PMDAAMDGPU(1)

💖 PMDAAMDGPU(1) 💖

okay bestie, welcome to the pmdaamdgpu era — that's the manual page for the AMD GPU metrics domain agent (PMDA) from the Performance Co-Pilot (PCP) crew ✨

---

🌸 NAME 🌸

pmdaamdgpu — AMD GPU metrics domain agent (PMDA) 🎀

---

✨ SYNOPSIS ✨

$PCP_PMDAS_DIR/amdgpu/pmdaamdgpu [-D debug] [-d domain] [-l logfile] [-t interval]

---

🦋 DESCRIPTION 🦋

pmdaamdgpu is a Performance Metrics Domain Agent (PMDA) — and what it does is pull out performance metrics that describe all the metrics available on AMD GPU cards, via the DRM library. no cap 💅

The amdgpu PMDA exports metrics that measure GPU activity, memory utilization, temperature, etc — on GCN 1.2+ AMD GPUs 👑

A brief description of the pmdaamdgpu command line options follows:

• -d — It is absolutely crucial that the performance metrics domain number you specify here is unique and consistent. Like, that means domain should be different for every PMDA on the one host, and the same domain number should be used for the same PMDA on all hosts. fr fr, don't play with this one 🙈
• -l — Location of the log file. By default, a log file named amdgpu.log gets written in the current directory of pmcd(1) when pmdaamdgpu is started, i.e. $PCP_LOG_DIR/pmcd. If the log file can't be created or isn't writable, output goes to standard error instead 💖
• -t — Enables and sets a sampling interval for automatic refreshing of metric values. This functionality is disabled by default, however this option lets you specify a time interval on which all values are sampled — and the effect of that is constantly updating the accumulating metrics, with the goal of helping client tools like pcp-atop(1) and pmlogger(1) observe sub-sample time changes in GPU and process state 🌈 Typically these tools have longer sampling intervals, so they can 'miss' activity happening during their sampling interval. not on our watch ✋

---

🎀 INSTALLATION 🎀

The amdgpu PMDA is not installed and available by default. If you want to undo the installation, do the following as root:

# cd $PCP_PMDAS_DIR/amdgpu
# ./Remove

If you want to establish access to the names, help text and values for the AMD GPU performance metrics once more, after removal, do the following as root:

# cd $PCP_PMDAS_DIR/amdgpu
# ./Install

pmdaamdgpu is launched by pmcd(1) and should never be executed directly 🚫 The Install and Remove scripts notify pmcd(1) when the agent is installed or removed. iconic behavior, honestly 👑

---

🍓 FILES 🍓

• $PCP_PMCDCONF_PATH — command line options used to launch pmdaamdgpu
• $PCP_PMDAS_DIR/amdgpu/help — default help text file for the amdgpu metrics
• $PCP_PMDAS_DIR/amdgpu/Install — installation script for the pmdaamdgpu agent
• $PCP_PMDAS_DIR/amdgpu/Remove — undo installation script for the pmdaamdgpu agent
• $PCP_LOG_DIR/pmcd/amdgpu.log — default log file for error messages and other information from pmdaamdgpu

---

🧚 PCP ENVIRONMENT 🧚

Environment variables with the prefix PCP_ are used to configure the file and directory names used by PCP. On each installation, the file /etc/pcp.conf contains the local values for these variables. The $PCP_CONF variable may be used to specify an alternative configuration file, as described in pcp.conf(5) ✨

---

💀 DEBUGGING OPTIONS 💀

The -D or --debug option enables the output of additional diagnostics on stderr to help triage problems — although ngl, the information is sometimes cryptic and primarily intended to give guidance for developers rather than for end-users 🙈 debug is a comma separated list of debugging options; use pmdbg(1) with the -l option to get a list of the available debugging options and their meaning 👑

Debugging options specific to pmdaamdgpu are as follows:

• appl0 — interaction with pmcd(1)
• appl2 — trace select(2) calls

---

🌈 SEE ALSO 🌈

PCPIntro(1), pcp-atop(1), pmcd(1), pmlogger(1), pcp.conf(5) and pcp.env(5) 🫶

---

and that's the whole page bestie, you ate 💅✨ this one came from the Performance Co-Pilot collection, tagged PCP, and it's the PMDAAMDGPU(1) page — mwah, go read it again if you need to, ily 🎀💖🦋