diff options
Diffstat (limited to 'kernel/cgroup')
| -rw-r--r-- | kernel/cgroup/cgroup-internal.h | 9 | ||||
| -rw-r--r-- | kernel/cgroup/cgroup-v1.c | 67 | ||||
| -rw-r--r-- | kernel/cgroup/cgroup.c | 376 | ||||
| -rw-r--r-- | kernel/cgroup/cpuset.c | 164 | ||||
| -rw-r--r-- | kernel/cgroup/freezer.c | 9 | ||||
| -rw-r--r-- | kernel/cgroup/pids.c | 11 | ||||
| -rw-r--r-- | kernel/cgroup/rstat.c | 48 |
7 files changed, 324 insertions, 360 deletions
diff --git a/kernel/cgroup/cgroup-internal.h b/kernel/cgroup/cgroup-internal.h index 809e34a3c017..bfbeabc17a9d 100644 --- a/kernel/cgroup/cgroup-internal.h +++ b/kernel/cgroup/cgroup-internal.h @@ -7,7 +7,7 @@ #include <linux/workqueue.h> #include <linux/list.h> #include <linux/refcount.h> -#include <linux/fs_context.h> +#include <linux/fs_parser.h> #define TRACE_CGROUP_PATH_LEN 1024 extern spinlock_t trace_cgroup_path_lock; @@ -231,9 +231,10 @@ int cgroup_migrate(struct task_struct *leader, bool threadgroup, int cgroup_attach_task(struct cgroup *dst_cgrp, struct task_struct *leader, bool threadgroup); -struct task_struct *cgroup_procs_write_start(char *buf, bool threadgroup) +struct task_struct *cgroup_procs_write_start(char *buf, bool threadgroup, + bool *locked) __acquires(&cgroup_threadgroup_rwsem); -void cgroup_procs_write_finish(struct task_struct *task) +void cgroup_procs_write_finish(struct task_struct *task, bool locked) __releases(&cgroup_threadgroup_rwsem); void cgroup_lock_and_drain_offline(struct cgroup *cgrp); @@ -264,7 +265,7 @@ extern const struct proc_ns_operations cgroupns_operations; */ extern struct cftype cgroup1_base_files[]; extern struct kernfs_syscall_ops cgroup1_kf_syscall_ops; -extern const struct fs_parameter_description cgroup1_fs_parameters; +extern const struct fs_parameter_spec cgroup1_fs_parameters[]; int proc_cgroupstats_show(struct seq_file *m, void *v); bool cgroup1_ssid_disabled(int ssid); diff --git a/kernel/cgroup/cgroup-v1.c b/kernel/cgroup/cgroup-v1.c index 88006be40ea3..be1a1c83cdd1 100644 --- a/kernel/cgroup/cgroup-v1.c +++ b/kernel/cgroup/cgroup-v1.c @@ -18,8 +18,6 @@ #include <trace/events/cgroup.h> -#define cg_invalf(fc, fmt, ...) invalf(fc, fmt, ## __VA_ARGS__) - /* * pidlists linger the following amount before being destroyed. The goal * is avoiding frequent destruction in the middle of consecutive read calls @@ -194,25 +192,6 @@ struct cgroup_pidlist { }; /* - * The following two functions "fix" the issue where there are more pids - * than kmalloc will give memory for; in such cases, we use vmalloc/vfree. - * TODO: replace with a kernel-wide solution to this problem - */ -#define PIDLIST_TOO_LARGE(c) ((c) * sizeof(pid_t) > (PAGE_SIZE * 2)) -static void *pidlist_allocate(int count) -{ - if (PIDLIST_TOO_LARGE(count)) - return vmalloc(array_size(count, sizeof(pid_t))); - else - return kmalloc_array(count, sizeof(pid_t), GFP_KERNEL); -} - -static void pidlist_free(void *p) -{ - kvfree(p); -} - -/* * Used to destroy all pidlists lingering waiting for destroy timer. None * should be left afterwards. */ @@ -244,7 +223,7 @@ static void cgroup_pidlist_destroy_work_fn(struct work_struct *work) */ if (!delayed_work_pending(dwork)) { list_del(&l->links); - pidlist_free(l->list); + kvfree(l->list); put_pid_ns(l->key.ns); tofree = l; } @@ -365,7 +344,7 @@ static int pidlist_array_load(struct cgroup *cgrp, enum cgroup_filetype type, * show up until sometime later on. */ length = cgroup_task_count(cgrp); - array = pidlist_allocate(length); + array = kvmalloc_array(length, sizeof(pid_t), GFP_KERNEL); if (!array) return -ENOMEM; /* now, populate the array */ @@ -390,12 +369,12 @@ static int pidlist_array_load(struct cgroup *cgrp, enum cgroup_filetype type, l = cgroup_pidlist_find_create(cgrp, type); if (!l) { - pidlist_free(array); + kvfree(array); return -ENOMEM; } /* store array, freeing old if necessary */ - pidlist_free(l->list); + kvfree(l->list); l->list = array; l->length = length; *lp = l; @@ -514,12 +493,13 @@ static ssize_t __cgroup1_procs_write(struct kernfs_open_file *of, struct task_struct *task; const struct cred *cred, *tcred; ssize_t ret; + bool locked; cgrp = cgroup_kn_lock_live(of->kn, false); if (!cgrp) return -ENODEV; - task = cgroup_procs_write_start(buf, threadgroup); + task = cgroup_procs_write_start(buf, threadgroup, &locked); ret = PTR_ERR_OR_ZERO(task); if (ret) goto out_unlock; @@ -541,7 +521,7 @@ static ssize_t __cgroup1_procs_write(struct kernfs_open_file *of, ret = cgroup_attach_task(cgrp, task, threadgroup); out_finish: - cgroup_procs_write_finish(task); + cgroup_procs_write_finish(task, locked); out_unlock: cgroup_kn_unlock(of->kn); @@ -905,7 +885,7 @@ enum cgroup1_param { Opt_xattr, }; -static const struct fs_parameter_spec cgroup1_param_specs[] = { +const struct fs_parameter_spec cgroup1_fs_parameters[] = { fsparam_flag ("all", Opt_all), fsparam_flag ("clone_children", Opt_clone_children), fsparam_flag ("cpuset_v2_mode", Opt_cpuset_v2_mode), @@ -917,11 +897,6 @@ static const struct fs_parameter_spec cgroup1_param_specs[] = { {} }; -const struct fs_parameter_description cgroup1_fs_parameters = { - .name = "cgroup1", - .specs = cgroup1_param_specs, -}; - int cgroup1_parse_param(struct fs_context *fc, struct fs_parameter *param) { struct cgroup_fs_context *ctx = cgroup_fc2context(fc); @@ -929,7 +904,7 @@ int cgroup1_parse_param(struct fs_context *fc, struct fs_parameter *param) struct fs_parse_result result; int opt, i; - opt = fs_parse(fc, &cgroup1_fs_parameters, param, &result); + opt = fs_parse(fc, cgroup1_fs_parameters, param, &result); if (opt == -ENOPARAM) { if (strcmp(param->key, "source") == 0) { fc->source = param->string; @@ -942,7 +917,7 @@ int cgroup1_parse_param(struct fs_context *fc, struct fs_parameter *param) ctx->subsys_mask |= (1 << i); return 0; } - return cg_invalf(fc, "cgroup1: Unknown subsys name '%s'", param->key); + return invalfc(fc, "Unknown subsys name '%s'", param->key); } if (opt < 0) return opt; @@ -970,7 +945,7 @@ int cgroup1_parse_param(struct fs_context *fc, struct fs_parameter *param) case Opt_release_agent: /* Specifying two release agents is forbidden */ if (ctx->release_agent) - return cg_invalf(fc, "cgroup1: release_agent respecified"); + return invalfc(fc, "release_agent respecified"); ctx->release_agent = param->string; param->string = NULL; break; @@ -980,9 +955,9 @@ int cgroup1_parse_param(struct fs_context *fc, struct fs_parameter *param) return -ENOENT; /* Can't specify an empty name */ if (!param->size) - return cg_invalf(fc, "cgroup1: Empty name"); + return invalfc(fc, "Empty name"); if (param->size > MAX_CGROUP_ROOT_NAMELEN - 1) - return cg_invalf(fc, "cgroup1: Name too long"); + return invalfc(fc, "Name too long"); /* Must match [\w.