perf: Fix perf bug in fork()

[sfrench/cifs-2.6.git] / kernel / events / core.c

/*
 * Performance events core code:
 *
 *  Copyright (C) 2008 Thomas Gleixner <tglx@linutronix.de>
 *  Copyright (C) 2008-2011 Red Hat, Inc., Ingo Molnar
 *  Copyright (C) 2008-2011 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com>
 *  Copyright  ©  2009 Paul Mackerras, IBM Corp. <paulus@au1.ibm.com>
 *
 * For licensing details see kernel-base/COPYING
 */

#include <linux/fs.h>
#include <linux/mm.h>
#include <linux/cpu.h>
#include <linux/smp.h>
#include <linux/idr.h>
#include <linux/file.h>
#include <linux/poll.h>
#include <linux/slab.h>
#include <linux/hash.h>
#include <linux/tick.h>
#include <linux/sysfs.h>
#include <linux/dcache.h>
#include <linux/percpu.h>
#include <linux/ptrace.h>
#include <linux/reboot.h>
#include <linux/vmstat.h>
#include <linux/device.h>
#include <linux/export.h>
#include <linux/vmalloc.h>
#include <linux/hardirq.h>
#include <linux/rculist.h>
#include <linux/uaccess.h>
#include <linux/syscalls.h>
#include <linux/anon_inodes.h>
#include <linux/kernel_stat.h>
#include <linux/perf_event.h>
#include <linux/ftrace_event.h>
#include <linux/hw_breakpoint.h>
#include <linux/mm_types.h>
#include <linux/cgroup.h>
#include <linux/module.h>
#include <linux/mman.h>
#include <linux/compat.h>

#include "internal.h"

#include <asm/irq_regs.h>

struct remote_function_call {
        struct task_struct      *p;
        int                     (*func)(void *info);
        void                    *info;
        int                     ret;
};

static void remote_function(void *data)
{
        struct remote_function_call *tfc = data;
        struct task_struct *p = tfc->p;

        if (p) {
                tfc->ret = -EAGAIN;
                if (task_cpu(p) != smp_processor_id() || !task_curr(p))
                        return;
        }

        tfc->ret = tfc->func(tfc->info);
}

/**
 * task_function_call - call a function on the cpu on which a task runs
 * @p:          the task to evaluate
 * @func:       the function to be called
 * @info:       the function call argument
 *
 * Calls the function @func when the task is currently running. This might
 * be on the current CPU, which just calls the function directly
 *
 * returns: @func return value, or
 *          -ESRCH  - when the process isn't running
 *          -EAGAIN - when the process moved away
 */
static int
task_function_call(struct task_struct *p, int (*func) (void *info), void *info)
{
        struct remote_function_call data = {
                .p      = p,
                .func   = func,
                .info   = info,
                .ret    = -ESRCH, /* No such (running) process */
        };

        if (task_curr(p))
                smp_call_function_single(task_cpu(p), remote_function, &data, 1);

        return data.ret;
}

/**
 * cpu_function_call - call a function on the cpu
 * @func:       the function to be called
 * @info:       the function call argument
 *
 * Calls the function @func on the remote cpu.
 *
 * returns: @func return value or -ENXIO when the cpu is offline
 */
static int cpu_function_call(int cpu, int (*func) (void *info), void *info)
{
        struct remote_function_call data = {
                .p      = NULL,
                .func   = func,
                .info   = info,
                .ret    = -ENXIO, /* No such CPU */
        };

        smp_call_function_single(cpu, remote_function, &data, 1);

        return data.ret;
}

#define PERF_FLAG_ALL (PERF_FLAG_FD_NO_GROUP |\
                       PERF_FLAG_FD_OUTPUT  |\
                       PERF_FLAG_PID_CGROUP |\
                       PERF_FLAG_FD_CLOEXEC)

/*
 * branch priv levels that need permission checks
 */
#define PERF_SAMPLE_BRANCH_PERM_PLM \
        (PERF_SAMPLE_BRANCH_KERNEL |\
         PERF_SAMPLE_BRANCH_HV)

enum event_type_t {
        EVENT_FLEXIBLE = 0x1,
        EVENT_PINNED = 0x2,
        EVENT_ALL = EVENT_FLEXIBLE | EVENT_PINNED,
};

/*
 * perf_sched_events : >0 events exist
 * perf_cgroup_events: >0 per-cpu cgroup events exist on this cpu
 */
struct static_key_deferred perf_sched_events __read_mostly;
static DEFINE_PER_CPU(atomic_t, perf_cgroup_events);
static DEFINE_PER_CPU(atomic_t, perf_branch_stack_events);

static atomic_t nr_mmap_events __read_mostly;
static atomic_t nr_comm_events __read_mostly;
static atomic_t nr_task_events __read_mostly;
static atomic_t nr_freq_events __read_mostly;

static LIST_HEAD(pmus);
static DEFINE_MUTEX(pmus_lock);
static struct srcu_struct pmus_srcu;

/*
 * perf event paranoia level:
 *  -1 - not paranoid at all
 *   0 - disallow raw tracepoint access for unpriv
 *   1 - disallow cpu events for unpriv
 *   2 - disallow kernel profiling for unpriv
 */
int sysctl_perf_event_paranoid __read_mostly = 1;

/* Minimum for 512 kiB + 1 user control page */
int sysctl_perf_event_mlock __read_mostly = 512 + (PAGE_SIZE / 1024); /* 'free' kiB per user */

/*
 * max perf event sample rate
 */
#define DEFAULT_MAX_SAMPLE_RATE         100000
#define DEFAULT_SAMPLE_PERIOD_NS        (NSEC_PER_SEC / DEFAULT_MAX_SAMPLE_RATE)
#define DEFAULT_CPU_TIME_MAX_PERCENT    25

int sysctl_perf_event_sample_rate __read_mostly = DEFAULT_MAX_SAMPLE_RATE;

static int max_samples_per_tick __read_mostly   = DIV_ROUND_UP(DEFAULT_MAX_SAMPLE_RATE, HZ);
static int perf_sample_period_ns __read_mostly  = DEFAULT_SAMPLE_PERIOD_NS;

static int perf_sample_allowed_ns __read_mostly =
        DEFAULT_SAMPLE_PERIOD_NS * DEFAULT_CPU_TIME_MAX_PERCENT / 100;

void update_perf_cpu_limits(void)
{
        u64 tmp = perf_sample_period_ns;

        tmp *= sysctl_perf_cpu_time_max_percent;
        do_div(tmp, 100);
        ACCESS_ONCE(perf_sample_allowed_ns) = tmp;
}

static int perf_rotate_context(struct perf_cpu_context *cpuctx);

int perf_proc_update_handler(struct ctl_table *table, int write,
                void __user *buffer, size_t *lenp,
                loff_t *ppos)
{
        int ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos);

        if (ret || !write)
                return ret;

        max_samples_per_tick = DIV_ROUND_UP(sysctl_perf_event_sample_rate, HZ);
        perf_sample_period_ns = NSEC_PER_SEC / sysctl_perf_event_sample_rate;
        update_perf_cpu_limits();

        return 0;
}

int sysctl_perf_cpu_time_max_percent __read_mostly = DEFAULT_CPU_TIME_MAX_PERCENT;

int perf_cpu_time_max_percent_handler(struct ctl_table *table, int write,
                                void __user *buffer, size_t *lenp,
                                loff_t *ppos)
{
        int ret = proc_dointvec(table, write, buffer, lenp, ppos);

        if (ret || !write)
                return ret;

        update_perf_cpu_limits();

        return 0;
}

/*
 * perf samples are done in some very critical code paths (NMIs).
 * If they take too much CPU time, the system can lock up and not
 * get any real work done.  This will drop the sample rate when
 * we detect that events are taking too long.
 */
#define NR_ACCUMULATED_SAMPLES 128
static DEFINE_PER_CPU(u64, running_sample_length);

static void perf_duration_warn(struct irq_work *w)
{
        u64 allowed_ns = ACCESS_ONCE(perf_sample_allowed_ns);
        u64 avg_local_sample_len;
        u64 local_samples_len;

        local_samples_len = __get_cpu_var(running_sample_length);
        avg_local_sample_len = local_samples_len/NR_ACCUMULATED_SAMPLES;

        printk_ratelimited(KERN_WARNING
                        "perf interrupt took too long (%lld > %lld), lowering "
                        "kernel.perf_event_max_sample_rate to %d\n",
                        avg_local_sample_len, allowed_ns >> 1,
                        sysctl_perf_event_sample_rate);
}

static DEFINE_IRQ_WORK(perf_duration_work, perf_duration_warn);

void perf_sample_event_took(u64 sample_len_ns)
{
        u64 allowed_ns = ACCESS_ONCE(perf_sample_allowed_ns);
        u64 avg_local_sample_len;
        u64 local_samples_len;

        if (allowed_ns == 0)
                return;

        /* decay the counter by 1 average sample */
        local_samples_len = __get_cpu_var(running_sample_length);
        local_samples_len -= local_samples_len/NR_ACCUMULATED_SAMPLES;
        local_samples_len += sample_len_ns;
        __get_cpu_var(running_sample_length) = local_samples_len;

        /*
         * note: this will be biased artifically low until we have
         * seen NR_ACCUMULATED_SAMPLES.  Doing it this way keeps us
         * from having to maintain a count.
         */
        avg_local_sample_len = local_samples_len/NR_ACCUMULATED_SAMPLES;

        if (avg_local_sample_len <= allowed_ns)
                return;

        if (max_samples_per_tick <= 1)
                return;

        max_samples_per_tick = DIV_ROUND_UP(max_samples_per_tick, 2);
        sysctl_perf_event_sample_rate = max_samples_per_tick * HZ;
        perf_sample_period_ns = NSEC_PER_SEC / sysctl_perf_event_sample_rate;

        update_perf_cpu_limits();

        if (!irq_work_queue(&perf_duration_work)) {
                early_printk("perf interrupt took too long (%lld > %lld), lowering "
                             "kernel.perf_event_max_sample_rate to %d\n",
                             avg_local_sample_len, allowed_ns >> 1,
                             sysctl_perf_event_sample_rate);
        }
}

static atomic64_t perf_event_id;

static void cpu_ctx_sched_out(struct perf_cpu_context *cpuctx,
                              enum event_type_t event_type);

static void cpu_ctx_sched_in(struct perf_cpu_context *cpuctx,
                             enum event_type_t event_type,
                             struct task_struct *task);

static void update_context_time(struct perf_event_context *ctx);
static u64 perf_event_time(struct perf_event *event);

void __weak perf_event_print_debug(void)        { }

extern __weak const char *perf_pmu_name(void)
{
        return "pmu";
}

static inline u64 perf_clock(void)
{
        return local_clock();
}

static inline struct perf_cpu_context *
__get_cpu_context(struct perf_event_context *ctx)
{
        return this_cpu_ptr(ctx->pmu->pmu_cpu_context);
}

static void perf_ctx_lock(struct perf_cpu_context *cpuctx,
                          struct perf_event_context *ctx)
{
        raw_spin_lock(&cpuctx->ctx.lock);
        if (ctx)
                raw_spin_lock(&ctx->lock);
}

static void perf_ctx_unlock(struct perf_cpu_context *cpuctx,
                            struct perf_event_context *ctx)
{
        if (ctx)
                raw_spin_unlock(&ctx->lock);
        raw_spin_unlock(&cpuctx->ctx.lock);
}

#ifdef CONFIG_CGROUP_PERF

/*
 * perf_cgroup_info keeps track of time_enabled for a cgroup.
 * This is a per-cpu dynamically allocated data structure.
 */
struct perf_cgroup_info {
        u64                             time;
        u64                             timestamp;
};

struct perf_cgroup {
        struct cgroup_subsys_state      css;
        struct perf_cgroup_info __percpu *info;
};

/*
 * Must ensure cgroup is pinned (css_get) before calling
 * this function. In other words, we cannot call this function
 * if there is no cgroup event for the current CPU context.
 */
static inline struct perf_cgroup *
perf_cgroup_from_task(struct task_struct *task)
{
        return container_of(task_css(task, perf_event_cgrp_id),
                            struct perf_cgroup, css);
}

static inline bool
perf_cgroup_match(struct perf_event *event)
{
        struct perf_event_context *ctx = event->ctx;
        struct perf_cpu_context *cpuctx = __get_cpu_context(ctx);

        /* @event doesn't care about cgroup */
        if (!event->cgrp)
                return true;

