Merge commit 'tracing/core' into tracing/kprobes

Conflicts:
	kernel/trace/trace_export.c
	kernel/trace/trace_kprobe.c

Merge reason: This topic branch lacks an important
build fix in tracing/core:

	0dd7b74787eaf7858c6c573353a83c3e2766e674:
	tracing: Fix double CPP substitution in TRACE_EVENT_FN

that prevents from multiple tracepoint headers inclusion crashes.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>

19 files changed, 543 insertions, 311 deletions

diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig
index fb5fbf75f279..e78dcbde1a81 100644
--- a/kernel/trace/Kconfig
+++ b/kernel/trace/Kconfig
@@ -60,9 +60,14 @@ config EVENT_TRACING
 	bool
 
 config CONTEXT_SWITCH_TRACER
-	select MARKERS
 	bool
 
+config RING_BUFFER_ALLOW_SWAP
+	bool
+	help
+	 Allow the use of ring_buffer_swap_cpu.
+	 Adds a very slight overhead to tracing when enabled.
+
 # All tracer options should select GENERIC_TRACER. For those options that are
 # enabled by all tracers (context switch and event tracer) they select TRACING.
 # This allows those options to appear when no other tracer is selected. But the
@@ -147,6 +152,7 @@ config IRQSOFF_TRACER
 	select TRACE_IRQFLAGS
 	select GENERIC_TRACER
 	select TRACER_MAX_TRACE
+	select RING_BUFFER_ALLOW_SWAP
 	help
 	  This option measures the time spent in irqs-off critical
 	  sections, with microsecond accuracy.
@@ -168,6 +174,7 @@ config PREEMPT_TRACER
 	depends on PREEMPT
 	select GENERIC_TRACER
 	select TRACER_MAX_TRACE
+	select RING_BUFFER_ALLOW_SWAP
 	help
 	  This option measures the time spent in preemption off critical
 	  sections, with microsecond accuracy.
diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c
index 1090b0aed9ba..3eb159c277c8 100644
--- a/kernel/trace/blktrace.c
+++ b/kernel/trace/blktrace.c
@@ -65,13 +65,15 @@ static void trace_note(struct blk_trace *bt, pid_t pid, int action,
 {
 	struct blk_io_trace *t;
 	struct ring_buffer_event *event = NULL;
+	struct ring_buffer *buffer = NULL;
 	int pc = 0;
 	int cpu = smp_processor_id();
 	bool blk_tracer = blk_tracer_enabled;
 
 	if (blk_tracer) {
+		buffer = blk_tr->buffer;
 		pc = preempt_count();
-		event = trace_buffer_lock_reserve(blk_tr, TRACE_BLK,
+		event = trace_buffer_lock_reserve(buffer, TRACE_BLK,
 						  sizeof(*t) + len,
 						  0, pc);
 		if (!event)
@@ -96,7 +98,7 @@ record_it:
 		memcpy((void *) t + sizeof(*t), data, len);
 
 		if (blk_tracer)
-			trace_buffer_unlock_commit(blk_tr, event, 0, pc);
+			trace_buffer_unlock_commit(buffer, event, 0, pc);
 	}
 }
 
@@ -179,6 +181,7 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
 {
 	struct task_struct *tsk = current;
 	struct ring_buffer_event *event = NULL;
+	struct ring_buffer *buffer = NULL;
 	struct blk_io_trace *t;
 	unsigned long flags = 0;
 	unsigned long *sequence;
@@ -204,8 +207,9 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
 	if (blk_tracer) {
 		tracing_record_cmdline(current);
 
+		buffer = blk_tr->buffer;
 		pc = preempt_count();
-		event = trace_buffer_lock_reserve(blk_tr, TRACE_BLK,
+		event = trace_buffer_lock_reserve(buffer, TRACE_BLK,
 						  sizeof(*t) + pdu_len,
 						  0, pc);
 		if (!event)
@@ -252,7 +256,7 @@ record_it:
 			memcpy((void *) t + sizeof(*t), pdu_data, pdu_len);
 
 		if (blk_tracer) {
-			trace_buffer_unlock_commit(blk_tr, event, 0, pc);
+			trace_buffer_unlock_commit(buffer, event, 0, pc);
 			return;
 		}
 	}
@@ -267,8 +271,8 @@ static void blk_trace_free(struct blk_trace *bt)
 {
 	debugfs_remove(bt->msg_file);
 	debugfs_remove(bt->dropped_file);
-	debugfs_remove(bt->dir);
 	relay_close(bt->rchan);
+	debugfs_remove(bt->dir);
 	free_percpu(bt->sequence);
 	free_percpu(bt->msg_data);
 	kfree(bt);
@@ -378,18 +382,8 @@ static int blk_subbuf_start_callback(struct rchan_buf *buf, void *subbuf,
 
 static int blk_remove_buf_file_callback(struct dentry *dentry)
 {
-	struct dentry *parent = dentry->d_parent;
 	debugfs_remove(dentry);
 