-]+ */ for (i = 0; i < param->size; i++) { char c = param->string[i]; @@ -990,11 +965,11 @@ int cgroup1_parse_param(struct fs_context *fc, struct fs_parameter *param) continue; if ((c == '.') || (c == '-') || (c == '_')) continue; - return cg_invalf(fc, "cgroup1: Invalid name"); + return invalfc(fc, "Invalid name"); } /* Specifying two names is forbidden */ if (ctx->name) - return cg_invalf(fc, "cgroup1: name respecified"); + return invalfc(fc, "name respecified"); ctx->name = param->string; param->string = NULL; break; @@ -1029,7 +1004,7 @@ static int check_cgroupfs_options(struct fs_context *fc) if (ctx->all_ss) { /* Mutually exclusive option 'all' + subsystem name */ if (ctx->subsys_mask) - return cg_invalf(fc, "cgroup1: subsys name conflicts with all"); + return invalfc(fc, "subsys name conflicts with all"); /* 'all' => select all the subsystems */ ctx->subsys_mask = enabled; } @@ -1039,7 +1014,7 @@ static int check_cgroupfs_options(struct fs_context *fc) * empty hierarchies must have a name). */ if (!ctx->subsys_mask && !ctx->name) - return cg_invalf(fc, "cgroup1: Need name or subsystem set"); + return invalfc(fc, "Need name or subsystem set"); /* * Option noprefix was introduced just for backward compatibility @@ -1047,11 +1022,11 @@ static int check_cgroupfs_options(struct fs_context *fc) * the cpuset subsystem. */ if ((ctx->flags & CGRP_ROOT_NOPREFIX) && (ctx->subsys_mask & mask)) - return cg_invalf(fc, "cgroup1: noprefix used incorrectly"); + return invalfc(fc, "noprefix used incorrectly"); /* Can't specify "none" and some subsystems */ if (ctx->subsys_mask && ctx->none) - return cg_invalf(fc, "cgroup1: none used incorrectly"); + return invalfc(fc, "none used incorrectly"); return 0; } @@ -1081,7 +1056,7 @@ int cgroup1_reconfigure(struct fs_context *fc) /* Don't allow flags or name to change at remount */ if ((ctx->flags ^ root->flags) || (ctx->name && strcmp(ctx->name, root->name))) { - cg_invalf(fc, "option or name mismatch, new: 0x%x \"%s\", old: 0x%x \"%s\"", + errorfc(fc, "option or name mismatch, new: 0x%x \"%s\", old: 0x%x \"%s\"", ctx->flags, ctx->name ?: "", root->flags, root->name); ret = -EINVAL; goto out_unlock; @@ -1198,7 +1173,7 @@ static int cgroup1_root_to_use(struct fs_context *fc) * can't create new one without subsys specification. */ if (!ctx->subsys_mask && !ctx->none) - return cg_invalf(fc, "cgroup1: No subsys list or none specified"); + return invalfc(fc, "No subsys list or none specified"); /* Hierarchies may only be created in the initial cgroup namespace. */ if (ctx->ns != &init_cgroup_ns) diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index 753afbca549f..75f687301bbf 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -488,7 +488,7 @@ static struct cgroup_subsys_state *cgroup_tryget_css(struct cgroup *cgrp, rcu_read_lock(); css = cgroup_css(cgrp, ss); - if (!css || !css_tryget_online(css)) + if (css && !css_tryget_online(css)) css = NULL; rcu_read_unlock(); @@ -899,8 +899,7 @@ static void css_set_move_task(struct task_struct *task, /* * We are synchronized through cgroup_threadgroup_rwsem * against PF_EXITING setting such that we can't race - * against cgroup_exit() changing the css_set to - * init_css_set and dropping the old one. + * against cgroup_exit()/cgroup_free() dropping the css_set. */ WARN_ON_ONCE(task->flags & PF_EXITING); @@ -1309,10 +1308,7 @@ static void cgroup_exit_root_id(struct cgroup_root *root) void cgroup_free_root(struct cgroup_root *root) { - if (root) { - idr_destroy(&root->cgroup_idr); - kfree(root); - } + kfree(root); } static void cgroup_destroy_root(struct cgroup_root *root) @@ -1374,6 +1370,8 @@ current_cgns_cgroup_from_root(struct cgroup_root *root) cset = current->nsproxy->cgroup_ns->root_cset; if (cset == &init_css_set) { res = &root->cgrp; + } else if (root == &cgrp_dfl_root) { + res = cset->dfl_cgrp; } else { struct cgrp_cset_link *link; @@ -1430,9 +1428,8 @@ struct cgroup *task_cgroup_from_root(struct task_struct *task, struct cgroup_root *root) { /* - * No need to lock the task - since we hold cgroup_mutex the - * task can't change groups, so the only thing that can happen - * is that it exits and its css is set back to init_css_set. + * No need to lock the task - since we hold css_set_lock the + * task can't change groups. */ return cset_cgroup_from_root(task_css_set(task), root); } @@ -1819,24 +1816,19 @@ enum cgroup2_param { nr__cgroup2_params }; -static const struct fs_parameter_spec cgroup2_param_specs[] = { +static const struct fs_parameter_spec cgroup2_fs_parameters[] = { fsparam_flag("nsdelegate", Opt_nsdelegate), fsparam_flag("memory_localevents", Opt_memory_localevents), {} }; -static const struct fs_parameter_description cgroup2_fs_parameters = { - .name = "cgroup2", - .specs = cgroup2_param_specs, -}; - static int cgroup2_parse_param(struct fs_context *fc, struct fs_parameter *param) { struct cgroup_fs_context *ctx = cgroup_fc2context(fc); struct fs_parse_result result; int opt; - opt = fs_parse(fc, &cgroup2_fs_parameters, param, &result); + opt = fs_parse(fc, cgroup2_fs_parameters, param, &result); if (opt < 0) return opt; @@ -1883,65 +1875,6 @@ static int cgroup_reconfigure(struct fs_context *fc) return 0; } -/* - * To reduce the fork() overhead for systems that are not actually using - * their cgroups capability, we don't maintain the lists running through - * each css_set to its tasks until we see the list actually used - in other - * words after the first mount. - */ -static bool use_task_css_set_links __read_mostly; - -static void cgroup_enable_task_cg_lists(void) -{ - struct task_struct *p, *g; - - /* - * We need tasklist_lock because RCU is not safe against - * while_each_thread(). Besides, a forking task that has passed - * cgroup_post_fork() without seeing use_task_css_set_links = 1 - * is not guaranteed to have its child immediately visible in the - * tasklist if we walk through it with RCU. - */ - read_lock(&tasklist_lock); - spin_lock_irq(&css_set_lock); - - if (use_task_css_set_links) - goto out_unlock; - - use_task_css_set_links = true; - - do_each_thread(g, p) { - WARN_ON_ONCE(!list_empty(&p->cg_list) || - task_css_set(p) != &init_css_set); - - /* - * We should check if the process is exiting, otherwise - * it will race with cgroup_exit() in that the list - * entry won't be deleted though the process has exited. - * Do it while holding siglock so that we don't end up - * racing against cgroup_exit(). - * - * Interrupts were already disabled while acquiring - * the css_set_lock, so we do not need to disable it - * again when acquiring the sighand->siglock here. - */ - spin_lock(&p->sighand->siglock); - if (!