diff options
| author |   Andrii Nakryiko <andriin@fb.com> | 2020-04-28 17:16:06 -0700 |
|---|---|---|
| committer |   Alexei Starovoitov <ast@kernel.org> | 2020-04-28 17:27:08 -0700 |
| commit | a3b80e1078943dc12553166fb08e258463dec013 (patch) | |
| tree | e022212949f8028eacf0c5d3ad77218d7f1a7b1a /include/uapi/linux/bpf.h | |
| parent | bpf: Refactor bpf_link update handling (diff) | |
| download | wireguard-linux-a3b80e1078943dc12553166fb08e258463dec013.tar.xz wireguard-linux-a3b80e1078943dc12553166fb08e258463dec013.zip | |
bpf: Allocate ID for bpf_link
Generate ID for each bpf_link using IDR, similarly to bpf_map and bpf_prog.
bpf_link creation, initialization, attachment, and exposing to user-space
through FD and ID is a complicated multi-step process, abstract it away
through bpf_link_primer and bpf_link_prime(), bpf_link_settle(), and
bpf_link_cleanup() internal API. They guarantee that until bpf_link is
properly attached, user-space won't be able to access partially-initialized
bpf_link either from FD or ID. All this allows to simplify bpf_link attachment
and error handling code.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200429001614.1544-3-andriin@fb.com
Diffstat (limited to 'include/uapi/linux/bpf.h')
| -rw-r--r-- | include/uapi/linux/bpf.h | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 4a6c47f3febe..6121aa487465 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -523,6 +523,7 @@ union bpf_attr { __u32 prog_id; __u32 map_id; __u32 btf_id; + __u32 link_id; }; __u32 next_id; __u32 open_flags; |