diff options
author | Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> | 2016-01-25 23:05:36 -0500 |
---|---|---|
committer | Michael Ellerman <mpe@ellerman.id.au> | 2016-03-10 06:56:55 -0500 |
commit | 2b206ee6b0df03a89783c6a9ada363122f918800 (patch) | |
tree | d31a1dca3ba011bbef8e3ddda9412d7e749bf4fe | |
parent | e5a5886d7ae32b7afebfffecca340e466e4be2d1 (diff) |
powerpc/perf/hv-24x7: Display change in counter values
For 24x7 counters, perf displays the raw value of the 24x7 counter, which
is a monotonically increasing value.
perf stat -C 0 -e \
'hv_24x7/HPM_0THRD_NON_IDLE_CCYC__PHYS_CORE,core=1/' \
sleep 1
Performance counter stats for 'CPU(s) 0':
9,105,403,170 hv_24x7/HPM_0THRD_NON_IDLE_CCYC__PHYS_CORE,core=1/
0.000425751 seconds time elapsed
In the typical usage of 'perf stat' this counter value is not as useful
as the _change_ in the counter value over the duration of the application.
Have h_24x7_event_init() set the event's prev_count to the raw value of
the 24x7 counter at the time of initialization. When the application
terminates, hv_24x7_event_read() will compute the change in value and
report to the perf tool. Similarly, for the transaction interface, clear
the event count to 0 at the beginning of the transaction.
perf stat -C 0 -e \
'hv_24x7/HPM_0THRD_NON_IDLE_CCYC__PHYS_CORE,core=1/' \
sleep 1
Performance counter stats for 'CPU(s) 0':
245,758 hv_24x7/HPM_0THRD_NON_IDLE_CCYC__PHYS_CORE,core=1/
1.006366383 seconds time elapsed
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
-rw-r--r-- | arch/powerpc/perf/hv-24x7.c | 13 |
1 files changed, 12 insertions, 1 deletions
diff --git a/arch/powerpc/perf/hv-24x7.c b/arch/powerpc/perf/hv-24x7.c index b7a9a03ca59d..77b958f213a0 100644 --- a/arch/powerpc/perf/hv-24x7.c +++ b/arch/powerpc/perf/hv-24x7.c | |||
@@ -1222,11 +1222,12 @@ static int h_24x7_event_init(struct perf_event *event) | |||
1222 | return -EACCES; | 1222 | return -EACCES; |
1223 | } | 1223 | } |
1224 | 1224 | ||
1225 | /* see if the event complains */ | 1225 | /* Get the initial value of the counter for this event */ |
1226 | if (single_24x7_request(event, &ct)) { | 1226 | if (single_24x7_request(event, &ct)) { |
1227 | pr_devel("test hcall failed\n"); | 1227 | pr_devel("test hcall failed\n"); |
1228 | return -EIO; | 1228 | return -EIO; |
1229 | } | 1229 | } |
1230 | (void)local64_xchg(&event->hw.prev_count, ct); | ||
1230 | 1231 | ||
1231 | return 0; | 1232 | return 0; |
1232 | } | 1233 | } |
@@ -1289,6 +1290,16 @@ static void h_24x7_event_read(struct perf_event *event) | |||
1289 | h24x7hw = &get_cpu_var(hv_24x7_hw); | 1290 | h24x7hw = &get_cpu_var(hv_24x7_hw); |
1290 | h24x7hw->events[i] = event; | 1291 | h24x7hw->events[i] = event; |
1291 | put_cpu_var(h24x7hw); | 1292 | put_cpu_var(h24x7hw); |
1293 | /* | ||
1294 | * Clear the event count so we can compute the _change_ | ||
1295 | * in the 24x7 raw counter value at the end of the txn. | ||
1296 | * | ||
1297 | * Note that we could alternatively read the 24x7 value | ||
1298 | * now and save its value in event->hw.prev_count. But | ||
1299 | * that would require issuing a hcall, which would then | ||
1300 | * defeat the purpose of using the txn interface. | ||
1301 | */ | ||
1302 | local64_set(&event->count, 0); | ||
1292 | } | 1303 | } |
1293 | 1304 | ||
1294 | put_cpu_var(hv_24x7_reqb); | 1305 | put_cpu_var(hv_24x7_reqb); |