diff options
Diffstat (limited to 'kernel')
58 files changed, 4061 insertions, 1170 deletions
diff --git a/kernel/bpf/inode.c b/kernel/bpf/inode.c index f2ece3c..8f94ca1 100644 --- a/kernel/bpf/inode.c +++ b/kernel/bpf/inode.c @@ -31,10 +31,10 @@ static void *bpf_any_get(void *raw, enum bpf_type type) { switch (type) { case BPF_TYPE_PROG: - atomic_inc(&((struct bpf_prog *)raw)->aux->refcnt); + raw = bpf_prog_inc(raw); break; case BPF_TYPE_MAP: - bpf_map_inc(raw, true); + raw = bpf_map_inc(raw, true); break; default: WARN_ON_ONCE(1); @@ -297,7 +297,8 @@ static void *bpf_obj_do_get(const struct filename *pathname, goto out; raw = bpf_any_get(inode->i_private, *type); - touch_atime(&path); + if (!IS_ERR(raw)) + touch_atime(&path); path_put(&path); return raw; diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c index 499d9e9..f5a1954 100644 --- a/kernel/bpf/stackmap.c +++ b/kernel/bpf/stackmap.c @@ -66,7 +66,7 @@ static struct bpf_map *stack_map_alloc(union bpf_attr *attr) /* check sanity of attributes */ if (attr->max_entries == 0 || attr->key_size != 4 || value_size < 8 || value_size % 8 || - value_size / 8 > PERF_MAX_STACK_DEPTH) + value_size / 8 > sysctl_perf_event_max_stack) return ERR_PTR(-EINVAL); /* hash table size must be power of 2 */ @@ -124,8 +124,8 @@ static u64 bpf_get_stackid(u64 r1, u64 r2, u64 flags, u64 r4, u64 r5) struct perf_callchain_entry *trace; struct stack_map_bucket *bucket, *new_bucket, *old_bucket; u32 max_depth = map->value_size / 8; - /* stack_map_alloc() checks that max_depth <= PERF_MAX_STACK_DEPTH */ - u32 init_nr = PERF_MAX_STACK_DEPTH - max_depth; + /* stack_map_alloc() checks that max_depth <= sysctl_perf_event_max_stack */ + u32 init_nr = sysctl_perf_event_max_stack - max_depth; u32 skip = flags & BPF_F_SKIP_FIELD_MASK; u32 hash, id, trace_nr, trace_len; bool user = flags & BPF_F_USER_STACK; @@ -143,7 +143,7 @@ static u64 bpf_get_stackid(u64 r1, u64 r2, u64 flags, u64 r4, u64 r5) return -EFAULT; /* get_perf_callchain() guarantees that trace->nr >= init_nr - * and trace-nr <= PERF_MAX_STACK_DEPTH, so trace_nr <= max_depth + * and trace-nr <= sysctl_perf_event_max_stack, so trace_nr <= max_depth */ trace_nr = trace->nr - init_nr; diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index adc5e4b..cf5e9f7 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -218,11 +218,18 @@ struct bpf_map *__bpf_map_get(struct fd f) return f.file->private_data; } -void bpf_map_inc(struct bpf_map *map, bool uref) +/* prog's and map's refcnt limit */ +#define BPF_MAX_REFCNT 32768 + +struct bpf_map *bpf_map_inc(struct bpf_map *map, bool uref) { - atomic_inc(&map->refcnt); + if (atomic_inc_return(&map->refcnt) > BPF_MAX_REFCNT) { + atomic_dec(&map->refcnt); + return ERR_PTR(-EBUSY); + } if (uref) atomic_inc(&map->usercnt); + return map; } struct bpf_map *bpf_map_get_with_uref(u32 ufd) @@ -234,7 +241,7 @@ struct bpf_map *bpf_map_get_with_uref(u32 ufd) if (IS_ERR(map)) return map; - bpf_map_inc(map, true); + map = bpf_map_inc(map, true); fdput(f); return map; @@ -658,6 +665,15 @@ static struct bpf_prog *__bpf_prog_get(struct fd f) return f.file->private_data; } +struct bpf_prog *bpf_prog_inc(struct bpf_prog *prog) +{ + if (atomic_inc_return(&prog->aux->refcnt) > BPF_MAX_REFCNT) { + atomic_dec(&prog->aux->refcnt); + return ERR_PTR(-EBUSY); + } + return prog; +} + /* called by sockets/tracing/seccomp before attaching program to an event * pairs with bpf_prog_put() */ @@ -670,7 +686,7 @@ struct bpf_prog *bpf_prog_get(u32 ufd) if (IS_ERR(prog)) return prog; - atomic_inc(&prog->aux->refcnt); + prog = bpf_prog_inc(prog); fdput(f); return prog; diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 2e08f8e..c5c17a6 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -239,16 +239,6 @@ static const char * const reg_type_str[] = { [CONST_IMM] = "imm", }; -static const struct { - int map_type; - int func_id; -} func_limit[] = { - {BPF_MAP_TYPE_PROG_ARRAY, BPF_FUNC_tail_call}, - {BPF_MAP_TYPE_PERF_EVENT_ARRAY, BPF_FUNC_perf_event_read}, - {BPF_MAP_TYPE_PERF_EVENT_ARRAY, BPF_FUNC_perf_event_output}, - {BPF_MAP_TYPE_STACK_TRACE, BPF_FUNC_get_stackid}, -}; - static void print_verifier_state(struct verifier_env *env) { enum bpf_reg_type t; @@ -921,27 +911,52 @@ static int check_func_arg(struct verifier_env *env, u32 regno, static int check_map_func_compatibility(struct bpf_map *map, int func_id) { - bool bool_map, bool_func; - int i; - if (!map) return 0; - for (i = 0; i < ARRAY_SIZE(func_limit); i++) { - bool_map = (map->map_type == func_limit[i].map_type); - bool_func = (func_id == func_limit[i].func_id); - /* only when map & func pair match it can continue. - * don't allow any other map type to be passed into - * the special func; - */ - if (bool_func && bool_map != bool_func) { - verbose("cannot pass map_type %d into func %d\n", - map->map_type, func_id); - return -EINVAL; - } + /* We need a two way check, first is from map perspective ... */ + switch (map->map_type) { + case BPF_MAP_TYPE_PROG_ARRAY: + if (func_id != BPF_FUNC_tail_call) + goto error; + break; + case BPF_MAP_TYPE_PERF_EVENT_ARRAY: + if (func_id != BPF_FUNC_perf_event_read && + func_id != BPF_FUNC_perf_event_output) + goto error; + break; + case BPF_MAP_TYPE_STACK_TRACE: + if (func_id != BPF_FUNC_get_stackid) + goto error; + break; + default: + break; + } + + /* ... and second from the function itself. */ + switch (func_id) { + case BPF_FUNC_tail_call: + if (map->map_type != BPF_MAP_TYPE_PROG_ARRAY) + goto error; + break; + case BPF_FUNC_perf_event_read: + case BPF_FUNC_perf_event_output: + if (map->map_type != BPF_MAP_TYPE_PERF_EVENT_ARRAY) + goto error; + break; + case BPF_FUNC_get_stackid: + if (map->map_type != BPF_MAP_TYPE_STACK_TRACE) + goto error; + break; + default: + break; } return 0; +error: + verbose("cannot pass map_type %d into func %d\n", + map->map_type, func_id); + return -EINVAL; } static int check_call(struct verifier_env *env, int func_id) @@ -1374,6 +1389,7 @@ static int check_ld_abs(struct verifier_env *env, struct bpf_insn *insn) } if (insn->dst_reg != BPF_REG_0 || insn->off != 0 || + BPF_SIZE(insn->code) == BPF_DW || (mode == BPF_ABS && insn->src_reg != BPF_REG_0)) { verbose("BPF_LD_ABS uses reserved fields\n"); return -EINVAL; @@ -2029,7 +2045,6 @@ static int replace_map_fd_with_map_ptr(struct verifier_env *env) if (IS_ERR(map)) { verbose("fd %d is not pointing to valid bpf_map\n", insn->imm); - fdput(f); return PTR_ERR(map); } @@ -2049,15 +2064,18 @@ static int replace_map_fd_with_map_ptr(struct verifier_env *env) return -E2BIG; } - /* remember this map */ - env->used_maps[env->used_map_cnt++] = map; - /* hold the map. If the program is rejected by verifier, * the map will be released by release_maps() or it * will be used by the valid program until it's unloaded * and all maps are released in free_bpf_prog_info() */ - bpf_map_inc(map, false); + map = bpf_map_inc(map, false); + if (IS_ERR(map)) { + fdput(f); + return PTR_ERR(map); + } + env->used_maps[env->used_map_cnt++] = map; + fdput(f); next_insn: insn++; diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 671dc05..86cb5c6 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -1215,6 +1215,41 @@ static void cgroup_destroy_root(struct cgroup_root *root) cgroup_free_root(root); } +/* + * look up cgroup associated with current task's cgroup namespace on the + * specified hierarchy + */ +static struct cgroup * +current_cgns_cgroup_from_root(struct cgroup_root *root) +{ + struct cgroup *res = NULL; + struct css_set *cset; + + lockdep_assert_held(&css_set_lock); + + rcu_read_lock(); + + cset = current->nsproxy->cgroup_ns->root_cset; + if (cset == &init_css_set) { + res = &root->cgrp; + } else { + struct cgrp_cset_link *link; + + list_for_each_entry(link, &cset->cgrp_links, cgrp_link) { + struct cgroup *c = link->cgrp; + + if (c->root == root) { + res = c; + break; + } + } + } + rcu_read_unlock(); + + BUG_ON(!res); + return res; +} + /* look up cgroup associated with given css_set on the specified hierarchy */ static struct cgroup *cset_cgroup_from_root(struct css_set *cset, struct cgroup_root *root) @@ -1593,6 +1628,33 @@ static int rebind_subsystems(struct cgroup_root *dst_root, u16 ss_mask) return 0; } +static int cgroup_show_path(struct seq_file *sf, struct kernfs_node *kf_node, + struct kernfs_root *kf_root) +{ + int len = 0; + char *buf = NULL; + struct cgroup_root *kf_cgroot = cgroup_root_from_kf(kf_root); + struct cgroup *ns_cgroup; + + buf = kmalloc(PATH_MAX, GFP_KERNEL); + if (!buf) + return -ENOMEM; + + spin_lock_bh(&css_set_lock); + ns_cgroup = current_cgns_cgroup_from_root(kf_cgroot); + len = kernfs_path_from_node(kf_node, ns_cgroup->kn, buf, PATH_MAX); + spin_unlock_bh(&css_set_lock); + + if (len >= PATH_MAX) + len = -ERANGE; + else if (len > 0) { + seq_escape(sf, buf, " \t\n\\"); + len = 0; + } + kfree(buf); + return len; +} + static int cgroup_show_options(struct seq_file *seq, struct kernfs_root *kf_root) { @@ -2825,9 +2887,10 @@ static ssize_t __cgroup_procs_write(struct kernfs_open_file *of, char *buf, size_t nbytes, loff_t off, bool threadgroup) { struct task_struct *tsk; + struct cgroup_subsys *ss; struct cgroup *cgrp; pid_t pid; - int ret; + int ssid, ret; if (kstrtoint(strstrip(buf), 0, &pid) || pid < 0) return -EINVAL; @@ -2875,8 +2938,10 @@ out_unlock_rcu: rcu_read_unlock(); out_unlock_threadgroup: percpu_up_write(&cgroup_threadgroup_rwsem); + for_each_subsys(ss, ssid) + if (ss->post_attach) + ss->post_attach(); cgroup_kn_unlock(of->kn); - cpuset_post_attach_flush(); return ret ?: nbytes; } @@ -5430,6 +5495,7 @@ static struct kernfs_syscall_ops cgroup_kf_syscall_ops = { .mkdir = cgroup_mkdir, .rmdir = cgroup_rmdir, .rename = cgroup_rename, + .show_path = cgroup_show_path, }; static void __init cgroup_init_subsys(struct cgroup_subsys *ss, bool early) diff --git a/kernel/cpu.c b/kernel/cpu.c index 6ea42e8..d948e44 100644 --- a/kernel/cpu.c +++ b/kernel/cpu.c @@ -36,6 +36,7 @@ * @target: The target state * @thread: Pointer to the hotplug thread * @should_run: Thread should execute + * @rollback: Perform a rollback * @cb_stat: The state for a single callback (install/uninstall) * @cb: Single callback function (install/uninstall) * @result: Result of the operation @@ -47,6 +48,7 @@ struct cpuhp_cpu_state { #ifdef CONFIG_SMP struct task_struct *thread; bool should_run; + bool rollback; enum cpuhp_state cb_state; int (*cb)(unsigned int cpu); int result; @@ -301,6 +303,11 @@ static int cpu_notify(unsigned long val, unsigned int cpu) return __cpu_notify(val, cpu, -1, NULL); } +static void cpu_notify_nofail(unsigned long val, unsigned int cpu) +{ + BUG_ON(cpu_notify(val, cpu)); +} + /* Notifier wrappers for transitioning to state machine */ static int notify_prepare(unsigned int cpu) { @@ -477,6 +484,16 @@ static void cpuhp_thread_fun(unsigned int cpu) } else { ret = cpuhp_invoke_callback(cpu, st->cb_state, st->cb); } + } else if (st->rollback) { + BUG_ON(st->state < CPUHP_AP_ONLINE_IDLE); + + undo_cpu_down(cpu, st, cpuhp_ap_states); + /* + * This is a momentary workaround to keep the notifier users + * happy. Will go away once we got rid of the notifiers. + */ + cpu_notify_nofail(CPU_DOWN_FAILED, cpu); + st->rollback = false; } else { /* Cannot happen .... */ BUG_ON(st->state < CPUHP_AP_ONLINE_IDLE); @@ -636,11 +653,6 @@ static inline void check_for_tasks(int dead_cpu) read_unlock(&tasklist_lock); } -static void cpu_notify_nofail(unsigned long val, unsigned int cpu) -{ - BUG_ON(cpu_notify(val, cpu)); -} - static int notify_down_prepare(unsigned int cpu) { int err, nr_calls = 0; @@ -691,21 +703,6 @@ static int takedown_cpu(unsigned int cpu) struct cpuhp_cpu_state *st = per_cpu_ptr(&cpuhp_state, cpu); int err; - /* - * By now we've cleared cpu_active_mask, wait for all preempt-disabled - * and RCU users of this state to go away such that all new such users - * will observe it. - * - * For CONFIG_PREEMPT we have preemptible RCU and its sync_rcu() might - * not imply sync_sched(), so wait for both. - * - * Do sync before park smpboot threads to take care the rcu boost case. - */ - if (IS_ENABLED(CONFIG_PREEMPT)) - synchronize_rcu_mult(call_rcu, call_rcu_sched); - else - synchronize_rcu(); - /* Park the smpboot threads */ kthread_park(per_cpu_ptr(&cpuhp_state, cpu)->thread); smpboot_park_threads(cpu); @@ -721,9 +718,10 @@ static int takedown_cpu(unsigned int cpu) */ err = stop_machine(take_cpu_down, NULL, cpumask_of(cpu)); if (err) { - /* CPU didn't die: tell everyone. Can't complain. */ - cpu_notify_nofail(CPU_DOWN_FAILED, cpu); + /* CPU refused to die */ irq_unlock_sparse(); + /* Unpark the hotplug thread so we can rollback there */ + kthread_unpark(per_cpu_ptr(&cpuhp_state, cpu)->thread); return err; } BUG_ON(cpu_online(cpu)); @@ -832,6 +830,11 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen, * to do the further cleanups. */ ret = cpuhp_down_callbacks(cpu, st, cpuhp_bp_states, target); + if (ret && st->state > CPUHP_TEARDOWN_CPU && st->state < prev_state) { + st->target = prev_state; + st->rollback = true; + cpuhp_kick_ap_work(cpu); + } hasdied = prev_state != st->state && st->state == CPUHP_OFFLINE; out: @@ -905,8 +908,6 @@ void cpuhp_online_idle(enum cpuhp_state state) st->state = CPUHP_AP_ONLINE_IDLE; - /* The cpu is marked online, set it active now */ - set_cpu_active(cpu, true); /* Unpark the stopper thread and the hotplug thread of this cpu */ stop_machine_unpark(cpu); kthread_unpark(st->thread); @@ -1218,6 +1219,12 @@ static struct cpuhp_step cpuhp_ap_states[] = { .name = "ap:offline", .cant_stop = true, }, + /* First state is scheduler control. Interrupts are disabled */ + [CPUHP_AP_SCHED_STARTING] = { + .name = "sched:starting", + .startup = sched_cpu_starting, + .teardown = sched_cpu_dying, + }, /* * Low level startup/teardown notifiers. Run with interrupts * disabled. Will be removed once the notifiers are converted to @@ -1249,12 +1256,22 @@ static struct cpuhp_step cpuhp_ap_states[] = { .name = "notify:online", .startup = notify_online, .teardown = notify_down_prepare, + .skip_onerr = true, }, #endif /* * The dynamically registered state space is here */ +#ifdef CONFIG_SMP + /* Last state is scheduler control setting the cpu active */ + [CPUHP_AP_ACTIVE] = { + .name = "sched:active", + .startup = sched_cpu_activate, + .teardown = sched_cpu_deactivate, + }, +#endif + /* CPU is fully up and running. */ [CPUHP_ONLINE] = { .name = "online", diff --git a/kernel/cpuset.c b/kernel/cpuset.c index 00ab5c2..1902956 100644 --- a/kernel/cpuset.c +++ b/kernel/cpuset.c @@ -58,7 +58,6 @@ #include <asm/uaccess.h> #include <linux/atomic.h> #include <linux/mutex.h> -#include <linux/workqueue.h> #include <linux/cgroup.h> #include <linux/wait.h> @@ -1016,7 +1015,7 @@ static void cpuset_migrate_mm(struct mm_struct *mm, const nodemask_t *from, } } -void cpuset_post_attach_flush(void) +static void cpuset_post_attach(void) { flush_workqueue(cpuset_migrate_mm_wq); } @@ -2087,6 +2086,7 @@ struct cgroup_subsys cpuset_cgrp_subsys = { .can_attach = cpuset_can_attach, .cancel_attach = cpuset_cancel_attach, .attach = cpuset_attach, + .post_attach = cpuset_post_attach, .bind = cpuset_bind, .legacy_cftypes = files, .early_init = true, diff --git a/kernel/events/callchain.c b/kernel/events/callchain.c index 343c22f..b9325e7 100644 --- a/kernel/events/callchain.c +++ b/kernel/events/callchain.c @@ -18,6 +18,14 @@ struct callchain_cpus_entries { struct perf_callchain_entry *cpu_entries[0]; }; +int sysctl_perf_event_max_stack __read_mostly = PERF_MAX_STACK_DEPTH; + +static inline size_t perf_callchain_entry__sizeof(void) +{ + return (sizeof(struct perf_callchain_entry) + + sizeof(__u64) * sysctl_perf_event_max_stack); +} + static DEFINE_PER_CPU(int, callchain_recursion[PERF_NR_CONTEXTS]); static atomic_t nr_callchain_events; static DEFINE_MUTEX(callchain_mutex); @@ -73,7 +81,7 @@ static int alloc_callchain_buffers(void) if (!entries) return -ENOMEM; - size = sizeof(struct perf_callchain_entry) * PERF_NR_CONTEXTS; + size = perf_callchain_entry__sizeof() * PERF_NR_CONTEXTS; for_each_possible_cpu(cpu) { entries->cpu_entries[cpu] = kmalloc_node(size, GFP_KERNEL, @@ -147,7 +155,8 @@ static struct perf_callchain_entry *get_callchain_entry(int *rctx) cpu = smp_processor_id(); - return &entries->cpu_entries[cpu][*rctx]; + return (((void *)entries->cpu_entries[cpu]) + + (*rctx * perf_callchain_entry__sizeof())); } static void @@ -215,3 +224,25 @@ exit_put: return entry; } + +int perf_event_max_stack_handler(struct ctl_table *table, int write, + void __user *buffer, size_t *lenp, loff_t *ppos) +{ + int new_value = sysctl_perf_event_max_stack, ret; + struct ctl_table new_table = *table; + + new_table.data = &new_value; + ret = proc_dointvec_minmax(&new_table, write, buffer, lenp, ppos); + if (ret || !write) + return ret; + + mutex_lock(&callchain_mutex); + if (atomic_read(&nr_callchain_events)) + ret = -EBUSY; + else + sysctl_perf_event_max_stack = new_value; + + mutex_unlock(&callchain_mutex); + + return ret; +} diff --git a/kernel/events/core.c b/kernel/events/core.c index 52bedc5..050a290 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -44,6 +44,8 @@ #include <linux/compat.h> #include <linux/bpf.h> #include <linux/filter.h> +#include <linux/namei.h> +#include <linux/parser.h> #include "internal.h" @@ -351,7 +353,7 @@ static struct srcu_struct pmus_srcu; * 1 - disallow cpu events for unpriv * 2 - disallow kernel profiling for unpriv */ -int sysctl_perf_event_paranoid __read_mostly = 1; +int sysctl_perf_event_paranoid __read_mostly = 2; /* Minimum for 512 kiB + 1 user control page */ int sysctl_perf_event_mlock __read_mostly = 512 + (PAGE_SIZE / 1024); /* 'free' kiB per user */ @@ -412,7 +414,8 @@ int perf_cpu_time_max_percent_handler(struct ctl_table *table, int write, if (ret || !write) return ret; - if (sysctl_perf_cpu_time_max_percent == 100) { + if (sysctl_perf_cpu_time_max_percent == 100 || + sysctl_perf_cpu_time_max_percent == 0) { printk(KERN_WARNING "perf: Dynamic interrupt throttling disabled, can hang your system!\n"); WRITE_ONCE(perf_sample_allowed_ns, 0); @@ -1105,6 +1108,7 @@ static void put_ctx(struct perf_event_context *ctx) * function. * * Lock order: + * cred_guard_mutex * task_struct::perf_event_mutex * perf_event_context::mutex * perf_event::child_mutex; @@ -1925,8 +1929,13 @@ event_sched_in(struct perf_event *event, if (event->state <= PERF_EVENT_STATE_OFF) return 0; - event->state = PERF_EVENT_STATE_ACTIVE; - event->oncpu = smp_processor_id(); + WRITE_ONCE(event->oncpu, smp_processor_id()); + /* + * Order event::oncpu write to happen before the ACTIVE state + * is visible. + */ + smp_wmb(); + WRITE_ONCE(event->state, PERF_EVENT_STATE_ACTIVE); /* * Unthrottle events, since we scheduled we might have missed several @@ -2358,6 +2367,112 @@ void perf_event_enable(struct perf_event *event) } EXPORT_SYMBOL_GPL(perf_event_enable); +struct stop_event_data { + struct perf_event *event; + unsigned int restart; +}; + +static int __perf_event_stop(void *info) +{ + struct stop_event_data *sd = info; + struct perf_event *event = sd->event; + + /* if it's already INACTIVE, do nothing */ + if (READ_ONCE(event->state) != PERF_EVENT_STATE_ACTIVE) + return 0; + + /* matches smp_wmb() in event_sched_in() */ + smp_rmb(); + + /* + * There is a window with interrupts enabled before we get here, + * so we need to check again lest we try to stop another CPU's event. + */ + if (READ_ONCE(event->oncpu) != smp_processor_id()) + return -EAGAIN; + + event->pmu->stop(event, PERF_EF_UPDATE); + + /* + * May race with the actual stop (through perf_pmu_output_stop()), + * but it is only used for events with AUX ring buffer, and such + * events will refuse to restart because of rb::aux_mmap_count==0, + * see comments in perf_aux_output_begin(). + * + * Since this is happening on a event-local CPU, no trace is lost + * while restarting. + */ + if (sd->restart) + event->pmu->start(event, PERF_EF_START); + + return 0; +} + +static int perf_event_restart(struct perf_event *event) +{ + struct stop_event_data sd = { + .event = event, + .restart = 1, + }; + int ret = 0; + + do { + if (READ_ONCE(event->state) != PERF_EVENT_STATE_ACTIVE) + return 0; + + /* matches smp_wmb() in event_sched_in() */ + smp_rmb(); + + /* + * We only want to restart ACTIVE events, so if the event goes + * inactive here (event->oncpu==-1), there's nothing more to do; + * fall through with ret==-ENXIO. + */ + ret = cpu_function_call(READ_ONCE(event->oncpu), + __perf_event_stop, &sd); + } while (ret == -EAGAIN); + + return ret; +} + +/* + * In order to contain the amount of racy and tricky in the address filter + * configuration management, it is a two part process: + * + * (p1) when userspace mappings change as a result of (1) or (2) or (3) below, + * we update the addresses of corresponding vmas in + * event::addr_filters_offs array and bump the event::addr_filters_gen; + * (p2) when an event is scheduled in (pmu::add), it calls + * perf_event_addr_filters_sync() which calls pmu::addr_filters_sync() + * if the generation has changed since the previous call. + * + * If (p1) happens while the event is active, we restart it to force (p2). + * + * (1) perf_addr_filters_apply(): adjusting filters' offsets based on + * pre-existing mappings, called once when new filters arrive via SET_FILTER + * ioctl; + * (2) perf_addr_filters_adjust(): adjusting filters' offsets based on newly + * registered mapping, called for every new mmap(), with mm::mmap_sem down + * for reading; + * (3) perf_event_addr_filters_exec(): clearing filters' offsets in the process + * of exec. + */ +void perf_event_addr_filters_sync(struct perf_event *event) +{ + struct perf_addr_filters_head *ifh = perf_event_addr_filters(event); + + if (!has_addr_filter(event)) + return; + + raw_spin_lock(&ifh->lock); + if (event->addr_filters_gen != event->hw.addr_filters_gen) { + event->pmu->addr_filters_sync(event); + event->hw.addr_filters_gen = event->addr_filters_gen; + } + raw_spin_unlock(&ifh->lock); +} +EXPORT_SYMBOL_GPL(perf_event_addr_filters_sync); + static int _perf_event_refresh(struct perf_event *event, int refresh) { /* @@ -3207,16 +3322,6 @@ out: put_ctx(clone_ctx); } -void perf_event_exec(void) -{ - int ctxn; - - rcu_read_lock(); - for_each_task_context_nr(ctxn) - perf_event_enable_on_exec(ctxn); - rcu_read_unlock(); -} - struct perf_read_data { struct perf_event *event; bool group; @@ -3420,7 +3525,6 @@ static struct task_struct * find_lively_task_by_vpid(pid_t vpid) { struct task_struct *task; - int err; rcu_read_lock(); if (!vpid) @@ -3434,16 +3538,7 @@ find_lively_task_by_vpid(pid_t vpid) if (!task) return ERR_PTR(-ESRCH); - /* Reuse ptrace permission checks for now. */ - err = -EACCES; - if (!ptrace_may_access(task, PTRACE_MODE_READ_REALCREDS)) - goto errout; - return task; -errout: - put_task_struct(task); - return ERR_PTR(err); - } /* @@ -3728,6 +3823,9 @@ static bool exclusive_event_installable(struct perf_event *event, return true; } +static void perf_addr_filters_splice(struct perf_event *event, + struct list_head *head); + static void _free_event(struct perf_event *event) { irq_work_sync(&event->pending); @@ -3755,6 +3853,8 @@ static void _free_event(struct perf_event *event) } perf_event_free_bpf_prog(event); + perf_addr_filters_splice(event, NULL); + kfree(event->addr_filters_offs); if (event->destroy) event->destroy(event); @@ -4351,6 +4451,19 @@ static long _perf_ioctl(struct perf_event *event, unsigned int cmd, unsigned lon case PERF_EVENT_IOC_SET_BPF: return perf_event_set_bpf_prog(event, arg); + case PERF_EVENT_IOC_PAUSE_OUTPUT: { + struct ring_buffer *rb; + + rcu_read_lock(); + rb = rcu_dereference(event->rb); + if (!rb || !rb->nr_pages) { + rcu_read_unlock(); + return -EINVAL; + } + rb_toggle_paused(rb, !!arg); + rcu_read_unlock(); + return 0; + } default: return -ENOTTY; } @@ -4667,6 +4780,8 @@ static void perf_mmap_open(struct vm_area_struct *vma) event->pmu->event_mapped(event); } +static void perf_pmu_output_stop(struct perf_event *event); + /* * A buffer can be mmap()ed multiple times; either directly through the same * event, or through other events by use of perf_event_set_output(). @@ -4694,10 +4809,22 @@ static void perf_mmap_close(struct vm_area_struct *vma) */ if (rb_has_aux(rb) && vma->vm_pgoff == rb->aux_pgoff && atomic_dec_and_mutex_lock(&rb->aux_mmap_count, &event->mmap_mutex)) { + /* + * Stop all AUX events that are writing to this buffer, + * so that we can free its AUX pages and corresponding PMU + * data. Note that after rb::aux_mmap_count dropped to zero, + * they won't start any more (see perf_aux_output_begin()). + */ + perf_pmu_output_stop(event); + + /* now it's safe to free the pages */ atomic_long_sub(rb->aux_nr_pages, &mmap_user->locked_vm); vma->vm_mm->pinned_vm -= rb->aux_mmap_locked; + /* this has to be the last one */ rb_free_aux(rb); + WARN_ON_ONCE(atomic_read(&rb->aux_refcount)); + mutex_unlock(&event->mmap_mutex); } @@ -5638,9 +5765,13 @@ void perf_prepare_sample(struct perf_event_header *header, } } -void perf_event_output(struct perf_event *event, - struct perf_sample_data *data, - struct pt_regs *regs) +static void __always_inline +__perf_event_output(struct perf_event *event, + struct perf_sample_data *data, + struct pt_regs *regs, + int (*output_begin)(struct perf_output_handle *, + struct perf_event *, + unsigned int)) { struct perf_output_handle handle; struct perf_event_header header; @@ -5650,7 +5781,7 @@ void perf_event_output(struct perf_event *event, perf_prepare_sample(&header, data, event, regs); - if (perf_output_begin(&handle, event, header.size)) + if (output_begin(&handle, event, header.size)) goto exit; perf_output_sample(&handle, &header, data, event); @@ -5661,6 +5792,30 @@ exit: rcu_read_unlock(); } +void +perf_event_output_forward(struct perf_event *event, + struct perf_sample_data *data, + struct pt_regs *regs) +{ + __perf_event_output(event, data, regs, perf_output_begin_forward); +} + +void +perf_event_output_backward(struct perf_event *event, + struct perf_sample_data *data, + struct pt_regs *regs) +{ + __perf_event_output(event, data, regs, perf_output_begin_backward); +} + +void +perf_event_output(struct perf_event *event, + struct perf_sample_data *data, + struct pt_regs *regs) +{ + __perf_event_output(event, data, regs, perf_output_begin); +} + /* * read event_id */ @@ -5706,15 +5861,18 @@ typedef void (perf_event_aux_output_cb)(struct perf_event *event, void *data); static void perf_event_aux_ctx(struct perf_event_context *ctx, perf_event_aux_output_cb output, - void *data) + void *data, bool all) { struct perf_event *event; list_for_each_entry_rcu(event, &ctx->event_list, event_entry) { - if (event->state < PERF_EVENT_STATE_INACTIVE) - continue; - if (!event_filter_match(event)) - continue; + if (!all) { + if (event->state < PERF_EVENT_STATE_INACTIVE) + continue; + if (!event_filter_match(event)) + continue; + } + output(event, data); } } @@ -5725,7 +5883,7 @@ perf_event_aux_task_ctx(perf_event_aux_output_cb output, void *data, { rcu_read_lock(); preempt_disable(); - perf_event_aux_ctx(task_ctx, output, data); + perf_event_aux_ctx(task_ctx, output, data, false); preempt_enable(); rcu_read_unlock(); } @@ -5755,13 +5913,13 @@ perf_event_aux(perf_event_aux_output_cb output, void *data, cpuctx = get_cpu_ptr(pmu->pmu_cpu_context); if (cpuctx->unique_pmu != pmu) goto next; - perf_event_aux_ctx(&cpuctx->ctx, output, data); + perf_event_aux_ctx(&cpuctx->ctx, output, data, false); ctxn = pmu->task_ctx_nr; if (ctxn < 0) goto next; ctx = rcu_dereference(current->perf_event_ctxp[ctxn]); if (ctx) - perf_event_aux_ctx(ctx, output, data); + perf_event_aux_ctx(ctx, output, data, false); next: put_cpu_ptr(pmu->pmu_cpu_context); } @@ -5769,6 +5927,134 @@ next: } /* + * Clear all file-based filters at exec, they'll have to be + * re-instated when/if these objects are mmapped again. + */ +static void perf_event_addr_filters_exec(struct perf_event *event, void *data) +{ + struct perf_addr_filters_head *ifh = perf_event_addr_filters(event); + struct perf_addr_filter *filter; + unsigned int restart = 0, count = 0; + unsigned long flags; + + if (!has_addr_filter(event)) + return; + + raw_spin_lock_irqsave(&ifh->lock, flags); + list_for_each_entry(filter, &ifh->list, entry) { + if (filter->inode) { + event->addr_filters_offs[count] = 0; + restart++; + } + + count++; + } + + if (restart) + event->addr_filters_gen++; + raw_spin_unlock_irqrestore(&ifh->lock, flags); + + if (restart) + perf_event_restart(event); +} + +void perf_event_exec(void) +{ + struct perf_event_context *ctx; + int ctxn; + + rcu_read_lock(); + for_each_task_context_nr(ctxn) { + ctx = current->perf_event_ctxp[ctxn]; + if (!ctx) + continue; + + perf_event_enable_on_exec(ctxn); + + perf_event_aux_ctx(ctx, perf_event_addr_filters_exec, NULL, + true); + } + rcu_read_unlock(); +} + +struct remote_output { + struct ring_buffer *rb; + int err; +}; + +static void __perf_event_output_stop(struct perf_event *event, void *data) +{ + struct perf_event *parent = event->parent; + struct remote_output *ro = data; + struct ring_buffer *rb = ro->rb; + struct stop_event_data sd = { + .event = event, + }; + + if (!has_aux(event)) + return; + + if (!parent) + parent = event; + + /* + * In case of inheritance, it will be the parent that links to the + * ring-buffer, but it will be the child that's actually using it: + */ + if (rcu_dereference(parent->rb) == rb) + ro->err = __perf_event_stop(&sd); +} + +static int __perf_pmu_output_stop(void *info) +{ + struct perf_event *event = info; + struct pmu *pmu = event->pmu; + struct perf_cpu_context *cpuctx = get_cpu_ptr(pmu->pmu_cpu_context); + struct remote_output ro = { + .rb = event->rb, + }; + + rcu_read_lock(); + perf_event_aux_ctx(&cpuctx->ctx, __perf_event_output_stop, &ro, false); + if (cpuctx->task_ctx) + perf_event_aux_ctx(cpuctx->task_ctx, __perf_event_output_stop, + &ro, false); + rcu_read_unlock(); + + return ro.err; +} + +static void perf_pmu_output_stop(struct perf_event *event) +{ + struct perf_event *iter; + int err, cpu; + +restart: + rcu_read_lock(); + list_for_each_entry_rcu(iter, &event->rb->event_list, rb_entry) { + /* + * For per-CPU events, we need to make sure that neither they + * nor their children are running; for cpu==-1 events it's + * sufficient to stop the event itself if it's active, since + * it can't have children. + */ + cpu = iter->cpu; + if (cpu == -1) + cpu = READ_ONCE(iter->oncpu); + + if (cpu == -1) + continue; + + err = cpu_function_call(cpu, __perf_pmu_output_stop, event); + if (err == -EAGAIN) { + rcu_read_unlock(); + goto restart; + } + } + rcu_read_unlock(); +} + +/* * task tracking -- fork/exit * * enabled by: attr.comm | attr.mmap | attr.mmap2 | attr.mmap_data | attr.task @@ -6177,6 +6463,87 @@ got_name: kfree(buf); } +/* + * Whether this @filter depends on a dynamic object which is not loaded + * yet or its load addresses are not known. + */ +static bool perf_addr_filter_needs_mmap(struct perf_addr_filter *filter) +{ + return filter->filter && filter->inode; +} + +/* + * Check whether inode and address range match filter criteria. + */ +static bool perf_addr_filter_match(struct perf_addr_filter *filter, + struct file *file, unsigned long offset, + unsigned long size) +{ + if (filter->inode != file->f_inode) + return false; + + if (filter->offset > offset + size) + return false; + + if (filter->offset + filter->size < offset) + return false; + + return true; +} + +static void __perf_addr_filters_adjust(struct perf_event *event, void *data) +{ + struct perf_addr_filters_head *ifh = perf_event_addr_filters(event); + struct vm_area_struct *vma = data; + unsigned long off = vma->vm_pgoff << PAGE_SHIFT, flags; + struct file *file = vma->vm_file; + struct perf_addr_filter *filter; + unsigned int restart = 0, count = 0; + + if (!has_addr_filter(event)) + return; + + if (!file) + return; + + raw_spin_lock_irqsave(&ifh->lock, flags); + list_for_each_entry(filter, &ifh->list, entry) { + if (perf_addr_filter_match(filter, file, off, + vma->vm_end - vma->vm_start)) { + event->addr_filters_offs[count] = vma->vm_start; + restart++; + } + + count++; + } + + if (restart) + event->addr_filters_gen++; + raw_spin_unlock_irqrestore(&ifh->lock, flags); + + if (restart) + perf_event_restart(event); +} + +/* + * Adjust all task's events' filters to the new vma + */ +static void perf_addr_filters_adjust(struct vm_area_struct *vma) +{ + struct perf_event_context *ctx; + int ctxn; + + rcu_read_lock(); + for_each_task_context_nr(ctxn) { + ctx = rcu_dereference(current->perf_event_ctxp[ctxn]); + if (!ctx) + continue; + + perf_event_aux_ctx(ctx, __perf_addr_filters_adjust, vma, true); + } + rcu_read_unlock(); +} + void perf_event_mmap(struct vm_area_struct *vma) { struct perf_mmap_event mmap_event; @@ -6208,6 +6575,7 @@ void perf_event_mmap(struct vm_area_struct *vma) /* .flags (attr_mmap2 only) */ }; + perf_addr_filters_adjust(vma); perf_event_mmap_event(&mmap_event); } @@ -6499,10 +6867,7 @@ static int __perf_event_overflow(struct perf_event *event, irq_work_queue(&event->pending); } - if (event->overflow_handler) - event->overflow_handler(event, data, regs); - else - perf_event_output(event, data, regs); + event->overflow_handler(event, data, regs); if (*perf_event_fasync(event) && event->pending_kill) { event->pending_wakeup = 1; @@ -7089,24 +7454,6 @@ static inline void perf_tp_register(void) perf_pmu_register(&perf_tracepoint, "tracepoint", PERF_TYPE_TRACEPOINT); } -static int perf_event_set_filter(struct perf_event *event, void __user *arg) -{ - char *filter_str; - int ret; - - if (event->attr.type != PERF_TYPE_TRACEPOINT) - return -EINVAL; - - filter_str = strndup_user(arg, PAGE_SIZE); - if (IS_ERR(filter_str)) - return PTR_ERR(filter_str); - - ret = ftrace_profile_set_filter(event, event->attr.config, filter_str); - - kfree(filter_str); - return ret; -} - static void perf_event_free_filter(struct perf_event *event) { ftrace_profile_free_filter(event); @@ -7161,11 +7508,6 @@ static inline void perf_tp_register(void) { } -static int perf_event_set_filter(struct perf_event *event, void __user *arg) -{ - return -ENOENT; -} - static void perf_event_free_filter(struct perf_event *event) { } @@ -7194,6 +7536,387 @@ void perf_bp_event(struct perf_event *bp, void *data) #endif /* + * Allocate a new address filter + */ +static struct perf_addr_filter * +perf_addr_filter_new(struct perf_event *event, struct list_head *filters) +{ + int node = cpu_to_node(event->cpu == -1 ? 