        /* wants specific cgroup scope but @cpuctx isn't associated with any */
        if (!cpuctx->cgrp)
                return false;

        /*
         * Cgroup scoping is recursive.  An event enabled for a cgroup is
         * also enabled for all its descendant cgroups.  If @cpuctx's
         * cgroup is a descendant of @event's (the test covers identity
         * case), it's a match.
         */
        return cgroup_is_descendant(cpuctx->cgrp->css.cgroup,
                                    event->cgrp->css.cgroup);
}

static inline void perf_put_cgroup(struct perf_event *event)
{
        css_put(&event->cgrp->css);
}

static inline void perf_detach_cgroup(struct perf_event *event)
{
        perf_put_cgroup(event);
        event->cgrp = NULL;
}

static inline int is_cgroup_event(struct perf_event *event)
{
        return event->cgrp != NULL;
}

static inline u64 perf_cgroup_event_time(struct perf_event *event)
{
        struct perf_cgroup_info *t;

        t = per_cpu_ptr(event->cgrp->info, event->cpu);
        return t->time;
}

static inline void __update_cgrp_time(struct perf_cgroup *cgrp)
{
        struct perf_cgroup_info *info;
        u64 now;

        now = perf_clock();

        info = this_cpu_ptr(cgrp->info);

        info->time += now - info->timestamp;
        info->timestamp = now;
}

static inline void update_cgrp_time_from_cpuctx(struct perf_cpu_context *cpuctx)
{
        struct perf_cgroup *cgrp_out = cpuctx->cgrp;
        if (cgrp_out)
                __update_cgrp_time(cgrp_out);
}

static inline void update_cgrp_time_from_event(struct perf_event *event)
{
        struct perf_cgroup *cgrp;

        /*
         * ensure we access cgroup data only when needed and
         * when we know the cgroup is pinned (css_get)
         */
        if (!is_cgroup_event(event))
                return;

        cgrp = perf_cgroup_from_task(current);
        /*
         * Do not update time when cgroup is not active
         */
        if (cgrp == event->cgrp)
                __update_cgrp_time(event->cgrp);
}

static inline void
perf_cgroup_set_timestamp(struct task_struct *task,
                          struct perf_event_context *ctx)
{
        struct perf_cgroup *cgrp;
        struct perf_cgroup_info *info;

        /*
         * ctx->lock held by caller
         * ensure we do not access cgroup data
         * unless we have the cgroup pinned (css_get)
         */
        if (!task || !ctx->nr_cgroups)
                return;

        cgrp = perf_cgroup_from_task(task);
        info = this_cpu_ptr(cgrp->info);
        info->timestamp = ctx->timestamp;
}

#define PERF_CGROUP_SWOUT       0x1 /* cgroup switch out every event */
#define PERF_CGROUP_SWIN        0x2 /* cgroup switch in events based on task */

/*
 * reschedule events based on the cgroup constraint of task.
 *
 * mode SWOUT : schedule out everything
 * mode SWIN : schedule in based on cgroup for next
 */
void perf_cgroup_switch(struct task_struct *task, int mode)
{
        struct perf_cpu_context *cpuctx;
        struct pmu *pmu;
        unsigned long flags;

        /*
         * disable interrupts to avoid geting nr_cgroup
         * changes via __perf_event_disable(). Also
         * avoids preemption.
         */
        local_irq_save(flags);

        /*
         * we reschedule only in the presence of cgroup
         * constrained events.
         */
        rcu_read_lock();

        list_for_each_entry_rcu(pmu, &pmus, entry) {
                cpuctx = this_cpu_ptr(pmu->pmu_cpu_context);
                if (cpuctx->unique_pmu != pmu)
                        continue; /* ensure we process each cpuctx once */

                /*
                 * perf_cgroup_events says at least one
                 * context on this CPU has cgroup events.
                 *
                 * ctx->nr_cgroups reports the number of cgroup
                 * events for a context.
                 */
                if (cpuctx->ctx.nr_cgroups > 0) {
                        perf_ctx_lock(cpuctx, cpuctx->task_ctx);
                        perf_pmu_disable(cpuctx->ctx.pmu);

                        if (mode & PERF_CGROUP_SWOUT) {
                                cpu_ctx_sched_out(cpuctx, EVENT_ALL);
                                /*
                                 * must not be done before ctxswout due
                                 * to event_filter_match() in event_sched_out()
                                 */
                                cpuctx->cgrp = NULL;
                        }

                        if (mode & PERF_CGROUP_SWIN) {
                                WARN_ON_ONCE(cpuctx->cgrp);
                                /*
                                 * set cgrp before ctxsw in to allow
                                 * event_filter_match() to not have to pass
                                 * task around
                                 */
                                cpuctx->cgrp = perf_cgroup_from_task(task);
                                cpu_ctx_sched_in(cpuctx, EVENT_ALL, task);
                        }
                        perf_pmu_enable(cpuctx->ctx.pmu);
                        perf_ctx_unlock(cpuctx, cpuctx->task_ctx);
                }
        }

        rcu_read_unlock();

        local_irq_restore(flags);
}

static inline void perf_cgroup_sched_out(struct task_struct *task,
                                         struct task_struct *next)
{
        struct perf_cgroup *cgrp1;
        struct perf_cgroup *cgrp2 = NULL;

        /*
         * we come here when we know perf_cgroup_events > 0
         */
        cgrp1 = perf_cgroup_from_task(task);

        /*
         * next is NULL when called from perf_event_enable_on_exec()
         * that will systematically cause a cgroup_switch()
         */
        if (next)
                cgrp2 = perf_cgroup_from_task(next);

        /*
         * only schedule out current cgroup events if we know
         * that we are switching to a different cgroup. Otherwise,
         * do no touch the cgroup events.
         */
        if (cgrp1 != cgrp2)
                perf_cgroup_switch(task, PERF_CGROUP_SWOUT);
}

static inline void perf_cgroup_sched_in(struct task_struct *prev,
                                        struct task_struct *task)
{
        struct perf_cgroup *cgrp1;
        struct perf_cgroup *cgrp2 = NULL;

        /*
         * we come here when we know perf_cgroup_events > 0
         */
        cgrp1 = perf_cgroup_from_task(task);

        /* prev can never be NULL */
        cgrp2 = perf_cgroup_from_task(prev);

        /*
         * only need to schedule in cgroup events if we are changing
         * cgroup during ctxsw. Cgroup events were not scheduled
         * out of ctxsw out if that was not the case.
         */
        if (cgrp1 != cgrp2)
                perf_cgroup_switch(task, PERF_CGROUP_SWIN);
}

static inline int perf_cgroup_connect(int fd, struct perf_event *event,
                                      struct perf_event_attr *attr,
                                      struct perf_event *group_leader)
{
        struct perf_cgroup *cgrp;
        struct cgroup_subsys_state *css;
        struct fd f = fdget(fd);
        int ret = 0;

        if (!f.file)
                return -EBADF;

        css = css_tryget_online_from_dir(f.file->f_dentry,
                                         &perf_event_cgrp_subsys);
        if (IS_ERR(css)) {
                ret = PTR_ERR(css);
                goto out;
        }

        cgrp = container_of(css, struct perf_cgroup, css);
        event->cgrp = cgrp;

        /*
         * all events in a group must monitor
         * the same cgroup because a task belongs
         * to only one perf cgroup at a time
         */
        if (group_leader && group_leader->cgrp != cgrp) {
                perf_detach_cgroup(event);
                ret = -EINVAL;
        }
out:
        fdput(f);
        return ret;
}

static inline void
perf_cgroup_set_shadow_time(struct perf_event *event, u64 now)
{
        struct perf_cgroup_info *t;
        t = per_cpu_ptr(event->cgrp->info, event->cpu);
        event->shadow_ctx_time = now - t->timestamp;
}

static inline void
perf_cgroup_defer_enabled(struct perf_event *event)
{
        /*
         * when the current task's perf cgroup does not match
         * the event's, we need to remember to call the
         * perf_mark_enable() function the first time a task with
         * a matching perf cgroup is scheduled in.
         */
        if (is_cgroup_event(event) && !perf_cgroup_match(event))
                event->cgrp_defer_enabled = 1;
}

static inline void
perf_cgroup_mark_enabled(struct perf_event *event,
                         struct perf_event_context *ctx)
{
        struct perf_event *sub;
        u64 tstamp = perf_event_time(event);

        if (!event->cgrp_defer_enabled)
                return;

        event->cgrp_defer_enabled = 0;

        event->tstamp_enabled = tstamp - event->total_time_enabled;
        list_for_each_entry(sub, &event->sibling_list, group_entry) {
                if (sub->state >= PERF_EVENT_STATE_INACTIVE) {
                        sub->tstamp_enabled = tstamp - sub->total_time_enabled;
                        sub->cgrp_defer_enabled = 0;
                }
        }
}
#else /* !CONFIG_CGROUP_PERF */

static inline bool
perf_cgroup_match(struct perf_event *event)
{
        return true;
}

static inline void perf_detach_cgroup(struct perf_event *event)
{}

static inline int is_cgroup_event(struct perf_event *event)
{
        return 0;
}

static inline u64 perf_cgroup_event_cgrp_time(struct perf_event *event)
{
        return 0;
}

static inline void update_cgrp_time_from_event(struct perf_event *event)
{
}

static inline void update_cgrp_time_from_cpuctx(struct perf_cpu_context *cpuctx)
{
}

static inline void perf_cgroup_sched_out(struct task_struct *task,
                                         struct task_struct *next)
{
}

static inline void perf_cgroup_sched_in(struct task_struct *prev,
                                        struct task_struct *task)
{
}

static inline int perf_cgroup_connect(pid_t pid, struct perf_event *event,
                                      struct perf_event_attr *attr,
                                      struct perf_event *group_leader)
{
        return -EINVAL;
}

static inline void
perf_cgroup_set_timestamp(struct task_struct *task,
                          struct perf_event_context *ctx)
{
}

void
perf_cgroup_switch(struct task_struct *task, struct task_struct *next)
{
}

static inline void
perf_cgroup_set_shadow_time(struct perf_event *event, u64 now)
{
}

static inline u64 perf_cgroup_event_time(struct perf_event *event)
{
        return 0;
}

static inline void
perf_cgroup_defer_enabled(struct perf_event *event)
{
}

static inline void
perf_cgroup_mark_enabled(struct perf_event *event,
                         struct perf_event_context *ctx)
{
}
#endif

/*
 * set default to be dependent on timer tick just
 * like original code
 */
#define PERF_CPU_HRTIMER (1000 / HZ)
/*
 * function must be called with interrupts disbled
 */
static enum hrtimer_restart perf_cpu_hrtimer_handler(struct hrtimer *hr)
{
        struct perf_cpu_context *cpuctx;
        enum hrtimer_restart ret = HRTIMER_NORESTART;
        int rotations = 0;

        WARN_ON(!irqs_disabled());

        cpuctx = container_of(hr, struct perf_cpu_context, hrtimer);

        rotations = perf_rotate_context(cpuctx);

        /*
         * arm timer if needed
         */
        if (rotations) {
                hrtimer_forward_now(hr, cpuctx->hrtimer_interval);
                ret = HRTIMER_RESTART;
        }

        return ret;
}

/* CPU is going down */
void perf_cpu_hrtimer_cancel(int cpu)
{
        struct perf_cpu_context *cpuctx;
        struct pmu *pmu;
        unsigned long flags;

        if (WARN_ON(cpu != smp_processor_id()))
                return;

        local_irq_save(flags);

        rcu_read_lock();

        list_for_each_entry_rcu(pmu, &pmus, entry) {
                cpuctx = this_cpu_ptr(pmu->pmu_cpu_context);

                if (pmu->task_ctx_nr == perf_sw_context)
                        continue;

                hrtimer_cancel(&cpuctx->hrtimer);
        }

        rcu_read_unlock();

        local_irq_restore(flags);
}

static void __perf_cpu_hrtimer_init(struct perf_cpu_context *cpuctx, int cpu)
{
        struct hrtimer *hr = &cpuctx->hrtimer;
        struct pmu *pmu = cpuctx->ctx.pmu;
        int timer;

        /* no multiplexing needed for SW PMU */
        if (pmu->task_ctx_nr == perf_sw_context)
                return;

