Merge tag 'sched-core-2025-03-22' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull scheduler updates from Ingo Molnar:
"Core & fair scheduler changes:

- Cancel the slice protection of the idle entity (Zihan Zhou)
- Reduce the default slice to avoid tasks getting an extra tick
(Zihan Zhou)
- Force propagating min_slice of cfs_rq when {en,de}queue tasks
(Tianchen Ding)
- Refactor can_migrate_task() to elimate looping (I Hsin Cheng)
- Add unlikey branch hints to several system calls (Colin Ian King)
- Optimize current_clr_polling() on certain architectures (Yujun
Dong)

Deadline scheduler: (Juri Lelli)
- Remove redundant dl_clear_root_domain call
- Move dl_rebuild_rd_accounting to cpuset.h

Uclamp:
- Use the uclamp_is_used() helper instead of open-coding it (Xuewen
Yan)
- Optimize sched_uclamp_used static key enabling (Xuewen Yan)

Scheduler topology support: (Juri Lelli)
- Ignore special tasks when rebuilding domains
- Add wrappers for sched_domains_mutex
- Generalize unique visiting of root domains
- Rebuild root domain accounting after every update
- Remove partition_and_rebuild_sched_domains
- Stop exposing partition_sched_domains_locked

RSEQ: (Michael Jeanson)
- Update kernel fields in lockstep with CONFIG_DEBUG_RSEQ=y
- Fix segfault on registration when rseq_cs is non-zero
- selftests: Add rseq syscall errors test
- selftests: Ensure the rseq ABI TLS is actually 1024 bytes

Membarriers:
- Fix redundant load of membarrier_state (Nysal Jan K.A.)

Scheduler debugging:
- Introduce and use preempt_model_str() (Sebastian Andrzej Siewior)
- Make CONFIG_SCHED_DEBUG unconditional (Ingo Molnar)

Fixes and cleanups:
- Always save/restore x86 TSC sched_clock() on suspend/resume
(Guilherme G. Piccoli)
- Misc fixes and cleanups (Thorsten Blum, Juri Lelli, Sebastian
Andrzej Siewior)"

* tag 'sched-core-2025-03-22' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (40 commits)
cpuidle, sched: Use smp_mb__after_atomic() in current_clr_polling()
sched/debug: Remove CONFIG_SCHED_DEBUG
sched/debug: Remove CONFIG_SCHED_DEBUG from self-test config files
sched/debug, Documentation: Remove (most) CONFIG_SCHED_DEBUG references from documentation
sched/debug: Make CONFIG_SCHED_DEBUG functionality unconditional
sched/debug: Make 'const_debug' tunables unconditional __read_mostly
sched/debug: Change SCHED_WARN_ON() to WARN_ON_ONCE()
rseq/selftests: Fix namespace collision with rseq UAPI header
include/{topology,cpuset}: Move dl_rebuild_rd_accounting to cpuset.h
sched/topology: Stop exposing partition_sched_domains_locked
cgroup/cpuset: Remove partition_and_rebuild_sched_domains
sched/topology: Remove redundant dl_clear_root_domain call
sched/deadline: Rebuild root domain accounting after every update
sched/deadline: Generalize unique visiting of root domains
sched/topology: Wrappers for sched_domains_mutex
sched/deadline: Ignore special tasks when rebuilding domains
tracing: Use preempt_model_str()
xtensa: Rely on generic printing of preemption model
x86: Rely on generic printing of preemption model
s390: Rely on generic printing of preemption model
...

Documentation/scheduler/sched-debug.rst

···
 Scheduler debugfs
 =================
 
-Booting a kernel with CONFIG_SCHED_DEBUG=y will give access to
+Booting a kernel with debugfs enabled will give access to
 scheduler specific debug files under /sys/kernel/debug/sched. Some of
 those files are described below.
 

Documentation/scheduler/sched-design-CFS.rst

···
 CFS uses nanosecond granularity accounting and does not rely on any jiffies or
 other HZ detail.  Thus the CFS scheduler has no notion of "timeslices" in the
 way the previous scheduler had, and has no heuristics whatsoever.  There is
-only one central tunable (you have to switch on CONFIG_SCHED_DEBUG):
+only one central tunable:
 
    /sys/kernel/debug/sched/base_slice_ns
 

Documentation/scheduler/sched-domains.rst

···
 for a given topology level by creating a sched_domain_topology_level array and
 calling set_sched_topology() with this array as the parameter.
 
-The sched-domains debugging infrastructure can be enabled by enabling
-CONFIG_SCHED_DEBUG and adding 'sched_verbose' to your cmdline. If you
-forgot to tweak your cmdline, you can also flip the
+The sched-domains debugging infrastructure can be enabled by 'sched_verbose'
+to your cmdline. If you forgot to tweak your cmdline, you can also flip the
 /sys/kernel/debug/sched/verbose knob. This enables an error checking parse of
 the sched domains which should catch most possible errors (described above). It
 also prints out the domain structure in a visual format.

Documentation/scheduler/sched-ext.rst

···
     nr_rejected   : 0
     enable_seq    : 1
 
-If ``CONFIG_SCHED_DEBUG`` is set, whether a given task is on sched_ext can
-be determined as follows:
+Whether a given task is on sched_ext can be determined as follows:
 
 .. code-block:: none
 

Documentation/scheduler/sched-stats.rst

···
 CONFIG_SMP is not defined, *no* domains are utilized and these lines
 will not appear in the output. <name> is an extension to the domain field
 that prints the name of the corresponding sched domain. It can appear in
-schedstat version 17 and above, and requires CONFIG_SCHED_DEBUG.)
+schedstat version 17 and above.
 
 domain<N> <name> <cpumask> 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
 

Documentation/translations/sp_SP/scheduler/sched-design-CFS.rst

···
 jiffy o detalles como HZ. De este modo, el gestor de tareas CFS no tiene
 noción de "ventanas de tiempo" de la forma en que tenía el gestor de
 tareas previo, y tampoco tiene heurísticos. Únicamente hay un parámetro
-central ajustable (se ha de cambiar en CONFIG_SCHED_DEBUG):
+central ajustable:
 
    /sys/kernel/debug/sched/base_slice_ns
 

arch/arm/kernel/traps.c

···
 	barrier();
 }
 
-#ifdef CONFIG_PREEMPT
-#define S_PREEMPT " PREEMPT"
-#elif defined(CONFIG_PREEMPT_RT)
-#define S_PREEMPT " PREEMPT_RT"
-#else
-#define S_PREEMPT ""
-#endif
 #ifdef CONFIG_SMP
 #define S_SMP " SMP"
 #else
···
 	static int die_counter;
 	int ret;
 
-	pr_emerg("Internal error: %s: %x [#%d]" S_PREEMPT S_SMP S_ISA "\n",
-	         str, err, ++die_counter);
+	pr_emerg("Internal error: %s: %x [#%d]" S_SMP S_ISA "\n",
+		 str, err, ++die_counter);
 
 	/* trap and error numbers are mostly meaningless on ARM */
 	ret = notify_die(DIE_OOPS, str, regs, err, tsk->thread.trap_no, SIGSEGV);

arch/arm64/kernel/traps.c

···
 	printk("%sCode: %s\n", lvl, str);
 }
 
-#ifdef CONFIG_PREEMPT
-#define S_PREEMPT " PREEMPT"
-#elif defined(CONFIG_PREEMPT_RT)
-#define S_PREEMPT " PREEMPT_RT"
-#else
-#define S_PREEMPT ""
-#endif
-
 #define S_SMP " SMP"
 
 static int __die(const char *str, long err, struct pt_regs *regs)
···
 	static int die_counter;
 	int ret;
 
-	pr_emerg("Internal error: %s: %016lx [#%d]" S_PREEMPT S_SMP "\n",
+	pr_emerg("Internal error: %s: %016lx [#%d] " S_SMP "\n",
 		 str, err, ++die_counter);
 
 	/* trap and error numbers are mostly meaningless on ARM */

arch/powerpc/kernel/traps.c

···
 {
 	printk("Oops: %s, sig: %ld [#%d]\n", str, err, ++die_counter);
 
-	printk("%s PAGE_SIZE=%luK%s%s%s%s%s%s %s\n",
+	printk("%s PAGE_SIZE=%luK%s %s%s%s%s %s\n",
 	       IS_ENABLED(CONFIG_CPU_LITTLE_ENDIAN) ? "LE" : "BE",
 	       PAGE_SIZE / 1024, get_mmu_str(),
-	       IS_ENABLED(CONFIG_PREEMPT) ? " PREEMPT" : "",
 	       IS_ENABLED(CONFIG_SMP) ? " SMP" : "",
 	       IS_ENABLED(CONFIG_SMP) ? (" NR_CPUS=" __stringify(NR_CPUS)) : "",
 	       debug_pagealloc_enabled() ? " DEBUG_PAGEALLOC" : "",

arch/s390/kernel/dumpstack.c

···
 	console_verbose();
 	spin_lock_irq(&die_lock);
 	bust_spinlocks(1);
-	printk("%s: %04x ilc:%d [#%d] ", str, regs->int_code & 0xffff,
+	printk("%s: %04x ilc:%d [#%d]", str, regs->int_code & 0xffff,
 	       regs->int_code >> 17, ++die_counter);
-#ifdef CONFIG_PREEMPT
-	pr_cont("PREEMPT ");
-#elif defined(CONFIG_PREEMPT_RT)
-	pr_cont("PREEMPT_RT ");
-#endif
 	pr_cont("SMP ");
 	if (debug_pagealloc_enabled())
 		pr_cont("DEBUG_PAGEALLOC");

arch/x86/kernel/dumpstack.c

···
 
 static void __die_header(const char *str, struct pt_regs *regs, long err)
 {
-	const char *pr = "";
-
 	/* Save the regs of the first oops for the executive summary later. */
 	if (!die_counter)
 		exec_summary_regs = *regs;
 
-	if (IS_ENABLED(CONFIG_PREEMPTION))
-		pr = IS_ENABLED(CONFIG_PREEMPT_RT) ? " PREEMPT_RT" : " PREEMPT";
-
 	printk(KERN_DEFAULT
-	       "Oops: %s: %04lx [#%d]%s%s%s%s%s\n", str, err & 0xffff,
-	       ++die_counter, pr,
+	       "Oops: %s: %04lx [#%d]%s%s%s%s\n", str, err & 0xffff,
+	       ++die_counter,
 	       IS_ENABLED(CONFIG_SMP)     ? " SMP"             : "",
 	       debug_pagealloc_enabled()  ? " DEBUG_PAGEALLOC" : "",
 	       IS_ENABLED(CONFIG_KASAN)   ? " KASAN"           : "",

arch/x86/kernel/tsc.c

···
 
 void tsc_save_sched_clock_state(void)
 {
-	if (!sched_clock_stable())
+	if (!static_branch_likely(&__use_tsc) && !sched_clock_stable())
 		return;
 
 	cyc2ns_suspend = sched_clock();
···
 	unsigned long flags;
 	int cpu;
 
-	if (!sched_clock_stable())
+	if (!static_branch_likely(&__use_tsc) && !sched_clock_stable())
 		return;
 
 	local_irq_save(flags);

arch/xtensa/kernel/traps.c

···
 void __noreturn die(const char * str, struct pt_regs * regs, long err)
 {
 	static int die_counter;
-	const char *pr = "";
-
-	if (IS_ENABLED(CONFIG_PREEMPTION))
-		pr = IS_ENABLED(CONFIG_PREEMPT_RT) ? " PREEMPT_RT" : " PREEMPT";
 
 	console_verbose();
 	spin_lock_irq(&die_lock);
 
-	pr_info("%s: sig: %ld [#%d]%s\n", str, err, ++die_counter, pr);
+	pr_info("%s: sig: %ld [#%d]\n", str, err, ++die_counter);
 	show_regs(regs);
 	if (!user_mode(regs))
 		show_stack(NULL, (unsigned long *)regs->areg[1], KERN_INFO);

fs/proc/base.c

···
 #endif
 
 
-#ifdef CONFIG_SCHED_DEBUG
 /*
  * Print out various scheduling related per-task fields:
  */
···
 	.llseek		= seq_lseek,
 	.release	= single_release,
 };
-
-#endif
 
