ring-buffer: Check if absolute timestamp goes backwards

Commit Message

Steven Rostedt Dec. 19, 2023, 3:16 a.m. UTC

From: "Steven Rostedt (Google)" <[e-mail]>

The check_buffer() which checks the timestamps of the ring buffer
sub-buffer page, when enabled, only checks if the adding of deltas of the
events from the last absolute timestamp or the timestamp of the sub-buffer
page adds up to the current event.

What it does not check is if the absolute timestamp causes the time of the
events to go backwards, as that can cause issues elsewhere.

Test for the timestamp going backwards too.

This also fixes a slight issue where if the warning triggers at boot up
(because of the resetting of the tsc), it will disable all further checks,
even those that are after boot Have it continue checking if the warning
was ignored during boot up.

Signed-off-by: Steven Rostedt (Google) <[e-mail]>
---
 kernel/trace/ring_buffer.c | 46 +++++++++++++++++++++++---------------
 1 file changed, 28 insertions(+), 18 deletions(-)

Patch

diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
index 73f142255274..3b1e2ce0dad5 100644
--- a/kernel/trace/ring_buffer.c
+++ b/kernel/trace/ring_buffer.c
@@ -3309,6 +3309,23 @@  static void dump_buffer_page(struct buffer_data_page *bpage,
 static DEFINE_PER_CPU(atomic_t, checking);
 static atomic_t ts_dump;
 
+#define buffer_warn_return(fmt, ...)					\
+	do {								\
+		/* If another report is happening, ignore this one */	\
+		if (atomic_inc_return(&ts_dump) != 1) {			\
+			atomic_dec(&ts_dump);				\
+			goto out;					\
+		}							\
+		atomic_inc(&cpu_buffer->record_disabled);		\
+		pr_warn(fmt, ##__VA_ARGS__);				\
+		dump_buffer_page(bpage, info, tail);			\
+		atomic_dec(&ts_dump);					\
+		/* There's some cases in boot up that this can happen */ \
+		if (WARN_ON_ONCE(system_state != SYSTEM_BOOTING))	\
+			/* Do not re-enable checking */			\
+			return;						\
+	} while (0)
+
 /*
  * Check if the current event time stamp matches the deltas on
  * the buffer page.
@@ -3362,7 +3379,12 @@  static void check_buffer(struct ring_buffer_per_cpu *cpu_buffer,
 
 		case RINGBUF_TYPE_TIME_STAMP:
 			delta = rb_event_time_stamp(event);
-			ts = rb_fix_abs_ts(delta, ts);
+			delta = rb_fix_abs_ts(delta, ts);
+			if (delta < ts) {
+				buffer_warn_return("[CPU: %d]ABSOLUTE TIME WENT BACKWARDS: last ts: %lld absolute ts: %lld\n",
+						   cpu_buffer->cpu, ts, delta);
+			}
+			ts = delta;
 			break;
 
 		case RINGBUF_TYPE_PADDING:
@@ -3379,23 +3401,11 @@  static void check_buffer(struct ring_buffer_per_cpu *cpu_buffer,
 	}
 	if ((full && ts > info->ts) ||
 	    (!full && ts + info->delta != info->ts)) {
-		/* If another report is happening, ignore this one */
-		if (atomic_inc_return(&ts_dump) != 1) {
-			atomic_dec(&ts_dump);
-			goto out;
-		}
-		atomic_inc(&cpu_buffer->record_disabled);
-		/* There's some cases in boot up that this can happen */
-		WARN_ON_ONCE(system_state != SYSTEM_BOOTING);
-		pr_warn("[CPU: %d]TIME DOES NOT MATCH expected:%lld actual:%lld delta:%lld before:%lld after:%lld%s context:%s\n",
-			cpu_buffer->cpu,
-			ts + info->delta, info->ts, info->delta,
-			info->before, info->after,
-			full ? " (full)" : "", show_interrupt_level());
-		dump_buffer_page(bpage, info, tail);
-		atomic_dec(&ts_dump);
-		/* Do not re-enable checking */
-		return;
+		buffer_warn_return("[CPU: %d]TIME DOES NOT MATCH expected:%lld actual:%lld delta:%lld before:%lld after:%lld%s context:%s\n",
+				   cpu_buffer->cpu,
+				   ts + info->delta, info->ts, info->delta,
+				   info->before, info->after,
+				   full ? " (full)" : "", show_interrupt_level());
 	}
 out:
 	atomic_dec(this_cpu_ptr(&checking));