Feature: PerfEvent support.

[Skrinikov]

• what problem are you trying to solve? (or if there's no problem, what's the motivation for this change?)
  • Adds the ability to attach perf events via attach_perf_event
• what changes does this pull request make?
  • Adds PerfEvents tobcc::core
  • Adds an example which uses perf events so to demonstrate how it works.
• are there any non-obvious implications of these changes? (does it break compatibility with previous versions, etc)
  1. Function signature is a bit heavy. (Python code has a bunch of default values)
  2. Not sure whether the enums PerfType, PerfHardwareConfig and PerfSoftwareConfig are added correctly.
  3. Maybe have a better error message handling (You can't run hardware perf on a VM)
  4. If no cpu is provided we use cpuonline::get per call to attach_perf_event (Some way to cache result)

[brayniac]

Overall, looks good. I'd like to have an example that uses some software event that we can run in CI. Maybe cpu migrations or context switches would make a good smoke test.

For the function signature, I was thinking of ways we can start moving to a builder pattern in this library. Does not need to be addressed in this PR

[brayniac]

There are a couple other lints that you might want to address: https://github.com/rust-bpf/rust-bcc/pull/106/checks?check_run_id=920461866#step:4:40

[Skrinikov]

There are a couple other lints that you might want to address: https://github.com/rust-bpf/rust-bcc/pull/106/checks?check_run_id=920461866#step:4:40

Changed everything besides the fn signature max argument count for now.

not sure why format fails right now.