hapd->conf might be NULL in case initialized failed, so better be
prepared for that when debug printing interface name in the deinit path.
hostapd.c:312:54: runtime error: member access within null pointer of type 'struct hostapd_bss_config'
hostapd.c:351:29: runtime error: member access within null pointer of type 'struct hostapd_bss_config'
hostapd.c:2158:18: runtime error: member access within null pointer of type 'struct hostapd_bss_config'
Signed-off-by: Jouni Malinen <j@w1.fi>
Add a constraint on the base64 encoded buffer length to avoid an integer
overflow in the output length calculation.
common.c:1087:16: runtime error: unsigned integer overflow: 0 - 1 cannot be represented in type 'size_t' (aka 'unsigned long')
Signed-off-by: Jouni Malinen <j@w1.fi>
Split the if/while loop condition into two independent steps so that
in_size-- happens only in the case in_size is nonzero. This gets rid of
unnecessary UBSan warnings.
common.c:1087:16: runtime error: unsigned integer overflow: 0 - 1 cannot be represented in type 'size_t' (aka 'unsigned long')
common.c:1076:16: runtime error: unsigned integer overflow: 0 - 1 cannot be represented in type 'size_t' (aka 'unsigned long')
common.c:1119:16: runtime error: unsigned integer overflow: 0 - 1 cannot be represented in type 'size_t' (aka 'unsigned long')
Signed-off-by: Jouni Malinen <j@w1.fi>
This results in an UBSan warning that can be avoided easily.
os_unix.c:524:3: runtime error: null pointer passed as argument 2, which is declared to never be null
Signed-off-by: Jouni Malinen <j@w1.fi>
ext_supp_rates_len would be 0 here, so decrementing it by 2 will result
in unsigned integer overflow even if that result is not actually used
anywhere. Avoid that to get rid of the UBSan warning.
tdls.c:1597:27: runtime error: unsigned integer overflow: 0 - 2 cannot be represented in type 'unsigned long'
Signed-off-by: Jouni Malinen <j@w1.fi>
This is needed to avoid an UBSan warning and since this struct is used
as part of a message construction, it needs to be packed anyway to
guarantee correct functionality.
ieee802_1x_kay.c:1021:3: runtime error: member access within misaligned address 0x0000031921e2 for type 'struct ieee802_1x_mka_peer_id', which requires 4 byte alignment
Signed-off-by: Jouni Malinen <j@w1.fi>
iface->num_bss is unsigned integer, so need to explicit typecast it to
unsigned before decrementation by one even when the result is stored in
an unsigned integer.
../src/ap/hostapd.c:2185:26: runtime error: unsigned integer overflow: 0 - 1 cannot be represented in type 'unsigned long'
Signed-off-by: Jouni Malinen <j@w1.fi>
UBSan testing with WPA_TRACE=y ended up hitting an unaligned access for
struct os_alloc_trace in os_program_deinit() because of the
dl_list_for_each() design that looked like dereferencing the member
element of the list head which is something that does not exist.
Get the first entry from the list using dl_list_first() so that the
empty list special case is covefred and compare item pointers instead of
struct dl_list pointers to check whether the end of the loop has been
reached.
Signed-off-by: Jouni Malinen <j@w1.fi>
This can result in compiler warnings due to the unexpected NULL pointer
as a source memory even when the length of the copied data is 0.
Signed-off-by: Jouni Malinen <j@w1.fi>
According to IEEE Std 802.11-2016, 9.4.2.25 when fields of an RSNE are
not included, the default values are used. The cipher suite defaults
were hardcoded to CCMP in the previous implementation, but the default
is actually different for DMG: GCMP (per 9.4.2.25.2).
It is not possible to find out from the RSNE if the network is non-DMG
or DMG, so callers of wpa_parse_wpa_ie_rsn() need to handle this case
based on context, which can be different for each caller.
In order to fix this issue, add flags to the wpa_ie_data indicating
whether pairwise/group ciphers were included in the RSNE. Callers can
check these flags and fill in the appropriate ciphers. The
wpa_parse_wpa_ie_rsn() function still initializes the ciphers to CCMP by
default so existing callers will not break. This change also fixes some
callers which need to handle the DMG network case.
Signed-off-by: Lior David <liord@codeaurora.org>
This new QCA vendor attribute adds provision to specify the
ethernet protocol id from userspace to the packets which are
offloaded to the driver/firmware (e.g., IPv4, IPv6).
Signed-off-by: Arun Kumar Khandavalli <akhandav@codeaurora.org>
The Wi-Fi Alliance Multi-AP Specification v1.0 allows onboarding of a
backhaul STA through WPS. To enable this, the WPS Registrar offers a
different set of credentials (backhaul credentials instead of fronthaul
credentials) when the Multi-AP subelement is present in the WFA vendor
extension element of the WSC M1 message.
