When strict group rekeying is in effect, every station that leaves will
cause a rekeying to happen 0.5 s after leaving. However, if a lot of
stations join/leave, the previous code could postpone this rekeying
forever, since it always re-registers the handling with a 0.5 s timeout.
Use eloop_deplete_timeout() to address that, only registering the
timeout from scratch if it wasn't pending.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
In order to test the WoWLAN GTK rekeying KRACK mitigation, add a
REKEY_GTK hostapd control interface command that can be used at certain
points of the test.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This allows hostapd testing functionality to be forced to send out a
plaintext EAPOL-Key frame with the RESEND_* command. That can be useful
in seeing how the station behaves if an unencrypted EAPOL frame is
received when TK is already configured.
This is not really perfect since there is no convenient way of sending
out a single unencrypted frame in the current nl80211 design. The
monitor interface could likely still do this, but that's not really
supposed to be used anymore. For now, clear and restore TK during this
operation. The restore part is not really working correctly, though,
since it ends up clearing the TSC value on the AP side and that shows up
as replay protection issues on the station. Anyway, this is sufficient
to generate sniffer captures to analyze station behavior.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
This adds a new hostapd configuration parameter
wpa_disable_eapol_key_retries=1 that can be used to disable
retransmission of EAPOL-Key frames that are used to install
keys (EAPOL-Key message 3/4 and group message 1/2). This is
similar to setting wpa_group_update_count=1 and
wpa_pairwise_update_count=1, but with no impact to message 1/4
retries and with extended timeout for messages 4/4 and group
message 2/2 to avoid causing issues with stations that may use
aggressive power saving have very long time in replying to the
EAPOL-Key messages.
This option can be used to work around key reinstallation attacks
on the station (supplicant) side in cases those station devices
cannot be updated for some reason. By removing the
retransmissions the attacker cannot cause key reinstallation with
a delayed frame transmission. This is related to the station side
vulnerabilities CVE-2017-13077, CVE-2017-13078, CVE-2017-13079,
CVE-2017-13080, and CVE-2017-13081.
This workaround might cause interoperability issues and reduced
robustness of key negotiation especially in environments with
heavy traffic load due to the number of attempts to perform the
key exchange is reduced significantly. As such, this workaround
is disabled by default (unless overridden in build
configuration). To enable this, set the parameter to 1.
It is also possible to enable this in the build by default by
adding the following to the build configuration:
CFLAGS += -DDEFAULT_WPA_DISABLE_EAPOL_KEY_RETRIES=1
Signed-off-by: Jouni Malinen <j@w1.fi>
The new hostapd control interface commands "RESEND_M1 <addr>" and
"RESEND_M3 <addr>" can be used to request a retransmission of the 4-Way
Handshake messages 1/4 and 3/4 witht he same or modified ANonce (in M1).
This functionality is for testing purposes and included only in builds
with CONFIG_TESTING_OPTIONS=y.
Signed-off-by: Jouni Malinen <j@w1.fi>
The new hostapd control interface command "RESEND_GROUP_M1 <addr>" can
be used to request a retransmission of the Group Key Handshake message
1/2 for the current GTK.
This functionality is for testing purposes and included only in builds
with CONFIG_TESTING_OPTIONS=y.
Signed-off-by: Jouni Malinen <j@w1.fi>
This was originally added to allow the IEEE 802.11 protocol to be
tested, but there are no known fully functional implementations based on
this nor any known deployments of PeerKey functionality. Furthermore,
PeerKey design in the IEEE Std 802.11-2016 standard has already been
marked as obsolete for DLS and it is being considered for complete
removal in REVmd.
This implementation did not really work, so it could not have been used
in practice. For example, key configuration was using incorrect
algorithm values (WPA_CIPHER_* instead of WPA_ALG_*) which resulted in
mapping to an invalid WPA_ALG_* value for the actual driver operation.
As such, the derived key could not have been successfully set for the
link.
Since there are bugs in this implementation and there does not seem to
be any future for the PeerKey design with DLS (TDLS being the future for
DLS), the best approach is to simply delete all this code to simplify
the EAPOL-Key handling design and to get rid of any potential issues if
these code paths were accidentially reachable.
Signed-off-by: Jouni Malinen <j@w1.fi>
The previous implementation ended up starting a new EAPOL-Key 4-way
handshake if the STA were to attempt to perform another association.
This resulted in immediate disconnection since the PTK was not ready for
configuring FILS TK at the point when EAPOL-Key msg 1/4 is sent out.
This is better than alloing the association to continue with the same TK
reconfigured, but not really ideal.
Address this potential sequence by not starting a new 4-way handshake on
the additional association attempt. Instead, allow the association to
complete, but do so without reconfiguring the TK to avoid potential
issues with PN reuse with the same TK.