-	/*
-	* this will fail for all but the last file, but that is ok. what we
-	* care about is the top level buts->name directory going away, when
-	* the last trace file is gone. Then we don't have to rmdir() that
-	* manually on trace stop, so it nicely solves the issue with
-	* force killing of running traces.
-	*/
-
-	debugfs_remove(parent);
 	return 0;
 }
 
diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
index 1993b7186cdb..8c804e24f96f 100644
--- a/kernel/trace/ftrace.c
+++ b/kernel/trace/ftrace.c
@@ -2222,7 +2222,11 @@ ftrace_regex_write(struct file *file, const char __user *ubuf,
 	read++;
 	cnt--;
 
-	if (!(iter->flags & ~FTRACE_ITER_CONT)) {
+	/*
+	 * If the parser haven't finished with the last write,
+	 * continue reading the user input without skipping spaces.
+	 */
+	if (!(iter->flags & FTRACE_ITER_CONT)) {
 		/* skip white space */
 		while (cnt && isspace(ch)) {
 			ret = get_user(ch, ubuf++);
@@ -2232,8 +2236,9 @@ ftrace_regex_write(struct file *file, const char __user *ubuf,
 			cnt--;
 		}
 
+		/* only spaces were written */
 		if (isspace(ch)) {
-			file->f_pos += read;
+			*ppos += read;
 			ret = read;
 			goto out;
 		}
@@ -2262,12 +2267,12 @@ ftrace_regex_write(struct file *file, const char __user *ubuf,
 		if (ret)
 			goto out;
 		iter->buffer_idx = 0;
-	} else
+	} else {
 		iter->flags |= FTRACE_ITER_CONT;
+		iter->buffer[iter->buffer_idx++] = ch;
+	}
 
-
-	file->f_pos += read;
-
+	*ppos += read;
 	ret = read;
  out:
 	mutex_unlock(&ftrace_regex_lock);
diff --git a/kernel/trace/kmemtrace.c b/kernel/trace/kmemtrace.c
index dda53ccf749b..81b1645c8549 100644
--- a/kernel/trace/kmemtrace.c
+++ b/kernel/trace/kmemtrace.c
@@ -183,11 +183,9 @@ static void kmemtrace_stop_probes(void)
 
 static int kmem_trace_init(struct trace_array *tr)
 {
-	int cpu;
 	kmemtrace_array = tr;
 
-	for_each_cpu(cpu, cpu_possible_mask)
-		tracing_reset(tr, cpu);
+	tracing_reset_online_cpus(tr);
 
 	kmemtrace_start_probes();
 
diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
index da2c59d8f486..454e74e718cf 100644
--- a/kernel/trace/ring_buffer.c
+++ b/kernel/trace/ring_buffer.c
@@ -218,17 +218,12 @@ enum {
 
 static inline int rb_null_event(struct ring_buffer_event *event)
 {
-	return event->type_len == RINGBUF_TYPE_PADDING
-			&& event->time_delta == 0;
-}
-
-static inline int rb_discarded_event(struct ring_buffer_event *event)
-{
-	return event->type_len == RINGBUF_TYPE_PADDING && event->time_delta;
+	return event->type_len == RINGBUF_TYPE_PADDING && !event->time_delta;
 }
 
 static void rb_event_set_padding(struct ring_buffer_event *event)
 {
+	/* padding has a NULL time_delta */
 	event->type_len = RINGBUF_TYPE_PADDING;
 	event->time_delta = 0;
 }
@@ -472,14 +467,19 @@ struct ring_buffer_iter {
 };
 
