Merge branch 'v2.6.37-rc2' into for-2.6.38/core

15 files changed, 825 insertions, 423 deletions

diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig
index 538501c6ea50..e04b8bcdef88 100644
--- a/kernel/trace/Kconfig
+++ b/kernel/trace/Kconfig
@@ -49,6 +49,11 @@ config HAVE_SYSCALL_TRACEPOINTS
 	help
 	  See Documentation/trace/ftrace-design.txt
 
+config HAVE_C_RECORDMCOUNT
+	bool
+	help
+	  C version of recordmcount available?
+
 config TRACER_MAX_TRACE
 	bool
 
@@ -121,7 +126,7 @@ if FTRACE
 config FUNCTION_TRACER
 	bool "Kernel Function Tracer"
 	depends on HAVE_FUNCTION_TRACER
-	select FRAME_POINTER
+	select FRAME_POINTER if (!ARM_UNWIND)
 	select KALLSYMS
 	select GENERIC_TRACER
 	select CONTEXT_SWITCH_TRACER
diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c
index 959f8d6c8cc1..7b8ec0281548 100644
--- a/kernel/trace/blktrace.c
+++ b/kernel/trace/blktrace.c
@@ -23,7 +23,6 @@
 #include <linux/mutex.h>
 #include <linux/slab.h>
 #include <linux/debugfs.h>
-#include <linux/smp_lock.h>
 #include <linux/time.h>
 #include <linux/uaccess.h>
 
@@ -169,7 +168,6 @@ static int act_log_check(struct blk_trace *bt, u32 what, sector_t sector,
 static const u32 ddir_act[2] = { BLK_TC_ACT(BLK_TC_READ),
 				 BLK_TC_ACT(BLK_TC_WRITE) };
 
-#define BLK_TC_HARDBARRIER	BLK_TC_BARRIER
 #define BLK_TC_RAHEAD		BLK_TC_AHEAD
 
 /* The ilog2() calls fall out because they're constant */
@@ -197,7 +195,6 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
 		return;
 
 	what |= ddir_act[rw & WRITE];
-	what |= MASK_TC_BIT(rw, HARDBARRIER);
 	what |= MASK_TC_BIT(rw, SYNC);
 	what |= MASK_TC_BIT(rw, RAHEAD);
 	what |= MASK_TC_BIT(rw, META);
@@ -326,6 +323,7 @@ static const struct file_operations blk_dropped_fops = {
 	.owner =	THIS_MODULE,
 	.open =		blk_dropped_open,
 	.read =		blk_dropped_read,
+	.llseek =	default_llseek,
 };
 
 static int blk_msg_open(struct inode *inode, struct file *filp)
@@ -365,6 +363,7 @@ static const struct file_operations blk_msg_fops = {
 	.owner =	THIS_MODULE,
 	.open =		blk_msg_open,
 	.write =	blk_msg_write,
+	.llseek =	noop_llseek,
 };
 
 /*
@@ -639,7 +638,6 @@ int blk_trace_ioctl(struct block_device *bdev, unsigned cmd, char __user *arg)
 	if (!q)
 		return -ENXIO;
 
-	lock_kernel();
 	mutex_lock(&bdev->bd_mutex);
 
 	switch (cmd) {
@@ -667,7 +665,6 @@ int blk_trace_ioctl(struct block_device *bdev, unsigned cmd, char __user *arg)
 	}
 
 	mutex_unlock(&bdev->bd_mutex);
-	unlock_kernel();
 	return ret;
 }
 
@@ -1652,10 +1649,9 @@ static ssize_t sysfs_blk_trace_attr_show(struct device *dev,
 	struct block_device *bdev;
 	ssize_t ret = -ENXIO;
 
-	lock_kernel();
 	bdev = bdget(part_devt(p));
 	if (bdev == NULL)
-		goto out_unlock_kernel;
+		goto out;
 
 	q = blk_trace_get_queue(bdev);
 	if (q == NULL)
@@ -1683,8 +1679,7 @@ out_unlock_bdev:
 	mutex_unlock(&bdev->bd_mutex);
 out_bdput:
 	bdput(bdev);
-out_unlock_kernel:
-	unlock_kernel();
+out:
 	return ret;
 }
 
@@ -1714,11 +1709,10 @@ static ssize_t sysfs_blk_trace_attr_store(struct device *dev,
 
 	ret = -ENXIO;
 