Signed-off-by: Jouni Malinen <j@w1.fi>
The Authenticator state machine path for PTK rekeying ended up bypassing
the AUTHENTICATION2 state where a new ANonce is generated when going
directly to the PTKSTART state since there is no need to try to
determine the PMK again in such a case. This is far from ideal since the
new PTK would depend on a new nonce only from the supplicant.
Fix this by generating a new ANonce when moving to the PTKSTART state
for the purpose of starting new 4-way handshake to rekey PTK.
Signed-off-by: Jouni Malinen <j@w1.fi>
Do not reinstall TK to the driver during Reassociation Response frame
processing if the first attempt of setting the TK succeeded. This avoids
issues related to clearing the TX/RX PN that could result in reusing
same PN values for transmitted frames (e.g., due to CCM nonce reuse and
also hitting replay protection on the receiver) and accepting replayed
frames on RX side.
This issue was introduced by the commit
0e84c25434 ('FT: Fix PTK configuration in
authenticator') which allowed wpa_ft_install_ptk() to be called multiple
times with the same PTK. While the second configuration attempt is
needed with some drivers, it must be done only if the first attempt
failed.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
This extends OWE support in hostapd to allow DH groups 20 and 21 to be
used in addition to the mandatory group 19 (NIST P-256).
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
While the Suite B AKM is not really going to be used with CCMP-128 or
GCMP-128 cipher, this corner case could be fixed if it is useful for
some testing purposes. Allow that special case to skip the HMAC-SHA1
check based on CCMP/GCMP cipher and use the following AKM-defined check
instead.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
This part is missing from IEEE Std 802.11ai-2016, but the lack of DHss
here means there would not be proper PFS for the case where PMKSA
caching is used with FILS SK+PFS authentication. This was not really the
intent of the FILS design and that issue was fixed during REVmd work
with the changes proposed in
https://mentor.ieee.org/802.11/dcn/17/11-17-0906-04-000m-fils-fixes.docx
that add DHss into FILS-Key-Data (and PTK, in practice) derivation for
the PMKSA caching case so that a unique ICK, KEK, and TK are derived
even when using the same PMK.
Note: This is not backwards compatible, i.e., this breaks PMKSA caching
with FILS SK+PFS if only STA or AP side implementation is updated.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
IEEE Std 802.11ai-2016 had missed a change in the Pairwise key hierarchy
clause (12.7.1.3 in IEEE Std 802.11-2016) and due to that, the previous
implementation ended up using HMAC-SHA-1 -based PMKID derivation. This
was not really the intent of the FILS design and that issue was fixed
during REVmd work with the changes proposed in
https://mentor.ieee.org/802.11/dcn/17/11-17-0906-04-000m-fils-fixes.docx
that change FILS cases to use HMAC-SHA-256 and HMAC-SHA-384 based on the
negotiated AKM.
Update the implementation to match the new design. This changes the
rsn_pmkid() function to take in the more generic AKMP identifier instead
of a boolean identifying whether SHA256 is used.
Note: This is not backwards compatible, i.e., this breaks PMKSA caching
based on the initial ERP key hierarchy setup if only STA or AP side
implementation is updated. PMKSA caching based on FILS authentication
exchange is not impacted by this, though.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Add PMKID into EAPOL-Key 1/4 when using SAE and fix the PMK-from-PMKSA
selection in some cases where PSK (from passphrase) could have been
used.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Previously, CONFIG_WNM enabled build that supports WNM for both
station mode and AP mode. However, in most wpa_supplicant cases only
station mode WNM is required and there is no need for AP mode WNM.
Add support to differentiate between station mode WNM and AP mode
WNM in wpa_supplicant builds by adding CONFIG_WNM_AP that should be
used when AP mode WNM support is required in addition to station mode
WNM. This allows binary size to be reduced for builds that require
only the station side WNM functionality.
Signed-off-by: Avraham Stern <avraham.stern@intel.com>
This allows external programs to generate and add PMKSA cache entries
into hostapd. The main use for this is to run external DPP processing
(network introduction) and testing.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
This new AKM is used with DPP when using the signed Connector to derive
a PMK. Since the KCK, KEK, and MIC lengths are variable within a single
AKM, this needs number of additional changes to get the PMK length
delivered to places that need to figure out the lengths of the PTK
components.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Derive PMK-R0 and the relevant key names when using FILS authentication
for initial FT mobility domain association.
Signed-off-by: Jouni Malinen <j@w1.fi>
This extends fils_pmk_to_ptk() to allow FILS-FT to be derived. The
callers do not yet use that capability; i.e., actual use will be added
in separate commits.
Signed-off-by: Jouni Malinen <j@w1.fi>
The conditional gSTA and gAP (DH public keys) were not previously
included in Key-Auth derivation, but they are needed for the PFS case.
