diff options
author | H. Peter Anvin <hpa@zytor.com> | 2009-08-27 00:17:51 (GMT) |
---|---|---|
committer | H. Peter Anvin <hpa@zytor.com> | 2009-08-27 00:24:28 (GMT) |
commit | b855192c08fcb14adbc5d3a7cab182022d433cca (patch) | |
tree | 3a10cafbfbf98cafacf667eb218c71300b351bfa /kernel | |
parent | d886c73cd4cf02a71e1650cbcb6176799d78aac1 (diff) | |
parent | 3e0e1e9c5a327d4dba8490d83ef55c0564e6e8a7 (diff) | |
download | linux-b855192c08fcb14adbc5d3a7cab182022d433cca.tar.xz |
Merge branch 'x86/urgent' into x86/pat
Reason: Change to is_new_memtype_allowed() in x86/urgent
Resolved semantic conflicts in:
arch/x86/mm/pat.c
arch/x86/mm/ioremap.c
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Diffstat (limited to 'kernel')
54 files changed, 1143 insertions, 609 deletions
diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 3737a68..b6eadfe 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -47,6 +47,7 @@ #include <linux/hash.h> #include <linux/namei.h> #include <linux/smp_lock.h> +#include <linux/pid_namespace.h> #include <asm/atomic.h> @@ -734,16 +735,28 @@ static void cgroup_d_remove_dir(struct dentry *dentry) * reference to css->refcnt. In general, this refcnt is expected to goes down * to zero, soon. * - * CGRP_WAIT_ON_RMDIR flag is modified under cgroup's inode->i_mutex; + * CGRP_WAIT_ON_RMDIR flag is set under cgroup's inode->i_mutex; */ DECLARE_WAIT_QUEUE_HEAD(cgroup_rmdir_waitq); -static void cgroup_wakeup_rmdir_waiters(const struct cgroup *cgrp) +static void cgroup_wakeup_rmdir_waiter(struct cgroup *cgrp) { - if (unlikely(test_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags))) + if (unlikely(test_and_clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags))) wake_up_all(&cgroup_rmdir_waitq); } +void cgroup_exclude_rmdir(struct cgroup_subsys_state *css) +{ + css_get(css); +} + +void cgroup_release_and_wakeup_rmdir(struct cgroup_subsys_state *css) +{ + cgroup_wakeup_rmdir_waiter(css->cgroup); + css_put(css); +} + + static int rebind_subsystems(struct cgroupfs_root *root, unsigned long final_bits) { @@ -960,6 +973,7 @@ static void init_cgroup_housekeeping(struct cgroup *cgrp) INIT_LIST_HEAD(&cgrp->children); INIT_LIST_HEAD(&cgrp->css_sets); INIT_LIST_HEAD(&cgrp->release_list); + INIT_LIST_HEAD(&cgrp->pids_list); init_rwsem(&cgrp->pids_mutex); } static void init_cgroup_root(struct cgroupfs_root *root) @@ -1357,7 +1371,7 @@ int cgroup_attach_task(struct cgroup *cgrp, struct task_struct *tsk) * wake up rmdir() waiter. the rmdir should fail since the cgroup * is no longer empty. */ - cgroup_wakeup_rmdir_waiters(cgrp); + cgroup_wakeup_rmdir_waiter(cgrp); return 0; } @@ -2201,12 +2215,30 @@ err: return ret; } +/* + * Cache pids for all threads in the same pid namespace that are + * opening the same "tasks" file. + */ +struct cgroup_pids { + /* The node in cgrp->pids_list */ + struct list_head list; + /* The cgroup those pids belong to */ + struct cgroup *cgrp; + /* The namepsace those pids belong to */ + struct pid_namespace *ns; + /* Array of process ids in the cgroup */ + pid_t *tasks_pids; + /* How many files are using the this tasks_pids array */ + int use_count; + /* Length of the current tasks_pids array */ + int length; +}; + static int cmppid(const void *a, const void *b) { return *(pid_t *)a - *(pid_t *)b; } - /* * seq_file methods for the "tasks" file. The seq_file position is the * next pid to display; the seq_file iterator is a pointer to the pid @@ -2221,45 +2253,47 @@ static void *cgroup_tasks_start(struct seq_file *s, loff_t *pos) * after a seek to the start). Use a binary-search to find the * next pid to display, if any */ - struct cgroup *cgrp = s->private; + struct cgroup_pids *cp = s->private; + struct cgroup *cgrp = cp->cgrp; int index = 0, pid = *pos; int *iter; down_read(&cgrp->pids_mutex); if (pid) { - int end = cgrp->pids_length; + int end = cp->length; while (index < end) { int mid = (index + end) / 2; - if (cgrp->tasks_pids[mid] == pid) { + if (cp->tasks_pids[mid] == pid) { index = mid; break; - } else if (cgrp->tasks_pids[mid] <= pid) + } else if (cp->tasks_pids[mid] <= pid) index = mid + 1; else end = mid; } } /* If we're off the end of the array, we're done */ - if (index >= cgrp->pids_length) + if (index >= cp->length) return NULL; /* Update the abstract position to be the actual pid that we found */ - iter = cgrp->tasks_pids + index; + iter = cp->tasks_pids + index; *pos = *iter; return iter; } static void cgroup_tasks_stop(struct seq_file *s, void *v) { - struct cgroup *cgrp = s->private; + struct cgroup_pids *cp = s->private; + struct cgroup *cgrp = cp->cgrp; up_read(&cgrp->pids_mutex); } static void *cgroup_tasks_next(struct seq_file *s, void *v, loff_t *pos) { - struct cgroup *cgrp = s->private; + struct cgroup_pids *cp = s->private; int *p = v; - int *end = cgrp->tasks_pids + cgrp->pids_length; + int *end = cp->tasks_pids + cp->length; /* * Advance to the next pid in the array. If this goes off the @@ -2286,26 +2320,33 @@ static struct seq_operations cgroup_tasks_seq_operations = { .show = cgroup_tasks_show, }; -static void release_cgroup_pid_array(struct cgroup *cgrp) +static void release_cgroup_pid_array(struct cgroup_pids *cp) { + struct cgroup *cgrp = cp->cgrp; + down_write(&cgrp->pids_mutex); - BUG_ON(!cgrp->pids_use_count); - if (!--cgrp->pids_use_count) { - kfree(cgrp->tasks_pids); - cgrp->tasks_pids = NULL; - cgrp->pids_length = 0; + BUG_ON(!cp->use_count); + if (!--cp->use_count) { + list_del(&cp->list); + put_pid_ns(cp->ns); + kfree(cp->tasks_pids); + kfree(cp); } up_write(&cgrp->pids_mutex); } static int cgroup_tasks_release(struct inode *inode, struct file *file) { - struct cgroup *cgrp = __d_cgrp(file->f_dentry->d_parent); + struct seq_file *seq; + struct cgroup_pids *cp; if (!(file->f_mode & FMODE_READ)) return 0; - release_cgroup_pid_array(cgrp); + seq = file->private_data; + cp = seq->private; + + release_cgroup_pid_array(cp); return seq_release(inode, file); } @@ -2324,6 +2365,8 @@ static struct file_operations cgroup_tasks_operations = { static int cgroup_tasks_open(struct inode *unused, struct file *file) { struct cgroup *cgrp = __d_cgrp(file->f_dentry->d_parent); + struct pid_namespace *ns = current->nsproxy->pid_ns; + struct cgroup_pids *cp; pid_t *pidarray; int npids; int retval; @@ -2350,20 +2393,37 @@ static int cgroup_tasks_open(struct inode *unused, struct file *file) * array if necessary */ down_write(&cgrp->pids_mutex); - kfree(cgrp->tasks_pids); - cgrp->tasks_pids = pidarray; - cgrp->pids_length = npids; - cgrp->pids_use_count++; + + list_for_each_entry(cp, &cgrp->pids_list, list) { + if (ns == cp->ns) + goto found; + } + + cp = kzalloc(sizeof(*cp), GFP_KERNEL); + if (!cp) { + up_write(&cgrp->pids_mutex); + kfree(pidarray); + return -ENOMEM; + } + cp->cgrp = cgrp; + cp->ns = ns; + get_pid_ns(ns); + list_add(&cp->list, &cgrp->pids_list); +found: + kfree(cp->tasks_pids); + cp->tasks_pids = pidarray; + cp->length = npids; + cp->use_count++; up_write(&cgrp->pids_mutex); file->f_op = &cgroup_tasks_operations; retval = seq_open(file, &cgroup_tasks_seq_operations); if (retval) { - release_cgroup_pid_array(cgrp); + release_cgroup_pid_array(cp); return retval; } - ((struct seq_file *)file->private_data)->private = cgrp; + ((struct seq_file *)file->private_data)->private = cp; return 0; } @@ -2696,33 +2756,42 @@ again: mutex_unlock(&cgroup_mutex); /* + * In general, subsystem has no css->refcnt after pre_destroy(). But + * in racy cases, subsystem may have to get css->refcnt after + * pre_destroy() and it makes rmdir return with -EBUSY. This sometimes + * make rmdir return -EBUSY too often. To avoid that, we use waitqueue + * for cgroup's rmdir. CGRP_WAIT_ON_RMDIR is for synchronizing rmdir + * and subsystem's reference count handling. Please see css_get/put + * and css_tryget() and cgroup_wakeup_rmdir_waiter() implementation. + */ + set_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); + + /* * Call pre_destroy handlers of subsys. Notify subsystems * that rmdir() request comes. */ ret = cgroup_call_pre_destroy(cgrp); - if (ret) + if (ret) { + clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); return ret; + } mutex_lock(&cgroup_mutex); parent = cgrp->parent; if (atomic_read(&cgrp->count) || !list_empty(&cgrp->children)) { + clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); mutex_unlock(&cgroup_mutex); return -EBUSY; } - /* - * css_put/get is provided for subsys to grab refcnt to css. In typical - * case, subsystem has no reference after pre_destroy(). But, under - * hierarchy management, some *temporal* refcnt can be hold. - * To avoid returning -EBUSY to a user, waitqueue is used. If subsys - * is really busy, it should return -EBUSY at pre_destroy(). wake_up - * is called when css_put() is called and refcnt goes down to 0. - */ - set_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); prepare_to_wait(&cgroup_rmdir_waitq, &wait, TASK_INTERRUPTIBLE); - if (!cgroup_clear_css_refs(cgrp)) { mutex_unlock(&cgroup_mutex); - schedule(); + /* + * Because someone may call cgroup_wakeup_rmdir_waiter() before + * prepare_to_wait(), we need to check this flag. + */ + if (test_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags)) + schedule(); finish_wait(&cgroup_rmdir_waitq, &wait); clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags); if (signal_pending(current)) @@ -3294,7 +3363,7 @@ void __css_put(struct cgroup_subsys_state *css) set_bit(CGRP_RELEASABLE, &cgrp->flags); check_for_release(cgrp); } - cgroup_wakeup_rmdir_waiters(cgrp); + cgroup_wakeup_rmdir_waiter(cgrp); } rcu_read_unlock(); } diff --git a/kernel/exit.c b/kernel/exit.c index 628d41f..869dc22 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -12,7 +12,6 @@ #include <linux/completion.h> #include <linux/personality.h> #include <linux/tty.h> -#include <linux/mnt_namespace.h> #include <linux/iocontext.h> #include <linux/key.h> #include <linux/security.h> diff --git a/kernel/fork.c b/kernel/fork.c index 467746b..021e113 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -17,7 +17,6 @@ #include <linux/module.h> #include <linux/vmalloc.h> #include <linux/completion.h> -#include <linux/mnt_namespace.h> #include <linux/personality.h> #include <linux/mempolicy.h> #include <linux/sem.h> @@ -427,6 +426,7 @@ static struct mm_struct * mm_init(struct mm_struct * mm, struct task_struct *p) init_rwsem(&mm->mmap_sem); INIT_LIST_HEAD(&mm->mmlist); mm->flags = (current->mm) ? current->mm->flags : default_dump_filter; + mm->oom_adj = (current->mm) ? current->mm->oom_adj : 0; mm->core_state = NULL; mm->nr_ptes = 0; set_mm_counter(mm, file_rss, 0); @@ -568,18 +568,18 @@ void mm_release(struct task_struct *tsk, struct mm_struct *mm) * the value intact in a core dump, and to save the unnecessary * trouble otherwise. Userland only wants this done for a sys_exit. */ - if (tsk->clear_child_tid - && !(tsk->flags & PF_SIGNALED) - && atomic_read(&mm->mm_users) > 1) { - u32 __user * tidptr = tsk->clear_child_tid; + if (tsk->clear_child_tid) { + if (!(tsk->flags & PF_SIGNALED) && + atomic_read(&mm->mm_users) > 1) { + /* + * We don't check the error code - if userspace has + * not set up a proper pointer then tough luck. + */ + put_user(0, tsk->clear_child_tid); + sys_futex(tsk->clear_child_tid, FUTEX_WAKE, + 1, NULL, NULL, 0); + } tsk->clear_child_tid = NULL; - - /* - * We don't check the error code - if userspace has - * not set up a proper pointer then tough luck. - */ - put_user(0, tidptr); - sys_futex(tidptr, FUTEX_WAKE, 1, NULL, NULL, 0); } } @@ -1269,6 +1269,7 @@ static struct task_struct *copy_process(unsigned long clone_flags, write_unlock_irq(&tasklist_lock); proc_fork_connector(p); cgroup_post_fork(p); + perf_counter_fork(p); return p; bad_fork_free_pid: @@ -1408,12 +1409,6 @@ long do_fork(unsigned long clone_flags, if (clone_flags & CLONE_VFORK) { p->vfork_done = &vfork; init_completion(&vfork); - } else if (!