        /*
         * check default is sane, if not set then force to
         * default interval (1/tick)
         */
        timer = pmu->hrtimer_interval_ms;
        if (timer < 1)
                timer = pmu->hrtimer_interval_ms = PERF_CPU_HRTIMER;

        cpuctx->hrtimer_interval = ns_to_ktime(NSEC_PER_MSEC * timer);

        hrtimer_init(hr, CLOCK_MONOTONIC, HRTIMER_MODE_REL_PINNED);
        hr->function = perf_cpu_hrtimer_handler;
}

static void perf_cpu_hrtimer_restart(struct perf_cpu_context *cpuctx)
{
        struct hrtimer *hr = &cpuctx->hrtimer;
        struct pmu *pmu = cpuctx->ctx.pmu;

        /* not for SW PMU */
        if (pmu->task_ctx_nr == perf_sw_context)
                return;

        if (hrtimer_active(hr))
                return;

        if (!hrtimer_callback_running(hr))
                __hrtimer_start_range_ns(hr, cpuctx->hrtimer_interval,
                                         0, HRTIMER_MODE_REL_PINNED, 0);
}

void perf_pmu_disable(struct pmu *pmu)
{
        int *count = this_cpu_ptr(pmu->pmu_disable_count);
        if (!(*count)++)
                pmu->pmu_disable(pmu);
}

void perf_pmu_enable(struct pmu *pmu)
{
        int *count = this_cpu_ptr(pmu->pmu_disable_count);
        if (!--(*count))
                pmu->pmu_enable(pmu);
}

static DEFINE_PER_CPU(struct list_head, rotation_list);

/*
 * perf_pmu_rotate_start() and perf_rotate_context() are fully serialized
 * because they're strictly cpu affine and rotate_start is called with IRQs
 * disabled, while rotate_context is called from IRQ context.
 */
static void perf_pmu_rotate_start(struct pmu *pmu)
{
        struct perf_cpu_context *cpuctx = this_cpu_ptr(pmu->pmu_cpu_context);
        struct list_head *head = &__get_cpu_var(rotation_list);

        WARN_ON(!irqs_disabled());

        if (list_empty(&cpuctx->rotation_list))
                list_add(&cpuctx->rotation_list, head);
}

static void get_ctx(struct perf_event_context *ctx)
{
        WARN_ON(!atomic_inc_not_zero(&ctx->refcount));
}

static void put_ctx(struct perf_event_context *ctx)
{
        if (atomic_dec_and_test(&ctx->refcount)) {
                if (ctx->parent_ctx)
                        put_ctx(ctx->parent_ctx);
                if (ctx->task)
                        put_task_struct(ctx->task);
                kfree_rcu(ctx, rcu_head);
        }
}

/*
 * This must be done under the ctx->lock, such as to serialize against
 * context_equiv(), therefore we cannot call put_ctx() since that might end up
 * calling scheduler related locks and ctx->lock nests inside those.
 */
static __must_check struct perf_event_context *
unclone_ctx(struct perf_event_context *ctx)
{
        struct perf_event_context *parent_ctx = ctx->parent_ctx;

        lockdep_assert_held(&ctx->lock);

        if (parent_ctx)
                ctx->parent_ctx = NULL;
        ctx->generation++;

        return parent_ctx;
}

static u32 perf_event_pid(struct perf_event *event, struct task_struct *p)
{
        /*
         * only top level events have the pid namespace they were created in
         */
        if (event->parent)
                event = event->parent;

        return task_tgid_nr_ns(p, event->ns);
}

static u32 perf_event_tid(struct perf_event *event, struct task_struct *p)
{
        /*
         * only top level events have the pid namespace they were created in
         */
        if (event->parent)
                event = event->parent;

        return task_pid_nr_ns(p, event->ns);
}

/*
 * If we inherit events we want to return the parent event id
 * to userspace.
 */
static u64 primary_event_id(struct perf_event *event)
{
        u64 id = event->id;

        if (event->parent)
                id = event->parent->id;

        return id;
}

/*
 * Get the perf_event_context for a task and lock it.
 * This has to cope with with the fact that until it is locked,
 * the context could get moved to another task.
 */
static struct perf_event_context *
perf_lock_task_context(struct task_struct *task, int ctxn, unsigned long *flags)
{
        struct perf_event_context *ctx;

retry:
        /*
         * One of the few rules of preemptible RCU is that one cannot do
         * rcu_read_unlock() while holding a scheduler (or nested) lock when
         * part of the read side critical section was preemptible -- see
         * rcu_read_unlock_special().
         *
         * Since ctx->lock nests under rq->lock we must ensure the entire read
         * side critical section is non-preemptible.
         */
        preempt_disable();
        rcu_read_lock();
        ctx = rcu_dereference(task->perf_event_ctxp[ctxn]);
        if (ctx) {
                /*
                 * If this context is a clone of another, it might
                 * get swapped for another underneath us by
                 * perf_event_task_sched_out, though the
                 * rcu_read_lock() protects us from any context
                 * getting freed.  Lock the context and check if it
                 * got swapped before we could get the lock, and retry
                 * if so.  If we locked the right context, then it
                 * can't get swapped on us any more.
                 */
                raw_spin_lock_irqsave(&ctx->lock, *flags);
                if (ctx != rcu_dereference(task->perf_event_ctxp[ctxn])) {
                        raw_spin_unlock_irqrestore(&ctx->lock, *flags);
                        rcu_read_unlock();
                        preempt_enable();
                        goto retry;
                }

                if (!atomic_inc_not_zero(&ctx->refcount)) {
                        raw_spin_unlock_irqrestore(&ctx->lock, *flags);
                        ctx = NULL;
                }
        }
        rcu_read_unlock();
        preempt_enable();
        return ctx;
}

/*
 * Get the context for a task and increment its pin_count so it
 * can't get swapped to another task.  This also increments its
 * reference count so that the context can't get freed.
 */
static struct perf_event_context *
perf_pin_task_context(struct task_struct *task, int ctxn)
{
        struct perf_event_context *ctx;
        unsigned long flags;

        ctx = perf_lock_task_context(task, ctxn, &flags);
        if (ctx) {
                ++ctx->pin_count;
                raw_spin_unlock_irqrestore(&ctx->lock, flags);
        }
        return ctx;
}

static void perf_unpin_context(struct perf_event_context *ctx)
{
        unsigned long flags;

        raw_spin_lock_irqsave(&ctx->lock, flags);
        --ctx->pin_count;
        raw_spin_unlock_irqrestore(&ctx->lock, flags);
}

/*
 * Update the record of the current time in a context.
 */
static void update_context_time(struct perf_event_context *ctx)
{
        u64 now = perf_clock();

        ctx->time += now - ctx->timestamp;
        ctx->timestamp = now;
}

static u64 perf_event_time(struct perf_event *event)
{
        struct perf_event_context *ctx = event->ctx;

        if (is_cgroup_event(event))
                return perf_cgroup_event_time(event);

        return ctx ? ctx->time : 0;
}

/*
 * Update the total_time_enabled and total_time_running fields for a event.
 * The caller of this function needs to hold the ctx->lock.
 */
static void update_event_times(struct perf_event *event)
{
        struct perf_event_context *ctx = event->ctx;
        u64 run_end;

        if (event->state < PERF_EVENT_STATE_INACTIVE ||
            event->group_leader->state < PERF_EVENT_STATE_INACTIVE)
                return;
        /*
         * in cgroup mode, time_enabled represents
         * the time the event was enabled AND active
         * tasks were in the monitored cgroup. This is
         * independent of the activity of the context as
         * there may be a mix of cgroup and non-cgroup events.
         *
         * That is why we treat cgroup events differently
         * here.
         */
        if (is_cgroup_event(event))
                run_end = perf_cgroup_event_time(event);
        else if (ctx->is_active)
                run_end = ctx->time;
        else
                run_end = event->tstamp_stopped;

        event->total_time_enabled = run_end - event->tstamp_enabled;

        if (event->state == PERF_EVENT_STATE_INACTIVE)
                run_end = event->tstamp_stopped;
        else
                run_end = perf_event_time(event);

        event->total_time_running = run_end - event->tstamp_running;

}

/*
 * Update total_time_enabled and total_time_running for all events in a group.
 */
static void update_group_times(struct perf_event *leader)
{
        struct perf_event *event;

        update_event_times(leader);
        list_for_each_entry(event, &leader->sibling_list, group_entry)
                update_event_times(event);
}

static struct list_head *
ctx_group_list(struct perf_event *event, struct perf_event_context *ctx)
{
        if (event->attr.pinned)
                return &ctx->pinned_groups;
        else
                return &ctx->flexible_groups;
}

/*
 * Add a event from the lists for its context.
 * Must be called with ctx->mutex and ctx->lock held.
 */
static void
list_add_event(struct perf_event *event, struct perf_event_context *ctx)
{
        WARN_ON_ONCE(event->attach_state & PERF_ATTACH_CONTEXT);
        event->attach_state |= PERF_ATTACH_CONTEXT;

        /*
         * If we're a stand alone event or group leader, we go to the context
         * list, group events are kept attached to the group so that
         * perf_group_detach can, at all times, locate all siblings.
         */
        if (event->group_leader == event) {
                struct list_head *list;

                if (is_software_event(event))
                        event->group_flags |= PERF_GROUP_SOFTWARE;

                list = ctx_group_list(event, ctx);
                list_add_tail(&event->group_entry, list);
        }

        if (is_cgroup_event(event))
                ctx->nr_cgroups++;

        if (has_branch_stack(event))
                ctx->nr_branch_stack++;

        list_add_rcu(&event->event_entry, &ctx->event_list);
        if (!ctx->nr_events)
                perf_pmu_rotate_start(ctx->pmu);
        ctx->nr_events++;
        if (event->attr.inherit_stat)
                ctx->nr_stat++;

        ctx->generation++;
}

/*
 * Initialize event state based on the perf_event_attr::disabled.
 */
static inline void perf_event__state_init(struct perf_event *event)
{
        event->state = event->attr.disabled ? PERF_EVENT_STATE_OFF :
                                              PERF_EVENT_STATE_INACTIVE;
}

/*
 * Called at perf_event creation and when events are attached/detached from a
 * group.
 */
static void perf_event__read_size(struct perf_event *event)
{
        int entry = sizeof(u64); /* value */
        int size = 0;
        int nr = 1;

        if (event->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED)
                size += sizeof(u64);

        if (event->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING)
                size += sizeof(u64);

        if (event->attr.read_format & PERF_FORMAT_ID)
                entry += sizeof(u64);

        if (event->attr.read_format & PERF_FORMAT_GROUP) {
                nr += event->group_leader->nr_siblings;
                size += sizeof(u64);
        }

        size += entry * nr;
        event->read_size = size;
}

static void perf_event__header_size(struct perf_event *event)
{
        struct perf_sample_data *data;
        u64 sample_type = event->attr.sample_type;
        u16 size = 0;

        perf_event__read_size(event);

        if (sample_type & PERF_SAMPLE_IP)
                size += sizeof(data->ip);

        if (sample_type & PERF_SAMPLE_ADDR)
                size += sizeof(data->addr);

        if (sample_type & PERF_SAMPLE_PERIOD)
                size += sizeof(data->period);

        if (sample_type & PERF_SAMPLE_WEIGHT)
                size += sizeof(data->weight);

        if (sample_type & PERF_SAMPLE_READ)
                size += event->read_size;

        if (sample_type & PERF_SAMPLE_DATA_SRC)
                size += sizeof(data->data_src.val);

        if (sample_type & PERF_SAMPLE_TRANSACTION)
                size += sizeof(data->txn);

        event->header_size = size;
}

static void perf_event__id_header_size(struct perf_event *event)
{
        struct perf_sample_data *data;
        u64 sample_type = event->attr.sample_type;
        u16 size = 0;

        if (sample_type & PERF_SAMPLE_TID)
                size += sizeof(data->tid_entry);

        if (sample_type & PERF_SAMPLE_TIME)
                size += sizeof(data->time);

        if (sample_type & PERF_SAMPLE_IDENTIFIER)
                size += sizeof(data->id);

        if (sample_type & PERF_SAMPLE_ID)
                size += sizeof(data->id);

        if (sample_type & PERF_SAMPLE_STREAM_ID)
                size += sizeof(data->stream_id);

        if (sample_type & PERF_SAMPLE_CPU)
                size += sizeof(data->cpu_entry);

        event->id_header_size = size;
}

static void perf_group_attach(struct perf_event *event)
{
        struct perf_event *group_leader = event->group_leader, *pos;

