From: Kan Liang <[email protected]>
perf_top__mmap_read has severe performance issue in
Knights Landing/Mill, when monitoring in heavy load system. It costs
several minutes to finish, which is unacceptable.
Currently, perf top is non overwrite mode. For non overwrite mode, it
tries to read everything in the ringbuffer and doesn't pause the
ringbuffer. Once there are lots of samples delivered persistently,
the processing time could be very long. Also, the latest samples could
be lost when the ringbuffer is full.
It's better to change it to overwrite mode, which takes a snapshot for
the system by pausing the ringbuffer and could significantly reducing
the processing time (from several minutes to several seconds).
Also, the overwrite mode always keep the latest samples.
Patch 1-3: Modify perf_mmap__read_catchup and perf_mmap__read_backward.
Make them ready for overwrite mode
Patch 4: Swith perf top to overwrite mode
Patch 5: The latency could be still higher than refresh time in some
extreme cases. Give user some hints to reduce the latency.
Kan Liang (5):
perf tools: remove stale perf evlist mmap read for backward
perf tools: rewrite perf mmap read for overwrite
perf tools: reuse perf_mmap__read_catchup in perf_mmap__push
perf top: switch to overwrite mode
perf top: check the latency of perf_top__mmap_read
tools/perf/builtin-top.c | 40 ++++++-----
tools/perf/tests/backward-ring-buffer.c | 9 ++-
tools/perf/ui/browsers/hists.c | 12 +++-
tools/perf/util/evlist.c | 17 -----
tools/perf/util/evlist.h | 4 --
tools/perf/util/mmap.c | 119 +++++++++++++++-----------------
tools/perf/util/mmap.h | 7 +-
7 files changed, 102 insertions(+), 106 deletions(-)
--
2.5.5