(clone_flags & CLONE_VM)) { - /* - * vfork will do an exec which will call - * set_task_comm() - */ - perf_counter_fork(p); } audit_finish_fork(p); diff --git a/kernel/freezer.c b/kernel/freezer.c index 2f4936c..bd1d42b 100644 --- a/kernel/freezer.c +++ b/kernel/freezer.c @@ -44,12 +44,19 @@ void refrigerator(void) recalc_sigpending(); /* We sent fake signal, clean it up */ spin_unlock_irq(¤t->sighand->siglock); + /* prevent accounting of that task to load */ + current->flags |= PF_FREEZING; + for (;;) { set_current_state(TASK_UNINTERRUPTIBLE); if (!frozen(current)) break; schedule(); } + + /* Remove the accounting blocker */ + current->flags &= ~PF_FREEZING; + pr_debug("%s left refrigerator\n", current->comm); __set_current_state(save); } diff --git a/kernel/futex.c b/kernel/futex.c index 794c862..e18cfbd 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -247,6 +247,7 @@ again: if (err < 0) return err; + page = compound_head(page); lock_page(page); if (!page->mapping) { unlock_page(page); @@ -1009,15 +1010,19 @@ void requeue_futex(struct futex_q *q, struct futex_hash_bucket *hb1, * requeue_pi_wake_futex() - Wake a task that acquired the lock during requeue * q: the futex_q * key: the key of the requeue target futex + * hb: the hash_bucket of the requeue target futex * * During futex_requeue, with requeue_pi=1, it is possible to acquire the * target futex if it is uncontended or via a lock steal. Set the futex_q key * to the requeue target futex so the waiter can detect the wakeup on the right * futex, but remove it from the hb and NULL the rt_waiter so it can detect - * atomic lock acquisition. Must be called with the q->lock_ptr held. + * atomic lock acquisition. Set the q->lock_ptr to the requeue target hb->lock + * to protect access to the pi_state to fixup the owner later. Must be called + * with both q->lock_ptr and hb->lock held. */ static inline -void requeue_pi_wake_futex(struct futex_q *q, union futex_key *key) +void requeue_pi_wake_futex(struct futex_q *q, union futex_key *key, + struct futex_hash_bucket *hb) { drop_futex_key_refs(&q->key); get_futex_key_refs(key); @@ -1029,6 +1034,11 @@ void requeue_pi_wake_futex(struct futex_q *q, union futex_key *key) WARN_ON(!q->rt_waiter); q->rt_waiter = NULL; + q->lock_ptr = &hb->lock; +#ifdef CONFIG_DEBUG_PI_LIST + q->list.plist.lock = &hb->lock; +#endif + wake_up_state(q->task, TASK_NORMAL); } @@ -1087,7 +1097,7 @@ static int futex_proxy_trylock_atomic(u32 __user *pifutex, ret = futex_lock_pi_atomic(pifutex, hb2, key2, ps, top_waiter->task, set_waiters); if (ret == 1) - requeue_pi_wake_futex(top_waiter, key2); + requeue_pi_wake_futex(top_waiter, key2, hb2); return ret; } @@ -1246,8 +1256,15 @@ retry_private: if (!match_futex(&this->key, &key1)) continue; - WARN_ON(!requeue_pi && this->rt_waiter); - WARN_ON(requeue_pi && !this->rt_waiter); + /* + * FUTEX_WAIT_REQEUE_PI and FUTEX_CMP_REQUEUE_PI should always + * be paired with each other and no other futex ops. + */ + if ((requeue_pi && !this->rt_waiter) || + (!requeue_pi && this->rt_waiter)) { + ret = -EINVAL; + break; + } /* * Wake nr_wake waiters. For requeue_pi, if we acquired the @@ -1272,7 +1289,7 @@ retry_private: this->task, 1); if (ret == 1) { /* We got the lock. */ - requeue_pi_wake_futex(this, &key2); + requeue_pi_wake_futex(this, &key2, hb2); continue; } else if (ret) { /* -EDEADLK */ diff --git a/kernel/futex_compat.c b/kernel/futex_compat.c index d607a5b..2357165 100644 --- a/kernel/futex_compat.c +++ b/kernel/futex_compat.c @@ -180,7 +180,8 @@ asmlinkage long compat_sys_futex(u32 __user *uaddr, int op, u32 val, int cmd = op & FUTEX_CMD_MASK; if (utime && (cmd == FUTEX_WAIT || cmd == FUTEX_LOCK_PI || - cmd == FUTEX_WAIT_BITSET)) { + cmd == FUTEX_WAIT_BITSET || + cmd == FUTEX_WAIT_REQUEUE_PI)) { if (get_compat_timespec(&ts, utime)) return -EFAULT; if (!timespec_valid(&ts)) @@ -191,7 +192,8 @@ asmlinkage long compat_sys_futex(u32 __user *uaddr, int op, u32 val, t = ktime_add_safe(ktime_get(), t); tp = &t; } - if (cmd == FUTEX_REQUEUE || cmd == FUTEX_CMP_REQUEUE) + if (cmd == FUTEX_REQUEUE || cmd == FUTEX_CMP_REQUEUE || + cmd == FUTEX_CMP_REQUEUE_PI || cmd == FUTEX_WAKE_OP) val2 = (int) (unsigned long) utime; return do_futex(uaddr, op, val, tp, uaddr2, val2, val3); diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c index 9002958..49da79a 100644 --- a/kernel/hrtimer.c +++ b/kernel/hrtimer.c @@ -191,6 +191,46 @@ struct hrtimer_clock_base *lock_hrtimer_base(const struct hrtimer *timer, } } + +/* + * Get the preferred target CPU for NOHZ + */ +static int hrtimer_get_target(int this_cpu, int pinned) +{ +#ifdef CONFIG_NO_HZ + if (!pinned && get_sysctl_timer_migration() && idle_cpu(this_cpu)) { + int preferred_cpu = get_nohz_load_balancer(); + + if (preferred_cpu >= 0) + return preferred_cpu; + } +#endif + return this_cpu; +} + +/* + * With HIGHRES=y we do not migrate the timer when it is expiring + * before the next event on the target cpu because we cannot reprogram + * the target cpu hardware and we would cause it to fire late. + * + * Called with cpu_base->lock of target cpu held. + */ +static int +hrtimer_check_target(struct hrtimer *timer, struct hrtimer_clock_base *new_base) +{ +#ifdef CONFIG_HIGH_RES_TIMERS + ktime_t expires; + + if (!new_base->cpu_base->hres_active) + return 0; + + expires = ktime_sub(hrtimer_get_expires(timer), new_base->offset); + return expires.tv64 <= new_base->cpu_base->expires_next.tv64; +#else + return 0; +#endif +} + /* * Switch the timer base to the current CPU when possible. */ @@ -200,16 +240,8 @@ switch_hrtimer_base(struct hrtimer *timer, struct hrtimer_clock_base *base, { struct hrtimer_clock_base *new_base; struct hrtimer_cpu_base *new_cpu_base; - int cpu, preferred_cpu = -1; - - cpu = smp_processor_id(); -#if defined(CONFIG_NO_HZ) && defined(CONFIG_SMP) - if (!pinned && get_sysctl_timer_migration() && idle_cpu(cpu)) { - preferred_cpu = get_nohz_load_balancer(); - if (preferred_cpu >= 0) - cpu = preferred_cpu; - } -#endif + int this_cpu = smp_processor_id(); + int cpu = hrtimer_get_target(this_cpu, pinned); again: new_cpu_base = &per_cpu(hrtimer_bases, cpu); @@ -217,7 +249,7 @@ again: if (base != new_base) { /* - * We are trying to schedule the timer on the local CPU. + * We are trying to move timer to new_base. * However we can't change timer's base while it is running, * so we keep it on the same CPU. No hassle vs. reprogramming * the event source in the high resolution case. The softirq @@ -233,38 +265,12 @@ again: spin_unlock(&base->cpu_base->lock); spin_lock(&new_base->cpu_base->lock); - /* Optimized away for NOHZ=n SMP=n */ - if (cpu == preferred_cpu) { - /* Calculate clock monotonic expiry time */ -#ifdef CONFIG_HIGH_RES_TIMERS - ktime_t expires = ktime_sub(hrtimer_get_expires(timer), - new_base->offset); -#else - ktime_t expires = hrtimer_get_expires(timer); -#endif - - /* - * Get the next event on target cpu from the - * clock events layer. - * This covers the highres=off nohz=on case as well. - */ - ktime_t next = clockevents_get_next_event(cpu); - - ktime_t delta = ktime_sub(expires, next); - - /* - * We do not migrate the timer when it is expiring - * before the next event on the target cpu because - * we cannot reprogram the target cpu hardware and - * we would cause it to fire late. - */ - if (delta.tv64 < 0) { - cpu = smp_processor_id(); - spin_unlock(&new_base->cpu_base->lock); - spin_lock(&base->cpu_base->lock); - timer->base = base; - goto again; - } + if (cpu != this_cpu && hrtimer_check_target(timer, new_base)) { + cpu = this_cpu; + spin_unlock(&new_base->cpu_base->lock); + spin_lock(&base->cpu_base->lock); + timer->base = base; + goto again; } timer->base = new_base; } @@ -1276,14 +1282,22 @@ void hrtimer_interrupt(struct clock_event_device *dev) expires_next.tv64 = KTIME_MAX; + spin_lock(&cpu_base->lock); + /* + * We set expires_next to KTIME_MAX here with cpu_base->lock + * held to prevent that a timer is enqueued in our queue via + * the migration code. This does not affect enqueueing of + * timers which run their callback and need to be requeued on + * this CPU. + */ + cpu_base->expires_next.tv64 = KTIME_MAX; + base = cpu_base->clock_base; for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) { ktime_t basenow; struct rb_node *node; - spin_lock(&cpu_base->lock); - basenow = ktime_add(now, base->offset); while ((node = base->first)) { @@ -1316,11 +1330,15 @@ void hrtimer_interrupt(struct clock_event_device *dev) __run_hrtimer(timer); } - spin_unlock(&cpu_base->lock); base++; } + /* + * Store the new expiry value so the migration code can verify + * against it. + */ cpu_base->expires_next = expires_next; + spin_unlock(&cpu_base->lock); /* Reprogramming necessary ? */ if (expires_next.tv64 != KTIME_MAX) { diff --git a/kernel/irq/internals.h b/kernel/irq/internals.h index 7346825..e70ed55 100644 --- a/kernel/irq/internals.h +++ b/kernel/irq/internals.h @@ -42,8 +42,7 @@ static inline void unregister_handler_proc(unsigned int irq, extern int irq_select_affinity_usr(unsigned int irq); -extern void -irq_set_thread_affinity(struct irq_desc *desc, const struct cpumask *cpumask); +extern void irq_set_thread_affinity(struct irq_desc *desc); /* * Debugging printout: diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index 50da676..d222515 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -80,14 +80,22 @@ int irq_can_set_affinity(unsigned int irq) return 1; } -void -irq_set_thread_affinity(struct irq_desc *desc, const struct cpumask *cpumask) +/** + * irq_set_thread_affinity - Notify irq threads to adjust affinity + * @desc: irq descriptor which has affitnity changed + * + * We just set IRQTF_AFFINITY and delegate the affinity setting + * to the interrupt thread itself. We can not call + * set_cpus_allowed_ptr() here as we hold desc->lock and this + * code can be called from hard interrupt context. + */ +void irq_set_thread_affinity(struct irq_desc *desc) { struct irqaction *action = desc->action; while (action) { if (action->thread) - set_cpus_allowed_ptr(action->thread, cpumask); + set_bit(IRQTF_AFFINITY, &action->thread_flags); action = action->next; } } @@ -112,7 +120,7 @@ int irq_set_affinity(unsigned int irq, const struct cpumask *cpumask) if (desc->status & IRQ_MOVE_PCNTXT) { if (!desc->chip->set_affinity(irq, cpumask)) { cpumask_copy(desc->affinity, cpumask); - irq_set_thread_affinity(desc, cpumask); + irq_set_thread_affinity(desc); } } else { @@ -122,7 +130,7 @@ int irq_set_affinity(unsigned int irq, const struct cpumask *cpumask) #else if (!desc->chip->set_affinity(irq, cpumask)) { cpumask_copy(desc->affinity, cpumask); - irq_set_thread_affinity(desc, cpumask); + irq_set_thread_affinity(desc); } #endif desc->status |= IRQ_AFFINITY_SET; @@ -176,7 +184,7 @@ int irq_select_affinity_usr(unsigned int irq) spin_lock_irqsave(&desc->lock, flags); ret = setup_affinity(irq, desc); if (!ret) - irq_set_thread_affinity(desc, desc->affinity); + irq_set_thread_affinity(desc); spin_unlock_irqrestore(&desc->lock, flags); return ret; @@ -443,6 +451,39 @@ static int irq_wait_for_interrupt(struct irqaction *action) return -1; } +#ifdef CONFIG_SMP +/* + * Check whether we need to change the affinity of the interrupt thread. + */ +static void +irq_thread_check_affinity(struct irq_desc *desc, struct irqaction *action) +{ + cpumask_var_t mask; + + if (!test_and_clear_bit(IRQTF_AFFINITY, &action->thread_flags)) + return; + + /* + * In case we are out of memory we set IRQTF_AFFINITY again and + * try again next time + */ + if (!alloc_cpumask_var(&mask, GFP_KERNEL)) { + set_bit(IRQTF_AFFINITY, &action->thread_flags); + return; + } + + spin_lock_irq(&desc->lock); + cpumask_copy(mask, desc->affinity); + spin_unlock_irq(&desc->lock); + + set_cpus_allowed_ptr(current, mask); + free_cpumask_var(mask); +} +#else +static inline void +irq_thread_check_affinity(struct irq_desc *desc, struct irqaction *action) { } +#endif + /* * Interrupt handler thread */ @@ -458,6 +499,8 @@ static int irq_thread(void *data) while (!irq_wait_for_interrupt(action)) { + irq_thread_check_affinity(desc, action); + atomic_inc(&desc->threads_active); spin_lock_irq(&desc->lock); @@ -718,7 +761,6 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id) { struct irq_desc *desc = irq_to_desc(irq); struct irqaction *action, **action_ptr; - struct task_struct *irqthread; unsigned long flags; WARN(in_interrupt(), "Trying to free IRQ %d from IRQ context!