0 : event->cpu); + struct perf_addr_filter *filter; + + filter = kzalloc_node(sizeof(*filter), GFP_KERNEL, node); + if (!filter) + return NULL; + + INIT_LIST_HEAD(&filter->entry); + list_add_tail(&filter->entry, filters); + + return filter; +} + +static void free_filters_list(struct list_head *filters) +{ + struct perf_addr_filter *filter, *iter; + + list_for_each_entry_safe(filter, iter, filters, entry) { + if (filter->inode) + iput(filter->inode); + list_del(&filter->entry); + kfree(filter); + } +} + +/* + * Free existing address filters and optionally install new ones + */ +static void perf_addr_filters_splice(struct perf_event *event, + struct list_head *head) +{ + unsigned long flags; + LIST_HEAD(list); + + if (!has_addr_filter(event)) + return; + + /* don't bother with children, they don't have their own filters */ + if (event->parent) + return; + + raw_spin_lock_irqsave(&event->addr_filters.lock, flags); + + list_splice_init(&event->addr_filters.list, &list); + if (head) + list_splice(head, &event->addr_filters.list); + + raw_spin_unlock_irqrestore(&event->addr_filters.lock, flags); + + free_filters_list(&list); +} + +/* + * Scan through mm's vmas and see if one of them matches the + * @filter; if so, adjust filter's address range. + * Called with mm::mmap_sem down for reading. + */ +static unsigned long perf_addr_filter_apply(struct perf_addr_filter *filter, + struct mm_struct *mm) +{ + struct vm_area_struct *vma; + + for (vma = mm->mmap; vma; vma = vma->vm_next) { + struct file *file = vma->vm_file; + unsigned long off = vma->vm_pgoff << PAGE_SHIFT; + unsigned long vma_size = vma->vm_end - vma->vm_start; + + if (!file) + continue; + + if (!perf_addr_filter_match(filter, file, off, vma_size)) + continue; + + return vma->vm_start; + } + + return 0; +} + +/* + * Update event's address range filters based on the + * task's existing mappings, if any. + */ +static void perf_event_addr_filters_apply(struct perf_event *event) +{ + struct perf_addr_filters_head *ifh = perf_event_addr_filters(event); + struct task_struct *task = READ_ONCE(event->ctx->task); + struct perf_addr_filter *filter; + struct mm_struct *mm = NULL; + unsigned int count = 0; + unsigned long flags; + + /* + * We may observe TASK_TOMBSTONE, which means that the event tear-down + * will stop on the parent's child_mutex that our caller is also holding + */ + if (task == TASK_TOMBSTONE) + return; + + mm = get_task_mm(event->ctx->task); + if (!mm) + goto restart; + + down_read(&mm->mmap_sem); + + raw_spin_lock_irqsave(&ifh->lock, flags); + list_for_each_entry(filter, &ifh->list, entry) { + event->addr_filters_offs[count] = 0; + + if (perf_addr_filter_needs_mmap(filter)) + event->addr_filters_offs[count] = + perf_addr_filter_apply(filter, mm); + + count++; + } + + event->addr_filters_gen++; + raw_spin_unlock_irqrestore(&ifh->lock, flags); + + up_read(&mm->mmap_sem); + + mmput(mm); + +restart: + perf_event_restart(event); +} + +/* + * Address range filtering: limiting the data to certain + * instruction address ranges. Filters are ioctl()ed to us from + * userspace as ascii strings. + * + * Filter string format: + * + * ACTION RANGE_SPEC + * where ACTION is one of the + * * "filter": limit the trace to this region + * * "start": start tracing from this address + * * "stop": stop tracing at this address/region; + * RANGE_SPEC is + * * for kernel addresses: <start address>[/<size>] + * * for object files: <start address>[/<size>]@</path/to/object/file> + * + * if <size> is not specified, the range is treated as a single address. + */ +enum { + IF_ACT_FILTER, + IF_ACT_START, + IF_ACT_STOP, + IF_SRC_FILE, + IF_SRC_KERNEL, + IF_SRC_FILEADDR, + IF_SRC_KERNELADDR, +}; + +enum { + IF_STATE_ACTION = 0, + IF_STATE_SOURCE, + IF_STATE_END, +}; + +static const match_table_t if_tokens = { + { IF_ACT_FILTER, "filter" }, + { IF_ACT_START, "start" }, + { IF_ACT_STOP, "stop" }, + { IF_SRC_FILE, "%u/%u@%s" }, + { IF_SRC_KERNEL, "%u/%u" }, + { IF_SRC_FILEADDR, "%u@%s" }, + { IF_SRC_KERNELADDR, "%u" }, +}; + +/* + * Address filter string parser + */ +static int +perf_event_parse_addr_filter(struct perf_event *event, char *fstr, + struct list_head *filters) +{ + struct perf_addr_filter *filter = NULL; + char *start, *orig, *filename = NULL; + struct path path; + substring_t args[MAX_OPT_ARGS]; + int state = IF_STATE_ACTION, token; + unsigned int kernel = 0; + int ret = -EINVAL; + + orig = fstr = kstrdup(fstr, GFP_KERNEL); + if (!fstr) + return -ENOMEM; + + while ((start = strsep(&fstr, " ,\n")) != NULL) { + ret = -EINVAL; + + if (!*start) + continue; + + /* filter definition begins */ + if (state == IF_STATE_ACTION) { + filter = perf_addr_filter_new(event, filters); + if (!filter) + goto fail; + } + + token = match_token(start, if_tokens, args); + switch (token) { + case IF_ACT_FILTER: + case IF_ACT_START: + filter->filter = 1; + + case IF_ACT_STOP: + if (state != IF_STATE_ACTION) + goto fail; + + state = IF_STATE_SOURCE; + break; + + case IF_SRC_KERNELADDR: + case IF_SRC_KERNEL: + kernel = 1; + + case IF_SRC_FILEADDR: + case IF_SRC_FILE: + if (state != IF_STATE_SOURCE) + goto fail; + + if (token == IF_SRC_FILE || token == IF_SRC_KERNEL) + filter->range = 1; + + *args[0].to = 0; + ret = kstrtoul(args[0].from, 0, &filter->offset); + if (ret) + goto fail; + + if (filter->range) { + *args[1].to = 0; + ret = kstrtoul(args[1].from, 0, &filter->size); + if (ret) + goto fail; + } + + if (token == IF_SRC_FILE) { + filename = match_strdup(&args[2]); + if (!filename) { + ret = -ENOMEM; + goto fail; + } + } + + state = IF_STATE_END; + break; + + default: + goto fail; + } + + /* + * Filter definition is fully parsed, validate and install it. + * Make sure that it doesn't contradict itself or the event's + * attribute. + */ + if (state == IF_STATE_END) { + if (kernel && event->attr.exclude_kernel) + goto fail; + + if (!kernel) { + if (!filename) + goto fail; + + /* look up the path and grab its inode */ + ret = kern_path(filename, LOOKUP_FOLLOW, &path); + if (ret) + goto fail_free_name; + + filter->inode = igrab(d_inode(path.dentry)); + path_put(&path); + kfree(filename); + filename = NULL; + + ret = -EINVAL; + if (!filter->inode || + !S_ISREG(filter->inode->i_mode)) + /* free_filters_list() will iput() */ + goto fail; + } + + /* ready to consume more filters */ + state = IF_STATE_ACTION; + filter = NULL; + } + } + + if (state != IF_STATE_ACTION) + goto fail; + + kfree(orig); + + return 0; + +fail_free_name: + kfree(filename); +fail: + free_filters_list(filters); + kfree(orig); + + return ret; +} + +static int +perf_event_set_addr_filter(struct perf_event *event, char *filter_str) +{ + LIST_HEAD(filters); + int ret; + + /* + * Since this is called in perf_ioctl() path, we're already holding + * ctx::mutex. + */ + lockdep_assert_held(&event->ctx->mutex); + + if (WARN_ON_ONCE(event->parent)) + return -EINVAL; + + /* + * For now, we only support filtering in per-task events; doing so + * for CPU-wide events requires additional context switching trickery, + * since same object code will be mapped at different virtual + * addresses in different processes. + */ + if (!event->ctx->task) + return -EOPNOTSUPP; + + ret = perf_event_parse_addr_filter(event, filter_str, &filters); + if (ret) + return ret; + + ret = event->pmu->addr_filters_validate(&filters); + if (ret) { + free_filters_list(&filters); + return ret; + } + + /* remove existing filters, if any */ + perf_addr_filters_splice(event, &filters); + + /* install new filters */ + perf_event_for_each_child(event, perf_event_addr_filters_apply); + + return ret; +} + +static int perf_event_set_filter(struct perf_event *event, void __user *arg) +{ + char *filter_str; + int ret = -EINVAL; + + if ((event->attr.type != PERF_TYPE_TRACEPOINT || + !IS_ENABLED(CONFIG_EVENT_TRACING)) && + !has_addr_filter(event)) + return -EINVAL; + + filter_str = strndup_user(arg, PAGE_SIZE); + if (IS_ERR(filter_str)) + return PTR_ERR(filter_str); + + if (IS_ENABLED(CONFIG_EVENT_TRACING) && + event->attr.type == PERF_TYPE_TRACEPOINT) + ret = ftrace_profile_set_filter(event, event->attr.config, + filter_str); + else if (has_addr_filter(event)) + ret = perf_event_set_addr_filter(event, filter_str); + + kfree(filter_str); + return ret; +} + +/* * hrtimer based swevent callback */ @@ -7550,6 +8273,20 @@ static void free_pmu_context(struct pmu *pmu) out: mutex_unlock(&pmus_lock); } + +/* + * Let userspace know that this PMU supports address range filtering: + */ +static ssize_t nr_addr_filters_show(struct device *dev, + struct device_attribute *attr, + char *page) +{ + struct pmu *pmu = dev_get_drvdata(dev); + + return snprintf(page, PAGE_SIZE - 1, "%d\n", pmu->nr_addr_filters); +} +DEVICE_ATTR_RO(nr_addr_filters); + static struct idr pmu_idr; static ssize_t @@ -7651,9 +8388,19 @@ static int pmu_dev_alloc(struct pmu *pmu) if (ret) goto free_dev; + /* For PMUs with address filters, throw in an extra attribute: */ + if (pmu->nr_addr_filters) + ret = device_create_file(pmu->dev, &dev_attr_nr_addr_filters); + + if (ret) + goto del_dev; + out: return ret; +del_dev: + device_del(pmu->dev); + free_dev: put_device(pmu->dev); goto out; @@ -7693,6 +8440,21 @@ int perf_pmu_register(struct pmu *pmu, const char *name, int type) } skip_type: + if (pmu->task_ctx_nr == perf_hw_context) { + static int hw_context_taken = 0; + + /* + * Other than systems with heterogeneous CPUs, it never makes + * sense for two PMUs to share perf_hw_context. PMUs which are + * uncore must use perf_invalid_context. + */ + if (WARN_ON_ONCE(hw_context_taken && + !(pmu->capabilities & PERF_PMU_CAP_HETEROGENEOUS_CPUS))) + pmu->task_ctx_nr = perf_invalid_context; + + hw_context_taken = 1; + } + pmu->pmu_cpu_context = find_pmu_context(pmu->task_ctx_nr); if (pmu->pmu_cpu_context) goto got_cpu_context; @@ -7780,6 +8542,8 @@ void perf_pmu_unregister(struct pmu *pmu) free_percpu(pmu->pmu_disable_count); if (pmu->type >= PERF_TYPE_MAX) idr_remove(&pmu_idr, pmu->type); + if (pmu->nr_addr_filters) + device_remove_file(pmu->dev, &dev_attr_nr_addr_filters); device_del(pmu->dev); put_device(pmu->dev); free_pmu_context(pmu); @@ -7973,6 +8737,7 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu, INIT_LIST_HEAD(&event->sibling_list); INIT_LIST_HEAD(&event->rb_entry); INIT_LIST_HEAD(&event->active_entry); + INIT_LIST_HEAD(&event->addr_filters.list); INIT_HLIST_NODE(&event->hlist_entry); @@ -7980,6 +8745,7 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu, init_irq_work(&event->pending, perf_pending_event); mutex_init(&event->mmap_mutex); + raw_spin_lock_init(&event->addr_filters.lock); atomic_long_set(&event->refcount, 1); event->cpu = cpu; @@ -8014,8 +8780,16 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu, context = parent_event->overflow_handler_context; } - event->overflow_handler = overflow_handler; - event->overflow_handler_context = context; + if (overflow_handler) { + event->overflow_handler = overflow_handler; + event->overflow_handler_context = context; + } else if (is_write_backward(event)){ + event->overflow_handler = perf_event_output_backward; + event->overflow_handler_context = NULL; + } else { + event->overflow_handler = perf_event_output_forward; + event->overflow_handler_context = NULL; + } perf_event__state_init(event); @@ -8056,11 +8830,22 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu, if (err) goto err_pmu; + if (has_addr_filter(event)) { + event->addr_filters_offs = kcalloc(pmu->nr_addr_filters, + sizeof(unsigned long), + GFP_KERNEL); + if (!event->addr_filters_offs) + goto err_per_task; + + /* force hw sync on the address filters */ + event->addr_filters_gen = 1; + } + if (!event->parent) { if (event->attr.sample_type & PERF_SAMPLE_CALLCHAIN) { err = get_callchain_buffers(); if (err) - goto err_per_task; + goto err_addr_filters; } } @@ -8069,6 +8854,9 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu, return event; +err_addr_filters: + kfree(event->addr_filters_offs); + err_per_task: exclusive_event_destroy(event); @@ -8248,6 +9036,13 @@ perf_event_set_output(struct perf_event *event, struct perf_event *output_event) goto out; /* + * Either writing ring buffer from beginning or from end. + * Mixing is not allowed. + */ + if (is_write_backward(output_event) != is_write_backward(event)) + goto out; + + /* * If both events generate aux data, they must be on the same PMU */ if (has_aux(event) && has_aux(output_event) && @@ -8413,6 +9208,24 @@ SYSCALL_DEFINE5(perf_event_open, get_online_cpus(); + if (task) { + err = mutex_lock_interruptible(&task->signal->cred_guard_mutex); + if (err) + goto err_cpus; + + /* + * Reuse ptrace permission checks for now. + * + * We must hold cred_guard_mutex across this and any potential + * perf_install_in_context() call for this new event to + * serialize against exec() altering our credentials (and the + * perf_event_exit_task() that could imply). + */ + err = -EACCES; + if (!ptrace_may_access(task, PTRACE_MODE_READ_REALCREDS)) + goto err_cred; + } + if (flags & PERF_FLAG_PID_CGROUP) cgroup_fd = pid; @@ -8420,7 +9233,7 @@ SYSCALL_DEFINE5(perf_event_open, NULL, NULL, cgroup_fd); if (IS_ERR(event)) { err = PTR_ERR(event); - goto err_cpus; + goto err_cred; } if (is_sampling_event(event)) { @@ -8479,11 +9292,6 @@ SYSCALL_DEFINE5(perf_event_open, goto err_context; } - if (task) { - put_task_struct(task); - task = NULL; - } - /* * Look up the group leader (we will attach this event to it): */ @@ -8581,6 +9389,11 @@ SYSCALL_DEFINE5(perf_event_open, WARN_ON_ONCE(ctx->parent_ctx); + /* + * This is the point on no return; we cannot fail hereafter. This is + * where we start modifying current state. + */ + if (move_group) { /* * See perf_event_ctx_lock() for comments on the details @@ -8652,6 +9465,11 @@ SYSCALL_DEFINE5(perf_event_open, mutex_unlock(&gctx->mutex); mutex_unlock(&ctx->mutex); + if (task) { + mutex_unlock(&task->signal->cred_guard_mutex); + put_task_struct(task); + } + put_online_cpus(); mutex_lock(¤t->perf_event_mutex); @@ -8684,6 +9502,9 @@ err_alloc: */ if (!event_file) free_event(event); +err_cred: + if (task) + mutex_unlock(&task->signal->cred_guard_mutex); err_cpus: put_online_cpus(); err_task: @@ -8968,6 +9789,9 @@ static void perf_event_exit_task_context(struct task_struct *child, int ctxn) /* * When a child task exits, feed back event values to parent events. + * + * Can be called with cred_guard_mutex held when called from + * install_exec_creds(). */ void perf_event_exit_task(struct task_struct *child) { diff --git a/kernel/events/internal.h b/kernel/events/internal.h index 4199b6d..05f9f6d 100644 --- a/kernel/events/internal.h +++ b/kernel/events/internal.h @@ -11,13 +11,13 @@ struct ring_buffer { atomic_t refcount; struct rcu_head rcu_head; - struct irq_work irq_work; #ifdef CONFIG_PERF_USE_VMALLOC struct work_struct work; int page_order; /* allocation order */ #endif int nr_pages; /* nr of data pages */ int overwrite; /* can overwrite itself */ + int paused; /* can write into ring buffer */ atomic_t poll; /* POLL_ for wakeups */ @@ -65,6 +65,14 @@ static inline void rb_free_rcu(struct rcu_head *rcu_head) rb_free(rb); } +static inline void rb_toggle_paused(struct ring_buffer *rb, bool pause) +{ + if (!pause && rb->nr_pages) + rb->paused = 0; + else + rb->paused = 1; +} + extern struct ring_buffer * rb_alloc(int nr_pages, long watermark, int cpu, int flags); extern void perf_event_wakeup(struct perf_event *event); diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c index c61f0cb..ae9b90d 100644 --- a/kernel/events/ring_buffer.c +++ b/kernel/events/ring_buffer.c @@ -102,8 +102,21 @@ out: preempt_enable(); } -int perf_output_begin(struct perf_output_handle *handle, - struct perf_event *event, unsigned int size) +static bool __always_inline +ring_buffer_has_space(unsigned long head, unsigned long tail, + unsigned long data_size, unsigned int size, + bool backward) +{ + if (!backward) + return CIRC_SPACE(head, tail, data_size) >= size; + else + return CIRC_SPACE(tail, head, data_size) >= size; +} + +static int __always_inline +__perf_output_begin(struct perf_output_handle *handle, + struct perf_event *event, unsigned int size, + bool backward) { struct ring_buffer *rb; unsigned long tail, offset, head; @@ -125,8 +138,11 @@ int perf_output_begin(struct perf_output_handle *handle, if (unlikely(!rb)) goto out; - if (unlikely(!rb->nr_pages)) + if (unlikely(rb->paused)) { + if (rb->nr_pages) + local_inc(&rb->lost); goto out; + } handle->rb = rb; handle->event = event; @@ -143,9 +159,12 @@ int perf_output_begin(struct perf_output_handle *handle, do { tail = READ_ONCE(rb->user_page->data_tail); offset = head = local_read(&rb->head); - if (!rb->overwrite && - unlikely(CIRC_SPACE(head, tail, perf_data_size(rb)) < size)) - goto fail; + if (!rb->overwrite) { + if (unlikely(!ring_buffer_has_space(head, tail, + perf_data_size(rb), + size, backward))) + goto fail; + } /* * The above forms a control dependency barrier separating the @@ -159,9 +178,17 @@ int perf_output_begin(struct perf_output_handle *handle, * See perf_output_put_handle(). */ - head += size; + if (!backward) + head += size; + else + head -= size; } while (local_cmpxchg(&rb->head, offset, head) != offset); + if (backward) { + offset = head; + head = (u64)(-head); + } + /* * We rely on the implied barrier() by local_cmpxchg() to ensure * none of the data stores below can be lifted up by the compiler. @@ -203,6 +230,26 @@ out: return -ENOSPC; } +int perf_output_begin_forward(struct perf_output_handle *handle, + struct perf_event *event, unsigned int size) +{ + return __perf_output_begin(handle, event, size, false); +} + +int perf_output_begin_backward(struct perf_output_handle *handle, + struct perf_event *event, unsigned int size) +{ + return __perf_output_begin(handle, event, size, true); +} + +int perf_output_begin(struct perf_output_handle *handle, + struct perf_event *event, unsigned int size) +{ + + return __perf_output_begin(handle, event, size, + unlikely(is_write_backward(event))); +} + unsigned int perf_output_copy(struct perf_output_handle *handle, const void *buf, unsigned int len) { @@ -221,8 +268,6 @@ void perf_output_end(struct perf_output_handle *handle) rcu_read_unlock(); } -static void rb_irq_work(struct irq_work *work); - static void ring_buffer_init(struct ring_buffer *rb, long watermark, int flags) { @@ -243,16 +288,13 @@ ring_buffer_init(struct ring_buffer *rb, long watermark, int flags) INIT_LIST_HEAD(&rb->event_list); spin_lock_init(&rb->event_lock); - init_irq_work(&rb->irq_work, rb_irq_work); -} -static void ring_buffer_put_async(struct ring_buffer *rb) -{ - if (!atomic_dec_and_test(&rb->refcount)) - return; - - rb->rcu_head.next = (void *)rb; - irq_work_queue(&rb->irq_work); + /* + * perf_output_begin() only checks rb->paused, therefore + * rb->paused must be true if we have no pages for output. + */ + if (!rb->nr_pages) + rb->paused = 1; } /* @@ -264,6 +306,10 @@ static void ring_buffer_put_async(struct ring_buffer *rb) * The ordering is similar to that of perf_output_{begin,end}, with * the exception of (B), which should be taken care of by the pmu * driver, since ordering rules will differ depending on hardware. + * + * Call this from pmu::start(); see the comment in perf_aux_output_end() + * about its use in pmu callbacks. Both can also be called from the PMI + * handler if needed. */ void *perf_aux_output_begin(struct perf_output_handle *handle, struct perf_event *event) @@ -288,6 +334,13 @@ void *perf_aux_output_begin(struct perf_output_handle *handle, goto err; /* + * If rb::aux_mmap_count is zero (and rb_has_aux() above went through), + * the aux buffer is in perf_mmap_close(), about to get freed. + */ + if (!atomic_read(&rb->aux_mmap_count)) + goto err_put; + + /* * Nesting is not supported for AUX area, make sure nested * writers are caught early */ @@ -328,10 +381,11 @@ void *perf_aux_output_begin(struct perf_output_handle *handle, return handle->rb->aux_priv; err_put: + /* can't be last */ rb_free_aux(rb); err: - ring_buffer_put_async(rb); + ring_buffer_put(rb); handle->event = NULL; return NULL; @@ -342,11 +396,16 @@ err: * aux_head and posting a PERF_RECORD_AUX into the perf buffer. It is the * pmu driver's responsibility to observe ordering rules of the hardware, * so that all the data is externally visible before this is called. + * + * Note: this has to be called from pmu::stop() callback, as the assumption + * of the AUX buffer management code is that after pmu::stop(), the AUX + * transaction must be stopped and therefore drop the AUX reference count. */ void perf_aux_output_end(struct perf_output_handle *handle, unsigned long size, bool truncated) { struct ring_buffer *rb = handle->rb; + bool wakeup = truncated; unsigned long aux_head; u64 flags = 0; @@ -375,14 +434,22 @@ void perf_aux_output_end(struct perf_output_handle *handle, unsigned long size, aux_head = rb->user_page->aux_head = local_read(&rb->aux_head); if (aux_head - local_read(&rb->aux_wakeup) >= rb->aux_watermark) { - perf_output_wakeup(handle); + wakeup = true; local_add(rb->aux_watermark, &rb->aux_wakeup); } + + if (wakeup) { + if (truncated) + handle->event->pending_disable = 1; + perf_output_wakeup(handle); + } + handle->event = NULL; local_set(&rb->aux_nest, 0); + /* can't be last */ rb_free_aux(rb); - ring_buffer_put_async(rb); + ring_buffer_put(rb); } /* @@ -463,6 +530,14 @@ static void __rb_free_aux(struct ring_buffer *rb) { int pg; + /* + * Should never happen, the last reference should be dropped from + * perf_mmap_close() path, which first stops aux transactions (which + * in turn are the atomic holders of aux_refcount) and then does the + * last rb_free_aux(). + */ + WARN_ON_ONCE(in_atomic()); + if (rb->aux_priv) { rb->free_aux(rb->aux_priv); rb->free_aux = NULL; @@ -574,18 +649,7 @@ out: void rb_free_aux(struct ring_buffer *rb) { if (atomic_dec_and_test(&rb->aux_refcount)) - irq_work_queue(&rb->irq_work); -} - -static void rb_irq_work(struct irq_work *work) -{ - struct ring_buffer *rb = container_of(work, struct ring_buffer, irq_work); - - if (!atomic_read(&rb->aux_refcount)) __rb_free_aux(rb); - - if (rb->rcu_head.next == (void *)rb) - call_rcu(&rb->rcu_head, rb_free_rcu); } #ifndef CONFIG_PERF_USE_VMALLOC diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c index 220fc17..7edc95e 100644 --- a/kernel/events/uprobes.c +++ b/kernel/events/uprobes.c @@ -321,7 +321,7 @@ retry: copy_to_page(new_page, vaddr, &opcode, UPROBE_SWBP_INSN_SIZE); ret = __replace_page(vma, vaddr, old_page, new_page); - page_cache_release(new_page); + put_page(new_page); put_old: put_page(old_page); @@ -539,14 +539,14 @@ static int __copy_insn(struct address_space *mapping, struct file *filp, * see uprobe_register(). */ if (mapping->a_ops->readpage) - page = read_mapping_page(mapping, offset >> PAGE_CACHE_SHIFT, filp); + page = read_mapping_page(mapping, offset >> PAGE_SHIFT, filp); else - page = shmem_read_mapping_page(mapping, offset >> PAGE_CACHE_SHIFT); + page = shmem_read_mapping_page(mapping, offset >> PAGE_SHIFT); if (IS_ERR(page)) return PTR_ERR(page); copy_from_page(page, offset, insn, nbytes); - page_cache_release(page); + put_page(page); return 0; } diff --git a/kernel/fork.c b/kernel/fork.c index d277e83..3e84515 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1494,7 +1494,7 @@ static struct task_struct *copy_process(unsigned long clone_flags, * sigaltstack should be cleared when sharing the same VM */ if ((clone_flags & (CLONE_VM|CLONE_VFORK)) == CLONE_VM) - p->sas_ss_sp = p->sas_ss_size = 0; + sas_ss_reset(p); /* * Syscall tracing and stepping should be turned off in the diff --git a/kernel/futex.c b/kernel/futex.c index a5d2e74..c20f06f 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -1295,10 +1295,20 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this, if (unlikely(should_fail_futex(true))) ret = -EFAULT; - if (cmpxchg_futex_value_locked(&curval, uaddr, uval, newval)) + if (cmpxchg_futex_value_locked(&curval, uaddr, uval, newval)) { ret = -EFAULT; - else if (curval != uval) - ret = -EINVAL; + } else if (curval != uval) { + /* + * If a unconditional UNLOCK_PI operation (user space did not + * try the TID->0 transition) raced with a waiter setting the + * FUTEX_WAITERS flag between get_user() and locking the hash + * bucket lock, retry the operation. + */ + if ((FUTEX_TID_MASK & curval) == uval) + ret = -EAGAIN; + else + ret = -EINVAL; + } if (ret) { raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); return ret; @@ -1525,8 +1535,8 @@ void requeue_futex(struct futex_q *q, struct futex_hash_bucket *hb1, if (likely(&hb1->chain != &hb2->chain)) { plist_del(&q->list, &hb1->chain); hb_waiters_dec(hb1); - plist_add(&q->list, &hb2->chain); hb_waiters_inc(hb2); + plist_add(&q->list, &hb2->chain); q->lock_ptr = &hb2->lock; } get_futex_key_refs(key2); @@ -2623,6 +2633,15 @@ retry: if (ret == -EFAULT) goto pi_faulted; /* + * A unconditional UNLOCK_PI op raced against a waiter + * setting the FUTEX_WAITERS bit. Try again. + */ + if (ret == -EAGAIN) { + spin_unlock(&hb->lock); + put_futex_key(&key); + goto retry; + } + /* * wake_futex_pi has detected invalid state. Tell user * space. */ diff --git a/kernel/irq/ipi.c b/kernel/irq/ipi.c index c37f34b..14777af 100644 --- a/kernel/irq/ipi.c +++ b/kernel/irq/ipi.c @@ -94,6 +94,7 @@ unsigned int irq_reserve_ipi(struct irq_domain *domain, data = irq_get_irq_data(virq + i); cpumask_copy(data->common->affinity, dest); data->common->ipi_offset = offset; + irq_set_status_flags(virq + i, IRQ_NO_BALANCING); } return virq; diff --git a/kernel/kcov.c b/kernel/kcov.c index 3efbee0..a02f2dd 100644 --- a/kernel/kcov.c +++ b/kernel/kcov.c @@ -1,5 +1,6 @@ #define pr_fmt(fmt) "kcov: " fmt +#define DISABLE_BRANCH_PROFILING #include <linux/compiler.h> #include <linux/types.h> #include <linux/file.h> @@ -43,7 +44,7 @@ struct kcov { * Entry point from instrumented code. * This is called once per basic-block/edge. */ -void __sanitizer_cov_trace_pc(void) +void notrace __sanitizer_cov_trace_pc(void) { struct task_struct *t; enum kcov_mode mode; diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c index 8d34308..1391d3e 100644 --- a/kernel/kexec_core.c +++ b/kernel/kexec_core.c @@ -1415,6 +1415,9 @@ static int __init crash_save_vmcoreinfo_init(void) VMCOREINFO_OFFSET(page, lru); VMCOREINFO_OFFSET(page, _mapcount); VMCOREINFO_OFFSET(page, private); + VMCOREINFO_OFFSET(page, compound_dtor); + VMCOREINFO_OFFSET(page, compound_order); + VMCOREINFO_OFFSET(page, compound_head); VMCOREINFO_OFFSET(pglist_data, node_zones); VMCOREINFO_OFFSET(pglist_data, nr_zones); #ifdef CONFIG_FLAT_NODE_MEM_MAP @@ -1447,8 +1450,8 @@ static int __init crash_save_vmcoreinfo_init(void) #ifdef CONFIG_X86 VMCOREINFO_NUMBER(KERNEL_IMAGE_SIZE); #endif -#ifdef CONFIG_HUGETLBFS - VMCOREINFO_SYMBOL(free_huge_page); +#ifdef CONFIG_HUGETLB_PAGE + VMCOREINFO_NUMBER(HUGETLB_PAGE_DTOR); #endif arch_crash_save_vmcoreinfo(); diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c index 2324ba5..81f1a71 100644 --- a/kernel/locking/lockdep.c +++ b/kernel/locking/lockdep.c @@ -45,6 +45,7 @@ #include <linux/bitops.h> #include <linux/gfp.h> #include <linux/kmemcheck.h> +#include <linux/random.h> #include <asm/sections.h> @@ -708,7 +709,7 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) * yet. Otherwise we look it up. We cache the result in the lock object * itself, so actual lookup of the hash should be once per lock object. */ -static inline struct lock_class * +static struct lock_class * register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) { struct lockdep_subclass_key *key; @@ -1999,6 +2000,7 @@ static inline int get_first_held_lock(struct task_struct *curr, return ++i; } +#ifdef CONFIG_DEBUG_LOCKDEP /* * Returns the next chain_key iteration */ @@ -2069,6 +2071,7 @@ static void print_collision(struct task_struct *curr, printk("\nstack backtrace:\n"); dump_stack(); } +#endif /* * Checks whether the chain and the current held locks are consistent @@ -2174,15 +2177,37 @@ cache_hit: chain->irq_context = hlock->irq_context; i = get_first_held_lock(curr, hlock); chain->depth = curr->lockdep_depth + 1 - i; + + BUILD_BUG_ON((1UL << 24) <= ARRAY_SIZE(chain_hlocks)); + BUILD_BUG_ON((1UL << 6) <= ARRAY_SIZE(curr->held_locks)); + BUILD_BUG_ON((1UL << 8*sizeof(chain_hlocks[0])) <= ARRAY_SIZE(lock_classes)); + if (likely(nr_chain_hlocks + chain->depth <= MAX_LOCKDEP_CHAIN_HLOCKS)) { chain->base = nr_chain_hlocks; - nr_chain_hlocks += chain->depth; for (j = 0; j < chain->depth - 1; j++, i++) { int lock_id = curr->held_locks[i].class_idx - 1; chain_hlocks[chain->base + j] = lock_id; } chain_hlocks[chain->base + j] = class - lock_classes; } + + if (nr_chain_hlocks < MAX_LOCKDEP_CHAIN_HLOCKS) + nr_chain_hlocks += chain->depth; + +#ifdef CONFIG_DEBUG_LOCKDEP + /* + * Important for check_no_collision(). + */ + if (unlikely(nr_chain_hlocks > MAX_LOCKDEP_CHAIN_HLOCKS)) { + if (debug_locks_off_graph_unlock()) + return 0; + + print_lockdep_off("BUG: MAX_LOCKDEP_CHAIN_HLOCKS too low!"); + dump_stack(); + return 0; + } +#endif + hlist_add_head_rcu(&chain->entry, hash_head); debug_atomic_inc(chain_lookup_misses); inc_chains(); @@ -2930,6 +2955,11 @@ static int mark_irqflags(struct task_struct *curr, struct held_lock *hlock) return 1; } +static inline unsigned int task_irq_context(struct task_struct *task) +{ + return 2 * !!task->hardirq_context + !!task->softirq_context; +} + static int separate_irq_context(struct task_struct *curr, struct held_lock *hlock) { @@ -2938,8 +2968,6 @@ static int separate_irq_context(struct task_struct *curr, /* * Keep track of points where we cross into an interrupt context: */ - hlock->irq_context = 2*(curr->hardirq_context ? 