 #ifdef CONFIG_SCHED_AUTOGROUP
 /*
···
 	ONE("status",     S_IRUGO, proc_pid_status),
 	ONE("personality", S_IRUSR, proc_pid_personality),
 	ONE("limits",	  S_IRUGO, proc_pid_limits),
-#ifdef CONFIG_SCHED_DEBUG
 	REG("sched",      S_IRUGO|S_IWUSR, proc_pid_sched_operations),
-#endif
 #ifdef CONFIG_SCHED_AUTOGROUP
 	REG("autogroup",  S_IRUGO|S_IWUSR, proc_pid_sched_autogroup_operations),
 #endif
···
 	ONE("status",    S_IRUGO, proc_pid_status),
 	ONE("personality", S_IRUSR, proc_pid_personality),
 	ONE("limits",	 S_IRUGO, proc_pid_limits),
-#ifdef CONFIG_SCHED_DEBUG
 	REG("sched",     S_IRUGO|S_IWUSR, proc_pid_sched_operations),
-#endif
 	NOD("comm",      S_IFREG|S_IRUGO|S_IWUSR,
 			 &proc_tid_comm_inode_operations,
 			 &proc_pid_set_comm_operations, {}),

include/linux/cpuset.h

···
 
 extern bool current_cpuset_is_being_rebound(void);
 
+extern void dl_rebuild_rd_accounting(void);
 extern void rebuild_sched_domains(void);
 
 extern void cpuset_print_current_mems_allowed(void);
+extern void cpuset_reset_sched_domains(void);
 
 /*
  * read_mems_allowed_begin is required when making decisions involving
···
 	return false;
 }
 
+static inline void dl_rebuild_rd_accounting(void)
+{
+}
+
 static inline void rebuild_sched_domains(void)
+{
+	partition_sched_domains(1, NULL, NULL);
+}
+
+static inline void cpuset_reset_sched_domains(void)
 {
 	partition_sched_domains(1, NULL, NULL);
 }

include/linux/energy_model.h

···
 	struct em_perf_state *ps;
 	int i;
 
-#ifdef CONFIG_SCHED_DEBUG
 	WARN_ONCE(!rcu_read_lock_held(), "EM: rcu read lock needed\n");
-#endif
 
 	if (!sum_util)
 		return 0;

include/linux/preempt.h

···
 	return IS_ENABLED(CONFIG_PREEMPT_RT);
 }
 
+extern const char *preempt_model_str(void);
+
 /*
  * Does the preemption model allow non-cooperative preemption?
  *

include/linux/sched.h

···
 #ifdef CONFIG_SMP
 extern struct root_domain def_root_domain;
 extern struct mutex sched_domains_mutex;
+extern void sched_domains_mutex_lock(void);
+extern void sched_domains_mutex_unlock(void);
+#else
+static inline void sched_domains_mutex_lock(void) { }
+static inline void sched_domains_mutex_unlock(void) { }
 #endif
 
 struct sched_param {

include/linux/sched/deadline.h

···
 struct root_domain;
 extern void dl_add_task_root_domain(struct task_struct *p);
 extern void dl_clear_root_domain(struct root_domain *rd);
+extern void dl_clear_root_domain_cpu(int cpu);
 
 #endif /* CONFIG_SMP */
+
+extern u64 dl_cookie;
+extern bool dl_bw_visited(int cpu, u64 cookie);
 
 #endif /* _LINUX_SCHED_DEADLINE_H */

include/linux/sched/debug.h

···
 
 extern void sched_show_task(struct task_struct *p);
 
-#ifdef CONFIG_SCHED_DEBUG
 struct seq_file;
 extern void proc_sched_show_task(struct task_struct *p,
 				 struct pid_namespace *ns, struct seq_file *m);
 extern void proc_sched_set_task(struct task_struct *p);
-#endif
 
 /* Attach to any functions which should be ignored in wchan output. */
 #define __sched		__section(".sched.text")

include/linux/sched/idle.h

···
 	return unlikely(tif_need_resched());
 }
 
+static __always_inline void current_clr_polling(void)
+{
+	__current_clr_polling();
+
+	/*
+	 * Ensure we check TIF_NEED_RESCHED after we clear the polling bit.
+	 * Once the bit is cleared, we'll get IPIs with every new
+	 * TIF_NEED_RESCHED and the IPI handler, scheduler_ipi(), will also
+	 * fold.
+	 */
+	smp_mb__after_atomic(); /* paired with resched_curr() */
+
+	preempt_fold_need_resched();
+}
+
 #else
 static inline void __current_set_polling(void) { }
 static inline void __current_clr_polling(void) { }
···
 {
 	return unlikely(tif_need_resched());
 }
-#endif
 
 static __always_inline void current_clr_polling(void)
 {
 	__current_clr_polling();
 
-	/*
-	 * Ensure we check TIF_NEED_RESCHED after we clear the polling bit.
-	 * Once the bit is cleared, we'll get IPIs with every new
-	 * TIF_NEED_RESCHED and the IPI handler, scheduler_ipi(), will also
-	 * fold.
-	 */
 	smp_mb(); /* paired with resched_curr() */
 
 	preempt_fold_need_resched();
 }
+#endif
 
 #endif /* _LINUX_SCHED_IDLE_H */

include/linux/sched/mm.h

···
 
 static inline void membarrier_mm_sync_core_before_usermode(struct mm_struct *mm)
 {
+	/*
+	 * The atomic_read() below prevents CSE. The following should
+	 * help the compiler generate more efficient code on architectures
+	 * where sync_core_before_usermode() is a no-op.
+	 */
+	if (!IS_ENABLED(CONFIG_ARCH_HAS_SYNC_CORE_BEFORE_USERMODE))
+		return;
 	if (current->mm != mm)
 		return;
 	if (likely(!(atomic_read(&mm->membarrier_state) &

include/linux/sched/topology.h

···
 };
 #undef SD_FLAG
 
-#ifdef CONFIG_SCHED_DEBUG
-
 struct sd_flag_debug {
 	unsigned int meta_flags;
 	char *name;
 };
 extern const struct sd_flag_debug sd_flag_debug[];
-
-#endif
 
 #ifdef CONFIG_SCHED_SMT
 static inline int cpu_smt_flags(void)
···
 	return to_cpumask(sd->span);
 }
 
-extern void partition_sched_domains_locked(int ndoms_new,
-					   cpumask_var_t doms_new[],
-					   struct sched_domain_attr *dattr_new);
-
 extern void partition_sched_domains(int ndoms_new, cpumask_var_t doms_new[],
 				    struct sched_domain_attr *dattr_new);
 
···
 #else /* CONFIG_SMP */
 
 struct sched_domain_attr;
-
-static inline void
-partition_sched_domains_locked(int ndoms_new, cpumask_var_t doms_new[],
-			       struct sched_domain_attr *dattr_new)
-{
-}
 
 static inline void
 partition_sched_domains(int ndoms_new, cpumask_var_t doms_new[],

include/trace/events/sched.h

···
 {
 	unsigned int state;
 
-#ifdef CONFIG_SCHED_DEBUG
 	BUG_ON(p != current);
-#endif /* CONFIG_SCHED_DEBUG */
 
 	/*
 	 * Preemption ignores task state, therefore preempted tasks are always

kernel/cgroup/cpuset.c

···
 	css_task_iter_end(&it);
 }
 
-static void dl_rebuild_rd_accounting(void)
+void dl_rebuild_rd_accounting(void)
 {
 	struct cpuset *cs = NULL;
 	struct cgroup_subsys_state *pos_css;
+	int cpu;
+	u64 cookie = ++dl_cookie;
 
 	lockdep_assert_held(&cpuset_mutex);
 	lockdep_assert_cpus_held();
···
 
 	rcu_read_lock();
 
-	/*
-	 * Clear default root domain DL accounting, it will be computed again
-	 * if a task belongs to it.
-	 */
-	dl_clear_root_domain(&def_root_domain);
+	for_each_possible_cpu(cpu) {
+		if (dl_bw_visited(cpu, cookie))
+			continue;
+
+		dl_clear_root_domain_cpu(cpu);
+	}
 
 	cpuset_for_each_descendant_pre(cs, pos_css, &top_cpuset) {
 
···
 		css_put(&cs->css);
 	}
 	rcu_read_unlock();
-}
-
-static void
-partition_and_rebuild_sched_domains(int ndoms_new, cpumask_var_t doms_new[],
-				    struct sched_domain_attr *dattr_new)
-{
-	mutex_lock(&sched_domains_mutex);
-	partition_sched_domains_locked(ndoms_new, doms_new, dattr_new);
-	dl_rebuild_rd_accounting();
-	mutex_unlock(&sched_domains_mutex);
 }
 
 /*
···
 	ndoms = generate_sched_domains(&doms, &attr);
 
 	/* Have scheduler rebuild the domains */
-	partition_and_rebuild_sched_domains(ndoms, doms, attr);
+	partition_sched_domains(ndoms, doms, attr);
 }
 #else /* !CONFIG_SMP */
 void rebuild_sched_domains_locked(void)
···
 	cpus_read_lock();
 	rebuild_sched_domains_cpuslocked();
 	cpus_read_unlock();
+}
+
+void cpuset_reset_sched_domains(void)
+{
+	mutex_lock(&cpuset_mutex);
+	partition_sched_domains(1, NULL, NULL);
+	mutex_unlock(&cpuset_mutex);
 }
 
 /**

kernel/rseq.c

···
 	return -EFAULT;
 }
 
-static void rseq_set_ro_fields(struct task_struct *t, u32 cpu_id_start, u32 cpu_id,
-			       u32 node_id, u32 mm_cid)
-{
-	rseq_kernel_fields(t)->cpu_id_start = cpu_id;
-	rseq_kernel_fields(t)->cpu_id = cpu_id;
-	rseq_kernel_fields(t)->node_id = node_id;
-	rseq_kernel_fields(t)->mm_cid = mm_cid;
-}
+/*
+ * Update an rseq field and its in-kernel copy in lock-step to keep a coherent
+ * state.
+ */
+#define rseq_unsafe_put_user(t, value, field, error_label)		\
+	do {								\
+		unsafe_put_user(value, &t->rseq->field, error_label);	\
+		rseq_kernel_fields(t)->field = value;			\
+	} while (0)
+
 #else
 static int rseq_validate_ro_fields(struct task_struct *t)
 {
 	return 0;
 }
 
-static void rseq_set_ro_fields(struct task_struct *t, u32 cpu_id_start, u32 cpu_id,
-			       u32 node_id, u32 mm_cid)
-{
-}
+#define rseq_unsafe_put_user(t, value, field, error_label)		\
+	unsafe_put_user(value, &t->rseq->field, error_label)
 #endif
 
 /*
···
 	WARN_ON_ONCE((int) mm_cid < 0);
 	if (!user_write_access_begin(rseq, t->rseq_len))
 		goto efault;
-	unsafe_put_user(cpu_id, &rseq->cpu_id_start, efault_end);
-	unsafe_put_user(cpu_id, &rseq->cpu_id, efault_end);
-	unsafe_put_user(node_id, &rseq->node_id, efault_end);
-	unsafe_put_user(mm_cid, &rseq->mm_cid, efault_end);
+
+	rseq_unsafe_put_user(t, cpu_id, cpu_id_start, efault_end);
+	rseq_unsafe_put_user(t, cpu_id, cpu_id, efault_end);
+	rseq_unsafe_put_user(t, node_id, node_id, efault_end);
+	rseq_unsafe_put_user(t, mm_cid, mm_cid, efault_end);
+
 	/*
 	 * Additional feature fields added after ORIG_RSEQ_SIZE
 	 * need to be conditionally updated only if
 	 * t->rseq_len != ORIG_RSEQ_SIZE.
 	 */
 	user_write_access_end();
-	rseq_set_ro_fields(t, cpu_id, cpu_id, node_id, mm_cid);
 	trace_rseq_update(t);
 	return 0;
 
···
 
 static int rseq_reset_rseq_cpu_node_id(struct task_struct *t)
 {
+	struct rseq __user *rseq = t->rseq;
 	u32 cpu_id_start = 0, cpu_id = RSEQ_CPU_ID_UNINITIALIZED, node_id = 0,
 	    mm_cid = 0;
 
