xref: /aosp_15_r20/external/bcc/tools/llcstat_example.txt (revision 387f9dfdfa2baef462e92476d413c7bc2470293e)

Demonstrations of llcstat.


llcstat traces cache reference and cache miss events system-wide, and summarizes
them by PID and CPU.

These events, defined in uapi/linux/perf_event.h, have different meanings on
different architecture. For x86-64, they mean misses and references to LLC.

Example output:

# ./llcstat.py 20 -c 5000
Running for 20 seconds or hit Ctrl-C to end.
PID      NAME             CPU     REFERENCE         MISS   HIT%
0        swapper/15       15        3515000       640000  81.79%
238      migration/38     38           5000            0 100.00%
4512     ntpd             11           5000            0 100.00%
150867   ipmitool         3           25000         5000  80.00%
150895   lscpu            17         280000        25000  91.07%
151807   ipmitool         15          15000         5000  66.67%
150757   awk              2           15000         5000  66.67%
151213   chef-client      5         1770000       240000  86.44%
151822   scribe-dispatch  12          15000            0 100.00%
123386   mysqld           5            5000            0 100.00%
[...]
Total References: 518920000 Total Misses: 90265000 Hit Rate: 82.61%

This shows each PID's cache hit rate during the 20 seconds run period.

A count of 5000 was used in this example, which means that one in every 5,000
events will trigger an in-kernel counter to be incremented. This is refactored
on the output, which is why it is always in multiples of 5,000.

We don't instrument every single event since the overhead would be prohibitive,
nor do we need to: this is a type of sampling profiler. Because of this, the
processes that trigger the 5,000'th cache reference or misses can happen to
some degree by chance. Overall it should make sense. But for low counts,
you might find a case where -- by chance -- a process has been tallied with
more misses than references, which would seem impossible.

# ./llcstat.py 10 -t
Running for 10 seconds or hit Ctrl-C to end.
PID      TID      NAME             CPU     REFERENCE         MISS    HIT%
170843   170845   docker           12           2700         1200  55.56%
298670   298670   kworker/15:0     15            500            0 100.00%
170254   170254   kworker/11:1     11           2500          400  84.00%
1046952  1046953  git              0            2600         1100  57.69%
170843   170849   docker           15           1000          400  60.00%
1027373  1027382  node             8            3500         2500  28.57%
0        0        swapper/7        7          173000         4200  97.57%
1028217  1028217  node             14          15600        22400   0.00%
[...]
Total References: 7139900 Total Misses: 1413900 Hit Rate: 80.20%

This shows each TID`s cache hit rate during the 10 seconds run period.

USAGE message:

# ./llcstat.py --help
usage: llcstat.py [-h] [-c SAMPLE_PERIOD] [duration]

Summarize cache references and misses by PID

positional arguments:
  duration                Duration, in seconds, to run

  optional arguments:
    -h, --help            show this help message and exit
    -c SAMPLE_PERIOD, --sample_period SAMPLE_PERIOD
                          Sample one in this many number of cache reference
                          and miss events
    -t, --tid             Summarize cache references and misses by PID/TID