\n", irq); @@ -766,9 +808,6 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id) desc->chip->disable(irq); } - irqthread = action->thread; - action->thread = NULL; - spin_unlock_irqrestore(&desc->lock, flags); unregister_handler_proc(irq, action); @@ -776,12 +815,6 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id) /* Make sure it's not being used on another CPU: */ synchronize_irq(irq); - if (irqthread) { - if (!test_bit(IRQTF_DIED, &action->thread_flags)) - kthread_stop(irqthread); - put_task_struct(irqthread); - } - #ifdef CONFIG_DEBUG_SHIRQ /* * It's a shared IRQ -- the driver ought to be prepared for an IRQ @@ -797,6 +830,13 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id) local_irq_restore(flags); } #endif + + if (action->thread) { + if (!test_bit(IRQTF_DIED, &action->thread_flags)) + kthread_stop(action->thread); + put_task_struct(action->thread); + } + return action; } diff --git a/kernel/irq/migration.c b/kernel/irq/migration.c index cfe767c..fcb6c96 100644 --- a/kernel/irq/migration.c +++ b/kernel/irq/migration.c @@ -45,7 +45,7 @@ void move_masked_irq(int irq) < nr_cpu_ids)) if (!desc->chip->set_affinity(irq, desc->pending_mask)) { cpumask_copy(desc->affinity, desc->pending_mask); - irq_set_thread_affinity(desc, desc->pending_mask); + irq_set_thread_affinity(desc); } cpumask_clear(desc->pending_mask); diff --git a/kernel/irq/numa_migrate.c b/kernel/irq/numa_migrate.c index 2f69bee..3fd3019 100644 --- a/kernel/irq/numa_migrate.c +++ b/kernel/irq/numa_migrate.c @@ -107,8 +107,8 @@ out_unlock: struct irq_desc *move_irq_desc(struct irq_desc *desc, int node) { - /* those all static, do move them */ - if (desc->irq < NR_IRQS_LEGACY) + /* those static or target node is -1, do not move them */ + if (desc->irq < NR_IRQS_LEGACY || node == -1) return desc; if (desc->node != node) diff --git a/kernel/kexec.c b/kernel/kexec.c index ae1c352..f336e21 100644 --- a/kernel/kexec.c +++ b/kernel/kexec.c @@ -1228,7 +1228,7 @@ static int __init parse_crashkernel_mem(char *cmdline, } while (*cur++ == ','); if (*crash_size > 0) { - while (*cur != ' ' && *cur != '@') + while (*cur && *cur != ' ' && *cur != '@') cur++; if (*cur == '@') { cur++; diff --git a/kernel/kmod.c b/kernel/kmod.c index 7e95bed..385c31a 100644 --- a/kernel/kmod.c +++ b/kernel/kmod.c @@ -24,7 +24,6 @@ #include <linux/unistd.h> #include <linux/kmod.h> #include <linux/slab.h> -#include <linux/mnt_namespace.h> #include <linux/completion.h> #include <linux/file.h> #include <linux/fdtable.h> diff --git a/kernel/kprobes.c b/kernel/kprobes.c index c0fa54b..0540948 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -237,13 +237,9 @@ static int __kprobes collect_garbage_slots(void) { struct kprobe_insn_page *kip; struct hlist_node *pos, *next; - int safety; /* Ensure no-one is preepmted on the garbages */ - mutex_unlock(&kprobe_insn_mutex); - safety = check_safety(); - mutex_lock(&kprobe_insn_mutex); - if (safety != 0) + if (check_safety()) return -EAGAIN; hlist_for_each_entry_safe(kip, pos, next, &kprobe_insn_pages, hlist) { @@ -698,7 +694,7 @@ int __kprobes register_kprobe(struct kprobe *p) p->addr = addr; preempt_disable(); - if (!__kernel_text_address((unsigned long) p->addr) || + if (!kernel_text_address((unsigned long) p->addr) || in_kprobes_functions((unsigned long) p->addr)) { preempt_enable(); return -EINVAL; diff --git a/kernel/kthread.c b/kernel/kthread.c index 9b1a7de..eb8751a 100644 --- a/kernel/kthread.c +++ b/kernel/kthread.c @@ -180,10 +180,12 @@ EXPORT_SYMBOL(kthread_bind); * @k: thread created by kthread_create(). * * Sets kthread_should_stop() for @k to return true, wakes it, and - * waits for it to exit. Your threadfn() must not call do_exit() - * itself if you use this function! This can also be called after - * kthread_create() instead of calling wake_up_process(): the thread - * will exit without calling threadfn(). + * waits for it to exit. This can also be called after kthread_create() + * instead of calling wake_up_process(): the thread will exit without + * calling threadfn(). + * + * If threadfn() may call do_exit() itself, the caller must ensure + * task_struct can't go away. * * Returns the result of threadfn(), or %-EINTR if wake_up_process() * was never called. diff --git a/kernel/lockdep_proc.c b/kernel/lockdep_proc.c index d7135aa..e94caa6 100644 --- a/kernel/lockdep_proc.c +++ b/kernel/lockdep_proc.c @@ -758,7 +758,8 @@ static int __init lockdep_proc_init(void) &proc_lockdep_stats_operations); #ifdef CONFIG_LOCK_STAT - proc_create("lock_stat", S_IRUSR, NULL, &proc_lock_stat_operations); + proc_create("lock_stat", S_IRUSR | S_IWUSR, NULL, + &proc_lock_stat_operations); #endif return 0; diff --git a/kernel/module.c b/kernel/module.c index 38928fc..fd141140 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -1068,7 +1068,8 @@ static inline int check_modstruct_version(Elf_Shdr *sechdrs, { const unsigned long *crc; - if (!find_symbol("module_layout", NULL, &crc, true, false)) + if (!find_symbol(MODULE_SYMBOL_PREFIX "module_layout", NULL, + &crc, true, false)) BUG(); return check_version(sechdrs, versindex, "module_layout", mod, crc); } @@ -2451,9 +2452,9 @@ SYSCALL_DEFINE3(init_module, void __user *, umod, return ret; } if (ret > 0) { - printk(KERN_WARNING "%s: '%s'->init suspiciously returned %d, " - "it should follow 0/-E convention\n" - KERN_WARNING "%s: loading module anyway...\n", + printk(KERN_WARNING +"%s: '%s'->init suspiciously returned %d, it should follow 0/-E convention\n" +"%s: loading module anyway...\n", __func__, mod->name, ret, __func__); dump_stack(); diff --git a/kernel/panic.c b/kernel/panic.c index 984b3ec..512ab73 100644 --- a/kernel/panic.c +++ b/kernel/panic.c @@ -301,6 +301,7 @@ int oops_may_print(void) */ void oops_enter(void) { + tracing_off(); /* can't trust the integrity of the kernel anymore: */ debug_locks_off(); do_oops_enter_exit(); diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c index d55a50d..534e20d 100644 --- a/kernel/perf_counter.c +++ b/kernel/perf_counter.c @@ -42,6 +42,7 @@ static int perf_overcommit __read_mostly = 1; static atomic_t nr_counters __read_mostly; static atomic_t nr_mmap_counters __read_mostly; static atomic_t nr_comm_counters __read_mostly; +static atomic_t nr_task_counters __read_mostly; /* * perf counter paranoia level: @@ -87,6 +88,7 @@ void __weak hw_perf_disable(void) { barrier(); } void __weak hw_perf_enable(void) { barrier(); } void __weak hw_perf_counter_setup(int cpu) { barrier(); } +void __weak hw_perf_counter_setup_online(int cpu) { barrier(); } int __weak hw_perf_group_sched_in(struct perf_counter *group_leader, @@ -146,6 +148,28 @@ static void put_ctx(struct perf_counter_context *ctx) } } +static void unclone_ctx(struct perf_counter_context *ctx) +{ + if (ctx->parent_ctx) { + put_ctx(ctx->parent_ctx); + ctx->parent_ctx = NULL; + } +} + +/* + * If we inherit counters we want to return the parent counter id + * to userspace. + */ +static u64 primary_counter_id(struct perf_counter *counter) +{ + u64 id = counter->id; + + if (counter->parent) + id = counter->parent->id; + + return id; +} + /* * Get the perf_counter_context for a task and lock it. * This has to cope with with the fact that until it is locked, @@ -283,6 +307,10 @@ counter_sched_out(struct perf_counter *counter, return; counter->state = PERF_COUNTER_STATE_INACTIVE; + if (counter->pending_disable) { + counter->pending_disable = 0; + counter->state = PERF_COUNTER_STATE_OFF; + } counter->tstamp_stopped = ctx->time; counter->pmu->disable(counter); counter->oncpu = -1; @@ -1081,7 +1109,7 @@ static void perf_counter_sync_stat(struct perf_counter_context *ctx, __perf_counter_sync_stat(counter, next_counter); counter = list_next_entry(counter, event_entry); - next_counter = list_next_entry(counter, event_entry); + next_counter = list_next_entry(next_counter, event_entry); } } @@ -1288,7 +1316,6 @@ static void perf_counter_cpu_sched_in(struct perf_cpu_context *cpuctx, int cpu) #define MAX_INTERRUPTS (~0ULL) static void perf_log_throttle(struct perf_counter *counter, int enable); -static void perf_log_period(struct perf_counter *counter, u64 period); static void perf_adjust_period(struct perf_counter *counter, u64 events) { @@ -1307,8 +1334,6 @@ static void perf_adjust_period(struct perf_counter *counter, u64 events) if (!sample_period) sample_period = 1; - perf_log_period(counter, sample_period); - hwc->sample_period = sample_period; } @@ -1463,10 +1488,8 @@ static void perf_counter_enable_on_exec(struct task_struct *task) /* * Unclone this context if we enabled any counter. */ - if (enabled && ctx->parent_ctx) { - put_ctx(ctx->parent_ctx); - ctx->parent_ctx = NULL; - } + if (enabled) + unclone_ctx(ctx); spin_unlock(&ctx->lock); @@ -1526,7 +1549,6 @@ __perf_counter_init_context(struct perf_counter_context *ctx, static struct perf_counter_context *find_get_context(pid_t pid, int cpu) { - struct perf_counter_context *parent_ctx; struct perf_counter_context *ctx; struct perf_cpu_context *cpuctx; struct task_struct *task; @@ -1586,11 +1608,7 @@ static struct perf_counter_context *find_get_context(pid_t pid, int cpu) retry: ctx = perf_lock_task_context(task, &flags); if (ctx) { - parent_ctx = ctx->parent_ctx; - if (parent_ctx) { - put_ctx(parent_ctx); - ctx->parent_ctx = NULL; /* no longer a clone */ - } + unclone_ctx(ctx); spin_unlock_irqrestore(&ctx->lock, flags); } @@ -1642,6 +1660,8 @@ static void free_counter(struct perf_counter *counter) atomic_dec(&nr_mmap_counters); if (counter->attr.comm) atomic_dec(&nr_comm_counters); + if (counter->attr.task) + atomic_dec(&nr_task_counters); } if (counter->destroy) @@ -1676,14 +1696,133 @@ static int perf_release(struct inode *inode, struct file *file) return 0; } +static int perf_counter_read_size(struct perf_counter *counter) +{ + int entry = sizeof(u64); /* value */ + int size = 0; + int nr = 1; + + if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) + size += sizeof(u64); + + if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) + size += sizeof(u64); + + if (counter->attr.read_format & PERF_FORMAT_ID) + entry += sizeof(u64); + + if (counter->attr.read_format & PERF_FORMAT_GROUP) { + nr += counter->group_leader->nr_siblings; + size += sizeof(u64); + } + + size += entry * nr; + + return size; +} + +static u64 perf_counter_read_value(struct perf_counter *counter) +{ + struct perf_counter *child; + u64 total = 0; + + total += perf_counter_read(counter); + list_for_each_entry(child, &counter->child_list, child_list) + total += perf_counter_read(child); + + return total; +} + +static int perf_counter_read_entry(struct perf_counter *counter, + u64 read_format, char __user *buf) +{ + int n = 0, count = 0; + u64 values[2]; + + values[n++] = perf_counter_read_value(counter); + if (read_format & PERF_FORMAT_ID) + values[n++] = primary_counter_id(counter); + + count = n * sizeof(u64); + + if (copy_to_user(buf, values, count)) + return -EFAULT; + + return count; +} + +static int perf_counter_read_group(struct perf_counter *counter, + u64 read_format, char __user *buf) +{ + struct perf_counter *leader = counter->group_leader, *sub; + int n = 0, size = 0, err = -EFAULT; + u64 values[3]; + + values[n++] = 1 + leader->nr_siblings; + if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) { + values[n++] = leader->total_time_enabled + + atomic64_read(&leader->child_total_time_enabled); + } + if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) { + values[n++] = leader->total_time_running + + atomic64_read(&leader->child_total_time_running); + } + + size = n * sizeof(u64); + + if (copy_to_user(buf, values, size)) + return -EFAULT; + + err = perf_counter_read_entry(leader, read_format, buf + size); + if (err < 0) + return err; + + size += err; + + list_for_each_entry(sub, &leader->sibling_list, list_entry) { + err = perf_counter_read_entry(counter, read_format, + buf + size); + if (err < 0) + return err; + + size += err; + } + + return size; +} + +static int perf_counter_read_one(struct perf_counter *counter, + u64 read_format, char __user *buf) +{ + u64 values[4]; + int n = 0; + + values[n++] = perf_counter_read_value(counter); + if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) { + values[n++] = counter->total_time_enabled + + atomic64_read(&counter->child_total_time_enabled); + } + if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) { + values[n++] = counter->total_time_running + + atomic64_read(&counter->child_total_time_running); + } + if (read_format & PERF_FORMAT_ID) + values[n++] = primary_counter_id(counter); + + if (copy_to_user(buf, values, n * sizeof(u64))) + return -EFAULT; + + return n * sizeof(u64); +} + /* * Read the performance counter - simple non blocking version for now */ static ssize_t perf_read_hw(struct perf_counter *counter, char __user *buf, size_t count) { - u64 values[4]; - int n; + u64 read_format = counter->attr.read_format; + int ret; /* * Return end-of-file for a read on a counter that is in @@ -1693,28 +1832,18 @@ perf_read_hw(struct perf_counter *counter, char __user *buf, size_t count) if (counter->state == PERF_COUNTER_STATE_ERROR) return 0; + if (count < perf_counter_read_size(counter)) + return -ENOSPC; + WARN_ON_ONCE(counter->ctx->parent_ctx); mutex_lock(&counter->child_mutex); - values[0] = perf_counter_read(counter); - n = 1; - if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) - values[n++] = counter->total_time_enabled + - atomic64_read(&counter->child_total_time_enabled); - if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) - values[n++] = counter->total_time_running + - atomic64_read(&counter->child_total_time_running); - if (counter->attr.read_format & PERF_FORMAT_ID) - values[n++] = counter->id; + if (read_format & PERF_FORMAT_GROUP) + ret = perf_counter_read_group(counter, read_format, buf); + else + ret = perf_counter_read_one(counter, read_format, buf); mutex_unlock(&counter->child_mutex); - if (count < n * sizeof(u64)) - return -EINVAL; - count = n * sizeof(u64); - - if (copy_to_user(buf, values, count)) - return -EFAULT; - - return count; + return ret; } static ssize_t @@ -1811,8 +1940,6 @@ static int perf_counter_period(struct perf_counter *counter, u64 __user *arg) counter->attr.sample_freq = value; } else { - perf_log_period(counter, value); - counter->attr.sample_period = value; counter->hw.sample_period = value; } @@ -2020,7 +2147,7 @@ fail: static void perf_mmap_free_page(unsigned long addr) { - struct page *page = virt_to_page(addr); + struct page *page = virt_to_page((void *)addr); page->mapping = NULL; __free_page(page); @@ -2220,7 +2347,7 @@ static void perf_pending_counter(struct perf_pending_entry *entry) if (counter->pending_disable) { counter->pending_disable = 0; - perf_counter_disable(counter); + __perf_counter_disable(counter); } if (counter->pending_wakeup) { @@ -2605,7 +2732,80 @@ static u32 perf_counter_tid(struct perf_counter *counter, struct task_struct *p) return task_pid_nr_ns(p, counter->ns); } -static void perf_counter_output(struct perf_counter *counter, int nmi, +static void perf_output_read_one(struct perf_output_handle *handle, + struct perf_counter *counter) +{ + u64 read_format = counter->attr.read_format; + u64 values[4]; + int n = 0; + + values[n++] = atomic64_read(&counter->count); + if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) { + values[n++] = counter->total_time_enabled + + atomic64_read(&counter->child_total_time_enabled); + } + if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) { + values[n++] = counter->total_time_running + + atomic64_read(&counter->child_total_time_running); + } + if (read_format & PERF_FORMAT_ID) + values[n++] = primary_counter_id(counter); + + perf_output_copy(handle, values, n * sizeof(u64)); +} + +/* + * XXX PERF_FORMAT_GROUP vs inherited counters seems difficult. + */ +static void perf_output_read_group(struct perf_output_handle *handle, + struct perf_counter *counter) +{ + struct perf_counter *leader = counter->group_leader, *sub; + u64 read_format = counter->attr.read_format; + u64 values[5]; + int n = 0; + + values[n++] = 1 + leader->nr_siblings; + + if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) + values[n++] = leader->total_time_enabled; + + if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) + values[n++] = leader->total_time_running; + + if (leader != counter) + leader->pmu->read(leader); + + values[n++] = atomic64_read(&leader->count); + if (read_format & PERF_FORMAT_ID) + values[n++] = primary_counter_id(leader); + + perf_output_copy(handle, values, n * sizeof(u64)); + + list_for_each_entry(sub, &leader->sibling_list, list_entry) { + n = 0; + + if (sub != counter) + sub->pmu->read(sub); + + values[n++] = atomic64_read(&sub->count); + if (read_format & PERF_FORMAT_ID) + values[n++] = primary_counter_id(sub); + + perf_output_copy(handle, values, n * sizeof(u64)); + } +} + +static void perf_output_read(struct perf_output_handle *handle, + struct perf_counter *counter) +{ + if (counter->attr.read_format & PERF_FORMAT_GROUP) + perf_output_read_group(handle, counter); + else + perf_output_read_one(handle, counter); +} + +void perf_counter_output(struct perf_counter *counter, int nmi, struct perf_sample_data *data) { int ret; @@ -2616,10 +2816,6 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, struct { u32 pid, tid; } tid_entry; - struct { - u64 id; - u64 counter; - } group_entry; struct perf_callchain_entry *callchain = NULL; int callchain_size = 0; u64 time; @@ -2661,19 +2857,21 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, if (sample_type & PERF_SAMPLE_ID) header.size += sizeof(u64); + if (sample_type & PERF_SAMPLE_STREAM_ID) + header.size += sizeof(u64); + if (sample_type & PERF_SAMPLE_CPU) { header.size += sizeof(cpu_entry); cpu_entry.cpu = raw_smp_processor_id(); + cpu_entry.reserved = 0; } if (sample_type & PERF_SAMPLE_PERIOD) header.size += sizeof(u64); - if (sample_type & PERF_SAMPLE_GROUP) { - header.size += sizeof(u64) + - counter->nr_siblings * sizeof(group_entry); - } + if (sample_type & PERF_SAMPLE_READ) + header.size += perf_counter_read_size(counter); if (sample_type & PERF_SAMPLE_CALLCHAIN) { callchain = perf_callchain(data->regs); @@ -2685,6 +2883,18 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, header.size += sizeof(u64); } + if (sample_type & PERF_SAMPLE_RAW) { + int size = sizeof(u32); + + if (data->raw) + size += data->raw->size; + else + size += sizeof(u32); + + WARN_ON_ONCE(size & (sizeof(u64)-1)); + header.size += size; + } + ret = perf_output_begin(&handle, counter, header.size, nmi, 1); if (ret) return; @@ -2703,7 +2913,13 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, if (sample_type & PERF_SAMPLE_ADDR) perf_output_put(&handle, data->addr); - if (sample_type & PERF_SAMPLE_ID) + if (sample_type & PERF_SAMPLE_ID) { + u64 id = primary_counter_id(counter); + + perf_output_put(&handle, id); + } + + if (sample_type & PERF_SAMPLE_STREAM_ID) perf_output_put(&handle, counter->id); if (sample_type & PERF_SAMPLE_CPU) @@ -2712,26 +2928,8 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, if (sample_type & PERF_SAMPLE_PERIOD) perf_output_put(&handle, data->period); - /* - * XXX PERF_SAMPLE_GROUP vs inherited counters seems difficult. - */ - if (sample_type & PERF_SAMPLE_GROUP) { - struct perf_counter *leader, *sub; - u64 nr = counter->nr_siblings; - - perf_output_put(&handle, nr); - - leader = counter->group_leader; - list_for_each_entry(sub, &leader->sibling_list, list_entry) { - if (sub != counter) - sub->pmu->read(sub); - - group_entry.id = sub->id; - group_entry.counter = atomic64_read(&sub->count); - - perf_output_put(&handle, group_entry); - } - } + if (sample_type & PERF_SAMPLE_READ) + perf_output_read(&handle, counter); if (sample_type & PERF_SAMPLE_CALLCHAIN) { if (callchain) @@ -2742,6 +2940,22 @@ static void perf_counter_output(struct perf_counter *counter, int nmi, } } + if (sample_type & PERF_SAMPLE_RAW) { + if (data->raw) { + perf_output_put(&handle, data->raw->size); + perf_output_copy(&handle, data->raw->data, data->raw->size); + } else { + struct { + u32 size; + u32 data; + } raw = { + .size = sizeof(u32), + .data = 0, + }; + perf_output_put(&handle, raw); + } + } + perf_output_end(&handle); } @@ -2754,8 +2968,6 @@ struct perf_read_event { u32 pid; u32 tid; - u64 value; - u64 format[3]; }; static void @@ -2767,87 +2979,74 @@ perf_counter_read_event(struct perf_counter *counter, .header = { .type = PERF_EVENT_READ, .misc = 0, - .size = sizeof(event) - sizeof(event.format), + .size = sizeof(event) + perf_counter_read_size(counter), }, .pid = perf_counter_pid(counter, task), .tid = perf_counter_tid(counter, task), - .value = atomic64_read(&counter->count), }; - int ret, i = 0; - - if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) { - event.header.size += sizeof(u64); - event.format[i++] = counter->total_time_enabled; - } - - if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) { - event.header.size += sizeof(u64); - event.format[i++] = counter->total_time_running; - } - - if (counter->attr.read_format & PERF_FORMAT_ID) { - u64 id; - - event.header.size += sizeof(u64); - if (counter->parent) - id = counter->parent->id; - else - id = counter->id; - - event.format[i++] = id; - } + int ret; ret = perf_output_begin(&handle, counter, event.header.size, 0, 0); if (ret) return; - perf_output_copy(&handle, &event, event.header.size); + perf_output_put(&handle, event); + perf_output_read(&handle, counter); + perf_output_end(&handle); } /* - * fork tracking + * task tracking -- fork/exit + * + * enabled by: attr.comm | attr.mmap | attr.task */ -struct perf_fork_event { - struct task_struct *task; +struct perf_task_event { + struct task_struct *task; + struct perf_counter_context *task_ctx; struct { struct perf_event_header header; u32 pid; u32 ppid; + u32 tid; + u32 ptid; } event; }; -static void perf_counter_fork_output(struct perf_counter *counter, - struct perf_fork_event *fork_event) +static void perf_counter_task_output(struct perf_counter *counter, + struct perf_task_event *task_event) { struct perf_output_handle handle; - int size = fork_event->event.header.size; - struct task_struct *task = fork_event->task; + int size = task_event->event.header.size; + struct task_struct *task = task_event->task; int ret = perf_output_begin(&handle, counter, size, 0, 0); if (ret) return; - fork_event->event.pid = perf_counter_pid(counter, task); - fork_event->event.ppid = perf_counter_pid(counter, task->real_parent); + task_event->event.pid = perf_counter_pid(counter, task); + task_event->event.ppid = perf_counter_pid(counter, current); + + task_event->event.tid = perf_counter_tid(counter, task); + task_event->event.ptid = perf_counter_tid(counter, current); - perf_output_put(&handle, fork_event->event); + perf_output_put(&handle, task_event->event); perf_output_end(&handle); } -static int perf_counter_fork_match(struct perf_counter *counter) +static int perf_counter_task_match(struct perf_counter *counter) { - if (counter->attr.comm || counter->attr.mmap) + if (counter->attr.comm || counter->attr.mmap || counter->attr.task) return 1; return 0; } -static void perf_counter_fork_ctx(struct perf_counter_context *ctx, - struct perf_fork_event *fork_event) +static void perf_counter_task_ctx(struct perf_counter_context *ctx, + struct perf_task_event *task_event) { struct perf_counter *counter; @@ -2856,51 +3055,62 @@ static void perf_counter_fork_ctx(struct perf_counter_context *ctx, rcu_read_lock(); list_for_each_entry_rcu(counter, &ctx->event_list, event_entry) { - if (perf_counter_fork_match(counter)) - perf_counter_fork_output(counter, fork_event); + if (perf_counter_task_match(counter)) + perf_counter_task_output(counter, task_event); } rcu_read_unlock(); } -static void perf_counter_fork_event(struct perf_fork_event *fork_event) +static void perf_counter_task_event(struct perf_task_event *task_event) { struct perf_cpu_context *cpuctx; - struct perf_counter_context *ctx; + struct perf_counter_context *ctx = task_event->task_ctx; cpuctx = &get_cpu_var(perf_cpu_context); - perf_counter_fork_ctx(&cpuctx->ctx, fork_event); + perf_counter_task_ctx(&cpuctx->ctx, task_event); put_cpu_var(perf_cpu_context); rcu_read_lock(); - /* - * doesn't really matter which of the child contexts the - * events ends up in. - */ - ctx = rcu_dereference(current->perf_counter_ctxp); + if (!ctx) + ctx = rcu_dereference(task_event->task->perf_counter_ctxp); if (ctx) - perf_counter_fork_ctx(ctx, fork_event); + perf_counter_task_ctx(ctx, task_event); rcu_read_unlock(); } -void perf_counter_fork(struct task_struct *task) +static void perf_counter_task(struct task_struct *task, + struct perf_counter_context *task_ctx, + int new) { - struct perf_fork_event fork_event; + struct perf_task_event task_event; if (!atomic_read(&nr_comm_counters) && - !atomic_read(&nr_mmap_counters)) + !atomic_read(&nr_mmap_counters) && + !atomic_read(&nr_task_counters)) return; - fork_event = (struct perf_fork_event){ - .task = task, - .event = { + task_event = (struct perf_task_event){ + .task = task, + .task_ctx = task_ctx, + .event = { .header = { - .type = PERF_EVENT_FORK, - .size = sizeof(fork_event.event), + .type = new ? PERF_EVENT_FORK : PERF_EVENT_EXIT, + .misc = 0, + .size = sizeof(task_event.event), }, + /* .pid */ + /* .ppid */ + /* .tid */ + /* .ptid */ }, }; - perf_counter_fork_event(&fork_event); + perf_counter_task_event(&task_event); +} + +void perf_counter_fork(struct task_struct *task) +{ + perf_counter_task(task, NULL, 1); } /* @@ -2968,8 +3178,10 @@ static void perf_counter_comm_event(struct perf_comm_event *comm_event) struct perf_cpu_context *cpuctx; struct perf_counter_context *ctx; unsigned int size; - char *comm = comm_event->task->comm; + char comm[TASK_COMM_LEN]; + memset(comm, 0, sizeof(comm)); + strncpy(comm, comm_event->task->comm, sizeof(comm)); size = ALIGN(strlen(comm)+1, sizeof(u64)); comm_event->comm = comm; @@ -3004,8 +3216,16 @@ void perf_counter_comm(struct task_struct *task) comm_event = (struct perf_comm_event){ .task = task, + /* .comm */ + /* .comm_size */ .event = { - .header = { .type = PERF_EVENT_COMM, }, + .header = { + .type = PERF_EVENT_COMM, + .misc = 0, + /* .size */ + }, + /* .pid */ + /* .tid */ }, }; @@ -3088,8 +3308,15 @@ static void perf_counter_mmap_event(struct perf_mmap_event *mmap_event) char *buf = NULL; const char *name; + memset(tmp, 0, sizeof(tmp)); + if (file) { - buf = kzalloc(PATH_MAX, GFP_KERNEL); + /* + * d_path works from the end of the buffer backwards, so we + * need to add enough zero bytes after the string to handle + * the 64bit alignment we do later. + */ + buf = kzalloc(PATH_MAX + sizeof(u64), GFP_KERNEL); if (!buf) { name = strncpy(tmp, "//enomem", sizeof(tmp)); goto got_name; @@ -3100,9 +3327,11 @@ static void perf_counter_mmap_event(struct