if (ret || !write)
return ret;
- if (sysctl_perf_cpu_time_max_percent == 100) {
+ if (sysctl_perf_cpu_time_max_percent == 100 ||
+ sysctl_perf_cpu_time_max_percent == 0) {
printk(KERN_WARNING
"perf: Dynamic interrupt throttling disabled, can hang your system!\n");
WRITE_ONCE(perf_sample_allowed_ns, 0);
* function.
*
* Lock order:
+ * cred_guard_mutex
* task_struct::perf_event_mutex
* perf_event_context::mutex
* perf_event::child_mutex;
if (event->state <= PERF_EVENT_STATE_OFF)
return 0;
- event->state = PERF_EVENT_STATE_ACTIVE;
- event->oncpu = smp_processor_id();
+ WRITE_ONCE(event->oncpu, smp_processor_id());
+ /*
+ * Order event::oncpu write to happen before the ACTIVE state
+ * is visible.
+ */
+ smp_wmb();
+ WRITE_ONCE(event->state, PERF_EVENT_STATE_ACTIVE);
/*
* Unthrottle events, since we scheduled we might have missed several
}
EXPORT_SYMBOL_GPL(perf_event_enable);
+static int __perf_event_stop(void *info)
+{
+ struct perf_event *event = info;
+
+ /* for AUX events, our job is done if the event is already inactive */
+ if (READ_ONCE(event->state) != PERF_EVENT_STATE_ACTIVE)
+ return 0;
+
+ /* matches smp_wmb() in event_sched_in() */
+ smp_rmb();
+
+ /*
+ * There is a window with interrupts enabled before we get here,
+ * so we need to check again lest we try to stop another CPU's event.
+ */
+ if (READ_ONCE(event->oncpu) != smp_processor_id())
+ return -EAGAIN;
+
+ event->pmu->stop(event, PERF_EF_UPDATE);
+
+ return 0;
+}
+
static int _perf_event_refresh(struct perf_event *event, int refresh)
{
/*
find_lively_task_by_vpid(pid_t vpid)
{
struct task_struct *task;
- int err;
rcu_read_lock();
if (!vpid)
if (!task)
return ERR_PTR(-ESRCH);
- /* Reuse ptrace permission checks for now. */
- err = -EACCES;
- if (!ptrace_may_access(task, PTRACE_MODE_READ_REALCREDS))
- goto errout;
-
return task;
-errout:
- put_task_struct(task);
- return ERR_PTR(err);
-
}
/*
case PERF_EVENT_IOC_SET_BPF:
return perf_event_set_bpf_prog(event, arg);
+ case PERF_EVENT_IOC_PAUSE_OUTPUT: {
+ struct ring_buffer *rb;
+
+ rcu_read_lock();
+ rb = rcu_dereference(event->rb);
+ if (!rb || !rb->nr_pages) {
+ rcu_read_unlock();
+ return -EINVAL;
+ }
+ rb_toggle_paused(rb, !!arg);
+ rcu_read_unlock();
+ return 0;
+ }
default:
return -ENOTTY;
}
event->pmu->event_mapped(event);
}
+static void perf_pmu_output_stop(struct perf_event *event);
+
/*
* A buffer can be mmap()ed multiple times; either directly through the same
* event, or through other events by use of perf_event_set_output().
*/
if (rb_has_aux(rb) && vma->vm_pgoff == rb->aux_pgoff &&
atomic_dec_and_mutex_lock(&rb->aux_mmap_count, &event->mmap_mutex)) {
+ /*
+ * Stop all AUX events that are writing to this buffer,
+ * so that we can free its AUX pages and corresponding PMU
+ * data. Note that after rb::aux_mmap_count dropped to zero,
+ * they won't start any more (see perf_aux_output_begin()).
