From 2d3791f116bb3d5b17571dadb8e085e12ae3a3cf Mon Sep 17 00:00:00 2001 From: Liu Xinpeng Date: Mon, 25 Oct 2021 11:46:25 +0800 Subject: psi: Remove repeated verbose comment Comment in function psi_task_switch,there are two same lines. ... * runtime state, the cgroup that contains both tasks * runtime state, the cgroup that contains both tasks ... Signed-off-by: Liu Xinpeng Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lore.kernel.org/r/1635133586-84611-1-git-send-email-liuxp11@chinatelecom.cn --- kernel/sched/psi.c | 1 - 1 file changed, 1 deletion(-) diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c index 1652f2bb54b7..526af84ab852 100644 --- a/kernel/sched/psi.c +++ b/kernel/sched/psi.c @@ -833,7 +833,6 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, /* * When switching between tasks that have an identical * runtime state, the cgroup that contains both tasks - * runtime state, the cgroup that contains both tasks * we reach the first common ancestor. Iterate @next's * ancestors only until we encounter @prev's ONCPU. */ -- cgit v1.2.3-70-g09d2 From 2fb75e1b642f49253d8848c9e47e8942f5366221 Mon Sep 17 00:00:00 2001 From: Liu Xinpeng Date: Mon, 25 Oct 2021 11:46:26 +0800 Subject: psi: Add a missing SPDX license header Add the missing SPDX license header to include/linux/psi.h include/linux/psi_types.h kernel/sched/psi.c Signed-off-by: Liu Xinpeng Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lore.kernel.org/r/1635133586-84611-2-git-send-email-liuxp11@chinatelecom.cn --- include/linux/psi.h | 1 + include/linux/psi_types.h | 1 + kernel/sched/psi.c | 1 + 3 files changed, 3 insertions(+) diff --git a/include/linux/psi.h b/include/linux/psi.h index 65eb1476ac70..a70ca833c6d7 100644 --- a/include/linux/psi.h +++ b/include/linux/psi.h @@ -1,3 +1,4 @@ +/* SPDX-License-Identifier: GPL-2.0 */ #ifndef _LINUX_PSI_H #define _LINUX_PSI_H diff --git a/include/linux/psi_types.h b/include/linux/psi_types.h index 0a23300d49af..bf50068d5d4b 100644 --- a/include/linux/psi_types.h +++ b/include/linux/psi_types.h @@ -1,3 +1,4 @@ +/* SPDX-License-Identifier: GPL-2.0 */ #ifndef _LINUX_PSI_TYPES_H #define _LINUX_PSI_TYPES_H diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c index 526af84ab852..3397fa001157 100644 --- a/kernel/sched/psi.c +++ b/kernel/sched/psi.c @@ -1,3 +1,4 @@ +// SPDX-License-Identifier: GPL-2.0 /* * Pressure stall information for CPU, memory and IO * -- cgit v1.2.3-70-g09d2 From 4feee7d12603deca8775f9f9ae5e121093837444 Mon Sep 17 00:00:00 2001 From: Josh Don Date: Mon, 18 Oct 2021 13:34:28 -0700 Subject: sched/core: Forced idle accounting Adds accounting for "forced idle" time, which is time where a cookie'd task forces its SMT sibling to idle, despite the presence of runnable tasks. Forced idle time is one means to measure the cost of enabling core scheduling (ie. the capacity lost due to the need to force idle). Forced idle time is attributed to the thread responsible for causing the forced idle. A few details: - Forced idle time is displayed via /proc/PID/sched. It also requires that schedstats is enabled. - Forced idle is only accounted when a sibling hyperthread is held idle despite the presence of runnable tasks. No time is charged if a sibling is idle but has no runnable tasks. - Tasks with 0 cookie are never charged forced idle. - For SMT > 2, we scale the amount of forced idle charged based on the number of forced idle siblings. Additionally, we split the time up and evenly charge it to all running tasks, as each is equally responsible for the forced idle. Signed-off-by: Josh Don Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/20211018203428.2025792-1-joshdon@google.com --- include/linux/sched.h | 4 +++ kernel/sched/core.c | 82 +++++++++++++++++++++++++++++++++++------------ kernel/sched/core_sched.c | 66 +++++++++++++++++++++++++++++++++++++- kernel/sched/debug.c | 4 +++ kernel/sched/fair.c | 2 +- kernel/sched/sched.h | 32 ++++++++++++++++-- 6 files changed, 166 insertions(+), 24 deletions(-) diff --git a/include/linux/sched.h b/include/linux/sched.h index 78c351e35fec..d2e261adb8ea 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -523,7 +523,11 @@ struct sched_statistics { u64 nr_wakeups_affine_attempts; u64 nr_wakeups_passive; u64 nr_wakeups_idle; + +#ifdef CONFIG_SCHED_CORE + u64 core_forceidle_sum; #endif +#endif /* CONFIG_SCHEDSTATS */ } ____cacheline_aligned; struct sched_entity { diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 3c9b0fda64ac..beaa8be6241e 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -144,7 +144,7 @@ static inline bool __sched_core_less(struct task_struct *a, struct task_struct * return false; /* flip prio, so high prio is leftmost */ - if (prio_less(b, a, task_rq(a)->core->core_forceidle)) + if (prio_less(b, a, !!task_rq(a)->core->core_forceidle_count)) return true; return false; @@ -181,15 +181,23 @@ void sched_core_enqueue(struct rq *rq, struct task_struct *p) rb_add(&p->core_node, &rq->core_tree, rb_sched_core_less); } -void sched_core_dequeue(struct rq *rq, struct task_struct *p) +void sched_core_dequeue(struct rq *rq, struct task_struct *p, int flags) { rq->core->core_task_seq++; - if (!sched_core_enqueued(p)) - return; + if (sched_core_enqueued(p)) { + rb_erase(&p->core_node, &rq->core_tree); + RB_CLEAR_NODE(&p->core_node); + } - rb_erase(&p->core_node, &rq->core_tree); - RB_CLEAR_NODE(&p->core_node); + /* + * Migrating the last task off the cpu, with the cpu in forced idle + * state. Reschedule to create an accounting edge for forced idle, + * and re-examine whether the core is still in forced idle state. + */ + if (!(flags & DEQUEUE_SAVE) && rq->nr_running == 1 && + rq->core->core_forceidle_count && rq->curr == rq->idle) + resched_curr(rq); } /* @@ -280,6 +288,8 @@ static void __sched_core_flip(bool enabled) for_each_cpu(t, smt_mask) cpu_rq(t)->core_enabled = enabled; + cpu_rq(cpu)->core->core_forceidle_start = 0; + sched_core_unlock(cpu, &flags); cpumask_andnot(&sched_core_mask, &sched_core_mask, smt_mask); @@ -364,7 +374,8 @@ void sched_core_put(void) #else /* !CONFIG_SCHED_CORE */ static inline void sched_core_enqueue(struct rq *rq, struct task_struct *p) { } -static inline void sched_core_dequeue(struct rq *rq, struct task_struct *p) { } +static inline void +sched_core_dequeue(struct rq *rq, struct task_struct *p, int flags) { } #endif /* CONFIG_SCHED_CORE */ @@ -2005,7 +2016,7 @@ static inline void enqueue_task(struct rq *rq, struct task_struct *p, int flags) static inline void dequeue_task(struct rq *rq, struct task_struct *p, int flags) { if (sched_core_enabled(rq)) - sched_core_dequeue(rq, p); + sched_core_dequeue(rq, p, flags); if (!