| Message ID | 9e77ae06243555a96a3fd5e854f61d24823110c9.1565040372.git.luto@kernel.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | bpf: A bit of progress toward unprivileged use | expand |
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 23f8f89d2a86..730afa2be786 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1649,8 +1649,7 @@ static int bpf_prog_load(union bpf_attr *attr, union bpf_attr __user *uattr) attr->insn_cnt > (capable(CAP_SYS_ADMIN) ? BPF_COMPLEXITY_LIMIT_INSNS : BPF_MAXINSNS)) return -E2BIG; if (type != BPF_PROG_TYPE_SOCKET_FILTER && - type != BPF_PROG_TYPE_CGROUP_SKB && - !capable(CAP_SYS_ADMIN)) + type != BPF_PROG_TYPE_CGROUP_SKB) return -EPERM; bpf_prog_load_fixup_attach_type(attr);
This doesn't let you *run* the programs except in test mode, so it should be safe. Famous last words. This assumes that the check-privilege-to-call-privileged-functions patch actually catches all the cases and that there's nothing else that should need privilege lurking in the type-specific verifiers. Signed-off-by: Andy Lutomirski <luto@kernel.org> --- kernel/bpf/syscall.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)