Merge tag 'trace-v4.2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace

Pull tracing updates from Steven Rostedt:
 "This patch series contains several clean ups and even a new trace
  clock "monitonic raw".  Also some enhancements to make the ring buffer
  even faster.  But the biggest and most noticeable change is the
  renaming of the ftrace* files, structures and variables that have to
  deal with trace events.

  Over the years I've had several developers tell me about their
  confusion with what ftrace is compared to events.  Technically,
  "ftrace" is the infrastructure to do the function hooks, which include
  tracing and also helps with live kernel patching.  But the trace
  events are a separate entity altogether, and the files that affect the
  trace events should not be named "ftrace".  These include:

    include/trace/ftrace.h         ->    include/trace/trace_events.h
    include/linux/ftrace_event.h   ->    include/linux/trace_events.h

  Also, functions that are specific for trace events have also been renamed:

    ftrace_print_*()               ->    trace_print_*()
    (un)register_ftrace_event()    ->    (un)register_trace_event()
    ftrace_event_name()            ->    trace_event_name()
    ftrace_trigger_soft_disabled() ->    trace_trigger_soft_disabled()
    ftrace_define_fields_##call()  ->    trace_define_fields_##call()
    ftrace_get_offsets_##call()    ->    trace_get_offsets_##call()

  Structures have been renamed:

    ftrace_event_file              ->    trace_event_file
    ftrace_event_{call,class}      ->    trace_event_{call,class}
    ftrace_event_buffer            ->    trace_event_buffer
    ftrace_subsystem_dir           ->    trace_subsystem_dir
    ftrace_event_raw_##call        ->    trace_event_raw_##call
    ftrace_event_data_offset_##call->    trace_event_data_offset_##call
    ftrace_event_type_funcs_##call ->    trace_event_type_funcs_##call

  And a few various variables and flags have also been updated.

  This has been sitting in linux-next for some time, and I have not
  heard a single complaint about this rename breaking anything.  Mostly
  because these functions, variables and structures are mostly internal
  to the tracing system and are seldom (if ever) used by anything
  external to that"

* tag 'trace-v4.2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace: (33 commits)
  ring_buffer: Allow to exit the ring buffer benchmark immediately
  ring-buffer-benchmark: Fix the wrong type
  ring-buffer-benchmark: Fix the wrong param in module_param
  ring-buffer: Add enum names for the context levels
  ring-buffer: Remove useless unused tracing_off_permanent()
  ring-buffer: Give NMIs a chance to lock the reader_lock
  ring-buffer: Add trace_recursive checks to ring_buffer_write()
  ring-buffer: Allways do the trace_recursive checks
  ring-buffer: Move recursive check to per_cpu descriptor
  ring-buffer: Add unlikelys to make fast path the default
  tracing: Rename ftrace_get_offsets_##call() to trace_event_get_offsets_##call()
  tracing: Rename ftrace_define_fields_##call() to trace_event_define_fields_##call()
  tracing: Rename ftrace_event_type_funcs_##call to trace_event_type_funcs_##call
  tracing: Rename ftrace_data_offset_##call to trace_event_data_offset_##call
  tracing: Rename ftrace_raw_##call event structures to trace_event_raw_##call
  tracing: Rename ftrace_trigger_soft_disabled() to trace_trigger_soft_disabled()
  tracing: Rename FTRACE_EVENT_FL_* flags to EVENT_FILE_FL_*
  tracing: Rename struct ftrace_subsystem_dir to trace_subsystem_dir
  tracing: Rename ftrace_event_name() to trace_event_name()
  tracing: Rename FTRACE_MAX_EVENT to TRACE_EVENT_TYPE_MAX
  ...