(flags & DEQUEUE_NOCLOCK)) update_rq_clock(rq); @@ -5244,6 +5255,7 @@ void scheduler_tick(void) if (sched_feat(LATENCY_WARN)) resched_latency = cpu_resched_latency(rq); calc_global_load_tick(rq); + sched_core_tick(rq); rq_unlock(rq, &rf); @@ -5656,6 +5668,7 @@ pick_next_task(struct rq *rq, struct task_struct *prev, struct rq_flags *rf) struct task_struct *next, *p, *max = NULL; const struct cpumask *smt_mask; bool fi_before = false; + bool core_clock_updated = (rq == rq->core); unsigned long cookie; int i, cpu, occ = 0; struct rq *rq_i; @@ -5708,10 +5721,18 @@ pick_next_task(struct rq *rq, struct task_struct *prev, struct rq_flags *rf) /* reset state */ rq->core->core_cookie = 0UL; - if (rq->core->core_forceidle) { + if (rq->core->core_forceidle_count) { + if (!core_clock_updated) { + update_rq_clock(rq->core); + core_clock_updated = true; + } + sched_core_account_forceidle(rq); + /* reset after accounting force idle */ + rq->core->core_forceidle_start = 0; + rq->core->core_forceidle_count = 0; + rq->core->core_forceidle_occupation = 0; need_sync = true; fi_before = true; - rq->core->core_forceidle = false; } /* @@ -5753,7 +5774,12 @@ pick_next_task(struct rq *rq, struct task_struct *prev, struct rq_flags *rf) for_each_cpu_wrap(i, smt_mask, cpu) { rq_i = cpu_rq(i); - if (i != cpu) + /* + * Current cpu always has its clock updated on entrance to + * pick_next_task(). If the current cpu is not the core, + * the core may also have been updated above. + */ + if (i != cpu && (rq_i != rq->core || !core_clock_updated)) update_rq_clock(rq_i); p = rq_i->core_pick = pick_task(rq_i); @@ -5783,7 +5809,7 @@ pick_next_task(struct rq *rq, struct task_struct *prev, struct rq_flags *rf) if (p == rq_i->idle) { if (rq_i->nr_running) { - rq->core->core_forceidle = true; + rq->core->core_forceidle_count++; if (!fi_before) rq->core->core_forceidle_seq++; } @@ -5792,6 +5818,12 @@ pick_next_task(struct rq *rq, struct task_struct *prev, struct rq_flags *rf) } } + if (schedstat_enabled() && rq->core->core_forceidle_count) { + if (cookie) + rq->core->core_forceidle_start = rq_clock(rq->core); + rq->core->core_forceidle_occupation = occ; + } + rq->core->core_pick_seq = rq->core->core_task_seq; next = rq->core_pick; rq->core_sched_seq = rq->core->core_pick_seq; @@ -5828,8 +5860,8 @@ pick_next_task(struct rq *rq, struct task_struct *prev, struct rq_flags *rf) * 1 0 1 * 1 1 0 */ - if (!(fi_before && rq->core->core_forceidle)) - task_vruntime_update(rq_i, rq_i->core_pick, rq->core->core_forceidle); + if (!(fi_before && rq->core->core_forceidle_count)) + task_vruntime_update(rq_i, rq_i->core_pick, !!rq->core->core_forceidle_count); rq_i->core_pick->core_occupation = occ; @@ -6033,11 +6065,19 @@ static void sched_core_cpu_deactivate(unsigned int cpu) goto unlock; /* copy the shared state to the new leader */ - core_rq->core_task_seq = rq->core_task_seq; - core_rq->core_pick_seq = rq->core_pick_seq; - core_rq->core_cookie = rq->core_cookie; - core_rq->core_forceidle = rq->core_forceidle; - core_rq->core_forceidle_seq = rq->core_forceidle_seq; + core_rq->core_task_seq = rq->core_task_seq; + core_rq->core_pick_seq = rq->core_pick_seq; + core_rq->core_cookie = rq->core_cookie; + core_rq->core_forceidle_count = rq->core_forceidle_count; + core_rq->core_forceidle_seq = rq->core_forceidle_seq; + core_rq->core_forceidle_occupation = rq->core_forceidle_occupation; + + /* + * Accounting edge for forced idle is handled in pick_next_task(). + * Don't need another one here, since the hotplug thread shouldn't + * have a cookie. + */ + core_rq->core_forceidle_start = 0; /* install new leader */ for_each_cpu(t, smt_mask) { @@ -9413,7 +9453,9 @@ void __init sched_init(void) rq->core_pick = NULL; rq->core_enabled = 0; rq->core_tree = RB_ROOT; - rq->core_forceidle = false; + rq->core_forceidle_count = 0; + rq->core_forceidle_occupation = 0; + rq->core_forceidle_start = 0; rq->core_cookie = 0UL; #endif diff --git a/kernel/sched/core_sched.c b/kernel/sched/core_sched.c index 517f72b008f5..1fb45672ec85 100644 --- a/kernel/sched/core_sched.c +++ b/kernel/sched/core_sched.c @@ -73,7 +73,7 @@ static unsigned long sched_core_update_cookie(struct task_struct *p, enqueued = sched_core_enqueued(p); if (enqueued) - sched_core_dequeue(rq, p); + sched_core_dequeue(rq, p, DEQUEUE_SAVE); old_cookie = p->core_cookie; p->core_cookie = cookie; @@ -85,6 +85,10 @@ static unsigned long sched_core_update_cookie(struct task_struct *p, * If task is currently running, it may not be compatible anymore after * the cookie change, so enter the scheduler on its CPU to schedule it * away. + * + * Note that it is possible that as a result of this cookie change, the + * core has now entered/left forced idle state. Defer accounting to the + * next scheduling edge, rather than always forcing a reschedule here. */ if (task_running(rq, p)) resched_curr(rq); @@ -232,3 +236,63 @@ out: return err; } +#ifdef CONFIG_SCHEDSTATS + +/* REQUIRES: rq->core's clock recently updated. */ +void __sched_core_account_forceidle(struct rq *rq) +{ + const struct cpumask *smt_mask = cpu_smt_mask(cpu_of(rq)); + u64 delta, now = rq_clock(rq->core); + struct rq *rq_i; + struct task_struct *p; + int i; + + lockdep_assert_rq_held(rq); + + WARN_ON_ONCE(!rq->core->core_forceidle_count); + + if (rq->core->core_forceidle_start == 0) + return; + + delta = now - rq->core->core_forceidle_start; + if (unlikely((s64)delta <= 0)) + return; + + rq->core->core_forceidle_start = now; + + if (WARN_ON_ONCE(!rq->core->core_forceidle_occupation)) { + /* can't be forced idle without a running task */ + } else if (rq->core->core_forceidle_count > 1 || + rq->core->core_forceidle_occupation > 1) { + /* + * For larger SMT configurations, we need to scale the charged + * forced idle amount since there can be more than one forced + * idle sibling and more than one running cookied task. + */ + delta *= rq->core->core_forceidle_count; + delta = div_u64(delta, rq->core->core_forceidle_occupation); + } + + for_each_cpu(i, smt_mask) { + rq_i = cpu_rq(i); + p = rq_i->core_pick ?: rq_i->curr; + + if (!p->core_cookie) + continue; + + __schedstat_add(p->stats.core_forceidle_sum, delta); + } +} + +void __sched_core_tick(struct rq *rq) +{ + if (!rq->core->core_forceidle_count) + return; + + if (rq != rq->core) + update_rq_clock(rq->core); + + __sched_core_account_forceidle(rq); +} + +#endif /* CONFIG_SCHEDSTATS */ diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c index 7dcbaa31c5d9..aa29211de1bf 100644 --- a/kernel/sched/debug.c +++ b/kernel/sched/debug.c @@ -1023,6 +1023,10 @@ void proc_sched_show_task(struct task_struct *p, struct pid_namespace *ns, __PN(avg_atom); __PN(avg_per_cpu); + +#ifdef CONFIG_SCHED_CORE + PN_SCHEDSTAT(core_forceidle_sum); +#endif } __P(nr_switches); diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 