···
 	 * Validate read-only rseq fields.
 	 */
 	if (rseq_validate_ro_fields(t))
-		return -EFAULT;
-	/*
-	 * Reset cpu_id_start to its initial state (0).
-	 */
-	if (put_user(cpu_id_start, &t->rseq->cpu_id_start))
-		return -EFAULT;
-	/*
-	 * Reset cpu_id to RSEQ_CPU_ID_UNINITIALIZED, so any user coming
-	 * in after unregistration can figure out that rseq needs to be
-	 * registered again.
-	 */
-	if (put_user(cpu_id, &t->rseq->cpu_id))
-		return -EFAULT;
-	/*
-	 * Reset node_id to its initial state (0).
-	 */
-	if (put_user(node_id, &t->rseq->node_id))
-		return -EFAULT;
-	/*
-	 * Reset mm_cid to its initial state (0).
-	 */
-	if (put_user(mm_cid, &t->rseq->mm_cid))
-		return -EFAULT;
+		goto efault;
 
-	rseq_set_ro_fields(t, cpu_id_start, cpu_id, node_id, mm_cid);
+	if (!user_write_access_begin(rseq, t->rseq_len))
+		goto efault;
+
+	/*
+	 * Reset all fields to their initial state.
+	 *
+	 * All fields have an initial state of 0 except cpu_id which is set to
+	 * RSEQ_CPU_ID_UNINITIALIZED, so that any user coming in after
+	 * unregistration can figure out that rseq needs to be registered
+	 * again.
+	 */
+	rseq_unsafe_put_user(t, cpu_id_start, cpu_id_start, efault_end);
+	rseq_unsafe_put_user(t, cpu_id, cpu_id, efault_end);
+	rseq_unsafe_put_user(t, node_id, node_id, efault_end);
+	rseq_unsafe_put_user(t, mm_cid, mm_cid, efault_end);
 
 	/*
 	 * Additional feature fields added after ORIG_RSEQ_SIZE
 	 * need to be conditionally reset only if
 	 * t->rseq_len != ORIG_RSEQ_SIZE.
 	 */
+	user_write_access_end();
+	return 0;
+
+efault_end:
+	user_write_access_end();
+efault:
+	return -EFAULT;
+}
+
+/*
+ * Get the user-space pointer value stored in the 'rseq_cs' field.
+ */
+static int rseq_get_rseq_cs_ptr_val(struct rseq __user *rseq, u64 *rseq_cs)
+{
+	if (!rseq_cs)
+		return -EFAULT;
+
+#ifdef CONFIG_64BIT
+	if (get_user(*rseq_cs, &rseq->rseq_cs))
+		return -EFAULT;
+#else
+	if (copy_from_user(rseq_cs, &rseq->rseq_cs, sizeof(*rseq_cs)))
+		return -EFAULT;
+#endif
+
 	return 0;
 }
 
+/*
+ * If the rseq_cs field of 'struct rseq' contains a valid pointer to
+ * user-space, copy 'struct rseq_cs' from user-space and validate its fields.
+ */
 static int rseq_get_rseq_cs(struct task_struct *t, struct rseq_cs *rseq_cs)
 {
 	struct rseq_cs __user *urseq_cs;
···
 	u32 sig;
 	int ret;
 
-#ifdef CONFIG_64BIT
-	if (get_user(ptr, &t->rseq->rseq_cs))
-		return -EFAULT;
-#else
-	if (copy_from_user(&ptr, &t->rseq->rseq_cs, sizeof(ptr)))
-		return -EFAULT;
-#endif
+	ret = rseq_get_rseq_cs_ptr_val(t->rseq, &ptr);
+	if (ret)
+		return ret;
+
+	/* If the rseq_cs pointer is NULL, return a cleared struct rseq_cs. */
 	if (!ptr) {
 		memset(rseq_cs, 0, sizeof(*rseq_cs));
 		return 0;
 	}
+	/* Check that the pointer value fits in the user-space process space. */
 	if (ptr >= TASK_SIZE)
 		return -EINVAL;
 	urseq_cs = (struct rseq_cs __user *)(unsigned long)ptr;
···
 	return !!event_mask;
 }
 
-static int clear_rseq_cs(struct task_struct *t)
+static int clear_rseq_cs(struct rseq __user *rseq)
 {
 	/*
 	 * The rseq_cs field is set to NULL on preemption or signal
···
 	 * Set rseq_cs to NULL.
 	 */
 #ifdef CONFIG_64BIT
-	return put_user(0UL, &t->rseq->rseq_cs);
+	return put_user(0UL, &rseq->rseq_cs);
 #else
-	if (clear_user(&t->rseq->rseq_cs, sizeof(t->rseq->rseq_cs)))
+	if (clear_user(&rseq->rseq_cs, sizeof(rseq->rseq_cs)))
 		return -EFAULT;
 	return 0;
 #endif
···
 	 * Clear the rseq_cs pointer and return.
 	 */
 	if (!in_rseq_cs(ip, &rseq_cs))
-		return clear_rseq_cs(t);
+		return clear_rseq_cs(t->rseq);
 	ret = rseq_need_restart(t, rseq_cs.flags);
 	if (ret <= 0)
 		return ret;
-	ret = clear_rseq_cs(t);
+	ret = clear_rseq_cs(t->rseq);
 	if (ret)
 		return ret;
 	trace_rseq_ip_fixup(ip, rseq_cs.start_ip, rseq_cs.post_commit_offset,
···
 		int, flags, u32, sig)
 {
 	int ret;
+	u64 rseq_cs;
 
 	if (flags & RSEQ_FLAG_UNREGISTER) {
 		if (flags & ~RSEQ_FLAG_UNREGISTER)
···
 		return -EINVAL;
 	if (!access_ok(rseq, rseq_len))
 		return -EFAULT;
+
+	/*
+	 * If the rseq_cs pointer is non-NULL on registration, clear it to
+	 * avoid a potential segfault on return to user-space. The proper thing
+	 * to do would have been to fail the registration but this would break
+	 * older libcs that reuse the rseq area for new threads without
+	 * clearing the fields.
+	 */
+	if (rseq_get_rseq_cs_ptr_val(rseq, &rseq_cs))
+	        return -EFAULT;
+	if (rseq_cs && clear_rseq_cs(rseq))
+		return -EFAULT;
+
 #ifdef CONFIG_DEBUG_RSEQ
 	/*
 	 * Initialize the in-kernel rseq fields copy for validation of

kernel/sched/build_utility.c

···
 # include "cpufreq_schedutil.c"
 #endif
 
-#ifdef CONFIG_SCHED_DEBUG
-# include "debug.c"
-#endif
+#include "debug.c"
 
 #ifdef CONFIG_SCHEDSTATS
 # include "stats.c"

kernel/sched/core.c

···
 #include "autogroup.h"
 #include "pelt.h"
 #include "smp.h"
-#include "stats.h"
 
 #include "../workqueue_internal.h"
 #include "../../io_uring/io-wq.h"
···
 
 DEFINE_PER_CPU_SHARED_ALIGNED(struct rq, runqueues);
 
-#ifdef CONFIG_SCHED_DEBUG
 /*
  * Debugging: various feature bits
  *
···
  */
 #define SCHED_FEAT(name, enabled)	\
 	(1UL << __SCHED_FEAT_##name) * enabled |
-const_debug unsigned int sysctl_sched_features =
+__read_mostly unsigned int sysctl_sched_features =
 #include "features.h"
 	0;
 #undef SCHED_FEAT
···
  */
 __read_mostly int sysctl_resched_latency_warn_ms = 100;
 __read_mostly int sysctl_resched_latency_warn_once = 1;
-#endif /* CONFIG_SCHED_DEBUG */
 
 /*
  * Number of tasks to iterate in a single balance run.
  * Limited because this is done with IRQs disabled.
  */
-const_debug unsigned int sysctl_sched_nr_migrate = SCHED_NR_MIGRATE_BREAK;
+__read_mostly unsigned int sysctl_sched_nr_migrate = SCHED_NR_MIGRATE_BREAK;
 
 __read_mostly int scheduler_running;
 
···
 	if (rq->clock_update_flags & RQCF_ACT_SKIP)
 		return;
 
-#ifdef CONFIG_SCHED_DEBUG
 	if (sched_feat(WARN_DOUBLE_CLOCK))
-		SCHED_WARN_ON(rq->clock_update_flags & RQCF_UPDATED);
+		WARN_ON_ONCE(rq->clock_update_flags & RQCF_UPDATED);
 	rq->clock_update_flags |= RQCF_UPDATED;
-#endif
+
 	clock = sched_clock_cpu(cpu_of(rq));
 	scx_rq_clock_update(rq, clock);
 
···
 
 	bucket = &uc_rq->bucket[uc_se->bucket_id];
 
-	SCHED_WARN_ON(!bucket->tasks);
+	WARN_ON_ONCE(!bucket->tasks);
 	if (likely(bucket->tasks))
 		bucket->tasks--;
 
···
 	 * Defensive programming: this should never happen. If it happens,
 	 * e.g. due to future modification, warn and fix up the expected value.
 	 */
-	SCHED_WARN_ON(bucket->value > rq_clamp);
+	WARN_ON_ONCE(bucket->value > rq_clamp);
 	if (bucket->value >= rq_clamp) {
 		bkt_clamp = uclamp_rq_max_value(rq, clamp_id, uc_se->value);
 		uclamp_rq_set(rq, clamp_id, bkt_clamp);
···
 	 * The condition is constructed such that a NOP is generated when
 	 * sched_uclamp_used is disabled.
 	 */
-	if (!static_branch_unlikely(&sched_uclamp_used))
+	if (!uclamp_is_used())
 		return;
 
 	if (unlikely(!p->sched_class->uclamp_enabled))
···
 	 * The condition is constructed such that a NOP is generated when
 	 * sched_uclamp_used is disabled.
 	 */
-	if (!static_branch_unlikely(&sched_uclamp_used))
+	if (!uclamp_is_used())
 		return;
 
 	if (unlikely(!p->sched_class->uclamp_enabled))
···
 	}
 
 	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();
 	}
 
···
 
 void deactivate_task(struct rq *rq, struct task_struct *p, int flags)
 {
-	SCHED_WARN_ON(flags & DEQUEUE_SLEEP);
+	WARN_ON_ONCE(flags & DEQUEUE_SLEEP);
 
 	WRITE_ONCE(p->on_rq, TASK_ON_RQ_MIGRATING);
 	ASSERT_EXCLUSIVE_WRITER(p->on_rq);
···
 	 * XXX do further audits, this smells like something putrid.
 	 */
 	if (ctx->flags & SCA_MIGRATE_DISABLE)
-		SCHED_WARN_ON(!p->on_cpu);
+		WARN_ON_ONCE(!p->on_cpu);
 	else
 		lockdep_assert_held(&p->pi_lock);
 
···
 
 void set_task_cpu(struct task_struct *p, unsigned int new_cpu)
 {
-#ifdef CONFIG_SCHED_DEBUG
 	unsigned int state = READ_ONCE(p->__state);
 
 	/*
···
 	WARN_ON_ONCE(!cpu_online(new_cpu));
 
 	WARN_ON_ONCE(is_migration_disabled(p));
-#endif
 
 	trace_sched_migrate_task(p, new_cpu);
 
···
 		 *  - we're serialized against set_special_state() by virtue of
 		 *    it disabling IRQs (this allows not taking ->pi_lock).
 		 */
-		SCHED_WARN_ON(p->se.sched_delayed);
+		WARN_ON_ONCE(p->se.sched_delayed);
 		if (!ttwu_state_match(p, state, &success))
 			goto out;
 
···
 	INIT_LIST_HEAD(&p->se.group_node);
 
 	/* A delayed task cannot be in clone(). */
-	SCHED_WARN_ON(p->se.sched_delayed);
+	WARN_ON_ONCE(p->se.sched_delayed);
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
 	p->se.cfs_rq			= NULL;
···
 	return ns;
 }
 
