Merge branch 'perfcounters-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

* 'perfcounters-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (27 commits)
  perf_counter: Zero dead bytes from ftrace raw samples size alignment
  perf_counter: Subtract the buffer size field from the event record size
  perf_counter: Require CAP_SYS_ADMIN for raw tracepoint data
  perf_counter: Correct PERF_SAMPLE_RAW output
  perf tools: callchain: Fix bad rounding of minimum rate
  perf_counter tools: Fix libbfd detection for systems with libz dependency
  perf: "Longum est iter per praecepta, breve et efficax per exempla"
  perf_counter: Fix a race on perf_counter_ctx
  perf_counter: Fix tracepoint sampling to be part of generic sampling
  perf_counter: Work around gcc warning by initializing tracepoint record unconditionally
  perf tools: callchain: Fix sum of percentages to be 100% by displaying amount of ignored chains in fractal mode
  perf tools: callchain: Fix 'perf report' display to be callchain by default
  perf tools: callchain: Fix spurious 'perf report' warnings: ignore empty callchains
  perf record: Fix the -A UI for empty or non-existent perf.data
  perf util: Fix do_read() to fail on EOF instead of busy-looping
  perf list: Fix the output to not include tracepoints without an id
  perf_counter/powerpc: Fix oops on cpus without perf_counter hardware support
  perf stat: Fix tool option consistency: rename -S/--scale to -c/--scale
  perf report: Add debug help for the finding of symbol bugs - show the symtab origin (DSO, build-id, kernel, etc)
  perf report: Fix per task mult-counter stat reporting
  ...

1 files changed, 145 insertions, 94 deletions

diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c
index 868102172aa..b0b20a07f39 100644
--- a/kernel/perf_counter.c
+++ b/kernel/perf_counter.c
@@ -2646,7 +2646,6 @@ static void perf_counter_output(struct perf_counter *counter, int nmi,
 		u64 counter;
 	} group_entry;
 	struct perf_callchain_entry *callchain = NULL;
-	struct perf_tracepoint_record *tp;
 	int callchain_size = 0;
 	u64 time;
 	struct {
@@ -2715,9 +2714,16 @@ static void perf_counter_output(struct perf_counter *counter, int nmi,
 			header.size += sizeof(u64);
 	}
 
-	if (sample_type & PERF_SAMPLE_TP_RECORD) {
-		tp = data->private;
-		header.size += tp->size;
+	if (sample_type & PERF_SAMPLE_RAW) {
+		int size = sizeof(u32);
+
+		if (data->raw)
+			size += data->raw->size;
+		else
+			size += sizeof(u32);
+
+		WARN_ON_ONCE(size & (sizeof(u64)-1));
+		header.size += size;
 	}
 
 	ret = perf_output_begin(&handle, counter, header.size, nmi, 1);
@@ -2783,8 +2789,21 @@ static void perf_counter_output(struct perf_counter *counter, int nmi,
 		}
 	}
 
-	if (sample_type & PERF_SAMPLE_TP_RECORD)
-		perf_output_copy(&handle, tp->record, tp->size);
+	if (sample_type & PERF_SAMPLE_RAW) {
+		if (data->raw) {
+			perf_output_put(&handle, data->raw->size);
+			perf_output_copy(&handle, data->raw->data, data->raw->size);
+		} else {
+			struct {
+				u32	size;
+				u32	data;
+			} raw = {
+				.size = sizeof(u32),
+				.data = 0,
+			};
+			perf_output_put(&handle, raw);
+		}
+	}
 
 	perf_output_end(&handle);
 }
@@ -2849,7 +2868,8 @@ perf_counter_read_event(struct perf_counter *counter,
  */
 
 struct perf_task_event {
-	struct task_struct	*task;
+	struct task_struct		*task;
+	struct perf_counter_context	*task_ctx;
 
