On AMD “Zen4” and “Zen5”-based processors, AMDuProfPcm supports monitoring and
reporting the pipeline utilization (pipeline_util) metrics. This feature provides
pipeline_util metrics to visualize the bottlenecks in the CPU pipeline. Use the
option -m pipeline_util to monitor and report the level-1 and
level-2 top-down metrics.
Level-1 Metrics
| Metric | Description |
|---|---|
| Total_Disp_Slots | Total dispatch slots; up to six instructions can be dispatched in one cycle. |
| SMT_Disp_contention | Unused dispatch slots as the other thread was selected. |
| Frontend_Bound | Dispatch slots that remained unused because the frontend did not supply appropriate instructions/ops. |
| Bad_Speculation | Dispatched operations that did not retire. |
| Backend_Bound | Dispatch slots that remained unused because of backend stalls. |
| Retiring | Dispatch slots used by operations that retired. |
Level-2 Metrics
| Metric | Description |
|---|---|
| Frontend_Bound.Latency | Unused dispatch slots due to latency bottleneck in the frontend, such as Instruction Cache or ITLB misses. |
| Frontend_Bound.BW | Unused dispatch slots due to bandwidth bottleneck in the frontend, such as decode bandwidth or Op Cache fetch bandwidth. |
| Bad_Speculation.Mispredicts | Dispatched operations that were flushed due to branch mis- predicts. |
| Bad_Speculation.Pipeline_Restarts | Dispatched operations that were flushed due to pipeline restarts (resyncs). |
| Backend_Bound.Memory | Dispatched slots that remained unused because of stalls due to memory subsystem. |
| Backend_Bound.CPU | Dispatched slots that remained unused because of stalls not related to the memory subsystem. |
| Retiring.Fastpath | Dispatch slots used by fastpath operations that retired. |
| Retiring.Microcode | Dispatch slots used by microcode operations that retired. |
Due to multiplexing, the reported metrics may be inconsistent. For better results, use taskset to bind the monitored application to a specific set of cores and monitor only the cores on which the monitored application is running.
Run the following command to collect the top-down metrics:
AMDuProfPCm -m pipeline_util --msr-A system -o /tmp/myapp-td.csv -- /usr/bin/taskset -c 0 myapp.exe--msr option requires root privileges. Run sudo
./AMDPcmSetCapability.sh, then open a new terminal tab or run without
--msr as noted
here:sudo AMDuProfPCm -m pipeline_util -c core=0 -A system -o /tmp/myapp-td.csv -- /usr/bin/ taskset -c 0 myapp.exeA sample report is as follows:
Examples
| Task Description | Command |
|---|---|
| Timeseries monitoring of level-1 and level-2 top-down metrics (pipeline utilization) of a single- threaded program |
# AMDuProfPcm -m
pipeline_util -c core=1 -o /tmp/td.csv -- /usr/bin/taskset -c 1
/tmp/myapp.exe
|
| Timeseries monitoring of level-1 and level-2 top-down metrics of a multi-threaded program running on all the cores: |
# AMDuProfPcm -m
pipeline_util -a -A system -o /tmp/td.csv --
/tmp/myapp.exe
|
| Cumulative monitoring of level-1 and level-2 top-down metrics of a multi-threaded program running on all the cores |
# AMDuProfPcm -m
pipeline_util -a -A system -C -o /tmp/td.csv --
/tmp/myapp.exe
|