diff options
Diffstat (limited to 'kernel/sched/ext.c')
| -rw-r--r-- | kernel/sched/ext.c | 5037 |
1 files changed, 2350 insertions, 2687 deletions
diff --git a/kernel/sched/ext.c b/kernel/sched/ext.c index 7fff1d045477..05f5a49e9649 100644 --- a/kernel/sched/ext.c +++ b/kernel/sched/ext.c @@ -6,853 +6,18 @@ * Copyright (c) 2022 Tejun Heo <tj@kernel.org> * Copyright (c) 2022 David Vernet <dvernet@meta.com> */ -#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_OPS_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_ops_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, - - /* - * CPU cgroup support flags - */ - SCX_OPS_HAS_CGROUP_WEIGHT = 1LLU << 16, /* cpu.weight */ - - SCX_OPS_ALL_FLAGS = SCX_OPS_KEEP_BUILTIN_IDLE | - SCX_OPS_ENQ_LAST | - SCX_OPS_ENQ_EXITING | - SCX_OPS_SWITCH_PARTIAL | - SCX_OPS_HAS_CGROUP_WEIGHT, -}; - -/* 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; -}; - -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 - * - * 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? - * - * 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 udpate 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_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 - * - * @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 @tg's weight to @weight. - */ - void (*cgroup_set_weight)(struct cgroup *cgrp, u32 weight); -#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]; -}; - -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), -}; - -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 - * bpf_scx_reenqueue_local() kfunc. If bpf_scx_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 */ -}; - -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_ops_enable_state { - SCX_OPS_ENABLING, - SCX_OPS_ENABLED, - SCX_OPS_DISABLING, - SCX_OPS_DISABLED, -}; - -static const char *scx_ops_enable_state_str[] = { - [SCX_OPS_ENABLING] = "enabling", - [SCX_OPS_ENABLED] = "enabled", - [SCX_OPS_DISABLING] = "disabling", - [SCX_OPS_DISABLED] = "disabled", -}; +#include <linux/btf_ids.h> +#include "ext_idle.h" /* - * 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. + * 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 + * caller is one of the tasks attached to SCX and explicit RCU dereference is + * necessary otherwise. Naked scx_root dereferences trigger sparse warnings but + * are used as temporary markers to indicate that the dereferences need to be + * updated to point to the associated scheduler instances rather than scx_root. */ -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) +static struct scx_sched __rcu *scx_root; /* * During exit, a task may schedule after losing its PIDs. When disabling the @@ -860,41 +25,22 @@ enum scx_ops_state { * guarantee system safety. Maintain a dedicated task list which contains every * task between its fork and eventual free. */ -static DEFINE_SPINLOCK(scx_tasks_lock); +static DEFINE_RAW_SPINLOCK(scx_tasks_lock); static LIST_HEAD(scx_tasks); /* ops enable/disable */ -static struct kthread_worker *scx_ops_helper; -static DEFINE_MUTEX(scx_ops_enable_mutex); -DEFINE_STATIC_KEY_FALSE(__scx_ops_enabled); +static DEFINE_MUTEX(scx_enable_mutex); +DEFINE_STATIC_KEY_FALSE(__scx_enabled); DEFINE_STATIC_PERCPU_RWSEM(scx_fork_rwsem); -static atomic_t scx_ops_enable_state_var = ATOMIC_INIT(SCX_OPS_DISABLED); -static unsigned long scx_in_softlockup; -static atomic_t scx_ops_breather_depth = ATOMIC_INIT(0); -static int scx_ops_bypass_depth; -static bool scx_ops_init_task_enabled; +static atomic_t scx_enable_state_var = ATOMIC_INIT(SCX_DISABLED); +static int scx_bypass_depth; +static cpumask_var_t scx_bypass_lb_donee_cpumask; +static cpumask_var_t scx_bypass_lb_resched_cpumask; +static bool scx_aborting; +static bool scx_init_task_enabled; static bool scx_switching_all; DEFINE_STATIC_KEY_FALSE(__scx_switched_all); -static struct sched_ext_ops scx_ops; -static bool scx_warned_zero_slice; - -static DEFINE_STATIC_KEY_FALSE(scx_ops_enq_last); -static DEFINE_STATIC_KEY_FALSE(scx_ops_enq_exiting); -static DEFINE_STATIC_KEY_FALSE(scx_ops_cpu_preempt); -static DEFINE_STATIC_KEY_FALSE(scx_builtin_idle_enabled); - -#ifdef CONFIG_SMP -static DEFINE_STATIC_KEY_FALSE(scx_selcpu_topo_llc); -static DEFINE_STATIC_KEY_FALSE(scx_selcpu_topo_numa); -#endif - -static struct static_key_false scx_has_op[SCX_OPI_END] = - { [0 ... SCX_OPI_END-1] = STATIC_KEY_FALSE_INIT }; - -static atomic_t scx_exit_kind = ATOMIC_INIT(SCX_EXIT_DONE); -static struct scx_exit_info *scx_exit_info; - static atomic_long_t scx_nr_rejected = ATOMIC_LONG_INIT(0); static atomic_long_t scx_hotplug_seq = ATOMIC_LONG_INIT(0); @@ -908,7 +54,7 @@ static atomic_long_t scx_enable_seq = ATOMIC_LONG_INIT(0); /* * The maximum amount of time in jiffies that a task may be runnable without * being scheduled on a CPU. If this timeout is exceeded, it will trigger - * scx_ops_error(). + * scx_error(). */ static unsigned long scx_watchdog_timeout; @@ -922,23 +68,19 @@ static unsigned long scx_watchdog_timestamp = INITIAL_JIFFIES; static struct delayed_work scx_watchdog_work; -/* idle tracking */ -#ifdef CONFIG_SMP -#ifdef CONFIG_CPUMASK_OFFSTACK -#define CL_ALIGNED_IF_ONSTACK -#else -#define CL_ALIGNED_IF_ONSTACK __cacheline_aligned_in_smp -#endif - -static struct { - cpumask_var_t cpu; - cpumask_var_t smt; -} idle_masks CL_ALIGNED_IF_ONSTACK; - -#endif /* CONFIG_SMP */ +/* + * For %SCX_KICK_WAIT: Each CPU has a pointer to an array of kick_sync sequence + * numbers. The arrays are allocated with kvzalloc() as size can exceed percpu + * allocator limits on large machines. O(nr_cpu_ids^2) allocation, allocated + * lazily when enabling and freed when disabling to avoid waste when sched_ext + * isn't active. + */ +struct scx_kick_syncs { + struct rcu_head rcu; + unsigned long syncs[]; +}; -/* for %SCX_KICK_WAIT */ -static unsigned long __percpu *scx_kick_cpus_pnt_seqs; +static DEFINE_PER_CPU(struct scx_kick_syncs __rcu *, scx_kick_syncs); /* * Direct dispatch marker. @@ -949,23 +91,12 @@ static unsigned long __percpu *scx_kick_cpus_pnt_seqs; */ static DEFINE_PER_CPU(struct task_struct *, direct_dispatch_task); -/* - * 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. - */ -static struct scx_dispatch_q **global_dsqs; - static const struct rhashtable_params dsq_hash_params = { - .key_len = 8, + .key_len = sizeof_field(struct scx_dispatch_q, id), .key_offset = offsetof(struct scx_dispatch_q, id), .head_offset = offsetof(struct scx_dispatch_q, hash_node), }; -static struct rhashtable dsq_hash; static LLIST_HEAD(dsqs_to_free); /* dispatch buf */ @@ -1012,27 +143,73 @@ static struct scx_dump_data scx_dump_data = { /* /sys/kernel/sched_ext interface */ static struct kset *scx_kset; -static struct kobject *scx_root_kobj; + +/* + * Parameters that can be adjusted through /sys/module/sched_ext/parameters. + * There usually is no reason to modify these as normal scheduler operation + * shouldn't be affected by them. The knobs are primarily for debugging. + */ +static u64 scx_slice_dfl = SCX_SLICE_DFL; +static unsigned int scx_slice_bypass_us = SCX_SLICE_BYPASS / NSEC_PER_USEC; +static unsigned int scx_bypass_lb_intv_us = SCX_BYPASS_LB_DFL_INTV_US; + +static int set_slice_us(const char *val, const struct kernel_param *kp) +{ + return param_set_uint_minmax(val, kp, 100, 100 * USEC_PER_MSEC); +} + +static const struct kernel_param_ops slice_us_param_ops = { + .set = set_slice_us, + .get = param_get_uint, +}; + +static int set_bypass_lb_intv_us(const char *val, const struct kernel_param *kp) +{ + return param_set_uint_minmax(val, kp, 0, 10 * USEC_PER_SEC); +} + +static const struct kernel_param_ops bypass_lb_intv_us_param_ops = { + .set = set_bypass_lb_intv_us, + .get = param_get_uint, +}; + +#undef MODULE_PARAM_PREFIX +#define MODULE_PARAM_PREFIX "sched_ext." + +module_param_cb(slice_bypass_us, &slice_us_param_ops, &scx_slice_bypass_us, 0600); +MODULE_PARM_DESC(slice_bypass_us, "bypass slice in microseconds, applied on [un]load (100us to 100ms)"); +module_param_cb(bypass_lb_intv_us, &bypass_lb_intv_us_param_ops, &scx_bypass_lb_intv_us, 0600); +MODULE_PARM_DESC(bypass_lb_intv_us, "bypass load balance interval in microseconds (0 (disable) to 10s)"); + +#undef MODULE_PARAM_PREFIX #define CREATE_TRACE_POINTS #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 __printf(3, 4) void scx_ops_exit_kind(enum scx_exit_kind kind, - s64 exit_code, - const char *fmt, ...); +static u32 reenq_local(struct rq *rq); +static void scx_kick_cpu(struct scx_sched *sch, s32 cpu, u64 flags); +static bool scx_vexit(struct scx_sched *sch, enum scx_exit_kind kind, + s64 exit_code, const char *fmt, va_list args); -#define scx_ops_error_kind(err, fmt, args...) \ - scx_ops_exit_kind((err), 0, fmt, ##args) +static __printf(4, 5) bool scx_exit(struct scx_sched *sch, + enum scx_exit_kind kind, s64 exit_code, + const char *fmt, ...) +{ + va_list args; + bool ret; + + va_start(args, fmt); + ret = scx_vexit(sch, kind, exit_code, fmt, args); + va_end(args); -#define scx_ops_exit(code, fmt, args...) \ - scx_ops_exit_kind(SCX_EXIT_UNREG_KERN, (code), fmt, ##args) + return ret; +} -#define scx_ops_error(fmt, args...) \ - scx_ops_error_kind(SCX_EXIT_ERROR, fmt, ##args) +#define scx_error(sch, fmt, args...) scx_exit((sch), SCX_EXIT_ERROR, 0, fmt, ##args) +#define scx_verror(sch, fmt, args) scx_vexit((sch), SCX_EXIT_ERROR, 0, fmt, args) -#define SCX_HAS_OP(op) static_branch_likely(&scx_has_op[SCX_OP_IDX(op)]) +#define SCX_HAS_OP(sch, op) test_bit(SCX_OP_IDX(op), (sch)->has_op) static long jiffies_delta_msecs(unsigned long at, unsigned long now) { @@ -1060,14 +237,23 @@ 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) +{ + return sch->global_dsqs[cpu_to_node(task_cpu(p))]; +} + +static struct scx_dispatch_q *find_user_dsq(struct scx_sched *sch, u64 dsq_id) { - return global_dsqs[cpu_to_node(task_cpu(p))]; + return rhashtable_lookup(&sch->dsq_hash, &dsq_id, dsq_hash_params); } -static struct scx_dispatch_q *find_user_dsq(u64 dsq_id) +static const struct sched_class *scx_setscheduler_class(struct task_struct *p) { - return rhashtable_lookup_fast(&dsq_hash, &dsq_id, dsq_hash_params); + if (p->sched_class == &stop_sched_class) + return &stop_sched_class; + + return __setscheduler_class(p->policy, p->prio); } /* @@ -1094,27 +280,56 @@ static void scx_kf_disallow(u32 mask) current->scx.kf_mask &= ~mask; } -#define SCX_CALL_OP(mask, op, args...) \ +/* + * Track the rq currently locked. + * + * This allows kfuncs to safely operate on rq from any scx ops callback, + * knowing which rq is already locked. + */ +DEFINE_PER_CPU(struct rq *, scx_locked_rq_state); + +static inline void update_locked_rq(struct rq *rq) +{ + /* + * Check whether @rq is actually locked. This can help expose bugs + * or incorrect assumptions about the context in which a kfunc or + * callback is executed. + */ + if (rq) + lockdep_assert_rq_held(rq); + __this_cpu_write(scx_locked_rq_state, rq); +} + +#define SCX_CALL_OP(sch, mask, op, rq, args...) \ do { \ + if (rq) \ + update_locked_rq(rq); \ if (mask) { \ scx_kf_allow(mask); \ - scx_ops.op(args); \ + (sch)->ops.op(args); \ scx_kf_disallow(mask); \ } else { \ - scx_ops.op(args); \ + (sch)->ops.op(args); \ } \ + if (rq) \ + update_locked_rq(NULL); \ } while (0) -#define SCX_CALL_OP_RET(mask, op, args...) \ +#define SCX_CALL_OP_RET(sch, mask, op, rq, args...) \ ({ \ - __typeof__(scx_ops.op(args)) __ret; \ + __typeof__((sch)->ops.op(args)) __ret; \ + \ + if (rq) \ + update_locked_rq(rq); \ if (mask) { \ scx_kf_allow(mask); \ - __ret = scx_ops.op(args); \ + __ret = (sch)->ops.op(args); \ scx_kf_disallow(mask); \ } else { \ - __ret = scx_ops.op(args); \ + __ret = (sch)->ops.op(args); \ } \ + if (rq) \ + update_locked_rq(NULL); \ __ret; \ }) @@ -1129,42 +344,42 @@ do { \ * scx_kf_allowed_on_arg_tasks() to test whether the invocation is allowed on * the specific task. */ -#define SCX_CALL_OP_TASK(mask, op, task, args...) \ +#define SCX_CALL_OP_TASK(sch, mask, op, rq, task, args...) \ do { \ BUILD_BUG_ON((mask) & ~__SCX_KF_TERMINAL); \ current->scx.kf_tasks[0] = task; \ - SCX_CALL_OP(mask, op, task, ##args); \ + SCX_CALL_OP((sch), mask, op, rq, task, ##args); \ current->scx.kf_tasks[0] = NULL; \ } while (0) -#define SCX_CALL_OP_TASK_RET(mask, op, task, args...) \ +#define SCX_CALL_OP_TASK_RET(sch, mask, op, rq, task, args...) \ ({ \ - __typeof__(scx_ops.op(task, ##args)) __ret; \ + __typeof__((sch)->ops.op(task, ##args)) __ret; \ BUILD_BUG_ON((mask) & ~__SCX_KF_TERMINAL); \ current->scx.kf_tasks[0] = task; \ - __ret = SCX_CALL_OP_RET(mask, op, task, ##args); \ + __ret = SCX_CALL_OP_RET((sch), mask, op, rq, task, ##args); \ current->scx.kf_tasks[0] = NULL; \ __ret; \ }) -#define SCX_CALL_OP_2TASKS_RET(mask, op, task0, task1, args...) \ +#define SCX_CALL_OP_2TASKS_RET(sch, mask, op, rq, task0, task1, args...) \ ({ \ - __typeof__(scx_ops.op(task0, task1, ##args)) __ret; \ + __typeof__((sch)->ops.op(task0, task1, ##args)) __ret; \ BUILD_BUG_ON((mask) & ~__SCX_KF_TERMINAL); \ current->scx.kf_tasks[0] = task0; \ current->scx.kf_tasks[1] = task1; \ - __ret = SCX_CALL_OP_RET(mask, op, task0, task1, ##args); \ + __ret = SCX_CALL_OP_RET((sch), mask, op, rq, task0, task1, ##args); \ current->scx.kf_tasks[0] = NULL; \ current->scx.kf_tasks[1] = NULL; \ __ret; \ }) /* @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_ops_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; } @@ -1177,13 +392,13 @@ 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_ops_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_ops_error("dispatch kfunc called from a nested operation"); + scx_error(sch, "dispatch kfunc called from a nested operation"); return false; } @@ -1191,29 +406,25 @@ 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_ops_error("called on a task not being operated on"); + scx_error(sch, "called on a task not being operated on"); return false; } return true; } -static bool scx_kf_allowed_if_unlocked(void) -{ - return !current->scx.kf_mask; -} - /** * nldsq_next_task - Iterate to the next task in a non-local DSQ - * @dsq: user dsq being interated + * @dsq: user dsq being iterated * @cur: current position, %NULL to start iteration * @rev: walk backwards * @@ -1290,10 +501,11 @@ 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; }; /** @@ -1310,26 +522,24 @@ struct scx_task_iter { * RCU read lock or obtaining a reference count. * * All tasks which existed when the iteration started are guaranteed to be - * visited as long as they still exist. + * visited as long as they are not dead. */ static void scx_task_iter_start(struct scx_task_iter *iter) { - BUILD_BUG_ON(__SCX_DSQ_ITER_ALL_FLAGS & - ((1U << __SCX_DSQ_LNODE_PRIV_SHIFT) - 1)); + memset(iter, 0, sizeof(*iter)); - spin_lock_irq(&scx_tasks_lock); + raw_spin_lock_irq(&scx_tasks_lock); iter->cursor = (struct sched_ext_entity){ .flags = SCX_TASK_CURSOR }; list_add(&iter->cursor.tasks_node, &scx_tasks); - iter->locked = 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; } } @@ -1339,24 +549,24 @@ 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; + raw_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) { + raw_spin_lock_irq(&scx_tasks_lock); + iter->list_locked = true; + } } /** @@ -1369,6 +579,7 @@ 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); } @@ -1378,20 +589,21 @@ static void scx_task_iter_stop(struct scx_task_iter *iter) * @iter: iterator to walk * * Visit the next task. See scx_task_iter_start() for details. Locks are dropped - * and re-acquired every %SCX_OPS_TASK_ITER_BATCH iterations to avoid causing - * stalls by holding scx_tasks_lock for too long. + * and re-acquired every %SCX_TASK_ITER_BATCH iterations to avoid causing stalls + * by holding scx_tasks_lock for too long. */ 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; - if (!(++iter->cnt % SCX_OPS_TASK_ITER_BATCH)) { + 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) { if (&pos->tasks_node == &scx_tasks) return NULL; @@ -1408,7 +620,6 @@ static struct task_struct *scx_task_iter_next(struct scx_task_iter *iter) /** * scx_task_iter_next_locked - Next non-idle task with its rq locked * @iter: iterator to walk - * @include_dead: Whether we should include dead tasks in the iteration * * Visit the non-idle task with its rq lock held. Allows callers to specify * whether they would like to filter out dead tasks. See scx_task_iter_start() @@ -1453,33 +664,77 @@ 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; } -static enum scx_ops_enable_state scx_ops_enable_state(void) +/** + * scx_add_event - Increase an event counter for 'name' by 'cnt' + * @sch: scx_sched to account events for + * @name: an event name defined in struct scx_event_stats + * @cnt: the number of the event occurred + * + * This can be used when preemption is not disabled. + */ +#define scx_add_event(sch, name, cnt) do { \ + this_cpu_add((sch)->pcpu->event_stats.name, (cnt)); \ + trace_sched_ext_event(#name, (cnt)); \ +} while(0) + +/** + * __scx_add_event - Increase an event counter for 'name' by 'cnt' + * @sch: scx_sched to account events for + * @name: an event name defined in struct scx_event_stats + * @cnt: the number of the event occurred + * + * This should be used only when preemption is disabled. + */ +#define __scx_add_event(sch, name, cnt) do { \ + __this_cpu_add((sch)->pcpu->event_stats.name, (cnt)); \ + trace_sched_ext_event(#name, cnt); \ +} while(0) + +/** + * scx_agg_event - Aggregate an event counter 'kind' from 'src_e' to 'dst_e' + * @dst_e: destination event stats + * @src_e: source event stats + * @kind: a kind of event to be aggregated + */ +#define scx_agg_event(dst_e, src_e, kind) do { \ + (dst_e)->kind += READ_ONCE((src_e)->kind); \ +} while(0) + +/** + * scx_dump_event - Dump an event 'kind' in 'events' to 's' + * @s: output seq_buf + * @events: event stats + * @kind: a kind of event to dump + */ +#define scx_dump_event(s, events, kind) do { \ + dump_line(&(s), "%40s: %16lld", #kind, (events)->kind); \ +} while (0) + + +static void scx_read_events(struct scx_sched *sch, + struct scx_event_stats *events); + +static enum scx_enable_state scx_enable_state(void) { - return atomic_read(&scx_ops_enable_state_var); + return atomic_read(&scx_enable_state_var); } -static enum scx_ops_enable_state -scx_ops_set_enable_state(enum scx_ops_enable_state to) +static enum scx_enable_state scx_set_enable_state(enum scx_enable_state to) { - return atomic_xchg(&scx_ops_enable_state_var, to); + return atomic_xchg(&scx_enable_state_var, to); } -static bool scx_ops_tryset_enable_state(enum scx_ops_enable_state to, - enum scx_ops_enable_state from) +static bool scx_tryset_enable_state(enum scx_enable_state to, + enum scx_enable_state from) { int from_v = from; - return atomic_try_cmpxchg(&scx_ops_enable_state_var, &from_v, to); -} - -static bool scx_rq_bypassing(struct rq *rq) -{ - return unlikely(rq->scx.flags & SCX_RQ_BYPASSING); + return atomic_try_cmpxchg(&scx_enable_state_var, &from_v, to); } /** @@ -1499,8 +754,14 @@ static void wait_ops_state(struct task_struct *p, unsigned long opss) } while (atomic_long_read_acquire(&p->scx.ops_state) == opss); } +static inline bool __cpu_valid(s32 cpu) +{ + return likely(cpu >= 0 && cpu < nr_cpu_ids && cpu_possible(cpu)); +} + /** - * ops_cpu_valid - Verify a cpu number + * ops_cpu_valid - Verify a cpu number, to be used on ops input args + * @sch: scx_sched to abort on error * @cpu: cpu number which came from a BPF ops * @where: extra information reported on error * @@ -1508,49 +769,52 @@ static void wait_ops_state(struct task_struct *p, unsigned long opss) * Verify that it is in range and one of the possible cpus. If invalid, trigger * an ops error. */ -static bool ops_cpu_valid(s32 cpu, const char *where) +static bool ops_cpu_valid(struct scx_sched *sch, s32 cpu, const char *where) { - if (likely(cpu >= 0 && cpu < nr_cpu_ids && cpu_possible(cpu))) { + if (__cpu_valid(cpu)) { return true; } else { - scx_ops_error("invalid CPU %d%s%s", cpu, - where ? " " : "", where ?: ""); + scx_error(sch, "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 * @err: -errno value to sanitize * - * Verify @err is a valid -errno. If not, trigger scx_ops_error() and return + * Verify @err is a valid -errno. If not, trigger scx_error() and return * -%EPROTO. This is necessary because returning a rogue -errno up the chain can * cause misbehaviors. For an example, a large negative return from * ops.init_task() triggers an oops when passed up the call chain because the * value fails IS_ERR() test after being encoded with ERR_PTR() and then is * handled as a pointer. */ -static int ops_sanitize_err(const char *ops_name, s32 err) +static int ops_sanitize_err(struct scx_sched *sch, const char *ops_name, s32 err) { if (err < 0 && err >= -MAX_ERRNO) return err; - scx_ops_error("ops.%s() returned an invalid errno %d", ops_name, err); + scx_error(sch, "ops.%s() returned an invalid errno %d", ops_name, err); return -EPROTO; } static void run_deferred(struct rq *rq) { process_ddsp_deferred_locals(rq); + + if (local_read(&rq->scx.reenq_local_deferred)) { + local_set(&rq->scx.reenq_local_deferred, 0); + reenq_local(rq); + } } -#ifdef CONFIG_SMP static void deferred_bal_cb_workfn(struct rq *rq) { run_deferred(rq); } -#endif static void deferred_irq_workfn(struct irq_work *irq_work) { @@ -1565,15 +829,30 @@ static void deferred_irq_workfn(struct irq_work *irq_work) * schedule_deferred - Schedule execution of deferred actions on an rq * @rq: target rq * - * Schedule execution of deferred actions on @rq. Must be called with @rq - * locked. Deferred actions are executed with @rq locked but unpinned, and thus - * can unlock @rq to e.g. migrate tasks to other rqs. + * Schedule execution of deferred actions on @rq. Deferred actions are executed + * with @rq locked but unpinned, and thus can unlock @rq to e.g. migrate tasks + * to other rqs. */ static void schedule_deferred(struct rq *rq) { + /* + * Queue an irq work. They are executed on IRQ re-enable which may take + * a bit longer than the scheduler hook in schedule_deferred_locked(). + */ + irq_work_queue(&rq->scx.deferred_irq_work); +} + +/** + * schedule_deferred_locked - Schedule execution of deferred actions on an rq + * @rq: target rq + * + * Schedule execution of deferred actions on @rq. Equivalent to + * schedule_deferred() but requires @rq to be locked and can be more efficient. + */ +static void schedule_deferred_locked(struct rq *rq) +{ lockdep_assert_rq_held(rq); -#ifdef CONFIG_SMP /* * If in the middle of waking up a task, task_woken_scx() will be called * afterwards which will then run the deferred actions, no need to @@ -1582,23 +861,32 @@ static void schedule_deferred(struct rq *rq) if (rq->scx.flags & SCX_RQ_IN_WAKEUP) return; + /* Don't do anything if there already is a deferred operation. */ + if (rq->scx.flags & SCX_RQ_BAL_CB_PENDING) + return; + /* * If in balance, the balance callbacks will be called before rq lock is * released. Schedule one. + * + * + * We can't directly insert the callback into the + * rq's list: The call can drop its lock and make the pending balance + * callback visible to unrelated code paths that call rq_pin_lock(). + * + * Just let balance_one() know that it must do it itself. */ if (rq->scx.flags & SCX_RQ_IN_BALANCE) { - queue_balance_callback(rq, &rq->scx.deferred_bal_cb, - deferred_bal_cb_workfn); + rq->scx.flags |= SCX_RQ_BAL_CB_PENDING; return; } -#endif + /* - * No scheduler hooks available. Queue an irq work. They are executed on - * IRQ re-enable which may take a bit longer than the scheduler hooks. - * The above WAKEUP and BALANCE paths should cover most of the cases and - * the time to IRQ re-enable shouldn't be long. + * No scheduler hooks available. Use the generic irq_work path. The + * above WAKEUP and BALANCE paths should cover most of the cases and the + * time to IRQ re-enable shouldn't be long. */ - irq_work_queue(&rq->scx.deferred_irq_work); + schedule_deferred(rq); } /** @@ -1643,7 +931,7 @@ static void touch_core_sched_dispatch(struct rq *rq, struct task_struct *p) lockdep_assert_rq_held(rq); #ifdef CONFIG_SCHED_CORE - if (SCX_HAS_OP(core_sched_before)) + if (unlikely(SCX_HAS_OP(scx_root, core_sched_before))) touch_core_sched(rq, p); #endif } @@ -1681,8 +969,14 @@ static void dsq_mod_nr(struct scx_dispatch_q *dsq, s32 delta) WRITE_ONCE(dsq->nr, dsq->nr + delta); } -static void dispatch_enqueue(struct scx_dispatch_q *dsq, struct task_struct *p, - u64 enq_flags) +static void refill_task_slice_dfl(struct scx_sched *sch, struct task_struct *p) +{ + p->scx.slice = READ_ONCE(scx_slice_dfl); + __scx_add_event(sch, SCX_EV_REFILL_SLICE_DFL, 1); +} + +static void dispatch_enqueue(struct scx_sched *sch, struct scx_dispatch_q *dsq, + struct task_struct *p, u64 enq_flags) { bool is_local = dsq->id == SCX_DSQ_LOCAL; @@ -1691,12 +985,14 @@ static void dispatch_enqueue(struct scx_dispatch_q *dsq, struct task_struct *p, !RB_EMPTY_NODE(&p->scx.dsq_priq)); if (!is_local) { - raw_spin_lock(&dsq->lock); + raw_spin_lock_nested(&dsq->lock, + (enq_flags & SCX_ENQ_NESTED) ? SINGLE_DEPTH_NESTING : 0); + if (unlikely(dsq->id == SCX_DSQ_INVALID)) { - scx_ops_error("attempting to dispatch to a destroyed 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); } } @@ -1710,7 +1006,7 @@ static void dispatch_enqueue(struct scx_dispatch_q *dsq, struct task_struct *p, * disallow any internal DSQ from doing vtime ordering of * tasks. */ - scx_ops_error("cannot use vtime ordering for built-in DSQs"); + scx_error(sch, "cannot use vtime ordering for built-in DSQs"); enq_flags &= ~SCX_ENQ_DSQ_PRIQ; } @@ -1724,8 +1020,8 @@ static void dispatch_enqueue(struct scx_dispatch_q *dsq, struct task_struct *p, */ if (unlikely(RB_EMPTY_ROOT(&dsq->priq) && nldsq_next_task(dsq, NULL, false))) - scx_ops_error("DSQ ID 0x%016llx already had FIFO-enqueued tasks", - dsq->id); + scx_error(sch, "DSQ ID 0x%016llx already had FIFO-enqueued tasks", + dsq->id); p->scx.dsq_flags |= SCX_TASK_DSQ_ON_PRIQ; rb_add(&p->scx.dsq_priq, &dsq->priq, scx_dsq_priq_less); @@ -1740,19 +1036,31 @@ static void dispatch_enqueue(struct scx_dispatch_q *dsq, struct task_struct *p, container_of(rbp, struct task_struct, scx.dsq_priq); list_add(&p->scx.dsq_list.node, &prev->scx.dsq_list.node); + /* first task unchanged - no update needed */ } else { list_add(&p->scx.dsq_list.node, &dsq->list); + /* not builtin and new task is at head - use fastpath */ + rcu_assign_pointer(dsq->first_task, p); } } else { /* a FIFO DSQ shouldn't be using PRIQ enqueuing */ if (unlikely(!RB_EMPTY_ROOT(&dsq->priq))) - scx_ops_error("DSQ ID 0x%016llx already had PRIQ-enqueued tasks", - dsq->id); + scx_error(sch, "DSQ ID 0x%016llx already had PRIQ-enqueued tasks", + dsq->id); - if (enq_flags & (SCX_ENQ_HEAD | SCX_ENQ_PREEMPT)) + if (enq_flags & (SCX_ENQ_HEAD | SCX_ENQ_PREEMPT)) { list_add(&p->scx.dsq_list.node, &dsq->list); - else + /* new task inserted at head - use fastpath */ + if (!