(p->flags & PF_EXITING)) { - struct css_set *cset = task_css_set(p); - - if (!css_set_populated(cset)) - css_set_update_populated(cset, true); - list_add_tail(&p->cg_list, &cset->tasks); - get_css_set(cset); - cset->nr_tasks++; - } - spin_unlock(&p->sighand->siglock); - } while_each_thread(g, p); -out_unlock: - spin_unlock_irq(&css_set_lock); - read_unlock(&tasklist_lock); -} - static void init_cgroup_housekeeping(struct cgroup *cgrp) { struct cgroup_subsys *ss; @@ -1976,7 +1909,6 @@ void init_cgroup_root(struct cgroup_fs_context *ctx) atomic_set(&root->nr_cgrps, 1); cgrp->root = root; init_cgroup_housekeeping(cgrp); - idr_init(&root->cgroup_idr); root->flags = ctx->flags; if (ctx->release_agent) @@ -1997,12 +1929,6 @@ int cgroup_setup_root(struct cgroup_root *root, u16 ss_mask) lockdep_assert_held(&cgroup_mutex); - ret = cgroup_idr_alloc(&root->cgroup_idr, root_cgrp, 1, 2, GFP_KERNEL); - if (ret < 0) - goto out; - root_cgrp->id = ret; - root_cgrp->ancestor_ids[0] = ret; - ret = percpu_ref_init(&root_cgrp->self.refcnt, css_release, 0, GFP_KERNEL); if (ret) @@ -2035,6 +1961,8 @@ int cgroup_setup_root(struct cgroup_root *root, u16 ss_mask) goto exit_root_id; } root_cgrp->kn = root->kf_root->kn; + WARN_ON_ONCE(cgroup_ino(root_cgrp) != 1); + root_cgrp->ancestor_ids[0] = cgroup_id(root_cgrp); ret = css_populate_dir(&root_cgrp->self); if (ret) @@ -2119,11 +2047,12 @@ int cgroup_do_get_tree(struct fs_context *fc) nsdentry = kernfs_node_dentry(cgrp->kn, sb); dput(fc->root); - fc->root = nsdentry; if (IS_ERR(nsdentry)) { - ret = PTR_ERR(nsdentry); deactivate_locked_super(sb); + ret = PTR_ERR(nsdentry); + nsdentry = NULL; } + fc->root = nsdentry; } if (!ctx->kfc.new_sb_created) @@ -2187,13 +2116,6 @@ static int cgroup_init_fs_context(struct fs_context *fc) if (!ctx) return -ENOMEM; - /* - * The first time anyone tries to mount a cgroup, enable the list - * linking each css_set to its tasks and fix up all existing tasks. - */ - if (!use_task_css_set_links) - cgroup_enable_task_cg_lists(); - ctx->ns = current->nsproxy->cgroup_ns; get_cgroup_ns(ctx->ns); fc->fs_private = &ctx->kfc; @@ -2229,7 +2151,7 @@ static void cgroup_kill_sb(struct super_block *sb) struct file_system_type cgroup_fs_type = { .name = "cgroup", .init_fs_context = cgroup_init_fs_context, - .parameters = &cgroup1_fs_parameters, + .parameters = cgroup1_fs_parameters, .kill_sb = cgroup_kill_sb, .fs_flags = FS_USERNS_MOUNT, }; @@ -2237,7 +2159,7 @@ struct file_system_type cgroup_fs_type = { static struct file_system_type cgroup2_fs_type = { .name = "cgroup2", .init_fs_context = cgroup_init_fs_context, - .parameters = &cgroup2_fs_parameters, + .parameters = cgroup2_fs_parameters, .kill_sb = cgroup_kill_sb, .fs_flags = FS_USERNS_MOUNT, }; @@ -2371,9 +2293,8 @@ static void cgroup_migrate_add_task(struct task_struct *task, if (task->flags & PF_EXITING) return; - /* leave @task alone if post_fork() hasn't linked it yet */ - if (list_empty(&task->cg_list)) - return; + /* cgroup_threadgroup_rwsem protects racing against forks */ + WARN_ON_ONCE(list_empty(&task->cg_list)); cset = task_css_set(task); if (!cset->mg_src_cgrp) @@ -2824,7 +2745,8 @@ int cgroup_attach_task(struct cgroup *dst_cgrp, struct task_struct *leader, return ret; } -struct task_struct *cgroup_procs_write_start(char *buf, bool threadgroup) +struct task_struct *cgroup_procs_write_start(char *buf, bool threadgroup, + bool *locked) __acquires(&cgroup_threadgroup_rwsem) { struct task_struct *tsk; @@ -2833,7 +2755,21 @@ struct task_struct *cgroup_procs_write_start(char *buf, bool threadgroup) if (kstrtoint(strstrip(buf), 0, &pid) || pid < 0) return ERR_PTR(-EINVAL); - percpu_down_write(&cgroup_threadgroup_rwsem); + /* + * If we migrate a single thread, we don't care about threadgroup + * stability. If the thread is `current`, it won't exit(2) under our + * hands or change PID through exec(2). We exclude + * cgroup_update_dfl_csses and other cgroup_{proc,thread}s_write + * callers by cgroup_mutex. + * Therefore, we can skip the global lock. + */ + lockdep_assert_held(&cgroup_mutex); + if (pid || threadgroup) { + percpu_down_write(&cgroup_threadgroup_rwsem); + *locked = true; + } else { + *locked = false; + } rcu_read_lock(); if (pid) { @@ -2864,13 +2800,16 @@ struct task_struct *cgroup_procs_write_start(char *buf, bool threadgroup) goto out_unlock_rcu; out_unlock_threadgroup: - percpu_up_write(&cgroup_threadgroup_rwsem); + if (*locked) { + percpu_up_write(&cgroup_threadgroup_rwsem); + *locked = false; + } out_unlock_rcu: rcu_read_unlock(); return tsk; } -void cgroup_procs_write_finish(struct task_struct *task) +void cgroup_procs_write_finish(struct task_struct *task, bool locked) __releases(&cgroup_threadgroup_rwsem) { struct cgroup_subsys *ss; @@ -2879,7 +2818,8 @@ void cgroup_procs_write_finish(struct task_struct *task) /* release reference from cgroup_procs_write_start() */ put_task_struct(task); - percpu_up_write(&cgroup_threadgroup_rwsem); + if (locked) + percpu_up_write(&cgroup_threadgroup_rwsem); for_each_subsys(ss, ssid) if (ss->post_attach) ss->post_attach(); @@ -2894,7 +2834,7 @@ static void cgroup_print_ss_mask(struct seq_file *seq, u16 ss_mask) do_each_subsys_mask(ss, ssid, ss_mask) { if (printed) seq_putc(seq, ' '); - seq_printf(seq, "%s", ss->name); + seq_puts(seq, ss->name); printed = true; } while_each_subsys_mask(); if (printed) @@ -3110,8 +3050,6 @@ static int cgroup_apply_control_enable(struct cgroup *cgrp) for_each_subsys(ss, ssid) { struct cgroup_subsys_state *css = cgroup_css(dsct, ss); - WARN_ON_ONCE(css && percpu_ref_is_dying(&css->refcnt)); - if (!