1 files changed, 89 insertions, 132 deletions

diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
index 0315d43176d8..6260717c18e3 100644
--- a/kernel/trace/ring_buffer.c
+++ b/kernel/trace/ring_buffer.c
@@ -3,7 +3,7 @@
  *
  * Copyright (C) 2008 Steven Rostedt <srostedt@redhat.com>
  */
-#include <linux/ftrace_event.h>
+#include <linux/trace_events.h>
 #include <linux/ring_buffer.h>
 #include <linux/trace_clock.h>
 #include <linux/trace_seq.h>
@@ -115,63 +115,11 @@ int ring_buffer_print_entry_header(struct trace_seq *s)
  *
  */
 
-/*
- * A fast way to enable or disable all ring buffers is to
- * call tracing_on or tracing_off. Turning off the ring buffers
- * prevents all ring buffers from being recorded to.
- * Turning this switch on, makes it OK to write to the
- * ring buffer, if the ring buffer is enabled itself.
- *
- * There's three layers that must be on in order to write
- * to the ring buffer.
- *
- * 1) This global flag must be set.
- * 2) The ring buffer must be enabled for recording.
- * 3) The per cpu buffer must be enabled for recording.
- *
- * In case of an anomaly, this global flag has a bit set that
- * will permantly disable all ring buffers.
- */
-
-/*
- * Global flag to disable all recording to ring buffers
- *  This has two bits: ON, DISABLED
- *
- *  ON   DISABLED
- * ---- ----------
- *   0      0        : ring buffers are off
- *   1      0        : ring buffers are on
- *   X      1        : ring buffers are permanently disabled
- */
-
-enum {
-	RB_BUFFERS_ON_BIT	= 0,
-	RB_BUFFERS_DISABLED_BIT	= 1,
-};
-
-enum {
-	RB_BUFFERS_ON		= 1 << RB_BUFFERS_ON_BIT,
-	RB_BUFFERS_DISABLED	= 1 << RB_BUFFERS_DISABLED_BIT,
-};
-
-static unsigned long ring_buffer_flags __read_mostly = RB_BUFFERS_ON;
-
 /* Used for individual buffers (after the counter) */
 #define RB_BUFFER_OFF		(1 << 20)
 
 #define BUF_PAGE_HDR_SIZE offsetof(struct buffer_data_page, data)
 
-/**
- * tracing_off_permanent - permanently disable ring buffers
- *
- * This function, once called, will disable all ring buffers
- * permanently.
- */
-void tracing_off_permanent(void)
-{
-	set_bit(RB_BUFFERS_DISABLED_BIT, &ring_buffer_flags);
-}
-
 #define RB_EVNT_HDR_SIZE (offsetof(struct ring_buffer_event, array))
 #define RB_ALIGNMENT		4U
 #define RB_MAX_SMALL_DATA	(RB_ALIGNMENT * RINGBUF_TYPE_DATA_TYPE_LEN_MAX)
@@ -452,6 +400,23 @@ struct rb_irq_work {
 };
 
 /*
+ * Used for which event context the event is in.
+ *  NMI     = 0
+ *  IRQ     = 1
+ *  SOFTIRQ = 2
+ *  NORMAL  = 3
+ *
+ * See trace_recursive_lock() comment below for more details.
+ */
+enum {
+	RB_CTX_NMI,
+	RB_CTX_IRQ,
+	RB_CTX_SOFTIRQ,
+	RB_CTX_NORMAL,
+	RB_CTX_MAX
+};
+
+/*
  * head_page == tail_page && head == tail then buffer is empty.
  */
 struct ring_buffer_per_cpu {
@@ -462,6 +427,7 @@ struct ring_buffer_per_cpu {
 	arch_spinlock_t			lock;
 	struct lock_class_key		lock_key;
 	unsigned int			nr_pages;
+	unsigned int			current_context;
 	struct list_head		*pages;
 	struct buffer_page		*head_page;	/* read from head */
 	struct buffer_page		*tail_page;	/* write to tail */
@@ -2224,7 +2190,7 @@ static unsigned rb_calculate_event_length(unsigned length)
 
