Re: [PATCH bpf-next 1/5] bpf: add memlock precharge check for cgroup_local_storage

[Song Liu <liu.song.a23@gmail.com>]

On Wed, May 29, 2019 at 6:05 PM Roman Gushchin <guro@fb.com> wrote:
>
> Cgroup local storage maps lack the memlock precharge check,
> which is performed before the memory allocation for
> most other bpf map types.
>
> Let's add it in order to unify all map types.
>
> Signed-off-by: Roman Gushchin <guro@fb.com>

Acked-by: Song Liu <songliubraving@fb.com>

> ---
>  kernel/bpf/local_storage.c | 11 +++++++++--
>  1 file changed, 9 insertions(+), 2 deletions(-)
>
> diff --git a/kernel/bpf/local_storage.c b/kernel/bpf/local_storage.c
> index 980e8f1f6cb5..e48302ecb389 100644
> --- a/kernel/bpf/local_storage.c
> +++ b/kernel/bpf/local_storage.c
> @@ -272,6 +272,8 @@ static struct bpf_map *cgroup_storage_map_alloc(union bpf_attr *attr)
>  {
>         int numa_node = bpf_map_attr_numa_node(attr);
>         struct bpf_cgroup_storage_map *map;
> +       u32 pages;
> +       int ret;
>
>         if (attr->key_size != sizeof(struct bpf_cgroup_storage_key))
>                 return ERR_PTR(-EINVAL);
> @@ -290,13 +292,18 @@ static struct bpf_map *cgroup_storage_map_alloc(union bpf_attr *attr)
>                 /* max_entries is not used and enforced to be 0 */
>                 return ERR_PTR(-EINVAL);
>
> +       pages = round_up(sizeof(struct bpf_cgroup_storage_map), PAGE_SIZE) >>
> +               PAGE_SHIFT;
> +       ret = bpf_map_precharge_memlock(pages);
> +       if (ret < 0)
> +               return ERR_PTR(ret);
> +
>         map = kmalloc_node(sizeof(struct bpf_cgroup_storage_map),
>                            __GFP_ZERO | GFP_USER, numa_node);
>         if (!map)
>                 return ERR_PTR(-ENOMEM);
>
> -       map->map.pages = round_up(sizeof(struct bpf_cgroup_storage_map),
> -                                 PAGE_SIZE) >> PAGE_SHIFT;
> +       map->map.pages = pages;
>
>         /* copy mandatory map attributes */
>         bpf_map_init_from_attr(&map->map, attr);
> --
> 2.20.1
>