(dsq->id & SCX_DSQ_FLAG_BUILTIN)) + rcu_assign_pointer(dsq->first_task, p); + } else { + bool was_empty; + + was_empty = list_empty(&dsq->list); list_add_tail(&p->scx.dsq_list.node, &dsq->list); + if (was_empty && !(dsq->id & SCX_DSQ_FLAG_BUILTIN)) + rcu_assign_pointer(dsq->first_task, p); + } } /* seq records the order tasks are queued, used by BPF DSQ iterator */ @@ -1809,6 +1117,13 @@ static void task_unlink_from_dsq(struct task_struct *p, list_del_init(&p->scx.dsq_list.node); dsq_mod_nr(dsq, -1); + + if (!(dsq->id & SCX_DSQ_FLAG_BUILTIN) && dsq->first_task == p) { + struct task_struct *first_task; + + first_task = nldsq_next_task(dsq, NULL, false); + rcu_assign_pointer(dsq->first_task, first_task); + } } static void dispatch_dequeue(struct rq *rq, struct task_struct *p) @@ -1816,6 +1131,8 @@ static void dispatch_dequeue(struct rq *rq, struct task_struct *p) struct scx_dispatch_q *dsq = p->scx.dsq; bool is_local = dsq == &rq->scx.local_dsq; + lockdep_assert_rq_held(rq); + if (!dsq) { /* * If !dsq && on-list, @p is on @rq's ddsp_deferred_locals. @@ -1862,7 +1179,22 @@ static void dispatch_dequeue(struct rq *rq, struct task_struct *p) raw_spin_unlock(&dsq->lock); } -static struct scx_dispatch_q *find_dsq_for_dispatch(struct rq *rq, u64 dsq_id, +/* + * Abbreviated version of dispatch_dequeue() that can be used when both @p's rq + * and dsq are locked. + */ +static void dispatch_dequeue_locked(struct task_struct *p, + struct scx_dispatch_q *dsq) +{ + lockdep_assert_rq_held(task_rq(p)); + lockdep_assert_held(&dsq->lock); + + task_unlink_from_dsq(p, dsq); + p->scx.dsq = NULL; +} + +static struct scx_dispatch_q *find_dsq_for_dispatch(struct scx_sched *sch, + struct rq *rq, u64 dsq_id, struct task_struct *p) { struct scx_dispatch_q *dsq; @@ -1873,27 +1205,28 @@ static struct scx_dispatch_q *find_dsq_for_dispatch(struct rq *rq, u64 dsq_id, if ((dsq_id & SCX_DSQ_LOCAL_ON) == SCX_DSQ_LOCAL_ON) { s32 cpu = dsq_id & SCX_DSQ_LOCAL_CPU_MASK; - if (!ops_cpu_valid(cpu, "in SCX_DSQ_LOCAL_ON dispatch verdict")) - return find_global_dsq(p); + if (!ops_cpu_valid(sch, cpu, "in SCX_DSQ_LOCAL_ON dispatch verdict")) + 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(dsq_id); + dsq = find_user_dsq(sch, dsq_id); if (unlikely(!dsq)) { - scx_ops_error("non-existent DSQ 0x%llx for %s[%d]", - dsq_id, p->comm, p->pid); - return find_global_dsq(p); + scx_error(sch, "non-existent DSQ 0x%llx for %s[%d]", + dsq_id, p->comm, p->pid); + 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) { @@ -1907,12 +1240,12 @@ 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_ops_error("%s[%d] already direct-dispatched", - p->comm, p->pid); + scx_error(sch, "%s[%d] already direct-dispatched", + p->comm, p->pid); else - scx_ops_error("scheduling for %s[%d] but trying to direct-dispatch %s[%d]", - ddsp_task->comm, ddsp_task->pid, - p->comm, p->pid); + 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; } @@ -1923,11 +1256,12 @@ static void mark_direct_dispatch(struct task_struct *ddsp_task, p->scx.ddsp_enq_flags = enq_flags; } -static void direct_dispatch(struct task_struct *p, u64 enq_flags) +static void direct_dispatch(struct scx_sched *sch, struct task_struct *p, + u64 enq_flags) { struct rq *rq = task_rq(p); struct scx_dispatch_q *dsq = - find_dsq_for_dispatch(rq, p->scx.ddsp_dsq_id, p); + find_dsq_for_dispatch(sch, rq, p->scx.ddsp_dsq_id, p); touch_core_sched_dispatch(rq, p); @@ -1964,11 +1298,12 @@ static void direct_dispatch(struct task_struct *p, u64 enq_flags) WARN_ON_ONCE(p->scx.dsq || !list_empty(&p->scx.dsq_list.node)); list_add_tail(&p->scx.dsq_list.node, &rq->scx.ddsp_deferred_locals); - schedule_deferred(rq); + schedule_deferred_locked(rq); return; } - dispatch_enqueue(dsq, p, p->scx.ddsp_enq_flags | SCX_ENQ_CLEAR_OPSS); + dispatch_enqueue(sch, dsq, p, + p->scx.ddsp_enq_flags | SCX_ENQ_CLEAR_OPSS); } static bool scx_rq_online(struct rq *rq) @@ -1986,7 +1321,9 @@ static bool scx_rq_online(struct rq *rq) static void do_enqueue_task(struct rq *rq, struct task_struct *p, u64 enq_flags, int sticky_cpu) { + struct scx_sched *sch = scx_root; struct task_struct **ddsp_taskp; + struct scx_dispatch_q *dsq; unsigned long qseq; WARN_ON_ONCE(!(p->scx.flags & SCX_TASK_QUEUED)); @@ -2003,18 +1340,29 @@ static void do_enqueue_task(struct rq *rq, struct task_struct *p, u64 enq_flags, if (!scx_rq_online(rq)) goto local; - if (scx_rq_bypassing(rq)) - goto global; + if (scx_rq_bypassing(rq)) { + __scx_add_event(sch, SCX_EV_BYPASS_DISPATCH, 1); + goto bypass; + } if (p->scx.ddsp_dsq_id != SCX_DSQ_INVALID) goto direct; /* see %SCX_OPS_ENQ_EXITING */ - if (!static_branch_unlikely(&scx_ops_enq_exiting) && - unlikely(p->flags & PF_EXITING)) + if (!(sch->ops.flags & SCX_OPS_ENQ_EXITING) && + unlikely(p->flags & PF_EXITING)) { + __scx_add_event(sch, SCX_EV_ENQ_SKIP_EXITING, 1); + goto local; + } + + /* see %SCX_OPS_ENQ_MIGRATION_DISABLED */ + if (!(sch->ops.flags & SCX_OPS_ENQ_MIGRATION_DISABLED) && + is_migration_disabled(p)) { + __scx_add_event(sch, SCX_EV_ENQ_SKIP_MIGRATION_DISABLED, 1); goto local; + } - if (!SCX_HAS_OP(enqueue)) + if (unlikely(!SCX_HAS_OP(sch, enqueue))) goto global; /* DSQ bypass didn't trigger, enqueue on the BPF scheduler */ @@ -2027,7 +1375,7 @@ static void do_enqueue_task(struct rq *rq, struct task_struct *p, u64 enq_flags, WARN_ON_ONCE(*ddsp_taskp); *ddsp_taskp = p; - SCX_CALL_OP_TASK(SCX_KF_ENQUEUE, enqueue, p, enq_flags); + SCX_CALL_OP_TASK(sch, SCX_KF_ENQUEUE, enqueue, rq, p, enq_flags); *ddsp_taskp = NULL; if (p->scx.ddsp_dsq_id != SCX_DSQ_INVALID) @@ -2041,25 +1389,30 @@ static void do_enqueue_task(struct rq *rq, struct task_struct *p, u64 enq_flags, return; direct: - direct_dispatch(p, enq_flags); + direct_dispatch(sch, p, enq_flags); + return; +local_norefill: + dispatch_enqueue(sch, &rq->scx.local_dsq, p, enq_flags); return; - local: + dsq = &rq->scx.local_dsq; + goto enqueue; +global: + dsq = find_global_dsq(sch, p); + goto enqueue; +bypass: + dsq = &task_rq(p)->scx.bypass_dsq; + goto enqueue; + +enqueue: /* * For task-ordering, slice refill must be treated as implying the end * of the current slice. Otherwise, the longer @p stays on the CPU, the * higher priority it becomes from scx_prio_less()'s POV. */ touch_core_sched(rq, p); - p->scx.slice = SCX_SLICE_DFL; -local_norefill: - dispatch_enqueue(&rq->scx.local_dsq, p, enq_flags); - return; - -global: - touch_core_sched(rq, p); /* see the comment in local: */ - p->scx.slice = SCX_SLICE_DFL; - dispatch_enqueue(find_global_dsq(p), p, enq_flags); + refill_task_slice_dfl(sch, p); + dispatch_enqueue(sch, dsq, p, enq_flags); } static bool task_runnable(const struct task_struct *p) @@ -2077,8 +1430,8 @@ static void set_task_runnable(struct rq *rq, struct task_struct *p) } /* - * list_add_tail() must be used. scx_ops_bypass() depends on tasks being - * appened to the runnable_list. + * list_add_tail() must be used. scx_bypass() depends on tasks being + * appended to the runnable_list. */ list_add_tail(&p->scx.runnable_node, &rq->scx.runnable_list); } @@ -2092,6 +1445,7 @@ static void clr_task_runnable(struct task_struct *p, bool reset_runnable_at) static void enqueue_task_scx(struct rq *rq, struct task_struct *p, int enq_flags) { + struct scx_sched *sch = scx_root; int sticky_cpu = p->scx.sticky_cpu; if (enq_flags & ENQUEUE_WAKEUP) @@ -2121,8 +1475,8 @@ static void enqueue_task_scx(struct rq *rq, struct task_struct *p, int enq_flags rq->scx.nr_running++; add_nr_running(rq, 1); - if (SCX_HAS_OP(runnable) && !task_on_rq_migrating(p)) - SCX_CALL_OP_TASK(SCX_KF_REST, runnable, p, enq_flags); + if (SCX_HAS_OP(sch, runnable) && !task_on_rq_migrating(p)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, runnable, rq, p, enq_flags); if (enq_flags & SCX_ENQ_WAKEUP) touch_core_sched(rq, p); @@ -2130,10 +1484,15 @@ static void enqueue_task_scx(struct rq *rq, struct task_struct *p, int enq_flags do_enqueue_task(rq, p, enq_flags, sticky_cpu); out: rq->scx.flags &= ~SCX_RQ_IN_WAKEUP; + + if ((enq_flags & SCX_ENQ_CPU_SELECTED) && + unlikely(cpu_of(rq) != p->scx.selected_cpu)) + __scx_add_event(sch, SCX_EV_SELECT_CPU_FALLBACK, 1); } -static void ops_dequeue(struct task_struct *p, u64 deq_flags) +static void ops_dequeue(struct rq *rq, struct task_struct *p, u64 deq_flags) { + struct scx_sched *sch = scx_root; unsigned long opss; /* dequeue is always temporary, don't reset runnable_at */ @@ -2152,8 +1511,9 @@ static void ops_dequeue(struct task_struct *p, u64 deq_flags) */ BUG(); case SCX_OPSS_QUEUED: - if (SCX_HAS_OP(dequeue)) - SCX_CALL_OP_TASK(SCX_KF_REST, dequeue, p, deq_flags); + if (SCX_HAS_OP(sch, dequeue)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, dequeue, rq, + p, deq_flags); if (atomic_long_try_cmpxchg(&p->scx.ops_state, &opss, SCX_OPSS_NONE)) @@ -2181,12 +1541,14 @@ static void ops_dequeue(struct task_struct *p, u64 deq_flags) static bool dequeue_task_scx(struct rq *rq, struct task_struct *p, int deq_flags) { + struct scx_sched *sch = scx_root; + if (!(p->scx.flags & SCX_TASK_QUEUED)) { WARN_ON_ONCE(task_runnable(p)); return true; } - ops_dequeue(p, deq_flags); + ops_dequeue(rq, p, deq_flags); /* * A currently running task which is going off @rq first gets dequeued @@ -2200,13 +1562,13 @@ static bool dequeue_task_scx(struct rq *rq, struct task_struct *p, int deq_flags * information meaningful to the BPF scheduler and can be suppressed by * skipping the callbacks if the task is !QUEUED. */ - if (SCX_HAS_OP(stopping) && task_current(rq, p)) { + if (SCX_HAS_OP(sch, stopping) && task_current(rq, p)) { update_curr_scx(rq); - SCX_CALL_OP_TASK(SCX_KF_REST, stopping, p, false); + SCX_CALL_OP_TASK(sch, SCX_KF_REST, stopping, rq, p, false); } - if (SCX_HAS_OP(quiescent) && !task_on_rq_migrating(p)) - SCX_CALL_OP_TASK(SCX_KF_REST, quiescent, p, deq_flags); + if (SCX_HAS_OP(sch, quiescent) && !task_on_rq_migrating(p)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, quiescent, rq, p, deq_flags); if (deq_flags & SCX_DEQ_SLEEP) p->scx.flags |= SCX_TASK_DEQD_FOR_SLEEP; @@ -2223,20 +1585,23 @@ static bool dequeue_task_scx(struct rq *rq, struct task_struct *p, int deq_flags static void yield_task_scx(struct rq *rq) { - struct task_struct *p = rq->curr; + struct scx_sched *sch = scx_root; + struct task_struct *p = rq->donor; - if (SCX_HAS_OP(yield)) - SCX_CALL_OP_2TASKS_RET(SCX_KF_REST, yield, p, NULL); + if (SCX_HAS_OP(sch, yield)) + SCX_CALL_OP_2TASKS_RET(sch, SCX_KF_REST, yield, rq, p, NULL); else p->scx.slice = 0; } static bool yield_to_task_scx(struct rq *rq, struct task_struct *to) { - struct task_struct *from = rq->curr; + struct scx_sched *sch = scx_root; + struct task_struct *from = rq->donor; - if (SCX_HAS_OP(yield)) - return SCX_CALL_OP_2TASKS_RET(SCX_KF_REST, yield, from, to); + if (SCX_HAS_OP(sch, yield)) + return SCX_CALL_OP_2TASKS_RET(sch, SCX_KF_REST, yield, rq, + from, to); else return false; } @@ -2262,7 +1627,6 @@ static void move_local_task_to_local_dsq(struct task_struct *p, u64 enq_flags, p->scx.dsq = dst_dsq; } -#ifdef CONFIG_SMP /** * move_remote_task_to_local_dsq - Move a task from a foreign rq to a local DSQ * @p: task to move @@ -2313,12 +1677,36 @@ static void move_remote_task_to_local_dsq(struct task_struct *p, u64 enq_flags, * * - The BPF scheduler is bypassed while the rq is offline and we can always say * no to the BPF scheduler initiated migrations while offline. + * + * The caller must ensure that @p and @rq are on different CPUs. */ -static bool task_can_run_on_remote_rq(struct task_struct *p, struct rq *rq, - bool trigger_error) +static bool task_can_run_on_remote_rq(struct scx_sched *sch, + struct task_struct *p, struct rq *rq, + bool enforce) { int cpu = cpu_of(rq); + WARN_ON_ONCE(task_cpu(p) == cpu); + + /* + * If @p has migration disabled, @p->cpus_ptr is updated to contain only + * the pinned CPU in migrate_disable_switch() while @p is being switched + * out. However, put_prev_task_scx() is called before @p->cpus_ptr is + * updated and thus another CPU may see @p on a DSQ inbetween leading to + * @p passing the below task_allowed_on_cpu() check while migration is + * disabled. + * + * Test the migration disabled state first as the race window is narrow + * and the BPF scheduler failing to check migration disabled state can + * easily be masked if task_allowed_on_cpu() is done first. + */ + if (unlikely(is_migration_disabled(p))) { + if (enforce) + scx_error(sch, "SCX_DSQ_LOCAL[_ON] cannot move migration disabled %s[%d] from CPU %d to %d", + p->comm, p->pid, task_cpu(p), cpu); + return false; + } + /* * We don't require the BPF scheduler to avoid dispatching to offline * CPUs mostly for convenience but also because CPUs can go offline @@ -2326,17 +1714,17 @@ static bool task_can_run_on_remote_rq(struct task_struct *p, struct rq *rq, * picked CPU is outside the allowed mask. */ if (!task_allowed_on_cpu(p, cpu)) { - if (trigger_error) - scx_ops_error("SCX_DSQ_LOCAL[_ON] verdict target cpu %d not allowed for %s[%d]", - cpu_of(rq), p->comm, p->pid); + if (enforce) + scx_error(sch, "SCX_DSQ_LOCAL[_ON] target CPU %d not allowed for %s[%d]", + cpu, p->comm, p->pid); return false; } - if (unlikely(is_migration_disabled(p))) - return false; - - if (!scx_rq_online(rq)) + if (!scx_rq_online(rq)) { + if (enforce) + __scx_add_event(sch, SCX_EV_DISPATCH_LOCAL_DSQ_OFFLINE, 1); return false; + } return true; } @@ -2404,14 +1792,10 @@ static bool consume_remote_task(struct rq *this_rq, struct task_struct *p, return false; } } -#else /* CONFIG_SMP */ -static inline void move_remote_task_to_local_dsq(struct task_struct *p, u64 enq_flags, struct rq *src_rq, struct rq *dst_rq) { WARN_ON_ONCE(1); } -static inline bool task_can_run_on_remote_rq(struct task_struct *p, struct rq *rq, bool trigger_error) { return false; } -static inline bool consume_remote_task(struct rq *this_rq, struct task_struct *p, struct scx_dispatch_q *dsq, struct rq *task_rq) { return false; } -#endif /* CONFIG_SMP */ /** * move_task_between_dsqs() - Move a task from one DSQ to another + * @sch: scx_sched being operated on * @p: target task * @enq_flags: %SCX_ENQ_* * @src_dsq: DSQ @p is currently on, must not be a local DSQ @@ -2425,7 +1809,8 @@ static inline bool consume_remote_task(struct rq *this_rq, struct task_struct *p * On return, @src_dsq is unlocked and only @p's new task_rq, which is the * return value, is locked. */ -static struct rq *move_task_between_dsqs(struct task_struct *p, u64 enq_flags, +static struct rq *move_task_between_dsqs(struct scx_sched *sch, + struct task_struct *p, u64 enq_flags, struct scx_dispatch_q *src_dsq, struct scx_dispatch_q *dst_dsq) { @@ -2437,8 +1822,9 @@ static struct rq *move_task_between_dsqs(struct task_struct *p, u64 enq_flags, if (dst_dsq->id == SCX_DSQ_LOCAL) { dst_rq = container_of(dst_dsq, struct rq, scx.local_dsq); - if (!task_can_run_on_remote_rq(p, dst_rq, true)) { - dst_dsq = find_global_dsq(p); + if (src_rq != dst_rq && + unlikely(!task_can_run_on_remote_rq(sch, p, dst_rq, true))) { + dst_dsq = find_global_dsq(sch, p); dst_rq = src_rq; } } else { @@ -2467,58 +1853,21 @@ static struct rq *move_task_between_dsqs(struct task_struct *p, u64 enq_flags, * @p is going from a non-local DSQ to a non-local DSQ. As * $src_dsq is already locked, do an abbreviated dequeue. */ - task_unlink_from_dsq(p, src_dsq); - p->scx.dsq = NULL; + dispatch_dequeue_locked(p, src_dsq); raw_spin_unlock(&src_dsq->lock); - dispatch_enqueue(dst_dsq, p, enq_flags); + dispatch_enqueue(sch, dst_dsq, p, enq_flags); } return dst_rq; } -/* - * A poorly behaving BPF scheduler can live-lock the system by e.g. incessantly - * banging on the same DSQ on a large NUMA system to the point where switching - * to the bypass mode can take a long time. Inject artifical delays while the - * bypass mode is switching to guarantee timely completion. - */ -static void scx_ops_breather(struct rq *rq) -{ - u64 until; - - lockdep_assert_rq_held(rq); - - if (likely(!atomic_read(&scx_ops_breather_depth))) - return; - - raw_spin_rq_unlock(rq); - - until = ktime_get_ns() + NSEC_PER_MSEC; - - do { - int cnt = 1024; - while (atomic_read(&scx_ops_breather_depth) && --cnt) - cpu_relax(); - } while (atomic_read(&scx_ops_breather_depth) && - time_before64(ktime_get_ns(), until)); - - raw_spin_rq_lock(rq); -} - -static bool consume_dispatch_q(struct rq *rq, struct scx_dispatch_q *dsq) +static bool consume_dispatch_q(struct scx_sched *sch, struct rq *rq, + struct scx_dispatch_q *dsq) { struct task_struct *p; retry: /* - * This retry loop can repeatedly race against scx_ops_bypass() - * dequeueing tasks from @dsq trying to put the system into the bypass - * mode. On some multi-socket machines (e.g. 2x Intel 8480c), this can - * live-lock the machine into soft lockups. Give a breather. - */ - scx_ops_breather(rq); - - /* * The caller can't expect to successfully consume a task if the task's * addition to @dsq isn't guaranteed to be visible somehow. Test * @dsq->list without locking and skip if it seems empty. @@ -2531,6 +1880,17 @@ retry: nldsq_for_each_task(p, dsq) { struct rq *task_rq = task_rq(p); + /* + * This loop can lead to multiple lockup scenarios, e.g. the BPF + * scheduler can put an enormous number of affinitized tasks into + * a contended DSQ, or the outer retry loop can repeatedly race + * against scx_bypass() dequeueing tasks from @dsq trying to put + * the system into the bypass mode. This can easily live-lock the + * machine. If aborting, exit from all non-bypass DSQs. + */ + if (unlikely(READ_ONCE(scx_aborting)) && dsq->id != SCX_DSQ_BYPASS) + break; + if (rq == task_rq) { task_unlink_from_dsq(p, dsq); move_local_task_to_local_dsq(p, 0, dsq, rq); @@ -2538,7 +1898,7 @@ retry: return true; } - if (task_can_run_on_remote_rq(p, rq, false)) { + if (task_can_run_on_remote_rq(sch, p, rq, false)) { if (likely(consume_remote_task(rq, p, dsq, task_rq))) return true; goto retry; @@ -2549,15 +1909,16 @@ retry: return false; } -static bool consume_global_dsq(struct rq *rq) +static bool consume_global_dsq(struct scx_sched *sch, struct rq *rq) { int node = cpu_to_node(cpu_of(rq)); - return consume_dispatch_q(rq, global_dsqs[node]); + return consume_dispatch_q(sch, rq, sch->global_dsqs[node]); } /** * dispatch_to_local_dsq - Dispatch a task to a local dsq + * @sch: scx_sched being operated on * @rq: current rq which is locked * @dst_dsq: destination DSQ * @p: task to dispatch @@ -2570,11 +1931,13 @@ static bool consume_global_dsq(struct rq *rq) * The caller must have exclusive ownership of @p (e.g. through * %SCX_OPSS_DISPATCHING). */ -static void dispatch_to_local_dsq(struct rq *rq, struct scx_dispatch_q *dst_dsq, +static void dispatch_to_local_dsq(struct scx_sched *sch, struct rq *rq, + struct scx_dispatch_q *dst_dsq, struct task_struct *p, u64 enq_flags) { struct rq *src_rq = task_rq(p); struct rq *dst_rq = container_of(dst_dsq, struct rq, scx.local_dsq); + struct rq *locked_rq = rq; /* * We're synchronized against dequeue through DISPATCHING. As @p can't @@ -2583,13 +1946,14 @@ static void dispatch_to_local_dsq(struct rq *rq, struct scx_dispatch_q *dst_dsq, * If dispatching to @rq that @p is already on, no lock dancing needed. */ if (rq == src_rq && rq == dst_rq) { - dispatch_enqueue(dst_dsq, p, enq_flags | SCX_ENQ_CLEAR_OPSS); + dispatch_enqueue(sch, dst_dsq, p, + enq_flags | SCX_ENQ_CLEAR_OPSS); return; } -#ifdef CONFIG_SMP - if (unlikely(!task_can_run_on_remote_rq(p, dst_rq, true))) { - dispatch_enqueue(find_global_dsq(p), p, + if (src_rq != dst_rq && + unlikely(!task_can_run_on_remote_rq(sch, p, dst_rq, true))) { + dispatch_enqueue(sch, find_global_dsq(sch, p), p, enq_flags | SCX_ENQ_CLEAR_OPSS); return; } @@ -2611,8 +1975,9 @@ static void dispatch_to_local_dsq(struct rq *rq, struct scx_dispatch_q *dst_dsq, atomic_long_set_release(&p->scx.ops_state, SCX_OPSS_NONE); /* switch to @src_rq lock */ - if (rq != src_rq) { - raw_spin_rq_unlock(rq); + if (locked_rq != src_rq) { + raw_spin_rq_unlock(locked_rq); + locked_rq = src_rq; raw_spin_rq_lock(src_rq); } @@ -2626,10 +1991,13 @@ static void dispatch_to_local_dsq(struct rq *rq, struct scx_dispatch_q *dst_dsq, */ if (src_rq == dst_rq) { p->scx.holding_cpu = -1; - dispatch_enqueue(&dst_rq->scx.local_dsq, p, enq_flags); + dispatch_enqueue(sch, &dst_rq->scx.local_dsq, p, + enq_flags); } else { move_remote_task_to_local_dsq(p, enq_flags, src_rq, dst_rq); + /* task has been moved to dst_rq, which is now locked */ + locked_rq = dst_rq; } /* if the destination CPU is idle, wake it up */ @@ -2638,13 +2006,10 @@ static void dispatch_to_local_dsq(struct rq *rq, struct scx_dispatch_q *dst_dsq, } /* switch back to @rq lock */ - if (rq != dst_rq) { - raw_spin_rq_unlock(dst_rq); + if (locked_rq != rq) { + raw_spin_rq_unlock(locked_rq); raw_spin_rq_lock(rq); } -#else /* CONFIG_SMP */ - BUG(); /* control can not reach here on UP */ -#endif /* CONFIG_SMP */ } /** @@ -2666,7 +2031,8 @@ static void dispatch_to_local_dsq(struct rq *rq, struct scx_dispatch_q *dst_dsq, * was valid in the first place. Make sure that the task is still owned by the * BPF scheduler and claim the ownership before dispatching. */ -static void finish_dispatch(struct rq *rq, struct task_struct *p, +static void finish_dispatch(struct scx_sched *sch, struct rq *rq, + struct task_struct *p, unsigned long qseq_at_dispatch, u64 dsq_id, u64 enq_flags) { @@ -2719,15 +2085,15 @@ retry: BUG_ON(!(p->scx.flags & SCX_TASK_QUEUED)); - dsq = find_dsq_for_dispatch(this_rq(), dsq_id, p); + dsq = find_dsq_for_dispatch(sch, this_rq(), dsq_id, p); if (dsq->id == SCX_DSQ_LOCAL) - dispatch_to_local_dsq(rq, dsq, p, enq_flags); + dispatch_to_local_dsq(sch, rq, dsq, p, enq_flags); else - dispatch_enqueue(dsq, p, enq_flags | SCX_ENQ_CLEAR_OPSS); + dispatch_enqueue(sch, dsq, p, enq_flags | SCX_ENQ_CLEAR_OPSS); } -static void flush_dispatch_buf(struct rq *rq) +static void flush_dispatch_buf(struct scx_sched *sch, struct rq *rq) { struct scx_dsp_ctx *dspc = this_cpu_ptr(scx_dsp_ctx); u32 u; @@ -2735,7 +2101,7 @@ static void flush_dispatch_buf(struct rq *rq) for (u = 0; u < dspc->cursor; u++) { struct scx_dsp_buf_ent *ent = &dspc->buf[u]; - finish_dispatch(rq, ent->task, ent->qseq, ent->dsq_id, + finish_dispatch(sch, rq, ent->task, ent->qseq, ent->dsq_id, ent->enq_flags); } @@ -2743,17 +2109,32 @@ static void flush_dispatch_buf(struct rq *rq) dspc->cursor = 0; } +static inline void maybe_queue_balance_callback(struct rq *rq) +{ + lockdep_assert_rq_held(rq); + + if (!