Merge tag 'trace-v5.7' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace

Pull tracing updates from Steven Rostedt:
 "New tracing features:

   - The ring buffer is no longer disabled when reading the trace file.

     The trace_pipe file was made to be used for live tracing and
     reading as it acted like the normal producer/consumer. As the trace
     file would not consume the data, the easy way of handling it was to
     just disable writes to the ring buffer.

     This came to a surprise to the BPF folks who complained about lost
     events due to reading. This is no longer an issue. If someone wants
     to keep the old disabling there's a new option "pause-on-trace"
     that can be set.

   - New set_ftrace_notrace_pid file. PIDs in this file will not be
     traced by the function tracer.

     Similar to set_ftrace_pid, which makes the function tracer only
     trace those tasks with PIDs in the file, the set_ftrace_notrace_pid
     does the reverse.

   - New set_event_notrace_pid file. PIDs in this file will cause events
     not to be traced if triggered by a task with a matching PID.

     Similar to the set_event_pid file but will not be traced. Note,
     sched_waking and sched_switch events may still be traced if one of
     the tasks referenced by those events contains a PID that is allowed
     to be traced.

  Tracing related features:

   - New bootconfig option, that is attached to the initrd file.

     If bootconfig is on the command line, then the initrd file is
     searched looking for a bootconfig appended at the end.

   - New GPU tracepoint infrastructure to help the gfx drivers to get
     off debugfs (acked by Greg Kroah-Hartman)

  And other minor updates and fixes"

* tag 'trace-v5.7' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace: (27 commits)
  tracing: Do not allocate buffer in trace_find_next_entry() in atomic
  tracing: Add documentation on set_ftrace_notrace_pid and set_event_notrace_pid
  selftests/ftrace: Add test to test new set_event_notrace_pid file
  selftests/ftrace: Add test to test new set_ftrace_notrace_pid file
  tracing: Create set_event_notrace_pid to not trace tasks
  ftrace: Create set_ftrace_notrace_pid to not trace tasks
  ftrace: Make function trace pid filtering a bit more exact
  ftrace/kprobe: Show the maxactive number on kprobe_events
  tracing: Have the document reflect that the trace file keeps tracing enabled
  ring-buffer/tracing: Have iterator acknowledge dropped events
  tracing: Do not disable tracing when reading the trace file
  ring-buffer: Do not disable recording when there is an iterator
  ring-buffer: Make resize disable per cpu buffer instead of total buffer
  ring-buffer: Optimize rb_iter_head_event()
  ring-buffer: Do not die if rb_iter_peek() fails more than thrice
  ring-buffer: Have rb_iter_head_event() handle concurrent writer
  ring-buffer: Add page_stamp to iterator for synchronization
  ring-buffer: Rename ring_buffer_read() to read_buffer_iter_advance()
  ring-buffer: Have ring_buffer_empty() not depend on tracing stopped
  tracing: Save off entry when peeking at next entry
  ...

1 files changed, 17 insertions, 7 deletions

diff --git a/kernel/trace/trace_hwlat.c b/kernel/trace/trace_hwlat.c
index a48808c43249..e2be7bb7ef7e 100644
--- a/kernel/trace/trace_hwlat.c
+++ b/kernel/trace/trace_hwlat.c
@@ -83,6 +83,7 @@ struct hwlat_sample {
 	u64			nmi_total_ts;	/* Total time spent in NMIs */
 	struct timespec64	timestamp;	/* wall time */
 	int			nmi_count;	/* # NMIs during this sample */
+	int			count;		/* # of iteratons over threash */
 };
 
 /* keep the global state somewhere. */
@@ -124,6 +125,7 @@ static void trace_hwlat_sample(struct hwlat_sample *sample)
 	entry->timestamp		= sample->timestamp;
 	entry->nmi_total_ts		= sample->nmi_total_ts;
 	entry->nmi_count		= sample->nmi_count;
+	entry->count			= sample->count;
 
 	if (!call_filter_check_discard(call, entry, buffer, event))
 		trace_buffer_unlock_commit_nostack(buffer, event);
@@ -167,12 +169,14 @@ void trace_hwlat_callback(bool enter)
 static int get_sample(void)
 {
 	struct trace_array *tr = hwlat_trace;
+	struct hwlat_sample s;
 	time_type start, t1, t2, last_t2;
-	s64 diff, total, last_total = 0;
+	s64 diff, outer_diff, total, last_total = 0;
 	u64 sample = 0;
 	u64 thresh = tracing_thresh;
 	u64 outer_sample = 0;
 	int ret = -1;
+	unsigned int count = 0;
 
 	do_div(thresh, NSEC_PER_USEC); /* modifies interval value */
 
@@ -186,6 +190,7 @@ static int get_sample(void)
 
 	init_time(last_t2, 0);
 	start = time_get(); /* start timestamp */
+	outer_diff = 0;
 
 	do {
 
@@ -194,14 +199,14 @@ static int get_sample(void)
 
 		if (time_u64(last_t2)) {
 			/* Check the delta from outer loop (t2 to next t1) */
-			diff = time_to_us(time_sub(t1, last_t2));
+			outer_diff = time_to_us(time_sub(t1, last_t2));
 			/* This shouldn't happen */
-			if (diff < 0) {
+			if (outer_diff < 0) {
 				pr_err(BANNER "time running backwards\n");
 				goto out;
 			}
-			if (diff > outer_sample)
-				outer_sample = diff;
+			if (outer_diff > outer_sample)
+				outer_sample = outer_diff;
 		}
 		last_t2 = t2;
 
@@ -217,6 +222,12 @@ static int get_sample(void)
 		/* This checks the inner loop (t1 to t2) */
 		diff = time_to_us(time_sub(t2, t1));     /* current diff */
 
+		if (diff > thresh || outer_diff > thresh) {
+			if (!count)
+				ktime_get_real_ts64(&s.timestamp);
+			count++;
+		}
+
 		/* This shouldn't happen */
 		if (diff < 0) {
 			pr_err(BANNER "time running backwards\n");
@@ -236,7 +247,6 @@ static int get_sample(void)
 
 	/* If we exceed the threshold value, we have found a hardware latency */
 	if (sample > thresh || outer_sample > thresh) {
-		struct hwlat_sample s;
 		u64 latency;
 
 		ret = 1;
@@ -249,9 +259,9 @@ static int get_sample(void)
 		s.seqnum = hwlat_data.count;
 		s.duration = sample;
 		s.outer_duration = outer_sample;
-		ktime_get_real_ts64(&s.timestamp);
 		s.nmi_total_ts = nmi_total_ts;
 		s.nmi_count = nmi_count;
+		s.count = count;
 		trace_hwlat_sample(&s);
 
 		latency = max(sample, outer_sample);