Commit graph

13187 commits

Author SHA1 Message Date
Saurav Babu
e360010c3a dbus: Add mesh to global capabilities
Signed-off-by: Saurav Babu <saurav.babu@samsung.com>
2018-04-01 21:52:56 +03:00
Mathy Vanhoef
9f917339cc tests: Fix clearing of IGTK PN in handshake retransmit test functions
When locating the position of the IGTK PN in the key data, we also need
to skip the KDE header, in addition to the keyid field. This fixes
hostapd RESEND_M3 and RESEND_GROUP_M1 behavior when PMF is negotiated
for the association. Previously, the IGTK KDE ended up getting
practically hidden since zeroing of the PN ended up clearing the KDE OUI
and Type fields.

Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
2018-04-01 21:47:16 +03:00
Jouni Malinen
20daa57b92 tests: RESEND_M3 and RESEND_GROUP_M1 with PMF in use
This covers an additional code path that has a bug in IGTK PN clearing.

Signed-off-by: Jouni Malinen <j@w1.fi>
2018-04-01 21:42:49 +03:00
Jouni Malinen
5a30fcf573 tests: EAP-ERROR-CODE
Signed-off-by: Jouni Malinen <j@w1.fi>
2018-03-31 12:02:25 +03:00
Ahmed ElArabawy
45f7574d31 Propagate the EAP method error code
In the current implementation, upon an EAP method failure, followed by
an EAP failure, the EAP Status is propagated up in wpa_supplicant with a
general failure parameter string "failure". This parameter is used for a
notification on the dbus.

This commit reports the EAP method failure error code in a separate
callback.

The solution in this commit is generic to all EAP methods, and can be
used by any method that need to pass its error code. However, this
commit only implements the reporting for EAP-SIM and EAP-AKA methods
where the Notification Code (in AT_NOTIFICATION) is used as the method
specific error code value.

Signed-off-by: Ahmed ElArabawy <arabawy@google.com>
2018-03-31 11:57:33 +03:00
Jouni Malinen
c9aba19bd3 tests: WPA2-Enterprise connection using EAP-PSK after MAC address change
This is a regression test for a sequence where wpa_supplicant interface
MAC address is changed externally and the ifdown-ifup sequence is
processed only after the interface has already been set UP.

Signed-off-by: Jouni Malinen <j@w1.fi>
2018-03-30 12:16:49 +03:00
Beniamino Galvani
77a020a118 wpa_supplicant: Fix auth failure when the MAC is updated externally
When connecting to a WPA-EAP network and the MAC address is changed
just before the association (for example by NetworkManager, which sets
a random MAC during scans), the authentication sometimes fails in the
following way ('####' logs added by me):

wpa_supplicant logs:
 wlan0: WPA: RX message 1 of 4-Way Handshake from 02:00:00:00:01:00 (ver=1)
 RSN: msg 1/4 key data - hexdump(len=22): dd 14 00 0f ac 04 d8 21 9d a5 73 98 88 26 ef 03 d2 ce f7 04 7d 23
 WPA: PMKID in EAPOL-Key - hexdump(len=22): dd 14 00 0f ac 04 d8 21 9d a5 73 98 88 26 ef 03 d2 ce f7 04 7d 23
 RSN: PMKID from Authenticator - hexdump(len=16): d8 21 9d a5 73 98 88 26 ef 03 d2 ce f7 04 7d 23
 wlan0: RSN: no matching PMKID found
 EAPOL: Successfully fetched key (len=32)
 WPA: PMK from EAPOL state machines - hexdump(len=32): [REMOVED]
 #### WPA: rsn_pmkid():
 #### WPA: aa              - hexdump(len=6): 02 00 00 00 01 00
 #### WPA: spa             - hexdump(len=6): 66 20 cf ab 8c dc
 #### WPA: PMK             - hexdump(len=32): b5 24 76 4f 6f 50 8c f6 a1 2e 24 b8 07 4e 9a 13 1b 94 c4 a8 1f 7e 22 d6 ed fc 7d 43 c7 77 b6 f7
 #### WPA: computed PMKID  - hexdump(len=16): ea 73 67 b1 8e 5f 18 43 58 24 e8 1c 47 23 87 71
 RSN: Replace PMKSA entry for the current AP and any PMKSA cache entry that was based on the old PMK
 nl80211: Delete PMKID for 02:00:00:00:01:00
 wlan0: RSN: PMKSA cache entry free_cb: 02:00:00:00:01:00 reason=1
 RSN: Added PMKSA cache entry for 02:00:00:00:01:00 network_ctx=0x5630bf85a270
 nl80211: Add PMKID for 02:00:00:00:01:00
 wlan0: RSN: PMKID mismatch - authentication server may have derived different MSK?!