(rq->scx.flags & SCX_RQ_BAL_CB_PENDING)) + return; + + queue_balance_callback(rq, &rq->scx.deferred_bal_cb, + deferred_bal_cb_workfn); + + rq->scx.flags &= ~SCX_RQ_BAL_CB_PENDING; +} + static int balance_one(struct rq *rq, struct task_struct *prev) { + struct scx_sched *sch = scx_root; struct scx_dsp_ctx *dspc = this_cpu_ptr(scx_dsp_ctx); bool prev_on_scx = prev->sched_class == &ext_sched_class; + bool prev_on_rq = prev->scx.flags & SCX_TASK_QUEUED; int nr_loops = SCX_DSP_MAX_LOOPS; lockdep_assert_rq_held(rq); rq->scx.flags |= SCX_RQ_IN_BALANCE; - rq->scx.flags &= ~(SCX_RQ_BAL_PENDING | SCX_RQ_BAL_KEEP); + rq->scx.flags &= ~SCX_RQ_BAL_KEEP; - if (static_branch_unlikely(&scx_ops_cpu_preempt) && + if ((sch->ops.flags & SCX_OPS_HAS_CPU_PREEMPT) && unlikely(rq->scx.cpu_released)) { /* * If the previous sched_class for the current CPU was not SCX, @@ -2761,8 +2142,9 @@ static int balance_one(struct rq *rq, struct task_struct *prev) * core. This callback complements ->cpu_release(), which is * emitted in switch_class(). */ - if (SCX_HAS_OP(cpu_acquire)) - SCX_CALL_OP(SCX_KF_REST, cpu_acquire, cpu_of(rq), NULL); + if (SCX_HAS_OP(sch, cpu_acquire)) + SCX_CALL_OP(sch, SCX_KF_REST, cpu_acquire, rq, + cpu_of(rq), NULL); rq->scx.cpu_released = false; } @@ -2776,11 +2158,10 @@ static int balance_one(struct rq *rq, struct task_struct *prev) * scheduler wants to handle this explicitly, it should * implement ->cpu_release(). * - * See scx_ops_disable_workfn() for the explanation on the - * bypassing test. + * See scx_disable_workfn() for the explanation on the bypassing + * test. */ - if ((prev->scx.flags & SCX_TASK_QUEUED) && - prev->scx.slice && !scx_rq_bypassing(rq)) { + if (prev_on_rq && prev->scx.slice && !scx_rq_bypassing(rq)) { rq->scx.flags |= SCX_RQ_BAL_KEEP; goto has_tasks; } @@ -2790,10 +2171,17 @@ static int balance_one(struct rq *rq, struct task_struct *prev) if (rq->scx.local_dsq.nr) goto has_tasks; - if (consume_global_dsq(rq)) + if (consume_global_dsq(sch, rq)) goto has_tasks; - if (!SCX_HAS_OP(dispatch) || scx_rq_bypassing(rq) || !scx_rq_online(rq)) + if (scx_rq_bypassing(rq)) { + if (consume_dispatch_q(sch, rq, &rq->scx.bypass_dsq)) + goto has_tasks; + else + goto no_tasks; + } + + if (unlikely(!SCX_HAS_OP(sch, dispatch)) || !scx_rq_online(rq)) goto no_tasks; dspc->rq = rq; @@ -2808,14 +2196,18 @@ static int balance_one(struct rq *rq, struct task_struct *prev) do { dspc->nr_tasks = 0; - SCX_CALL_OP(SCX_KF_DISPATCH, dispatch, cpu_of(rq), - prev_on_scx ? prev : NULL); + SCX_CALL_OP(sch, SCX_KF_DISPATCH, dispatch, rq, + cpu_of(rq), prev_on_scx ? prev : NULL); - flush_dispatch_buf(rq); + flush_dispatch_buf(sch, rq); + if (prev_on_rq && prev->scx.slice) { + rq->scx.flags |= SCX_RQ_BAL_KEEP; + goto has_tasks; + } if (rq->scx.local_dsq.nr) goto has_tasks; - if (consume_global_dsq(rq)) + if (consume_global_dsq(sch, rq)) goto has_tasks; /* @@ -2825,10 +2217,10 @@ 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); @@ -2838,10 +2230,10 @@ no_tasks: * Didn't find another task to run. Keep running @prev unless * %SCX_OPS_ENQ_LAST is in effect. */ - if ((prev->scx.flags & SCX_TASK_QUEUED) && - (!static_branch_unlikely(&scx_ops_enq_last) || - scx_rq_bypassing(rq))) { + if (prev_on_rq && + (!(sch->ops.flags & SCX_OPS_ENQ_LAST) || scx_rq_bypassing(rq))) { rq->scx.flags |= SCX_RQ_BAL_KEEP; + __scx_add_event(sch, SCX_EV_DISPATCH_KEEP_LAST, 1); goto has_tasks; } rq->scx.flags &= ~SCX_RQ_IN_BALANCE; @@ -2852,40 +2244,6 @@ has_tasks: return true; } -static int balance_scx(struct rq *rq, struct task_struct *prev, - struct rq_flags *rf) -{ - int ret; - - rq_unpin_lock(rq, rf); - - ret = balance_one(rq, prev); - -#ifdef CONFIG_SCHED_SMT - /* - * When core-sched is enabled, this ops.balance() call will be followed - * by pick_task_scx() on this CPU and the SMT siblings. Balance the - * siblings too. - */ - if (sched_core_enabled(rq)) { - const struct cpumask *smt_mask = cpu_smt_mask(cpu_of(rq)); - int scpu; - - for_each_cpu_andnot(scpu, smt_mask, cpumask_of(cpu_of(rq))) { - struct rq *srq = cpu_rq(scpu); - struct task_struct *sprev = srq->curr; - - WARN_ON_ONCE(__rq_lockp(rq) != __rq_lockp(srq)); - update_rq_clock(srq); - balance_one(srq, sprev); - } - } -#endif - rq_repin_lock(rq, rf); - - return ret; -} - static void process_ddsp_deferred_locals(struct rq *rq) { struct task_struct *p; @@ -2901,32 +2259,36 @@ static void process_ddsp_deferred_locals(struct rq *rq) */ while ((p = list_first_entry_or_null(&rq->scx.ddsp_deferred_locals, struct task_struct, scx.dsq_list.node))) { + struct scx_sched *sch = scx_root; struct scx_dispatch_q *dsq; list_del_init(&p->scx.dsq_list.node); - dsq = find_dsq_for_dispatch(rq, p->scx.ddsp_dsq_id, p); + dsq = find_dsq_for_dispatch(sch, rq, p->scx.ddsp_dsq_id, p); if (!WARN_ON_ONCE(dsq->id != SCX_DSQ_LOCAL)) - dispatch_to_local_dsq(rq, dsq, p, p->scx.ddsp_enq_flags); + dispatch_to_local_dsq(sch, rq, dsq, p, + p->scx.ddsp_enq_flags); } } static void set_next_task_scx(struct rq *rq, struct task_struct *p, bool first) { + struct scx_sched *sch = scx_root; + if (p->scx.flags & SCX_TASK_QUEUED) { /* * Core-sched might decide to execute @p before it is * dispatched. Call ops_dequeue() to notify the BPF scheduler. */ - ops_dequeue(p, SCX_DEQ_CORE_SCHED_EXEC); + ops_dequeue(rq, p, SCX_DEQ_CORE_SCHED_EXEC); dispatch_dequeue(rq, p); } p->se.exec_start = rq_clock_task(rq); /* see dequeue_task_scx() on why we skip when !QUEUED */ - if (SCX_HAS_OP(running) && (p->scx.flags & SCX_TASK_QUEUED)) - SCX_CALL_OP_TASK(SCX_KF_REST, running, p); + if (SCX_HAS_OP(sch, running) && (p->scx.flags & SCX_TASK_QUEUED)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, running, rq, p); clr_task_runnable(p, true); @@ -2956,10 +2318,8 @@ static void set_next_task_scx(struct rq *rq, struct task_struct *p, bool first) static enum scx_cpu_preempt_reason preempt_reason_from_class(const struct sched_class *class) { -#ifdef CONFIG_SMP if (class == &stop_sched_class) return SCX_CPU_PREEMPT_STOP; -#endif if (class == &dl_sched_class) return SCX_CPU_PREEMPT_DL; if (class == &rt_sched_class) @@ -2969,17 +2329,10 @@ preempt_reason_from_class(const struct sched_class *class) static void switch_class(struct rq *rq, struct task_struct *next) { + struct scx_sched *sch = scx_root; const struct sched_class *next_class = next->sched_class; -#ifdef CONFIG_SMP - /* - * Pairs with the smp_load_acquire() issued by a CPU in - * kick_cpus_irq_workfn() who is waiting for this CPU to perform a - * resched. - */ - smp_store_release(&rq->scx.pnt_seq, rq->scx.pnt_seq + 1); -#endif - if (!static_branch_unlikely(&scx_ops_cpu_preempt)) + if (!(sch->ops.flags & SCX_OPS_HAS_CPU_PREEMPT)) return; /* @@ -3001,14 +2354,14 @@ static void switch_class(struct rq *rq, struct task_struct *next) * next time that balance_scx() is invoked. */ if (!rq->scx.cpu_released) { - if (SCX_HAS_OP(cpu_release)) { + if (SCX_HAS_OP(sch, cpu_release)) { struct scx_cpu_release_args args = { .reason = preempt_reason_from_class(next_class), .task = next, }; - SCX_CALL_OP(SCX_KF_CPU_RELEASE, - cpu_release, cpu_of(rq), &args); + SCX_CALL_OP(sch, SCX_KF_CPU_RELEASE, cpu_release, rq, + cpu_of(rq), &args); } rq->scx.cpu_released = true; } @@ -3017,11 +2370,16 @@ static void switch_class(struct rq *rq, struct task_struct *next) static void put_prev_task_scx(struct rq *rq, struct task_struct *p, struct task_struct *next) { + struct scx_sched *sch = scx_root; + + /* see kick_cpus_irq_workfn() */ + smp_store_release(&rq->scx.kick_sync, rq->scx.kick_sync + 1); + update_curr_scx(rq); /* see dequeue_task_scx() on why we skip when !QUEUED */ - if (SCX_HAS_OP(stopping) && (p->scx.flags & SCX_TASK_QUEUED)) - SCX_CALL_OP_TASK(SCX_KF_REST, stopping, p, true); + if (SCX_HAS_OP(sch, stopping) && (p->scx.flags & SCX_TASK_QUEUED)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, stopping, rq, p, true); if (p->scx.flags & SCX_TASK_QUEUED) { set_task_runnable(rq, p); @@ -3033,8 +2391,9 @@ static void put_prev_task_scx(struct rq *rq, struct task_struct *p, * DSQ. */ if (p->scx.slice && !scx_rq_bypassing(rq)) { - dispatch_enqueue(&rq->scx.local_dsq, p, SCX_ENQ_HEAD); - return; + dispatch_enqueue(sch, &rq->scx.local_dsq, p, + SCX_ENQ_HEAD); + goto switch_class; } /* @@ -3044,13 +2403,14 @@ static void put_prev_task_scx(struct rq *rq, struct task_struct *p, * which should trigger an explicit follow-up scheduling event. */ if (sched_class_above(&ext_sched_class, next->sched_class)) { - WARN_ON_ONCE(!static_branch_unlikely(&scx_ops_enq_last)); + WARN_ON_ONCE(!(sch->ops.flags & SCX_OPS_ENQ_LAST)); do_enqueue_task(rq, p, SCX_ENQ_LAST, -1); } else { do_enqueue_task(rq, p, 0, -1); } } +switch_class: if (next && next->sched_class != &ext_sched_class) switch_class(rq, next); } @@ -3061,39 +2421,38 @@ static struct task_struct *first_local_task(struct rq *rq) struct task_struct, scx.dsq_list.node); } -static struct task_struct *pick_task_scx(struct rq *rq) +static struct task_struct * +do_pick_task_scx(struct rq *rq, struct rq_flags *rf, bool force_scx) { struct task_struct *prev = rq->curr; + bool keep_prev, kick_idle = false; struct task_struct *p; - bool prev_on_scx = prev->sched_class == &ext_sched_class; - bool keep_prev = rq->scx.flags & SCX_RQ_BAL_KEEP; - bool kick_idle = false; + + /* see kick_cpus_irq_workfn() */ + smp_store_release(&rq->scx.kick_sync, rq->scx.kick_sync + 1); + + rq_modified_clear(rq); + + rq_unpin_lock(rq, rf); + balance_one(rq, prev); + rq_repin_lock(rq, rf); + maybe_queue_balance_callback(rq); /* - * WORKAROUND: + * If any higher-priority sched class enqueued a runnable task on + * this rq during balance_one(), abort and return RETRY_TASK, so + * that the scheduler loop can restart. * - * %SCX_RQ_BAL_KEEP should be set iff $prev is on SCX as it must just - * have gone through balance_scx(). Unfortunately, there currently is a - * bug where fair could say yes on balance() but no on pick_task(), - * which then ends up calling pick_task_scx() without preceding - * balance_scx(). - * - * Keep running @prev if possible and avoid stalling from entering idle - * without balancing. - * - * Once fair is fixed, remove the workaround and trigger WARN_ON_ONCE() - * if pick_task_scx() is called without preceding balance_scx(). + * If @force_scx is true, always try to pick a SCHED_EXT task, + * regardless of any higher-priority sched classes activity. */ - if (unlikely(rq->scx.flags & SCX_RQ_BAL_PENDING)) { - if (prev_on_scx) { - keep_prev = true; - } else { - keep_prev = false; - kick_idle = true; - } - } else if (unlikely(keep_prev && !prev_on_scx)) { - /* only allowed during transitions */ - WARN_ON_ONCE(scx_ops_enable_state() == SCX_OPS_ENABLED); + if (!force_scx && rq_modified_above(rq, &ext_sched_class)) + return RETRY_TASK; + + keep_prev = rq->scx.flags & SCX_RQ_BAL_KEEP; + if (unlikely(keep_prev && + prev->sched_class != &ext_sched_class)) { + WARN_ON_ONCE(scx_enable_state() == SCX_ENABLED); keep_prev = false; } @@ -3105,33 +2464,42 @@ static struct task_struct *pick_task_scx(struct rq *rq) if (keep_prev) { p = prev; if (!p->scx.slice) - p->scx.slice = SCX_SLICE_DFL; + 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)) { - if (!scx_rq_bypassing(rq) && !scx_warned_zero_slice) { + 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__); - scx_warned_zero_slice = true; + sch->warned_zero_slice = true; } - p->scx.slice = SCX_SLICE_DFL; + refill_task_slice_dfl(sch, p); } } return p; } +static struct task_struct *pick_task_scx(struct rq *rq, struct rq_flags *rf) +{ + return do_pick_task_scx(rq, rf, false); +} + #ifdef CONFIG_SCHED_CORE /** * scx_prio_less - Task ordering for core-sched * @a: task A * @b: task B + * @in_fi: in forced idle state * * Core-sched is implemented as an additional scheduling layer on top of the * usual sched_class'es and needs to find out the expected task ordering. For @@ -3139,7 +2507,7 @@ static struct task_struct *pick_task_scx(struct rq *rq) * * Unless overridden by ops.core_sched_before(), @p->scx.core_sched_at is used * to implement the default task ordering. The older the timestamp, the higher - * prority the task - the global FIFO ordering matching the default scheduling + * priority the task - the global FIFO ordering matching the default scheduling * behavior. * * When ops.core_sched_before() is enabled, @p->scx.core_sched_at is used to @@ -3148,13 +2516,17 @@ static struct task_struct *pick_task_scx(struct rq *rq) bool scx_prio_less(const struct task_struct *a, const struct task_struct *b, bool in_fi) { + struct scx_sched *sch = scx_root; + /* * The const qualifiers are dropped from task_struct pointers when * calling ops.core_sched_before(). Accesses are controlled by the * verifier. */ - if (SCX_HAS_OP(core_sched_before) && !scx_rq_bypassing(task_rq(a))) - return SCX_CALL_OP_2TASKS_RET(SCX_KF_REST, core_sched_before, + if (SCX_HAS_OP(sch, core_sched_before) && + !scx_rq_bypassing(task_rq(a))) + return SCX_CALL_OP_2TASKS_RET(sch, SCX_KF_REST, core_sched_before, + NULL, (struct task_struct *)a, (struct task_struct *)b); else @@ -3162,356 +2534,11 @@ bool scx_prio_less(const struct task_struct *a, const struct task_struct *b, } #endif /* CONFIG_SCHED_CORE */ -#ifdef CONFIG_SMP - -static bool test_and_clear_cpu_idle(int cpu) -{ -#ifdef CONFIG_SCHED_SMT - /* - * SMT mask should be cleared whether we can claim @cpu or not. The SMT - * cluster is not wholly idle either way. This also prevents - * scx_pick_idle_cpu() from getting caught in an infinite loop. - */ - if (sched_smt_active()) { - const struct cpumask *smt = cpu_smt_mask(cpu); - - /* - * If offline, @cpu is not its own sibling and - * scx_pick_idle_cpu() can get caught in an infinite loop as - * @cpu is never cleared from idle_masks.smt. Ensure that @cpu - * is eventually cleared. - */ - if (cpumask_intersects(smt, idle_masks.smt)) - cpumask_andnot(idle_masks.smt, idle_masks.smt, smt); - else if (cpumask_test_cpu(cpu, idle_masks.smt)) - __cpumask_clear_cpu(cpu, idle_masks.smt); - } -#endif - return cpumask_test_and_clear_cpu(cpu, idle_masks.cpu); -} - -static s32 scx_pick_idle_cpu(const struct cpumask *cpus_allowed, u64 flags) -{ - int cpu; - -retry: - if (sched_smt_active()) { - cpu = cpumask_any_and_distribute(idle_masks.smt, cpus_allowed); - if (cpu < nr_cpu_ids) - goto found; - - if (flags & SCX_PICK_IDLE_CORE) - return -EBUSY; - } - - cpu = cpumask_any_and_distribute(idle_masks.cpu, cpus_allowed); - if (cpu >= nr_cpu_ids) - return -EBUSY; - -found: - if (test_and_clear_cpu_idle(cpu)) - return cpu; - else - goto retry; -} - -/* - * Return true if the LLC domains do not perfectly overlap with the NUMA - * domains, false otherwise. - */ -static bool llc_numa_mismatch(void) -{ - int cpu; - - /* - * We need to scan all online CPUs to verify whether their scheduling - * domains overlap. - * - * While it is rare to encounter architectures with asymmetric NUMA - * topologies, CPU hotplugging or virtualized environments can result - * in asymmetric configurations. - * - * For example: - * - * NUMA 0: - * - LLC 0: cpu0..cpu7 - * - LLC 1: cpu8..cpu15 [offline] - * - * NUMA 1: - * - LLC 0: cpu16..cpu23 - * - LLC 1: cpu24..cpu31 - * - * In this case, if we only check the first online CPU (cpu0), we might - * incorrectly assume that the LLC and NUMA domains are fully - * overlapping, which is incorrect (as NUMA 1 has two distinct LLC - * domains). - */ - for_each_online_cpu(cpu) { - const struct cpumask *numa_cpus; - struct sched_domain *sd; - - sd = rcu_dereference(per_cpu(sd_llc, cpu)); - if (!sd) - return true; - - numa_cpus = cpumask_of_node(cpu_to_node(cpu)); - if (sd->span_weight != cpumask_weight(numa_cpus)) - return true; - } - - return false; -} - -/* - * Initialize topology-aware scheduling. - * - * Detect if the system has multiple LLC or multiple NUMA domains and enable - * cache-aware / NUMA-aware scheduling optimizations in the default CPU idle - * selection policy. - * - * Assumption: the kernel's internal topology representation assumes that each - * CPU belongs to a single LLC domain, and that each LLC domain is entirely - * contained within a single NUMA node. - */ -static void update_selcpu_topology(void) -{ - bool enable_llc = false, enable_numa = false; - struct sched_domain *sd; - const struct cpumask *cpus; - s32 cpu = cpumask_first(cpu_online_mask); - - /* - * Enable LLC domain optimization only when there are multiple LLC - * domains among the online CPUs. If all online CPUs are part of a - * single LLC domain, the idle CPU selection logic can choose any - * online CPU without bias. - * - * Note that it is sufficient to check the LLC domain of the first - * online CPU to determine whether a single LLC domain includes all - * CPUs. - */ - rcu_read_lock(); - sd = rcu_dereference(per_cpu(sd_llc, cpu)); - if (sd) { - if (sd->span_weight < num_online_cpus()) - enable_llc = true; - } - - /* - * Enable NUMA optimization only when there are multiple NUMA domains - * among the online CPUs and the NUMA domains don't perfectly overlaps - * with the LLC domains. - * - * If all CPUs belong to the same NUMA node and the same LLC domain, - * enabling both NUMA and LLC optimizations is unnecessary, as checking - * for an idle CPU in the same domain twice is redundant. - */ - cpus = cpumask_of_node(cpu_to_node(cpu)); - if ((cpumask_weight(cpus) < num_online_cpus()) && llc_numa_mismatch()) - enable_numa = true; - rcu_read_unlock(); - - pr_debug("sched_ext: LLC idle selection %s\n", - enable_llc ? "enabled" : "disabled"); - pr_debug("sched_ext: NUMA idle selection %s\n", - enable_numa ? "enabled" : "disabled"); - - if (enable_llc) - static_branch_enable_cpuslocked(&scx_selcpu_topo_llc); - else - static_branch_disable_cpuslocked(&scx_selcpu_topo_llc); - if (enable_numa) - static_branch_enable_cpuslocked(&scx_selcpu_topo_numa); - else - static_branch_disable_cpuslocked(&scx_selcpu_topo_numa); -} - -/* - * Built-in CPU idle selection policy: - * - * 1. Prioritize full-idle cores: - * - always prioritize CPUs from fully idle cores (both logical CPUs are - * idle) to avoid interference caused by SMT. - * - * 2. Reuse the same CPU: - * - prefer the last used CPU to take advantage of cached data (L1, L2) and - * branch prediction optimizations. - * - * 3. Pick a CPU within the same LLC (Last-Level Cache): - * - if the above conditions aren't met, pick a CPU that shares the same LLC - * to maintain cache locality. - * - * 4. Pick a CPU within the same NUMA node, if enabled: - * - choose a CPU from the same NUMA node to reduce memory access latency. - * - * Step 3 and 4 are performed only if the system has, respectively, multiple - * LLC domains / multiple NUMA nodes (see scx_selcpu_topo_llc and - * scx_selcpu_topo_numa). - * - * NOTE: tasks that can only run on 1 CPU are excluded by this logic, because - * we never call ops.select_cpu() for them, see select_task_rq(). - */ -static s32 scx_select_cpu_dfl(struct task_struct *p, s32 prev_cpu, - u64 wake_flags, bool *found) -{ - const struct cpumask *llc_cpus = NULL; - const struct cpumask *numa_cpus = NULL; - s32 cpu; - - *found = false; - - - /* - * This is necessary to protect llc_cpus. - */ - rcu_read_lock(); - - /* - * Determine the scheduling domain only if the task is allowed to run - * on all CPUs. - * - * This is done primarily for efficiency, as it avoids the overhead of - * updating a cpumask every time we need to select an idle CPU (which - * can be costly in large SMP systems), but it also aligns logically: - * if a task's scheduling domain is restricted by user-space (through - * CPU affinity), the task will simply use the flat scheduling domain - * defined by user-space. - */ - if (p->nr_cpus_allowed >= num_possible_cpus()) { - if (static_branch_maybe(CONFIG_NUMA, &scx_selcpu_topo_numa)) - numa_cpus = cpumask_of_node(cpu_to_node(prev_cpu)); - - if (static_branch_maybe(CONFIG_SCHED_MC, &scx_selcpu_topo_llc)) { - struct sched_domain *sd; - - sd = rcu_dereference(per_cpu(sd_llc, prev_cpu)); - if (sd) - llc_cpus = sched_domain_span(sd); - } - } - - /* - * If WAKE_SYNC, try to migrate the wakee to the waker's CPU. - */ - if (wake_flags & SCX_WAKE_SYNC) { - cpu = smp_processor_id(); - - /* - * If the waker's CPU is cache affine and prev_cpu is idle, - * then avoid a migration. - */ - if (cpus_share_cache(cpu, prev_cpu) && - test_and_clear_cpu_idle(prev_cpu)) { - cpu = prev_cpu; - goto cpu_found; - } - - /* - * If the waker's local DSQ is empty, and the system is under - * utilized, try to wake up @p to the local DSQ of the waker. - * - * Checking only for an empty local DSQ is insufficient as it - * could give the wakee an unfair advantage when the system is - * oversaturated. - * - * Checking only for the presence of idle CPUs is also - * insufficient as the local DSQ of the waker could have tasks - * piled up on it even if there is an idle core elsewhere on - * the system. - */ - if (!cpumask_empty(idle_masks.cpu) && - !