-	lock_kernel();
 	p = dev_to_part(dev);
 	bdev = bdget(part_devt(p));
 	if (bdev == NULL)
-		goto out_unlock_kernel;
+		goto out;
 
 	q = blk_trace_get_queue(bdev);
 	if (q == NULL)
@@ -1753,8 +1747,6 @@ out_unlock_bdev:
 	mutex_unlock(&bdev->bd_mutex);
 out_bdput:
 	bdput(bdev);
-out_unlock_kernel:
-	unlock_kernel();
 out:
 	return ret ? ret : count;
 }
@@ -1813,8 +1805,6 @@ void blk_fill_rwbs(char *rwbs, u32 rw, int bytes)
 
 	if (rw & REQ_RAHEAD)
 		rwbs[i++] = 'A';
-	if (rw & REQ_HARDBARRIER)
-		rwbs[i++] = 'B';
 	if (rw & REQ_SYNC)
 		rwbs[i++] = 'S';
 	if (rw & REQ_META)
diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
index fa7ece649fe1..f3dadae83883 100644
--- a/kernel/trace/ftrace.c
+++ b/kernel/trace/ftrace.c
@@ -800,6 +800,7 @@ static const struct file_operations ftrace_profile_fops = {
 	.open		= tracing_open_generic,
 	.read		= ftrace_profile_read,
 	.write		= ftrace_profile_write,
+	.llseek		= default_llseek,
 };
 
 /* used to initialize the real stat files */
@@ -884,10 +885,8 @@ enum {
 	FTRACE_ENABLE_CALLS		= (1 << 0),
 	FTRACE_DISABLE_CALLS		= (1 << 1),
 	FTRACE_UPDATE_TRACE_FUNC	= (1 << 2),
-	FTRACE_ENABLE_MCOUNT		= (1 << 3),
-	FTRACE_DISABLE_MCOUNT		= (1 << 4),
-	FTRACE_START_FUNC_RET		= (1 << 5),
-	FTRACE_STOP_FUNC_RET		= (1 << 6),
+	FTRACE_START_FUNC_RET		= (1 << 3),
+	FTRACE_STOP_FUNC_RET		= (1 << 4),
 };
 
 static int ftrace_filtered;
@@ -1226,8 +1225,6 @@ static void ftrace_shutdown(int command)
 
 static void ftrace_startup_sysctl(void)
 {
-	int command = FTRACE_ENABLE_MCOUNT;
-
 	if (unlikely(ftrace_disabled))
 		return;
 
@@ -1235,23 +1232,17 @@ static void ftrace_startup_sysctl(void)
 	saved_ftrace_func = NULL;
 	/* ftrace_start_up is true if we want ftrace running */
 	if (ftrace_start_up)
-		command |= FTRACE_ENABLE_CALLS;
-
-	ftrace_run_update_code(command);
+		ftrace_run_update_code(FTRACE_ENABLE_CALLS);
 }
 
 static void ftrace_shutdown_sysctl(void)
 {
-	int command = FTRACE_DISABLE_MCOUNT;
-
 	if (unlikely(ftrace_disabled))
 		return;
 
 	/* ftrace_start_up is true if ftrace is running */
 	if (ftrace_start_up)
-		command |= FTRACE_DISABLE_CALLS;
-
-	ftrace_run_update_code(command);
+		ftrace_run_update_code(FTRACE_DISABLE_CALLS);
 }
 
 static cycle_t		ftrace_update_time;
@@ -1368,24 +1359,29 @@ enum {
 #define FTRACE_BUFF_MAX (KSYM_SYMBOL_LEN+4) /* room for wildcards */
 
 struct ftrace_iterator {
-	struct ftrace_page	*pg;
-	int			hidx;
-	int			idx;
-	unsigned		flags;
-	struct trace_parser	parser;
+	loff_t				pos;
+	loff_t				func_pos;
+	struct ftrace_page		*pg;
+	struct dyn_ftrace		*func;
+	struct ftrace_func_probe	*probe;
+	struct trace_parser		parser;
+	int				hidx;
+	int				idx;
+	unsigned			flags;
 };
 
 static void *
-t_hash_next(struct seq_file *m, void *v, loff_t *pos)
+t_hash_next(struct seq_file *m, loff_t *pos)
 {
 	struct ftrace_iterator *iter = m->private;
-	struct hlist_node *hnd = v;
+	struct hlist_node *hnd = NULL;
 	struct hlist_head *hhd;
 
-	WARN_ON(!(iter->flags & FTRACE_ITER_HASH));
-
 	(*pos)++;
+	iter->pos = *pos;
 
+	if (iter->probe)
+		hnd = &iter->probe->node;
  retry:
 	if (iter->hidx >= FTRACE_FUNC_HASHSIZE)
 		return NULL;
@@ -1408,7 +1404,12 @@ t_hash_next(struct seq_file *m, void *v, loff_t *pos)
 		}
 	}
 