hostapd logs:
 WPA: PMK from EAPOL state machine (MSK len=64 PMK len=32)
 WPA: 02:00:00:00:00:00 WPA_PTK entering state PTKSTART
 wlan1: STA 02:00:00:00:00:00 WPA: sending 1/4 msg of 4-Way Handshake
 #### WPA: rsn_pmkid():
 #### WPA: aa              - hexdump(len=6): 02 00 00 00 01 00
 #### WPA: spa             - hexdump(len=6): 02 00 00 00 00 00
 #### WPA: PMK             - hexdump(len=32): b5 24 76 4f 6f 50 8c f6 a1 2e 24 b8 07 4e 9a 13 1b 94 c4 a8 1f 7e 22 d6 ed fc 7d 43 c7 77 b6 f7
 #### WPA: computed PMKID  - hexdump(len=16): d8 21 9d a5 73 98 88 26 ef 03 d2 ce f7 04 7d 23
 WPA: Send EAPOL(version=1 secure=0 mic=0 ack=1 install=0 pairwise=1 kde_len=22 keyidx=0 encr=0)

That's because wpa_supplicant computed the PMKID using the wrong (old)
MAC address used during the scan. wpa_supplicant updates own_addr when
the interface goes up, as the MAC can only change while the interface
is down. However, drivers don't report all interface state changes:
for example the nl80211 driver may ignore a down-up cycle if the down
message is processed later, when the interface is already up. In such
cases, wpa_supplicant (and in particular, the EAP state machine) would
continue to use the old MAC.

Add a new driver event that notifies of MAC address changes while the
interface is active.

Signed-off-by: Beniamino Galvani <bgalvani@redhat.com>
2018-03-30 12:16:42 +03:00
Alejandro Pérez Méndez
2ff9696d3b Add definitions for RADIUS attributes standardised in RFC 7055
hostap code is used by the Moonshot software (an implementation of the
GSS EAP mechanism - RFC 7055), and those definitions are required but
missing.

Signed-off-by: Alejandro Perez <alex.perez-mendez@jisc.ac.uk>
2018-03-30 11:04:29 +03:00
Dmitry Lebed
c17915462a nl80211: Add DFS offload support using upstream nl80211 definitions
Add generic DFS offload support using the nl80211 feature that was
recently added to the mac80211-next tree. This uses the already
available DFS offload infrastructure that was previously used with
vendor specific definitions and just sets necessary flags (DFS_OFFLOAD
ext_feature) and forawrds CAC_STARTED event for processing.

Signed-off-by: Dmitry Lebed <lebed.dmitry@gmail.com>
2018-03-30 10:56:05 +03:00
Jouni Malinen
1841086adf Sync with mac80211-next.git include/uapi/linux/nl80211.h
This brings in nl80211 definitions as of 2018-03-26.

Signed-off-by: Jouni Malinen <j@w1.fi>
2018-03-30 10:54:57 +03:00
Kabilan Kannan
7cfe2f0011 Add SAR V2 power selection capability (QCA vendor attributes)
Add changes to select SAR V2 power limits using the existing QCA vendor
command.

