Eu tenho Intel Xeon Phi 7210
executando CentOS 7
com o kernel 3.10.0-514.el7.x86_64
. Estou comparando quais etapas de frequência estão disponíveis quando eu inicializo o sistema em intel p-state
vs quando o sistema usa acpi cpufreq
. Eu corro sudo cpupower frequency-info
para seguir dois logs.
Intel P-State
analyzing CPU 0:
driver: intel_pstate
CPUs which run at the same hardware frequency: 0
CPUs which need to have their frequency coordinated by software: 0
maximum transition latency: Cannot determine or is not supported.
hardware limits: 1000 MHz - 1.50 GHz
available cpufreq governors: performance powersave
current policy: frequency should be within 1.40 GHz and 1.40 GHz.
The governor "performance" may decide which speed to use
within this range.
current CPU frequency: 1.30 GHz (asserted by call to hardware)
boost state support:
Supported: yes
Active: yes
CPUFREQ da ACPI
analyzing CPU 0:
driver: acpi-cpufreq
CPUs which run at the same hardware frequency: 0
CPUs which need to have their frequency coordinated by software: 0
maximum transition latency: 10.0 us
hardware limits: 1000 MHz - 1.30 GHz
available frequency steps: 1.30 GHz, 1.30 GHz, 1.20 GHz, 1.10 GHz, 1000 MHz
available cpufreq governors: conservative userspace powersave ondemand performance
current policy: frequency should be within 1000 MHz and 1.30 GHz.
The governor "conservative" may decide which speed to use
within this range.
current CPU frequency: 1000 MHz (asserted by call to hardware)
boost state support:
Supported: yes
Active: yes
Por que dois drivers de escala diferentes mostram limites de frequência de hardware diferentes? Para intel p-state
, vejo limites de frequência de 1GHz
a 1.5GHz
e, por outro lado, acpi cpufreq
mostra 1GHz
a 1.3GHz
.
Como por folha de dados, o processador acima pode ser executado em 1.5GHz
com turbo boost, é por isso que ele não está disponível para acpi cpufreq
?
Alguma razão específica para isso?
Obrigado.
Tags intel acpi linux cpu-frequency