Reading perf events (as listed by perf list) in BPF Programs

Hi, I would like to access (some of the) hardware/PMC counter events as identified by perf list, in order to be able to perform microarchitectural characterization on a per thread basis using a python/c bcc based tool that has been hacked from llcstat to undertake tracing.

Have I missed an appropriate example on how to configure events that are not listed under help(bcc.PerfHWConfig) help(bcc.PerfSWConfig) in python. Any pointers on which example to look at to achieve this. I'm currently using the stable packages on ubuntu 16.04, kernel 4.13.0-38-generic.

Cheers,
Andy

Start by reading the llcstat-based tracing tool and the Python help for bcc.PerfHWConfig and bcc.PerfSWConfig, then compare those entry points with the events exposed by perf list on the stated kernel. The scope and completion criteria need to be established first; a usable outcome would need a defined way to access the requested per-thread hardware events and an example or validation showing it works.

Written by the indexing model from the issue text.