When bpf_program has only one instance, don't create a subdirectory with per-instance pin files (<prog>/0). Instead, just create a single pin file for that single instance. This simplifies object pinning by not creating unnecessary subdirectories. This can potentially break existing users that depend on the case where '/0' is always created. However, I couldn't find any serious usage of bpf_program__pin inside the kernel tree and I suppose there should be none outside. Signed-off-by: Stanislav Fomichev <sdf@xxxxxxxxxx> --- tools/lib/bpf/libbpf.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index db84c85554e7..8407a880acbe 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -1761,6 +1761,11 @@ int bpf_program__pin(struct bpf_program *prog, const char *path) return -EINVAL; } + if (prog->instances.nr == 1) { + /* don't create subdirs when pinning single instance */ + return bpf_program__pin_instance(prog, path, 0); + } + err = make_dir(path); if (err) return err; @@ -1823,6 +1828,11 @@ int bpf_program__unpin(struct bpf_program *prog, const char *path) return -EINVAL; } + if (prog->instances.nr == 1) { + /* don't create subdirs when pinning single instance */ + return bpf_program__unpin_instance(prog, path, 0); + } + for (i = 0; i < prog->instances.nr; i++) { char buf[PATH_MAX]; int len; -- 2.19.1.930.g4563a0d9d0-goog