Re: [PATCH bpf-next v3] libbpf: introduce legacy kprobe events support

[Rafael David Tinoco <rafaeldtinoco@gmail.com>]

> Allow kprobe tracepoint events creation through legacy interface, as the
> kprobe dynamic PMUs support, used by default, was only created in v4.17.
> 
> This enables CO.RE support for older kernels.
> 
> Signed-off-by: Rafael David Tinoco <rafaeldtinoco@gmail.com>

Related to:
https://github.com/libbpf/libbpf/issues/317

> ---
> tools/lib/bpf/libbpf.c | 125 ++++++++++++++++++++++++++++++++++++++++-
> 1 file changed, 123 insertions(+), 2 deletions(-)
> 
> diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
> index 1e04ce724240..72a22c4d8295 100644
> --- a/tools/lib/bpf/libbpf.c
> +++ b/tools/lib/bpf/libbpf.c

[snip]

> static int bpf_link__detach_perf_event(struct bpf_link *link)
> {
> 	int err;
> @@ -10152,6 +10197,12 @@ static int bpf_link__detach_perf_event(struct bpf_link *link)
> 		err = -errno;
> 
> 	close(link->fd);

It needed the perf event fd closure for the ‘kprobe_events’ to allow releasing.

> +
> +	if (link->legacy.name) {
> +		remove_kprobe_event_legacy(link->legacy.name, link->legacy.retprobe);
> +		free(link->legacy.name);
> +	}
> +
> 	return libbpf_err(err);
> }

[snip]

Tested with: https://github.com/rafaeldtinoco/portablebpf (w/ CO.RE) in kernels 5.8 and 4.15.