 /* buffer may be either ring_buffer or ring_buffer_per_cpu */
-#define RB_WARN_ON(buffer, cond)				\
-	({							\
-		int _____ret = unlikely(cond);			\
-		if (_____ret) {					\
-			atomic_inc(&buffer->record_disabled);	\
-			WARN_ON(1);				\
-		}						\
-		_____ret;					\
+#define RB_WARN_ON(b, cond)						\
+	({								\
+		int _____ret = unlikely(cond);				\
+		if (_____ret) {						\
+			if (__same_type(*(b), struct ring_buffer_per_cpu)) { \
+				struct ring_buffer_per_cpu *__b =	\
+					(void *)b;			\
+				atomic_inc(&__b->buffer->record_disabled); \
+			} else						\
+				atomic_inc(&b->record_disabled);	\
+			WARN_ON(1);					\
+		}							\
+		_____ret;						\
 	})
 
 /* Up this if you want to test the TIME_EXTENTS and normalization */
@@ -1778,9 +1778,6 @@ rb_reset_tail(struct ring_buffer_per_cpu *cpu_buffer,
 	event->type_len = RINGBUF_TYPE_PADDING;
 	/* time delta must be non zero */
 	event->time_delta = 1;
-	/* Account for this as an entry */
-	local_inc(&tail_page->entries);
-	local_inc(&cpu_buffer->entries);
 
 	/* Set write to end of buffer */
 	length = (tail + length) - BUF_PAGE_SIZE;
@@ -2076,7 +2073,8 @@ static void rb_end_commit(struct ring_buffer_per_cpu *cpu_buffer)
 }
 
 static struct ring_buffer_event *
-rb_reserve_next_event(struct ring_buffer_per_cpu *cpu_buffer,
+rb_reserve_next_event(struct ring_buffer *buffer,
+		      struct ring_buffer_per_cpu *cpu_buffer,
 		      unsigned long length)
 {
 	struct ring_buffer_event *event;
@@ -2086,6 +2084,21 @@ rb_reserve_next_event(struct ring_buffer_per_cpu *cpu_buffer,
 
 	rb_start_commit(cpu_buffer);
 
+#ifdef CONFIG_RING_BUFFER_ALLOW_SWAP
+	/*
+	 * Due to the ability to swap a cpu buffer from a buffer
+	 * it is possible it was swapped before we committed.
+	 * (committing stops a swap). We check for it here and
+	 * if it happened, we have to fail the write.
+	 */
+	barrier();
+	if (unlikely(ACCESS_ONCE(cpu_buffer->buffer) != buffer)) {
+		local_dec(&cpu_buffer->committing);
+		local_dec(&cpu_buffer->commits);
+		return NULL;
+	}
+#endif
+
 	length = rb_calculate_event_length(length);
  again:
 	/*
@@ -2246,7 +2259,7 @@ ring_buffer_lock_reserve(struct ring_buffer *buffer, unsigned long length)
 	if (length > BUF_MAX_DATA_SIZE)
 		goto out;
 
-	event = rb_reserve_next_event(cpu_buffer, length);
+	event = rb_reserve_next_event(buffer, cpu_buffer, length);
 	if (!event)
 		goto out;
 
@@ -2269,18 +2282,23 @@ ring_buffer_lock_reserve(struct ring_buffer *buffer, unsigned long length)
 }
 EXPORT_SYMBOL_GPL(ring_buffer_lock_reserve);
 
-static void rb_commit(struct ring_buffer_per_cpu *cpu_buffer,
+static void
+rb_update_write_stamp(struct ring_buffer_per_cpu *cpu_buffer,
 		      struct ring_buffer_event *event)
 {
-	local_inc(&cpu_buffer->entries);
-
 	/*
 	 * The event first in the commit queue updates the
 	 * time stamp.
 	 */
 	if (rb_event_is_commit(cpu_buffer, event))
 		cpu_buffer->write_stamp += event->time_delta;
+}
 
+static void rb_commit(struct ring_buffer_per_cpu *cpu_buffer,
+		      struct ring_buffer_event *event)
+{
+	local_inc(&cpu_buffer->entries);
+	rb_update_write_stamp(cpu_buffer, event);
 	rb_end_commit(cpu_buffer);
 }
 
@@ -2327,32 +2345,57 @@ static inline void rb_event_discard(struct ring_buffer_event *event)
 		event->time_delta = 1;
 }
 
