> > what does "rdmsr 0x64D" return when there are [now 4] fixed counters? OK, so I rebooted after a power cycle, and here's the results with 4 fixed counters (i.e., it sees the "psys domain"): $ sudo rdmsr 0x64D a6e824 $ dmesg | fgrep RAPL [ 1.761697] RAPL PMU: API unit is 2^-32 Joules, 4 fixed counters, 655360 ms ovfl timer [ 1.761703] RAPL PMU: hw unit of domain pp0-core 2^-14 Joules [ 1.761706] RAPL PMU: hw unit of domain package 2^-14 Joules [ 1.761709] RAPL PMU: hw unit of domain pp1-gpu 2^-14 Joules [ 1.761712] RAPL PMU: hw unit of domain psys 2^-14 Joules [ 3.119434] intel_rapl_common: Found RAPL domain package [ 4.007495] intel_rapl_common: Found RAPL domain package [ 4.009635] intel_rapl_common: Found RAPL domain core [ 4.011749] intel_rapl_common: Found RAPL domain uncore > > if it returns 0, which I think it is the case, does reading the msr > > from other cpus return non-zero value, by using rdmsr -p parameter? Here's the "4 counters" case: MSR value on CPU 0 for 0x64D is: 1079240 MSR value on CPU 1 for 0x64D is: 10793e2 MSR value on CPU 2 for 0x64D is: 1079512 MSR value on CPU 3 for 0x64D is: 10796e4 MSR value on CPU 4 for 0x64D is: 107987d MSR value on CPU 5 for 0x64D is: 1079a23 MSR value on CPU 6 for 0x64D is: 1079ab2 MSR value on CPU 7 for 0x64D is: 1079b4e > > > > if that is not the problem, please attach the output of "grep . > > > > /sys/class/powercap/intel-rapl*/name" in both cases. Here's the "4 counters case" (same running kernel as above): $ sudo egrep -r . /sys/class/powercap/intel-rapl*/name /sys/class/powercap/intel-rapl:0:0/name:core /sys/class/powercap/intel-rapl:0:1/name:uncore /sys/class/powercap/intel-rapl:0/name:package-0 /sys/class/powercap/intel-rapl:1/name:psys /sys/class/powercap/intel-rapl-mmio:0/name:package-0 $ -Kenny -- Kenneth R. Crudup Sr. SW Engineer, Scott County Consulting, Silicon Valley
