Because the "tsc" clock isn't in nanoseconds, the ring buffer must be
reset when changing clocks so that incomparable timestamps don't end up
in the same trace.
Tested: Confirmed switching clocks resets the trace buffer.
Google-Bug-Id: 6980623
Signed-off-by: David Sharp <[email protected]>
Cc: Steven Rostedt <[email protected]>
Cc: Masami Hiramatsu <[email protected]>
---
kernel/trace/trace.c | 8 ++++++++
1 files changed, 8 insertions(+), 0 deletions(-)
diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index 9aed7f5..84aa489 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -4017,6 +4017,14 @@ static ssize_t tracing_clock_write(struct file *filp, const char __user *ubuf,
if (max_tr.buffer)
ring_buffer_set_clock(max_tr.buffer, trace_clocks[i].func);
+ /*
+ * New clock may not be consistent with the previous clock.
+ * Reset the buffer so that it doesn't have incomparable timestamps.
+ */
+ tracing_reset_online_cpus(&global_trace);
+ if (max_tr.buffer)
+ tracing_reset_online_cpus(&max_tr);
+
mutex_unlock(&trace_types_lock);
*fpos += cnt;
--
1.7.7.3