On Linux, use the script AMDPcmsetCapability.sh
to run the msr mode without root
privilege. This option collects Core, L3, and DF PMC events on AMD “Zen”-based
processors. The newer processors may require the latest kernel for Perf mode
support.
Examples
| Task Description | Command |
|---|---|
| Timeseries monitoring of IPC of a benchmark, system aggregate level |
$ AMDuProfPcm -m ipc -O /tmp --
/tmp/myapp.exe
|
| Timeseries monitoring of IPC of a benchmark, aggregate metrics per processor package |
$ AMDuProfPcm -m ipc -A package -O /tmp --
/tmp/ myapp.exe
|
| Timeseries monitoring of IPC of a benchmark, aggregate metrics at system level |
$ AMDuProfPcm -m ipc -A system -O /tmp --
/tmp/ myapp.exe
|
| Cumulative reporting of IPC metrics at the end of the benchmark execution |
$ AMDuProfPcm -m ipc -C -O /tmp --
/tmp/myapp.exe
|
| Cumulative reporting of IPC metrics at the end of the benchmark execution, aggregate metrics per processor package |
$ AMDuProfPcm -m ipc -C -A package -O /tmp --
/tmp/ myapp.exe
|
| Cumulative reporting of IPC metrics at the end of the benchmark execution, aggregate metrics at system level |
$ AMDuProfPcm -m ipc -C -A system -O /tmp --
/tmp/ myapp.exe
|
| Timeseries monitoring of memory bandwidth reporting at package and memory channels level |
$ AMDuProfPcm -m memory -a -A system,package
-O /tmp
|
| Timeseries monitoring of level-1 and level-2 top-down metrics (pipeline utilization) |
$ AMDuProfPcm -m pipeline_util -A system -O
/tmp -- / tmp/myapp.exe
|
| Cumulative reporting of level-1 and level-2 top-down metrics (pipeline utilization) |
$ AMDuProfPcm -m pipeline_util -C -A system -O
/tmp -- /tmp/myapp.exe
|
To load the uncore module, use the shell command modprobe amd_uncore which enables DF and L3 PMCs.