1 : 0) + - curr->softirq_context; if (depth) { struct held_lock *prev_hlock; @@ -2971,6 +2999,11 @@ static inline int mark_irqflags(struct task_struct *curr, return 1; } +static inline unsigned int task_irq_context(struct task_struct *task) +{ + return 0; +} + static inline int separate_irq_context(struct task_struct *curr, struct held_lock *hlock) { @@ -3239,6 +3272,7 @@ static int __lock_acquire(struct lockdep_map *lock, unsigned int subclass, hlock->acquire_ip = ip; hlock->instance = lock; hlock->nest_lock = nest_lock; + hlock->irq_context = task_irq_context(curr); hlock->trylock = trylock; hlock->read = read; hlock->check = check; @@ -3552,7 +3586,35 @@ static int __lock_is_held(struct lockdep_map *lock) return 0; } -static void __lock_pin_lock(struct lockdep_map *lock) +static struct pin_cookie __lock_pin_lock(struct lockdep_map *lock) +{ + struct pin_cookie cookie = NIL_COOKIE; + struct task_struct *curr = current; + int i; + + if (unlikely(!debug_locks)) + return cookie; + + for (i = 0; i < curr->lockdep_depth; i++) { + struct held_lock *hlock = curr->held_locks + i; + + if (match_held_lock(hlock, lock)) { + /* + * Grab 16bits of randomness; this is sufficient to not + * be guessable and still allows some pin nesting in + * our u32 pin_count. + */ + cookie.val = 1 + (prandom_u32() >> 16); + hlock->pin_count += cookie.val; + return cookie; + } + } + + WARN(1, "pinning an unheld lock\n"); + return cookie; +} + +static void __lock_repin_lock(struct lockdep_map *lock, struct pin_cookie cookie) { struct task_struct *curr = current; int i; @@ -3564,7 +3626,7 @@ static void __lock_pin_lock(struct lockdep_map *lock) struct held_lock *hlock = curr->held_locks + i; if (match_held_lock(hlock, lock)) { - hlock->pin_count++; + hlock->pin_count += cookie.val; return; } } @@ -3572,7 +3634,7 @@ static void __lock_pin_lock(struct lockdep_map *lock) WARN(1, "pinning an unheld lock\n"); } -static void __lock_unpin_lock(struct lockdep_map *lock) +static void __lock_unpin_lock(struct lockdep_map *lock, struct pin_cookie cookie) { struct task_struct *curr = current; int i; @@ -3587,7 +3649,11 @@ static void __lock_unpin_lock(struct lockdep_map *lock) if (WARN(!hlock->pin_count, "unpinning an unpinned lock\n")) return; - hlock->pin_count--; + hlock->pin_count -= cookie.val; + + if (WARN((int)hlock->pin_count < 0, "pin count corrupted\n")) + hlock->pin_count = 0; + return; } } @@ -3718,24 +3784,44 @@ int lock_is_held(struct lockdep_map *lock) } EXPORT_SYMBOL_GPL(lock_is_held); -void lock_pin_lock(struct lockdep_map *lock) +struct pin_cookie lock_pin_lock(struct lockdep_map *lock) { + struct pin_cookie cookie = NIL_COOKIE; unsigned long flags; if (unlikely(current->lockdep_recursion)) - return; + return cookie; raw_local_irq_save(flags); check_flags(flags); current->lockdep_recursion = 1; - __lock_pin_lock(lock); + cookie = __lock_pin_lock(lock); current->lockdep_recursion = 0; raw_local_irq_restore(flags); + + return cookie; } EXPORT_SYMBOL_GPL(lock_pin_lock); -void lock_unpin_lock(struct lockdep_map *lock) +void lock_repin_lock(struct lockdep_map *lock, struct pin_cookie cookie) +{ + unsigned long flags; + + if (unlikely(current->lockdep_recursion)) + return; + + raw_local_irq_save(flags); + check_flags(flags); + + current->lockdep_recursion = 1; + __lock_repin_lock(lock, cookie); + current->lockdep_recursion = 0; + raw_local_irq_restore(flags); +} +EXPORT_SYMBOL_GPL(lock_repin_lock); + +void lock_unpin_lock(struct lockdep_map *lock, struct pin_cookie cookie) { unsigned long flags; @@ -3746,7 +3832,7 @@ void lock_unpin_lock(struct lockdep_map *lock) check_flags(flags); current->lockdep_recursion = 1; - __lock_unpin_lock(lock); + __lock_unpin_lock(lock, cookie); current->lockdep_recursion = 0; raw_local_irq_restore(flags); } diff --git a/kernel/locking/lockdep_proc.c b/kernel/locking/lockdep_proc.c index dbb61a3..a0f61ef 100644 --- a/kernel/locking/lockdep_proc.c +++ b/kernel/locking/lockdep_proc.c @@ -141,6 +141,8 @@ static int lc_show(struct seq_file *m, void *v) int i; if (v == SEQ_START_TOKEN) { + if (nr_chain_hlocks > MAX_LOCKDEP_CHAIN_HLOCKS) + seq_printf(m, "(buggered) "); seq_printf(m, "all lock chains:\n"); return 0; } diff --git a/kernel/locking/locktorture.c b/kernel/locking/locktorture.c index 8ef1919..f8c5af5 100644 --- a/kernel/locking/locktorture.c +++ b/kernel/locking/locktorture.c @@ -75,12 +75,7 @@ struct lock_stress_stats { long n_lock_acquired; }; -#if defined(MODULE) -#define LOCKTORTURE_RUNNABLE_INIT 1 -#else -#define LOCKTORTURE_RUNNABLE_INIT 0 -#endif -int torture_runnable = LOCKTORTURE_RUNNABLE_INIT; +int torture_runnable = IS_ENABLED(MODULE); module_param(torture_runnable, int, 0444); MODULE_PARM_DESC(torture_runnable, "Start locktorture at module init"); @@ -394,12 +389,12 @@ static void torture_rtmutex_boost(struct torture_random_state *trsp) if (!rt_task(current)) { /* - * (1) Boost priority once every ~50k operations. When the + * Boost priority once every ~50k operations. When the * task tries to take the lock, the rtmutex it will account * for the new priority, and do any corresponding pi-dance. */ - if (!(torture_random(trsp) % - (cxt.nrealwriters_stress * factor))) { + if (trsp && !(torture_random(trsp) % + (cxt.nrealwriters_stress * factor))) { policy = SCHED_FIFO; param.sched_priority = MAX_RT_PRIO - 1; } else /* common case, do nothing */ @@ -748,6 +743,15 @@ static void lock_torture_cleanup(void) if (torture_cleanup_begin()) return; + /* + * Indicates early cleanup, meaning that the test has not run, + * such as when passing bogus args when loading the module. As + * such, only perform the underlying torture-specific cleanups, + * and avoid anything related to locktorture. + */ + if (!cxt.lwsa) + goto end; + if (writer_tasks) { for (i = 0; i < cxt.nrealwriters_stress; i++) torture_stop_kthread(lock_torture_writer, @@ -776,6 +780,7 @@ static void lock_torture_cleanup(void) else lock_torture_print_module_parms(cxt.cur_ops, "End of test: SUCCESS"); +end: torture_cleanup_end(); } @@ -870,6 +875,7 @@ static int __init lock_torture_init(void) VERBOSE_TOROUT_STRING("cxt.lrsa: Out of memory"); firsterr = -ENOMEM; kfree(cxt.lwsa); + cxt.lwsa = NULL; goto unwind; } @@ -878,6 +884,7 @@ static int __init lock_torture_init(void) cxt.lrsa[i].n_lock_acquired = 0; } } + lock_torture_print_module_parms(cxt.cur_ops, "Start of test"); /* Prepare torture context. */ diff --git a/kernel/locking/qspinlock_stat.h b/kernel/locking/qspinlock_stat.h index eb2a2c9..22e0253 100644 --- a/kernel/locking/qspinlock_stat.h +++ b/kernel/locking/qspinlock_stat.h @@ -136,10 +136,12 @@ static ssize_t qstat_read(struct file *file, char __user *user_buf, } if (counter == qstat_pv_hash_hops) { - u64 frac; + u64 frac = 0; - frac = 100ULL * do_div(stat, kicks); - frac = DIV_ROUND_CLOSEST_ULL(frac, kicks); + if (kicks) { + frac = 100ULL * do_div(stat, kicks); + frac = DIV_ROUND_CLOSEST_ULL(frac, kicks); + } /* * Return a X.XX decimal number @@ -189,8 +191,6 @@ static ssize_t qstat_write(struct file *file, const char __user *user_buf, for (i = 0 ; i < qstat_num; i++) WRITE_ONCE(ptr[i], 0); - for (i = 0 ; i < qstat_num; i++) - WRITE_ONCE(ptr[i], 0); } return count; } @@ -212,10 +212,8 @@ static int __init init_qspinlock_stat(void) struct dentry *d_qstat = debugfs_create_dir("qlockstat", NULL); int i; - if (!d_qstat) { - pr_warn("Could not create 'qlockstat' debugfs directory\n"); - return 0; - } + if (!d_qstat) + goto out; /* * Create the debugfs files @@ -225,12 +223,20 @@ static int __init init_qspinlock_stat(void) * performance. */ for (i = 0; i < qstat_num; i++) - debugfs_create_file(qstat_names[i], 0400, d_qstat, - (void *)(long)i, &fops_qstat); + if (!debugfs_create_file(qstat_names[i], 0400, d_qstat, + (void *)(long)i, &fops_qstat)) + goto fail_undo; + + if (!debugfs_create_file(qstat_names[qstat_reset_cnts], 0200, d_qstat, + (void *)(long)qstat_reset_cnts, &fops_qstat)) + goto fail_undo; - debugfs_create_file(qstat_names[qstat_reset_cnts], 0200, d_qstat, - (void *)(long)qstat_reset_cnts, &fops_qstat); return 0; +fail_undo: + debugfs_remove_recursive(d_qstat); +out: + pr_warn("Could not create 'qlockstat' debugfs entries\n"); + return -ENOMEM; } fs_initcall(init_qspinlock_stat); diff --git a/kernel/locking/rwsem-spinlock.c b/kernel/locking/rwsem-spinlock.c index 3a50485..1591f6b 100644 --- a/kernel/locking/rwsem-spinlock.c +++ b/kernel/locking/rwsem-spinlock.c @@ -191,11 +191,12 @@ int __down_read_trylock(struct rw_semaphore *sem) /* * get a write lock on the semaphore */ -void __sched __down_write_nested(struct rw_semaphore *sem, int subclass) +int __sched __down_write_common(struct rw_semaphore *sem, int state) { struct rwsem_waiter waiter; struct task_struct *tsk; unsigned long flags; + int ret = 0; raw_spin_lock_irqsave(&sem->wait_lock, flags); @@ -215,21 +216,33 @@ void __sched __down_write_nested(struct rw_semaphore *sem, int subclass) */ if (sem->count == 0) break; - set_task_state(tsk, TASK_UNINTERRUPTIBLE); + if (signal_pending_state(state, current)) { + ret = -EINTR; + goto out; + } + set_task_state(tsk, state); raw_spin_unlock_irqrestore(&sem->wait_lock, flags); schedule(); raw_spin_lock_irqsave(&sem->wait_lock, flags); } /* got the lock */ sem->count = -1; +out: list_del(&waiter.list); raw_spin_unlock_irqrestore(&sem->wait_lock, flags); + + return ret; } void __sched __down_write(struct rw_semaphore *sem) { - __down_write_nested(sem, 0); + __down_write_common(sem, TASK_UNINTERRUPTIBLE); +} + +int __sched __down_write_killable(struct rw_semaphore *sem) +{ + return __down_write_common(sem, TASK_KILLABLE); } /* diff --git a/kernel/locking/rwsem-xadd.c b/kernel/locking/rwsem-xadd.c index a4d4de0..09e30c6 100644 --- a/kernel/locking/rwsem-xadd.c +++ b/kernel/locking/rwsem-xadd.c @@ -433,12 +433,13 @@ static inline bool rwsem_has_spinner(struct rw_semaphore *sem) /* * Wait until we successfully acquire the write lock */ -__visible -struct rw_semaphore __sched *rwsem_down_write_failed(struct rw_semaphore *sem) +static inline struct rw_semaphore * +__rwsem_down_write_failed_common(struct rw_semaphore *sem, int state) { long count; bool waiting = true; /* any queued threads before us */ struct rwsem_waiter waiter; + struct rw_semaphore *ret = sem; /* undo write bias from down_write operation, stop active locking */ count = rwsem_atomic_update(-RWSEM_ACTIVE_WRITE_BIAS, sem); @@ -478,7 +479,7 @@ struct rw_semaphore __sched *rwsem_down_write_failed(struct rw_semaphore *sem) count = rwsem_atomic_update(RWSEM_WAITING_BIAS, sem); /* wait until we successfully acquire the lock */ - set_current_state(TASK_UNINTERRUPTIBLE); + set_current_state(state); while (true) { if (rwsem_try_write_lock(count, sem)) break; @@ -486,21 +487,48 @@ struct rw_semaphore __sched *rwsem_down_write_failed(struct rw_semaphore *sem) /* Block until there are no active lockers. */ do { + if (signal_pending_state(state, current)) + goto out_nolock; + schedule(); - set_current_state(TASK_UNINTERRUPTIBLE); + set_current_state(state); } while ((count = sem->count) & RWSEM_ACTIVE_MASK); raw_spin_lock_irq(&sem->wait_lock); } __set_current_state(TASK_RUNNING); + list_del(&waiter.list); + raw_spin_unlock_irq(&sem->wait_lock); + return ret; + +out_nolock: + __set_current_state(TASK_RUNNING); + raw_spin_lock_irq(&sem->wait_lock); list_del(&waiter.list); + if (list_empty(&sem->wait_list)) + rwsem_atomic_update(-RWSEM_WAITING_BIAS, sem); + else + __rwsem_do_wake(sem, RWSEM_WAKE_ANY); raw_spin_unlock_irq(&sem->wait_lock); - return sem; + return ERR_PTR(-EINTR); +} + +__visible struct rw_semaphore * __sched +rwsem_down_write_failed(struct rw_semaphore *sem) +{ + return __rwsem_down_write_failed_common(sem, TASK_UNINTERRUPTIBLE); } EXPORT_SYMBOL(rwsem_down_write_failed); +__visible struct rw_semaphore * __sched +rwsem_down_write_failed_killable(struct rw_semaphore *sem) +{ + return __rwsem_down_write_failed_common(sem, TASK_KILLABLE); +} +EXPORT_SYMBOL(rwsem_down_write_failed_killable); + /* * handle waking up a waiter on the semaphore * - up_read/up_write has decremented the active part of count if we come here diff --git a/kernel/locking/rwsem.c b/kernel/locking/rwsem.c index 205be0c..c817216 100644 --- a/kernel/locking/rwsem.c +++ b/kernel/locking/rwsem.c @@ -55,6 +55,25 @@ void __sched down_write(struct rw_semaphore *sem) EXPORT_SYMBOL(down_write); /* + * lock for writing + */ +int __sched down_write_killable(struct rw_semaphore *sem) +{ + might_sleep(); + rwsem_acquire(&sem->dep_map, 0, 0, _RET_IP_); + + if (LOCK_CONTENDED_RETURN(sem, __down_write_trylock, __down_write_killable)) { + rwsem_release(&sem->dep_map, 1, _RET_IP_); + return -EINTR; + } + + rwsem_set_owner(sem); + return 0; +} + +EXPORT_SYMBOL(down_write_killable); + +/* * trylock for writing -- returns 1 if successful, 0 if contention */ int down_write_trylock(struct rw_semaphore *sem) diff --git a/kernel/power/swap.c b/kernel/power/swap.c index 12cd989..160e100 100644 --- a/kernel/power/swap.c +++ b/kernel/power/swap.c @@ -37,6 +37,14 @@ #define HIBERNATE_SIG "S1SUSPEND" /* + * When reading an {un,}compressed image, we may restore pages in place, + * in which case some architectures need these pages cleaning before they + * can be executed. We don't know which pages these may be, so clean the lot. + */ +static bool clean_pages_on_read; +static bool clean_pages_on_decompress; + +/* * The swap map is a data structure used for keeping track of each page * written to a swap partition. It consists of many swap_map_page * structures that contain each an array of MAP_PAGE_ENTRIES swap entries. @@ -241,6 +249,9 @@ static void hib_end_io(struct bio *bio) if (bio_data_dir(bio) == WRITE) put_page(page); + else if (clean_pages_on_read) + flush_icache_range((unsigned long)page_address(page), + (unsigned long)page_address(page) + PAGE_SIZE); if (bio->bi_error && !hb->error) hb->error = bio->bi_error; @@ -1049,6 +1060,7 @@ static int load_image(struct swap_map_handle *handle, hib_init_batch(&hb); + clean_pages_on_read = true; printk(KERN_INFO "PM: Loading image data pages (%u pages)...\n", nr_to_read); m = nr_to_read / 10; @@ -1124,6 +1136,10 @@ static int lzo_decompress_threadfn(void *data) d->unc_len = LZO_UNC_SIZE; d->ret = lzo1x_decompress_safe(d->cmp + LZO_HEADER, d->cmp_len, d->unc, &d->unc_len); + if (clean_pages_on_decompress) + flush_icache_range((unsigned long)d->unc, + (unsigned long)d->unc + d->unc_len); + atomic_set(&d->stop, 1); wake_up(&d->done); } @@ -1189,6 +1205,8 @@ static int load_image_lzo(struct swap_map_handle *handle, } memset(crc, 0, offsetof(struct crc_data, go)); + clean_pages_on_decompress = true; + /* * Start the decompression threads. */ diff --git a/kernel/rcu/Makefile b/kernel/rcu/Makefile index 032b2c0..18dfc48 100644 --- a/kernel/rcu/Makefile +++ b/kernel/rcu/Makefile @@ -5,6 +5,7 @@ KCOV_INSTRUMENT := n obj-y += update.o sync.o obj-$(CONFIG_SRCU) += srcu.o obj-$(CONFIG_RCU_TORTURE_TEST) += rcutorture.o +obj-$(CONFIG_RCU_PERF_TEST) += rcuperf.o obj-$(CONFIG_TREE_RCU) += tree.o obj-$(CONFIG_PREEMPT_RCU) += tree.o obj-$(CONFIG_TREE_RCU_TRACE) += tree_trace.o diff --git a/kernel/rcu/rcuperf.c b/kernel/rcu/rcuperf.c new file mode 100644 index 0000000..3cee0d8 --- /dev/null +++ b/kernel/rcu/rcuperf.c @@ -0,0 +1,655 @@ +/* + * Read-Copy Update module-based performance-test facility + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, you can access it online at + * http://www.gnu.org/licenses/gpl-2.0.html. + * + * Copyright (C) IBM Corporation, 2015 + * + * Authors: Paul E. McKenney <paulmck@us.ibm.com> + */ +#include <linux/types.h> +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/module.h> +#include <linux/kthread.h> +#include <linux/err.h> +#include <linux/spinlock.h> +#include <linux/smp.h> +#include <linux/rcupdate.h> +#include <linux/interrupt.h> +#include <linux/sched.h> +#include <linux/atomic.h> +#include <linux/bitops.h> +#include <linux/completion.h> +#include <linux/moduleparam.h> +#include <linux/percpu.h> +#include <linux/notifier.h> +#include <linux/reboot.h> +#include <linux/freezer.h> +#include <linux/cpu.h> +#include <linux/delay.h> +#include <linux/stat.h> +#include <linux/srcu.h> +#include <linux/slab.h> +#include <asm/byteorder.h> +#include <linux/torture.h> +#include <linux/vmalloc.h> + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Paul E. McKenney <paulmck@linux.vnet.ibm.com>"); + +#define PERF_FLAG "-perf:" +#define PERFOUT_STRING(s) \ + pr_alert("%s" PERF_FLAG s "\n", perf_type) +#define VERBOSE_PERFOUT_STRING(s) \ + do { if (verbose) pr_alert("%s" PERF_FLAG " %s\n", perf_type, s); } while (0) +#define VERBOSE_PERFOUT_ERRSTRING(s) \ + do { if (verbose) pr_alert("%s" PERF_FLAG "!!! %s\n", perf_type, s); } while (0) + +torture_param(bool, gp_exp, true, "Use expedited GP wait primitives"); +torture_param(int, holdoff, 10, "Holdoff time before test start (s)"); +torture_param(int, nreaders, -1, "Number of RCU reader threads"); +torture_param(int, nwriters, -1, "Number of RCU updater threads"); +torture_param(bool, shutdown, false, "Shutdown at end of performance tests."); +torture_param(bool, verbose, true, "Enable verbose debugging printk()s"); + +static char *perf_type = "rcu"; +module_param(perf_type, charp, 0444); +MODULE_PARM_DESC(perf_type, "Type of RCU to performance-test (rcu, rcu_bh, ...)"); + +static int nrealreaders; +static int nrealwriters; +static struct task_struct **writer_tasks; +static struct task_struct **reader_tasks; +static struct task_struct *shutdown_task; + +static u64 **writer_durations; +static int *writer_n_durations; +static atomic_t n_rcu_perf_reader_started; +static atomic_t n_rcu_perf_writer_started; +static atomic_t n_rcu_perf_writer_finished; +static wait_queue_head_t shutdown_wq; +static u64 t_rcu_perf_writer_started; +static u64 t_rcu_perf_writer_finished; +static unsigned long b_rcu_perf_writer_started; +static unsigned long b_rcu_perf_writer_finished; + +static int rcu_perf_writer_state; +#define RTWS_INIT 0 +#define RTWS_EXP_SYNC 1 +#define RTWS_SYNC 2 +#define RTWS_IDLE 2 +#define RTWS_STOPPING 3 + +#define MAX_MEAS 10000 +#define MIN_MEAS 100 + +#if defined(MODULE) || defined(CONFIG_RCU_PERF_TEST_RUNNABLE) +#define RCUPERF_RUNNABLE_INIT 1 +#else +#define RCUPERF_RUNNABLE_INIT 0 +#endif +static int perf_runnable = RCUPERF_RUNNABLE_INIT; +module_param(perf_runnable, int, 0444); +MODULE_PARM_DESC(perf_runnable, "Start rcuperf at boot"); + +/* + * Operations vector for selecting different types of tests. + */ + +struct rcu_perf_ops { + int ptype; + void (*init)(void); + void (*cleanup)(void); + int (*readlock)(void); + void (*readunlock)(int idx); + unsigned long (*started)(void); + unsigned long (*completed)(void); + unsigned long (*exp_completed)(void); + void (*sync)(void); + void (*exp_sync)(void); + const char *name; +}; + +static struct rcu_perf_ops *cur_ops; + +/* + * Definitions for rcu perf testing. + */ + +static int rcu_perf_read_lock(void) __acquires(RCU) +{ + rcu_read_lock(); + return 0; +} + +static void rcu_perf_read_unlock(int idx) __releases(RCU) +{ + rcu_read_unlock(); +} + +static unsigned long __maybe_unused rcu_no_completed(void) +{ + return 0; +} + +static void rcu_sync_perf_init(void) +{ +} + +static struct rcu_perf_ops rcu_ops = { + .ptype = RCU_FLAVOR, + .init = rcu_sync_perf_init, + .readlock = rcu_perf_read_lock, + .readunlock = rcu_perf_read_unlock, + .started = rcu_batches_started, + .completed = rcu_batches_completed, + .exp_completed = rcu_exp_batches_completed, + .sync = synchronize_rcu, + .exp_sync = synchronize_rcu_expedited, + .name = "rcu" +}; + +/* + * Definitions for rcu_bh perf testing. + */ + +static int rcu_bh_perf_read_lock(void) __acquires(RCU_BH) +{ + rcu_read_lock_bh(); + return 0; +} + +static void rcu_bh_perf_read_unlock(int idx) __releases(RCU_BH) +{ + rcu_read_unlock_bh(); +} + +static struct rcu_perf_ops rcu_bh_ops = { + .ptype = RCU_BH_FLAVOR, + .init = rcu_sync_perf_init, + .readlock = rcu_bh_perf_read_lock, + .readunlock = rcu_bh_perf_read_unlock, + .started = rcu_batches_started_bh, + .completed = rcu_batches_completed_bh, + .exp_completed = rcu_exp_batches_completed_sched, + .sync = synchronize_rcu_bh, + .exp_sync = synchronize_rcu_bh_expedited, + .name = "rcu_bh" +}; + +/* + * Definitions for srcu perf testing. + */ + +DEFINE_STATIC_SRCU(srcu_ctl_perf); +static struct srcu_struct *srcu_ctlp = &srcu_ctl_perf; + +static int srcu_perf_read_lock(void) __acquires(srcu_ctlp) +{ + return srcu_read_lock(srcu_ctlp); +} + +static void srcu_perf_read_unlock(int idx) __releases(srcu_ctlp) +{ + srcu_read_unlock(srcu_ctlp, idx); +} + +static unsigned long srcu_perf_completed(void) +{ + return srcu_batches_completed(srcu_ctlp); +} + +static void srcu_perf_synchronize(void) +{ + synchronize_srcu(srcu_ctlp); +} + +static void srcu_perf_synchronize_expedited(void) +{ + synchronize_srcu_expedited(srcu_ctlp); +} + +static struct rcu_perf_ops srcu_ops = { + .ptype = SRCU_FLAVOR, + .init = rcu_sync_perf_init, + .readlock = srcu_perf_read_lock, + .readunlock = srcu_perf_read_unlock, + .started = NULL, + .completed = srcu_perf_completed, + .exp_completed = srcu_perf_completed, + .sync = srcu_perf_synchronize, + .exp_sync = srcu_perf_synchronize_expedited, + .name = "srcu" +}; + +/* + * Definitions for sched perf testing. + */ + +static int sched_perf_read_lock(void) +{ + preempt_disable(); + return 0; +} + +static void sched_perf_read_unlock(int idx) +{ + preempt_enable(); +} + +static struct rcu_perf_ops sched_ops = { + .ptype = RCU_SCHED_FLAVOR, + .init = rcu_sync_perf_init, + .readlock = sched_perf_read_lock, + .readunlock = sched_perf_read_unlock, + .started = rcu_batches_started_sched, + .completed = rcu_batches_completed_sched, + .exp_completed = rcu_exp_batches_completed_sched, + .sync = synchronize_sched, + .exp_sync = synchronize_sched_expedited, + .name = "sched" +}; + +#ifdef CONFIG_TASKS_RCU + +/* + * Definitions for RCU-tasks perf testing. + */ + +static int tasks_perf_read_lock(void) +{ + return 0; +} + +static void tasks_perf_read_unlock(int idx) +{ +} + +static struct rcu_perf_ops tasks_ops = { + .ptype = RCU_TASKS_FLAVOR, + .init = rcu_sync_perf_init, + .readlock = tasks_perf_read_lock, + .readunlock = tasks_perf_read_unlock, + .started = rcu_no_completed, + .completed = rcu_no_completed, + .sync = synchronize_rcu_tasks, + .exp_sync = synchronize_rcu_tasks, + .name = "tasks" +}; + +#define RCUPERF_TASKS_OPS &tasks_ops, + +static bool __maybe_unused torturing_tasks(void) +{ + return cur_ops == &tasks_ops; +} + +#else /* #ifdef CONFIG_TASKS_RCU */ + +#define RCUPERF_TASKS_OPS + +static bool __maybe_unused torturing_tasks(void) +{ + return false; +} + +#endif /* #else #ifdef CONFIG_TASKS_RCU */ + +/* + * If performance tests complete, wait for shutdown to commence. + */ +static void rcu_perf_wait_shutdown(void) +{ + cond_resched_rcu_qs(); + if (atomic_read(&n_rcu_perf_writer_finished) < nrealwriters) + return; + while (!torture_must_stop()) + schedule_timeout_uninterruptible(1); +} + +/* + * RCU perf reader kthread. Repeatedly does empty RCU read-side + * critical section, minimizing update-side interference. + */ +static int +rcu_perf_reader(void *arg) +{ + unsigned long flags; + int idx; + long me = (long)arg; + + VERBOSE_PERFOUT_STRING("rcu_perf_reader task started"); + set_cpus_allowed_ptr(current, cpumask_of(me % nr_cpu_ids)); + set_user_nice(current, MAX_NICE); + atomic_inc(&n_rcu_perf_reader_started); + + do { + local_irq_save(flags); + idx = cur_ops->readlock(); + cur_ops->readunlock(idx); + local_irq_restore(flags); + rcu_perf_wait_shutdown(); + } while (!torture_must_stop()); + torture_kthread_stopping("rcu_perf_reader"); + return 0; +} + +/* + * RCU perf writer kthread. Repeatedly does a grace period. + */ +static int +rcu_perf_writer(void *arg) +{ + int i = 0; + int i_max; + long me = (long)arg; + struct sched_param sp; + bool started = false, done = false, alldone = false; + u64 t; + u64 *wdp; + u64 *wdpp = writer_durations[me]; + + VERBOSE_PERFOUT_STRING("rcu_perf_writer task started"); + WARN_ON(rcu_gp_is_expedited() && !rcu_gp_is_normal() && !gp_exp); + WARN_ON(rcu_gp_is_normal() && gp_exp); + WARN_ON(!wdpp); + set_cpus_allowed_ptr(current, cpumask_of(me % nr_cpu_ids)); + sp.sched_priority = 1; + sched_setscheduler_nocheck(current, SCHED_FIFO, &sp); + + if (holdoff) + schedule_timeout_uninterruptible(holdoff * HZ); + + t = ktime_get_mono_fast_ns(); + if (atomic_inc_return(&n_rcu_perf_writer_started) >= nrealwriters) { + t_rcu_perf_writer_started = t; + if (gp_exp) { + b_rcu_perf_writer_started = + cur_ops->exp_completed() / 2; + } else { + b_rcu_perf_writer_started = + cur_ops->completed(); + } + } + + do { + wdp = &wdpp[i]; + *wdp = ktime_get_mono_fast_ns(); + if (gp_exp) { + rcu_perf_writer_state = RTWS_EXP_SYNC; + cur_ops->exp_sync(); + } else { + rcu_perf_writer_state = RTWS_SYNC; + cur_ops->sync(); + } + rcu_perf_writer_state = RTWS_IDLE; + t = ktime_get_mono_fast_ns(); + *wdp = t - *wdp; + i_max = i; + if (!started && + atomic_read(&n_rcu_perf_writer_started) >= nrealwriters) + started = true; + if (!done && i >= MIN_MEAS) { + done = true; + sp.sched_priority = 0; + sched_setscheduler_nocheck(current, + SCHED_NORMAL, &sp); + pr_alert("%s" PERF_FLAG + "rcu_perf_writer %ld has %d measurements\n", + perf_type, me, MIN_MEAS); + if (atomic_inc_return(&n_rcu_perf_writer_finished) >= + nrealwriters) { + schedule_timeout_interruptible(10); + rcu_ftrace_dump(DUMP_ALL); + PERFOUT_STRING("Test complete"); + t_rcu_perf_writer_finished = t; + if (gp_exp) { + b_rcu_perf_writer_finished = + cur_ops->exp_completed() / 2; + } else { + b_rcu_perf_writer_finished = + cur_ops->completed(); + } + if (shutdown) { + smp_mb(); /* Assign before wake. */ + wake_up(&shutdown_wq); + } + } + } + if (done && !alldone && + atomic_read(&n_rcu_perf_writer_finished) >= nrealwriters) + alldone = true; + if (started && !alldone && i < MAX_MEAS - 1) + i++; + rcu_perf_wait_shutdown(); + } while (!torture_must_stop()); + rcu_perf_writer_state = RTWS_STOPPING; + writer_n_durations[me] = i_max; + torture_kthread_stopping("rcu_perf_writer"); + return 0; +} + +static inline void +rcu_perf_print_module_parms(struct rcu_perf_ops *cur_ops, const char *tag) +{ + pr_alert("%s" PERF_FLAG + "--- %s: nreaders=%d nwriters=%d verbose=%d shutdown=%d\n", + perf_type, tag, nrealreaders, nrealwriters, verbose, shutdown); +} + +static void +rcu_perf_cleanup(void) +{ + int i; + int j; + int ngps = 0; + u64 *wdp; + u64 *wdpp; + + if (torture_cleanup_begin()) + return; + + if (reader_tasks) { + for (i = 0; i < nrealreaders; i++) + torture_stop_kthread(rcu_perf_reader, + reader_tasks[i]); + kfree(reader_tasks); + } + + if (writer_tasks) { + for (i = 0; i < nrealwriters; i++) { + torture_stop_kthread(rcu_perf_writer, + writer_tasks[i]); + if (!writer_n_durations) + continue; + j = writer_n_durations[i]; + pr_alert("%s%s writer %d gps: %d\n", + perf_type, PERF_FLAG, i, j); + ngps += j; + } + pr_alert("%s%s start: %llu end: %llu duration: %llu gps: %d batches: %ld\n", + perf_type, PERF_FLAG, + t_rcu_perf_writer_started, t_rcu_perf_writer_finished, + t_rcu_perf_writer_finished - + t_rcu_perf_writer_started, + ngps, + b_rcu_perf_writer_finished - + b_rcu_perf_writer_started); + for (i = 0; i < nrealwriters; i++) { + if (!writer_durations) + break; + if (!writer_n_durations) + continue; + wdpp = writer_durations[i]; + if (!wdpp) + continue; + for (j = 0; j <= writer_n_durations[i]; j++) { + wdp = &wdpp[j]; + pr_alert("%s%s %4d writer-duration: %5d %llu\n", + perf_type, PERF_FLAG, + i, j, *wdp); + if (j % 100 == 0) + schedule_timeout_uninterruptible(1); + } + kfree(writer_durations[i]); + } + kfree(writer_tasks); + kfree(writer_durations); + kfree(writer_n_durations); + } + + /* Do flavor-specific cleanup operations. */ + if (cur_ops->cleanup != NULL) + cur_ops->cleanup(); + + torture_cleanup_end(); +} + +/* + * Return the number if non-negative. If -1, the number of CPUs. + * If less than -1, that much less than the number of CPUs, but + * at least one. + */ +static int compute_real(int n) +{ + int nr; + + if (n >= 0) { + nr = n; + } else { + nr = num_online_cpus() + 1 + n; + if (nr <= 0) + nr = 1; + } + return nr; +} + +/* + * RCU perf shutdown kthread. Just waits to be awakened, then shuts + * down system. + */ +static int +rcu_perf_shutdown(void *arg) +{ + do { + wait_event(shutdown_wq, + atomic_read(&n_rcu_perf_writer_finished) >= + nrealwriters); + } while (atomic_read(&n_rcu_perf_writer_finished) < nrealwriters); + smp_mb(); /* Wake before output. */ + rcu_perf_cleanup(); + kernel_power_off(); + return -EINVAL; +} + +static int __init +rcu_perf_init(void) +{ + long i; + int firsterr = 0; + static struct rcu_perf_ops *perf_ops[] = { + &rcu_ops, &rcu_bh_ops, &srcu_ops, &sched_ops, + RCUPERF_TASKS_OPS + }; + + if (!torture_init_begin(perf_type, verbose, &perf_runnable)) + return -EBUSY; + + /* Process args and tell the world that the perf'er is on the job. */ + for (i = 0; i < ARRAY_SIZE(perf_ops); i++) { + cur_ops = perf_ops[i]; + if (strcmp(perf_type, cur_ops->name) == 0) + break; + } + if (i == ARRAY_SIZE(perf_ops)) { + pr_alert("rcu-perf: invalid perf type: \"%s\"\n", + perf_type); + pr_alert("rcu-perf types:"); + for (i = 0; i < ARRAY_SIZE(perf_ops); i++) + pr_alert(" %s", perf_ops[i]->name); + pr_alert("\n"); + firsterr = -EINVAL; + goto unwind; + } + if (cur_ops->init) + cur_ops->init(); + + nrealwriters = compute_real(nwriters); + nrealreaders = compute_real(nreaders); + atomic_set(&n_rcu_perf_reader_started, 0); + atomic_set(&n_rcu_perf_writer_started, 0); + atomic_set(&n_rcu_perf_writer_finished, 0); + rcu_perf_print_module_parms(cur_ops, "Start of test"); + + /* Start up the kthreads. */ + + if (shutdown) { + init_waitqueue_head(&shutdown_wq); + firsterr = torture_create_kthread(rcu_perf_shutdown, NULL, + shutdown_task); + if (firsterr) + goto unwind; + schedule_timeout_uninterruptible(1); + } + reader_tasks = kcalloc(nrealreaders, sizeof(reader_tasks[0]), + GFP_KERNEL); + if (reader_tasks == NULL) { + VERBOSE_PERFOUT_ERRSTRING("out of memory"); + firsterr = -ENOMEM; + goto unwind; + } + for (i = 0; i < nrealreaders; i++) { + firsterr = torture_create_kthread(rcu_perf_reader, (void *)i, + reader_tasks[i]); + if (firsterr) + goto unwind; + } + while (atomic_read(&n_rcu_perf_reader_started) < nrealreaders) + schedule_timeout_uninterruptible(1); + writer_tasks = kcalloc(nrealwriters, sizeof(reader_tasks[0]), + GFP_KERNEL); + writer_durations = kcalloc(nrealwriters, sizeof(*writer_durations), + GFP_KERNEL); + writer_n_durations = + kcalloc(nrealwriters, sizeof(*writer_n_durations), + GFP_KERNEL); + if (!writer_tasks || !writer_durations || !writer_n_durations) { + VERBOSE_PERFOUT_ERRSTRING("out of memory"); + firsterr = -ENOMEM; + goto unwind; + } + for (i = 0; i < nrealwriters; i++) { + writer_durations[i] = + kcalloc(MAX_MEAS, sizeof(*writer_durations[i]), + GFP_KERNEL); + if (!writer_durations[i]) + goto unwind; + firsterr = torture_create_kthread(rcu_perf_writer, (void *)i, + writer_tasks[i]); + if (firsterr) + goto unwind; + } + torture_init_end(); + return 0; + +unwind: + torture_init_end(); + rcu_perf_cleanup(); + return firsterr; +} + +module_init(rcu_perf_init); +module_exit(rcu_perf_cleanup); diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c index 250ea67..084a28a 100644 --- a/kernel/rcu/rcutorture.c +++ b/kernel/rcu/rcutorture.c @@ -130,8 +130,8 @@ static struct rcu_torture __rcu *rcu_torture_current; static unsigned long rcu_torture_current_version; static struct rcu_torture rcu_tortures[10 * RCU_TORTURE_PIPE_LEN]; static DEFINE_SPINLOCK(rcu_torture_lock); -static DEFINE_PER_CPU(long [RCU_TORTURE_PIPE_LEN + 1], rcu_torture_count) = { 0 }; -static DEFINE_PER_CPU(long [RCU_TORTURE_PIPE_LEN + 1], rcu_torture_batch) = { 0 }; +static DEFINE_PER_CPU(long [RCU_TORTURE_PIPE_LEN + 1], rcu_torture_count); +static DEFINE_PER_CPU(long [RCU_TORTURE_PIPE_LEN + 1], rcu_torture_batch); static atomic_t rcu_torture_wcount[RCU_TORTURE_PIPE_LEN + 1]; static atomic_t n_rcu_torture_alloc; static atomic_t n_rcu_torture_alloc_fail; @@ -916,7 +916,7 @@ rcu_torture_fqs(void *arg) static int rcu_torture_writer(void *arg) { - bool can_expedite = !rcu_gp_is_expedited(); + bool can_expedite = !rcu_gp_is_expedited() && !rcu_gp_is_normal(); int expediting = 0; unsigned long gp_snap; bool gp_cond1 = gp_cond, gp_exp1 = gp_exp, gp_normal1 = gp_normal; @@ -932,7 +932,7 @@ rcu_torture_writer(void *arg) VERBOSE_TOROUT_STRING("rcu_torture_writer task started"); if (!can_expedite) { pr_alert("%s" TORTURE_FLAG - " Grace periods expedited from boot/sysfs for %s,\n", + " GP expediting controlled from boot/sysfs for %s,\n", torture_type, cur_ops->name); pr_alert("%s" TORTURE_FLAG " Disabled dynamic grace-period expediting.