ap: only accept message 4 after receiving message 2

ap: only accept message 4 after receiving message 2

[Mathy Vanhoef]

We found that when IWD is operating in AP mode, it doesn't verify
whether message 2 was already received before processing message 4.

I've included a candidate commit message and patch below. My mailing
client might destroy the spacing, but the small patch should
illustrate the possible fix.

---

When operating as an AP, drop message 4 of the 4-way handshake if the AP
has not yet received message 2. Otherwise an attacker can skip message 2
and immediately sent message 4 to bypass authentication (the AP would be
using an all-zero ptk to verify the authenticity of message 4).
---
 src/eapol.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/src/eapol.c b/src/eapol.c
index bad4bbb7..3ce14d5c 100644
--- a/src/eapol.c
+++ b/src/eapol.c
@@ -2092,6 +2092,10 @@ static void eapol_handle_ptk_4_of_4(struct eapol_sm *sm,
     if (L_BE64_TO_CPU(ek->key_replay_counter) != sm->replay_counter)
         return;

+    /* Ensure we received Message 2 and thus have a PTK to verify MIC */
+    if (!sm->handshake->have_snonce)
+        return;
+
     kck = handshake_state_get_kck(sm->handshake);

     if (!eapol_verify_mic(sm->handshake->akm_suite, kck, ek,
-- 
2.43.0

Re: ap: only accept message 4 after receiving message 2

[Denis Kenzior]

Hi Mathy,

On 1/29/24 10:11, Mathy Vanhoef wrote:
> We found that when IWD is operating in AP mode, it doesn't verify
> whether message 2 was already received before processing message 4.
> 
> I've included a candidate commit message and patch below. My mailing
> client might destroy the spacing, but the small patch should
> illustrate the possible fix.

Ah nice catch.  Thank you.  I created a commit based on your patch / 
description.  Applied now as:
6415420f1c92 ("ap: only accept ptk 4/4 after receiving ptk 2/4")

Regards,
-Denis