by David Sharp

[permalink] [raw]

Subject: [tip:perf/core] tracing: Reset ring buffer when changing trace_clocks

Commit-ID: 60303ed3f4b9332b9aa9bc17c68bc174e7343e2d
Gitweb: http://git.kernel.org/tip/60303ed3f4b9332b9aa9bc17c68bc174e7343e2d
Author: David Sharp <[email protected]>
AuthorDate: Thu, 11 Oct 2012 16:27:52 -0700
Committer: Steven Rostedt <[email protected]>
CommitDate: Fri, 2 Nov 2012 10:21:47 -0400

tracing: Reset ring buffer when changing trace_clocks

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
Link: http://lkml.kernel.org/r/[email protected]

Cc: Masami Hiramatsu <[email protected]>
Signed-off-by: David Sharp <[email protected]>
Signed-off-by: Steven Rostedt <[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 88111b0..6ed6013 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -4073,6 +4073,14 @@ 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;

2012-11-14 07:04:10

by Yoshihiro YUNOMAE

[permalink] [raw]

Subject: [tip:perf/core] ring-buffer: Change unsigned long type of ring_buffer_oldest_event_ts() to u64

Commit-ID: 50ecf2c3afead23a05227ab004e4212eca08c207
Gitweb: http://git.kernel.org/tip/50ecf2c3afead23a05227ab004e4212eca08c207
Author: Yoshihiro YUNOMAE <[email protected]>
AuthorDate: Thu, 11 Oct 2012 16:27:54 -0700
Committer: Steven Rostedt <[email protected]>
CommitDate: Fri, 2 Nov 2012 10:21:48 -0400

ring-buffer: Change unsigned long type of ring_buffer_oldest_event_ts() to u64

ring_buffer_oldest_event_ts() should return a value of u64 type, because
ring_buffer_per_cpu->buffer_page->buffer_data_page->time_stamp is u64 type.

Link: http://lkml.kernel.org/r/[email protected]

Cc: Frederic Weisbecker <[email protected]>
Cc: Vaibhav Nagarnaik <[email protected]>
Signed-off-by: Yoshihiro YUNOMAE <[email protected]>
Signed-off-by: David Sharp <[email protected]>
Signed-off-by: Steven Rostedt <[email protected]>
---
include/linux/ring_buffer.h | 2 +-
kernel/trace/ring_buffer.c | 4 ++--
2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/include/linux/ring_buffer.h b/include/linux/ring_buffer.h
index 2007375..519777e 100644
--- a/include/linux/ring_buffer.h
+++ b/include/linux/ring_buffer.h
@@ -159,7 +159,7 @@ int ring_buffer_record_is_on(struct ring_buffer *buffer);
void ring_buffer_record_disable_cpu(struct ring_buffer *buffer, int cpu);
void ring_buffer_record_enable_cpu(struct ring_buffer *buffer, int cpu);

-unsigned long ring_buffer_oldest_event_ts(struct ring_buffer *buffer, int cpu);
+u64 ring_buffer_oldest_event_ts(struct ring_buffer *buffer, int cpu);
unsigned long ring_buffer_bytes_cpu(struct ring_buffer *buffer, int cpu);
unsigned long ring_buffer_entries(struct ring_buffer *buffer);
unsigned long ring_buffer_overruns(struct ring_buffer *buffer);
diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
index 23a384b..3c7834c 100644
--- a/kernel/trace/ring_buffer.c
+++ b/kernel/trace/ring_buffer.c
@@ -2932,12 +2932,12 @@ rb_num_of_entries(struct ring_buffer_per_cpu *cpu_buffer)
* @buffer: The ring buffer
* @cpu: The per CPU buffer to read from.
*/
-unsigned long ring_buffer_oldest_event_ts(struct ring_buffer *buffer, int cpu)
+u64 ring_buffer_oldest_event_ts(struct ring_buffer *buffer, int cpu)
{
unsigned long flags;
struct ring_buffer_per_cpu *cpu_buffer;
struct buffer_page *bpage;
- unsigned long ret;
+ u64 ret;

if (!cpumask_test_cpu(cpu, buffer->cpumask))
return 0;