diff options
Diffstat (limited to 'kernel/sched')
| -rw-r--r-- | kernel/sched/build_policy.c | 1 | ||||
| -rw-r--r-- | kernel/sched/core.c | 76 | ||||
| -rw-r--r-- | kernel/sched/deadline.c | 124 | ||||
| -rw-r--r-- | kernel/sched/debug.c | 6 | ||||
| -rw-r--r-- | kernel/sched/ext.c | 1566 | ||||
| -rw-r--r-- | kernel/sched/ext.h | 25 | ||||
| -rw-r--r-- | kernel/sched/ext_idle.c | 174 | ||||
| -rw-r--r-- | kernel/sched/ext_internal.h | 1078 | ||||
| -rw-r--r-- | kernel/sched/fair.c | 498 | ||||
| -rw-r--r-- | kernel/sched/pelt.h | 4 | ||||
| -rw-r--r-- | kernel/sched/rq-offsets.c | 12 | ||||
| -rw-r--r-- | kernel/sched/sched.h | 48 | ||||
| -rw-r--r-- | kernel/sched/topology.c | 75 |
13 files changed, 2030 insertions, 1657 deletions
diff --git a/kernel/sched/build_policy.c b/kernel/sched/build_policy.c index c4a488e67aa7..755883faf751 100644 --- a/kernel/sched/build_policy.c +++ b/kernel/sched/build_policy.c @@ -58,6 +58,7 @@ #include "deadline.c" #ifdef CONFIG_SCHED_CLASS_EXT +# include "ext_internal.h" # include "ext.c" # include "ext_idle.c" #endif diff --git a/kernel/sched/core.c b/kernel/sched/core.c index be00629f0ba4..198d2dd45f59 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -7,6 +7,8 @@ * Copyright (C) 1991-2002 Linus Torvalds * Copyright (C) 1998-2024 Ingo Molnar, Red Hat */ +#define INSTANTIATE_EXPORTED_MIGRATE_DISABLE +#include <linux/sched.h> #include <linux/highmem.h> #include <linux/hrtimer_api.h> #include <linux/ktime_api.h> @@ -917,7 +919,7 @@ void hrtick_start(struct rq *rq, u64 delay) * doesn't make sense and can cause timer DoS. */ delta = max_t(s64, delay, 10000LL); - rq->hrtick_time = ktime_add_ns(timer->base->get_time(), delta); + rq->hrtick_time = ktime_add_ns(hrtimer_cb_get_time(timer), delta); if (rq == this_rq()) __hrtick_restart(rq); @@ -2381,28 +2383,7 @@ static void migrate_disable_switch(struct rq *rq, struct task_struct *p) __do_set_cpus_allowed(p, &ac); } -void migrate_disable(void) -{ - struct task_struct *p = current; - - if (p->migration_disabled) { -#ifdef CONFIG_DEBUG_PREEMPT - /* - *Warn about overflow half-way through the range. - */ - WARN_ON_ONCE((s16)p->migration_disabled < 0); -#endif - p->migration_disabled++; - return; - } - - guard(preempt)(); - this_rq()->nr_pinned++; - p->migration_disabled = 1; -} -EXPORT_SYMBOL_GPL(migrate_disable); - -void migrate_enable(void) +void ___migrate_enable(void) { struct task_struct *p = current; struct affinity_context ac = { @@ -2410,35 +2391,19 @@ void migrate_enable(void) .flags = SCA_MIGRATE_ENABLE, }; -#ifdef CONFIG_DEBUG_PREEMPT - /* - * Check both overflow from migrate_disable() and superfluous - * migrate_enable(). - */ - if (WARN_ON_ONCE((s16)p->migration_disabled <= 0)) - return; -#endif + __set_cpus_allowed_ptr(p, &ac); +} +EXPORT_SYMBOL_GPL(___migrate_enable); - if (p->migration_disabled > 1) { - p->migration_disabled--; - return; - } +void migrate_disable(void) +{ + __migrate_disable(); +} +EXPORT_SYMBOL_GPL(migrate_disable); - /* - * Ensure stop_task runs either before or after this, and that - * __set_cpus_allowed_ptr(SCA_MIGRATE_ENABLE) doesn't schedule(). - */ - guard(preempt)(); - if (p->cpus_ptr != &p->cpus_mask) - __set_cpus_allowed_ptr(p, &ac); - /* - * Mustn't clear migration_disabled() until cpus_ptr points back at the - * regular cpus_mask, otherwise things that race (eg. - * select_fallback_rq) get confused. - */ - barrier(); - p->migration_disabled = 0; - this_rq()->nr_pinned--; +void migrate_enable(void) +{ + __migrate_enable(); } EXPORT_SYMBOL_GPL(migrate_enable); @@ -4472,7 +4437,7 @@ int wake_up_state(struct task_struct *p, unsigned int state) * __sched_fork() is basic setup which is also used by sched_init() to * initialize the boot CPU's idle task. */ -static void __sched_fork(unsigned long clone_flags, struct task_struct *p) +static void __sched_fork(u64 clone_flags, struct task_struct *p) { p->on_rq = 0; @@ -4490,6 +4455,9 @@ static void __sched_fork(unsigned long clone_flags, struct task_struct *p) #ifdef CONFIG_FAIR_GROUP_SCHED p->se.cfs_rq = NULL; +#ifdef CONFIG_CFS_BANDWIDTH + init_cfs_throttle_work(p); +#endif #endif #ifdef CONFIG_SCHEDSTATS @@ -4707,7 +4675,7 @@ late_initcall(sched_core_sysctl_init); /* * fork()/clone()-time setup: */ -int sched_fork(unsigned long clone_flags, struct task_struct *p) +int sched_fork(u64 clone_flags, struct task_struct *p) { __sched_fork(clone_flags, p); /* @@ -9362,8 +9330,6 @@ static void cpu_cgroup_attach(struct cgroup_taskset *tset) cgroup_taskset_for_each(task, css, tset) sched_move_task(task, false); - - scx_cgroup_finish_attach(); } static void cpu_cgroup_cancel_attach(struct cgroup_taskset *tset) @@ -9551,7 +9517,7 @@ static unsigned long tg_weight(struct task_group *tg) #ifdef CONFIG_FAIR_GROUP_SCHED return scale_load_down(tg->shares); #else - return sched_weight_from_cgroup(tg->scx_weight); + return sched_weight_from_cgroup(tg->scx.weight); #endif } diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c index e2d51f4306b3..615411a0a881 100644 --- a/kernel/sched/deadline.c +++ b/kernel/sched/deadline.c @@ -875,7 +875,7 @@ static void replenish_dl_entity(struct sched_dl_entity *dl_se) */ if (dl_se->dl_defer && !dl_se->dl_defer_running && dl_time_before(rq_clock(dl_se->rq), dl_se->deadline - dl_se->runtime)) { - if (!is_dl_boosted(dl_se) && dl_se->server_has_tasks(dl_se)) { + if (!is_dl_boosted(dl_se)) { /* * Set dl_se->dl_defer_armed and dl_throttled variables to @@ -1152,8 +1152,6 @@ static void __push_dl_task(struct rq *rq, struct rq_flags *rf) /* a defer timer will not be reset if the runtime consumed was < dl_server_min_res */ static const u64 dl_server_min_res = 1 * NSEC_PER_MSEC; -static bool dl_server_stopped(struct sched_dl_entity *dl_se); - static enum hrtimer_restart dl_server_timer(struct hrtimer *timer, struct sched_dl_entity *dl_se) { struct rq *rq = rq_of_dl_se(dl_se); @@ -1171,12 +1169,6 @@ static enum hrtimer_restart dl_server_timer(struct hrtimer *timer, struct sched_ if (!dl_se->dl_runtime) return HRTIMER_NORESTART; - if (!dl_se->server_has_tasks(dl_se)) { - replenish_dl_entity(dl_se); - dl_server_stopped(dl_se); - return HRTIMER_NORESTART; - } - if (dl_se->dl_defer_armed) { /* * First check if the server could consume runtime in background. @@ -1496,10 +1488,12 @@ throttle: } if (unlikely(is_dl_boosted(dl_se) || !start_dl_timer(dl_se))) { - if (dl_server(dl_se)) - enqueue_dl_entity(dl_se, ENQUEUE_REPLENISH); - else + if (dl_server(dl_se)) { + replenish_dl_new_period(dl_se, rq); + start_dl_timer(dl_se); + } else { enqueue_task_dl(rq, dl_task_of(dl_se), ENQUEUE_REPLENISH); + } } if (!is_leftmost(dl_se, &rq->dl)) @@ -1577,10 +1571,8 @@ void dl_server_update_idle_time(struct rq *rq, struct task_struct *p) void dl_server_update(struct sched_dl_entity *dl_se, s64 delta_exec) { /* 0 runtime = fair server disabled */ - if (dl_se->dl_runtime) { - dl_se->dl_server_idle = 0; + if (dl_se->dl_runtime) update_curr_dl_se(dl_se->rq, dl_se, delta_exec); - } } void dl_server_start(struct sched_dl_entity *dl_se) @@ -1608,26 +1600,10 @@ void dl_server_stop(struct sched_dl_entity *dl_se) dl_se->dl_server_active = 0; } -static bool dl_server_stopped(struct sched_dl_entity *dl_se) -{ - if (!dl_se->dl_server_active) - return false; - - if (dl_se->dl_server_idle) { - dl_server_stop(dl_se); - return true; - } - - dl_se->dl_server_idle = 1; - return false; -} - void dl_server_init(struct sched_dl_entity *dl_se, struct rq *rq, - dl_server_has_tasks_f has_tasks, dl_server_pick_f pick_task) { dl_se->rq = rq; - dl_se->server_has_tasks = has_tasks; dl_se->server_pick_task = pick_task; } @@ -1849,7 +1825,9 @@ void inc_dl_tasks(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq) u64 deadline = dl_se->deadline; dl_rq->dl_nr_running++; - add_nr_running(rq_of_dl_rq(dl_rq), 1); + + if (!dl_server(dl_se)) + add_nr_running(rq_of_dl_rq(dl_rq), 1); inc_dl_deadline(dl_rq, deadline); } @@ -1859,7 +1837,9 @@ void dec_dl_tasks(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq) { WARN_ON(!dl_rq->dl_nr_running); dl_rq->dl_nr_running--; - sub_nr_running(rq_of_dl_rq(dl_rq), 1); + + if (!dl_server(dl_se)) + sub_nr_running(rq_of_dl_rq(dl_rq), 1); dec_dl_deadline(dl_rq, dl_se->deadline); } @@ -2388,10 +2368,7 @@ again: if (dl_server(dl_se)) { p = dl_se->server_pick_task(dl_se); if (!p) { - if (!dl_server_stopped(dl_se)) { - dl_se->dl_yielded = 1; - update_curr_dl_se(rq, dl_se, 0); - } + dl_server_stop(dl_se); goto again; } rq->dl_server = dl_se; @@ -2574,6 +2551,25 @@ static int find_later_rq(struct task_struct *task) return -1; } +static struct task_struct *pick_next_pushable_dl_task(struct rq *rq) +{ + struct task_struct *p; + + if (!has_pushable_dl_tasks(rq)) + return NULL; + + p = __node_2_pdl(rb_first_cached(&rq->dl.pushable_dl_tasks_root)); + + WARN_ON_ONCE(rq->cpu != task_cpu(p)); + WARN_ON_ONCE(task_current(rq, p)); + WARN_ON_ONCE(p->nr_cpus_allowed <= 1); + + WARN_ON_ONCE(!task_on_rq_queued(p)); + WARN_ON_ONCE(!dl_task(p)); + + return p; +} + /* Locks the rq it finds */ static struct rq *find_lock_later_rq(struct task_struct *task, struct rq *rq) { @@ -2601,12 +2597,37 @@ static struct rq *find_lock_later_rq(struct task_struct *task, struct rq *rq) /* Retry if something changed. */ if (double_lock_balance(rq, later_rq)) { - if (unlikely(task_rq(task) != rq || + /* + * double_lock_balance had to release rq->lock, in the + * meantime, task may no longer be fit to be migrated. + * Check the following to ensure that the task is + * still suitable for migration: + * 1. It is possible the task was scheduled, + * migrate_disabled was set and then got preempted, + * so we must check the task migration disable + * flag. + * 2. The CPU picked is in the task's affinity. + * 3. For throttled task (dl_task_offline_migration), + * check the following: + * - the task is not on the rq anymore (it was + * migrated) + * - the task is not on CPU anymore + * - the task is still a dl task + * - the task is not queued on the rq anymore + * 4. For the non-throttled task (push_dl_task), the + * check to ensure that this task is still at the + * head of the pushable tasks list is enough. + */ + if (unlikely(is_migration_disabled(task) || !cpumask_test_cpu(later_rq->cpu, &task->cpus_mask) || - task_on_cpu(rq, task) || - !dl_task(task) || - is_migration_disabled(task) || - !task_on_rq_queued(task))) { + (task->dl.dl_throttled && + (task_rq(task) != rq || + task_on_cpu(rq, task) || + !dl_task(task) || + !task_on_rq_queued(task))) || + (!task->dl.dl_throttled && + task != pick_next_pushable_dl_task(rq)))) { + double_unlock_balance(rq, later_rq); later_rq = NULL; break; @@ -2629,25 +2650,6 @@ static struct rq *find_lock_later_rq(struct task_struct *task, struct rq *rq) return later_rq; } -static struct task_struct *pick_next_pushable_dl_task(struct rq *rq) -{ - struct task_struct *p; - - if (!has_pushable_dl_tasks(rq)) - return NULL; - - p = __node_2_pdl(rb_first_cached(&rq->dl.pushable_dl_tasks_root)); - - WARN_ON_ONCE(rq->cpu != task_cpu(p)); - WARN_ON_ONCE(task_current(rq, p)); - WARN_ON_ONCE(p->nr_cpus_allowed <= 1); - - WARN_ON_ONCE(!task_on_rq_queued(p)); - WARN_ON_ONCE(!dl_task(p)); - - return p; -} - /* * See if the non running -deadline tasks on this rq * can be sent to some other CPU where they can preempt diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c index 3f06ab84d53f..02e16b70a790 100644 --- a/kernel/sched/debug.c +++ b/kernel/sched/debug.c @@ -376,10 +376,8 @@ static ssize_t sched_fair_server_write(struct file *filp, const char __user *ubu return -EINVAL; } - if (rq->cfs.h_nr_queued) { - update_rq_clock(rq); - dl_server_stop(&rq->fair_server); - } + update_rq_clock(rq); + dl_server_stop(&rq->fair_server); retval = dl_server_apply_params(&rq->fair_server, runtime, period, 0); if (retval) diff --git a/kernel/sched/ext.c b/kernel/sched/ext.c index 7dedc9a16281..2b0e88206d07 100644 --- a/kernel/sched/ext.c +++ b/kernel/sched/ext.c @@ -9,1040 +9,6 @@ #include <linux/btf_ids.h> #include "ext_idle.h" -#define SCX_OP_IDX(op) (offsetof(struct sched_ext_ops, op) / sizeof(void (*)(void))) - -enum scx_consts { - SCX_DSP_DFL_MAX_BATCH = 32, - SCX_DSP_MAX_LOOPS = 32, - SCX_WATCHDOG_MAX_TIMEOUT = 30 * HZ, - - SCX_EXIT_BT_LEN = 64, - SCX_EXIT_MSG_LEN = 1024, - SCX_EXIT_DUMP_DFL_LEN = 32768, - - SCX_CPUPERF_ONE = SCHED_CAPACITY_SCALE, - - /* - * Iterating all tasks may take a while. Periodically drop - * scx_tasks_lock to avoid causing e.g. CSD and RCU stalls. - */ - SCX_TASK_ITER_BATCH = 32, -}; - -enum scx_exit_kind { - SCX_EXIT_NONE, - SCX_EXIT_DONE, - - SCX_EXIT_UNREG = 64, /* user-space initiated unregistration */ - SCX_EXIT_UNREG_BPF, /* BPF-initiated unregistration */ - SCX_EXIT_UNREG_KERN, /* kernel-initiated unregistration */ - SCX_EXIT_SYSRQ, /* requested by 'S' sysrq */ - - SCX_EXIT_ERROR = 1024, /* runtime error, error msg contains details */ - SCX_EXIT_ERROR_BPF, /* ERROR but triggered through scx_bpf_error() */ - SCX_EXIT_ERROR_STALL, /* watchdog detected stalled runnable tasks */ -}; - -/* - * An exit code can be specified when exiting with scx_bpf_exit() or scx_exit(), - * corresponding to exit_kind UNREG_BPF and UNREG_KERN respectively. The codes - * are 64bit of the format: - * - * Bits: [63 .. 48 47 .. 32 31 .. 0] - * [ SYS ACT ] [ SYS RSN ] [ USR ] - * - * SYS ACT: System-defined exit actions - * SYS RSN: System-defined exit reasons - * USR : User-defined exit codes and reasons - * - * Using the above, users may communicate intention and context by ORing system - * actions and/or system reasons with a user-defined exit code. - */ -enum scx_exit_code { - /* Reasons */ - SCX_ECODE_RSN_HOTPLUG = 1LLU << 32, - - /* Actions */ - SCX_ECODE_ACT_RESTART = 1LLU << 48, -}; - -/* - * scx_exit_info is passed to ops.exit() to describe why the BPF scheduler is - * being disabled. - */ -struct scx_exit_info { - /* %SCX_EXIT_* - broad category of the exit reason */ - enum scx_exit_kind kind; - - /* exit code if gracefully exiting */ - s64 exit_code; - - /* textual representation of the above */ - const char *reason; - - /* backtrace if exiting due to an error */ - unsigned long *bt; - u32 bt_len; - - /* informational message */ - char *msg; - - /* debug dump */ - char *dump; -}; - -/* sched_ext_ops.flags */ -enum scx_ops_flags { - /* - * Keep built-in idle tracking even if ops.update_idle() is implemented. - */ - SCX_OPS_KEEP_BUILTIN_IDLE = 1LLU << 0, - - /* - * By default, if there are no other task to run on the CPU, ext core - * keeps running the current task even after its slice expires. If this - * flag is specified, such tasks are passed to ops.enqueue() with - * %SCX_ENQ_LAST. See the comment above %SCX_ENQ_LAST for more info. - */ - SCX_OPS_ENQ_LAST = 1LLU << 1, - - /* - * An exiting task may schedule after PF_EXITING is set. In such cases, - * bpf_task_from_pid() may not be able to find the task and if the BPF - * scheduler depends on pid lookup for dispatching, the task will be - * lost leading to various issues including RCU grace period stalls. - * - * To mask this problem, by default, unhashed tasks are automatically - * dispatched to the local DSQ on enqueue. If the BPF scheduler doesn't - * depend on pid lookups and wants to handle these tasks directly, the - * following flag can be used. - */ - SCX_OPS_ENQ_EXITING = 1LLU << 2, - - /* - * If set, only tasks with policy set to SCHED_EXT are attached to - * sched_ext. If clear, SCHED_NORMAL tasks are also included. - */ - SCX_OPS_SWITCH_PARTIAL = 1LLU << 3, - - /* - * A migration disabled task can only execute on its current CPU. By - * default, such tasks are automatically put on the CPU's local DSQ with - * the default slice on enqueue. If this ops flag is set, they also go - * through ops.enqueue(). - * - * A migration disabled task never invokes ops.select_cpu() as it can - * only select the current CPU. Also, p->cpus_ptr will only contain its - * current CPU while p->nr_cpus_allowed keeps tracking p->user_cpus_ptr - * and thus may disagree with cpumask_weight(p->cpus_ptr). - */ - SCX_OPS_ENQ_MIGRATION_DISABLED = 1LLU << 4, - - /* - * Queued wakeup (ttwu_queue) is a wakeup optimization that invokes - * ops.enqueue() on the ops.select_cpu() selected or the wakee's - * previous CPU via IPI (inter-processor interrupt) to reduce cacheline - * transfers. When this optimization is enabled, ops.select_cpu() is - * skipped in some cases (when racing against the wakee switching out). - * As the BPF scheduler may depend on ops.select_cpu() being invoked - * during wakeups, queued wakeup is disabled by default. - * - * If this ops flag is set, queued wakeup optimization is enabled and - * the BPF scheduler must be able to handle ops.enqueue() invoked on the - * wakee's CPU without preceding ops.select_cpu() even for tasks which - * may be executed on multiple CPUs. - */ - SCX_OPS_ALLOW_QUEUED_WAKEUP = 1LLU << 5, - - /* - * If set, enable per-node idle cpumasks. If clear, use a single global - * flat idle cpumask. - */ - SCX_OPS_BUILTIN_IDLE_PER_NODE = 1LLU << 6, - - /* - * CPU cgroup support flags - */ - SCX_OPS_HAS_CGROUP_WEIGHT = 1LLU << 16, /* DEPRECATED, will be removed on 6.18 */ - - SCX_OPS_ALL_FLAGS = SCX_OPS_KEEP_BUILTIN_IDLE | - SCX_OPS_ENQ_LAST | - SCX_OPS_ENQ_EXITING | - SCX_OPS_ENQ_MIGRATION_DISABLED | - SCX_OPS_ALLOW_QUEUED_WAKEUP | - SCX_OPS_SWITCH_PARTIAL | - SCX_OPS_BUILTIN_IDLE_PER_NODE | - SCX_OPS_HAS_CGROUP_WEIGHT, - - /* high 8 bits are internal, don't include in SCX_OPS_ALL_FLAGS */ - __SCX_OPS_INTERNAL_MASK = 0xffLLU << 56, - - SCX_OPS_HAS_CPU_PREEMPT = 1LLU << 56, -}; - -/* argument container for ops.init_task() */ -struct scx_init_task_args { - /* - * Set if ops.init_task() is being invoked on the fork path, as opposed - * to the scheduler transition path. - */ - bool fork; -#ifdef CONFIG_EXT_GROUP_SCHED - /* the cgroup the task is joining */ - struct cgroup *cgroup; -#endif -}; - -/* argument container for ops.exit_task() */ -struct scx_exit_task_args { - /* Whether the task exited before running on sched_ext. */ - bool cancelled; -}; - -/* argument container for ops->cgroup_init() */ -struct scx_cgroup_init_args { - /* the weight of the cgroup [1..10000] */ - u32 weight; - - /* bandwidth control parameters from cpu.max and cpu.max.burst */ - u64 bw_period_us; - u64 bw_quota_us; - u64 bw_burst_us; -}; - -enum scx_cpu_preempt_reason { - /* next task is being scheduled by &sched_class_rt */ - SCX_CPU_PREEMPT_RT, - /* next task is being scheduled by &sched_class_dl */ - SCX_CPU_PREEMPT_DL, - /* next task is being scheduled by &sched_class_stop */ - SCX_CPU_PREEMPT_STOP, - /* unknown reason for SCX being preempted */ - SCX_CPU_PREEMPT_UNKNOWN, -}; - -/* - * Argument container for ops->cpu_acquire(). Currently empty, but may be - * expanded in the future. - */ -struct scx_cpu_acquire_args {}; - -/* argument container for ops->cpu_release() */ -struct scx_cpu_release_args { - /* the reason the CPU was preempted */ - enum scx_cpu_preempt_reason reason; - - /* the task that's going to be scheduled on the CPU */ - struct task_struct *task; -}; - -/* - * Informational context provided to dump operations. - */ -struct scx_dump_ctx { - enum scx_exit_kind kind; - s64 exit_code; - const char *reason; - u64 at_ns; - u64 at_jiffies; -}; - -/** - * struct sched_ext_ops - Operation table for BPF scheduler implementation - * - * A BPF scheduler can implement an arbitrary scheduling policy by - * implementing and loading operations in this table. Note that a userland - * scheduling policy can also be implemented using the BPF scheduler - * as a shim layer. - */ -struct sched_ext_ops { - /** - * @select_cpu: Pick the target CPU for a task which is being woken up - * @p: task being woken up - * @prev_cpu: the cpu @p was on before sleeping - * @wake_flags: SCX_WAKE_* - * - * Decision made here isn't final. @p may be moved to any CPU while it - * is getting dispatched for execution later. However, as @p is not on - * the rq at this point, getting the eventual execution CPU right here - * saves a small bit of overhead down the line. - * - * If an idle CPU is returned, the CPU is kicked and will try to - * dispatch. While an explicit custom mechanism can be added, - * select_cpu() serves as the default way to wake up idle CPUs. - * - * @p may be inserted into a DSQ directly by calling - * scx_bpf_dsq_insert(). If so, the ops.enqueue() will be skipped. - * Directly inserting into %SCX_DSQ_LOCAL will put @p in the local DSQ - * of the CPU returned by this operation. - * - * Note that select_cpu() is never called for