On 08/08/2017 08:05 PM, Laurent Dufour wrote:
quoted hunk ↗ jump to hunk
Add support for the new speculative faults events.
Signed-off-by: Laurent Dufour <redacted>
---
tools/include/uapi/linux/perf_event.h | 2 ++
tools/perf/util/evsel.c | 2 ++
tools/perf/util/parse-events.c | 8 ++++++++
tools/perf/util/parse-events.l | 2 ++
tools/perf/util/python.c | 2 ++
5 files changed, 16 insertions(+)
diff --git a/tools/include/uapi/linux/perf_event.h b/tools/include/uapi/linux/perf_event.h
index b1c0b187acfe..fbfb03dff334 100644
--- a/tools/include/uapi/linux/perf_event.h
+++ b/tools/include/uapi/linux/perf_event.h
@@ -111,6 +111,8 @@ enum perf_sw_ids {
PERF_COUNT_SW_EMULATION_FAULTS = 8,
PERF_COUNT_SW_DUMMY = 9,
PERF_COUNT_SW_BPF_OUTPUT = 10,
+ PERF_COUNT_SW_SPF_DONE = 11,
+ PERF_COUNT_SW_SPF_FAILED = 12,
PERF_COUNT_SW_SPF_FAULTS makes sense but not the FAILED one. IIRC,
there are no error path counting in perf SW events at the moment.
SPF_FAULTS and SPF_FAILS are VM internal events like THP collapse
etc. IMHO it should be added as a VM statistics counter or as a
trace point event instead.