6e476f6d9435..884f29d07963 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -11068,7 +11068,7 @@ static inline void task_tick_core(struct rq *rq, struct task_struct *curr) * MIN_NR_TASKS_DURING_FORCEIDLE - 1 tasks and use that to check * if we need to give up the CPU. */ - if (rq->core->core_forceidle && rq->cfs.nr_running == 1 && + if (rq->core->core_forceidle_count && rq->cfs.nr_running == 1 && __entity_slice_used(&curr->se, MIN_NR_TASKS_DURING_FORCEIDLE)) resched_curr(rq); } diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 0e66749486e7..eb971151e7e4 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -1111,8 +1111,10 @@ struct rq { unsigned int core_task_seq; unsigned int core_pick_seq; unsigned long core_cookie; - unsigned char core_forceidle; + unsigned int core_forceidle_count; unsigned int core_forceidle_seq; + unsigned int core_forceidle_occupation; + u64 core_forceidle_start; #endif }; @@ -1253,7 +1255,7 @@ static inline bool sched_core_enqueued(struct task_struct *p) } extern void sched_core_enqueue(struct rq *rq, struct task_struct *p); -extern void sched_core_dequeue(struct rq *rq, struct task_struct *p); +extern void sched_core_dequeue(struct rq *rq, struct task_struct *p, int flags); extern void sched_core_get(void); extern void sched_core_put(void); @@ -1854,6 +1856,32 @@ static inline void flush_smp_call_function_from_idle(void) { } #include "stats.h" #include "autogroup.h" +#if defined(CONFIG_SCHED_CORE) && defined(CONFIG_SCHEDSTATS) + +extern void __sched_core_account_forceidle(struct rq *rq); + +static inline void sched_core_account_forceidle(struct rq *rq) +{ + if (schedstat_enabled()) + __sched_core_account_forceidle(rq); +} + +extern void __sched_core_tick(struct rq *rq); + +static inline void sched_core_tick(struct rq *rq) +{ + if (sched_core_enabled(rq) && schedstat_enabled()) + __sched_core_tick(rq); +} + +#else + +static inline void sched_core_account_forceidle(struct rq *rq) {} + +static inline void sched_core_tick(struct rq *rq) {} + +#endif /* CONFIG_SCHED_CORE && CONFIG_SCHEDSTATS */ + #ifdef CONFIG_CGROUP_SCHED /* -- cgit v1.2.3-70-g09d2 From cb0e52b7748737b2cf6481fdd9b920ce7e1ebbdf Mon Sep 17 00:00:00 2001 From: Brian Chen Date: Wed, 10 Nov 2021 21:33:12 +0000 Subject: psi: Fix PSI_MEM_FULL state when tasks are in memstall and doing reclaim We've noticed cases where tasks in a cgroup are stalled on memory but there is little memory FULL pressure since tasks stay on the runqueue in reclaim. A simple example involves a single threaded program that keeps leaking and touching large amounts of memory. It runs in a cgroup with swap enabled, memory.high set at 10M and cpu.max ratio set at 5%. Though there is significant CPU pressure and memory SOME, there is barely any memory FULL since the task enters reclaim and stays on the runqueue. However, this memory-bound task is effectively stalled on memory and we expect memory FULL to match memory SOME in this scenario. The code is confused about memstall && running, thinking there is a stalled task and a productive task when there's only one task: a reclaimer that's counted as both. To fix this, we redefine the condition for PSI_MEM_FULL to check that all running tasks are in an active memstall instead of checking that there are no running tasks. case PSI_MEM_FULL: - return unlikely(tasks[NR_MEMSTALL] && !tasks[NR_RUNNING]); + return unlikely(tasks[NR_MEMSTALL] && + tasks[NR_RUNNING] == tasks[NR_MEMSTALL_RUNNING]); This will capture reclaimers. It will also capture tasks that called psi_memstall_enter() and are about to sleep, but this should be negligible noise. Signed-off-by: Brian Chen Signed-off-by: Peter Zijlstra (Intel) Acked-by: Johannes Weiner Link: https://lore.kernel.org/r/20211110213312.310243-1-brianchen118@gmail.com --- include/linux/psi_types.h | 13 ++++++++++++- kernel/sched/psi.c | 45 ++++++++++++++++++++++++++++----------------- kernel/sched/stats.h | 5 ++++- 3 files changed, 44 insertions(+), 19 deletions(-) diff --git a/include/linux/psi_types.h b/include/linux/psi_types.h index bf50068d5d4b..516c0fe836fd 100644 --- a/include/linux/psi_types.h +++ b/include/linux/psi_types.h @@ -22,7 +22,17 @@ enum psi_task_count { * don't have to special case any state tracking for it. */ NR_ONCPU, - NR_PSI_TASK_COUNTS = 4, + /* + * For IO and CPU stalls the presence of running/oncpu tasks + * in the domain means a partial rather than a full stall. + * For memory it's not so simple because of page reclaimers: + * they are running/oncpu while representing a stall. To tell + * whether a domain has productivity left or not, we need to + * distinguish between regular running (i.e. productive) + * threads and memstall ones. + */ + NR_MEMSTALL_RUNNING, + NR_PSI_TASK_COUNTS = 5, }; /* Task state bitmasks */ @@ -30,6 +40,7 @@ enum psi_task_count { #define TSK_MEMSTALL (1 << NR_MEMSTALL) #define TSK_RUNNING (1 << NR_RUNNING) #define TSK_ONCPU (1 << NR_ONCPU) +#define TSK_MEMSTALL_RUNNING (1 << NR_MEMSTALL_RUNNING) /* Resources that workloads could be stalled on */ enum psi_res { diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c index 3397fa001157..a679613a7cb7 100644 --- a/kernel/sched/psi.c +++ b/kernel/sched/psi.c @@ -35,13 +35,19 @@ * delayed on that resource such that nobody is advancing and the CPU * goes idle. This leaves both workload and CPU unproductive. * - * Naturally, the FULL state doesn't exist for the CPU resource at the - * system level, but exist at the cgroup level, means all non-idle tasks - * in a cgroup are delayed on the CPU resource which used by others outside - * of the cgroup or throttled by the cgroup cpu.max configuration. - * * SOME = nr_delayed_tasks != 0 - * FULL = nr_delayed_tasks != 0 && nr_running_tasks == 0 + * FULL = nr_delayed_tasks != 0 && nr_productive_tasks == 0 + * + * What it means for a task to be productive is defined differently + * for each resource. For IO, productive means a running task. For + * memory, productive means a running task that isn't a reclaimer. For + * CPU, productive means an oncpu task. + * + * Naturally, the FULL state doesn't exist for the CPU resource at the + * system level, but exist at the cgroup level. At the cgroup level, + * FULL means all non-idle tasks in the cgroup are delayed on the CPU + * resource which is being used by others outside of the cgroup or + * throttled by the cgroup cpu.max configuration. * * The percentage of wallclock time spent in those compound stall * states gives pressure numbers between 0 and 100 for each resource, @@ -82,13 +88,13 @@ * * threads = min(nr_nonidle_tasks, nr_cpus) * SOME = min(nr_delayed_tasks / threads, 1) - * FULL = (threads - min(nr_running_tasks, threads)) / threads + * FULL = (threads - min(nr_productive_tasks, threads)) / threads * * For the 257 number crunchers on 256 CPUs, this yields: * * threads = min(257, 256) * SOME = min(1 / 256, 1) = 0.4% - * FULL = (256 - min(257, 256)) / 256 = 0% + * FULL = (256 - min(256, 256)) / 256 = 0% * * For the 1 out of 4 memory-delayed tasks, this yields: * @@ -113,7 +119,7 @@ * For each runqueue, we track: * * tSOME[cpu] = time(nr_delayed_tasks[cpu] != 0) - * tFULL[cpu] = time(nr_delayed_tasks[cpu] && !nr_running_tasks[cpu]) + * tFULL[cpu] = time(nr_delayed_tasks[cpu] && !nr_productive_tasks[cpu]) * tNONIDLE[cpu] = time(nr_nonidle_tasks[cpu] != 0) * * and then periodically aggregate: @@ -234,7 +240,8 @@ static bool test_state(unsigned int *tasks, enum psi_states state) case PSI_MEM_SOME: return unlikely(tasks[NR_MEMSTALL]); case PSI_MEM_FULL: - return unlikely(tasks[NR_MEMSTALL] && !tasks[NR_RUNNING]); + return unlikely(tasks[NR_MEMSTALL] && + tasks[NR_RUNNING] == tasks[NR_MEMSTALL_RUNNING]); case PSI_CPU_SOME: return unlikely(tasks[NR_RUNNING] > tasks[NR_ONCPU]); case PSI_CPU_FULL: @@ -711,10 +718,11 @@ static void psi_group_change(struct psi_group *group, int cpu, if (groupc->tasks[t]) { groupc->tasks[t]--; } else if (!psi_bug) { - printk_deferred(KERN_ERR "psi: task underflow! cpu=%d t=%d tasks=[%u %u %u %u] clear=%x set=%x\n", + printk_deferred(KERN_ERR "psi: task underflow! cpu=%d t=%d tasks=[%u %u %u %u %u] clear=%x set=%x\n", cpu, t, groupc->tasks[0], groupc->tasks[1], groupc->tasks[2], - groupc->tasks[3], clear, set); + groupc->tasks[3], groupc->tasks[4], + clear, set); psi_bug = 1; } } @@ -854,12 +862,15 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next, int clear = TSK_ONCPU, set = 0; /* - * When we're going to sleep, psi_dequeue() lets us handle - * TSK_RUNNING and TSK_IOWAIT here, where we can combine it - * with TSK_ONCPU and save walking common ancestors twice. + * When we're going to sleep, psi_dequeue() lets us + * handle TSK_RUNNING, TSK_MEMSTALL_RUNNING and + * TSK_IOWAIT here, where we can combine it with + * TSK_ONCPU and save walking common ancestors twice. */ if (sleep) { clear |= TSK_RUNNING; + if (prev->in_memstall) + clear |= TSK_MEMSTALL_RUNNING; if (prev->in_iowait) set |= TSK_IOWAIT; } @@ -908,7 +919,7 @@ void psi_memstall_enter(unsigned long *flags) rq = this_rq_lock_irq(&rf); current->in_memstall = 1; - psi_task_change(current, 0, TSK_MEMSTALL); + psi_task_change(current, 0, TSK_MEMSTALL | TSK_MEMSTALL_RUNNING); rq_unlock_irq(rq, &rf); } @@ -937,7 +948,7 @@ void psi_memstall_leave(unsigned long *flags) rq = this_rq_lock_irq(&rf); current->in_memstall = 0; - psi_task_change(current, TSK_MEMSTALL, 0); + psi_task_change(current, TSK_MEMSTALL | TSK_MEMSTALL_RUNNING, 0); rq_unlock_irq(rq, &rf); } diff --git a/kernel/sched/stats.h b/kernel/sched/stats.h index cfb0893a83d4..3a3c826dd83a 100644 --- a/kernel/sched/stats.h +++ b/kernel/sched/stats.h @@ -118,6 +118,9 @@ static inline void psi_enqueue(struct task_struct *p, bool wakeup) if (static_branch_likely(&psi_disabled)) return; + if (p->in_memstall) + set |= TSK_MEMSTALL_RUNNING; + if (!wakeup || p->sched_psi_wake_requeue) { if (p->in_memstall) set |= TSK_MEMSTALL; @@ -148,7 +151,7 @@ static inline void psi_dequeue(struct task_struct *p, bool sleep) return; if (p->in_memstall) - clear |= TSK_MEMSTALL; + clear |= (TSK_MEMSTALL | TSK_MEMSTALL_RUNNING); psi_task_change(p, clear, 0); } -- cgit v1.2.3-70-g09d2 From 9731698ecb9c851f353ce2496292ff9fcea39dff Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Mon, 15 Nov 2021 19:46:04 +0300 Subject: cputime, cpuacct: Include guest time in user time in cpuacct.stat cpuacct.stat in no-root cgroups shows user time without guest time included int it. This doesn't match with user time shown in root cpuacct.stat and /proc//stat. This also affects cgroup2's cpu.stat in the same way. Make account_guest_time() to add user time to cgroup's cpustat to fix this. Fixes: ef12fefabf94 ("cpuacct: add per-cgroup utime/stime statistics") Signed-off-by: Andrey Ryabinin Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Daniel Jordan Acked-by: Tejun Heo Cc: Link: https://lore.kernel.org/r/20211115164607.23784-1-arbn@yandex-team.com --- kernel/sched/cputime.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c index 872e481d5098..042a6dbce8f3 100644 --- a/kernel/sched/cputime.c +++ b/kernel/sched/cputime.c @@ -148,10 +148,10 @@ void account_guest_time(struct task_struct *p, u64 cputime) /* Add guest time to cpustat. */ if (task_nice(p) > 0) { - cpustat[CPUTIME_NICE] += cputime; + task_group_account_field(p, CPUTIME_NICE, cputime); cpustat[CPUTIME_GUEST_NICE] += cputime; } else { - cpustat[CPUTIME_USER] += cputime; + task_group_account_field(p, CPUTIME_USER, cputime); cpustat[CPUTIME_GUEST] += cputime; } } -- cgit v1.2.3-70-g09d2 From c7ccbf4b6174e32c130892570db06d0f496cfef0 Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Mon, 15 Nov 2021 19:46:05 +0300 Subject: cpuacct: Convert BUG_ON() to WARN_ON_ONCE() Replace fatal BUG_ON() with more safe WARN_ON_ONCE() in cpuacct_cpuusage_read(). Signed-off-by: Andrey Ryabinin Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Daniel Jordan Acked-by: Tejun Heo Link: https://lore.kernel.org/r/20211115164607.23784-2-arbn@yandex-team.com --- kernel/sched/cpuacct.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/kernel/sched/cpuacct.c b/kernel/sched/cpuacct.c index 893eece65bfd..f347cf9e4634 100644 --- a/kernel/sched/cpuacct.c +++ b/kernel/sched/cpuacct.c @@ -106,7 +106,8 @@ static u64 cpuacct_cpuusage_read(struct cpuacct *ca, int cpu, * We allow index == CPUACCT_STAT_NSTATS here to read * the sum of usages. */ - BUG_ON(index > CPUACCT_STAT_NSTATS); + if (WARN_ON_ONCE(index > CPUACCT_STAT_NSTATS)) + return 0; #ifndef CONFIG_64BIT /* -- cgit v1.2.3-70-g09d2 From dd02d4234c9a2214a81c57a16484304a1a51872a Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Mon, 15 Nov 2021 19:46:06 +0300 Subject: sched/cpuacct: Fix user/system in shown cpuacct.usage* cpuacct has 2 different ways of accounting and showing user and system times. The first one uses cpuacct_account_field() to account times and cpuacct.stat file to expose them. And this one seems to work ok. The second one is uses cpuacct_charge() function for accounting and set of cpuacct.usage* files to show times. Despite some attempts to fix it in the past it still doesn't work. Sometimes while running KVM guest the cpuacct_charge() accounts most of the guest time as system time. This doesn't match with user&system times shown in cpuacct.stat or proc//stat. Demonstration: # git clone https://github.com/aryabinin/kvmsample # make # mkdir /sys/fs/cgroup/cpuacct/test # echo $$ > /sys/fs/cgroup/cpuacct/test/tasks # ./kvmsample & # for i in {1..5}; do cat /sys/fs/cgroup/cpuacct/test/cpuacct.usage_sys; sleep 