perf: Use generic sample reordering in perf kmem

Use the new generic sample events reordering from perf kmem,
this drops the need of multiplexing the buffers on record time,
improving the scalability of perf kmem.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Masami Hiramatsu <mhiramat@redhat.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: Li Zefan <lizf@cn.fujitsu.com>

tools/perf/builtin-kmem.c

@@ -335,8 +335,9 @@ static int process_sample_event(event_t *event, struct perf_session *session)
 }
 
 static struct perf_event_ops event_ops = {
-	.sample	= process_sample_event,
-	.comm	= event__process_comm,
+	.sample			= process_sample_event,
+	.comm			= event__process_comm,
+	.ordered_samples	= true,
 };
 
 static double fragmentation(unsigned long n_req, unsigned long n_alloc)
@@ -730,7 +731,6 @@ static const char *record_args[] = {
 	"record",
 	"-a",
 	"-R",
-	"-M",
 	"-f",
 	"-c", "1",
 	"-e", "kmem:kmalloc",