Message-ID: <6c94bf74-594c-4e20-958e-057a80518e53@...2889...> Date: Sat, 23 Aug 2025 23:28:35 +0200 MIME-Version: 1.0 References: <5c374002a1dfb9b99a32f5715d21121d05d7912a-HTML@...2715...> Content-Language: en-US, sv-SE, nl From: Walter Doekes In-Reply-To: <5c374002a1dfb9b99a32f5715d21121d05d7912a-HTML@...2715...> Subject: Re: [Openvpn-devel] [M] Change in openvpn[master]: Check message id/acked ids too when doing sessionid cookie checks List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit To: openvpn-devel@lists.sourceforge.net Thanks! Can we get this backported to 2.6 as well? Walter On 23-08-2025 18:01, cron2 (Code Review) wrote: > cron2 *submitted* this change. > > View Change > > Check message id/acked ids too when doing sessionid cookie checks > > This fixes that control packets on a floating client can trigger > creating a new session in special circumstances: > > To trigger this circumstance a connection needs to > > - starts on IP A > - successfully floats to IP B by data packet > - then has a control packet from IP A before any > data packet can trigger the float back to IP A > > and all of this needs to happen in the 60s time > that hmac cookie is valid in the default > configuration. > > In this scenario we would trigger a new connection as the HMAC > session id would be valid. > > This patch adds checking also of the message-id and acked ids to > discern packet from the initial three-way handshake where these > ids are 0 or 1 from any later packet. > > This will now trigger (at verb 4 or higher) a messaged like: > > Packet (P_ACK_V1) with invalid or missing SID > > instead. > > Also remove a few duplicated free_tls_pre_decrypt_state in test_ssl. > > Reported-By: Walter Doekes > Tested-By: Walter Doekes > > Change-Id: I6752dcd5aff3e5cea2b439366479e86751a1c403 > Signed-off-by: Arne Schwabe > Acked-by: MaxF > Message-Id: <20250819212214.16218-1-gert@...1296...> > URL: https://www.mail-archive.com/openvpn-devel@lists.sourceforge.net/msg32626.html > Signed-off-by: Gert Doering > --- > M src/openvpn/mudp.c > M src/openvpn/ssl_pkt.c > M src/openvpn/ssl_pkt.h > M tests/unit_tests/openvpn/test_pkt.c > 4 files changed, 129 insertions(+), 24 deletions(-) > > diff --git a/src/openvpn/mudp.c b/src/openvpn/mudp.c > index 7259a4b..31134be 100644 > --- a/src/openvpn/mudp.c > +++ b/src/openvpn/mudp.c > @@ -151,7 +151,8 @@ > * need to contain the peer id */ > struct gc_arena gc = gc_new(); > > - bool ret = check_session_id_hmac(state, from, hmac, handwindow); > + bool pkt_is_ack = (verdict == VERDICT_VALID_ACK_V1); > + bool ret = check_session_hmac_and_pkt_id(state, from, hmac, > handwindow, pkt_is_ack); > > const char *peer = print_link_socket_actual(&m->top.c2.from, &gc); > uint8_t pkt_firstbyte = *BPTR(&m->top.c2.buf); > @@ -159,7 +160,8 @@ > > if (!ret) > { > - msg(D_MULTI_MEDIUM, "Packet (%s) with invalid or missing SID from %s", > + msg(D_MULTI_MEDIUM, "Packet (%s) with invalid or missing SID from" > + " %s or wrong packet id", > packet_opcode_name(op), peer); > } > else > diff --git a/src/openvpn/ssl_pkt.c b/src/openvpn/ssl_pkt.c > index b901f87..6ec05a7 100644 > --- a/src/openvpn/ssl_pkt.c > +++ b/src/openvpn/ssl_pkt.c > @@ -496,8 +496,11 @@ > } > > bool > -check_session_id_hmac(struct tls_pre_decrypt_state *state, const struct > openvpn_sockaddr *from, > - hmac_ctx_t *hmac, int handwindow) > +check_session_hmac_and_pkt_id(struct tls_pre_decrypt_state *state, > + const struct openvpn_sockaddr *from, > + hmac_ctx_t *hmac, > + int handwindow, > + bool pkt_is_ack) > { > if (!from) > { > @@ -512,6 +515,36 @@ > return false; > } > > + /* Check if the packet ID of the packet or ACKED packet is <= 1 */ > + for (int i = 0; i < ack.len; i++) > + { > + /* This packet ACKs a packet that has a higher packet id than the > + * ones expected in the three-way handshake, consider it as invalid > + * for the session */ > + if (ack.packet_id[i] > 1) > + { > + return false; > + } > + } > + > + if (!pkt_is_ack) > + { > + packet_id_type message_id; > + /* Extract the packet ID from the packet */ > + if (!reliable_ack_read_packet_id(&buf, &message_id)) > + { > + return false; > + } > + > + /* similar check. Anything larger than 1 is not considered part of the > + * three-way handshake */ > + if (message_id > 1) > + { > + return false; > + } > + } > + > + > /* check adjacent timestamps too */ > for (int offset = -2; offset <= 1; offset++) > { > diff --git a/src/openvpn/ssl_pkt.h b/src/openvpn/ssl_pkt.h > index 8fe4880..96cdd68 100644 > --- a/src/openvpn/ssl_pkt.h > +++ b/src/openvpn/ssl_pkt.h > @@ -178,14 +178,20 @@ > /** > * Checks if a control packet has a correct HMAC server session id > * > + * This will also consider packets that have a packet id higher > + * than 1 or ack packets higher than 1 to be invalid as they are > + * not part of the initial three way handshake of OpenVPN and should > + * not create a new connection. > + * > * @param state session information > * @param from link_socket from the client > * @param hmac the hmac context to use for the calculation > * @param handwindow the quantisation of the current time > + * @param pkt_is_ack the packet being checked is a P_ACK_V1 > * @return the expected server session id > */ > -bool check_session_id_hmac(struct tls_pre_decrypt_state *state, const > struct openvpn_sockaddr *from, > - hmac_ctx_t *hmac, int handwindow); > +bool check_session_hmac_and_pkt_id(struct tls_pre_decrypt_state *state, > const struct openvpn_sockaddr *from, > + hmac_ctx_t *hmac, int handwindow, bool pkt_is_ack); > > /* > * Write a control channel authentication record. > diff --git a/tests/unit_tests/openvpn/test_pkt.c b/tests/unit_tests/ > openvpn/test_pkt.c > index 65b31e7..5122766 100644 > --- a/tests/unit_tests/openvpn/test_pkt.c > +++ b/tests/unit_tests/openvpn/test_pkt.c > @@ -139,6 +139,27 @@ > 0xc8, 0x01, 0x00, 0x00, 0x00, 0x00, 0xdd, 0x85, > 0xdb, 0x53, 0x56, 0x23, 0xb0, 0x2e }; > > +/* no tls-auth, P_ACK_V1, acks 0,1, and 2 */ > +const uint8_t client_ack_123_none_random_id[] = { > + 0x28, > + 0xae, 0xb9, 0xaf, 0xe1, 0xf0, 0x1d, 0x79, 0xc8, > + 0x03, > + 0x00, 0x00, 0x00, 0x00, > + 0x00, 0x00, 0x00, 0x01, > + 0x00, 0x00, 0x00, 0x02, > + 0xdd, 0x85, 0xdb, 0x53, 0x56, 0x23, 0xb0, 0x2e > +}; > + > +/* no tls-auth, P_CONTROL_V1, acks 0, msg-id 2 */ > +const uint8_t client_control_none_random_id[] = { > + 0x20, > + 0xae, 0xb9, 0xaf, 0xe1, 0xf0, 0x1d, 0x79, 0xc8, > + 0x01, > + 0x00, 0x00, 0x00, 0x00, > + 0x02 > +}; > + > + > struct tls_auth_standalone > init_tas_auth(int key_direction) > { > @@ -256,12 +277,10 @@ > assert_int_equal(verdict, VERDICT_VALID_RESET_V2); > free_tls_pre_decrypt_state(&state); > > - free_tls_pre_decrypt_state(&state); > /* The pre decrypt function should not modify the buffer, so calling it > * again should have the same result */ > verdict = tls_pre_decrypt_lite(&tas, &state, &from, &buf); > assert_int_equal(verdict, VERDICT_VALID_RESET_V2); > - free_tls_pre_decrypt_state(&state); > > /* and buf memory should be equal */ > assert_memory_equal(BPTR(&buf), client_reset_v2_tls_auth, > sizeof(client_reset_v2_tls_auth)); > @@ -279,7 +298,6 @@ > assert_int_equal(verdict, VERDICT_INVALID); > free_tls_pre_decrypt_state(&state); > > - free_tls_pre_decrypt_state(&state); > /* Wrong key direction gives a wrong hmac key and should not validate */ > free_key_ctx_bi(&tas.tls_wrap.opt.key_ctx_bi); > free_tas(&tas); > @@ -319,15 +337,12 @@ > assert_int_equal(verdict, VERDICT_VALID_RESET_V2); > free_tls_pre_decrypt_state(&state); > > - free_tls_pre_decrypt_state(&state); > buf_reset_len(&buf); > buf_write(&buf, client_reset_v2_tls_crypt, sizeof(client_reset_v2_none)); > verdict = tls_pre_decrypt_lite(&tas, &state, &from, &buf); > assert_int_equal(verdict, VERDICT_VALID_RESET_V2); > free_tls_pre_decrypt_state(&state); > > - free_tls_pre_decrypt_state(&state); > - > /* This is not a reset packet and should trigger the other response */ > buf_reset_len(&buf); > buf_write(&buf, client_ack_tls_auth_randomid, > sizeof(client_ack_tls_auth_randomid)); > @@ -405,7 +420,7 @@ > assert_int_equal(verdict, VERDICT_VALID_CONTROL_V1); > > /* This is a valid packet but containing