Add extra_flag option to bpf_attach_perf_event_raw

[palmtenor]

The bpf_attach_perf_event_raw API is designed to provide maximum flexibility for people to use advanced features of Kernel Perf Events with BPF. Some times specifying flags is necessary, such as if we want to use PERF_FLAG_PID_CGROUP to profile a container. This commit adds extra_flag option to C and C++ interface

I fee bad for the interface change after introducing the interface. However it was not so long time ago so I guess it should be find...:)

[yonghong-song]

@palmtenor do you think this extra_flags will be also useful for python scripts using raw_tracepoints? Adding this to python API as well to keep API parity between C++ and python?

[palmtenor]

@yonghong-song I think this is mainly targeting on the (hardware) Perf Event, not Raw Tracepoint (although it also uses Perf Event interface). We discussed that it might be harder for Python users to fine-tune the perf_event_attr struct, and that's not the normal use case in a Python script tool. I still feel the same way, what do you think?

[yonghong-song]

Sorry. I messed raw_tracepoints with attach_perf_event. Okay, let us only add the extra flags in C++ now. We can add to python API later if there is ever a need. Thanks!