        /*
         * We can have double attach due to group movement in perf_event_open.
         */
        if (event->attach_state & PERF_ATTACH_GROUP)
                return;

        event->attach_state |= PERF_ATTACH_GROUP;

        if (group_leader == event)
                return;

        if (group_leader->group_flags & PERF_GROUP_SOFTWARE &&
                        !is_software_event(event))
                group_leader->group_flags &= ~PERF_GROUP_SOFTWARE;

        list_add_tail(&event->group_entry, &group_leader->sibling_list);
        group_leader->nr_siblings++;

        perf_event__header_size(group_leader);

        list_for_each_entry(pos, &group_leader->sibling_list, group_entry)
                perf_event__header_size(pos);
}

/*
 * Remove a event from the lists for its context.
 * Must be called with ctx->mutex and ctx->lock held.
 */
static void
list_del_event(struct perf_event *event, struct perf_event_context *ctx)
{
        struct perf_cpu_context *cpuctx;
        /*
         * We can have double detach due to exit/hot-unplug + close.
         */
        if (!(event->attach_state & PERF_ATTACH_CONTEXT))
                return;

        event->attach_state &= ~PERF_ATTACH_CONTEXT;

        if (is_cgroup_event(event)) {
                ctx->nr_cgroups--;
                cpuctx = __get_cpu_context(ctx);
                /*
                 * if there are no more cgroup events
                 * then cler cgrp to avoid stale pointer
                 * in update_cgrp_time_from_cpuctx()
                 */
                if (!ctx->nr_cgroups)
                        cpuctx->cgrp = NULL;
        }

        if (has_branch_stack(event))
                ctx->nr_branch_stack--;

        ctx->nr_events--;
        if (event->attr.inherit_stat)
                ctx->nr_stat--;

        list_del_rcu(&event->event_entry);

        if (event->group_leader == event)
                list_del_init(&event->group_entry);

        update_group_times(event);

        /*
         * If event was in error state, then keep it
         * that way, otherwise bogus counts will be
         * returned on read(). The only way to get out
         * of error state is by explicit re-enabling
         * of the event
         */
        if (event->state > PERF_EVENT_STATE_OFF)
                event->state = PERF_EVENT_STATE_OFF;

        ctx->generation++;
}

static void perf_group_detach(struct perf_event *event)
{
        struct perf_event *sibling, *tmp;
        struct list_head *list = NULL;

        /*
         * We can have double detach due to exit/hot-unplug + close.
         */
        if (!(event->attach_state & PERF_ATTACH_GROUP))
                return;

        event->attach_state &= ~PERF_ATTACH_GROUP;

        /*
         * If this is a sibling, remove it from its group.
         */
        if (event->group_leader != event) {
                list_del_init(&event->group_entry);
                event->group_leader->nr_siblings--;
                goto out;
        }

        if (!list_empty(&event->group_entry))
                list = &event->group_entry;

        /*
         * If this was a group event with sibling events then
         * upgrade the siblings to singleton events by adding them
         * to whatever list we are on.
         */
        list_for_each_entry_safe(sibling, tmp, &event->sibling_list, group_entry) {
                if (list)
                        list_move_tail(&sibling->group_entry, list);
                sibling->group_leader = sibling;

                /* Inherit group flags from the previous leader */
                sibling->group_flags = event->group_flags;
        }

out:
        perf_event__header_size(event->group_leader);

        list_for_each_entry(tmp, &event->group_leader->sibling_list, group_entry)
                perf_event__header_size(tmp);
}

static inline int
event_filter_match(struct perf_event *event)
{
        return (event->cpu == -1 || event->cpu == smp_processor_id())
            && perf_cgroup_match(event);
}

static void
event_sched_out(struct perf_event *event,
                  struct perf_cpu_context *cpuctx,
                  struct perf_event_context *ctx)
{
        u64 tstamp = perf_event_time(event);
        u64 delta;
        /*
         * An event which could not be activated because of
         * filter mismatch still needs to have its timings
         * maintained, otherwise bogus information is return
         * via read() for time_enabled, time_running:
         */
        if (event->state == PERF_EVENT_STATE_INACTIVE
            && !event_filter_match(event)) {
                delta = tstamp - event->tstamp_stopped;
                event->tstamp_running += delta;
                event->tstamp_stopped = tstamp;
        }

        if (event->state != PERF_EVENT_STATE_ACTIVE)
                return;

        perf_pmu_disable(event->pmu);

        event->state = PERF_EVENT_STATE_INACTIVE;
        if (event->pending_disable) {
                event->pending_disable = 0;
                event->state = PERF_EVENT_STATE_OFF;
        }
        event->tstamp_stopped = tstamp;
        event->pmu->del(event, 0);
        event->oncpu = -1;

        if (!is_software_event(event))
                cpuctx->active_oncpu--;
        ctx->nr_active--;
        if (event->attr.freq && event->attr.sample_freq)
                ctx->nr_freq--;
        if (event->attr.exclusive || !cpuctx->active_oncpu)
                cpuctx->exclusive = 0;

        perf_pmu_enable(event->pmu);
}

static void
group_sched_out(struct perf_event *group_event,
                struct perf_cpu_context *cpuctx,
                struct perf_event_context *ctx)
{
        struct perf_event *event;
        int state = group_event->state;

        event_sched_out(group_event, cpuctx, ctx);

        /*
         * Schedule out siblings (if any):
         */
        list_for_each_entry(event, &group_event->sibling_list, group_entry)
                event_sched_out(event, cpuctx, ctx);

        if (state == PERF_EVENT_STATE_ACTIVE && group_event->attr.exclusive)
                cpuctx->exclusive = 0;
}

struct remove_event {
        struct perf_event *event;
        bool detach_group;
};

/*
 * Cross CPU call to remove a performance event
 *
 * We disable the event on the hardware level first. After that we
 * remove it from the context list.
 */
static int __perf_remove_from_context(void *info)
{
        struct remove_event *re = info;
        struct perf_event *event = re->event;
        struct perf_event_context *ctx = event->ctx;
        struct perf_cpu_context *cpuctx = __get_cpu_context(ctx);

        raw_spin_lock(&ctx->lock);
        event_sched_out(event, cpuctx, ctx);
        if (re->detach_group)
                perf_group_detach(event);
        list_del_event(event, ctx);
        if (!ctx->nr_events && cpuctx->task_ctx == ctx) {
                ctx->is_active = 0;
                cpuctx->task_ctx = NULL;
        }
        raw_spin_unlock(&ctx->lock);

        return 0;
}


/*
 * Remove the event from a task's (or a CPU's) list of events.
 *
 * CPU events are removed with a smp call. For task events we only
 * call when the task is on a CPU.
 *
 * If event->ctx is a cloned context, callers must make sure that
 * every task struct that event->ctx->task could possibly point to
 * remains valid.  This is OK when called from perf_release since
 * that only calls us on the top-level context, which can't be a clone.
 * When called from perf_event_exit_task, it's OK because the
 * context has been detached from its task.
 */
static void perf_remove_from_context(struct perf_event *event, bool detach_group)
{
        struct perf_event_context *ctx = event->ctx;
        struct task_struct *task = ctx->task;
        struct remove_event re = {
                .event = event,
                .detach_group = detach_group,
        };

        lockdep_assert_held(&ctx->mutex);

        if (!task) {
                /*
                 * Per cpu events are removed via an smp call and
                 * the removal is always successful.
                 */
                cpu_function_call(event->cpu, __perf_remove_from_context, &re);
                return;
        }

retry:
        if (!task_function_call(task, __perf_remove_from_context, &re))
                return;

        raw_spin_lock_irq(&ctx->lock);
        /*
         * If we failed to find a running task, but find the context active now
         * that we've acquired the ctx->lock, retry.
         */
        if (ctx->is_active) {
                raw_spin_unlock_irq(&ctx->lock);
                /*
                 * Reload the task pointer, it might have been changed by
                 * a concurrent perf_event_context_sched_out().
                 */
                task = ctx->task;
                goto retry;
        }

        /*
         * Since the task isn't running, its safe to remove the event, us
         * holding the ctx->lock ensures the task won't get scheduled in.
         */
        if (detach_group)
                perf_group_detach(event);
        list_del_event(event, ctx);
        raw_spin_unlock_irq(&ctx->lock);
}

/*
 * Cross CPU call to disable a performance event
 */
int __perf_event_disable(void *info)
{
        struct perf_event *event = info;
        struct perf_event_context *ctx = event->ctx;
        struct perf_cpu_context *cpuctx = __get_cpu_context(ctx);

        /*
         * If this is a per-task event, need to check whether this
         * event's task is the current task on this cpu.
         *
         * Can trigger due to concurrent perf_event_context_sched_out()
         * flipping contexts around.
         */
        if (ctx->task && cpuctx->task_ctx != ctx)
                return -EINVAL;

        raw_spin_lock(&ctx->lock);

        /*
         * If the event is on, turn it off.
         * If it is in error state, leave it in error state.
         */
        if (event->state >= PERF_EVENT_STATE_INACTIVE) {
                update_context_time(ctx);
                update_cgrp_time_from_event(event);
                update_group_times(event);
                if (event == event->group_leader)
                        group_sched_out(event, cpuctx, ctx);
                else
                        event_sched_out(event, cpuctx, ctx);
                event->state = PERF_EVENT_STATE_OFF;
        }

        raw_spin_unlock(&ctx->lock);

        return 0;
}

/*
 * Disable a event.
 *
 * If event->ctx is a cloned context, callers must make sure that
 * every task struct that event->ctx->task could possibly point to
 * remains valid.  This condition is satisifed when called through
 * perf_event_for_each_child or perf_event_for_each because they
 * hold the top-level event's child_mutex, so any descendant that
 * goes to exit will block in sync_child_event.
 * When called from perf_pending_event it's OK because event->ctx
 * is the current context on this CPU and preemption is disabled,
 * hence we can't get into perf_event_task_sched_out for this context.
 */
void perf_event_disable(struct perf_event *event)
{
        struct perf_event_context *ctx = event->ctx;
        struct task_struct *task = ctx->task;

        if (!task) {
                /*
                 * Disable the event on the cpu that it's on
                 */
                cpu_function_call(event->cpu, __perf_event_disable, event);
                return;
        }

retry:
        if (!task_function_call(task, __perf_event_disable, event))
                return;

        raw_spin_lock_irq(&ctx->lock);
        /*
         * If the event is still active, we need to retry the cross-call.
         */
        if (event->state == PERF_EVENT_STATE_ACTIVE) {
                raw_spin_unlock_irq(&ctx->lock);
                /*
                 * Reload the task pointer, it might have been changed by
                 * a concurrent perf_event_context_sched_out().
                 */
                task = ctx->task;
                goto retry;
        }

        /*
         * Since we have the lock this context can't be scheduled
         * in, so we can change the state safely.
         */
        if (event->state == PERF_EVENT_STATE_INACTIVE) {
                update_group_times(event);
                event->state = PERF_EVENT_STATE_OFF;
        }
        raw_spin_unlock_irq(&ctx->lock);
}
EXPORT_SYMBOL_GPL(perf_event_disable);

static void perf_set_shadow_time(struct perf_event *event,
                                 struct perf_event_context *ctx,
                                 u64 tstamp)
{
        /*
         * use the correct time source for the time snapshot
         *
         * We could get by without this by leveraging the
         * fact that to get to this function, the caller
         * has most likely already called update_context_time()
         * and update_cgrp_time_xx() and thus both timestamp
         * are identical (or very close). Given that tstamp is,
         * already adjusted for cgroup, we could say that:
         *    tstamp - ctx->timestamp
         * is equivalent to
         *    tstamp - cgrp->timestamp.
         *
         * Then, in perf_output_read(), the calculation would
         * work with no changes because:
         * - event is guaranteed scheduled in
         * - no scheduled out in between
         * - thus the timestamp would be the same
         *
         * But this is a bit hairy.
         *
         * So instead, we have an explicit cgroup call to remain
         * within the time time source all along. We believe it
         * is cleaner and simpler to understand.
         */
        if (is_cgroup_event(event))
                perf_cgroup_set_shadow_time(event, tstamp);
        else
                event->shadow_ctx_time = tstamp - ctx->timestamp;
}

#define MAX_INTERRUPTS (~0ULL)

static void perf_log_throttle(struct perf_event *event, int enable);

static int
event_sched_in(struct perf_event *event,
                 struct perf_cpu_context *cpuctx,
                 struct perf_event_context *ctx)
{
        u64 tstamp = perf_event_time(event);
        int ret = 0;

        lockdep_assert_held(&ctx->lock);

        if (event->state <= PERF_EVENT_STATE_OFF)
                return 0;

        event->state = PERF_EVENT_STATE_ACTIVE;
        event->oncpu = smp_processor_id();