a random id instead of an HMAC id*/ > - bool valid = check_session_id_hmac(&state, &from.dest, hmac, 30); > + bool valid = check_session_hmac_and_pkt_id(&state, &from.dest, hmac, > 30, false); > assert_false(valid); > > free_tls_pre_decrypt_state(&state); > @@ -436,7 +451,7 @@ > verdict = tls_pre_decrypt_lite(&tas, &state, &from, &buf); > assert_int_equal(verdict, VERDICT_VALID_ACK_V1); > > - bool valid = check_session_id_hmac(&state, &from.dest, hmac, 30); > + bool valid = check_session_hmac_and_pkt_id(&state, &from.dest, hmac, > 30, true); > assert_true(valid); > > free_tls_pre_decrypt_state(&state); > @@ -445,6 +460,51 @@ > hmac_ctx_free(hmac); > } > > +static void > +test_verify_hmac_none_out_of_range_ack(void **ut_state) > +{ > + hmac_ctx_t *hmac = session_id_hmac_init(); > + > + struct link_socket_actual from = { 0 }; > + from.dest.addr.sa.sa_family = AF_INET; > + > + struct tls_auth_standalone tas = { 0 }; > + struct tls_pre_decrypt_state state = { 0 }; > + > + struct buffer buf = alloc_buf(1024); > + enum first_packet_verdict verdict; > + > + tas.tls_wrap.mode = TLS_WRAP_NONE; > + > + buf_reset_len(&buf); > + buf_write(&buf, client_ack_123_none_random_id, > sizeof(client_ack_123_none_random_id)); > + > + > + verdict = tls_pre_decrypt_lite(&tas, &state, &from, &buf); > + assert_int_equal(verdict, VERDICT_VALID_ACK_V1); > + > + /* should fail because it acks 2 */ > + bool valid = check_session_hmac_and_pkt_id(&state, &from.dest, hmac, > 30, true); > + assert_false(valid); > + free_tls_pre_decrypt_state(&state); > + > + /* Try test with the control with a too high message id now */ > + buf_reset_len(&buf); > + buf_write(&buf, client_control_none_random_id, > sizeof(client_control_none_random_id)); > + > + verdict = tls_pre_decrypt_lite(&tas, &state, &from, &buf); > + assert_int_equal(verdict, VERDICT_VALID_CONTROL_V1); > + > + /* should fail because it has message id 2 */ > + valid = check_session_hmac_and_pkt_id(&state, &from.dest, hmac, 30, true); > + assert_false(valid); > + > + free_tls_pre_decrypt_state(&state); > + free_buf(&buf); > + hmac_ctx_cleanup(hmac); > + hmac_ctx_free(hmac); > +} > + > static hmac_ctx_t * > init_static_hmac(void) > { > @@ -634,16 +694,20 @@ > main(void) > { > openvpn_unit_test_setup(); > - const struct CMUnitTest tests[] = > { cmocka_unit_test(test_tls_decrypt_lite_none), > - cmocka_unit_test(test_tls_decrypt_lite_auth), > - cmocka_unit_test(test_tls_decrypt_lite_crypt), > - cmocka_unit_test(test_parse_ack), > - cmocka_unit_test(test_calc_session_id_hmac_static), > - cmocka_unit_test(test_verify_hmac_none), > - cmocka_unit_test(test_verify_hmac_tls_auth), > - cmocka_unit_test(test_generate_reset_packet_plain), > - cmocka_unit_test(test_generate_reset_packet_tls_auth), > - cmocka_unit_test(test_extract_control_message) }; > + > + const struct CMUnitTest tests[] = { > + cmocka_unit_test(test_tls_decrypt_lite_none), > + cmocka_unit_test(test_tls_decrypt_lite_auth), > + cmocka_unit_test(test_tls_decrypt_lite_crypt), > + cmocka_unit_test(test_parse_ack), > + cmocka_unit_test(test_calc_session_id_hmac_static), > + cmocka_unit_test(test_verify_hmac_none), > + cmocka_unit_test(test_verify_hmac_tls_auth), > + cmocka_unit_test(test_verify_hmac_none_out_of_range_ack), > + cmocka_unit_test(test_generate_reset_packet_plain), > + cmocka_unit_test(test_generate_reset_packet_tls_auth), > + cmocka_unit_test(test_extract_control_message) > + }; > > #if defined(ENABLE_CRYPTO_OPENSSL) > OpenSSL_add_all_algorithms(); > > To view, visit change 1067 usp=email>. To unsubscribe, or for help writing mail filters, visit > settings . > > Gerrit-Project: openvpn > Gerrit-Branch: master > Gerrit-Change-Id: I6752dcd5aff3e5cea2b439366479e86751a1c403 > Gerrit-Change-Number: 1067 > Gerrit-PatchSet: 7 > Gerrit-Owner: plaisthos > Gerrit-Reviewer: MaxF > Gerrit-Reviewer: flichtenheld > Gerrit-CC: openvpn-devel > Gerrit-MessageType: merged > > > _______________________________________________ > Openvpn-devel mailing list > Openvpn-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/openvpn-devel