(cgroup_ss_mask(dsct) & (1 << ss->id))) continue; @@ -3121,6 +3059,8 @@ static int cgroup_apply_control_enable(struct cgroup *cgrp) return PTR_ERR(css); } + WARN_ON_ONCE(percpu_ref_is_dying(&css->refcnt)); + if (css_visible(css)) { ret = css_populate_dir(css); if (ret) @@ -3156,11 +3096,11 @@ static void cgroup_apply_control_disable(struct cgroup *cgrp) for_each_subsys(ss, ssid) { struct cgroup_subsys_state *css = cgroup_css(dsct, ss); - WARN_ON_ONCE(css && percpu_ref_is_dying(&css->refcnt)); - if (!css) continue; + WARN_ON_ONCE(percpu_ref_is_dying(&css->refcnt)); + if (css->parent && !(cgroup_ss_mask(dsct) & (1 << ss->id))) { kill_css(css); @@ -3447,7 +3387,8 @@ static ssize_t cgroup_type_write(struct kernfs_open_file *of, char *buf, if (strcmp(strstrip(buf), "threaded")) return -EINVAL; - cgrp = cgroup_kn_lock_live(of->kn, false); + /* drain dying csses before we re-apply (threaded) subtree control */ + cgrp = cgroup_kn_lock_live(of->kn, true); if (!cgrp) return -ENOENT; @@ -3600,22 +3541,22 @@ static int cpu_stat_show(struct seq_file *seq, void *v) #ifdef CONFIG_PSI static int cgroup_io_pressure_show(struct seq_file *seq, void *v) { - struct cgroup *cgroup = seq_css(seq)->cgroup; - struct psi_group *psi = cgroup->id == 1 ? &psi_system : &cgroup->psi; + struct cgroup *cgrp = seq_css(seq)->cgroup; + struct psi_group *psi = cgroup_id(cgrp) == 1 ? &psi_system : &cgrp->psi; return psi_show(seq, psi, PSI_IO); } static int cgroup_memory_pressure_show(struct seq_file *seq, void *v) { - struct cgroup *cgroup = seq_css(seq)->cgroup; - struct psi_group *psi = cgroup->id == 1 ? &psi_system : &cgroup->psi; + struct cgroup *cgrp = seq_css(seq)->cgroup; + struct psi_group *psi = cgroup_id(cgrp) == 1 ? &psi_system : &cgrp->psi; return psi_show(seq, psi, PSI_MEM); } static int cgroup_cpu_pressure_show(struct seq_file *seq, void *v) { - struct cgroup *cgroup = seq_css(seq)->cgroup; - struct psi_group *psi = cgroup->id == 1 ? &psi_system : &cgroup->psi; + struct cgroup *cgrp = seq_css(seq)->cgroup; + struct psi_group *psi = cgroup_id(cgrp) == 1 ? &psi_system : &cgrp->psi; return psi_show(seq, psi, PSI_CPU); } @@ -4567,9 +4508,6 @@ repeat: void css_task_iter_start(struct cgroup_subsys_state *css, unsigned int flags, struct css_task_iter *it) { - /* no one should try to iterate before mounting cgroups */ - WARN_ON_ONCE(!use_task_css_set_links); - memset(it, 0, sizeof(*it)); spin_lock_irq(&css_set_lock); @@ -4754,12 +4692,13 @@ static ssize_t cgroup_procs_write(struct kernfs_open_file *of, struct cgroup *src_cgrp, *dst_cgrp; struct task_struct *task; ssize_t ret; + bool locked; dst_cgrp = cgroup_kn_lock_live(of->kn, false); if (!dst_cgrp) return -ENODEV; - task = cgroup_procs_write_start(buf, true); + task = cgroup_procs_write_start(buf, true, &locked); ret = PTR_ERR_OR_ZERO(task); if (ret) goto out_unlock; @@ -4777,7 +4716,7 @@ static ssize_t cgroup_procs_write(struct kernfs_open_file *of, ret = cgroup_attach_task(dst_cgrp, task, true); out_finish: - cgroup_procs_write_finish(task); + cgroup_procs_write_finish(task, locked); out_unlock: cgroup_kn_unlock(of->kn); @@ -4795,6 +4734,7 @@ static ssize_t cgroup_threads_write(struct kernfs_open_file *of, struct cgroup *src_cgrp, *dst_cgrp; struct task_struct *task; ssize_t ret; + bool locked; buf = strstrip(buf); @@ -4802,7 +4742,7 @@ static ssize_t cgroup_threads_write(struct kernfs_open_file *of, if (!dst_cgrp) return -ENODEV; - task = cgroup_procs_write_start(buf, false); + task = cgroup_procs_write_start(buf, false, &locked); ret = PTR_ERR_OR_ZERO(task); if (ret) goto out_unlock; @@ -4826,7 +4766,7 @@ static ssize_t cgroup_threads_write(struct kernfs_open_file *of, ret = cgroup_attach_task(dst_cgrp, task, false); out_finish: - cgroup_procs_write_finish(task); + cgroup_procs_write_finish(task, locked); out_unlock: cgroup_kn_unlock(of->kn); @@ -5036,9 +4976,6 @@ static void css_release_work_fn(struct work_struct *work) tcgrp->nr_dying_descendants--; spin_unlock_irq(&css_set_lock); - cgroup_idr_remove(&cgrp->root->cgroup_idr, cgrp->id); - cgrp->id = -1; - /* * There are two control paths which try to determine * cgroup from dentry without going through kernfs - @@ -5203,10 +5140,12 @@ err_free_css: * it isn't associated with its kernfs_node and doesn't have the control * mask applied. */ -static struct cgroup *cgroup_create(struct cgroup *parent) +static struct cgroup *cgroup_create(struct cgroup *parent, const char *name, + umode_t mode) { struct cgroup_root *root = parent->root; struct cgroup *cgrp, *tcgrp; + struct kernfs_node *kn; int level = parent->level + 1; int ret; @@ -5226,15 +5165,13 @@ static struct cgroup *cgroup_create(struct cgroup *parent) goto out_cancel_ref; } - /* - * Temporarily set the pointer to NULL, so idr_find() won't return - * a half-baked cgroup. - */ - cgrp->id = cgroup_idr_alloc(&root->cgroup_idr, NULL, 2, 0, GFP_KERNEL); - if (cgrp->id < 0) { - ret = -ENOMEM; + /* create the directory */ + kn = kernfs_create_dir(parent->kn, name, mode, cgrp); + if (IS_ERR(kn)) { + ret = PTR_ERR(kn); goto out_stat_exit; } + cgrp->kn = kn; init_cgroup_housekeeping(cgrp); @@ -5244,7 +5181,7 @@ static struct cgroup *cgroup_create(struct cgroup *parent) ret = psi_cgroup_alloc(cgrp); if (ret) - goto out_idr_free; + goto out_kernfs_remove; ret = cgroup_bpf_inherit(cgrp); if (ret) @@ -5255,12 +5192,20 @@ static struct cgroup *cgroup_create(struct cgroup *parent) * if the parent has to be frozen, the child has too. */ cgrp->freezer.e_freeze = parent->freezer.e_freeze; - if (cgrp->freezer.e_freeze) + if (cgrp->freezer.e_freeze) { + /* + * Set the CGRP_FREEZE flag, so when a process will be + * attached to the child cgroup, it will become frozen. + * At this point the new cgroup is unpopulated, so we can + * consider it frozen immediately. + */ + set_bit(CGRP_FREEZE, &cgrp->flags); set_bit(CGRP_FROZEN, &cgrp->flags); + } spin_lock_irq(&css_set_lock); for (tcgrp = cgrp; tcgrp; tcgrp = cgroup_parent(tcgrp)) { - cgrp->ancestor_ids[tcgrp->level] = tcgrp->id; + cgrp->ancestor_ids[tcgrp->level] = cgroup_id(tcgrp); if (tcgrp != cgrp) { tcgrp->nr_descendants++; @@ -5290,12 +5235,6 @@ static struct cgroup *cgroup_create(struct cgroup *parent) cgroup_get_live(parent); /* - * @cgrp is now fully operational. If something fails after this - * point, it'll be released via the normal destruction path. - */ - cgroup_idr_replace(&root->cgroup_idr, cgrp, cgrp->id); - - /* * On the default hierarchy, a child doesn't automatically inherit * subtree_control from the parent. Each is configured manually. */ @@ -5308,8 +5247,8 @@ static struct cgroup *cgroup_create(struct cgroup *parent) out_psi_free: psi_cgroup_free(cgrp); -out_idr_free: - cgroup_idr_remove(&root->cgroup_idr, cgrp->id); +out_kernfs_remove: + kernfs_remove(cgrp->kn); out_stat_exit: if (cgroup_on_dfl(parent)) cgroup_rstat_exit(cgrp); @@ -5346,7 +5285,6 @@ fail: int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, umode_t mode) { struct cgroup *parent, *cgrp; - struct kernfs_node *kn; int ret; /* do not accept '\n' to prevent making /proc/<pid>/cgroup unparsable */ @@ -5362,27 +5300,19 @@ int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, umode_t mode) goto out_unlock; } - cgrp = cgroup_create(parent); + cgrp = cgroup_create(parent, name, mode); if (IS_ERR(cgrp)) { ret = PTR_ERR(cgrp); goto out_unlock; } - /* create the directory */ - kn = kernfs_create_dir(parent->kn, name, mode, cgrp); - if (IS_ERR(kn)) { - ret = PTR_ERR(kn); - goto out_destroy; - } - cgrp->kn = kn; - /* * This extra ref will be put in cgroup_free_fn() and guarantees * that @cgrp->kn is always accessible. */ - kernfs_get(kn); + kernfs_get(cgrp->kn); - ret = cgroup_kn_set_ugid(kn); + ret = cgroup_kn_set_ugid(cgrp->kn); if (ret) goto out_destroy; @@ -5397,7 +5327,7 @@ int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, umode_t mode) TRACE_CGROUP_PATH(mkdir, cgrp); /* let's create and online css's */ - kernfs_activate(kn); + kernfs_activate(cgrp->kn); ret = 0; goto out_unlock; @@ -5827,12 +5757,11 @@ static int __init cgroup_wq_init(void) } core_initcall(cgroup_wq_init); -void cgroup_path_from_kernfs_id(const union kernfs_node_id *id, - char *buf, size_t buflen) +void cgroup_path_from_kernfs_id(u64 id, char *buf, size_t buflen) { struct kernfs_node *kn; - kn = kernfs_get_node_by_id(cgrp_dfl_root.kf_root, id); + kn = kernfs_find_and_get_node_by_id(cgrp_dfl_root.kf_root, id); if (!kn) return; kernfs_path(kn, buf, buflen); @@ -5993,62 +5922,41 @@ void cgroup_cancel_fork(struct task_struct *child) void cgroup_post_fork(struct task_struct *child) { struct cgroup_subsys *ss; + struct css_set *cset; int i; + spin_lock_irq(&css_set_lock); + + /* init tasks are special, only link regular threads */ + if (likely(child->pid)) { + WARN_ON_ONCE(!list_empty(&child->cg_list)); + cset = task_css_set(current); /* current is @child's parent */ + get_css_set(cset); + cset->nr_tasks++; + css_set_move_task(child, NULL, cset, false); + } + /* - * This may race against cgroup_enable_task_cg_lists(). As that - * function sets use_task_css_set_links before grabbing - * tasklist_lock and we just went through tasklist_lock to add - * @child, it's guaranteed that either we see the set - * use_task_css_set_links or cgroup_enable_task_cg_lists() sees - * @child during its iteration. - * - * If we won the race, @child is associated with %current's - * css_set. Grabbing css_set_lock guarantees both that the - * association is stable, and, on completion of the parent's - * migration, @child is visible in the source of migration or - * already in the destination cgroup. This guarantee is necessary - * when implementing operations which need to migrate all tasks of - * a cgroup to another. - * - * Note that if we lose to cgroup_enable_task_cg_lists(), @child - * will remain in init_css_set. This is safe because all tasks are - * in the init_css_set before cg_links is enabled and there's no - * operation which transfers all tasks out of init_css_set. + * If the cgroup has to be frozen, the new task has too. Let's set + * the JOBCTL_TRAP_FREEZE jobctl bit to get the task into the + * frozen state. */ - if (use_task_css_set_links) { - struct css_set *cset; - - spin_lock_irq(&css_set_lock); - cset = task_css_set(current); - if (list_empty(&child->cg_list)) { - get_css_set(cset); - cset->nr_tasks++; - css_set_move_task(child, NULL, cset, false); - } + if (unlikely(cgroup_task_freeze(child))) { + spin_lock(&child->sighand->siglock); + WARN_ON_ONCE(child->frozen); + child->jobctl |= JOBCTL_TRAP_FREEZE; + spin_unlock(&child->sighand->siglock); /* - * If the cgroup has to be frozen, the new task has too. - * Let's set the JOBCTL_TRAP_FREEZE jobctl bit to get - * the task into the frozen state. + * Calling cgroup_update_frozen() isn't required here, + * because it will be called anyway a bit later from + * do_freezer_trap(). So we avoid cgroup's transient switch + * from the frozen state and back. */ - if (unlikely(cgroup_task_freeze(child))) { - spin_lock(&child->sighand->siglock); - WARN_ON_ONCE(child->frozen); - child->jobctl |= JOBCTL_TRAP_FREEZE; - spin_unlock(&child->sighand->siglock); - - /* - * Calling cgroup_update_frozen() isn't required here, - * because it will be called anyway a bit later - * from do_freezer_trap(). So we avoid cgroup's - * transient switch from the frozen state and back. - */ - } - - spin_unlock_irq(&css_set_lock); } + spin_unlock_irq(&css_set_lock); + /* * Call ss->fork(). This must happen after @child is linked on * css_set; otherwise, @child might change state between ->fork() @@ -6063,20 +5971,8 @@ void cgroup_post_fork(struct task_struct *child) * cgroup_exit - detach cgroup from exiting task * @tsk: pointer to task_struct of exiting process * - * Description: Detach cgroup from @tsk and release it. + * Description: Detach cgroup from @tsk. * - * Note that cgroups marked notify_on_release force every task in - * them to take the global cgroup_mutex mutex when exiting. - * This could impact scaling on very large systems. Be reluctant to - * use notify_on_release cgroups where very high task exit scaling - * is required on large systems. - * - * We set the exiting tasks cgroup to the root cgroup (top_cgroup). We - * call cgroup_exit() while the task is still competent to handle - * notify_on_release(), then leave the task attached to the root cgroup in - * each hierarchy for the remainder of its exit. No need to bother with - * init_css_set refcnting. init_css_set never goes away and we can't race - * with migration path - PF_EXITING is visible to migration path. */ void cgroup_exit(struct task_struct *tsk) { @@ -6084,26 +5980,19 @@ void cgroup_exit(struct task_struct *tsk) struct css_set *cset; int i; - /* - * Unlink from @tsk from its css_set. As migration path can't race - * with us, we can check css_set and cg_list without synchronization. - */ - cset = task_css_set(tsk); + spin_lock_irq(&css_set_lock); - if (!list_empty(&tsk->cg_list)) { - spin_lock_irq(&css_set_lock); - css_set_move_task(tsk, cset, NULL, false); - list_add_tail(&tsk->cg_list, &cset->dying_tasks); - cset->nr_tasks--; + WARN_ON_ONCE(list_empty(&tsk->cg_list)); + cset = task_css_set(tsk); + css_set_move_task(tsk, cset, NULL, false); + list_add_tail(&tsk->cg_list, &cset->dying_tasks); + cset->nr_tasks--; - WARN_ON_ONCE(cgroup_task_frozen(tsk)); - if (unlikely(cgroup_task_freeze(tsk))) - cgroup_update_frozen(task_dfl_cgroup(tsk)); + WARN_ON_ONCE(cgroup_task_frozen(tsk)); + if (unlikely(cgroup_task_freeze(tsk))) + cgroup_update_frozen(task_dfl_cgroup(tsk)); - spin_unlock_irq(&css_set_lock); - } else { - get_css_set(cset); - } + spin_unlock_irq(&css_set_lock); /* see cgroup_post_fork() for details */ do_each_subsys_mask(ss, i, have_exit_callback) { @@ -6120,12 +6009,10 @@ void cgroup_release(struct task_struct *task) ss->release(task); } while_each_subsys_mask(); - if (use_task_css_set_links) { - spin_lock_irq(&css_set_lock); - css_set_skip_task_iters(task_css_set(task), task); - list_del_init(&task->cg_list); - spin_unlock_irq(&css_set_lock); - } + spin_lock_irq(&css_set_lock); + css_set_skip_task_iters(task_css_set(task), task); + list_del_init(&task->cg_list); + spin_unlock_irq(&css_set_lock); } void cgroup_free(struct task_struct *task) @@ -6400,12 +6287,13 @@ void cgroup_sk_free(struct sock_cgroup_data *skcd) #ifdef CONFIG_CGROUP_BPF int cgroup_bpf_attach(struct cgroup *cgrp, struct bpf_prog *prog, - enum bpf_attach_type type, u32 flags) + struct bpf_prog *replace_prog, enum bpf_attach_type type, + u32 flags) { int ret; mutex_lock(&cgroup_mutex); - ret = __cgroup_bpf_attach(cgrp, prog, type, flags); + ret = __cgroup_bpf_attach(cgrp, prog, replace_prog, type, flags); mutex_unlock(&cgroup_mutex); return ret; } diff --git a/kernel/cgroup/cpuset.c b/kernel/cgroup/cpuset.c index 5aa37531ce76..58f5073acff7 100644 --- a/kernel/cgroup/cpuset.c +++ b/kernel/cgroup/cpuset.c @@ -45,6 +45,7 @@ #include <linux/proc_fs.h> #include <linux/rcupdate.h> #include <linux/sched.h> +#include <linux/sched/deadline.h> #include <linux/sched/mm.h> #include <linux/sched/task.h> #include <linux/seq_file.h> @@ -332,7 +333,18 @@ static struct cpuset top_cpuset = { * guidelines for accessing subsystem state in kernel/cgroup.c */ -static DEFINE_MUTEX(cpuset_mutex); +DEFINE_STATIC_PERCPU_RWSEM(cpuset_rwsem); + +void cpuset_read_lock(void) +{ + percpu_down_read(&cpuset_rwsem); +} + +void cpuset_read_unlock(void) +{ + percpu_up_read(&cpuset_rwsem); +} + static DEFINE_SPINLOCK(callback_lock); static struct workqueue_struct *cpuset_migrate_mm_wq; @@ -786,7 +798,8 @@ static int generate_sched_domains(cpumask_var_t **domains, cpumask_subset(cp->cpus_allowed, top_cpuset.effective_cpus)) continue; - if (is_sched_load_balance(cp)) + if (is_sched_load_balance(cp) && + !cpumask_empty(cp->effective_cpus)) csa[csn++] = cp; /* skip @cp's subtree if not a partition root */ @@ -894,6 +907,65 @@ done: return ndoms; } +static void update_tasks_root_domain(struct cpuset *cs) +{ + struct css_task_iter it; + struct task_struct *task; + + css_task_iter_start(&cs->css, 0, &it); + + while ((task = css_task_iter_next(&it))) + dl_add_task_root_domain(task); + + css_task_iter_end(&it); +} + +static void rebuild_root_domains(void) +{ + struct cpuset *cs = NULL; + struct cgroup_subsys_state *pos_css; + + percpu_rwsem_assert_held(&cpuset_rwsem); + lockdep_assert_cpus_held(); + lockdep_assert_held(&sched_domains_mutex); + + rcu_read_lock(); + + /* + * Clear default root domain DL accounting, it will be computed again + * if a task belongs to it. + */ + dl_clear_root_domain(&def_root_domain); + + cpuset_for_each_descendant_pre(cs, pos_css, &top_cpuset) { + + if (cpumask_empty(cs->effective_cpus)) { + pos_css = css_rightmost_descendant(pos_css); + continue; + } + + css_get(&cs->css); + + rcu_read_unlock(); + + update_tasks_root_domain(cs); + + rcu_read_lock(); + css_put(&cs->css); + } + rcu_read_unlock(); +} + +static void +partition_and_rebuild_sched_domains(int ndoms_new, cpumask_var_t doms_new[], + struct sched_domain_attr *dattr_new) +{ + mutex_lock(&sched_domains_mutex); + partition_sched_domains_locked(ndoms_new, doms_new, dattr_new); + rebuild_root_domains(); + mutex_unlock(&sched_domains_mutex); +} + /* * Rebuild scheduler domains. * @@ -911,8 +983,8 @@ static void rebuild_sched_domains_locked(void) cpumask_var_t *doms; int ndoms; - lockdep_assert_held(&cpuset_mutex); - get_online_cpus(); + lockdep_assert_cpus_held(); + percpu_rwsem_assert_held(&cpuset_rwsem); /* * We have raced with CPU hotplug. Don't do anything to avoid @@ -921,19 +993,17 @@ static void rebuild_sched_domains_locked(void) */ if (!top_cpuset.nr_subparts_cpus && !cpumask_equal(top_cpuset.effective_cpus, cpu_active_mask)) - goto out; + return; if (top_cpuset.nr_subparts_cpus && !cpumask_subset(top_cpuset.effective_cpus, cpu_active_mask)) - goto out; + return; /* Generate domain masks and attrs */ ndoms = generate_sched_domains(&doms, &attr); /* Have scheduler rebuild the domains */ - partition_sched_domains(ndoms, doms, attr); -out: - put_online_cpus(); + partition_and_rebuild_sched_domains(ndoms, doms, attr); } #else /* !CONFIG_SMP */ static void rebuild_sched_domains_locked(void) @@ -943,9 +1013,11 @@ static void rebuild_sched_domains_locked(void) void rebuild_sched_domains(void) { - mutex_lock(&cpuset_mutex); + get_online_cpus(); + percpu_down_write(&cpuset_rwsem); rebuild_sched_domains_locked(); - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); + put_online_cpus(); } /** @@ -1051,7 +1123,7 @@ static int