1; done 1976535645 2979839428 3979832704 4983603153 5983604157 Use cpustats accounted in cpuacct_account_field() as the source of user/sys times for cpuacct.usage* files. Make cpuacct_charge() to account only summary execution time. Fixes: d740037fac70 ("sched/cpuacct: Split usage accounting into user_usage and sys_usage") Signed-off-by: Andrey Ryabinin Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Daniel Jordan Acked-by: Tejun Heo Cc: Link: https://lore.kernel.org/r/20211115164607.23784-3-arbn@yandex-team.com --- kernel/sched/cpuacct.c | 79 ++++++++++++++++++++------------------------------ 1 file changed, 32 insertions(+), 47 deletions(-) diff --git a/kernel/sched/cpuacct.c b/kernel/sched/cpuacct.c index f347cf9e4634..9de7dd51beb0 100644 --- a/kernel/sched/cpuacct.c +++ b/kernel/sched/cpuacct.c @@ -21,15 +21,11 @@ static const char * const cpuacct_stat_desc[] = { [CPUACCT_STAT_SYSTEM] = "system", }; -struct cpuacct_usage { - u64 usages[CPUACCT_STAT_NSTATS]; -}; - /* track CPU usage of a group of tasks and its child groups */ struct cpuacct { struct cgroup_subsys_state css; /* cpuusage holds pointer to a u64-type object on every CPU */ - struct cpuacct_usage __percpu *cpuusage; + u64 __percpu *cpuusage; struct kernel_cpustat __percpu *cpustat; }; @@ -49,7 +45,7 @@ static inline struct cpuacct *parent_ca(struct cpuacct *ca) return css_ca(ca->css.parent); } -static DEFINE_PER_CPU(struct cpuacct_usage, root_cpuacct_cpuusage); +static DEFINE_PER_CPU(u64, root_cpuacct_cpuusage); static struct cpuacct root_cpuacct = { .cpustat = &kernel_cpustat, .cpuusage = &root_cpuacct_cpuusage, @@ -68,7 +64,7 @@ cpuacct_css_alloc(struct cgroup_subsys_state *parent_css) if (!ca) goto out; - ca->cpuusage = alloc_percpu(struct cpuacct_usage); + ca->cpuusage = alloc_percpu(u64); if (!ca->cpuusage) goto out_free_ca; @@ -99,7 +95,8 @@ static void cpuacct_css_free(struct cgroup_subsys_state *css) static u64 cpuacct_cpuusage_read(struct cpuacct *ca, int cpu, enum cpuacct_stat_index index) { - struct cpuacct_usage *cpuusage = per_cpu_ptr(ca->cpuusage, cpu); + u64 *cpuusage = per_cpu_ptr(ca->cpuusage, cpu); + u64 *cpustat = per_cpu_ptr(ca->cpustat, cpu)->cpustat; u64 data; /* @@ -116,14 +113,17 @@ static u64 cpuacct_cpuusage_read(struct cpuacct *ca, int cpu, raw_spin_rq_lock_irq(cpu_rq(cpu)); #endif - if (index == CPUACCT_STAT_NSTATS) { - int i = 0; - - data = 0; - for (i = 0; i < CPUACCT_STAT_NSTATS; i++) - data += cpuusage->usages[i]; - } else { - data = cpuusage->usages[index]; + switch (index) { + case CPUACCT_STAT_USER: + data = cpustat[CPUTIME_USER] + cpustat[CPUTIME_NICE]; + break; + case CPUACCT_STAT_SYSTEM: + data = cpustat[CPUTIME_SYSTEM] + cpustat[CPUTIME_IRQ] + + cpustat[CPUTIME_SOFTIRQ]; + break; + case CPUACCT_STAT_NSTATS: + data = *cpuusage; + break; } #ifndef CONFIG_64BIT @@ -133,10 +133,14 @@ static u64 cpuacct_cpuusage_read(struct cpuacct *ca, int cpu, return data; } -static void cpuacct_cpuusage_write(struct cpuacct *ca, int cpu, u64 val) +static void cpuacct_cpuusage_write(struct cpuacct *ca, int cpu) { - struct cpuacct_usage *cpuusage = per_cpu_ptr(ca->cpuusage, cpu); - int i; + u64 *cpuusage = per_cpu_ptr(ca->cpuusage, cpu); + u64 *cpustat = per_cpu_ptr(ca->cpustat, cpu)->cpustat; + + /* Don't allow to reset global kernel_cpustat */ + if (ca == &root_cpuacct) + return; #ifndef CONFIG_64BIT /* @@ -144,9 +148,10 @@ static void cpuacct_cpuusage_write(struct cpuacct *ca, int cpu, u64 val) */ raw_spin_rq_lock_irq(cpu_rq(cpu)); #endif - - for (i = 0; i < CPUACCT_STAT_NSTATS; i++) - cpuusage->usages[i] = val; + *cpuusage = 0; + cpustat[CPUTIME_USER] = cpustat[CPUTIME_NICE] = 0; + cpustat[CPUTIME_SYSTEM] = cpustat[CPUTIME_IRQ] = 0; + cpustat[CPUTIME_SOFTIRQ] = 0; #ifndef CONFIG_64BIT raw_spin_rq_unlock_irq(cpu_rq(cpu)); @@ -197,7 +202,7 @@ static int cpuusage_write(struct cgroup_subsys_state *css, struct cftype *cft, return -EINVAL; for_each_possible_cpu(cpu) - cpuacct_cpuusage_write(ca, cpu, 0); + cpuacct_cpuusage_write(ca, cpu); return 0; } @@ -244,25 +249,10 @@ static int cpuacct_all_seq_show(struct seq_file *m, void *V) seq_puts(m, "\n"); for_each_possible_cpu(cpu) { - struct cpuacct_usage *cpuusage = per_cpu_ptr(ca->cpuusage, cpu); - seq_printf(m, "%d", cpu); - - for (index = 0; index < CPUACCT_STAT_NSTATS; index++) { -#ifndef CONFIG_64BIT - /* - * Take rq->lock to make 64-bit read safe on 32-bit - * platforms. - */ - raw_spin_rq_lock_irq(cpu_rq(cpu)); -#endif - - seq_printf(m, " %llu", cpuusage->usages[index]); - -#ifndef CONFIG_64BIT - raw_spin_rq_unlock_irq(cpu_rq(cpu)); -#endif - } + for (index = 0; index < CPUACCT_STAT_NSTATS; index++) + seq_printf(m, " %llu", + cpuacct_cpuusage_read(ca, cpu, index)); seq_puts(m, "\n"); } return 0; @@ -340,16 +330,11 @@ static struct cftype files[] = { void cpuacct_charge(struct task_struct *tsk, u64 cputime) { struct cpuacct *ca; - int index = CPUACCT_STAT_SYSTEM; - struct pt_regs *regs = get_irq_regs() ? : task_pt_regs(tsk); - - if (regs && user_mode(regs)) - index = CPUACCT_STAT_USER; rcu_read_lock(); for (ca = task_ca(tsk); ca; ca = parent_ca(ca)) - __this_cpu_add(ca->cpuusage->usages[index], cputime); + __this_cpu_add(*ca->cpuusage, cputime); rcu_read_unlock(); } -- cgit v1.2.3-70-g09d2 From 8c92606ab81086db00cbb73347d124b4eb169b7e Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Mon, 15 Nov 2021 19:46:07 +0300 Subject: sched/cpuacct: Make user/system times in cpuacct.stat more precise cpuacct.stat shows user time based on raw random precision tick based counters. Use cputime_addjust() to scale these values against the total runtime accounted by the scheduler, like we already do for user/system times in /proc//stat. Signed-off-by: Andrey Ryabinin Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Daniel Jordan Acked-by: Tejun Heo Link: https://lore.kernel.org/r/20211115164607.23784-4-arbn@yandex-team.com --- kernel/sched/cpuacct.c | 25 +++++++++++++++---------- 1 file changed, 15 insertions(+), 10 deletions(-) diff --git a/kernel/sched/cpuacct.c b/kernel/sched/cpuacct.c index 9de7dd51beb0..3d06c5e4220d 100644 --- a/kernel/sched/cpuacct.c +++ b/kernel/sched/cpuacct.c @@ -261,25 +261,30 @@ static int cpuacct_all_seq_show(struct seq_file *m, void *V) static int cpuacct_stats_show(struct seq_file *sf, void *v) { struct cpuacct *ca = css_ca(seq_css(sf)); - s64 val[CPUACCT_STAT_NSTATS]; + struct task_cputime cputime; + u64 val[CPUACCT_STAT_NSTATS]; int cpu; int stat; - memset(val, 0, sizeof(val)); + memset(&cputime, 0, sizeof(cputime)); for_each_possible_cpu(cpu) { u64 *cpustat = per_cpu_ptr(ca->cpustat, cpu)->cpustat; - val[CPUACCT_STAT_USER] += cpustat[CPUTIME_USER]; - val[CPUACCT_STAT_USER] += cpustat[CPUTIME_NICE]; - val[CPUACCT_STAT_SYSTEM] += cpustat[CPUTIME_SYSTEM]; - val[CPUACCT_STAT_SYSTEM] += cpustat[CPUTIME_IRQ]; - val[CPUACCT_STAT_SYSTEM] += cpustat[CPUTIME_SOFTIRQ]; + cputime.utime += cpustat[CPUTIME_USER]; + cputime.utime += cpustat[CPUTIME_NICE]; + cputime.stime += cpustat[CPUTIME_SYSTEM]; + cputime.stime += cpustat[CPUTIME_IRQ]; + cputime.stime += cpustat[CPUTIME_SOFTIRQ]; + + cputime.sum_exec_runtime += *per_cpu_ptr(ca->cpuusage, cpu); } + cputime_adjust(&cputime, &seq_css(sf)->cgroup->prev_cputime, + &val[CPUACCT_STAT_USER], &val[CPUACCT_STAT_SYSTEM]); + for (stat = 0; stat < CPUACCT_STAT_NSTATS; stat++) { - seq_printf(sf, "%s %lld\n", - cpuacct_stat_desc[stat], - (long long)nsec_to_clock_t(val[stat])); + seq_printf(sf, "%s %llu\n", cpuacct_stat_desc[stat], + nsec_to_clock_t(val[stat])); } return 0; -- cgit v1.2.3-70-g09d2 From 8b4e74ccb582797f6f0b0a50372ebd9fd2372a27 Mon Sep 17 00:00:00 2001 From: Vincent Donnefort Date: Wed, 1 Dec 2021 14:34:50 +0000 Subject: sched/fair: Fix detection of per-CPU kthreads waking a task select_idle_sibling() has a special case for tasks woken up by a per-CPU kthread, where the selected CPU is the previous one. However, the current condition for this exit path is incomplete. A task can wake up from an interrupt context (e.g. hrtimer), while a per-CPU kthread is running. A such scenario would spuriously trigger the special case described above. Also, a recent change made the idle task like a regular per-CPU kthread, hence making that situation more likely to happen (is_per_cpu_kthread(swapper) being true now). Checking for task context makes sure select_idle_sibling() will not interpret a wake up from any other context as a wake up by a per-CPU kthread. Fixes: 52262ee567ad ("sched/fair: Allow a per-CPU kthread waking a task to stack on the same CPU, to fix XFS performance regression") Signed-off-by: Vincent Donnefort Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Vincent Guittot Reviewed-by: Valentin Schneider Link: https://lore.kernel.org/r/20211201143450.479472-1-vincent.donnefort@arm.com --- kernel/sched/fair.c | 1 + 1 file changed, 1 insertion(+) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 884f29d07963..5cd27986b43e 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -6398,6 +6398,7 @@ static int select_idle_sibling(struct task_struct *p, int prev, int target) * pattern is IO completions. */ if (is_per_cpu_kthread(current) && + in_task() && prev == smp_processor_id() && this_rq()->nr_running <= 1) { return prev; -- cgit v1.2.3-70-g09d2 From 014ba44e8184e1acf93e0cbb7089ee847802f8f0 Mon Sep 17 00:00:00 2001 From: Vincent Donnefort Date: Mon, 29 Nov 2021 17:31:15 +0000 Subject: sched/fair: Fix per-CPU kthread and wakee stacking for asym CPU capacity select_idle_sibling() has a special case for tasks woken up by a per-CPU kthread where the selected CPU is the previous one. For asymmetric CPU capacity systems, the assumption was that the wakee couldn't have a bigger utilization during task placement than it used to have during the last activation. That was not considering uclamp.min which can completely change between two task activations and as a consequence mandates the fitness criterion asym_fits_capacity(), even for the exit path described above. Fixes: b4c9c9f15649 ("sched/fair: Prefer prev cpu in asymmetric wakeup path") Signed-off-by: Vincent Donnefort Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Valentin Schneider Reviewed-by: Dietmar Eggemann Link: https://lkml.kernel.org/r/20211129173115.4006346-1-vincent.donnefort@arm.com --- kernel/sched/fair.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 5cd27986b43e..06722188df49 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -6400,7 +6400,8 @@ static int select_idle_sibling(struct task_struct *p, int prev, int target) if (is_per_cpu_kthread(current) && in_task() && prev == smp_processor_id() && - this_rq()->nr_running <= 1) { + this_rq()->nr_running <= 1 && + asym_fits_capacity(task_util, prev)) { return prev; } -- cgit v1.2.3-70-g09d2 From 2917406c352757642c3c1a13a4c99c96e6d22fde Mon Sep 17 00:00:00 2001 From: Barry Song Date: Sat, 16 Oct 2021 19:11:09 +0800 Subject: sched/fair: Document the slow path and fast path in select_task_rq_fair All People I know including myself took a long time to figure out that typical wakeup will always go to fast path and never go to slow path except WF_FORK and WF_EXEC. Vincent reminded me once in a linaro meeting and made me understand slow path won't happen for WF_TTWU. But my other friends repeatedly wasted a lot of time on testing this path like me before I reminded them. So obviously the code needs some document. Signed-off-by: Barry Song Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/20211016111109.5559-1-21cnbao@gmail.com --- kernel/sched/fair.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 06722188df49..f34f2f344fe9 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -6917,6 +6917,11 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int wake_flags) break; } + /* + * Usually only true for WF_EXEC and WF_FORK, as sched_domains + * usually do not have SD_BALANCE_WAKE set. That means wakeup + * will usually go to the fast path. + */ if (tmp->flags & sd_flag) sd = tmp; else if (!want_affine) -- cgit v1.2.3-70-g09d2 From 9b58e976b3b391c0cf02e038d53dd0478ed3013c Mon Sep 17 00:00:00 2001 From: Li Hua Date: Fri, 3 Dec 2021 03:36:18 +0000 Subject: sched/rt: Try to restart rt period timer when rt runtime exceeded When rt_runtime is modified from -1 to a valid control value, it may cause the task to be throttled all the time. Operations like the following will trigger the bug. E.g: 1. echo -1 > /proc/sys/kernel/sched_rt_runtime_us 2. Run a FIFO task named A that executes while(1) 3. echo 950000 > /proc/sys/kernel/sched_rt_runtime_us When rt_runtime is -1, The rt period timer will not be activated when task A enqueued. And then the task will be throttled after setting rt_runtime to 950,000. The task will always be throttled because the rt period timer is not activated. Fixes: d0b27fa77854 ("sched: rt-group: synchonised bandwidth period") Reported-by: Hulk Robot Signed-off-by: Li Hua Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/20211203033618.11895-1-hucool.lihua@huawei.com --- kernel/sched/rt.c | 23 ++++++++++++++++++----- 1 file changed, 18 insertions(+), 5 deletions(-) diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index b48baaba2fc2..7b4f4fbbb404 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -52,11 +52,8 @@ void init_rt_bandwidth(struct rt_bandwidth *rt_b, u64 period, u64 runtime) rt_b->rt_period_timer.function = sched_rt_period_timer; } -static void start_rt_bandwidth(struct rt_bandwidth *rt_b) +static inline void do_start_rt_bandwidth(struct rt_bandwidth *rt_b) { - if (!rt_bandwidth_enabled() || rt_b->rt_runtime == RUNTIME_INF) - return; - raw_spin_lock(&rt_b->rt_runtime_lock); if (!rt_b->rt_period_active) { rt_b->rt_period_active = 1; @@ -75,6 +72,14 @@ static void start_rt_bandwidth(struct rt_bandwidth *rt_b) raw_spin_unlock(&rt_b->rt_runtime_lock); } +static void start_rt_bandwidth(struct rt_bandwidth *rt_b) +{ + if (!rt_bandwidth_enabled() || rt_b->rt_runtime == RUNTIME_INF) + return; + + do_start_rt_bandwidth(rt_b); +} + void init_rt_rq(struct rt_rq *rt_rq) { struct rt_prio_array *array; @@ -1031,13 +1036,17 @@ static void update_curr_rt(struct rq *rq) for_each_sched_rt_entity(rt_se) { struct rt_rq *rt_rq = rt_rq_of_se(rt_se); + int exceeded; if (sched_rt_runtime(rt_rq) != RUNTIME_INF) { raw_spin_lock(&rt_rq->rt_runtime_lock); rt_rq->rt_time += delta_exec; - if (sched_rt_runtime_exceeded(rt_rq)) + exceeded = sched_rt_runtime_exceeded(rt_rq); + if (exceeded) resched_curr(rq); raw_spin_unlock(&rt_rq->rt_runtime_lock); + if (exceeded) + do_start_rt_bandwidth(sched_rt_bandwidth(rt_rq)); } } } @@ -2911,8 +2920,12 @@ static int sched_rt_global_validate(void) static void sched_rt_do_global(void) { + unsigned long flags; + + raw_spin_lock_irqsave(&def_rt_bandwidth.rt_runtime_lock, flags); def_rt_bandwidth.rt_runtime = global_rt_runtime(); def_rt_bandwidth.rt_period = ns_to_ktime(global_rt_period()); + raw_spin_unlock_irqrestore(&def_rt_bandwidth.rt_runtime_lock, flags); } int sched_rt_handler(struct ctl_table *table, int write, void *buffer, -- cgit v1.2.3-70-g09d2 From ef8df9798d469b7c45c66664550e93469749f1e8 Mon Sep 17 00:00:00 2001 From: Vincent Donnefort Date: Tue, 7 Dec 2021 09:57:55 +0000 Subject: sched/fair: Cleanup task_util and capacity type task_util and capacity are comparable unsigned long values. There is no need for an intermidiate implicit signed cast. Signed-off-by: Vincent Donnefort Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/20211207095755.859972-1-vincent.donnefort@arm.com --- kernel/sched/fair.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index f34f2f344fe9..ac5e55441cab 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -4070,7 +4070,8 @@ done: trace_sched_util_est_se_tp(&p->se); } -static inline int task_fits_capacity(struct task_struct *p, long capacity) +static inline int task_fits_capacity(struct task_struct *p, + unsigned long capacity) { return fits_capacity(uclamp_task_util(p), capacity); } @@ -6345,7 +6346,7 @@ select_idle_capacity(struct task_struct *p, struct sched_domain *sd, int target) return best_cpu; } -static inline bool asym_fits_capacity(int task_util, int cpu) +static inline bool asym_fits_capacity(unsigned long task_util, int cpu) { if (static_branch_unlikely(&sched_asym_cpucapacity)) return fits_capacity(task_util, capacity_of(cpu)); -- cgit v1.2.3-70-g09d2 From 82762d2af31a60081162890983a83499c9c7dd74 Mon Sep 17 00:00:00 2001 From: Dietmar Eggemann Date: Thu, 18 Nov 2021 17:42:40 +0100 Subject: sched/fair: Replace CFS internal cpu_util() with cpu_util_cfs() cpu_util_cfs() was created by commit d4edd662ac16 ("sched/cpufreq: Use the DEADLINE utilization signal") to enable the access to CPU utilization from the Schedutil CPUfreq governor. Commit a07630b8b2c1 ("sched/cpufreq/schedutil: Use util_est for OPP selection") added util_est support later. The only thing cpu_util() is doing on top of what cpu_util_cfs() already does is to clamp the return value to the [0..capacity_orig] capacity range of the CPU. Integrating this into cpu_util_cfs() is not harming the existing users (Schedutil and CPUfreq cooling (latter via sched_cpu_util() wrapper)). For straightforwardness, prefer to keep using `int cpu` as the function parameter over using `struct rq *rq` which might avoid some calls to cpu_rq(cpu) -> per_cpu(runqueues, cpu) -> RELOC_HIDE(). Update cfs_util()'s documentation and reuse it for cpu_util_cfs(). Remove cpu_util(). Signed-off-by: Dietmar Eggemann Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Vincent Guittot Link: https://lore.kernel.org/r/20211118164240.623551-1-dietmar.eggemann@arm.com --- kernel/sched/core.c | 2 +- kernel/sched/cpufreq_schedutil.c | 2 +- kernel/sched/fair.c | 71 +++++----------------------------------- kernel/sched/sched.h | 44 ++++++++++++++++++++++--- 4 files changed, 50 insertions(+), 69 deletions(-) diff --git a/kernel/sched/core.c b/kernel/sched/core.c index beaa8be6241e..fe53e510e711 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -7166,7 +7166,7 @@ unsigned long effective_cpu_util(int cpu, unsigned long util_cfs, unsigned long sched_cpu_util(int cpu, unsigned long max) { - return effective_cpu_util(cpu, cpu_util_cfs(cpu_rq(cpu)), max, + return effective_cpu_util(cpu, cpu_util_cfs(cpu), max, ENERGY_UTIL, NULL); } #endif /* CONFIG_SMP */ diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index e7af18857371..26778884d9ab 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -168,7 +168,7 @@ static void sugov_get_util(struct sugov_cpu *sg_cpu) sg_cpu->max = max; sg_cpu->bw_dl = cpu_bw_dl(rq); - sg_cpu->util = effective_cpu_util(sg_cpu->cpu, cpu_util_cfs(rq), max, + sg_cpu->util = effective_cpu_util(sg_cpu->cpu, cpu_util_cfs(sg_cpu->cpu), max, FREQUENCY_UTIL, NULL); } diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index ac5e55441cab..095b0aa378df 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -1502,7 +1502,6 @@ struct task_numa_env { static unsigned long cpu_load(struct rq *rq); static unsigned long cpu_runnable(struct rq *rq); -static unsigned long cpu_util(int cpu); static inline long adjust_numa_imbalance(int imbalance, int dst_running, int dst_weight); @@ -1569,7 +1568,7 @@ static void update_numa_stats(struct task_numa_env *env, ns->load += cpu_load(rq); ns->runnable += cpu_runnable(rq); - ns->util += cpu_util(cpu); + ns->util += cpu_util_cfs(cpu); ns->nr_running += rq->cfs.h_nr_running; ns->compute_capacity += capacity_of(cpu); @@ -3240,7 +3239,7 @@ static inline void cfs_rq_util_change(struct cfs_rq *cfs_rq, int flags) * As is, the util number is not freq-invariant (we'd have to * implement arch_scale_freq_capacity() for that). * - * See cpu_util(). + * See cpu_util_cfs(). */ cpufreq_update_util(rq, flags); } @@ -5510,11 +5509,9 @@ static inline void hrtick_update(struct rq *rq) #endif #ifdef CONFIG_SMP -static inline unsigned long cpu_util(int cpu); - static inline bool cpu_overutilized(int cpu) { - return !fits_capacity(cpu_util(cpu), capacity_of(cpu)); + return !fits_capacity(cpu_util_cfs(cpu), capacity_of(cpu)); } static inline void update_overutilized_status(struct rq *rq) @@ -6459,58 +6456,6 @@ static int select_idle_sibling(struct task_struct *p, int prev, int target) return target; } -/** - * cpu_util - Estimates the amount of capacity of a CPU used