        /*
         * Unthrottle events, since we scheduled we might have missed several
         * ticks already, also for a heavily scheduling task there is little
         * guarantee it'll get a tick in a timely manner.
         */
        if (unlikely(event->hw.interrupts == MAX_INTERRUPTS)) {
                perf_log_throttle(event, 1);
                event->hw.interrupts = 0;
        }

        /*
         * The new state must be visible before we turn it on in the hardware:
         */
        smp_wmb();

        perf_pmu_disable(event->pmu);

        if (event->pmu->add(event, PERF_EF_START)) {
                event->state = PERF_EVENT_STATE_INACTIVE;
                event->oncpu = -1;
                ret = -EAGAIN;
                goto out;
        }

        event->tstamp_running += tstamp - event->tstamp_stopped;

        perf_set_shadow_time(event, ctx, tstamp);

        if (!is_software_event(event))
                cpuctx->active_oncpu++;
        ctx->nr_active++;
        if (event->attr.freq && event->attr.sample_freq)
                ctx->nr_freq++;

        if (event->attr.exclusive)
                cpuctx->exclusive = 1;

out:
        perf_pmu_enable(event->pmu);

        return ret;
}

static int
group_sched_in(struct perf_event *group_event,
               struct perf_cpu_context *cpuctx,
               struct perf_event_context *ctx)
{
        struct perf_event *event, *partial_group = NULL;
        struct pmu *pmu = ctx->pmu;
        u64 now = ctx->time;
        bool simulate = false;

        if (group_event->state == PERF_EVENT_STATE_OFF)
                return 0;

        pmu->start_txn(pmu);

        if (event_sched_in(group_event, cpuctx, ctx)) {
                pmu->cancel_txn(pmu);
                perf_cpu_hrtimer_restart(cpuctx);
                return -EAGAIN;
        }

        /*
         * Schedule in siblings as one group (if any):
         */
        list_for_each_entry(event, &group_event->sibling_list, group_entry) {
                if (event_sched_in(event, cpuctx, ctx)) {
                        partial_group = event;
                        goto group_error;
                }
        }

        if (!pmu->commit_txn(pmu))
                return 0;

group_error:
        /*
         * Groups can be scheduled in as one unit only, so undo any
         * partial group before returning:
         * The events up to the failed event are scheduled out normally,
         * tstamp_stopped will be updated.
         *
         * The failed events and the remaining siblings need to have
         * their timings updated as if they had gone thru event_sched_in()
         * and event_sched_out(). This is required to get consistent timings
         * across the group. This also takes care of the case where the group
         * could never be scheduled by ensuring tstamp_stopped is set to mark
         * the time the event was actually stopped, such that time delta
         * calculation in update_event_times() is correct.
         */
        list_for_each_entry(event, &group_event->sibling_list, group_entry) {
                if (event == partial_group)
                        simulate = true;

                if (simulate) {
                        event->tstamp_running += now - event->tstamp_stopped;
                        event->tstamp_stopped = now;
                } else {
                        event_sched_out(event, cpuctx, ctx);
                }
        }
        event_sched_out(group_event, cpuctx, ctx);

        pmu->cancel_txn(pmu);

        perf_cpu_hrtimer_restart(cpuctx);

        return -EAGAIN;
}

/*
 * Work out whether we can put this event group on the CPU now.
 */
static int group_can_go_on(struct perf_event *event,
                           struct perf_cpu_context *cpuctx,
                           int can_add_hw)
{
        /*
         * Groups consisting entirely of software events can always go on.
         */
        if (event->group_flags & PERF_GROUP_SOFTWARE)
                return 1;
        /*
         * If an exclusive group is already on, no other hardware
         * events can go on.
         */
        if (cpuctx->exclusive)
                return 0;
        /*
         * If this group is exclusive and there are already
         * events on the CPU, it can't go on.
         */
        if (event->attr.exclusive && cpuctx->active_oncpu)
                return 0;
        /*
         * Otherwise, try to add it if all previous groups were able
         * to go on.
         */
        return can_add_hw;
}

static void add_event_to_ctx(struct perf_event *event,
                               struct perf_event_context *ctx)
{
        u64 tstamp = perf_event_time(event);

        list_add_event(event, ctx);
        perf_group_attach(event);
        event->tstamp_enabled = tstamp;
        event->tstamp_running = tstamp;
        event->tstamp_stopped = tstamp;
}

static void task_ctx_sched_out(struct perf_event_context *ctx);
static void
ctx_sched_in(struct perf_event_context *ctx,
             struct perf_cpu_context *cpuctx,
             enum event_type_t event_type,
             struct task_struct *task);

static void perf_event_sched_in(struct perf_cpu_context *cpuctx,
                                struct perf_event_context *ctx,
                                struct task_struct *task)
{
        cpu_ctx_sched_in(cpuctx, EVENT_PINNED, task);
        if (ctx)
                ctx_sched_in(ctx, cpuctx, EVENT_PINNED, task);
        cpu_ctx_sched_in(cpuctx, EVENT_FLEXIBLE, task);
        if (ctx)
                ctx_sched_in(ctx, cpuctx, EVENT_FLEXIBLE, task);
}

/*
 * Cross CPU call to install and enable a performance event
 *
 * Must be called with ctx->mutex held
 */
static int  __perf_install_in_context(void *info)
{
        struct perf_event *event = info;
        struct perf_event_context *ctx = event->ctx;
        struct perf_cpu_context *cpuctx = __get_cpu_context(ctx);
        struct perf_event_context *task_ctx = cpuctx->task_ctx;
        struct task_struct *task = current;

        perf_ctx_lock(cpuctx, task_ctx);
        perf_pmu_disable(cpuctx->ctx.pmu);

        /*
         * If there was an active task_ctx schedule it out.
         */
        if (task_ctx)
                task_ctx_sched_out(task_ctx);

        /*
         * If the context we're installing events in is not the
         * active task_ctx, flip them.
         */
        if (ctx->task && task_ctx != ctx) {
                if (task_ctx)
                        raw_spin_unlock(&task_ctx->lock);
                raw_spin_lock(&ctx->lock);
                task_ctx = ctx;
        }

        if (task_ctx) {
                cpuctx->task_ctx = task_ctx;
                task = task_ctx->task;
        }

        cpu_ctx_sched_out(cpuctx, EVENT_ALL);

        update_context_time(ctx);
        /*
         * update cgrp time only if current cgrp
         * matches event->cgrp. Must be done before
         * calling add_event_to_ctx()
         */
        update_cgrp_time_from_event(event);

        add_event_to_ctx(event, ctx);

        /*
         * Schedule everything back in
         */
        perf_event_sched_in(cpuctx, task_ctx, task);

        perf_pmu_enable(cpuctx->ctx.pmu);
        perf_ctx_unlock(cpuctx, task_ctx);

        return 0;
}

/*
 * Attach a performance event to a context
 *
 * First we add the event to the list with the hardware enable bit
 * in event->hw_config cleared.
 *
 * If the event is attached to a task which is on a CPU we use a smp
 * call to enable it in the task context. The task might have been
 * scheduled away, but we check this in the smp call again.
 */
static void
perf_install_in_context(struct perf_event_context *ctx,
                        struct perf_event *event,
                        int cpu)
{
        struct task_struct *task = ctx->task;

        lockdep_assert_held(&ctx->mutex);

        event->ctx = ctx;
        if (event->cpu != -1)
                event->cpu = cpu;

        if (!task) {
                /*
                 * Per cpu events are installed via an smp call and
                 * the install is always successful.
                 */
                cpu_function_call(cpu, __perf_install_in_context, event);
                return;
        }

retry:
        if (!task_function_call(task, __perf_install_in_context, event))
                return;

        raw_spin_lock_irq(&ctx->lock);
        /*
         * If we failed to find a running task, but find the context active now
         * that we've acquired the ctx->lock, retry.
         */
        if (ctx->is_active) {
                raw_spin_unlock_irq(&ctx->lock);
                /*
                 * Reload the task pointer, it might have been changed by
                 * a concurrent perf_event_context_sched_out().
                 */
                task = ctx->task;
                goto retry;
        }

        /*
         * Since the task isn't running, its safe to add the event, us holding
         * the ctx->lock ensures the task won't get scheduled in.
         */
        add_event_to_ctx(event, ctx);
        raw_spin_unlock_irq(&ctx->lock);
}

/*
 * Put a event into inactive state and update time fields.
 * Enabling the leader of a group effectively enables all
 * the group members that aren't explicitly disabled, so we
 * have to update their ->tstamp_enabled also.
 * Note: this works for group members as well as group leaders
 * since the non-leader members' sibling_lists will be empty.
 */
static void __perf_event_mark_enabled(struct perf_event *event)
{
        struct perf_event *sub;
        u64 tstamp = perf_event_time(event);

        event->state = PERF_EVENT_STATE_INACTIVE;
        event->tstamp_enabled = tstamp - event->total_time_enabled;
        list_for_each_entry(sub, &event->sibling_list, group_entry) {
                if (sub->state >= PERF_EVENT_STATE_INACTIVE)
                        sub->tstamp_enabled = tstamp - sub->total_time_enabled;
        }
}

/*
 * Cross CPU call to enable a performance event
 */
static int __perf_event_enable(void *info)
{
        struct perf_event *event = info;
        struct perf_event_context *ctx = event->ctx;
        struct perf_event *leader = event->group_leader;
        struct perf_cpu_context *cpuctx = __get_cpu_context(ctx);
        int err;

        /*
         * There's a time window between 'ctx->is_active' check
         * in perf_event_enable function and this place having:
         *   - IRQs on
         *   - ctx->lock unlocked
         *
         * where the task could be killed and 'ctx' deactivated
         * by perf_event_exit_task.
         */
        if (!ctx->is_active)
                return -EINVAL;

        raw_spin_lock(&ctx->lock);
        update_context_time(ctx);

        if (event->state >= PERF_EVENT_STATE_INACTIVE)
                goto unlock;

        /*
         * set current task's cgroup time reference point
         */
        perf_cgroup_set_timestamp(current, ctx);

        __perf_event_mark_enabled(event);

        if (!event_filter_match(event)) {
                if (is_cgroup_event(event))
                        perf_cgroup_defer_enabled(event);
                goto unlock;
        }

        /*
         * If the event is in a group and isn't the group leader,
         * then don't put it on unless the group is on.
         */
        if (leader != event && leader->state != PERF_EVENT_STATE_ACTIVE)
                goto unlock;

        if (!group_can_go_on(event, cpuctx, 1)) {
                err = -EEXIST;
        } else {
                if (event == leader)
                        err = group_sched_in(event, cpuctx, ctx);
                else
                        err = event_sched_in(event, cpuctx, ctx);
        }

        if (err) {
                /*
                 * If this event can't go on and it's part of a
                 * group, then the whole group has to come off.
                 */
                if (leader != event) {
                        group_sched_out(leader, cpuctx, ctx);
                        perf_cpu_hrtimer_restart(cpuctx);
                }
                if (leader->attr.pinned) {
                        update_group_times(leader);
                        leader->state = PERF_EVENT_STATE_ERROR;
                }
        }

unlock:
        raw_spin_unlock(&ctx->lock);

        return 0;
}

/*
 * Enable a event.
 *
 * If event->ctx is a cloned context, callers must make sure that
 * every task struct that event->ctx->task could possibly point to
 * remains valid.  This condition is satisfied when called through
 * perf_event_for_each_child or perf_event_for_each as described
 * for perf_event_disable.
 */
void perf_event_enable(struct perf_event *event)
{
        struct perf_event_context *ctx = event->ctx;
        struct task_struct *task = ctx->task;

        if (!task) {
                /*
                 * Enable the event on the cpu that it's on
                 */
                cpu_function_call(event->cpu, __perf_event_enable, event);
                return;
        }

        raw_spin_lock_irq(&ctx->lock);
        if (event->state >= PERF_EVENT_STATE_INACTIVE)
                goto out;