tasks that can only run - * on a single CPU or tasks with migration disabled, as they don't have - * the option to select a different CPU. See select_task_rq() for - * details. - */ - s32 (*select_cpu)(struct task_struct *p, s32 prev_cpu, u64 wake_flags); - - /** - * @enqueue: Enqueue a task on the BPF scheduler - * @p: task being enqueued - * @enq_flags: %SCX_ENQ_* - * - * @p is ready to run. Insert directly into a DSQ by calling - * scx_bpf_dsq_insert() or enqueue on the BPF scheduler. If not directly - * inserted, the bpf scheduler owns @p and if it fails to dispatch @p, - * the task will stall. - * - * If @p was inserted into a DSQ from ops.select_cpu(), this callback is - * skipped. - */ - void (*enqueue)(struct task_struct *p, u64 enq_flags); - - /** - * @dequeue: Remove a task from the BPF scheduler - * @p: task being dequeued - * @deq_flags: %SCX_DEQ_* - * - * Remove @p from the BPF scheduler. This is usually called to isolate - * the task while updating its scheduling properties (e.g. priority). - * - * The ext core keeps track of whether the BPF side owns a given task or - * not and can gracefully ignore spurious dispatches from BPF side, - * which makes it safe to not implement this method. However, depending - * on the scheduling logic, this can lead to confusing behaviors - e.g. - * scheduling position not being updated across a priority change. - */ - void (*dequeue)(struct task_struct *p, u64 deq_flags); - - /** - * @dispatch: Dispatch tasks from the BPF scheduler and/or user DSQs - * @cpu: CPU to dispatch tasks for - * @prev: previous task being switched out - * - * Called when a CPU's local dsq is empty. The operation should dispatch - * one or more tasks from the BPF scheduler into the DSQs using - * scx_bpf_dsq_insert() and/or move from user DSQs into the local DSQ - * using scx_bpf_dsq_move_to_local(). - * - * The maximum number of times scx_bpf_dsq_insert() can be called - * without an intervening scx_bpf_dsq_move_to_local() is specified by - * ops.dispatch_max_batch. See the comments on top of the two functions - * for more details. - * - * When not %NULL, @prev is an SCX task with its slice depleted. If - * @prev is still runnable as indicated by set %SCX_TASK_QUEUED in - * @prev->scx.flags, it is not enqueued yet and will be enqueued after - * ops.dispatch() returns. To keep executing @prev, return without - * dispatching or moving any tasks. Also see %SCX_OPS_ENQ_LAST. - */ - void (*dispatch)(s32 cpu, struct task_struct *prev); - - /** - * @tick: Periodic tick - * @p: task running currently - * - * This operation is called every 1/HZ seconds on CPUs which are - * executing an SCX task. Setting @p->scx.slice to 0 will trigger an - * immediate dispatch cycle on the CPU. - */ - void (*tick)(struct task_struct *p); - - /** - * @runnable: A task is becoming runnable on its associated CPU - * @p: task becoming runnable - * @enq_flags: %SCX_ENQ_* - * - * This and the following three functions can be used to track a task's - * execution state transitions. A task becomes ->runnable() on a CPU, - * and then goes through one or more ->running() and ->stopping() pairs - * as it runs on the CPU, and eventually becomes ->quiescent() when it's - * done running on the CPU. - * - * @p is becoming runnable on the CPU because it's - * - * - waking up (%SCX_ENQ_WAKEUP) - * - being moved from another CPU - * - being restored after temporarily taken off the queue for an - * attribute change. - * - * This and ->enqueue() are related but not coupled. This operation - * notifies @p's state transition and may not be followed by ->enqueue() - * e.g. when @p is being dispatched to a remote CPU, or when @p is - * being enqueued on a CPU experiencing a hotplug event. Likewise, a - * task may be ->enqueue()'d without being preceded by this operation - * e.g. after exhausting its slice. - */ - void (*runnable)(struct task_struct *p, u64 enq_flags); - - /** - * @running: A task is starting to run on its associated CPU - * @p: task starting to run - * - * Note that this callback may be called from a CPU other than the - * one the task is going to run on. This can happen when a task - * property is changed (i.e., affinity), since scx_next_task_scx(), - * which triggers this callback, may run on a CPU different from - * the task's assigned CPU. - * - * Therefore, always use scx_bpf_task_cpu(@p) to determine the - * target CPU the task is going to use. - * - * See ->runnable() for explanation on the task state notifiers. - */ - void (*running)(struct task_struct *p); - - /** - * @stopping: A task is stopping execution - * @p: task stopping to run - * @runnable: is task @p still runnable? - * - * Note that this callback may be called from a CPU other than the - * one the task was running on. This can happen when a task - * property is changed (i.e., affinity), since dequeue_task_scx(), - * which triggers this callback, may run on a CPU different from - * the task's assigned CPU. - * - * Therefore, always use scx_bpf_task_cpu(@p) to retrieve the CPU - * the task was running on. - * - * See ->runnable() for explanation on the task state notifiers. If - * !@runnable, ->quiescent() will be invoked after this operation - * returns. - */ - void (*stopping)(struct task_struct *p, bool runnable); - - /** - * @quiescent: A task is becoming not runnable on its associated CPU - * @p: task becoming not runnable - * @deq_flags: %SCX_DEQ_* - * - * See ->runnable() for explanation on the task state notifiers. - * - * @p is becoming quiescent on the CPU because it's - * - * - sleeping (%SCX_DEQ_SLEEP) - * - being moved to another CPU - * - being temporarily taken off the queue for an attribute change - * (%SCX_DEQ_SAVE) - * - * This and ->dequeue() are related but not coupled. This operation - * notifies @p's state transition and may not be preceded by ->dequeue() - * e.g. when @p is being dispatched to a remote CPU. - */ - void (*quiescent)(struct task_struct *p, u64 deq_flags); - - /** - * @yield: Yield CPU - * @from: yielding task - * @to: optional yield target task - * - * If @to is NULL, @from is yielding the CPU to other runnable tasks. - * The BPF scheduler should ensure that other available tasks are - * dispatched before the yielding task. Return value is ignored in this - * case. - * - * If @to is not-NULL, @from wants to yield the CPU to @to. If the bpf - * scheduler can implement the request, return %true; otherwise, %false. - */ - bool (*yield)(struct task_struct *from, struct task_struct *to); - - /** - * @core_sched_before: Task ordering for core-sched - * @a: task A - * @b: task B - * - * Used by core-sched to determine the ordering between two tasks. See - * Documentation/admin-guide/hw-vuln/core-scheduling.rst for details on - * core-sched. - * - * Both @a and @b are runnable and may or may not currently be queued on - * the BPF scheduler. Should return %true if @a should run before @b. - * %false if there's no required ordering or @b should run before @a. - * - * If not specified, the default is ordering them according to when they - * became runnable. - */ - bool (*core_sched_before)(struct task_struct *a, struct task_struct *b); - - /** - * @set_weight: Set task weight - * @p: task to set weight for - * @weight: new weight [1..10000] - * - * Update @p's weight to @weight. - */ - void (*set_weight)(struct task_struct *p, u32 weight); - - /** - * @set_cpumask: Set CPU affinity - * @p: task to set CPU affinity for - * @cpumask: cpumask of cpus that @p can run on - * - * Update @p's CPU affinity to @cpumask. - */ - void (*set_cpumask)(struct task_struct *p, - const struct cpumask *cpumask); - - /** - * @update_idle: Update the idle state of a CPU - * @cpu: CPU to update the idle state for - * @idle: whether entering or exiting the idle state - * - * This operation is called when @rq's CPU goes or leaves the idle - * state. By default, implementing this operation disables the built-in - * idle CPU tracking and the following helpers become unavailable: - * - * - scx_bpf_select_cpu_dfl() - * - scx_bpf_select_cpu_and() - * - scx_bpf_test_and_clear_cpu_idle() - * - scx_bpf_pick_idle_cpu() - * - * The user also must implement ops.select_cpu() as the default - * implementation relies on scx_bpf_select_cpu_dfl(). - * - * Specify the %SCX_OPS_KEEP_BUILTIN_IDLE flag to keep the built-in idle - * tracking. - */ - void (*update_idle)(s32 cpu, bool idle); - - /** - * @cpu_acquire: A CPU is becoming available to the BPF scheduler - * @cpu: The CPU being acquired by the BPF scheduler. - * @args: Acquire arguments, see the struct definition. - * - * A CPU that was previously released from the BPF scheduler is now once - * again under its control. - */ - void (*cpu_acquire)(s32 cpu, struct scx_cpu_acquire_args *args); - - /** - * @cpu_release: A CPU is taken away from the BPF scheduler - * @cpu: The CPU being released by the BPF scheduler. - * @args: Release arguments, see the struct definition. - * - * The specified CPU is no longer under the control of the BPF - * scheduler. This could be because it was preempted by a higher - * priority sched_class, though there may be other reasons as well. The - * caller should consult @args->reason to determine the cause. - */ - void (*cpu_release)(s32 cpu, struct scx_cpu_release_args *args); - - /** - * @init_task: Initialize a task to run in a BPF scheduler - * @p: task to initialize for BPF scheduling - * @args: init arguments, see the struct definition - * - * Either we're loading a BPF scheduler or a new task is being forked. - * Initialize @p for BPF scheduling. This operation may block and can - * be used for allocations, and is called exactly once for a task. - * - * Return 0 for success, -errno for failure. An error return while - * loading will abort loading of the BPF scheduler. During a fork, it - * will abort that specific fork. - */ - s32 (*init_task)(struct task_struct *p, struct scx_init_task_args *args); - - /** - * @exit_task: Exit a previously-running task from the system - * @p: task to exit - * @args: exit arguments, see the struct definition - * - * @p is exiting or the BPF scheduler is being unloaded. Perform any - * necessary cleanup for @p. - */ - void (*exit_task)(struct task_struct *p, struct scx_exit_task_args *args); - - /** - * @enable: Enable BPF scheduling for a task - * @p: task to enable BPF scheduling for - * - * Enable @p for BPF scheduling. enable() is called on @p any time it - * enters SCX, and is always paired with a matching disable(). - */ - void (*enable)(struct task_struct *p); - - /** - * @disable: Disable BPF scheduling for a task - * @p: task to disable BPF scheduling for - * - * @p is exiting, leaving SCX or the BPF scheduler is being unloaded. - * Disable BPF scheduling for @p. A disable() call is always matched - * with a prior enable() call. - */ - void (*disable)(struct task_struct *p); - - /** - * @dump: Dump BPF scheduler state on error - * @ctx: debug dump context - * - * Use scx_bpf_dump() to generate BPF scheduler specific debug dump. - */ - void (*dump)(struct scx_dump_ctx *ctx); - - /** - * @dump_cpu: Dump BPF scheduler state for a CPU on error - * @ctx: debug dump context - * @cpu: CPU to generate debug dump for - * @idle: @cpu is currently idle without any runnable tasks - * - * Use scx_bpf_dump() to generate BPF scheduler specific debug dump for - * @cpu. If @idle is %true and this operation doesn't produce any - * output, @cpu is skipped for dump. - */ - void (*dump_cpu)(struct scx_dump_ctx *ctx, s32 cpu, bool idle); - - /** - * @dump_task: Dump BPF scheduler state for a runnable task on error - * @ctx: debug dump context - * @p: runnable task to generate debug dump for - * - * Use scx_bpf_dump() to generate BPF scheduler specific debug dump for - * @p. - */ - void (*dump_task)(struct scx_dump_ctx *ctx, struct task_struct *p); - -#ifdef CONFIG_EXT_GROUP_SCHED - /** - * @cgroup_init: Initialize a cgroup - * @cgrp: cgroup being initialized - * @args: init arguments, see the struct definition - * - * Either the BPF scheduler is being loaded or @cgrp created, initialize - * @cgrp for sched_ext. This operation may block. - * - * Return 0 for success, -errno for failure. An error return while - * loading will abort loading of the BPF scheduler. During cgroup - * creation, it will abort the specific cgroup creation. - */ - s32 (*cgroup_init)(struct cgroup *cgrp, - struct scx_cgroup_init_args *args); - - /** - * @cgroup_exit: Exit a cgroup - * @cgrp: cgroup being exited - * - * Either the BPF scheduler is being unloaded or @cgrp destroyed, exit - * @cgrp for sched_ext. This operation my block. - */ - void (*cgroup_exit)(struct cgroup *cgrp); - - /** - * @cgroup_prep_move: Prepare a task to be moved to a different cgroup - * @p: task being moved - * @from: cgroup @p is being moved from - * @to: cgroup @p is being moved to - * - * Prepare @p for move from cgroup @from to @to. This operation may - * block and can be used for allocations. - * - * Return 0 for success, -errno for failure. An error return aborts the - * migration. - */ - s32 (*cgroup_prep_move)(struct task_struct *p, - struct cgroup *from, struct cgroup *to); - - /** - * @cgroup_move: Commit cgroup move - * @p: task being moved - * @from: cgroup @p is being moved from - * @to: cgroup @p is being moved to - * - * Commit the move. @p is dequeued during this operation. - */ - void (*cgroup_move)(struct task_struct *p, - struct cgroup *from, struct cgroup *to); - - /** - * @cgroup_cancel_move: Cancel cgroup move - * @p: task whose cgroup move is being canceled - * @from: cgroup @p was being moved from - * @to: cgroup @p was being moved to - * - * @p was cgroup_prep_move()'d but failed before reaching cgroup_move(). - * Undo the preparation. - */ - void (*cgroup_cancel_move)(struct task_struct *p, - struct cgroup *from, struct cgroup *to); - - /** - * @cgroup_set_weight: A cgroup's weight is being changed - * @cgrp: cgroup whose weight is being updated - * @weight: new weight [1..10000] - * - * Update @cgrp's weight to @weight. - */ - void (*cgroup_set_weight)(struct cgroup *cgrp, u32 weight); - - /** - * @cgroup_set_bandwidth: A cgroup's bandwidth is being changed - * @cgrp: cgroup whose bandwidth is being updated - * @period_us: bandwidth control period - * @quota_us: bandwidth control quota - * @burst_us: bandwidth control burst - * - * Update @cgrp's bandwidth control parameters. This is from the cpu.max - * cgroup interface. - * - * @quota_us / @period_us determines the CPU bandwidth @cgrp is entitled - * to. For example, if @period_us is 1_000_000 and @quota_us is - * 2_500_000. @cgrp is entitled to 2.5 CPUs. @burst_us can be - * interpreted in the same fashion and specifies how much @cgrp can - * burst temporarily. The specific control mechanism and thus the - * interpretation of @period_us and burstiness is upto to the BPF - * scheduler. - */ - void (*cgroup_set_bandwidth)(struct cgroup *cgrp, - u64 period_us, u64 quota_us, u64 burst_us); - -#endif /* CONFIG_EXT_GROUP_SCHED */ - - /* - * All online ops must come before ops.cpu_online(). - */ - - /** - * @cpu_online: A CPU became online - * @cpu: CPU which just came up - * - * @cpu just came online. @cpu will not call ops.enqueue() or - * ops.dispatch(), nor run tasks associated with other CPUs beforehand. - */ - void (*cpu_online)(s32 cpu); - - /** - * @cpu_offline: A CPU is going offline - * @cpu: CPU which is going offline - * - * @cpu is going offline. @cpu will not call ops.enqueue() or - * ops.dispatch(), nor run tasks associated with other CPUs afterwards. - */ - void (*cpu_offline)(s32 cpu); - - /* - * All CPU hotplug ops must come before ops.init(). - */ - - /** - * @init: Initialize the BPF scheduler - */ - s32 (*init)(void); - - /** - * @exit: Clean up after the BPF scheduler - * @info: Exit info - * - * ops.exit() is also called on ops.init() failure, which is a bit - * unusual. This is to allow rich reporting through @info on how - * ops.init() failed. - */ - void (*exit)(struct scx_exit_info *info); - - /** - * @dispatch_max_batch: Max nr of tasks that dispatch() can dispatch - */ - u32 dispatch_max_batch; - - /** - * @flags: %SCX_OPS_* flags - */ - u64 flags; - - /** - * @timeout_ms: The maximum amount of time, in milliseconds, that a - * runnable task should be able to wait before being scheduled. The - * maximum timeout may not exceed the default timeout of 30 seconds. - * - * Defaults to the maximum allowed timeout value of 30 seconds. - */ - u32 timeout_ms; - - /** - * @exit_dump_len: scx_exit_info.dump buffer length. If 0, the default - * value of 32768 is used. - */ - u32 exit_dump_len; - - /** - * @hotplug_seq: A sequence number that may be set by the scheduler to - * detect when a hotplug event has occurred during the loading process. - * If 0, no detection occurs. Otherwise, the scheduler will fail to - * load if the sequence number does not match @scx_hotplug_seq on the - * enable path. - */ - u64 hotplug_seq; - - /** - * @name: BPF scheduler's name - * - * Must be a non-zero valid BPF object name including only isalnum(), - * '_' and '.' chars. Shows up in kernel.sched_ext_ops sysctl while the - * BPF scheduler is enabled. - */ - char name[SCX_OPS_NAME_LEN]; - - /* internal use only, must be NULL */ - void *priv; -}; - -enum scx_opi { - SCX_OPI_BEGIN = 0, - SCX_OPI_NORMAL_BEGIN = 0, - SCX_OPI_NORMAL_END = SCX_OP_IDX(cpu_online), - SCX_OPI_CPU_HOTPLUG_BEGIN = SCX_OP_IDX(cpu_online), - SCX_OPI_CPU_HOTPLUG_END = SCX_OP_IDX(init), - SCX_OPI_END = SCX_OP_IDX(init), -}; - -/* - * Collection of event counters. Event types are placed in descending order. - */ -struct scx_event_stats { - /* - * If ops.select_cpu() returns a CPU which can't be used by the task, - * the core scheduler code silently picks a fallback CPU. - */ - s64 SCX_EV_SELECT_CPU_FALLBACK; - - /* - * When dispatching to a local DSQ, the CPU may have gone offline in - * the meantime. In this case, the task is bounced to the global DSQ. - */ - s64 SCX_EV_DISPATCH_LOCAL_DSQ_OFFLINE; - - /* - * If SCX_OPS_ENQ_LAST is not set, the number of times that a task - * continued to run because there were no other tasks on the CPU. - */ - s64 SCX_EV_DISPATCH_KEEP_LAST; - - /* - * If SCX_OPS_ENQ_EXITING is not set, the number of times that a task - * is dispatched to a local DSQ when exiting. - */ - s64 SCX_EV_ENQ_SKIP_EXITING; - - /* - * If SCX_OPS_ENQ_MIGRATION_DISABLED is not set, the number of times a - * migration disabled task skips ops.enqueue() and is dispatched to its - * local DSQ. - */ - s64 SCX_EV_ENQ_SKIP_MIGRATION_DISABLED; - - /* - * Total number of times a task's time slice was refilled with the - * default value (SCX_SLICE_DFL). - */ - s64 SCX_EV_REFILL_SLICE_DFL; - - /* - * The total duration of bypass modes in nanoseconds. - */ - s64 SCX_EV_BYPASS_DURATION; - - /* - * The number of tasks dispatched in the bypassing mode. - */ - s64 SCX_EV_BYPASS_DISPATCH; - - /* - * The number of times the bypassing mode has been activated. - */ - s64 SCX_EV_BYPASS_ACTIVATE; -}; - -struct scx_sched { - struct sched_ext_ops ops; - DECLARE_BITMAP(has_op, SCX_OPI_END); - - /* - * Dispatch queues. - * - * The global DSQ (%SCX_DSQ_GLOBAL) is split per-node for scalability. - * This is to avoid live-locking in bypass mode where all tasks are - * dispatched to %SCX_DSQ_GLOBAL and all CPUs consume from it. If - * per-node split isn't sufficient, it can be further split. - */ - struct rhashtable dsq_hash; - struct scx_dispatch_q **global_dsqs; - - /* - * The event counters are in a per-CPU variable to minimize the - * accounting overhead. A system-wide view on the event counter is - * constructed when requested by scx_bpf_events(). - */ - struct scx_event_stats __percpu *event_stats_cpu; - - bool warned_zero_slice; - - atomic_t exit_kind; - struct scx_exit_info *exit_info; - - struct kobject kobj; - - struct kthread_worker *helper; - struct irq_work error_irq_work; - struct kthread_work disable_work; - struct rcu_work rcu_work; -}; - -enum scx_wake_flags { - /* expose select WF_* flags as enums */ - SCX_WAKE_FORK = WF_FORK, - SCX_WAKE_TTWU = WF_TTWU, - SCX_WAKE_SYNC = WF_SYNC, -}; - -enum scx_enq_flags { - /* expose select ENQUEUE_* flags as enums */ - SCX_ENQ_WAKEUP = ENQUEUE_WAKEUP, - SCX_ENQ_HEAD = ENQUEUE_HEAD, - SCX_ENQ_CPU_SELECTED = ENQUEUE_RQ_SELECTED, - - /* high 32bits are SCX specific */ - - /* - * Set the following to trigger preemption when calling - * scx_bpf_dsq_insert() with a local dsq as the target. The slice of the - * current task is cleared to