\n", @@ -1082,17 +1082,6 @@ rcu_torture_fakewriter(void *arg) return 0; } -static void rcutorture_trace_dump(void) -{ - static atomic_t beenhere = ATOMIC_INIT(0); - - if (atomic_read(&beenhere)) - return; - if (atomic_xchg(&beenhere, 1) != 0) - return; - ftrace_dump(DUMP_ALL); -} - /* * RCU torture reader from timer handler. Dereferences rcu_torture_current, * incrementing the corresponding element of the pipeline array. The @@ -1142,7 +1131,7 @@ static void rcu_torture_timer(unsigned long unused) if (pipe_count > 1) { do_trace_rcu_torture_read(cur_ops->name, &p->rtort_rcu, ts, started, completed); - rcutorture_trace_dump(); + rcu_ftrace_dump(DUMP_ALL); } __this_cpu_inc(rcu_torture_count[pipe_count]); completed = completed - started; @@ -1215,7 +1204,7 @@ rcu_torture_reader(void *arg) if (pipe_count > 1) { do_trace_rcu_torture_read(cur_ops->name, &p->rtort_rcu, ts, started, completed); - rcutorture_trace_dump(); + rcu_ftrace_dump(DUMP_ALL); } __this_cpu_inc(rcu_torture_count[pipe_count]); completed = completed - started; @@ -1333,7 +1322,7 @@ rcu_torture_stats_print(void) rcu_torture_writer_state, gpnum, completed, flags); show_rcu_gp_kthreads(); - rcutorture_trace_dump(); + rcu_ftrace_dump(DUMP_ALL); } rtcv_snap = rcu_torture_current_version; } @@ -1489,7 +1478,9 @@ static int rcu_torture_barrier_cbs(void *arg) * The above smp_load_acquire() ensures barrier_phase load * is ordered before the folloiwng ->call(). */ + local_irq_disable(); /* Just to test no-irq call_rcu(). */ cur_ops->call(&rcu, rcu_torture_barrier_cbf); + local_irq_enable(); if (atomic_dec_and_test(&barrier_cbs_count)) wake_up(&barrier_wq); } while (!torture_must_stop()); @@ -1596,7 +1587,7 @@ static int rcutorture_cpu_notify(struct notifier_block *self, { long cpu = (long)hcpu; - switch (action) { + switch (action & ~CPU_TASKS_FROZEN) { case CPU_ONLINE: case CPU_DOWN_FAILED: (void)rcutorture_booster_init(cpu); diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 9a535a8..c7f1bc4 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -102,6 +102,8 @@ struct rcu_state sname##_state = { \ .barrier_mutex = __MUTEX_INITIALIZER(sname##_state.barrier_mutex), \ .name = RCU_STATE_NAME(sname), \ .abbr = sabbr, \ + .exp_mutex = __MUTEX_INITIALIZER(sname##_state.exp_mutex), \ + .exp_wake_mutex = __MUTEX_INITIALIZER(sname##_state.exp_wake_mutex), \ } RCU_STATE_INITIALIZER(rcu_sched, 's', call_rcu_sched); @@ -370,6 +372,21 @@ void rcu_all_qs(void) rcu_momentary_dyntick_idle(); local_irq_restore(flags); } + if (unlikely(raw_cpu_read(rcu_sched_data.cpu_no_qs.b.exp))) { + /* + * Yes, we just checked a per-CPU variable with preemption + * enabled, so we might be migrated to some other CPU at + * this point. That is OK because in that case, the + * migration will supply the needed quiescent state. + * We might end up needlessly disabling preemption and + * invoking rcu_sched_qs() on the destination CPU, but + * the probability and cost are both quite low, so this + * should not be a problem in practice. + */ + preempt_disable(); + rcu_sched_qs(); + preempt_enable(); + } this_cpu_inc(rcu_qs_ctr); barrier(); /* Avoid RCU read-side critical sections leaking up. */ } @@ -385,9 +402,11 @@ module_param(qlowmark, long, 0444); static ulong jiffies_till_first_fqs = ULONG_MAX; static ulong jiffies_till_next_fqs = ULONG_MAX; +static bool rcu_kick_kthreads; module_param(jiffies_till_first_fqs, ulong, 0644); module_param(jiffies_till_next_fqs, ulong, 0644); +module_param(rcu_kick_kthreads, bool, 0644); /* * How long the grace period must be before we start recruiting @@ -460,6 +479,28 @@ unsigned long rcu_batches_completed_bh(void) EXPORT_SYMBOL_GPL(rcu_batches_completed_bh); /* + * Return the number of RCU expedited batches completed thus far for + * debug & stats. Odd numbers mean that a batch is in progress, even + * numbers mean idle. The value returned will thus be roughly double + * the cumulative batches since boot. + */ +unsigned long rcu_exp_batches_completed(void) +{ + return rcu_state_p->expedited_sequence; +} +EXPORT_SYMBOL_GPL(rcu_exp_batches_completed); + +/* + * Return the number of RCU-sched expedited batches completed thus far + * for debug & stats. Similar to rcu_exp_batches_completed(). + */ +unsigned long rcu_exp_batches_completed_sched(void) +{ + return rcu_sched_state.expedited_sequence; +} +EXPORT_SYMBOL_GPL(rcu_exp_batches_completed_sched); + +/* * Force a quiescent state. */ void rcu_force_quiescent_state(void) @@ -637,7 +678,7 @@ static void rcu_eqs_enter_common(long long oldval, bool user) idle_task(smp_processor_id()); trace_rcu_dyntick(TPS("Error on entry: not idle task"), oldval, 0); - ftrace_dump(DUMP_ORIG); + rcu_ftrace_dump(DUMP_ORIG); WARN_ONCE(1, "Current pid: %d comm: %s / Idle pid: %d comm: %s", current->pid, current->comm, idle->pid, idle->comm); /* must be idle task! */ @@ -799,7 +840,7 @@ static void rcu_eqs_exit_common(long long oldval, int user) trace_rcu_dyntick(TPS("Error on exit: not idle task"), oldval, rdtp->dynticks_nesting); - ftrace_dump(DUMP_ORIG); + rcu_ftrace_dump(DUMP_ORIG); WARN_ONCE(1, "Current pid: %d comm: %s / Idle pid: %d comm: %s", current->pid, current->comm, idle->pid, idle->comm); /* must be idle task! */ @@ -1224,8 +1265,10 @@ static void rcu_check_gp_kthread_starvation(struct rcu_state *rsp) rsp->gp_flags, gp_state_getname(rsp->gp_state), rsp->gp_state, rsp->gp_kthread ? rsp->gp_kthread->state : ~0); - if (rsp->gp_kthread) + if (rsp->gp_kthread) { sched_show_task(rsp->gp_kthread); + wake_up_process(rsp->gp_kthread); + } } } @@ -1249,6 +1292,25 @@ static void rcu_dump_cpu_stacks(struct rcu_state *rsp) } } +/* + * If too much time has passed in the current grace period, and if + * so configured, go kick the relevant kthreads. + */ +static void rcu_stall_kick_kthreads(struct rcu_state *rsp) +{ + unsigned long j; + + if (!rcu_kick_kthreads) + return; + j = READ_ONCE(rsp->jiffies_kick_kthreads); + if (time_after(jiffies, j) && rsp->gp_kthread) { + WARN_ONCE(1, "Kicking %s grace-period kthread\n", rsp->name); + rcu_ftrace_dump(DUMP_ALL); + wake_up_process(rsp->gp_kthread); + WRITE_ONCE(rsp->jiffies_kick_kthreads, j + HZ); + } +} + static void print_other_cpu_stall(struct rcu_state *rsp, unsigned long gpnum) { int cpu; @@ -1260,6 +1322,11 @@ static void print_other_cpu_stall(struct rcu_state *rsp, unsigned long gpnum) struct rcu_node *rnp = rcu_get_root(rsp); long totqlen = 0; + /* Kick and suppress, if so configured. */ + rcu_stall_kick_kthreads(rsp); + if (rcu_cpu_stall_suppress) + return; + /* Only let one CPU complain about others per time interval. */ raw_spin_lock_irqsave_rcu_node(rnp, flags); @@ -1333,6 +1400,11 @@ static void print_cpu_stall(struct rcu_state *rsp) struct rcu_node *rnp = rcu_get_root(rsp); long totqlen = 0; + /* Kick and suppress, if so configured. */ + rcu_stall_kick_kthreads(rsp); + if (rcu_cpu_stall_suppress) + return; + /* * OK, time to rat on ourselves... * See Documentation/RCU/stallwarn.txt for info on how to debug @@ -1377,8 +1449,10 @@ static void check_cpu_stall(struct rcu_state *rsp, struct rcu_data *rdp) unsigned long js; struct rcu_node *rnp; - if (rcu_cpu_stall_suppress || !rcu_gp_in_progress(rsp)) + if ((rcu_cpu_stall_suppress && !rcu_kick_kthreads) || + !rcu_gp_in_progress(rsp)) return; + rcu_stall_kick_kthreads(rsp); j = jiffies; /* @@ -2117,8 +2191,11 @@ static int __noreturn rcu_gp_kthread(void *arg) } ret = 0; for (;;) { - if (!ret) + if (!ret) { rsp->jiffies_force_qs = jiffies + j; + WRITE_ONCE(rsp->jiffies_kick_kthreads, + jiffies + 3 * j); + } trace_rcu_grace_period(rsp->name, READ_ONCE(rsp->gpnum), TPS("fqswait")); @@ -2144,6 +2221,15 @@ static int __noreturn rcu_gp_kthread(void *arg) TPS("fqsend")); cond_resched_rcu_qs(); WRITE_ONCE(rsp->gp_activity, jiffies); + ret = 0; /* Force full wait till next FQS. */ + j = jiffies_till_next_fqs; + if (j > HZ) { + j = HZ; + jiffies_till_next_fqs = HZ; + } else if (j < 1) { + j = 1; + jiffies_till_next_fqs = 1; + } } else { /* Deal with stray signal. */ cond_resched_rcu_qs(); @@ -2152,14 +2238,12 @@ static int __noreturn rcu_gp_kthread(void *arg) trace_rcu_grace_period(rsp->name, READ_ONCE(rsp->gpnum), TPS("fqswaitsig")); - } - j = jiffies_till_next_fqs; - if (j > HZ) { - j = HZ; - jiffies_till_next_fqs = HZ; - } else if (j < 1) { - j = 1; - jiffies_till_next_fqs = 1; + ret = 1; /* Keep old FQS timing. */ + j = jiffies; + if (time_after(jiffies, rsp->jiffies_force_qs)) + j = 1; + else + j = rsp->jiffies_force_qs - j; } } @@ -3376,8 +3460,12 @@ static void rcu_exp_gp_seq_end(struct rcu_state *rsp) } static unsigned long rcu_exp_gp_seq_snap(struct rcu_state *rsp) { + unsigned long s; + smp_mb(); /* Caller's modifications seen first by other CPUs. */ - return rcu_seq_snap(&rsp->expedited_sequence); + s = rcu_seq_snap(&rsp->expedited_sequence); + trace_rcu_exp_grace_period(rsp->name, s, TPS("snap")); + return s; } static bool rcu_exp_gp_seq_done(struct rcu_state *rsp, unsigned long s) { @@ -3469,7 +3557,7 @@ static void __maybe_unused sync_exp_reset_tree(struct rcu_state *rsp) * for the current expedited grace period. Works only for preemptible * RCU -- other RCU implementation use other means. * - * Caller must hold the root rcu_node's exp_funnel_mutex. + * Caller must hold the rcu_state's exp_mutex. */ static int sync_rcu_preempt_exp_done(struct rcu_node *rnp) { @@ -3485,8 +3573,8 @@ static int sync_rcu_preempt_exp_done(struct rcu_node *rnp) * recursively up the tree. (Calm down, calm down, we do the recursion * iteratively!) * - * Caller must hold the root rcu_node's exp_funnel_mutex and the - * specified rcu_node structure's ->lock. + * Caller must hold the rcu_state's exp_mutex and the specified rcu_node + * structure's ->lock. */ static void __rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp, bool wake, unsigned long flags) @@ -3523,7 +3611,7 @@ static void __rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp, * Report expedited quiescent state for specified node. This is a * lock-acquisition wrapper function for __rcu_report_exp_rnp(). * - * Caller must hold the root rcu_node's exp_funnel_mutex. + * Caller must hold the rcu_state's exp_mutex. */ static void __maybe_unused rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp, bool wake) @@ -3536,8 +3624,8 @@ static void __maybe_unused rcu_report_exp_rnp(struct rcu_state *rsp, /* * Report expedited quiescent state for multiple CPUs, all covered by the - * specified leaf rcu_node structure. Caller must hold the root - * rcu_node's exp_funnel_mutex. + * specified leaf rcu_node structure. Caller must hold the rcu_state's + * exp_mutex. */ static void rcu_report_exp_cpu_mult(struct rcu_state *rsp, struct rcu_node *rnp, unsigned long mask, bool wake) @@ -3555,7 +3643,6 @@ static void rcu_report_exp_cpu_mult(struct rcu_state *rsp, struct rcu_node *rnp, /* * Report expedited quiescent state for specified rcu_data (CPU). - * Caller must hold the root rcu_node's exp_funnel_mutex. */ static void rcu_report_exp_rdp(struct rcu_state *rsp, struct rcu_data *rdp, bool wake) @@ -3564,15 +3651,11 @@ static void rcu_report_exp_rdp(struct rcu_state *rsp, struct rcu_data *rdp, } /* Common code for synchronize_{rcu,sched}_expedited() work-done checking. */ -static bool sync_exp_work_done(struct rcu_state *rsp, struct rcu_node *rnp, - struct rcu_data *rdp, - atomic_long_t *stat, unsigned long s) +static bool sync_exp_work_done(struct rcu_state *rsp, atomic_long_t *stat, + unsigned long s) { if (rcu_exp_gp_seq_done(rsp, s)) { - if (rnp) - mutex_unlock(&rnp->exp_funnel_mutex); - else if (rdp) - mutex_unlock(&rdp->exp_funnel_mutex); + trace_rcu_exp_grace_period(rsp->name, s, TPS("done")); /* Ensure test happens before caller kfree(). */ smp_mb__before_atomic(); /* ^^^ */ atomic_long_inc(stat); @@ -3582,59 +3665,65 @@ static bool sync_exp_work_done(struct rcu_state *rsp, struct rcu_node *rnp, } /* - * Funnel-lock acquisition for expedited grace periods. Returns a - * pointer to the root rcu_node structure, or NULL if some other - * task did the expedited grace period for us. + * Funnel-lock acquisition for expedited grace periods. Returns true + * if some other task completed an expedited grace period that this task + * can piggy-back on, and with no mutex held. Otherwise, returns false + * with the mutex held, indicating that the caller must actually do the + * expedited grace period. */ -static struct rcu_node *exp_funnel_lock(struct rcu_state *rsp, unsigned long s) +static bool exp_funnel_lock(struct rcu_state *rsp, unsigned long s) { struct rcu_data *rdp = per_cpu_ptr(rsp->rda, raw_smp_processor_id()); - struct rcu_node *rnp0; - struct rcu_node *rnp1 = NULL; + struct rcu_node *rnp = rdp->mynode; + struct rcu_node *rnp_root = rcu_get_root(rsp); + + /* Low-contention fastpath. */ + if (ULONG_CMP_LT(READ_ONCE(rnp->exp_seq_rq), s) && + (rnp == rnp_root || + ULONG_CMP_LT(READ_ONCE(rnp_root->exp_seq_rq), s)) && + !mutex_is_locked(&rsp->exp_mutex) && + mutex_trylock(&rsp->exp_mutex)) + goto fastpath; /* - * First try directly acquiring the root lock in order to reduce - * latency in the common case where expedited grace periods are - * rare. We check mutex_is_locked() to avoid pathological levels of - * memory contention on ->exp_funnel_mutex in the heavy-load case. + * Each pass through the following loop works its way up + * the rcu_node tree, returning if others have done the work or + * otherwise falls through to acquire rsp->exp_mutex. The mapping + * from CPU to rcu_node structure can be inexact, as it is just + * promoting locality and is not strictly needed for correctness. */ - rnp0 = rcu_get_root(rsp); - if (!mutex_is_locked(&rnp0->exp_funnel_mutex)) { - if (mutex_trylock(&rnp0->exp_funnel_mutex)) { - if (sync_exp_work_done(rsp, rnp0, NULL, - &rdp->expedited_workdone0, s)) - return NULL; - return rnp0; + for (; rnp != NULL; rnp = rnp->parent) { + if (sync_exp_work_done(rsp, &rdp->exp_workdone1, s)) + return true; + + /* Work not done, either wait here or go up. */ + spin_lock(&rnp->exp_lock); + if (ULONG_CMP_GE(rnp->exp_seq_rq, s)) { + + /* Someone else doing GP, so wait for them. */ + spin_unlock(&rnp->exp_lock); + trace_rcu_exp_funnel_lock(rsp->name, rnp->level, + rnp->grplo, rnp->grphi, + TPS("wait")); + wait_event(rnp->exp_wq[(s >> 1) & 0x3], + sync_exp_work_done(rsp, + &rdp->exp_workdone2, s)); + return true; } + rnp->exp_seq_rq = s; /* Followers can wait on us. */ + spin_unlock(&rnp->exp_lock); + trace_rcu_exp_funnel_lock(rsp->name, rnp->level, rnp->grplo, + rnp->grphi, TPS("nxtlvl")); } - - /* - * Each pass through the following loop works its way - * up the rcu_node tree, returning if others have done the - * work or otherwise falls through holding the root rnp's - * ->exp_funnel_mutex. The mapping from CPU to rcu_node structure - * can be inexact, as it is just promoting locality and is not - * strictly needed for correctness. - */ - if (sync_exp_work_done(rsp, NULL, NULL, &rdp->expedited_workdone1, s)) - return NULL; - mutex_lock(&rdp->exp_funnel_mutex); - rnp0 = rdp->mynode; - for (; rnp0 != NULL; rnp0 = rnp0->parent) { - if (sync_exp_work_done(rsp, rnp1, rdp, - &rdp->expedited_workdone2, s)) - return NULL; - mutex_lock(&rnp0->exp_funnel_mutex); - if (rnp1) - mutex_unlock(&rnp1->exp_funnel_mutex); - else - mutex_unlock(&rdp->exp_funnel_mutex); - rnp1 = rnp0; + mutex_lock(&rsp->exp_mutex); +fastpath: + if (sync_exp_work_done(rsp, &rdp->exp_workdone3, s)) { + mutex_unlock(&rsp->exp_mutex); + return true; } - if (sync_exp_work_done(rsp, rnp1, rdp, - &rdp->expedited_workdone3, s)) - return NULL; - return rnp1; + rcu_exp_gp_seq_start(rsp); + trace_rcu_exp_grace_period(rsp->name, s, TPS("start")); + return false; } /* Invoked on each online non-idle CPU for expedited quiescent state. */ @@ -3649,6 +3738,11 @@ static void sync_sched_exp_handler(void *data) if (!(READ_ONCE(rnp->expmask) & rdp->grpmask) || __this_cpu_read(rcu_sched_data.cpu_no_qs.b.exp)) return; + if (rcu_is_cpu_rrupt_from_idle()) { + rcu_report_exp_rdp(&rcu_sched_state, + this_cpu_ptr(&rcu_sched_data), true); + return; + } __this_cpu_write(rcu_sched_data.cpu_no_qs.b.exp, true); resched_cpu(smp_processor_id()); } @@ -3773,7 +3867,7 @@ static void synchronize_sched_expedited_wait(struct rcu_state *rsp) rsp->name); ndetected = 0; rcu_for_each_leaf_node(rsp, rnp) { - ndetected = rcu_print_task_exp_stall(rnp); + ndetected += rcu_print_task_exp_stall(rnp); mask = 1; for (cpu = rnp->grplo; cpu <= rnp->grphi; cpu++, mask <<= 1) { struct rcu_data *rdp; @@ -3783,7 +3877,7 @@ static void synchronize_sched_expedited_wait(struct rcu_state *rsp) ndetected++; rdp = per_cpu_ptr(rsp->rda, cpu); pr_cont(" %d-%c%c%c", cpu, - "O."[cpu_online(cpu)], + "O."[!!cpu_online(cpu)], "o."[!!(rdp->grpmask & rnp->expmaskinit)], "N."[!!(rdp->grpmask & rnp->expmaskinitnext)]); } @@ -3792,7 +3886,7 @@ static void synchronize_sched_expedited_wait(struct rcu_state *rsp) pr_cont(" } %lu jiffies s: %lu root: %#lx/%c\n", jiffies - jiffies_start, rsp->expedited_sequence, rnp_root->expmask, ".T"[!!rnp_root->exp_tasks]); - if (!ndetected) { + if (ndetected) { pr_err("blocking rcu_node structures:"); rcu_for_each_node_breadth_first(rsp, rnp) { if (rnp == rnp_root) @@ -3818,6 +3912,41 @@ static void synchronize_sched_expedited_wait(struct rcu_state *rsp) } } +/* + * Wait for the current expedited grace period to complete, and then + * wake up everyone who piggybacked on the just-completed expedited + * grace period. Also update all the ->exp_seq_rq counters as needed + * in order to avoid counter-wrap problems. + */ +static void rcu_exp_wait_wake(struct rcu_state *rsp, unsigned long s) +{ + struct rcu_node *rnp; + + synchronize_sched_expedited_wait(rsp); + rcu_exp_gp_seq_end(rsp); + trace_rcu_exp_grace_period(rsp->name, s, TPS("end")); + + /* + * Switch over to wakeup mode, allowing the next GP, but -only- the + * next GP, to proceed. + */ + mutex_lock(&rsp->exp_wake_mutex); + mutex_unlock(&rsp->exp_mutex); + + rcu_for_each_node_breadth_first(rsp, rnp) { + if (ULONG_CMP_LT(READ_ONCE(rnp->exp_seq_rq), s)) { + spin_lock(&rnp->exp_lock); + /* Recheck, avoid hang in case someone just arrived. */ + if (ULONG_CMP_LT(rnp->exp_seq_rq, s)) + rnp->exp_seq_rq = s; + spin_unlock(&rnp->exp_lock); + } + wake_up_all(&rnp->exp_wq[(rsp->expedited_sequence >> 1) & 0x3]); + } + trace_rcu_exp_grace_period(rsp->name, s, TPS("endwake")); + mutex_unlock(&rsp->exp_wake_mutex); +} + /** * synchronize_sched_expedited - Brute-force RCU-sched grace period * @@ -3837,7 +3966,6 @@ static void synchronize_sched_expedited_wait(struct rcu_state *rsp) void synchronize_sched_expedited(void) { unsigned long s; - struct rcu_node *rnp; struct rcu_state *rsp = &rcu_sched_state; /* If only one CPU, this is automatically a grace period. */ @@ -3852,17 +3980,14 @@ void synchronize_sched_expedited(void) /* Take a snapshot of the sequence number. */ s = rcu_exp_gp_seq_snap(rsp); - - rnp = exp_funnel_lock(rsp, s); - if (rnp == NULL) + if (exp_funnel_lock(rsp, s)) return; /* Someone else did our work for us. */ - rcu_exp_gp_seq_start(rsp); + /* Initialize the rcu_node tree in preparation for the wait. */ sync_rcu_exp_select_cpus(rsp, sync_sched_exp_handler); - synchronize_sched_expedited_wait(rsp); - rcu_exp_gp_seq_end(rsp); - mutex_unlock(&rnp->exp_funnel_mutex); + /* Wait and clean up, including waking everyone. */ + rcu_exp_wait_wake(rsp, s); } EXPORT_SYMBOL_GPL(synchronize_sched_expedited); @@ -4162,7 +4287,6 @@ rcu_boot_init_percpu_data(int cpu, struct rcu_state *rsp) WARN_ON_ONCE(atomic_read(&rdp->dynticks->dynticks) != 1); rdp->cpu = cpu; rdp->rsp = rsp; - mutex_init(&rdp->exp_funnel_mutex); rcu_boot_init_nocb_percpu_data(rdp); raw_spin_unlock_irqrestore_rcu_node(rnp, flags); } @@ -4420,10 +4544,8 @@ static void __init rcu_init_one(struct rcu_state *rsp) { static const char * const buf[] = RCU_NODE_NAME_INIT; static const char * const fqs[] = RCU_FQS_NAME_INIT; - static const char * const exp[] = RCU_EXP_NAME_INIT; static struct lock_class_key rcu_node_class[RCU_NUM_LVLS]; static struct lock_class_key rcu_fqs_class[RCU_NUM_LVLS]; - static struct lock_class_key rcu_exp_class[RCU_NUM_LVLS]; static u8 fl_mask = 0x1; int levelcnt[RCU_NUM_LVLS]; /* # nodes in each level. */ @@ -4482,9 +4604,11 @@ static void __init rcu_init_one(struct rcu_state *rsp) rnp->level = i; INIT_LIST_HEAD(&rnp->blkd_tasks); rcu_init_one_nocb(rnp); - mutex_init(&rnp->exp_funnel_mutex); - lockdep_set_class_and_name(&rnp->exp_funnel_mutex, - &rcu_exp_class[i], exp[i]); + init_waitqueue_head(&rnp->exp_wq[0]); + init_waitqueue_head(&rnp->exp_wq[1]); + init_waitqueue_head(&rnp->exp_wq[2]); + init_waitqueue_head(&rnp->exp_wq[3]); + spin_lock_init(&rnp->exp_lock); } } diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h index df668c0..e3959f5 100644 --- a/kernel/rcu/tree.h +++ b/kernel/rcu/tree.h @@ -70,7 +70,6 @@ # define NUM_RCU_LVL_INIT { NUM_RCU_LVL_0 } # define RCU_NODE_NAME_INIT { "rcu_node_0" } # define RCU_FQS_NAME_INIT { "rcu_node_fqs_0" } -# define RCU_EXP_NAME_INIT { "rcu_node_exp_0" } #elif NR_CPUS <= RCU_FANOUT_2 # define RCU_NUM_LVLS 2 # define NUM_RCU_LVL_0 1 @@ -79,7 +78,6 @@ # define NUM_RCU_LVL_INIT { NUM_RCU_LVL_0, NUM_RCU_LVL_1 } # define RCU_NODE_NAME_INIT { "rcu_node_0", "rcu_node_1" } # define RCU_FQS_NAME_INIT { "rcu_node_fqs_0", "rcu_node_fqs_1" } -# define RCU_EXP_NAME_INIT { "rcu_node_exp_0", "rcu_node_exp_1" } #elif NR_CPUS <= RCU_FANOUT_3 # define RCU_NUM_LVLS 3 # define NUM_RCU_LVL_0 1 @@ -89,7 +87,6 @@ # define NUM_RCU_LVL_INIT { NUM_RCU_LVL_0, NUM_RCU_LVL_1, NUM_RCU_LVL_2 } # define RCU_NODE_NAME_INIT { "rcu_node_0", "rcu_node_1", "rcu_node_2" } # define RCU_FQS_NAME_INIT { "rcu_node_fqs_0", "rcu_node_fqs_1", "rcu_node_fqs_2" } -# define RCU_EXP_NAME_INIT { "rcu_node_exp_0", "rcu_node_exp_1", "rcu_node_exp_2" } #elif NR_CPUS <= RCU_FANOUT_4 # define RCU_NUM_LVLS 4 # define NUM_RCU_LVL_0 1 @@ -100,7 +97,6 @@ # define NUM_RCU_LVL_INIT { NUM_RCU_LVL_0, NUM_RCU_LVL_1, NUM_RCU_LVL_2, NUM_RCU_LVL_3 } # define RCU_NODE_NAME_INIT { "rcu_node_0", "rcu_node_1", "rcu_node_2", "rcu_node_3" } # define RCU_FQS_NAME_INIT { "rcu_node_fqs_0", "rcu_node_fqs_1", "rcu_node_fqs_2", "rcu_node_fqs_3" } -# define RCU_EXP_NAME_INIT { "rcu_node_exp_0", "rcu_node_exp_1", "rcu_node_exp_2", "rcu_node_exp_3" } #else # error "CONFIG_RCU_FANOUT insufficient for NR_CPUS" #endif /* #if (NR_CPUS) <= RCU_FANOUT_1 */ @@ -252,7 +248,9 @@ struct rcu_node { /* Counts of upcoming no-CB GP requests. */ raw_spinlock_t fqslock ____cacheline_internodealigned_in_smp; - struct mutex exp_funnel_mutex ____cacheline_internodealigned_in_smp; + spinlock_t exp_lock ____cacheline_internodealigned_in_smp; + unsigned long exp_seq_rq; + wait_queue_head_t exp_wq[4]; } ____cacheline_internodealigned_in_smp; /* @@ -387,11 +385,9 @@ struct rcu_data { #ifdef CONFIG_RCU_FAST_NO_HZ struct rcu_head oom_head; #endif /* #ifdef CONFIG_RCU_FAST_NO_HZ */ - struct mutex exp_funnel_mutex; - atomic_long_t expedited_workdone0; /* # done by others #0. */ - atomic_long_t expedited_workdone1; /* # done by others #1. */ - atomic_long_t expedited_workdone2; /* # done by others #2. */ - atomic_long_t expedited_workdone3; /* # done by others #3. */ + atomic_long_t exp_workdone1; /* # done by others #1. */ + atomic_long_t exp_workdone2; /* # done by others #2. */ + atomic_long_t exp_workdone3; /* # done by others #3. */ /* 7) Callback offloading. */ #ifdef CONFIG_RCU_NOCB_CPU @@ -505,6 +501,8 @@ struct rcu_state { /* _rcu_barrier(). */ /* End of fields guarded by barrier_mutex. */ + struct mutex exp_mutex; /* Serialize expedited GP. */ + struct mutex exp_wake_mutex; /* Serialize wakeup. */ unsigned long expedited_sequence; /* Take a ticket. */ atomic_long_t expedited_normal; /* # fallbacks to normal. */ atomic_t expedited_need_qs; /* # CPUs left to check in. */ @@ -513,6 +511,8 @@ struct rcu_state { unsigned long jiffies_force_qs; /* Time at which to invoke */ /* force_quiescent_state(). */ + unsigned long jiffies_kick_kthreads; /* Time at which to kick */ + /* kthreads, if configured. */ unsigned long n_force_qs; /* Number of calls to */ /* force_quiescent_state(). */ unsigned long n_force_qs_lh; /* ~Number of calls leaving */ diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h index efdf7b6..ff1cd4e 100644 --- a/kernel/rcu/tree_plugin.h +++ b/kernel/rcu/tree_plugin.h @@ -722,18 +722,22 @@ static void sync_rcu_exp_handler(void *info) * synchronize_rcu_expedited - Brute-force RCU grace period * * Wait for an RCU-preempt grace period, but expedite it. The basic - * idea is to invoke synchronize_sched_expedited() to push all the tasks to - * the ->blkd_tasks lists and wait for this list to drain. This consumes - * significant time on all CPUs and is unfriendly to real-time workloads, - * so is thus not recommended for any sort of common-case code. - * In fact, if you are using synchronize_rcu_expedited() in a loop, - * please restructure your code to batch your updates, and then Use a - * single synchronize_rcu() instead. + * idea is to IPI all non-idle non-nohz online CPUs. The IPI handler + * checks whether the CPU is in an RCU-preempt critical section, and + * if so, it sets a flag that causes the outermost rcu_read_unlock() + * to report the quiescent state. On the other hand, if the CPU is + * not in an RCU read-side critical section, the IPI handler reports + * the quiescent state immediately. + * + * Although this is a greate improvement over previous expedited + * implementations, it is still unfriendly to real-time workloads, so is + * thus not recommended for any sort of common-case code. In fact, if + * you are using synchronize_rcu_expedited() in a loop, please restructure + * your code to batch your updates, and then Use a single synchronize_rcu() + * instead. */ void synchronize_rcu_expedited(void) { - struct rcu_node *rnp; - struct rcu_node *rnp_unlock; struct rcu_state *rsp = rcu_state_p; unsigned long s; @@ -744,23 +748,14 @@ void synchronize_rcu_expedited(void) } s = rcu_exp_gp_seq_snap(rsp); - - rnp_unlock = exp_funnel_lock(rsp, s); - if (rnp_unlock == NULL) + if (exp_funnel_lock(rsp, s)) return; /* Someone else did our work for us. */ - rcu_exp_gp_seq_start(rsp); - /* Initialize the rcu_node tree in preparation for the wait. */ sync_rcu_exp_select_cpus(rsp, sync_rcu_exp_handler); - /* Wait for snapshotted ->blkd_tasks lists to drain. */ - rnp = rcu_get_root(rsp); - synchronize_sched_expedited_wait(rsp); - - /* Clean up and exit. */ - rcu_exp_gp_seq_end(rsp); - mutex_unlock(&rnp_unlock->exp_funnel_mutex); + /* Wait for ->blkd_tasks lists to drain, then wake everyone up. */ + rcu_exp_wait_wake(rsp, s); } EXPORT_SYMBOL_GPL(synchronize_rcu_expedited); diff --git a/kernel/rcu/tree_trace.c b/kernel/rcu/tree_trace.c index 1088e64..86782f9a 100644 --- a/kernel/rcu/tree_trace.c +++ b/kernel/rcu/tree_trace.c @@ -185,17 +185,16 @@ static int show_rcuexp(struct seq_file *m, void *v) int cpu; struct rcu_state *rsp = (struct rcu_state *)m->private; struct rcu_data *rdp; - unsigned long s0 = 0, s1 = 0, s2 = 0, s3 = 0; + unsigned long s1 = 0, s2 = 0, s3 = 0; for_each_possible_cpu(cpu) { rdp = per_cpu_ptr(rsp->rda, cpu); - s0 += atomic_long_read(&rdp->expedited_workdone0); - s1 += atomic_long_read(&rdp->expedited_workdone1); - s2 += atomic_long_read(&rdp->expedited_workdone2); - s3 += atomic_long_read(&rdp->expedited_workdone3); + s1 += atomic_long_read(&rdp->exp_workdone1); + s2 += atomic_long_read(&rdp->exp_workdone2); + s3 += atomic_long_read(&rdp->exp_workdone3); } - seq_printf(m, "s=%lu wd0=%lu wd1=%lu wd2=%lu wd3=%lu n=%lu enq=%d sc=%lu\n", - rsp->expedited_sequence, s0, s1, s2, s3, + seq_printf(m, "s=%lu wd1=%lu wd2=%lu wd3=%lu n=%lu enq=%d sc=%lu\n", + rsp->expedited_sequence, s1, s2, s3, atomic_long_read(&rsp->expedited_normal), atomic_read(&rsp->expedited_need_qs), rsp->expedited_sequence / 2); diff --git a/kernel/rcu/update.c b/kernel/rcu/update.c index ca828b4..3ccdc8e 100644 --- a/kernel/rcu/update.c +++ b/kernel/rcu/update.c @@ -67,7 +67,7 @@ static int rcu_normal_after_boot; module_param(rcu_normal_after_boot, int, 0); #endif /* #ifndef CONFIG_TINY_RCU */ -#if defined(CONFIG_DEBUG_LOCK_ALLOC) && defined(CONFIG_PREEMPT_COUNT) +#ifdef CONFIG_DEBUG_LOCK_ALLOC /** * rcu_read_lock_sched_held() - might we be in RCU-sched read-side critical section? * @@ -111,7 +111,7 @@ int rcu_read_lock_sched_held(void) return 0; if (debug_locks) lockdep_opinion = lock_is_held(&rcu_sched_lock_map); - return lockdep_opinion || preempt_count() != 0 || irqs_disabled(); + return lockdep_opinion || !preemptible(); } EXPORT_SYMBOL(rcu_read_lock_sched_held); #endif diff --git a/kernel/resource.c b/kernel/resource.c index 2e78ead..9b5f044 100644 --- a/kernel/resource.c +++ b/kernel/resource.c @@ -105,16 +105,25 @@ static int r_show(struct seq_file *m, void *v) { struct resource *root = m->private; struct resource *r = v, *p; + unsigned long long start, end; int width = root->end < 0x10000 ? 