Signed-off-by: Kabilan Kannan <kabilank@codeaurora.org>
2018-03-30 10:49:11 +03:00
Jouni Malinen
4ae39c121f tests: sigma_dut DPP protocol testing - Stop at TX on Initiator/Enrollee
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-30 10:44:38 +03:00
Jouni Malinen
f68caa35e3 tests: GAS/ANQP and Venue URL (hostapd venue_url)
Test the hostapd venue_url configuration parameter. In addition, fix the
previous defined gas_anqp_venue_url test case to use correct encoding of
the Venue URL ANQP-element payload (URLs were missing and Venue Number
was off-by-one).

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-26 16:11:54 +03:00
Jouni Malinen
7e1d3ee963 Add hostapd.conf venue_url to set Venue URL ANQP-element
The new venue_url parameter can now be used to set the Venue URL ANQP
information instead of having to construct the data and use
anqp_elem=277:<hexdump> to set the raw value.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-26 16:10:47 +03:00
Jouni Malinen
64624f31cf OWE: Fix CONFIG_OWE=y build without CONFIG_IEEE80211R=y
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-26 12:34:36 +03:00
Jouni Malinen
d555c39c41 FILS: Fix CONFIG_FILS=y build without CONFIG_IEEE80211R=y
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-26 12:33:52 +03:00
Jouni Malinen
bf001e42bf tests: FILS SK using ERP and FT initial mobility domain association
Verify use of KCK2/KEK2 and new FTE MIC calculation with FT-FILS-SHA256.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-26 12:31:04 +03:00
Jouni Malinen
f6de72534e Add NOTE control interface command for hostapd
This does the same as the matching command in wpa_supplicant, i.e., add
a note in the debug log.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-26 12:31:04 +03:00
Jouni Malinen
2f37387812 FILS: Add more complete support for FT-FILS use cases
This extends the original IEEE Std 802.11ai-2016 functionality with the
changes added in REVmd to describe how additional keys are derived to
protect the FT protocol using keys derived through FILS authentication.

This allows key_mgmt=FT-FILS-SHA256 to be used with FT protocol since
the FTE MIC can now be calculated following the changes in REVmd. The
FT-FILS-SHA384 case is still unsupported (it needs support for variable
length MIC field in FTE).

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-26 12:31:04 +03:00
Jouni Malinen
4ddca8142e FT: Derive PMK-R1 locally if requested PMKR0Name is found
Derive PMK-R1 locally if the derived PMKR1Name is not found from the
local cache, but the request is for a key that was originally generated
locally (R0KH-ID matches) and the PMKR0Name is found in the local cache.
This was apparently not hit in the previously used FT sequences, but
this is useful to have available if a PMK-R1 entry is dropped from the
local cache before PMK-R0.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-24 18:53:14 +02:00
Jouni Malinen
1dc0945cc9 FT: Do not send PMK-R1 pull request to own R0KH address
If the requested key is not available locally, there is no point in
trying to send a pull request back to self for the key.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-24 18:53:14 +02:00
Jouni Malinen
1778f1e9a4 SAE: Fix PTK derivation to use KDF-SHA256
The previous implementation ended up defaulting to using PRF-SHA1 for
deriving PTK from PMK when SAE was used. This is not correct since the
SAE AKM is defined to be using SHA-256 -based KDF instead. Fix that.

Note: This change is not backwards compatible. Both the AP and station
side implementations will need to be updated at the same time to
maintain functionality.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-23 18:44:48 +02:00
Jouni Malinen
9d94e4bb6b SAE: Fix PMKID in EAPOL-Key msg 1/4
Previously, the association that used SAE authentication ended up
recalculating the PMKID for EAPOL-Key msg 1/4 using incorrect
PMK-to-PMKID derivation instead of using the previously derived PMKID
from SAE. The correct PMKID was used only when going through PMKSA
caching exchange with a previously derived PMKSA from SAE.