 	/* zero length can cause confusions */
 	if (!length)
-		length = 1;
+		length++;
 
 	if (length > RB_MAX_SMALL_DATA || RB_FORCE_8BYTE_ALIGNMENT)
 		length += sizeof(event.array[0]);
@@ -2636,8 +2602,6 @@ rb_reserve_next_event(struct ring_buffer *buffer,
 	return NULL;
 }
 
-#ifdef CONFIG_TRACING
-
 /*
  * The lock and unlock are done within a preempt disable section.
  * The current_context per_cpu variable can only be modified
@@ -2675,44 +2639,38 @@ rb_reserve_next_event(struct ring_buffer *buffer,
  * just so happens that it is the same bit corresponding to
  * the current context.
  */
-static DEFINE_PER_CPU(unsigned int, current_context);
 
-static __always_inline int trace_recursive_lock(void)
+static __always_inline int
+trace_recursive_lock(struct ring_buffer_per_cpu *cpu_buffer)
 {
-	unsigned int val = __this_cpu_read(current_context);
+	unsigned int val = cpu_buffer->current_context;
 	int bit;
 
 	if (in_interrupt()) {
 		if (in_nmi())
-			bit = 0;
+			bit = RB_CTX_NMI;
 		else if (in_irq())
-			bit = 1;
+			bit = RB_CTX_IRQ;
 		else
-			bit = 2;
+			bit = RB_CTX_SOFTIRQ;
 	} else
-		bit = 3;
+		bit = RB_CTX_NORMAL;
 
 	if (unlikely(val & (1 << bit)))
 		return 1;
 
 	val |= (1 << bit);
-	__this_cpu_write(current_context, val);
+	cpu_buffer->current_context = val;
 
 	return 0;
 }
 
-static __always_inline void trace_recursive_unlock(void)
+static __always_inline void
+trace_recursive_unlock(struct ring_buffer_per_cpu *cpu_buffer)
 {
-	__this_cpu_and(current_context, __this_cpu_read(current_context) - 1);
+	cpu_buffer->current_context &= cpu_buffer->current_context - 1;
 }
 
-#else
-
-#define trace_recursive_lock()		(0)
-#define trace_recursive_unlock()	do { } while (0)
-
-#endif
-
 /**
  * ring_buffer_lock_reserve - reserve a part of the buffer
  * @buffer: the ring buffer to reserve from
@@ -2735,41 +2693,37 @@ ring_buffer_lock_reserve(struct ring_buffer *buffer, unsigned long length)
 	struct ring_buffer_event *event;
 	int cpu;
 
-	if (ring_buffer_flags != RB_BUFFERS_ON)
-		return NULL;
-
 	/* If we are tracing schedule, we don't want to recurse */
 	preempt_disable_notrace();
 
-	if (atomic_read(&buffer->record_disabled))
-		goto out_nocheck;
-
-	if (trace_recursive_lock())
-		goto out_nocheck;
+	if (unlikely(atomic_read(&buffer->record_disabled)))
+		goto out;
 
 	cpu = raw_smp_processor_id();
 
-	if (!cpumask_test_cpu(cpu, buffer->cpumask))
+	if (unlikely(!cpumask_test_cpu(cpu, buffer->cpumask)))
 		goto out;
 
 	cpu_buffer = buffer->buffers[cpu];
 
-	if (atomic_read(&cpu_buffer->record_disabled))
+	if (unlikely(atomic_read(&cpu_buffer->record_disabled)))
 		goto out;
 
-	if (length > BUF_MAX_DATA_SIZE)
+	if (unlikely(length > BUF_MAX_DATA_SIZE))
+		goto out;
+
+	if (unlikely(trace_recursive_lock(cpu_buffer)))
 		goto out;
 
 	event = rb_reserve_next_event(buffer, cpu_buffer, length);
 	if (!event)
-		goto out;
+		goto out_unlock;
 
 	return event;
 