update_parent_subparts_cpumask(struct cpuset *cpuset, int cmd, int deleting; /* Moving cpus from subparts_cpus to effective_cpus */ bool part_error = false; /* Partition error? */ - lockdep_assert_held(&cpuset_mutex); + percpu_rwsem_assert_held(&cpuset_rwsem); /* * The parent must be a partition root. @@ -2039,7 +2111,7 @@ static int cpuset_can_attach(struct cgroup_taskset *tset) cpuset_attach_old_cs = task_cs(cgroup_taskset_first(tset, &css)); cs = css_cs(css); - mutex_lock(&cpuset_mutex); + percpu_down_write(&cpuset_rwsem); /* allow moving tasks into an empty cpuset if on default hierarchy */ ret = -ENOSPC; @@ -2063,7 +2135,7 @@ static int cpuset_can_attach(struct cgroup_taskset *tset) cs->attach_in_progress++; ret = 0; out_unlock: - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); return ret; } @@ -2073,9 +2145,9 @@ static void cpuset_cancel_attach(struct cgroup_taskset *tset) cgroup_taskset_first(tset, &css); - mutex_lock(&cpuset_mutex); + percpu_down_write(&cpuset_rwsem); css_cs(css)->attach_in_progress--; - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); } /* @@ -2098,7 +2170,7 @@ static void cpuset_attach(struct cgroup_taskset *tset) cgroup_taskset_first(tset, &css); cs = css_cs(css); - mutex_lock(&cpuset_mutex); + percpu_down_write(&cpuset_rwsem); /* prepare for attach */ if (cs == &top_cpuset) @@ -2152,7 +2224,7 @@ static void cpuset_attach(struct cgroup_taskset *tset) if (!cs->attach_in_progress) wake_up(&cpuset_attach_wq); - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); } /* The various types of files and directories in a cpuset file system */ @@ -2183,7 +2255,8 @@ static int cpuset_write_u64(struct cgroup_subsys_state *css, struct cftype *cft, cpuset_filetype_t type = cft->private; int retval = 0; - mutex_lock(&cpuset_mutex); + get_online_cpus(); + percpu_down_write(&cpuset_rwsem); if (!is_cpuset_online(cs)) { retval = -ENODEV; goto out_unlock; @@ -2219,7 +2292,8 @@ static int cpuset_write_u64(struct cgroup_subsys_state *css, struct cftype *cft, break; } out_unlock: - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); + put_online_cpus(); return retval; } @@ -2230,7 +2304,8 @@ static int cpuset_write_s64(struct cgroup_subsys_state *css, struct cftype *cft, cpuset_filetype_t type = cft->private; int retval = -ENODEV; - mutex_lock(&cpuset_mutex); + get_online_cpus(); + percpu_down_write(&cpuset_rwsem); if (!is_cpuset_online(cs)) goto out_unlock; @@ -2243,7 +2318,8 @@ static int cpuset_write_s64(struct cgroup_subsys_state *css, struct cftype *cft, break; } out_unlock: - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); + put_online_cpus(); return retval; } @@ -2282,7 +2358,8 @@ static ssize_t cpuset_write_resmask(struct kernfs_open_file *of, kernfs_break_active_protection(of->kn); flush_work(&cpuset_hotplug_work); - mutex_lock(&cpuset_mutex); + get_online_cpus(); + percpu_down_write(&cpuset_rwsem); if (!is_cpuset_online(cs)) goto out_unlock; @@ -2306,7 +2383,8 @@ static ssize_t cpuset_write_resmask(struct kernfs_open_file *of, free_cpuset(trialcs); out_unlock: - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); + put_online_cpus(); kernfs_unbreak_active_protection(of->kn); css_put(&cs->css); flush_workqueue(cpuset_migrate_mm_wq); @@ -2437,13 +2515,15 @@ static ssize_t sched_partition_write(struct kernfs_open_file *of, char *buf, return -EINVAL; css_get(&cs->css); - mutex_lock(&cpuset_mutex); + get_online_cpus(); + percpu_down_write(&cpuset_rwsem); if (!is_cpuset_online(cs)) goto out_unlock; retval = update_prstate(cs, val); out_unlock: - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); + put_online_cpus(); css_put(&cs->css); return retval ?: nbytes; } @@ -2649,7 +2729,8 @@ static int cpuset_css_online(struct cgroup_subsys_state *css) if (!parent) return 0; - mutex_lock(&cpuset_mutex); + get_online_cpus(); + percpu_down_write(&cpuset_rwsem); set_bit(CS_ONLINE, &cs->flags); if (is_spread_page(parent)) @@ -2700,7 +2781,8 @@ static int cpuset_css_online(struct cgroup_subsys_state *css) cpumask_copy(cs->effective_cpus, parent->cpus_allowed); spin_unlock_irq(&callback_lock); out_unlock: - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); + put_online_cpus(); return 0; } @@ -2719,7 +2801,8 @@ static void cpuset_css_offline(struct cgroup_subsys_state *css) { struct cpuset *cs = css_cs(css); - mutex_lock(&cpuset_mutex); + get_online_cpus(); + percpu_down_write(&cpuset_rwsem); if (is_partition_root(cs)) update_prstate(cs, 0); @@ -2738,7 +2821,8 @@ static void cpuset_css_offline(struct cgroup_subsys_state *css) cpuset_dec(); clear_bit(CS_ONLINE, &cs->flags); - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); + put_online_cpus(); } static void cpuset_css_free(struct cgroup_subsys_state *css) @@ -2750,7 +2834,7 @@ static void cpuset_css_free(struct cgroup_subsys_state *css) static void cpuset_bind(struct cgroup_subsys_state *root_css) { - mutex_lock(&cpuset_mutex); + percpu_down_write(&cpuset_rwsem); spin_lock_irq(&callback_lock); if (is_in_v2_mode()) { @@ -2763,7 +2847,7 @@ static void cpuset_bind(struct cgroup_subsys_state *root_css) } spin_unlock_irq(&callback_lock); - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); } /* @@ -2805,6 +2889,8 @@ struct cgroup_subsys cpuset_cgrp_subsys = { int __init cpuset_init(void) { + BUG_ON(percpu_init_rwsem(&cpuset_rwsem)); + BUG_ON(!alloc_cpumask_var(&top_cpuset.cpus_allowed, GFP_KERNEL)); BUG_ON(!alloc_cpumask_var(&top_cpuset.effective_cpus, GFP_KERNEL)); BUG_ON(!zalloc_cpumask_var(&top_cpuset.subparts_cpus, GFP_KERNEL)); @@ -2876,7 +2962,7 @@ hotplug_update_tasks_legacy(struct cpuset *cs, is_empty = cpumask_empty(cs->cpus_allowed) || nodes_empty(cs->mems_allowed); - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); /* * Move tasks to the nearest ancestor with execution resources, @@ -2886,7 +2972,7 @@ hotplug_update_tasks_legacy(struct cpuset *cs, if (is_empty) remove_tasks_in_empty_cpuset(cs); - mutex_lock(&cpuset_mutex); + percpu_down_write(&cpuset_rwsem); } static void @@ -2936,14 +3022,14 @@ static void cpuset_hotplug_update_tasks(struct cpuset *cs, struct tmpmasks *tmp) retry: wait_event(cpuset_attach_wq, cs->attach_in_progress == 0); - mutex_lock(&cpuset_mutex); + percpu_down_write(&cpuset_rwsem); /* * We have raced with task