-/**
- * ring_buffer_event_discard - discard any event in the ring buffer
- * @event: the event to discard
- *
- * Sometimes a event that is in the ring buffer needs to be ignored.
- * This function lets the user discard an event in the ring buffer
- * and then that event will not be read later.
- *
- * Note, it is up to the user to be careful with this, and protect
- * against races. If the user discards an event that has been consumed
- * it is possible that it could corrupt the ring buffer.
+/*
+ * Decrement the entries to the page that an event is on.
+ * The event does not even need to exist, only the pointer
+ * to the page it is on. This may only be called before the commit
+ * takes place.
  */
-void ring_buffer_event_discard(struct ring_buffer_event *event)
+static inline void
+rb_decrement_entry(struct ring_buffer_per_cpu *cpu_buffer,
+		   struct ring_buffer_event *event)
 {
-	rb_event_discard(event);
+	unsigned long addr = (unsigned long)event;
+	struct buffer_page *bpage = cpu_buffer->commit_page;
+	struct buffer_page *start;
+
+	addr &= PAGE_MASK;
+
+	/* Do the likely case first */
+	if (likely(bpage->page == (void *)addr)) {
+		local_dec(&bpage->entries);
+		return;
+	}
+
+	/*
+	 * Because the commit page may be on the reader page we
+	 * start with the next page and check the end loop there.
+	 */
+	rb_inc_page(cpu_buffer, &bpage);
+	start = bpage;
+	do {
+		if (bpage->page == (void *)addr) {
+			local_dec(&bpage->entries);
+			return;
+		}
+		rb_inc_page(cpu_buffer, &bpage);
+	} while (bpage != start);
+
+	/* commit not part of this buffer?? */
+	RB_WARN_ON(cpu_buffer, 1);
 }
-EXPORT_SYMBOL_GPL(ring_buffer_event_discard);
 
 /**
  * ring_buffer_commit_discard - discard an event that has not been committed
  * @buffer: the ring buffer
  * @event: non committed event to discard
  *
- * This is similar to ring_buffer_event_discard but must only be
- * performed on an event that has not been committed yet. The difference
- * is that this will also try to free the event from the ring buffer
+ * Sometimes an event that is in the ring buffer needs to be ignored.
+ * This function lets the user discard an event in the ring buffer
+ * and then that event will not be read later.
+ *
+ * This function only works if it is called before the the item has been
+ * committed. It will try to free the event from the ring buffer
  * if another event has not been added behind it.
  *
  * If another event has been added behind it, it will set the event
@@ -2380,14 +2423,15 @@ void ring_buffer_discard_commit(struct ring_buffer *buffer,
 	 */
 	RB_WARN_ON(buffer, !local_read(&cpu_buffer->committing));
 
+	rb_decrement_entry(cpu_buffer, event);
 	if (rb_try_to_discard(cpu_buffer, event))
 		goto out;
 
 	/*
 	 * The commit is still visible by the reader, so we
-	 * must increment entries.
+	 * must still update the timestamp.
 	 */
-	local_inc(&cpu_buffer->entries);
+	rb_update_write_stamp(cpu_buffer, event);
  out:
 	rb_end_commit(cpu_buffer);
 
@@ -2448,7 +2492,7 @@ int ring_buffer_write(struct ring_buffer *buffer,
 	if (length > BUF_MAX_DATA_SIZE)
 		goto out;
 
-	event = rb_reserve_next_event(cpu_buffer, length);
+	event = rb_reserve_next_event(buffer, cpu_buffer, length);
 	if (!event)
 		goto out;
 
@@ -2899,8 +2943,7 @@ static void rb_advance_reader(struct ring_buffer_per_cpu *cpu_buffer)
 
 	event = rb_reader_event(cpu_buffer);
 
-	if (event->type_len <= RINGBUF_TYPE_DATA_TYPE_LEN_MAX
-			|| rb_discarded_event(event))
+	if (event->type_len <= RINGBUF_TYPE_DATA_TYPE_LEN_MAX)
 		cpu_buffer->read++;
 
 	rb_update_read_stamp(cpu_buffer, event);
@@ -3132,10 +3175,8 @@ ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts)
 		spin_unlock(&cpu_buffer->reader_lock);
 	local_irq_restore(flags);
 