perf_mmap_event *mmap_event) goto got_name; } } else { - name = arch_vma_name(mmap_event->vma); - if (name) + if (arch_vma_name(mmap_event->vma)) { + name = strncpy(tmp, arch_vma_name(mmap_event->vma), + sizeof(tmp)); goto got_name; + } if (!vma->vm_mm) { name = strncpy(tmp, "[vdso]", sizeof(tmp)); @@ -3147,8 +3376,16 @@ void __perf_counter_mmap(struct vm_area_struct *vma) mmap_event = (struct perf_mmap_event){ .vma = vma, + /* .file_name */ + /* .file_size */ .event = { - .header = { .type = PERF_EVENT_MMAP, }, + .header = { + .type = PERF_EVENT_MMAP, + .misc = 0, + /* .size */ + }, + /* .pid */ + /* .tid */ .start = vma->vm_start, .len = vma->vm_end - vma->vm_start, .pgoff = vma->vm_pgoff, @@ -3159,49 +3396,6 @@ void __perf_counter_mmap(struct vm_area_struct *vma) } /* - * Log sample_period changes so that analyzing tools can re-normalize the - * event flow. - */ - -struct freq_event { - struct perf_event_header header; - u64 time; - u64 id; - u64 period; -}; - -static void perf_log_period(struct perf_counter *counter, u64 period) -{ - struct perf_output_handle handle; - struct freq_event event; - int ret; - - if (counter->hw.sample_period == period) - return; - - if (counter->attr.sample_type & PERF_SAMPLE_PERIOD) - return; - - event = (struct freq_event) { - .header = { - .type = PERF_EVENT_PERIOD, - .misc = 0, - .size = sizeof(event), - }, - .time = sched_clock(), - .id = counter->id, - .period = period, - }; - - ret = perf_output_begin(&handle, counter, sizeof(event), 1, 0); - if (ret) - return; - - perf_output_put(&handle, event); - perf_output_end(&handle); -} - -/* * IRQ throttle logging */ @@ -3214,16 +3408,21 @@ static void perf_log_throttle(struct perf_counter *counter, int enable) struct perf_event_header header; u64 time; u64 id; + u64 stream_id; } throttle_event = { .header = { - .type = PERF_EVENT_THROTTLE + 1, + .type = PERF_EVENT_THROTTLE, .misc = 0, .size = sizeof(throttle_event), }, - .time = sched_clock(), - .id = counter->id, + .time = sched_clock(), + .id = primary_counter_id(counter), + .stream_id = counter->id, }; + if (enable) + throttle_event.header.type = PERF_EVENT_UNTHROTTLE; + ret = perf_output_begin(&handle, counter, sizeof(throttle_event), 1, 0); if (ret) return; @@ -3300,125 +3499,111 @@ int perf_counter_overflow(struct perf_counter *counter, int nmi, * Generic software counter infrastructure */ -static void perf_swcounter_update(struct perf_counter *counter) +/* + * We directly increment counter->count and keep a second value in + * counter->hw.period_left to count intervals. This period counter + * is kept in the range [-sample_period, 0] so that we can use the + * sign as trigger. + */ + +static u64 perf_swcounter_set_period(struct perf_counter *counter) { struct hw_perf_counter *hwc = &counter->hw; - u64 prev, now; - s64 delta; + u64 period = hwc->last_period; + u64 nr, offset; + s64 old, val; + + hwc->last_period = hwc->sample_period; again: - prev = atomic64_read(&hwc->prev_count); - now = atomic64_read(&hwc->count); - if (atomic64_cmpxchg(&hwc->prev_count, prev, now) != prev) - goto again; + old = val = atomic64_read(&hwc->period_left); + if (val < 0) + return 0; - delta = now - prev; + nr = div64_u64(period + val, period); + offset = nr * period; + val -= offset; + if (atomic64_cmpxchg(&hwc->period_left, old, val) != old) + goto again; - atomic64_add(delta, &counter->count); - atomic64_sub(delta, &hwc->period_left); + return nr; } -static void perf_swcounter_set_period(struct perf_counter *counter) +static void perf_swcounter_overflow(struct perf_counter *counter, + int nmi, struct perf_sample_data *data) { struct hw_perf_counter *hwc = &counter->hw; - s64 left = atomic64_read(&hwc->period_left); - s64 period = hwc->sample_period; + u64 overflow; - if (unlikely(left <= -period)) { - left = period; - atomic64_set(&hwc->period_left, left); - hwc->last_period = period; - } + data->period = counter->hw.last_period; + overflow = perf_swcounter_set_period(counter); - if (unlikely(left <= 0)) { - left += period; - atomic64_add(period, &hwc->period_left); - hwc->last_period = period; - } + if (hwc->interrupts == MAX_INTERRUPTS) + return; - atomic64_set(&hwc->prev_count, -left); - atomic64_set(&hwc->count, -left); + for (; overflow; overflow--) { + if (perf_counter_overflow(counter, nmi, data)) { + /* + * We inhibit the overflow from happening when + * hwc->interrupts == MAX_INTERRUPTS. + */ + break; + } + } } -static enum hrtimer_restart perf_swcounter_hrtimer(struct hrtimer *hrtimer) +static void perf_swcounter_unthrottle(struct perf_counter *counter) { - enum hrtimer_restart ret = HRTIMER_RESTART; - struct perf_sample_data data; - struct perf_counter *counter; - u64 period; - - counter = container_of(hrtimer, struct perf_counter, hw.hrtimer); - counter->pmu->read(counter); - - data.addr = 0; - data.regs = get_irq_regs(); /* - * In case we exclude kernel IPs or are somehow not in interrupt - * context, provide the next best thing, the user IP. + * Nothing to do, we already reset hwc->interrupts. */ - if ((counter->attr.exclude_kernel || !data.regs) && - !counter->attr.exclude_user) - data.regs = task_pt_regs(current); +} - if (data.regs) { - if (perf_counter_overflow(counter, 0, &data)) - ret = HRTIMER_NORESTART; - } +static void perf_swcounter_add(struct perf_counter *counter, u64 nr, + int nmi, struct perf_sample_data *data) +{ + struct hw_perf_counter *hwc = &counter->hw; - period = max_t(u64, 10000, counter->hw.sample_period); - hrtimer_forward_now(hrtimer, ns_to_ktime(period)); + atomic64_add(nr, &counter->count); - return ret; -} + if (!hwc->sample_period) + return; -static void perf_swcounter_overflow(struct perf_counter *counter, - int nmi, struct perf_sample_data *data) -{ - data->period = counter->hw.last_period; + if (!data->regs) + return; - perf_swcounter_update(counter); - perf_swcounter_set_period(counter); - if (perf_counter_overflow(counter, nmi, data)) - /* soft-disable the counter */ - ; + if (!atomic64_add_negative(nr, &hwc->period_left)) + perf_swcounter_overflow(counter, nmi, data); } static int perf_swcounter_is_counting(struct perf_counter *counter) { - struct perf_counter_context *ctx; - unsigned long flags; - int count; - + /* + * The counter is active, we're good! + */ if (counter->state == PERF_COUNTER_STATE_ACTIVE) return 1; + /* + * The counter is off/error, not counting. + */ if (counter->state != PERF_COUNTER_STATE_INACTIVE) return 0; /* - * If the counter is inactive, it could be just because - * its task is scheduled out, or because it's in a group - * which could not go on the PMU. We want to count in - * the first case but not the second. If the context is - * currently active then an inactive software counter must - * be the second case. If it's not currently active then - * we need to know whether the counter was active when the - * context was last active, which we can determine by - * comparing counter->tstamp_stopped with ctx->time. - * - * We are within an RCU read-side critical section, - * which protects the existence of *ctx. + * The counter is inactive, if the context is active + * we're part of a group that didn't make it on the 'pmu', + * not counting. */ - ctx = counter->ctx; - spin_lock_irqsave(&ctx->lock, flags); - count = 1; - /* Re-check state now we have the lock */ - if (counter->state < PERF_COUNTER_STATE_INACTIVE || - counter->ctx->is_active || - counter->tstamp_stopped < ctx->time) - count = 0; - spin_unlock_irqrestore(&ctx->lock, flags); - return count; + if (counter->ctx->is_active) + return 0; + + /* + * We're inactive and the context is too, this means the + * task is scheduled out, we're counting events that happen + * to us, like migration events. + */ + return 1; } static int perf_swcounter_match(struct perf_counter *counter, @@ -3444,15 +3629,6 @@ static int perf_swcounter_match(struct perf_counter *counter, return 1; } -static void perf_swcounter_add(struct perf_counter *counter, u64 nr, - int nmi, struct perf_sample_data *data) -{ - int neg = atomic64_add_negative(nr, &counter->hw.count); - - if (counter->hw.sample_period && !neg && data->regs) - perf_swcounter_overflow(counter, nmi, data); -} - static void perf_swcounter_ctx_event(struct perf_counter_context *ctx, enum perf_type_id type, u32 event, u64 nr, int nmi, @@ -3531,27 +3707,66 @@ void __perf_swcounter_event(u32 event, u64 nr, int nmi, static void perf_swcounter_read(struct perf_counter *counter) { - perf_swcounter_update(counter); } static int perf_swcounter_enable(struct perf_counter *counter) { - perf_swcounter_set_period(counter); + struct hw_perf_counter *hwc = &counter->hw; + + if (hwc->sample_period) { + hwc->last_period = hwc->sample_period; + perf_swcounter_set_period(counter); + } return 0; } static void perf_swcounter_disable(struct perf_counter *counter) { - perf_swcounter_update(counter); } static const struct pmu perf_ops_generic = { .enable = perf_swcounter_enable, .disable = perf_swcounter_disable, .read = perf_swcounter_read, + .unthrottle = perf_swcounter_unthrottle, }; /* + * hrtimer based swcounter callback + */ + +static enum hrtimer_restart perf_swcounter_hrtimer(struct hrtimer *hrtimer) +{ + enum hrtimer_restart ret = HRTIMER_RESTART; + struct perf_sample_data data; + struct perf_counter *counter; + u64 period; + + counter = container_of(hrtimer, struct perf_counter, hw.hrtimer); + counter->pmu->read(counter); + + data.addr = 0; + data.regs = get_irq_regs(); + /* + * In case we exclude kernel IPs or are somehow not in interrupt + * context, provide the next best thing, the user IP. + */ + if ((counter->attr.exclude_kernel || !data.regs) && + !counter->attr.exclude_user) + data.regs = task_pt_regs(current); + + if (data.regs) { + if (perf_counter_overflow(counter, 0, &data)) + ret = HRTIMER_NORESTART; + } + + period = max_t(u64, 10000, counter->hw.sample_period); + hrtimer_forward_now(hrtimer, ns_to_ktime(period)); + + return ret; +} + +/* * Software counter: cpu wall time clock */ @@ -3668,17 +3883,24 @@ static const struct pmu perf_ops_task_clock = { }; #ifdef CONFIG_EVENT_PROFILE -void perf_tpcounter_event(int event_id) +void perf_tpcounter_event(int event_id, u64 addr, u64 count, void *record, + int entry_size) { + struct perf_raw_record raw = { + .size = entry_size, + .data = record, + }; + struct perf_sample_data data = { - .regs = get_irq_regs(); - .addr = 0, + .regs = get_irq_regs(), + .addr = addr, + .raw = &raw, }; if (!data.regs) data.regs = task_pt_regs(current); - do_perf_swcounter_event(PERF_TYPE_TRACEPOINT, event_id, 1, 1, &data); + do_perf_swcounter_event(PERF_TYPE_TRACEPOINT, event_id, count, 1, &data); } EXPORT_SYMBOL_GPL(perf_tpcounter_event); @@ -3687,16 +3909,20 @@ extern void ftrace_profile_disable(int); static void tp_perf_counter_destroy(struct perf_counter *counter) { - ftrace_profile_disable(perf_event_id(&counter->attr)); + ftrace_profile_disable(counter->attr.config); } static const struct pmu *tp_perf_counter_init(struct perf_counter *counter) { - int event_id = perf_event_id(&counter->attr); - int ret; + /* + * Raw tracepoint data is a severe data leak, only allow root to + * have these. + */ + if ((counter->attr.sample_type & PERF_SAMPLE_RAW) && + !capable(CAP_SYS_ADMIN)) + return ERR_PTR(-EPERM); - ret = ftrace_profile_enable(event_id); - if (ret) + if (ftrace_profile_enable(counter->attr.config)) return NULL; counter->destroy = tp_perf_counter_destroy; @@ -3829,9 +4055,9 @@ perf_counter_alloc(struct perf_counter_attr *attr, atomic64_set(&hwc->period_left, hwc->sample_period); /* - * we currently do not support PERF_SAMPLE_GROUP on inherited counters + * we currently do not support PERF_FORMAT_GROUP on inherited counters */ - if (attr->inherit && (attr->sample_type & PERF_SAMPLE_GROUP)) + if (attr->inherit && (attr->read_format & PERF_FORMAT_GROUP)) goto done; switch (attr->type) { @@ -3874,6 +4100,8 @@ done: atomic_inc(&nr_mmap_counters); if (counter->attr.comm) atomic_inc(&nr_comm_counters); + if (counter->attr.task) + atomic_inc(&nr_task_counters); } return counter; @@ -4235,8 +4463,10 @@ void perf_counter_exit_task(struct task_struct *child) struct perf_counter_context *child_ctx; unsigned long flags; - if (likely(!child->perf_counter_ctxp)) + if (likely(!child->perf_counter_ctxp)) { + perf_counter_task(child, NULL, 0); return; + } local_irq_save(flags); /* @@ -4255,17 +4485,20 @@ void perf_counter_exit_task(struct task_struct *child) */ spin_lock(&child_ctx->lock); child->perf_counter_ctxp = NULL; - if (child_ctx->parent_ctx) { - /* - * This context is a clone; unclone it so it can't get - * swapped to another process while we're removing all - * the counters from it. - */ - put_ctx(child_ctx->parent_ctx); - child_ctx->parent_ctx = NULL; - } - spin_unlock(&child_ctx->lock); - local_irq_restore(flags); + /* + * If this context is a clone; unclone it so it can't get + * swapped to another process while we're removing all + * the counters from it. + */ + unclone_ctx(child_ctx); + spin_unlock_irqrestore(&child_ctx->lock, flags); + + /* + * Report the task dead after unscheduling the counters so that we + * won't get any samples after PERF_EVENT_EXIT. We can however still + * get a few PERF_EVENT_READ events. + */ + perf_counter_task(child, child_ctx, 0); /* * We can recurse on the same lock type through: @@ -4486,6 +4719,11 @@ perf_cpu_notify(struct notifier_block *self, unsigned long action, void *hcpu) perf_counter_init_cpu(cpu); break; + case CPU_ONLINE: + case CPU_ONLINE_FROZEN: + hw_perf_counter_setup_online(cpu); + break; + case CPU_DOWN_PREPARE: case CPU_DOWN_PREPARE_FROZEN: perf_counter_exit_cpu(cpu); @@ -4510,6 +4748,8 @@ void __init perf_counter_init(void) { perf_cpu_notify(&perf_cpu_nb, (unsigned long)CPU_UP_PREPARE, (void *)(long)smp_processor_id()); + perf_cpu_notify(&perf_cpu_nb, (unsigned long)CPU_ONLINE, + (void *)(long)smp_processor_id()); register_cpu_notifier(&perf_cpu_nb); } diff --git a/kernel/pid.c b/kernel/pid.c index 5fa1db4..31310b5 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -36,7 +36,6 @@ #include <linux/pid_namespace.h> #include <linux/init_task.h> #include <linux/syscalls.h> -#include <linux/kmemleak.h> #define pid_hashfn(nr, ns) \ hash_long((unsigned long)nr + (unsigned long)ns, pidhash_shift) @@ -513,12 +512,6 @@ void __init pidhash_init(void) pid_hash = alloc_bootmem(pidhash_size * sizeof(*(pid_hash))); if (!pid_hash) panic("Could not alloc pidhash!