Missing ref_ctr_offset arg in bpf_attach_uprobe python def

`bpf_attach_uprobe` is defined as:
https://github.com/iovisor/bcc/blob/db5b63ff876d3346021871e2189a354bfc6d510e/src/cc/libbpf.h#L111-L113

but its python definition seems to be missing the last arg (`uint32_t ref_ctr_offset`):
https://github.com/iovisor/bcc/blob/db5b63ff876d3346021871e2189a354bfc6d510e/src/python/bcc/libbcc.py#L113-L115

It seems this results in garbage being passed to the C call when attaching probes from
https://github.com/iovisor/bcc/blob/db5b63ff876d3346021871e2189a354bfc6d510e/src/python/bcc/__init__.py#L1425
https://github.com/iovisor/bcc/blob/db5b63ff876d3346021871e2189a354bfc6d510e/src/python/bcc/__init__.py#L1457

I've noticed attaching a uprobe/uretprobe pair would fail for no apparent reason after minor changes in the python code. It seems `bpf_attach_uprobe` can take any value for `ref_ctr_offset` and it works fine as long as the value is the same for the same probe location.

What would be the appropriate fix for this ? Always passing 0 to `ref_ctr_offset` ?

Compare the bpf_attach_uprobe declaration in src/cc/libbpf.h with its Python binding in src/python/bcc/libbcc.py, then trace the calls in src/python/bcc/__init__.py at the referenced locations. Check existing tests for uprobe and uretprobe attachment behavior. Done means the binding and C signature agree and attaching the probe pair no longer receives an unintended argument value.

Written by the indexing model from the issue text.