attaching. We wait until attaching * is finished, so we won't attach a task to an empty cpuset. */ if (cs->attach_in_progress) { - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); goto retry; } @@ -3011,7 +3097,7 @@ update_tasks: hotplug_update_tasks_legacy(cs, &new_cpus, &new_mems, cpus_updated, mems_updated); - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); } /** @@ -3041,7 +3127,7 @@ static void cpuset_hotplug_workfn(struct work_struct *work) if (on_dfl && !alloc_cpumasks(NULL, &tmp)) ptmp = &tmp; - mutex_lock(&cpuset_mutex); + percpu_down_write(&cpuset_rwsem); /* fetch the available cpus/mems and find out which changed how */ cpumask_copy(&new_cpus, cpu_active_mask); @@ -3091,7 +3177,7 @@ static void cpuset_hotplug_workfn(struct work_struct *work) update_tasks_nodemask(&top_cpuset); } - mutex_unlock(&cpuset_mutex); + percpu_up_write(&cpuset_rwsem); /* if cpus or mems changed, we need to propagate to descendants */ if (cpus_updated || mems_updated) { diff --git a/kernel/cgroup/freezer.c b/kernel/cgroup/freezer.c index 8cf010680678..3984dd6b8ddb 100644 --- a/kernel/cgroup/freezer.c +++ b/kernel/cgroup/freezer.c @@ -231,6 +231,15 @@ void cgroup_freezer_migrate_task(struct task_struct *task, return; /* + * It's not necessary to do changes if both of the src and dst cgroups + * are not freezing and task is not frozen. + */ + if (!test_bit(CGRP_FREEZE, &src->flags) && + !test_bit(CGRP_FREEZE, &dst->flags) && + !task->frozen) + return; + + /* * Adjust counters of freezing and frozen tasks. * Note, that if the task is frozen, but the destination cgroup is not * frozen, we bump both counters to keep them balanced. diff --git a/kernel/cgroup/pids.c b/kernel/cgroup/pids.c index 8e513a573fe9..138059eb730d 100644 --- a/kernel/cgroup/pids.c +++ b/kernel/cgroup/pids.c @@ -45,7 +45,7 @@ struct pids_cgroup { * %PIDS_MAX = (%PID_MAX_LIMIT + 1). */ atomic64_t counter; - int64_t limit; + atomic64_t limit; /* Handle for "pids.events" */ struct cgroup_file events_file; @@ -73,8 +73,8 @@ pids_css_alloc(struct cgroup_subsys_state *parent) if (!pids) return ERR_PTR(-ENOMEM); - pids->limit = PIDS_MAX; atomic64_set(&pids->counter, 0); + atomic64_set(&pids->limit, PIDS_MAX); atomic64_set(&pids->events_limit, 0); return &pids->css; } @@ -146,13 +146,14 @@ static int pids_try_charge(struct pids_cgroup *pids, int num) for (p = pids; parent_pids(p); p = parent_pids(p)) { int64_t new = atomic64_add_return(num, &p->counter); + int64_t limit = atomic64_read(&p->limit); /* * Since new is capped to the maximum number of pid_t, if * p->limit is %PIDS_MAX then we know that this test will never * fail. */ - if (new > p->limit) + if (new > limit) goto revert; } @@ -277,7 +278,7 @@ set_limit: * Limit updates don't need to be mutex'd, since it isn't * critical that any racing fork()s follow the new limit. */ - pids->limit = limit; + atomic64_set(&pids->limit, limit); return nbytes; } @@ -285,7 +286,7 @@ static int pids_max_show(struct seq_file *sf, void *v) { struct cgroup_subsys_state *css = seq_css(sf); struct pids_cgroup *pids = css_pids(css); - int64_t limit = pids->limit; + int64_t limit = atomic64_read(&pids->limit); if (limit >= PIDS_MAX) seq_printf(sf, "%s\n", PIDS_MAX_STR); diff --git a/kernel/cgroup/rstat.c b/kernel/cgroup/rstat.c index ca19b4c8acf5..6f87352f8219 100644 --- a/kernel/cgroup/rstat.c +++ b/kernel/cgroup/rstat.c @@ -33,7 +33,7 @@ void cgroup_rstat_updated(struct cgroup *cgrp, int cpu) return; /* - * Paired with the one in cgroup_rstat_cpu_pop_upated(). Either we + * Paired with the one in cgroup_rstat_cpu_pop_updated(). Either we * see NULL updated_next or they see our updated stat. */ smp_mb(); @@ -304,44 +304,48 @@ void __init cgroup_rstat_boot(void) * Functions for cgroup basic resource statistics implemented on top of * rstat. */ -static void cgroup_base_stat_accumulate(struct cgroup_base_stat *dst_bstat, - struct cgroup_base_stat *src_bstat) +static void cgroup_base_stat_add(struct cgroup_base_stat *dst_bstat, + struct cgroup_base_stat *src_bstat) { dst_bstat->cputime.utime += src_bstat->cputime.utime; dst_bstat->cputime.stime += src_bstat->cputime.stime; dst_bstat->cputime.sum_exec_runtime += src_bstat->cputime.sum_exec_runtime; } +static void cgroup_base_stat_sub(struct cgroup_base_stat *dst_bstat, + struct cgroup_base_stat *src_bstat) +{ + dst_bstat->cputime.utime -= src_bstat->cputime.utime; + dst_bstat->cputime.stime -= src_bstat->cputime.stime; + dst_bstat->cputime.sum_exec_runtime -= src_bstat->cputime.sum_exec_runtime; +} + static void cgroup_base_stat_flush(struct cgroup *cgrp, int cpu) { struct cgroup *parent = cgroup_parent(cgrp); struct cgroup_rstat_cpu *rstatc = cgroup_rstat_cpu(cgrp, cpu); - struct task_cputime *last_cputime = &rstatc->last_bstat.cputime; - struct task_cputime cputime; - struct cgroup_base_stat delta; + struct cgroup_base_stat cur, delta; unsigned seq; /* fetch the current per-cpu values */ do { seq = __u64_stats_fetch_begin(&rstatc->bsync); - cputime = rstatc->bstat.cputime; + cur.cputime = rstatc->bstat.cputime; } while (__u64_stats_fetch_retry(&rstatc->bsync, seq)); - /* calculate the delta to propgate */ - delta.cputime.utime = cputime.utime - last_cputime->utime; - delta.cputime.stime = cputime.stime - last_cputime->stime; - delta.cputime.sum_exec_runtime = cputime.sum_exec_runtime - - last_cputime->sum_exec_runtime; - *last_cputime = cputime; - - /* transfer the pending stat into delta */ - cgroup_base_stat_accumulate(&delta, &cgrp->pending_bstat); - memset(&cgrp->pending_bstat, 0, sizeof(cgrp->pending_bstat)); - - /* propagate delta into the global stat and the parent's pending */ - cgroup_base_stat_accumulate(&cgrp->bstat, &delta); - if (parent) - cgroup_base_stat_accumulate(&parent->pending_bstat, &delta); + /* propagate percpu delta to global */ + delta = cur; + cgroup_base_stat_sub(&delta, &rstatc->last_bstat); + cgroup_base_stat_add(&cgrp->bstat, &delta); + cgroup_base_stat_add(&rstatc->last_bstat, &delta); + + /* propagate global delta to parent */ + if (parent) { + delta = cgrp->bstat; + cgroup_base_stat_sub(&delta, &cgrp->last_bstat); + cgroup_base_stat_add(&parent->bstat, &delta); + cgroup_base_stat_add(&cgrp->last_bstat, &delta); + } } static struct cgroup_rstat_cpu * |
