From: Andi Kleen <a...@linux.intel.com> Describe the new uncore wildcard PMU match in the perf list manpage.
Also fix a typo in an existing example. Signed-off-by: Andi Kleen <a...@linux.intel.com> --- tools/perf/Documentation/perf-list.txt | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt index f709de54707b..b6eaa2b34921 100644 --- a/tools/perf/Documentation/perf-list.txt +++ b/tools/perf/Documentation/perf-list.txt @@ -134,12 +134,22 @@ all the CPUs in the same socket. This example measures memory bandwidth every second on the first memory controller on socket 0 of a Intel Xeon system - perf stat -C 0 -a uncore_imc_0/cas_count_read/,uncore_imc_0/cas_count_write/ -I 1000 ... + perf stat -C 0 -e uncore_imc_0/cas_count_read/,uncore_imc_0/cas_count_write/ -I 1000 ... Each memory controller has its own PMU. Measuring the complete system bandwidth would require specifying all imc PMUs (see perf list output), and adding the values together. +perf supports wildcard matches on uncore PMUs. All "imc" boxes can be +specified with + + perf stat ... -e uncore_imc/cas_count_read/ -I 1000 + +and perf will automatically measure all "imc" PMUs and sum them together. +For the special case of a uncore prefix this can be also shortened to + + perf stat ... -e imc/cas_count_read/ -I 1000 + This example measures the combined core power every second perf stat -I 1000 -e power/energy-cores/ -a -- 2.9.3
