Merge tag 'sched-core-2024-07-16' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull scheduler updates from Ingo Molnar:

 - Update Daniel Bristot de Oliveira's entry in MAINTAINERS,
   and credit him in CREDITS

 - Harmonize the lock-yielding behavior on dynamically selected
   preemption models with static ones

 - Reorganize the code a bit: split out sched/syscalls.c to reduce
   the size of sched/core.c

 - Micro-optimize psi_group_change()

 - Fix set_load_weight() for SCHED_IDLE tasks

 - Misc cleanups & fixes

* tag 'sched-core-2024-07-16' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched: Update MAINTAINERS and CREDITS
  sched/fair: set_load_weight() must also call reweight_task() for SCHED_IDLE tasks
  sched/psi: Optimise psi_group_change a bit
  sched/core: Drop spinlocks on contention iff kernel is preemptible
  sched/core: Move preempt_model_*() helpers from sched.h to preempt.h
  sched/balance: Skip unnecessary updates to idle load balancer's flags
  idle: Remove stale RCU comment
  sched/headers: Move struct pre-declarations to the beginning of the header
  sched/core: Clean up kernel/sched/sched.h a bit
  sched/core: Simplify prefetch_curr_exec_start()
  sched: Fix spelling in comments
  sched/syscalls: Split out kernel/sched/syscalls.c from kernel/sched/core.c

1 files changed, 30 insertions, 30 deletions

diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c
index 507d7b8d79af..020d58967d4e 100644
--- a/kernel/sched/psi.c
+++ b/kernel/sched/psi.c
@@ -41,7 +41,7 @@
  * What it means for a task to be productive is defined differently
  * for each resource. For IO, productive means a running task. For
  * memory, productive means a running task that isn't a reclaimer. For
- * CPU, productive means an oncpu task.
+ * CPU, productive means an on-CPU task.
  *
  * Naturally, the FULL state doesn't exist for the CPU resource at the
  * system level, but exist at the cgroup level. At the cgroup level,
@@ -49,7 +49,7 @@
  * resource which is being used by others outside of the cgroup or
  * throttled by the cgroup cpu.max configuration.
  *
- * The percentage of wallclock time spent in those compound stall
+ * The percentage of wall clock time spent in those compound stall
  * states gives pressure numbers between 0 and 100 for each resource,
  * where the SOME percentage indicates workload slowdowns and the FULL
  * percentage indicates reduced CPU utilization:
@@ -218,28 +218,32 @@ void __init psi_init(void)
 	group_init(&psi_system);
 }
 
-static bool test_state(unsigned int *tasks, enum psi_states state, bool oncpu)
+static u32 test_states(unsigned int *tasks, u32 state_mask)
 {
-	switch (state) {
-	case PSI_IO_SOME:
-		return unlikely(tasks[NR_IOWAIT]);
-	case PSI_IO_FULL:
-		return unlikely(tasks[NR_IOWAIT] && !tasks[NR_RUNNING]);
-	case PSI_MEM_SOME:
-		return unlikely(tasks[NR_MEMSTALL]);
-	case PSI_MEM_FULL:
-		return unlikely(tasks[NR_MEMSTALL] &&
-			tasks[NR_RUNNING] == tasks[NR_MEMSTALL_RUNNING]);
-	case PSI_CPU_SOME:
-		return unlikely(tasks[NR_RUNNING] > oncpu);
-	case PSI_CPU_FULL:
-		return unlikely(tasks[NR_RUNNING] && !oncpu);
-	case PSI_NONIDLE:
-		return tasks[NR_IOWAIT] || tasks[NR_MEMSTALL] ||
-			tasks[NR_RUNNING];
-	default:
-		return false;
+	const bool oncpu = state_mask & PSI_ONCPU;
+
+	if (tasks[NR_IOWAIT]) {
+		state_mask |= BIT(PSI_IO_SOME);
+		if (!tasks[NR_RUNNING])
+			state_mask |= BIT(PSI_IO_FULL);
+	}
+
+	if (tasks[NR_MEMSTALL]) {
+		state_mask |= BIT(PSI_MEM_SOME);
+		if (tasks[NR_RUNNING] == tasks[NR_MEMSTALL_RUNNING])
+			state_mask |= BIT(PSI_MEM_FULL);
 	}
+
+	if (tasks[NR_RUNNING] > oncpu)
+		state_mask |= BIT(PSI_CPU_SOME);
+
+	if (tasks[NR_RUNNING] && !oncpu)
+		state_mask |= BIT(PSI_CPU_FULL);
+
+	if (tasks[NR_IOWAIT] || tasks[NR_MEMSTALL] || tasks[NR_RUNNING])
+		state_mask |= BIT(PSI_NONIDLE);
+
+	return state_mask;
 }
 
 static void get_recent_times(struct psi_group *group, int cpu,
@@ -345,7 +349,7 @@ static void collect_percpu_times(struct psi_group *group,
 
 	/*
 	 * Collect the per-cpu time buckets and average them into a
-	 * single time sample that is normalized to wallclock time.
+	 * single time sample that is normalized to wall clock time.
 	 *
 	 * For averaging, each CPU is weighted by its non-idle time in
 	 * the sampling period. This eliminates artifacts from uneven
@@ -770,7 +774,6 @@ static void psi_group_change(struct psi_group *group, int cpu,
 {
 	struct psi_group_cpu *groupc;
 	unsigned int t, m;
-	enum psi_states s;
 	u32 state_mask;
 
 	lockdep_assert_rq_held(cpu_rq(cpu));
@@ -842,10 +845,7 @@ static void psi_group_change(struct psi_group *group, int cpu,
 		return;
 	}
 
-	for (s = 0; s < NR_PSI_STATES; s++) {
-		if (test_state(groupc->tasks, s, state_mask & PSI_ONCPU))
-			state_mask |= (1 << s);
-	}
+	state_mask = test_states(groupc->tasks, state_mask);
 
 	/*
 	 * Since we care about lost potential, a memstall is FULL
@@ -1205,7 +1205,7 @@ void psi_cgroup_restart(struct psi_group *group)
 	/*
 	 * After we disable psi_group->enabled, we don't actually
 	 * stop percpu tasks accounting in each psi_group_cpu,
-	 * instead only stop test_state() loop, record_times()
+	 * instead only stop test_states() loop, record_times()
 	 * and averaging worker, see psi_group_change() for details.
 	 *
 	 * When disable cgroup PSI, this function has nothing to sync
@@ -1213,7 +1213,7 @@ void psi_cgroup_restart(struct psi_group *group)
 	 * would see !psi_group->enabled and only do task accounting.
 	 *
 	 * When re-enable cgroup PSI, this function use psi_group_change()
-	 * to get correct state mask from test_state() loop on tasks[],
+	 * to get correct state mask from test_states() loop on tasks[],
 	 * and restart groupc->state_start from now, use .clear = .set = 0
 	 * here since no task status really changed.
 	 */