+ out_unlock:
+	trace_recursive_unlock(cpu_buffer);
  out:
-	trace_recursive_unlock();
-
- out_nocheck:
 	preempt_enable_notrace();
 	return NULL;
 }
@@ -2859,7 +2813,7 @@ int ring_buffer_unlock_commit(struct ring_buffer *buffer,
 
 	rb_wakeups(buffer, cpu_buffer);
 
-	trace_recursive_unlock();
+	trace_recursive_unlock(cpu_buffer);
 
 	preempt_enable_notrace();
 
@@ -2970,7 +2924,7 @@ void ring_buffer_discard_commit(struct ring_buffer *buffer,
  out:
 	rb_end_commit(cpu_buffer);
 
-	trace_recursive_unlock();
+	trace_recursive_unlock(cpu_buffer);
 
 	preempt_enable_notrace();
 
@@ -3000,9 +2954,6 @@ int ring_buffer_write(struct ring_buffer *buffer,
 	int ret = -EBUSY;
 	int cpu;
 
-	if (ring_buffer_flags != RB_BUFFERS_ON)
-		return -EBUSY;
-
 	preempt_disable_notrace();
 
 	if (atomic_read(&buffer->record_disabled))
@@ -3021,9 +2972,12 @@ int ring_buffer_write(struct ring_buffer *buffer,
 	if (length > BUF_MAX_DATA_SIZE)
 		goto out;
 
+	if (unlikely(trace_recursive_lock(cpu_buffer)))
+		goto out;
+
 	event = rb_reserve_next_event(buffer, cpu_buffer, length);
 	if (!event)
-		goto out;
+		goto out_unlock;
 
 	body = rb_event_data(event);
 
@@ -3034,6 +2988,10 @@ int ring_buffer_write(struct ring_buffer *buffer,
 	rb_wakeups(buffer, cpu_buffer);
 
 	ret = 0;
+
+ out_unlock:
+	trace_recursive_unlock(cpu_buffer);
+
  out:
 	preempt_enable_notrace();
 
@@ -3860,19 +3818,36 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts)
 }
 EXPORT_SYMBOL_GPL(ring_buffer_iter_peek);
 
-static inline int rb_ok_to_lock(void)
+static inline bool rb_reader_lock(struct ring_buffer_per_cpu *cpu_buffer)
 {
+	if (likely(!in_nmi())) {
+		raw_spin_lock(&cpu_buffer->reader_lock);
+		return true;
+	}
+
 	/*
 	 * If an NMI die dumps out the content of the ring buffer
-	 * do not grab locks. We also permanently disable the ring
-	 * buffer too. A one time deal is all you get from reading
-	 * the ring buffer from an NMI.
+	 * trylock must be used to prevent a deadlock if the NMI
+	 * preempted a task that holds the ring buffer locks. If
+	 * we get the lock then all is fine, if not, then continue
+	 * to do the read, but this can corrupt the ring buffer,
+	 * so it must be permanently disabled from future writes.
+	 * Reading from NMI is a oneshot deal.
 	 */
-	if (likely(!in_nmi()))
-		return 1;
+	if (raw_spin_trylock(&cpu_buffer->reader_lock))
+		return true;
 
-	tracing_off_permanent();
-	return 0;
+	/* Continue without locking, but disable the ring buffer */
+	atomic_inc(&cpu_buffer->record_disabled);
+	return false;
+}
+
+static inline void
+rb_reader_unlock(struct ring_buffer_per_cpu *cpu_buffer, bool locked)
+{
+	if (likely(locked))
+		raw_spin_unlock(&cpu_buffer->reader_lock);
+	return;
 }
 
 /**
@@ -3892,21 +3867,18 @@ ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts,
 	struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu];
 	struct ring_buffer_event *event;
 	unsigned long flags;
-	int dolock;
+	bool dolock;
 
 	if (!cpumask_test_cpu(cpu, buffer->cpumask))
 		return NULL;
 