Fix this by storing the SAE PMKID into the state machine entry for the
initial SAE authentication case when there is no explicit PMKSA entry
attached to the station.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-23 18:44:48 +02:00
Jouni Malinen
a03f9d17ea SAE: Fix FT-SAE key derivation for a case where PMKID in msg 1/4 matches
Previously, matching PMKSA cache entry ended up clearing XXKey. However,
that XXKey is needed in the specific case where FT-SAE goes through the
initial mobility domain association with SAE authentication. FT-SAE
worked previously since the hostapd side generation of the particular
PMKID value in msg 1/4 was broken, but once that PMKID is fixed,
wpa_supplicant will need this fix to allow FT-SAE to be used.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-23 18:44:48 +02:00
Jouni Malinen
0fa669bcae Fix a resource leak on hostapd maclist parsing error path
The open file needs to be closed in error case. The conversion to using
a new helper function (hostapd_add_acl_maclist) somehow managed to
remove the neede fclose(f) call. Bring it back to fix this.

Fixes: 3988046de5 ("hostapd: Dynamic MAC ACL management over control interface")
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-21 22:34:09 +02:00
Jouni Malinen
80da433507 tests: Overlapping BSS scan report handling no overlap
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-21 22:15:29 +02:00
Jouni Malinen
10949e2480 Do not disable 40 MHz based on co-ex report with matching primary channel
When processing 20/40 BSS Coexistence Management frames that do not
explicitly require 40 MHz to be disabled, check whether the reported
channels in 20/40 BSS Intolerant Channel Report element match the
current primary channel. If so, allow 40 MHz operation to continue. This
makes the during-operation updates for 20/40 Operation Permitted more
consistent with the scans during initial BSS startup.

The received 20/40 BSS Intolerant Channel Report channels are to be used
in the OT set in the during-operation determination and the P == OT_i
exception was ignored in the previous implementation which could result
in the AP first starting with 40 MHz and then dropping to 20 MHz on
first received 20/40 BSS Coexistence Management frame even though there
was no change in the neighboring BSSs.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-21 16:58:05 +02:00
Jouni Malinen
8cbd9c3ed4 Support multiple 20/40 BSS Intolerant Channel Report elements
This extends 20/40 BSS Coexistence Management frame processing to
iterate over all the included 20/40 BSS Intolerant Channel Report
elements instead of using only the first one.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-21 16:53:50 +02:00
Jouni Malinen
8127a0ac02 Ignore intra-BSS 20/40 BSS Coexistence Management from not-associated STA
The 20 MHz BSS Width Request field is set to 1 only for intra-BSS
reports. As such, ignore the frame if such a claim is made by a
transmitter that is not currently associated with the AP.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-21 16:35:15 +02:00
Jouni Malinen
8aa599d45a Add more debug prints for 20/40 BSS Coexistence Management frame Rx
This makes it easier to understand what kind of information a STA is
reporting about 20/40 MHz coexistence requirements.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-21 16:32:35 +02:00
Rajkumar Manoharan
d58c3bd8b7 hostapd: Ignore LOW_ACK event for co-operative steering clients
Ignore hostapd_event_sta_low_ack for a station which has agreed to
steering by checking the agreed_to_steer flag. This flag will be set
whenever a station accepts the BSS transition request from the AP.
Without this ignoring of the LOW_ACK event, the steering in-progress
might be affected due to disassociation. In this way AP will allow some
time (two seconds) for the station to move away and reset the flag after
the timeout.

Co-Developed-by: Tamizh Chelvam <tamizhr@codeaurora.org>
Signed-off-by: Rajkumar Manoharan <rmanohar@codeaurora.org>
Signed-off-by: Tamizh chelvam <tamizhr@codeaurora.org>
2018-03-19 20:16:32 +02:00
Tamizh chelvam
ec2b5173ce Make STA opmode change event available to upper layers
Add an event callback for EVENT_STATION_OPMODE_CHANGED to allow
user/application to get the notification whenever there is a change in a
station's HT/VHT op mode.

The new events:
STA-OPMODE-MAX-BW-CHANGED <addr> <20(no-HT)|20|40|80|80+80|160>
STA-OPMODE-SMPS-MODE-CHANGED <addr> <automatic|off|dynamic|static>
STA-OPMODE-N_SS-CHANGED <addr> <N_SS>

Signed-off-by: Tamizh chelvam <tamizhr@codeaurora.org>
2018-03-19 20:07:22 +02:00
Tamizh chelvam
e8ada1600f nl80211: Add support for STA opmode change events
The nl80211 driver can report STA_OPMODE notification event as soon as
it receives an HT/VHT Action frame about modification of station's SMPS
mode/bandwidth/RX NSS. Add support to parse such events.