Add new configuration options to specify the backhaul credentials for
the hostapd internal registrar: multi_ap_backhaul_ssid,
multi_ap_backhaul_wpa_psk, multi_ap_backhaul_wpa_passphrase. These are
only relevant for a fronthaul SSID, i.e., where multi_ap is set to 2 or
3. When these options are set, pass the backhaul credentials instead of
the normal credentials when the Multi-AP subelement is present.
Ignore the Multi-AP subelement if the backhaul config options are not
set. Note that for an SSID which is fronthaul and backhaul at the same
time (i.e., multi_ap == 3), this results in the correct credentials
being sent anyway.
The security to be used for the backaul BSS is fixed to WPA2PSK. The
Multi-AP Specification only allows Open and WPA2PSK networks to be
configured. Although not stated explicitly, the backhaul link is
intended to be always encrypted, hence WPA2PSK.
To build the credentials, the credential-building code is essentially
copied and simplified. Indeed, the backhaul credentials are always
WPA2PSK and never use per-device PSK. All the options set for the
fronthaul BSS WPS are simply ignored.
Signed-off-by: Davina Lu <ylu@quantenna.com>
Signed-off-by: Igor Mitsyanko <igor.mitsyanko.os@quantenna.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Marianna Carrera <marianna.carrera.so@quantenna.com>
The Wi-Fi Alliance Multi-AP Specification v1.0 allows onboarding of a
backhaul STA through WPS. To enable this, the backhaul STA needs to add
a Multi-AP IE to the WFA vendor extension element in the WSC M1 message
that indicates it supports the Multi-AP backhaul STA role. The Registrar
(if it support Multi-AP onboarding) will respond to that with a WSC M8
message that also contains the Multi-AP IE, and that contains the
credentials for the backhaul SSID (which may be different from the SSID
on which WPS is performed).
Introduce a new parameter to wpas_wps_start_pbc() and allow it to be
set via control interface's new multi_ap=1 parameter of WPS_PBC call.
multi_ap_backhaul_sta is set to 1 in the automatically created SSID.
Thus, if the AP does not support Multi-AP, association will fail and
WPS will be terminated.
Only wps_pbc is supported.
This commit adds the multi_ap argument only to the control socket
interface, not to the D-Bus interface.
Since WPS associates with the fronthaul BSS instead of the backhaul BSS,
we should not drop association if the AP announces fronthaul-only BSS.
Still, we should only do that in the specific case of WPS. Therefore,
add a check to multi_ap_process_assoc_resp() to allow association with a
fronthaul-only BSS if and only if key_mgmt contains WPS.
Signed-off-by: Davina Lu <ylu@quantenna.com>
Signed-off-by: Igor Mitsyanko <igor.mitsyanko.os@quantenna.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Cc: Marianna Carrera <marianna.carrera.so@quantenna.com>
The Multi-AP specification adds a new subelement to the WFA extension
element in the WPS exchange. Add an additional parameter to
wps_build_wfa_ext() to add this subelement. The subelement is only added
if the parameter is nonzero. Note that we don't reuse the existing
MULTI_AP_SUB_ELEM_TYPE definition here, but rather define a new
WFA_ELEM_MULTI_AP, to make sure the enum of WFA subelement types for WPS
vendor extension remains complete.
For now, all callers set the multi_ap_subelem parameter to 0.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
The Multi-AP specification only specifies that information elements have
to be added to the Association Request and Association Response frame;
it doesn't specify anything about what should be done in case they are
missing. Previously, we rejected non-backhaul associations on a
backhaul-only BSS, and non-fronthaul associations on a fronthaul-only
BSS.
However, this makes WPS fail when fronthaul and backhaul are separate
SSIDs. Indeed, WPS for the backhaul link is performed on the *fronthaul*
SSID. Thus, the Association Request frmae used for WPS *will* contain
the Multi-AP IE indicating a backhaul STA. Rejecting that association
makes WPS fail.
Therefore, accept a multi-AP backhaul STA Association Request frame on a
fronthaul-only BSS. Still issue a warning about it, but only at level
DEBUG intead of INFO. Also change the condition checking to make it
clearer.
While we're at it, also fix the handling of unexpected bits in the
Multi-AP IE. 4 bits are reserved in the specification, so these
certainly have to be ignored. The specification also doesn't say that
setting one of the other bits is not allowed. Therefore, only report
unexpected values in the Multi-AP IE, don't reject because of it. Note
that a malformed IE (containing more than one byte) still triggers a
rejection.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
str_token() can return NULL for the name if the remaining token contains
only the delimiter. Fix this to avoid NULL pointer dereference with a
corner case of an invalid value used in the configuration.
Fixes: ec5c39a557 ("AP: Allow identifying which passphrase station used with wpa_psk_file")
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
The new sae_password parameter [|vlanid=<VLAN ID>] can now be used to
assign stations to a specific VLAN based on which SAE Password
Identifier they use. This is similar to the WPA2-Enterprise case where
the RADIUS server can assign stations to different VLANs and the
WPA2-Personal case where vlanid parameter in wpa_psk_file is used.