(current->flags & PF_EXITING) && - cpu_rq(cpu)->scx.local_dsq.nr == 0) { - if (cpumask_test_cpu(cpu, p->cpus_ptr)) - goto cpu_found; - } - } - - /* - * If CPU has SMT, any wholly idle CPU is likely a better pick than - * partially idle @prev_cpu. - */ - if (sched_smt_active()) { - /* - * Keep using @prev_cpu if it's part of a fully idle core. - */ - if (cpumask_test_cpu(prev_cpu, idle_masks.smt) && - test_and_clear_cpu_idle(prev_cpu)) { - cpu = prev_cpu; - goto cpu_found; - } - - /* - * Search for any fully idle core in the same LLC domain. - */ - if (llc_cpus) { - cpu = scx_pick_idle_cpu(llc_cpus, SCX_PICK_IDLE_CORE); - if (cpu >= 0) - goto cpu_found; - } - - /* - * Search for any fully idle core in the same NUMA node. - */ - if (numa_cpus) { - cpu = scx_pick_idle_cpu(numa_cpus, SCX_PICK_IDLE_CORE); - if (cpu >= 0) - goto cpu_found; - } - - /* - * Search for any full idle core usable by the task. - */ - cpu = scx_pick_idle_cpu(p->cpus_ptr, SCX_PICK_IDLE_CORE); - if (cpu >= 0) - goto cpu_found; - } - - /* - * Use @prev_cpu if it's idle. - */ - if (test_and_clear_cpu_idle(prev_cpu)) { - cpu = prev_cpu; - goto cpu_found; - } - - /* - * Search for any idle CPU in the same LLC domain. - */ - if (llc_cpus) { - cpu = scx_pick_idle_cpu(llc_cpus, 0); - if (cpu >= 0) - goto cpu_found; - } - - /* - * Search for any idle CPU in the same NUMA node. - */ - if (numa_cpus) { - cpu = scx_pick_idle_cpu(numa_cpus, 0); - if (cpu >= 0) - goto cpu_found; - } - - /* - * Search for any idle CPU usable by the task. - */ - cpu = scx_pick_idle_cpu(p->cpus_ptr, 0); - if (cpu >= 0) - goto cpu_found; - - rcu_read_unlock(); - return prev_cpu; - -cpu_found: - rcu_read_unlock(); - - *found = true; - return cpu; -} - static int select_task_rq_scx(struct task_struct *p, int prev_cpu, int wake_flags) { + struct scx_sched *sch = scx_root; + bool rq_bypass; + /* * sched_exec() calls with %WF_EXEC when @p is about to exec(2) as it * can be a good migration opportunity with low cache and memory @@ -3525,7 +2552,8 @@ static int select_task_rq_scx(struct task_struct *p, int prev_cpu, int wake_flag if (unlikely(wake_flags & WF_EXEC)) return prev_cpu; - if (SCX_HAS_OP(select_cpu) && !scx_rq_bypassing(task_rq(p))) { + rq_bypass = scx_rq_bypassing(task_rq(p)); + if (likely(SCX_HAS_OP(sch, select_cpu)) && !rq_bypass) { s32 cpu; struct task_struct **ddsp_taskp; @@ -3533,22 +2561,30 @@ static int select_task_rq_scx(struct task_struct *p, int prev_cpu, int wake_flag WARN_ON_ONCE(*ddsp_taskp); *ddsp_taskp = p; - cpu = SCX_CALL_OP_TASK_RET(SCX_KF_ENQUEUE | SCX_KF_SELECT_CPU, - select_cpu, p, prev_cpu, wake_flags); + cpu = SCX_CALL_OP_TASK_RET(sch, + SCX_KF_ENQUEUE | SCX_KF_SELECT_CPU, + select_cpu, NULL, p, prev_cpu, + wake_flags); + p->scx.selected_cpu = cpu; *ddsp_taskp = NULL; - if (ops_cpu_valid(cpu, "from ops.select_cpu()")) + if (ops_cpu_valid(sch, cpu, "from ops.select_cpu()")) return cpu; else return prev_cpu; } else { - bool found; s32 cpu; - cpu = scx_select_cpu_dfl(p, prev_cpu, wake_flags, &found); - if (found) { - p->scx.slice = SCX_SLICE_DFL; + cpu = scx_select_cpu_dfl(p, prev_cpu, wake_flags, NULL, 0); + if (cpu >= 0) { + refill_task_slice_dfl(sch, p); p->scx.ddsp_dsq_id = SCX_DSQ_LOCAL; + } else { + cpu = prev_cpu; } + p->scx.selected_cpu = cpu; + + if (rq_bypass) + __scx_add_event(sch, SCX_EV_BYPASS_DISPATCH, 1); return cpu; } } @@ -3561,6 +2597,8 @@ static void task_woken_scx(struct rq *rq, struct task_struct *p) static void set_cpus_allowed_scx(struct task_struct *p, struct affinity_context *ac) { + struct scx_sched *sch = scx_root; + set_cpus_allowed_common(p, ac); /* @@ -3571,74 +2609,38 @@ static void set_cpus_allowed_scx(struct task_struct *p, * Fine-grained memory write control is enforced by BPF making the const * designation pointless. Cast it away when calling the operation. */ - if (SCX_HAS_OP(set_cpumask)) - SCX_CALL_OP_TASK(SCX_KF_REST, set_cpumask, p, - (struct cpumask *)p->cpus_ptr); -} - -static void reset_idle_masks(void) -{ - /* - * Consider all online cpus idle. Should converge to the actual state - * quickly. - */ - cpumask_copy(idle_masks.cpu, cpu_online_mask); - cpumask_copy(idle_masks.smt, cpu_online_mask); -} - -void __scx_update_idle(struct rq *rq, bool idle) -{ - int cpu = cpu_of(rq); - - if (SCX_HAS_OP(update_idle) && !scx_rq_bypassing(rq)) { - SCX_CALL_OP(SCX_KF_REST, update_idle, cpu_of(rq), idle); - if (!static_branch_unlikely(&scx_builtin_idle_enabled)) - return; - } - - if (idle) - cpumask_set_cpu(cpu, idle_masks.cpu); - else - cpumask_clear_cpu(cpu, idle_masks.cpu); - -#ifdef CONFIG_SCHED_SMT - if (sched_smt_active()) { - const struct cpumask *smt = cpu_smt_mask(cpu); - - if (idle) { - /* - * idle_masks.smt handling is racy but that's fine as - * it's only for optimization and self-correcting. - */ - for_each_cpu(cpu, smt) { - if (!cpumask_test_cpu(cpu, idle_masks.cpu)) - return; - } - cpumask_or(idle_masks.smt, idle_masks.smt, smt); - } else { - cpumask_andnot(idle_masks.smt, idle_masks.smt, smt); - } - } -#endif + if (SCX_HAS_OP(sch, set_cpumask)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, set_cpumask, NULL, + p, (struct cpumask *)p->cpus_ptr); } static void handle_hotplug(struct rq *rq, bool online) { + struct scx_sched *sch = scx_root; int cpu = cpu_of(rq); atomic_long_inc(&scx_hotplug_seq); + /* + * scx_root updates are protected by cpus_read_lock() and will stay + * stable here. Note that we can't depend on scx_enabled() test as the + * hotplug ops need to be enabled before __scx_enabled is set. + */ + if (unlikely(!sch)) + return; + if (scx_enabled()) - update_selcpu_topology(); + scx_idle_update_selcpu_topology(&sch->ops); - if (online && SCX_HAS_OP(cpu_online)) - SCX_CALL_OP(SCX_KF_UNLOCKED, cpu_online, cpu); - else if (!online && SCX_HAS_OP(cpu_offline)) - SCX_CALL_OP(SCX_KF_UNLOCKED, cpu_offline, cpu); + if (online && SCX_HAS_OP(sch, cpu_online)) + SCX_CALL_OP(sch, SCX_KF_UNLOCKED, cpu_online, NULL, cpu); + else if (!online && SCX_HAS_OP(sch, cpu_offline)) + SCX_CALL_OP(sch, SCX_KF_UNLOCKED, cpu_offline, NULL, cpu); else - scx_ops_exit(SCX_ECODE_ACT_RESTART | SCX_ECODE_RSN_HOTPLUG, - "cpu %d going %s, exiting scheduler", cpu, - online ? "online" : "offline"); + scx_exit(sch, SCX_EXIT_UNREG_KERN, + SCX_ECODE_ACT_RESTART | SCX_ECODE_RSN_HOTPLUG, + "cpu %d going %s, exiting scheduler", cpu, + online ? "online" : "offline"); } void scx_rq_activate(struct rq *rq) @@ -3661,21 +2663,19 @@ static void rq_offline_scx(struct rq *rq) rq->scx.flags &= ~SCX_RQ_ONLINE; } -#else /* CONFIG_SMP */ - -static bool test_and_clear_cpu_idle(int cpu) { return false; } -static s32 scx_pick_idle_cpu(const struct cpumask *cpus_allowed, u64 flags) { return -EBUSY; } -static void reset_idle_masks(void) {} - -#endif /* CONFIG_SMP */ static bool check_rq_for_timeouts(struct rq *rq) { + struct scx_sched *sch; struct task_struct *p; struct rq_flags rf; bool timed_out = false; rq_lock_irqsave(rq, &rf); + sch = rcu_dereference_bh(scx_root); + if (unlikely(!sch)) + goto out_unlock; + list_for_each_entry(p, &rq->scx.runnable_list, scx.runnable_node) { unsigned long last_runnable = p->scx.runnable_at; @@ -3683,16 +2683,15 @@ static bool check_rq_for_timeouts(struct rq *rq) last_runnable + scx_watchdog_timeout))) { u32 dur_ms = jiffies_to_msecs(jiffies - last_runnable); - scx_ops_error_kind(SCX_EXIT_ERROR_STALL, - "%s[%d] failed to run for %u.%03us", - p->comm, p->pid, - dur_ms / 1000, dur_ms % 1000); + scx_exit(sch, SCX_EXIT_ERROR_STALL, 0, + "%s[%d] failed to run for %u.%03us", + p->comm, p->pid, dur_ms / 1000, dur_ms % 1000); timed_out = true; break; } } +out_unlock: rq_unlock_irqrestore(rq, &rf); - return timed_out; } @@ -3714,19 +2713,24 @@ static void scx_watchdog_workfn(struct work_struct *work) void scx_tick(struct rq *rq) { + struct scx_sched *sch; unsigned long last_check; if (!scx_enabled()) return; + sch = rcu_dereference_bh(scx_root); + if (unlikely(!sch)) + return; + last_check = READ_ONCE(scx_watchdog_timestamp); if (unlikely(time_after(jiffies, last_check + READ_ONCE(scx_watchdog_timeout)))) { u32 dur_ms = jiffies_to_msecs(jiffies - last_check); - scx_ops_error_kind(SCX_EXIT_ERROR_STALL, - "watchdog failed to check in for %u.%03us", - dur_ms / 1000, dur_ms % 1000); + scx_exit(sch, SCX_EXIT_ERROR_STALL, 0, + "watchdog failed to check in for %u.%03us", + dur_ms / 1000, dur_ms % 1000); } update_other_load_avgs(rq); @@ -3734,6 +2738,8 @@ void scx_tick(struct rq *rq) static void task_tick_scx(struct rq *rq, struct task_struct *curr, int queued) { + struct scx_sched *sch = scx_root; + update_curr_scx(rq); /* @@ -3743,8 +2749,8 @@ static void task_tick_scx(struct rq *rq, struct task_struct *curr, int queued) if (scx_rq_bypassing(rq)) { curr->scx.slice = 0; touch_core_sched(rq, curr); - } else if (SCX_HAS_OP(tick)) { - SCX_CALL_OP(SCX_KF_REST, tick, curr); + } else if (SCX_HAS_OP(sch, tick)) { + SCX_CALL_OP_TASK(sch, SCX_KF_REST, tick, rq, curr); } if (!curr->scx.slice) @@ -3809,21 +2815,23 @@ static void scx_set_task_state(struct task_struct *p, enum scx_task_state state) p->scx.flags |= state << SCX_TASK_STATE_SHIFT; } -static int scx_ops_init_task(struct task_struct *p, struct task_group *tg, bool fork) +static int scx_init_task(struct task_struct *p, struct task_group *tg, bool fork) { + struct scx_sched *sch = scx_root; int ret; p->scx.disallow = false; - if (SCX_HAS_OP(init_task)) { + if (SCX_HAS_OP(sch, init_task)) { struct scx_init_task_args args = { SCX_INIT_TASK_ARGS_CGROUP(tg) .fork = fork, }; - ret = SCX_CALL_OP_RET(SCX_KF_UNLOCKED, init_task, p, &args); + ret = SCX_CALL_OP_RET(sch, SCX_KF_UNLOCKED, init_task, NULL, + p, &args); if (unlikely(ret)) { - ret = ops_sanitize_err("init_task", ret); + ret = ops_sanitize_err(sch, "init_task", ret); return ret; } } @@ -3851,8 +2859,8 @@ static int scx_ops_init_task(struct task_struct *p, struct task_group *tg, bool task_rq_unlock(rq, p, &rf); } else if (p->policy == SCHED_EXT) { - scx_ops_error("ops.init_task() set task->scx.disallow for %s[%d] during fork", - p->comm, p->pid); + scx_error(sch, "ops.init_task() set task->scx.disallow for %s[%d] during fork", + p->comm, p->pid); } } @@ -3860,11 +2868,13 @@ static int scx_ops_init_task(struct task_struct *p, struct task_group *tg, bool return 0; } -static void scx_ops_enable_task(struct task_struct *p) +static void scx_enable_task(struct task_struct *p) { + struct scx_sched *sch = scx_root; + struct rq *rq = task_rq(p); u32 weight; - lockdep_assert_rq_held(task_rq(p)); + lockdep_assert_rq_held(rq); /* * Set the weight before calling ops.enable() so that the scheduler @@ -3877,26 +2887,31 @@ static void scx_ops_enable_task(struct task_struct *p) p->scx.weight = sched_weight_to_cgroup(weight); - if (SCX_HAS_OP(enable)) - SCX_CALL_OP_TASK(SCX_KF_REST, enable, p); + if (SCX_HAS_OP(sch, enable)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, enable, rq, p); scx_set_task_state(p, SCX_TASK_ENABLED); - if (SCX_HAS_OP(set_weight)) - SCX_CALL_OP_TASK(SCX_KF_REST, set_weight, p, p->scx.weight); + if (SCX_HAS_OP(sch, set_weight)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, set_weight, rq, + p, p->scx.weight); } -static void scx_ops_disable_task(struct task_struct *p) +static void scx_disable_task(struct task_struct *p) { - lockdep_assert_rq_held(task_rq(p)); + struct scx_sched *sch = scx_root; + struct rq *rq = task_rq(p); + + lockdep_assert_rq_held(rq); WARN_ON_ONCE(scx_get_task_state(p) != SCX_TASK_ENABLED); - if (SCX_HAS_OP(disable)) - SCX_CALL_OP(SCX_KF_REST, disable, p); + if (SCX_HAS_OP(sch, disable)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, disable, rq, p); scx_set_task_state(p, SCX_TASK_READY); } -static void scx_ops_exit_task(struct task_struct *p) +static void scx_exit_task(struct task_struct *p) { + struct scx_sched *sch = scx_root; struct scx_exit_task_args args = { .cancelled = false, }; @@ -3912,15 +2927,16 @@ static void scx_ops_exit_task(struct task_struct *p) case SCX_TASK_READY: break; case SCX_TASK_ENABLED: - scx_ops_disable_task(p); + scx_disable_task(p); break; default: WARN_ON_ONCE(true); return; } - if (SCX_HAS_OP(exit_task)) - SCX_CALL_OP(SCX_KF_REST, exit_task, p, &args); + if (SCX_HAS_OP(sch, exit_task)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, exit_task, task_rq(p), + p, &args); scx_set_task_state(p, SCX_TASK_NONE); } @@ -3934,7 +2950,7 @@ void init_scx_entity(struct sched_ext_entity *scx) INIT_LIST_HEAD(&scx->runnable_node); scx->runnable_at = jiffies; scx->ddsp_dsq_id = SCX_DSQ_INVALID; - scx->slice = SCX_SLICE_DFL; + scx->slice = READ_ONCE(scx_slice_dfl); } void scx_pre_fork(struct task_struct *p) @@ -3952,15 +2968,15 @@ int scx_fork(struct task_struct *p) { percpu_rwsem_assert_held(&scx_fork_rwsem); - if (scx_ops_init_task_enabled) - return scx_ops_init_task(p, task_group(p), true); + if (scx_init_task_enabled) + return scx_init_task(p, task_group(p), true); else return 0; } void scx_post_fork(struct task_struct *p) { - if (scx_ops_init_task_enabled) { + if (scx_init_task_enabled) { scx_set_task_state(p, SCX_TASK_READY); /* @@ -3973,14 +2989,14 @@ void scx_post_fork(struct task_struct *p) struct rq *rq; rq = task_rq_lock(p, &rf); - scx_ops_enable_task(p); + scx_enable_task(p); task_rq_unlock(rq, p, &rf); } } - spin_lock_irq(&scx_tasks_lock); + raw_spin_lock_irq(&scx_tasks_lock); list_add_tail(&p->scx.tasks_node, &scx_tasks); - spin_unlock_irq(&scx_tasks_lock); + raw_spin_unlock_irq(&scx_tasks_lock); percpu_up_read(&scx_fork_rwsem); } @@ -3993,31 +3009,31 @@ void scx_cancel_fork(struct task_struct *p) rq = task_rq_lock(p, &rf); WARN_ON_ONCE(scx_get_task_state(p) >= SCX_TASK_READY); - scx_ops_exit_task(p); + scx_exit_task(p); task_rq_unlock(rq, p, &rf); } percpu_up_read(&scx_fork_rwsem); } -void sched_ext_free(struct task_struct *p) +void sched_ext_dead(struct task_struct *p) { unsigned long flags; - spin_lock_irqsave(&scx_tasks_lock, flags); + raw_spin_lock_irqsave(&scx_tasks_lock, flags); list_del_init(&p->scx.tasks_node); - spin_unlock_irqrestore(&scx_tasks_lock, flags); + raw_spin_unlock_irqrestore(&scx_tasks_lock, flags); /* - * @p is off scx_tasks and wholly ours. scx_ops_enable()'s READY -> - * ENABLED transitions can't race us. Disable ops for @p. + * @p is off scx_tasks and wholly ours. scx_enable()'s READY -> ENABLED + * transitions can't race us. Disable ops for @p. */ if (scx_get_task_state(p) != SCX_TASK_NONE) { struct rq_flags rf; struct rq *rq; rq = task_rq_lock(p, &rf); - scx_ops_exit_task(p); + scx_exit_task(p); task_rq_unlock(rq, p, &rf); } } @@ -4025,33 +3041,38 @@ void sched_ext_free(struct task_struct *p) static void reweight_task_scx(struct rq *rq, struct task_struct *p, const struct load_weight *lw) { + struct scx_sched *sch = scx_root; + lockdep_assert_rq_held(task_rq(p)); p->scx.weight = sched_weight_to_cgroup(scale_load_down(lw->weight)); - if (SCX_HAS_OP(set_weight)) - SCX_CALL_OP_TASK(SCX_KF_REST, set_weight, p, p->scx.weight); + if (SCX_HAS_OP(sch, set_weight)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, set_weight, rq, + p, p->scx.weight); } -static void prio_changed_scx(struct rq *rq, struct task_struct *p, int oldprio) +static void prio_changed_scx(struct rq *rq, struct task_struct *p, u64 oldprio) { } static void switching_to_scx(struct rq *rq, struct task_struct *p) { - scx_ops_enable_task(p); + struct scx_sched *sch = scx_root; + + scx_enable_task(p); /* * set_cpus_allowed_scx() is not called while @p is associated with a * different scheduler class. Keep the BPF scheduler up-to-date. */ - if (SCX_HAS_OP(set_cpumask)) - SCX_CALL_OP_TASK(SCX_KF_REST, set_cpumask, p, - (struct cpumask *)p->cpus_ptr); + if (SCX_HAS_OP(sch, set_cpumask)) + SCX_CALL_OP_TASK(sch, SCX_KF_REST, set_cpumask, rq, + p, (struct cpumask *)p->cpus_ptr); } static void switched_from_scx(struct rq *rq, struct task_struct *p) { - scx_ops_disable_task(p); + scx_disable_task(p); } static void wakeup_preempt_scx(struct rq *rq, struct task_struct *p,int wake_flags) {} @@ -4091,90 +3112,66 @@ 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; -static bool cgroup_warned_missing_weight; -static bool cgroup_warned_missing_idle; -static void scx_cgroup_warn_missing_weight(struct task_group *tg) +void scx_tg_init(struct task_group *tg) { - if (scx_ops_enable_state() == SCX_OPS_DISABLED || - cgroup_warned_missing_weight) - return; - - if ((scx_ops.flags & SCX_OPS_HAS_CGROUP_WEIGHT) || !tg->css.parent) - return; - - pr_warn("sched_ext: \"%s\" does not implement cgroup cpu.weight\n", - scx_ops.name); - cgroup_warned_missing_weight = true; -} - -static void scx_cgroup_warn_missing_idle(struct task_group *tg) -{ - if (!scx_cgroup_enabled || cgroup_warned_missing_idle) - return; - - if (!tg->idle) - return; - - pr_warn("sched_ext: \"%s\" does not implement cgroup cpu.idle\n", - scx_ops.name); - cgroup_warned_missing_idle = true; + tg->scx.weight = CGROUP_WEIGHT_DFL; + tg->scx.bw_period_us = default_bw_period_us(); + tg->scx.bw_quota_us = RUNTIME_INF; + tg->scx.idle = false; } int scx_tg_online(struct task_group *tg) { + struct scx_sched *sch = scx_root; int ret = 0; - WARN_ON_ONCE(tg->scx_flags & (SCX_TG_ONLINE | SCX_TG_INITED)); - - percpu_down_read(&scx_cgroup_rwsem); - - scx_cgroup_warn_missing_weight(tg); + WARN_ON_ONCE(tg->scx.flags & (SCX_TG_ONLINE | SCX_TG_INITED)); if (scx_cgroup_enabled) { - if (SCX_HAS_OP(cgroup_init)) { + if (SCX_HAS_OP(sch, cgroup_init)) { struct scx_cgroup_init_args args = - { .weight = tg->scx_weight }; + { .weight = tg->scx.weight, + .bw_period_us = tg->scx.bw_period_us, + .bw_quota_us = tg->scx.bw_quota_us, + .bw_burst_us = tg->scx.bw_burst_us }; - ret = SCX_CALL_OP_RET(SCX_KF_UNLOCKED, cgroup_init, - tg->css.cgroup, &args); + ret = SCX_CALL_OP_RET(sch, SCX_KF_UNLOCKED, cgroup_init, + NULL, tg->css.cgroup, &args); if (ret) - ret = ops_sanitize_err("cgroup_init", ret); + ret = ops_sanitize_err(sch, "cgroup_init", ret); } if (ret == 0) - tg->scx_flags |= SCX_TG_ONLINE | SCX_TG_INITED; + tg->scx.flags |= SCX_TG_ONLINE | SCX_TG_INITED; } else { - tg->scx_flags |= SCX_TG_ONLINE; + tg->scx.flags |= SCX_TG_ONLINE; } - percpu_up_read(&scx_cgroup_rwsem); return ret; } void scx_tg_offline(struct task_group *tg) { - WARN_ON_ONCE(!(tg->scx_flags & SCX_TG_ONLINE)); - - percpu_down_read(&scx_cgroup_rwsem); + struct scx_sched *sch = scx_root; - if (SCX_HAS_OP(cgroup_exit) && (tg->scx_flags & SCX_TG_INITED)) - SCX_CALL_OP(SCX_KF_UNLOCKED, cgroup_exit, tg->css.cgroup); - tg->scx_flags &= ~(SCX_TG_ONLINE | SCX_TG_INITED); + WARN_ON_ONCE(!(tg->scx.flags & SCX_TG_ONLINE)); - percpu_up_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); } int scx_cgroup_can_attach(struct cgroup_taskset *tset) { + struct scx_sched *sch = scx_root; struct cgroup_subsys_state *css; struct task_struct *p; int ret; - /* released in scx_finish/cancel_attach() */ - percpu_down_read(&scx_cgroup_rwsem); - if (!scx_cgroup_enabled) return 0; @@ -4192,8 +3189,9 @@ int scx_cgroup_can_attach(struct cgroup_taskset *tset) if (from == to) continue; - if (SCX_HAS_OP(cgroup_prep_move)) { - ret = SCX_CALL_OP_RET(SCX_KF_UNLOCKED, cgroup_prep_move, + if (SCX_HAS_OP(sch, cgroup_prep_move)) { + ret = SCX_CALL_OP_RET(sch, SCX_KF_UNLOCKED, + cgroup_prep_move, NULL, p, from, css->cgroup); if (ret) goto err; @@ -4206,102 +3204,122 @@ int scx_cgroup_can_attach(struct cgroup_taskset *tset) err: cgroup_taskset_for_each(p, css, tset) { - if (SCX_HAS_OP(cgroup_cancel_move) && p->scx.cgrp_moving_from) - SCX_CALL_OP(SCX_KF_UNLOCKED, cgroup_cancel_move, p, - p->scx.cgrp_moving_from, css->cgroup); + if (SCX_HAS_OP(sch, cgroup_cancel_move) && + p->scx.cgrp_moving_from) + SCX_CALL_OP(sch, SCX_KF_UNLOCKED, cgroup_cancel_move, NULL, + p, p->scx.cgrp_moving_from, css->cgroup); p->scx.cgrp_moving_from = NULL; } - percpu_up_read(&scx_cgroup_rwsem); - return ops_sanitize_err("cgroup_prep_move", ret); + return ops_sanitize_err(sch, "cgroup_prep_move", ret); } -void scx_move_task(struct task_struct *p) +void scx_cgroup_move_task(struct task_struct *p) { - if (!scx_cgroup_enabled) - return; + struct scx_sched *sch = scx_root; - /* - * We're called from sched_move_task() which handles both cgroup and - * autogroup moves. Ignore the latter. - * - * Also ignore exiting tasks, because in the exit path tasks transition - * from the autogroup to the root group, so task_group_is_autogroup() - * alone isn't able to catch exiting autogroup tasks. This is safe for - * cgroup_move(), because cgroup migrations never happen for PF_EXITING - * tasks. - */ - if (task_group_is_autogroup(task_group(p)) || (p->flags & PF_EXITING)) + if (!scx_cgroup_enabled) return; /* * @p must have ops.cgroup_prep_move() called on it and thus * cgrp_moving_from set. */ - if (SCX_HAS_OP(cgroup_move) && !WARN_ON_ONCE(!p->scx.cgrp_moving_from)) - SCX_CALL_OP_TASK(SCX_KF_UNLOCKED, cgroup_move, p, - p->scx.cgrp_moving_from, tg_cgrp(task_group(p))); + if (SCX_HAS_OP(sch, cgroup_move) && + !WARN_ON_ONCE(!p->scx.cgrp_moving_from)) + SCX_CALL_OP_TASK(sch, SCX_KF_UNLOCKED, cgroup_move, NULL, + p, p->scx.cgrp_moving_from, + tg_cgrp(task_group(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; struct cgroup_subsys_state *css; struct task_struct *p; if (!scx_cgroup_enabled) - goto out_unlock; + return; cgroup_taskset_for_each(p, css, tset) { - if (SCX_HAS_OP(cgroup_cancel_move) && p->scx.cgrp_moving_from) - SCX_CALL_OP(SCX_KF_UNLOCKED, cgroup_cancel_move, p, - p->scx.cgrp_moving_from, css->cgroup); + if (SCX_HAS_OP(sch, cgroup_cancel_move) && + p->scx.cgrp_moving_from) + SCX_CALL_OP(sch, SCX_KF_UNLOCKED, cgroup_cancel_move, NULL, + 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) { - percpu_down_read(&scx_cgroup_rwsem); + struct scx_sched *sch = scx_root; - if (scx_cgroup_enabled && tg->scx_weight != weight) { - if (SCX_HAS_OP(cgroup_set_weight)) - SCX_CALL_OP(SCX_KF_UNLOCKED, cgroup_set_weight, - tg_cgrp(tg), weight); - tg->scx_weight = weight; - } + percpu_down_read(&scx_cgroup_ops_rwsem); + + if (scx_cgroup_enabled && SCX_HAS_OP(sch, cgroup_set_weight) && + tg->scx.weight != weight) + SCX_CALL_OP(sch, SCX_KF_UNLOCKED, cgroup_set_weight, NULL, + tg_cgrp(tg), weight); - percpu_up_read(&scx_cgroup_rwsem); + tg->scx.weight = weight; + + percpu_up_read(&scx_cgroup_ops_rwsem); } void scx_group_set_idle(struct task_group *tg, bool idle) { - percpu_down_read(&scx_cgroup_rwsem); - scx_cgroup_warn_missing_idle(tg); - percpu_up_read(&scx_cgroup_rwsem); + struct scx_sched *sch = scx_root; + + percpu_down_read(&scx_cgroup_ops_rwsem); + + if (scx_cgroup_enabled && SCX_HAS_OP(sch, cgroup_set_idle)) + SCX_CALL_OP(sch, SCX_KF_UNLOCKED, cgroup_set_idle, NULL, + tg_cgrp(tg), idle); + + /* Update the task group's idle state */ + tg->scx.idle = idle; + + percpu_up_read(&scx_cgroup_ops_rwsem); +} + +void scx_group_set_bandwidth(struct task_group *tg, + u64 period_us, u64 quota_us, u64 burst_us) +{ + struct scx_sched *sch = scx_root; + + 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 || + tg->scx.bw_quota_us != quota_us || + tg->scx.bw_burst_us != burst_us)) + SCX_CALL_OP(sch, SCX_KF_UNLOCKED, cgroup_set_bandwidth, NULL, + tg_cgrp(tg), period_us, quota_us, burst_us); + + tg->scx.bw_period_us = period_us; + tg->scx.bw_quota_us = quota_us; + tg->scx.bw_burst_us = burst_us; + + 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 */ @@ -4318,6 +3336,8 @@ static inline void scx_cgroup_unlock(void) {} * their current sched_class. Call them directly from sched core instead. */ DEFINE_SCHED_CLASS(ext) = { + .queue_mask = 1, + .enqueue_task = enqueue_task_scx, .dequeue_task = dequeue_task_scx, .yield_task = yield_task_scx, @@ -4325,20 +3345,17 @@ DEFINE_SCHED_CLASS(ext) = { .wakeup_preempt = wakeup_preempt_scx, - .balance = balance_scx, .pick_task = pick_task_scx, .put_prev_task = put_prev_task_scx, .set_next_task = set_next_task_scx, -#ifdef CONFIG_SMP .select_task_rq = select_task_rq_scx, .task_woken = task_woken_scx, .set_cpus_allowed = set_cpus_allowed_scx, .rq_online = rq_online_scx, .rq_offline = rq_offline_scx, -#endif .task_tick = task_tick_scx, @@ -4364,29 +3381,6 @@ static void init_dsq(struct scx_dispatch_q *dsq, u64 dsq_id) dsq->id = dsq_id; } -static struct scx_dispatch_q *create_dsq(u64 dsq_id, int node) -{ - struct scx_dispatch_q *dsq; - int ret; - - if (dsq_id & SCX_DSQ_FLAG_BUILTIN) - return ERR_PTR(-EINVAL); - - dsq = kmalloc_node(sizeof(*dsq), GFP_KERNEL, node); - if (!dsq) - return ERR_PTR(-ENOMEM); - - init_dsq(dsq, dsq_id); - - ret = rhashtable_insert_fast(&dsq_hash, &dsq->hash_node, - dsq_hash_params); - if (ret) { - kfree(dsq); - return ERR_PTR(ret); - } - return dsq; -} - static void free_dsq_irq_workfn(struct irq_work *irq_work) { struct llist_node *to_free = llist_del_all(&dsqs_to_free); @@ -4398,26 +3392,27 @@ static void free_dsq_irq_workfn(struct irq_work *irq_work) static DEFINE_IRQ_WORK(free_dsq_irq_work, free_dsq_irq_workfn); -static void destroy_dsq(u64 dsq_id) +static void destroy_dsq(struct scx_sched *sch, u64 dsq_id) { struct scx_dispatch_q *dsq; unsigned long flags; rcu_read_lock(); - dsq = find_user_dsq(dsq_id); + dsq = find_user_dsq(sch, dsq_id); if (!dsq) goto out_unlock_rcu; raw_spin_lock_irqsave(&dsq->lock, flags); if (dsq->nr) { - scx_ops_error("attempting to destroy in-use dsq 0x%016llx (nr=%u)", - dsq->id, dsq->nr); + scx_error(sch, "attempting to destroy in-use dsq 0x%016llx (nr=%u)", + dsq->id, dsq->nr); goto out_unlock_dsq; } - if (rhashtable_remove_fast(&dsq_hash, &dsq->hash_node, dsq_hash_params)) + if (rhashtable_remove_fast(&sch->dsq_hash, &dsq->hash_node, + dsq_hash_params)) goto out_unlock_dsq; /* @@ -4437,89 +3432,67 @@ out_unlock_rcu: } #ifdef CONFIG_EXT_GROUP_SCHED -static void scx_cgroup_exit(void) +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); - if (!(tg->scx_flags & SCX_TG_INITED)) - continue; - tg->scx_flags &= ~SCX_TG_INITED; - - if (!scx_ops.cgroup_exit) + if (!