-	if (event && event->type_len == RINGBUF_TYPE_PADDING) {
-		cpu_relax();
+	if (event && event->type_len == RINGBUF_TYPE_PADDING)
 		goto again;
-	}
 
 	return event;
 }
@@ -3160,10 +3201,8 @@ ring_buffer_iter_peek(struct ring_buffer_iter *iter, u64 *ts)
 	event = rb_iter_peek(iter, ts);
 	spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
 
-	if (event && event->type_len == RINGBUF_TYPE_PADDING) {
-		cpu_relax();
+	if (event && event->type_len == RINGBUF_TYPE_PADDING)
 		goto again;
-	}
 
 	return event;
 }
@@ -3209,10 +3248,8 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts)
  out:
 	preempt_enable();
 
-	if (event && event->type_len == RINGBUF_TYPE_PADDING) {
-		cpu_relax();
+	if (event && event->type_len == RINGBUF_TYPE_PADDING)
 		goto again;
-	}
 
 	return event;
 }
@@ -3292,21 +3329,19 @@ ring_buffer_read(struct ring_buffer_iter *iter, u64 *ts)
 	struct ring_buffer_per_cpu *cpu_buffer = iter->cpu_buffer;
 	unsigned long flags;
 
- again:
 	spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
+ again:
 	event = rb_iter_peek(iter, ts);
 	if (!event)
 		goto out;
 
+	if (event->type_len == RINGBUF_TYPE_PADDING)
+		goto again;
+
 	rb_advance_iter(iter);
  out:
 	spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
 
-	if (event && event->type_len == RINGBUF_TYPE_PADDING) {
-		cpu_relax();
-		goto again;
-	}
-
 	return event;
 }
 EXPORT_SYMBOL_GPL(ring_buffer_read);
@@ -3373,12 +3408,16 @@ void ring_buffer_reset_cpu(struct ring_buffer *buffer, int cpu)
 
 	spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
 
+	if (RB_WARN_ON(cpu_buffer, local_read(&cpu_buffer->committing)))
+		goto out;
+
 	__raw_spin_lock(&cpu_buffer->lock);
 
 	rb_reset_cpu(cpu_buffer);
 
 	__raw_spin_unlock(&cpu_buffer->lock);
 
+ out:
 	spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
 
 	atomic_dec(&cpu_buffer->record_disabled);
@@ -3461,6 +3500,7 @@ int ring_buffer_empty_cpu(struct ring_buffer *buffer, int cpu)
 }
 EXPORT_SYMBOL_GPL(ring_buffer_empty_cpu);
 
+#ifdef CONFIG_RING_BUFFER_ALLOW_SWAP
 /**
  * ring_buffer_swap_cpu - swap a CPU buffer between two ring buffers
  * @buffer_a: One buffer to swap with
@@ -3515,20 +3555,28 @@ int ring_buffer_swap_cpu(struct ring_buffer *buffer_a,
 	atomic_inc(&cpu_buffer_a->record_disabled);
 	atomic_inc(&cpu_buffer_b->record_disabled);
 
+	ret = -EBUSY;
+	if (local_read(&cpu_buffer_a->committing))
+		goto out_dec;
+	if (local_read(&cpu_buffer_b->committing))
+		goto out_dec;
+
 	buffer_a->buffers[cpu] = cpu_buffer_b;
 	buffer_b->buffers[cpu] = cpu_buffer_a;
 
 	cpu_buffer_b->buffer = buffer_a;
 	cpu_buffer_a->buffer = buffer_b;
 
+	ret = 0;
+
+out_dec:
 	atomic_dec(&cpu_buffer_a->record_disabled);
 	atomic_dec(&cpu_buffer_b->record_disabled);
-
-	ret = 0;
 out:
 	return ret;
 }
 EXPORT_SYMBOL_GPL(ring_buffer_swap_cpu);
+#endif /* CONFIG_RING_BUFFER_ALLOW_SWAP */
 
 /**
  * ring_buffer_alloc_read_page - allocate a page to read from buffer
diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index 63dbc7ff213f..5c75deeefe30 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -43,9 +43,6 @@
 
 #define TRACE_BUFFER_FLAGS	(RB_FL_OVERWRITE)
 
-unsigned long __read_mostly	tracing_max_latency;
-unsigned long __read_mostly	tracing_thresh;
-
 /*
  * On boot up, the ring buffer is set to the minimum size, so that
  * we do not waste memory on systems that are not using tracing.
@@ -172,10 +169,11 @@ static struct trace_array	global_trace;
 
 static DEFINE_PER_CPU(struct trace_array_cpu, global_trace_cpu);
 