zero and the CPU is kicked into the - * scheduling path. Implies %SCX_ENQ_HEAD. - */ - SCX_ENQ_PREEMPT = 1LLU << 32, - - /* - * The task being enqueued was previously enqueued on the current CPU's - * %SCX_DSQ_LOCAL, but was removed from it in a call to the - * scx_bpf_reenqueue_local() kfunc. If scx_bpf_reenqueue_local() was - * invoked in a ->cpu_release() callback, and the task is again - * dispatched back to %SCX_LOCAL_DSQ by this current ->enqueue(), the - * task will not be scheduled on the CPU until at least the next invocation - * of the ->cpu_acquire() callback. - */ - SCX_ENQ_REENQ = 1LLU << 40, - - /* - * The task being enqueued is the only task available for the cpu. By - * default, ext core keeps executing such tasks but when - * %SCX_OPS_ENQ_LAST is specified, they're ops.enqueue()'d with the - * %SCX_ENQ_LAST flag set. - * - * The BPF scheduler is responsible for triggering a follow-up - * scheduling event. Otherwise, Execution may stall. - */ - SCX_ENQ_LAST = 1LLU << 41, - - /* high 8 bits are internal */ - __SCX_ENQ_INTERNAL_MASK = 0xffLLU << 56, - - SCX_ENQ_CLEAR_OPSS = 1LLU << 56, - SCX_ENQ_DSQ_PRIQ = 1LLU << 57, -}; - -enum scx_deq_flags { - /* expose select DEQUEUE_* flags as enums */ - SCX_DEQ_SLEEP = DEQUEUE_SLEEP, - - /* high 32bits are SCX specific */ - - /* - * The generic core-sched layer decided to execute the task even though - * it hasn't been dispatched yet. Dequeue from the BPF side. - */ - SCX_DEQ_CORE_SCHED_EXEC = 1LLU << 32, -}; - -enum scx_pick_idle_cpu_flags { - SCX_PICK_IDLE_CORE = 1LLU << 0, /* pick a CPU whose SMT siblings are also idle */ - SCX_PICK_IDLE_IN_NODE = 1LLU << 1, /* pick a CPU in the same target NUMA node */ -}; - -enum scx_kick_flags { - /* - * Kick the target CPU if idle. Guarantees that the target CPU goes - * through at least one full scheduling cycle before going idle. If the - * target CPU can be determined to be currently not idle and going to go - * through a scheduling cycle before going idle, noop. - */ - SCX_KICK_IDLE = 1LLU << 0, - - /* - * Preempt the current task and execute the dispatch path. If the - * current task of the target CPU is an SCX task, its ->scx.slice is - * cleared to zero before the scheduling path is invoked so that the - * task expires and the dispatch path is invoked. - */ - SCX_KICK_PREEMPT = 1LLU << 1, - - /* - * Wait for the CPU to be rescheduled. The scx_bpf_kick_cpu() call will - * return after the target CPU finishes picking the next task. - */ - SCX_KICK_WAIT = 1LLU << 2, -}; - -enum scx_tg_flags { - SCX_TG_ONLINE = 1U << 0, - SCX_TG_INITED = 1U << 1, -}; - -enum scx_enable_state { - SCX_ENABLING, - SCX_ENABLED, - SCX_DISABLING, - SCX_DISABLED, -}; - -static const char *scx_enable_state_str[] = { - [SCX_ENABLING] = "enabling", - [SCX_ENABLED] = "enabled", - [SCX_DISABLING] = "disabling", - [SCX_DISABLED] = "disabled", -}; - -/* - * sched_ext_entity->ops_state - * - * Used to track the task ownership between the SCX core and the BPF scheduler. - * State transitions look as follows: - * - * NONE -> QUEUEING -> QUEUED -> DISPATCHING - * ^ | | - * | v v - * \-------------------------------/ - * - * QUEUEING and DISPATCHING states can be waited upon. See wait_ops_state() call - * sites for explanations on the conditions being waited upon and why they are - * safe. Transitions out of them into NONE or QUEUED must store_release and the - * waiters should load_acquire. - * - * Tracking scx_ops_state enables sched_ext core to reliably determine whether - * any given task can be dispatched by the BPF scheduler at all times and thus - * relaxes the requirements on the BPF scheduler. This allows the BPF scheduler - * to try to dispatch any task anytime regardless of its state as the SCX core - * can safely reject invalid dispatches. - */ -enum scx_ops_state { - SCX_OPSS_NONE, /* owned by the SCX core */ - SCX_OPSS_QUEUEING, /* in transit to the BPF scheduler */ - SCX_OPSS_QUEUED, /* owned by the BPF scheduler */ - SCX_OPSS_DISPATCHING, /* in transit back to the SCX core */ - - /* - * QSEQ brands each QUEUED instance so that, when dispatch races - * dequeue/requeue, the dispatcher can tell whether it still has a claim - * on the task being dispatched. - * - * As some 32bit archs can't do 64bit store_release/load_acquire, - * p->scx.ops_state is atomic_long_t which leaves 30 bits for QSEQ on - * 32bit machines. The dispatch race window QSEQ protects is very narrow - * and runs with IRQ disabled. 30 bits should be sufficient. - */ - SCX_OPSS_QSEQ_SHIFT = 2, -}; - -/* Use macros to ensure that the type is unsigned long for the masks */ -#define SCX_OPSS_STATE_MASK ((1LU << SCX_OPSS_QSEQ_SHIFT) - 1) -#define SCX_OPSS_QSEQ_MASK (~SCX_OPSS_STATE_MASK) - /* * NOTE: sched_ext is in the process of growing multiple scheduler support and * scx_root usage is in a transitional state. Naked dereferences are safe if the @@ -1170,7 +136,7 @@ static struct kset *scx_kset; #include <trace/events/sched_ext.h> static void process_ddsp_deferred_locals(struct rq *rq); -static void scx_bpf_kick_cpu(s32 cpu, u64 flags); +static void scx_kick_cpu(struct scx_sched *sch, s32 cpu, u64 flags); static void scx_vexit(struct scx_sched *sch, enum scx_exit_kind kind, s64 exit_code, const char *fmt, va_list args); @@ -1185,24 +151,7 @@ static __printf(4, 5) void scx_exit(struct scx_sched *sch, va_end(args); } -static __printf(3, 4) void scx_kf_exit(enum scx_exit_kind kind, s64 exit_code, - const char *fmt, ...) -{ - struct scx_sched *sch; - va_list args; - - rcu_read_lock(); - sch = rcu_dereference(scx_root); - if (sch) { - va_start(args, fmt); - scx_vexit(sch, kind, exit_code, fmt, args); - va_end(args); - } - rcu_read_unlock(); -} - #define scx_error(sch, fmt, args...) scx_exit((sch), SCX_EXIT_ERROR, 0, fmt, ##args) -#define scx_kf_error(fmt, args...) scx_kf_exit(SCX_EXIT_ERROR, 0, fmt, ##args) #define SCX_HAS_OP(sch, op) test_bit(SCX_OP_IDX(op), (sch)->has_op) @@ -1232,10 +181,9 @@ static bool u32_before(u32 a, u32 b) return (s32)(a - b) < 0; } -static struct scx_dispatch_q *find_global_dsq(struct task_struct *p) +static struct scx_dispatch_q *find_global_dsq(struct scx_sched *sch, + struct task_struct *p) { - struct scx_sched *sch = scx_root; - return sch->global_dsqs[cpu_to_node(task_cpu(p))]; } @@ -1363,11 +311,11 @@ do { \ }) /* @mask is constant, always inline to cull unnecessary branches */ -static __always_inline bool scx_kf_allowed(u32 mask) +static __always_inline bool scx_kf_allowed(struct scx_sched *sch, u32 mask) { if (unlikely(!(current->scx.kf_mask & mask))) { - scx_kf_error("kfunc with mask 0x%x called from an operation only allowing 0x%x", - mask, current->scx.kf_mask); + scx_error(sch, "kfunc with mask 0x%x called from an operation only allowing 0x%x", + mask, current->scx.kf_mask); return false; } @@ -1380,13 +328,13 @@ static __always_inline bool scx_kf_allowed(u32 mask) */ if (unlikely(highest_bit(mask) == SCX_KF_CPU_RELEASE && (current->scx.kf_mask & higher_bits(SCX_KF_CPU_RELEASE)))) { - scx_kf_error("cpu_release kfunc called from a nested operation"); + scx_error(sch, "cpu_release kfunc called from a nested operation"); return false; } if (unlikely(highest_bit(mask) == SCX_KF_DISPATCH && (current->scx.kf_mask & higher_bits(SCX_KF_DISPATCH)))) { - scx_kf_error("dispatch kfunc called from a nested operation"); + scx_error(sch, "dispatch kfunc called from a nested operation"); return false; } @@ -1394,15 +342,16 @@ static __always_inline bool scx_kf_allowed(u32 mask) } /* see SCX_CALL_OP_TASK() */ -static __always_inline bool scx_kf_allowed_on_arg_tasks(u32 mask, +static __always_inline bool scx_kf_allowed_on_arg_tasks(struct scx_sched *sch, + u32 mask, struct task_struct *p) { - if (!scx_kf_allowed(mask)) + if (!scx_kf_allowed(sch, mask)) return false; if (unlikely((p != current->scx.kf_tasks[0] && p != current->scx.kf_tasks[1]))) { - scx_kf_error("called on a task not being operated on"); + scx_error(sch, "called on a task not being operated on"); return false; } @@ -1488,10 +437,11 @@ struct bpf_iter_scx_dsq { */ struct scx_task_iter { struct sched_ext_entity cursor; - struct task_struct *locked; + struct task_struct *locked_task; struct rq *rq; struct rq_flags rf; u32 cnt; + bool list_locked; }; /** @@ -1519,15 +469,16 @@ static void scx_task_iter_start(struct scx_task_iter *iter) iter->cursor = (struct sched_ext_entity){ .flags = SCX_TASK_CURSOR }; list_add(&iter->cursor.tasks_node, &scx_tasks); - iter->locked = NULL; + iter->locked_task = NULL; iter->cnt = 0; + iter->list_locked = true; } static void __scx_task_iter_rq_unlock(struct scx_task_iter *iter) { - if (iter->locked) { - task_rq_unlock(iter->rq, iter->locked, &iter->rf); - iter->locked = NULL; + if (iter->locked_task) { + task_rq_unlock(iter->rq, iter->locked_task, &iter->rf); + iter->locked_task = NULL; } } @@ -1537,24 +488,24 @@ static void __scx_task_iter_rq_unlock(struct scx_task_iter *iter) * * If @iter is in the middle of a locked iteration, it may be locking the rq of * the task currently being visited in addition to scx_tasks_lock. Unlock both. - * This function can be safely called anytime during an iteration. + * This function can be safely called anytime during an iteration. The next + * iterator operation will automatically restore the necessary locking. */ static void scx_task_iter_unlock(struct scx_task_iter *iter) { __scx_task_iter_rq_unlock(iter); - spin_unlock_irq(&scx_tasks_lock); + if (iter->list_locked) { + iter->list_locked = false; + spin_unlock_irq(&scx_tasks_lock); + } } -/** - * scx_task_iter_relock - Lock scx_tasks_lock released by scx_task_iter_unlock() - * @iter: iterator to re-lock - * - * Re-lock scx_tasks_lock unlocked by scx_task_iter_unlock(). Note that it - * doesn't re-lock the rq lock. Must be called before other iterator operations. - */ -static void scx_task_iter_relock(struct scx_task_iter *iter) +static void __scx_task_iter_maybe_relock(struct scx_task_iter *iter) { - spin_lock_irq(&scx_tasks_lock); + if (!iter->list_locked) { + spin_lock_irq(&scx_tasks_lock); + iter->list_locked = true; + } } /** @@ -1567,6 +518,7 @@ static void scx_task_iter_relock(struct scx_task_iter *iter) */ static void scx_task_iter_stop(struct scx_task_iter *iter) { + __scx_task_iter_maybe_relock(iter); list_del_init(&iter->cursor.tasks_node); scx_task_iter_unlock(iter); } @@ -1584,10 +536,12 @@ static struct task_struct *scx_task_iter_next(struct scx_task_iter *iter) struct list_head *cursor = &iter->cursor.tasks_node; struct sched_ext_entity *pos; + __scx_task_iter_maybe_relock(iter); + if (!(++iter->cnt % SCX_TASK_ITER_BATCH)) { scx_task_iter_unlock(iter); cond_resched(); - scx_task_iter_relock(iter); + __scx_task_iter_maybe_relock(iter); } list_for_each_entry(pos, cursor, tasks_node) { @@ -1650,7 +604,7 @@ static struct task_struct *scx_task_iter_next_locked(struct scx_task_iter *iter) return NULL; iter->rq = task_rq_lock(p, &iter->rf); - iter->locked = p; + iter->locked_task = p; return p; } @@ -1664,7 +618,7 @@ static struct task_struct *scx_task_iter_next_locked(struct scx_task_iter *iter) * This can be used when preemption is not disabled. */ #define scx_add_event(sch, name, cnt) do { \ - this_cpu_add((sch)->event_stats_cpu->name, (cnt)); \ + this_cpu_add((sch)->pcpu->event_stats.name, (cnt)); \ trace_sched_ext_event(#name, (cnt)); \ } while(0) @@ -1677,7 +631,7 @@ static struct task_struct *scx_task_iter_next_locked(struct scx_task_iter *iter) * This should be used only when preemption is disabled. */ #define __scx_add_event(sch, name, cnt) do { \ - __this_cpu_add((sch)->event_stats_cpu->name, (cnt)); \ + __this_cpu_add((sch)->pcpu->event_stats.name, (cnt)); \ trace_sched_ext_event(#name, cnt); \ } while(0) @@ -1766,23 +720,6 @@ static bool ops_cpu_valid(struct scx_sched *sch, s32 cpu, const char *where) } /** - * kf_cpu_valid - Verify a CPU number, to be used on kfunc input args - * @cpu: cpu number which came from a BPF ops - * @where: extra information reported on error - * - * The same as ops_cpu_valid() but @sch is implicit. - */ -static bool kf_cpu_valid(u32 cpu, const char *where) -{ - if (__cpu_valid(cpu)) { - return true; - } else { - scx_kf_error("invalid CPU %d%s%s", cpu, where ? " " : "", where ?: ""); - return false; - } -} - -/** * ops_sanitize_err - Sanitize a -errno value * @sch: scx_sched to error out on error * @ops_name: operation to blame on failure @@ -1942,10 +879,10 @@ static void dsq_mod_nr(struct scx_dispatch_q *dsq, s32 delta) WRITE_ONCE(dsq->nr, dsq->nr + delta); } -static void refill_task_slice_dfl(struct task_struct *p) +static void refill_task_slice_dfl(struct scx_sched *sch, struct task_struct *p) { p->scx.slice = SCX_SLICE_DFL; - __scx_add_event(scx_root, SCX_EV_REFILL_SLICE_DFL, 1); + __scx_add_event(sch, SCX_EV_REFILL_SLICE_DFL, 1); } static void dispatch_enqueue(struct scx_sched *sch, struct scx_dispatch_q *dsq, @@ -1963,7 +900,7 @@ static void dispatch_enqueue(struct scx_sched *sch, struct scx_dispatch_q *dsq, scx_error(sch, "attempting to dispatch to a destroyed dsq"); /* fall back to the global dsq */ raw_spin_unlock(&dsq->lock); - dsq = find_global_dsq(p); + dsq = find_global_dsq(sch, p); raw_spin_lock(&dsq->lock); } } @@ -2142,26 +1079,27 @@ static struct scx_dispatch_q *find_dsq_for_dispatch(struct scx_sched *sch, s32 cpu = dsq_id & SCX_DSQ_LOCAL_CPU_MASK; if (!ops_cpu_valid(sch, cpu, "in SCX_DSQ_LOCAL_ON dispatch verdict")) - return find_global_dsq(p); + return find_global_dsq(sch, p); return &cpu_rq(cpu)->scx.local_dsq; } if (dsq_id == SCX_DSQ_GLOBAL) - dsq = find_global_dsq(p); + dsq = find_global_dsq(sch, p); else dsq = find_user_dsq(sch, dsq_id); if (unlikely(!dsq)) { scx_error(sch, "non-existent DSQ 0x%llx for %s[%d]", dsq_id, p->comm, p->pid); - return find_global_dsq(p); + return find_global_dsq(sch, p); } return dsq; } -static void mark_direct_dispatch(struct task_struct *ddsp_task, +static void mark_direct_dispatch(struct scx_sched *sch, + struct task_struct *ddsp_task, struct task_struct *p, u64 dsq_id, u64 enq_flags) { @@ -2175,10 +1113,10 @@ static void mark_direct_dispatch(struct task_struct *ddsp_task, /* @p must match the task on the enqueue path */ if (unlikely(p != ddsp_task)) { if (IS_ERR(ddsp_task)) - scx_kf_error("%s[%d] already direct-dispatched", + scx_error(sch, "%s[%d] already direct-dispatched", p->comm, p->pid); else - scx_kf_error("scheduling for %s[%d] but trying to direct-dispatch %s[%d]", + scx_error(sch, "scheduling for %s[%d] but trying to direct-dispatch %s[%d]", ddsp_task->comm, ddsp_task->pid, p->comm, p->pid); return; @@ -2333,15 +1271,15 @@ local: * higher priority it becomes from scx_prio_less()'s POV. */ touch_core_sched(rq, p); - refill_task_slice_dfl(p); + refill_task_slice_dfl(sch, p); local_norefill: dispatch_enqueue(sch, &rq->scx.local_dsq, p, enq_flags); return; global: touch_core_sched(rq, p); /* see the comment in local: */ - refill_task_slice_dfl(p); - dispatch_enqueue(sch, find_global_dsq(p), p, enq_flags); + refill_task_slice_dfl(sch, p); + dispatch_enqueue(sch, find_global_dsq(sch, p), p, enq_flags); } static bool task_runnable(const struct task_struct *p) @@ -2651,8 +1589,7 @@ static bool task_can_run_on_remote_rq(struct scx_sched *sch, if (!scx_rq_online(rq)) { if (enforce) - __scx_add_event(scx_root, - SCX_EV_DISPATCH_LOCAL_DSQ_OFFLINE, 1); + __scx_add_event(sch, SCX_EV_DISPATCH_LOCAL_DSQ_OFFLINE, 1); return false; } @@ -2754,7 +1691,7 @@ static struct rq *move_task_between_dsqs(struct scx_sched *sch, dst_rq = container_of(dst_dsq, struct rq, scx.local_dsq); if (src_rq != dst_rq && unlikely(!task_can_run_on_remote_rq(sch, p, dst_rq, true))) { - dst_dsq = find_global_dsq(p); + dst_dsq = find_global_dsq(sch, p); dst_rq = src_rq; } } else { @@ -2910,7 +1847,7 @@ static void dispatch_to_local_dsq(struct scx_sched *sch, struct rq *rq, if (src_rq != dst_rq && unlikely(!task_can_run_on_remote_rq(sch, p, dst_rq, true))) { - dispatch_enqueue(sch, find_global_dsq(p), p, + dispatch_enqueue(sch, find_global_dsq(sch, p), p, enq_flags | SCX_ENQ_CLEAR_OPSS); return; } @@ -3155,10 +2092,10 @@ static int balance_one(struct rq *rq, struct task_struct *prev) * balance(), we want to complete this scheduling cycle and then * start a new one. IOW, we want to call resched_curr() on the * next, most likely idle, task, not the current one. Use - * scx_bpf_kick_cpu() for deferred kicking. + * scx_kick_cpu() for deferred kicking. */ if (unlikely(!--nr_loops)) { - scx_bpf_kick_cpu(cpu_of(rq), 0); + scx_kick_cpu(sch, cpu_of(rq), 0); break; } } while (dspc->nr_tasks); @@ -3442,24 +2379,25 @@ static struct task_struct *pick_task_scx(struct rq *rq) if (keep_prev) { p = prev; if (!p->scx.slice) - refill_task_slice_dfl(p); + refill_task_slice_dfl(rcu_dereference_sched(scx_root), p); } else { p = first_local_task(rq); if (!p) { if (kick_idle) - scx_bpf_kick_cpu(cpu_of(rq), SCX_KICK_IDLE); + scx_kick_cpu(rcu_dereference_sched(scx_root), + cpu_of(rq), SCX_KICK_IDLE); return NULL; } if (unlikely(!p->scx.slice)) { - struct scx_sched *sch = scx_root; + struct scx_sched *sch = rcu_dereference_sched(scx_root); if (!scx_rq_bypassing(rq) && !sch->warned_zero_slice) { printk_deferred(KERN_WARNING "sched_ext: %s[%d] has zero slice in %s()\n", p->comm, p->pid, __func__); sch->warned_zero_slice = true; } - refill_task_slice_dfl(p); + refill_task_slice_dfl(sch, p); } } @@ -3548,7 +2486,7 @@ static int select_task_rq_scx(struct task_struct *p, int prev_cpu, int wake_flag cpu = scx_select_cpu_dfl(p, prev_cpu, wake_flags, NULL, 0); if (cpu >= 0) { - refill_task_slice_dfl(p); + refill_task_slice_dfl(sch, p); p->scx.ddsp_dsq_id = SCX_DSQ_LOCAL; } else { cpu = prev_cpu; @@ -4084,7 +3022,7 @@ bool scx_can_stop_tick(struct rq *rq) #ifdef CONFIG_EXT_GROUP_SCHED -DEFINE_STATIC_PERCPU_RWSEM(scx_cgroup_rwsem); +DEFINE_STATIC_PERCPU_RWSEM(scx_cgroup_ops_rwsem); static bool scx_cgroup_enabled; void scx_tg_init(struct task_group *tg) @@ -4101,8 +3039,6 @@ int scx_tg_online(struct task_group *tg) WARN_ON_ONCE(tg->scx.flags & (SCX_TG_ONLINE | SCX_TG_INITED)); - percpu_down_read(&scx_cgroup_rwsem); - if (scx_cgroup_enabled) { if (SCX_HAS_OP(sch, cgroup_init)) { struct scx_cgroup_init_args args = @@ -4122,7 +3058,6 @@ int scx_tg_online(struct task_group *tg) tg->scx.flags |= SCX_TG_ONLINE; } - percpu_up_read(&scx_cgroup_rwsem); return ret; } @@ -4132,15 +3067,11 @@ void scx_tg_offline(struct task_group *tg) WARN_ON_ONCE(!