A practical reference for exploring Linux perf (perf_events) in performance analysis and debugging of Linux systems, with a range of event sources.
The page collects examples of using Linux perf, also called perf_events, for performance analysis and debugging on Linux systems.
The event sources cited include tracepoints, kprobes and uprobes, hardware performance counters (PMCs), and software events.
To begin, choose a measurable question, such as where execution time is spent. Select a suitable event source and take a short capture before expanding profiling or tracing.
When studying or adapting the examples with AI, remove personal data and secrets from logs, commands, and configurations. Send only the context needed and follow your organization's policy for data that could identify people or systems.