-	return hnd;
+	if (WARN_ON_ONCE(!hnd))
+		return NULL;
+
+	iter->probe = hlist_entry(hnd, struct ftrace_func_probe, node);
+
+	return iter;
 }
 
 static void *t_hash_start(struct seq_file *m, loff_t *pos)
@@ -1417,26 +1418,32 @@ static void *t_hash_start(struct seq_file *m, loff_t *pos)
 	void *p = NULL;
 	loff_t l;
 
-	if (!(iter->flags & FTRACE_ITER_HASH))
-		*pos = 0;
-
-	iter->flags |= FTRACE_ITER_HASH;
+	if (iter->func_pos > *pos)
+		return NULL;
 
 	iter->hidx = 0;
-	for (l = 0; l <= *pos; ) {
-		p = t_hash_next(m, p, &l);
+	for (l = 0; l <= (*pos - iter->func_pos); ) {
+		p = t_hash_next(m, &l);
 		if (!p)
 			break;
 	}
-	return p;
+	if (!p)
+		return NULL;
+
+	/* Only set this if we have an item */
+	iter->flags |= FTRACE_ITER_HASH;
+
+	return iter;
 }
 
-static int t_hash_show(struct seq_file *m, void *v)
+static int
+t_hash_show(struct seq_file *m, struct ftrace_iterator *iter)
 {
 	struct ftrace_func_probe *rec;
-	struct hlist_node *hnd = v;
 
-	rec = hlist_entry(hnd, struct ftrace_func_probe, node);
+	rec = iter->probe;
+	if (WARN_ON_ONCE(!rec))
+		return -EIO;
 
 	if (rec->ops->print)
 		return rec->ops->print(m, rec->ip, rec->ops, rec->data);
@@ -1457,12 +1464,13 @@ t_next(struct seq_file *m, void *v, loff_t *pos)
 	struct dyn_ftrace *rec = NULL;
 
 	if (iter->flags & FTRACE_ITER_HASH)
-		return t_hash_next(m, v, pos);
+		return t_hash_next(m, pos);
 
 	(*pos)++;
+	iter->pos = *pos;
 
 	if (iter->flags & FTRACE_ITER_PRINTALL)
-		return NULL;
+		return t_hash_start(m, pos);
 
  retry:
 	if (iter->idx >= iter->pg->index) {
@@ -1491,7 +1499,20 @@ t_next(struct seq_file *m, void *v, loff_t *pos)
 		}
 	}
 
-	return rec;
+	if (!rec)
+		return t_hash_start(m, pos);
+
+	iter->func_pos = *pos;
+	iter->func = rec;
+
+	return iter;
+}
+
+static void reset_iter_read(struct ftrace_iterator *iter)
+{
+	iter->pos = 0;
+	iter->func_pos = 0;
+	iter->flags &= ~(FTRACE_ITER_PRINTALL & FTRACE_ITER_HASH);
 }
 
 static void *t_start(struct seq_file *m, loff_t *pos)
@@ -1502,6 +1523,12 @@ static void *t_start(struct seq_file *m, loff_t *pos)
 
 	mutex_lock(&ftrace_lock);
 	/*
+	 * If an lseek was done, then reset and start from beginning.
+	 */
+	if (*pos < iter->pos)
+		reset_iter_read(iter);
+
+	/*
 	 * For set_ftrace_filter reading, if we have the filter
 	 * off, we can short cut and just print out that all
 	 * functions are enabled.
@@ -1518,6 +1545,11 @@ static void *t_start(struct seq_file *m, loff_t *pos)
 	if (iter->flags & FTRACE_ITER_HASH)
 		return t_hash_start(m, pos);
 
+	/*
+	 * Unfortunately, we need to restart at ftrace_pages_start
+	 * every time we let go of the ftrace_mutex. This is because
+	 * those pointers can change without the lock.
+	 */
 	iter->pg = ftrace_pages_start;
 	iter->idx = 0;
 	for (l = 0; l <= *pos; ) {
@@ -1526,10 +1558,14 @@ static void *t_start(struct seq_file *m, loff_t *pos)
 			break;
 	}
 
-	if (!p && iter->flags & FTRACE_ITER_FILTER)
-		return t_hash_start(m, pos);
+	if (!p) {
+		if (iter->flags & FTRACE_ITER_FILTER)
+			return t_hash_start(m, pos);
+
+		return NULL;
+	}
 
-	return p;
+	return iter;
 }
 
 static void t_stop(struct seq_file *m, void *p)
@@ -1540,16 +1576,18 @@ static void t_stop(struct seq_file *m, void *p)
 static int t_show(struct seq_file *m, void *v)
 {
 	struct ftrace_iterator *iter = m->private;
-	struct dyn_ftrace *rec = v;
+	struct dyn_ftrace *rec;
 
 	if (iter->flags & FTRACE_ITER_HASH)
-		return t_hash_show(m, v);
+		return t_hash_show(m, iter);
 
 	if (iter->flags & FTRACE_ITER_PRINTALL) {
 		seq_printf(m, "#### all functions enabled ####\n");
 		return 0;
 	}
 