(tg->scx.flags & SCX_TG_ONLINE)); - percpu_down_read(&scx_cgroup_rwsem); - if (scx_cgroup_enabled && SCX_HAS_OP(sch, cgroup_exit) && (tg->scx.flags & SCX_TG_INITED)) SCX_CALL_OP(sch, SCX_KF_UNLOCKED, cgroup_exit, NULL, tg->css.cgroup); tg->scx.flags &= ~(SCX_TG_ONLINE | SCX_TG_INITED); - - percpu_up_read(&scx_cgroup_rwsem); } int scx_cgroup_can_attach(struct cgroup_taskset *tset) @@ -4150,9 +3081,6 @@ int scx_cgroup_can_attach(struct cgroup_taskset *tset) struct task_struct *p; int ret; - /* released in scx_finish/cancel_attach() */ - percpu_down_read(&scx_cgroup_rwsem); - if (!scx_cgroup_enabled) return 0; @@ -4192,7 +3120,6 @@ err: p->scx.cgrp_moving_from = NULL; } - percpu_up_read(&scx_cgroup_rwsem); return ops_sanitize_err(sch, "cgroup_prep_move", ret); } @@ -4215,11 +3142,6 @@ void scx_cgroup_move_task(struct task_struct *p) p->scx.cgrp_moving_from = NULL; } -void scx_cgroup_finish_attach(void) -{ - percpu_up_read(&scx_cgroup_rwsem); -} - void scx_cgroup_cancel_attach(struct cgroup_taskset *tset) { struct scx_sched *sch = scx_root; @@ -4227,7 +3149,7 @@ void scx_cgroup_cancel_attach(struct cgroup_taskset *tset) struct task_struct *p; if (!scx_cgroup_enabled) - goto out_unlock; + return; cgroup_taskset_for_each(p, css, tset) { if (SCX_HAS_OP(sch, cgroup_cancel_move) && @@ -4236,15 +3158,13 @@ void scx_cgroup_cancel_attach(struct cgroup_taskset *tset) p, p->scx.cgrp_moving_from, css->cgroup); p->scx.cgrp_moving_from = NULL; } -out_unlock: - percpu_up_read(&scx_cgroup_rwsem); } void scx_group_set_weight(struct task_group *tg, unsigned long weight) { struct scx_sched *sch = scx_root; - percpu_down_read(&scx_cgroup_rwsem); + percpu_down_read(&scx_cgroup_ops_rwsem); if (scx_cgroup_enabled && SCX_HAS_OP(sch, cgroup_set_weight) && tg->scx.weight != weight) @@ -4253,7 +3173,7 @@ void scx_group_set_weight(struct task_group *tg, unsigned long weight) tg->scx.weight = weight; - percpu_up_read(&scx_cgroup_rwsem); + percpu_up_read(&scx_cgroup_ops_rwsem); } void scx_group_set_idle(struct task_group *tg, bool idle) @@ -4266,7 +3186,7 @@ void scx_group_set_bandwidth(struct task_group *tg, { struct scx_sched *sch = scx_root; - percpu_down_read(&scx_cgroup_rwsem); + percpu_down_read(&scx_cgroup_ops_rwsem); if (scx_cgroup_enabled && SCX_HAS_OP(sch, cgroup_set_bandwidth) && (tg->scx.bw_period_us != period_us || @@ -4279,23 +3199,25 @@ void scx_group_set_bandwidth(struct task_group *tg, tg->scx.bw_quota_us = quota_us; tg->scx.bw_burst_us = burst_us; - percpu_up_read(&scx_cgroup_rwsem); + percpu_up_read(&scx_cgroup_ops_rwsem); } static void scx_cgroup_lock(void) { - percpu_down_write(&scx_cgroup_rwsem); + percpu_down_write(&scx_cgroup_ops_rwsem); + cgroup_lock(); } static void scx_cgroup_unlock(void) { - percpu_up_write(&scx_cgroup_rwsem); + cgroup_unlock(); + percpu_up_write(&scx_cgroup_ops_rwsem); } #else /* CONFIG_EXT_GROUP_SCHED */ -static inline void scx_cgroup_lock(void) {} -static inline void scx_cgroup_unlock(void) {} +static void scx_cgroup_lock(void) {} +static void scx_cgroup_unlock(void) {} #endif /* CONFIG_EXT_GROUP_SCHED */ @@ -4411,15 +3333,12 @@ static void scx_cgroup_exit(struct scx_sched *sch) { struct cgroup_subsys_state *css; - percpu_rwsem_assert_held(&scx_cgroup_rwsem); - scx_cgroup_enabled = false; /* - * scx_tg_on/offline() are excluded through scx_cgroup_rwsem. If we walk + * scx_tg_on/offline() are excluded through cgroup_lock(). If we walk * cgroups and exit all the inited ones, all online cgroups are exited. */ - rcu_read_lock(); css_for_each_descendant_post(css, &root_task_group.css) { struct task_group *tg = css_tg(css); @@ -4430,17 +3349,9 @@ static void scx_cgroup_exit(struct scx_sched *sch) if (!sch->ops.cgroup_exit) continue; - if (WARN_ON_ONCE(!css_tryget(css))) - continue; - rcu_read_unlock(); - SCX_CALL_OP(sch, SCX_KF_UNLOCKED, cgroup_exit, NULL, css->cgroup); - - rcu_read_lock(); - css_put(css); } - rcu_read_unlock(); } static int scx_cgroup_init(struct scx_sched *sch) @@ -4448,13 +3359,10 @@ static int scx_cgroup_init(struct scx_sched *sch) struct cgroup_subsys_state *css; int ret; - percpu_rwsem_assert_held(&scx_cgroup_rwsem); - /* - * scx_tg_on/offline() are excluded through scx_cgroup_rwsem. If we walk + * scx_tg_on/offline() are excluded through cgroup_lock(). If we walk * cgroups and init, all online cgroups are initialized. */ - rcu_read_lock(); css_for_each_descendant_pre(css, &root_task_group.css) { struct task_group *tg = css_tg(css); struct scx_cgroup_init_args args = { @@ -4473,10 +3381,6 @@ static int scx_cgroup_init(struct scx_sched *sch) continue; } - if (WARN_ON_ONCE(!css_tryget(css))) - continue; - rcu_read_unlock(); - ret = SCX_CALL_OP_RET(sch, SCX_KF_UNLOCKED, cgroup_init, NULL, css->cgroup, &args); if (ret) { @@ -4485,11 +3389,7 @@ static int scx_cgroup_init(struct scx_sched *sch) return ret; } tg->scx.flags |= SCX_TG_INITED; - - rcu_read_lock(); - css_put(css); } - rcu_read_unlock(); WARN_ON_ONCE(scx_cgroup_enabled); scx_cgroup_enabled = true; @@ -4572,7 +3472,7 @@ static void scx_sched_free_rcu_work(struct work_struct *work) int node; kthread_stop(sch->helper->task); - free_percpu(sch->event_stats_cpu); + free_percpu(sch->pcpu); for_each_node_state(node, N_POSSIBLE) kfree(sch->global_dsqs[node]); @@ -4671,9 +3571,22 @@ bool task_should_scx(int policy) bool scx_allow_ttwu_queue(const struct task_struct *p) { - return !scx_enabled() || - (scx_root->ops.flags & SCX_OPS_ALLOW_QUEUED_WAKEUP) || - p->sched_class != &ext_sched_class; + struct scx_sched *sch; + + if (!scx_enabled()) + return true; + + sch = rcu_dereference_sched(scx_root); + if (unlikely(!sch)) + return true; + + if (sch->ops.flags & SCX_OPS_ALLOW_QUEUED_WAKEUP) + return true; + + if (unlikely(p->sched_class != &ext_sched_class)) + return true; + + return false; } /** @@ -4789,7 +3702,7 @@ static void scx_clear_softlockup(void) * * - pick_next_task() suppresses zero slice warning. * - * - scx_bpf_kick_cpu() is disabled to avoid irq_work malfunction during PM + * - scx_kick_cpu() is disabled to avoid irq_work malfunction during PM * operations. * * - scx_prio_less() reverts to the default core_sched_at order. @@ -5234,7 +4147,8 @@ static void scx_dump_task(struct seq_buf *s, struct scx_dump_ctx *dctx, p->scx.sticky_cpu, p->scx.holding_cpu, dsq_id_buf); dump_line(s, " dsq_vtime=%llu slice=%llu weight=%u", p->scx.dsq_vtime, p->scx.slice, p->scx.weight); - dump_line(s, " cpus=%*pb", cpumask_pr_args(p->cpus_ptr)); + dump_line(s, " cpus=%*pb no_mig=%u", cpumask_pr_args(p->cpus_ptr), + p->migration_disabled); if (SCX_HAS_OP(sch, dump_task)) { ops_dump_init(s, " "); @@ -5473,13 +4387,13 @@ static struct scx_sched *scx_alloc_and_add_sched(struct sched_ext_ops *ops) sch->global_dsqs[node] = dsq; } - sch->event_stats_cpu = alloc_percpu(struct scx_event_stats); - if (!sch->event_stats_cpu) + sch->pcpu = alloc_percpu(struct scx_sched_pcpu); + if (!sch->pcpu) goto err_free_gdsqs; sch->helper = kthread_run_worker(0, "sched_ext_helper"); if (!sch->helper) - goto err_free_event_stats; + goto err_free_pcpu; sched_set_fifo(sch->helper->task); atomic_set(&sch->exit_kind, SCX_EXIT_NONE); @@ -5497,8 +4411,8 @@ static struct scx_sched *scx_alloc_and_add_sched(struct sched_ext_ops *ops) err_stop_helper: kthread_stop(sch->helper->task); -err_free_event_stats: - free_percpu(sch->event_stats_cpu); +err_free_pcpu: + free_percpu(sch->pcpu); err_free_gdsqs: for_each_node_state(node, N_POSSIBLE) kfree(sch->global_dsqs[node]); @@ -5621,6 +4535,7 @@ static int scx_enable(struct sched_ext_ops *ops, struct bpf_link *link) scx_error(sch, "ops.init() failed (%d)", ret); goto err_disable; } + sch->exit_info->flags |= SCX_EFLAG_INITIALIZED; } for (i = SCX_OPI_CPU_HOTPLUG_BEGIN; i < SCX_OPI_CPU_HOTPLUG_END; i++) @@ -5713,7 +4628,6 @@ static int scx_enable(struct sched_ext_ops *ops, struct bpf_link *link) ret = scx_init_task(p, task_group(p), false); if (ret) { put_task_struct(p); - scx_task_iter_relock(&sti); scx_task_iter_stop(&sti); scx_error(sch, "ops.init_task() failed (%d) for %s[%d]", ret, p->comm, p->pid); @@ -5723,7 +4637,6 @@ static int scx_enable(struct sched_ext_ops *ops, struct bpf_link *link) scx_set_task_state(p, SCX_TASK_READY); put_task_struct(p); - scx_task_iter_relock(&sti); } scx_task_iter_stop(&sti); scx_cgroup_unlock(); @@ -5749,6 +4662,9 @@ static int scx_enable(struct sched_ext_ops *ops, struct bpf_link *link) __setscheduler_class(p->policy, p->prio); struct sched_enq_and_set_ctx ctx; + if (!tryget_task_struct(p)) + continue; + if (old_class != new_class && p->se.sched_delayed) dequeue_task(task_rq(p), p, DEQUEUE_SLEEP | DEQUEUE_DELAYED); @@ -5761,6 +4677,7 @@ static int scx_enable(struct sched_ext_ops *ops, struct bpf_link *link) sched_enq_and_set_task(&ctx); check_class_changed(task_rq(p), p, old_class, p->prio); + put_task_struct(p); } scx_task_iter_stop(&sti); percpu_up_write(&scx_fork_rwsem); @@ -5791,7 +4708,7 @@ err_unlock: err_disable_unlock_all: scx_cgroup_unlock(); percpu_up_write(&scx_fork_rwsem); - scx_bypass(false); + /* we'll soon enter disable path, keep bypass on */ err_disable: mutex_unlock(&scx_enable_mutex); /* @@ -6324,40 +5241,41 @@ void __init init_sched_ext_class(void) /******************************************************************************** * Helpers that can be called from the BPF scheduler. */ -static bool scx_dsq_insert_preamble(struct task_struct *p, u64 enq_flags) +static bool scx_dsq_insert_preamble(struct scx_sched *sch, struct task_struct *p, + u64 enq_flags) { - if (!scx_kf_allowed(SCX_KF_ENQUEUE | SCX_KF_DISPATCH)) + if (!scx_kf_allowed(sch, SCX_KF_ENQUEUE | SCX_KF_DISPATCH)) return false; lockdep_assert_irqs_disabled(); if (unlikely(!p)) { - scx_kf_error("called with NULL task"); + scx_error(sch, "called with NULL task"); return false; } if (unlikely(enq_flags & __SCX_ENQ_INTERNAL_MASK)) { - scx_kf_error("invalid enq_flags 0x%llx", enq_flags); + scx_error(sch, "invalid enq_flags 0x%llx", enq_flags); return false; } return true; } -static void scx_dsq_insert_commit(struct task_struct *p, u64 dsq_id, - u64 enq_flags) +static void scx_dsq_insert_commit(struct scx_sched *sch, struct task_struct *p, + u64 dsq_id, u64 enq_flags) { struct scx_dsp_ctx *dspc = this_cpu_ptr(scx_dsp_ctx); struct task_struct *ddsp_task; ddsp_task = __this_cpu_read(direct_dispatch_task); if (ddsp_task) { - mark_direct_dispatch(ddsp_task, p, dsq_id, enq_flags); + mark_direct_dispatch(sch, ddsp_task, p, dsq_id, enq_flags); return; } if (unlikely(dspc->cursor >= scx_dsp_max_batch)) { - scx_kf_error("dispatch buffer overflow"); + scx_error(sch, "dispatch buffer overflow"); return; } @@ -6409,7 +5327,14 @@ __bpf_kfunc_start_defs(); __bpf_kfunc void scx_bpf_dsq_insert(struct task_struct *p, u64 dsq_id, u64 slice, u64 enq_flags) { - if (!scx_dsq_insert_preamble(p, enq_flags)) + struct scx_sched *sch; + + guard(rcu)(); + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return; + + if (!scx_dsq_insert_preamble(sch, p, enq_flags)) return; if (slice) @@ -6417,7 +5342,7 @@ __bpf_kfunc void scx_bpf_dsq_insert(struct task_struct *p, u64 dsq_id, u64 slice else p->scx.slice = p->scx.slice ?: 1; - scx_dsq_insert_commit(p, dsq_id, enq_flags); + scx_dsq_insert_commit(sch, p, dsq_id, enq_flags); } /** @@ -6444,7 +5369,14 @@ __bpf_kfunc void scx_bpf_dsq_insert(struct task_struct *p, u64 dsq_id, u64 slice __bpf_kfunc void scx_bpf_dsq_insert_vtime(struct task_struct *p, u64 dsq_id, u64 slice, u64 vtime, u64 enq_flags) { - if (!scx_dsq_insert_preamble(p, enq_flags)) + struct scx_sched *sch; + + guard(rcu)(); + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return; + + if (!scx_dsq_insert_preamble(sch, p, enq_flags)) return; if (slice) @@ -6454,7 +5386,7 @@ __bpf_kfunc void scx_bpf_dsq_insert_vtime(struct task_struct *p, u64 dsq_id, p->scx.dsq_vtime = vtime; - scx_dsq_insert_commit(p, dsq_id, enq_flags | SCX_ENQ_DSQ_PRIQ); + scx_dsq_insert_commit(sch, p, dsq_id, enq_flags | SCX_ENQ_DSQ_PRIQ); } __bpf_kfunc_end_defs(); @@ -6479,7 +5411,8 @@ static bool scx_dsq_move(struct bpf_iter_scx_dsq_kern *kit, bool in_balance; unsigned long flags; - if (!scx_kf_allowed_if_unlocked() && !scx_kf_allowed(SCX_KF_DISPATCH)) + if (!scx_kf_allowed_if_unlocked() && + !scx_kf_allowed(sch, SCX_KF_DISPATCH)) return false; /* @@ -6564,7 +5497,15 @@ __bpf_kfunc_start_defs(); */ __bpf_kfunc u32 scx_bpf_dispatch_nr_slots(void) { - if (!scx_kf_allowed(SCX_KF_DISPATCH)) + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return 0; + + if (!scx_kf_allowed(sch, SCX_KF_DISPATCH)) return 0; return scx_dsp_max_batch - __this_cpu_read(scx_dsp_ctx->cursor); @@ -6579,14 +5520,21 @@ __bpf_kfunc u32 scx_bpf_dispatch_nr_slots(void) __bpf_kfunc void scx_bpf_dispatch_cancel(void) { struct scx_dsp_ctx *dspc = this_cpu_ptr(scx_dsp_ctx); + struct scx_sched *sch; + + guard(rcu)(); - if (!scx_kf_allowed(SCX_KF_DISPATCH)) + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return; + + if (!scx_kf_allowed(sch, SCX_KF_DISPATCH)) return; if (dspc->cursor > 0) dspc->cursor--; else - scx_kf_error("dispatch buffer underflow"); + scx_error(sch, "dispatch buffer underflow"); } /** @@ -6605,11 +5553,17 @@ __bpf_kfunc void scx_bpf_dispatch_cancel(void) */ __bpf_kfunc bool scx_bpf_dsq_move_to_local(u64 dsq_id) { - struct scx_sched *sch = scx_root; struct scx_dsp_ctx *dspc = this_cpu_ptr(scx_dsp_ctx); struct scx_dispatch_q *dsq; + struct scx_sched *sch; - if (!scx_kf_allowed(SCX_KF_DISPATCH)) + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return false; + + if (!scx_kf_allowed(sch, SCX_KF_DISPATCH)) return false; flush_dispatch_buf(sch, dspc->rq); @@ -6756,12 +5710,18 @@ __bpf_kfunc_start_defs(); */ __bpf_kfunc u32 scx_bpf_reenqueue_local(void) { + struct scx_sched *sch; LIST_HEAD(tasks); u32 nr_enqueued = 0; struct rq *rq; struct task_struct *p, *n; - if (!scx_kf_allowed(SCX_KF_CPU_RELEASE)) + guard(rcu)(); + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return 0; + + if (!scx_kf_allowed(sch, SCX_KF_CPU_RELEASE)) return 0; rq = cpu_rq(smp_processor_id()); @@ -6784,12 +5744,8 @@ __bpf_kfunc u32 scx_bpf_reenqueue_local(void) * CPUs disagree, they use %ENQUEUE_RESTORE which is bypassed to * the current local DSQ for running tasks and thus are not * visible to the BPF scheduler. - * - * Also skip re-enqueueing tasks that can only run on this - * CPU, as they would just be re-added to the same local - * DSQ without any benefit. */ - if (p->migration_pending || is_migration_disabled(p) || p->nr_cpus_allowed == 1) + if (p->migration_pending) continue; dispatch_dequeue(rq, p); @@ -6877,22 +5833,12 @@ static const struct btf_kfunc_id_set scx_kfunc_set_unlocked = { __bpf_kfunc_start_defs(); -/** - * scx_bpf_kick_cpu - Trigger reschedule on a CPU - * @cpu: cpu to kick - * @flags: %SCX_KICK_* flags - * - * Kick @cpu into rescheduling. This can be used to wake up an idle CPU or - * trigger rescheduling on a busy CPU. This can be called from any online - * scx_ops operation and the actual kicking is performed asynchronously through - * an irq work. - */ -__bpf_kfunc void scx_bpf_kick_cpu(s32 cpu, u64 flags) +static void scx_kick_cpu(struct scx_sched *sch, s32 cpu, u64 flags) { struct rq *this_rq; unsigned long irq_flags; - if (!kf_cpu_valid(cpu, NULL)) + if (!ops_cpu_valid(sch, cpu, NULL)) return; local_irq_save(irq_flags); @@ -6916,7 +5862,7 @@ __bpf_kfunc void scx_bpf_kick_cpu(s32 cpu, u64 flags) struct rq *target_rq = cpu_rq(cpu); if (unlikely(flags & (SCX_KICK_PREEMPT | SCX_KICK_WAIT))) - scx_kf_error("PREEMPT/WAIT cannot be used with SCX_KICK_IDLE"); + scx_error(sch, "PREEMPT/WAIT cannot be used with SCX_KICK_IDLE"); if (raw_spin_rq_trylock(target_rq)) { if (can_skip_idle_kick(target_rq)) { @@ -6941,6 +5887,26 @@ out: } /** + * scx_bpf_kick_cpu - Trigger reschedule on a CPU + * @cpu: cpu to kick + * @flags: %SCX_KICK_* flags + * + * Kick @cpu into rescheduling. This can be used to wake up an idle CPU or + * trigger rescheduling on a busy CPU. This can be called from any online + * scx_ops operation and the actual kicking is performed asynchronously through + * an irq work. + */ +__bpf_kfunc void scx_bpf_kick_cpu(s32 cpu, u64 flags) +{ + struct scx_sched *sch; + + guard(rcu)(); + sch = rcu_dereference(scx_root); + if (likely(sch)) + scx_kick_cpu(sch, cpu, flags); +} + +/** * scx_bpf_dsq_nr_queued - Return the number of queued tasks * @dsq_id: id of the DSQ * @@ -7120,28 +6086,29 @@ __bpf_kfunc void bpf_iter_scx_dsq_destroy(struct bpf_iter_scx_dsq *it) __bpf_kfunc_end_defs(); -static s32 __bstr_format(u64 *data_buf, char *line_buf, size_t line_size, - char *fmt, unsigned long long *data, u32 data__sz) +static s32 __bstr_format(struct scx_sched *sch, u64 *data_buf, char *line_buf, + size_t line_size, char *fmt, unsigned long long *data, + u32 data__sz) { struct bpf_bprintf_data bprintf_data = { .get_bin_args = true }; s32 ret; if (data__sz % 8 || data__sz > MAX_BPRINTF_VARARGS * 8 || (data__sz && !data)) { - scx_kf_error("invalid data=%p and data__sz=%u", (void *)data, data__sz); + scx_error(sch, "invalid data=%p and data__sz=%u", (void *)data, data__sz); return -EINVAL; } ret = copy_from_kernel_nofault(data_buf, data, data__sz); if (ret < 0) { - scx_kf_error("failed to read data fields (%d)", ret); + scx_error(sch, "failed to read data fields (%d)", ret); return ret; } ret = bpf_bprintf_prepare(fmt, UINT_MAX, data_buf, data__sz / 8, &bprintf_data); if (ret < 0) { - scx_kf_error("format preparation failed (%d)", ret); + scx_error(sch, "format preparation failed (%d)", ret); return ret; } @@ -7149,17 +6116,17 @@ static s32 __bstr_format(u64 *data_buf, char *line_buf, size_t line_size, bprintf_data.bin_args); bpf_bprintf_cleanup(&bprintf_data); if (ret < 0) { - scx_kf_error("(\"%s\", %p, %u) failed to format", fmt, data, data__sz); + scx_error(sch, "(\"%s\", %p, %u) failed to format", fmt, data, data__sz); return ret; } return ret; } -static s32 bstr_format(struct scx_bstr_buf *buf, +static s32 bstr_format(struct scx_sched *sch, struct scx_bstr_buf *buf, char *fmt, unsigned long long *data, u32 data__sz) { - return __bstr_format(buf->data, buf->line, sizeof(buf->line), + return __bstr_format(sch, buf->data, buf->line, sizeof(buf->line), fmt, data, data__sz); } @@ -7178,11 +6145,14 @@ __bpf_kfunc_start_defs(); __bpf_kfunc void scx_bpf_exit_bstr(s64 exit_code, char *fmt, unsigned long long *data, u32 data__sz) { + struct scx_sched *sch; unsigned long flags; raw_spin_lock_irqsave(&scx_exit_bstr_buf_lock, flags); - if (bstr_format(&scx_exit_bstr_buf, fmt, data, data__sz) >= 0) - scx_kf_exit(SCX_EXIT_UNREG_BPF, exit_code, "%s", scx_exit_bstr_buf.line); + sch = rcu_dereference_bh(scx_root); + if (likely(sch) && + bstr_format(sch, &scx_exit_bstr_buf, fmt, data, data__sz) >= 0) + scx_exit(sch, SCX_EXIT_UNREG_BPF, exit_code, "%s", scx_exit_bstr_buf.line); raw_spin_unlock_irqrestore(&scx_exit_bstr_buf_lock, flags); } @@ -7198,11 +6168,14 @@ __bpf_kfunc void scx_bpf_exit_bstr(s64 exit_code, char *fmt, __bpf_kfunc void scx_bpf_error_bstr(char *fmt, unsigned long long *data, u32 data__sz) { + struct scx_sched *sch; unsigned long flags; raw_spin_lock_irqsave(&scx_exit_bstr_buf_lock, flags); - if (bstr_format(&scx_exit_bstr_buf, fmt, data, data__sz) >= 0) - scx_kf_exit(SCX_EXIT_ERROR_BPF, 0, "%s", scx_exit_bstr_buf.line); + sch = rcu_dereference_bh(scx_root); + if (likely(sch) && + bstr_format(sch, &scx_exit_bstr_buf, fmt, data, data__sz) >= 0) + scx_exit(sch, SCX_EXIT_ERROR_BPF, 0, "%s", scx_exit_bstr_buf.line); raw_spin_unlock_irqrestore(&scx_exit_bstr_buf_lock, flags); } @@ -7221,17 +6194,24 @@ __bpf_kfunc void scx_bpf_error_bstr(char *fmt, unsigned long long *data, __bpf_kfunc void scx_bpf_dump_bstr(char *fmt, unsigned long long *data, u32 data__sz) { + struct scx_sched *sch; struct scx_dump_data *dd = &scx_dump_data; struct scx_bstr_buf *buf = &dd->buf; s32 ret; + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return; + if (raw_smp_processor_id() != dd->cpu) { - scx_kf_error("scx_bpf_dump() must only be called from ops.dump() and friends"); + scx_error(sch, "scx_bpf_dump() must only be called from ops.dump() and friends"); return; } /* append the formatted string to the line buf */ - ret = __bstr_format(buf->data, buf->line + dd->cursor, + ret = __bstr_format(sch, buf->data, buf->line + dd->cursor, sizeof(buf->line) - dd->cursor, fmt, data, data__sz); if (ret < 0) { dump_line(dd->s, "%s[!] (\"%s\", %p, %u) failed to format (%d)", @@ -7267,7 +6247,12 @@ __bpf_kfunc void scx_bpf_dump_bstr(char *fmt, unsigned long long *data, */ __bpf_kfunc u32 scx_bpf_cpuperf_cap(s32 cpu) { - if (kf_cpu_valid(cpu, NULL)) + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (likely(sch) && ops_cpu_valid(sch, cpu, NULL)) return arch_scale_cpu_capacity(cpu); else return SCX_CPUPERF_ONE; @@ -7289,7 +6274,12 @@ __bpf_kfunc u32 scx_bpf_cpuperf_cap(s32 cpu) */ __bpf_kfunc u32 scx_bpf_cpuperf_cur(s32 cpu) { - if (kf_cpu_valid(cpu, NULL)) + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (likely(sch) && ops_cpu_valid(sch, cpu, NULL)) return arch_scale_freq_capacity(cpu); else return SCX_CPUPERF_ONE; @@ -7311,12 +6301,20 @@ __bpf_kfunc u32 scx_bpf_cpuperf_cur(s32 cpu) */ __bpf_kfunc void scx_bpf_cpuperf_set(s32 cpu, u32 perf) { + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(sch); + if (unlikely(!sch)) + return; + if (unlikely(perf > SCX_CPUPERF_ONE)) { - scx_kf_error("Invalid cpuperf target %u for CPU %d", perf, cpu); + scx_error(sch, "Invalid cpuperf target %u for CPU %d", perf, cpu); return; } - if (kf_cpu_valid(cpu, NULL)) { + if (ops_cpu_valid(sch, cpu, NULL)) { struct rq *rq = cpu_rq(cpu), *locked_rq = scx_locked_rq(); struct rq_flags rf; @@ -7325,7 +6323,7 @@ __bpf_kfunc void scx_bpf_cpuperf_set(s32 cpu, u32 perf) * to the corresponding CPU to prevent ABBA deadlocks. */ if (locked_rq && rq != locked_rq) { - scx_kf_error("Invalid target CPU %d", cpu); + scx_error(sch, "Invalid target CPU %d", cpu); return; } @@ -7420,13 +6418,76 @@ __bpf_kfunc s32 scx_bpf_task_cpu(const struct task_struct *p) */ __bpf_kfunc struct rq *scx_bpf_cpu_rq(s32 cpu) { - if (!kf_cpu_valid(cpu, NULL)) + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) return NULL; + if (!ops_cpu_valid(sch, cpu, NULL)) + return NULL; + + if (!sch->warned_deprecated_rq) { + printk_deferred(KERN_WARNING "sched_ext: %s() is deprecated; " + "use scx_bpf_locked_rq() when holding rq lock " + "or scx_bpf_cpu_curr() to read remote curr safely.