-#ifdef CONFIG_SCHED_DEBUG
 static u64 cpu_resched_latency(struct rq *rq)
 {
 	int latency_warn_ms = READ_ONCE(sysctl_resched_latency_warn_ms);
···
 	return 1;
 }
 __setup("resched_latency_warn_ms=", setup_resched_latency_warn_ms);
-#else
-static inline u64 cpu_resched_latency(struct rq *rq) { return 0; }
-#endif /* CONFIG_SCHED_DEBUG */
 
 /*
  * This function gets called by the timer code, with HZ frequency.
···
 			 * we are always sure that there is no proxy (only a
 			 * single task is running).
 			 */
-			SCHED_WARN_ON(rq->curr != rq->donor);
+			WARN_ON_ONCE(rq->curr != rq->donor);
 			update_rq_clock(rq);
 
 			if (!is_idle_task(curr)) {
···
 		preempt_count_set(PREEMPT_DISABLED);
 	}
 	rcu_sleep_check();
-	SCHED_WARN_ON(ct_state() == CT_STATE_USER);
+	WARN_ON_ONCE(ct_state() == CT_STATE_USER);
 
 	profile_hit(SCHED_PROFILING, __builtin_return_address(0));
 
···
 picked:
 	clear_tsk_need_resched(prev);
 	clear_preempt_need_resched();
-#ifdef CONFIG_SCHED_DEBUG
 	rq->last_seen_need_resched_ns = 0;
-#endif
 
 	if (likely(prev != next)) {
 		rq->nr_switches++;
···
 	 * deadlock if the callback attempts to acquire a lock which is
 	 * already acquired.
 	 */
-	SCHED_WARN_ON(current->__state & TASK_RTLOCK_WAIT);
+	WARN_ON_ONCE(current->__state & TASK_RTLOCK_WAIT);
 
 	/*
 	 * If we are going to sleep and we have plugged IO queued,
···
 int default_wake_function(wait_queue_entry_t *curr, unsigned mode, int wake_flags,
 			  void *key)
 {
-	WARN_ON_ONCE(IS_ENABLED(CONFIG_SCHED_DEBUG) && wake_flags & ~(WF_SYNC|WF_CURRENT_CPU));
+	WARN_ON_ONCE(wake_flags & ~(WF_SYNC|WF_CURRENT_CPU));
 	return try_to_wake_up(curr->private, mode, wake_flags);
 }
 EXPORT_SYMBOL(default_wake_function);
···
 
 #else /* !CONFIG_PREEMPT_DYNAMIC: */
 
+#define preempt_dynamic_mode -1
+
 static inline void preempt_dynamic_init(void) { }
 
 #endif /* CONFIG_PREEMPT_DYNAMIC */
+
+const char *preempt_modes[] = {
+	"none", "voluntary", "full", "lazy", NULL,
+};
+
+const char *preempt_model_str(void)
+{
+	bool brace = IS_ENABLED(CONFIG_PREEMPT_RT) &&
+		(IS_ENABLED(CONFIG_PREEMPT_DYNAMIC) ||
+		 IS_ENABLED(CONFIG_PREEMPT_LAZY));
+	static char buf[128];
+
+	if (IS_ENABLED(CONFIG_PREEMPT_BUILD)) {
+		struct seq_buf s;
+
+		seq_buf_init(&s, buf, sizeof(buf));
+		seq_buf_puts(&s, "PREEMPT");
+
+		if (IS_ENABLED(CONFIG_PREEMPT_RT))
+			seq_buf_printf(&s, "%sRT%s",
+				       brace ? "_{" : "_",
+				       brace ? "," : "");
+
+		if (IS_ENABLED(CONFIG_PREEMPT_DYNAMIC)) {
+			seq_buf_printf(&s, "(%s)%s",
+				       preempt_dynamic_mode > 0 ?
+				       preempt_modes[preempt_dynamic_mode] : "undef",
+				       brace ? "}" : "");
+			return seq_buf_str(&s);
+		}
+
+		if (IS_ENABLED(CONFIG_PREEMPT_LAZY)) {
+			seq_buf_printf(&s, "LAZY%s",
+				       brace ? "}" : "");
+			return seq_buf_str(&s);
+		}
+
+		return seq_buf_str(&s);
+	}
+
+	if (IS_ENABLED(CONFIG_PREEMPT_VOLUNTARY_BUILD))
+		return "VOLUNTARY";
+
+	return "NONE";
+}
 
 int io_schedule_prepare(void)
 {
···
 			sched_show_task(p);
 	}
 
-#ifdef CONFIG_SCHED_DEBUG
 	if (!state_filter)
 		sysrq_sched_debug_show();
-#endif
+
 	rcu_read_unlock();
 	/*
 	 * Only show locks if all tasks are dumped:
···
 		 * operation in the resume sequence, just build a single sched
 		 * domain, ignoring cpusets.
 		 */
-		partition_sched_domains(1, NULL, NULL);
+		cpuset_reset_sched_domains();
 		if (--num_cpus_frozen)
 			return;
 		/*
···
 		cpuset_update_active_cpus();
 	} else {
 		num_cpus_frozen++;
-		partition_sched_domains(1, NULL, NULL);
+		cpuset_reset_sched_domains();
 	}
 }
 
···
 	 * CPU masks are stable and all blatant races in the below code cannot
 	 * happen.
 	 */
-	mutex_lock(&sched_domains_mutex);
+	sched_domains_mutex_lock();
 	sched_init_domains(cpu_active_mask);
-	mutex_unlock(&sched_domains_mutex);
+	sched_domains_mutex_unlock();
 
 	/* Move init over to a non-isolated CPU */
 	if (set_cpus_allowed_ptr(current, housekeeping_cpumask(HK_TYPE_DOMAIN)) < 0)
···
 	unsigned int clamps;
 
 	lockdep_assert_held(&uclamp_mutex);
-	SCHED_WARN_ON(!rcu_read_lock_held());
+	WARN_ON_ONCE(!rcu_read_lock_held());
 
 	css_for_each_descendant_pre(css, top_css) {
 		uc_parent = css_tg(css)->parent
···
 	if (req.ret)
 		return req.ret;
 
-	static_branch_enable(&sched_uclamp_used);
+	sched_uclamp_enable();
 
 	guard(mutex)(&uclamp_mutex);
 	guard(rcu)();
···
 	struct mm_struct *mm;
 	int weight, cpu;
 
-	SCHED_WARN_ON(t != container_of(work, struct task_struct, cid_work));
+	WARN_ON_ONCE(t != container_of(work, struct task_struct, cid_work));
 
 	work->next = work;	/* Prevent double-add */
 	if (t->flags & PF_EXITING)

kernel/sched/core_sched.c

···
 	 * a cookie until after we've removed it, we must have core scheduling
 	 * enabled here.
 	 */
-	SCHED_WARN_ON((p->core_cookie || cookie) && !sched_core_enabled(rq));
+	WARN_ON_ONCE((p->core_cookie || cookie) && !sched_core_enabled(rq));
 
 	if (sched_core_enqueued(p))
 		sched_core_dequeue(rq, p, DEQUEUE_SAVE);

kernel/sched/deadline.c

···
 	}
 }
 
-static inline bool dl_bw_visited(int cpu, u64 gen)
+bool dl_bw_visited(int cpu, u64 cookie)
 {
 	struct root_domain *rd = cpu_rq(cpu)->rd;
 
-	if (rd->visit_gen == gen)
+	if (rd->visit_cookie == cookie)
 		return true;
 
-	rd->visit_gen = gen;
+	rd->visit_cookie = cookie;
 	return false;
 }
 
···
 	return SCHED_CAPACITY_SCALE;
 }
 
-static inline bool dl_bw_visited(int cpu, u64 gen)
+bool dl_bw_visited(int cpu, u64 cookie)
 {
 	return false;
 }
···
 
 	lockdep_assert_rq_held(rq_of_dl_rq(dl_rq));
 	dl_rq->running_bw += dl_bw;
-	SCHED_WARN_ON(dl_rq->running_bw < old); /* overflow */
-	SCHED_WARN_ON(dl_rq->running_bw > dl_rq->this_bw);
+	WARN_ON_ONCE(dl_rq->running_bw < old); /* overflow */
+	WARN_ON_ONCE(dl_rq->running_bw > dl_rq->this_bw);
 	/* kick cpufreq (see the comment in kernel/sched/sched.h). */
 	cpufreq_update_util(rq_of_dl_rq(dl_rq), 0);
 }
···
 
 	lockdep_assert_rq_held(rq_of_dl_rq(dl_rq));
 	dl_rq->running_bw -= dl_bw;
-	SCHED_WARN_ON(dl_rq->running_bw > old); /* underflow */
+	WARN_ON_ONCE(dl_rq->running_bw > old); /* underflow */
 	if (dl_rq->running_bw > old)
 		dl_rq->running_bw = 0;
 	/* kick cpufreq (see the comment in kernel/sched/sched.h). */
···
 
 	lockdep_assert_rq_held(rq_of_dl_rq(dl_rq));
 	dl_rq->this_bw += dl_bw;
-	SCHED_WARN_ON(dl_rq->this_bw < old); /* overflow */
+	WARN_ON_ONCE(dl_rq->this_bw < old); /* overflow */
 }
 
 static inline
···
 
 	lockdep_assert_rq_held(rq_of_dl_rq(dl_rq));
 	dl_rq->this_bw -= dl_bw;
-	SCHED_WARN_ON(dl_rq->this_bw > old); /* underflow */
+	WARN_ON_ONCE(dl_rq->this_bw > old); /* underflow */
 	if (dl_rq->this_bw > old)
 		dl_rq->this_bw = 0;
-	SCHED_WARN_ON(dl_rq->running_bw > dl_rq->this_bw);
+	WARN_ON_ONCE(dl_rq->running_bw > dl_rq->this_bw);
 }
 
 static inline
···
 	struct dl_bw *dl_b;
 
 	raw_spin_lock_irqsave(&p->pi_lock, rf.flags);
-	if (!dl_task(p)) {
+	if (!dl_task(p) || dl_entity_is_special(&p->dl)) {
 		raw_spin_unlock_irqrestore(&p->pi_lock, rf.flags);
 		return;
 	}
···
 	rd->dl_bw.total_bw = 0;
 
 	/*
-	 * dl_server bandwidth is only restored when CPUs are attached to root
-	 * domains (after domains are created or CPUs moved back to the
-	 * default root doamin).
+	 * dl_servers are not tasks. Since dl_add_task_root_domain ignores
+	 * them, we need to account for them here explicitly.
 	 */
 	for_each_cpu(i, rd->span) {
 		struct sched_dl_entity *dl_se = &cpu_rq(i)->fair_server;
 
 		if (dl_server(dl_se) && cpu_active(i))
-			rd->dl_bw.total_bw += dl_se->dl_bw;
+			__dl_add(&rd->dl_bw, dl_se->dl_bw, dl_bw_cpus(i));
 	}
+}
+
+void dl_clear_root_domain_cpu(int cpu)
+{
+	dl_clear_root_domain(cpu_rq(cpu)->rd);
 }
 
 #endif /* CONFIG_SMP */
···
 #endif
 };
 
-/* Used for dl_bw check and update, used under sched_rt_handler()::mutex */
-static u64 dl_generation;
+/*
+ * Used for dl_bw check and update, used under sched_rt_handler()::mutex and
+ * sched_domains_mutex.
+ */
+u64 dl_cookie;
 
 int sched_dl_global_validate(void)
 {
 	u64 runtime = global_rt_runtime();
 	u64 period = global_rt_period();
 	u64 new_bw = to_ratio(period, runtime);
-	u64 gen = ++dl_generation;
+	u64 cookie = ++dl_cookie;
 	struct dl_bw *dl_b;
 	int cpu, cpus, ret = 0;
 	unsigned long flags;
···
 	for_each_online_cpu(cpu) {
 		rcu_read_lock_sched();
 
-		if (dl_bw_visited(cpu, gen))
+		if (dl_bw_visited(cpu, cookie))
 			goto next;
 
 		dl_b = dl_bw_of(cpu);
···
 void sched_dl_do_global(void)
 {
 	u64 new_bw = -1;
-	u64 gen = ++dl_generation;
+	u64 cookie = ++dl_cookie;
 	struct dl_bw *dl_b;
 	int cpu;
 	unsigned long flags;
···
 	for_each_possible_cpu(cpu) {
 		rcu_read_lock_sched();
 