-int filter_current_check_discard(struct ftrace_event_call *call, void *rec,
+int filter_current_check_discard(struct ring_buffer *buffer,
+				 struct ftrace_event_call *call, void *rec,
 				 struct ring_buffer_event *event)
 {
-	return filter_check_discard(call, rec, global_trace.buffer, event);
+	return filter_check_discard(call, rec, buffer, event);
 }
 EXPORT_SYMBOL_GPL(filter_current_check_discard);
 
@@ -266,6 +264,9 @@ unsigned long trace_flags = TRACE_ITER_PRINT_PARENT | TRACE_ITER_PRINTK |
 	TRACE_ITER_ANNOTATE | TRACE_ITER_CONTEXT_INFO | TRACE_ITER_SLEEP_TIME |
 	TRACE_ITER_GRAPH_TIME;
 
+static int trace_stop_count;
+static DEFINE_SPINLOCK(tracing_start_lock);
+
 /**
  * trace_wake_up - wake up tasks waiting for trace input
  *
@@ -338,45 +339,6 @@ static struct {
 
 int trace_clock_id;
 
-/*
- * ftrace_max_lock is used to protect the swapping of buffers
- * when taking a max snapshot. The buffers themselves are
- * protected by per_cpu spinlocks. But the action of the swap
- * needs its own lock.
- *
- * This is defined as a raw_spinlock_t in order to help
- * with performance when lockdep debugging is enabled.
- */
-static raw_spinlock_t ftrace_max_lock =
-	(raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED;
-
-/*
- * Copy the new maximum trace into the separate maximum-trace
- * structure. (this way the maximum trace is permanently saved,
- * for later retrieval via /sys/kernel/debug/tracing/latency_trace)
- */
-static void
-__update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu)
-{
-	struct trace_array_cpu *data = tr->data[cpu];
-
-	max_tr.cpu = cpu;
-	max_tr.time_start = data->preempt_timestamp;
-
-	data = max_tr.data[cpu];
-	data->saved_latency = tracing_max_latency;
-
-	memcpy(data->comm, tsk->comm, TASK_COMM_LEN);
-	data->pid = tsk->pid;
-	data->uid = task_uid(tsk);
-	data->nice = tsk->static_prio - 20 - MAX_RT_PRIO;
-	data->policy = tsk->policy;
-	data->rt_priority = tsk->rt_priority;
-
-	/* record this tasks comm */
-	tracing_record_cmdline(tsk);
-}
-
 ssize_t trace_seq_to_user(struct trace_seq *s, char __user *ubuf, size_t cnt)
 {
 	int len;
@@ -420,6 +382,56 @@ static ssize_t trace_seq_to_buffer(struct trace_seq *s, void *buf, size_t cnt)
 	return cnt;
 }
 
+/*
+ * ftrace_max_lock is used to protect the swapping of buffers
+ * when taking a max snapshot. The buffers themselves are
+ * protected by per_cpu spinlocks. But the action of the swap
+ * needs its own lock.
+ *
+ * This is defined as a raw_spinlock_t in order to help
+ * with performance when lockdep debugging is enabled.
+ *
+ * It is also used in other places outside the update_max_tr
+ * so it needs to be defined outside of the
+ * CONFIG_TRACER_MAX_TRACE.
+ */
+static raw_spinlock_t ftrace_max_lock =
+	(raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED;
+
+#ifdef CONFIG_TRACER_MAX_TRACE
+unsigned long __read_mostly	tracing_max_latency;
+unsigned long __read_mostly	tracing_thresh;
+
+/*
+ * Copy the new maximum trace into the separate maximum-trace
+ * structure. (this way the maximum trace is permanently saved,
+ * for later retrieval via /sys/kernel/debug/tracing/latency_trace)
+ */
+static void
+__update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu)
+{
+	struct trace_array_cpu *data = tr->data[cpu];
+	struct trace_array_cpu *max_data = tr->data[cpu];
+
+	max_tr.cpu = cpu;
+	max_tr.time_start = data->preempt_timestamp;
+
+	max_data = max_tr.data[cpu];
+	max_data->saved_latency = tracing_max_latency;
+	max_data->critical_start = data->critical_start;
+	max_data->critical_end = data->critical_end;
+
+	memcpy(data->comm, tsk->comm, TASK_COMM_LEN);
+	max_data->pid = tsk->pid;
+	max_data->uid = task_uid(tsk);
+	max_data->nice = tsk->static_prio - 20 - MAX_RT_PRIO;
+	max_data->policy = tsk->policy;
+	max_data->rt_priority = tsk->rt_priority;
+
+	/* record this tasks comm */
+	tracing_record_cmdline(tsk);
+}
+
 /**
  * update_max_tr - snapshot all trace buffers from global_trace to max_tr
  * @tr: tracer
@@ -434,16 +446,15 @@ update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu)
 {
 	struct ring_buffer *buf = tr->buffer;
 
