Ben Greear [Wed, 18 Oct 2017 22:58:30 +0000 (15:58 -0700)]
Fix test build breakage when not compiling with mesh support
Build breakage was introduced by commit 16579769ff7bb255e101c6b77fb2c47d3735a883 ('Add testing functionality for
resetting PN/IPN for configured keys') for some CONFIG_TESTING_OPTIONS=y
builds.
Signed-off-by: Ben Greear <greearb@candelatech.com>
Jouni Malinen [Fri, 20 Oct 2017 15:18:53 +0000 (18:18 +0300)]
Fix the notes on EAPOL-Key testing procedures
The extra sanity check for replay protection in these procedures ended
up breaking the tests. RESET_PN cannot be used before RESEND_* commands
since that would prevent the DUT from accepting the retransmitted
EAPOL-Key frames.
Jouni Malinen [Fri, 20 Oct 2017 14:44:07 +0000 (17:44 +0300)]
WNM: Ignore BSS Transition Management frames in bss_transition=0 case
The hostapd bss_transition parameter was previously used to control
advertisement of BSS Transition Management support, but it was not used
when processing BSS Transition Management Query/Response frames. Add an
explicit check during frame processing as well so that any misbehaving
station is ignored. In addition to bss_transition=1, allow mbo=1 to be
used to mark the functionality enabled.
Jouni Malinen [Fri, 20 Oct 2017 14:39:42 +0000 (17:39 +0300)]
WNM: Ignore WNM-Sleep Mode Request in wnm_sleep_mode=0 case
The hostapd wnm_sleep_mode parameter was previously used to control
advertisement of WNM-Sleep Mode support, but it was not used when
processing a request to use WNM-Sleep Mode. Add an explicit check during
request processing as well so that any misbehaving station is ignored.
Jouni Malinen [Thu, 19 Oct 2017 09:16:18 +0000 (12:16 +0300)]
Extend RESEND_* test commands to allow forcing plaintext TX
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.
FILS: Send updated connection parameters to drivers if needed
After an initial connection wpa_supplicant derives ERP information which
can be used in doing eventual authentications in the same realm. This
information can be used by drivers with offloaded FILS support to do
driver/firmware initiated roamings. Add support to send this updated
information to such drivers.
Update the replay counter after a roam for all cases. This restores the
design back to what it was before commit 01ef320f192daa074c7055a44a03b6b5b811d6bd ('FILS: Update ERP next
sequence number with driver offload').
nl80211: Add support to send updated connection parameters
After an initial connection certain connection parameters may be
updated. It may be necessary to send these parameters to drivers since
these will be used in driver-initiated roaming cases. This commit
defines the driver_ops call for this and implements the needed
functionality for the nl80211 driver interface.
Add attributes to support roam+auth vendor event for FILS
Add additional attributes to specify the PMK, PMKID, and the ERP next
sequence number to the vendor subcommand
QCA_NL80211_VENDOR_SUBCMD_KEY_MGMT_ROAM_AUTH. These are needed in case
of an offloaded FILS roaming.
Jouni Malinen [Mon, 16 Oct 2017 22:15:24 +0000 (01:15 +0300)]
Clear BSSID information in supplicant state machine on disconnection
This fixes a corner case where RSN pre-authentication candidate from
scan results was ignored if the station was associated with that BSS
just before running the new scan for the connection.
Jouni Malinen [Mon, 16 Oct 2017 21:01:11 +0000 (00:01 +0300)]
Additional consistentcy checks for PTK component lengths
Verify that TK, KCK, and KEK lengths are set to consistent values within
struct wpa_ptk before using them in supplicant. This is an additional
layer of protection against unexpected states.
Jouni Malinen [Mon, 16 Oct 2017 15:37:43 +0000 (18:37 +0300)]
Optional AP side workaround for key reinstallation attacks
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:
Jouni Malinen [Sun, 15 Oct 2017 20:25:55 +0000 (23:25 +0300)]
Allow last (Re)Association Request frame to be replayed for testing
The new wpa_supplicant RESEND_ASSOC command can be used to request the
last (Re)Association Request frame to be sent to the AP to test FT
protocol behavior.
This functionality is for testing purposes and included only in builds
with CONFIG_TESTING_OPTIONS=y.
Jouni Malinen [Sat, 14 Oct 2017 14:23:59 +0000 (17:23 +0300)]
Allow EAPOL-Key messages 1/4 and 3/4 to be retransmitted for testing
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.
Jouni Malinen [Sat, 14 Oct 2017 13:53:27 +0000 (16:53 +0300)]
Allow group key handshake message 1/2 to be retransmitted for testing
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.
Jouni Malinen [Sat, 14 Oct 2017 10:41:08 +0000 (13:41 +0300)]
Add testing functionality for resetting PN/IPN for configured keys
This can be used to test replay protection. The "RESET_PN" command in
wpa_supplicant and "RESET_PN <addr>" command in hostapd resets the local
counters to zero for the last configured key. For hostapd, the address
parameter specifies which STA this operation is for or selects GTK
("ff:ff:ff:ff:ff:ff") or IGTK ("ff:ff:ff:ff:ff:ff IGTK").
