Re: [PATCH bpf-next v2 1/6] libbpf: Unpin auto-pinned maps if loading fails

[Andrii Nakryiko <andrii.nakryiko@gmail.com>]

On Fri, Nov 8, 2019 at 3:33 PM Toke Høiland-Jørgensen <toke@redhat.com> wrote:
>
> Andrii Nakryiko <andrii.nakryiko@gmail.com> writes:
>
> > On Fri, Nov 8, 2019 at 1:33 PM Toke Høiland-Jørgensen <toke@redhat.com> wrote:
> >>
> >> From: Toke Høiland-Jørgensen <toke@redhat.com>
> >>
> >> Since the automatic map-pinning happens during load, it will leave pinned
> >> maps around if the load fails at a later stage. Fix this by unpinning any
> >> pinned maps on cleanup. To avoid unpinning pinned maps that were reused
> >> rather than newly pinned, add a new boolean property on struct bpf_map to
> >> keep track of whether that map was reused or not; and only unpin those maps
> >> that were not reused.
> >>
> >> Fixes: 57a00f41644f ("libbpf: Add auto-pinning of maps when loading BPF objects")
> >> Acked-by: Song Liu <songliubraving@fb.com>
> >> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
> >> ---
> >>  tools/lib/bpf/libbpf.c |   16 +++++++++++++---
> >>  1 file changed, 13 insertions(+), 3 deletions(-)
> >>
> >> diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
> >> index be4af95d5a2c..cea61b2ec9d3 100644
> >> --- a/tools/lib/bpf/libbpf.c
> >> +++ b/tools/lib/bpf/libbpf.c
> >> @@ -229,6 +229,7 @@ struct bpf_map {
> >>         enum libbpf_map_type libbpf_type;
> >>         char *pin_path;
> >>         bool pinned;
> >> +       bool was_reused;
> >
> > nit: just reused, similar to pinned?
> >
> >>  };
> >>
> >>  struct bpf_secdata {
> >> @@ -1995,6 +1996,7 @@ int bpf_map__reuse_fd(struct bpf_map *map, int fd)
> >>         map->def.map_flags = info.map_flags;
> >>         map->btf_key_type_id = info.btf_key_type_id;
> >>         map->btf_value_type_id = info.btf_value_type_id;
> >> +       map->was_reused = true;
> >>
> >>         return 0;
> >>
> >> @@ -4007,15 +4009,18 @@ bpf_object__open_buffer(const void *obj_buf, size_t obj_buf_sz,
> >>         return bpf_object__open_mem(obj_buf, obj_buf_sz, &opts);
> >>  }
> >>
> >> -int bpf_object__unload(struct bpf_object *obj)
> >> +static int __bpf_object__unload(struct bpf_object *obj, bool unpin)
> >>  {
> >>         size_t i;
> >>
> >>         if (!obj)
> >>                 return -EINVAL;
> >>
> >> -       for (i = 0; i < obj->nr_maps; i++)
> >> +       for (i = 0; i < obj->nr_maps; i++) {
> >>                 zclose(obj->maps[i].fd);
> >> +               if (unpin && obj->maps[i].pinned && !obj->maps[i].was_reused)
> >> +                       bpf_map__unpin(&obj->maps[i], NULL);
> >> +       }
> >>
> >>         for (i = 0; i < obj->nr_programs; i++)
> >>                 bpf_program__unload(&obj->programs[i]);
> >> @@ -4023,6 +4028,11 @@ int bpf_object__unload(struct bpf_object *obj)
> >>         return 0;
> >>  }
> >>
> >> +int bpf_object__unload(struct bpf_object *obj)
> >> +{
> >> +       return __bpf_object__unload(obj, false);
> >> +}
> >> +
> >>  int bpf_object__load_xattr(struct bpf_object_load_attr *attr)
> >>  {
> >>         struct bpf_object *obj;
> >> @@ -4047,7 +4057,7 @@ int bpf_object__load_xattr(struct bpf_object_load_attr *attr)
> >>
> >>         return 0;
> >>  out:
> >> -       bpf_object__unload(obj);
> >> +       __bpf_object__unload(obj, true);
> >
> > giving this is the only (special) case of auto-unpinning auto-pinned
> > maps, why not do a trivial loop here, instead of having this extra
> > unpin flag and extra __bpf_object__unload function?
>
> Oh, you mean just do a loop in addition to the call to __unload? Sure, I
> guess we can do that instead...

I think that's cleaner, because it's custom clean up logic in one
place, rather than supported feature of unload.

>
> -Toke