+	rec = iter->func;
+
 	if (!rec)
 		return 0;
 
@@ -1601,8 +1639,8 @@ ftrace_failures_open(struct inode *inode, struct file *file)
 
 	ret = ftrace_avail_open(inode, file);
 	if (!ret) {
-		m = (struct seq_file *)file->private_data;
-		iter = (struct ftrace_iterator *)m->private;
+		m = file->private_data;
+		iter = m->private;
 		iter->flags = FTRACE_ITER_FAILURES;
 	}
 
@@ -2418,7 +2456,7 @@ static const struct file_operations ftrace_filter_fops = {
 	.open = ftrace_filter_open,
 	.read = seq_read,
 	.write = ftrace_filter_write,
-	.llseek = no_llseek,
+	.llseek = ftrace_regex_lseek,
 	.release = ftrace_filter_release,
 };
 
@@ -2632,6 +2670,7 @@ static const struct file_operations ftrace_graph_fops = {
 	.read		= seq_read,
 	.write		= ftrace_graph_write,
 	.release	= ftrace_graph_release,
+	.llseek		= seq_lseek,
 };
 #endif /* CONFIG_FUNCTION_GRAPH_TRACER */
 
diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
index bca96377fd4e..9ed509a015d8 100644
--- a/kernel/trace/ring_buffer.c
+++ b/kernel/trace/ring_buffer.c
@@ -224,6 +224,9 @@ enum {
 	RB_LEN_TIME_STAMP = 16,
 };
 
+#define skip_time_extend(event) \
+	((struct ring_buffer_event *)((char *)event + RB_LEN_TIME_EXTEND))
+
 static inline int rb_null_event(struct ring_buffer_event *event)
 {
 	return event->type_len == RINGBUF_TYPE_PADDING && !event->time_delta;
@@ -248,8 +251,12 @@ rb_event_data_length(struct ring_buffer_event *event)
 	return length + RB_EVNT_HDR_SIZE;
 }
 
-/* inline for ring buffer fast paths */
-static unsigned
+/*
+ * Return the length of the given event. Will return
+ * the length of the time extend if the event is a
+ * time extend.
+ */
+static inline unsigned
 rb_event_length(struct ring_buffer_event *event)
 {
 	switch (event->type_len) {
@@ -274,13 +281,41 @@ rb_event_length(struct ring_buffer_event *event)
 	return 0;
 }
 
+/*
+ * Return total length of time extend and data,
+ *   or just the event length for all other events.
+ */
+static inline unsigned
+rb_event_ts_length(struct ring_buffer_event *event)
+{
+	unsigned len = 0;
+
+	if (event->type_len == RINGBUF_TYPE_TIME_EXTEND) {
+		/* time extends include the data event after it */
+		len = RB_LEN_TIME_EXTEND;
+		event = skip_time_extend(event);
+	}
+	return len + rb_event_length(event);
+}
+
 /**
  * ring_buffer_event_length - return the length of the event
  * @event: the event to get the length of
+ *
+ * Returns the size of the data load of a data event.
+ * If the event is something other than a data event, it
+ * returns the size of the event itself. With the exception
+ * of a TIME EXTEND, where it still returns the size of the
+ * data load of the data event after it.
  */
 unsigned ring_buffer_event_length(struct ring_buffer_event *event)
 {
-	unsigned length = rb_event_length(event);
+	unsigned length;
+
+	if (event->type_len == RINGBUF_TYPE_TIME_EXTEND)
+		event = skip_time_extend(event);
+
+	length = rb_event_length(event);
 	if (event->type_len > RINGBUF_TYPE_DATA_TYPE_LEN_MAX)
 		return length;
 	length -= RB_EVNT_HDR_SIZE;
@@ -294,6 +329,8 @@ EXPORT_SYMBOL_GPL(ring_buffer_event_length);
 static void *
 rb_event_data(struct ring_buffer_event *event)
 {
+	if (event->type_len == RINGBUF_TYPE_TIME_EXTEND)
+		event = skip_time_extend(event);
 	BUG_ON(event->type_len > RINGBUF_TYPE_DATA_TYPE_LEN_MAX);
 	/* If length is in len field, then array[0] has the data */
 	if (event->type_len)
@@ -404,9 +441,6 @@ static inline int test_time_stamp(u64 delta)
 /* Max payload is BUF_PAGE_SIZE - header (8bytes) */
 #define BUF_MAX_DATA_SIZE (BUF_PAGE_SIZE - (sizeof(u32) * 2))
 
