forked from luck/tmp_suning_uos_patched
perf stat: Fix --interval_clear option
Currently we display extra header line, like:
# perf stat -I 1000 -a --interval-clear
# time counts unit events
insn per cycle branch-misses of all branches
2.964917103 3855.349912 cpu-clock (msec) # 3.855 CPUs utilized
2.964917103 23,993 context-switches # 0.006 M/sec
2.964917103 1,301 cpu-migrations # 0.329 K/sec
...
Fixing the condition and getting proper:
# perf stat -I 1000 -a --interval-clear
# time counts unit events
2.359048938 1432.492228 cpu-clock (msec) # 1.432 CPUs utilized
2.359048938 7,613 context-switches # 0.002 M/sec
2.359048938 419 cpu-migrations # 0.133 K/sec
...
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Fixes: 9660e08ee8
("perf stat: Add --interval-clear option")
Link: http://lkml.kernel.org/r/20180702134202.17745-2-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This commit is contained in:
parent
a09603f851
commit
c818cc0630
|
@ -1742,7 +1742,7 @@ static void print_interval(char *prefix, struct timespec *ts)
|
|||
}
|
||||
}
|
||||
|
||||
if ((num_print_interval == 0 && metric_only) || interval_clear)
|
||||
if ((num_print_interval == 0 || interval_clear) && metric_only)
|
||||
print_metric_headers(" ", true);
|
||||
if (++num_print_interval == 25)
|
||||
num_print_interval = 0;
|
||||
|
|
Loading…
Reference in New Issue
Block a user