Commit 81f6d053 authored by Eduard Zingerman's avatar Eduard Zingerman Committed by Alexei Starovoitov
Browse files

bpf: check changes_pkt_data property for extension programs



When processing calls to global sub-programs, verifier decides whether
to invalidate all packet pointers in current state depending on the
changes_pkt_data property of the global sub-program.

Because of this, an extension program replacing a global sub-program
must be compatible with changes_pkt_data property of the sub-program
being replaced.

This commit:
- adds changes_pkt_data flag to struct bpf_prog_aux:
  - this flag is set in check_cfg() for main sub-program;
  - in jit_subprogs() for other sub-programs;
- modifies bpf_check_attach_btf_id() to check changes_pkt_data flag;
- moves call to check_attach_btf_id() after the call to check_cfg(),
  because it needs changes_pkt_data flag to be set:

    bpf_check:
      ...                             ...
    - check_attach_btf_id             resolve_pseudo_ldimm64
      resolve_pseudo_ldimm64   -->    bpf_prog_is_offloaded
      bpf_prog_is_offloaded           check_cfg
      check_cfg                     + check_attach_btf_id
      ...                             ...

The following fields are set by check_attach_btf_id():
- env->ops
- prog->aux->attach_btf_trace
- prog->aux->attach_func_name
- prog->aux->attach_func_proto
- prog->aux->dst_trampoline
- prog->aux->mod
- prog->aux->saved_dst_attach_type
- prog->aux->saved_dst_prog_type
- prog->expected_attach_type

Neither of these fields are used by resolve_pseudo_ldimm64() or
bpf_prog_offload_verifier_prep() (for netronome and netdevsim
drivers), so the reordering is safe.

Suggested-by: default avatarAlexei Starovoitov <alexei.starovoitov@gmail.com>
Signed-off-by: default avatarEduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20241210041100.1898468-6-eddyz87@gmail.com


Signed-off-by: default avatarAlexei Starovoitov <ast@kernel.org>
parent 3f23ee55
Loading
Loading
Loading
Loading
+1 −0
Original line number Diff line number Diff line
@@ -1527,6 +1527,7 @@ struct bpf_prog_aux {
	bool is_extended; /* true if extended by freplace program */
	bool jits_use_priv_stack;
	bool priv_stack_requested;
	bool changes_pkt_data;
	u64 prog_array_member_cnt; /* counts how many times as member of prog_array */
	struct mutex ext_mutex; /* mutex for is_extended and prog_array_member_cnt */
	struct bpf_arena *arena;
+12 −4
Original line number Diff line number Diff line
@@ -16872,6 +16872,7 @@ static int check_cfg(struct bpf_verifier_env *env)
		}
	}
	ret = 0; /* cfg looks good */
	env->prog->aux->changes_pkt_data = env->subprog_info[0].changes_pkt_data;
err_free:
	kvfree(insn_state);
@@ -20361,6 +20362,7 @@ static int jit_subprogs(struct bpf_verifier_env *env)
		func[i]->aux->num_exentries = num_exentries;
		func[i]->aux->tail_call_reachable = env->subprog_info[i].tail_call_reachable;
		func[i]->aux->exception_cb = env->subprog_info[i].is_exception_cb;
		func[i]->aux->changes_pkt_data = env->subprog_info[i].changes_pkt_data;
		if (!i)
			func[i]->aux->exception_boundary = env->seen_exception;
		func[i] = bpf_int_jit_compile(func[i]);
@@ -22225,6 +22227,12 @@ int bpf_check_attach_target(struct bpf_verifier_log *log,
					"Extension programs should be JITed\n");
				return -EINVAL;
			}
			if (prog->aux->changes_pkt_data &&
			    !aux->func[subprog]->aux->changes_pkt_data) {
				bpf_log(log,
					"Extension program changes packet data, while original does not\n");
				return -EINVAL;
			}
		}
		if (!tgt_prog->jited) {
			bpf_log(log, "Can attach to only JITed progs\n");
@@ -22690,10 +22698,6 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, bpfptr_t uattr, __u3
	if (ret < 0)
		goto skip_full_check;
	ret = check_attach_btf_id(env);
	if (ret)
		goto skip_full_check;
	ret = resolve_pseudo_ldimm64(env);
	if (ret < 0)
		goto skip_full_check;
@@ -22708,6 +22712,10 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, bpfptr_t uattr, __u3
	if (ret < 0)
		goto skip_full_check;
	ret = check_attach_btf_id(env);
	if (ret)
		goto skip_full_check;
	ret = mark_fastcall_patterns(env);
	if (ret < 0)
		goto skip_full_check;