        /*
         * If the event is in error state, clear that first.
         * That way, if we see the event in error state below, we
         * know that it has gone back into error state, as distinct
         * from the task having been scheduled away before the
         * cross-call arrived.
         */
        if (event->state == PERF_EVENT_STATE_ERROR)
                event->state = PERF_EVENT_STATE_OFF;

retry:
        if (!ctx->is_active) {
                __perf_event_mark_enabled(event);
                goto out;
        }

        raw_spin_unlock_irq(&ctx->lock);

        if (!task_function_call(task, __perf_event_enable, event))
                return;

        raw_spin_lock_irq(&ctx->lock);

        /*
         * If the context is active and the event is still off,
         * we need to retry the cross-call.
         */
        if (ctx->is_active && event->state == PERF_EVENT_STATE_OFF) {
                /*
                 * task could have been flipped by a concurrent
                 * perf_event_context_sched_out()
                 */
                task = ctx->task;
                goto retry;
        }

out:
        raw_spin_unlock_irq(&ctx->lock);
}
EXPORT_SYMBOL_GPL(perf_event_enable);

int perf_event_refresh(struct perf_event *event, int refresh)
{
        /*
         * not supported on inherited events
         */
        if (event->attr.inherit || !is_sampling_event(event))
                return -EINVAL;

        atomic_add(refresh, &event->event_limit);
        perf_event_enable(event);

        return 0;
}
EXPORT_SYMBOL_GPL(perf_event_refresh);

static void ctx_sched_out(struct perf_event_context *ctx,
                          struct perf_cpu_context *cpuctx,
                          enum event_type_t event_type)
{
        struct perf_event *event;
        int is_active = ctx->is_active;

        ctx->is_active &= ~event_type;
        if (likely(!ctx->nr_events))
                return;

        update_context_time(ctx);
        update_cgrp_time_from_cpuctx(cpuctx);
        if (!ctx->nr_active)
                return;

        perf_pmu_disable(ctx->pmu);
        if ((is_active & EVENT_PINNED) && (event_type & EVENT_PINNED)) {
                list_for_each_entry(event, &ctx->pinned_groups, group_entry)
                        group_sched_out(event, cpuctx, ctx);
        }

        if ((is_active & EVENT_FLEXIBLE) && (event_type & EVENT_FLEXIBLE)) {
                list_for_each_entry(event, &ctx->flexible_groups, group_entry)
                        group_sched_out(event, cpuctx, ctx);
        }
        perf_pmu_enable(ctx->pmu);
}

/*
 * Test whether two contexts are equivalent, i.e. whether they have both been
 * cloned from the same version of the same context.
 *
 * Equivalence is measured using a generation number in the context that is
 * incremented on each modification to it; see unclone_ctx(), list_add_event()
 * and list_del_event().
 */
static int context_equiv(struct perf_event_context *ctx1,
                         struct perf_event_context *ctx2)
{
        lockdep_assert_held(&ctx1->lock);
        lockdep_assert_held(&ctx2->lock);

        /* Pinning disables the swap optimization */
        if (ctx1->pin_count || ctx2->pin_count)
                return 0;

        /* If ctx1 is the parent of ctx2 */
        if (ctx1 == ctx2->parent_ctx && ctx1->generation == ctx2->parent_gen)
                return 1;

        /* If ctx2 is the parent of ctx1 */
        if (ctx1->parent_ctx == ctx2 && ctx1->parent_gen == ctx2->generation)
                return 1;

        /*
         * If ctx1 and ctx2 have the same parent; we flatten the parent
         * hierarchy, see perf_event_init_context().
         */
        if (ctx1->parent_ctx && ctx1->parent_ctx == ctx2->parent_ctx &&
                        ctx1->parent_gen == ctx2->parent_gen)
                return 1;

        /* Unmatched */
        return 0;
}

static void __perf_event_sync_stat(struct perf_event *event,
                                     struct perf_event *next_event)
{
        u64 value;

        if (!event->attr.inherit_stat)
                return;

        /*
         * Update the event value, we cannot use perf_event_read()
         * because we're in the middle of a context switch and have IRQs
         * disabled, which upsets smp_call_function_single(), however
         * we know the event must be on the current CPU, therefore we
         * don't need to use it.
         */
        switch (event->state) {
        case PERF_EVENT_STATE_ACTIVE:
                event->pmu->read(event);
                /* fall-through */

        case PERF_EVENT_STATE_INACTIVE:
                update_event_times(event);
                break;

        default:
                break;
        }

        /*
         * In order to keep per-task stats reliable we need to flip the event
         * values when we flip the contexts.
         */
        value = local64_read(&next_event->count);
        value = local64_xchg(&event->count, value);
        local64_set(&next_event->count, value);

        swap(event->total_time_enabled, next_event->total_time_enabled);
        swap(event->total_time_running, next_event->total_time_running);

        /*
         * Since we swizzled the values, update the user visible data too.
         */
        perf_event_update_userpage(event);
        perf_event_update_userpage(next_event);
}

static void perf_event_sync_stat(struct perf_event_context *ctx,
                                   struct perf_event_context *next_ctx)
{
        struct perf_event *event, *next_event;

        if (!ctx->nr_stat)
                return;

        update_context_time(ctx);

        event = list_first_entry(&ctx->event_list,
                                   struct perf_event, event_entry);

        next_event = list_first_entry(&next_ctx->event_list,
                                        struct perf_event, event_entry);

        while (&event->event_entry != &ctx->event_list &&
               &next_event->event_entry != &next_ctx->event_list) {

                __perf_event_sync_stat(event, next_event);

                event = list_next_entry(event, event_entry);
                next_event = list_next_entry(next_event, event_entry);
        }
}

static void perf_event_context_sched_out(struct task_struct *task, int ctxn,
                                         struct task_struct *next)
{
        struct perf_event_context *ctx = task->perf_event_ctxp[ctxn];
        struct perf_event_context *next_ctx;
        struct perf_event_context *parent, *next_parent;
        struct perf_cpu_context *cpuctx;
        int do_switch = 1;

        if (likely(!ctx))
                return;

        cpuctx = __get_cpu_context(ctx);
        if (!cpuctx->task_ctx)
                return;

        rcu_read_lock();
        next_ctx = next->perf_event_ctxp[ctxn];
        if (!next_ctx)
                goto unlock;

        parent = rcu_dereference(ctx->parent_ctx);
        next_parent = rcu_dereference(next_ctx->parent_ctx);

        /* If neither context have a parent context; they cannot be clones. */
        if (!parent || !next_parent)
                goto unlock;

        if (next_parent == ctx || next_ctx == parent || next_parent == parent) {
                /*
                 * Looks like the two contexts are clones, so we might be
                 * able to optimize the context switch.  We lock both
                 * contexts and check that they are clones under the
                 * lock (including re-checking that neither has been
                 * uncloned in the meantime).  It doesn't matter which
                 * order we take the locks because no other cpu could
                 * be trying to lock both of these tasks.
                 */
                raw_spin_lock(&ctx->lock);
                raw_spin_lock_nested(&next_ctx->lock, SINGLE_DEPTH_NESTING);
                if (context_equiv(ctx, next_ctx)) {
                        /*
                         * XXX do we need a memory barrier of sorts
                         * wrt to rcu_dereference() of perf_event_ctxp
                         */
                        task->perf_event_ctxp[ctxn] = next_ctx;
                        next->perf_event_ctxp[ctxn] = ctx;
                        ctx->task = next;
                        next_ctx->task = task;
                        do_switch = 0;

                        perf_event_sync_stat(ctx, next_ctx);
                }
                raw_spin_unlock(&next_ctx->lock);
                raw_spin_unlock(&ctx->lock);
        }
unlock:
        rcu_read_unlock();

        if (do_switch) {
                raw_spin_lock(&ctx->lock);
                ctx_sched_out(ctx, cpuctx, EVENT_ALL);
                cpuctx->task_ctx = NULL;
                raw_spin_unlock(&ctx->lock);
        }
}

#define for_each_task_context_nr(ctxn)                                  \
        for ((ctxn) = 0; (ctxn) < perf_nr_task_contexts; (ctxn)++)

/*
 * Called from scheduler to remove the events of the current task,
 * with interrupts disabled.
 *
 * We stop each event and update the event value in event->count.
 *
 * This does not protect us against NMI, but disable()
 * sets the disabled bit in the control field of event _before_
 * accessing the event control register. If a NMI hits, then it will
 * not restart the event.
 */
void __perf_event_task_sched_out(struct task_struct *task,
                                 struct task_struct *next)
{
        int ctxn;

        for_each_task_context_nr(ctxn)
                perf_event_context_sched_out(task, ctxn, next);

        /*
         * if cgroup events exist on this CPU, then we need
         * to check if we have to switch out PMU state.
         * cgroup event are system-wide mode only
         */
        if (atomic_read(&__get_cpu_var(perf_cgroup_events)))
                perf_cgroup_sched_out(task, next);
}

static void task_ctx_sched_out(struct perf_event_context *ctx)
{
        struct perf_cpu_context *cpuctx = __get_cpu_context(ctx);

        if (!cpuctx->task_ctx)
                return;

        if (WARN_ON_ONCE(ctx != cpuctx->task_ctx))
                return;

        ctx_sched_out(ctx, cpuctx, EVENT_ALL);
        cpuctx->task_ctx = NULL;
}

/*
 * Called with IRQs disabled
 */
static void cpu_ctx_sched_out(struct perf_cpu_context *cpuctx,
                              enum event_type_t event_type)
{
        ctx_sched_out(&cpuctx->ctx, cpuctx, event_type);
}

static void
ctx_pinned_sched_in(struct perf_event_context *ctx,
                    struct perf_cpu_context *cpuctx)
{
        struct perf_event *event;

        list_for_each_entry(event, &ctx->pinned_groups, group_entry) {
                if (event->state <= PERF_EVENT_STATE_OFF)
                        continue;
                if (!event_filter_match(event))
                        continue;

                /* may need to reset tstamp_enabled */
                if (is_cgroup_event(event))
                        perf_cgroup_mark_enabled(event, ctx);

                if (group_can_go_on(event, cpuctx, 1))
                        group_sched_in(event, cpuctx, ctx);

                /*
                 * If this pinned group hasn't been scheduled,
                 * put it in error state.
                 */
                if (event->state == PERF_EVENT_STATE_INACTIVE) {
                        update_group_times(event);
                        event->state = PERF_EVENT_STATE_ERROR;
                }
        }
}

static void
ctx_flexible_sched_in(struct perf_event_context *ctx,
                      struct perf_cpu_context *cpuctx)
{
        struct perf_event *event;
        int can_add_hw = 1;

        list_for_each_entry(event, &ctx->flexible_groups, group_entry) {
                /* Ignore events in OFF or ERROR state */
                if (event->state <= PERF_EVENT_STATE_OFF)
                        continue;
                /*
                 * Listen to the 'cpu' scheduling filter constraint
                 * of events:
                 */
                if (!event_filter_match(event))
                        continue;

                /* may need to reset tstamp_enabled */
                if (is_cgroup_event(event))
                        perf_cgroup_mark_enabled(event, ctx);

                if (group_can_go_on(event, cpuctx, can_add_hw)) {
                        if (group_sched_in(event, cpuctx, ctx))
                                can_add_hw = 0;
                }
        }
}

static void
ctx_sched_in(struct perf_event_context *ctx,
             struct perf_cpu_context *cpuctx,
             enum event_type_t event_type,
             struct task_struct *task)
{
        u64 now;
        int is_active = ctx->is_active;

        ctx->is_active |= event_type;
        if (likely(!ctx->nr_events))
                return;

        now = perf_clock();
        ctx->timestamp = now;
        perf_cgroup_set_timestamp(task, ctx);
        /*
         * First go through the list and put on any pinned groups
         * in order to give them the best chance of going on.
         */
        if (!(is_active & EVENT_PINNED) && (event_type & EVENT_PINNED))
                ctx_pinned_sched_in(ctx, cpuctx);

        /* Then walk through the lower prio flexible groups */
        if (!(is_active & EVENT_FLEXIBLE) && (event_type & EVENT_FLEXIBLE))
                ctx_flexible_sched_in(ctx, cpuctx);
}

static void cpu_ctx_sched_in(struct perf_cpu_context *cpuctx,
                             enum event_type_t event_type,
                             struct task_struct *task)
{
        struct perf_event_context *ctx = &cpuctx->ctx;

        ctx_sched_in(ctx, cpuctx, event_type, task);
}

static void perf_event_context_sched_in(struct perf_event_context *ctx,
                                        struct task_struct *task)
{
        struct perf_cpu_context *cpuctx;

        cpuctx = __get_cpu_context(ctx);
        if (cpuctx->task_ctx == ctx)
                return;

        perf_ctx_lock(cpuctx, ctx);
        perf_pmu_disable(ctx->pmu);
        /*
         * We want to keep the following priority order:
         * cpu pinned (that don't need to move), task pinned,
         * cpu flexible, task flexible.
         */
        cpu_ctx_sched_out(cpuctx, EVENT_FLEXIBLE);

        if (ctx->nr_events)
                cpuctx->task_ctx = ctx;

        perf_event_sched_in(cpuctx, cpuctx->task_ctx, task);

        perf_pmu_enable(ctx->pmu);
        perf_ctx_unlock(cpuctx, ctx);