(tg->scx.flags & SCX_TG_INITED)) continue; + tg->scx.flags &= ~SCX_TG_INITED; - if (WARN_ON_ONCE(!css_tryget(css))) + if (!sch->ops.cgroup_exit) continue; - rcu_read_unlock(); - SCX_CALL_OP(SCX_KF_UNLOCKED, cgroup_exit, css->cgroup); - - rcu_read_lock(); - css_put(css); + SCX_CALL_OP(sch, SCX_KF_UNLOCKED, cgroup_exit, NULL, + css->cgroup); } - rcu_read_unlock(); } -static int scx_cgroup_init(void) +static int scx_cgroup_init(struct scx_sched *sch) { struct cgroup_subsys_state *css; int ret; - percpu_rwsem_assert_held(&scx_cgroup_rwsem); - - cgroup_warned_missing_weight = false; - cgroup_warned_missing_idle = false; - /* - * scx_tg_on/offline() are excluded thorugh 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 = { .weight = tg->scx_weight }; - - scx_cgroup_warn_missing_weight(tg); - scx_cgroup_warn_missing_idle(tg); + struct scx_cgroup_init_args args = { + .weight = tg->scx.weight, + .bw_period_us = tg->scx.bw_period_us, + .bw_quota_us = tg->scx.bw_quota_us, + .bw_burst_us = tg->scx.bw_burst_us, + }; - if ((tg->scx_flags & + if ((tg->scx.flags & (SCX_TG_ONLINE | SCX_TG_INITED)) != SCX_TG_ONLINE) continue; - if (!scx_ops.cgroup_init) { - tg->scx_flags |= SCX_TG_INITED; + if (!sch->ops.cgroup_init) { + tg->scx.flags |= SCX_TG_INITED; continue; } - if (WARN_ON_ONCE(!css_tryget(css))) - continue; - rcu_read_unlock(); - - ret = SCX_CALL_OP_RET(SCX_KF_UNLOCKED, cgroup_init, + ret = SCX_CALL_OP_RET(sch, SCX_KF_UNLOCKED, cgroup_init, NULL, css->cgroup, &args); if (ret) { css_put(css); - scx_ops_error("ops.cgroup_init() failed (%d)", ret); + scx_error(sch, "ops.cgroup_init() failed (%d)", ret); return ret; } - tg->scx_flags |= SCX_TG_INITED; - - rcu_read_lock(); - css_put(css); + tg->scx.flags |= SCX_TG_INITED; } - rcu_read_unlock(); WARN_ON_ONCE(scx_cgroup_enabled); scx_cgroup_enabled = true; @@ -4528,8 +3501,8 @@ static int scx_cgroup_init(void) } #else -static void scx_cgroup_exit(void) {} -static int scx_cgroup_init(void) { return 0; } +static void scx_cgroup_exit(struct scx_sched *sch) {} +static int scx_cgroup_init(struct scx_sched *sch) { return 0; } #endif @@ -4546,8 +3519,7 @@ static int scx_cgroup_init(void) { return 0; } static ssize_t scx_attr_state_show(struct kobject *kobj, struct kobj_attribute *ka, char *buf) { - return sysfs_emit(buf, "%s\n", - scx_ops_enable_state_str[scx_ops_enable_state()]); + return sysfs_emit(buf, "%s\n", scx_enable_state_str[scx_enable_state()]); } SCX_ATTR(state); @@ -4592,20 +3564,84 @@ static const struct attribute_group scx_global_attr_group = { .attrs = scx_global_attrs, }; +static void free_exit_info(struct scx_exit_info *ei); + +static void scx_sched_free_rcu_work(struct work_struct *work) +{ + struct rcu_work *rcu_work = to_rcu_work(work); + struct scx_sched *sch = container_of(rcu_work, struct scx_sched, rcu_work); + struct rhashtable_iter rht_iter; + struct scx_dispatch_q *dsq; + int node; + + irq_work_sync(&sch->error_irq_work); + kthread_stop(sch->helper->task); + + free_percpu(sch->pcpu); + + for_each_node_state(node, N_POSSIBLE) + kfree(sch->global_dsqs[node]); + kfree(sch->global_dsqs); + + rhashtable_walk_enter(&sch->dsq_hash, &rht_iter); + do { + rhashtable_walk_start(&rht_iter); + + while ((dsq = rhashtable_walk_next(&rht_iter)) && !IS_ERR(dsq)) + destroy_dsq(sch, dsq->id); + + rhashtable_walk_stop(&rht_iter); + } while (dsq == ERR_PTR(-EAGAIN)); + rhashtable_walk_exit(&rht_iter); + + rhashtable_free_and_destroy(&sch->dsq_hash, NULL, NULL); + free_exit_info(sch->exit_info); + kfree(sch); +} + static void scx_kobj_release(struct kobject *kobj) { - kfree(kobj); + struct scx_sched *sch = container_of(kobj, struct scx_sched, kobj); + + INIT_RCU_WORK(&sch->rcu_work, scx_sched_free_rcu_work); + queue_rcu_work(system_unbound_wq, &sch->rcu_work); } static ssize_t scx_attr_ops_show(struct kobject *kobj, struct kobj_attribute *ka, char *buf) { - return sysfs_emit(buf, "%s\n", scx_ops.name); + return sysfs_emit(buf, "%s\n", scx_root->ops.name); } SCX_ATTR(ops); +#define scx_attr_event_show(buf, at, events, kind) ({ \ + sysfs_emit_at(buf, at, "%s %llu\n", #kind, (events)->kind); \ +}) + +static ssize_t scx_attr_events_show(struct kobject *kobj, + struct kobj_attribute *ka, char *buf) +{ + struct scx_sched *sch = container_of(kobj, struct scx_sched, kobj); + struct scx_event_stats events; + int at = 0; + + scx_read_events(sch, &events); + at += scx_attr_event_show(buf, at, &events, SCX_EV_SELECT_CPU_FALLBACK); + at += scx_attr_event_show(buf, at, &events, SCX_EV_DISPATCH_LOCAL_DSQ_OFFLINE); + at += scx_attr_event_show(buf, at, &events, SCX_EV_DISPATCH_KEEP_LAST); + at += scx_attr_event_show(buf, at, &events, SCX_EV_ENQ_SKIP_EXITING); + at += scx_attr_event_show(buf, at, &events, SCX_EV_ENQ_SKIP_MIGRATION_DISABLED); + at += scx_attr_event_show(buf, at, &events, SCX_EV_REFILL_SLICE_DFL); + at += scx_attr_event_show(buf, at, &events, SCX_EV_BYPASS_DURATION); + at += scx_attr_event_show(buf, at, &events, SCX_EV_BYPASS_DISPATCH); + at += scx_attr_event_show(buf, at, &events, SCX_EV_BYPASS_ACTIVATE); + return at; +} +SCX_ATTR(events); + static struct attribute *scx_sched_attrs[] = { &scx_attr_ops.attr, + &scx_attr_events.attr, NULL, }; ATTRIBUTE_GROUPS(scx_sched); @@ -4618,7 +3654,7 @@ static const struct kobj_type scx_ktype = { static int scx_uevent(const struct kobject *kobj, struct kobj_uevent_env *env) { - return add_uevent_var(env, "SCXOPS=%s", scx_ops.name); + return add_uevent_var(env, "SCXOPS=%s", scx_root->ops.name); } static const struct kset_uevent_ops scx_uevent_ops = { @@ -4631,59 +3667,330 @@ static const struct kset_uevent_ops scx_uevent_ops = { */ bool task_should_scx(int policy) { - if (!scx_enabled() || - unlikely(scx_ops_enable_state() == SCX_OPS_DISABLING)) + if (!scx_enabled() || unlikely(scx_enable_state() == SCX_DISABLING)) return false; if (READ_ONCE(scx_switching_all)) return true; return policy == SCHED_EXT; } +bool scx_allow_ttwu_queue(const struct task_struct *p) +{ + 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; +} + +/** + * handle_lockup - sched_ext common lockup handler + * @fmt: format string + * + * Called on system stall or lockup condition and initiates abort of sched_ext + * if enabled, which may resolve the reported lockup. + * + * Returns %true if sched_ext is enabled and abort was initiated, which may + * resolve the lockup. %false if sched_ext is not enabled or abort was already + * initiated by someone else. + */ +static __printf(1, 2) bool handle_lockup(const char *fmt, ...) +{ + struct scx_sched *sch; + va_list args; + bool ret; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return false; + + switch (scx_enable_state()) { + case SCX_ENABLING: + case SCX_ENABLED: + va_start(args, fmt); + ret = scx_verror(sch, fmt, args); + va_end(args); + return ret; + default: + return false; + } +} + +/** + * scx_rcu_cpu_stall - sched_ext RCU CPU stall handler + * + * While there are various reasons why RCU CPU stalls can occur on a system + * that may not be caused by the current BPF scheduler, try kicking out the + * current scheduler in an attempt to recover the system to a good state before + * issuing panics. + * + * Returns %true if sched_ext is enabled and abort was initiated, which may + * resolve the reported RCU stall. %false if sched_ext is not enabled or someone + * else already initiated abort. + */ +bool scx_rcu_cpu_stall(void) +{ + return handle_lockup("RCU CPU stall detected!"); +} + /** * scx_softlockup - sched_ext softlockup handler + * @dur_s: number of seconds of CPU stuck due to soft lockup * * On some multi-socket setups (e.g. 2x Intel 8480c), the BPF scheduler can * live-lock the system by making many CPUs target the same DSQ to the point * where soft-lockup detection triggers. This function is called from * soft-lockup watchdog when the triggering point is close and tries to unjam - * the system by enabling the breather and aborting the BPF scheduler. + * the system and aborting the BPF scheduler. */ void scx_softlockup(u32 dur_s) { - switch (scx_ops_enable_state()) { - case SCX_OPS_ENABLING: - case SCX_OPS_ENABLED: - break; - default: + if (!handle_lockup("soft lockup - CPU %d stuck for %us", smp_processor_id(), dur_s)) return; + + printk_deferred(KERN_ERR "sched_ext: Soft lockup - CPU %d stuck for %us, disabling BPF scheduler\n", + smp_processor_id(), dur_s); +} + +/** + * scx_hardlockup - sched_ext hardlockup handler + * + * A poorly behaving BPF scheduler can trigger hard lockup by e.g. putting + * numerous affinitized tasks in a single queue and directing all CPUs at it. + * Try kicking out the current scheduler in an attempt to recover the system to + * a good state before taking more drastic actions. + * + * Returns %true if sched_ext is enabled and abort was initiated, which may + * resolve the reported hardlockdup. %false if sched_ext is not enabled or + * someone else already initiated abort. + */ +bool scx_hardlockup(int cpu) +{ + if (!handle_lockup("hard lockup - CPU %d", cpu)) + return false; + + printk_deferred(KERN_ERR "sched_ext: Hard lockup - CPU %d, disabling BPF scheduler\n", + cpu); + return true; +} + +static u32 bypass_lb_cpu(struct scx_sched *sch, struct rq *rq, + struct cpumask *donee_mask, struct cpumask *resched_mask, + u32 nr_donor_target, u32 nr_donee_target) +{ + struct scx_dispatch_q *donor_dsq = &rq->scx.bypass_dsq; + struct task_struct *p, *n; + struct scx_dsq_list_node cursor = INIT_DSQ_LIST_CURSOR(cursor, 0, 0); + s32 delta = READ_ONCE(donor_dsq->nr) - nr_donor_target; + u32 nr_balanced = 0, min_delta_us; + + /* + * All we want to guarantee is reasonable forward progress. No reason to + * fine tune. Assuming every task on @donor_dsq runs their full slice, + * consider offloading iff the total queued duration is over the + * threshold. + */ + min_delta_us = scx_bypass_lb_intv_us / SCX_BYPASS_LB_MIN_DELTA_DIV; + if (delta < DIV_ROUND_UP(min_delta_us, scx_slice_bypass_us)) + return 0; + + raw_spin_rq_lock_irq(rq); + raw_spin_lock(&donor_dsq->lock); + list_add(&cursor.node, &donor_dsq->list); +resume: + n = container_of(&cursor, struct task_struct, scx.dsq_list); + n = nldsq_next_task(donor_dsq, n, false); + + while ((p = n)) { + struct rq *donee_rq; + struct scx_dispatch_q *donee_dsq; + int donee; + + n = nldsq_next_task(donor_dsq, n, false); + + if (donor_dsq->nr <= nr_donor_target) + break; + + if (cpumask_empty(donee_mask)) + break; + + donee = cpumask_any_and_distribute(donee_mask, p->cpus_ptr); + if (donee >= nr_cpu_ids) + continue; + + donee_rq = cpu_rq(donee); + donee_dsq = &donee_rq->scx.bypass_dsq; + + /* + * $p's rq is not locked but $p's DSQ lock protects its + * scheduling properties making this test safe. + */ + if (!task_can_run_on_remote_rq(sch, p, donee_rq, false)) + continue; + + /* + * Moving $p from one non-local DSQ to another. The source rq + * and DSQ are already locked. Do an abbreviated dequeue and + * then perform enqueue without unlocking $donor_dsq. + * + * We don't want to drop and reacquire the lock on each + * iteration as @donor_dsq can be very long and potentially + * highly contended. Donee DSQs are less likely to be contended. + * The nested locking is safe as only this LB moves tasks + * between bypass DSQs. + */ + dispatch_dequeue_locked(p, donor_dsq); + dispatch_enqueue(sch, donee_dsq, p, SCX_ENQ_NESTED); + + /* + * $donee might have been idle and need to be woken up. No need + * to be clever. Kick every CPU that receives tasks. + */ + cpumask_set_cpu(donee, resched_mask); + + if (READ_ONCE(donee_dsq->nr) >= nr_donee_target) + cpumask_clear_cpu(donee, donee_mask); + + nr_balanced++; + if (!(nr_balanced % SCX_BYPASS_LB_BATCH) && n) { + list_move_tail(&cursor.node, &n->scx.dsq_list.node); + raw_spin_unlock(&donor_dsq->lock); + raw_spin_rq_unlock_irq(rq); + cpu_relax(); + raw_spin_rq_lock_irq(rq); + raw_spin_lock(&donor_dsq->lock); + goto resume; + } } - /* allow only one instance, cleared at the end of scx_ops_bypass() */ - if (test_and_set_bit(0, &scx_in_softlockup)) - return; + list_del_init(&cursor.node); + raw_spin_unlock(&donor_dsq->lock); + raw_spin_rq_unlock_irq(rq); + + return nr_balanced; +} + +static void bypass_lb_node(struct scx_sched *sch, int node) +{ + const struct cpumask *node_mask = cpumask_of_node(node); + struct cpumask *donee_mask = scx_bypass_lb_donee_cpumask; + struct cpumask *resched_mask = scx_bypass_lb_resched_cpumask; + u32 nr_tasks = 0, nr_cpus = 0, nr_balanced = 0; + u32 nr_target, nr_donor_target; + u32 before_min = U32_MAX, before_max = 0; + u32 after_min = U32_MAX, after_max = 0; + int cpu; - printk_deferred(KERN_ERR "sched_ext: Soft lockup - CPU%d stuck for %us, disabling \"%s\"\n", - smp_processor_id(), dur_s, scx_ops.name); + /* count the target tasks and CPUs */ + for_each_cpu_and(cpu, cpu_online_mask, node_mask) { + u32 nr = READ_ONCE(cpu_rq(cpu)->scx.bypass_dsq.nr); + + nr_tasks += nr; + nr_cpus++; + + before_min = min(nr, before_min); + before_max = max(nr, before_max); + } + + if (!nr_cpus) + return; /* - * Some CPUs may be trapped in the dispatch paths. Enable breather - * immediately; otherwise, we might even be able to get to - * scx_ops_bypass(). + * We don't want CPUs to have more than $nr_donor_target tasks and + * balancing to fill donee CPUs upto $nr_target. Once targets are + * calculated, find the donee CPUs. */ - atomic_inc(&scx_ops_breather_depth); + nr_target = DIV_ROUND_UP(nr_tasks, nr_cpus); + nr_donor_target = DIV_ROUND_UP(nr_target * SCX_BYPASS_LB_DONOR_PCT, 100); + + cpumask_clear(donee_mask); + for_each_cpu_and(cpu, cpu_online_mask, node_mask) { + if (READ_ONCE(cpu_rq(cpu)->scx.bypass_dsq.nr) < nr_target) + cpumask_set_cpu(cpu, donee_mask); + } + + /* iterate !donee CPUs and see if they should be offloaded */ + cpumask_clear(resched_mask); + for_each_cpu_and(cpu, cpu_online_mask, node_mask) { + struct rq *rq = cpu_rq(cpu); + struct scx_dispatch_q *donor_dsq = &rq->scx.bypass_dsq; + + if (cpumask_empty(donee_mask)) + break; + if (cpumask_test_cpu(cpu, donee_mask)) + continue; + if (READ_ONCE(donor_dsq->nr) <= nr_donor_target) + continue; + + nr_balanced += bypass_lb_cpu(sch, rq, donee_mask, resched_mask, + nr_donor_target, nr_target); + } + + for_each_cpu(cpu, resched_mask) { + struct rq *rq = cpu_rq(cpu); + + raw_spin_rq_lock_irq(rq); + resched_curr(rq); + raw_spin_rq_unlock_irq(rq); + } + + for_each_cpu_and(cpu, cpu_online_mask, node_mask) { + u32 nr = READ_ONCE(cpu_rq(cpu)->scx.bypass_dsq.nr); + + after_min = min(nr, after_min); + after_max = max(nr, after_max); + + } - scx_ops_error("soft lockup - CPU#%d stuck for %us", - smp_processor_id(), dur_s); + trace_sched_ext_bypass_lb(node, nr_cpus, nr_tasks, nr_balanced, + before_min, before_max, after_min, after_max); } -static void scx_clear_softlockup(void) +/* + * In bypass mode, all tasks are put on the per-CPU bypass DSQs. If the machine + * is over-saturated and the BPF scheduler skewed tasks into few CPUs, some + * bypass DSQs can be overloaded. If there are enough tasks to saturate other + * lightly loaded CPUs, such imbalance can lead to very high execution latency + * on the overloaded CPUs and thus to hung tasks and RCU stalls. To avoid such + * outcomes, a simple load balancing mechanism is implemented by the following + * timer which runs periodically while bypass mode is in effect. + */ +static void scx_bypass_lb_timerfn(struct timer_list *timer) { - if (test_and_clear_bit(0, &scx_in_softlockup)) - atomic_dec(&scx_ops_breather_depth); + struct scx_sched *sch; + int node; + u32 intv_us; + + sch = rcu_dereference_all(scx_root); + if (unlikely(!sch) || !READ_ONCE(scx_bypass_depth)) + return; + + for_each_node_with_cpus(node) + bypass_lb_node(sch, node); + + intv_us = READ_ONCE(scx_bypass_lb_intv_us); + if (intv_us) + mod_timer(timer, jiffies + usecs_to_jiffies(intv_us)); } +static DEFINE_TIMER(scx_bypass_lb_timer, scx_bypass_lb_timerfn); + /** - * scx_ops_bypass - [Un]bypass scx_ops and guarantee forward progress + * scx_bypass - [Un]bypass scx_ops and guarantee forward progress + * @bypass: true for bypass, false for unbypass * * Bypassing guarantees that all runnable tasks make forward progress without * trusting the BPF scheduler. We can't grab any mutexes or rwsems as they might @@ -4707,32 +4014,51 @@ 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. */ -static void scx_ops_bypass(bool bypass) +static void scx_bypass(bool bypass) { static DEFINE_RAW_SPINLOCK(bypass_lock); - int cpu; + static unsigned long bypass_timestamp; + struct scx_sched *sch; unsigned long flags; + int cpu; raw_spin_lock_irqsave(&bypass_lock, flags); + sch = rcu_dereference_bh(scx_root); + if (bypass) { - scx_ops_bypass_depth++; - WARN_ON_ONCE(scx_ops_bypass_depth <= 0); - if (scx_ops_bypass_depth != 1) + u32 intv_us; + + WRITE_ONCE(scx_bypass_depth, scx_bypass_depth + 1); + WARN_ON_ONCE(scx_bypass_depth <= 0); + if (scx_bypass_depth != 1) goto unlock; + WRITE_ONCE(scx_slice_dfl, scx_slice_bypass_us * NSEC_PER_USEC); + bypass_timestamp = ktime_get_ns(); + if (sch) + scx_add_event(sch, SCX_EV_BYPASS_ACTIVATE, 1); + + intv_us = READ_ONCE(scx_bypass_lb_intv_us); + if (intv_us && !timer_pending(&scx_bypass_lb_timer)) { + scx_bypass_lb_timer.expires = + jiffies + usecs_to_jiffies(intv_us); + add_timer_global(&scx_bypass_lb_timer); + } } else { - scx_ops_bypass_depth--; - WARN_ON_ONCE(scx_ops_bypass_depth < 0); - if (scx_ops_bypass_depth != 0) + WRITE_ONCE(scx_bypass_depth, scx_bypass_depth - 1); + WARN_ON_ONCE(scx_bypass_depth < 0); + if (scx_bypass_depth != 0) goto unlock; + WRITE_ONCE(scx_slice_dfl, SCX_SLICE_DFL); + if (sch) + scx_add_event(sch, SCX_EV_BYPASS_DURATION, + ktime_get_ns() - bypass_timestamp); } - atomic_inc(&scx_ops_breather_depth); - /* * No task property is changing. We just need to make sure all currently * queued tasks are re-queued according to the new scx_rq_bypassing() @@ -4744,10 +4070,9 @@ static void scx_ops_bypass(bool bypass) */ for_each_possible_cpu(cpu) { struct rq *rq = cpu_rq(cpu); - struct rq_flags rf; struct task_struct *p, *n; - rq_lock(rq, &rf); + raw_spin_rq_lock(rq); if (bypass) { WARN_ON_ONCE(rq->scx.flags & SCX_RQ_BYPASSING); @@ -4763,7 +4088,7 @@ static void scx_ops_bypass(bool bypass) * sees scx_rq_bypassing() before moving tasks to SCX. */ if (!scx_enabled()) { - rq_unlock_irqrestore(rq, &rf); + raw_spin_rq_unlock(rq); continue; } @@ -4776,28 +4101,26 @@ static void scx_ops_bypass(bool bypass) */ list_for_each_entry_safe_reverse(p, n, &rq->scx.runnable_list, scx.runnable_node) { - struct sched_enq_and_set_ctx ctx; - /* cycling deq/enq is enough, see the function comment */ - sched_deq_and_put_task(p, DEQUEUE_SAVE | DEQUEUE_MOVE, &ctx); - sched_enq_and_set_task(&ctx); + scoped_guard (sched_change, p, DEQUEUE_SAVE | DEQUEUE_MOVE) { + /* nothing */ ; + } } - rq_unlock(rq, &rf); - /* resched to restore ticks and idle state */ - resched_cpu(cpu); + if (cpu_online(cpu) || cpu == smp_processor_id()) + resched_curr(rq); + + raw_spin_rq_unlock(rq); } - atomic_dec(&scx_ops_breather_depth); unlock: raw_spin_unlock_irqrestore(&bypass_lock, flags); - scx_clear_softlockup(); } static void free_exit_info(struct scx_exit_info *ei) { - kfree(ei->dump); + kvfree(ei->dump); kfree(ei->msg); kfree(ei->bt); kfree(ei); @@ -4813,7 +4136,7 @@ static struct scx_exit_info *alloc_exit_info(size_t exit_dump_len) ei->bt = kcalloc(SCX_EXIT_BT_LEN, sizeof(ei->bt[0]), GFP_KERNEL); ei->msg = kzalloc(SCX_EXIT_MSG_LEN, GFP_KERNEL); - ei->dump = kzalloc(exit_dump_len, GFP_KERNEL); + ei->dump = kvzalloc(exit_dump_len, GFP_KERNEL); if (!ei->bt || !ei->msg || !ei->dump) { free_exit_info(ei); @@ -4845,42 +4168,51 @@ static const char *scx_exit_reason(enum scx_exit_kind kind) } } -static void scx_ops_disable_workfn(struct kthread_work *work) +static void free_kick_syncs(void) +{ + int cpu; + + for_each_possible_cpu(cpu) { + struct scx_kick_syncs **ksyncs = per_cpu_ptr(&scx_kick_syncs, cpu); + struct scx_kick_syncs *to_free; + + to_free = rcu_replace_pointer(*ksyncs, NULL, true); + if (to_free) + kvfree_rcu(to_free, rcu); + } +} + +static void scx_disable_workfn(struct kthread_work *work) { - struct scx_exit_info *ei = scx_exit_info; + struct scx_sched *sch = container_of(work, struct scx_sched, disable_work); + struct scx_exit_info *ei = sch->exit_info; struct scx_task_iter sti; struct task_struct *p; - struct rhashtable_iter rht_iter; - struct scx_dispatch_q *dsq; - int i, kind; + int kind, cpu; - kind = atomic_read(&scx_exit_kind); + kind = atomic_read(&sch->exit_kind); while (true) { - /* - * NONE indicates that a new scx_ops has been registered since - * disable was scheduled - don't kill the new ops. DONE - * indicates that the ops has already been disabled. - */ - if (kind == SCX_EXIT_NONE || kind == SCX_EXIT_DONE) + if (kind == SCX_EXIT_DONE) /* already disabled? */ return; - if (atomic_try_cmpxchg(&scx_exit_kind, &kind, SCX_EXIT_DONE)) + WARN_ON_ONCE(kind == SCX_EXIT_NONE); + if (atomic_try_cmpxchg(&sch->exit_kind, &kind, SCX_EXIT_DONE)) break; } ei->kind = kind; ei->reason = scx_exit_reason(ei->kind); /* guarantee forward progress by bypassing scx_ops */ - scx_ops_bypass(true); + scx_bypass(true); + WRITE_ONCE(scx_aborting, false); - switch (scx_ops_set_enable_state(SCX_OPS_DISABLING)) { - case SCX_OPS_DISABLING: + switch (scx_set_enable_state(SCX_DISABLING)) { + case SCX_DISABLING: WARN_ONCE(true, "sched_ext: duplicate disabling instance?"); break; - case SCX_OPS_DISABLED: + case SCX_DISABLED: pr_warn("sched_ext: ops error detected without ops (%s)\n", - scx_exit_info->msg); - WARN_ON_ONCE(scx_ops_set_enable_state(SCX_OPS_DISABLED) != - SCX_OPS_DISABLING); + sch->exit_info->msg); + WARN_ON_ONCE(scx_set_enable_state(SCX_DISABLED) != SCX_DISABLING); goto done; default: break; @@ -4891,17 +4223,17 @@ static void scx_ops_disable_workfn(struct kthread_work *work) * we can safely use blocking synchronization constructs. Actually * disable ops. */ - mutex_lock(&scx_ops_enable_mutex); + mutex_lock(&scx_enable_mutex); static_branch_disable(&__scx_switched_all); WRITE_ONCE(scx_switching_all, false); /* * Shut down cgroup support before tasks so that the cgroup attach path - * doesn't race against scx_ops_exit_task(). + * doesn't race against scx_exit_task(). */ scx_cgroup_lock(); - scx_cgroup_exit(); + scx_cgroup_exit(sch); scx_cgroup_unlock(); /* @@ -4910,122 +4242,119 @@ static void scx_ops_disable_workfn(struct kthread_work *work) */ percpu_down_write(&scx_fork_rwsem); - scx_ops_init_task_enabled = false; + scx_init_task_enabled = false; scx_task_iter_start(&sti); while ((p = scx_task_iter_next_locked(&sti))) { + unsigned int queue_flags = DEQUEUE_SAVE | DEQUEUE_MOVE | DEQUEUE_NOCLOCK; const struct sched_class *old_class = p->sched_class; - const struct sched_class *new_class = - __setscheduler_class(p->policy, p->prio); - struct sched_enq_and_set_ctx ctx; - - if (old_class != new_class && p->se.sched_delayed) - dequeue_task(task_rq(p), p, DEQUEUE_SLEEP | DEQUEUE_DELAYED); + const struct sched_class *new_class = scx_setscheduler_class(p); - sched_deq_and_put_task(p, DEQUEUE_SAVE | DEQUEUE_MOVE, &ctx); + update_rq_clock(task_rq(p)); - p->sched_class = new_class; - check_class_changing(task_rq(p), p, old_class); + if (old_class != new_class) + queue_flags |= DEQUEUE_CLASS; - sched_enq_and_set_task(&ctx); + scoped_guard (sched_change, p, queue_flags) { + p->sched_class = new_class; + } - check_class_changed(task_rq(p), p, old_class, p->prio); - scx_ops_exit_task(p); + scx_exit_task(p); } scx_task_iter_stop(&sti); percpu_up_write(&scx_fork_rwsem); + /* + * Invalidate all the rq clocks to prevent getting outdated + * rq clocks from a previous scx scheduler. + */ + for_each_possible_cpu(cpu) { + struct rq *rq = cpu_rq(cpu); + scx_rq_clock_invalidate(rq); + } + /* no task is on scx, turn off all the switches and flush in-progress calls */ - static_branch_disable(&__scx_ops_enabled); - for (i = SCX_OPI_BEGIN; i < SCX_OPI_END; i++) - static_branch_disable(&scx_has_op[i]); - static_branch_disable(&scx_ops_enq_last); - static_branch_disable(&scx_ops_enq_exiting); - static_branch_disable(&scx_ops_cpu_preempt); - static_branch_disable(&scx_builtin_idle_enabled); + static_branch_disable(&__scx_enabled); + bitmap_zero(sch->has_op, SCX_OPI_END); + scx_idle_disable(); synchronize_rcu(); if (ei->kind >= SCX_EXIT_ERROR) { pr_err("sched_ext: BPF scheduler \"%s\" disabled (%s)\n", - scx_ops.name, ei->reason); + sch->ops.name, ei->reason); if (ei->msg[0] != '\0') - pr_err("sched_ext: %s: %s\n", scx_ops.name, ei->msg); + pr_err("sched_ext: %s: %s\n", sch->ops.name, ei->msg); #ifdef CONFIG_STACKTRACE stack_trace_print(ei->bt, ei->bt_len, 2); #endif } else { pr_info("sched_ext: BPF scheduler \"%s\" disabled (%s)\n", - scx_ops.name, ei->reason); + sch->ops.name, ei->reason); } - if (scx_ops.exit) - SCX_CALL_OP(SCX_KF_UNLOCKED, exit, ei); + if (sch->ops.exit) + SCX_CALL_OP(sch, SCX_KF_UNLOCKED, exit, NULL, ei); cancel_delayed_work_sync(&scx_watchdog_work); /* - * Delete the kobject from the hierarchy eagerly in addition to just - * dropping a reference. Otherwise, if the object is deleted - * asynchronously, sysfs could observe an object of the same name still - * in the hierarchy when another scheduler is loaded. + * scx_root clearing must be inside cpus_read_lock(). See + * handle_hotplug(). */ - kobject_del(scx_root_kobj); - kobject_put(scx_root_kobj); - scx_root_kobj = NULL; - - memset(&scx_ops, 0, sizeof(scx_ops)); - - rhashtable_walk_enter(&dsq_hash, &rht_iter); - do { - rhashtable_walk_start(&rht_iter); - - while ((dsq = rhashtable_walk_next(&rht_iter)) && !IS_ERR(dsq)) - destroy_dsq(dsq->id); + cpus_read_lock(); + RCU_INIT_POINTER(scx_root, NULL); + cpus_read_unlock(); - rhashtable_walk_stop(&rht_iter); - } while (dsq == ERR_PTR(-EAGAIN)); - rhashtable_walk_exit(&rht_iter); + /* + * Delete the kobject from the hierarchy synchronously. Otherwise, sysfs + * could observe an object of the same name still in the hierarchy when + * the next scheduler is loaded. + */ + kobject_del(&sch->kobj); free_percpu(scx_dsp_ctx); scx_dsp_ctx = NULL; scx_dsp_max_batch = 0; + free_kick_syncs(); - free_exit_info(scx_exit_info); - scx_exit_info = NULL; - - mutex_unlock(&scx_ops_enable_mutex); + mutex_unlock(&scx_enable_mutex); - WARN_ON_ONCE(scx_ops_set_enable_state(SCX_OPS_DISABLED) != - SCX_OPS_DISABLING); + WARN_ON_ONCE(scx_set_enable_state(SCX_DISABLED) != SCX_DISABLING); done: - scx_ops_bypass(false); + scx_bypass(false); } -static DEFINE_KTHREAD_WORK(scx_ops_disable_work, scx_ops_disable_workfn); - -static void schedule_scx_ops_disable_work(void) +static bool scx_claim_exit(struct scx_sched *sch, enum scx_exit_kind kind) { - struct kthread_worker *helper = READ_ONCE(scx_ops_helper); + int none = SCX_EXIT_NONE; + + if (!atomic_try_cmpxchg(&sch->exit_kind, &none, kind)) + return false; /* - * We may be called spuriously before the first bpf_sched_ext_reg(). If - * scx_ops_helper isn't set up yet, there's nothing to do. + * Some CPUs may be trapped in the dispatch paths. Set the aborting + * flag to break potential live-lock scenarios, ensuring we can + * successfully reach scx_bypass(). */ - if (helper) - kthread_queue_work(helper, &scx_ops_disable_work); + WRITE_ONCE(scx_aborting, true); + return true; } -static void scx_ops_disable(enum scx_exit_kind kind) +static void scx_disable(enum scx_exit_kind kind) { - int none = SCX_EXIT_NONE; + struct scx_sched *sch; if (WARN_ON_ONCE(kind == SCX_EXIT_NONE || kind == SCX_EXIT_DONE)) kind = SCX_EXIT_ERROR; - atomic_try_cmpxchg(&scx_exit_kind, &none, kind); - - schedule_scx_ops_disable_work(); + rcu_read_lock(); + sch = rcu_dereference(scx_root); + if (sch) { + scx_claim_exit(sch, kind); + kthread_queue_work(sch->helper, &sch->disable_work); + } + rcu_read_unlock(); } static void dump_newline(struct seq_buf *s) @@ -5143,6 +4472,7 @@ static void scx_dump_task(struct seq_buf *s, struct scx_dump_ctx *dctx, struct task_struct *p, char marker) { static unsigned long bt[SCX_EXIT_BT_LEN]; + struct scx_sched *sch = scx_root; char dsq_id_buf[19] = "(n/a)"; unsigned long ops_state = atomic_long_read(&p->scx.ops_state); unsigned int bt_len = 0; @@ -5159,14 +4489,16 @@ static void scx_dump_task(struct seq_buf *s, struct scx_dump_ctx *dctx, scx_get_task_state(p), p->scx.flags & ~SCX_TASK_STATE_MASK, p->scx.dsq_flags, ops_state & SCX_OPSS_STATE_MASK, ops_state >> SCX_OPSS_QSEQ_SHIFT); - dump_line(s, " sticky/holding_cpu=%d/%d dsq_id=%s dsq_vtime=%llu", - p->scx.sticky_cpu, p->scx.holding_cpu, dsq_id_buf, - p->scx.dsq_vtime); - dump_line(s, " cpus=%*pb", cpumask_pr_args(p->cpus_ptr)); - - if (SCX_HAS_OP(dump_task)) { + dump_line(s, " sticky/holding_cpu=%d/%d dsq_id=%s", + 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 no_mig=%u", cpumask_pr_args(p->cpus_ptr), + p->migration_disabled); + + if (SCX_HAS_OP(sch, dump_task)) { ops_dump_init(s, " "); - SCX_CALL_OP(SCX_KF_REST, dump_task, dctx, p); + SCX_CALL_OP(sch, SCX_KF_REST, dump_task, NULL, dctx, p); ops_dump_exit(); } @@ -5183,6 +4515,7 @@ static void scx_dump_state(struct scx_exit_info *ei, size_t dump_len) { static DEFINE_SPINLOCK(dump_lock); static const char trunc_marker[] = "\n\n~~~~ TRUNCATED ~~~~\n"; + struct scx_sched *sch = scx_root; struct scx_dump_ctx dctx = { .kind = ei->kind, .exit_code = ei->exit_code, @@ -5191,6 +4524,7 @@ static void scx_dump_state(struct scx_exit_info *ei, size_t dump_len) .at_jiffies = jiffies, }; struct seq_buf s; + struct scx_event_stats events; unsigned long flags; char *buf; int cpu; @@ -5210,9 +4544,9 @@ static void scx_dump_state(struct scx_exit_info *ei, size_t dump_len) dump_stack_trace(&s, " ", ei->bt, ei->bt_len); } - if (SCX_HAS_OP(dump)) { + if (SCX_HAS_OP(sch, dump)) { ops_dump_init(&s, ""); - SCX_CALL_OP(SCX_KF_UNLOCKED, dump, &dctx); + SCX_CALL_OP(sch, SCX_KF_UNLOCKED, dump, NULL, &dctx); ops_dump_exit(); } @@ -5228,12 +4562,12 @@ static void scx_dump_state(struct scx_exit_info *ei, size_t dump_len) size_t avail, used; bool idle; - rq_lock(rq, &rf); + rq_lock_irqsave(rq, &rf); idle = list_empty(&rq->scx.runnable_list) && rq->curr->sched_class == &idle_sched_class; - if (idle && !SCX_HAS_OP(dump_cpu)) + if (idle && !SCX_HAS_OP(sch, dump_cpu)) goto next; /* @@ -5246,10 +4580,10 @@ static void scx_dump_state(struct scx_exit_info *ei, size_t dump_len) seq_buf_init(&ns, buf, avail); dump_newline(&ns); - dump_line(&ns, "CPU %-4d: nr_run=%u flags=0x%x cpu_rel=%d ops_qseq=%lu pnt_seq=%lu", + dump_line(&ns, "CPU %-4d: nr_run=%u flags=0x%x cpu_rel=%d ops_qseq=%lu ksync=%lu", cpu, rq->scx.nr_running, rq->scx.flags, rq->scx.cpu_released, rq->scx.ops_qseq, - rq->scx.pnt_seq); + rq->scx.kick_sync); dump_line(&ns, " curr=%s[%d] class=%ps", rq->curr->comm, rq->curr->pid, rq->curr->sched_class); @@ -5267,9 +4601,10 @@ static void scx_dump_state(struct scx_exit_info *ei, size_t dump_len) cpumask_pr_args(rq->scx.cpus_to_wait)); used = seq_buf_used(&ns); - if (SCX_HAS_OP(dump_cpu)) { + if (SCX_HAS_OP(sch, dump_cpu)) { ops_dump_init(&ns, " "); - SCX_CALL_OP(SCX_KF_REST, dump_cpu, &dctx, cpu, idle); + SCX_CALL_OP(sch, SCX_KF_REST, dump_cpu, NULL, + &dctx, cpu, idle); ops_dump_exit(); } @@ -5296,9 +4631,24 @@ static void scx_dump_state(struct scx_exit_info *ei, size_t dump_len) list_for_each_entry(p, &rq->scx.runnable_list, scx.runnable_node) scx_dump_task(&s, &dctx, p, ' '); next: - rq_unlock(rq, &rf); + rq_unlock_irqrestore(rq, &rf); } + dump_newline(&s); + dump_line(&s, "Event counters"); + dump_line(&s, "--------------"); + + scx_read_events(sch, &events); + scx_dump_event(s, &events, SCX_EV_SELECT_CPU_FALLBACK); + scx_dump_event(s, &events, SCX_EV_DISPATCH_LOCAL_DSQ_OFFLINE); + scx_dump_event(s, &events, SCX_EV_DISPATCH_KEEP_LAST); + scx_dump_event(s, &events, SCX_EV_ENQ_SKIP_EXITING); + scx_dump_event(s, &events, SCX_EV_ENQ_SKIP_MIGRATION_DISABLED); + scx_dump_event(s, &events, SCX_EV_REFILL_SLICE_DFL); + scx_dump_event(s, &events, SCX_EV_BYPASS_DURATION); + scx_dump_event(s, &events, SCX_EV_BYPASS_DISPATCH); + scx_dump_event(s, &events, SCX_EV_BYPASS_ACTIVATE); + if (seq_buf_has_overflowed(&s) && dump_len >= sizeof(trunc_marker)) memcpy(ei->dump + dump_len - sizeof(trunc_marker), trunc_marker, sizeof(trunc_marker)); @@ -5306,59 +4656,154 @@ static void scx_dump_state(struct scx_exit_info *ei, size_t dump_len) spin_unlock_irqrestore(&dump_lock, flags); } -static void scx_ops_error_irq_workfn(struct irq_work *irq_work) +static void scx_error_irq_workfn(struct irq_work *irq_work) { - struct scx_exit_info *ei = scx_exit_info; + struct scx_sched *sch = container_of(irq_work, struct scx_sched, error_irq_work); + struct scx_exit_info *ei = sch->exit_info; if (ei->kind >= SCX_EXIT_ERROR) - scx_dump_state(ei, scx_ops.exit_dump_len); + scx_dump_state(ei, sch->ops.exit_dump_len); - schedule_scx_ops_disable_work(); + kthread_queue_work(sch->helper, &sch->disable_work); } -static DEFINE_IRQ_WORK(scx_ops_error_irq_work, scx_ops_error_irq_workfn); - -static __printf(3, 4) void scx_ops_exit_kind(enum scx_exit_kind kind, - s64 exit_code, - const char *fmt, ...) +static bool scx_vexit(struct scx_sched *sch, + enum scx_exit_kind kind, s64 exit_code, + const char *fmt, va_list args) { - struct scx_exit_info *ei = scx_exit_info; - int none = SCX_EXIT_NONE; - va_list args; + struct scx_exit_info *ei = sch->exit_info; - if (!atomic_try_cmpxchg(&scx_exit_kind, &none, kind)) - return; + if (!scx_claim_exit(sch, kind)) + return false; ei->exit_code = exit_code; #ifdef CONFIG_STACKTRACE if (kind >= SCX_EXIT_ERROR) ei->bt_len = stack_trace_save(ei->bt, SCX_EXIT_BT_LEN, 1); #endif - va_start(args, fmt); vscnprintf(ei->msg, SCX_EXIT_MSG_LEN, fmt, args); - va_end(args); /* * Set ei->kind and ->reason for scx_dump_state(). They'll be set again - * in scx_ops_disable_workfn(). + * in scx_disable_workfn(). */ ei->kind = kind; ei->reason = scx_exit_reason(ei->kind); - irq_work_queue(&scx_ops_error_irq_work); + irq_work_queue(&sch->error_irq_work); + return true; } -static struct kthread_worker *scx_create_rt_helper(const char *name) +static int alloc_kick_syncs(void) { - struct kthread_worker *helper; + int cpu; + + /* + * Allocate per-CPU arrays sized by nr_cpu_ids. Use kvzalloc as size + * can exceed percpu allocator limits on large machines. + */ + for_each_possible_cpu(cpu) { + struct scx_kick_syncs **ksyncs = per_cpu_ptr(&scx_kick_syncs, cpu); + struct scx_kick_syncs *new_ksyncs; + + WARN_ON_ONCE(rcu_access_pointer(*ksyncs)); - helper = kthread_create_worker(0, name); - if (helper) - sched_set_fifo(helper->task); - return helper; + new_ksyncs = kvzalloc_node(struct_size(new_ksyncs, syncs, nr_cpu_ids), + GFP_KERNEL, cpu_to_node(cpu)); + if (!new_ksyncs) { + free_kick_syncs(); + return -ENOMEM; + } + + rcu_assign_pointer(*ksyncs, new_ksyncs); + } + + return 0; } -static void check_hotplug_seq(const struct sched_ext_ops *ops) +static struct scx_sched *scx_alloc_and_add_sched(struct sched_ext_ops *ops) +{ + struct scx_sched *sch; + int node, ret; + + sch = kzalloc(sizeof(*sch), GFP_KERNEL); + if (!sch) + return ERR_PTR(-ENOMEM); + + sch->exit_info = alloc_exit_info(ops->exit_dump_len); + if (!sch->exit_info) { + ret = -ENOMEM; + goto err_free_sch; + } + + ret = rhashtable_init(&sch->dsq_hash, &dsq_hash_params); + if (ret < 0) + goto err_free_ei; + + sch->global_dsqs = kcalloc(nr_node_ids, sizeof(sch->global_dsqs[0]), + GFP_KERNEL); + if (!sch->global_dsqs) { + ret = -ENOMEM; + goto err_free_hash; + } + + for_each_node_state(node, N_POSSIBLE) { + struct scx_dispatch_q *dsq; + + dsq = kzalloc_node(sizeof(*dsq), GFP_KERNEL, node); + if (!dsq) { + ret = -ENOMEM; + goto err_free_gdsqs; + } + + init_dsq(dsq, SCX_DSQ_GLOBAL); + sch->global_dsqs[node] = dsq; + } + + 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 (IS_ERR(sch->helper)) { + ret = PTR_ERR(sch->helper); + goto err_free_pcpu; + } + + sched_set_fifo(sch->helper->task); + + atomic_set(&sch->exit_kind, SCX_EXIT_NONE); + init_irq_work(&sch->error_irq_work, scx_error_irq_workfn); + kthread_init_work(&sch->disable_work, scx_disable_workfn); + sch->ops = *ops; + ops->priv = sch; + + sch->kobj.kset = scx_kset; + ret = kobject_init_and_add(&sch->kobj, &scx_ktype, NULL, "root"); + if (ret < 0) + goto err_stop_helper; + + return sch; + +err_stop_helper: + kthread_stop(sch->helper->task); +err_free_pcpu: + free_percpu(sch->pcpu); +err_free_gdsqs: + for_each_node_state(node, N_POSSIBLE) + kfree(sch->global_dsqs[node]); + kfree(sch->global_dsqs); +err_free_hash: + rhashtable_free_and_destroy(&sch->dsq_hash, NULL, NULL); +err_free_ei: + free_exit_info(sch->exit_info); +err_free_sch: + kfree(sch); + return ERR_PTR(ret); +} + +static int check_hotplug_seq(struct scx_sched *sch, + const struct sched_ext_ops *ops) { unsigned long long global_hotplug_seq; @@ -5370,33 +4815,54 @@ static void check_hotplug_seq(const struct sched_ext_ops *ops) if (ops->hotplug_seq) { global_hotplug_seq = atomic_long_read(&scx_hotplug_seq); if (ops->hotplug_seq != global_hotplug_seq) { - scx_ops_exit(SCX_ECODE_ACT_RESTART | SCX_ECODE_RSN_HOTPLUG, - "expected hotplug seq %llu did not match actual %llu", - ops->hotplug_seq, global_hotplug_seq); + scx_exit(sch, SCX_EXIT_UNREG_KERN, + SCX_ECODE_ACT_RESTART | SCX_ECODE_RSN_HOTPLUG, + "expected hotplug seq %llu did not match actual %llu", + ops->hotplug_seq, global_hotplug_seq); + return -EBUSY; } } + + return 0; } -static int validate_ops(const struct sched_ext_ops *ops) +static int validate_ops(struct scx_sched *sch, const struct sched_ext_ops *ops) { /* * It doesn't make sense to specify the SCX_OPS_ENQ_LAST flag if the * ops.enqueue() callback isn't implemented. */ if ((ops->flags & SCX_OPS_ENQ_LAST) && !ops->enqueue) { - scx_ops_error("SCX_OPS_ENQ_LAST requires ops.enqueue() to be implemented"); + scx_error(sch, "SCX_OPS_ENQ_LAST requires ops.enqueue() to be implemented"); return -EINVAL; } + /* + * SCX_OPS_BUILTIN_IDLE_PER_NODE requires built-in CPU idle + * selection policy to be enabled. + */ + if ((ops->flags & SCX_OPS_BUILTIN_IDLE_PER_NODE) && + (ops->update_idle && !(ops->flags & SCX_OPS_KEEP_BUILTIN_IDLE))) { + scx_error(sch, "SCX_OPS_BUILTIN_IDLE_PER_NODE requires CPU idle selection enabled"); + return -EINVAL; + } + + if (ops->flags & SCX_OPS_HAS_CGROUP_WEIGHT) + pr_warn("SCX_OPS_HAS_CGROUP_WEIGHT is deprecated and a noop\n"); + + if (ops->cpu_acquire || ops->cpu_release) + pr_warn("ops->cpu_acquire/release() are deprecated, use sched_switch TP instead\n"); + return 0; } -static int scx_ops_enable(struct sched_ext_ops *ops, struct bpf_link *link) +static int scx_enable(struct sched_ext_ops *ops, struct bpf_link *link) { + struct scx_sched *sch; struct scx_task_iter sti; struct task_struct *p; unsigned long timeout; - int i, cpu, node, ret; + int i, cpu, ret; if (!cpumask_equal(housekeeping_cpumask(HK_TYPE_DOMAIN), cpu_possible_mask)) { @@ -5404,78 +4870,31 @@ static int scx_ops_enable(struct sched_ext_ops *ops, struct bpf_link *link) return -EINVAL; } - mutex_lock(&scx_ops_enable_mutex); - - if (!scx_ops_helper) { - WRITE_ONCE(scx_ops_helper, - scx_create_rt_helper("sched_ext_ops_helper")); - if (!scx_ops_helper) { - ret = -ENOMEM; - goto err_unlock; - } - } - - if (!global_dsqs) { - struct scx_dispatch_q **dsqs; - - dsqs = kcalloc(nr_node_ids, sizeof(dsqs[0]), GFP_KERNEL); - if (!dsqs) { - ret = -ENOMEM; - goto err_unlock; - } - - for_each_node_state(node, N_POSSIBLE) { - struct scx_dispatch_q *dsq; - - dsq = kzalloc_node(sizeof(*dsq), GFP_KERNEL, node); - if (!dsq) { - for_each_node_state(node, N_POSSIBLE) - kfree(dsqs[node]); - kfree(dsqs); - ret = -ENOMEM; - goto err_unlock; - } - - init_dsq(dsq, SCX_DSQ_GLOBAL); - dsqs[node] = dsq; - } - - global_dsqs = dsqs; - } + mutex_lock(&scx_enable_mutex); - if (scx_ops_enable_state() != SCX_OPS_DISABLED) { + if (scx_enable_state() != SCX_DISABLED) { ret = -EBUSY; goto err_unlock; } - scx_root_kobj = kzalloc(sizeof(*scx_root_kobj), GFP_KERNEL); - if (!scx_root_kobj) { - ret = -ENOMEM; + ret = alloc_kick_syncs(); + if (ret) goto err_unlock; - } - - scx_root_kobj->kset = scx_kset; - ret = kobject_init_and_add(scx_root_kobj, &scx_ktype, NULL, "root"); - if (ret < 0) - goto err; - scx_exit_info = alloc_exit_info(ops->exit_dump_len); - if (!scx_exit_info) { - ret = -ENOMEM; - goto err_del; + sch = scx_alloc_and_add_sched(ops); + if (IS_ERR(sch)) { + ret = PTR_ERR(sch); + goto err_free_ksyncs; } /* - * Set scx_ops, transition to ENABLING and clear exit info to arm the - * disable path. Failure triggers full disabling from here on. + * Transition to ENABLING and clear exit info to arm the disable path. + * Failure triggers full disabling from here on. */ - scx_ops = *ops; - - WARN_ON_ONCE(scx_ops_set_enable_state(SCX_OPS_ENABLING) != - SCX_OPS_DISABLED); - - atomic_set(&scx_exit_kind, SCX_EXIT_NONE); - scx_warned_zero_slice = false; + WARN_ON_ONCE(scx_set_enable_state(SCX_ENABLING) != SCX_DISABLED); + WARN_ON_ONCE(scx_root); + if (WARN_ON_ONCE(READ_ONCE(scx_aborting))) + WRITE_ONCE(scx_aborting, false); atomic_long_set(&scx_nr_rejected, 0); @@ -5488,27 +4907,39 @@ static int scx_ops_enable(struct sched_ext_ops *ops, struct bpf_link *link) */ cpus_read_lock(); - if (scx_ops.init) { - ret = SCX_CALL_OP_RET(SCX_KF_UNLOCKED, init); + /* + * Make the scheduler instance visible. Must be inside cpus_read_lock(). + * See handle_hotplug(). + */ + rcu_assign_pointer(scx_root, sch); + + scx_idle_enable(ops); + + if (sch->ops.init) { + ret = SCX_CALL_OP_RET(sch, SCX_KF_UNLOCKED, init, NULL); if (ret) { - ret = ops_sanitize_err("init", ret); + ret = ops_sanitize_err(sch, "init", ret); cpus_read_unlock(); - scx_ops_error("ops.init() failed (%d)", ret); + 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++) if (((void (**)(void))ops)[i]) - static_branch_enable_cpuslocked(&scx_has_op[i]); + set_bit(i, sch->has_op); + + ret = check_hotplug_seq(sch, ops); + if (ret) { + cpus_read_unlock(); + goto err_disable; + } + scx_idle_update_selcpu_topology(ops); - check_hotplug_seq(ops); -#ifdef CONFIG_SMP - update_selcpu_topology(); -#endif cpus_read_unlock(); - ret = validate_ops(ops); + ret = validate_ops(sch, ops); if (ret) goto err_disable; @@ -5533,31 +4964,19 @@ static int scx_ops_enable(struct sched_ext_ops *ops, struct bpf_link *link) scx_watchdog_timeout / 2); /* - * Once __scx_ops_enabled is set, %current can be switched to SCX - * anytime. This can lead to stalls as some BPF schedulers (e.g. - * userspace scheduling) may not function correctly before all tasks are - * switched. Init in bypass mode to guarantee forward progress. + * Once __scx_enabled is set, %current can be switched to SCX anytime. + * This can lead to stalls as some BPF schedulers (e.g. userspace + * scheduling) may not function correctly before all tasks are switched. + * Init in bypass mode to guarantee forward progress. */ - scx_ops_bypass(true); + scx_bypass(true); for (i = SCX_OPI_NORMAL_BEGIN; i < SCX_OPI_NORMAL_END; i++) if (((void (**)(void))ops)[i]) - static_branch_enable(&scx_has_op[i]); + set_bit(i, sch->has_op); - if (ops->flags & SCX_OPS_ENQ_LAST) - static_branch_enable(&scx_ops_enq_last); - - if (ops->flags & SCX_OPS_ENQ_EXITING) - static_branch_enable(&scx_ops_enq_exiting); - if (scx_ops.cpu_acquire || scx_ops.cpu_release) - static_branch_enable(&scx_ops_cpu_preempt); - - if (!ops->update_idle || (ops->flags & SCX_OPS_KEEP_BUILTIN_IDLE)) { - reset_idle_masks(); - static_branch_enable(&scx_builtin_idle_enabled); - } else { - static_branch_disable(&scx_builtin_idle_enabled); - } + if (sch->ops.cpu_acquire || sch->ops.cpu_release) + sch->ops.flags |= SCX_OPS_HAS_CPU_PREEMPT; /* * Lock out forks, cgroup on/offlining and moves before opening the @@ -5565,8 +4984,8 @@ static int scx_ops_enable(struct sched_ext_ops *ops, struct bpf_link *link) */ percpu_down_write(&scx_fork_rwsem); - WARN_ON_ONCE(scx_ops_init_task_enabled); - scx_ops_init_task_enabled = true; + WARN_ON_ONCE(scx_init_task_enabled); + scx_init_task_enabled = true; /* * Enable ops for every task. Fork is excluded by scx_fork_rwsem @@ -5575,14 +4994,14 @@ static int scx_ops_enable(struct sched_ext_ops *ops, struct bpf_link *link) * tasks. Prep all tasks first and then enable them with preemption * disabled. * - * All cgroups should be initialized before scx_ops_init_task() so that - * the BPF scheduler can reliably track each task's cgroup membership - * from scx_ops_init_task(). Lock out cgroup on/offlining and task - * migrations while tasks are being initialized so that - * scx_cgroup_can_attach() never sees uninitialized tasks. + * All cgroups should be initialized before scx_init_task() so that the + * BPF scheduler can reliably track each task's cgroup membership from + * scx_init_task(). Lock out cgroup on/offlining and task migrations + * while tasks are being initialized so that scx_cgroup_can_attach() + * never sees uninitialized tasks. */ scx_cgroup_lock(); - ret = scx_cgroup_init(); + ret = scx_cgroup_init(sch); if (ret) goto err_disable_unlock_all; @@ -5598,20 +5017,18 @@ static int scx_ops_enable(struct sched_ext_ops *ops, struct bpf_link *link) scx_task_iter_unlock(&sti); - ret = scx_ops_init_task(p, task_group(p), false); + 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_ops_error("ops.init_task() failed (%d) for %s[%d]", - ret, p->comm, p->pid); + scx_error(sch, "ops.init_task() failed (%d) for %s[%d]", + ret, p->comm, p->pid); goto err_disable_unlock_all; } 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(); @@ -5622,7 +5039,7 @@ static int scx_ops_enable(struct sched_ext_ops *ops, struct bpf_link *link) * all eligible tasks. */ WRITE_ONCE(scx_switching_all, !(ops->flags & SCX_OPS_SWITCH_PARTIAL)); - static_branch_enable(&__scx_ops_enabled); + static_branch_enable(&__scx_enabled); /* * We're fully committed and can't fail. The task READY -> ENABLED @@ -5632,31 +5049,28 @@ static int scx_ops_enable(struct sched_ext_ops *ops, struct bpf_link *link) percpu_down_write(&scx_fork_rwsem); scx_task_iter_start(&sti); while ((p = scx_task_iter_next_locked(&sti))) { + unsigned int queue_flags = DEQUEUE_SAVE | DEQUEUE_MOVE; const struct sched_class *old_class = p->sched_class; - const struct sched_class *new_class = - __setscheduler_class(p->policy, p->prio); - struct sched_enq_and_set_ctx ctx; - - if (old_class != new_class && p->se.sched_delayed) - dequeue_task(task_rq(p), p, DEQUEUE_SLEEP | DEQUEUE_DELAYED); - - sched_deq_and_put_task(p, DEQUEUE_SAVE | DEQUEUE_MOVE, &ctx); + const struct sched_class *new_class = scx_setscheduler_class(p); - p->scx.slice = SCX_SLICE_DFL; - p->sched_class = new_class; - check_class_changing(task_rq(p), p, old_class); + if (scx_get_task_state(p) != SCX_TASK_READY) + continue; - sched_enq_and_set_task(&ctx); + if (old_class != new_class) + queue_flags |= DEQUEUE_CLASS; - check_class_changed(task_rq(p), p, old_class, p->prio); + scoped_guard (sched_change, p, queue_flags) { + p->scx.slice = READ_ONCE(scx_slice_dfl); + p->sched_class = new_class; + } } scx_task_iter_stop(&sti); percpu_up_write(&scx_fork_rwsem); - scx_ops_bypass(false); + scx_bypass(false); - if (!scx_ops_tryset_enable_state(SCX_OPS_ENABLED, SCX_OPS_ENABLING)) { - WARN_ON_ONCE(atomic_read(&scx_exit_kind) == SCX_EXIT_NONE); + if (!scx_tryset_enable_state(SCX_ENABLED, SCX_ENABLING)) { + WARN_ON_ONCE(atomic_read(&sch->exit_kind) == SCX_EXIT_NONE); goto err_disable; } @@ -5664,44 +5078,37 @@ static int scx_ops_enable(struct sched_ext_ops *ops, struct bpf_link *link) static_branch_enable(&__scx_switched_all); pr_info("sched_ext: BPF scheduler \"%s\" enabled%s\n", - scx_ops.name, scx_switched_all() ? "" : " (partial)"); - kobject_uevent(scx_root_kobj, KOBJ_ADD); - mutex_unlock(&scx_ops_enable_mutex); + sch->ops.name, scx_switched_all() ? "" : " (partial)"); + kobject_uevent(&sch->kobj, KOBJ_ADD); + mutex_unlock(&scx_enable_mutex); atomic_long_inc(&scx_enable_seq); return 0; -err_del: - kobject_del(scx_root_kobj); -err: - kobject_put(scx_root_kobj); - scx_root_kobj = NULL; - if (scx_exit_info) { - free_exit_info(scx_exit_info); - scx_exit_info = NULL; - } +err_free_ksyncs: + free_kick_syncs(); err_unlock: - mutex_unlock(&scx_ops_enable_mutex); + mutex_unlock(&scx_enable_mutex); return ret; err_disable_unlock_all: scx_cgroup_unlock(); percpu_up_write(&scx_fork_rwsem); - scx_ops_bypass(false); + /* we'll soon enter disable path, keep bypass on */ err_disable: - mutex_unlock(&scx_ops_enable_mutex); + mutex_unlock(&scx_enable_mutex); /* * Returning an error code here would not pass all the error information - * to userspace. Record errno using scx_ops_error() for cases - * scx_ops_error() wasn't already invoked and exit indicating success so - * that the error is notified through ops.exit() with all the details. + * to userspace. Record errno using scx_error() for cases scx_error() + * wasn't already invoked and exit indicating success so that the error + * is notified through ops.exit() with all the details. * - * Flush scx_ops_disable_work to ensure that error is reported before - * init completion. + * Flush scx_disable_work to ensure that error is reported before init + * completion. sch's base reference will be put by bpf_scx_unreg(). */ - scx_ops_error("scx_ops_enable() failed (%d)", ret); - kthread_flush_work(&scx_ops_disable_work); + scx_error(sch, "scx_enable() failed (%d)", ret); + kthread_flush_work(&sch->disable_work); return 0; } @@ -5752,21 +5159,8 @@ static int bpf_scx_btf_struct_access(struct bpf_verifier_log *log, return -EACCES; } -static const struct bpf_func_proto * -bpf_scx_get_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) -{ - switch (func_id) { - case BPF_FUNC_task_storage_get: - return &bpf_task_storage_get_proto; - case BPF_FUNC_task_storage_delete: - return &bpf_task_storage_delete_proto; - default: - return bpf_base_func_proto(func_id, prog); - } -} - static const struct bpf_verifier_ops bpf_scx_verifier_ops = { - .get_func_proto = bpf_scx_get_func_proto, + .get_func_proto = bpf_base_func_proto, .is_valid_access = bpf_scx_is_valid_access, .btf_struct_access = bpf_scx_btf_struct_access, }; @@ -5845,13 +5239,17 @@ static int bpf_scx_check_member(const struct btf_type *t, static int bpf_scx_reg(void *kdata, struct bpf_link *link) { - return scx_ops_enable(kdata, link); + return scx_enable(kdata, link); } static void bpf_scx_unreg(void *kdata, struct bpf_link *link) { - scx_ops_disable(SCX_EXIT_UNREG); - kthread_flush_work(&scx_ops_disable_work); + struct sched_ext_ops *ops = kdata; + struct scx_sched *sch = ops->priv; + + scx_disable(SCX_EXIT_UNREG); + kthread_flush_work(&sch->disable_work); + kobject_put(&sch->kobj); } static int bpf_scx_init(struct btf *btf) @@ -5905,6 +5303,8 @@ static s32 sched_ext_ops__cgroup_prep_move(struct task_struct *p, struct cgroup static void sched_ext_ops__cgroup_move(struct task_struct *p, struct cgroup *from, struct cgroup *to) {} static void sched_ext_ops__cgroup_cancel_move(struct task_struct *p, struct cgroup *from, struct cgroup *to) {} static void sched_ext_ops__cgroup_set_weight(struct cgroup *cgrp, u32 weight) {} +static void sched_ext_ops__cgroup_set_bandwidth(struct cgroup *cgrp, u64 period_us, u64 quota_us, u64 burst_us) {} +static void sched_ext_ops__cgroup_set_idle(struct cgroup *cgrp, bool idle) {} #endif static void sched_ext_ops__cpu_online(s32 cpu) {} static void sched_ext_ops__cpu_offline(s32 cpu) {} @@ -5942,6 +5342,8 @@ static struct sched_ext_ops __bpf_ops_sched_ext_ops = { .cgroup_move = sched_ext_ops__cgroup_move, .cgroup_cancel_move = sched_ext_ops__cgroup_cancel_move, .cgroup_set_weight = sched_ext_ops__cgroup_set_weight, + .cgroup_set_bandwidth = sched_ext_ops__cgroup_set_bandwidth, + .cgroup_set_idle = sched_ext_ops__cgroup_set_idle, #endif .cpu_online = sched_ext_ops__cpu_online, .cpu_offline = sched_ext_ops__cpu_offline, @@ -5973,10 +5375,7 @@ static struct bpf_struct_ops bpf_sched_ext_ops = { static void sysrq_handle_sched_ext_reset(u8 key) { - if (scx_ops_helper) - scx_ops_disable(SCX_EXIT_SYSRQ); - else - pr_info("sched_ext: BPF scheduler not yet used\n"); + scx_disable(SCX_EXIT_SYSRQ); } static const struct sysrq_key_op sysrq_sched_ext_reset_op = { @@ -6018,29 +5417,38 @@ static bool can_skip_idle_kick(struct rq *rq) return !is_idle_task(rq->curr) && !