by CFS tasks. - * @cpu: the CPU to get the utilization of - * - * The unit of the return value must be the one of capacity so we can compare - * the utilization with the capacity of the CPU that is available for CFS task - * (ie cpu_capacity). - * - * cfs_rq.avg.util_avg is the sum of running time of runnable tasks plus the - * recent utilization of currently non-runnable tasks on a CPU. It represents - * the amount of utilization of a CPU in the range [0..capacity_orig] where - * capacity_orig is the cpu_capacity available at the highest frequency - * (arch_scale_freq_capacity()). - * The utilization of a CPU converges towards a sum equal to or less than the - * current capacity (capacity_curr <= capacity_orig) of the CPU because it is - * the running time on this CPU scaled by capacity_curr. - * - * The estimated utilization of a CPU is defined to be the maximum between its - * cfs_rq.avg.util_avg and the sum of the estimated utilization of the tasks - * currently RUNNABLE on that CPU. - * This allows to properly represent the expected utilization of a CPU which - * has just got a big task running since a long sleep period. At the same time - * however it preserves the benefits of the "blocked utilization" in - * describing the potential for other tasks waking up on the same CPU. - * - * Nevertheless, cfs_rq.avg.util_avg can be higher than capacity_curr or even - * higher than capacity_orig because of unfortunate rounding in - * cfs.avg.util_avg or just after migrating tasks and new task wakeups until - * the average stabilizes with the new running time. We need to check that the - * utilization stays within the range of [0..capacity_orig] and cap it if - * necessary. Without utilization capping, a group could be seen as overloaded - * (CPU0 utilization at 121% + CPU1 utilization at 80%) whereas CPU1 has 20% of - * available capacity. We allow utilization to overshoot capacity_curr (but not - * capacity_orig) as it useful for predicting the capacity required after task - * migrations (scheduler-driven DVFS). - * - * Return: the (estimated) utilization for the specified CPU - */ -static inline unsigned long cpu_util(int cpu) -{ - struct cfs_rq *cfs_rq; - unsigned int util; - - cfs_rq = &cpu_rq(cpu)->cfs; - util = READ_ONCE(cfs_rq->avg.util_avg); - - if (sched_feat(UTIL_EST)) - util = max(util, READ_ONCE(cfs_rq->avg.util_est.enqueued)); - - return min_t(unsigned long, util, capacity_orig_of(cpu)); -} - /* * cpu_util_without: compute cpu utilization without any contributions from *p * @cpu: the CPU which utilization is requested @@ -6531,7 +6476,7 @@ static unsigned long cpu_util_without(int cpu, struct task_struct *p) /* Task has no contribution or is new */ if (cpu != task_cpu(p) || !READ_ONCE(p->se.avg.last_update_time)) - return cpu_util(cpu); + return cpu_util_cfs(cpu); cfs_rq = &cpu_rq(cpu)->cfs; util = READ_ONCE(cfs_rq->avg.util_avg); @@ -6595,7 +6540,7 @@ static unsigned long cpu_util_without(int cpu, struct task_struct *p) /* * Utilization (estimated) can exceed the CPU capacity, thus let's * clamp to the maximum CPU capacity to ensure consistency with - * the cpu_util call. + * cpu_util. */ return min_t(unsigned long, util, capacity_orig_of(cpu)); } @@ -6627,7 +6572,7 @@ static unsigned long cpu_util_next(int cpu, struct task_struct *p, int dst_cpu) * During wake-up, the task isn't enqueued yet and doesn't * appear in the cfs_rq->avg.util_est.enqueued of any rq, * so just add it (if needed) to "simulate" what will be - * cpu_util() after the task has been enqueued. + * cpu_util after the task has been enqueued. */ if (dst_cpu == cpu) util_est += _task_util_est(p); @@ -8689,7 +8634,7 @@ static inline void update_sg_lb_stats(struct lb_env *env, struct rq *rq = cpu_rq(i); sgs->group_load += cpu_load(rq); - sgs->group_util += cpu_util(i); + sgs->group_util += cpu_util_cfs(i); sgs->group_runnable += cpu_runnable(rq); sgs->sum_h_nr_running += rq->cfs.h_nr_running; @@ -9707,7 +9652,7 @@ static struct rq *find_busiest_queue(struct lb_env *env, break; case migrate_util: - util = cpu_util(cpu_of(rq)); + util = cpu_util_cfs(i); /* * Don't try to pull utilization from a CPU with one diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index eb971151e7e4..de53be905739 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -2966,16 +2966,52 @@ static inline unsigned long cpu_util_dl(struct rq *rq) return READ_ONCE(rq->avg_dl.util_avg); } -static inline unsigned long cpu_util_cfs(struct rq *rq) +/** + * cpu_util_cfs() - Estimates the amount of CPU capacity used by CFS tasks. + * @cpu: the CPU to get the utilization for. + * + * The unit of the return value must be the same as the one of CPU capacity + * so that CPU utilization can be compared with CPU capacity. + * + * CPU utilization is the sum of running time of runnable tasks plus the + * recent utilization of currently non-runnable tasks on that CPU. + * It represents the amount of CPU capacity currently used by CFS tasks in + * the range [0..max CPU capacity] with max CPU capacity being the CPU + * capacity at f_max. + * + * The estimated CPU utilization is defined as the maximum between CPU + * utilization and sum of the estimated utilization of the currently + * runnable tasks on that CPU. It preserves a utilization "snapshot" of + * previously-executed tasks, which helps better deduce how busy a CPU will + * be when a long-sleeping task wakes up. The contribution to CPU utilization + * of such a task would be significantly decayed at this point of time. + * + * CPU utilization can be higher than the current CPU capacity + * (f_curr/f_max * max CPU capacity) or even the max CPU capacity because + * of rounding errors as well as task migrations or wakeups of new tasks. + * CPU utilization has to be capped to fit into the [0..max CPU capacity] + * range. Otherwise a group of CPUs (CPU0 util = 121% + CPU1 util = 80%) + * could be seen as over-utilized even though CPU1 has 20% of spare CPU + * capacity. CPU utilization is allowed to overshoot current CPU capacity + * though since this is useful for predicting the CPU capacity required + * after task migrations (scheduler-driven DVFS). + * + * Return: (Estimated) utilization for the specified CPU. + */ +static inline unsigned long cpu_util_cfs(int cpu) { - unsigned long util = READ_ONCE(rq->cfs.avg.util_avg); + struct cfs_rq *cfs_rq; + unsigned long util; + + cfs_rq = &cpu_rq(cpu)->cfs; + util = READ_ONCE(cfs_rq->avg.util_avg); if (sched_feat(UTIL_EST)) { util = max_t(unsigned long, util, - READ_ONCE(rq->cfs.avg.util_est.enqueued)); + READ_ONCE(cfs_rq->avg.util_est.enqueued)); } - return util; + return min(util, capacity_orig_of(cpu)); } static inline unsigned long cpu_util_rt(struct rq *rq) -- cgit v1.2.3-70-g09d2