+ */
+ perf_pmu_output_stop(event);
+
+ /* now it's safe to free the pages */
atomic_long_sub(rb->aux_nr_pages, &mmap_user->locked_vm);
vma->vm_mm->pinned_vm -= rb->aux_mmap_locked;
+ /* this has to be the last one */
rb_free_aux(rb);
+ WARN_ON_ONCE(atomic_read(&rb->aux_refcount));
+
mutex_unlock(&event->mmap_mutex);
}
}
}
-void perf_event_output(struct perf_event *event,
- struct perf_sample_data *data,
- struct pt_regs *regs)
+static void __always_inline
+__perf_event_output(struct perf_event *event,
+ struct perf_sample_data *data,
+ struct pt_regs *regs,
+ int (*output_begin)(struct perf_output_handle *,
+ struct perf_event *,
+ unsigned int))
{
struct perf_output_handle handle;
struct perf_event_header header;
perf_prepare_sample(&header, data, event, regs);
- if (perf_output_begin(&handle, event, header.size))
+ if (output_begin(&handle, event, header.size))
goto exit;
perf_output_sample(&handle, &header, data, event);
rcu_read_unlock();
}
+void
+perf_event_output_forward(struct perf_event *event,
+ struct perf_sample_data *data,
+ struct pt_regs *regs)
+{
+ __perf_event_output(event, data, regs, perf_output_begin_forward);
+}
+
+void
+perf_event_output_backward(struct perf_event *event,
+ struct perf_sample_data *data,
+ struct pt_regs *regs)
+{
+ __perf_event_output(event, data, regs, perf_output_begin_backward);
+}
+
+void
+perf_event_output(struct perf_event *event,
+ struct perf_sample_data *data,
+ struct pt_regs *regs)
+{
+ __perf_event_output(event, data, regs, perf_output_begin);
+}
+
/*
* read event_id
*/
rcu_read_unlock();
}
+struct remote_output {
+ struct ring_buffer *rb;
+ int err;
+};
+
+static void __perf_event_output_stop(struct perf_event *event, void *data)
+{
+ struct perf_event *parent = event->parent;
+ struct remote_output *ro = data;
+ struct ring_buffer *rb = ro->rb;
+
+ if (!has_aux(event))
+ return;
+
+ if (!parent)
+ parent = event;
+
+ /*
+ * In case of inheritance, it will be the parent that links to the
+ * ring-buffer, but it will be the child that's actually using it:
+ */
+ if (rcu_dereference(parent->rb) == rb)
+ ro->err = __perf_event_stop(event);
+}
+
+static int __perf_pmu_output_stop(void *info)
+{
+ struct perf_event *event = info;
+ struct pmu *pmu = event->pmu;
+ struct perf_cpu_context *cpuctx = get_cpu_ptr(pmu->pmu_cpu_context);
+ struct remote_output ro = {
+ .rb = event->rb,
+ };
+
+ rcu_read_lock();
+ perf_event_aux_ctx(&cpuctx->ctx, __perf_event_output_stop, &ro);
+ if (cpuctx->task_ctx)
+ perf_event_aux_ctx(cpuctx->task_ctx, __perf_event_output_stop,
+ &ro);
+ rcu_read_unlock();
+
+ return ro.err;
+}
+
+static void perf_pmu_output_stop(struct perf_event *event)
+{
+ struct perf_event *iter;
+ int err, cpu;
+
+restart:
+ rcu_read_lock();
+ list_for_each_entry_rcu(iter, &event->rb->event_list, rb_entry) {
+ /*
+ * For per-CPU events, we need to make sure that neither they
+ * nor their children are running; for cpu==-1 events it's
+ * sufficient to stop the event itself if it's active, since
+ * it can't have children.