+	if (trace_stop_count)
+		return;
+
 	WARN_ON_ONCE(!irqs_disabled());
 	__raw_spin_lock(&ftrace_max_lock);
 
 	tr->buffer = max_tr.buffer;
 	max_tr.buffer = buf;
 
-	ftrace_disable_cpu();
-	ring_buffer_reset(tr->buffer);
-	ftrace_enable_cpu();
-
 	__update_max_tr(tr, tsk, cpu);
 	__raw_spin_unlock(&ftrace_max_lock);
 }
@@ -461,21 +472,35 @@ update_max_tr_single(struct trace_array *tr, struct task_struct *tsk, int cpu)
 {
 	int ret;
 
+	if (trace_stop_count)
+		return;
+
 	WARN_ON_ONCE(!irqs_disabled());
 	__raw_spin_lock(&ftrace_max_lock);
 
 	ftrace_disable_cpu();
 
-	ring_buffer_reset(max_tr.buffer);
 	ret = ring_buffer_swap_cpu(max_tr.buffer, tr->buffer, cpu);
 
+	if (ret == -EBUSY) {
+		/*
+		 * We failed to swap the buffer due to a commit taking
+		 * place on this CPU. We fail to record, but we reset
+		 * the max trace buffer (no one writes directly to it)
+		 * and flag that it failed.
+		 */
+		trace_array_printk(&max_tr, _THIS_IP_,
+			"Failed to swap buffers due to commit in progress\n");
+	}
+
 	ftrace_enable_cpu();
 
-	WARN_ON_ONCE(ret && ret != -EAGAIN);
+	WARN_ON_ONCE(ret && ret != -EAGAIN && ret != -EBUSY);
 
 	__update_max_tr(tr, tsk, cpu);
 	__raw_spin_unlock(&ftrace_max_lock);
 }
+#endif /* CONFIG_TRACER_MAX_TRACE */
 
 /**
  * register_tracer - register a tracer with the ftrace system.
@@ -532,7 +557,6 @@ __acquires(kernel_lock)
 	if (type->selftest && !tracing_selftest_disabled) {
 		struct tracer *saved_tracer = current_trace;
 		struct trace_array *tr = &global_trace;
-		int i;
 
 		/*
 		 * Run a selftest on this tracer.
@@ -541,8 +565,7 @@ __acquires(kernel_lock)
 		 * internal tracing to verify that everything is in order.
 		 * If we fail, we do not register this tracer.
 		 */
-		for_each_tracing_cpu(i)
-			tracing_reset(tr, i);
+		tracing_reset_online_cpus(tr);
 
 		current_trace = type;
 		/* the test is responsible for initializing and enabling */
@@ -555,8 +578,7 @@ __acquires(kernel_lock)
 			goto out;
 		}
 		/* Only reset on passing, to avoid touching corrupted buffers */
-		for_each_tracing_cpu(i)
-			tracing_reset(tr, i);
+		tracing_reset_online_cpus(tr);
 
 		printk(KERN_CONT "PASSED\n");
 	}
@@ -631,21 +653,42 @@ void unregister_tracer(struct tracer *type)
 	mutex_unlock(&trace_types_lock);
 }
 
-void tracing_reset(struct trace_array *tr, int cpu)
+static void __tracing_reset(struct trace_array *tr, int cpu)
 {
 	ftrace_disable_cpu();
 	ring_buffer_reset_cpu(tr->buffer, cpu);
 	ftrace_enable_cpu();
 }
 