\n"); - /* - * pid_hash contains references to allocated struct pid objects and it - * must be scanned by kmemleak to avoid false positives. - */ - kmemleak_alloc(pid_hash, pidhash_size * sizeof(*(pid_hash)), 0, - GFP_KERNEL); for (i = 0; i < pidhash_size; i++) INIT_HLIST_HEAD(&pid_hash[i]); } diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index bece7c0..e33a21c 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -521,11 +521,12 @@ void posix_cpu_timers_exit(struct task_struct *tsk) } void posix_cpu_timers_exit_group(struct task_struct *tsk) { - struct task_cputime cputime; + struct signal_struct *const sig = tsk->signal; - thread_group_cputimer(tsk, &cputime); cleanup_timers(tsk->signal->cpu_timers, - cputime.utime, cputime.stime, cputime.sum_exec_runtime); + cputime_add(tsk->utime, sig->utime), + cputime_add(tsk->stime, sig->stime), + tsk->se.sum_exec_runtime + sig->sum_sched_runtime); } static void clear_dead_task(struct k_itimer *timer, union cpu_time_count now) diff --git a/kernel/posix-timers.c b/kernel/posix-timers.c index 052ec4d..d089d05 100644 --- a/kernel/posix-timers.c +++ b/kernel/posix-timers.c @@ -202,6 +202,12 @@ static int no_timer_create(struct k_itimer *new_timer) return -EOPNOTSUPP; } +static int no_nsleep(const clockid_t which_clock, int flags, + struct timespec *tsave, struct timespec __user *rmtp) +{ + return -EOPNOTSUPP; +} + /* * Return nonzero if we know a priori this clockid_t value is bogus. */ @@ -254,6 +260,7 @@ static __init int init_posix_timers(void) .clock_get = posix_get_monotonic_raw, .clock_set = do_posix_clock_nosettime, .timer_create = no_timer_create, + .nsleep = no_nsleep, }; register_posix_clock(CLOCK_REALTIME, &clock_realtime); diff --git a/kernel/power/user.c b/kernel/power/user.c index ed97375..bf0014d 100644 --- a/kernel/power/user.c +++ b/kernel/power/user.c @@ -23,7 +23,6 @@ #include <linux/console.h> #include <linux/cpu.h> #include <linux/freezer.h> -#include <linux/smp_lock.h> #include <scsi/scsi_scan.h> #include <asm/uaccess.h> diff --git a/kernel/profile.c b/kernel/profile.c index 69911b5..419250e 100644 --- a/kernel/profile.c +++ b/kernel/profile.c @@ -117,11 +117,12 @@ int __ref profile_init(void) cpumask_copy(prof_cpu_mask, cpu_possible_mask); - prof_buffer = kzalloc(buffer_bytes, GFP_KERNEL); + prof_buffer = kzalloc(buffer_bytes, GFP_KERNEL|__GFP_NOWARN); if (prof_buffer) return 0; - prof_buffer = alloc_pages_exact(buffer_bytes, GFP_KERNEL|__GFP_ZERO); + prof_buffer = alloc_pages_exact(buffer_bytes, + GFP_KERNEL|__GFP_ZERO|__GFP_NOWARN); if (prof_buffer) return 0; diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 61c78b2..082c320 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -181,8 +181,8 @@ int ptrace_attach(struct task_struct *task) * interference; SUID, SGID and LSM creds get determined differently * under ptrace. */ - retval = mutex_lock_interruptible(&task->cred_guard_mutex); - if (retval < 0) + retval = -ERESTARTNOINTR; + if (mutex_lock_interruptible(&task->cred_guard_mutex)) goto out; task_lock(task); diff --git a/kernel/rcutree.c b/kernel/rcutree.c index 0dccfbb..7717b95 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -1533,7 +1533,7 @@ void __init __rcu_init(void) int j; struct rcu_node *rnp; - printk(KERN_WARNING "Experimental hierarchical RCU implementation.\n"); + printk(KERN_INFO "Hierarchical RCU implementation.\n"); #ifdef CONFIG_RCU_CPU_STALL_DETECTOR printk(KERN_INFO "RCU-based detection of stalled CPUs is enabled.\n"); #endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ @@ -1546,7 +1546,6 @@ void __init __rcu_init(void) rcu_cpu_notify(&rcu_nb, CPU_UP_PREPARE, (void *)(long)i); /* Register notifier for non-boot CPUs */ register_cpu_notifier(&rcu_nb); - printk(KERN_WARNING "Experimental hierarchical RCU init done.\n"); } module_param(blimit, int, 0); diff --git a/kernel/rtmutex.c b/kernel/rtmutex.c index fcd107a..29bd4ba 100644 --- a/kernel/rtmutex.c +++ b/kernel/rtmutex.c @@ -1039,16 +1039,14 @@ int rt_mutex_start_proxy_lock(struct rt_mutex *lock, if (!rt_mutex_owner(lock) || try_to_steal_lock(lock, task)) { /* We got the lock for task. */ debug_rt_mutex_lock(lock); - rt_mutex_set_owner(lock, task, 0); - + spin_unlock(&lock->wait_lock); rt_mutex_deadlock_account_lock(lock, task); return 1; } ret = task_blocks_on_rt_mutex(lock, waiter, task, detect_deadlock); - if (ret && !waiter->task) { /* * Reset the return value. We might have diff --git a/kernel/sched.c b/kernel/sched.c index 7c9098d..1b59e26 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -493,6 +493,7 @@ struct rt_rq { #endif #ifdef CONFIG_SMP unsigned long rt_nr_migratory; + unsigned long rt_nr_total; int overloaded; struct plist_head pushable_tasks; #endif @@ -2571,15 +2572,37 @@ static void __sched_fork(struct task_struct *p) p->se.avg_wakeup = sysctl_sched_wakeup_granularity; #ifdef CONFIG_SCHEDSTATS - p->se.wait_start = 0; - p->se.sum_sleep_runtime = 0; - p->se.sleep_start = 0; - p->se.block_start = 0; - p->se.sleep_max = 0; - p->se.block_max = 0; - p->se.exec_max = 0; - p->se.slice_max = 0; - p->se.wait_max = 0; + p->se.wait_start = 0; + p->se.wait_max = 0; + p->se.wait_count = 0; + p->se.wait_sum = 0; + + p->se.sleep_start = 0; + p->se.sleep_max = 0; + p->se.sum_sleep_runtime = 0; + + p->se.block_start = 0; + p->se.block_max = 0; + p->se.exec_max = 0; + p->se.slice_max = 0; + + p->se.nr_migrations_cold = 0; + p->se.nr_failed_migrations_affine = 0; + p->se.nr_failed_migrations_running = 0; + p->se.nr_failed_migrations_hot = 0; + p->se.nr_forced_migrations = 0; + p->se.nr_forced2_migrations = 0; + + p->se.nr_wakeups = 0; + p->se.nr_wakeups_sync = 0; + p->se.nr_wakeups_migrate = 0; + p->se.nr_wakeups_local = 0; + p->se.nr_wakeups_remote = 0; + p->se.nr_wakeups_affine = 0; + p->se.nr_wakeups_affine_attempts = 0; + p->se.nr_wakeups_passive = 0; + p->se.nr_wakeups_idle = 0; + #endif INIT_LIST_HEAD(&p->rt.run_list); @@ -6541,6 +6564,11 @@ SYSCALL_DEFINE0(sched_yield) return 0; } +static inline int should_resched(void) +{ + return need_resched() && !(preempt_count() & PREEMPT_ACTIVE); +} + static void __cond_resched(void) { #ifdef CONFIG_DEBUG_SPINLOCK_SLEEP @@ -6560,8 +6588,7 @@ static void __cond_resched(void) int __sched _cond_resched(void) { - if (need_resched() && !(preempt_count() & PREEMPT_ACTIVE) && - system_state == SYSTEM_RUNNING) { + if (should_resched()) { __cond_resched(); return 1; } @@ -6579,12 +6606,12 @@ EXPORT_SYMBOL(_cond_resched); */ int cond_resched_lock(spinlock_t *lock) { - int resched = need_resched() && system_state == SYSTEM_RUNNING; + int resched = should_resched(); int ret = 0; if (spin_needbreak(lock) || resched) { spin_unlock(lock); - if (resched && need_resched()) + if (resched) __cond_resched(); else cpu_relax(); @@ -6599,7 +6626,7 @@ int __sched cond_resched_softirq(void) { BUG_ON(!in_softirq()); - if (need_resched() && system_state == SYSTEM_RUNNING) { + if (should_resched()) { local_bh_enable(); __cond_resched(); local_bh_disable(); @@ -7262,6 +7289,7 @@ static void migrate_dead_tasks(unsigned int dead_cpu) static void calc_global_load_remove(struct rq *rq) { atomic_long_sub(rq->calc_load_active, &calc_load_tasks); + rq->calc_load_active = 0; } #endif /* CONFIG_HOTPLUG_CPU */ @@ -7488,6 +7516,7 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) task_rq_unlock(rq, &flags); get_task_struct(p); cpu_rq(cpu)->migration_thread = p; + rq->calc_load_update = calc_load_update; break; case CPU_ONLINE: @@ -7498,8 +7527,6 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) /* Update our root-domain */ rq = cpu_rq(cpu); spin_lock_irqsave(&rq->lock, flags); - rq->calc_load_update = calc_load_update; - rq->calc_load_active = 0; if (rq->rd) { BUG_ON(!cpumask_test_cpu(cpu, rq->rd->span)); @@ -9070,7 +9097,7 @@ static void init_rt_rq(struct rt_rq *rt_rq, struct rq *rq) #ifdef CONFIG_SMP rt_rq->rt_nr_migratory = 0; rt_rq->overloaded = 0; - plist_head_init(&rq->rt.pushable_tasks, &rq->lock); + plist_head_init(&rt_rq->pushable_tasks, &rq->lock); #endif rt_rq->rt_time = 0; diff --git a/kernel/sched_cpupri.c b/kernel/sched_cpupri.c index e6c2517..d014efb 100644 --- a/kernel/sched_cpupri.c +++ b/kernel/sched_cpupri.c @@ -81,8 +81,21 @@ int cpupri_find(struct cpupri *cp, struct task_struct *p, if (cpumask_any_and(&p->cpus_allowed, vec->mask) >= nr_cpu_ids) continue; - if (lowest_mask) + if (lowest_mask) { cpumask_and(lowest_mask, &p->cpus_allowed, vec->mask); + + /* + * We have to ensure that we have at least one bit + * still set in the array, since the map could have + * been concurrently emptied between the first and + * second reads of vec->mask. If we hit this + * condition, simply act as though we never hit this + * priority level and continue on. + */ + if (cpumask_any(lowest_mask) >= nr_cpu_ids) + continue; + } + return 1; } diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index ba7fd6e..652e8bd 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -266,6 +266,12 @@ static inline u64 min_vruntime(u64 min_vruntime, u64 vruntime) return min_vruntime; } +static inline int entity_before(struct sched_entity *a, + struct sched_entity *b) +{ + return (s64)(a->vruntime - b->vruntime) < 0; +} + static inline s64 entity_key(struct cfs_rq *cfs_rq, struct sched_entity *se) { return se->vruntime - cfs_rq->min_vruntime; @@ -605,9 +611,13 @@ account_entity_dequeue(struct cfs_rq *cfs_rq, struct sched_entity *se) static void enqueue_sleeper(struct cfs_rq *cfs_rq, struct sched_entity *se) { #ifdef CONFIG_SCHEDSTATS + struct task_struct *tsk = NULL; + + if (entity_is_task(se)) + tsk = task_of(se); + if (se->sleep_start) { u64 delta = rq_of(cfs_rq)->clock - se->sleep_start; - struct task_struct *tsk = task_of(se); if ((s64)delta < 0) delta = 0; @@ -618,11 +628,11 @@ static void enqueue_sleeper(struct cfs_rq *cfs_rq, struct sched_entity *se) se->sleep_start = 0; se->sum_sleep_runtime += delta; - account_scheduler_latency(tsk, delta >> 10, 1); + if (tsk) + account_scheduler_latency(tsk, delta >> 10, 1); } if (se->block_start) { u64 delta = rq_of(cfs_rq)->clock - se->block_start; - struct task_struct *tsk = task_of(se); if ((s64)delta < 0) delta = 0; @@ -633,17 +643,19 @@ static void enqueue_sleeper(struct cfs_rq *cfs_rq, struct sched_entity *se) se->block_start = 0; se->sum_sleep_runtime += delta; - /* - * Blocking time is in units of nanosecs, so shift by 20 to - * get a milliseconds-range estimation of the amount of - * time that the task spent sleeping: - */ - if (unlikely(prof_on == SLEEP_PROFILING)) { - - profile_hits(SLEEP_PROFILING, (void *)get_wchan(tsk), - delta >> 20); + if (tsk) { + /* + * Blocking time is in units of nanosecs, so shift by + * 20 to get a milliseconds-range estimation of the + * amount of time that the task spent sleeping: + */ + if (unlikely(prof_on == SLEEP_PROFILING)) { + profile_hits(SLEEP_PROFILING, + (void *)get_wchan(tsk), + delta >> 20); + } + account_scheduler_latency(tsk, delta >> 10, 0); } - account_scheduler_latency(tsk, delta >> 10, 0); } #endif } @@ -687,7 +699,8 @@ place_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int initial) * all of which have the same weight. */ if (sched_feat(NORMALIZED_SLEEPER) && - task_of(se)->policy != SCHED_IDLE) + (!entity_is_task(se) || + task_of(se)->policy != SCHED_IDLE)) thresh = calc_delta_fair(thresh, se); vruntime -= thresh; @@ -1016,7 +1029,7 @@ static void yield_task_fair(struct rq *rq) /* * Already in the rightmost position? */ - if (unlikely(!rightmost || rightmost->vruntime < se->vruntime)) + if (unlikely(!rightmost || entity_before(rightmost, se))) return; /* @@ -1712,7 +1725,7 @@ static void task_new_fair(struct rq *rq, struct task_struct *p) /* 'curr' will be NULL if the child belongs to a different group */ if (sysctl_sched_child_runs_first && this_cpu == task_cpu(p) && - curr && curr->vruntime < se->vruntime) { + curr && entity_before(curr, se)) { /* * Upon rescheduling, sched_class::put_prev_task() will place * 'current' within the tree based on its new key value. diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c index 9bf0d2a..3918e01 100644 --- a/kernel/sched_rt.c +++ b/kernel/sched_rt.c @@ -10,6 +10,8 @@ static inline struct task_struct *rt_task_of(struct sched_rt_entity *rt_se) #ifdef CONFIG_RT_GROUP_SCHED +#define rt_entity_is_task(rt_se) (!