bpf: consider that tail calls invalidate packet pointers

Tail-called programs could execute any of the helpers that invalidate
packet pointers. Hence, conservatively assume that each tail call
invalidates packet pointers.

Making the change in bpf_helper_changes_pkt_data() automatically makes
use of check_cfg() logic that computes 'changes_pkt_data' effect for
global sub-programs, such that the following program could be
rejected:

    int tail_call(struct __sk_buff *sk)
    {
    	bpf_tail_call_static(sk, &jmp_table, 0);
    	return 0;
    }

    SEC("tc")
    int not_safe(struct __sk_buff *sk)
    {
    	int *p = (void *)(long)sk->data;
    	... make p valid ...
    	tail_call(sk);
    	*p = 42; /* this is unsafe */
    	...
    }

The tc_bpf2bpf.c:subprog_tc() needs change: mark it as a function that
can invalidate packet pointers. Otherwise, it can't be freplaced with
tailcall_freplace.c:entry_freplace() that does a tail call.

Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20241210041100.1898468-8-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

1 files changed, 2 insertions, 0 deletions

diff --git a/net/core/filter.c b/net/core/filter.c
index efb75eed2e35..21131ec25f24 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -7924,6 +7924,8 @@ bool bpf_helper_changes_pkt_data(enum bpf_func_id func_id)
 	case BPF_FUNC_xdp_adjust_head:
 	case BPF_FUNC_xdp_adjust_meta:
 	case BPF_FUNC_xdp_adjust_tail:
+	/* tail-called program could call any of the above */
+	case BPF_FUNC_tail_call:
 		return true;
 	default:
 		return false;