-/* Max number of timestamps that can fit on a page */
-#define RB_TIMESTAMPS_PER_PAGE	(BUF_PAGE_SIZE / RB_LEN_TIME_EXTEND)
-
 int ring_buffer_print_page_header(struct trace_seq *s)
 {
 	struct buffer_data_page field;
@@ -1546,6 +1580,25 @@ static void rb_inc_iter(struct ring_buffer_iter *iter)
 	iter->head = 0;
 }
 
+/* Slow path, do not inline */
+static noinline struct ring_buffer_event *
+rb_add_time_stamp(struct ring_buffer_event *event, u64 delta)
+{
+	event->type_len = RINGBUF_TYPE_TIME_EXTEND;
+
+	/* Not the first event on the page? */
+	if (rb_event_index(event)) {
+		event->time_delta = delta & TS_MASK;
+		event->array[0] = delta >> TS_SHIFT;
+	} else {
+		/* nope, just zero it */
+		event->time_delta = 0;
+		event->array[0] = 0;
+	}
+
+	return skip_time_extend(event);
+}
+
 /**
  * ring_buffer_update_event - update event type and data
  * @event: the even to update
@@ -1558,28 +1611,31 @@ static void rb_inc_iter(struct ring_buffer_iter *iter)
  * data field.
  */
 static void
-rb_update_event(struct ring_buffer_event *event,
-			 unsigned type, unsigned length)
+rb_update_event(struct ring_buffer_per_cpu *cpu_buffer,
+		struct ring_buffer_event *event, unsigned length,
+		int add_timestamp, u64 delta)
 {
-	event->type_len = type;
-
-	switch (type) {
-
-	case RINGBUF_TYPE_PADDING:
-	case RINGBUF_TYPE_TIME_EXTEND:
-	case RINGBUF_TYPE_TIME_STAMP:
-		break;
+	/* Only a commit updates the timestamp */
+	if (unlikely(!rb_event_is_commit(cpu_buffer, event)))
+		delta = 0;
 
-	case 0:
-		length -= RB_EVNT_HDR_SIZE;
-		if (length > RB_MAX_SMALL_DATA || RB_FORCE_8BYTE_ALIGNMENT)
-			event->array[0] = length;
-		else
-			event->type_len = DIV_ROUND_UP(length, RB_ALIGNMENT);
-		break;
-	default:
-		BUG();
+	/*
+	 * If we need to add a timestamp, then we
+	 * add it to the start of the resevered space.
+	 */
+	if (unlikely(add_timestamp)) {
+		event = rb_add_time_stamp(event, delta);
+		length -= RB_LEN_TIME_EXTEND;
+		delta = 0;
 	}
+
+	event->time_delta = delta;
+	length -= RB_EVNT_HDR_SIZE;
+	if (length > RB_MAX_SMALL_DATA || RB_FORCE_8BYTE_ALIGNMENT) {
+		event->type_len = 0;
+		event->array[0] = length;
+	} else
+		event->type_len = DIV_ROUND_UP(length, RB_ALIGNMENT);
 }
 
 /*
@@ -1823,10 +1879,13 @@ rb_reset_tail(struct ring_buffer_per_cpu *cpu_buffer,
 	local_sub(length, &tail_page->write);
 }
 
-static struct ring_buffer_event *
+/*
+ * This is the slow path, force gcc not to inline it.
+ */
+static noinline struct ring_buffer_event *
 rb_move_tail(struct ring_buffer_per_cpu *cpu_buffer,
 	     unsigned long length, unsigned long tail,
-	     struct buffer_page *tail_page, u64 *ts)
+	     struct buffer_page *tail_page, u64 ts)
 {
 	struct buffer_page *commit_page = cpu_buffer->commit_page;
 	struct ring_buffer *buffer = cpu_buffer->buffer;
@@ -1909,8 +1968,8 @@ rb_move_tail(struct ring_buffer_per_cpu *cpu_buffer,
 		 * Nested commits always have zero deltas, so
 		 * just reread the time stamp
 		 */
-		*ts = rb_time_stamp(buffer);
-		next_page->page->time_stamp = *ts;
+		ts = rb_time_stamp(buffer);
+		next_page->page->time_stamp = ts;
 	}
 
  out_again:
@@ -1929,12 +1988,21 @@ rb_move_tail(struct ring_buffer_per_cpu *cpu_buffer,
 
 static struct ring_buffer_event *
 __rb_reserve_next(struct ring_buffer_per_cpu *cpu_buffer,
-		  unsigned type, unsigned long length, u64 *ts)
+		  unsigned long length, u64 ts,
+		  u64 delta, int add_timestamp)
 {
 	struct buffer_page *tail_page;
 	struct ring_buffer_event *event;
 	unsigned long tail, write;
 