4 : 8; int depth; for (depth = 0, p = r; depth < MAX_IORES_LEVEL; depth++, p = p->parent) if (p->parent == root) break; + + if (file_ns_capable(m->file, &init_user_ns, CAP_SYS_ADMIN)) { + start = r->start; + end = r->end; + } else { + start = end = 0; + } + seq_printf(m, "%*s%0*llx-%0*llx : %s\n", depth * 2, "", - width, (unsigned long long) r->start, - width, (unsigned long long) r->end, + width, start, + width, end, r->name ? r->name : "<BAD>"); return 0; } diff --git a/kernel/sched/Makefile b/kernel/sched/Makefile index 414d9c1..5e59b83 100644 --- a/kernel/sched/Makefile +++ b/kernel/sched/Makefile @@ -24,3 +24,4 @@ obj-$(CONFIG_SCHEDSTATS) += stats.o obj-$(CONFIG_SCHED_DEBUG) += debug.o obj-$(CONFIG_CGROUP_CPUACCT) += cpuacct.o obj-$(CONFIG_CPU_FREQ) += cpufreq.o +obj-$(CONFIG_CPU_FREQ_GOV_SCHEDUTIL) += cpufreq_schedutil.o diff --git a/kernel/sched/clock.c b/kernel/sched/clock.c index fedb967..e85a725 100644 --- a/kernel/sched/clock.c +++ b/kernel/sched/clock.c @@ -318,6 +318,7 @@ u64 sched_clock_cpu(int cpu) return clock; } +EXPORT_SYMBOL_GPL(sched_clock_cpu); void sched_clock_tick(void) { @@ -363,39 +364,6 @@ void sched_clock_idle_wakeup_event(u64 delta_ns) } EXPORT_SYMBOL_GPL(sched_clock_idle_wakeup_event); -/* - * As outlined at the top, provides a fast, high resolution, nanosecond - * time source that is monotonic per cpu argument and has bounded drift - * between cpus. - * - * ######################### BIG FAT WARNING ########################## - * # when comparing cpu_clock(i) to cpu_clock(j) for i != j, time can # - * # go backwards !! # - * #################################################################### - */ -u64 cpu_clock(int cpu) -{ - if (!sched_clock_stable()) - return sched_clock_cpu(cpu); - - return sched_clock(); -} - -/* - * Similar to cpu_clock() for the current cpu. Time will only be observed - * to be monotonic if care is taken to only compare timestampt taken on the - * same CPU. - * - * See cpu_clock(). - */ -u64 local_clock(void) -{ - if (!sched_clock_stable()) - return sched_clock_cpu(raw_smp_processor_id()); - - return sched_clock(); -} - #else /* CONFIG_HAVE_UNSTABLE_SCHED_CLOCK */ void sched_clock_init(void) @@ -410,22 +378,8 @@ u64 sched_clock_cpu(int cpu) return sched_clock(); } - -u64 cpu_clock(int cpu) -{ - return sched_clock(); -} - -u64 local_clock(void) -{ - return sched_clock(); -} - #endif /* CONFIG_HAVE_UNSTABLE_SCHED_CLOCK */ -EXPORT_SYMBOL_GPL(cpu_clock); -EXPORT_SYMBOL_GPL(local_clock); - /* * Running clock - returns the time that has elapsed while a guest has been * running. diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 8b489fc..404c078 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -33,7 +33,7 @@ #include <linux/init.h> #include <linux/uaccess.h> #include <linux/highmem.h> -#include <asm/mmu_context.h> +#include <linux/mmu_context.h> #include <linux/interrupt.h> #include <linux/capability.h> #include <linux/completion.h> @@ -170,6 +170,71 @@ static struct rq *this_rq_lock(void) return rq; } +/* + * __task_rq_lock - lock the rq @p resides on. + */ +struct rq *__task_rq_lock(struct task_struct *p, struct rq_flags *rf) + __acquires(rq->lock) +{ + struct rq *rq; + + lockdep_assert_held(&p->pi_lock); + + for (;;) { + rq = task_rq(p); + raw_spin_lock(&rq->lock); + if (likely(rq == task_rq(p) && !task_on_rq_migrating(p))) { + rf->cookie = lockdep_pin_lock(&rq->lock); + return rq; + } + raw_spin_unlock(&rq->lock); + + while (unlikely(task_on_rq_migrating(p))) + cpu_relax(); + } +} + +/* + * task_rq_lock - lock p->pi_lock and lock the rq @p resides on. + */ +struct rq *task_rq_lock(struct task_struct *p, struct rq_flags *rf) + __acquires(p->pi_lock) + __acquires(rq->lock) +{ + struct rq *rq; + + for (;;) { + raw_spin_lock_irqsave(&p->pi_lock, rf->flags); + rq = task_rq(p); + raw_spin_lock(&rq->lock); + /* + * move_queued_task() task_rq_lock() + * + * ACQUIRE (rq->lock) + * [S] ->on_rq = MIGRATING [L] rq = task_rq() + * WMB (__set_task_cpu()) ACQUIRE (rq->lock); + * [S] ->cpu = new_cpu [L] task_rq() + * [L] ->on_rq + * RELEASE (rq->lock) + * + * If we observe the old cpu in task_rq_lock, the acquire of + * the old rq->lock will fully serialize against the stores. + * + * If we observe the new cpu in task_rq_lock, the acquire will + * pair with the WMB to ensure we must then also see migrating. + */ + if (likely(rq == task_rq(p) && !task_on_rq_migrating(p))) { + rf->cookie = lockdep_pin_lock(&rq->lock); + return rq; + } + raw_spin_unlock(&rq->lock); + raw_spin_unlock_irqrestore(&p->pi_lock, rf->flags); + + while (unlikely(task_on_rq_migrating(p))) + cpu_relax(); + } +} + #ifdef CONFIG_SCHED_HRTICK /* * Use HR-timers to deliver accurate preemption points. @@ -249,29 +314,6 @@ void hrtick_start(struct rq *rq, u64 delay) } } -static int -hotplug_hrtick(struct notifier_block *nfb, unsigned long action, void *hcpu) -{ - int cpu = (int)(long)hcpu; - - switch (action) { - case CPU_UP_CANCELED: - case CPU_UP_CANCELED_FROZEN: - case CPU_DOWN_PREPARE: - case CPU_DOWN_PREPARE_FROZEN: - case CPU_DEAD: - case CPU_DEAD_FROZEN: - hrtick_clear(cpu_rq(cpu)); - return NOTIFY_OK; - } - - return NOTIFY_DONE; -} - -static __init void init_hrtick(void) -{ - hotcpu_notifier(hotplug_hrtick, 0); -} #else /* * Called to set the hrtick timer state. @@ -288,10 +330,6 @@ void hrtick_start(struct rq *rq, u64 delay) hrtimer_start(&rq->hrtick_timer, ns_to_ktime(delay), HRTIMER_MODE_REL_PINNED); } - -static inline void init_hrtick(void) -{ -} #endif /* CONFIG_SMP */ static void init_rq_hrtick(struct rq *rq) @@ -315,10 +353,6 @@ static inline void hrtick_clear(struct rq *rq) static inline void init_rq_hrtick(struct rq *rq) { } - -static inline void init_hrtick(void) -{ -} #endif /* CONFIG_SCHED_HRTICK */ /* @@ -400,7 +434,7 @@ void wake_q_add(struct wake_q_head *head, struct task_struct *task) * wakeup due to that. * * This cmpxchg() implies a full barrier, which pairs with the write - * barrier implied by the wakeup in wake_up_list(). + * barrier implied by the wakeup in wake_up_q(). */ if (cmpxchg(&node->next, NULL, WAKE_Q_TAIL)) return; @@ -499,7 +533,10 @@ int get_nohz_timer_target(void) rcu_read_lock(); for_each_domain(cpu, sd) { for_each_cpu(i, sched_domain_span(sd)) { - if (!idle_cpu(i) && is_housekeeping_cpu(cpu)) { + if (cpu == i) + continue; + + if (!idle_cpu(i) && is_housekeeping_cpu(i)) { cpu = i; goto unlock; } @@ -596,17 +633,8 @@ bool sched_can_stop_tick(struct rq *rq) return false; /* - * FIFO realtime policy runs the highest priority task (after DEADLINE). - * Other runnable tasks are of a lower priority. The scheduler tick - * isn't needed. - */ - fifo_nr_running = rq->rt.rt_nr_running - rq->rt.rr_nr_running; - if (fifo_nr_running) - return true; - - /* - * Round-robin realtime tasks time slice with other tasks at the same - * realtime priority. + * If there are more than one RR tasks, we need the tick to effect the + * actual RR behaviour. */ if (rq->rt.rr_nr_running) { if (rq->rt.rr_nr_running == 1) @@ -615,8 +643,20 @@ bool sched_can_stop_tick(struct rq *rq) return false; } - /* Normal multitasking need periodic preemption checks */ - if (rq->cfs.nr_running > 1) + /* + * If there's no RR tasks, but FIFO tasks, we can skip the tick, no + * forced preemption between FIFO tasks. + */ + fifo_nr_running = rq->rt.rt_nr_running - rq->rt.rr_nr_running; + if (fifo_nr_running) + return true; + + /* + * If there are no DL,RR/FIFO tasks, there must only be CFS tasks left; + * if there's more than one we need the tick for involuntary + * preemption. + */ + if (rq->nr_running > 1) return false; return true; @@ -1082,12 +1122,20 @@ void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask) static int __set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask, bool check) { - unsigned long flags; - struct rq *rq; + const struct cpumask *cpu_valid_mask = cpu_active_mask; unsigned int dest_cpu; + struct rq_flags rf; + struct rq *rq; int ret = 0; - rq = task_rq_lock(p, &flags); + rq = task_rq_lock(p, &rf); + + if (p->flags & PF_KTHREAD) { + /* + * Kernel threads are allowed on online && !active CPUs + */ + cpu_valid_mask = cpu_online_mask; + } /* * Must re-check here, to close a race against __kthread_bind(), @@ -1101,22 +1149,32 @@ static int __set_cpus_allowed_ptr(struct task_struct *p, if (cpumask_equal(&p->cpus_allowed, new_mask)) goto out; - if (!cpumask_intersects(new_mask, cpu_active_mask)) { + if (!cpumask_intersects(new_mask, cpu_valid_mask)) { ret = -EINVAL; goto out; } do_set_cpus_allowed(p, new_mask); + if (p->flags & PF_KTHREAD) { + /* + * For kernel threads that do indeed end up on online && + * !active we want to ensure they are strict per-cpu threads. + */ + WARN_ON(cpumask_intersects(new_mask, cpu_online_mask) && + !cpumask_intersects(new_mask, cpu_active_mask) && + p->nr_cpus_allowed != 1); + } + /* Can the task run on the task's current CPU? If so, we're done */ if (cpumask_test_cpu(task_cpu(p), new_mask)) goto out; - dest_cpu = cpumask_any_and(cpu_active_mask, new_mask); + dest_cpu = cpumask_any_and(cpu_valid_mask, new_mask); if (task_running(rq, p) || p->state == TASK_WAKING) { struct migration_arg arg = { p, dest_cpu }; /* Need help from migration thread: drop lock and wait. */ - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); stop_one_cpu(cpu_of(rq), migration_cpu_stop, &arg); tlb_migrate_finish(p->mm); return 0; @@ -1125,12 +1183,12 @@ static int __set_cpus_allowed_ptr(struct task_struct *p, * OK, since we're going to drop the lock immediately * afterwards anyway. */ - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, rf.cookie); rq = move_queued_task(rq, p, dest_cpu); - lockdep_pin_lock(&rq->lock); + lockdep_repin_lock(&rq->lock, rf.cookie); } out: - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); return ret; } @@ -1314,8 +1372,8 @@ out: */ unsigned long wait_task_inactive(struct task_struct *p, long match_state) { - unsigned long flags; int running, queued; + struct rq_flags rf; unsigned long ncsw; struct rq *rq; @@ -1350,14 +1408,14 @@ unsigned long wait_task_inactive(struct task_struct *p, long match_state) * lock now, to be *sure*. If we're wrong, we'll * just go back and repeat. */ - rq = task_rq_lock(p, &flags); + rq = task_rq_lock(p, &rf); trace_sched_wait_task(p); running = task_running(rq, p); queued = task_on_rq_queued(p); ncsw = 0; if (!match_state || p->state == match_state) ncsw = p->nvcsw | LONG_MIN; /* sets MSB */ - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); /* * If it changed from the expected state, bail out now. @@ -1431,6 +1489,25 @@ EXPORT_SYMBOL_GPL(kick_process); /* * ->cpus_allowed is protected by both rq->lock and p->pi_lock + * + * A few notes on cpu_active vs cpu_online: + * + * - cpu_active must be a subset of cpu_online + * + * - on cpu-up we allow per-cpu kthreads on the online && !active cpu, + * see __set_cpus_allowed_ptr(). At this point the newly online + * cpu isn't yet part of the sched domains, and balancing will not + * see it. + * + * - on cpu-down we clear cpu_active() to mask the sched domains and + * avoid the load balancer to place new tasks on the to be removed + * cpu. Existing tasks will remain running there and will be taken + * off. + * + * This means that fallback selection must not select !active CPUs. + * And can assume that any active CPU must be online. Conversely + * select_task_rq() below may allow selection of !active CPUs in order + * to satisfy the above rules. */ static int select_fallback_rq(int cpu, struct task_struct *p) { @@ -1449,8 +1526,6 @@ static int select_fallback_rq(int cpu, struct task_struct *p) /* Look for allowed, online CPU in same node. */ for_each_cpu(dest_cpu, nodemask) { - if (!cpu_online(dest_cpu)) - continue; if (!cpu_active(dest_cpu)) continue; if (cpumask_test_cpu(dest_cpu, tsk_cpus_allowed(p))) @@ -1461,8 +1536,6 @@ static int select_fallback_rq(int cpu, struct task_struct *p) for (;;) { /* Any allowed, online CPU? */ for_each_cpu(dest_cpu, tsk_cpus_allowed(p)) { - if (!cpu_online(dest_cpu)) - continue; if (!cpu_active(dest_cpu)) continue; goto out; @@ -1512,8 +1585,10 @@ int select_task_rq(struct task_struct *p, int cpu, int sd_flags, int wake_flags) { lockdep_assert_held(&p->pi_lock); - if (p->nr_cpus_allowed > 1) + if (tsk_nr_cpus_allowed(p) > 1) cpu = p->sched_class->select_task_rq(p, cpu, sd_flags, wake_flags); + else + cpu = cpumask_any(tsk_cpus_allowed(p)); /* * In order not to call set_task_cpu() on a blocking task we need @@ -1601,8 +1676,8 @@ static inline void ttwu_activate(struct rq *rq, struct task_struct *p, int en_fl /* * Mark the task runnable and perform wakeup-preemption. */ -static void -ttwu_do_wakeup(struct rq *rq, struct task_struct *p, int wake_flags) +static void ttwu_do_wakeup(struct rq *rq, struct task_struct *p, int wake_flags, + struct pin_cookie cookie) { check_preempt_curr(rq, p, wake_flags); p->state = TASK_RUNNING; @@ -1614,9 +1689,9 @@ ttwu_do_wakeup(struct rq *rq, struct task_struct *p, int wake_flags) * Our task @p is fully woken up and running; so its safe to * drop the rq->lock, hereafter rq is only used for statistics. */ - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, cookie); p->sched_class->task_woken(rq, p); - lockdep_pin_lock(&rq->lock); + lockdep_repin_lock(&rq->lock, cookie); } if (rq->idle_stamp) { @@ -1634,17 +1709,23 @@ ttwu_do_wakeup(struct rq *rq, struct task_struct *p, int wake_flags) } static void -ttwu_do_activate(struct rq *rq, struct task_struct *p, int wake_flags) +ttwu_do_activate(struct rq *rq, struct task_struct *p, int wake_flags, + struct pin_cookie cookie) { + int en_flags = ENQUEUE_WAKEUP; + lockdep_assert_held(&rq->lock); #ifdef CONFIG_SMP if (p->sched_contributes_to_load) rq->nr_uninterruptible--; + + if (wake_flags & WF_MIGRATED) + en_flags |= ENQUEUE_MIGRATED; #endif - ttwu_activate(rq, p, ENQUEUE_WAKEUP | ENQUEUE_WAKING); - ttwu_do_wakeup(rq, p, wake_flags); + ttwu_activate(rq, p, en_flags); + ttwu_do_wakeup(rq, p, wake_flags, cookie); } /* @@ -1655,17 +1736,18 @@ ttwu_do_activate(struct rq *rq, struct task_struct *p, int wake_flags) */ static int ttwu_remote(struct task_struct *p, int wake_flags) { + struct rq_flags rf; struct rq *rq; int ret = 0; - rq = __task_rq_lock(p); + rq = __task_rq_lock(p, &rf); if (task_on_rq_queued(p)) { /* check_preempt_curr() may use rq clock */ update_rq_clock(rq); - ttwu_do_wakeup(rq, p, wake_flags); + ttwu_do_wakeup(rq, p, wake_flags, rf.cookie); ret = 1; } - __task_rq_unlock(rq); + __task_rq_unlock(rq, &rf); return ret; } @@ -1675,6 +1757,7 @@ void sched_ttwu_pending(void) { struct rq *rq = this_rq(); struct llist_node *llist = llist_del_all(&rq->wake_list); + struct pin_cookie cookie; struct task_struct *p; unsigned long flags; @@ -1682,15 +1765,19 @@ void sched_ttwu_pending(void) return; raw_spin_lock_irqsave(&rq->lock, flags); - lockdep_pin_lock(&rq->lock); + cookie = lockdep_pin_lock(&rq->lock); while (llist) { p = llist_entry(llist, struct task_struct, wake_entry); llist = llist_next(llist); - ttwu_do_activate(rq, p, 0); + /* + * See ttwu_queue(); we only call ttwu_queue_remote() when + * its a x-cpu wakeup. + */ + ttwu_do_activate(rq, p, WF_MIGRATED, cookie); } - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, cookie); raw_spin_unlock_irqrestore(&rq->lock, flags); } @@ -1774,9 +1861,10 @@ bool cpus_share_cache(int this_cpu, int that_cpu) } #endif /* CONFIG_SMP */ -static void ttwu_queue(struct task_struct *p, int cpu) +static void ttwu_queue(struct task_struct *p, int cpu, int wake_flags) { struct rq *rq = cpu_rq(cpu); + struct pin_cookie cookie; #if defined(CONFIG_SMP) if (sched_feat(TTWU_QUEUE) && !cpus_share_cache(smp_processor_id(), cpu)) { @@ -1787,9 +1875,9 @@ static void ttwu_queue(struct task_struct *p, int cpu) #endif raw_spin_lock(&rq->lock); - lockdep_pin_lock(&rq->lock); - ttwu_do_activate(rq, p, 0); - lockdep_unpin_lock(&rq->lock); + cookie = lockdep_pin_lock(&rq->lock); + ttwu_do_activate(rq, p, wake_flags, cookie); + lockdep_unpin_lock(&rq->lock, cookie); raw_spin_unlock(&rq->lock); } @@ -1958,9 +2046,6 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags) p->sched_contributes_to_load = !!task_contributes_to_load(p); p->state = TASK_WAKING; - if (p->sched_class->task_waking) - p->sched_class->task_waking(p); - cpu = select_task_rq(p, p->wake_cpu, SD_BALANCE_WAKE, wake_flags); if (task_cpu(p) != cpu) { wake_flags |= WF_MIGRATED; @@ -1968,7 +2053,7 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags) } #endif /* CONFIG_SMP */ - ttwu_queue(p, cpu); + ttwu_queue(p, cpu, wake_flags); stat: if (schedstat_enabled()) ttwu_stat(p, cpu, wake_flags); @@ -1986,7 +2071,7 @@ out: * ensure that this_rq() is locked, @p is bound to this_rq() and not * the current task. */ -static void try_to_wake_up_local(struct task_struct *p) +static void try_to_wake_up_local(struct task_struct *p, struct pin_cookie cookie) { struct rq *rq = task_rq(p); @@ -2003,11 +2088,11 @@ static void try_to_wake_up_local(struct task_struct *p) * disabled avoiding further scheduler activity on it and we've * not yet picked a replacement task. */ - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, cookie); raw_spin_unlock(&rq->lock); raw_spin_lock(&p->pi_lock); raw_spin_lock(&rq->lock); - lockdep_pin_lock(&rq->lock); + lockdep_repin_lock(&rq->lock, cookie); } if (!(p->state & TASK_NORMAL)) @@ -2018,7 +2103,7 @@ static void try_to_wake_up_local(struct task_struct *p) if (!task_on_rq_queued(p)) ttwu_activate(rq, p, ENQUEUE_WAKEUP); - ttwu_do_wakeup(rq, p, 0); + ttwu_do_wakeup(rq, p, 0, cookie); if (schedstat_enabled()) ttwu_stat(p, smp_processor_id(), 0); out: @@ -2378,7 +2463,8 @@ static int dl_overflow(struct task_struct *p, int policy, u64 new_bw = dl_policy(policy) ? to_ratio(period, runtime) : 0; int cpus, err = -1; - if (new_bw == p->dl.dl_bw) + /* !deadline task may carry old deadline bandwidth */ + if (new_bw == p->dl.dl_bw && task_has_dl_policy(p)) return 0; /* @@ -2417,12 +2503,12 @@ extern void init_dl_bw(struct dl_bw *dl_b); */ void wake_up_new_task(struct task_struct *p) { - unsigned long flags; + struct rq_flags rf; struct rq *rq; - raw_spin_lock_irqsave(&p->pi_lock, flags); /* Initialize new task's runnable average */ init_entity_runnable_average(&p->se); + raw_spin_lock_irqsave(&p->pi_lock, rf.flags); #ifdef CONFIG_SMP /* * Fork balancing, do it here and not earlier because: @@ -2431,8 +2517,10 @@ void wake_up_new_task(struct task_struct *p) */ set_task_cpu(p, select_task_rq(p, task_cpu(p), SD_BALANCE_FORK, 0)); #endif + /* Post initialize new task's util average when its cfs_rq is set */ + post_init_entity_util_avg(&p->se); - rq = __task_rq_lock(p); + rq = __task_rq_lock(p, &rf); activate_task(rq, p, 0); p->on_rq = TASK_ON_RQ_QUEUED; trace_sched_wakeup_new(p); @@ -2443,12 +2531,12 @@ void wake_up_new_task(struct task_struct *p) * Nothing relies on rq->lock after this, so its fine to * drop it. */ - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, rf.cookie); p->sched_class->task_woken(rq, p); - lockdep_pin_lock(&rq->lock); + lockdep_repin_lock(&rq->lock, rf.cookie); } #endif - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); } #ifdef CONFIG_PREEMPT_NOTIFIERS @@ -2710,7 +2798,7 @@ asmlinkage __visible void schedule_tail(struct task_struct *prev) */ static __always_inline struct rq * context_switch(struct rq *rq, struct task_struct *prev, - struct task_struct *next) + struct task_struct *next, struct pin_cookie cookie) { struct mm_struct *mm, *oldmm; @@ -2730,7 +2818,7 @@ context_switch(struct rq *rq, struct task_struct *prev, atomic_inc(&oldmm->mm_count); enter_lazy_tlb(oldmm, next); } else - switch_mm(oldmm, mm, next); + switch_mm_irqs_off(oldmm, mm, next); if (!prev->mm) { prev->active_mm = NULL; @@ -2742,7 +2830,7 @@ context_switch(struct rq *rq, struct task_struct *prev, * of the scheduler it's an obvious special-case), so we * do an early lockdep release here: */ - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, cookie); spin_release(&rq->lock.dep_map, 1, _THIS_IP_); /* Here we just switch the register state and the stack. */ @@ -2864,7 +2952,7 @@ EXPORT_PER_CPU_SYMBOL(kernel_cpustat); */ unsigned long long task_sched_runtime(struct task_struct *p) { - unsigned long flags; + struct rq_flags rf; struct rq *rq; u64 ns; @@ -2884,7 +2972,7 @@ unsigned long long task_sched_runtime(struct task_struct *p) return p->se.sum_exec_runtime; #endif - rq = task_rq_lock(p, &flags); + rq = task_rq_lock(p, &rf); /* * Must be ->curr _and_ ->on_rq. If dequeued, we would * project cycles that may never be accounted to this @@ -2895,7 +2983,7 @@ unsigned long long task_sched_runtime(struct task_struct *p) p->sched_class->update_curr(rq); } ns = p->se.sum_exec_runtime; - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); return ns; } @@ -2915,7 +3003,7 @@ void scheduler_tick(void) raw_spin_lock(&rq->lock); update_rq_clock(rq); curr->sched_class->task_tick(rq, curr, 0); - update_cpu_load_active(rq); + cpu_load_update_active(rq); calc_global_load_tick(rq); raw_spin_unlock(&rq->lock); @@ -2958,6 +3046,20 @@ u64 scheduler_tick_max_deferment(void) #if defined(CONFIG_PREEMPT) && (defined(CONFIG_DEBUG_PREEMPT) || \ defined(CONFIG_PREEMPT_TRACER)) +/* + * If the value passed in is equal to the current preempt count + * then we just disabled preemption. Start timing the latency. + */ +static inline void preempt_latency_start(int val) +{ + if (preempt_count() == val) { + unsigned long ip = get_lock_parent_ip(); +#ifdef CONFIG_DEBUG_PREEMPT + current->preempt_disable_ip = ip; +#endif + trace_preempt_off(CALLER_ADDR0, ip); + } +} void preempt_count_add(int val) { @@ -2976,17 +3078,21 @@ void preempt_count_add(int val) DEBUG_LOCKS_WARN_ON((preempt_count() & PREEMPT_MASK) >= PREEMPT_MASK - 10); #endif - if (preempt_count() == val) { - unsigned long ip = get_lock_parent_ip(); -#ifdef CONFIG_DEBUG_PREEMPT - current->preempt_disable_ip = ip; -#endif - trace_preempt_off(CALLER_ADDR0, ip); - } + preempt_latency_start(val); } EXPORT_SYMBOL(preempt_count_add); NOKPROBE_SYMBOL(preempt_count_add); +/* + * If the value passed in equals to the current preempt count + * then we just enabled preemption. Stop timing the latency. + */ +static inline void preempt_latency_stop(int val) +{ + if (preempt_count() == val) + trace_preempt_on(CALLER_ADDR0, get_lock_parent_ip()); +} + void preempt_count_sub(int val) { #ifdef CONFIG_DEBUG_PREEMPT @@ -3003,13 +3109,15 @@ void preempt_count_sub(int val) return; #endif - if (preempt_count() == val) - trace_preempt_on(CALLER_ADDR0, get_lock_parent_ip()); + preempt_latency_stop(val); __preempt_count_sub(val); } EXPORT_SYMBOL(preempt_count_sub); NOKPROBE_SYMBOL(preempt_count_sub); +#else +static inline void preempt_latency_start(int val) { } +static inline void preempt_latency_stop(int val) { } #endif /* @@ -3062,7 +3170,7 @@ static inline void schedule_debug(struct task_struct *prev) * Pick up the highest-prio task: */ static inline struct task_struct * -pick_next_task(struct rq *rq, struct task_struct *prev) +pick_next_task(struct rq *rq, struct task_struct *prev, struct pin_cookie cookie) { const struct sched_class *class = &fair_sched_class; struct task_struct *p; @@ -3073,20 +3181,20 @@ pick_next_task(struct rq *rq, struct task_struct *prev) */ if (likely(prev->sched_class == class && rq->nr_running == rq->cfs.h_nr_running)) { - p = fair_sched_class.pick_next_task(rq, prev); + p = fair_sched_class.pick_next_task(rq, prev, cookie); if (unlikely(p == RETRY_TASK)) goto again; /* assumes fair_sched_class->next == idle_sched_class */ if (unlikely(!p)) - p = idle_sched_class.pick_next_task(rq, prev); + p = idle_sched_class.pick_next_task(rq, prev, cookie); return p; } again: for_each_class(class) { - p = class->pick_next_task(rq, prev); + p = class->pick_next_task(rq, prev, cookie); if (p) { if (unlikely(p == RETRY_TASK)) goto again; @@ -3140,6 +3248,7 @@ static void __sched notrace __schedule(bool preempt) { struct task_struct *prev, *next; unsigned long *switch_count; + struct pin_cookie cookie; struct rq *rq; int cpu; @@ -3173,7 +3282,7 @@ static void __sched notrace __schedule(bool preempt) */ smp_mb__before_spinlock(); raw_spin_lock(&rq->lock); - lockdep_pin_lock(&rq->lock); + cookie = lockdep_pin_lock(&rq->lock); rq->clock_skip_update <<= 1; /* promote REQ to ACT */ @@ -3195,7 +3304,7 @@ static void __sched notrace __schedule(bool preempt) to_wakeup = wq_worker_sleeping(prev); if (to_wakeup) - try_to_wake_up_local(to_wakeup); + try_to_wake_up_local(to_wakeup, cookie); } } switch_count = &prev->nvcsw; @@ -3204,7 +3313,7 @@ static void __sched notrace __schedule(bool preempt) if (task_on_rq_queued(prev)) update_rq_clock(rq); - next = pick_next_task(rq, prev); + next = pick_next_task(rq, prev, cookie); clear_tsk_need_resched(prev); clear_preempt_need_resched(); rq->clock_skip_update = 0; @@ -3215,9 +3324,9 @@ static void __sched notrace __schedule(bool preempt) ++*switch_count; trace_sched_switch(preempt, prev, next); - rq = context_switch(rq, prev, next); /* unlocks the rq */ + rq = context_switch(rq, prev, next, cookie); /* unlocks the rq */ } else { - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, cookie); raw_spin_unlock_irq(&rq->lock); } @@ -3284,8 +3393,23 @@ void __sched schedule_preempt_disabled(void) static void __sched notrace preempt_schedule_common(void) { do { + /* + * Because the function tracer can trace preempt_count_sub() + * and it also uses preempt_enable/disable_notrace(), if + * NEED_RESCHED is set, the preempt_enable_notrace() called + * by the function tracer will call this function again and + * cause infinite recursion. + * + * Preemption must be disabled here before the function + * tracer can trace. Break up preempt_disable() into two + * calls. One to disable preemption without fear of being + * traced. The other to still record the preemption latency, + * which can also be traced by the function tracer. + */ preempt_disable_notrace(); + preempt_latency_start(1); __schedule(true); + preempt_latency_stop(1); preempt_enable_no_resched_notrace(); /* @@ -3337,7 +3461,21 @@ asmlinkage __visible void __sched notrace preempt_schedule_notrace(void) return; do { + /* + * Because the function tracer can trace preempt_count_sub() + * and it also uses preempt_enable/disable_notrace(), if + * NEED_RESCHED is set, the preempt_enable_notrace() called + * by the function tracer will call this function again and + * cause infinite recursion. + * + * Preemption must be disabled here before the function + * tracer can trace. Break up preempt_disable() into two + * calls. One to disable preemption without fear of being + * traced. The other to still record the preemption latency, + * which can also be traced by the function tracer. + */ preempt_disable_notrace(); + preempt_latency_start(1); /* * Needs preempt disabled in case user_exit() is traced * and the tracer calls preempt_enable_notrace() causing @@ -3347,6 +3485,7 @@ asmlinkage __visible void __sched notrace preempt_schedule_notrace(void) __schedule(true); exception_exit(prev_ctx); + preempt_latency_stop(1); preempt_enable_no_resched_notrace(); } while (need_resched()); } @@ -3403,12 +3542,13 @@ EXPORT_SYMBOL(default_wake_function); void rt_mutex_setprio(struct task_struct *p, int prio) { int oldprio, queued, running, queue_flag = DEQUEUE_SAVE | DEQUEUE_MOVE; - struct rq *rq; const struct sched_class *prev_class; + struct rq_flags rf; + struct rq *rq; BUG_ON(prio > MAX_PRIO); - rq = __task_rq_lock(p); + rq = __task_rq_lock(p, &rf); /* * Idle task boosting is a nono in general. There is one @@ -3484,7 +3624,7 @@ void rt_mutex_setprio(struct task_struct *p, int prio) check_class_changed(rq, p, prev_class, oldprio); out_unlock: preempt_disable(); /* avoid rq from going away on us */ - __task_rq_unlock(rq); + __task_rq_unlock(rq, &rf); balance_callback(rq); preempt_enable(); @@ -3494,7 +3634,7 @@ out_unlock: void set_user_nice(struct task_struct *p, long nice) { int old_prio, delta, queued; - unsigned long flags; + struct rq_flags rf; struct rq *rq; if (task_nice(p) == nice || nice < MIN_NICE || nice > MAX_NICE) @@ -3503,7 +3643,7 @@ void set_user_nice(struct task_struct *p, long nice) * We have to be careful, if called from sys_setpriority(), * the task might be in the middle of scheduling on another CPU. */ - rq = task_rq_lock(p, &flags); + rq = task_rq_lock(p, &rf); /* * The RT priorities are set via sched_setscheduler(), but we still * allow the 'normal' nice value to be set - but as expected @@ -3534,7 +3674,7 @@ void set_user_nice(struct task_struct *p, long nice) resched_curr(rq); } out_unlock: - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); } EXPORT_SYMBOL(set_user_nice); @@ -3831,11 +3971,11 @@ static int __sched_setscheduler(struct task_struct *p, MAX_RT_PRIO - 1 - attr->sched_priority; int retval, oldprio, oldpolicy = -1, queued, running; int new_effective_prio, policy = attr->sched_policy; - unsigned long flags; const struct sched_class *prev_class; - struct rq *rq; + struct rq_flags rf; int reset_on_fork; int queue_flags = DEQUEUE_SAVE | DEQUEUE_MOVE; + struct rq *rq; /* may grab non-irq protected spin_locks */ BUG_ON(in_interrupt()); @@ -3930,13 +4070,13 @@ recheck: * To be able to change p->policy safely, the appropriate * runqueue lock must be held. */ - rq = task_rq_lock(p, &flags); + rq = task_rq_lock(p, &rf); /* * Changing the policy of the stop threads its a very bad idea */ if (p == rq->stop) { - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); return -EINVAL; } @@ -3953,7 +4093,7 @@ recheck: goto change; p->sched_reset_on_fork = reset_on_fork; - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); return 0; } change: @@ -3967,7 +4107,7 @@ change: if (rt_bandwidth_enabled() && rt_policy(policy) && task_group(p)->rt_bandwidth.rt_runtime == 0 && !task_group_is_autogroup(task_group(p))) { - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); return -EPERM; } #endif @@ -3982,7 +4122,7 @@ change: */ if (!cpumask_subset(span, &p->cpus_allowed) || rq->rd->dl_bw.bw == 0) { - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); return -EPERM; } } @@ -3992,7 +4132,7 @@ change: /* recheck policy now with rq lock held */ if (unlikely(oldpolicy != -1 && oldpolicy != p->policy)) { policy = oldpolicy = -1; - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); goto recheck; } @@ -4002,7 +4142,7 @@ change: * is available. */ if ((dl_policy(policy) || dl_task(p)) && dl_overflow(p, policy, attr)) { - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); return -EBUSY; } @@ -4047,7 +4187,7 @@ change: check_class_changed(rq, p, prev_class, oldprio); preempt_disable(); /* avoid rq from going away on us */ - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); if (pi) rt_mutex_adjust_pi(p); @@ -4900,10 +5040,10 @@ SYSCALL_DEFINE2(sched_rr_get_interval, pid_t, pid, { struct task_struct *p; unsigned int time_slice; - unsigned long flags; + struct rq_flags rf; + struct timespec t; struct rq *rq; int retval; - struct timespec t; if (pid < 0) return -EINVAL; @@ -4918,11 +5058,11 @@ SYSCALL_DEFINE2(sched_rr_get_interval, pid_t, pid, if (retval) goto out_unlock; - rq = task_rq_lock(p, &flags); + rq = task_rq_lock(p, &rf); time_slice = 0; if (p->sched_class->get_rr_interval) time_slice = p->sched_class->get_rr_interval(rq, p); - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); rcu_read_unlock(); jiffies_to_timespec(time_slice, &t); @@ -4998,7 +5138,8 @@ void show_state_filter(unsigned long state_filter) touch_all_softlockup_watchdogs(); #ifdef CONFIG_SCHED_DEBUG - sysrq_sched_debug_show(); + if (!state_filter) + sysrq_sched_debug_show(); #endif rcu_read_unlock(); /* @@ -5160,6 +5301,8 @@ out: #ifdef CONFIG_SMP +static bool sched_smp_initialized __read_mostly; + #ifdef CONFIG_NUMA_BALANCING /* Migrate current task p to target_cpu */ int migrate_task_to(struct task_struct *p, int target_cpu) @@ -5185,11 +5328,11 @@ int migrate_task_to(struct task_struct *p, int target_cpu) */ void sched_setnuma(struct task_struct *p, int nid) { - struct rq *rq; - unsigned long flags; bool queued, running; + struct rq_flags rf; + struct rq *rq; - rq = task_rq_lock(p, &flags); + rq = task_rq_lock(p, &rf); queued = task_on_rq_queued(p); running = task_current(rq, p); @@ -5204,7 +5347,7 @@ void sched_setnuma(struct task_struct *p, int nid) p->sched_class->set_curr_task(rq); if (queued) enqueue_task(rq, p, ENQUEUE_RESTORE); - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); } #endif /* CONFIG_NUMA_BALANCING */ @@ -5220,7 +5363,7 @@ void idle_task_exit(void) BUG_ON(cpu_online(smp_processor_id())); if (mm != &init_mm) { - switch_mm(mm, &init_mm, current); + switch_mm_irqs_off(mm, &init_mm, current); finish_arch_post_lock_switch(); } mmdrop(mm); @@ -5268,6 +5411,7 @@ static void migrate_tasks(struct rq *dead_rq) { struct rq *rq = dead_rq; struct task_struct *next, *stop = rq->stop; + struct pin_cookie cookie; int dest_cpu; /* @@ -5299,8 +5443,8 @@ static void migrate_tasks(struct rq *dead_rq) /* * pick_next_task assumes pinned rq->lock. */ - lockdep_pin_lock(&rq->lock); - next = pick_next_task(rq, &fake_task); + cookie = lockdep_pin_lock(&rq->lock); + next = pick_next_task(rq, &fake_task, cookie); BUG_ON(!next); next->sched_class->put_prev_task(rq, next); @@ -5313,7 +5457,7 @@ static void migrate_tasks(struct rq *dead_rq) * because !cpu_active at this point, which means load-balance * will not interfere. Also, stop-machine. */ - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, cookie); raw_spin_unlock(&rq->lock); raw_spin_lock(&next->pi_lock); raw_spin_lock(&rq->lock); @@ -5374,127 +5518,13 @@ static void set_rq_offline(struct rq *rq) } } -/* - * migration_call - callback that gets triggered when a CPU is added. - * Here we can start up the necessary migration thread for the new CPU. - */ -static int -migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) +static void set_cpu_rq_start_time(unsigned int cpu) { - int cpu = (long)hcpu; - unsigned long flags; struct rq *rq = cpu_rq(cpu); - switch (action & ~CPU_TASKS_FROZEN) { - - case CPU_UP_PREPARE: - rq->calc_load_update = calc_load_update; - account_reset_rq(rq); - break; - - case CPU_ONLINE: - /* Update our root-domain */ - raw_spin_lock_irqsave(&rq->lock, flags); - if (rq->rd) { - BUG_ON(!cpumask_test_cpu(cpu, rq->rd->span)); - - set_rq_online(rq); - } - raw_spin_unlock_irqrestore(&rq->lock, flags); - break; - -#ifdef CONFIG_HOTPLUG_CPU - case CPU_DYING: - sched_ttwu_pending(); - /* Update our root-domain */ - raw_spin_lock_irqsave(&rq->lock, flags); - if (rq->rd) { - BUG_ON(!cpumask_test_cpu(cpu, rq->rd->span)); - set_rq_offline(rq); - } - migrate_tasks(rq); - BUG_ON(rq->nr_running != 1); /* the migration thread */ - raw_spin_unlock_irqrestore(&rq->lock, flags); - break; - - case CPU_DEAD: - calc_load_migrate(rq); - break; -#endif - } - - update_max_interval(); - - return NOTIFY_OK; -} - -/* - * Register at high priority so that task migration (migrate_all_tasks) - * happens