(rt_se)->my_q) + static inline struct rq *rq_of_rt_rq(struct rt_rq *rt_rq) { return rt_rq->rq; @@ -22,6 +24,8 @@ static inline struct rt_rq *rt_rq_of_se(struct sched_rt_entity *rt_se) #else /* CONFIG_RT_GROUP_SCHED */ +#define rt_entity_is_task(rt_se) (1) + static inline struct rq *rq_of_rt_rq(struct rt_rq *rt_rq) { return container_of(rt_rq, struct rq, rt); @@ -73,7 +77,7 @@ static inline void rt_clear_overload(struct rq *rq) static void update_rt_migration(struct rt_rq *rt_rq) { - if (rt_rq->rt_nr_migratory && (rt_rq->rt_nr_running > 1)) { + if (rt_rq->rt_nr_migratory && rt_rq->rt_nr_total > 1) { if (!rt_rq->overloaded) { rt_set_overload(rq_of_rt_rq(rt_rq)); rt_rq->overloaded = 1; @@ -86,6 +90,12 @@ static void update_rt_migration(struct rt_rq *rt_rq) static void inc_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq) { + if (!rt_entity_is_task(rt_se)) + return; + + rt_rq = &rq_of_rt_rq(rt_rq)->rt; + + rt_rq->rt_nr_total++; if (rt_se->nr_cpus_allowed > 1) rt_rq->rt_nr_migratory++; @@ -94,6 +104,12 @@ static void inc_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq) static void dec_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq) { + if (!rt_entity_is_task(rt_se)) + return; + + rt_rq = &rq_of_rt_rq(rt_rq)->rt; + + rt_rq->rt_nr_total--; if (rt_se->nr_cpus_allowed > 1) rt_rq->rt_nr_migratory--; diff --git a/kernel/signal.c b/kernel/signal.c index ccf1cee..64c5dee 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -2454,11 +2454,9 @@ do_sigaltstack (const stack_t __user *uss, stack_t __user *uoss, unsigned long s stack_t oss; int error; - if (uoss) { - 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_sp = (void __user *) current->sas_ss_sp; + oss.ss_size = current->sas_ss_size; + oss.ss_flags = sas_ss_flags(sp); if (uss) { void __user *ss_sp; @@ -2466,10 +2464,12 @@ do_sigaltstack (const stack_t __user *uss, stack_t __user *uoss, unsigned long s int ss_flags; error = -EFAULT; - if (!access_ok(VERIFY_READ, uss, sizeof(*uss)) - || __get_user(ss_sp, &uss->ss_sp) - || __get_user(ss_flags, &uss->ss_flags) - || __get_user(ss_size, &uss->ss_size)) + if (!access_ok(VERIFY_READ, uss, sizeof(*uss))) + goto out; + error = __get_user(ss_sp, &uss->ss_sp) | + __get_user(ss_flags, &uss->ss_flags) | + __get_user(ss_size, &uss->ss_size); + if (error) goto out; error = -EPERM; @@ -2501,13 +2501,16 @@ do_sigaltstack (const stack_t __user *uss, stack_t __user *uoss, unsigned long s current->sas_ss_size = ss_size; } + error = 0; if (uoss) { error = -EFAULT; - if (copy_to_user(uoss, &oss, sizeof(oss))) + if (!access_ok(VERIFY_WRITE, uoss, sizeof(*uoss))) goto out; + error = __put_user(oss.ss_sp, &uoss->ss_sp) | + __put_user(oss.ss_size, &uoss->ss_size) | + __put_user(oss.ss_flags, &uoss->ss_flags); } - error = 0; out: return error; } diff --git a/kernel/smp.c b/kernel/smp.c index 2accdf6..8e21850 100644 --- a/kernel/smp.c +++ b/kernel/smp.c @@ -57,7 +57,7 @@ hotplug_cfd(struct notifier_block *nfb, unsigned long action, void *hcpu) return NOTIFY_BAD; break; -#ifdef CONFIG_CPU_HOTPLUG +#ifdef CONFIG_HOTPLUG_CPU case CPU_UP_CANCELED: case CPU_UP_CANCELED_FROZEN: diff --git a/kernel/softirq.c b/kernel/softirq.c index 3a94905..eb5e131 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -345,7 +345,9 @@ void open_softirq(int nr, void (*action)(struct softirq_action *)) softirq_vec[nr].action = action; } -/* Tasklets */ +/* + * Tasklets + */ struct tasklet_head { struct tasklet_struct *head; @@ -493,6 +495,66 @@ void tasklet_kill(struct tasklet_struct *t) EXPORT_SYMBOL(tasklet_kill); +/* + * tasklet_hrtimer + */ + +/* + * The trampoline is called when the hrtimer expires. If this is + * called from the hrtimer interrupt then we schedule the tasklet as + * the timer callback function expects to run in softirq context. If + * it's called in softirq context anyway (i.e. high resolution timers + * disabled) then the hrtimer callback is called right away. + */ +static enum hrtimer_restart __hrtimer_tasklet_trampoline(struct hrtimer *timer) +{ + struct tasklet_hrtimer *ttimer = + container_of(timer, struct tasklet_hrtimer, timer); + + if (hrtimer_is_hres_active(timer)) { + tasklet_hi_schedule(&ttimer->tasklet); + return HRTIMER_NORESTART; + } + return ttimer->function(timer); +} + +/* + * Helper function which calls the hrtimer callback from + * tasklet/softirq context + */ +static void __tasklet_hrtimer_trampoline(unsigned long data) +{ + struct tasklet_hrtimer *ttimer = (void *)data; + enum hrtimer_restart restart; + + restart = ttimer->function(&ttimer->timer); + if (restart != HRTIMER_NORESTART) + hrtimer_restart(&ttimer->timer); +} + +/** + * tasklet_hrtimer_init - Init a tasklet/hrtimer combo for softirq callbacks + * @ttimer: tasklet_hrtimer which is initialized + * @function: hrtimer callback funtion which gets called from softirq context + * @which_clock: clock id (CLOCK_MONOTONIC/CLOCK_REALTIME) + * @mode: hrtimer mode (HRTIMER_MODE_ABS/HRTIMER_MODE_REL) + */ +void tasklet_hrtimer_init(struct tasklet_hrtimer *ttimer, + enum hrtimer_restart (*function)(struct hrtimer *), + clockid_t which_clock, enum hrtimer_mode mode) +{ + hrtimer_init(&ttimer->timer, which_clock, mode); + ttimer->timer.function = __hrtimer_tasklet_trampoline; + tasklet_init(&ttimer->tasklet, __tasklet_hrtimer_trampoline, + (unsigned long)ttimer); + ttimer->function = function; +} +EXPORT_SYMBOL_GPL(tasklet_hrtimer_init); + +/* + * Remote softirq bits + */ + DEFINE_PER_CPU(struct list_head [NR_SOFTIRQS], softirq_work_list); EXPORT_PER_CPU_SYMBOL(softirq_work_list); diff --git a/kernel/time/clockevents.c b/kernel/time/clockevents.c index 1ad6dd4..a6dcd67 100644 --- a/kernel/time/clockevents.c +++ b/kernel/time/clockevents.c @@ -254,15 +254,4 @@ void clockevents_notify(unsigned long reason, void *arg) spin_unlock(&clockevents_lock); } EXPORT_SYMBOL_GPL(clockevents_notify); - -ktime_t clockevents_get_next_event(int cpu) -{ - struct tick_device *td; - struct clock_event_device *dev; - - td = &per_cpu(tick_cpu_device, cpu); - dev = td->evtdev; - - return dev->next_event; -} #endif diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index 592bf58..7466cb8 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -513,7 +513,7 @@ static ssize_t sysfs_override_clocksource(struct sys_device *dev, * Check to make sure we don't switch to a non-highres capable * clocksource if the tick code is in oneshot mode (highres or nohz) */ - if (tick_oneshot_mode_active() && + if (tick_oneshot_mode_active() && ovr && !(ovr->flags & CLOCK_SOURCE_VALID_FOR_HRES)) { printk(KERN_WARNING "%s clocksource is not HRT compatible. " "Cannot switch while in HRT/NOHZ mode\n", ovr->name); diff --git a/kernel/timer.c b/kernel/timer.c index 0b36b9e..a7f07d5 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -714,7 +714,7 @@ int mod_timer(struct timer_list *timer, unsigned long expires) * networking code - if the timer is re-modified * to be the same thing then just return: */ - if (timer->expires == expires && timer_pending(timer)) + if (timer_pending(timer) && timer->expires == expires) return 1; return __mod_timer(timer, expires, false, TIMER_NOT_PINNED); diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 1551f47..019f380 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -226,13 +226,13 @@ config BOOT_TRACER the timings of the initcalls and traces key events and the identity of tasks that can cause boot delays, such as context-switches. - Its aim is to be parsed by the /scripts/bootgraph.pl tool to + Its aim is to be parsed by the scripts/bootgraph.pl tool to produce pretty graphics about boot inefficiencies, giving a visual representation of the delays during initcalls - but the raw /debug/tracing/trace text output is readable too. - You must pass in ftrace=initcall to the kernel command line - to enable this on bootup. + You must pass in initcall_debug and ftrace=initcall to the kernel + command line to enable this on bootup. config TRACE_BRANCH_PROFILING bool diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 39af8af..7a34cb5 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -22,6 +22,7 @@ #include <linux/init.h> #include <linux/mutex.h> #include <linux/debugfs.h> +#include <linux/smp_lock.h> #include <linux/time.h> #include <linux/uaccess.h> @@ -266,8 +267,8 @@ static void blk_trace_free(struct blk_trace *bt) { debugfs_remove(bt->msg_file); debugfs_remove(bt->dropped_file); - debugfs_remove(bt->dir); relay_close(bt->rchan); + debugfs_remove(bt->dir); free_percpu(bt->sequence); free_percpu(bt->msg_data); kfree(bt); @@ -377,18 +378,8 @@ static int blk_subbuf_start_callback(struct rchan_buf *buf, void *subbuf, static int blk_remove_buf_file_callback(struct dentry *dentry) { - struct dentry *parent = dentry->d_parent; debugfs_remove(dentry); - /* - * this will fail for all but the last file, but that is ok. what we - * care about is the top level buts->name directory going away, when - * the last trace file is gone. Then we don't have to rmdir() that - * manually on trace stop, so it nicely solves the issue with - * force killing of running traces. - */ - - debugfs_remove(parent); return 0; } diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index f3716bf..1e1d23c 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -768,7 +768,7 @@ static struct tracer_stat function_stats __initdata = { .stat_show = function_stat_show }; -static void ftrace_profile_debugfs(struct dentry *d_tracer) +static __init void ftrace_profile_debugfs(struct dentry *d_tracer) { struct ftrace_profile_stat *stat; struct dentry *entry; @@ -786,7 +786,6 @@ static void ftrace_profile_debugfs(struct dentry *d_tracer) * The files created are permanent, if something happens * we still do not free memory. */ - kfree(stat); WARN(1, "Could not allocate stat file for cpu %d\n", cpu); @@ -813,7 +812,7 @@ static void ftrace_profile_debugfs(struct dentry *d_tracer) } #else /* CONFIG_FUNCTION_PROFILER */ -static void ftrace_profile_debugfs(struct dentry *d_tracer) +static __init void ftrace_profile_debugfs(struct dentry *d_tracer) { } #endif /* CONFIG_FUNCTION_PROFILER */ @@ -1663,7 +1662,7 @@ ftrace_regex_open(struct inode *inode, struct file *file, int enable) mutex_lock(&ftrace_regex_lock); if ((file->f_mode & FMODE_WRITE) && - !(file->f_flags & O_APPEND)) + (file->f_flags & O_TRUNC)) ftrace_filter_reset(enable); if (file->f_mode & FMODE_READ) { @@ -2578,7 +2577,7 @@ ftrace_graph_open(struct inode *inode, struct file *file) mutex_lock(&graph_lock); if ((file->f_mode & FMODE_WRITE) && - !