 	struct {
 		struct perf_event_header	header;
@@ -2909,24 +2929,23 @@ static void perf_counter_task_ctx(struct perf_counter_context *ctx,
 static void perf_counter_task_event(struct perf_task_event *task_event)
 {
 	struct perf_cpu_context *cpuctx;
-	struct perf_counter_context *ctx;
+	struct perf_counter_context *ctx = task_event->task_ctx;
 
 	cpuctx = &get_cpu_var(perf_cpu_context);
 	perf_counter_task_ctx(&cpuctx->ctx, task_event);
 	put_cpu_var(perf_cpu_context);
 
 	rcu_read_lock();
-	/*
-	 * doesn't really matter which of the child contexts the
-	 * events ends up in.
-	 */
-	ctx = rcu_dereference(current->perf_counter_ctxp);
+	if (!ctx)
+		ctx = rcu_dereference(task_event->task->perf_counter_ctxp);
 	if (ctx)
 		perf_counter_task_ctx(ctx, task_event);
 	rcu_read_unlock();
 }
 
-static void perf_counter_task(struct task_struct *task, int new)
+static void perf_counter_task(struct task_struct *task,
+			      struct perf_counter_context *task_ctx,
+			      int new)
 {
 	struct perf_task_event task_event;
 
@@ -2936,8 +2955,9 @@ static void perf_counter_task(struct task_struct *task, int new)
 		return;
 
 	task_event = (struct perf_task_event){
-		.task	= task,
-		.event  = {
+		.task	  = task,
+		.task_ctx = task_ctx,
+		.event    = {
 			.header = {
 				.type = new ? PERF_EVENT_FORK : PERF_EVENT_EXIT,
 				.misc = 0,
@@ -2955,7 +2975,7 @@ static void perf_counter_task(struct task_struct *task, int new)
 
 void perf_counter_fork(struct task_struct *task)
 {
-	perf_counter_task(task, 1);
+	perf_counter_task(task, NULL, 1);
 }
 
 /*
@@ -3344,87 +3364,81 @@ int perf_counter_overflow(struct perf_counter *counter, int nmi,
  * Generic software counter infrastructure
  */
 
-static void perf_swcounter_update(struct perf_counter *counter)
+/*
+ * We directly increment counter->count and keep a second value in
+ * counter->hw.period_left to count intervals. This period counter
+ * is kept in the range [-sample_period, 0] so that we can use the
+ * sign as trigger.
+ */
+
+static u64 perf_swcounter_set_period(struct perf_counter *counter)
 {
 	struct hw_perf_counter *hwc = &counter->hw;
-	u64 prev, now;
-	s64 delta;
+	u64 period = hwc->last_period;
+	u64 nr, offset;
+	s64 old, val;
+
+	hwc->last_period = hwc->sample_period;
 
 again:
-	prev = atomic64_read(&hwc->prev_count);
-	now = atomic64_read(&hwc->count);
-	if (atomic64_cmpxchg(&hwc->prev_count, prev, now) != prev)
-		goto again;
+	old = val = atomic64_read(&hwc->period_left);
+	if (val < 0)
+		return 0;
 
-	delta = now - prev;
+	nr = div64_u64(period + val, period);
+	offset = nr * period;
+	val -= offset;
+	if (atomic64_cmpxchg(&hwc->period_left, old, val) != old)
+		goto again;
 
-	atomic64_add(delta, &counter->count);
-	atomic64_sub(delta, &hwc->period_left);
+	return nr;
 }
 
-static void perf_swcounter_set_period(struct perf_counter *counter)
+static void perf_swcounter_overflow(struct perf_counter *counter,
+				    int nmi, struct perf_sample_data *data)
 {
 	struct hw_perf_counter *hwc = &counter->hw;
-	s64 left = atomic64_read(&hwc->period_left);
-	s64 period = hwc->sample_period;
+	u64 overflow;
 
-	if (unlikely(left <= -period)) {
-		left = period;
-		atomic64_set(&hwc->period_left, left);
-		hwc->last_period = period;
-	}
+	data->period = counter->hw.last_period;
+	overflow = perf_swcounter_set_period(counter);
 