before everything else. This has to be lower priority than - * the notifier in the perf_event subsystem, though. - */ -static struct notifier_block migration_notifier = { - .notifier_call = migration_call, - .priority = CPU_PRI_MIGRATION, -}; - -static void set_cpu_rq_start_time(void) -{ - int cpu = smp_processor_id(); - struct rq *rq = cpu_rq(cpu); rq->age_stamp = sched_clock_cpu(cpu); } -static int sched_cpu_active(struct notifier_block *nfb, - unsigned long action, void *hcpu) -{ - int cpu = (long)hcpu; - - switch (action & ~CPU_TASKS_FROZEN) { - case CPU_STARTING: - set_cpu_rq_start_time(); - return NOTIFY_OK; - - case CPU_DOWN_FAILED: - set_cpu_active(cpu, true); - return NOTIFY_OK; - - default: - return NOTIFY_DONE; - } -} - -static int sched_cpu_inactive(struct notifier_block *nfb, - unsigned long action, void *hcpu) -{ - switch (action & ~CPU_TASKS_FROZEN) { - case CPU_DOWN_PREPARE: - set_cpu_active((long)hcpu, false); - return NOTIFY_OK; - default: - return NOTIFY_DONE; - } -} - -static int __init migration_init(void) -{ - void *cpu = (void *)(long)smp_processor_id(); - int err; - - /* Initialize migration for the boot CPU */ - err = migration_call(&migration_notifier, CPU_UP_PREPARE, cpu); - BUG_ON(err == NOTIFY_BAD); - migration_call(&migration_notifier, CPU_ONLINE, cpu); - register_cpu_notifier(&migration_notifier); - - /* Register cpu active notifiers */ - cpu_notifier(sched_cpu_active, CPU_PRI_SCHED_ACTIVE); - cpu_notifier(sched_cpu_inactive, CPU_PRI_SCHED_INACTIVE); - - return 0; -} -early_initcall(migration_init); - static cpumask_var_t sched_domains_tmpmask; /* sched_domains_mutex */ #ifdef CONFIG_SCHED_DEBUG @@ -6642,10 +6672,10 @@ static void sched_init_numa(void) init_numa_topology_type(); } -static void sched_domains_numa_masks_set(int cpu) +static void sched_domains_numa_masks_set(unsigned int cpu) { - int i, j; int node = cpu_to_node(cpu); + int i, j; for (i = 0; i < sched_domains_numa_levels; i++) { for (j = 0; j < nr_node_ids; j++) { @@ -6655,51 +6685,20 @@ static void sched_domains_numa_masks_set(int cpu) } } -static void sched_domains_numa_masks_clear(int cpu) +static void sched_domains_numa_masks_clear(unsigned int cpu) { int i, j; + for (i = 0; i < sched_domains_numa_levels; i++) { for (j = 0; j < nr_node_ids; j++) cpumask_clear_cpu(cpu, sched_domains_numa_masks[i][j]); } } -/* - * Update sched_domains_numa_masks[level][node] array when new cpus - * are onlined. - */ -static int sched_domains_numa_masks_update(struct notifier_block *nfb, - unsigned long action, - void *hcpu) -{ - int cpu = (long)hcpu; - - switch (action & ~CPU_TASKS_FROZEN) { - case CPU_ONLINE: - sched_domains_numa_masks_set(cpu); - break; - - case CPU_DEAD: - sched_domains_numa_masks_clear(cpu); - break; - - default: - return NOTIFY_DONE; - } - - return NOTIFY_OK; -} #else -static inline void sched_init_numa(void) -{ -} - -static int sched_domains_numa_masks_update(struct notifier_block *nfb, - unsigned long action, - void *hcpu) -{ - return 0; -} +static inline void sched_init_numa(void) { } +static void sched_domains_numa_masks_set(unsigned int cpu) { } +static void sched_domains_numa_masks_clear(unsigned int cpu) { } #endif /* CONFIG_NUMA */ static int __sdt_alloc(const struct cpumask *cpu_map) @@ -7089,13 +7088,9 @@ static int num_cpus_frozen; /* used to mark begin/end of suspend/resume */ * If we come here as part of a suspend/resume, don't touch cpusets because we * want to restore it back to its original state upon resume anyway. */ -static int cpuset_cpu_active(struct notifier_block *nfb, unsigned long action, - void *hcpu) +static void cpuset_cpu_active(void) { - switch (action) { - case CPU_ONLINE_FROZEN: - case CPU_DOWN_FAILED_FROZEN: - + if (cpuhp_tasks_frozen) { /* * num_cpus_frozen tracks how many CPUs are involved in suspend * resume sequence. As long as this is not the last online @@ -7105,35 +7100,25 @@ static int cpuset_cpu_active(struct notifier_block *nfb, unsigned long action, num_cpus_frozen--; if (likely(num_cpus_frozen)) { partition_sched_domains(1, NULL, NULL); - break; + return; } - /* * This is the last CPU online operation. So fall through and * restore the original sched domains by considering the * cpuset configurations. */ - - case CPU_ONLINE: - cpuset_update_active_cpus(true); - break; - default: - return NOTIFY_DONE; } - return NOTIFY_OK; + cpuset_update_active_cpus(true); } -static int cpuset_cpu_inactive(struct notifier_block *nfb, unsigned long action, - void *hcpu) +static int cpuset_cpu_inactive(unsigned int cpu) { unsigned long flags; - long cpu = (long)hcpu; struct dl_bw *dl_b; bool overflow; int cpus; - switch (action) { - case CPU_DOWN_PREPARE: + if (!cpuhp_tasks_frozen) { rcu_read_lock_sched(); dl_b = dl_bw_of(cpu); @@ -7145,18 +7130,119 @@ static int cpuset_cpu_inactive(struct notifier_block *nfb, unsigned long action, rcu_read_unlock_sched(); if (overflow) - return notifier_from_errno(-EBUSY); + return -EBUSY; cpuset_update_active_cpus(false); - break; - case CPU_DOWN_PREPARE_FROZEN: + } else { num_cpus_frozen++; partition_sched_domains(1, NULL, NULL); - break; - default: - return NOTIFY_DONE; } - return NOTIFY_OK; + return 0; +} + +int sched_cpu_activate(unsigned int cpu) +{ + struct rq *rq = cpu_rq(cpu); + unsigned long flags; + + set_cpu_active(cpu, true); + + if (sched_smp_initialized) { + sched_domains_numa_masks_set(cpu); + cpuset_cpu_active(); + } + + /* + * Put the rq online, if not already. This happens: + * + * 1) In the early boot process, because we build the real domains + * after all cpus have been brought up. + * + * 2) At runtime, if cpuset_cpu_active() fails to rebuild the + * domains. + */ + raw_spin_lock_irqsave(&rq->lock, flags); + if (rq->rd) { + BUG_ON(!cpumask_test_cpu(cpu, rq->rd->span)); + set_rq_online(rq); + } + raw_spin_unlock_irqrestore(&rq->lock, flags); + + update_max_interval(); + + return 0; +} + +int sched_cpu_deactivate(unsigned int cpu) +{ + int ret; + + set_cpu_active(cpu, false); + /* + * We've cleared cpu_active_mask, wait for all preempt-disabled and RCU + * users of this state to go away such that all new such users will + * observe it. + * + * For CONFIG_PREEMPT we have preemptible RCU and its sync_rcu() might + * not imply sync_sched(), so wait for both. + * + * Do sync before park smpboot threads to take care the rcu boost case. + */ + if (IS_ENABLED(CONFIG_PREEMPT)) + synchronize_rcu_mult(call_rcu, call_rcu_sched); + else + synchronize_rcu(); + + if (!sched_smp_initialized) + return 0; + + ret = cpuset_cpu_inactive(cpu); + if (ret) { + set_cpu_active(cpu, true); + return ret; + } + sched_domains_numa_masks_clear(cpu); + return 0; +} + +static void sched_rq_cpu_starting(unsigned int cpu) +{ + struct rq *rq = cpu_rq(cpu); + + rq->calc_load_update = calc_load_update; + account_reset_rq(rq); + update_max_interval(); +} + +int sched_cpu_starting(unsigned int cpu) +{ + set_cpu_rq_start_time(cpu); + sched_rq_cpu_starting(cpu); + return 0; +} + +#ifdef CONFIG_HOTPLUG_CPU +int sched_cpu_dying(unsigned int cpu) +{ + struct rq *rq = cpu_rq(cpu); + unsigned long flags; + + /* Handle pending wakeups and then migrate everything off */ + sched_ttwu_pending(); + raw_spin_lock_irqsave(&rq->lock, flags); + if (rq->rd) { + BUG_ON(!cpumask_test_cpu(cpu, rq->rd->span)); + set_rq_offline(rq); + } + migrate_tasks(rq); + BUG_ON(rq->nr_running != 1); + raw_spin_unlock_irqrestore(&rq->lock, flags); + calc_load_migrate(rq); + update_max_interval(); + nohz_balance_exit_idle(cpu); + hrtick_clear(rq); + return 0; } +#endif void __init sched_init_smp(void) { @@ -7179,12 +7265,6 @@ void __init sched_init_smp(void) cpumask_set_cpu(smp_processor_id(), non_isolated_cpus); mutex_unlock(&sched_domains_mutex); - hotcpu_notifier(sched_domains_numa_masks_update, CPU_PRI_SCHED_ACTIVE); - hotcpu_notifier(cpuset_cpu_active, CPU_PRI_CPUSET_ACTIVE); - hotcpu_notifier(cpuset_cpu_inactive, CPU_PRI_CPUSET_INACTIVE); - - init_hrtick(); - /* Move init over to a non-isolated CPU */ if (set_cpus_allowed_ptr(current, non_isolated_cpus) < 0) BUG(); @@ -7193,7 +7273,16 @@ void __init sched_init_smp(void) init_sched_rt_class(); init_sched_dl_class(); + sched_smp_initialized = true; +} + +static int __init migration_init(void) +{ + sched_rq_cpu_starting(smp_processor_id()); + return 0; } +early_initcall(migration_init); + #else void __init sched_init_smp(void) { @@ -7328,8 +7417,6 @@ void __init sched_init(void) for (j = 0; j < CPU_LOAD_IDX_MAX; j++) rq->cpu_load[j] = 0; - rq->last_load_update_tick = jiffies; - #ifdef CONFIG_SMP rq->sd = NULL; rq->rd = NULL; @@ -7348,12 +7435,13 @@ void __init sched_init(void) rq_attach_root(rq, &def_root_domain); #ifdef CONFIG_NO_HZ_COMMON + rq->last_load_update_tick = jiffies; rq->nohz_flags = 0; #endif #ifdef CONFIG_NO_HZ_FULL rq->last_sched_tick = 0; #endif -#endif +#endif /* CONFIG_SMP */ init_rq_hrtick(rq); atomic_set(&rq->nr_iowait, 0); } @@ -7391,7 +7479,7 @@ void __init sched_init(void) if (cpu_isolated_map == NULL) zalloc_cpumask_var(&cpu_isolated_map, GFP_NOWAIT); idle_thread_set_boot_cpu(); - set_cpu_rq_start_time(); + set_cpu_rq_start_time(smp_processor_id()); #endif init_sched_fair_class(); @@ -7636,10 +7724,10 @@ void sched_move_task(struct task_struct *tsk) { struct task_group *tg; int queued, running; - unsigned long flags; + struct rq_flags rf; struct rq *rq; - rq = task_rq_lock(tsk, &flags); + rq = task_rq_lock(tsk, &rf); running = task_current(rq, tsk); queued = task_on_rq_queued(tsk); @@ -7671,7 +7759,7 @@ void sched_move_task(struct task_struct *tsk) if (queued) enqueue_task(rq, tsk, ENQUEUE_RESTORE | ENQUEUE_MOVE); - task_rq_unlock(rq, tsk, &flags); + task_rq_unlock(rq, tsk, &rf); } #endif /* CONFIG_CGROUP_SCHED */ @@ -7891,7 +7979,7 @@ static int sched_rt_can_attach(struct task_group *tg, struct task_struct *tsk) static int sched_rt_global_constraints(void) { unsigned long flags; - int i, ret = 0; + int i; raw_spin_lock_irqsave(&def_rt_bandwidth.rt_runtime_lock, flags); for_each_possible_cpu(i) { @@ -7903,7 +7991,7 @@ static int sched_rt_global_constraints(void) } raw_spin_unlock_irqrestore(&def_rt_bandwidth.rt_runtime_lock, flags); - return ret; + return 0; } #endif /* CONFIG_RT_GROUP_SCHED */ diff --git a/kernel/sched/cpuacct.c b/kernel/sched/cpuacct.c index 4a81120..41f85c4 100644 --- a/kernel/sched/cpuacct.c +++ b/kernel/sched/cpuacct.c @@ -25,11 +25,22 @@ enum cpuacct_stat_index { CPUACCT_STAT_NSTATS, }; +enum cpuacct_usage_index { + CPUACCT_USAGE_USER, /* ... user mode */ + CPUACCT_USAGE_SYSTEM, /* ... kernel mode */ + + CPUACCT_USAGE_NRUSAGE, +}; + +struct cpuacct_usage { + u64 usages[CPUACCT_USAGE_NRUSAGE]; +}; + /* 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 */ - u64 __percpu *cpuusage; + struct cpuacct_usage __percpu *cpuusage; struct kernel_cpustat __percpu *cpustat; }; @@ -49,7 +60,7 @@ static inline struct cpuacct *parent_ca(struct cpuacct *ca) return css_ca(ca->css.parent); } -static DEFINE_PER_CPU(u64, root_cpuacct_cpuusage); +static DEFINE_PER_CPU(struct cpuacct_usage, root_cpuacct_cpuusage); static struct cpuacct root_cpuacct = { .cpustat = &kernel_cpustat, .cpuusage = &root_cpuacct_cpuusage, @@ -68,7 +79,7 @@ cpuacct_css_alloc(struct cgroup_subsys_state *parent_css) if (!ca) goto out; - ca->cpuusage = alloc_percpu(u64); + ca->cpuusage = alloc_percpu(struct cpuacct_usage); if (!ca->cpuusage) goto out_free_ca; @@ -96,20 +107,37 @@ static void cpuacct_css_free(struct cgroup_subsys_state *css) kfree(ca); } -static u64 cpuacct_cpuusage_read(struct cpuacct *ca, int cpu) +static u64 cpuacct_cpuusage_read(struct cpuacct *ca, int cpu, + enum cpuacct_usage_index index) { - u64 *cpuusage = per_cpu_ptr(ca->cpuusage, cpu); + struct cpuacct_usage *cpuusage = per_cpu_ptr(ca->cpuusage, cpu); u64 data; + /* + * We allow index == CPUACCT_USAGE_NRUSAGE here to read + * the sum of suages. + */ + BUG_ON(index > CPUACCT_USAGE_NRUSAGE); + #ifndef CONFIG_64BIT /* * Take rq->lock to make 64-bit read safe on 32-bit platforms. */ raw_spin_lock_irq(&cpu_rq(cpu)->lock); - data = *cpuusage; +#endif + + if (index == CPUACCT_USAGE_NRUSAGE) { + int i = 0; + + data = 0; + for (i = 0; i < CPUACCT_USAGE_NRUSAGE; i++) + data += cpuusage->usages[i]; + } else { + data = cpuusage->usages[index]; + } + +#ifndef CONFIG_64BIT raw_spin_unlock_irq(&cpu_rq(cpu)->lock); -#else - data = *cpuusage; #endif return data; @@ -117,69 +145,103 @@ static u64 cpuacct_cpuusage_read(struct cpuacct *ca, int cpu) static void cpuacct_cpuusage_write(struct cpuacct *ca, int cpu, u64 val) { - u64 *cpuusage = per_cpu_ptr(ca->cpuusage, cpu); + struct cpuacct_usage *cpuusage = per_cpu_ptr(ca->cpuusage, cpu); + int i; #ifndef CONFIG_64BIT /* * Take rq->lock to make 64-bit write safe on 32-bit platforms. */ raw_spin_lock_irq(&cpu_rq(cpu)->lock); - *cpuusage = val; +#endif + + for (i = 0; i < CPUACCT_USAGE_NRUSAGE; i++) + cpuusage->usages[i] = val; + +#ifndef CONFIG_64BIT raw_spin_unlock_irq(&cpu_rq(cpu)->lock); -#else - *cpuusage = val; #endif } /* return total cpu usage (in nanoseconds) of a group */ -static u64 cpuusage_read(struct cgroup_subsys_state *css, struct cftype *cft) +static u64 __cpuusage_read(struct cgroup_subsys_state *css, + enum cpuacct_usage_index index) { struct cpuacct *ca = css_ca(css); u64 totalcpuusage = 0; int i; - for_each_present_cpu(i) - totalcpuusage += cpuacct_cpuusage_read(ca, i); + for_each_possible_cpu(i) + totalcpuusage += cpuacct_cpuusage_read(ca, i, index); return totalcpuusage; } +static u64 cpuusage_user_read(struct cgroup_subsys_state *css, + struct cftype *cft) +{ + return __cpuusage_read(css, CPUACCT_USAGE_USER); +} + +static u64 cpuusage_sys_read(struct cgroup_subsys_state *css, + struct cftype *cft) +{ + return __cpuusage_read(css, CPUACCT_USAGE_SYSTEM); +} + +static u64 cpuusage_read(struct cgroup_subsys_state *css, struct cftype *cft) +{ + return __cpuusage_read(css, CPUACCT_USAGE_NRUSAGE); +} + static int cpuusage_write(struct cgroup_subsys_state *css, struct cftype *cft, u64 val) { struct cpuacct *ca = css_ca(css); - int err = 0; - int i; + int cpu; /* * Only allow '0' here to do a reset. */ - if (val) { - err = -EINVAL; - goto out; - } + if (val) + return -EINVAL; - for_each_present_cpu(i) - cpuacct_cpuusage_write(ca, i, 0); + for_each_possible_cpu(cpu) + cpuacct_cpuusage_write(ca, cpu, 0); -out: - return err; + return 0; } -static int cpuacct_percpu_seq_show(struct seq_file *m, void *V) +static int __cpuacct_percpu_seq_show(struct seq_file *m, + enum cpuacct_usage_index index) { struct cpuacct *ca = css_ca(seq_css(m)); u64 percpu; int i; - for_each_present_cpu(i) { - percpu = cpuacct_cpuusage_read(ca, i); + for_each_possible_cpu(i) { + percpu = cpuacct_cpuusage_read(ca, i, index); seq_printf(m, "%llu ", (unsigned long long) percpu); } seq_printf(m, "\n"); return 0; } +static int cpuacct_percpu_user_seq_show(struct seq_file *m, void *V) +{ + return __cpuacct_percpu_seq_show(m, CPUACCT_USAGE_USER); +} + +static int cpuacct_percpu_sys_seq_show(struct seq_file *m, void *V) +{ + return __cpuacct_percpu_seq_show(m, CPUACCT_USAGE_SYSTEM); +} + +static int cpuacct_percpu_seq_show(struct seq_file *m, void *V) +{ + return __cpuacct_percpu_seq_show(m, CPUACCT_USAGE_NRUSAGE); +} + static const char * const cpuacct_stat_desc[] = { [CPUACCT_STAT_USER] = "user", [CPUACCT_STAT_SYSTEM] = "system", @@ -191,7 +253,7 @@ static int cpuacct_stats_show(struct seq_file *sf, void *v) int cpu; s64 val = 0; - for_each_online_cpu(cpu) { + for_each_possible_cpu(cpu) { struct kernel_cpustat *kcpustat = per_cpu_ptr(ca->cpustat, cpu); val += kcpustat->cpustat[CPUTIME_USER]; val += kcpustat->cpustat[CPUTIME_NICE]; @@ -200,7 +262,7 @@ static int cpuacct_stats_show(struct seq_file *sf, void *v) seq_printf(sf, "%s %lld\n", cpuacct_stat_desc[CPUACCT_STAT_USER], val); val = 0; - for_each_online_cpu(cpu) { + for_each_possible_cpu(cpu) { struct kernel_cpustat *kcpustat = per_cpu_ptr(ca->cpustat, cpu); val += kcpustat->cpustat[CPUTIME_SYSTEM]; val += kcpustat->cpustat[CPUTIME_IRQ]; @@ -220,10 +282,26 @@ static struct cftype files[] = { .write_u64 = cpuusage_write, }, { + .name = "usage_user", + .read_u64 = cpuusage_user_read, + }, + { + .name = "usage_sys", + .read_u64 = cpuusage_sys_read, + }, + { .name = "usage_percpu", .seq_show = cpuacct_percpu_seq_show, }, { + .name = "usage_percpu_user", + .seq_show = cpuacct_percpu_user_seq_show, + }, + { + .name = "usage_percpu_sys", + .seq_show = cpuacct_percpu_sys_seq_show, + }, + { .name = "stat", .seq_show = cpuacct_stats_show, }, @@ -238,10 +316,17 @@ static struct cftype files[] = { void cpuacct_charge(struct task_struct *tsk, u64 cputime) { struct cpuacct *ca; + int index = CPUACCT_USAGE_SYSTEM; + struct pt_regs *regs = task_pt_regs(tsk); + + if (regs && user_mode(regs)) + index = CPUACCT_USAGE_USER; rcu_read_lock(); + for (ca = task_ca(tsk); ca; ca = parent_ca(ca)) - *this_cpu_ptr(ca->cpuusage) += cputime; + this_cpu_ptr(ca->cpuusage)->usages[index] += cputime; + rcu_read_unlock(); } diff --git a/kernel/sched/cpudeadline.c b/kernel/sched/cpudeadline.c index 5a75b08..5be5882 100644 --- a/kernel/sched/cpudeadline.c +++ b/kernel/sched/cpudeadline.c @@ -103,10 +103,10 @@ int cpudl_find(struct cpudl *cp, struct task_struct *p, const struct sched_dl_entity *dl_se = &p->dl; if (later_mask && - cpumask_and(later_mask, cp->free_cpus, &p->cpus_allowed)) { + cpumask_and(later_mask, cp->free_cpus, tsk_cpus_allowed(p))) { best_cpu = cpumask_any(later_mask); goto out; - } else if (cpumask_test_cpu(cpudl_maximum(cp), &p->cpus_allowed) && + } else if (cpumask_test_cpu(cpudl_maximum(cp), tsk_cpus_allowed(p)) && dl_time_before(dl_se->deadline, cp->elements[0].dl)) { best_cpu = cpudl_maximum(cp); if (later_mask) diff --git a/kernel/sched/cpufreq.c b/kernel/sched/cpufreq.c index 928c4ba..1141954 100644 --- a/kernel/sched/cpufreq.c +++ b/kernel/sched/cpufreq.c @@ -14,24 +14,50 @@ DEFINE_PER_CPU(struct update_util_data *, cpufreq_update_util_data); /** - * cpufreq_set_update_util_data - Populate the CPU's update_util_data pointer. + * cpufreq_add_update_util_hook - Populate the CPU's update_util_data pointer. * @cpu: The CPU to set the pointer for. * @data: New pointer value. + * @func: Callback function to set for the CPU. * - * Set and publish the update_util_data pointer for the given CPU. That pointer - * points to a struct update_util_data object containing a callback function - * to call from cpufreq_update_util(). That function will be called from an RCU - * read-side critical section, so it must not sleep. + * Set and publish the update_util_data pointer for the given CPU. * - * Callers must use RCU-sched callbacks to free any memory that might be - * accessed via the old update_util_data pointer or invoke synchronize_sched() - * right after this function to avoid use-after-free. + * The update_util_data pointer of @cpu is set to @data and the callback + * function pointer in the target struct update_util_data is set to @func. + * That function will be called by cpufreq_update_util() from RCU-sched + * read-side critical sections, so it must not sleep. @data will always be + * passed to it as the first argument which allows the function to get to the + * target update_util_data structure and its container. + * + * The update_util_data pointer of @cpu must be NULL when this function is + * called or it will WARN() and return with no effect. */ -void cpufreq_set_update_util_data(int cpu, struct update_util_data *data) +void cpufreq_add_update_util_hook(int cpu, struct update_util_data *data, + void (*func)(struct update_util_data *data, u64 time, + unsigned long util, unsigned long max)) { - if (WARN_ON(data && !data->func)) + if (WARN_ON(!data || !func)) return; + if (WARN_ON(per_cpu(cpufreq_update_util_data, cpu))) + return; + + data->func = func; rcu_assign_pointer(per_cpu(cpufreq_update_util_data, cpu), data); } -EXPORT_SYMBOL_GPL(cpufreq_set_update_util_data); +EXPORT_SYMBOL_GPL(cpufreq_add_update_util_hook); + +/** + * cpufreq_remove_update_util_hook - Clear the CPU's update_util_data pointer. + * @cpu: The CPU to clear the pointer for. + * + * Clear the update_util_data pointer for the given CPU. + * + * Callers must use RCU-sched callbacks to free any memory that might be + * accessed via the old update_util_data pointer or invoke synchronize_sched() + * right after this function to avoid use-after-free. + */ +void cpufreq_remove_update_util_hook(int cpu) +{ + rcu_assign_pointer(per_cpu(cpufreq_update_util_data, cpu), NULL); +} +EXPORT_SYMBOL_GPL(cpufreq_remove_update_util_hook); diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c new file mode 100644 index 0000000..14c4aa2 --- /dev/null +++ b/kernel/sched/cpufreq_schedutil.c @@ -0,0 +1,532 @@ +/* + * CPUFreq governor based on scheduler-provided CPU utilization data. + * + * Copyright (C) 2016, Intel Corporation + * Author: Rafael J. Wysocki <rafael.j.wysocki@intel.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/cpufreq.h> +#include <linux/module.h> +#include <linux/slab.h> +#include <trace/events/power.h> + +#include "sched.h" + +struct sugov_tunables { + struct gov_attr_set attr_set; + unsigned int rate_limit_us; +}; + +struct sugov_policy { + struct cpufreq_policy *policy; + + struct sugov_tunables *tunables; + struct list_head tunables_hook; + + raw_spinlock_t update_lock; /* For shared policies */ + u64 last_freq_update_time; + s64 freq_update_delay_ns; + unsigned int next_freq; + + /* The next fields are only needed if fast switch cannot be used. */ + struct irq_work irq_work; + struct work_struct work; + struct mutex work_lock; + bool work_in_progress; + + bool need_freq_update; +}; + +struct sugov_cpu { + struct update_util_data update_util; + struct sugov_policy *sg_policy; + + /* The fields below are only needed when sharing a policy. */ + unsigned long util; + unsigned long max; + u64 last_update; +}; + +static DEFINE_PER_CPU(struct sugov_cpu, sugov_cpu); + +/************************ Governor internals ***********************/ + +static bool sugov_should_update_freq(struct sugov_policy *sg_policy, u64 time) +{ + s64 delta_ns; + + if (sg_policy->work_in_progress) + return false; + + if (unlikely(sg_policy->need_freq_update)) { + sg_policy->need_freq_update = false; + /* + * This happens when limits change, so forget the previous + * next_freq value and force an update. + */ + sg_policy->next_freq = UINT_MAX; + return true; + } + + delta_ns = time - sg_policy->last_freq_update_time; + return delta_ns >= sg_policy->freq_update_delay_ns; +} + +static void sugov_update_commit(struct sugov_policy *sg_policy, u64 time, + unsigned int next_freq) +{ + struct cpufreq_policy *policy = sg_policy->policy; + + sg_policy->last_freq_update_time = time; + + if (policy->fast_switch_enabled) { + if (sg_policy->next_freq == next_freq) { + trace_cpu_frequency(policy->cur, smp_processor_id()); + return; + } + sg_policy->next_freq = next_freq; + next_freq = cpufreq_driver_fast_switch(policy, next_freq); + if (next_freq == CPUFREQ_ENTRY_INVALID) + return; + + policy->cur = next_freq; + trace_cpu_frequency(next_freq, smp_processor_id()); + } else if (sg_policy->next_freq != next_freq) { + sg_policy->next_freq = next_freq; + sg_policy->work_in_progress = true; + irq_work_queue(&sg_policy->irq_work); + } +} + +/** + * get_next_freq - Compute a new frequency for a given cpufreq policy. + * @policy: cpufreq policy object to compute the new frequency for. + * @util: Current CPU utilization. + * @max: CPU capacity. + * + * If the utilization is frequency-invariant, choose the new frequency to be + * proportional to it, that is + * + * next_freq = C * max_freq * util / max + * + * Otherwise, approximate the would-be frequency-invariant utilization by + * util_raw * (curr_freq / max_freq) which leads to + * + * next_freq = C * curr_freq * util_raw / max + * + * Take C = 1.25 for the frequency tipping point at (util / max) = 0.8. + */ +static unsigned int get_next_freq(struct cpufreq_policy *policy, + unsigned long util, unsigned long max) +{ + unsigned int freq = arch_scale_freq_invariant() ? + policy->cpuinfo.max_freq : policy->cur; + + return (freq + (freq >> 2)) * util / max; +} + +static void sugov_update_single(struct update_util_data *hook, u64 time, + unsigned long util, unsigned long max) +{ + struct sugov_cpu *sg_cpu = container_of(hook, struct sugov_cpu, update_util); + struct sugov_policy *sg_policy = sg_cpu->sg_policy; + struct cpufreq_policy *policy = sg_policy->policy; + unsigned int next_f; + + if (!sugov_should_update_freq(sg_policy, time)) + return; + + next_f = util == ULONG_MAX ? policy->cpuinfo.max_freq : + get_next_freq(policy, util, max); + sugov_update_commit(sg_policy, time, next_f); +} + +static unsigned int sugov_next_freq_shared(struct sugov_policy *sg_policy, + unsigned long util, unsigned long max) +{ + struct cpufreq_policy *policy = sg_policy->policy; + unsigned int max_f = policy->cpuinfo.max_freq; + u64 last_freq_update_time = sg_policy->last_freq_update_time; + unsigned int j; + + if (util == ULONG_MAX) + return max_f; + + for_each_cpu(j, policy->cpus) { + struct sugov_cpu *j_sg_cpu; + unsigned long j_util, j_max; + s64 delta_ns; + + if (j == smp_processor_id()) + continue; + + j_sg_cpu = &per_cpu(sugov_cpu, j); + /* + * If the CPU utilization was last updated before the previous + * frequency update and the time elapsed between the last update + * of the CPU utilization and the last frequency update is long + * enough, don't take the CPU into account as it probably is + * idle now. + */ + delta_ns = last_freq_update_time - j_sg_cpu->last_update; + if (delta_ns > TICK_NSEC) + continue; + + j_util = j_sg_cpu->util; + if (j_util == ULONG_MAX) + return max_f; + + j_max = j_sg_cpu->max; + if (j_util * max > j_max * util) { + util = j_util; + max = j_max; + } + } + + return get_next_freq(policy, util, max); +} + +static void sugov_update_shared(struct update_util_data *hook, u64 time, + unsigned long util, unsigned long max) +{ + struct sugov_cpu *sg_cpu = container_of(hook, struct sugov_cpu, update_util); + struct sugov_policy *sg_policy = sg_cpu->sg_policy; + unsigned int next_f; + + raw_spin_lock(&sg_policy->update_lock); + + sg_cpu->util = util; + sg_cpu->max = max; + sg_cpu->last_update = time; + + if (sugov_should_update_freq(sg_policy, time)) { + next_f = sugov_next_freq_shared(sg_policy, util, max); + sugov_update_commit(sg_policy, time, next_f); + } + + raw_spin_unlock(&sg_policy->update_lock); +} + +static void sugov_work(struct work_struct *work) +{ + struct sugov_policy *sg_policy = container_of(work, struct sugov_policy, work); + + mutex_lock(&sg_policy->work_lock); + __cpufreq_driver_target(sg_policy->policy, sg_policy->next_freq, + CPUFREQ_RELATION_L); + mutex_unlock(&sg_policy->work_lock); + + sg_policy->work_in_progress = false; +} + +static void sugov_irq_work(struct irq_work *irq_work) +{ + struct sugov_policy *sg_policy; + + sg_policy = container_of(irq_work, struct sugov_policy, irq_work); + schedule_work_on(smp_processor_id(), &sg_policy->work); +} + +/************************** sysfs interface ************************/ + +static struct sugov_tunables *global_tunables; +static DEFINE_MUTEX(global_tunables_lock); + +static inline struct sugov_tunables *to_sugov_tunables(struct gov_attr_set *attr_set) +{ + return container_of(attr_set, struct sugov_tunables, attr_set); +} + +static ssize_t rate_limit_us_show(struct gov_attr_set *attr_set, char *buf) +{ + struct sugov_tunables *tunables = to_sugov_tunables(attr_set); + + return sprintf(buf, "%u\n", tunables->rate_limit_us); +} + +static ssize_t rate_limit_us_store(struct gov_attr_set *attr_set, const char *buf, + size_t count) +{ + struct sugov_tunables *tunables = to_sugov_tunables(attr_set); + struct sugov_policy *sg_policy; + unsigned int rate_limit_us; + + if (kstrtouint(buf, 10, &rate_limit_us)) + return -EINVAL; + + tunables->rate_limit_us = rate_limit_us; + + list_for_each_entry(sg_policy, &attr_set->policy_list, tunables_hook) + sg_policy->freq_update_delay_ns = rate_limit_us * NSEC_PER_USEC; + + return count; +} + +static struct governor_attr rate_limit_us = __ATTR_RW(rate_limit_us); + +static struct attribute *sugov_attributes[] = { + &rate_limit_us.attr, + NULL +}; + +static struct kobj_type sugov_tunables_ktype = { + .default_attrs = sugov_attributes, + .sysfs_ops = &governor_sysfs_ops, +}; + +/********************** cpufreq governor interface *********************/ + +static struct cpufreq_governor schedutil_gov; + +static struct sugov_policy *sugov_policy_alloc(struct cpufreq_policy *policy) +{ + struct sugov_policy *sg_policy; + + sg_policy = kzalloc(sizeof(*sg_policy), GFP_KERNEL); + if (!sg_policy) + return NULL; + + sg_policy->policy = policy; + init_irq_work(&sg_policy->irq_work, sugov_irq_work); + INIT_WORK(&sg_policy->work, sugov_work); + mutex_init(&sg_policy->work_lock); + raw_spin_lock_init(&sg_policy->update_lock); + return sg_policy; +} + +static void sugov_policy_free(struct sugov_policy *sg_policy) +{ + mutex_destroy(&sg_policy->work_lock); + kfree(sg_policy); +} + +static struct sugov_tunables *sugov_tunables_alloc(struct sugov_policy *sg_policy) +{ + struct sugov_tunables *tunables; + + tunables = kzalloc(sizeof(*tunables), GFP_KERNEL); + if (tunables) { + gov_attr_set_init(&tunables->attr_set, &sg_policy->tunables_hook); + if (!have_governor_per_policy()) + global_tunables = tunables; + } + return tunables; +} + +static void sugov_tunables_free(struct sugov_tunables *tunables) +{ + if (!have_governor_per_policy()) + global_tunables = NULL; + + kfree(tunables); +} + +static int sugov_init(struct cpufreq_policy *policy) +{ + struct sugov_policy *sg_policy; + struct sugov_tunables *tunables; + unsigned int lat; + int ret = 0; + + /* State should be equivalent to EXIT */ + if (policy->governor_data) + return -EBUSY; + + sg_policy = sugov_policy_alloc(policy); + if (!sg_policy) + return -ENOMEM; + + mutex_lock(&global_tunables_lock); + + if (global_tunables) { + if (WARN_ON(have_governor_per_policy())) { + ret = -EINVAL; + goto free_sg_policy; + } + policy->governor_data = sg_policy; + sg_policy->tunables = global_tunables; + + gov_attr_set_get(&global_tunables->attr_set, &sg_policy->tunables_hook); + goto out; + } + + tunables = sugov_tunables_alloc(sg_policy); + if (!tunables) { + ret = -ENOMEM; + goto free_sg_policy; + } + + tunables->rate_limit_us = LATENCY_MULTIPLIER; + lat = policy->cpuinfo.transition_latency / NSEC_PER_USEC; + if (lat) + tunables->rate_limit_us *= lat; + + policy->governor_data = sg_policy; + sg_policy->tunables = tunables; + + ret = kobject_init_and_add(&tunables->attr_set.kobj, &sugov_tunables_ktype, + get_governor_parent_kobj(policy), "%s", + schedutil_gov.name); + if (ret) + goto fail; + + out: + mutex_unlock(&global_tunables_lock); + + cpufreq_enable_fast_switch(policy); + return 0; + + fail: + policy->governor_data = NULL; + sugov_tunables_free(tunables); + + free_sg_policy: + mutex_unlock(&global_tunables_lock); + + sugov_policy_free(sg_policy); + pr_err("initialization failed (error %d)\n", ret); + return ret; +} + +static int sugov_exit(struct cpufreq_policy *policy) +{ + struct sugov_policy *sg_policy = policy->governor_data; + struct sugov_tunables *tunables = sg_policy->tunables; + unsigned int count; + + cpufreq_disable_fast_switch(policy); + + mutex_lock(&global_tunables_lock); + + count = gov_attr_set_put(&tunables->attr_set, &sg_policy->tunables_hook); + policy->governor_data = NULL; + if (!count) + sugov_tunables_free(tunables); + + mutex_unlock(&global_tunables_lock); + + sugov_policy_free(sg_policy); + return 0; +} + +static int sugov_start(struct cpufreq_policy *policy) +{ + struct sugov_policy *sg_policy = policy->governor_data; + unsigned int cpu; + + sg_policy->freq_update_delay_ns = sg_policy->tunables->rate_limit_us * NSEC_PER_USEC; + sg_policy->last_freq_update_time = 0; + sg_policy->next_freq = UINT_MAX; + sg_policy->work_in_progress = false; + sg_policy->need_freq_update = false; + + for_each_cpu(cpu, policy->cpus) { + struct sugov_cpu *sg_cpu = &per_cpu(sugov_cpu, cpu); + + sg_cpu->sg_policy = sg_policy; + if (policy_is_shared(policy)) { + sg_cpu->util = ULONG_MAX; + sg_cpu->max = 0; + sg_cpu->last_update = 0; + cpufreq_add_update_util_hook(cpu, &sg_cpu->update_util, + sugov_update_shared); + } else { + cpufreq_add_update_util_hook(cpu, &sg_cpu->update_util, + sugov_update_single); + } + } + return 0; +} + +static int sugov_stop(struct cpufreq_policy *policy) +{ + struct sugov_policy *sg_policy = policy->governor_data; + unsigned int cpu; + + for_each_cpu(cpu, policy->cpus) + cpufreq_remove_update_util_hook(cpu); + + synchronize_sched(); + + irq_work_sync(&sg_policy->irq_work); + cancel_work_sync(&sg_policy->work); + return 0; +} + +static int sugov_limits(struct cpufreq_policy *policy) +{ + struct sugov_policy *sg_policy = policy->governor_data; + + if (!policy->fast_switch_enabled) { + mutex_lock(&sg_policy->work_lock); + + if (policy->max < policy->cur) + __cpufreq_driver_target(policy, policy->max, + CPUFREQ_RELATION_H); + else if (policy->min > policy->cur) + __cpufreq_driver_target(policy, policy->min, + CPUFREQ_RELATION_L); + + mutex_unlock(&sg_policy->work_lock); + } + + sg_policy->need_freq_update = true; + return 0; +} + +int sugov_governor(struct cpufreq_policy *policy, unsigned int event) +{ + if (event == CPUFREQ_GOV_POLICY_INIT) { + return sugov_init(policy); + } else if (policy->governor_data) { + switch (event) { + case CPUFREQ_GOV_POLICY_EXIT: + return sugov_exit(policy); + case CPUFREQ_GOV_START: + return sugov_start(policy); + case CPUFREQ_GOV_STOP: + return sugov_stop(policy); + case CPUFREQ_GOV_LIMITS: + return sugov_limits(policy); + } + } + return -EINVAL; +} + +static struct cpufreq_governor schedutil_gov = { + .name = "schedutil", + .governor = sugov_governor, + .owner = THIS_MODULE, +}; + +static int __init sugov_module_init(void) +{ + return cpufreq_register_governor(&schedutil_gov); +} + +static void __exit sugov_module_exit(void) +{ + cpufreq_unregister_governor(&schedutil_gov); +} + +MODULE_AUTHOR("Rafael J. Wysocki <rafael.j.wysocki@intel.com>"); +MODULE_DESCRIPTION("Utilization-based CPU frequency selection"); +MODULE_LICENSE("GPL"); + +#ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_SCHEDUTIL +struct cpufreq_governor *cpufreq_default_governor(void) +{ + return &schedutil_gov; +} + +fs_initcall(sugov_module_init); +#else +module_init(sugov_module_init); +#endif +module_exit(sugov_module_exit); diff --git a/kernel/sched/cpupri.c b/kernel/sched/cpupri.c index 981fcd7..11e9705 100644 --- a/kernel/sched/cpupri.c +++ b/kernel/sched/cpupri.c @@ -103,11 +103,11 @@ int cpupri_find(struct cpupri *cp, struct task_struct *p, if (skip) continue; - if (cpumask_any_and(&p->cpus_allowed, vec->mask) >= nr_cpu_ids) + if (cpumask_any_and(tsk_cpus_allowed(p), vec->mask) >= nr_cpu_ids) continue; if (lowest_mask) { - cpumask_and(lowest_mask, &p->cpus_allowed, vec->mask); + cpumask_and(lowest_mask, tsk_cpus_allowed(p), vec->mask); /* * We have to ensure that we have at least one bit diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c index affd97e..fcb7f02 100644 --- a/kernel/sched/deadline.c +++ b/kernel/sched/deadline.c @@ -134,7 +134,7 @@ static void inc_dl_migration(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq) { struct task_struct *p = dl_task_of(dl_se); - if (p->nr_cpus_allowed > 1) + if (tsk_nr_cpus_allowed(p) > 1) dl_rq->dl_nr_migratory++; update_dl_migration(dl_rq); @@ -144,7 +144,7 @@ static void dec_dl_migration(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq) { struct task_struct *p = dl_task_of(dl_se); - if (p->nr_cpus_allowed > 1) + if (tsk_nr_cpus_allowed(p) > 1) dl_rq->dl_nr_migratory--; update_dl_migration(dl_rq); @@ -591,10 +591,10 @@ static enum hrtimer_restart dl_task_timer(struct hrtimer *timer) struct sched_dl_entity, dl_timer); struct task_struct *p = dl_task_of(dl_se); - unsigned long flags; + struct rq_flags rf; struct rq *rq; - rq = task_rq_lock(p, &flags); + rq = task_rq_lock(p, &rf); /* * The task might have changed its scheduling policy to something @@ -670,14 +670,14 @@ static enum hrtimer_restart dl_task_timer(struct hrtimer *timer) * Nothing relies on rq->lock after this, so its safe to drop * rq->lock. */ - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, rf.cookie); push_dl_task(rq); - lockdep_pin_lock(&rq->lock); + lockdep_repin_lock(&rq->lock, rf.cookie); } #endif unlock: - task_rq_unlock(rq, p, &flags); + task_rq_unlock(rq, p, &rf); /* * This can free the task_struct, including this hrtimer, do not touch @@ -717,10 +717,6 @@ static void update_curr_dl(struct rq *rq) if (!dl_task(curr) || !on_dl_rq(dl_se)) return; - /* Kick cpufreq (see the comment in linux/cpufreq.h). */ - if (cpu_of(rq) == smp_processor_id()) - cpufreq_trigger_update(rq_clock(rq)); - /* * Consumed budget is computed considering the time as * observed by schedulable tasks (excluding time spent @@ -736,6 +732,10 @@ static void update_curr_dl(struct rq *rq) return; } + /* kick cpufreq (see the comment in linux/cpufreq.h). */ + if (cpu_of(rq) == smp_processor_id()) + cpufreq_trigger_update(rq_clock(rq)); + schedstat_set(curr->se.statistics.exec_max, max(curr->se.statistics.exec_max, delta_exec)); @@ -966,7 +966,7 @@ static void enqueue_task_dl(struct rq *rq, struct task_struct *p, int flags) enqueue_dl_entity(&p->dl, pi_se, flags); - if (!task_current(rq, p) && p->nr_cpus_allowed > 1) + if (!task_current(rq, p) && tsk_nr_cpus_allowed(p) > 1) enqueue_pushable_dl_task(rq, p); } @@ -1040,9 +1040,9 @@ select_task_rq_dl(struct task_struct *p, int cpu, int sd_flag, int flags) * try to make it stay here, it might be important. */ if (unlikely(dl_task(curr)) && - (curr->nr_cpus_allowed < 2 || + (tsk_nr_cpus_allowed(curr) < 2 || !dl_entity_preempt(&p->dl, &curr->dl)) && - (p->nr_cpus_allowed > 1)) { + (tsk_nr_cpus_allowed(p) > 1)) { int target = find_later_rq(p); if (target != -1 && @@ -1063,7 +1063,7 @@ static void check_preempt_equal_dl(struct rq *rq, struct task_struct *p) * Current can't be migrated, useless to reschedule, * let's hope p can move out. */ - if (rq->curr->nr_cpus_allowed == 1 || + if (tsk_nr_cpus_allowed(rq->curr) == 1 || cpudl_find(&rq->rd->cpudl, rq->curr, NULL) == -1) return; @@ -1071,7 +1071,7 @@ static void check_preempt_equal_dl(struct rq *rq, struct task_struct *p) * p is migratable, so let's not schedule it and * see if it is pushed or pulled somewhere else. */ - if (p->nr_cpus_allowed != 1 && + if (tsk_nr_cpus_allowed(p) != 1 && cpudl_find(&rq->rd->cpudl, p, NULL) != -1) return; @@ -1125,7 +1125,8 @@ static struct sched_dl_entity *pick_next_dl_entity(struct rq *rq, return rb_entry(left, struct sched_dl_entity, rb_node); } -struct task_struct *pick_next_task_dl(struct rq *rq, struct task_struct *prev) +struct task_struct * +pick_next_task_dl(struct rq *rq, struct task_struct *prev, struct pin_cookie cookie) { struct sched_dl_entity *dl_se; struct task_struct *p; @@ -1140,9 +1141,9 @@ struct task_struct *pick_next_task_dl(struct rq *rq, struct task_struct *prev) * disabled avoiding further scheduler activity on it and we're * being very careful to re-start the picking loop. */ - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, cookie); pull_dl_task(rq); - lockdep_pin_lock(&rq->lock); + lockdep_repin_lock(&rq->lock, cookie); /* * pull_rt_task() can drop (and re-acquire) rq->lock; this * means a stop task can slip in, in which case we need to @@ -1185,7 +1186,7 @@ static void put_prev_task_dl(struct rq *rq, struct task_struct *p) { update_curr_dl(rq); - if (on_dl_rq(&p->dl) && p->nr_cpus_allowed > 1) + if (on_dl_rq(&p->dl) && tsk_nr_cpus_allowed(p) > 1) enqueue_pushable_dl_task(rq, p); } @@ -1286,7 +1287,7 @@ static int find_later_rq(struct task_struct *task) if (unlikely(!later_mask)) return -1; - if (task->nr_cpus_allowed == 1) + if (tsk_nr_cpus_allowed(task) == 1) return -1; /* @@ -1392,8 +1393,9 @@ static struct rq *find_lock_later_rq(struct task_struct *task, struct rq *rq) if (double_lock_balance(rq, later_rq)) { if (unlikely(task_rq(task) != rq || !cpumask_test_cpu(later_rq->cpu, - &task->cpus_allowed) || + tsk_cpus_allowed(task)) || task_running(rq, task) || + !dl_task(task) || !task_on_rq_queued(task))) { double_unlock_balance(rq, later_rq); later_rq = NULL; @@ -1431,7 +1433,7 @@ static struct task_struct *pick_next_pushable_dl_task(struct rq *rq) BUG_ON(rq->cpu != task_cpu(p)); BUG_ON(task_current(rq, p)); - BUG_ON(p->nr_cpus_allowed <= 1); + BUG_ON(tsk_nr_cpus_allowed(p) <= 1); BUG_ON(!task_on_rq_queued(p)); BUG_ON(!dl_task(p)); @@ -1470,7 +1472,7 @@ retry: */ if (dl_task(rq->curr) && dl_time_before(next_task->dl.deadline, rq->curr->dl.deadline) && - rq->curr->nr_cpus_allowed > 1) { + tsk_nr_cpus_allowed(rq->curr) > 1) { resched_curr(rq); return 0; } @@ -1617,9 +1619,9 @@ static void task_woken_dl(struct rq *rq, struct task_struct *p) { if (!task_running(rq, p) && !test_tsk_need_resched(rq->curr) && - p->nr_cpus_allowed > 1 && + tsk_nr_cpus_allowed(p) > 1 && dl_task(rq->curr) && - (rq->curr->nr_cpus_allowed < 2 || + (tsk_nr_cpus_allowed(rq->curr) < 2 || !dl_entity_preempt(&p->dl, &rq->curr->dl))) { push_dl_tasks(rq); } @@ -1723,7 +1725,7 @@ static void switched_to_dl(struct rq *rq, struct task_struct *p) if (task_on_rq_queued(p) && rq->curr != p) { #ifdef CONFIG_SMP - if (p->nr_cpus_allowed > 1 && rq->dl.overloaded) + if (tsk_nr_cpus_allowed(p) > 1 && rq->dl.overloaded) queue_push_tasks(rq); #else if (dl_task(rq->curr)) diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c index 4fbc3bd..cf905f6 100644 --- a/kernel/sched/debug.c +++ b/kernel/sched/debug.c @@ -626,15 +626,16 @@ do { \ #undef P #undef PN -#ifdef CONFIG_SCHEDSTATS -#define P(n) SEQ_printf(m, " .%-30s: %d\n", #n, rq->n); -#define P64(n) SEQ_printf(m, " .%-30s: %Ld\n", #n, rq->n); - #ifdef CONFIG_SMP +#define P64(n) SEQ_printf(m, " .%-30s: %Ld\n", #n, rq->n); P64(avg_idle); P64(max_idle_balance_cost); +#undef P64 #endif +#ifdef CONFIG_SCHEDSTATS +#define P(n) SEQ_printf(m, " .%-30s: %d\n", #n, rq->n); + if (schedstat_enabled()) { P(yld_count); P(sched_count); @@ -644,7 +645,6 @@ do { \ } #undef P -#undef P64 #endif spin_lock_irqsave(&sched_debug_lock, flags); print_cfs_stats(m, cpu); diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 0fe30e6..218f8e8 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -204,7 +204,7 @@ static void __update_inv_weight(struct load_weight *lw) * OR * (delta_exec * (weight * lw->inv_weight)) >> WMULT_SHIFT * - * Either weight := NICE_0_LOAD and lw \e prio_to_wmult[], in which case + * Either weight := NICE_0_LOAD and lw \e sched_prio_to_wmult[], in which case * we're guaranteed shift stays positive because inv_weight is guaranteed to * fit 32 bits, and NICE_0_LOAD gives another 10 bits; therefore shift >= 22. * @@ -682,17 +682,68 @@ void init_entity_runnable_average(struct sched_entity *se) sa->period_contrib = 1023; sa->load_avg = scale_load_down(se->load.weight); sa->load_sum = sa->load_avg * LOAD_AVG_MAX; - sa->util_avg = scale_load_down(SCHED_LOAD_SCALE); - sa->util_sum = sa->util_avg * LOAD_AVG_MAX; + /* + * At this point, util_avg won't be used in select_task_rq_fair anyway + */ + sa->util_avg = 0; + sa->util_sum = 0; /* when this task enqueue'ed, it will contribute to its cfs_rq's load_avg */ } +/* + * With new tasks being created, their initial util_avgs are extrapolated + * based on the cfs_rq's current util_avg: + * + * util_avg = cfs_rq->util_avg / (cfs_rq->load_avg + 1) * se.load.weight + * + * However, in many cases, the above util_avg does not give a desired + * value. Moreover, the sum of the util_avgs may be divergent, such + * as when the series is a harmonic series. + * + * To solve this problem, we also cap the util_avg of successive tasks to + * only 1/2 of the left utilization budget: + * + * util_avg_cap = (1024 - cfs_rq->avg.util_avg) / 2^n + * + * where n denotes the nth task. + * + * For example, a simplest series from the beginning would be like: + * + * task util_avg: 512, 256, 128, 64, 32, 16, 8, ... + * cfs_rq util_avg: 512, 768, 896, 960, 992, 1008, 1016, ... + * + * Finally, that extrapolated util_avg is clamped to the cap (util_avg_cap) + * if util_avg > util_avg_cap. + */ +void post_init_entity_util_avg(struct sched_entity *se) +{ + struct cfs_rq *cfs_rq = cfs_rq_of(se); + struct sched_avg *sa = &se->avg; + long cap = (long)(SCHED_CAPACITY_SCALE - cfs_rq->avg.util_avg) / 2; + + if (cap > 0) { + if (cfs_rq->avg.util_avg != 0) { + sa->util_avg = cfs_rq->avg.util_avg * se->load.weight; + sa->util_avg /= (cfs_rq->avg.load_avg + 1); + + if (sa->util_avg > cap) + sa->util_avg = cap; + } else { + sa->util_avg = cap; + } + sa->util_sum = sa->util_avg * LOAD_AVG_MAX; + } +} + static inline unsigned long cfs_rq_runnable_load_avg(struct cfs_rq *cfs_rq); static inline unsigned long cfs_rq_load_avg(struct cfs_rq *cfs_rq); #else void init_entity_runnable_average(struct sched_entity *se) { } +void post_init_entity_util_avg(struct sched_entity *se) +{ +} #endif /* @@ -2437,10 +2488,12 @@ account_entity_dequeue(struct cfs_rq *cfs_rq, struct sched_entity *se) update_load_sub(&cfs_rq->load, se->load.weight); if (!parent_entity(se)) update_load_sub(&rq_of(cfs_rq)->load, se->load.weight); +#ifdef CONFIG_SMP if (entity_is_task(se)) { account_numa_dequeue(rq_of(cfs_rq), task_of(se)); list_del_init(&se->group_node); } +#endif cfs_rq->nr_running--; } @@ -2550,6 +2603,16 @@ static const u32 runnable_avg_yN_sum[] = { }; /* + * Precomputed \Sum y^k { 1<=k<=n, where n%32=0). Values are rolled down to + * lower integers. See Documentation/scheduler/sched-avg.txt how these + * were generated: + */ +static const u32 __accumulated_sum_N32[] = { + 0, 23371, 35056, 40899, 43820, 45281, + 46011, 46376, 46559, 46650, 46696, 46719, +}; + +/* * Approximate: * val * y^n, where y^32 ~= 0.5 (~1 scheduling period) */ @@ -2597,22 +2660,13 @@ static u32 __compute_runnable_contrib(u64 n) else if (unlikely(n >= LOAD_AVG_MAX_N)) return LOAD_AVG_MAX; - /* Compute \Sum k^n combining precomputed values for k^i, \Sum k^j */ - do { - contrib /= 2; /* y^LOAD_AVG_PERIOD = 1/2 */ - contrib += runnable_avg_yN_sum[LOAD_AVG_PERIOD]; - - n -= LOAD_AVG_PERIOD; - } while (n > LOAD_AVG_PERIOD); - + /* Since n < LOAD_AVG_MAX_N, n/LOAD_AVG_PERIOD < 11 */ + contrib = __accumulated_sum_N32[n/LOAD_AVG_PERIOD]; + n %= LOAD_AVG_PERIOD; contrib = decay_load(contrib, n); return contrib + runnable_avg_yN_sum[n]; } -#if (SCHED_LOAD_SHIFT - SCHED_LOAD_RESOLUTION) != 10 || SCHED_CAPACITY_SHIFT != 10 -#error "load tracking assumes 2^10 as unit" -#endif - #define cap_scale(v, s) ((v)*(s) >> SCHED_CAPACITY_SHIFT) /* @@ -2821,23 +2875,54 @@ static inline void update_tg_load_avg(struct cfs_rq *cfs_rq, int force) {} static inline u64 cfs_rq_clock_task(struct cfs_rq *cfs_rq); +static inline void cfs_rq_util_change(struct cfs_rq *cfs_rq) +{ + struct rq *rq = rq_of(cfs_rq); + int cpu = cpu_of(rq); + + if (cpu == smp_processor_id() && &rq->cfs == cfs_rq) { + unsigned long max = rq->cpu_capacity_orig; + + /* + * There are a few boundary cases this might miss but it should + * get called often enough that that should (hopefully) not be + * a real problem -- added to that it only calls on the local + * CPU, so if we enqueue remotely we'll miss an update, but + * the next tick/schedule should update. + * + * It will not get called when we go idle, because the idle + * thread is a different class (!fair), nor will the utilization + * number include things like RT tasks. + * + * As is, the util number is not freq-invariant (we'd have to + * implement arch_scale_freq_capacity() for that). + * + * See cpu_util(). + */ + cpufreq_update_util(rq_clock(rq), + min(cfs_rq->avg.util_avg, max), max); + } +} + /* Group cfs_rq's load_avg is used for task_h_load and update_cfs_share */ -static inline int update_cfs_rq_load_avg(u64 now, struct cfs_rq *cfs_rq) +static inline int +update_cfs_rq_load_avg(u64 now, struct cfs_rq *cfs_rq, bool update_freq) { struct sched_avg *sa = &cfs_rq->avg; - int decayed, removed = 0; + int decayed, removed_load = 0, removed_util = 0; if (atomic_long_read(&cfs_rq->removed_load_avg)) { s64 r = atomic_long_xchg(&cfs_rq->removed_load_avg, 0); sa->load_avg = max_t(long, sa->load_avg - r, 0); sa->load_sum = max_t(s64, sa->load_sum - r * LOAD_AVG_MAX, 0); - removed = 1; + removed_load = 1; } if (atomic_long_read(&cfs_rq->removed_util_avg)) { long r = atomic_long_xchg(&cfs_rq->removed_util_avg, 0); sa->util_avg = max_t(long, sa->util_avg - r, 0); sa->util_sum = max_t(s32, sa->util_sum - r * LOAD_AVG_MAX, 0); + removed_util = 1; } decayed = __update_load_avg(now, cpu_of(rq_of(cfs_rq)), sa, @@ -2848,7 +2933,10 @@ static inline int update_cfs_rq_load_avg(u64 now, struct cfs_rq *cfs_rq) cfs_rq->load_last_update_time_copy = sa->last_update_time; #endif - return decayed || removed; + if (update_freq && (decayed || removed_util)) + cfs_rq_util_change(cfs_rq); + + return decayed || removed_load; } /* Update task and its cfs_rq load average */ @@ -2867,31 +2955,8 @@ static inline void update_load_avg(struct sched_entity *se, int update_tg) se->on_rq * scale_load_down(se->load.weight), cfs_rq->curr == se, NULL); - if (update_cfs_rq_load_avg(now, cfs_rq) && update_tg) + if (update_cfs_rq_load_avg(now, cfs_rq, true) && update_tg) update_tg_load_avg(cfs_rq, 0); - - if (cpu == smp_processor_id() && &rq->cfs == cfs_rq) { - unsigned long max = rq->cpu_capacity_orig; - - /* - * There are a few boundary cases this might miss but it should - * get called often enough that that should (hopefully) not be - * a real problem -- added to that it only calls on the local - * CPU, so if we enqueue remotely we'll miss an update, but - * the next tick/schedule should update. - * - * It will not get called when we go idle, because the idle - * thread is a different class (!fair), nor will the utilization - * number include things like RT tasks. - * - * As is, the util number is not freq-invariant (we'd have to - * implement arch_scale_freq_capacity() for that). - * - * See cpu_util(). - */ - cpufreq_update_util(rq_clock(rq), - min(cfs_rq->avg.util_avg, max), max); - } } static void attach_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se) @@ -2919,6 +2984,8 @@ skip_aging: cfs_rq->avg.load_sum += se->avg.load_sum; cfs_rq->avg.util_avg += se->avg.util_avg; cfs_rq->avg.util_sum += se->avg.util_sum; + + cfs_rq_util_change(cfs_rq); } static void detach_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se) @@ -2931,6 +2998,8 @@ static void detach_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *s cfs_rq->avg.load_sum = max_t(s64, cfs_rq->avg.load_sum - se->avg.load_sum, 0); cfs_rq->avg.util_avg = max_t(long, cfs_rq->avg.util_avg - se->avg.util_avg, 0); cfs_rq->avg.util_sum = max_t(s32, cfs_rq->avg.util_sum - se->avg.util_sum, 0); + + cfs_rq_util_change(cfs_rq); } /* Add the load generated by se into cfs_rq's load average */ @@ -2948,7 +3017,7 @@ enqueue_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se) cfs_rq->curr == se, NULL); } - decayed = update_cfs_rq_load_avg(now, cfs_rq); + decayed = update_cfs_rq_load_avg(now, cfs_rq, !migrated); cfs_rq->runnable_load_avg += sa->load_avg; cfs_rq->runnable_load_sum += sa->load_sum; @@ -3030,7 +3099,14 @@ static int idle_balance(struct rq *this_rq); #else /* CONFIG_SMP */ -static inline void update_load_avg(struct sched_entity *se, int update_tg) {} +static inline void update_load_avg(struct sched_entity *se, int not_used) +{ + struct cfs_rq *cfs_rq = cfs_rq_of(se); + struct rq *rq = rq_of(cfs_rq); + + cpufreq_trigger_update(rq_clock(rq)); +} + static inline void enqueue_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se) {} static inline void @@ -3178,10 +3254,41 @@ static inline void check_schedstat_required(void) #endif } + +/* + * MIGRATION + * + * dequeue + * update_curr() + * update_min_vruntime() + * vruntime -= min_vruntime + * + * enqueue + * update_curr() + * update_min_vruntime() + * vruntime += min_vruntime + * + * this way the vruntime transition between RQs is done when both + * min_vruntime are up-to-date. + * + * WAKEUP (remote) + * + * ->migrate_task_rq_fair() (p->state == TASK_WAKING) + * vruntime -= min_vruntime + * + * enqueue + * update_curr() + * update_min_vruntime() + * vruntime += min_vruntime + * + * this way we don't have the most up-to-date min_vruntime on the originating + * CPU and an up-to-date min_vruntime on the destination CPU. + */ + static void enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags) { - bool renorm = !(flags & ENQUEUE_WAKEUP) || (flags & ENQUEUE_WAKING); + bool renorm = !(flags & ENQUEUE_WAKEUP) || (flags & ENQUEUE_MIGRATED); bool curr = cfs_rq->curr == se; /* @@ -3195,7 +3302,9 @@ enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags) /* * Otherwise, renormalise after, such that we're placed at the current - * moment in time, instead of some random moment in the past. + * moment in time, instead of some random moment in the past. Being + * placed in the past could significantly boost this task to the + * fairness detriment of existing tasks. */ if (renorm && !curr) se->vruntime += cfs_rq->min_vruntime; @@ -4423,7 +4532,7 @@ static void dequeue_task_fair(struct rq *rq, struct task_struct *p, int flags) } #ifdef CONFIG_SMP - +#ifdef CONFIG_NO_HZ_COMMON /* * per rq 'load' arrray crap; XXX kill this. */ @@ -4489,13 +4598,13 @@ decay_load_missed(unsigned long load, unsigned long missed_updates, int idx) } return load; } +#endif /* CONFIG_NO_HZ_COMMON */ /** - * __update_cpu_load - update the rq->cpu_load[] statistics + * __cpu_load_update - update the rq->cpu_load[] statistics * @this_rq: The rq to update statistics for * @this_load: The current load * @pending_updates: The number of missed updates - * @active: !0 for NOHZ_FULL * * Update rq->cpu_load[] statistics. This function is usually called every * scheduler tick (TICK_NSEC). @@ -4524,12 +4633,12 @@ decay_load_missed(unsigned long load, unsigned long missed_updates, int idx) * load[i]_n = (1 - 1/2^i)^n * load[i]_0 * * see decay_load_misses(). For NOHZ_FULL we get to subtract and add the extra - * term. See the @active paramter. + * term. */ -static void __update_cpu_load(struct rq *this_rq, unsigned long this_load, - unsigned long pending_updates, int active) +static void cpu_load_update(struct rq *this_rq, unsigned long this_load, + unsigned long pending_updates) { - unsigned long tickless_load = active ? this_rq->cpu_load[0] : 0; + unsigned long __maybe_unused tickless_load = this_rq->cpu_load[0]; int i, scale; this_rq->nr_load_updates++; @@ -4542,6 +4651,7 @@ static void __update_cpu_load(struct rq *this_rq, unsigned long this_load, /* scale is effectively 1 << i now, and >> i divides by scale */ old_load = this_rq->cpu_load[i]; +#ifdef CONFIG_NO_HZ_COMMON old_load = decay_load_missed(old_load, pending_updates - 1, i); if (tickless_load) { old_load -= decay_load_missed(tickless_load, pending_updates - 1, i); @@ -4552,6 +4662,7 @@ static void __update_cpu_load(struct rq *this_rq, unsigned long this_load, */ old_load += tickless_load; } +#endif new_load = this_load; /* * Round up the averaging division if load is increasing. This @@ -4574,10 +4685,23 @@ static unsigned long weighted_cpuload(const int cpu) } #ifdef CONFIG_NO_HZ_COMMON -static void __update_cpu_load_nohz(struct rq *this_rq, - unsigned long curr_jiffies, - unsigned long load, - int active) +/* + * There is no sane way to deal with nohz on smp when using jiffies because the + * cpu doing the jiffies update might drift wrt the cpu doing the jiffy reading + * causing off-by-one errors in observed deltas; {0,2} instead of {1,1}. + * + * Therefore we need to avoid the delta approach from the regular tick when + * possible since that would seriously skew the load calculation. This is why we + * use cpu_load_update_periodic() for CPUs out of nohz. However we'll rely on + * jiffies deltas for updates happening while in nohz mode (idle ticks, idle + * loop exit, nohz_idle_balance, nohz full exit...) + * + * This means we might still be one tick off for nohz periods. + */ + +static void cpu_load_update_nohz(struct rq *this_rq, + unsigned long curr_jiffies, + unsigned long load) { unsigned long pending_updates; @@ -4589,28 +4713,15 @@ static void __update_cpu_load_nohz(struct rq *this_rq, * In the NOHZ_FULL case, we were non-idle, we should consider * its weighted load. */ - __update_cpu_load(this_rq, load, pending_updates, active); + cpu_load_update(this_rq, load, pending_updates); } } /* - * There is no sane way to deal with nohz on smp when using jiffies because the - * cpu doing the jiffies update might drift wrt the cpu doing the jiffy reading - * causing off-by-one errors in observed deltas; {0,2} instead of {1,1}. - * - * Therefore we cannot use the delta approach from the regular tick since that - * would seriously skew the load calculation. However we'll make do for those - * updates happening while idle (nohz_idle_balance) or coming out of idle - * (tick_nohz_idle_exit). - * - * This means we might still be one tick off for nohz periods. - */ - -/* * Called from nohz_idle_balance() to update the load ratings before doing the * idle balance. */ -static void update_cpu_load_idle(struct rq *this_rq) +static void cpu_load_update_idle(struct rq *this_rq) { /* * bail if there's load or we're actually up-to-date. @@ -4618,38 +4729,71 @@ static void update_cpu_load_idle(struct rq *this_rq) if (weighted_cpuload(cpu_of(this_rq))) return; - __update_cpu_load_nohz(this_rq, READ_ONCE(jiffies), 0, 0); + cpu_load_update_nohz(this_rq, READ_ONCE(jiffies), 0); } /* - * Called from tick_nohz_idle_exit() -- try and fix up the ticks we missed. + * Record CPU load on nohz entry so we know the tickless load to account + * on nohz exit. cpu_load[0] happens then to be updated more frequently + * than other cpu_load[idx] but it should be fine as cpu_load readers + * shouldn't rely into synchronized cpu_load[*] updates. */ -void update_cpu_load_nohz(int active) +void cpu_load_update_nohz_start(void) { struct rq *this_rq = this_rq(); + + /* + * This is all lockless but should be fine. If weighted_cpuload changes + * concurrently we'll exit nohz. And cpu_load write can race with + * cpu_load_update_idle() but both updater would be writing the same. + */ + this_rq->cpu_load[0] = weighted_cpuload(cpu_of(this_rq)); +} + +/* + * Account the tickless load in the end of a nohz frame. + */ +void cpu_load_update_nohz_stop(void) +{ unsigned long curr_jiffies = READ_ONCE(jiffies); - unsigned long load = active ? weighted_cpuload(cpu_of(this_rq)) : 0; + struct rq *this_rq = this_rq(); + unsigned long load; if (curr_jiffies == this_rq->last_load_update_tick) return; + load = weighted_cpuload(cpu_of(this_rq)); raw_spin_lock(&this_rq->lock); - __update_cpu_load_nohz(this_rq, curr_jiffies, load, active); + update_rq_clock(this_rq); + cpu_load_update_nohz(this_rq, curr_jiffies, load); raw_spin_unlock(&this_rq->lock); } -#endif /* CONFIG_NO_HZ */ +#else /* !CONFIG_NO_HZ_COMMON */ +static inline void cpu_load_update_nohz(struct rq *this_rq, + unsigned long curr_jiffies, + unsigned long load) { } +#endif /* CONFIG_NO_HZ_COMMON */ + +static void cpu_load_update_periodic(struct rq *this_rq, unsigned long load) +{ +#ifdef CONFIG_NO_HZ_COMMON + /* See the mess around cpu_load_update_nohz(). */ + this_rq->last_load_update_tick = READ_ONCE(jiffies); +#endif + cpu_load_update(this_rq, load, 1); +} /* * Called from scheduler_tick() */ -void update_cpu_load_active(struct rq *this_rq) +void cpu_load_update_active(struct rq *this_rq) { unsigned long load = weighted_cpuload(cpu_of(this_rq)); - /* - * See the mess around update_cpu_load_idle() / update_cpu_load_nohz(). - */ - this_rq->last_load_update_tick = jiffies; - __update_cpu_load(this_rq, load, 1, 1); + + if (tick_nohz_tick_stopped()) + cpu_load_update_nohz(this_rq, READ_ONCE(jiffies), load); + else + cpu_load_update_periodic(this_rq, load); } /* @@ -4707,46 +4851,6 @@ static unsigned long cpu_avg_load_per_task(int cpu) return 0; } -static void record_wakee(struct task_struct *p) -{ - /* - * Rough decay (wiping) for cost saving, don't worry - * about the boundary, really active task won't care - * about the loss. - */ - if (time_after(jiffies, current->wakee_flip_decay_ts + HZ)) { - current->wakee_flips >>= 1; - current->wakee_flip_decay_ts = jiffies; - } - - if (current->last_wakee != p) { - current->last_wakee = p; - current->wakee_flips++; - } -} - -static void task_waking_fair(struct task_struct *p) -{ - struct sched_entity *se = &p->se; - struct cfs_rq *cfs_rq = cfs_rq_of(se); - u64 min_vruntime; - -#ifndef CONFIG_64BIT - u64 min_vruntime_copy; - - do { - min_vruntime_copy = cfs_rq->min_vruntime_copy; - smp_rmb(); - min_vruntime = cfs_rq->min_vruntime; - } while (min_vruntime != min_vruntime_copy); -#else - min_vruntime = cfs_rq->min_vruntime; -#endif - - se->vruntime -= min_vruntime; - record_wakee(p); -} - #ifdef CONFIG_FAIR_GROUP_SCHED /* * effective_load() calculates the load change as seen from the root_task_group @@ -4862,17 +4966,39 @@ static long effective_load(struct task_group *tg, int cpu, long wl, long wg) #endif +static void record_wakee(struct task_struct *p) +{ + /* + * Only decay a single time; tasks that have less then 1 wakeup per + * jiffy will not have built up many flips. + */ + if (time_after(jiffies, current->wakee_flip_decay_ts + HZ)) { + current->wakee_flips >>= 1; + current->wakee_flip_decay_ts = jiffies; + } + + if (current->last_wakee != p) { + current->last_wakee = p; + current->wakee_flips++; + } +} + /* * Detect M:N waker/wakee relationships via a switching-frequency heuristic. + * * A waker of many should wake a different task than the one last awakened - * at a frequency roughly N times higher than one of its wakees. In order - * to determine whether we should let the load spread vs consolodating to - * shared cache, we look for a minimum 'flip' frequency of llc_size in one - * partner, and a factor of lls_size higher frequency in the other. With - * both conditions met, we can be relatively sure that the relationship is - * non-monogamous, with partner count exceeding socket size. Waker/wakee - * being client/server, worker/dispatcher, interrupt source or whatever is - * irrelevant, spread criteria is apparent partner count exceeds socket size. + * at a frequency roughly N times higher than one of its wakees. + * + * In order to determine whether we should let the load spread vs consolidating + * to shared cache, we look for a minimum 'flip' frequency of llc_size in one + * partner, and a factor of lls_size higher frequency in the other. + * + * With both conditions met, we can be relatively sure that the relationship is + * non-monogamous, with partner count exceeding socket size. + * + * Waker/wakee being client/server, worker/dispatcher, interrupt source or + * whatever is irrelevant, spread criteria is apparent partner count exceeds + * socket size. */ static int wake_wide(struct task_struct *p) { @@ -5177,8 +5303,10 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int sd_flag, int wake_f int want_affine = 0; int sync = wake_flags & WF_SYNC; - if (sd_flag & SD_BALANCE_WAKE) + if (sd_flag & SD_BALANCE_WAKE) { + record_wakee(p); want_affine = !wake_wide(p) && cpumask_test_cpu(cpu, tsk_cpus_allowed(p)); + } rcu_read_lock(); for_each_domain(cpu, tmp) { @@ -5258,6 +5386,32 @@ select_task_rq_fair(struct task_struct *p, int prev_cpu, int sd_flag, int wake_f static void migrate_task_rq_fair(struct task_struct *p) { /* + * As blocked tasks retain absolute vruntime the migration needs to + * deal with this by subtracting the old and adding the new + * min_vruntime -- the latter is done by enqueue_entity() when placing + * the task on the new runqueue. + */ + if (p->state == TASK_WAKING) { + struct sched_entity *se = &p->se; + struct cfs_rq *cfs_rq = cfs_rq_of(se); + u64 min_vruntime; + +#ifndef CONFIG_64BIT + u64 min_vruntime_copy; + + do { + min_vruntime_copy = cfs_rq->min_vruntime_copy; + smp_rmb(); + min_vruntime = cfs_rq->min_vruntime; + } while (min_vruntime != min_vruntime_copy); +#else + min_vruntime = cfs_rq->min_vruntime; +#endif + + se->vruntime -= min_vruntime; + } + + /* * We are supposed to update the task to "current" time, then its up to date * and ready to go to new CPU/cfs_rq. But we have difficulty in getting * what current time is, so simply throw away the out-of-date time. This @@ -5440,7 +5594,7 @@ preempt: } static struct task_struct * -pick_next_task_fair(struct rq *rq, struct task_struct *prev) +pick_next_task_fair(struct rq *rq, struct task_struct *prev, struct pin_cookie cookie) { struct cfs_rq *cfs_rq = &rq->cfs; struct sched_entity *se; @@ -5553,9 +5707,9 @@ idle: * further scheduler activity on it and we're being very careful to * re-start the picking loop. */ - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, cookie); new_tasks = idle_balance(rq); - lockdep_pin_lock(&rq->lock); + lockdep_repin_lock(&rq->lock, cookie); /* * Because idle_balance() releases (and re-acquires) rq->lock, it is * possible for any higher priority task to appear. In that case we @@ -5654,7 +5808,7 @@ static bool yield_to_task_fair(struct rq *rq, struct task_struct *p, bool preemp * W_i,0 = \Sum_j w_i,j (2) * * Where w_i,j is the weight of the j-th runnable task on cpu i. This weight - * is derived from the nice value as per prio_to_weight[]. + * is derived from the nice value as per sched_prio_to_weight[]. * * The weight average is an exponential decay average of the instantaneous * weight: @@ -6156,7 +6310,7 @@ static void update_blocked_averages(int cpu) if (throttled_hierarchy(cfs_rq)) continue; - if (update_cfs_rq_load_avg(cfs_rq_clock_task(cfs_rq), cfs_rq)) + if (update_cfs_rq_load_avg(cfs_rq_clock_task(cfs_rq), cfs_rq, true)) update_tg_load_avg(cfs_rq, 0); } raw_spin_unlock_irqrestore(&rq->lock, flags); @@ -6217,7 +6371,7 @@ static inline void update_blocked_averages(int cpu) raw_spin_lock_irqsave(&rq->lock, flags); update_rq_clock(rq); - update_cfs_rq_load_avg(cfs_rq_clock_task(cfs_rq), cfs_rq); + update_cfs_rq_load_avg(cfs_rq_clock_task(cfs_rq), cfs_rq, true); raw_spin_unlock_irqrestore(&rq->lock, flags); } @@ -6626,6 +6780,9 @@ static bool update_sd_pick_busiest(struct lb_env *env, if (!