diff options
| author | Hubert Chrzaniuk <[email protected]> | 2015-09-14 11:31:00 +0000 |
|---|---|---|
| committer | Len Brown <[email protected]> | 2015-09-26 04:50:54 +0000 |
| commit | b2b34dfe4d9aa4c468fc363b3b666974783ed1f9 (patch) | |
| tree | 1613d67821a25330cee5cdb348c41675f0cb177d /net/switchdev/switchdev.c | |
| parent | tools/power turbostat: IVB Xeon: fix --debug regression (diff) | |
| download | kernel-b2b34dfe4d9aa4c468fc363b3b666974783ed1f9.tar.gz kernel-b2b34dfe4d9aa4c468fc363b3b666974783ed1f9.zip | |
tools/power turbostat: KNL workaround for %Busy and Avg_MHz
KNL increments APERF and MPERF every 1024 clocks.
This is compliant with the architecture specification,
which requires that only the ratio of APERF/MPERF need be valid.
However, turbostat takes advantage of the fact that these
two MSRs increment every un-halted clock
at the actual and base frequency:
AVG_MHz = APERF_delta/measurement_interval
%Busy = MPERF_delta/TSC_delta
This quirk is needed for these calculations to also work on KNL,
which would otherwise show a value 1024x smaller than expected.
Signed-off-by: Hubert Chrzaniuk <[email protected]>
Signed-off-by: Len Brown <[email protected]>
Diffstat (limited to 'net/switchdev/switchdev.c')
0 files changed, 0 insertions, 0 deletions