This functionality is for testing purposes and included only in builds
with CONFIG_TESTING_OPTIONS=y.
Jouni Malinen [Sat, 14 Oct 2017 09:15:52 +0000 (12:15 +0300)]
wlantest: Do not update RSC on replays
This changes wlantest behavior to mark CCMP/TKIP replays for more cases
in case a device is resetting its TSC. Previously, the RSC check got
cleared on the first marked replay and the following packets were not
marked as replays if they continued incrementing the PN even if that PN
was below the highest value received with this key at some point in the
past.
Jouni Malinen [Sun, 8 Oct 2017 10:18:02 +0000 (13:18 +0300)]
Clear PMK length and check for this when deriving PTK
Instead of setting the default PMK length for the cleared PMK, set the
length to 0 and explicitly check for this when deriving PTK to avoid
unexpected key derivation with an all-zeroes key should it be possible
to somehow trigger PTK derivation to happen before PMK derivation.
Mathy Vanhoef [Thu, 5 Oct 2017 21:53:01 +0000 (23:53 +0200)]
WPA: Extra defense against PTK reinstalls in 4-way handshake
Currently, reinstallations of the PTK are prevented by (1) assuring the
same TPTK is only set once as the PTK, and (2) that one particular PTK
is only installed once. This patch makes it more explicit that point (1)
is required to prevent key reinstallations. At the same time, this patch
hardens wpa_supplicant such that future changes do not accidentally
break this property.
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.
FILS: Do not allow multiple (Re)Association Response frames
The driver is expected to not report a second association event without
the station having explicitly request a new association. As such, this
case should not be reachable. However, since reconfiguring the same
pairwise or group keys to the driver could result in nonce reuse issues,
be extra careful here and do an additional state check to avoid this
even if the local driver ends up somehow accepting an unexpected
(Re)Association Response frame.
tests: Fix wnm_action_proto_no_pmf to have active WNM_SLEEP operation
The previous designed worked since wpa_supplicant did not track pending
request state. With such tracking added, this test case needs to make
sure there is a pending operation when injecting the invalid response.
Jouni Malinen [Fri, 25 Aug 2017 13:24:18 +0000 (16:24 +0300)]
FILS: Accept another (Re)Association Request frame during an association
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.
Jouni Malinen [Fri, 4 Aug 2017 10:14:57 +0000 (13:14 +0300)]
Add MGMT_TX_STATUS_PROCESS command for testing purposes
This allows ext_mgmt_frame_handling=1 cases with hostapd to process TX
status events based on external processing. This is useful for increased
test coverage of management frame processing.
FT: Do not allow multiple Reassociation Response frames
The driver is expected to not report a second association event without
the station having explicitly request a new association. As such, this
case should not be reachable. However, since reconfiguring the same
pairwise or group keys to the driver could result in nonce reuse issues,
be extra careful here and do an additional state check to avoid this
even if the local driver ends up somehow accepting an unexpected
Reassociation Response frame.
WNM: Ignore WNM-Sleep Mode Response without pending request
Commit 03ed0a52393710be6bdae657d1b36efa146520e5 ('WNM: Ignore WNM-Sleep
Mode Response if WNM-Sleep Mode has not been used') started ignoring the
response when no WNM-Sleep Mode Request had been used during the
association. This can be made tighter by clearing the used flag when
successfully processing a response. This adds an additional layer of
protection against unexpected retransmissions of the response frame.
Do not try to reconfigure the same TPK-TK to the driver after it has
been successfully configured. This is an explicit check to avoid issues
related to resetting the TX/RX packet number. There was already a check
for this for TPK M2 (retries of that message are ignored completely), so
that behavior does not get modified.
For TPK M3, the TPK-TK could have been reconfigured, but that was
followed by immediate teardown of the link due to an issue in updating
the STA entry. Furthermore, for TDLS with any real security (i.e.,
ignoring open/WEP), the TPK message exchange is protected on the AP path
and simple replay attacks are not feasible.
As an additional corner case, make sure the local nonce gets updated if
the peer uses a very unlikely "random nonce" of all zeros.
Jouni Malinen [Sun, 1 Oct 2017 09:32:57 +0000 (12:32 +0300)]
Fix PTK rekeying to generate a new ANonce
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.
Properly track whether a PTK has already been installed to the driver
and the TK part cleared from memory. This prevents an attacker from
trying to trick the client into installing an all-zero TK.
This fixes the earlier fix in commit ad00d64e7d8827b3cebd665a0ceb08adabf15e1e ('Fix TK configuration to the
driver in EAPOL-Key 3/4 retry case') which did not take into account
possibility of an extra message 1/4 showing up between retries of
message 3/4.