\n", __func__); + sch->warned_deprecated_rq = true; + } + return cpu_rq(cpu); } /** + * scx_bpf_locked_rq - Return the rq currently locked by SCX + * + * Returns the rq if a rq lock is currently held by SCX. + * Otherwise emits an error and returns NULL. + */ +__bpf_kfunc struct rq *scx_bpf_locked_rq(void) +{ + struct scx_sched *sch; + struct rq *rq; + + guard(preempt)(); + + sch = rcu_dereference_sched(scx_root); + if (unlikely(!sch)) + return NULL; + + rq = scx_locked_rq(); + if (!rq) { + scx_error(sch, "accessing rq without holding rq lock"); + return NULL; + } + + return rq; +} + +/** + * scx_bpf_cpu_curr - Return remote CPU's curr task + * @cpu: CPU of interest + * + * Callers must hold RCU read lock (KF_RCU). + */ +__bpf_kfunc struct task_struct *scx_bpf_cpu_curr(s32 cpu) +{ + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return NULL; + + if (!ops_cpu_valid(sch, cpu, NULL)) + return NULL; + + return rcu_dereference(cpu_rq(cpu)->curr); +} + +/** * scx_bpf_task_cgroup - Return the sched cgroup of a task * @p: task of interest * @@ -7442,8 +6503,15 @@ __bpf_kfunc struct cgroup *scx_bpf_task_cgroup(struct task_struct *p) { struct task_group *tg = p->sched_task_group; struct cgroup *cgrp = &cgrp_dfl_root.cgrp; + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + goto out; - if (!scx_kf_allowed_on_arg_tasks(__SCX_KF_RQ_LOCKED, p)) + if (!scx_kf_allowed_on_arg_tasks(sch, __SCX_KF_RQ_LOCKED, p)) goto out; cgrp = tg_cgrp(tg); @@ -7524,7 +6592,7 @@ static void scx_read_events(struct scx_sched *sch, struct scx_event_stats *event /* Aggregate per-CPU event counters into @events. */ memset(events, 0, sizeof(*events)); for_each_possible_cpu(cpu) { - e_cpu = per_cpu_ptr(sch->event_stats_cpu, cpu); + e_cpu = &per_cpu_ptr(sch->pcpu, cpu)->event_stats; scx_agg_event(events, e_cpu, SCX_EV_SELECT_CPU_FALLBACK); scx_agg_event(events, e_cpu, SCX_EV_DISPATCH_LOCAL_DSQ_OFFLINE); scx_agg_event(events, e_cpu, SCX_EV_DISPATCH_KEEP_LAST); @@ -7590,6 +6658,8 @@ BTF_ID_FLAGS(func, scx_bpf_put_cpumask, KF_RELEASE) BTF_ID_FLAGS(func, scx_bpf_task_running, KF_RCU) BTF_ID_FLAGS(func, scx_bpf_task_cpu, KF_RCU) BTF_ID_FLAGS(func, scx_bpf_cpu_rq) +BTF_ID_FLAGS(func, scx_bpf_locked_rq, KF_RET_NULL) +BTF_ID_FLAGS(func, scx_bpf_cpu_curr, KF_RET_NULL | KF_RCU_PROTECTED) #ifdef CONFIG_CGROUP_SCHED BTF_ID_FLAGS(func, scx_bpf_task_cgroup, KF_RCU | KF_ACQUIRE) #endif diff --git a/kernel/sched/ext.h b/kernel/sched/ext.h index 292bb41a242e..43429b33e52c 100644 --- a/kernel/sched/ext.h +++ b/kernel/sched/ext.h @@ -8,29 +8,6 @@ */ #ifdef CONFIG_SCHED_CLASS_EXT -static inline bool scx_kf_allowed_if_unlocked(void) -{ - return !current->scx.kf_mask; -} - -static inline bool scx_rq_bypassing(struct rq *rq) -{ - return unlikely(rq->scx.flags & SCX_RQ_BYPASSING); -} - -DECLARE_STATIC_KEY_FALSE(scx_ops_allow_queued_wakeup); - -DECLARE_PER_CPU(struct rq *, scx_locked_rq_state); - -/* - * Return the rq currently locked from an scx callback, or NULL if no rq is - * locked. - */ -static inline struct rq *scx_locked_rq(void) -{ - return __this_cpu_read(scx_locked_rq_state); -} - void scx_tick(struct rq *rq); void init_scx_entity(struct sched_ext_entity *scx); void scx_pre_fork(struct task_struct *p); @@ -100,7 +77,6 @@ int scx_tg_online(struct task_group *tg); void scx_tg_offline(struct task_group *tg); int scx_cgroup_can_attach(struct cgroup_taskset *tset); void scx_cgroup_move_task(struct task_struct *p); -void scx_cgroup_finish_attach(void); void scx_cgroup_cancel_attach(struct cgroup_taskset *tset); void scx_group_set_weight(struct task_group *tg, unsigned long cgrp_weight); void scx_group_set_idle(struct task_group *tg, bool idle); @@ -111,7 +87,6 @@ static inline int scx_tg_online(struct task_group *tg) { return 0; } static inline void scx_tg_offline(struct task_group *tg) {} static inline int scx_cgroup_can_attach(struct cgroup_taskset *tset) { return 0; } static inline void scx_cgroup_move_task(struct task_struct *p) {} -static inline void scx_cgroup_finish_attach(void) {} static inline void scx_cgroup_cancel_attach(struct cgroup_taskset *tset) {} static inline void scx_group_set_weight(struct task_group *tg, unsigned long cgrp_weight) {} static inline void scx_group_set_idle(struct task_group *tg, bool idle) {} diff --git a/kernel/sched/ext_idle.c b/kernel/sched/ext_idle.c index 7174e1c1a392..d2434c954848 100644 --- a/kernel/sched/ext_idle.c +++ b/kernel/sched/ext_idle.c @@ -819,10 +819,10 @@ void scx_idle_disable(void) * Helpers that can be called from the BPF scheduler. */ -static int validate_node(int node) +static int validate_node(struct scx_sched *sch, int node) { if (!static_branch_likely(&scx_builtin_idle_per_node)) { - scx_kf_error("per-node idle tracking is disabled"); + scx_error(sch, "per-node idle tracking is disabled"); return -EOPNOTSUPP; } @@ -832,13 +832,13 @@ static int validate_node(int node) /* Make sure node is in a valid range */ if (node < 0 || node >= nr_node_ids) { - scx_kf_error("invalid node %d", node); + scx_error(sch, "invalid node %d", node); return -EINVAL; } /* Make sure the node is part of the set of possible nodes */ if (!node_possible(node)) { - scx_kf_error("unavailable node %d", node); + scx_error(sch, "unavailable node %d", node); return -EINVAL; } @@ -847,26 +847,53 @@ static int validate_node(int node) __bpf_kfunc_start_defs(); -static bool check_builtin_idle_enabled(void) +static bool check_builtin_idle_enabled(struct scx_sched *sch) { if (static_branch_likely(&scx_builtin_idle_enabled)) return true; - scx_kf_error("built-in idle tracking is disabled"); + scx_error(sch, "built-in idle tracking is disabled"); return false; } -static s32 select_cpu_from_kfunc(struct task_struct *p, s32 prev_cpu, u64 wake_flags, +/* + * Determine whether @p is a migration-disabled task in the context of BPF + * code. + * + * We can't simply check whether @p->migration_disabled is set in a + * sched_ext callback, because migration is always disabled for the current + * task while running BPF code. + * + * The prolog (__bpf_prog_enter) and epilog (__bpf_prog_exit) respectively + * disable and re-enable migration. For this reason, the current task + * inside a sched_ext callback is always a migration-disabled task. + * + * Therefore, when @p->migration_disabled == 1, check whether @p is the + * current task or not: if it is, then migration was not disabled before + * entering the callback, otherwise migration was disabled. + * + * Returns true if @p is migration-disabled, false otherwise. + */ +static bool is_bpf_migration_disabled(const struct task_struct *p) +{ + if (p->migration_disabled == 1) + return p != current; + else + return p->migration_disabled; +} + +static s32 select_cpu_from_kfunc(struct scx_sched *sch, struct task_struct *p, + s32 prev_cpu, u64 wake_flags, const struct cpumask *allowed, u64 flags) { struct rq *rq; struct rq_flags rf; s32 cpu; - if (!kf_cpu_valid(prev_cpu, NULL)) + if (!ops_cpu_valid(sch, prev_cpu, NULL)) return -EINVAL; - if (!check_builtin_idle_enabled()) + if (!check_builtin_idle_enabled(sch)) return -EBUSY; /* @@ -879,7 +906,7 @@ static s32 select_cpu_from_kfunc(struct task_struct *p, s32 prev_cpu, u64 wake_f if (scx_kf_allowed_if_unlocked()) { rq = task_rq_lock(p, &rf); } else { - if (!scx_kf_allowed(SCX_KF_SELECT_CPU | SCX_KF_ENQUEUE)) + if (!scx_kf_allowed(sch, SCX_KF_SELECT_CPU | SCX_KF_ENQUEUE)) return -EPERM; rq = scx_locked_rq(); } @@ -898,7 +925,7 @@ static s32 select_cpu_from_kfunc(struct task_struct *p, s32 prev_cpu, u64 wake_f * selection optimizations and simply check whether the previously * used CPU is idle and within the allowed cpumask. */ - if (p->nr_cpus_allowed == 1 || is_migration_disabled(p)) { + if (p->nr_cpus_allowed == 1 || is_bpf_migration_disabled(p)) { if (cpumask_test_cpu(prev_cpu, allowed ?: p->cpus_ptr) && scx_idle_test_and_clear_cpu(prev_cpu)) cpu = prev_cpu; @@ -922,9 +949,13 @@ static s32 select_cpu_from_kfunc(struct task_struct *p, s32 prev_cpu, u64 wake_f */ __bpf_kfunc int scx_bpf_cpu_node(s32 cpu) { - if (!kf_cpu_valid(cpu, NULL)) - return NUMA_NO_NODE; + struct scx_sched *sch; + + guard(rcu)(); + sch = rcu_dereference(scx_root); + if (unlikely(!sch) || !ops_cpu_valid(sch, cpu, NULL)) + return NUMA_NO_NODE; return cpu_to_node(cpu); } @@ -946,15 +977,21 @@ __bpf_kfunc int scx_bpf_cpu_node(s32 cpu) __bpf_kfunc s32 scx_bpf_select_cpu_dfl(struct task_struct *p, s32 prev_cpu, u64 wake_flags, bool *is_idle) { + struct scx_sched *sch; s32 cpu; - cpu = select_cpu_from_kfunc(p, prev_cpu, wake_flags, NULL, 0); + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return -ENODEV; + + cpu = select_cpu_from_kfunc(sch, p, prev_cpu, wake_flags, NULL, 0); if (cpu >= 0) { *is_idle = true; return cpu; } *is_idle = false; - return prev_cpu; } @@ -981,7 +1018,16 @@ __bpf_kfunc s32 scx_bpf_select_cpu_dfl(struct task_struct *p, s32 prev_cpu, __bpf_kfunc s32 scx_bpf_select_cpu_and(struct task_struct *p, s32 prev_cpu, u64 wake_flags, const struct cpumask *cpus_allowed, u64 flags) { - return select_cpu_from_kfunc(p, prev_cpu, wake_flags, cpus_allowed, flags); + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return -ENODEV; + + return select_cpu_from_kfunc(sch, p, prev_cpu, wake_flags, + cpus_allowed, flags); } /** @@ -995,7 +1041,15 @@ __bpf_kfunc s32 scx_bpf_select_cpu_and(struct task_struct *p, s32 prev_cpu, u64 */ __bpf_kfunc const struct cpumask *scx_bpf_get_idle_cpumask_node(int node) { - node = validate_node(node); + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return cpu_none_mask; + + node = validate_node(sch, node); if (node < 0) return cpu_none_mask; @@ -1011,12 +1065,20 @@ __bpf_kfunc const struct cpumask *scx_bpf_get_idle_cpumask_node(int node) */ __bpf_kfunc const struct cpumask *scx_bpf_get_idle_cpumask(void) { + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return cpu_none_mask; + if (static_branch_unlikely(&scx_builtin_idle_per_node)) { - scx_kf_error("SCX_OPS_BUILTIN_IDLE_PER_NODE enabled"); + scx_error(sch, "SCX_OPS_BUILTIN_IDLE_PER_NODE enabled"); return cpu_none_mask; } - if (!check_builtin_idle_enabled()) + if (!check_builtin_idle_enabled(sch)) return cpu_none_mask; return idle_cpumask(NUMA_NO_NODE)->cpu; @@ -1034,7 +1096,15 @@ __bpf_kfunc const struct cpumask *scx_bpf_get_idle_cpumask(void) */ __bpf_kfunc const struct cpumask *scx_bpf_get_idle_smtmask_node(int node) { - node = validate_node(node); + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return cpu_none_mask; + + node = validate_node(sch, node); if (node < 0) return cpu_none_mask; @@ -1054,12 +1124,20 @@ __bpf_kfunc const struct cpumask *scx_bpf_get_idle_smtmask_node(int node) */ __bpf_kfunc const struct cpumask *scx_bpf_get_idle_smtmask(void) { + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return cpu_none_mask; + if (static_branch_unlikely(&scx_builtin_idle_per_node)) { - scx_kf_error("SCX_OPS_BUILTIN_IDLE_PER_NODE enabled"); + scx_error(sch, "SCX_OPS_BUILTIN_IDLE_PER_NODE enabled"); return cpu_none_mask; } - if (!check_builtin_idle_enabled()) + if (!check_builtin_idle_enabled(sch)) return cpu_none_mask; if (sched_smt_active()) @@ -1095,10 +1173,18 @@ __bpf_kfunc void scx_bpf_put_idle_cpumask(const struct cpumask *idle_mask) */ __bpf_kfunc bool scx_bpf_test_and_clear_cpu_idle(s32 cpu) { - if (!check_builtin_idle_enabled()) + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) return false; - if (!kf_cpu_valid(cpu, NULL)) + if (!check_builtin_idle_enabled(sch)) + return false; + + if (!ops_cpu_valid(sch, cpu, NULL)) return false; return scx_idle_test_and_clear_cpu(cpu); @@ -1126,7 +1212,15 @@ __bpf_kfunc bool scx_bpf_test_and_clear_cpu_idle(s32 cpu) __bpf_kfunc s32 scx_bpf_pick_idle_cpu_node(const struct cpumask *cpus_allowed, int node, u64 flags) { - node = validate_node(node); + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return -ENODEV; + + node = validate_node(sch, node); if (node < 0) return node; @@ -1158,12 +1252,20 @@ __bpf_kfunc s32 scx_bpf_pick_idle_cpu_node(const struct cpumask *cpus_allowed, __bpf_kfunc s32 scx_bpf_pick_idle_cpu(const struct cpumask *cpus_allowed, u64 flags) { + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return -ENODEV; + if (static_branch_maybe(CONFIG_NUMA, &scx_builtin_idle_per_node)) { - scx_kf_error("per-node idle tracking is enabled"); + scx_error(sch, "per-node idle tracking is enabled"); return -EBUSY; } - if (!check_builtin_idle_enabled()) + if (!check_builtin_idle_enabled(sch)) return -EBUSY; return scx_pick_idle_cpu(cpus_allowed, NUMA_NO_NODE, flags); @@ -1193,9 +1295,16 @@ __bpf_kfunc s32 scx_bpf_pick_idle_cpu(const struct cpumask *cpus_allowed, __bpf_kfunc s32 scx_bpf_pick_any_cpu_node(const struct cpumask *cpus_allowed, int node, u64 flags) { + struct scx_sched *sch; s32 cpu; - node = validate_node(node); + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return -ENODEV; + + node = validate_node(sch, node); if (node < 0) return node; @@ -1233,10 +1342,17 @@ __bpf_kfunc s32 scx_bpf_pick_any_cpu_node(const struct cpumask *cpus_allowed, __bpf_kfunc s32 scx_bpf_pick_any_cpu(const struct cpumask *cpus_allowed, u64 flags) { + struct scx_sched *sch; s32 cpu; + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return -ENODEV; + if (static_branch_maybe(CONFIG_NUMA, &scx_builtin_idle_per_node)) { - scx_kf_error("per-node idle tracking is enabled"); + scx_error(sch, "per-node idle tracking is enabled"); return -EBUSY; } diff --git a/kernel/sched/ext_internal.h b/kernel/sched/ext_internal.h new file mode 100644 index 000000000000..b3617abed510 --- /dev/null +++ b/kernel/sched/ext_internal.h @@ -0,0 +1,1078 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * BPF extensible scheduler class: Documentation/scheduler/sched-ext.rst + * + * Copyright (c) 2025 Meta Platforms, Inc. and affiliates. + * Copyright (c) 2025 Tejun Heo <tj@kernel.org> + */ +#define SCX_OP_IDX(op) (offsetof(struct sched_ext_ops, op) / sizeof(void (*)(void))) + +enum scx_consts { + SCX_DSP_DFL_MAX_BATCH = 32, + SCX_DSP_MAX_LOOPS = 32, + SCX_WATCHDOG_MAX_TIMEOUT = 30 * HZ, + + SCX_EXIT_BT_LEN = 64, + SCX_EXIT_MSG_LEN = 1024, + SCX_EXIT_DUMP_DFL_LEN = 32768, + + SCX_CPUPERF_ONE = SCHED_CAPACITY_SCALE, + + /* + * Iterating all tasks may take a while. Periodically drop + * scx_tasks_lock to avoid causing e.g. CSD and RCU stalls. + */ + SCX_TASK_ITER_BATCH = 32, +}; + +enum scx_exit_kind { + SCX_EXIT_NONE, + SCX_EXIT_DONE, + + SCX_EXIT_UNREG = 64, /* user-space initiated unregistration */ + SCX_EXIT_UNREG_BPF, /* BPF-initiated unregistration */ + SCX_EXIT_UNREG_KERN, /* kernel-initiated unregistration */ + SCX_EXIT_SYSRQ, /* requested by 'S' sysrq */ + + SCX_EXIT_ERROR = 1024, /* runtime error, error msg contains details */ + SCX_EXIT_ERROR_BPF, /* ERROR but triggered through scx_bpf_error() */ + SCX_EXIT_ERROR_STALL, /* watchdog detected stalled runnable tasks */ +}; + +/* + * An exit code can be specified when exiting with scx_bpf_exit() or scx_exit(), + * corresponding to exit_kind UNREG_BPF and UNREG_KERN respectively. The codes + * are 64bit of the format: + * + * Bits: [63 .. 48 47 .. 32 31 .. 0] + * [ SYS ACT ] [ SYS RSN ] [ USR ] + * + * SYS ACT: System-defined exit actions + * SYS RSN: System-defined exit reasons + * USR : User-defined exit codes and reasons + * + * Using the above, users may communicate intention and context by ORing system + * actions and/or system reasons with a user-defined exit code. + */ +enum scx_exit_code { + /* Reasons */ + SCX_ECODE_RSN_HOTPLUG = 1LLU << 32, + + /* Actions */ + SCX_ECODE_ACT_RESTART = 1LLU << 48, +}; + +enum scx_exit_flags { + /* + * ops.exit() may be called even if the loading failed before ops.init() + * finishes successfully. This is because ops.exit() allows rich exit + * info communication. The following flag indicates whether ops.init() + * finished successfully. + */ + SCX_EFLAG_INITIALIZED, +}; + +/* + * scx_exit_info is passed to ops.exit() to describe why the BPF scheduler is + * being disabled. + */ +struct scx_exit_info { + /* %SCX_EXIT_* - broad category of the exit reason */ + enum scx_exit_kind kind; + + /* exit code if gracefully exiting */ + s64 exit_code; + + /* %SCX_EFLAG_* */ + u64 flags; + + /* textual representation of the above */ + const char *reason; + + /* backtrace if exiting due to an error */ + unsigned long *bt; + u32 bt_len; + + /* informational message */ + char *msg; + + /* debug dump */ + char *dump; +}; + +/* sched_ext_ops.flags */ +enum scx_ops_flags { + /* + * Keep built-in idle tracking even if ops.update_idle() is implemented. + */ + SCX_OPS_KEEP_BUILTIN_IDLE = 1LLU << 0, + + /* + * By default, if there are no other task to run on the CPU, ext core + * keeps running the current task even after its slice expires. If this + * flag is specified, such tasks are passed to ops.enqueue() with + * %SCX_ENQ_LAST. See the comment above %SCX_ENQ_LAST for more info. + */ + SCX_OPS_ENQ_LAST = 1LLU << 1, + + /* + * An exiting task may schedule after PF_EXITING is set. In such cases, + * bpf_task_from_pid() may not be able to find the task and if the BPF + * scheduler depends on pid lookup for dispatching, the task will be + * lost leading to various issues including RCU grace period stalls. + * + * To mask this problem, by default, unhashed tasks are automatically + * dispatched to the local DSQ on enqueue. If the BPF scheduler doesn't + * depend on pid lookups and wants to handle these tasks directly, the + * following flag can be used. + */ + SCX_OPS_ENQ_EXITING = 1LLU << 2, + + /* + * If set, only tasks with policy set to SCHED_EXT are attached to + * sched_ext. If clear, SCHED_NORMAL tasks are also included. + */ + SCX_OPS_SWITCH_PARTIAL = 1LLU << 3, + + /* + * A migration disabled task can only execute on its current CPU. By + * default, such tasks are automatically put on the CPU's local DSQ with + * the default slice on enqueue. If this ops flag is set, they also go + * through ops.enqueue(). + * + * A migration disabled task never invokes ops.select_cpu() as it can + * only select the current CPU. Also, p->cpus_ptr will only contain its + * current CPU while p->nr_cpus_allowed keeps tracking p->user_cpus_ptr + * and thus may disagree with cpumask_weight(p->cpus_ptr). + */ + SCX_OPS_ENQ_MIGRATION_DISABLED = 1LLU << 4, + + /* + * Queued wakeup (ttwu_queue) is a wakeup optimization that invokes + * ops.enqueue() on the ops.select_cpu() selected or the wakee's + * previous CPU via IPI (inter-processor interrupt) to reduce cacheline + * transfers. When this optimization is enabled, ops.select_cpu() is + * skipped in some cases (when racing against the wakee switching out). + * As the BPF scheduler may depend on ops.select_cpu() being invoked + * during wakeups, queued wakeup is disabled by default. + * + * If this ops flag is set, queued wakeup optimization is enabled and + * the BPF scheduler must be able to handle ops.enqueue() invoked on the + * wakee's CPU without preceding ops.select_cpu() even for tasks which + * may be executed on multiple CPUs. + */ + SCX_OPS_ALLOW_QUEUED_WAKEUP = 1LLU << 5, + + /* + * If set, enable per-node idle cpumasks. If clear, use a single global + * flat idle cpumask. + */ + SCX_OPS_BUILTIN_IDLE_PER_NODE = 1LLU << 6, + + /* + * CPU cgroup support flags + */ + SCX_OPS_HAS_CGROUP_WEIGHT = 1LLU << 16, /* DEPRECATED, will be removed on 6.18 */ + + SCX_OPS_ALL_FLAGS = SCX_OPS_KEEP_BUILTIN_IDLE | + SCX_OPS_ENQ_LAST | + SCX_OPS_ENQ_EXITING | + SCX_OPS_ENQ_MIGRATION_DISABLED | + SCX_OPS_ALLOW_QUEUED_WAKEUP | + SCX_OPS_SWITCH_PARTIAL | + SCX_OPS_BUILTIN_IDLE_PER_NODE | + SCX_OPS_HAS_CGROUP_WEIGHT, + + /* high 8 bits are internal, don't include in SCX_OPS_ALL_FLAGS */ + __SCX_OPS_INTERNAL_MASK = 0xffLLU << 56, + + SCX_OPS_HAS_CPU_PREEMPT = 1LLU << 56, +}; + +/* argument container for ops.init_task() */ +struct scx_init_task_args { + /* + * Set if ops.init_task() is being invoked on the fork path, as opposed + * to the scheduler transition path. + */ + bool fork; +#ifdef CONFIG_EXT_GROUP_SCHED + /* the cgroup the task is joining */ + struct cgroup *cgroup; +#endif +}; + +/* argument container for ops.exit_task() */ +struct scx_exit_task_args { + /* Whether the task exited before running on sched_ext. */ + bool cancelled; +}; + +/* argument container for ops->cgroup_init() */ +struct scx_cgroup_init_args { + /* the weight of the cgroup [1..10000] */ + u32 weight; + + /* bandwidth control parameters from cpu.max and cpu.max.burst */ + u64 bw_period_us; + u64 bw_quota_us; + u64 bw_burst_us; +}; + +enum scx_cpu_preempt_reason { + /* next task is being scheduled by &sched_class_rt */ + SCX_CPU_PREEMPT_RT, + /* next task is being scheduled by &sched_class_dl */ + SCX_CPU_PREEMPT_DL, + /* next task is being scheduled by &sched_class_stop */ + SCX_CPU_PREEMPT_STOP, + /* unknown reason for SCX being preempted */ + SCX_CPU_PREEMPT_UNKNOWN, +}; + +/* + * Argument