Increase the maximum command length from 4 kB to 8 kB mainly to allow
larger certificate blobs to be configured. Use heap memory to avoid
inconveniently large stack requirements. In addition, reject potentially
truncated commands instead of trying to process them.
The maximum length of the request can now be determined with
"GET max_command_len".
Signed-off-by: Jouni Malinen <j@w1.fi>
Python 3.8 removed the previously used default of MD5 algorithm, so
provide the explicit digestmod=hashlib.md5 parameter to the couple of
places that were missing it.
Signed-off-by: Jouni Malinen <j@w1.fi>
Randomize GAS dialog token field, instead of using an incremental
counter with predictable values. This change will make this field
unuseful for user fingerprinting.
Signed-off-by: Hai Shalom <haishalom@google.com>
IEEE 802.11 defines reason codes for indicating 4-way handshake and
group key handshake timeouts. Use those reason codes instead of the more
generic one for these particular cases in the Authenticator state
machine.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
The ap_beacon_rate_* test cases could end up terminating before the very
first Beacon frame is captured, so wait a bit after the connection to
make sure at least one Beacon frame is going to be in the capture file.
Signed-off-by: Jouni Malinen <j@w1.fi>
Convert driver indications to internal events to allow invalid Beacon
frames to be reported to the AP.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Update frame processing registration for DPP Public Action frames to
include multicast reception when in DPP listen mode and the driver
indicates support for this type of explicit request.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This test case was using sigma_dut to set sae_pwe=2, but that value was
not cleared in the end. This could result in following SAE test cases
failing due to unexpected behavior.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Documentation of the return values for wpa_parse_vendor_specific() and
wpa_parse_generic() were not accurate and the parsing results from these
were not really handled appropriately. There is no point in calling
wpa_parse_vendor_specific() if wpa_parse_generic() recognizes a KDE. Not
that this would break anything in practice, but still, it looks
confusing.
The part about handling wpa_parse_vendor_specific() return value can, at
least in theory, break some cases where an unexpectedly short KDE/vendor
specific element were present and something would need to be recognized
after it. That does not really happen with any standard compliant
implementation and this is unlikely to cause any real harm, but it is
clearer to handle this more appropriately even for any theoretical case,
including misbehavior of a peer device.
Instead of stopping parsing on too short vendor specific element,
continue parsing the following KDEs/IEs. Skip the
wpa_parse_vendor_specific() call when a KDE has been recognized. Also
fix the return value documentation for wpa_parse_generic() and remove
the useless return value from wpa_parse_vendor_specific().
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Do not use a random MAC address for the GAS exchange that is used as
part of the DPP protocol exchange since that would break DPP.
Configurator expects the same MAC address to be used for DPP
Authentication exchange and DPP Configuration exchange (GAS).
Since the DPP Authentication exchange has already used the MAC address
configured on the interface, use of a random address for the GAS
exchange would not provide any additional privacy protection. If a
random MAC address needs to be used for this type of an exchange, that
random address would need to be first configured on the interface before
starting DPP exchange.
This does not change GAS query behavior for any other use case, i.e.,
the gas_rand_mac_addr configuration continues to apply to all the
Interworking/Hotspot 2.0 uses.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
$(LIBS) got included twice into the linker command line from $^ and
$(LIBS). The former ended up getting converted based on host CPU rather
than target. Fix that by removing duplication and -lcrypto from $(LIBS).
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
We have practically started requiring some C99 features, so might as
well finally go ahead and bring in the C99 bool as well.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
EVP_PKEY_get0_EC_KEY() compatibility wrapper is used only within
CONFIG_DPP2 blocks, so define it with matching condition.
Signed-off-by: Jouni Malinen <j@w1.fi>
The static function hostapd_ctrl_check_freq_params() was called only
within #ifdef NEED_AP_MLME block so the function needs to be defined
under matching condition.
Signed-off-by: Jouni Malinen <j@w1.fi>
This has no changes to the function itself, i.e., it is just moved to a
location that is above the first caller.
Signed-off-by: Jouni Malinen <j@w1.fi>
If the driver supports control port for EAPOL RX, do not register
l2_packet for RX to minimize unnecessary operation load.
Signed-off-by: Jouni Malinen <j@w1.fi>
drv->eapol_sock is used only for receiving EAPOL frames in AP mode, so
it is not needed when using control port for EAPOL frame RX.
Signed-off-by: Jouni Malinen <j@w1.fi>
The eloop_register_read_sock() call in i802_init() will be skipped if
the driver supports control port for EAPOL RX, so need to skip this part
of the test case conditionally.
Signed-off-by: Jouni Malinen <j@w1.fi>
There is no need for this to be outside the if block that is used to
check whether l2_packet is used for the interface.
Signed-off-by: Jouni Malinen <j@w1.fi>
This allows wpa_supplicant to avoid the packet socket workaround for
cases where a bridge interface is used if the driver supports control
port for EAPOL frame TX and RX.
Signed-off-by: Jouni Malinen <j@w1.fi>
Since this nl80211 mechanism for sending EAPOL frames does not currently
support TX status notification, disable it by default of AP mode where
the Authenticator state machine uses those notifications to optimize
retransmission. The control port TX can be enabled for AP mode with
driver param control_port_ap=1.
Signed-off-by: Jouni Malinen <j@w1.fi>
Use and/or set socker owner where necessary to allow EAPOL frames to be
received using the nl80211 control port. This is done when the driver
indicates support for the control port without previous hardcoded
reception of RSN preauth frames.
Use methods which set or use the connection owner nl_sock * where
necessary. Initial operations need to register with the SOCKET_OWNER
attribute set (e.g., connect for STA mode). Final operations need to use
the socket which holds the owner attribute (e.g., disconnect for STA
mode).
Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
This commit creates a connect nl_sock * for every bss unconditionally.
It is used in the next commit for nl80211 control port RX.
Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
nl80211 control port event handling previously did not differentiate
between EAPOL and RSN preauth ethertypes. Add checking of the ethertype
and report unexpected frames (only EAPOL frames are supposed to be
delivered through this path).
Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
This is needed since the initial implementation of the control port
mechanism in the kernel mixed in RSN pre-authentication ethertype
unconditionally (and IMHO, incorrectly) into the control port.
Signed-off-by: Jouni Malinen <j@w1.fi>