-	if (unlikely(left <= 0)) {
-		left += period;
-		atomic64_add(period, &hwc->period_left);
-		hwc->last_period = period;
-	}
+	if (hwc->interrupts == MAX_INTERRUPTS)
+		return;
 
-	atomic64_set(&hwc->prev_count, -left);
-	atomic64_set(&hwc->count, -left);
+	for (; overflow; overflow--) {
+		if (perf_counter_overflow(counter, nmi, data)) {
+			/*
+			 * We inhibit the overflow from happening when
+			 * hwc->interrupts == MAX_INTERRUPTS.
+			 */
+			break;
+		}
+	}
 }
 
-static enum hrtimer_restart perf_swcounter_hrtimer(struct hrtimer *hrtimer)
+static void perf_swcounter_unthrottle(struct perf_counter *counter)
 {
-	enum hrtimer_restart ret = HRTIMER_RESTART;
-	struct perf_sample_data data;
-	struct perf_counter *counter;
-	u64 period;
-
-	counter	= container_of(hrtimer, struct perf_counter, hw.hrtimer);
-	counter->pmu->read(counter);
-
-	data.addr = 0;
-	data.regs = get_irq_regs();
 	/*
-	 * In case we exclude kernel IPs or are somehow not in interrupt
-	 * context, provide the next best thing, the user IP.
+	 * Nothing to do, we already reset hwc->interrupts.
 	 */
-	if ((counter->attr.exclude_kernel || !data.regs) &&
-			!counter->attr.exclude_user)
-		data.regs = task_pt_regs(current);
+}
 
-	if (data.regs) {
-		if (perf_counter_overflow(counter, 0, &data))
-			ret = HRTIMER_NORESTART;
-	}
+static void perf_swcounter_add(struct perf_counter *counter, u64 nr,
+			       int nmi, struct perf_sample_data *data)
+{
+	struct hw_perf_counter *hwc = &counter->hw;
 
-	period = max_t(u64, 10000, counter->hw.sample_period);
-	hrtimer_forward_now(hrtimer, ns_to_ktime(period));
+	atomic64_add(nr, &counter->count);
 
-	return ret;
-}
+	if (!hwc->sample_period)
+		return;
 
-static void perf_swcounter_overflow(struct perf_counter *counter,
-				    int nmi, struct perf_sample_data *data)
-{
-	data->period = counter->hw.last_period;
+	if (!data->regs)
+		return;
 
-	perf_swcounter_update(counter);
-	perf_swcounter_set_period(counter);
-	if (perf_counter_overflow(counter, nmi, data))
-		/* soft-disable the counter */
-		;
+	if (!atomic64_add_negative(nr, &hwc->period_left))
+		perf_swcounter_overflow(counter, nmi, data);
 }
 
 static int perf_swcounter_is_counting(struct perf_counter *counter)
@@ -3488,15 +3502,6 @@ static int perf_swcounter_match(struct perf_counter *counter,
 	return 1;
 }
 
-static void perf_swcounter_add(struct perf_counter *counter, u64 nr,
-			       int nmi, struct perf_sample_data *data)
-{
-	int neg = atomic64_add_negative(nr, &counter->hw.count);
-
-	if (counter->hw.sample_period && !neg && data->regs)
-		perf_swcounter_overflow(counter, nmi, data);
-}
-
 static void perf_swcounter_ctx_event(struct perf_counter_context *ctx,
 				     enum perf_type_id type,
 				     u32 event, u64 nr, int nmi,
@@ -3575,27 +3580,66 @@ void __perf_swcounter_event(u32 event, u64 nr, int nmi,
 
 static void perf_swcounter_read(struct perf_counter *counter)
 {
-	perf_swcounter_update(counter);
 }
 
 static int perf_swcounter_enable(struct perf_counter *counter)
 {
-	perf_swcounter_set_period(counter);
+	struct hw_perf_counter *hwc = &counter->hw;
+
+	if (hwc->sample_period) {
+		hwc->last_period = hwc->sample_period;
+		perf_swcounter_set_period(counter);
+	}
 	return 0;
 }
 
 static void perf_swcounter_disable(struct perf_counter *counter)
 {
-	perf_swcounter_update(counter);
 }
 
 static const struct pmu perf_ops_generic = {
 	.enable		= perf_swcounter_enable,
 	.disable	= perf_swcounter_disable,
 	.read		= perf_swcounter_read,
+	.unthrottle	= perf_swcounter_unthrottle,
 };
 