Signed-off-by: Jouni Malinen <j@w1.fi>
Enable use of FT RRB without configuring each other AP locally. Instead,
broadcast messages are exchanged to discover APs within the local
network.
When an R0KH or R1KH is discovered, it is cached for one day.
When a station uses an invalid or offline r0kh_id, requests are always
broadcast. In order to avoid this, if r0kh does not reply, a temporary
blacklist entry is added to r0kh_list.
To avoid blocking a valid r0kh when a non-existing pmk_r0_name is
requested, r0kh is required to always reply using a NAK. Resend requests
a few times to ensure blacklisting does not happen due to small packet
loss.
To free newly created stations later, the r*kh_list start pointer in
conf needs to be updateable from wpa_auth_ft.c, where only wconf is
accessed.
Signed-off-by: Michael Braun <michael-dev@fami-braun.de>
This adds a counter and adds sequence numbering to FT RRB packets. The
sequence number is checked against r0kh/r1kh sequence number cache.
Special attention is needed in case the remote AP reboots and thus loses
its state. I prefer it to recover automatically even without synchronized
clocks. Therefore an identifier called dom is generated randomly along the
initial sequence number. If the dom transmitted does not match or the
sequence number is not in the range currently expected, the sender is asked
for a fresh confirmation of its currently used sequence numbers. The packet
that triggered this is cached and processed again later.
Additionally, in order to ensure freshness, the remote AP includes an
timestamp with its messages. It is then verified that the received
messages are indeed fresh by comparing it to the older timestamps
received and the time elapsed since then. Therefore FT_RRB_TIMESTAMP is
no longer needed.
This assigns new OUI 00:13:74 vendor-specific subtype 0x0001 subtypes:
4 (SEQ_REQ) and 5 (SEQ_RESP).
This breaks backward compatibility, i.e., hostapd needs to be updated
on all APs at the same time to allow FT to remain functional.
Signed-off-by: Michael Braun <michael-dev@fami-braun.de>
This is used with partial AP SME in driver cases to enable FILS
association (AES-SIV) processing.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
MDE was already added with RSNE, but FTE needed to be added to the FILS
Authentication frame for the FT initial mobility domain association
using FILS authentication case.
Signed-off-by: Jouni Malinen <j@w1.fi>
This fixes the following compiler warning:
wpa_auth.c:4249:34: error: address of array 'a->conf.fils_cache_id'
will always evaluate to 'true' [-Werror,-Wpointer-bool-conversion]
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
draft-harkins-owe-07.txt does not specify these parameters, so need to
pick something sensible to use for the experimental implementation. The
Suite B 128-bit level AKM 00-0F-AC:11 has reasonable parameters for the
DH group 19 case (i.e., SHA256 hash), so use it for now. This can be
updated if the OWE RFC becomes clearer on the appropriate parameters
(KEK/KCK/MIC length, PRF/KDF algorithm, and key-wrap algorithm).
Signed-off-by: Jouni Malinen <j@w1.fi>
This leads to cleaner code overall, and also reduces the size
of the hostapd and wpa_supplicant binaries (in hwsim test build
on x86_64) by about 2.5 and 3.5KiB respectively.
The mechanical conversions all over the code were done with
the following spatch:
@@
expression SIZE, SRC;
expression a;
@@
-a = os_malloc(SIZE);
+a = os_memdup(SRC, SIZE);
<...
if (!a) {...}
...>
-os_memcpy(a, SRC, SIZE);
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This allows PMKSA cache entries to be shared between all the BSSs
operated by the same hostapd process when those BSSs use the same FILS
Cache Identifier value.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
While the FILS authentication cases were already using the proper PMK
length (48 octets instead of the old hardcoded 32 octet), the initial
association case had not yet been updated to cover the new FILS SHA384
AKM and ended up using only a 32-octet PMK. Fix that to use 48-octet PMK
when using FILS SHA384 AKM.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Previously, any potential (even if very unlikely) local operation error
was ignored. Now these will result in aborting the negotiation.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
The struct wpa_stsl_negotiation seemed to have been for some kind of
tracking of state of PeerKey negotiations within hostapd. However,
nothing is actually adding any entries to wpa_auth->stsl_negotiations or
using this state. Since PeerKey does not look like something that would
be deployed in practice, there is no justification to spend time on
making this any more complete. Remove the dead code now instead of
trying to figure out what it might be used for.
Signed-off-by: Jouni Malinen <j@w1.fi>
It can happen if the station is unreachable or sleeping longer than
the actual total GTK rekey timeout. To fix the latter case
wpa_group_update_count may be increased.
Signed-off-by: Günther Kelleter <guenther.kelleter@devolo.de>
wpa_group_update_count and wpa_pairwise_update_count can now be used to
set the GTK and PTK rekey retry limits (dot11RSNAConfigGroupUpdateCount
and dot11RSNAConfigPairwiseUpdateCount). Defaults set to current
hardcoded value (4).
