bpf: Validate BPF object in BPF_OBJ_PIN before calling LSM

Do a sanity check whether provided file-to-be-pinned is actually a BPF
object (prog, map, btf) before calling security_path_mknod LSM hook. If
it's not, LSM hook doesn't have to be triggered, as the operation has no
chance of succeeding anyways.

Suggested-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Christian Brauner <brauner@kernel.org>
Link: https://lore.kernel.org/bpf/20230522232917.2454595-2-andrii@kernel.org

diff --git a/kernel/bpf/inode.c b/kernel/bpf/inode.c
index 9948b542a470ebfa5a7f3bbc7f15d57d63f50678..329f27d5cacfca29462fbd559fd8e150f505d614 100644 (file)
--- a/kernel/bpf/inode.c
+++ b/kernel/bpf/inode.c
@@ -448,18 +448,17 @@ static int bpf_obj_do_pin(const char __user *pathname, void *raw,
        if (IS_ERR(dentry))
                return PTR_ERR(dentry);
 
-       mode = S_IFREG | ((S_IRUSR | S_IWUSR) & ~current_umask());
-
-       ret = security_path_mknod(&path, dentry, mode, 0);
-       if (ret)
-               goto out;
-
        dir = d_inode(path.dentry);
        if (dir->i_op != &bpf_dir_iops) {
                ret = -EPERM;
                goto out;
        }
 
+       mode = S_IFREG | ((S_IRUSR | S_IWUSR) & ~current_umask());
+       ret = security_path_mknod(&path, dentry, mode, 0);
+       if (ret)
+               goto out;
+
        switch (type) {
        case BPF_TYPE_PROG:
                ret = vfs_mkobj(dentry, mode, bpf_mkprog, raw);