container for ops->cpu_acquire(). Currently empty, but may be + * expanded in the future. + */ +struct scx_cpu_acquire_args {}; + +/* argument container for ops->cpu_release() */ +struct scx_cpu_release_args { + /* the reason the CPU was preempted */ + enum scx_cpu_preempt_reason reason; + + /* the task that's going to be scheduled on the CPU */ + struct task_struct *task; +}; + +/* + * Informational context provided to dump operations. + */ +struct scx_dump_ctx { + enum scx_exit_kind kind; + s64 exit_code; + const char *reason; + u64 at_ns; + u64 at_jiffies; +}; + +/** + * struct sched_ext_ops - Operation table for BPF scheduler implementation + * + * A BPF scheduler can implement an arbitrary scheduling policy by + * implementing and loading operations in this table. Note that a userland + * scheduling policy can also be implemented using the BPF scheduler + * as a shim layer. + */ +struct sched_ext_ops { + /** + * @select_cpu: Pick the target CPU for a task which is being woken up + * @p: task being woken up + * @prev_cpu: the cpu @p was on before sleeping + * @wake_flags: SCX_WAKE_* + * + * Decision made here isn't final. @p may be moved to any CPU while it + * is getting dispatched for execution later. However, as @p is not on + * the rq at this point, getting the eventual execution CPU right here + * saves a small bit of overhead down the line. + * + * If an idle CPU is returned, the CPU is kicked and will try to + * dispatch. While an explicit custom mechanism can be added, + * select_cpu() serves as the default way to wake up idle CPUs. + * + * @p may be inserted into a DSQ directly by calling + * scx_bpf_dsq_insert(). If so, the ops.enqueue() will be skipped. + * Directly inserting into %SCX_DSQ_LOCAL will put @p in the local DSQ + * of the CPU returned by this operation. + * + * Note that select_cpu() is never called for tasks that can only run + * on a single CPU or tasks with migration disabled, as they don't have + * the option to select a different CPU. See select_task_rq() for + * details. + */ + s32 (*select_cpu)(struct task_struct *p, s32 prev_cpu, u64 wake_flags); + + /** + * @enqueue: Enqueue a task on the BPF scheduler + * @p: task being enqueued + * @enq_flags: %SCX_ENQ_* + * + * @p is ready to run. Insert directly into a DSQ by calling + * scx_bpf_dsq_insert() or enqueue on the BPF scheduler. If not directly + * inserted, the bpf scheduler owns @p and if it fails to dispatch @p, + * the task will stall. + * + * If @p was inserted into a DSQ from ops.select_cpu(), this callback is + * skipped. + */ + void (*enqueue)(struct task_struct *p, u64 enq_flags); + + /** + * @dequeue: Remove a task from the BPF scheduler + * @p: task being dequeued + * @deq_flags: %SCX_DEQ_* + * + * Remove @p from the BPF scheduler. This is usually called to isolate + * the task while updating its scheduling properties (e.g. priority). + * + * The ext core keeps track of whether the BPF side owns a given task or + * not and can gracefully ignore spurious dispatches from BPF side, + * which makes it safe to not implement this method. However, depending + * on the scheduling logic, this can lead to confusing behaviors - e.g. + * scheduling position not being updated across a priority change. + */ + void (*dequeue)(struct task_struct *p, u64 deq_flags); + + /** + * @dispatch: Dispatch tasks from the BPF scheduler and/or user DSQs + * @cpu: CPU to dispatch tasks for + * @prev: previous task being switched out + * + * Called when a CPU's local dsq is empty. The operation should dispatch + * one or more tasks from the BPF scheduler into the DSQs using + * scx_bpf_dsq_insert() and/or move from user DSQs into the local DSQ + * using scx_bpf_dsq_move_to_local(). + * + * The maximum number of times scx_bpf_dsq_insert() can be called + * without an intervening scx_bpf_dsq_move_to_local() is specified by + * ops.dispatch_max_batch. See the comments on top of the two functions + * for more details. + * + * When not %NULL, @prev is an SCX task with its slice depleted. If + * @prev is still runnable as indicated by set %SCX_TASK_QUEUED in + * @prev->scx.flags, it is not enqueued yet and will be enqueued after + * ops.dispatch() returns. To keep executing @prev, return without + * dispatching or moving any tasks. Also see %SCX_OPS_ENQ_LAST. + */ + void (*dispatch)(s32 cpu, struct task_struct *prev); + + /** + * @tick: Periodic tick + * @p: task running currently + * + * This operation is called every 1/HZ seconds on CPUs which are + * executing an SCX task. Setting @p->scx.slice to 0 will trigger an + * immediate dispatch cycle on the CPU. + */ + void (*tick)(struct task_struct *p); + + /** + * @runnable: A task is becoming runnable on its associated CPU + * @p: task becoming runnable + * @enq_flags: %SCX_ENQ_* + * + * This and the following three functions can be used to track a task's + * execution state transitions. A task becomes ->runnable() on a CPU, + * and then goes through one or more ->running() and ->stopping() pairs + * as it runs on the CPU, and eventually becomes ->quiescent() when it's + * done running on the CPU. + * + * @p is becoming runnable on the CPU because it's + * + * - waking up (%SCX_ENQ_WAKEUP) + * - being moved from another CPU + * - being restored after temporarily taken off the queue for an + * attribute change. + * + * This and ->enqueue() are related but not coupled. This operation + * notifies @p's state transition and may not be followed by ->enqueue() + * e.g. when @p is being dispatched to a remote CPU, or when @p is + * being enqueued on a CPU experiencing a hotplug event. Likewise, a + * task may be ->enqueue()'d without being preceded by this operation + * e.g. after exhausting its slice. + */ + void (*runnable)(struct task_struct *p, u64 enq_flags); + + /** + * @running: A task is starting to run on its associated CPU + * @p: task starting to run + * + * Note that this callback may be called from a CPU other than the + * one the task is going to run on. This can happen when a task + * property is changed (i.e., affinity), since scx_next_task_scx(), + * which triggers this callback, may run on a CPU different from + * the task's assigned CPU. + * + * Therefore, always use scx_bpf_task_cpu(@p) to determine the + * target CPU the task is going to use. + * + * See ->runnable() for explanation on the task state notifiers. + */ + void (*running)(struct task_struct *p); + + /** + * @stopping: A task is stopping execution + * @p: task stopping to run + * @runnable: is task @p still runnable? + * + * Note that this callback may be called from a CPU other than the + * one the task was running on. This can happen when a task + * property is changed (i.e., affinity), since dequeue_task_scx(), + * which triggers this callback, may run on a CPU different from + * the task's assigned CPU. + * + * Therefore, always use scx_bpf_task_cpu(@p) to retrieve the CPU + * the task was running on. + * + * See ->runnable() for explanation on the task state notifiers. If + * !@runnable, ->quiescent() will be invoked after this operation + * returns. + */ + void (*stopping)(struct task_struct *p, bool runnable); + + /** + * @quiescent: A task is becoming not runnable on its associated CPU + * @p: task becoming not runnable + * @deq_flags: %SCX_DEQ_* + * + * See ->runnable() for explanation on the task state notifiers. + * + * @p is becoming quiescent on the CPU because it's + * + * - sleeping (%SCX_DEQ_SLEEP) + * - being moved to another CPU + * - being temporarily taken off the queue for an attribute change + * (%SCX_DEQ_SAVE) + * + * This and ->dequeue() are related but not coupled. This operation + * notifies @p's state transition and may not be preceded by ->dequeue() + * e.g. when @p is being dispatched to a remote CPU. + */ + void (*quiescent)(struct task_struct *p, u64 deq_flags); + + /** + * @yield: Yield CPU + * @from: yielding task + * @to: optional yield target task + * + * If @to is NULL, @from is yielding the CPU to other runnable tasks. + * The BPF scheduler should ensure that other available tasks are + * dispatched before the yielding task. Return value is ignored in this + * case. + * + * If @to is not-NULL, @from wants to yield the CPU to @to. If the bpf + * scheduler can implement the request, return %true; otherwise, %false. + */ + bool (*yield)(struct task_struct *from, struct task_struct *to); + + /** + * @core_sched_before: Task ordering for core-sched + * @a: task A + * @b: task B + * + * Used by core-sched to determine the ordering between two tasks. See + * Documentation/admin-guide/hw-vuln/core-scheduling.rst for details on + * core-sched. + * + * Both @a and @b are runnable and may or may not currently be queued on + * the BPF scheduler. Should return %true if @a should run before @b. + * %false if there's no required ordering or @b should run before @a. + * + * If not specified, the default is ordering them according to when they + * became runnable. + */ + bool (*core_sched_before)(struct task_struct *a, struct task_struct *b); + + /** + * @set_weight: Set task weight + * @p: task to set weight for + * @weight: new weight [1..10000] + * + * Update @p's weight to @weight. + */ + void (*set_weight)(struct task_struct *p, u32 weight); + + /** + * @set_cpumask: Set CPU affinity + * @p: task to set CPU affinity for + * @cpumask: cpumask of cpus that @p can run on + * + * Update @p's CPU affinity to @cpumask. + */ + void (*set_cpumask)(struct task_struct *p, + const struct cpumask *cpumask); + + /** + * @update_idle: Update the idle state of a CPU + * @cpu: CPU to update the idle state for + * @idle: whether entering or exiting the idle state + * + * This operation is called when @rq's CPU goes or leaves the idle + * state. By default, implementing this operation disables the built-in + * idle CPU tracking and the following helpers become unavailable: + * + * - scx_bpf_select_cpu_dfl() + * - scx_bpf_select_cpu_and() + * - scx_bpf_test_and_clear_cpu_idle() + * - scx_bpf_pick_idle_cpu() + * + * The user also must implement ops.select_cpu() as the default + * implementation relies on scx_bpf_select_cpu_dfl(). + * + * Specify the %SCX_OPS_KEEP_BUILTIN_IDLE flag to keep the built-in idle + * tracking. + */ + void (*update_idle)(s32 cpu, bool idle); + + /** + * @cpu_acquire: A CPU is becoming available to the BPF scheduler + * @cpu: The CPU being acquired by the BPF scheduler. + * @args: Acquire arguments, see the struct definition. + * + * A CPU that was previously released from the BPF scheduler is now once + * again under its control. + */ + void (*cpu_acquire)(s32 cpu, struct scx_cpu_acquire_args *args); + + /** + * @cpu_release: A CPU is taken away from the BPF scheduler + * @cpu: The CPU being released by the BPF scheduler. + * @args: Release arguments, see the struct definition. + * + * The specified CPU is no longer under the control of the BPF + * scheduler. This could be because it was preempted by a higher + * priority sched_class, though there may be other reasons as well. The + * caller should consult @args->reason to determine the cause. + */ + void (*cpu_release)(s32 cpu, struct scx_cpu_release_args *args); + + /** + * @init_task: Initialize a task to run in a BPF scheduler + * @p: task to initialize for BPF scheduling + * @args: init arguments, see the struct definition + * + * Either we're loading a BPF scheduler or a new task is being forked. + * Initialize @p for BPF scheduling. This operation may block and can + * be used for allocations, and is called exactly once for a task. + * + * Return 0 for success, -errno for failure. An error return while + * loading will abort loading of the BPF scheduler. During a fork, it + * will abort that specific fork. + */ + s32 (*init_task)(struct task_struct *p, struct scx_init_task_args *args); + + /** + * @exit_task: Exit a previously-running task from the system + * @p: task to exit + * @args: exit arguments, see the struct definition + * + * @p is exiting or the BPF scheduler is being unloaded. Perform any + * necessary cleanup for @p. + */ + void (*exit_task)(struct task_struct *p, struct scx_exit_task_args *args); + + /** + * @enable: Enable BPF scheduling for a task + * @p: task to enable BPF scheduling for + * + * Enable @p for BPF scheduling. enable() is called on @p any time it + * enters SCX, and is always paired with a matching disable(). + */ + void (*enable)(struct task_struct *p); + + /** + * @disable: Disable BPF scheduling for a task + * @p: task to disable BPF scheduling for + * + * @p is exiting, leaving SCX or the BPF scheduler is being unloaded. + * Disable BPF scheduling for @p. A disable() call is always matched + * with a prior enable() call. + */ + void (*disable)(struct task_struct *p); + + /** + * @dump: Dump BPF scheduler state on error + * @ctx: debug dump context + * + * Use scx_bpf_dump() to generate BPF scheduler specific debug dump. + */ + void (*dump)(struct scx_dump_ctx *ctx); + + /** + * @dump_cpu: Dump BPF scheduler state for a CPU on error + * @ctx: debug dump context + * @cpu: CPU to generate debug dump for + * @idle: @cpu is currently idle without any runnable tasks + * + * Use scx_bpf_dump() to generate BPF scheduler specific debug dump for + * @cpu. If @idle is %true and this operation doesn't produce any + * output, @cpu is skipped for dump. + */ + void (*dump_cpu)(struct scx_dump_ctx *ctx, s32 cpu, bool idle); + + /** + * @dump_task: Dump BPF scheduler state for a runnable task on error + * @ctx: debug dump context + * @p: runnable task to generate debug dump for + * + * Use scx_bpf_dump() to generate BPF scheduler specific debug dump for + * @p. + */ + void (*dump_task)(struct scx_dump_ctx *ctx, struct task_struct *p); + +#ifdef CONFIG_EXT_GROUP_SCHED + /** + * @cgroup_init: Initialize a cgroup + * @cgrp: cgroup being initialized + * @args: init arguments, see the struct definition + * + * Either the BPF scheduler is being loaded or @cgrp created, initialize + * @cgrp for sched_ext. This operation may block. + * + * Return 0 for success, -errno for failure. An error return while + * loading will abort loading of the BPF scheduler. During cgroup + * creation, it will abort the specific cgroup creation. + */ + s32 (*cgroup_init)(struct cgroup *cgrp, + struct scx_cgroup_init_args *args); + + /** + * @cgroup_exit: Exit a cgroup + * @cgrp: cgroup being exited + * + * Either the BPF scheduler is being unloaded or @cgrp destroyed, exit + * @cgrp for sched_ext. This operation my block. + */ + void (*cgroup_exit)(struct cgroup *cgrp); + + /** + * @cgroup_prep_move: Prepare a task to be moved to a different cgroup + * @p: task being moved + * @from: cgroup @p is being moved from + * @to: cgroup @p is being moved to + * + * Prepare @p for move from cgroup @from to @to. This operation may + * block and can be used for allocations. + * + * Return 0 for success, -errno for failure. An error return aborts the + * migration. + */ + s32 (*cgroup_prep_move)(struct task_struct *p, + struct cgroup *from, struct cgroup *to); + + /** + * @cgroup_move: Commit cgroup move + * @p: task being moved + * @from: cgroup @p is being moved from + * @to: cgroup @p is being moved to + * + * Commit the move. @p is dequeued during this operation. + */ + void (*cgroup_move)(struct task_struct *p, + struct cgroup *from, struct cgroup *to); + + /** + * @cgroup_cancel_move: Cancel cgroup move + * @p: task whose cgroup move is being canceled + * @from: cgroup @p was being moved from + * @to: cgroup @p was being moved to + * + * @p was cgroup_prep_move()'d but failed before reaching cgroup_move(). + * Undo the preparation. + */ + void (*cgroup_cancel_move)(struct task_struct *p, + struct cgroup *from, struct cgroup *to); + + /** + * @cgroup_set_weight: A cgroup's weight is being changed + * @cgrp: cgroup whose weight is being updated + * @weight: new weight [1..10000] + * + * Update @cgrp's weight to @weight. + */ + void (*cgroup_set_weight)(struct cgroup *cgrp, u32 weight); + + /** + * @cgroup_set_bandwidth: A cgroup's bandwidth is being changed + * @cgrp: cgroup whose bandwidth is being updated + * @period_us: bandwidth control period + * @quota_us: bandwidth control quota + * @burst_us: bandwidth control burst + * + * Update @cgrp's bandwidth control parameters. This is from the cpu.max + * cgroup interface. + * + * @quota_us / @period_us determines the CPU bandwidth @cgrp is entitled + * to. For example, if @period_us is 1_000_000 and @quota_us is + * 2_500_000. @cgrp is entitled to 2.5 CPUs. @burst_us can be + * interpreted in the same fashion and specifies how much @cgrp can + * burst temporarily. The specific control mechanism and thus the + * interpretation of @period_us and burstiness is upto to the BPF + * scheduler. + */ + void (*cgroup_set_bandwidth)(struct cgroup *cgrp, + u64 period_us, u64 quota_us, u64 burst_us); + +#endif /* CONFIG_EXT_GROUP_SCHED */ + + /* + * All online ops must come before ops.cpu_online(). + */ + + /** + * @cpu_online: A CPU became online + * @cpu: CPU which just came up + * + * @cpu just came online. @cpu will not call ops.enqueue() or + * ops.dispatch(), nor run tasks associated with other CPUs beforehand. + */ + void (*cpu_online)(s32 cpu); + + /** + * @cpu_offline: A CPU is going offline + * @cpu: CPU which is going offline + * + * @cpu is going offline. @cpu will not call ops.enqueue() or + * ops.dispatch(), nor run tasks associated with other CPUs afterwards. + */ + void (*cpu_offline)(s32 cpu); + + /* + * All CPU hotplug ops must come before ops.init(). + */ + + /** + * @init: Initialize the BPF scheduler + */ + s32 (*init)(void); + + /** + * @exit: Clean up after the BPF scheduler + * @info: Exit info + * + * ops.exit() is also called on ops.init() failure, which is a bit + * unusual. This is to allow rich reporting through @info on how + * ops.init() failed. + */ + void (*exit)(struct scx_exit_info *info); + + /** + * @dispatch_max_batch: Max nr of tasks that dispatch() can dispatch + */ + u32 dispatch_max_batch; + + /** + * @flags: %SCX_OPS_* flags + */ + u64 flags; + + /** + * @timeout_ms: The maximum amount of time, in milliseconds, that a + * runnable task should be able to wait before being scheduled. The + * maximum timeout may not exceed the default timeout of 30 seconds. + * + * Defaults to the maximum allowed timeout value of 30 seconds. + */ + u32 timeout_ms; + + /** + * @exit_dump_len: scx_exit_info.dump buffer length. If 0, the default + * value of 32768 is used. + */ + u32 exit_dump_len; + + /** + * @hotplug_seq: A sequence number that may be set by the scheduler to + * detect when a hotplug event has occurred during the loading process. + * If 0, no detection occurs. Otherwise, the scheduler will fail to + * load if the sequence number does not match @scx_hotplug_seq on the + * enable path. + */ + u64 hotplug_seq; + + /** + * @name: BPF scheduler's name + * + * Must be a non-zero valid BPF object name including only isalnum(), + * '_' and '.' chars. Shows up in kernel.sched_ext_ops sysctl while the + * BPF scheduler is enabled. + */ + char name[SCX_OPS_NAME_LEN]; + + /* internal use only, must be NULL */ + void *priv; +}; + +enum scx_opi { + SCX_OPI_BEGIN = 0, + SCX_OPI_NORMAL_BEGIN = 0, + SCX_OPI_NORMAL_END = SCX_OP_IDX(cpu_online), + SCX_OPI_CPU_HOTPLUG_BEGIN = SCX_OP_IDX(cpu_online), + SCX_OPI_CPU_HOTPLUG_END = SCX_OP_IDX(init), + SCX_OPI_END = SCX_OP_IDX(init), +}; + +/* + * Collection of event counters. Event types are placed in descending order. + */ +struct scx_event_stats { + /* + * If ops.select_cpu() returns a CPU which can't be used by the task, + * the core scheduler code silently picks a fallback CPU. + */ + s64 SCX_EV_SELECT_CPU_FALLBACK; + + /* + * When dispatching to a local DSQ, the CPU may have gone offline in + * the meantime. In this case, the task is bounced to the global DSQ. + */ + s64 SCX_EV_DISPATCH_LOCAL_DSQ_OFFLINE; + + /* + * If SCX_OPS_ENQ_LAST is not set, the number of times that a task + * continued to run because there were no other tasks on the CPU. + */ + s64 SCX_EV_DISPATCH_KEEP_LAST; + + /* + * If SCX_OPS_ENQ_EXITING is not set, the number of times that a task + * is dispatched to a local DSQ when exiting. + */ + s64 SCX_EV_ENQ_SKIP_EXITING; + + /* + * If SCX_OPS_ENQ_MIGRATION_DISABLED is not set, the number of times a + * migration disabled task skips ops.enqueue() and is dispatched to its + * local DSQ. + */ + s64 SCX_EV_ENQ_SKIP_MIGRATION_DISABLED; + + /* + * Total number of times a task's time slice was refilled with the + * default