Some stations may suffer from frequent deauthentications due to GTK
rekey failures: EAPOL 1/2 frame is not answered during the total timeout
period of currently ~3.5 seconds. For example, a Galaxy S6 with Android
6.0.1 appears to go into power save mode for up to 5 seconds. Increasing
wpa_group_update_count to 6 fixed this issue.
Signed-off-by: Günther Kelleter <guenther.kelleter@devolo.de>
P802.11i/D3.0 described the Key Length as having value 16 for the group
key handshake. However, this was changed to 0 in the published IEEE Std
802.11i-2004 amendment (and still remains 0 in the current standard IEEE
Std 802.11-2016). We need to maintain the non-zero value for WPA (v1)
cases, but the RSN case can be changed to 0 to be closer to the current
standard.
Signed-off-by: Jouni Malinen <j@w1.fi>
While the IEEE Std 802.11-2016 is still indicating that GNonce would be
exchanged in EAPOL-Key messages (see, e.g., Figure 12-52 showing the
Send EAPOL-Key operation in the REKEYNEGOTIATING state or the sample
group key handshake in Figure 12-47), there are also examples of
describing this field as having value zero (e.g., 12.7.7.2 Group key
handshake message 1).
GNonce is used only with the Authenticator and the Supplicant does not
have any use for it, so it is better not to expose that internal value.
Hardcode the Key Nonce field to 0 in EAPOL-Key group message 1/2.
Signed-off-by: Jouni Malinen <j@w1.fi>
sm->Pair needs to be initialized to TRUE since unicast cipher is
supported and this is an ESS. However, the normal place for setting this
(WPA_PTK::INITIALIZE) is skipped with using FT protocol or FILS
authentication, so need to do that separately when forcing PTKINITDONE.
Signed-off-by: Jouni Malinen <j@w1.fi>
The hostapd processing of the AES-SIV AAD was incorrect. The design for
the AAD changed between P802.11ai/D7.0 and D8.0 from a single vector
with concatenated data to separate vectors. The change in the
implementation had missed the change in the aes_siv_decrypt() call for
the num_elem parameter. This happened to work with the mac80211
implementation due to a similar error there.
Fix this by using the correct numbers of vectors in the SIV AAD so that
all the vectors get checked. The last vector was also 14 octets too long
due to incorrect starting pointer, so fix that as well. The changes here
are not backwards compatible, i.e., a similar fix in the Linux mac80211
is needed to make things interoperate again.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
The new dhcp_server configuration parameter can now be used to configure
hostapd to act as a DHCP relay for DHCPDISCOVER messages received as
FILS HLP requests. The dhcp_rapid_commit_proxy=1 parameter can be used
to configure hostapd to convert 4 message DHCP exchange into a 2 message
exchange in case the DHCP server does not support DHCP rapid commit
option.
The fils_hlp_wait_time parameter can be used to set the time hostapd
waits for an HLP response. This matches the dot11HLPWaitTime in IEEE Std
802.11ai-2016.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Instead of copying the struct wpa_auth_callbacks, just keep a pointer to
it, keep the context pointer separate, and let the user just provide a
static const structure. This reduces the attack surface of heap
overwrites, since the function pointers move elsewhere.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
These commnds are mesh version of PMKSA_GET/ADD commands. So the usage
and security risk is similar to them. Refer to
commit 3459381dd2 ('External persistent
storage for PMKSA cache entries') also.
The MESH_PMKSA_GET command requires peer MAC address or "any" as an
argument and outputs appropriate stored PMKSA cache. And the
MESH_PMKSA_ADD command receives an output of MESH_PMKSA_GET and re-store
the PMKSA cache into wpa_supplicant. By using re-stored PMKSA cache,
wpa_supplicant can skip commit message creation which can use
significant CPU resources.
The output of the MESH_PMKSA_GET command uses the following format:
<BSSID> <PMKID> <PMK> <expiration in seconds>
The example of MESH_PMKSA_ADD command is this.
MESH_PMKSA_ADD 02:00:00:00:03:00 231dc1c9fa2eed0354ea49e8ff2cc2dc cb0f6c9cab358a8146488566ca155421ab4f3ea4a6de2120050c149b797018fe 42930
MESH_PMKSA_ADD 02:00:00:00:04:00 d7e595916611640d3e4e8eac02909c3c eb414a33c74831275f25c2357b3c12e3d8bd2f2aab6cf781d6ade706be71321a 43180
This functionality is disabled by default and can be enabled with
CONFIG_PMKSA_CACHE_EXTERNAL=y build configuration option.
Signed-off-by: Masashi Honma <masashi.honma@gmail.com>