(file->f_flags & O_APPEND)) { + (file->f_flags & O_TRUNC)) { ftrace_graph_count = 0; memset(ftrace_graph_funcs, 0, sizeof(ftrace_graph_funcs)); } @@ -2597,6 +2596,14 @@ ftrace_graph_open(struct inode *inode, struct file *file) } static int +ftrace_graph_release(struct inode *inode, struct file *file) +{ + if (file->f_mode & FMODE_READ) + seq_release(inode, file); + return 0; +} + +static int ftrace_set_func(unsigned long *array, int *idx, char *buffer) { struct dyn_ftrace *rec; @@ -2725,9 +2732,10 @@ ftrace_graph_write(struct file *file, const char __user *ubuf, } static const struct file_operations ftrace_graph_fops = { - .open = ftrace_graph_open, - .read = seq_read, - .write = ftrace_graph_write, + .open = ftrace_graph_open, + .read = seq_read, + .write = ftrace_graph_write, + .release = ftrace_graph_release, }; #endif /* CONFIG_FUNCTION_GRAPH_TRACER */ @@ -3160,10 +3168,10 @@ ftrace_enable_sysctl(struct ctl_table *table, int write, ret = proc_dointvec(table, write, file, buffer, lenp, ppos); - if (ret || !write || (last_ftrace_enabled == ftrace_enabled)) + if (ret || !write || (last_ftrace_enabled == !!ftrace_enabled)) goto out; - last_ftrace_enabled = ftrace_enabled; + last_ftrace_enabled = !!ftrace_enabled; if (ftrace_enabled) { diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index bf27bb7..a330513 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -735,6 +735,7 @@ ring_buffer_free(struct ring_buffer *buffer) put_online_cpus(); + kfree(buffer->buffers); free_cpumask_var(buffer->cpumask); kfree(buffer); @@ -1785,7 +1786,7 @@ void ring_buffer_discard_commit(struct ring_buffer *buffer, */ RB_WARN_ON(buffer, !local_read(&cpu_buffer->committing)); - if (!rb_try_to_discard(cpu_buffer, event)) + if (rb_try_to_discard(cpu_buffer, event)) goto out; /* @@ -2383,7 +2384,6 @@ rb_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts) * the box. Return the padding, and we will release * the current locks, and try again. */ - rb_advance_reader(cpu_buffer); return event; case RINGBUF_TYPE_TIME_EXTEND: @@ -2486,7 +2486,7 @@ static inline int rb_ok_to_lock(void) * buffer too. A one time deal is all you get from reading * the ring buffer from an NMI. */ - if (likely(!in_nmi() && !oops_in_progress)) + if (likely(!in_nmi())) return 1; tracing_off_permanent(); @@ -2519,6 +2519,8 @@ ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts) if (dolock) spin_lock(&cpu_buffer->reader_lock); event = rb_buffer_peek(buffer, cpu, ts); + if (event && event->type_len == RINGBUF_TYPE_PADDING) + rb_advance_reader(cpu_buffer); if (dolock) spin_unlock(&cpu_buffer->reader_lock); local_irq_restore(flags); @@ -2590,12 +2592,9 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts) spin_lock(&cpu_buffer->reader_lock); event = rb_buffer_peek(buffer, cpu, ts); - if (!event) - goto out_unlock; - - rb_advance_reader(cpu_buffer); + if (event) + rb_advance_reader(cpu_buffer); - out_unlock: if (dolock) spin_unlock(&cpu_buffer->reader_lock); local_irq_restore(flags); diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 3aa0a0d..c22b40f 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -17,6 +17,7 @@ #include <linux/writeback.h> #include <linux/kallsyms.h> #include <linux/seq_file.h> +#include <linux/smp_lock.h> #include <linux/notifier.h> #include <linux/irqflags.h> #include <linux/debugfs.h> @@ -847,6 +848,7 @@ tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags, ((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) | (need_resched() ? TRACE_FLAG_NEED_RESCHED : 0); } +EXPORT_SYMBOL_GPL(tracing_generic_entry_update); struct ring_buffer_event *trace_buffer_lock_reserve(struct trace_array *tr, int type, @@ -2030,7 +2032,7 @@ static int tracing_open(struct inode *inode, struct file *file) /* If this file was open for write, then erase contents */ if ((file->f_mode & FMODE_WRITE) && - !(file->f_flags & O_APPEND)) { + (file->f_flags & O_TRUNC)) { long cpu = (long) inode->i_private; if (cpu == TRACE_PIPE_ALL_CPU) @@ -3084,7 +3086,8 @@ tracing_fill_pipe_page(size_t rem, struct trace_iterator *iter) break; } - trace_consume(iter); + if (ret != TRACE_TYPE_NO_CONSUME) + trace_consume(iter); rem -= count; if (!find_next_entry_inc(iter)) { rem = 0; @@ -4232,8 +4235,11 @@ static void __ftrace_dump(bool disable_tracing) iter.pos = -1; if (find_next_entry_inc(&iter) != NULL) { - print_trace_line(&iter); - trace_consume(&iter); + int ret; + + ret = print_trace_line(&iter); + if (ret != TRACE_TYPE_NO_CONSUME) + trace_consume(&iter); } trace_printk_seq(&iter.seq); diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 3548ae5..8b9f4f6 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -438,10 +438,6 @@ struct trace_entry *tracing_get_trace_entry(struct trace_array *tr, struct trace_entry *trace_find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts); -void tracing_generic_entry_update(struct trace_entry *entry, - unsigned long flags, - int pc); - void default_wait_pipe(struct trace_iterator *iter); void poll_wait_pipe(struct trace_iterator *iter); diff --git a/kernel/trace/trace_event_profile.c b/kernel/trace/trace_event_profile.c index 5b5895a..11ba5bb 100644 --- a/kernel/trace/trace_event_profile.c +++ b/kernel/trace/trace_event_profile.c @@ -14,7 +14,7 @@ int ftrace_profile_enable(int event_id) mutex_lock(&event_mutex); list_for_each_entry(event, &ftrace_events, list) { - if (event->id == event_id) { + if (event->id == event_id && event->profile_enable) { ret = event->profile_enable(event); break; } diff --git a/kernel/trace/trace_event_types.h b/kernel/trace/trace_event_types.h index 5e32e37..6db005e 100644 --- a/kernel/trace/trace_event_types.h +++ b/kernel/trace/trace_event_types.h @@ -26,6 +26,9 @@ TRACE_EVENT_FORMAT(funcgraph_exit, TRACE_GRAPH_RET, ftrace_graph_ret_entry, ignore, TRACE_STRUCT( TRACE_FIELD(unsigned long, ret.func, func) + TRACE_FIELD(unsigned long long, ret.calltime, calltime) + TRACE_FIELD(unsigned long long, ret.rettime, rettime) + TRACE_FIELD(unsigned long, ret.overrun, overrun) TRACE_FIELD(int, ret.depth, depth) ), TP_RAW_FMT("<-- %lx (%d)") diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 53c8fd3..e75276a 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -376,7 +376,7 @@ ftrace_event_seq_open(struct inode *inode, struct file *file) const struct seq_operations *seq_ops; if ((file->f_mode & FMODE_WRITE) && - !(file->f_flags & O_APPEND)) + (file->f_flags & O_TRUNC)) ftrace_clear_events(); seq_ops = inode->i_private; @@ -940,7 +940,7 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, entry = trace_create_file("enable", 0644, call->dir, call, enable); - if (call->id) + if (call->id && call->profile_enable) entry = trace_create_file("id", 0444, call->dir, call, id); diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 936c621..f32dc9d 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -624,9 +624,6 @@ static int filter_add_subsystem_pred(struct filter_parse_state *ps, return -ENOSPC; } - filter->preds[filter->n_preds] = pred; - filter->n_preds++; - list_for_each_entry(call, &ftrace_events, list) { if (!call->define_fields) @@ -643,6 +640,9 @@ static int filter_add_subsystem_pred(struct filter_parse_state *ps, } replace_filter_string(call->filter, filter_string); } + + filter->preds[filter->n_preds] = pred; + filter->n_preds++; out: return err; } @@ -1029,12 +1029,17 @@ static int replace_preds(struct event_subsystem *system, if (elt->op == OP_AND || elt->op == OP_OR) { pred = create_logical_pred(elt->op); + if (!pred) + return -ENOMEM; if (call) { err = filter_add_pred(ps, call, pred); filter_free_pred(pred); - } else + } else { err = filter_add_subsystem_pred(ps, system, pred, filter_string); + if (err) + filter_free_pred(pred); + } if (err) return err; @@ -1048,12 +1053,17 @@ static int replace_preds(struct event_subsystem *system, } pred = create_pred(elt->op, operand1, operand2); + if (!pred) + return -ENOMEM; if (call) { err = filter_add_pred(ps, call, pred); filter_free_pred(pred); - } else + } else { err = filter_add_subsystem_pred(ps, system, pred, filter_string); + if (err) + filter_free_pred(pred); + } if (err) return err; diff --git a/kernel/trace/trace_functions.c b/kernel/trace/trace_functions.c index 7402144..75ef000 100644 --- a/kernel/trace/trace_functions.c +++ b/kernel/trace/trace_functions.c @@ -363,7 +363,7 @@ ftrace_trace_onoff_callback(char *glob, char *cmd, char *param, int enable) out_reg: ret = register_ftrace_function_probe(glob, ops, count); - return ret; + return ret < 0 ? ret : 0; } static struct ftrace_func_command ftrace_traceon_cmd = { diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index d2249ab..420ec34 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -843,9 +843,16 @@ print_graph_function(struct trace_iterator *iter) switch (entry->type) { case TRACE_GRAPH_ENT: { - struct ftrace_graph_ent_entry *field; + /* + * print_graph_entry() may consume the current event, + * thus @field may become invalid, so we need to save it. + * sizeof(struct ftrace_graph_ent_entry) is very small, + * it can be safely saved at the stack. + */ + struct ftrace_graph_ent_entry *field, saved; trace_assign_type(field, entry); - return print_graph_entry(field, s, iter); + saved = *field; + return print_graph_entry(&saved, s, iter); } case TRACE_GRAPH_RET: { struct ftrace_graph_ret_entry *field; diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index 7938f3a..e0c2545 100644 --- a/kernel/trace/trace_output.c +++ b/kernel/trace/trace_output.c @@ -27,8 +27,7 @@ void trace_print_seq(struct seq_file *m, struct trace_seq *s) { int len = s->len >= PAGE_SIZE ? PAGE_SIZE - 1 : s->len; - s->buffer[len] = 0; - seq_puts(m, s->buffer); + seq_write(m, s->buffer, len); trace_seq_init(s); } diff --git a/kernel/trace/trace_printk.c b/kernel/trace/trace_printk.c index 7b62781..687699d 100644 --- a/kernel/trace/trace_printk.c +++ b/kernel/trace/trace_printk.c @@ -176,7 +176,7 @@ static int t_show(struct seq_file *m, void *v) const char *str = *fmt; int i; - seq_printf(m, "0x%lx : \"", (unsigned long)fmt); + seq_printf(m, "0x%lx : \"", *(unsigned long *)fmt); /* * Tabs and new lines need to be converted. diff --git a/kernel/trace/trace_stack.c b/kernel/trace/trace_stack.c index 2d7aebd..6a2a9d4 100644 --- a/kernel/trace/trace_stack.c +++ b/kernel/trace/trace_stack.c @@ -301,17 +301,14 @@ static const struct seq_operations stack_trace_seq_ops = { static int stack_trace_open(struct inode *inode, struct file *file) { - int ret; - - ret = seq_open(file, &stack_trace_seq_ops); - - return ret; + return seq_open(file, &stack_trace_seq_ops); } static const struct file_operations stack_trace_fops = { .open = stack_trace_open, .read = seq_read, .llseek = seq_lseek, + .release = seq_release, }; int @@ -326,10 +323,10 @@ stack_trace_sysctl(struct ctl_table *table, int write, ret = proc_dointvec(table, write, file, buffer, lenp, ppos); if (ret || !write || - (last_stack_tracer_enabled == stack_tracer_enabled)) + (last_stack_tracer_enabled == !!stack_tracer_enabled)) goto out; - last_stack_tracer_enabled = stack_tracer_enabled; + last_stack_tracer_enabled = !!stack_tracer_enabled; if (stack_tracer_enabled) register_ftrace_function(&trace_ops); diff --git a/kernel/trace/trace_stat.c b/kernel/trace/trace_stat.c index e66f5e4..aea321c 100644 --- a/kernel/trace/trace_stat.c +++ b/kernel/trace/trace_stat.c @@ -73,7 +73,7 @@ static struct rb_node *release_next(struct rb_node *node) } } -static void reset_stat_session(struct stat_session *session) +static void __reset_stat_session(struct stat_session *session) { struct rb_node *node = session->stat_root.rb_node; @@ -83,10 +83,17 @@ static void reset_stat_session(struct stat_session *session) session->stat_root = RB_ROOT; } +static void reset_stat_session(struct stat_session *session) +{ + mutex_lock(&session->stat_mutex); + __reset_stat_session(session); + mutex_unlock(&session->stat_mutex); +} + static void destroy_session(struct stat_session *session) { debugfs_remove(session->file); - reset_stat_session(session); + __reset_stat_session(session); mutex_destroy(&session->stat_mutex); kfree(session); } @@ -150,7 +157,7 @@ static int stat_seq_init(struct stat_session *session) int i; mutex_lock(&session->stat_mutex); - reset_stat_session(session); + __reset_stat_session(session); if (!ts->stat_cmp) ts->stat_cmp = dummy_cmp; @@ -183,7 +190,7 @@ exit: return ret; exit_free_rbtree: - reset_stat_session(session); + __reset_stat_session(session); mutex_unlock(&session->stat_mutex); return ret; } @@ -250,16 +257,21 @@ static const struct seq_operations trace_stat_seq_ops = { static int tracing_stat_open(struct inode *inode, struct file *file) { int ret; - + struct seq_file *m; struct stat_session *session = inode->i_private; + ret = stat_seq_init(session); + if (ret) + return ret; + ret = seq_open(file, &trace_stat_seq_ops); - if (!ret) { - struct seq_file *m = file->private_data; - m->private = session; - ret = stat_seq_init(session); + if (ret) { + reset_stat_session(session); + return ret; } + m = file->private_data; + m->private = session; return ret; } @@ -270,11 +282,9 @@ static int tracing_stat_release(struct inode *i, struct file *f) { struct stat_session *session = i->i_private; - mutex_lock(&session->stat_mutex); reset_stat_session(session); - mutex_unlock(&session->stat_mutex); - return 0; + return seq_release(i, f); } static const struct file_operations tracing_stat_fops = { diff --git a/kernel/wait.c b/kernel/wait.c index ea7c3b4..c4bd3d8 100644 --- a/kernel/wait.c +++ b/kernel/wait.c @@ -10,13 +10,14 @@ #include <linux/wait.h> #include <linux/hash.h> -void init_waitqueue_head(wait_queue_head_t *q) +void __init_waitqueue_head(wait_queue_head_t *q, struct lock_class_key *key) { spin_lock_init(&q->lock); + lockdep_set_class(&q->lock, key); INIT_LIST_HEAD(&q->task_list); } -EXPORT_SYMBOL(init_waitqueue_head); +EXPORT_SYMBOL(__init_waitqueue_head); void add_wait_queue(wait_queue_head_t *q, wait_queue_t *wait) { |