+	/*
+	 * If the time delta since the last event is too big to
+	 * hold in the time field of the event, then we append a
+	 * TIME EXTEND event ahead of the data event.
+	 */
+	if (unlikely(add_timestamp))
+		length += RB_LEN_TIME_EXTEND;
+
 	tail_page = cpu_buffer->tail_page;
 	write = local_add_return(length, &tail_page->write);
 
@@ -1943,7 +2011,7 @@ __rb_reserve_next(struct ring_buffer_per_cpu *cpu_buffer,
 	tail = write - length;
 
 	/* See if we shot pass the end of this buffer page */
-	if (write > BUF_PAGE_SIZE)
+	if (unlikely(write > BUF_PAGE_SIZE))
 		return rb_move_tail(cpu_buffer, length, tail,
 				    tail_page, ts);
 
@@ -1951,18 +2019,16 @@ __rb_reserve_next(struct ring_buffer_per_cpu *cpu_buffer,
 
 	event = __rb_page_index(tail_page, tail);
 	kmemcheck_annotate_bitfield(event, bitfield);
-	rb_update_event(event, type, length);
+	rb_update_event(cpu_buffer, event, length, add_timestamp, delta);
 
-	/* The passed in type is zero for DATA */
-	if (likely(!type))
-		local_inc(&tail_page->entries);
+	local_inc(&tail_page->entries);
 
 	/*
 	 * If this is the first commit on the page, then update
 	 * its timestamp.
 	 */
 	if (!tail)
-		tail_page->page->time_stamp = *ts;
+		tail_page->page->time_stamp = ts;
 
 	return event;
 }
@@ -1977,7 +2043,7 @@ rb_try_to_discard(struct ring_buffer_per_cpu *cpu_buffer,
 	unsigned long addr;
 
 	new_index = rb_event_index(event);
-	old_index = new_index + rb_event_length(event);
+	old_index = new_index + rb_event_ts_length(event);
 	addr = (unsigned long)event;
 	addr &= PAGE_MASK;
 
@@ -2003,76 +2069,13 @@ rb_try_to_discard(struct ring_buffer_per_cpu *cpu_buffer,
 	return 0;
 }
 
-static int
-rb_add_time_stamp(struct ring_buffer_per_cpu *cpu_buffer,
-		  u64 *ts, u64 *delta)
-{
-	struct ring_buffer_event *event;
-	int ret;
-
-	WARN_ONCE(*delta > (1ULL << 59),
-		  KERN_WARNING "Delta way too big! %llu ts=%llu write stamp = %llu\n",
-		  (unsigned long long)*delta,
-		  (unsigned long long)*ts,
-		  (unsigned long long)cpu_buffer->write_stamp);
-
-	/*
-	 * The delta is too big, we to add a
-	 * new timestamp.
-	 */
-	event = __rb_reserve_next(cpu_buffer,
-				  RINGBUF_TYPE_TIME_EXTEND,
-				  RB_LEN_TIME_EXTEND,
-				  ts);
-	if (!event)
-		return -EBUSY;
-
-	if (PTR_ERR(event) == -EAGAIN)
-		return -EAGAIN;
-
-	/* Only a commited time event can update the write stamp */
-	if (rb_event_is_commit(cpu_buffer, event)) {
-		/*
-		 * If this is the first on the page, then it was
-		 * updated with the page itself. Try to discard it
-		 * and if we can't just make it zero.
-		 */
-		if (rb_event_index(event)) {
-			event->time_delta = *delta & TS_MASK;
-			event->array[0] = *delta >> TS_SHIFT;
-		} else {
-			/* try to discard, since we do not need this */
-			if (!rb_try_to_discard(cpu_buffer, event)) {
-				/* nope, just zero it */
-				event->time_delta = 0;
-				event->array[0] = 0;
-			}
-		}
-		cpu_buffer->write_stamp = *ts;
-		/* let the caller know this was the commit */
-		ret = 1;
-	} else {
-		/* Try to discard the event */
-		if (!rb_try_to_discard(cpu_buffer, event)) {
-			/* Darn, this is just wasted space */
-			event->time_delta = 0;
-			event->array[0] = 0;
-		}
-		ret = 0;
-	}
-
-	*delta = 0;
-
-	return ret;
-}
-
 static void rb_start_commit(struct ring_buffer_per_cpu *cpu_buffer)
 {
 	local_inc(&cpu_buffer->committing);
 	local_inc(&cpu_buffer->commits);
 }
 