Jouni Malinen [Sun, 1 Oct 2017 09:12:24 +0000 (12:12 +0300)]
Extend protection of GTK/IGTK reinstallation of WNM-Sleep Mode cases
This extends the protection to track last configured GTK/IGTK value
separately from EAPOL-Key frames and WNM-Sleep Mode frames to cover a
corner case where these two different mechanisms may get used when the
GTK/IGTK has changed and tracking a single value is not sufficient to
detect a possible key reconfiguration.
Prevent reinstallation of an already in-use group key
Track the current GTK and IGTK that is in use and when receiving a
(possibly retransmitted) Group Message 1 or WNM-Sleep Mode Response, do
not install the given key if it is already in use. This prevents an
attacker from trying to trick the client into resetting or lowering the
sequence counter associated to the group key.
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 0e84c25434e6a1f283c7b4e62e483729085b78d2 ('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.
Jouni Malinen [Wed, 11 Oct 2017 20:09:16 +0000 (23:09 +0300)]
SAE: Allow SAE password to be configured separately (STA)
The new sae_password network profile parameter can now be used to set
the SAE password instead of the previously used psk parameter. This
allows shorter than 8 characters and longer than 63 characters long
passwords to be used.
Jouni Malinen [Wed, 11 Oct 2017 20:07:08 +0000 (23:07 +0300)]
SAE: Allow SAE password to be configured separately (AP)
The new sae_password hostapd configuration parameter can now be used to
set the SAE password instead of the previously used wpa_passphrase
parameter. This allows shorter than 8 characters and longer than 63
characters long passwords to be used. In addition, this makes it
possible to configure a BSS with both WPA-PSK and SAE enabled to use
different passphrase/password based on which AKM is selected.
Jouni Malinen [Wed, 11 Oct 2017 15:19:03 +0000 (18:19 +0300)]
DPP: Fix static analyzer warnings in key generation and JWK construction
Memory allocation failures could have resulted in error paths that
dereference a NULL pointer or double-freeing memory. Fix this by
explicitly clearing the freed pointer and checking allocation results.
Sunil Dutt [Sun, 8 Oct 2017 05:33:21 +0000 (11:03 +0530)]
P2P: Prefer 5/60 GHz band over 2.4 GHz during GO configuration
Previously, wpas_p2p_select_go_freq_no_pref() ended up selecting a 2.4
GHz band channel first before even considering 5 or 60 GHz channels.
This was likely done more or less by accident rather than by design when
the 5 GHz and 60 GHz band extensions were added. It seems reasonable to
enhance this by reordering the code to start with 5 and 60 GHz operating
classes and move to 2.4 GHz band only if no channel was available in 5
or 60 GHz bands for P2P GO use.
This does have some potential interop issues with 2.4 GHz only peer
devices when starting up an autonomous GO (i.e., without there being
prior knowledge of channels that the peers support). Upper layers are
expected to enforce 2.4 GHz selection if that is needed for some use
cases.
Jouni Malinen [Tue, 10 Oct 2017 16:00:57 +0000 (19:00 +0300)]
OWE: Allow set of enabled DH groups to be limited on AP
The new hostapd configuration parameter owe_groups can be used to
specify a subset of the allowed DH groups as a space separated list of
group identifiers.
Jouni Malinen [Tue, 10 Oct 2017 15:26:29 +0000 (18:26 +0300)]
OWE: Allow DH Parameters element to be overridden for testing purposes
This allows CONFIG_TESTING_OPTIONS=y builds of wpa_supplicant to
override the OWE DH Parameters element in (Re)Association Request frames
with arbitrary data specified with the "VENDOR_ELEM_ADD 13 <IE>"
command. This is only for testing purposes.
Jouni Malinen [Mon, 9 Oct 2017 10:38:15 +0000 (13:38 +0300)]
OWE: Transition mode information based on BSS ifname
The owe_transition_bssid and owe_transition_ssid parameters can now be
replace with owe_transition_ifname to clone the BSSID/SSID information
automatically in case the same hostapd process manages both the OWE and
open BSS for transition mode.
Jouni Malinen [Mon, 9 Oct 2017 09:35:14 +0000 (12:35 +0300)]
OWE: Support station SME-in-driver case
Previously, only the SME-in-wpa_supplicant case was supported. This
extends that to cover the drivers that implement SME internally (e.g.,
through the cfg80211 Connect command).
Jouni Malinen [Sun, 8 Oct 2017 13:39:08 +0000 (16:39 +0300)]
OWE: Support DH groups 20 (NIST P-384) and 21 (NIST P-521) in station
This extends OWE support in wpa_supplicant to allow DH groups 20 and 21
to be used in addition to the mandatory group 19 (NIST P-256). The group
is configured using the new network profile parameter owe_group.
Jouni Malinen [Sun, 8 Oct 2017 10:49:45 +0000 (13:49 +0300)]
OWE: Include RSNE in (Re)Association Response frame
This is not normally done in RSN, but RFC 8110 seems to imply that AP
has to include OWE AKM in the RSNE within these frames. So, add the RSNE
to (Re)Association Response frames when OWE is being negotiated.
projects / thirdparty / hostap.git / log