 /*
+ * hrtimer based swcounter callback
+ */
+
+static enum hrtimer_restart perf_swcounter_hrtimer(struct hrtimer *hrtimer)
+{
+	enum hrtimer_restart ret = HRTIMER_RESTART;
+	struct perf_sample_data data;
+	struct perf_counter *counter;
+	u64 period;
+
+	counter	= container_of(hrtimer, struct perf_counter, hw.hrtimer);
+	counter->pmu->read(counter);
+
+	data.addr = 0;
+	data.regs = get_irq_regs();
+	/*
+	 * In case we exclude kernel IPs or are somehow not in interrupt
+	 * context, provide the next best thing, the user IP.
+	 */
+	if ((counter->attr.exclude_kernel || !data.regs) &&
+			!counter->attr.exclude_user)
+		data.regs = task_pt_regs(current);
+
+	if (data.regs) {
+		if (perf_counter_overflow(counter, 0, &data))
+			ret = HRTIMER_NORESTART;
+	}
+
+	period = max_t(u64, 10000, counter->hw.sample_period);
+	hrtimer_forward_now(hrtimer, ns_to_ktime(period));
+
+	return ret;
+}
+
+/*
  * Software counter: cpu wall time clock
  */
 
@@ -3715,15 +3759,15 @@ static const struct pmu perf_ops_task_clock = {
 void perf_tpcounter_event(int event_id, u64 addr, u64 count, void *record,
 			  int entry_size)
 {
-	struct perf_tracepoint_record tp = {
+	struct perf_raw_record raw = {
 		.size = entry_size,
-		.record = record,
+		.data = record,
 	};
 
 	struct perf_sample_data data = {
 		.regs = get_irq_regs(),
 		.addr = addr,
-		.private = &tp,
+		.raw = &raw,
 	};
 
 	if (!data.regs)
@@ -3743,6 +3787,14 @@ static void tp_perf_counter_destroy(struct perf_counter *counter)
 
 static const struct pmu *tp_perf_counter_init(struct perf_counter *counter)
 {
+	/*
+	 * Raw tracepoint data is a severe data leak, only allow root to
+	 * have these.
+	 */
+	if ((counter->attr.sample_type & PERF_SAMPLE_RAW) &&
+			!capable(CAP_SYS_ADMIN))
+		return ERR_PTR(-EPERM);
+
 	if (ftrace_profile_enable(counter->attr.config))
 		return NULL;
 
@@ -4285,7 +4337,7 @@ void perf_counter_exit_task(struct task_struct *child)
 	unsigned long flags;
 
 	if (likely(!child->perf_counter_ctxp)) {
-		perf_counter_task(child, 0);
+		perf_counter_task(child, NULL, 0);
 		return;
 	}
 
@@ -4305,6 +4357,7 @@ void perf_counter_exit_task(struct task_struct *child)
 	 * incremented the context's refcount before we do put_ctx below.
 	 */
 	spin_lock(&child_ctx->lock);
+	child->perf_counter_ctxp = NULL;
 	/*
 	 * If this context is a clone; unclone it so it can't get
 	 * swapped to another process while we're removing all
@@ -4318,9 +4371,7 @@ void perf_counter_exit_task(struct task_struct *child)
 	 * won't get any samples after PERF_EVENT_EXIT. We can however still
 	 * get a few PERF_EVENT_READ events.
 	 */
-	perf_counter_task(child, 0);
-
-	child->perf_counter_ctxp = NULL;
+	perf_counter_task(child, child_ctx, 0);
 
 	/*
 	 * We can recurse on the same lock type through: