diff mbox series

[bpf-next] bpf: btf: Fix arg verification in btf_ctx_access()

Message ID 20200330144246.338-1-kpsingh@chromium.org
State Accepted
Delegated to: BPF Maintainers
Headers show
Series [bpf-next] bpf: btf: Fix arg verification in btf_ctx_access() | expand

Commit Message

KP Singh March 30, 2020, 2:42 p.m. UTC 
From: KP Singh <kpsingh@google.com>

The bounds checking for the arguments accessed in the BPF program breaks
when the expected_attach_type is not BPF_TRACE_FEXIT, BPF_LSM_MAC or
BPF_MODIFY_RETURN resulting in no check being done for the default case
(the programs which do not receive the return value of the attached
function in its arguments) when the index of the argument being accessed
is equal to the number of arguments (nr_args).

This was a result of a misplaced "else if" block  introduced by the
Commit 6ba43b761c41 ("bpf: Attachment verification for
BPF_MODIFY_RETURN")

Signed-off-by: KP Singh <kpsingh@google.com>
Fixes: 6ba43b761c41 ("bpf: Attachment verification for BPF_MODIFY_RETURN")
Reported-by: Jann Horn <jannh@google.com>
---
 kernel/bpf/btf.c | 24 +++++++++++++++++-------
 1 file changed, 17 insertions(+), 7 deletions(-)

Comments

Alexei Starovoitov March 30, 2020, 8:32 p.m. UTC | #1 
On Mon, Mar 30, 2020 at 7:43 AM KP Singh <kpsingh@chromium.org> wrote:
>
> From: KP Singh <kpsingh@google.com>
>
> The bounds checking for the arguments accessed in the BPF program breaks
> when the expected_attach_type is not BPF_TRACE_FEXIT, BPF_LSM_MAC or
> BPF_MODIFY_RETURN resulting in no check being done for the default case
> (the programs which do not receive the return value of the attached
> function in its arguments) when the index of the argument being accessed
> is equal to the number of arguments (nr_args).
>
> This was a result of a misplaced "else if" block  introduced by the
> Commit 6ba43b761c41 ("bpf: Attachment verification for
> BPF_MODIFY_RETURN")
>
> Signed-off-by: KP Singh <kpsingh@google.com>
> Fixes: 6ba43b761c41 ("bpf: Attachment verification for BPF_MODIFY_RETURN")
> Reported-by: Jann Horn <jannh@google.com>

Applied. Thanks
diff mbox series

Patch

diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c
index de335cd386f0..3b6dcfb6ea49 100644
--- a/kernel/bpf/btf.c
+++ b/kernel/bpf/btf.c
@@ -3709,9 +3709,16 @@  bool btf_ctx_access(int off, int size, enum bpf_access_type type,
 		nr_args--;
 	}
 
+	if (arg > nr_args) {
+		bpf_log(log, "func '%s' doesn't have %d-th argument\n",
+			tname, arg + 1);
+		return false;
+	}
+
 	if (arg == nr_args) {
-		if (prog->expected_attach_type == BPF_TRACE_FEXIT ||
-		    prog->expected_attach_type == BPF_LSM_MAC) {
+		switch (prog->expected_attach_type) {
+		case BPF_LSM_MAC:
+		case BPF_TRACE_FEXIT:
 			/* When LSM programs are attached to void LSM hooks
 			 * they use FEXIT trampolines and when attached to
 			 * int LSM hooks, they use MODIFY_RETURN trampolines.
@@ -3728,7 +3735,8 @@  bool btf_ctx_access(int off, int size, enum bpf_access_type type,
 			if (!t)
 				return true;
 			t = btf_type_by_id(btf, t->type);
-		} else if (prog->expected_attach_type == BPF_MODIFY_RETURN) {
+			break;
+		case BPF_MODIFY_RETURN:
 			/* For now the BPF_MODIFY_RETURN can only be attached to
 			 * functions that return an int.
 			 */
@@ -3742,17 +3750,19 @@  bool btf_ctx_access(int off, int size, enum bpf_access_type type,
 					btf_kind_str[BTF_INFO_KIND(t->info)]);
 				return false;
 			}
+			break;
+		default:
+			bpf_log(log, "func '%s' doesn't have %d-th argument\n",
+				tname, arg + 1);
+			return false;
 		}
-	} else if (arg >= nr_args) {
-		bpf_log(log, "func '%s' doesn't have %d-th argument\n",
-			tname, arg + 1);
-		return false;
 	} else {
 		if (!t)
 			/* Default prog with 5 args */
 			return true;
 		t = btf_type_by_id(btf, args[arg].type);
 	}
+
 	/* skip modifiers */
 	while (btf_type_is_modifier(t))
 		t = btf_type_by_id(btf, t->type);