        /*
         * Since these rotations are per-cpu, we need to ensure the
         * cpu-context we got scheduled on is actually rotating.
         */
        perf_pmu_rotate_start(ctx->pmu);
}

/*
 * When sampling the branck stack in system-wide, it may be necessary
 * to flush the stack on context switch. This happens when the branch
 * stack does not tag its entries with the pid of the current task.
 * Otherwise it becomes impossible to associate a branch entry with a
 * task. This ambiguity is more likely to appear when the branch stack
 * supports priv level filtering and the user sets it to monitor only
 * at the user level (which could be a useful measurement in system-wide
 * mode). In that case, the risk is high of having a branch stack with
 * branch from multiple tasks. Flushing may mean dropping the existing
 * entries or stashing them somewhere in the PMU specific code layer.
 *
 * This function provides the context switch callback to the lower code
 * layer. It is invoked ONLY when there is at least one system-wide context
 * with at least one active event using taken branch sampling.
 */
static void perf_branch_stack_sched_in(struct task_struct *prev,
                                       struct task_struct *task)
{
        struct perf_cpu_context *cpuctx;
        struct pmu *pmu;
        unsigned long flags;

        /* no need to flush branch stack if not changing task */
        if (prev == task)
                return;

        local_irq_save(flags);

        rcu_read_lock();

        list_for_each_entry_rcu(pmu, &pmus, entry) {
                cpuctx = this_cpu_ptr(pmu->pmu_cpu_context);

                /*
                 * check if the context has at least one
                 * event using PERF_SAMPLE_BRANCH_STACK
                 */
                if (cpuctx->ctx.nr_branch_stack > 0
                    && pmu->flush_branch_stack) {

                        perf_ctx_lock(cpuctx, cpuctx->task_ctx);

                        perf_pmu_disable(pmu);

                        pmu->flush_branch_stack();

                        perf_pmu_enable(pmu);

                        perf_ctx_unlock(cpuctx, cpuctx->task_ctx);
                }
        }

        rcu_read_unlock();

        local_irq_restore(flags);
}

/*
 * Called from scheduler to add the events of the current task
 * with interrupts disabled.
 *
 * We restore the event value and then enable it.
 *
 * This does not protect us against NMI, but enable()
 * sets the enabled bit in the control field of event _before_
 * accessing the event control register. If a NMI hits, then it will
 * keep the event running.
 */
void __perf_event_task_sched_in(struct task_struct *prev,
                                struct task_struct *task)
{
        struct perf_event_context *ctx;
        int ctxn;

        for_each_task_context_nr(ctxn) {
                ctx = task->perf_event_ctxp[ctxn];
                if (likely(!ctx))
                        continue;

                perf_event_context_sched_in(ctx, task);
        }
        /*
         * if cgroup events exist on this CPU, then we need
         * to check if we have to switch in PMU state.
         * cgroup event are system-wide mode only
         */
        if (atomic_read(&__get_cpu_var(perf_cgroup_events)))
                perf_cgroup_sched_in(prev, task);

        /* check for system-wide branch_stack events */
        if (atomic_read(&__get_cpu_var(perf_branch_stack_events)))
                perf_branch_stack_sched_in(prev, task);
}

static u64 perf_calculate_period(struct perf_event *event, u64 nsec, u64 count)
{
        u64 frequency = event->attr.sample_freq;
        u64 sec = NSEC_PER_SEC;
        u64 divisor, dividend;

        int count_fls, nsec_fls, frequency_fls, sec_fls;

        count_fls = fls64(count);
        nsec_fls = fls64(nsec);
        frequency_fls = fls64(frequency);
        sec_fls = 30;

        /*
         * We got @count in @nsec, with a target of sample_freq HZ
         * the target period becomes:
         *
         *             @count * 10^9
         * period = -------------------
         *          @nsec * sample_freq
         *
         */

        /*
         * Reduce accuracy by one bit such that @a and @b converge
         * to a similar magnitude.
         */
#define REDUCE_FLS(a, b)                \
do {                                    \
        if (a##_fls > b##_fls) {        \
                a >>= 1;                \
                a##_fls--;              \
        } else {                        \
                b >>= 1;                \
                b##_fls--;              \
        }                               \
} while (0)

        /*
         * Reduce accuracy until either term fits in a u64, then proceed with
         * the other, so that finally we can do a u64/u64 division.
         */
        while (count_fls + sec_fls > 64 && nsec_fls + frequency_fls > 64) {
                REDUCE_FLS(nsec, frequency);
                REDUCE_FLS(sec, count);
        }

        if (count_fls + sec_fls > 64) {
                divisor = nsec * frequency;

                while (count_fls + sec_fls > 64) {
                        REDUCE_FLS(count, sec);
                        divisor >>= 1;
                }

                dividend = count * sec;
        } else {
                dividend = count * sec;

                while (nsec_fls + frequency_fls > 64) {
                        REDUCE_FLS(nsec, frequency);
                        dividend >>= 1;
                }

                divisor = nsec * frequency;
        }

        if (!divisor)
                return dividend;

        return div64_u64(dividend, divisor);
}

static DEFINE_PER_CPU(int, perf_throttled_count);
static DEFINE_PER_CPU(u64, perf_throttled_seq);

static void perf_adjust_period(struct perf_event *event, u64 nsec, u64 count, bool disable)
{
        struct hw_perf_event *hwc = &event->hw;
        s64 period, sample_period;
        s64 delta;

        period = perf_calculate_period(event, nsec, count);

        delta = (s64)(period - hwc->sample_period);
        delta = (delta + 7) / 8; /* low pass filter */

        sample_period = hwc->sample_period + delta;

        if (!sample_period)
                sample_period = 1;

        hwc->sample_period = sample_period;

        if (local64_read(&hwc->period_left) > 8*sample_period) {
                if (disable)
                        event->pmu->stop(event, PERF_EF_UPDATE);

                local64_set(&hwc->period_left, 0);

                if (disable)
                        event->pmu->start(event, PERF_EF_RELOAD);
        }
}

/*
 * combine freq adjustment with unthrottling to avoid two passes over the
 * events. At the same time, make sure, having freq events does not change
 * the rate of unthrottling as that would introduce bias.
 */
static void perf_adjust_freq_unthr_context(struct perf_event_context *ctx,
                                           int needs_unthr)
{
        struct perf_event *event;
        struct hw_perf_event *hwc;
        u64 now, period = TICK_NSEC;
        s64 delta;

        /*
         * only need to iterate over all events iff:
         * - context have events in frequency mode (needs freq adjust)
         * - there are events to unthrottle on this cpu
         */
        if (!(ctx->nr_freq || needs_unthr))
                return;

        raw_spin_lock(&ctx->lock);
        perf_pmu_disable(ctx->pmu);

        list_for_each_entry_rcu(event, &ctx->event_list, event_entry) {
                if (event->state != PERF_EVENT_STATE_ACTIVE)
                        continue;

                if (!event_filter_match(event))
                        continue;

                perf_pmu_disable(event->pmu);

                hwc = &event->hw;

                if (hwc->interrupts == MAX_INTERRUPTS) {
                        hwc->interrupts = 0;
                        perf_log_throttle(event, 1);
                        event->pmu->start(event, 0);
                }

                if (!event->attr.freq || !event->attr.sample_freq)
                        goto next;

                /*
                 * stop the event and update event->count
                 */
                event->pmu->stop(event, PERF_EF_UPDATE);

                now = local64_read(&event->count);
                delta = now - hwc->freq_count_stamp;
                hwc->freq_count_stamp = now;

                /*
                 * restart the event
                 * reload only if value has changed
                 * we have stopped the event so tell that
                 * to perf_adjust_period() to avoid stopping it
                 * twice.
                 */
                if (delta > 0)
                        perf_adjust_period(event, period, delta, false);

                event->pmu->start(event, delta > 0 ? PERF_EF_RELOAD : 0);
        next:
                perf_pmu_enable(event->pmu);
        }

        perf_pmu_enable(ctx->pmu);
        raw_spin_unlock(&ctx->lock);
}

/*
 * Round-robin a context's events:
 */
static void rotate_ctx(struct perf_event_context *ctx)
{
        /*
         * Rotate the first entry last of non-pinned groups. Rotation might be
         * disabled by the inheritance code.
         */
        if (!ctx->rotate_disable)
                list_rotate_left(&ctx->flexible_groups);
}

/*
 * perf_pmu_rotate_start() and perf_rotate_context() are fully serialized
 * because they're strictly cpu affine and rotate_start is called with IRQs
 * disabled, while rotate_context is called from IRQ context.
 */
static int perf_rotate_context(struct perf_cpu_context *cpuctx)
{
        struct perf_event_context *ctx = NULL;
        int rotate = 0, remove = 1;

        if (cpuctx->ctx.nr_events) {
                remove = 0;
                if (cpuctx->ctx.nr_events != cpuctx->ctx.nr_active)
                        rotate = 1;
        }

        ctx = cpuctx->task_ctx;
        if (ctx && ctx->nr_events) {
                remove = 0;
                if (ctx->nr_events != ctx->nr_active)
                        rotate = 1;
        }

        if (!rotate)
                goto done;

        perf_ctx_lock(cpuctx, cpuctx->task_ctx);
        perf_pmu_disable(cpuctx->ctx.pmu);

        cpu_ctx_sched_out(cpuctx, EVENT_FLEXIBLE);
        if (ctx)
                ctx_sched_out(ctx, cpuctx, EVENT_FLEXIBLE);

        rotate_ctx(&cpuctx->ctx);
        if (ctx)
                rotate_ctx(ctx);

        perf_event_sched_in(cpuctx, ctx, current);

        perf_pmu_enable(cpuctx->ctx.pmu);
        perf_ctx_unlock(cpuctx, cpuctx->task_ctx);
done:
        if (remove)
                list_del_init(&cpuctx->rotation_list);

        return rotate;
}

#ifdef CONFIG_NO_HZ_FULL
bool perf_event_can_stop_tick(void)
{
        if (atomic_read(&nr_freq_events) ||
            __this_cpu_read(perf_throttled_count))
                return false;
        else
                return true;
}
#endif

void perf_event_task_tick(void)
{
        struct list_head *head = &__get_cpu_var(rotation_list);
        struct perf_cpu_context *cpuctx, *tmp;
        struct perf_event_context *ctx;
        int throttled;

        WARN_ON(!irqs_disabled());

        __this_cpu_inc(perf_throttled_seq);
        throttled = __this_cpu_xchg(perf_throttled_count, 0);

        list_for_each_entry_safe(cpuctx, tmp, head, rotation_list) {
                ctx = &cpuctx->ctx;
                perf_adjust_freq_unthr_context(ctx, throttled);

                ctx = cpuctx->task_ctx;
                if (ctx)
                        perf_adjust_freq_unthr_context(ctx, throttled);
        }
}

static int event_enable_on_exec(struct perf_event *event,
                                struct perf_event_context *ctx)
{
        if (!event->attr.enable_on_exec)
                return 0;

        event->attr.enable_on_exec = 0;
        if (event->state >= PERF_EVENT_STATE_INACTIVE)
                return 0;

        __perf_event_mark_enabled(event);

        return 1;
}

/*
 * Enable all of a task's events that have been marked enable-on-exec.
 * This expects task == current.
 */
static void perf_event_enable_on_exec(struct perf_event_context *ctx)
{
        struct perf_event_context *clone_ctx = NULL;
        struct perf_event *event;
        unsigned long flags;
        int enabled = 0;
        int ret;

        local_irq_save(flags);
        if (!ctx || !ctx->nr_events)
                goto out;