-		if (dl_bw_visited(cpu, gen)) {
+		if (dl_bw_visited(cpu, cookie)) {
 			rcu_read_unlock_sched();
 			continue;
 		}
···
 }
 #endif
 
-#ifdef CONFIG_SCHED_DEBUG
 void print_dl_stats(struct seq_file *m, int cpu)
 {
 	print_dl_rq(m, cpu, &cpu_rq(cpu)->dl);
 }
-#endif /* CONFIG_SCHED_DEBUG */

kernel/sched/debug.c

···
 
 static int sched_dynamic_show(struct seq_file *m, void *v)
 {
-	static const char * preempt_modes[] = {
-		"none", "voluntary", "full", "lazy",
-	};
-	int j = ARRAY_SIZE(preempt_modes) - !IS_ENABLED(CONFIG_ARCH_HAS_PREEMPT_LAZY);
 	int i = IS_ENABLED(CONFIG_PREEMPT_RT) * 2;
+	int j;
+
+	/* Count entries in NULL terminated preempt_modes */
+	for (j = 0; preempt_modes[j]; j++)
+		;
+	j -= !IS_ENABLED(CONFIG_ARCH_HAS_PREEMPT_LAZY);
 
 	for (; i < j; i++) {
 		if (preempt_dynamic_mode == i)
···
 	bool orig;
 
 	cpus_read_lock();
-	mutex_lock(&sched_domains_mutex);
+	sched_domains_mutex_lock();
 
 	orig = sched_debug_verbose;
 	result = debugfs_write_file_bool(filp, ubuf, cnt, ppos);
···
 		sd_dentry = NULL;
 	}
 
-	mutex_unlock(&sched_domains_mutex);
+	sched_domains_mutex_unlock();
 	cpus_read_unlock();
 
 	return result;
···
 	debugfs_create_u32("migration_cost_ns", 0644, debugfs_sched, &sysctl_sched_migration_cost);
 	debugfs_create_u32("nr_migrate", 0644, debugfs_sched, &sysctl_sched_nr_migrate);
 
-	mutex_lock(&sched_domains_mutex);
+	sched_domains_mutex_lock();
 	update_sched_domain_debugfs();
-	mutex_unlock(&sched_domains_mutex);
+	sched_domains_mutex_unlock();
 #endif
 
 #ifdef CONFIG_NUMA_BALANCING

kernel/sched/ext.c

···
 {
 	int cpu = cpu_of(rq);
 
-	SCHED_WARN_ON(task_cpu(p) == cpu);
+	WARN_ON_ONCE(task_cpu(p) == cpu);
 
 	/*
 	 * If @p has migration disabled, @p->cpus_ptr is updated to contain only

kernel/sched/fair.c

···
 /*
  * Minimal preemption granularity for CPU-bound tasks:
  *
- * (default: 0.75 msec * (1 + ilog(ncpus)), units: nanoseconds)
+ * (default: 0.70 msec * (1 + ilog(ncpus)), units: nanoseconds)
  */
-unsigned int sysctl_sched_base_slice			= 750000ULL;
-static unsigned int normalized_sysctl_sched_base_slice	= 750000ULL;
+unsigned int sysctl_sched_base_slice			= 700000ULL;
+static unsigned int normalized_sysctl_sched_base_slice	= 700000ULL;
 
-const_debug unsigned int sysctl_sched_migration_cost	= 500000UL;
+__read_mostly unsigned int sysctl_sched_migration_cost	= 500000UL;
 
 static int __init setup_sched_thermal_decay_shift(char *str)
 {
···
 
 static inline void assert_list_leaf_cfs_rq(struct rq *rq)
 {
-	SCHED_WARN_ON(rq->tmp_alone_branch != &rq->leaf_cfs_rq_list);
+	WARN_ON_ONCE(rq->tmp_alone_branch != &rq->leaf_cfs_rq_list);
 }
 
 /* Iterate through all leaf cfs_rq's on a runqueue */
···
 {
 	s64 vlag, limit;
 
-	SCHED_WARN_ON(!se->on_rq);
+	WARN_ON_ONCE(!se->on_rq);
 
 	vlag = avg_vruntime(cfs_rq) - se->vruntime;
 	limit = calc_delta_fair(max_t(u64, 2*se->slice, TICK_NSEC), se);
···
 }
 
 /*
+ * HACK, stash a copy of deadline at the point of pick in vlag,
+ * which isn't used until dequeue.
+ */
+static inline void set_protect_slice(struct sched_entity *se)
+{
+	se->vlag = se->deadline;
+}
+
+static inline bool protect_slice(struct sched_entity *se)
+{
+	return se->vlag == se->deadline;
+}
+
+static inline void cancel_protect_slice(struct sched_entity *se)
+{
+	if (protect_slice(se))
+		se->vlag = se->deadline + 1;
+}
+
+/*
  * Earliest Eligible Virtual Deadline First
  *
  * In order to provide latency guarantees for different request sizes
···
 	if (curr && (!curr->on_rq || !entity_eligible(cfs_rq, curr)))
 		curr = NULL;
 
-	/*
-	 * Once selected, run a task until it either becomes non-eligible or
-	 * until it gets a new slice. See the HACK in set_next_entity().
-	 */
-	if (sched_feat(RUN_TO_PARITY) && curr && curr->vlag == curr->deadline)
+	if (sched_feat(RUN_TO_PARITY) && curr && protect_slice(curr))
 		return curr;
 
 	/* Pick the leftmost entity if it's eligible */
···
 	return best;
 }
 
-#ifdef CONFIG_SCHED_DEBUG
 struct sched_entity *__pick_last_entity(struct cfs_rq *cfs_rq)
 {
 	struct rb_node *last = rb_last(&cfs_rq->tasks_timeline.rb_root);
···
 
 	return 0;
 }
-#endif
 #endif
 
 static void clear_buddies(struct cfs_rq *cfs_rq, struct sched_entity *se);
···
 	bool vma_pids_skipped;
 	bool vma_pids_forced = false;
 
-	SCHED_WARN_ON(p != container_of(work, struct task_struct, numa_work));
+	WARN_ON_ONCE(p != container_of(work, struct task_struct, numa_work));
 
 	work->next = work;
 	/*
···
 	 * Make sure that rounding and/or propagation of PELT values never
 	 * break this.
 	 */
-	SCHED_WARN_ON(sa->load_avg ||
+	WARN_ON_ONCE(sa->load_avg ||
 		      sa->util_avg ||
 		      sa->runnable_avg);
 
···
 	clear_buddies(cfs_rq, se);
 
 	if (flags & DEQUEUE_DELAYED) {
-		SCHED_WARN_ON(!se->sched_delayed);
+		WARN_ON_ONCE(!se->sched_delayed);
 	} else {
 		bool delay = sleep;
 		/*
···
 		if (flags & DEQUEUE_SPECIAL)
 			delay = false;
 
-		SCHED_WARN_ON(delay && se->sched_delayed);
+		WARN_ON_ONCE(delay && se->sched_delayed);
 
 		if (sched_feat(DELAY_DEQUEUE) && delay &&
 		    !entity_eligible(cfs_rq, se)) {
···
 		update_stats_wait_end_fair(cfs_rq, se);
 		__dequeue_entity(cfs_rq, se);
 		update_load_avg(cfs_rq, se, UPDATE_TG);
-		/*
-		 * HACK, stash a copy of deadline at the point of pick in vlag,
-		 * which isn't used until dequeue.
-		 */
-		se->vlag = se->deadline;
+
+		set_protect_slice(se);
 	}
 
 	update_stats_curr_start(cfs_rq, se);
-	SCHED_WARN_ON(cfs_rq->curr);
+	WARN_ON_ONCE(cfs_rq->curr);
 	cfs_rq->curr = se;
 
 	/*
···
 	if (sched_feat(PICK_BUDDY) &&
 	    cfs_rq->next && entity_eligible(cfs_rq, cfs_rq->next)) {
 		/* ->next will never be delayed */
-		SCHED_WARN_ON(cfs_rq->next->sched_delayed);
+		WARN_ON_ONCE(cfs_rq->next->sched_delayed);
 		return cfs_rq->next;
 	}
 
···
 		/* in !on_rq case, update occurred at dequeue */
 		update_load_avg(cfs_rq, prev, 0);
 	}
-	SCHED_WARN_ON(cfs_rq->curr != prev);
+	WARN_ON_ONCE(cfs_rq->curr != prev);
 	cfs_rq->curr = NULL;
 }
 
···
 
 			cfs_rq->throttled_clock_self = 0;
 
-			if (SCHED_WARN_ON((s64)delta < 0))
+			if (WARN_ON_ONCE((s64)delta < 0))
 				delta = 0;
 
 			cfs_rq->throttled_clock_self_time += delta;
···
 		cfs_rq->throttled_clock_pelt = rq_clock_pelt(rq);
 		list_del_leaf_cfs_rq(cfs_rq);
 
-		SCHED_WARN_ON(cfs_rq->throttled_clock_self);
+		WARN_ON_ONCE(cfs_rq->throttled_clock_self);
 		if (cfs_rq->nr_queued)
 			cfs_rq->throttled_clock_self = rq_clock(rq);
 	}
···
 	 * throttled-list.  rq->lock protects completion.
 	 */
 	cfs_rq->throttled = 1;
-	SCHED_WARN_ON(cfs_rq->throttled_clock);
+	WARN_ON_ONCE(cfs_rq->throttled_clock);
 	if (cfs_rq->nr_queued)
 		cfs_rq->throttled_clock = rq_clock(rq);
 	return true;
···
 	}
 
 	/* Already enqueued */
-	if (SCHED_WARN_ON(!list_empty(&cfs_rq->throttled_csd_list)))
+	if (WARN_ON_ONCE(!list_empty(&cfs_rq->throttled_csd_list)))
 		return;
 
 	first = list_empty(&rq->cfsb_csd_list);
···
 {
 	lockdep_assert_rq_held(rq_of(cfs_rq));
 
-	if (SCHED_WARN_ON(!cfs_rq_throttled(cfs_rq) ||
+	if (WARN_ON_ONCE(!cfs_rq_throttled(cfs_rq) ||
 	    cfs_rq->runtime_remaining <= 0))
 		return;
 
···
 			goto next;
 
 		/* By the above checks, this should never be true */
-		SCHED_WARN_ON(cfs_rq->runtime_remaining > 0);
+		WARN_ON_ONCE(cfs_rq->runtime_remaining > 0);
 
 		raw_spin_lock(&cfs_b->lock);
 		runtime = -cfs_rq->runtime_remaining + 1;
···
 				 * We currently only expect to be unthrottling
 				 * a single cfs_rq locally.
 				 */
-				SCHED_WARN_ON(!list_empty(&local_unthrottle));
+				WARN_ON_ONCE(!list_empty(&local_unthrottle));
 				list_add_tail(&cfs_rq->throttled_csd_list,
 					      &local_unthrottle);
 			}
···
 
 		rq_unlock_irqrestore(rq, &rf);
 	}
-	SCHED_WARN_ON(!list_empty(&local_unthrottle));
+	WARN_ON_ONCE(!list_empty(&local_unthrottle));
 
 	rcu_read_unlock();
 
···
 {
 	struct sched_entity *se = &p->se;
 