Signed-off-by: Tamizh chelvam <tamizhr@codeaurora.org>
2018-03-19 19:57:25 +02:00
Bhagavathi Perumal S
72123a84cb hostapd: Add last_ack_rssi into ctrl iface cmd STA
This allows external application to get last ACK signal strength of the
last transmitted frame if the driver makes this information
(NL80211_STA_INFO_ACK_SIGNAL) available.

Signed-off-by: Bhagavathi Perumal S <bperumal@codeaurora.org>
Signed-off-by: Venkateswara Naralasetty <vnaralas@codeaurora.org>
2018-03-19 19:23:26 +02:00
Bhagavathi Perumal S
2df73f52e8 Add hostapd_cli poll_sta command
This uses the already existing POLL_STA control interface to poll an
associated station to check connectivity.

Signed-off-by: Bhagavathi Perumal S <bperumal@codeaurora.org>
Signed-off-by: Venkateswara Naralasetty <vnaralas@codeaurora.org>
2018-03-19 19:14:40 +02:00
Ashok Ponnaiah
f5701cc66e OWE: Clean up pointer check in a testing code path
Check wpa_auth_write_assoc_resp_owe() return value to keep static
analyzers happier. The code path where this could happen is not really
reachable due to the separate hapd->conf->own_ie_override check and
wpa_auth_write_assoc_resp_owe() returning NULL only in an error case in
the override path. Furthermore, clean up the pointer return value to use
a proper pointer (NULL vs. 0).

Signed-off-by: Ashok Ponnaiah <aponnaia@codeaurora.org>
2018-03-19 19:08:34 +02:00
Jouni Malinen
ebe61e564c Sync with mac80211-next.git include/uapi/linux/nl80211.h
This brings in nl80211 definitions as of 2018-02-13.

Signed-off-by: Jouni Malinen <j@w1.fi>
2018-03-19 19:08:15 +02:00
Jouni Malinen
0c061630ad tests: DPP Configurator reconfiguration
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-16 20:56:05 +02:00
Purushottam Kushwaha
8179ae3a2a DPP: Support retrieving of configurator's private key
To retain configurator information across hostapd/wpa_supplicant
restart, private key need to be maintained to generate a valid pair of
authentication keys (connector, netaccess_key, csign) for new enrollees
in the network.

Add a DPP_CONFIGURATOR_GET_KEY control interface API through which the
private key of an existing configurator can be fetched.

Command format:
DPP_CONFIGURATOR_GET_KEY <configurator_id>

The output from this command can then be used with
"DPP_CONFIGURATOR_ADD key=<hexdump>" to create the same key again.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-16 20:55:32 +02:00
Jouni Malinen
4bc801ab42 SAE: Fix EAPOL-Key integrity and key-wrap algorithm selection
The SAE AKM 00-0F-AC:8 is supposed to use EAPOL-Key Key Descriptor
Version 0 (AKM-defined) with AES-128-CMAC and NIST AES Key Wrap.
However, the previous implementation ended up using Key Descriptor
Version 2 (HMAC-SHA-1-128 and NIST AES Key Wrap). Fix this by using the
appropriate Key Descriptor Version and integrity algorithm. Use helper
functions to keep the selection clearer and more consistent between
wpa_supplicant and hostapd uses.

Note: This change is not backwards compatible. Both the AP and station
side implementations will need to be updated at the same time to
maintain functionality.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-16 13:36:42 +02:00
Jouni Malinen
c63e69c379 tests: DPP protocol testing - stop when transmitting Auth Conf
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-13 01:04:03 +02:00
Srinivas Dasari
d74963d411 DPP: Extend dpp_test 89 functionality to transmit side
This extends dpp_test functionality to allow DPP exchanges to be stopped
after authentication is completed on the Initiator, i.e., after sending
out the Authentication Confirm message. Previously, dpp_test=89 was used
only on the Responder side to stop after receiving the Authentication
Confirm message. The main use case for this extended functionality is to
be able to stop the protocol exchange on a device that acts as
authentication Initiator and Enrollee.

Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-13 01:03:37 +02:00
Purushottam Kushwaha
f8bfc9cbde Use correct WPA_ALG_* values to compare for enum wpa_alg
enum wpa_alg was being compared with WPA_CIPHER_* values. That does not
work here and strict compilers will report this as an error. Fix the
comparision to use proper WPA_ALG_* values. This fixes testing
capability for resetting IPN for BIP.

Fixes: 16579769ff ("Add testing functionality for resetting PN/IPN for configured keys")
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
2018-03-13 00:32:43 +02:00
Jouni Malinen
cce16e43f8 mka: Mark ieee802_1x_kay_create_mka() ckn and cak arguments const
These structures are not modified or freed (i.e., only data from them is
copied), so mark the arguments const to document this a bit more clearly
now that there was a memory leak in one of the callers to this function.

Signed-off-by: Jouni Malinen <j@w1.fi>
2018-03-11 17:04:34 +02:00
Davide Caratti
22151b111b wpa_supplicant: Fix memory leaks in ieee802_1x_create_preshared_mka()
In case MKA is initialized successfully, local copies of CAK and CKN
were allocated, but never freed. Ensure that such memory is released
also when ieee802_1x_kay_create_mka() returns a valid pointer.

Fixes: ad51731abf ("wpa_supplicant: Allow pre-shared (CAK,CKN) pair for MKA")
Signed-off-by: Davide Caratti <davide.caratti@gmail.com>
2018-03-11 17:01:14 +02:00
Mike Siedzik
3a52f6b387 mka: Do not print contents of SAK to debug log
Log newly generated SAKs as well as unwrapped SAKs with wpa_hexdump_key()
rather than wpa_hexdump(). By default, the wpa_hexdump_key() function
will not display sensitive key data.

Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
2018-03-11 16:52:56 +02:00
Mike Siedzik
77977b3d5d mka: Detect duplicate MAC addresses during key server election
In the unlikely event the local KaY and the elected peer have the same
actor priority as well as the same MAC address, log a warning message
and do not elect a key server. Resolution is for network administrator
to reconfigure MAC address.

Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
2018-03-11 16:49:00 +02:00
Mike Siedzik
5762855abd mka: Loss of live peers to result in connect PENDING not AUTHENTICATED
When the number of live peers becomes 0 the KaY was setting
kay->authenticated true and telling the CP to connect AUTHENTICATED.
Per IEEE Std 802.1X-2010 Clause 12.2, MKA.authenticated means "the Key
Server has proved mutual authentication but has determined that
Controlled Port communication should proceed without the use of MACsec",
which means port traffic will be passed in the clear.

When the number of live peers becomes 0 the KaY must instead set
kay->authenticated false and tell the CP to connect PENDING. Per Clause
12.3 connect PENDING will "prevent connectivity by clearing the
controlledPortEnabled parameter."

Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
2018-03-11 16:43:12 +02:00
Mike Siedzik
8fb546d8e6 mka: Ignore MACsec SAK Use Old Key parameter if we don't have our old key
Upon receipt of the "MACsec MKPDU SAK Use parameter set" the KaY verifies
that both the latest key and the old key are valid. If the local system
reboots or is reinitialized, the KaY won't have a copy of its old key.
Therefore if the KaY does not have a copy of its old key it should not
reject MKPDUs that contain old key data in the MACsec SAK Use parameter.

Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
2018-03-11 16:30:29 +02:00
Mike Siedzik
b54b53e644 mka: When matching CKNs ensure that lengths are identical
KaY looks up participants using CAK Name (CKN). Per IEEE Std 802.1X-2010
Clause 9.3.1 CAK identification, the CKN is an integral number of
octets, between 1 and 32 (inclusive). This fix will ensure that the KaY
does not inadvertently match CKNs such as 'myCakNamedFoo' and
'myCakNamedFooBar'.

Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
2018-03-11 13:33:39 +02:00