-	dolock = rb_ok_to_lock();
  again:
 	local_irq_save(flags);
-	if (dolock)
-		raw_spin_lock(&cpu_buffer->reader_lock);
+	dolock = rb_reader_lock(cpu_buffer);
 	event = rb_buffer_peek(cpu_buffer, ts, lost_events);
 	if (event && event->type_len == RINGBUF_TYPE_PADDING)
 		rb_advance_reader(cpu_buffer);
-	if (dolock)
-		raw_spin_unlock(&cpu_buffer->reader_lock);
+	rb_reader_unlock(cpu_buffer, dolock);
 	local_irq_restore(flags);
 
 	if (event && event->type_len == RINGBUF_TYPE_PADDING)
@@ -3959,9 +3931,7 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts,
 	struct ring_buffer_per_cpu *cpu_buffer;
 	struct ring_buffer_event *event = NULL;
 	unsigned long flags;
-	int dolock;
-
-	dolock = rb_ok_to_lock();
+	bool dolock;
 
  again:
 	/* might be called in atomic */
@@ -3972,8 +3942,7 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts,
 
 	cpu_buffer = buffer->buffers[cpu];
 	local_irq_save(flags);
-	if (dolock)
-		raw_spin_lock(&cpu_buffer->reader_lock);
+	dolock = rb_reader_lock(cpu_buffer);
 
 	event = rb_buffer_peek(cpu_buffer, ts, lost_events);
 	if (event) {
@@ -3981,8 +3950,7 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts,
 		rb_advance_reader(cpu_buffer);
 	}
 
-	if (dolock)
-		raw_spin_unlock(&cpu_buffer->reader_lock);
+	rb_reader_unlock(cpu_buffer, dolock);
 	local_irq_restore(flags);
 
  out:
@@ -4263,21 +4231,17 @@ int ring_buffer_empty(struct ring_buffer *buffer)
 {
 	struct ring_buffer_per_cpu *cpu_buffer;
 	unsigned long flags;
-	int dolock;
+	bool dolock;
 	int cpu;
 	int ret;
 
-	dolock = rb_ok_to_lock();
-
 	/* yes this is racy, but if you don't like the race, lock the buffer */
 	for_each_buffer_cpu(buffer, cpu) {
 		cpu_buffer = buffer->buffers[cpu];
 		local_irq_save(flags);
-		if (dolock)
-			raw_spin_lock(&cpu_buffer->reader_lock);
+		dolock = rb_reader_lock(cpu_buffer);
 		ret = rb_per_cpu_empty(cpu_buffer);
-		if (dolock)
-			raw_spin_unlock(&cpu_buffer->reader_lock);
+		rb_reader_unlock(cpu_buffer, dolock);
 		local_irq_restore(flags);
 
 		if (!ret)
@@ -4297,21 +4261,17 @@ int ring_buffer_empty_cpu(struct ring_buffer *buffer, int cpu)
 {
 	struct ring_buffer_per_cpu *cpu_buffer;
 	unsigned long flags;
-	int dolock;
+	bool dolock;
 	int ret;
 
 	if (!cpumask_test_cpu(cpu, buffer->cpumask))
 		return 1;
 
-	dolock = rb_ok_to_lock();
-
 	cpu_buffer = buffer->buffers[cpu];
 	local_irq_save(flags);
-	if (dolock)
-		raw_spin_lock(&cpu_buffer->reader_lock);
+	dolock = rb_reader_lock(cpu_buffer);
 	ret = rb_per_cpu_empty(cpu_buffer);
-	if (dolock)
-		raw_spin_unlock(&cpu_buffer->reader_lock);
+	rb_reader_unlock(cpu_buffer, dolock);
 	local_irq_restore(flags);
 
 	return ret;
@@ -4349,9 +4309,6 @@ int ring_buffer_swap_cpu(struct ring_buffer *buffer_a,
 
 	ret = -EAGAIN;
 
-	if (ring_buffer_flags != RB_BUFFERS_ON)
-		goto out;
-
 	if (atomic_read(&buffer_a->record_disabled))
 		goto out;