default) {
-		static_branch_enable(&sched_uclamp_used);
+		sched_uclamp_enable();
 		uclamp_sync_util_min_rt_default();
 	}
 
@@ -9294,7 +9294,7 @@ static ssize_t cpu_uclamp_write(struct kernfs_open_file *of, char *buf,
 	if (req.ret)
 		return req.ret;
 
-	static_branch_enable(&sched_uclamp_used);
+	sched_uclamp_enable();
 
 	guard(mutex)(&uclamp_mutex);
 	guard(rcu)();

kernel/sched/sched.h

@@ -3407,6 +3407,18 @@ static inline bool uclamp_is_used(void)
 	return static_branch_likely(&sched_uclamp_used);
 }
 
+/*
+ * Enabling static branches would get the cpus_read_lock(),
+ * check whether uclamp_is_used before enable it to avoid always
+ * calling cpus_read_lock(). Because we never disable this
+ * static key once enable it.
+ */
+static inline void sched_uclamp_enable(void)
+{
+	if (!uclamp_is_used())
+		static_branch_enable(&sched_uclamp_used);
+}
+
 static inline unsigned long uclamp_rq_get(struct rq *rq,
 					  enum uclamp_id clamp_id)
 {
@@ -3486,6 +3498,8 @@ static inline bool uclamp_is_used(void)
 	return false;
 }
 
+static inline void sched_uclamp_enable(void) {}
+
 static inline unsigned long
 uclamp_rq_get(struct rq *rq, enum uclamp_id clamp_id)
 {

kernel/sched/syscalls.c

@@ -368,7 +368,7 @@ static int uclamp_validate(struct task_struct *p,
 	 * blocking operation which obviously cannot be done while holding
 	 * scheduler locks.
 	 */
-	static_branch_enable(&sched_uclamp_used);
+	sched_uclamp_enable();
 
 	return 0;
 }