-	SCHED_WARN_ON(task_rq(p) != rq);
+	WARN_ON_ONCE(task_rq(p) != rq);
 
 	if (rq->cfs.h_nr_queued > 1) {
 		u64 ran = se->sum_exec_runtime - se->prev_sum_exec_runtime;
···
 	 * Because a delayed entity is one that is still on
 	 * the runqueue competing until elegibility.
 	 */
-	SCHED_WARN_ON(!se->sched_delayed);
-	SCHED_WARN_ON(!se->on_rq);
+	WARN_ON_ONCE(!se->sched_delayed);
+	WARN_ON_ONCE(!se->on_rq);
 
 	if (sched_feat(DELAY_ZERO)) {
 		update_entity_lag(cfs_rq, se);
···
 		update_cfs_group(se);
 
 		se->slice = slice;
+		if (se != cfs_rq->curr)
+			min_vruntime_cb_propagate(&se->run_node, NULL);
 		slice = cfs_rq_min_slice(cfs_rq);
 
 		cfs_rq->h_nr_runnable += h_nr_runnable;
···
 		update_cfs_group(se);
 
 		se->slice = slice;
+		if (se != cfs_rq->curr)
+			min_vruntime_cb_propagate(&se->run_node, NULL);
 		slice = cfs_rq_min_slice(cfs_rq);
 
 		cfs_rq->h_nr_runnable -= h_nr_runnable;
···
 		rq->next_balance = jiffies;
 
 	if (p && task_delayed) {
-		SCHED_WARN_ON(!task_sleep);
-		SCHED_WARN_ON(p->on_rq != 1);
+		WARN_ON_ONCE(!task_sleep);
+		WARN_ON_ONCE(p->on_rq != 1);
 
 		/* Fix-up what dequeue_task_fair() skipped */
 		hrtick_update(rq);
···
 static void set_next_buddy(struct sched_entity *se)
 {
 	for_each_sched_entity(se) {
-		if (SCHED_WARN_ON(!se->on_rq))
+		if (WARN_ON_ONCE(!se->on_rq))
 			return;
 		if (se_is_idle(se))
 			return;
···
 	 * Preempt an idle entity in favor of a non-idle entity (and don't preempt
 	 * in the inverse case).
 	 */
-	if (cse_is_idle && !pse_is_idle)
+	if (cse_is_idle && !pse_is_idle) {
+		/*
+		 * When non-idle entity preempt an idle entity,
+		 * don't give idle entity slice protection.
+		 */
+		cancel_protect_slice(se);
 		goto preempt;
+	}
+
 	if (cse_is_idle != pse_is_idle)
 		return;
 
···
 	 * Note that even if @p does not turn out to be the most eligible
 	 * task at this moment, current's slice protection will be lost.
 	 */
-	if (do_preempt_short(cfs_rq, pse, se) && se->vlag == se->deadline)
-		se->vlag = se->deadline + 1;
+	if (do_preempt_short(cfs_rq, pse, se))
+		cancel_protect_slice(se);
 
 	/*
 	 * If @p has become the most eligible task, force preemption.
···
 			return 0;
 
 		/* Prevent to re-select dst_cpu via env's CPUs: */
-		for_each_cpu_and(cpu, env->dst_grpmask, env->cpus) {
-			if (cpumask_test_cpu(cpu, p->cpus_ptr)) {
-				env->flags |= LBF_DST_PINNED;
-				env->new_dst_cpu = cpu;
-				break;
-			}
+		cpu = cpumask_first_and_and(env->dst_grpmask, env->cpus, p->cpus_ptr);
+
+		if (cpu < nr_cpu_ids) {
+			env->flags |= LBF_DST_PINNED;
+			env->new_dst_cpu = cpu;
 		}
 
 		return 0;
···
 
 void nohz_balance_exit_idle(struct rq *rq)
 {
-	SCHED_WARN_ON(rq != this_rq());
+	WARN_ON_ONCE(rq != this_rq());
 
 	if (likely(!rq->nohz_tick_stopped))
 		return;
···
 {
 	struct rq *rq = cpu_rq(cpu);
 
-	SCHED_WARN_ON(cpu != smp_processor_id());
+	WARN_ON_ONCE(cpu != smp_processor_id());
 
 	/* If this CPU is going down, then nothing needs to be done: */
 	if (!cpu_active(cpu))
···
 	int balance_cpu;
 	struct rq *rq;
 
-	SCHED_WARN_ON((flags & NOHZ_KICK_MASK) == NOHZ_BALANCE_KICK);
+	WARN_ON_ONCE((flags & NOHZ_KICK_MASK) == NOHZ_BALANCE_KICK);
 
 	/*
 	 * We assume there will be no idle load after this update and clear
···
 	struct cfs_rq *cfs_rqb;
 	s64 delta;
 
-	SCHED_WARN_ON(task_rq(b)->core != rq->core);
+	WARN_ON_ONCE(task_rq(b)->core != rq->core);
 
 #ifdef CONFIG_FAIR_GROUP_SCHED
 	/*
···
 
 static void switched_to_fair(struct rq *rq, struct task_struct *p)
 {
-	SCHED_WARN_ON(p->se.sched_delayed);
+	WARN_ON_ONCE(p->se.sched_delayed);
 
 	attach_task_cfs_rq(p);
 
···
 	if (!first)
 		return;
 
-	SCHED_WARN_ON(se->sched_delayed);
+	WARN_ON_ONCE(se->sched_delayed);
 
 	if (hrtick_enabled_fair(rq))
 		hrtick_start_fair(rq, p);
···
 #endif
 };
 
-#ifdef CONFIG_SCHED_DEBUG
 void print_cfs_stats(struct seq_file *m, int cpu)
 {
 	struct cfs_rq *cfs_rq, *pos;
···
 	rcu_read_unlock();
 }
 #endif /* CONFIG_NUMA_BALANCING */
-#endif /* CONFIG_SCHED_DEBUG */
 
 __init void init_sched_fair_class(void)
 {

kernel/sched/rt.c

···
 
 static inline struct task_struct *rt_task_of(struct sched_rt_entity *rt_se)
 {
-#ifdef CONFIG_SCHED_DEBUG
 	WARN_ON_ONCE(!rt_entity_is_task(rt_se));
-#endif
+
 	return container_of(rt_se, struct task_struct, rt);
 }
 
···
 	BUG_ON(idx >= MAX_RT_PRIO);
 
 	queue = array->queue + idx;
-	if (SCHED_WARN_ON(list_empty(queue)))
+	if (WARN_ON_ONCE(list_empty(queue)))
 		return NULL;
 	next = list_entry(queue->next, struct sched_rt_entity, run_list);
 
···
 	int ret;
 
 	mutex_lock(&mutex);
+	sched_domains_mutex_lock();
 	old_period = sysctl_sched_rt_period;
 	old_runtime = sysctl_sched_rt_runtime;
 
···
 		sysctl_sched_rt_period = old_period;
 		sysctl_sched_rt_runtime = old_runtime;
 	}
+	sched_domains_mutex_unlock();
 	mutex_unlock(&mutex);
 
 	return ret;
···
 }
 #endif /* CONFIG_SYSCTL */
 
-#ifdef CONFIG_SCHED_DEBUG
 void print_rt_stats(struct seq_file *m, int cpu)
 {
 	rt_rq_iter_t iter;
···
 		print_rt_rq(m, cpu, rt_rq);
 	rcu_read_unlock();
 }
-#endif /* CONFIG_SCHED_DEBUG */

kernel/sched/sched.h

···
 #include "cpupri.h"
 #include "cpudeadline.h"
 
-#ifdef CONFIG_SCHED_DEBUG
-# define SCHED_WARN_ON(x)      WARN_ONCE(x, #x)
-#else
-# define SCHED_WARN_ON(x)      ({ (void)(x), 0; })
-#endif
-
 /* task_struct::on_rq states: */
 #define TASK_ON_RQ_QUEUED	1
 #define TASK_ON_RQ_MIGRATING	2
···
 	 * Also, some corner cases, like 'wrap around' is dangerous, but given
 	 * that u64 is 'big enough'. So that shouldn't be a concern.
 	 */
-	u64 visit_gen;
+	u64 visit_cookie;
 
 #ifdef HAVE_RT_PUSH_IPI
 	/*
···
 
 	atomic_t		nr_iowait;
 
-#ifdef CONFIG_SCHED_DEBUG
 	u64 last_seen_need_resched_ns;
 	int ticks_without_resched;
-#endif
 
 #ifdef CONFIG_MEMBARRIER
 	int membarrier_state;
···
 
 static inline struct task_struct *task_of(struct sched_entity *se)
 {
-	SCHED_WARN_ON(!entity_is_task(se));
+	WARN_ON_ONCE(!entity_is_task(se));
 	return container_of(se, struct task_struct, se);
 }
 
···
 	 * The only reason for not seeing a clock update since the
 	 * last rq_pin_lock() is if we're currently skipping updates.
 	 */
-	SCHED_WARN_ON(rq->clock_update_flags < RQCF_ACT_SKIP);
+	WARN_ON_ONCE(rq->clock_update_flags < RQCF_ACT_SKIP);
 }
 
 static inline u64 rq_clock(struct rq *rq)
···
 static inline void rq_clock_start_loop_update(struct rq *rq)
 {
 	lockdep_assert_rq_held(rq);
-	SCHED_WARN_ON(rq->clock_update_flags & RQCF_ACT_SKIP);
+	WARN_ON_ONCE(rq->clock_update_flags & RQCF_ACT_SKIP);
 	rq->clock_update_flags |= RQCF_ACT_SKIP;
 }
 
···
 struct rq_flags {
 	unsigned long flags;
 	struct pin_cookie cookie;
-#ifdef CONFIG_SCHED_DEBUG
 	/*
 	 * A copy of (rq::clock_update_flags & RQCF_UPDATED) for the
 	 * current pin context is stashed here in case it needs to be
 	 * restored in rq_repin_lock().
 	 */
 	unsigned int clock_update_flags;
-#endif
 };
 
 extern struct balance_callback balance_push_callback;
···
 {
 	rf->cookie = lockdep_pin_lock(__rq_lockp(rq));
 
-#ifdef CONFIG_SCHED_DEBUG
 	rq->clock_update_flags &= (RQCF_REQ_SKIP|RQCF_ACT_SKIP);
 	rf->clock_update_flags = 0;
-# ifdef CONFIG_SMP
-	SCHED_WARN_ON(rq->balance_callback && rq->balance_callback != &balance_push_callback);
-# endif
+#ifdef CONFIG_SMP
+	WARN_ON_ONCE(rq->balance_callback && rq->balance_callback != &balance_push_callback);
 #endif
 }
 
 static inline void rq_unpin_lock(struct rq *rq, struct rq_flags *rf)
 {
-#ifdef CONFIG_SCHED_DEBUG
 	if (rq->clock_update_flags > RQCF_ACT_SKIP)
 		rf->clock_update_flags = RQCF_UPDATED;
-#endif
+
 	scx_rq_clock_invalidate(rq);
 	lockdep_unpin_lock(__rq_lockp(rq), rf->cookie);
 }
···
 {
 	lockdep_repin_lock(__rq_lockp(rq), rf->cookie);
 
-#ifdef CONFIG_SCHED_DEBUG
 	/*
 	 * Restore the value we stashed in @rf for this pin context.
 	 */
 	rq->clock_update_flags |= rf->clock_update_flags;
-#endif
 }
 
 extern
···
 	unsigned long		next_update;
 	int			imbalance;		/* XXX unrelated to capacity but shared group state */
 
-#ifdef CONFIG_SCHED_DEBUG
 	int			id;
-#endif
 
 	unsigned long		cpumask[];		/* Balance mask */
 };
···
 
 extern int group_balance_cpu(struct sched_group *sg);
 
-#ifdef CONFIG_SCHED_DEBUG
 extern void update_sched_domain_debugfs(void);
 extern void dirty_sched_domain_sysctl(int cpu);
-#else
-static inline void update_sched_domain_debugfs(void) { }
-static inline void dirty_sched_domain_sysctl(int cpu) { }
-#endif
 
 extern int sched_update_scaling(void);
 
···
 }
 
 /*
- * Tunables that become constants when CONFIG_SCHED_DEBUG is off:
+ * Tunables:
  */
-#ifdef CONFIG_SCHED_DEBUG
-# define const_debug __read_mostly
-#else
-# define const_debug const
-#endif
 
 #define SCHED_FEAT(name, enabled)	\
 	__SCHED_FEAT_##name ,
···
 
 #undef SCHED_FEAT
 
-#ifdef CONFIG_SCHED_DEBUG
-
 /*
  * To support run-time toggling of sched features, all the translation units
  * (but core.c) reference the sysctl_sched_features defined in core.c.
  */
-extern const_debug unsigned int sysctl_sched_features;
+extern __read_mostly unsigned int sysctl_sched_features;
 
 #ifdef CONFIG_JUMP_LABEL
 
···
 #define sched_feat(x) (sysctl_sched_features & (1UL << __SCHED_FEAT_##x))
 