(rq->scx.flags & SCX_RQ_IN_BALANCE); } -static bool kick_one_cpu(s32 cpu, struct rq *this_rq, unsigned long *pseqs) +static bool kick_one_cpu(s32 cpu, struct rq *this_rq, unsigned long *ksyncs) { struct rq *rq = cpu_rq(cpu); struct scx_rq *this_scx = &this_rq->scx; + const struct sched_class *cur_class; bool should_wait = false; unsigned long flags; raw_spin_rq_lock_irqsave(rq, flags); + cur_class = rq->curr->sched_class; /* * During CPU hotplug, a CPU may depend on kicking itself to make - * forward progress. Allow kicking self regardless of online state. + * forward progress. Allow kicking self regardless of online state. If + * @cpu is running a higher class task, we have no control over @cpu. + * Skip kicking. */ - if (cpu_online(cpu) || cpu == cpu_of(this_rq)) { + if ((cpu_online(cpu) || cpu == cpu_of(this_rq)) && + !sched_class_above(cur_class, &ext_sched_class)) { if (cpumask_test_cpu(cpu, this_scx->cpus_to_preempt)) { - if (rq->curr->sched_class == &ext_sched_class) + if (cur_class == &ext_sched_class) rq->curr->scx.slice = 0; cpumask_clear_cpu(cpu, this_scx->cpus_to_preempt); } if (cpumask_test_cpu(cpu, this_scx->cpus_to_wait)) { - pseqs[cpu] = rq->scx.pnt_seq; - should_wait = true; + if (cur_class == &ext_sched_class) { + ksyncs[cpu] = rq->scx.kick_sync; + should_wait = true; + } else { + cpumask_clear_cpu(cpu, this_scx->cpus_to_wait); + } } resched_curr(rq); @@ -6072,12 +5480,20 @@ static void kick_cpus_irq_workfn(struct irq_work *irq_work) { struct rq *this_rq = this_rq(); struct scx_rq *this_scx = &this_rq->scx; - unsigned long *pseqs = this_cpu_ptr(scx_kick_cpus_pnt_seqs); + struct scx_kick_syncs __rcu *ksyncs_pcpu = __this_cpu_read(scx_kick_syncs); bool should_wait = false; + unsigned long *ksyncs; s32 cpu; + if (unlikely(!ksyncs_pcpu)) { + pr_warn_once("kick_cpus_irq_workfn() called with NULL scx_kick_syncs"); + return; + } + + ksyncs = rcu_dereference_bh(ksyncs_pcpu)->syncs; + for_each_cpu(cpu, this_scx->cpus_to_kick) { - should_wait |= kick_one_cpu(cpu, this_rq, pseqs); + should_wait |= kick_one_cpu(cpu, this_rq, ksyncs); cpumask_clear_cpu(cpu, this_scx->cpus_to_kick); cpumask_clear_cpu(cpu, this_scx->cpus_to_kick_if_idle); } @@ -6091,20 +5507,21 @@ static void kick_cpus_irq_workfn(struct irq_work *irq_work) return; for_each_cpu(cpu, this_scx->cpus_to_wait) { - unsigned long *wait_pnt_seq = &cpu_rq(cpu)->scx.pnt_seq; + unsigned long *wait_kick_sync = &cpu_rq(cpu)->scx.kick_sync; - if (cpu != cpu_of(this_rq)) { - /* - * Pairs with smp_store_release() issued by this CPU in - * switch_class() on the resched path. - * - * We busy-wait here to guarantee that no other task can - * be scheduled on our core before the target CPU has - * entered the resched path. - */ - while (smp_load_acquire(wait_pnt_seq) == pseqs[cpu]) - cpu_relax(); - } + /* + * Busy-wait until the task running at the time of kicking is no + * longer running. This can be used to implement e.g. core + * scheduling. + * + * smp_cond_load_acquire() pairs with store_releases in + * pick_task_scx() and put_prev_task_scx(). The former breaks + * the wait if SCX's scheduling path is entered even if the same + * task is picked subsequently. The latter is necessary to break + * the wait when $cpu is taken by a higher sched class. + */ + if (cpu != cpu_of(this_rq)) + smp_cond_load_acquire(wait_kick_sync, VAL != ksyncs[cpu]); cpumask_clear_cpu(cpu, this_scx->cpus_to_wait); } @@ -6124,13 +5541,14 @@ static void kick_cpus_irq_workfn(struct irq_work *irq_work) */ void print_scx_info(const char *log_lvl, struct task_struct *p) { - enum scx_ops_enable_state state = scx_ops_enable_state(); + struct scx_sched *sch = scx_root; + enum scx_enable_state state = scx_enable_state(); const char *all = READ_ONCE(scx_switching_all) ? "+all" : ""; char runnable_at_buf[22] = "?"; struct sched_class *class; unsigned long runnable_at; - if (state == SCX_OPS_DISABLED) + if (state == SCX_DISABLED) return; /* @@ -6139,8 +5557,8 @@ void print_scx_info(const char *log_lvl, struct task_struct *p) */ if (copy_from_kernel_nofault(&class, &p->sched_class, sizeof(class)) || class != &ext_sched_class) { - printk("%sSched_ext: %s (%s%s)", log_lvl, scx_ops.name, - scx_ops_enable_state_str[state], all); + printk("%sSched_ext: %s (%s%s)", log_lvl, sch->ops.name, + scx_enable_state_str[state], all); return; } @@ -6151,7 +5569,7 @@ void print_scx_info(const char *log_lvl, struct task_struct *p) /* print everything onto one line to conserve console space */ printk("%sSched_ext: %s (%s%s), task: runnable_at=%s", - log_lvl, scx_ops.name, scx_ops_enable_state_str[state], all, + log_lvl, sch->ops.name, scx_enable_state_str[state], all, runnable_at_buf); } @@ -6167,12 +5585,12 @@ static int scx_pm_handler(struct notifier_block *nb, unsigned long event, void * case PM_HIBERNATION_PREPARE: case PM_SUSPEND_PREPARE: case PM_RESTORE_PREPARE: - scx_ops_bypass(true); + scx_bypass(true); break; case PM_POST_HIBERNATION: case PM_POST_SUSPEND: case PM_POST_RESTORE: - scx_ops_bypass(false); + scx_bypass(false); break; } @@ -6195,29 +5613,23 @@ void __init init_sched_ext_class(void) WRITE_ONCE(v, SCX_ENQ_WAKEUP | SCX_DEQ_SLEEP | SCX_KICK_PREEMPT | SCX_TG_ONLINE); - BUG_ON(rhashtable_init(&dsq_hash, &dsq_hash_params)); -#ifdef CONFIG_SMP - BUG_ON(!alloc_cpumask_var(&idle_masks.cpu, GFP_KERNEL)); - BUG_ON(!alloc_cpumask_var(&idle_masks.smt, GFP_KERNEL)); -#endif - scx_kick_cpus_pnt_seqs = - __alloc_percpu(sizeof(scx_kick_cpus_pnt_seqs[0]) * nr_cpu_ids, - __alignof__(scx_kick_cpus_pnt_seqs[0])); - BUG_ON(!scx_kick_cpus_pnt_seqs); + scx_idle_init_masks(); for_each_possible_cpu(cpu) { struct rq *rq = cpu_rq(cpu); + int n = cpu_to_node(cpu); init_dsq(&rq->scx.local_dsq, SCX_DSQ_LOCAL); + init_dsq(&rq->scx.bypass_dsq, SCX_DSQ_BYPASS); INIT_LIST_HEAD(&rq->scx.runnable_list); INIT_LIST_HEAD(&rq->scx.ddsp_deferred_locals); - BUG_ON(!zalloc_cpumask_var(&rq->scx.cpus_to_kick, GFP_KERNEL)); - BUG_ON(!zalloc_cpumask_var(&rq->scx.cpus_to_kick_if_idle, GFP_KERNEL)); - BUG_ON(!zalloc_cpumask_var(&rq->scx.cpus_to_preempt, GFP_KERNEL)); - BUG_ON(!zalloc_cpumask_var(&rq->scx.cpus_to_wait, GFP_KERNEL)); - init_irq_work(&rq->scx.deferred_irq_work, deferred_irq_workfn); - init_irq_work(&rq->scx.kick_cpus_irq_work, kick_cpus_irq_workfn); + BUG_ON(!zalloc_cpumask_var_node(&rq->scx.cpus_to_kick, GFP_KERNEL, n)); + BUG_ON(!zalloc_cpumask_var_node(&rq->scx.cpus_to_kick_if_idle, GFP_KERNEL, n)); + BUG_ON(!zalloc_cpumask_var_node(&rq->scx.cpus_to_preempt, GFP_KERNEL, n)); + BUG_ON(!zalloc_cpumask_var_node(&rq->scx.cpus_to_wait, GFP_KERNEL, n)); + rq->scx.deferred_irq_work = IRQ_WORK_INIT_HARD(deferred_irq_workfn); + rq->scx.kick_cpus_irq_work = IRQ_WORK_INIT_HARD(kick_cpus_irq_workfn); if (cpu_online(cpu)) cpu_rq(cpu)->scx.flags |= SCX_RQ_ONLINE; @@ -6232,89 +5644,41 @@ void __init init_sched_ext_class(void) /******************************************************************************** * Helpers that can be called from the BPF scheduler. */ -#include <linux/btf_ids.h> - -__bpf_kfunc_start_defs(); - -/** - * scx_bpf_select_cpu_dfl - The default implementation of ops.select_cpu() - * @p: task_struct to select a CPU for - * @prev_cpu: CPU @p was on previously - * @wake_flags: %SCX_WAKE_* flags - * @is_idle: out parameter indicating whether the returned CPU is idle - * - * Can only be called from ops.select_cpu() if the built-in CPU selection is - * enabled - ops.update_idle() is missing or %SCX_OPS_KEEP_BUILTIN_IDLE is set. - * @p, @prev_cpu and @wake_flags match ops.select_cpu(). - * - * Returns the picked CPU with *@is_idle indicating whether the picked CPU is - * currently idle and thus a good candidate for direct dispatching. - */ -__bpf_kfunc s32 scx_bpf_select_cpu_dfl(struct task_struct *p, s32 prev_cpu, - u64 wake_flags, bool *is_idle) -{ - if (!static_branch_likely(&scx_builtin_idle_enabled)) { - scx_ops_error("built-in idle tracking is disabled"); - goto prev_cpu; - } - - if (!scx_kf_allowed(SCX_KF_SELECT_CPU)) - goto prev_cpu; - -#ifdef CONFIG_SMP - return scx_select_cpu_dfl(p, prev_cpu, wake_flags, is_idle); -#endif - -prev_cpu: - *is_idle = false; - return prev_cpu; -} - -__bpf_kfunc_end_defs(); - -BTF_KFUNCS_START(scx_kfunc_ids_select_cpu) -BTF_ID_FLAGS(func, scx_bpf_select_cpu_dfl, KF_RCU) -BTF_KFUNCS_END(scx_kfunc_ids_select_cpu) - -static const struct btf_kfunc_id_set scx_kfunc_set_select_cpu = { - .owner = THIS_MODULE, - .set = &scx_kfunc_ids_select_cpu, -}; - -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_ops_error("called with NULL task"); + scx_error(sch, "called with NULL task"); return false; } if (unlikely(enq_flags & __SCX_ENQ_INTERNAL_MASK)) { - scx_ops_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_ops_error("dispatch buffer overflow"); + scx_error(sch, "dispatch buffer overflow"); return; } @@ -6340,9 +5704,7 @@ __bpf_kfunc_start_defs(); * ops.select_cpu(), and ops.dispatch(). * * When called from ops.select_cpu() or ops.enqueue(), it's for direct dispatch - * and @p must match the task being enqueued. Also, %SCX_DSQ_LOCAL_ON can't be - * used to target the local DSQ of a CPU other than the enqueueing one. Use - * ops.select_cpu() to be on the target CPU in the first place. + * and @p must match the task being enqueued. * * When called from ops.select_cpu(), @enq_flags and @dsp_id are stored, and @p * will be directly inserted into the corresponding dispatch queue after @@ -6354,7 +5716,8 @@ __bpf_kfunc_start_defs(); * When called from ops.dispatch(), there are no restrictions on @p or @dsq_id * and this function can be called upto ops.dispatch_max_batch times to insert * multiple tasks. scx_bpf_dispatch_nr_slots() returns the number of the - * remaining slots. scx_bpf_consume() flushes the batch and resets the counter. + * remaining slots. scx_bpf_dsq_move_to_local() flushes the batch and resets the + * counter. * * This function doesn't have any locking restrictions and may be called under * BPF locks (in the future when BPF introduces more flexible locking). @@ -6363,81 +5726,139 @@ __bpf_kfunc_start_defs(); * exhaustion. If zero, the current residual slice is maintained. If * %SCX_SLICE_INF, @p never expires and the BPF scheduler must kick the CPU with * scx_bpf_kick_cpu() to trigger scheduling. + * + * Returns %true on successful insertion, %false on failure. On the root + * scheduler, %false return triggers scheduler abort and the caller doesn't need + * to check the return value. */ -__bpf_kfunc void scx_bpf_dsq_insert(struct task_struct *p, u64 dsq_id, u64 slice, - u64 enq_flags) +__bpf_kfunc bool scx_bpf_dsq_insert___v2(struct task_struct *p, u64 dsq_id, + u64 slice, u64 enq_flags) { - if (!scx_dsq_insert_preamble(p, enq_flags)) - return; + struct scx_sched *sch; + + guard(rcu)(); + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return false; + + if (!scx_dsq_insert_preamble(sch, p, enq_flags)) + return false; if (slice) p->scx.slice = 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); + + return true; } -/* for backward compatibility, will be removed in v6.15 */ -__bpf_kfunc void scx_bpf_dispatch(struct task_struct *p, u64 dsq_id, u64 slice, - u64 enq_flags) +/* + * COMPAT: Will be removed in v6.23 along with the ___v2 suffix. + */ +__bpf_kfunc void scx_bpf_dsq_insert(struct task_struct *p, u64 dsq_id, + u64 slice, u64 enq_flags) +{ + scx_bpf_dsq_insert___v2(p, dsq_id, slice, enq_flags); +} + +static bool scx_dsq_insert_vtime(struct scx_sched *sch, struct task_struct *p, + u64 dsq_id, u64 slice, u64 vtime, u64 enq_flags) { - printk_deferred_once(KERN_WARNING "sched_ext: scx_bpf_dispatch() renamed to scx_bpf_dsq_insert()"); - scx_bpf_dsq_insert(p, dsq_id, slice, enq_flags); + if (!scx_dsq_insert_preamble(sch, p, enq_flags)) + return false; + + if (slice) + p->scx.slice = slice; + else + p->scx.slice = p->scx.slice ?: 1; + + p->scx.dsq_vtime = vtime; + + scx_dsq_insert_commit(sch, p, dsq_id, enq_flags | SCX_ENQ_DSQ_PRIQ); + + return true; } +struct scx_bpf_dsq_insert_vtime_args { + /* @p can't be packed together as KF_RCU is not transitive */ + u64 dsq_id; + u64 slice; + u64 vtime; + u64 enq_flags; +}; + /** - * scx_bpf_dsq_insert_vtime - Insert a task into the vtime priority queue of a DSQ + * __scx_bpf_dsq_insert_vtime - Arg-wrapped vtime DSQ insertion * @p: task_struct to insert - * @dsq_id: DSQ to insert into - * @slice: duration @p can run for in nsecs, 0 to keep the current value - * @vtime: @p's ordering inside the vtime-sorted queue of the target DSQ - * @enq_flags: SCX_ENQ_* + * @args: struct containing the rest of the arguments + * @args->dsq_id: DSQ to insert into + * @args->slice: duration @p can run for in nsecs, 0 to keep the current value + * @args->vtime: @p's ordering inside the vtime-sorted queue of the target DSQ + * @args->enq_flags: SCX_ENQ_* + * + * Wrapper kfunc that takes arguments via struct to work around BPF's 5 argument + * limit. BPF programs should use scx_bpf_dsq_insert_vtime() which is provided + * as an inline wrapper in common.bpf.h. * - * Insert @p into the vtime priority queue of the DSQ identified by @dsq_id. - * Tasks queued into the priority queue are ordered by @vtime. All other aspects - * are identical to scx_bpf_dsq_insert(). + * Insert @p into the vtime priority queue of the DSQ identified by + * @args->dsq_id. Tasks queued into the priority queue are ordered by + * @args->vtime. All other aspects are identical to scx_bpf_dsq_insert(). * - * @vtime ordering is according to time_before64() which considers wrapping. A - * numerically larger vtime may indicate an earlier position in the ordering and - * vice-versa. + * @args->vtime ordering is according to time_before64() which considers + * wrapping. A numerically larger vtime may indicate an earlier position in the + * ordering and vice-versa. * * A DSQ can only be used as a FIFO or priority queue at any given time and this * function must not be called on a DSQ which already has one or more FIFO tasks * queued and vice-versa. Also, the built-in DSQs (SCX_DSQ_LOCAL and * SCX_DSQ_GLOBAL) cannot be used as priority queues. + * + * Returns %true on successful insertion, %false on failure. On the root + * scheduler, %false return triggers scheduler abort and the caller doesn't need + * to check the return value. */ -__bpf_kfunc void scx_bpf_dsq_insert_vtime(struct task_struct *p, u64 dsq_id, - u64 slice, u64 vtime, u64 enq_flags) +__bpf_kfunc bool +__scx_bpf_dsq_insert_vtime(struct task_struct *p, + struct scx_bpf_dsq_insert_vtime_args *args) { - if (!scx_dsq_insert_preamble(p, enq_flags)) - return; + struct scx_sched *sch; - if (slice) - p->scx.slice = slice; - else - p->scx.slice = p->scx.slice ?: 1; + guard(rcu)(); - p->scx.dsq_vtime = vtime; + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return false; - scx_dsq_insert_commit(p, dsq_id, enq_flags | SCX_ENQ_DSQ_PRIQ); + return scx_dsq_insert_vtime(sch, p, args->dsq_id, args->slice, + args->vtime, args->enq_flags); } -/* for backward compatibility, will be removed in v6.15 */ -__bpf_kfunc void scx_bpf_dispatch_vtime(struct task_struct *p, u64 dsq_id, - u64 slice, u64 vtime, u64 enq_flags) +/* + * COMPAT: Will be removed in v6.23. + */ +__bpf_kfunc void scx_bpf_dsq_insert_vtime(struct task_struct *p, u64 dsq_id, + u64 slice, u64 vtime, u64 enq_flags) { - printk_deferred_once(KERN_WARNING "sched_ext: scx_bpf_dispatch_vtime() renamed to scx_bpf_dsq_insert_vtime()"); - scx_bpf_dsq_insert_vtime(p, dsq_id, slice, vtime, enq_flags); + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return; + + scx_dsq_insert_vtime(sch, p, dsq_id, slice, vtime, enq_flags); } __bpf_kfunc_end_defs(); BTF_KFUNCS_START(scx_kfunc_ids_enqueue_dispatch) BTF_ID_FLAGS(func, scx_bpf_dsq_insert, KF_RCU) +BTF_ID_FLAGS(func, scx_bpf_dsq_insert___v2, KF_RCU) +BTF_ID_FLAGS(func, __scx_bpf_dsq_insert_vtime, KF_RCU) BTF_ID_FLAGS(func, scx_bpf_dsq_insert_vtime, KF_RCU) -BTF_ID_FLAGS(func, scx_bpf_dispatch, KF_RCU) -BTF_ID_FLAGS(func, scx_bpf_dispatch_vtime, KF_RCU) BTF_KFUNCS_END(scx_kfunc_ids_enqueue_dispatch) static const struct btf_kfunc_id_set scx_kfunc_set_enqueue_dispatch = { @@ -6448,13 +5869,22 @@ static const struct btf_kfunc_id_set scx_kfunc_set_enqueue_dispatch = { static bool scx_dsq_move(struct bpf_iter_scx_dsq_kern *kit, struct task_struct *p, u64 dsq_id, u64 enq_flags) { + struct scx_sched *sch = scx_root; struct scx_dispatch_q *src_dsq = kit->dsq, *dst_dsq; struct rq *this_rq, *src_rq, *locked_rq; bool dispatched = false; 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; + + /* + * If the BPF scheduler keeps calling this function repeatedly, it can + * cause similar live-lock conditions as consume_dispatch_q(). + */ + if (unlikely(READ_ONCE(scx_aborting))) return false; /* @@ -6477,13 +5907,6 @@ static bool scx_dsq_move(struct bpf_iter_scx_dsq_kern *kit, raw_spin_rq_lock(src_rq); } - /* - * If the BPF scheduler keeps calling this function repeatedly, it can - * cause similar live-lock conditions as consume_dispatch_q(). Insert a - * breather if necessary. - */ - scx_ops_breather(src_rq); - locked_rq = src_rq; raw_spin_lock(&src_dsq->lock); @@ -6500,7 +5923,7 @@ static bool scx_dsq_move(struct bpf_iter_scx_dsq_kern *kit, } /* @p is still on $src_dsq and stable, determine the destination */ - dst_dsq = find_dsq_for_dispatch(this_rq, dsq_id, p); + dst_dsq = find_dsq_for_dispatch(sch, this_rq, dsq_id, p); /* * Apply vtime and slice updates before moving so that the new time is @@ -6513,7 +5936,7 @@ static bool scx_dsq_move(struct bpf_iter_scx_dsq_kern *kit, p->scx.slice = kit->slice; /* execute move */ - locked_rq = move_task_between_dsqs(p, enq_flags, src_dsq, dst_dsq); + locked_rq = move_task_between_dsqs(sch, p, enq_flags, src_dsq, dst_dsq); dispatched = true; out: if (in_balance) { @@ -6539,7 +5962,15 @@ __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); @@ -6554,14 +5985,21 @@ __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; - if (!scx_kf_allowed(SCX_KF_DISPATCH)) + guard(rcu)(); + + 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_ops_error("dispatch buffer underflow"); + scx_error(sch, "dispatch buffer underflow"); } /** @@ -6582,19 +6020,26 @@ __bpf_kfunc bool scx_bpf_dsq_move_to_local(u64 dsq_id) { struct scx_dsp_ctx *dspc = this_cpu_ptr(scx_dsp_ctx); struct scx_dispatch_q *dsq; + struct scx_sched *sch; + + guard(rcu)(); + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return false; - if (!scx_kf_allowed(SCX_KF_DISPATCH)) + if (!scx_kf_allowed(sch, SCX_KF_DISPATCH)) return false; - flush_dispatch_buf(dspc->rq); + flush_dispatch_buf(sch, dspc->rq); - dsq = find_user_dsq(dsq_id); + dsq = find_user_dsq(sch, dsq_id); if (unlikely(!dsq)) { - scx_ops_error("invalid DSQ ID 0x%016llx", dsq_id); + scx_error(sch, "invalid DSQ ID 0x%016llx", dsq_id); return false; } - if (consume_dispatch_q(dspc->rq, dsq)) { + if (consume_dispatch_q(sch, dspc->rq, dsq)) { /* * A successfully consumed task can be dequeued before it starts * running while the CPU is trying to migrate other dispatched @@ -6608,13 +6053,6 @@ __bpf_kfunc bool scx_bpf_dsq_move_to_local(u64 dsq_id) } } -/* for backward compatibility, will be removed in v6.15 */ -__bpf_kfunc bool scx_bpf_consume(u64 dsq_id) -{ - printk_deferred_once(KERN_WARNING "sched_ext: scx_bpf_consume() renamed to scx_bpf_dsq_move_to_local()"); - return scx_bpf_dsq_move_to_local(dsq_id); -} - /** * scx_bpf_dsq_move_set_slice - Override slice when moving between DSQs * @it__iter: DSQ iterator in progress @@ -6633,14 +6071,6 @@ __bpf_kfunc void scx_bpf_dsq_move_set_slice(struct bpf_iter_scx_dsq *it__iter, kit->cursor.flags |= __SCX_DSQ_ITER_HAS_SLICE; } -/* for backward compatibility, will be removed in v6.15 */ -__bpf_kfunc void scx_bpf_dispatch_from_dsq_set_slice( - struct bpf_iter_scx_dsq *it__iter, u64 slice) -{ - printk_deferred_once(KERN_WARNING "sched_ext: scx_bpf_dispatch_from_dsq_set_slice() renamed to scx_bpf_dsq_move_set_slice()"); - scx_bpf_dsq_move_set_slice(it__iter, slice); -} - /** * scx_bpf_dsq_move_set_vtime - Override vtime when moving between DSQs * @it__iter: DSQ iterator in progress @@ -6660,14 +6090,6 @@ __bpf_kfunc void scx_bpf_dsq_move_set_vtime(struct bpf_iter_scx_dsq *it__iter, kit->cursor.flags |= __SCX_DSQ_ITER_HAS_VTIME; } -/* for backward compatibility, will be removed in v6.15 */ -__bpf_kfunc void scx_bpf_dispatch_from_dsq_set_vtime( - struct bpf_iter_scx_dsq *it__iter, u64 vtime) -{ - printk_deferred_once(KERN_WARNING "sched_ext: scx_bpf_dispatch_from_dsq_set_vtime() renamed to scx_bpf_dsq_move_set_vtime()"); - scx_bpf_dsq_move_set_vtime(it__iter, vtime); -} - /** * scx_bpf_dsq_move - Move a task from DSQ iteration to a DSQ * @it__iter: DSQ iterator in progress @@ -6689,8 +6111,9 @@ __bpf_kfunc void scx_bpf_dispatch_from_dsq_set_vtime( * Can be called from ops.dispatch() or any BPF context which doesn't hold a rq * lock (e.g. BPF timers or SYSCALL programs). * - * Returns %true if @p has been consumed, %false if @p had already been consumed - * or dequeued. + * Returns %true if @p has been consumed, %false if @p had already been + * consumed, dequeued, or, for sub-scheds, @dsq_id points to a disallowed local + * DSQ. */ __bpf_kfunc bool scx_bpf_dsq_move(struct bpf_iter_scx_dsq *it__iter, struct task_struct *p, u64 dsq_id, @@ -6700,15 +6123,6 @@ __bpf_kfunc bool scx_bpf_dsq_move(struct bpf_iter_scx_dsq *it__iter, p, dsq_id, enq_flags); } -/* for backward compatibility, will be removed in v6.15 */ -__bpf_kfunc bool scx_bpf_dispatch_from_dsq(struct bpf_iter_scx_dsq *it__iter, - struct task_struct *p, u64 dsq_id, - u64 enq_flags) -{ - printk_deferred_once(KERN_WARNING "sched_ext: scx_bpf_dispatch_from_dsq() renamed to scx_bpf_dsq_move()"); - return scx_bpf_dsq_move(it__iter, p, dsq_id, enq_flags); -} - /** * scx_bpf_dsq_move_vtime - Move a task from DSQ iteration to a PRIQ DSQ * @it__iter: DSQ iterator in progress @@ -6734,30 +6148,16 @@ __bpf_kfunc bool scx_bpf_dsq_move_vtime(struct bpf_iter_scx_dsq *it__iter, p, dsq_id, enq_flags | SCX_ENQ_DSQ_PRIQ); } -/* for backward compatibility, will be removed in v6.15 */ -__bpf_kfunc bool scx_bpf_dispatch_vtime_from_dsq(struct bpf_iter_scx_dsq *it__iter, - struct task_struct *p, u64 dsq_id, - u64 enq_flags) -{ - printk_deferred_once(KERN_WARNING "sched_ext: scx_bpf_dispatch_from_dsq_vtime() renamed to scx_bpf_dsq_move_vtime()"); - return scx_bpf_dsq_move_vtime(it__iter, p, dsq_id, enq_flags); -} - __bpf_kfunc_end_defs(); BTF_KFUNCS_START(scx_kfunc_ids_dispatch) BTF_ID_FLAGS(func, scx_bpf_dispatch_nr_slots) BTF_ID_FLAGS(func, scx_bpf_dispatch_cancel) BTF_ID_FLAGS(func, scx_bpf_dsq_move_to_local) -BTF_ID_FLAGS(func, scx_bpf_consume) -BTF_ID_FLAGS(func, scx_bpf_dsq_move_set_slice) -BTF_ID_FLAGS(func, scx_bpf_dsq_move_set_vtime) +BTF_ID_FLAGS(func, scx_bpf_dsq_move_set_slice, KF_RCU) +BTF_ID_FLAGS(func, scx_bpf_dsq_move_set_vtime, KF_RCU) BTF_ID_FLAGS(func, scx_bpf_dsq_move, KF_RCU) BTF_ID_FLAGS(func, scx_bpf_dsq_move_vtime, KF_RCU) -BTF_ID_FLAGS(func, scx_bpf_dispatch_from_dsq_set_slice) -BTF_ID_FLAGS(func, scx_bpf_dispatch_from_dsq_set_vtime) -BTF_ID_FLAGS(func, scx_bpf_dispatch_from_dsq, KF_RCU) -BTF_ID_FLAGS(func, scx_bpf_dispatch_vtime_from_dsq, KF_RCU) BTF_KFUNCS_END(scx_kfunc_ids_dispatch) static const struct btf_kfunc_id_set scx_kfunc_set_dispatch = { @@ -6765,26 +6165,12 @@ static const struct btf_kfunc_id_set scx_kfunc_set_dispatch = { .set = &scx_kfunc_ids_dispatch, }; -__bpf_kfunc_start_defs(); - -/** - * scx_bpf_reenqueue_local - Re-enqueue tasks on a local DSQ - * - * Iterate over all of the tasks currently enqueued on the local DSQ of the - * caller's CPU, and re-enqueue them in the BPF scheduler. Returns the number of - * processed tasks. Can only be called from ops.cpu_release(). - */ -__bpf_kfunc u32 scx_bpf_reenqueue_local(void) +static u32 reenq_local(struct rq *rq) { LIST_HEAD(tasks); u32 nr_enqueued = 0; - struct rq *rq; struct task_struct *p, *n; - if (!scx_kf_allowed(SCX_KF_CPU_RELEASE)) - return 0; - - rq = cpu_rq(smp_processor_id()); lockdep_assert_rq_held(rq); /* @@ -6821,6 +6207,37 @@ __bpf_kfunc u32 scx_bpf_reenqueue_local(void) return nr_enqueued; } +__bpf_kfunc_start_defs(); + +/** + * scx_bpf_reenqueue_local - Re-enqueue tasks on a local DSQ + * + * Iterate over all of the tasks currently enqueued on the local DSQ of the + * caller's CPU, and re-enqueue them in the BPF scheduler. Returns the number of + * processed tasks. Can only be called from ops.cpu_release(). + * + * COMPAT: Will be removed in v6.23 along with the ___v2 suffix on the void + * returning variant that can be called from anywhere. + */ +__bpf_kfunc u32 scx_bpf_reenqueue_local(void) +{ + struct scx_sched *sch; + struct rq *rq; + + 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()); + lockdep_assert_rq_held(rq); + + return reenq_local(rq); +} + __bpf_kfunc_end_defs(); BTF_KFUNCS_START(scx_kfunc_ids_cpu_release) @@ -6844,24 +6261,46 @@ __bpf_kfunc_start_defs(); */ __bpf_kfunc s32 scx_bpf_create_dsq(u64 dsq_id, s32 node) { + struct scx_dispatch_q *dsq; + struct scx_sched *sch; + s32 ret; + if (unlikely(node >= (int)nr_node_ids || (node < 0 && node != NUMA_NO_NODE))) return -EINVAL; - return PTR_ERR_OR_ZERO(create_dsq(dsq_id, node)); + + if (unlikely(dsq_id & SCX_DSQ_FLAG_BUILTIN)) + return -EINVAL; + + dsq = kmalloc_node(sizeof(*dsq), GFP_KERNEL, node); + if (!dsq) + return -ENOMEM; + + init_dsq(dsq, dsq_id); + + rcu_read_lock(); + + sch = rcu_dereference(scx_root); + if (sch) + ret = rhashtable_lookup_insert_fast(&sch->dsq_hash, &dsq->hash_node, + dsq_hash_params); + else + ret = -ENODEV; + + rcu_read_unlock(); + if (ret) + kfree(dsq); + return ret; } __bpf_kfunc_end_defs(); BTF_KFUNCS_START(scx_kfunc_ids_unlocked) BTF_ID_FLAGS(func, scx_bpf_create_dsq, KF_SLEEPABLE) -BTF_ID_FLAGS(func, scx_bpf_dsq_move_set_slice) -BTF_ID_FLAGS(func, scx_bpf_dsq_move_set_vtime) +BTF_ID_FLAGS(func, scx_bpf_dsq_move_set_slice, KF_RCU) +BTF_ID_FLAGS(func, scx_bpf_dsq_move_set_vtime, KF_RCU) BTF_ID_FLAGS(func, scx_bpf_dsq_move, KF_RCU) BTF_ID_FLAGS(func, scx_bpf_dsq_move_vtime, KF_RCU) -BTF_ID_FLAGS(func, scx_bpf_dispatch_from_dsq_set_slice) -BTF_ID_FLAGS(func, scx_bpf_dispatch_from_dsq_set_vtime) -BTF_ID_FLAGS(func, scx_bpf_dispatch_from_dsq, KF_RCU) -BTF_ID_FLAGS(func, scx_bpf_dispatch_vtime_from_dsq, KF_RCU) BTF_KFUNCS_END(scx_kfunc_ids_unlocked) static const struct btf_kfunc_id_set scx_kfunc_set_unlocked = { @@ -6872,21 +6311,39 @@ 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 + * scx_bpf_task_set_slice - Set task's time slice + * @p: task of interest + * @slice: time slice to set in nsecs * - * 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. + * Set @p's time slice to @slice. Returns %true on success, %false if the + * calling scheduler doesn't have authority over @p. */ -__bpf_kfunc void scx_bpf_kick_cpu(s32 cpu, u64 flags) +__bpf_kfunc bool scx_bpf_task_set_slice(struct task_struct *p, u64 slice) +{ + p->scx.slice = slice; + return true; +} + +/** + * scx_bpf_task_set_dsq_vtime - Set task's virtual time for DSQ ordering + * @p: task of interest + * @vtime: virtual time to set + * + * Set @p's virtual time to @vtime. Returns %true on success, %false if the + * calling scheduler doesn't have authority over @p. + */ +__bpf_kfunc bool scx_bpf_task_set_dsq_vtime(struct task_struct *p, u64 vtime) +{ + p->scx.dsq_vtime = vtime; + return true; +} + +static void scx_kick_cpu(struct scx_sched *sch, s32 cpu, u64 flags) { struct rq *this_rq; unsigned long irq_flags; - if (!ops_cpu_valid(cpu, NULL)) + if (!ops_cpu_valid(sch, cpu, NULL)) return; local_irq_save(irq_flags); @@ -6910,7 +6367,7 @@ __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_ops_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)) { @@ -6935,6 +6392,26 @@ 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 * @@ -6943,23 +6420,30 @@ out: */ __bpf_kfunc s32 scx_bpf_dsq_nr_queued(u64 dsq_id) { + struct scx_sched *sch; struct scx_dispatch_q *dsq; s32 ret; preempt_disable(); + sch = rcu_dereference_sched(scx_root); + if (unlikely(!sch)) { + ret = -ENODEV; + goto out; + } + if (dsq_id == SCX_DSQ_LOCAL) { ret = READ_ONCE(this_rq()->scx.local_dsq.nr); goto out; } else if ((dsq_id & SCX_DSQ_LOCAL_ON) == SCX_DSQ_LOCAL_ON) { s32 cpu = dsq_id & SCX_DSQ_LOCAL_CPU_MASK; - if (ops_cpu_valid(cpu, NULL)) { + if (ops_cpu_valid(sch, cpu, NULL)) { ret = READ_ONCE(cpu_rq(cpu)->scx.local_dsq.nr); goto out; } } else { - dsq = find_user_dsq(dsq_id); + dsq = find_user_dsq(sch, dsq_id); if (dsq) { ret = READ_ONCE(dsq->nr); goto out; @@ -6982,7 +6466,13 @@ out: */ __bpf_kfunc void scx_bpf_destroy_dsq(u64 dsq_id) { - destroy_dsq(dsq_id); + struct scx_sched *sch; + + rcu_read_lock(); + sch = rcu_dereference(scx_root); + if (sch) + destroy_dsq(sch, dsq_id); + rcu_read_unlock(); } /** @@ -6999,22 +6489,34 @@ __bpf_kfunc int bpf_iter_scx_dsq_new(struct bpf_iter_scx_dsq *it, u64 dsq_id, u64 flags) { struct bpf_iter_scx_dsq_kern *kit = (void *)it; + struct scx_sched *sch; BUILD_BUG_ON(sizeof(struct bpf_iter_scx_dsq_kern) > sizeof(struct bpf_iter_scx_dsq)); BUILD_BUG_ON(__alignof__(struct bpf_iter_scx_dsq_kern) != __alignof__(struct bpf_iter_scx_dsq)); + BUILD_BUG_ON(__SCX_DSQ_ITER_ALL_FLAGS & + ((1U << __SCX_DSQ_LNODE_PRIV_SHIFT) - 1)); + + /* + * next() and destroy() will be called regardless of the return value. + * Always clear $kit->dsq. + */ + kit->dsq = NULL; + + sch = rcu_dereference_check(scx_root, rcu_read_lock_bh_held()); + if (unlikely(!sch)) + return -ENODEV; if (flags & ~__SCX_DSQ_ITER_USER_FLAGS) return -EINVAL; - kit->dsq = find_user_dsq(dsq_id); + kit->dsq = find_user_dsq(sch, dsq_id); if (!kit->dsq) return -ENOENT; - INIT_LIST_HEAD(&kit->cursor.node); - kit->cursor.flags |= SCX_DSQ_LNODE_ITER_CURSOR | flags; - kit->cursor.priv = READ_ONCE(kit->dsq->seq); + kit->cursor = INIT_DSQ_LIST_CURSOR(kit->cursor, flags, + READ_ONCE(kit->dsq->seq)); return 0; } @@ -7088,31 +6590,65 @@ __bpf_kfunc void bpf_iter_scx_dsq_destroy(struct bpf_iter_scx_dsq *it) kit->dsq = NULL; } +/** + * scx_bpf_dsq_peek - Lockless peek at the first element. + * @dsq_id: DSQ to examine. + * + * Read the first element in the DSQ. This is semantically equivalent to using + * the DSQ iterator, but is lockfree. Of course, like any lockless operation, + * this provides only a point-in-time snapshot, and the contents may change + * by the time any subsequent locking operation reads the queue. + * + * Returns the pointer, or NULL indicates an empty queue OR internal error. + */ +__bpf_kfunc struct task_struct *scx_bpf_dsq_peek(u64 dsq_id) +{ + struct scx_sched *sch; + struct scx_dispatch_q *dsq; + + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return NULL; + + if (unlikely(dsq_id & SCX_DSQ_FLAG_BUILTIN)) { + scx_error(sch, "peek disallowed on builtin DSQ 0x%llx", dsq_id); + return NULL; + } + + dsq = find_user_dsq(sch, dsq_id); + if (unlikely(!dsq)) { + scx_error(sch, "peek on non-existent DSQ 0x%llx", dsq_id); + return NULL; + } + + return rcu_dereference(dsq->first_task); +} + __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_ops_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_ops_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_ops_error("format preparation failed (%d)", ret); + scx_error(sch, "format preparation failed (%d)", ret); return ret; } @@ -7120,18 +6656,17 @@ 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_ops_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); } @@ -7150,12 +6685,14 @@ __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_ops_exit_kind(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); } @@ -7171,17 +6708,19 @@ __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_ops_exit_kind(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); } /** - * scx_bpf_dump - Generate extra debug dump specific to the BPF scheduler + * scx_bpf_dump_bstr - Generate extra debug dump specific to the BPF scheduler * @fmt: format string * @data: format string parameters packaged using ___bpf_fill() macro * @data__sz: @data len, must end in '__sz' for the verifier @@ -7195,17 +6734,24 @@ __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_ops_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)", @@ -7232,6 +6778,24 @@ __bpf_kfunc void scx_bpf_dump_bstr(char *fmt, unsigned long long *data, } /** + * scx_bpf_reenqueue_local - Re-enqueue tasks on a local DSQ + * + * Iterate over all of the tasks currently enqueued on the local DSQ of the + * caller's CPU, and re-enqueue them in the BPF scheduler. Can be called from + * anywhere. + */ +__bpf_kfunc void scx_bpf_reenqueue_local___v2(void) +{ + struct rq *rq; + + guard(preempt)(); + + rq = this_rq(); + local_set(&rq->scx.reenq_local_deferred, 1); + schedule_deferred(rq); +} + +/** * scx_bpf_cpuperf_cap - Query the maximum relative capacity of a CPU * @cpu: CPU of interest * @@ -7241,7 +6805,12 @@ __bpf_kfunc void scx_bpf_dump_bstr(char *fmt, unsigned long long *data, */ __bpf_kfunc u32 scx_bpf_cpuperf_cap(s32 cpu) { - if (ops_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; @@ -7263,7 +6832,12 @@ __bpf_kfunc u32 scx_bpf_cpuperf_cap(s32 cpu) */ __bpf_kfunc u32 scx_bpf_cpuperf_cur(s32 cpu) { - if (ops_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; @@ -7273,7 +6847,6 @@ __bpf_kfunc u32 scx_bpf_cpuperf_cur(s32 cpu) * scx_bpf_cpuperf_set - Set the relative performance target of a CPU * @cpu: CPU of interest * @perf: target performance level [0, %SCX_CPUPERF_ONE] - * @flags: %SCX_CPUPERF_* flags * * Set the target performance level of @cpu to @perf. @perf is in linear * relative scale between 0 and %SCX_CPUPERF_ONE. This determines how the @@ -7286,23 +6859,60 @@ __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(scx_root); + if (unlikely(!sch)) + return; + if (unlikely(perf > SCX_CPUPERF_ONE)) { - scx_ops_error("Invalid cpuperf target %u for CPU %d", perf, cpu); + scx_error(sch, "Invalid cpuperf target %u for CPU %d", perf, cpu); return; } - if (ops_cpu_valid(cpu, NULL)) { - struct rq *rq = cpu_rq(cpu); + if (ops_cpu_valid(sch, cpu, NULL)) { + struct rq *rq = cpu_rq(cpu), *locked_rq = scx_locked_rq(); + struct rq_flags rf; + + /* + * When called with an rq lock held, restrict the operation + * to the corresponding CPU to prevent ABBA deadlocks. + */ + if (locked_rq && rq != locked_rq) { + scx_error(sch, "Invalid target CPU %d", cpu); + return; + } + + /* + * If no rq lock is held, allow to operate on any CPU by + * acquiring the corresponding rq lock. + */ + if (!locked_rq) { + rq_lock_irqsave(rq, &rf); + update_rq_clock(rq); + } rq->scx.cpuperf_target = perf; + cpufreq_update_util(rq, 0); - rcu_read_lock_sched_notrace(); - cpufreq_update_util(cpu_rq(cpu), 0); - rcu_read_unlock_sched_notrace(); + if (!locked_rq) + rq_unlock_irqrestore(rq, &rf); } } /** + * scx_bpf_nr_node_ids - Return the number of possible node IDs + * + * All valid node IDs in the system are smaller than the returned value. + */ +__bpf_kfunc u32 scx_bpf_nr_node_ids(void) +{ + return nr_node_ids; +} + +/** * scx_bpf_nr_cpu_ids - Return the number of possible CPU IDs * * All valid CPU IDs in the system are smaller than the returned value. @@ -7343,176 +6953,96 @@ __bpf_kfunc void scx_bpf_put_cpumask(const struct cpumask *cpumask) } /** - * scx_bpf_get_idle_cpumask - Get a referenced kptr to the idle-tracking - * per-CPU cpumask. - * - * Returns NULL if idle tracking is not enabled, or running on a UP kernel. + * scx_bpf_task_running - Is task currently running? + * @p: task of interest */ -__bpf_kfunc const struct cpumask *scx_bpf_get_idle_cpumask(void) +__bpf_kfunc bool scx_bpf_task_running(const struct task_struct *p) { - if (!static_branch_likely(&scx_builtin_idle_enabled)) { - scx_ops_error("built-in idle tracking is disabled"); - return cpu_none_mask; - } - -#ifdef CONFIG_SMP - return idle_masks.cpu; -#else - return cpu_none_mask; -#endif + return task_rq(p)->curr == p; } /** - * scx_bpf_get_idle_smtmask - Get a referenced kptr to the idle-tracking, - * per-physical-core cpumask. Can be used to determine if an entire physical - * core is free. - * - * Returns NULL if idle tracking is not enabled, or running on a UP kernel. + * scx_bpf_task_cpu - CPU a task is currently associated with + * @p: task of interest */ -__bpf_kfunc const struct cpumask *scx_bpf_get_idle_smtmask(void) +__bpf_kfunc s32 scx_bpf_task_cpu(const struct task_struct *p) { - if (!static_branch_likely(&scx_builtin_idle_enabled)) { - scx_ops_error("built-in idle tracking is disabled"); - return cpu_none_mask; - } - -#ifdef CONFIG_SMP - if (sched_smt_active()) - return idle_masks.smt; - else - return idle_masks.cpu; -#else - return cpu_none_mask; -#endif + return task_cpu(p); } /** - * scx_bpf_put_idle_cpumask - Release a previously acquired referenced kptr to - * either the percpu, or SMT idle-tracking cpumask. + * scx_bpf_cpu_rq - Fetch the rq of a CPU + * @cpu: CPU of the rq */ -__bpf_kfunc void scx_bpf_put_idle_cpumask(const struct cpumask *idle_mask) +__bpf_kfunc struct rq *scx_bpf_cpu_rq(s32 cpu) { - /* - * Empty function body because we aren't actually acquiring or releasing - * a reference to a global idle cpumask, which is read-only in the - * caller and is never released. The acquire / release semantics here - * are just used to make the cpumask a trusted pointer in the caller. - */ -} + struct scx_sched *sch; -/** - * scx_bpf_test_and_clear_cpu_idle - Test and clear @cpu's idle state - * @cpu: cpu to test and clear idle for - * - * Returns %true if @cpu was idle and its idle state was successfully cleared. - * %false otherwise. - * - * Unavailable if ops.update_idle() is implemented and - * %SCX_OPS_KEEP_BUILTIN_IDLE is not set. - */ -__bpf_kfunc bool scx_bpf_test_and_clear_cpu_idle(s32 cpu) -{ - if (!static_branch_likely(&scx_builtin_idle_enabled)) { - scx_ops_error("built-in idle tracking is disabled"); - return false; - } + guard(rcu)(); - if (ops_cpu_valid(cpu, NULL)) - return test_and_clear_cpu_idle(cpu); - else - return false; -} + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + return NULL; -/** - * scx_bpf_pick_idle_cpu - Pick and claim an idle cpu - * @cpus_allowed: Allowed cpumask - * @flags: %SCX_PICK_IDLE_CPU_* flags - * - * Pick and claim an idle cpu in @cpus_allowed. Returns the picked idle cpu - * number on success. -%EBUSY if no matching cpu was found. - * - * Idle CPU tracking may race against CPU scheduling state transitions. For - * example, this function may return -%EBUSY as CPUs are transitioning into the - * idle state. If the caller then assumes that there will be dispatch events on - * the CPUs as they were all busy, the scheduler may end up stalling with CPUs - * idling while there are pending tasks. Use scx_bpf_pick_any_cpu() and - * scx_bpf_kick_cpu() to guarantee that there will be at least one dispatch - * event in the near future. - * - * Unavailable if ops.update_idle() is implemented and - * %SCX_OPS_KEEP_BUILTIN_IDLE is not set. - */ -__bpf_kfunc s32 scx_bpf_pick_idle_cpu(const struct cpumask *cpus_allowed, - u64 flags) -{ - if (!static_branch_likely(&scx_builtin_idle_enabled)) { - scx_ops_error("built-in idle tracking is disabled"); - return -EBUSY; + 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 scx_pick_idle_cpu(cpus_allowed, flags); + return cpu_rq(cpu); } /** - * scx_bpf_pick_any_cpu - Pick and claim an idle cpu if available or pick any CPU - * @cpus_allowed: Allowed cpumask - * @flags: %SCX_PICK_IDLE_CPU_* flags + * scx_bpf_locked_rq - Return the rq currently locked by SCX * - * Pick and claim an idle cpu in @cpus_allowed. If none is available, pick any - * CPU in @cpus_allowed. Guaranteed to succeed and returns the picked idle cpu - * number if @cpus_allowed is not empty. -%EBUSY is returned if @cpus_allowed is - * empty. - * - * If ops.update_idle() is implemented and %SCX_OPS_KEEP_BUILTIN_IDLE is not - * set, this function can't tell which CPUs are idle and will always pick any - * CPU. + * Returns the rq if a rq lock is currently held by SCX. + * Otherwise emits an error and returns NULL. */ -__bpf_kfunc s32 scx_bpf_pick_any_cpu(const struct cpumask *cpus_allowed, - u64 flags) +__bpf_kfunc struct rq *scx_bpf_locked_rq(void) { - s32 cpu; + struct scx_sched *sch; + struct rq *rq; - if (static_branch_likely(&scx_builtin_idle_enabled)) { - cpu = scx_pick_idle_cpu(cpus_allowed, flags); - if (cpu >= 0) - return cpu; + 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; } - cpu = cpumask_any_distribute(cpus_allowed); - if (cpu < nr_cpu_ids) - return cpu; - else - return -EBUSY; + return rq; } /** - * scx_bpf_task_running - Is task currently running? - * @p: task of interest + * scx_bpf_cpu_curr - Return remote CPU's curr task + * @cpu: CPU of interest + * + * Callers must hold RCU read lock (KF_RCU). */ -__bpf_kfunc bool scx_bpf_task_running(const struct task_struct *p) +__bpf_kfunc struct task_struct *scx_bpf_cpu_curr(s32 cpu) { - return task_rq(p)->curr == p; -} + struct scx_sched *sch; -/** - * scx_bpf_task_cpu - CPU a task is currently associated with - * @p: task of interest - */ -__bpf_kfunc s32 scx_bpf_task_cpu(const struct task_struct *p) -{ - return task_cpu(p); -} + guard(rcu)(); -/** - * scx_bpf_cpu_rq - Fetch the rq of a CPU - * @cpu: CPU of the rq - */ -__bpf_kfunc struct rq *scx_bpf_cpu_rq(s32 cpu) -{ - if (!ops_cpu_valid(cpu, NULL)) + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) return NULL; - return cpu_rq(cpu); + if (!ops_cpu_valid(sch, cpu, NULL)) + return NULL; + + return rcu_dereference(cpu_rq(cpu)->curr); } /** @@ -7531,8 +7061,15 @@ __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)(); - if (!scx_kf_allowed_on_arg_tasks(__SCX_KF_RQ_LOCKED, p)) + sch = rcu_dereference(scx_root); + if (unlikely(!sch)) + goto out; + + if (!scx_kf_allowed_on_arg_tasks(sch, __SCX_KF_RQ_LOCKED, p)) goto out; cgrp = tg_cgrp(tg); @@ -7543,37 +7080,153 @@ out: } #endif +/** + * scx_bpf_now - Returns a high-performance monotonically non-decreasing + * clock for the current CPU. The clock returned is in nanoseconds. + * + * It provides the following properties: + * + * 1) High performance: Many BPF schedulers call bpf_ktime_get_ns() frequently + * to account for execution time and track tasks' runtime properties. + * Unfortunately, in some hardware platforms, bpf_ktime_get_ns() -- which + * eventually reads a hardware timestamp counter -- is neither performant nor + * scalable. scx_bpf_now() aims to provide a high-performance clock by + * using the rq clock in the scheduler core whenever possible. + * + * 2) High enough resolution for the BPF scheduler use cases: In most BPF + * scheduler use cases, the required clock resolution is lower than the most + * accurate hardware clock (e.g., rdtsc in x86). scx_bpf_now() basically + * uses the rq clock in the scheduler core whenever it is valid. It considers + * that the rq clock is valid from the time the rq clock is updated + * (update_rq_clock) until the rq is unlocked (rq_unpin_lock). + * + * 3) Monotonically non-decreasing clock for the same CPU: scx_bpf_now() + * guarantees the clock never goes backward when comparing them in the same + * CPU. On the other hand, when comparing clocks in different CPUs, there + * is no such guarantee -- the clock can go backward. It provides a + * monotonically *non-decreasing* clock so that it would provide the same + * clock values in two different scx_bpf_now() calls in the same CPU + * during the same period of when the rq clock is valid. + */ +__bpf_kfunc u64 scx_bpf_now(void) +{ + struct rq *rq; + u64 clock; + + preempt_disable(); + + rq = this_rq(); + if (smp_load_acquire(&rq->scx.flags) & SCX_RQ_CLK_VALID) { + /* + * If the rq clock is valid, use the cached rq clock. + * + * Note that scx_bpf_now() is re-entrant between a process + * context and an interrupt context (e.g., timer interrupt). + * However, we don't need to consider the race between them + * because such race is not observable from a caller. + */ + clock = READ_ONCE(rq->scx.clock); + } else { + /* + * Otherwise, return a fresh rq clock. + * + * The rq clock is updated outside of the rq lock. + * In this case, keep the updated rq clock invalid so the next + * kfunc call outside the rq lock gets a fresh rq clock. + */ + clock = sched_clock_cpu(cpu_of(rq)); + } + + preempt_enable(); + + return clock; +} + +static void scx_read_events(struct scx_sched *sch, struct scx_event_stats *events) +{ + struct scx_event_stats *e_cpu; + int cpu; + + /* Aggregate per-CPU event counters into @events. */ + memset(events, 0, sizeof(*events)); + for_each_possible_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); + scx_agg_event(events, e_cpu, SCX_EV_ENQ_SKIP_EXITING); + scx_agg_event(events, e_cpu, SCX_EV_ENQ_SKIP_MIGRATION_DISABLED); + scx_agg_event(events, e_cpu, SCX_EV_REFILL_SLICE_DFL); + scx_agg_event(events, e_cpu, SCX_EV_BYPASS_DURATION); + scx_agg_event(events, e_cpu, SCX_EV_BYPASS_DISPATCH); + scx_agg_event(events, e_cpu, SCX_EV_BYPASS_ACTIVATE); + } +} + +/* + * scx_bpf_events - Get a system-wide event counter to + * @events: output buffer from a BPF program + * @events__sz: @events len, must end in '__sz'' for the verifier + */ +__bpf_kfunc void scx_bpf_events(struct scx_event_stats *events, + size_t events__sz) +{ + struct scx_sched *sch; + struct scx_event_stats e_sys; + + rcu_read_lock(); + sch = rcu_dereference(scx_root); + if (sch) + scx_read_events(sch, &e_sys); + else + memset(&e_sys, 0, sizeof(e_sys)); + rcu_read_unlock(); + + /* + * We cannot entirely trust a BPF-provided size since a BPF program + * might be compiled against a different vmlinux.h, of which + * scx_event_stats would be larger (a newer vmlinux.h) or smaller + * (an older vmlinux.h). Hence, we use the smaller size to avoid + * memory corruption. + */ + events__sz = min(events__sz, sizeof(*events)); + memcpy(events, &e_sys, events__sz); +} + __bpf_kfunc_end_defs(); BTF_KFUNCS_START(scx_kfunc_ids_any) +BTF_ID_FLAGS(func, scx_bpf_task_set_slice, KF_RCU); +BTF_ID_FLAGS(func, scx_bpf_task_set_dsq_vtime, KF_RCU); BTF_ID_FLAGS(func, scx_bpf_kick_cpu) BTF_ID_FLAGS(func, scx_bpf_dsq_nr_queued) BTF_ID_FLAGS(func, scx_bpf_destroy_dsq) +BTF_ID_FLAGS(func, scx_bpf_dsq_peek, KF_RCU_PROTECTED | KF_RET_NULL) BTF_ID_FLAGS(func, bpf_iter_scx_dsq_new, KF_ITER_NEW | KF_RCU_PROTECTED) BTF_ID_FLAGS(func, bpf_iter_scx_dsq_next, KF_ITER_NEXT | KF_RET_NULL) BTF_ID_FLAGS(func, bpf_iter_scx_dsq_destroy, KF_ITER_DESTROY) BTF_ID_FLAGS(func, scx_bpf_exit_bstr, KF_TRUSTED_ARGS) BTF_ID_FLAGS(func, scx_bpf_error_bstr, KF_TRUSTED_ARGS) BTF_ID_FLAGS(func, scx_bpf_dump_bstr, KF_TRUSTED_ARGS) +BTF_ID_FLAGS(func, scx_bpf_reenqueue_local___v2) BTF_ID_FLAGS(func, scx_bpf_cpuperf_cap) BTF_ID_FLAGS(func, scx_bpf_cpuperf_cur) BTF_ID_FLAGS(func, scx_bpf_cpuperf_set) +BTF_ID_FLAGS(func, scx_bpf_nr_node_ids) BTF_ID_FLAGS(func, scx_bpf_nr_cpu_ids) BTF_ID_FLAGS(func, scx_bpf_get_possible_cpumask, KF_ACQUIRE) BTF_ID_FLAGS(func, scx_bpf_get_online_cpumask, KF_ACQUIRE) BTF_ID_FLAGS(func, scx_bpf_put_cpumask, KF_RELEASE) -BTF_ID_FLAGS(func, scx_bpf_get_idle_cpumask, KF_ACQUIRE) -BTF_ID_FLAGS(func, scx_bpf_get_idle_smtmask, KF_ACQUIRE) -BTF_ID_FLAGS(func, scx_bpf_put_idle_cpumask, KF_RELEASE) -BTF_ID_FLAGS(func, scx_bpf_test_and_clear_cpu_idle) -BTF_ID_FLAGS(func, scx_bpf_pick_idle_cpu, KF_RCU) -BTF_ID_FLAGS(func, scx_bpf_pick_any_cpu, KF_RCU) 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 +BTF_ID_FLAGS(func, scx_bpf_now) +BTF_ID_FLAGS(func, scx_bpf_events, KF_TRUSTED_ARGS) BTF_KFUNCS_END(scx_kfunc_ids_any) static const struct btf_kfunc_id_set scx_kfunc_set_any = { @@ -7597,8 +7250,6 @@ static int __init scx_init(void) * check using scx_kf_allowed(). */ if ((ret = register_btf_kfunc_id_set(BPF_PROG_TYPE_STRUCT_OPS, - &scx_kfunc_set_select_cpu)) || - (ret = register_btf_kfunc_id_set(BPF_PROG_TYPE_STRUCT_OPS, &scx_kfunc_set_enqueue_dispatch)) || (ret = register_btf_kfunc_id_set(BPF_PROG_TYPE_STRUCT_OPS, &scx_kfunc_set_dispatch)) || @@ -7618,6 +7269,12 @@ static int __init scx_init(void) return ret; } + ret = scx_idle_init(); + if (ret) { + pr_err("sched_ext: Failed to initialize idle tracking (%d)\n", ret); + return ret; + } + ret = register_bpf_struct_ops(&bpf_sched_ext_ops, sched_ext_ops); if (ret) { pr_err("sched_ext: Failed to register struct_ops (%d)\n", ret); @@ -7642,6 +7299,12 @@ static int __init scx_init(void) return ret; } + if (!alloc_cpumask_var(&scx_bypass_lb_donee_cpumask, GFP_KERNEL) || + !alloc_cpumask_var(&scx_bypass_lb_resched_cpumask, GFP_KERNEL)) { + pr_err("sched_ext: Failed to allocate cpumasks\n"); + return -ENOMEM; + } + return 0; } __initcall(scx_init); |