value (SCX_SLICE_DFL). + */ + s64 SCX_EV_REFILL_SLICE_DFL; + + /* + * The total duration of bypass modes in nanoseconds. + */ + s64 SCX_EV_BYPASS_DURATION; + + /* + * The number of tasks dispatched in the bypassing mode. + */ + s64 SCX_EV_BYPASS_DISPATCH; + + /* + * The number of times the bypassing mode has been activated. + */ + s64 SCX_EV_BYPASS_ACTIVATE; +}; + +struct scx_sched_pcpu { + /* + * The event counters are in a per-CPU variable to minimize the + * accounting overhead. A system-wide view on the event counter is + * constructed when requested by scx_bpf_events(). + */ + struct scx_event_stats event_stats; +}; + +struct scx_sched { + struct sched_ext_ops ops; + DECLARE_BITMAP(has_op, SCX_OPI_END); + + /* + * Dispatch queues. + * + * The global DSQ (%SCX_DSQ_GLOBAL) is split per-node for scalability. + * This is to avoid live-locking in bypass mode where all tasks are + * dispatched to %SCX_DSQ_GLOBAL and all CPUs consume from it. If + * per-node split isn't sufficient, it can be further split. + */ + struct rhashtable dsq_hash; + struct scx_dispatch_q **global_dsqs; + struct scx_sched_pcpu __percpu *pcpu; + + bool warned_zero_slice:1; + bool warned_deprecated_rq:1; + + atomic_t exit_kind; + struct scx_exit_info *exit_info; + + struct kobject kobj; + + struct kthread_worker *helper; + struct irq_work error_irq_work; + struct kthread_work disable_work; + struct rcu_work rcu_work; +}; + +enum scx_wake_flags { + /* expose select WF_* flags as enums */ + SCX_WAKE_FORK = WF_FORK, + SCX_WAKE_TTWU = WF_TTWU, + SCX_WAKE_SYNC = WF_SYNC, +}; + +enum scx_enq_flags { + /* expose select ENQUEUE_* flags as enums */ + SCX_ENQ_WAKEUP = ENQUEUE_WAKEUP, + SCX_ENQ_HEAD = ENQUEUE_HEAD, + SCX_ENQ_CPU_SELECTED = ENQUEUE_RQ_SELECTED, + + /* high 32bits are SCX specific */ + + /* + * Set the following to trigger preemption when calling + * scx_bpf_dsq_insert() with a local dsq as the target. The slice of the + * current task is cleared to zero and the CPU is kicked into the + * scheduling path. Implies %SCX_ENQ_HEAD. + */ + SCX_ENQ_PREEMPT = 1LLU << 32, + + /* + * The task being enqueued was previously enqueued on the current CPU's + * %SCX_DSQ_LOCAL, but was removed from it in a call to the + * scx_bpf_reenqueue_local() kfunc. If scx_bpf_reenqueue_local() was + * invoked in a ->cpu_release() callback, and the task is again + * dispatched back to %SCX_LOCAL_DSQ by this current ->enqueue(), the + * task will not be scheduled on the CPU until at least the next invocation + * of the ->cpu_acquire() callback. + */ + SCX_ENQ_REENQ = 1LLU << 40, + + /* + * The task being enqueued is the only task available for the cpu. By + * default, ext core keeps executing such tasks but when + * %SCX_OPS_ENQ_LAST is specified, they're ops.enqueue()'d with the + * %SCX_ENQ_LAST flag set. + * + * The BPF scheduler is responsible for triggering a follow-up + * scheduling event. Otherwise, Execution may stall. + */ + SCX_ENQ_LAST = 1LLU << 41, + + /* high 8 bits are internal */ + __SCX_ENQ_INTERNAL_MASK = 0xffLLU << 56, + + SCX_ENQ_CLEAR_OPSS = 1LLU << 56, + SCX_ENQ_DSQ_PRIQ = 1LLU << 57, +}; + +enum scx_deq_flags { + /* expose select DEQUEUE_* flags as enums */ + SCX_DEQ_SLEEP = DEQUEUE_SLEEP, + + /* high 32bits are SCX specific */ + + /* + * The generic core-sched layer decided to execute the task even though + * it hasn't been dispatched yet. Dequeue from the BPF side. + */ + SCX_DEQ_CORE_SCHED_EXEC = 1LLU << 32, +}; + +enum scx_pick_idle_cpu_flags { + SCX_PICK_IDLE_CORE = 1LLU << 0, /* pick a CPU whose SMT siblings are also idle */ + SCX_PICK_IDLE_IN_NODE = 1LLU << 1, /* pick a CPU in the same target NUMA node */ +}; + +enum scx_kick_flags { + /* + * Kick the target CPU if idle. Guarantees that the target CPU goes + * through at least one full scheduling cycle before going idle. If the + * target CPU can be determined to be currently not idle and going to go + * through a scheduling cycle before going idle, noop. + */ + SCX_KICK_IDLE = 1LLU << 0, + + /* + * Preempt the current task and execute the dispatch path. If the + * current task of the target CPU is an SCX task, its ->scx.slice is + * cleared to zero before the scheduling path is invoked so that the + * task expires and the dispatch path is invoked. + */ + SCX_KICK_PREEMPT = 1LLU << 1, + + /* + * Wait for the CPU to be rescheduled. The scx_bpf_kick_cpu() call will + * return after the target CPU finishes picking the next task. + */ + SCX_KICK_WAIT = 1LLU << 2, +}; + +enum scx_tg_flags { + SCX_TG_ONLINE = 1U << 0, + SCX_TG_INITED = 1U << 1, +}; + +enum scx_enable_state { + SCX_ENABLING, + SCX_ENABLED, + SCX_DISABLING, + SCX_DISABLED, +}; + +static const char *scx_enable_state_str[] = { + [SCX_ENABLING] = "enabling", + [SCX_ENABLED] = "enabled", + [SCX_DISABLING] = "disabling", + [SCX_DISABLED] = "disabled", +}; + +/* + * sched_ext_entity->ops_state + * + * Used to track the task ownership between the SCX core and the BPF scheduler. + * State transitions look as follows: + * + * NONE -> QUEUEING -> QUEUED -> DISPATCHING + * ^ | | + * | v v + * \-------------------------------/ + * + * QUEUEING and DISPATCHING states can be waited upon. See wait_ops_state() call + * sites for explanations on the conditions being waited upon and why they are + * safe. Transitions out of them into NONE or QUEUED must store_release and the + * waiters should load_acquire. + * + * Tracking scx_ops_state enables sched_ext core to reliably determine whether + * any given task can be dispatched by the BPF scheduler at all times and thus + * relaxes the requirements on the BPF scheduler. This allows the BPF scheduler + * to try to dispatch any task anytime regardless of its state as the SCX core + * can safely reject invalid dispatches. + */ +enum scx_ops_state { + SCX_OPSS_NONE, /* owned by the SCX core */ + SCX_OPSS_QUEUEING, /* in transit to the BPF scheduler */ + SCX_OPSS_QUEUED, /* owned by the BPF scheduler */ + SCX_OPSS_DISPATCHING, /* in transit back to the SCX core */ + + /* + * QSEQ brands each QUEUED instance so that, when dispatch races + * dequeue/requeue, the dispatcher can tell whether it still has a claim + * on the task being dispatched. + * + * As some 32bit archs can't do 64bit store_release/load_acquire, + * p->scx.ops_state is atomic_long_t which leaves 30 bits for QSEQ on + * 32bit machines. The dispatch race window QSEQ protects is very narrow + * and runs with IRQ disabled. 30 bits should be sufficient. + */ + SCX_OPSS_QSEQ_SHIFT = 2, +}; + +/* Use macros to ensure that the type is unsigned long for the masks */ +#define SCX_OPSS_STATE_MASK ((1LU << SCX_OPSS_QSEQ_SHIFT) - 1) +#define SCX_OPSS_QSEQ_MASK (~SCX_OPSS_STATE_MASK) + +DECLARE_PER_CPU(struct rq *, scx_locked_rq_state); + +/* + * Return the rq currently locked from an scx callback, or NULL if no rq is + * locked. + */ +static inline struct rq *scx_locked_rq(void) +{ + return __this_cpu_read(scx_locked_rq_state); +} + +static inline bool scx_kf_allowed_if_unlocked(void) +{ + return !current->scx.kf_mask; +} + +static inline bool scx_rq_bypassing(struct rq *rq) +{ + return unlikely(rq->scx.flags & SCX_RQ_BYPASSING); +} diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index b173a059315c..3a89f949e307 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -3542,7 +3542,7 @@ out: } } -void init_numa_balancing(unsigned long clone_flags, struct task_struct *p) +void init_numa_balancing(u64 clone_flags, struct task_struct *p) { int mm_users = 0; struct mm_struct *mm = p->mm; @@ -3957,9 +3957,6 @@ static void update_cfs_group(struct sched_entity *se) if (!gcfs_rq || !gcfs_rq->load.weight) return; - if (throttled_hierarchy(gcfs_rq)) - return; - shares = calc_group_shares(gcfs_rq); if (unlikely(se->load.weight != shares)) reweight_entity(cfs_rq_of(se), se, shares); @@ -5291,18 +5288,16 @@ enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags) if (cfs_rq->nr_queued == 1) { check_enqueue_throttle(cfs_rq); - if (!throttled_hierarchy(cfs_rq)) { - list_add_leaf_cfs_rq(cfs_rq); - } else { + list_add_leaf_cfs_rq(cfs_rq); #ifdef CONFIG_CFS_BANDWIDTH + if (cfs_rq->pelt_clock_throttled) { struct rq *rq = rq_of(cfs_rq); - if (cfs_rq_throttled(cfs_rq) && !cfs_rq->throttled_clock) - cfs_rq->throttled_clock = rq_clock(rq); - if (!cfs_rq->throttled_clock_self) - cfs_rq->throttled_clock_self = rq_clock(rq); -#endif + cfs_rq->throttled_clock_pelt_time += rq_clock_pelt(rq) - + cfs_rq->throttled_clock_pelt; + cfs_rq->pelt_clock_throttled = 0; } +#endif } } @@ -5341,8 +5336,6 @@ static void set_delayed(struct sched_entity *se) struct cfs_rq *cfs_rq = cfs_rq_of(se); cfs_rq->h_nr_runnable--; - if (cfs_rq_throttled(cfs_rq)) - break; } } @@ -5363,8 +5356,6 @@ static void clear_delayed(struct sched_entity *se) struct cfs_rq *cfs_rq = cfs_rq_of(se); cfs_rq->h_nr_runnable++; - if (cfs_rq_throttled(cfs_rq)) - break; } } @@ -5392,7 +5383,7 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags) * DELAY_DEQUEUE relies on spurious wakeups, special task * states must not suffer spurious wakeups, excempt them. */ - if (flags & DEQUEUE_SPECIAL) + if (flags & (DEQUEUE_SPECIAL | DEQUEUE_THROTTLE)) delay = false; WARN_ON_ONCE(delay && se->sched_delayed); @@ -5450,8 +5441,18 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags) if (flags & DEQUEUE_DELAYED) finish_delayed_dequeue_entity(se); - if (cfs_rq->nr_queued == 0) + if (cfs_rq->nr_queued == 0) { update_idle_cfs_rq_clock_pelt(cfs_rq); +#ifdef CONFIG_CFS_BANDWIDTH + if (throttled_hierarchy(cfs_rq)) { + struct rq *rq = rq_of(cfs_rq); + + list_del_leaf_cfs_rq(cfs_rq); + cfs_rq->throttled_clock_pelt = rq_clock_pelt(rq); + cfs_rq->pelt_clock_throttled = 1; + } +#endif + } return true; } @@ -5725,74 +5726,253 @@ static inline int cfs_rq_throttled(struct cfs_rq *cfs_rq) return cfs_bandwidth_used() && cfs_rq->throttled; } +static inline bool cfs_rq_pelt_clock_throttled(struct cfs_rq *cfs_rq) +{ + return cfs_bandwidth_used() && cfs_rq->pelt_clock_throttled; +} + /* check whether cfs_rq, or any parent, is throttled */ static inline int throttled_hierarchy(struct cfs_rq *cfs_rq) { return cfs_bandwidth_used() && cfs_rq->throttle_count; } +static inline int lb_throttled_hierarchy(struct task_struct *p, int dst_cpu) +{ + return throttled_hierarchy(task_group(p)->cfs_rq[dst_cpu]); +} + +static inline bool task_is_throttled(struct task_struct *p) +{ + return cfs_bandwidth_used() && p->throttled; +} + +static bool dequeue_task_fair(struct rq *rq, struct task_struct *p, int flags); +static void throttle_cfs_rq_work(struct callback_head *work) +{ + struct task_struct *p = container_of(work, struct task_struct, sched_throttle_work); + struct sched_entity *se; + struct cfs_rq *cfs_rq; + struct rq *rq; + + WARN_ON_ONCE(p != current); + p->sched_throttle_work.next = &p->sched_throttle_work; + + /* + * If task is exiting, then there won't be a return to userspace, so we + * don't have to bother with any of this. + */ + if ((p->flags & PF_EXITING)) + return; + + scoped_guard(task_rq_lock, p) { + se = &p->se; + cfs_rq = cfs_rq_of(se); + + /* Raced, forget */ + if (p->sched_class != &fair_sched_class) + return; + + /* + * If not in limbo, then either replenish has happened or this + * task got migrated out of the throttled cfs_rq, move along. + */ + if (!cfs_rq->throttle_count) + return; + rq = scope.rq; + update_rq_clock(rq); + WARN_ON_ONCE(p->throttled || !list_empty(&p->throttle_node)); + dequeue_task_fair(rq, p, DEQUEUE_SLEEP | DEQUEUE_THROTTLE); + list_add(&p->throttle_node, &cfs_rq->throttled_limbo_list); + /* + * Must not set throttled before dequeue or dequeue will + * mistakenly regard this task as an already throttled one. + */ + p->throttled = true; + resched_curr(rq); + } +} + +void init_cfs_throttle_work(struct task_struct *p) +{ + init_task_work(&p->sched_throttle_work, throttle_cfs_rq_work); + /* Protect against double add, see throttle_cfs_rq() and throttle_cfs_rq_work() */ + p->sched_throttle_work.next = &p->sched_throttle_work; + INIT_LIST_HEAD(&p->throttle_node); +} + /* - * Ensure that neither of the group entities corresponding to src_cpu or - * dest_cpu are members of a throttled hierarchy when performing group - * load-balance operations. + * Task is throttled and someone wants to dequeue it again: + * it could be sched/core when core needs to do things like + * task affinity change, task group change, task sched class + * change etc. and in these cases, DEQUEUE_SLEEP is not set; + * or the task is blocked after throttled due to freezer etc. + * and in these cases, DEQUEUE_SLEEP is set. */ -static inline int throttled_lb_pair(struct task_group *tg, - int src_cpu, int dest_cpu) +static void detach_task_cfs_rq(struct task_struct *p); +static void dequeue_throttled_task(struct task_struct *p, int flags) +{ + WARN_ON_ONCE(p->se.on_rq); + list_del_init(&p->throttle_node); + + /* task blocked after throttled */ + if (flags & DEQUEUE_SLEEP) { + p->throttled = false; + return; + } + + /* + * task is migrating off its old cfs_rq, detach + * the task's load from its old cfs_rq. + */ + if (task_on_rq_migrating(p)) + detach_task_cfs_rq(p); +} + +static bool enqueue_throttled_task(struct task_struct *p) { - struct cfs_rq *src_cfs_rq, *dest_cfs_rq; + struct cfs_rq *cfs_rq = cfs_rq_of(&p->se); - src_cfs_rq = tg->cfs_rq[src_cpu]; - dest_cfs_rq = tg->cfs_rq[dest_cpu]; + /* @p should have gone through dequeue_throttled_task() first */ + WARN_ON_ONCE(!list_empty(&p->throttle_node)); + + /* + * If the throttled task @p is enqueued to a throttled cfs_rq, + * take the fast path by directly putting the task on the + * target cfs_rq's limbo list. + * + * Do not do that when @p is current because the following race can + * cause @p's group_node to be incorectly re-insterted in its rq's + * cfs_tasks list, despite being throttled: + * + * cpuX cpuY + * p ret2user + * throttle_cfs_rq_work() sched_move_task(p) + * LOCK task_rq_lock + * dequeue_task_fair(p) + * UNLOCK task_rq_lock + * LOCK task_rq_lock + * task_current_donor(p) == true + * task_on_rq_queued(p) == true + * dequeue_task(p) + * put_prev_task(p) + * sched_change_group() + * enqueue_task(p) -> p's new cfs_rq + * is throttled, go + * fast path and skip + * actual enqueue + * set_next_task(p) + * list_move(&se->group_node, &rq->cfs_tasks); // bug + * schedule() + * + * In the above race case, @p current cfs_rq is in the same rq as + * its previous cfs_rq because sched_move_task() only moves a task + * to a different group from the same rq, so we can use its current + * cfs_rq to derive rq and test if the task is current. + */ + if (throttled_hierarchy(cfs_rq) && + !task_current_donor(rq_of(cfs_rq), p)) { + list_add(&p->throttle_node, &cfs_rq->throttled_limbo_list); + return true; + } - return throttled_hierarchy(src_cfs_rq) || - throttled_hierarchy(dest_cfs_rq); + /* we can't take the fast path, do an actual enqueue*/ + p->throttled = false; + return false; } +static void enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags); static int tg_unthrottle_up(struct task_group *tg, void *data) { struct rq *rq = data; struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)]; + struct task_struct *p, *tmp; + + if (--cfs_rq->throttle_count) + return 0; - cfs_rq->throttle_count--; - if (!cfs_rq->throttle_count) { + if (cfs_rq->pelt_clock_throttled) { cfs_rq->throttled_clock_pelt_time += rq_clock_pelt(rq) - cfs_rq->throttled_clock_pelt; + cfs_rq->pelt_clock_throttled = 0; + } - /* Add cfs_rq with load or one or more already running entities to the list */ - if (!cfs_rq_is_decayed(cfs_rq)) - list_add_leaf_cfs_rq(cfs_rq); + if (cfs_rq->throttled_clock_self) { + u64 delta = rq_clock(rq) - cfs_rq->throttled_clock_self; - if (cfs_rq->throttled_clock_self) { - u64 delta = rq_clock(rq) - cfs_rq->throttled_clock_self; + cfs_rq->throttled_clock_self = 0; - cfs_rq->throttled_clock_self = 0; + if (WARN_ON_ONCE((s64)delta < 0)) + delta = 0; - if (WARN_ON_ONCE((s64)delta < 0)) - delta = 0; + cfs_rq->throttled_clock_self_time += delta; + } - cfs_rq->throttled_clock_self_time += delta; - } + /* Re-enqueue the tasks that have been throttled at this level. */ + list_for_each_entry_safe(p, tmp, &cfs_rq->throttled_limbo_list, throttle_node) { + list_del_init(&p->throttle_node); + p->throttled = false; + enqueue_task_fair(rq_of(cfs_rq), p, ENQUEUE_WAKEUP); } + /* Add cfs_rq with load or one or more already running entities to the list */ + if (!cfs_rq_is_decayed(cfs_rq)) + list_add_leaf_cfs_rq(cfs_rq); + return 0; } +static inline bool task_has_throttle_work(struct task_struct *p) +{ + return p->sched_throttle_work.next != &p->sched_throttle_work; +} + +static inline void task_throttle_setup_work(struct task_struct *p) +{ + if (task_has_throttle_work(p)) + return; + + /* + * Kthreads and exiting tasks don't return to userspace, so adding the + * work is pointless + */ + if ((p->flags & (PF_EXITING | PF_KTHREAD))) + return; + + task_work_add(p, &p->sched_throttle_work, TWA_RESUME); +} + +static void record_throttle_clock(struct cfs_rq *cfs_rq) +{ + struct rq *rq = rq_of(cfs_rq); + + if (cfs_rq_throttled(cfs_rq) && !cfs_rq->throttled_clock) + cfs_rq->throttled_clock = rq_clock(rq); + + if (!cfs_rq->throttled_clock_self) + cfs_rq->throttled_clock_self = rq_clock(rq); +} + static int tg_throttle_down(struct task_group *tg, void *data) { struct rq *rq = data; struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)]; - /* group is entering throttled state, stop time */ - if (!cfs_rq->throttle_count) { - cfs_rq->throttled_clock_pelt = rq_clock_pelt(rq); - list_del_leaf_cfs_rq(cfs_rq); + if (cfs_rq->throttle_count++) + return 0; - WARN_ON_ONCE(cfs_rq->throttled_clock_self); - if (cfs_rq->nr_queued) - cfs_rq->throttled_clock_self = rq_clock(rq); + /* + * For cfs_rqs that still have entities enqueued, PELT clock + * stop happens at dequeue time when all entities are dequeued. + */ + if (!cfs_rq->nr_queued) { + list_del_leaf_cfs_rq(cfs_rq); + cfs_rq->throttled_clock_pelt = rq_clock_pelt(rq); + cfs_rq->pelt_clock_throttled = 1; } - cfs_rq->throttle_count++; + WARN_ON_ONCE(cfs_rq->throttled_clock_self); + WARN_ON_ONCE(!list_empty(&cfs_rq->throttled_limbo_list)); return 0; } @@ -5800,8 +5980,7 @@ static bool throttle_cfs_rq(struct cfs_rq *cfs_rq) { struct rq *rq = rq_of(cfs_rq); struct cfs_bandwidth *cfs_b = tg_cfs_bandwidth(cfs_rq->tg); - struct sched_entity *se; - long queued_delta, runnable_delta, idle_delta, dequeue = 1; + int dequeue = 1; raw_spin_lock(&cfs_b->lock); /* This will start the period timer if necessary */ @@ -5824,76 +6003,17 @@ static bool throttle_cfs_rq(struct cfs_rq *cfs_rq) if (!dequeue) return false; /* Throttle no longer required. */ - se = cfs_rq->tg->se[cpu_of(rq_of(cfs_rq))]; - /* freeze hierarchy runnable averages while throttled */ rcu_read_lock(); walk_tg_tree_from(cfs_rq->tg, tg_throttle_down, tg_nop, (void *)rq); rcu_read_unlock(); - queued_delta = cfs_rq->h_nr_queued; - runnable_delta = cfs_rq->h_nr_runnable; - idle_delta = cfs_rq->h_nr_idle; - for_each_sched_entity(se) { - struct cfs_rq *qcfs_rq = cfs_rq_of(se); - int flags; - - /* throttled entity or throttle-on-deactivate */ - if (!se->on_rq) - goto done; - - /* - * Abuse SPECIAL to avoid delayed dequeue in this instance. - * This avoids teaching dequeue_entities() about throttled - * entities and keeps things relatively simple. - */ - flags = DEQUEUE_SLEEP | DEQUEUE_SPECIAL; - if (se->sched_delayed) - flags |= DEQUEUE_DELAYED; - dequeue_entity(qcfs_rq, se, flags); - - if (cfs_rq_is_idle(group_cfs_rq(se))) - idle_delta = cfs_rq->h_nr_queued; - - qcfs_rq->h_nr_queued -= queued_delta; - qcfs_rq->h_nr_runnable -= runnable_delta; - qcfs_rq->h_nr_idle -= idle_delta; - - if (qcfs_rq->load.weight) { - /* Avoid re-evaluating load for this entity: */ - se = parent_entity(se); - break; - } - } - - for_each_sched_entity(se) { - struct cfs_rq *qcfs_rq = cfs_rq_of(se); - /* throttled entity or throttle-on-deactivate */ - if (!se->on_rq) - goto done; - - update_load_avg(qcfs_rq, se, 0); - se_update_runnable(se); - - if (cfs_rq_is_idle(group_cfs_rq(se))) - idle_delta = cfs_rq->h_nr_queued; - - qcfs_rq->h_nr_queued -= queued_delta; - qcfs_rq->h_nr_runnable -= runnable_delta; - qcfs_rq->h_nr_idle -= idle_delta; - } - - /* At this point se is NULL and we are at root level*/ - sub_nr_running(rq, queued_delta); -done: /* * Note: distribution will already see us throttled via the * throttled-list. rq->lock protects completion. */ cfs_rq->throttled = 1; WARN_ON_ONCE(cfs_rq->throttled_clock); - if (cfs_rq->nr_queued) - cfs_rq->throttled_clock = rq_clock(rq); return true; } @@ -5901,9 +6021,20 @@ void unthrottle_cfs_rq(struct cfs_rq *cfs_rq) { struct rq *rq = rq_of(cfs_rq); struct cfs_bandwidth *cfs_b = tg_cfs_bandwidth(cfs_rq->tg); - struct sched_entity *se; - long queued_delta, runnable_delta, idle_delta; - long rq_h_nr_queued = rq->cfs.h_nr_queued; + struct sched_entity *se = cfs_rq->tg->se[cpu_of(rq)]; + + /* + * It's possible we are called with !runtime_remaining due to things + * like user changed quota setting(see tg_set_cfs_bandwidth()) or async + * unthrottled us with a positive runtime_remaining but other still + * running entities consumed those runtime before we reached here. + * + * Anyway, we can't unthrottle this cfs_rq without any runtime remaining + * because any enqueue in tg_unthrottle_up() will immediately trigger a + * throttle, which is not supposed to happen on unthrottle path. + */ + if (cfs_rq->runtime_enabled && cfs_rq->runtime_remaining <= 0) + return; se = cfs_rq->tg->se[cpu_of(rq)]; @@ -5933,62 +6064,8 @@ void unthrottle_cfs_rq(struct cfs_rq *cfs_rq) if (list_add_leaf_cfs_rq(cfs_rq_of(se))) break; } - goto unthrottle_throttle; - } - - queued_delta = cfs_rq->h_nr_queued; - runnable_delta = cfs_rq->h_nr_runnable; - idle_delta = cfs_rq->h_nr_idle; - for_each_sched_entity(se) { - struct cfs_rq *qcfs_rq = cfs_rq_of(se); - - /* Handle any unfinished DELAY_DEQUEUE business first. */ - if (se->sched_delayed) { - int flags = DEQUEUE_SLEEP | DEQUEUE_DELAYED; - - dequeue_entity(qcfs_rq, se, flags); - } else if (se->on_rq) - break; - enqueue_entity(qcfs_rq, se, ENQUEUE_WAKEUP); - - if (cfs_rq_is_idle(group_cfs_rq(se))) - idle_delta = cfs_rq->h_nr_queued; - - qcfs_rq->h_nr_queued += queued_delta; - qcfs_rq->h_nr_runnable += runnable_delta; - qcfs_rq->h_nr_idle += idle_delta; - - /* end evaluation on encountering a throttled cfs_rq */ - if (cfs_rq_throttled(qcfs_rq)) - goto unthrottle_throttle; - } - - for_each_sched_entity(se) { - struct cfs_rq *qcfs_rq = cfs_rq_of(se); - - update_load_avg(qcfs_rq, se, UPDATE_TG); - se_update_runnable(se); - - if (cfs_rq_is_idle(group_cfs_rq(se))) - idle_delta = cfs_rq->h_nr_queued; - - qcfs_rq->h_nr_queued += queued_delta; - qcfs_rq->h_nr_runnable += runnable_delta; - qcfs_rq->h_nr_idle += idle_delta; - - /* end evaluation on encountering a throttled cfs_rq */ - if (cfs_rq_throttled(qcfs_rq)) - goto unthrottle_throttle; } - /* Start the fair server if un-throttling resulted in new runnable tasks */ - if (!rq_h_nr_queued && rq->cfs.h_nr_queued) - dl_server_start(&rq->fair_server); - - /* At this point se is NULL and we are at root level*/ - add_nr_running(rq, queued_delta); - -unthrottle_throttle: assert_list_leaf_cfs_rq(rq); /* Determine whether we need to wake up potentially idle CPU: */ @@ -6472,6 +6549,7 @@ static void init_cfs_rq_runtime(struct cfs_rq *cfs_rq) cfs_rq->runtime_enabled = 0; INIT_LIST_HEAD(&cfs_rq->throttled_list); INIT_LIST_HEAD(&cfs_rq->throttled_csd_list); + INIT_LIST_HEAD(&cfs_rq->throttled_limbo_list); } void start_cfs_bandwidth(struct cfs_bandwidth *cfs_b) @@ -6639,19 +6717,28 @@ static bool check_cfs_rq_runtime(struct cfs_rq *cfs_rq) { return false; } static void check_enqueue_throttle(struct cfs_rq *cfs_rq) {} static inline void sync_throttle(struct task_group *tg, int cpu) {} static __always_inline void return_cfs_rq_runtime(struct cfs_rq *cfs_rq) {} +static void task_throttle_setup_work(struct task_struct *p) {} +static bool task_is_throttled(struct task_struct *p) { return false; } +static void dequeue_throttled_task(struct task_struct *p, int flags) {} +static bool enqueue_throttled_task(struct task_struct *p) { return false; } +static void record_throttle_clock(struct cfs_rq *cfs_rq) {} static inline int cfs_rq_throttled(struct cfs_rq *cfs_rq) { return 0; } +static inline bool cfs_rq_pelt_clock_throttled(struct cfs_rq *cfs_rq) +{ + return false; +} + static inline int throttled_hierarchy(struct cfs_rq *cfs_rq) { return 0; } -static inline int throttled_lb_pair(struct task_group *tg, - int src_cpu, int dest_cpu) +static inline int lb_throttled_hierarchy(struct task_struct *p, int dst_cpu) { return 0; } @@ -6831,6 +6918,9 @@ enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags) int rq_h_nr_queued = rq->cfs.h_nr_queued; u64 slice = 0; + if (task_is_throttled(p) && enqueue_throttled_task(p)) + return; + /* * The code below (indirectly) updates schedutil which looks at * the cfs_rq utilization to select a frequency. @@ -6883,10 +6973,6 @@ enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags) if (cfs_rq_is_idle(cfs_rq)) h_nr_idle = 1; - /* end evaluation on encountering a throttled cfs_rq */ - if (cfs_rq_throttled(cfs_rq)) - goto enqueue_throttle; - flags = ENQUEUE_WAKEUP; } @@ -6908,10 +6994,6 @@ enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags) if (cfs_rq_is_idle(cfs_rq)) h_nr_idle = 1; - - /* end evaluation on encountering a throttled cfs_rq */ - if (cfs_rq_throttled(cfs_rq)) - goto enqueue_throttle; } if (!rq_h_nr_queued && rq->cfs.h_nr_queued) { @@ -6941,7 +7023,6 @@ enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags) if (!task_new) check_update_overutilized_status(rq); -enqueue_throttle: assert_list_leaf_cfs_rq(rq); hrtick_update(rq); @@ -6963,6 +7044,7 @@ static int dequeue_entities(struct rq *rq, struct sched_entity *se, int flags) bool was_sched_idle = sched_idle_rq(rq); bool task_sleep = flags & DEQUEUE_SLEEP; bool task_delayed = flags & DEQUEUE_DELAYED; + bool task_throttled = flags & DEQUEUE_THROTTLE; struct task_struct *p = NULL; int h_nr_idle = 0; int h_nr_queued = 0; @@ -6996,9 +7078,8 @@ static int dequeue_entities(struct rq *rq, struct sched_entity *se, int flags) if (cfs_rq_is_idle(cfs_rq)) h_nr_idle = h_nr_queued; - /* end evaluation on encountering a throttled cfs_rq */ - if (cfs_rq_throttled(cfs_rq)) - return 0; + if (throttled_hierarchy(cfs_rq) && task_throttled) + record_throttle_clock(cfs_rq); /* Don't dequeue parent if it has other entities besides us */ if (cfs_rq->load.weight) { @@ -7010,7 +7091,7 @@ static int dequeue_entities(struct rq *rq, struct sched_entity *se, int flags) * Bias pick_next to pick a task from this cfs_rq, as * p is sleeping when it is within its sched_slice. */ - if (task_sleep && se && !throttled_hierarchy(cfs_rq)) + if (task_sleep && se) set_next_buddy(se); break; } @@ -7037,9 +7118,8 @@ static int dequeue_entities(struct rq *rq, struct sched_entity *se, int flags) if (cfs_rq_is_idle(cfs_rq)) h_nr_idle = h_nr_queued; - /* end evaluation on encountering a throttled cfs_rq */ - if (cfs_rq_throttled(cfs_rq)) - return 0; + if (throttled_hierarchy(cfs_rq) && task_throttled) + record_throttle_clock(cfs_rq); } sub_nr_running(rq, h_nr_queued); @@ -7073,6 +7153,11 @@ static int dequeue_entities(struct rq *rq, struct sched_entity *se, int flags) */ static bool dequeue_task_fair(struct rq *rq, struct task_struct *p, int flags) { + if (task_is_throttled(p)) { + dequeue_throttled_task(p, flags); + return true; + } + if (!p->se.sched_delayed) util_est_dequeue(&rq->cfs, p); @@ -8660,7 +8745,7 @@ static void check_preempt_wakeup_fair(struct rq *rq, struct task_struct *p, int * lead to a throttle). This both saves work and prevents false * next-buddy nomination below. */ - if (unlikely(throttled_hierarchy(cfs_rq_of(pse)))) + if (task_is_throttled(p)) return; if (sched_feat(NEXT_BUDDY) && !(wake_flags & WF_FORK) && !pse->sched_delayed) { @@ -8741,19 +8826,22 @@ static struct task_struct *pick_task_fair(struct rq *rq) { struct sched_entity *se; struct cfs_rq *cfs_rq; + struct task_struct *p; + bool throttled; again: cfs_rq = &rq->cfs; if (!cfs_rq->nr_queued) return NULL; + throttled = false; + do { /* Might not have done put_prev_entity() */ if (cfs_rq->curr && cfs_rq->curr->on_rq) update_curr(cfs_rq); - if (unlikely(check_cfs_rq_runtime(cfs_rq))) - goto again; + throttled |= check_cfs_rq_runtime(cfs_rq); se = pick_next_entity(rq, cfs_rq); if (!se) @@ -8761,7 +8849,10 @@ again: cfs_rq = group_cfs_rq(se); } while (cfs_rq); - return task_of(se); + p = task_of(se); + if (unlikely(throttled)) + task_throttle_setup_work(p); + return p; } static void __set_next_task_fair(struct rq *rq, struct task_struct *p, bool first); @@ -8859,11 +8950,6 @@ static struct task_struct *__pick_next_task_fair(struct rq *rq, struct task_stru return pick_next_task_fair(rq, prev, NULL); } -static bool fair_server_has_tasks(struct sched_dl_entity *dl_se) -{ - return !!dl_se->rq->cfs.nr_queued; -} - static struct task_struct *fair_server_pick_task(struct sched_dl_entity *dl_se) { return pick_task_fair(dl_se->rq); @@ -8875,7 +8961,7 @@ void fair_server_init(struct rq *rq) init_dl_entity(dl_se); - dl_server_init(dl_se, rq, fair_server_has_tasks, fair_server_pick_task); + dl_server_init(dl_se, rq, fair_server_pick_task); } /* @@ -8928,8 +9014,8 @@ static bool yield_to_task_fair(struct rq *rq, struct task_struct *p) { struct sched_entity *se = &p->se; - /* throttled hierarchies are not runnable */ - if (!se->on_rq || throttled_hierarchy(cfs_rq_of(se))) + /* !se->on_rq also covers throttled task */ + if (!se->on_rq) return false; /* Tell the scheduler that we'd really like se to run next. */ @@ -9288,7 +9374,7 @@ int can_migrate_task(struct task_struct *p, struct lb_env *env) /* * We do not migrate tasks that are: * 1) delayed dequeued unless we migrate load, or - * 2) throttled_lb_pair, or + * 2) target cfs_rq is in throttled hierarchy, or * 3) cannot be migrated to this CPU due to cpus_ptr, or * 4) running (obviously), or * 5) are cache-hot on their current CPU, or @@ -9297,7 +9383,7 @@ int can_migrate_task(struct task_struct *p, struct lb_env *env) if ((p->se.sched_delayed) && (env->migration_type != migrate_load)) return 0; - if (throttled_lb_pair(task_group(p), env->src_cpu, env->dst_cpu)) + if (lb_throttled_hierarchy(p, env->dst_cpu)) return 0; /* @@ -13081,10 +13167,13 @@ static void propagate_entity_cfs_rq(struct sched_entity *se) { struct cfs_rq *cfs_rq = cfs_rq_of(se); - if (cfs_rq_throttled(cfs_rq)) - return; - - if (!throttled_hierarchy(cfs_rq)) + /* + * If a task gets attached to this cfs_rq and before being queued, + * it gets migrated to another CPU due to reasons like affinity + * change, make sure this cfs_rq stays on leaf cfs_rq list to have + * that removed load decayed or it can cause faireness problem. + */ + if (!cfs_rq_pelt_clock_throttled(cfs_rq)) list_add_leaf_cfs_rq(cfs_rq); /* Start to propagate at parent */ @@ -13095,10 +13184,7 @@ static void propagate_entity_cfs_rq(struct sched_entity *se) update_load_avg(cfs_rq, se, UPDATE_TG); - if (cfs_rq_throttled(cfs_rq)) - break; - - if (!throttled_hierarchy(cfs_rq)) + if (!cfs_rq_pelt_clock_throttled(cfs_rq)) list_add_leaf_cfs_rq(cfs_rq); } } diff --git a/kernel/sched/pelt.h b/kernel/sched/pelt.h index 62c3fa543c0f..f921302dc40f 100644 --- a/kernel/sched/pelt.h +++ b/kernel/sched/pelt.h @@ -162,7 +162,7 @@ static inline void update_idle_cfs_rq_clock_pelt(struct cfs_rq *cfs_rq) { u64 throttled; - if (unlikely(cfs_rq->throttle_count)) + if (unlikely(cfs_rq->pelt_clock_throttled)) throttled = U64_MAX; else throttled = cfs_rq->throttled_clock_pelt_time; @@ -173,7 +173,7 @@ static inline void update_idle_cfs_rq_clock_pelt(struct cfs_rq *cfs_rq) /* rq->task_clock normalized against any time this cfs_rq has spent throttled */ static inline u64 cfs_rq_clock_pelt(struct cfs_rq *cfs_rq) { - if (unlikely(cfs_rq->throttle_count)) + if (unlikely(cfs_rq->pelt_clock_throttled)) return cfs_rq->throttled_clock_pelt - cfs_rq->throttled_clock_pelt_time; return rq_clock_pelt(rq_of(cfs_rq)) - cfs_rq->throttled_clock_pelt_time; diff --git a/kernel/sched/rq-offsets.c b/kernel/sched/rq-offsets.c new file mode 100644 index 000000000000..a23747bbe25b --- /dev/null +++ b/kernel/sched/rq-offsets.c @@ -0,0 +1,12 @@ +// SPDX-License-Identifier: GPL-2.0 +#define COMPILE_OFFSETS +#include <linux/kbuild.h> +#include <linux/types.h> +#include "sched.h" + +int main(void) +{ + DEFINE(RQ_nr_pinned, offsetof(struct rq, nr_pinned)); + + return 0; +} diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index be9745d104f7..1f5d07067f60 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -365,25 +365,50 @@ extern s64 dl_scaled_delta_exec(struct rq *rq, struct sched_dl_entity *dl_se, s6 * * dl_se::rq -- runqueue we belong to. * - * dl_se::server_has_tasks() -- used on bandwidth enforcement; we 'stop' the - * server when it runs out of tasks to run. - * * dl_se::server_pick() -- nested pick_next_task(); we yield the period if this * returns NULL. * * dl_server_update() -- called from update_curr_common(), propagates runtime * to the server. * - * dl_server_start() - * dl_server_stop() -- start/stop the server when it has (no) tasks. + * dl_server_start() -- start the server when it has tasks; it will stop + * automatically when there are no more tasks, per + * dl_se::server_pick() returning NULL. + * + * dl_server_stop() -- (force) stop the server; use when updating + * parameters. * * dl_server_init() -- initializes the server. + * + * When started the dl_server will (per dl_defer) schedule a timer for its + * zero-laxity point -- that is, unlike regular EDF tasks which run ASAP, a + * server will run at the very end of its period. + * + * This is done such that any runtime from the target class can be accounted + * against the server -- through dl_server_update() above -- such that when it + * becomes time to run, it might already be out of runtime and get deferred + * until the next period. In this case dl_server_timer() will alternate + * between defer and replenish but never actually enqueue the server. + * + * Only when the target class does not manage to exhaust the server's runtime + * (there's actualy starvation in the given period), will the dl_server get on + * the runqueue. Once queued it will pick tasks from the target class and run + * them until either its runtime is exhaused, at which point its back to + * dl_server_timer, or until there are no more tasks to run, at which point + * the dl_server stops itself. + * + * By stopping at this point the dl_server retains bandwidth, which, if a new + * task wakes up imminently (starting the server again), can be used -- + * subject to CBS wakeup rules -- without having to wait for the next period. + * + * Additionally, because of the dl_defer behaviour the start/stop behaviour is + * naturally thottled to once per period, avoiding high context switch + * workloads from spamming the hrtimer program/cancel paths. */ extern void dl_server_update(struct sched_dl_entity *dl_se, s64 delta_exec); extern void dl_server_start(struct sched_dl_entity *dl_se); extern void dl_server_stop(struct sched_dl_entity *dl_se); extern void dl_server_init(struct sched_dl_entity *dl_se, struct rq *rq, - dl_server_has_tasks_f has_tasks, dl_server_pick_f pick_task); extern void sched_init_dl_servers(void); @@ -735,10 +760,12 @@ struct cfs_rq { u64 throttled_clock_pelt_time; u64 throttled_clock_self; u64 throttled_clock_self_time; - int throttled; + bool throttled:1; + bool pelt_clock_throttled:1; int throttle_count; struct list_head throttled_list; struct list_head throttled_csd_list; + struct list_head throttled_limbo_list; #endif /* CONFIG_CFS_BANDWIDTH */ #endif /* CONFIG_FAIR_GROUP_SCHED */ }; @@ -1935,12 +1962,12 @@ extern void sched_setnuma(struct task_struct *p, int node); extern int migrate_task_to(struct task_struct *p, int cpu); extern int migrate_swap(struct task_struct *p, struct task_struct *t, int cpu, int scpu); -extern void init_numa_balancing(unsigned long clone_flags, struct task_struct *p); +extern void init_numa_balancing(u64 clone_flags, struct task_struct *p); #else /* !CONFIG_NUMA_BALANCING: */ static inline void -init_numa_balancing(unsigned long clone_flags, struct task_struct *p) +init_numa_balancing(u64 clone_flags, struct task_struct *p) { } @@ -2342,6 +2369,7 @@ extern const u32 sched_prio_to_wmult[40]; #define DEQUEUE_SPECIAL 0x10 #define DEQUEUE_MIGRATING 0x100 /* Matches ENQUEUE_MIGRATING */ #define DEQUEUE_DELAYED 0x200 /* Matches ENQUEUE_DELAYED */ +#define DEQUEUE_THROTTLE 0x800 #define ENQUEUE_WAKEUP 0x01 #define ENQUEUE_RESTORE 0x02 @@ -2658,6 +2686,8 @@ extern bool sched_rt_bandwidth_account(struct rt_rq *rt_rq); extern void init_dl_entity(struct sched_dl_entity *dl_se); +extern void init_cfs_throttle_work(struct task_struct *p); + #define BW_SHIFT 20 #define BW_UNIT (1 << BW_SHIFT) #define RATIO_SHIFT 8 diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c index 977e133bb8a4..444bdfdab731 100644 --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -1591,7 +1591,6 @@ static void claim_allocations(int cpu, struct sched_domain *sd) enum numa_topology_type sched_numa_topology_type; static int sched_domains_numa_levels; -static int sched_domains_curr_level; int sched_max_numa_distance; static int *sched_domains_numa_distance; @@ -1632,14 +1631,7 @@ sd_init(struct sched_domain_topology_level *tl, int sd_id, sd_weight, sd_flags = 0; struct cpumask *sd_span; -#ifdef CONFIG_NUMA - /* - * Ugly hack to pass state to sd_numa_mask()... - */ - sched_domains_curr_level = tl->numa_level; -#endif - - sd_weight = cpumask_weight(tl->mask(cpu)); + sd_weight = cpumask_weight(tl->mask(tl, cpu)); if (tl->sd_flags) sd_flags = (*tl->sd_flags)(); @@ -1677,7 +1669,7 @@ sd_init(struct sched_domain_topology_level *tl, }; sd_span = sched_domain_span(sd); - cpumask_and(sd_span, cpu_map, tl->mask(cpu)); + cpumask_and(sd_span, cpu_map, tl->mask(tl, cpu)); sd_id = cpumask_first(sd_span); sd->flags |= asym_cpu_capacity_classify(sd_span, cpu_map); @@ -1732,22 +1724,63 @@ sd_init(struct sched_domain_topology_level *tl, return sd; } +#ifdef CONFIG_SCHED_SMT +int cpu_smt_flags(void) +{ + return SD_SHARE_CPUCAPACITY | SD_SHARE_LLC; +} + +const struct cpumask *tl_smt_mask(struct sched_domain_topology_level *tl, int cpu) +{ + return cpu_smt_mask(cpu); +} +#endif + +#ifdef CONFIG_SCHED_CLUSTER +int cpu_cluster_flags(void) +{ + return SD_CLUSTER | SD_SHARE_LLC; +} + +const struct cpumask *tl_cls_mask(struct sched_domain_topology_level *tl, int cpu) +{ + return cpu_clustergroup_mask(cpu); +} +#endif + +#ifdef CONFIG_SCHED_MC +int cpu_core_flags(void) +{ + return SD_SHARE_LLC; +} + +const struct cpumask *tl_mc_mask(struct sched_domain_topology_level *tl, int cpu) +{ + return cpu_coregroup_mask(cpu); +} +#endif + +const struct cpumask *tl_pkg_mask(struct sched_domain_topology_level *tl, int cpu) +{ + return cpu_node_mask(cpu); +} + /* * Topology list, bottom-up. */ static struct sched_domain_topology_level default_topology[] = { #ifdef CONFIG_SCHED_SMT - SDTL_INIT(cpu_smt_mask, cpu_smt_flags, SMT), + SDTL_INIT(tl_smt_mask, cpu_smt_flags, SMT), #endif #ifdef CONFIG_SCHED_CLUSTER - SDTL_INIT(cpu_clustergroup_mask, cpu_cluster_flags, CLS), + SDTL_INIT(tl_cls_mask, cpu_cluster_flags, CLS), #endif #ifdef CONFIG_SCHED_MC - SDTL_INIT(cpu_coregroup_mask, cpu_core_flags, MC), + SDTL_INIT(tl_mc_mask, cpu_core_flags, MC), #endif - SDTL_INIT(cpu_cpu_mask, NULL, PKG), + SDTL_INIT(tl_pkg_mask, NULL, PKG), { NULL, }, }; @@ -1768,10 +1801,14 @@ void __init set_sched_topology(struct sched_domain_topology_level *tl) } #ifdef CONFIG_NUMA +static int cpu_numa_flags(void) +{ + return SD_NUMA; +} -static const struct cpumask *sd_numa_mask(int cpu) +static const struct cpumask *sd_numa_mask(struct sched_domain_topology_level *tl, int cpu) { - return sched_domains_numa_masks[sched_domains_curr_level][cpu_to_node(cpu)]; + return sched_domains_numa_masks[tl->numa_level][cpu_to_node(cpu)]; } static void sched_numa_warn(const char *str) @@ -2201,6 +2238,8 @@ int sched_numa_find_nth_cpu(const struct cpumask *cpus, int cpu, int node) goto unlock; hop_masks = bsearch(&k, k.masks, sched_domains_numa_levels, sizeof(k.masks[0]), hop_cmp); + if (!hop_masks) + goto unlock; hop = hop_masks - k.masks; ret = hop ? @@ -2411,7 +2450,7 @@ static bool topology_span_sane(const struct cpumask *cpu_map) * breaks the linking done for an earlier span. */ for_each_cpu(cpu, cpu_map) { - const struct cpumask *tl_cpu_mask = tl->mask(cpu); + const struct cpumask *tl_cpu_mask = tl->mask(tl, cpu); int id; /* lowest bit set in this mask is used as a unique id */ @@ -2419,7 +2458,7 @@ static bool topology_span_sane(const struct cpumask *cpu_map) if (cpumask_test_cpu(id, id_seen)) { /* First CPU has already been seen, ensure identical spans */ - if (!cpumask_equal(tl->mask(id), tl_cpu_mask)) + if (!cpumask_equal(tl->mask(tl, id), tl_cpu_mask)) return false; } else { /* First CPU hasn't been seen before, ensure it's a completely new span */ |