 #endif /* !CONFIG_JUMP_LABEL */
-
-#else /* !SCHED_DEBUG: */
-
-/*
- * Each translation unit has its own copy of sysctl_sched_features to allow
- * constants propagation at compile time and compiler optimization based on
- * features default.
- */
-#define SCHED_FEAT(name, enabled)	\
-	(1UL << __SCHED_FEAT_##name) * enabled |
-static const_debug __maybe_unused unsigned int sysctl_sched_features =
-#include "features.h"
-	0;
-#undef SCHED_FEAT
-
-#define sched_feat(x) !!(sysctl_sched_features & (1UL << __SCHED_FEAT_##x))
-
-#endif /* !SCHED_DEBUG */
 
 extern struct static_key_false sched_numa_balancing;
 extern struct static_key_false sched_schedstats;
···
 
 static inline struct cpuidle_state *idle_get_state(struct rq *rq)
 {
-	SCHED_WARN_ON(!rcu_read_lock_held());
+	WARN_ON_ONCE(!rcu_read_lock_held());
 
 	return rq->idle_state;
 }
···
 # define SCHED_NR_MIGRATE_BREAK 32
 #endif
 
-extern const_debug unsigned int sysctl_sched_nr_migrate;
-extern const_debug unsigned int sysctl_sched_migration_cost;
+extern __read_mostly unsigned int sysctl_sched_nr_migrate;
+extern __read_mostly unsigned int sysctl_sched_migration_cost;
 
 extern unsigned int sysctl_sched_base_slice;
 
-#ifdef CONFIG_SCHED_DEBUG
 extern int sysctl_resched_latency_warn_ms;
 extern int sysctl_resched_latency_warn_once;
 
···
 extern unsigned int sysctl_numa_balancing_scan_period_max;
 extern unsigned int sysctl_numa_balancing_scan_size;
 extern unsigned int sysctl_numa_balancing_hot_threshold;
-#endif
 
 #ifdef CONFIG_SCHED_HRTICK
 
···
 }
 #endif
 
-#ifdef CONFIG_SCHED_DEBUG
 /*
  * In double_lock_balance()/double_rq_lock(), we use raw_spin_rq_lock() to
  * acquire rq lock instead of rq_lock(). So at the end of these two functions
···
 	rq2->clock_update_flags &= (RQCF_REQ_SKIP|RQCF_ACT_SKIP);
 #endif
 }
-#else
-static inline void double_rq_clock_clear_update(struct rq *rq1, struct rq *rq2) { }
-#endif
 
 #define DEFINE_LOCK_GUARD_2(name, type, _lock, _unlock, ...)				\
 __DEFINE_UNLOCK_GUARD(name, type, _unlock, type *lock2; __VA_ARGS__)			\
···
 extern struct sched_entity *__pick_first_entity(struct cfs_rq *cfs_rq);
 extern struct sched_entity *__pick_last_entity(struct cfs_rq *cfs_rq);
 
-#ifdef	CONFIG_SCHED_DEBUG
 extern bool sched_debug_verbose;
 
 extern void print_cfs_stats(struct seq_file *m, int cpu);
···
 extern void print_dl_rq(struct seq_file *m, int cpu, struct dl_rq *dl_rq);
 
 extern void resched_latency_warn(int cpu, u64 latency);
-# ifdef CONFIG_NUMA_BALANCING
+#ifdef CONFIG_NUMA_BALANCING
 extern void show_numa_stats(struct task_struct *p, struct seq_file *m);
 extern void
 print_numa_stats(struct seq_file *m, int node, unsigned long tsf,
 		 unsigned long tpf, unsigned long gsf, unsigned long gpf);
-# endif /* CONFIG_NUMA_BALANCING */
-#else /* !CONFIG_SCHED_DEBUG: */
-static inline void resched_latency_warn(int cpu, u64 latency) { }
-#endif /* !CONFIG_SCHED_DEBUG */
+#endif /* CONFIG_NUMA_BALANCING */
 
 extern void init_cfs_rq(struct cfs_rq *cfs_rq);
 extern void init_rt_rq(struct rt_rq *rt_rq);
···
 
 unsigned long uclamp_eff_value(struct task_struct *p, enum uclamp_id clamp_id);
 
+/*
+ * When uclamp is compiled in, the aggregation at rq level is 'turned off'
+ * by default in the fast path and only gets turned on once userspace performs
+ * an operation that requires it.
+ *
+ * Returns true if userspace opted-in to use uclamp and aggregation at rq level
+ * hence is active.
+ */
+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)
 {
···
 	unsigned long rq_util;
 	unsigned long max_util;
 
-	if (!static_branch_likely(&sched_uclamp_used))
+	if (!uclamp_is_used())
 		return false;
 
 	rq_util = cpu_util_cfs(cpu_of(rq)) + cpu_util_rt(rq);
 	max_util = READ_ONCE(rq->uclamp[UCLAMP_MAX].value);
 
 	return max_util != SCHED_CAPACITY_SCALE && rq_util >= max_util;
-}
-
-/*
- * When uclamp is compiled in, the aggregation at rq level is 'turned off'
- * by default in the fast path and only gets turned on once userspace performs
- * an operation that requires it.
- *
- * Returns true if userspace opted-in to use uclamp and aggregation at rq level
- * hence is active.
- */
-static inline bool uclamp_is_used(void)
-{
-	return static_branch_likely(&sched_uclamp_used);
 }
 
 #define for_each_clamp_id(clamp_id) \
···
 {
 	return false;
 }
+
+static inline void sched_uclamp_enable(void) {}
 
 static inline unsigned long
 uclamp_rq_get(struct rq *rq, enum uclamp_id clamp_id)
···
 extern int sched_dynamic_mode(const char *str);
 extern void sched_dynamic_update(int mode);
 #endif
+extern const char *preempt_modes[];
 
 #ifdef CONFIG_SCHED_MM_CID
 

kernel/sched/stats.h

···
 
 	if (p->se.sched_delayed) {
 		/* CPU migration of "sleeping" task */
-		SCHED_WARN_ON(!(flags & ENQUEUE_MIGRATED));
+		WARN_ON_ONCE(!(flags & ENQUEUE_MIGRATED));
 		if (p->in_memstall)
 			set |= TSK_MEMSTALL;
 		if (p->in_iowait)

kernel/sched/syscalls.c

···
 	 * blocking operation which obviously cannot be done while holding
 	 * scheduler locks.
 	 */
-	static_branch_enable(&sched_uclamp_used);
+	sched_uclamp_enable();
 
 	return 0;
 }
···
 {
 	struct sched_param lparam;
 
-	if (!param || pid < 0)
+	if (unlikely(!param || pid < 0))
 		return -EINVAL;
 	if (copy_from_user(&lparam, param, sizeof(struct sched_param)))
 		return -EFAULT;
···
 	struct sched_attr attr;
 	int retval;
 
-	if (!uattr || pid < 0 || flags)
+	if (unlikely(!uattr || pid < 0 || flags))
 		return -EINVAL;
 
 	retval = sched_copy_attr(uattr, &attr);
···
 	struct task_struct *p;
 	int retval;
 
-	if (!param || pid < 0)
+	if (unlikely(!param || pid < 0))
 		return -EINVAL;
 
 	scoped_guard (rcu) {
···
 	struct task_struct *p;
 	int retval;
 
-	if (!uattr || pid < 0 || usize > PAGE_SIZE ||
-	    usize < SCHED_ATTR_SIZE_VER0 || flags)
+	if (unlikely(!uattr || pid < 0 || usize > PAGE_SIZE ||
+		      usize < SCHED_ATTR_SIZE_VER0 || flags))
 		return -EINVAL;
 
 	scoped_guard (rcu) {

kernel/sched/topology.c

···
 #include <linux/bsearch.h>
 
 DEFINE_MUTEX(sched_domains_mutex);
+void sched_domains_mutex_lock(void)
+{
+	mutex_lock(&sched_domains_mutex);
+}
+void sched_domains_mutex_unlock(void)
+{
+	mutex_unlock(&sched_domains_mutex);
+}
 
 /* Protected by sched_domains_mutex: */
 static cpumask_var_t sched_domains_tmpmask;
 static cpumask_var_t sched_domains_tmpmask2;
-
-#ifdef CONFIG_SCHED_DEBUG
 
 static int __init sched_debug_setup(char *str)
 {
···
 			break;
 	}
 }
-#else /* !CONFIG_SCHED_DEBUG */
-
-# define sched_debug_verbose 0
-# define sched_domain_debug(sd, cpu) do { } while (0)
-static inline bool sched_debug(void)
-{
-	return false;
-}
-#endif /* CONFIG_SCHED_DEBUG */
 
 /* Generate a mask of SD flags with the SDF_NEEDS_GROUPS metaflag */
 #define SD_FLAG(name, mflags) (name * !!((mflags) & SDF_NEEDS_GROUPS)) |
···
 	rd->rto_push_work = IRQ_WORK_INIT_HARD(rto_push_irq_work_func);
 #endif
 
-	rd->visit_gen = 0;
+	rd->visit_cookie = 0;
 	init_dl_bw(&rd->dl_bw);
 	if (cpudl_init(&rd->cpudl) != 0)
 		goto free_rto_mask;
···
 			if (!sgc)
 				return -ENOMEM;
 
-#ifdef CONFIG_SCHED_DEBUG
 			sgc->id = j;
-#endif
 
 			*per_cpu_ptr(sdd->sgc, j) = sgc;
 		}
···
  *
  * Call with hotplug lock and sched_domains_mutex held
  */
-void partition_sched_domains_locked(int ndoms_new, cpumask_var_t doms_new[],
+static void partition_sched_domains_locked(int ndoms_new, cpumask_var_t doms_new[],
 				    struct sched_domain_attr *dattr_new)
 {
 	bool __maybe_unused has_eas = false;
···
 	for (i = 0; i < ndoms_cur; i++) {
 		for (j = 0; j < n && !new_topology; j++) {
 			if (cpumask_equal(doms_cur[i], doms_new[j]) &&
-			    dattrs_equal(dattr_cur, i, dattr_new, j)) {
-				struct root_domain *rd;
-
-				/*
-				 * This domain won't be destroyed and as such
-				 * its dl_bw->total_bw needs to be cleared.
-				 * Tasks contribution will be then recomputed
-				 * in function dl_update_tasks_root_domain(),
-				 * dl_servers contribution in function
-				 * dl_restore_server_root_domain().
-				 */
-				rd = cpu_rq(cpumask_any(doms_cur[i]))->rd;
-				dl_clear_root_domain(rd);
+			    dattrs_equal(dattr_cur, i, dattr_new, j))
 				goto match1;
-			}
 		}
 		/* No match - a current sched domain not in new doms_new[] */
 		detach_destroy_domains(doms_cur[i]);
···
 	ndoms_cur = ndoms_new;
 
 	update_sched_domain_debugfs();
+	dl_rebuild_rd_accounting();
 }
 
 /*
···
 void partition_sched_domains(int ndoms_new, cpumask_var_t doms_new[],
 			     struct sched_domain_attr *dattr_new)
 {
-	mutex_lock(&sched_domains_mutex);
+	sched_domains_mutex_lock();
 	partition_sched_domains_locked(ndoms_new, doms_new, dattr_new);
-	mutex_unlock(&sched_domains_mutex);
+	sched_domains_mutex_unlock();
 }

kernel/trace/trace.c

···
 		   entries,
 		   total,
 		   buf->cpu,
-		   preempt_model_none()      ? "server" :
-		   preempt_model_voluntary() ? "desktop" :
-		   preempt_model_full()      ? "preempt" :
-		   preempt_model_lazy()	     ? "lazy"    :
-		   preempt_model_rt()        ? "preempt_rt" :
-		   "unknown",
+		   preempt_model_str(),
 		   /* These are reserved for later use */
 		   0, 0, 0, 0);
 #ifdef CONFIG_SMP

lib/Kconfig.debug

···
 
 menu "Scheduler Debugging"
 