+ */
+ cpu = iter->cpu;
+ if (cpu == -1)
+ cpu = READ_ONCE(iter->oncpu);
+
+ if (cpu == -1)
+ continue;
+
+ err = cpu_function_call(cpu, __perf_pmu_output_stop, event);
+ if (err == -EAGAIN) {
+ rcu_read_unlock();
+ goto restart;
+ }
+ }
+ rcu_read_unlock();
+}
+
/*
* task tracking -- fork/exit
*
irq_work_queue(&event->pending);
}
- if (event->overflow_handler)
- event->overflow_handler(event, data, regs);
- else
- perf_event_output(event, data, regs);
+ event->overflow_handler(event, data, regs);
if (*perf_event_fasync(event) && event->pending_kill) {
event->pending_wakeup = 1;
}
skip_type:
+ if (pmu->task_ctx_nr == perf_hw_context) {
+ static int hw_context_taken = 0;
+
+ if (WARN_ON_ONCE(hw_context_taken))
+ pmu->task_ctx_nr = perf_invalid_context;
+
+ hw_context_taken = 1;
+ }
+
pmu->pmu_cpu_context = find_pmu_context(pmu->task_ctx_nr);
if (pmu->pmu_cpu_context)
goto got_cpu_context;
context = parent_event->overflow_handler_context;
}
- event->overflow_handler = overflow_handler;
- event->overflow_handler_context = context;
+ if (overflow_handler) {
+ event->overflow_handler = overflow_handler;
+ event->overflow_handler_context = context;
+ } else if (is_write_backward(event)){
+ event->overflow_handler = perf_event_output_backward;
+ event->overflow_handler_context = NULL;
+ } else {
+ event->overflow_handler = perf_event_output_forward;
+ event->overflow_handler_context = NULL;
+ }
perf_event__state_init(event);
if (output_event->clock != event->clock)
goto out;
+ /*
+ * Either writing ring buffer from beginning or from end.
+ * Mixing is not allowed.
+ */
+ if (is_write_backward(output_event) != is_write_backward(event))
+ goto out;
+
/*
* If both events generate aux data, they must be on the same PMU
*/
get_online_cpus();
+ if (task) {
+ err = mutex_lock_interruptible(&task->signal->cred_guard_mutex);
+ if (err)
+ goto err_cpus;
+
+ /*
+ * Reuse ptrace permission checks for now.
+ *
+ * We must hold cred_guard_mutex across this and any potential
+ * perf_install_in_context() call for this new event to
+ * serialize against exec() altering our credentials (and the
+ * perf_event_exit_task() that could imply).
+ */
+ err = -EACCES;
+ if (!ptrace_may_access(task, PTRACE_MODE_READ_REALCREDS))
+ goto err_cred;
+ }
+
if (flags & PERF_FLAG_PID_CGROUP)
cgroup_fd = pid;
NULL, NULL, cgroup_fd);
if (IS_ERR(event)) {
err = PTR_ERR(event);
- goto err_cpus;
+ goto err_cred;
}
if (is_sampling_event(event)) {
goto err_context;
}
- if (task) {
- put_task_struct(task);
- task = NULL;
- }
-
/*
* Look up the group leader (we will attach this event to it):
*/
WARN_ON_ONCE(ctx->parent_ctx);
+ /*
+ * This is the point on no return; we cannot fail hereafter. This is
+ * where we start modifying current state.
+ */
+
if (move_group) {
/*
* See perf_event_ctx_lock() for comments on the details
mutex_unlock(&gctx->mutex);
mutex_unlock(&ctx->mutex);
+ if (task) {
+ mutex_unlock(&task->signal->cred_guard_mutex);
+ put_task_struct(task);
+ }
+
put_online_cpus();
mutex_lock(¤t->perf_event_mutex);
*/
if (!event_file)
free_event(event);
+err_cred:
+ if (task)
+ mutex_unlock(&task->signal->cred_guard_mutex);
err_cpus:
put_online_cpus();
err_task:
/*
* When a child task exits, feed back event values to parent events.
+ *
+ * Can be called with cred_guard_mutex held when called from
+ * install_exec_creds().
*/
void perf_event_exit_task(struct task_struct *child)
{
projects / deliverable / linux.git / blobdiff