2018-09-28 13:35:13

by Roman Gushchin

Subject: [PATCH bpf] bpf: harden flags check in cgroup_storage_update_elem()

cgroup_storage_update_elem() shouldn't accept any flags
argument values except BPF_ANY and BPF_EXIST to guarantee
the backward compatibility, had a new flag value been added.

Fixes: de9cbbaadba5 ("bpf: introduce cgroup storage maps")
Signed-off-by: Roman Gushchin <[email protected]>
Cc: Alexei Starovoitov <[email protected]>
Cc: Daniel Borkmann <[email protected]>
---
kernel/bpf/local_storage.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/bpf/local_storage.c b/kernel/bpf/local_storage.c
index f8746e6a9a35..2a7680106320 100644
--- a/kernel/bpf/local_storage.c
+++ b/kernel/bpf/local_storage.c
@@ -130,7 +130,7 @@ static int cgroup_storage_update_elem(struct bpf_map *map, void *_key,
struct bpf_cgroup_storage *storage;
struct bpf_storage_buffer *new;

- if (flags & BPF_NOEXIST)
+ if (flags != BPF_ANY && flags != BPF_EXIST)
return -EINVAL;

storage = cgroup_storage_lookup((struct bpf_cgroup_storage_map *)map,
--
2.17.1

2018-09-28 13:53:54

by Daniel Borkmann

Subject: Re: [PATCH bpf] bpf: harden flags check in cgroup_storage_update_elem()

On 09/28/2018 03:33 PM, Roman Gushchin wrote:
> cgroup_storage_update_elem() shouldn't accept any flags
> argument values except BPF_ANY and BPF_EXIST to guarantee
> the backward compatibility, had a new flag value been added.
>
> Fixes: de9cbbaadba5 ("bpf: introduce cgroup storage maps")
> Signed-off-by: Roman Gushchin <[email protected]>
> Cc: Alexei Starovoitov <[email protected]>
> Cc: Daniel Borkmann <[email protected]>
> ---
> kernel/bpf/local_storage.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/kernel/bpf/local_storage.c b/kernel/bpf/local_storage.c
> index f8746e6a9a35..2a7680106320 100644
> --- a/kernel/bpf/local_storage.c
> +++ b/kernel/bpf/local_storage.c
> @@ -130,7 +130,7 @@ static int cgroup_storage_update_elem(struct bpf_map *map, void *_key,
> struct bpf_cgroup_storage *storage;
> struct bpf_storage_buffer *new;
>
> - if (flags & BPF_NOEXIST)
> + if (flags != BPF_ANY && flags != BPF_EXIST)
> return -EINVAL;
>
> storage = cgroup_storage_lookup((struct bpf_cgroup_storage_map *)map,
>

Applied to bpf, thanks!