-config SCHED_DEBUG
-	bool "Collect scheduler debugging info"
-	depends on DEBUG_KERNEL && DEBUG_FS
-	default y
-	help
-	  If you say Y here, the /sys/kernel/debug/sched file will be provided
-	  that can help debug the scheduler. The runtime overhead of this
-	  option is minimal.
-
 config SCHED_INFO
 	bool
 	default n

lib/dump_stack.c

···
  */
 void dump_stack_print_info(const char *log_lvl)
 {
-	printk("%sCPU: %d UID: %u PID: %d Comm: %.20s %s%s %s %.*s" BUILD_ID_FMT "\n",
+	printk("%sCPU: %d UID: %u PID: %d Comm: %.20s %s%s %s %.*s %s " BUILD_ID_FMT "\n",
 	       log_lvl, raw_smp_processor_id(),
 	       __kuid_val(current_real_cred()->euid),
 	       current->pid, current->comm,
···
 	       print_tainted(),
 	       init_utsname()->release,
 	       (int)strcspn(init_utsname()->version, " "),
-	       init_utsname()->version, BUILD_ID_VAL);
+	       init_utsname()->version, preempt_model_str(), BUILD_ID_VAL);
 
 	if (get_taint())
 		printk("%s%s\n", log_lvl, print_tainted_verbose());

tools/testing/selftests/rseq/.gitignore

···
 param_test_mm_cid
 param_test_mm_cid_benchmark
 param_test_mm_cid_compare_twice
+syscall_errors_test

tools/testing/selftests/rseq/Makefile

···
 
 TEST_GEN_PROGS = basic_test basic_percpu_ops_test basic_percpu_ops_mm_cid_test param_test \
 		param_test_benchmark param_test_compare_twice param_test_mm_cid \
-		param_test_mm_cid_benchmark param_test_mm_cid_compare_twice
+		param_test_mm_cid_benchmark param_test_mm_cid_compare_twice \
+		syscall_errors_test
 
 TEST_GEN_PROGS_EXTENDED = librseq.so
 
-TEST_PROGS = run_param_test.sh
+TEST_PROGS = run_param_test.sh run_syscall_errors_test.sh
 
 TEST_FILES := settings
 
···
 $(OUTPUT)/param_test_mm_cid_compare_twice: param_test.c $(TEST_GEN_PROGS_EXTENDED) \
 					rseq.h rseq-*.h
 	$(CC) $(CFLAGS) -DBUILDOPT_RSEQ_PERCPU_MM_CID -DRSEQ_COMPARE_TWICE $< $(LDLIBS) -lrseq -o $@
+
+$(OUTPUT)/syscall_errors_test: syscall_errors_test.c $(TEST_GEN_PROGS_EXTENDED) \
+					rseq.h rseq-*.h
+	$(CC) $(CFLAGS) $< $(LDLIBS) -lrseq -o $@

tools/testing/selftests/rseq/rseq.c

···
 /* Original struct rseq allocation size is 32 bytes. */
 #define ORIG_RSEQ_ALLOC_SIZE		32
 
+/*
+ * Use a union to ensure we allocate a TLS area of 1024 bytes to accomodate an
+ * rseq registration that is larger than the current rseq ABI.
+ */
+union rseq_tls {
+	struct rseq_abi abi;
+	char dummy[RSEQ_THREAD_AREA_ALLOC_SIZE];
+};
+
 static
-__thread struct rseq_abi __rseq_abi __attribute__((tls_model("initial-exec"), aligned(RSEQ_THREAD_AREA_ALLOC_SIZE))) = {
-	.cpu_id = RSEQ_ABI_CPU_ID_UNINITIALIZED,
+__thread union rseq_tls __rseq __attribute__((tls_model("initial-exec"))) = {
+	.abi = {
+		.cpu_id = RSEQ_ABI_CPU_ID_UNINITIALIZED,
+	},
 };
 
 static int sys_rseq(struct rseq_abi *rseq_abi, uint32_t rseq_len,
···
 	return syscall(__NR_getcpu, cpu, node, NULL);
 }
 
-int rseq_available(void)
+bool rseq_available(void)
 {
 	int rc;
 
···
 		abort();
 	switch (errno) {
 	case ENOSYS:
-		return 0;
+		return false;
 	case EINVAL:
-		return 1;
+		return true;
 	default:
 		abort();
 	}
···
 		/* Treat libc's ownership as a successful registration. */
 		return 0;
 	}
-	rc = sys_rseq(&__rseq_abi, get_rseq_min_alloc_size(), 0, RSEQ_SIG);
+	rc = sys_rseq(&__rseq.abi, get_rseq_min_alloc_size(), 0, RSEQ_SIG);
 	if (rc) {
 		/*
 		 * After at least one thread has registered successfully
···
 		/* Treat libc's ownership as a successful unregistration. */
 		return 0;
 	}
-	rc = sys_rseq(&__rseq_abi, get_rseq_min_alloc_size(), RSEQ_ABI_FLAG_UNREGISTER, RSEQ_SIG);
+	rc = sys_rseq(&__rseq.abi, get_rseq_min_alloc_size(), RSEQ_ABI_FLAG_UNREGISTER, RSEQ_SIG);
 	if (rc)
 		return -1;
 	return 0;
···
 	rseq_ownership = 1;
 
 	/* Calculate the offset of the rseq area from the thread pointer. */
-	rseq_offset = (void *)&__rseq_abi - rseq_thread_pointer();
+	rseq_offset = (void *)&__rseq.abi - rseq_thread_pointer();
 
 	/* rseq flags are deprecated, always set to 0. */
 	rseq_flags = 0;

tools/testing/selftests/rseq/rseq.h

···
 int32_t rseq_fallback_current_node(void);
 
 /*
+ * Returns true if rseq is supported.
+ */
+bool rseq_available(void);
+
+/*
  * Values returned can be either the current CPU number, -1 (rseq is
  * uninitialized), or -2 (rseq initialization has failed).
  */

tools/testing/selftests/rseq/run_syscall_errors_test.sh

···
+#!/bin/bash
+# SPDX-License-Identifier: MIT
+# SPDX-FileCopyrightText: 2024 Michael Jeanson <[e-mail]>
+
+GLIBC_TUNABLES="${GLIBC_TUNABLES:-}:glibc.pthread.rseq=0" ./syscall_errors_test

tools/testing/selftests/rseq/syscall_errors_test.c

···
+// SPDX-License-Identifier: MIT
+// SPDX-FileCopyrightText: 2024 Michael Jeanson <[e-mail]>
+
+#ifndef _GNU_SOURCE
+#define _GNU_SOURCE
+#endif
+
+#include <assert.h>
+#include <stdint.h>
+#include <syscall.h>
+#include <string.h>
+#include <unistd.h>
+
+#include "rseq.h"
+
+static int sys_rseq(void *rseq_abi, uint32_t rseq_len,
+		    int flags, uint32_t sig)
+{
+	return syscall(__NR_rseq, rseq_abi, rseq_len, flags, sig);
+}
+
+/*
+ * Check the value of errno on some expected failures of the rseq syscall.
+ */
+
+int main(void)
+{
+	struct rseq_abi *global_rseq = rseq_get_abi();
+	int ret;
+	int errno_copy;
+
+	if (!rseq_available()) {
+		fprintf(stderr, "rseq syscall unavailable");
+		goto error;
+	}
+
+	/* The current thread is NOT registered. */
+
+	/* EINVAL */
+	errno = 0;
+	ret = sys_rseq(global_rseq, 32, -1, RSEQ_SIG);
+	errno_copy = errno;
+	fprintf(stderr, "Registration with invalid flag fails with errno set to EINVAL (ret = %d, errno = %s)\n", ret, strerrorname_np(errno_copy));
+	if (ret == 0 || errno_copy != EINVAL)
+		goto error;
+
+	errno = 0;
+	ret = sys_rseq((char *) global_rseq + 1, 32, 0, RSEQ_SIG);
+	errno_copy = errno;
+	fprintf(stderr, "Registration with unaligned rseq_abi fails with errno set to EINVAL (ret = %d, errno = %s)\n", ret, strerrorname_np(errno_copy));
+	if (ret == 0 || errno_copy != EINVAL)
+		goto error;
+
+	errno = 0;
+	ret = sys_rseq(global_rseq, 31, 0, RSEQ_SIG);
+	errno_copy = errno;
+	fprintf(stderr, "Registration with invalid size fails with errno set to EINVAL (ret = %d, errno = %s)\n", ret, strerrorname_np(errno_copy));
+	if (ret == 0 || errno_copy != EINVAL)
+		goto error;
+
+
+#if defined(__LP64__) && (!defined(__s390__) && !defined(__s390x__))
+	/*
+	 * We haven't found a reliable way to find an invalid address when
+	 * running a 32bit userspace on a 64bit kernel, so only run this test
+	 * on 64bit builds for the moment.
+	 *
+	 * Also exclude architectures that select
+	 * CONFIG_ALTERNATE_USER_ADDRESS_SPACE where the kernel and userspace
+	 * have their own address space and this failure can't happen.
+	 */
+
+	/* EFAULT */
+	errno = 0;
+	ret = sys_rseq((void *) -4096UL, 32, 0, RSEQ_SIG);
+	errno_copy = errno;
+	fprintf(stderr, "Registration with invalid address fails with errno set to EFAULT (ret = %d, errno = %s)\n", ret, strerrorname_np(errno_copy));
+	if (ret == 0 || errno_copy != EFAULT)
+		goto error;
+#endif
+
+	errno = 0;
+	ret = sys_rseq(global_rseq, 32, 0, RSEQ_SIG);
+	errno_copy = errno;
+	fprintf(stderr, "Registration succeeds for the current thread (ret = %d, errno = %s)\n", ret, strerrorname_np(errno_copy));
+	if (ret != 0 && errno != 0)
+		goto error;
+
+	/* The current thread is registered. */
+
+	/* EBUSY */
+	errno = 0;
+	ret = sys_rseq(global_rseq, 32, 0, RSEQ_SIG);
+	errno_copy = errno;
+	fprintf(stderr, "Double registration fails with errno set to EBUSY (ret = %d, errno = %s)\n", ret, strerrorname_np(errno_copy));
+	if (ret == 0 || errno_copy != EBUSY)
+		goto error;
+
+	/* EPERM */
+	errno = 0;
+	ret = sys_rseq(global_rseq, 32, RSEQ_ABI_FLAG_UNREGISTER, RSEQ_SIG + 1);
+	errno_copy = errno;
+	fprintf(stderr, "Unregistration with wrong RSEQ_SIG fails with errno to EPERM (ret = %d, errno = %s)\n", ret, strerrorname_np(errno_copy));
+	if (ret == 0 || errno_copy != EPERM)
+		goto error;
+
+	errno = 0;
+	ret = sys_rseq(global_rseq, 32, RSEQ_ABI_FLAG_UNREGISTER, RSEQ_SIG);
+	errno_copy = errno;
+	fprintf(stderr, "Unregistration succeeds for the current thread (ret = %d, errno = %s)\n", ret, strerrorname_np(errno_copy));
+	if (ret != 0)
+		goto error;
+
+	errno = 0;
+	ret = sys_rseq(global_rseq, 32, RSEQ_ABI_FLAG_UNREGISTER, RSEQ_SIG);
+	errno_copy = errno;
+	fprintf(stderr, "Double unregistration fails with errno set to EINVAL (ret = %d, errno = %s)\n", ret, strerrorname_np(errno_copy));
+	if (ret == 0 || errno_copy != EINVAL)
+		goto error;
+
+	return 0;
+error:
+	return -1;
+}

tools/testing/selftests/sched/config

···
-CONFIG_SCHED_DEBUG=y
+# empty

tools/testing/selftests/sched_ext/config

···
-CONFIG_SCHED_DEBUG=y
 CONFIG_SCHED_CLASS_EXT=y
 CONFIG_CGROUPS=y
 CONFIG_CGROUP_SCHED=y

tools/testing/selftests/wireguard/qemu/debug.config

···
 CONFIG_DEBUG_STACK_USAGE=y
 CONFIG_DEBUG_SHIRQ=y
 CONFIG_WQ_WATCHDOG=y
-CONFIG_SCHED_DEBUG=y
 CONFIG_SCHED_INFO=y
 CONFIG_SCHEDSTATS=y
 CONFIG_SCHED_STACK_END_CHECK=y