-static void rb_end_commit(struct ring_buffer_per_cpu *cpu_buffer)
+static inline void rb_end_commit(struct ring_buffer_per_cpu *cpu_buffer)
 {
 	unsigned long commits;
 
@@ -2110,9 +2113,10 @@ rb_reserve_next_event(struct ring_buffer *buffer,
 		      unsigned long length)
 {
 	struct ring_buffer_event *event;
-	u64 ts, delta = 0;
-	int commit = 0;
+	u64 ts, delta;
 	int nr_loops = 0;
+	int add_timestamp;
+	u64 diff;
 
 	rb_start_commit(cpu_buffer);
 
@@ -2133,6 +2137,9 @@ rb_reserve_next_event(struct ring_buffer *buffer,
 
 	length = rb_calculate_event_length(length);
  again:
+	add_timestamp = 0;
+	delta = 0;
+
 	/*
 	 * We allow for interrupts to reenter here and do a trace.
 	 * If one does, it will cause this original code to loop
@@ -2146,56 +2153,32 @@ rb_reserve_next_event(struct ring_buffer *buffer,
 		goto out_fail;
 
 	ts = rb_time_stamp(cpu_buffer->buffer);
+	diff = ts - cpu_buffer->write_stamp;
 
-	/*
-	 * Only the first commit can update the timestamp.
-	 * Yes there is a race here. If an interrupt comes in
-	 * just after the conditional and it traces too, then it
-	 * will also check the deltas. More than one timestamp may
-	 * also be made. But only the entry that did the actual
-	 * commit will be something other than zero.
-	 */
-	if (likely(cpu_buffer->tail_page == cpu_buffer->commit_page &&
-		   rb_page_write(cpu_buffer->tail_page) ==
-		   rb_commit_index(cpu_buffer))) {
-		u64 diff;
-
-		diff = ts - cpu_buffer->write_stamp;
-
-		/* make sure this diff is calculated here */
-		barrier();
-
-		/* Did the write stamp get updated already? */
-		if (unlikely(ts < cpu_buffer->write_stamp))
-			goto get_event;
+	/* make sure this diff is calculated here */
+	barrier();
 
+	/* Did the write stamp get updated already? */
+	if (likely(ts >= cpu_buffer->write_stamp)) {
 		delta = diff;
 		if (unlikely(test_time_stamp(delta))) {
-
-			commit = rb_add_time_stamp(cpu_buffer, &ts, &delta);
-			if (commit == -EBUSY)
-				goto out_fail;
-
-			if (commit == -EAGAIN)
-				goto again;
-
-			RB_WARN_ON(cpu_buffer, commit < 0);
+			WARN_ONCE(delta > (1ULL << 59),
+				  KERN_WARNING "Delta way too big! %llu ts=%llu write stamp = %llu\n",
+				  (unsigned long long)delta,
+				  (unsigned long long)ts,
+				  (unsigned long long)cpu_buffer->write_stamp);
+			add_timestamp = 1;
 		}
 	}
 
- get_event:
-	event = __rb_reserve_next(cpu_buffer, 0, length, &ts);
+	event = __rb_reserve_next(cpu_buffer, length, ts,
+				  delta, add_timestamp);
 	if (unlikely(PTR_ERR(event) == -EAGAIN))
 		goto again;
 
 	if (!event)
 		goto out_fail;
 
-	if (!rb_event_is_commit(cpu_buffer, event))
-		delta = 0;
-
-	event->time_delta = delta;
-
 	return event;
 
  out_fail:
@@ -2207,13 +2190,9 @@ rb_reserve_next_event(struct ring_buffer *buffer,
 
 #define TRACE_RECURSIVE_DEPTH 16
 
-static int trace_recursive_lock(void)
+/* Keep this code out of the fast path cache */
+static noinline void trace_recursive_fail(void)
 {
-	current->trace_recursion++;
-
-	if (likely(current->trace_recursion < TRACE_RECURSIVE_DEPTH))
-		return 0;
-
 	/* Disable all tracing before we do anything else */
 	tracing_off_permanent();
 
@@ -2225,10 +2204,21 @@ static int trace_recursive_lock(void)
 		    in_nmi());
 
 	WARN_ON_ONCE(1);
+}
+
+static inline int trace_recursive_lock(void)
+{
+	current->trace_recursion++;
+
+	if (likely(current->trace_recursion < TRACE_RECURSIVE_DEPTH))
+		return 0;
+
+	trace_recursive_fail();
+
 	return -1;
 }
 
-static void trace_recursive_unlock(void)
+static inline void trace_recursive_unlock(void)
 {
 	WARN_ON_ONCE(!current->trace_recursion);
 
