On Thu, Oct 31, 2019 at 10:37 AM Andrii Nakryiko
<andrii.nakryiko@xxxxxxxxx> wrote:
>
> On Tue, Oct 29, 2019 at 12:39 PM Toke Høiland-Jørgensen <toke@xxxxxxxxxx> wrote:
> >
> > From: Toke Høiland-Jørgensen <toke@xxxxxxxxxx>
> >
> > This adds support to libbpf for setting map pinning information as part of
> > the BTF map declaration, to get automatic map pinning (and reuse) on load.
> > The pinning type currently only supports a single PIN_BY_NAME mode, where
> > each map will be pinned by its name in a path that can be overridden, but
> > defaults to /sys/fs/bpf.
> >
> > Since auto-pinning only does something if any maps actually have a
> > 'pinning' BTF attribute set, we default the new option to enabled, on the
> > assumption that seamless pinning is what most callers want.
> >
> > When a map has a pin_path set at load time, libbpf will compare the map
> > pinned at that location (if any), and if the attributes match, will re-use
> > that map instead of creating a new one. If no existing map is found, the
> > newly created map will instead be pinned at the location.
> >
> > Programs wanting to customise the pinning can override the pinning paths
> > using bpf_map__set_pin_path() before calling bpf_object__load() (including
> > setting it to NULL to disable pinning of a particular map).
> >
> > Signed-off-by: Toke Høiland-Jørgensen <toke@xxxxxxxxxx>
> > ---
>
> Please fix unconditional pin_path setting, with that:
>
> Acked-by: Andrii Nakryiko <andriin@xxxxxx>
>
> > tools/lib/bpf/bpf_helpers.h | 6 ++
> > tools/lib/bpf/libbpf.c | 144 +++++++++++++++++++++++++++++++++++++++++--
> > tools/lib/bpf/libbpf.h | 13 ++++
> > 3 files changed, 154 insertions(+), 9 deletions(-)
> >
> > diff --git a/tools/lib/bpf/bpf_helpers.h b/tools/lib/bpf/bpf_helpers.h
> > index 2203595f38c3..0c7d28292898 100644
> > --- a/tools/lib/bpf/bpf_helpers.h
> > +++ b/tools/lib/bpf/bpf_helpers.h
> > @@ -38,4 +38,10 @@ struct bpf_map_def {
> > unsigned int map_flags;
> > };
> >
>
> [...]
>
> > @@ -1270,6 +1292,28 @@ static int bpf_object__init_user_btf_map(struct bpf_object *obj,
> > }
> > map->def.value_size = sz;
> > map->btf_value_type_id = t->type;
> > + } else if (strcmp(name, "pinning") == 0) {
> > + __u32 val;
> > + int err;
> > +
> > + if (!get_map_field_int(map_name, obj->btf, def, m,
> > + &val))
> > + return -EINVAL;
> > + pr_debug("map '%s': found pinning = %u.\n",
> > + map_name, val);
> > +
> > + if (val != LIBBPF_PIN_NONE &&
> > + val != LIBBPF_PIN_BY_NAME) {
> > + pr_warn("map '%s': invalid pinning value %u.\n",
> > + map_name, val);
> > + return -EINVAL;
> > + }
> > + err = build_map_pin_path(map, pin_root_path);
>
> uhm... only if (val == LIBBPF_PIN_BY_NAME)?.. maybe extend tests with
> a mix if auto-pinned and never pinned map to catch issue like this?
I was wondering why your selftest didn't catch this, got puzzled for a
bit. It's because this code path will be executed only when map
defintion has __uint(pinning, LIBBPF_PIN_NONE), can you please add
that to selftest as well?
>
> > + if (err) {
> > + pr_warn("map '%s': couldn't build pin path.\n",
> > + map_name);
> > + return err;
> > + }
> > } else {
> > if (strict) {
> > pr_warn("map '%s': unknown field '%s'.\n",
> > @@ -1289,7 +1333,8 @@ static int bpf_object__init_user_btf_map(struct bpf_object *obj,
> > return 0;
> > }
> >
>
> [...]