(env->sd->flags & SD_ASYM_PACKING)) return true; + /* No ASYM_PACKING if target cpu is already busy */ + if (env->idle == CPU_NOT_IDLE) + return true; /* * ASYM_PACKING needs to move all the work to the lowest * numbered CPUs in the group, therefore mark all groups @@ -6635,7 +6792,8 @@ static bool update_sd_pick_busiest(struct lb_env *env, if (!sds->busiest) return true; - if (group_first_cpu(sds->busiest) > group_first_cpu(sg)) + /* Prefer to move from highest possible cpu's work */ + if (group_first_cpu(sds->busiest) < group_first_cpu(sg)) return true; } @@ -6781,6 +6939,9 @@ static int check_asym_packing(struct lb_env *env, struct sd_lb_stats *sds) if (!(env->sd->flags & SD_ASYM_PACKING)) return 0; + if (env->idle == CPU_NOT_IDLE) + return 0; + if (!sds->busiest) return 0; @@ -6889,9 +7050,10 @@ static inline void calculate_imbalance(struct lb_env *env, struct sd_lb_stats *s } /* - * In the presence of smp nice balancing, certain scenarios can have - * max load less than avg load(as we skip the groups at or below - * its cpu_capacity, while calculating max_load..) + * Avg load of busiest sg can be less and avg load of local sg can + * be greater than avg load across all sgs of sd because avg load + * factors in sg capacity and sgs with smaller group_type are + * skipped when updating the busiest sg: */ if (busiest->avg_load <= sds->avg_load || local->avg_load >= sds->avg_load) { @@ -6904,11 +7066,12 @@ static inline void calculate_imbalance(struct lb_env *env, struct sd_lb_stats *s */ if (busiest->group_type == group_overloaded && local->group_type == group_overloaded) { - load_above_capacity = busiest->sum_nr_running * - SCHED_LOAD_SCALE; - if (load_above_capacity > busiest->group_capacity) + load_above_capacity = busiest->sum_nr_running * SCHED_CAPACITY_SCALE; + if (load_above_capacity > busiest->group_capacity) { load_above_capacity -= busiest->group_capacity; - else + load_above_capacity *= NICE_0_LOAD; + load_above_capacity /= busiest->group_capacity; + } else load_above_capacity = ~0UL; } @@ -6916,9 +7079,8 @@ static inline void calculate_imbalance(struct lb_env *env, struct sd_lb_stats *s * We're trying to get all the cpus to the average_load, so we don't * want to push ourselves above the average load, nor do we wish to * reduce the max loaded cpu below the average load. At the same time, - * we also don't want to reduce the group load below the group capacity - * (so that we can implement power-savings policies etc). Thus we look - * for the minimum possible imbalance. + * we also don't want to reduce the group load below the group + * capacity. Thus we look for the minimum possible imbalance. */ max_pull = min(busiest->avg_load - sds->avg_load, load_above_capacity); @@ -6942,10 +7104,7 @@ static inline void calculate_imbalance(struct lb_env *env, struct sd_lb_stats *s /** * find_busiest_group - Returns the busiest group within the sched_domain - * if there is an imbalance. If there isn't an imbalance, and - * the user has opted for power-savings, it returns a group whose - * CPUs can be put to idle by rebalancing those tasks elsewhere, if - * such a group exists. + * if there is an imbalance. * * Also calculates the amount of weighted load which should be moved * to restore balance. @@ -6953,9 +7112,6 @@ static inline void calculate_imbalance(struct lb_env *env, struct sd_lb_stats *s * @env: The load balancing environment. * * Return: - The busiest group if imbalance exists. - * - If no imbalance and user has opted for power-savings balance, - * return the least loaded group whose CPUs can be - * put to idle by rebalancing its tasks onto our group. */ static struct sched_group *find_busiest_group(struct lb_env *env) { @@ -6973,8 +7129,7 @@ static struct sched_group *find_busiest_group(struct lb_env *env) busiest = &sds.busiest_stat; /* ASYM feature bypasses nice load balance check */ - if ((env->idle == CPU_IDLE || env->idle == CPU_NEWLY_IDLE) && - check_asym_packing(env, &sds)) + if (check_asym_packing(env, &sds)) return sds.busiest; /* There is no busy sibling group to pull tasks from */ @@ -7399,10 +7554,7 @@ more_balance: &busiest->active_balance_work); } - /* - * We've kicked active balancing, reset the failure - * counter. - */ + /* We've kicked active balancing, force task migration. */ sd->nr_balance_failed = sd->cache_nice_tries+1; } } else @@ -7637,10 +7789,13 @@ static int active_load_balance_cpu_stop(void *data) schedstat_inc(sd, alb_count); p = detach_one_task(&env); - if (p) + if (p) { schedstat_inc(sd, alb_pushed); - else + /* Active balancing done, reset the failure counter. */ + sd->nr_balance_failed = 0; + } else { schedstat_inc(sd, alb_failed); + } } rcu_read_unlock(); out_unlock: @@ -7711,7 +7866,7 @@ static void nohz_balancer_kick(void) return; } -static inline void nohz_balance_exit_idle(int cpu) +void nohz_balance_exit_idle(unsigned int cpu) { if (unlikely(test_bit(NOHZ_TICK_STOPPED, nohz_flags(cpu)))) { /* @@ -7784,18 +7939,6 @@ void nohz_balance_enter_idle(int cpu) atomic_inc(&nohz.nr_cpus); set_bit(NOHZ_TICK_STOPPED, nohz_flags(cpu)); } - -static int sched_ilb_notifier(struct notifier_block *nfb, - unsigned long action, void *hcpu) -{ - switch (action & ~CPU_TASKS_FROZEN) { - case CPU_DYING: - nohz_balance_exit_idle(smp_processor_id()); - return NOTIFY_OK; - default: - return NOTIFY_DONE; - } -} #endif static DEFINE_SPINLOCK(balancing); @@ -7957,7 +8100,7 @@ static void nohz_idle_balance(struct rq *this_rq, enum cpu_idle_type idle) if (time_after_eq(jiffies, rq->next_balance)) { raw_spin_lock_irq(&rq->lock); update_rq_clock(rq); - update_cpu_load_idle(rq); + cpu_load_update_idle(rq); raw_spin_unlock_irq(&rq->lock); rebalance_domains(rq, CPU_IDLE); } @@ -8382,6 +8525,7 @@ int alloc_fair_sched_group(struct task_group *tg, struct task_group *parent) init_cfs_rq(cfs_rq); init_tg_cfs_entry(tg, cfs_rq, se, i, parent->se[i]); init_entity_runnable_average(se); + post_init_entity_util_avg(se); } return 1; @@ -8538,7 +8682,6 @@ const struct sched_class fair_sched_class = { .rq_online = rq_online_fair, .rq_offline = rq_offline_fair, - .task_waking = task_waking_fair, .task_dead = task_dead_fair, .set_cpus_allowed = set_cpus_allowed_common, #endif @@ -8600,7 +8743,6 @@ __init void init_sched_fair_class(void) #ifdef CONFIG_NO_HZ_COMMON nohz.next_balance = jiffies; zalloc_cpumask_var(&nohz.idle_cpus_mask, GFP_NOWAIT); - cpu_notifier(sched_ilb_notifier, 0); #endif #endif /* SMP */ diff --git a/kernel/sched/idle_task.c b/kernel/sched/idle_task.c index 47ce949..2ce5458 100644 --- a/kernel/sched/idle_task.c +++ b/kernel/sched/idle_task.c @@ -24,7 +24,7 @@ static void check_preempt_curr_idle(struct rq *rq, struct task_struct *p, int fl } static struct task_struct * -pick_next_task_idle(struct rq *rq, struct task_struct *prev) +pick_next_task_idle(struct rq *rq, struct task_struct *prev, struct pin_cookie cookie) { put_prev_task(rq, prev); diff --git a/kernel/sched/loadavg.c b/kernel/sched/loadavg.c index ef71590..b0b93fd 100644 --- a/kernel/sched/loadavg.c +++ b/kernel/sched/loadavg.c @@ -99,10 +99,13 @@ long calc_load_fold_active(struct rq *this_rq) static unsigned long calc_load(unsigned long load, unsigned long exp, unsigned long active) { - load *= exp; - load += active * (FIXED_1 - exp); - load += 1UL << (FSHIFT - 1); - return load >> FSHIFT; + unsigned long newload; + + newload = load * exp + active * (FIXED_1 - exp); + if (active >= load) + newload += FIXED_1-1; + + return newload / FIXED_1; } #ifdef CONFIG_NO_HZ_COMMON diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c index c41ea7a..d5690b7 100644 --- a/kernel/sched/rt.c +++ b/kernel/sched/rt.c @@ -334,7 +334,7 @@ static void inc_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq) rt_rq = &rq_of_rt_rq(rt_rq)->rt; rt_rq->rt_nr_total++; - if (p->nr_cpus_allowed > 1) + if (tsk_nr_cpus_allowed(p) > 1) rt_rq->rt_nr_migratory++; update_rt_migration(rt_rq); @@ -351,7 +351,7 @@ static void dec_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq) rt_rq = &rq_of_rt_rq(rt_rq)->rt; rt_rq->rt_nr_total--; - if (p->nr_cpus_allowed > 1) + if (tsk_nr_cpus_allowed(p) > 1) rt_rq->rt_nr_migratory--; update_rt_migration(rt_rq); @@ -953,14 +953,14 @@ static void update_curr_rt(struct rq *rq) if (curr->sched_class != &rt_sched_class) return; - /* Kick cpufreq (see the comment in linux/cpufreq.h). */ - if (cpu_of(rq) == smp_processor_id()) - cpufreq_trigger_update(rq_clock(rq)); - delta_exec = rq_clock_task(rq) - curr->se.exec_start; if (unlikely((s64)delta_exec <= 0)) return; + /* Kick cpufreq (see the comment in linux/cpufreq.h). */ + if (cpu_of(rq) == smp_processor_id()) + cpufreq_trigger_update(rq_clock(rq)); + schedstat_set(curr->se.statistics.exec_max, max(curr->se.statistics.exec_max, delta_exec)); @@ -1324,7 +1324,7 @@ enqueue_task_rt(struct rq *rq, struct task_struct *p, int flags) enqueue_rt_entity(rt_se, flags); - if (!task_current(rq, p) && p->nr_cpus_allowed > 1) + if (!task_current(rq, p) && tsk_nr_cpus_allowed(p) > 1) enqueue_pushable_task(rq, p); } @@ -1413,7 +1413,7 @@ select_task_rq_rt(struct task_struct *p, int cpu, int sd_flag, int flags) * will have to sort it out. */ if (curr && unlikely(rt_task(curr)) && - (curr->nr_cpus_allowed < 2 || + (tsk_nr_cpus_allowed(curr) < 2 || curr->prio <= p->prio)) { int target = find_lowest_rq(p); @@ -1437,7 +1437,7 @@ static void check_preempt_equal_prio(struct rq *rq, struct task_struct *p) * Current can't be migrated, useless to reschedule, * let's hope p can move out. */ - if (rq->curr->nr_cpus_allowed == 1 || + if (tsk_nr_cpus_allowed(rq->curr) == 1 || !cpupri_find(&rq->rd->cpupri, rq->curr, NULL)) return; @@ -1445,7 +1445,7 @@ static void check_preempt_equal_prio(struct rq *rq, struct task_struct *p) * p is migratable, so let's not schedule it and * see if it is pushed or pulled somewhere else. */ - if (p->nr_cpus_allowed != 1 + if (tsk_nr_cpus_allowed(p) != 1 && cpupri_find(&rq->rd->cpupri, p, NULL)) return; @@ -1524,7 +1524,7 @@ static struct task_struct *_pick_next_task_rt(struct rq *rq) } static struct task_struct * -pick_next_task_rt(struct rq *rq, struct task_struct *prev) +pick_next_task_rt(struct rq *rq, struct task_struct *prev, struct pin_cookie cookie) { struct task_struct *p; struct rt_rq *rt_rq = &rq->rt; @@ -1536,9 +1536,9 @@ pick_next_task_rt(struct rq *rq, struct task_struct *prev) * disabled avoiding further scheduler activity on it and we're * being very careful to re-start the picking loop. */ - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, cookie); pull_rt_task(rq); - lockdep_pin_lock(&rq->lock); + lockdep_repin_lock(&rq->lock, cookie); /* * pull_rt_task() can drop (and re-acquire) rq->lock; this * means a dl or stop task can slip in, in which case we need @@ -1579,7 +1579,7 @@ static void put_prev_task_rt(struct rq *rq, struct task_struct *p) * The previous task needs to be made eligible for pushing * if it is still active */ - if (on_rt_rq(&p->rt) && p->nr_cpus_allowed > 1) + if (on_rt_rq(&p->rt) && tsk_nr_cpus_allowed(p) > 1) enqueue_pushable_task(rq, p); } @@ -1629,7 +1629,7 @@ static int find_lowest_rq(struct task_struct *task) if (unlikely(!lowest_mask)) return -1; - if (task->nr_cpus_allowed == 1) + if (tsk_nr_cpus_allowed(task) == 1) return -1; /* No other targets possible */ if (!cpupri_find(&task_rq(task)->rd->cpupri, task, lowest_mask)) @@ -1729,6 +1729,7 @@ static struct rq *find_lock_lowest_rq(struct task_struct *task, struct rq *rq) !cpumask_test_cpu(lowest_rq->cpu, tsk_cpus_allowed(task)) || task_running(rq, task) || + !rt_task(task) || !task_on_rq_queued(task))) { double_unlock_balance(rq, lowest_rq); @@ -1761,7 +1762,7 @@ static struct task_struct *pick_next_pushable_task(struct rq *rq) BUG_ON(rq->cpu != task_cpu(p)); BUG_ON(task_current(rq, p)); - BUG_ON(p->nr_cpus_allowed <= 1); + BUG_ON(tsk_nr_cpus_allowed(p) <= 1); BUG_ON(!task_on_rq_queued(p)); BUG_ON(!rt_task(p)); @@ -2121,9 +2122,9 @@ static void task_woken_rt(struct rq *rq, struct task_struct *p) { if (!task_running(rq, p) && !test_tsk_need_resched(rq->curr) && - p->nr_cpus_allowed > 1 && + tsk_nr_cpus_allowed(p) > 1 && (dl_task(rq->curr) || rt_task(rq->curr)) && - (rq->curr->nr_cpus_allowed < 2 || + (tsk_nr_cpus_allowed(rq->curr) < 2 || rq->curr->prio <= p->prio)) push_rt_tasks(rq); } @@ -2196,7 +2197,7 @@ static void switched_to_rt(struct rq *rq, struct task_struct *p) */ if (task_on_rq_queued(p) && rq->curr != p) { #ifdef CONFIG_SMP - if (p->nr_cpus_allowed > 1 && rq->rt.overloaded) + if (tsk_nr_cpus_allowed(p) > 1 && rq->rt.overloaded) queue_push_tasks(rq); #else if (p->prio < rq->curr->prio) diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index ec2e8d2..72f1f30 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -31,9 +31,9 @@ extern void calc_global_load_tick(struct rq *this_rq); extern long calc_load_fold_active(struct rq *this_rq); #ifdef CONFIG_SMP -extern void update_cpu_load_active(struct rq *this_rq); +extern void cpu_load_update_active(struct rq *this_rq); #else -static inline void update_cpu_load_active(struct rq *this_rq) { } +static inline void cpu_load_update_active(struct rq *this_rq) { } #endif /* @@ -49,25 +49,32 @@ static inline void update_cpu_load_active(struct rq *this_rq) { } * and does not change the user-interface for setting shares/weights. * * We increase resolution only if we have enough bits to allow this increased - * resolution (i.e. BITS_PER_LONG > 32). The costs for increasing resolution - * when BITS_PER_LONG <= 32 are pretty high and the returns do not justify the - * increased costs. + * resolution (i.e. 64bit). The costs for increasing resolution when 32bit are + * pretty high and the returns do not justify the increased costs. + * + * Really only required when CONFIG_FAIR_GROUP_SCHED is also set, but to + * increase coverage and consistency always enable it on 64bit platforms. */ -#if 0 /* BITS_PER_LONG > 32 -- currently broken: it increases power usage under light load */ -# define SCHED_LOAD_RESOLUTION 10 -# define scale_load(w) ((w) << SCHED_LOAD_RESOLUTION) -# define scale_load_down(w) ((w) >> SCHED_LOAD_RESOLUTION) +#ifdef CONFIG_64BIT +# define NICE_0_LOAD_SHIFT (SCHED_FIXEDPOINT_SHIFT + SCHED_FIXEDPOINT_SHIFT) +# define scale_load(w) ((w) << SCHED_FIXEDPOINT_SHIFT) +# define scale_load_down(w) ((w) >> SCHED_FIXEDPOINT_SHIFT) #else -# define SCHED_LOAD_RESOLUTION 0 +# define NICE_0_LOAD_SHIFT (SCHED_FIXEDPOINT_SHIFT) # define scale_load(w) (w) # define scale_load_down(w) (w) #endif -#define SCHED_LOAD_SHIFT (10 + SCHED_LOAD_RESOLUTION) -#define SCHED_LOAD_SCALE (1L << SCHED_LOAD_SHIFT) - -#define NICE_0_LOAD SCHED_LOAD_SCALE -#define NICE_0_SHIFT SCHED_LOAD_SHIFT +/* + * Task weight (visible to users) and its load (invisible to users) have + * independent resolution, but they should be well calibrated. We use + * scale_load() and scale_load_down(w) to convert between them. The + * following must be true: + * + * scale_load(sched_prio_to_weight[USER_PRIO(NICE_TO_PRIO(0))]) == NICE_0_LOAD + * + */ +#define NICE_0_LOAD (1L << NICE_0_LOAD_SHIFT) /* * Single value that decides SCHED_DEADLINE internal math precision. @@ -585,11 +592,13 @@ struct rq { #endif #define CPU_LOAD_IDX_MAX 5 unsigned long cpu_load[CPU_LOAD_IDX_MAX]; - unsigned long last_load_update_tick; #ifdef CONFIG_NO_HZ_COMMON +#ifdef CONFIG_SMP + unsigned long last_load_update_tick; +#endif /* CONFIG_SMP */ u64 nohz_stamp; unsigned long nohz_flags; -#endif +#endif /* CONFIG_NO_HZ_COMMON */ #ifdef CONFIG_NO_HZ_FULL unsigned long last_sched_tick; #endif @@ -854,7 +863,7 @@ DECLARE_PER_CPU(struct sched_domain *, sd_asym); struct sched_group_capacity { atomic_t ref; /* - * CPU capacity of this group, SCHED_LOAD_SCALE being max capacity + * CPU capacity of this group, SCHED_CAPACITY_SCALE being max capacity * for a single CPU. */ unsigned int capacity; @@ -1159,7 +1168,7 @@ extern const u32 sched_prio_to_wmult[40]; * * ENQUEUE_HEAD - place at front of runqueue (tail if not specified) * ENQUEUE_REPLENISH - CBS (replenish runtime and postpone deadline) - * ENQUEUE_WAKING - sched_class::task_waking was called + * ENQUEUE_MIGRATED - the task was migrated during wakeup * */ @@ -1174,9 +1183,9 @@ extern const u32 sched_prio_to_wmult[40]; #define ENQUEUE_HEAD 0x08 #define ENQUEUE_REPLENISH 0x10 #ifdef CONFIG_SMP -#define ENQUEUE_WAKING 0x20 +#define ENQUEUE_MIGRATED 0x20 #else -#define ENQUEUE_WAKING 0x00 +#define ENQUEUE_MIGRATED 0x00 #endif #define RETRY_TASK ((void *)-1UL) @@ -1200,14 +1209,14 @@ struct sched_class { * tasks. */ struct task_struct * (*pick_next_task) (struct rq *rq, - struct task_struct *prev); + struct task_struct *prev, + struct pin_cookie cookie); void (*put_prev_task) (struct rq *rq, struct task_struct *p); #ifdef CONFIG_SMP int (*select_task_rq)(struct task_struct *p, int task_cpu, int sd_flag, int flags); void (*migrate_task_rq)(struct task_struct *p); - void (*task_waking) (struct task_struct *task); void (*task_woken) (struct rq *this_rq, struct task_struct *task); void (*set_cpus_allowed)(struct task_struct *p, @@ -1313,6 +1322,7 @@ extern void init_dl_task_timer(struct sched_dl_entity *dl_se); unsigned long to_ratio(u64 period, u64 runtime); extern void init_entity_runnable_average(struct sched_entity *se); +extern void post_init_entity_util_avg(struct sched_entity *se); #ifdef CONFIG_NO_HZ_FULL extern bool sched_can_stop_tick(struct rq *rq); @@ -1448,86 +1458,32 @@ static inline void sched_rt_avg_update(struct rq *rq, u64 rt_delta) { } static inline void sched_avg_update(struct rq *rq) { } #endif -/* - * __task_rq_lock - lock the rq @p resides on. - */ -static inline struct rq *__task_rq_lock(struct task_struct *p) - __acquires(rq->lock) -{ - struct rq *rq; - - lockdep_assert_held(&p->pi_lock); - - for (;;) { - rq = task_rq(p); - raw_spin_lock(&rq->lock); - if (likely(rq == task_rq(p) && !task_on_rq_migrating(p))) { - lockdep_pin_lock(&rq->lock); - return rq; - } - raw_spin_unlock(&rq->lock); - - while (unlikely(task_on_rq_migrating(p))) - cpu_relax(); - } -} +struct rq_flags { + unsigned long flags; + struct pin_cookie cookie; +}; -/* - * task_rq_lock - lock p->pi_lock and lock the rq @p resides on. - */ -static inline struct rq *task_rq_lock(struct task_struct *p, unsigned long *flags) +struct rq *__task_rq_lock(struct task_struct *p, struct rq_flags *rf) + __acquires(rq->lock); +struct rq *task_rq_lock(struct task_struct *p, struct rq_flags *rf) __acquires(p->pi_lock) - __acquires(rq->lock) -{ - struct rq *rq; - - for (;;) { - raw_spin_lock_irqsave(&p->pi_lock, *flags); - rq = task_rq(p); - raw_spin_lock(&rq->lock); - /* - * move_queued_task() task_rq_lock() - * - * ACQUIRE (rq->lock) - * [S] ->on_rq = MIGRATING [L] rq = task_rq() - * WMB (__set_task_cpu()) ACQUIRE (rq->lock); - * [S] ->cpu = new_cpu [L] task_rq() - * [L] ->on_rq - * RELEASE (rq->lock) - * - * If we observe the old cpu in task_rq_lock, the acquire of - * the old rq->lock will fully serialize against the stores. - * - * If we observe the new cpu in task_rq_lock, the acquire will - * pair with the WMB to ensure we must then also see migrating. - */ - if (likely(rq == task_rq(p) && !task_on_rq_migrating(p))) { - lockdep_pin_lock(&rq->lock); - return rq; - } - raw_spin_unlock(&rq->lock); - raw_spin_unlock_irqrestore(&p->pi_lock, *flags); + __acquires(rq->lock); - while (unlikely(task_on_rq_migrating(p))) - cpu_relax(); - } -} - -static inline void __task_rq_unlock(struct rq *rq) +static inline void __task_rq_unlock(struct rq *rq, struct rq_flags *rf) __releases(rq->lock) { - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, rf->cookie); raw_spin_unlock(&rq->lock); } static inline void -task_rq_unlock(struct rq *rq, struct task_struct *p, unsigned long *flags) +task_rq_unlock(struct rq *rq, struct task_struct *p, struct rq_flags *rf) __releases(rq->lock) __releases(p->pi_lock) { - lockdep_unpin_lock(&rq->lock); + lockdep_unpin_lock(&rq->lock, rf->cookie); raw_spin_unlock(&rq->lock); - raw_spin_unlock_irqrestore(&p->pi_lock, *flags); + raw_spin_unlock_irqrestore(&p->pi_lock, rf->flags); } #ifdef CONFIG_SMP @@ -1743,6 +1699,10 @@ enum rq_nohz_flag_bits { }; #define nohz_flags(cpu) (&cpu_rq(cpu)->nohz_flags) + +extern void nohz_balance_exit_idle(unsigned int cpu); +#else +static inline void nohz_balance_exit_idle(unsigned int cpu) { } #endif #ifdef CONFIG_IRQ_TIME_ACCOUNTING @@ -1842,6 +1802,14 @@ static inline void cpufreq_update_util(u64 time, unsigned long util, unsigned lo static inline void cpufreq_trigger_update(u64 time) {} #endif /* CONFIG_CPU_FREQ */ +#ifdef arch_scale_freq_capacity +#ifndef arch_scale_freq_invariant +#define arch_scale_freq_invariant() (true) +#endif +#else /* arch_scale_freq_capacity */ +#define arch_scale_freq_invariant() (false) +#endif + static inline void account_reset_rq(struct rq *rq) { #ifdef CONFIG_IRQ_TIME_ACCOUNTING diff --git a/kernel/sched/stop_task.c b/kernel/sched/stop_task.c index cbc67da..604297a 100644 --- a/kernel/sched/stop_task.c +++ b/kernel/sched/stop_task.c @@ -24,7 +24,7 @@ check_preempt_curr_stop(struct rq *rq, struct task_struct *p, int flags) } static struct task_struct * -pick_next_task_stop(struct rq *rq, struct task_struct *prev) +pick_next_task_stop(struct rq *rq, struct task_struct *prev, struct pin_cookie cookie) { struct task_struct *stop = rq->stop; diff --git a/kernel/signal.c b/kernel/signal.c index aa9bf00..ab122a2 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -3099,12 +3099,14 @@ do_sigaltstack (const stack_t __user *uss, stack_t __user *uoss, unsigned long s oss.ss_sp = (void __user *) current->sas_ss_sp; oss.ss_size = current->sas_ss_size; - oss.ss_flags = sas_ss_flags(sp); + oss.ss_flags = sas_ss_flags(sp) | + (current->sas_ss_flags & SS_FLAG_BITS); if (uss) { void __user *ss_sp; size_t ss_size; - int ss_flags; + unsigned ss_flags; + int ss_mode; error = -EFAULT; if (!access_ok(VERIFY_READ, uss, sizeof(*uss))) @@ -3119,18 +3121,13 @@ do_sigaltstack (const stack_t __user *uss, stack_t __user *uoss, unsigned long s if (on_sig_stack(sp)) goto out; + ss_mode = ss_flags & ~SS_FLAG_BITS; error = -EINVAL; - /* - * Note - this code used to test ss_flags incorrectly: - * old code may have been written using ss_flags==0 - * to mean ss_flags==SS_ONSTACK (as this was the only - * way that worked) - this fix preserves that older - * mechanism. - */ - if (ss_flags != SS_DISABLE && ss_flags != SS_ONSTACK && ss_flags != 0) + if (ss_mode != SS_DISABLE && ss_mode != SS_ONSTACK && + ss_mode != 0) goto out; - if (ss_flags == SS_DISABLE) { + if (ss_mode == SS_DISABLE) { ss_size = 0; ss_sp = NULL; } else { @@ -3141,6 +3138,7 @@ do_sigaltstack (const stack_t __user *uss, stack_t __user *uoss, unsigned long s current->sas_ss_sp = (unsigned long) ss_sp; current->sas_ss_size = ss_size; + current->sas_ss_flags = ss_flags; } error = 0; @@ -3171,9 +3169,14 @@ int restore_altstack(const stack_t __user *uss) int __save_altstack(stack_t __user *uss, unsigned long sp) { struct task_struct *t = current; - return __put_user((void __user *)t->sas_ss_sp, &uss->ss_sp) | - __put_user(sas_ss_flags(sp), &uss->ss_flags) | + int err = __put_user((void __user *)t->sas_ss_sp, &uss->ss_sp) | + __put_user(t->sas_ss_flags, &uss->ss_flags) | __put_user(t->sas_ss_size, &uss->ss_size); + if (err) + return err; + if (t->sas_ss_flags & SS_AUTODISARM) + sas_ss_reset(t); + return 0; } #ifdef CONFIG_COMPAT diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 725587f..c8b3186 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -130,6 +130,9 @@ static int one_thousand = 1000; #ifdef CONFIG_PRINTK static int ten_thousand = 10000; #endif +#ifdef CONFIG_PERF_EVENTS +static int six_hundred_forty_kb = 640 * 1024; +#endif /* this is needed for the proc_doulongvec_minmax of vm_dirty_bytes */ static unsigned long dirty_bytes_min = 2 * PAGE_SIZE; @@ -1144,6 +1147,15 @@ static struct ctl_table kern_table[] = { .extra1 = &zero, .extra2 = &one_hundred, }, + { + .procname = "perf_event_max_stack", + .data = NULL, /* filled in by handler */ + .maxlen = sizeof(sysctl_perf_event_max_stack), + .mode = 0644, + .proc_handler = perf_event_max_stack_handler, + .extra1 = &zero, + .extra2 = &six_hundred_forty_kb, + }, #endif #ifdef CONFIG_KMEMCHECK { diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c index 58e3310..536ada8 100644 --- a/kernel/time/tick-sched.c +++ b/kernel/time/tick-sched.c @@ -262,7 +262,7 @@ static void tick_nohz_dep_set_all(atomic_t *dep, { int prev; - prev = atomic_fetch_or(dep, BIT(bit)); + prev = atomic_fetch_or(BIT(bit), dep); if (!prev) tick_nohz_full_kick_all(); } @@ -292,7 +292,7 @@ void tick_nohz_dep_set_cpu(int cpu, enum tick_dep_bits bit) ts = per_cpu_ptr(&tick_cpu_sched, cpu); - prev = atomic_fetch_or(&ts->tick_dep_mask, BIT(bit)); + prev = atomic_fetch_or(BIT(bit), &ts->tick_dep_mask); if (!prev) { preempt_disable(); /* Perf needs local kick that is NMI safe */ @@ -776,6 +776,7 @@ static ktime_t tick_nohz_stop_sched_tick(struct tick_sched *ts, if (!ts->tick_stopped) { nohz_balance_enter_idle(cpu); calc_load_enter_idle(); + cpu_load_update_nohz_start(); ts->last_tick = hrtimer_get_expires(&ts->sched_timer); ts->tick_stopped = 1; @@ -802,11 +803,11 @@ out: return tick; } -static void tick_nohz_restart_sched_tick(struct tick_sched *ts, ktime_t now, int active) +static void tick_nohz_restart_sched_tick(struct tick_sched *ts, ktime_t now) { /* Update jiffies first */ tick_do_update_jiffies64(now); - update_cpu_load_nohz(active); + cpu_load_update_nohz_stop(); calc_load_exit_idle(); touch_softlockup_watchdog_sched(); @@ -833,7 +834,7 @@ static void tick_nohz_full_update_tick(struct tick_sched *ts) if (can_stop_full_tick(ts)) tick_nohz_stop_sched_tick(ts, ktime_get(), cpu); else if (ts->tick_stopped) - tick_nohz_restart_sched_tick(ts, ktime_get(), 1); + tick_nohz_restart_sched_tick(ts, ktime_get()); #endif } @@ -1024,7 +1025,7 @@ void tick_nohz_idle_exit(void) tick_nohz_stop_idle(ts, now); if (ts->tick_stopped) { - tick_nohz_restart_sched_tick(ts, now, 0); + tick_nohz_restart_sched_tick(ts, now); tick_nohz_account_idle_ticks(ts); } diff --git a/kernel/torture.c b/kernel/torture.c index 44aa462..fa0bdee 100644 --- a/kernel/torture.c +++ b/kernel/torture.c @@ -451,6 +451,7 @@ static int torture_shutdown(void *arg) torture_shutdown_hook(); else VERBOSE_TOROUT_STRING("No torture_shutdown_hook(), skipping."); + ftrace_dump(DUMP_ALL); kernel_power_off(); /* Shut down the system. */ return 0; } @@ -602,8 +603,9 @@ bool torture_init_begin(char *ttype, bool v, int *runnable) { mutex_lock(&fullstop_mutex); if (torture_type != NULL) { - pr_alert("torture_init_begin: refusing %s init: %s running", + pr_alert("torture_init_begin: Refusing %s init: %s running.\n", ttype, torture_type); + pr_alert("torture_init_begin: One torture test at a time!\n"); mutex_unlock(&fullstop_mutex); return false; } diff --git a/kernel/trace/power-traces.c b/kernel/trace/power-traces.c index 81b8745..0c7dee2 100644 --- a/kernel/trace/power-traces.c +++ b/kernel/trace/power-traces.c @@ -15,5 +15,6 @@ EXPORT_TRACEPOINT_SYMBOL_GPL(suspend_resume); EXPORT_TRACEPOINT_SYMBOL_GPL(cpu_idle); +EXPORT_TRACEPOINT_SYMBOL_GPL(cpu_frequency); EXPORT_TRACEPOINT_SYMBOL_GPL(powernv_throttle); diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 00df25f..e11108f 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -47,6 +47,9 @@ static int perf_trace_event_perm(struct trace_event_call *tp_event, if (perf_paranoid_tracepoint_raw() && !capable(CAP_SYS_ADMIN)) return -EPERM; + if (!is_sampling_event(p_event)) + return 0; + /* * We don't allow user space callchains for function trace * event, due to issues with page faults while tracing page diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 05ddc08..6f96586 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -2095,8 +2095,13 @@ event_create_dir(struct dentry *parent, struct trace_event_file *file) trace_create_file("filter", 0644, file->dir, file, &ftrace_event_filter_fops); - trace_create_file("trigger", 0644, file->dir, file, - &event_trigger_fops); + /* + * Only event directories that can be enabled should have + * triggers. + */ + if (!(call->flags & TRACE_EVENT_FL_IGNORE_ENABLE)) + trace_create_file("trigger", 0644, file->dir, file, + &event_trigger_fops); trace_create_file("format", 0444, file->dir, call, &ftrace_event_format_fops); diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 2232ae3..5f5068e 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -666,6 +666,35 @@ static void set_work_pool_and_clear_pending(struct work_struct *work, */ smp_wmb(); set_work_data(work, (unsigned long)pool_id << WORK_OFFQ_POOL_SHIFT, 0); + /* + * The following mb guarantees that previous clear of a PENDING bit + * will not be reordered with any speculative LOADS or STORES from + * work->current_func, which is executed afterwards. This possible + * reordering can lead to a missed execution on attempt to qeueue + * the same @work. E.g. consider this case: + * + * CPU#0 CPU#1 + * ---------------------------- -------------------------------- + * + * 1 STORE event_indicated + * 2 queue_work_on() { + * 3 test_and_set_bit(PENDING) + * 4 } set_..._and_clear_pending() { + * 5 set_work_data() # clear bit + * 6 smp_mb() + * 7 work->current_func() { + * 8 LOAD event_indicated + * } + * + * Without an explicit full barrier speculative LOAD on line 8 can + * be executed before CPU#0 does STORE on line 1. If that happens, + * CPU#0 observes the PENDING bit is still set and new execution of + * a @work is not queued in a hope, that CPU#1 will eventually + * finish the queued @work. Meanwhile CPU#1 does not see + * event_indicated is set, because speculative LOAD was executed + * before actual STORE. + */ + smp_mb(); } static void clear_work_data(struct work_struct *work) @@ -4525,6 +4554,17 @@ static void rebind_workers(struct worker_pool *pool) pool->attrs->cpumask) < 0); spin_lock_irq(&pool->lock); + + /* + * XXX: CPU hotplug notifiers are weird and can call DOWN_FAILED + * w/o preceding DOWN_PREPARE. Work around it. CPU hotplug is + * being reworked and this can go away in time. + */ + if (!(pool->flags & POOL_DISASSOCIATED)) { + spin_unlock_irq(&pool->lock); + return; + } + pool->flags &= ~POOL_DISASSOCIATED; for_each_pool_worker(worker, pool) { |