@@ -2308,12 +2298,28 @@ static void
 rb_update_write_stamp(struct ring_buffer_per_cpu *cpu_buffer,
 		      struct ring_buffer_event *event)
 {
+	u64 delta;
+
 	/*
 	 * 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;
+	if (rb_event_is_commit(cpu_buffer, event)) {
+		/*
+		 * A commit event that is first on a page
+		 * updates the write timestamp with the page stamp
+		 */
+		if (!rb_event_index(event))
+			cpu_buffer->write_stamp =
+				cpu_buffer->commit_page->page->time_stamp;
+		else if (event->type_len == RINGBUF_TYPE_TIME_EXTEND) {
+			delta = event->array[0];
+			delta <<= TS_SHIFT;
+			delta += event->time_delta;
+			cpu_buffer->write_stamp += delta;
+		} else
+			cpu_buffer->write_stamp += event->time_delta;
+	}
 }
 
 static void rb_commit(struct ring_buffer_per_cpu *cpu_buffer,
@@ -2353,6 +2359,9 @@ EXPORT_SYMBOL_GPL(ring_buffer_unlock_commit);
 
 static inline void rb_event_discard(struct ring_buffer_event *event)
 {
+	if (event->type_len == RINGBUF_TYPE_TIME_EXTEND)
+		event = skip_time_extend(event);
+
 	/* array[0] holds the actual length for the discarded event */
 	event->array[0] = rb_event_data_length(event) - RB_EVNT_HDR_SIZE;
 	event->type_len = RINGBUF_TYPE_PADDING;
@@ -2606,6 +2615,19 @@ void ring_buffer_record_enable_cpu(struct ring_buffer *buffer, int cpu)
 }
 EXPORT_SYMBOL_GPL(ring_buffer_record_enable_cpu);
 
+/*
+ * The total entries in the ring buffer is the running counter
+ * of entries entered into the ring buffer, minus the sum of
+ * the entries read from the ring buffer and the number of
+ * entries that were overwritten.
+ */
+static inline unsigned long
+rb_num_of_entries(struct ring_buffer_per_cpu *cpu_buffer)
+{
+	return local_read(&cpu_buffer->entries) -
+		(local_read(&cpu_buffer->overrun) + cpu_buffer->read);
+}
+
 /**
  * ring_buffer_entries_cpu - get the number of entries in a cpu buffer
  * @buffer: The ring buffer
@@ -2614,16 +2636,13 @@ EXPORT_SYMBOL_GPL(ring_buffer_record_enable_cpu);
 unsigned long ring_buffer_entries_cpu(struct ring_buffer *buffer, int cpu)
 {
 	struct ring_buffer_per_cpu *cpu_buffer;
-	unsigned long ret;
 
 	if (!cpumask_test_cpu(cpu, buffer->cpumask))
 		return 0;
 
 	cpu_buffer = buffer->buffers[cpu];
-	ret = (local_read(&cpu_buffer->entries) - local_read(&cpu_buffer->overrun))
-		- cpu_buffer->read;
 
-	return ret;
+	return rb_num_of_entries(cpu_buffer);
 }
 EXPORT_SYMBOL_GPL(ring_buffer_entries_cpu);
 
@@ -2684,8 +2703,7 @@ unsigned long ring_buffer_entries(struct ring_buffer *buffer)
 	/* if you care about this being correct, lock the buffer */
 	for_each_buffer_cpu(buffer, cpu) {
 		cpu_buffer = buffer->buffers[cpu];
-		entries += (local_read(&cpu_buffer->entries) -
-			    local_read(&cpu_buffer->overrun)) - cpu_buffer->read;
+		entries += rb_num_of_entries(cpu_buffer);
 	}
 
 	return entries;
@@ -3040,12 +3058,12 @@ rb_buffer_peek(struct ring_buffer_per_cpu *cpu_buffer, u64 *ts,
 
  again:
 	/*
-	 * We repeat when a timestamp is encountered. It is possible
-	 * to get multiple timestamps from an interrupt entering just
-	 * as one timestamp is about to be written, or from discarded
-	 * commits. The most that we can have is the number on a single page.
+	 * We repeat when a time extend is encountered.
+	 * Since the time extend is always attached to a data event,
+	 * we should never loop more than once.
+	 * (We never hit the following condition more than twice).
 	 */
-	if (RB_WARN_ON(cpu_buffer, ++nr_loops > RB_TIMESTAMPS_PER_PAGE))
+	if (RB_WARN_ON(cpu_buffer, ++nr_loops > 2))
 		return NULL;
 
 	reader = rb_get_reader_page(cpu_buffer);
@@ -3121,14 +3139,12 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts)
 		return NULL;
 
 	/*
-	 * We repeat when a timest