+void tracing_reset(struct trace_array *tr, int cpu)
+{
+	struct ring_buffer *buffer = tr->buffer;
+
+	ring_buffer_record_disable(buffer);
+
+	/* Make sure all commits have finished */
+	synchronize_sched();
+	__tracing_reset(tr, cpu);
+
+	ring_buffer_record_enable(buffer);
+}
+
 void tracing_reset_online_cpus(struct trace_array *tr)
 {
+	struct ring_buffer *buffer = tr->buffer;
 	int cpu;
 
+	ring_buffer_record_disable(buffer);
+
+	/* Make sure all commits have finished */
+	synchronize_sched();
+
 	tr->time_start = ftrace_now(tr->cpu);
 
 	for_each_online_cpu(cpu)
-		tracing_reset(tr, cpu);
+		__tracing_reset(tr, cpu);
+
+	ring_buffer_record_enable(buffer);
 }
 
 void tracing_reset_current(int cpu)
@@ -676,9 +719,6 @@ static void trace_init_cmdlines(void)
 	cmdline_idx = 0;
 }
 
-static int trace_stop_count;
-static DEFINE_SPINLOCK(tracing_start_lock);
-
 /**
  * ftrace_off_permanent - disable all ftrace code permanently
  *
@@ -859,14 +899,15 @@ tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags,
 }
 EXPORT_SYMBOL_GPL(tracing_generic_entry_update);
 
-struct ring_buffer_event *trace_buffer_lock_reserve(struct trace_array *tr,
-						    int type,
-						    unsigned long len,
-						    unsigned long flags, int pc)
+struct ring_buffer_event *
+trace_buffer_lock_reserve(struct ring_buffer *buffer,
+			  int type,
+			  unsigned long len,
+			  unsigned long flags, int pc)
 {
 	struct ring_buffer_event *event;
 
-	event = ring_buffer_lock_reserve(tr->buffer, len);
+	event = ring_buffer_lock_reserve(buffer, len);
 	if (event != NULL) {
 		struct trace_entry *ent = ring_buffer_event_data(event);
 
@@ -877,53 +918,59 @@ struct ring_buffer_event *trace_buffer_lock_reserve(struct trace_array *tr,
 	return event;
 }
 
-static inline void __trace_buffer_unlock_commit(struct trace_array *tr,
-					struct ring_buffer_event *event,
-					unsigned long flags, int pc,
-					int wake)
+static inline void
+__trace_buffer_unlock_commit(struct ring_buffer *buffer,
+			     struct ring_buffer_event *event,
+			     unsigned long flags, int pc,
+			     int wake)
 {
-	ring_buffer_unlock_commit(tr->buffer, event);
+	ring_buffer_unlock_commit(buffer, event);
 
-	ftrace_trace_stack(tr, flags, 6, pc);
-	ftrace_trace_userstack(tr, flags, pc);
+	ftrace_trace_stack(buffer, flags, 6, pc);
+	ftrace_trace_userstack(buffer, flags, pc);
 
 	if (wake)
 		trace_wake_up();
 }
 
-void trace_buffer_unlock_commit(struct trace_array *tr,
-					struct ring_buffer_event *event,
-					unsigned long flags, int pc)
+void trace_buffer_unlock_commit(struct ring_buffer *buffer,
+				struct ring_buffer_event *event,
+				unsigned long flags, int pc)
 {
-	__trace_buffer_unlock_commit(tr, event, flags, pc, 1);
+	__trace_buffer_unlock_commit(buffer, event, flags, pc, 1);
 }
 
 struct ring_buffer_event *
-trace_current_buffer_lock_reserve(int type, unsigned long len,
+trace_current_buffer_lock_reserve(struct ring_buffer **current_rb,
+				  int type, unsigned long len,
 				  unsigned long flags, int pc)
 {
-	return trace_buffer_lock_reserve(&global_trace,
+	*current_rb = global_trace.buffer;
+	return trace_buffer_lock_reserve(*current_rb,
 					 type, len, flags, pc);
 }
 EXPORT_SYMBOL_GPL(trace_current_buffer_lock_reserve);
 
-void trace_current_buffer_unlock_commit(struct ring_buffer_event *event,
+void trace_current_buffer_unlock_commit(struct ring_buffer *buffer,
+					struct ring_buffer_event *event,
 					unsigned long flags, int pc)
 {
-	__trace_buffer_unlock_commit(&global_trace, event, flags, pc, 1);
+	__trace_buffer_unlock_com