        /*
         * We must ctxsw out cgroup events to avoid conflict
         * when invoking perf_task_event_sched_in() later on
         * in this function. Otherwise we end up trying to
         * ctxswin cgroup events which are already scheduled
         * in.
         */
        perf_cgroup_sched_out(current, NULL);

        raw_spin_lock(&ctx->lock);
        task_ctx_sched_out(ctx);

        list_for_each_entry(event, &ctx->event_list, event_entry) {
                ret = event_enable_on_exec(event, ctx);
                if (ret)
                        enabled = 1;
        }

        /*
         * Unclone this context if we enabled any event.
         */
        if (enabled)
                clone_ctx = unclone_ctx(ctx);

        raw_spin_unlock(&ctx->lock);

        /*
         * Also calls ctxswin for cgroup events, if any:
         */
        perf_event_context_sched_in(ctx, ctx->task);
out:
        local_irq_restore(flags);

        if (clone_ctx)
                put_ctx(clone_ctx);
}

void perf_event_exec(void)
{
        struct perf_event_context *ctx;
        int ctxn;

        rcu_read_lock();
        for_each_task_context_nr(ctxn) {
                ctx = current->perf_event_ctxp[ctxn];
                if (!ctx)
                        continue;

                perf_event_enable_on_exec(ctx);
        }
        rcu_read_unlock();
}

/*
 * Cross CPU call to read the hardware event
 */
static void __perf_event_read(void *info)
{
        struct perf_event *event = info;
        struct perf_event_context *ctx = event->ctx;
        struct perf_cpu_context *cpuctx = __get_cpu_context(ctx);

        /*
         * If this is a task context, we need to check whether it is
         * the current task context of this cpu.  If not it has been
         * scheduled out before the smp call arrived.  In that case
         * event->count would have been updated to a recent sample
         * when the event was scheduled out.
         */
        if (ctx->task && cpuctx->task_ctx != ctx)
                return;

        raw_spin_lock(&ctx->lock);
        if (ctx->is_active) {
                update_context_time(ctx);
                update_cgrp_time_from_event(event);
        }
        update_event_times(event);
        if (event->state == PERF_EVENT_STATE_ACTIVE)
                event->pmu->read(event);
        raw_spin_unlock(&ctx->lock);
}

static inline u64 perf_event_count(struct perf_event *event)
{
        return local64_read(&event->count) + atomic64_read(&event->child_count);
}

static u64 perf_event_read(struct perf_event *event)
{
        /*
         * If event is enabled and currently active on a CPU, update the
         * value in the event structure:
         */
        if (event->state == PERF_EVENT_STATE_ACTIVE) {
                smp_call_function_single(event->oncpu,
                                         __perf_event_read, event, 1);
        } else if (event->state == PERF_EVENT_STATE_INACTIVE) {
                struct perf_event_context *ctx = event->ctx;
                unsigned long flags;

                raw_spin_lock_irqsave(&ctx->lock, flags);
                /*
                 * may read while context is not active
                 * (e.g., thread is blocked), in that case
                 * we cannot update context time
                 */
                if (ctx->is_active) {
                        update_context_time(ctx);
                        update_cgrp_time_from_event(event);
                }
                update_event_times(event);
                raw_spin_unlock_irqrestore(&ctx->lock, flags);
        }

        return perf_event_count(event);
}

/*
 * Initialize the perf_event context in a task_struct:
 */
static void __perf_event_init_context(struct perf_event_context *ctx)
{
        raw_spin_lock_init(&ctx->lock);
        mutex_init(&ctx->mutex);
        INIT_LIST_HEAD(&ctx->pinned_groups);
        INIT_LIST_HEAD(&ctx->flexible_groups);
        INIT_LIST_HEAD(&ctx->event_list);
        atomic_set(&ctx->refcount, 1);
}

static struct perf_event_context *
alloc_perf_context(struct pmu *pmu, struct task_struct *task)
{
        struct perf_event_context *ctx;

        ctx = kzalloc(sizeof(struct perf_event_context), GFP_KERNEL);
        if (!ctx)
                return NULL;

        __perf_event_init_context(ctx);
        if (task) {
                ctx->task = task;
                get_task_struct(task);
        }
        ctx->pmu = pmu;

        return ctx;
}

static struct task_struct *
find_lively_task_by_vpid(pid_t vpid)
{
        struct task_struct *task;
        int err;

        rcu_read_lock();
        if (!vpid)
                task = current;
        else
                task = find_task_by_vpid(vpid);
        if (task)
                get_task_struct(task);
        rcu_read_unlock();

        if (!task)
                return ERR_PTR(-ESRCH);

        /* Reuse ptrace permission checks for now. */
        err = -EACCES;
        if (!ptrace_may_access(task, PTRACE_MODE_READ))
                goto errout;

        return task;
errout:
        put_task_struct(task);
        return ERR_PTR(err);

}

/*
 * Returns a matching context with refcount and pincount.
 */
static struct perf_event_context *
find_get_context(struct pmu *pmu, struct task_struct *task, int cpu)
{
        struct perf_event_context *ctx, *clone_ctx = NULL;
        struct perf_cpu_context *cpuctx;
        unsigned long flags;
        int ctxn, err;

        if (!task) {
                /* Must be root to operate on a CPU event: */
                if (perf_paranoid_cpu() && !capable(CAP_SYS_ADMIN))
                        return ERR_PTR(-EACCES);

                /*
                 * We could be clever and allow to attach a event to an
                 * offline CPU and activate it when the CPU comes up, but
                 * that's for later.
                 */
                if (!cpu_online(cpu))
                        return ERR_PTR(-ENODEV);

                cpuctx = per_cpu_ptr(pmu->pmu_cpu_context, cpu);
                ctx = &cpuctx->ctx;
                get_ctx(ctx);
                ++ctx->pin_count;

                return ctx;
        }

        err = -EINVAL;
        ctxn = pmu->task_ctx_nr;
        if (ctxn < 0)
                goto errout;

retry:
        ctx = perf_lock_task_context(task, ctxn, &flags);
        if (ctx) {
                clone_ctx = unclone_ctx(ctx);
                ++ctx->pin_count;
                raw_spin_unlock_irqrestore(&ctx->lock, flags);

                if (clone_ctx)
                        put_ctx(clone_ctx);
        } else {
                ctx = alloc_perf_context(pmu, task);
                err = -ENOMEM;
                if (!ctx)
                        goto errout;

                err = 0;
                mutex_lock(&task->perf_event_mutex);
                /*
                 * If it has already passed perf_event_exit_task().
                 * we must see PF_EXITING, it takes this mutex too.
                 */
                if (task->flags & PF_EXITING)
                        err = -ESRCH;
                else if (task->perf_event_ctxp[ctxn])
                        err = -EAGAIN;
                else {
                        get_ctx(ctx);
                        ++ctx->pin_count;
                        rcu_assign_pointer(task->perf_event_ctxp[ctxn], ctx);
                }
                mutex_unlock(&task->perf_event_mutex);

                if (unlikely(err)) {
                        put_ctx(ctx);

                        if (err == -EAGAIN)
                                goto retry;
                        goto errout;
                }
        }

        return ctx;

errout:
        return ERR_PTR(err);
}

static void perf_event_free_filter(struct perf_event *event);

static void free_event_rcu(struct rcu_head *head)
{
        struct perf_event *event;

        event = container_of(head, struct perf_event, rcu_head);
        if (event->ns)
                put_pid_ns(event->ns);
        perf_event_free_filter(event);
        kfree(event);
}

static void ring_buffer_put(struct ring_buffer *rb);
static void ring_buffer_attach(struct perf_event *event,
                               struct ring_buffer *rb);

static void unaccount_event_cpu(struct perf_event *event, int cpu)
{
        if (event->parent)
                return;

        if (has_branch_stack(event)) {
                if (!(event->attach_state & PERF_ATTACH_TASK))
                        atomic_dec(&per_cpu(perf_branch_stack_events, cpu));
        }
        if (is_cgroup_event(event))
                atomic_dec(&per_cpu(perf_cgroup_events, cpu));
}

static void unaccount_event(struct perf_event *event)
{
        if (event->parent)
                return;

        if (event->attach_state & PERF_ATTACH_TASK)
                static_key_slow_dec_deferred(&perf_sched_events);
        if (event->attr.mmap || event->attr.mmap_data)
                atomic_dec(&nr_mmap_events);
        if (event->attr.comm)
                atomic_dec(&nr_comm_events);
        if (event->attr.task)
                atomic_dec(&nr_task_events);
        if (event->attr.freq)
                atomic_dec(&nr_freq_events);
        if (is_cgroup_event(event))
                static_key_slow_dec_deferred(&perf_sched_events);
        if (has_branch_stack(event))
                static_key_slow_dec_deferred(&perf_sched_events);

        unaccount_event_cpu(event, event->cpu);
}

static void __free_event(struct perf_event *event)
{
        if (!event->parent) {
                if (event->attr.sample_type & PERF_SAMPLE_CALLCHAIN)
                        put_callchain_buffers();
        }

        if (event->destroy)
                event->destroy(event);

        if (event->ctx)
                put_ctx(event->ctx);

        if (event->pmu)
                module_put(event->pmu->module);

        call_rcu(&event->rcu_head, free_event_rcu);
}

static void _free_event(struct perf_event *event)
{
        irq_work_sync(&event->pending);

        unaccount_event(event);

        if (event->rb) {
                /*
                 * Can happen when we close an event with re-directed output.
                 *
                 * Since we have a 0 refcount, perf_mmap_close() will skip
                 * over us; possibly making our ring_buffer_put() the last.
                 */
                mutex_lock(&event->mmap_mutex);
                ring_buffer_attach(event, NULL);
                mutex_unlock(&event->mmap_mutex);
        }

        if (is_cgroup_event(event))
                perf_detach_cgroup(event);

        __free_event(event);
}

/*
 * Used to free events which have a known refcount of 1, such as in error paths
 * where the event isn't exposed yet and inherited events.
 */
static void free_event(struct perf_event *event)
{
        if (WARN(atomic_long_cmpxchg(&event->refcount, 1, 0) != 1,
                                "unexpected event refcount: %ld; ptr=%p\n",
                                atomic_long_read(&event->refcount), event)) {
                /* leak to avoid use-after-free */
                return;
        }

        _free_event(event);
}

/*
 * Called when the last reference to the file is gone.
 */
static void put_event(struct perf_event *event)
{
        struct perf_event_context *ctx = event->ctx;
        struct task_struct *owner;

        if (!atomic_long_dec_and_test(&event->refcount))
                return;

        rcu_read_lock();
        owner = ACCESS_ONCE(event->owner);
        /*
         * Matches the smp_wmb() in perf_event_exit_task(). If we observe
         * !owner it means the list deletion is complete and we can indeed
         * free this event, otherwise we need to serialize on
         * owner->perf_event_mutex.
         */
        smp_read_barrier_depends();
        if (owner) {
                /*
                 * Since delayed_put_task_struct() also drops the last
                 * task reference we can safely take a new reference
                 * while holding the rcu_read_lock().
                 */
                get_task_struct(owner);
        }
        rcu_read_unlock();

        if (owner) {
                mutex_lock(&owner->perf_event_mutex);
                /*
                 * We have to re-check the event->owner field, if it is cleared
                 * we raced with perf_event_exit_task(), acquiring the mutex
                 * ensured they're done, and we can proceed with freeing the
                 * event.
                 */
                if (event->owner)
                        list_del_init(&event->owner_entry);
                mutex_unlock(&owner->perf_event_mutex);
                put_task_struct(owner);
        }

        WARN_ON_ONCE(ctx->parent_ctx);
        /*
         * There are two ways this annotation is useful:
         *
         *  1) there is a lock recursion from perf_event_exit_task
         *     see the comment there.
         *
         *  2) there is a lock-inversion with mmap_sem through
         *     perf_event_read_group(), which takes faults while
         *     holding ctx->mutex, however this is called after
         *     the last filedesc died, so there is no possibility
         *     to trigger the AB-BA case.
         */
        mutex_lock_nested(&ctx->mutex, SINGLE_DEPTH_NESTING);
        perf_remove_from_context(event, true);
        mutex_unlock(&ctx->mutex);

        _free_event(event);
}

int perf_event_release_kernel(struct perf_event *event)
{
        put_event(event);
        return 0;
}
EXPORT_SYMBOL_GPL(perf_event_release_kernel);

static int perf_release(struct inode *inode, struct file *file)
{
        put_event(file->private_data);
        return 0;
}

u64 perf_event_read_value(struct perf_event *event, u64 *enabled, u64 *running)
{
        struct perf_event *child;
        u64 total = 0;