Signed-off-by: Jouni Malinen <j@w1.fi>
This speeds up P2P responses to frames received on an operating channel
in case there is an ongoing P2P listen operation on another channel.
This is applicable to drivers that support multiple channels in
concurrently.
This addresses an issue showing up in the
p2ps_channel_active_go_and_station_different_mcc test case where the
Provision Discovery Request frame can be received on the operating
channel of a group instead of the Listen channel. The response was
delayed until the listen operation timed out and this took too long time
for the peer to receive the response.
Signed-off-by: Jouni Malinen <j@w1.fi>
Allow user space applications to check whether wlan firmware is alive
through an nl80211 vendor command.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Process NL80211_ATTR_REQ_IE from the NL80211_CMD_ASSOCIATE event to
allow request IEs to be made available for the SME-in-wpa_supplicant
case similarly to how this is done with SME-in-driver with
NL80211_CMD_CONNECT.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Extend wpa_psk_file to allow an optional VLAN ID to be specified with
"vlanid=<VLAN ID>" prefix on the line. If VLAN ID is specified and the
particular wpa_psk_file entry is used for a station, that station is
bound to the specified VLAN. This can be used to operate a single
WPA2-Personal BSS with multiple VLANs based on the used passphrase/PSK.
This is similar to the WPA2-Enterprise case where the RADIUS server can
assign stations to different VLANs.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
If the ap_handle_timer() timeout is reached for a not-associated STA, do
not default to disassociating that STA first since Disassociation frame
is not really appropriate to send to a STA that is not in associated
state. Instead, skip directly to deauthentication and STA entry removal.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
The flags variable needs to be initialized to 0 if check_crl is 0 in
the updated configuration.
Fixes: 159a7fbdea ("crl_reload_interval: Add CRL reloading support")
Signed-off-by: Jouni Malinen <j@w1.fi>
The change is bigger because here we need to catch the error
condition if the last element doesn't fit.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Rather than always iterating elements from frames with pure
u8 pointers, add a type "struct element" that encapsulates
the id/datalen/data format of them.
Then, add the element iteration macros
* for_each_element
* for_each_element_id
* for_each_element_extid
which take, as their first 'argument', such a structure and
iterate through a given u8 array interpreting it as elements.
While at it also add
* for_each_subelement
* for_each_subelement_id
* for_each_subelement_extid
which instead of taking data/length just take an outer element
and use its data/datalen.
Also add for_each_element_completed() to determine if any of
the loops above completed, i.e., it was able to parse all of
the elements successfully and no data remained.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Add test-eapol program that can be used for fuzzing the EAPOL-Key
Supplicant and Authenticator implementations. This tool can write
Supplicant or Authenticator messages into a file as an initialization
step and for the fuzzing step, that file (with potential modifications)
can be used to replace the internally generated message contents.
The TEST_FUZZ=y build parameter is used to make a special build where a
hardcoded random number generator and hardcoded timestamp are used to
force deterministic behavior for the EAPOL-Key operations. This will
also make the implementation ignore Key MIC and AES keywrap errors to
allow processing of modified messages to continue further.
Signed-off-by: Jouni Malinen <j@w1.fi>
There is no need to schedule the postponed RSN preauthentication start
if there are no candidates. Avoid wasting eloop resources for this.
This is most useful for fuzz testing of the 4-way handshake
implementation to avoid getting stuck waiting for this unnecessary one
second time when using eloop to coordinate the Authenticator and
Supplicant state machines.
Signed-off-by: Jouni Malinen <j@w1.fi>
If none of the supported name attributes are present, the name string
was nul terminated only at the end. Add an explicit nul termination at
the end of the last written (or beginning of the buffer, if nothing is
written) to avoid writing uninitialized data to debug log.
Signed-off-by: Jouni Malinen <j@w1.fi>
Explicitly check the remaining buffer length before trying to read the
ASN.1 header values. Attempt to parse an ASN.1 header when there was not
enough buffer room for it would have started by reading one or two
octets beyond the end of the buffer before reporting invalid data at the
following explicit check for buffer room.
Signed-off-by: Jouni Malinen <j@w1.fi>
tlsv1_record_receive() did not return error here and as such, &alert was
not set and must not be used. Report internal error instead to avoid use
of uninitialized memory.
Signed-off-by: Jouni Malinen <j@w1.fi>
Add test-tls program that can be used for fuzzing the internal TLS
client and server implementations. This tool can write client or server
messages into a file as an initialization step and for the fuzzing step,
that file (with potential modifications) can be used to replace the
internally generated message contents.
The TEST_FUZZ=y build parameter is used to make a special build where a
hardcoded random number generator and hardcoded timestamp are used to
force deterministic behavior for the TLS operations.
Signed-off-by: Jouni Malinen <j@w1.fi>
conn->cred might be NULL here, so check for that explicitly before
checking whether conn->cred->cert_probe is set. This fixes a potential
NULL pointer dereference when going through peer certificates with
event_cb functionality enabled.
Signed-off-by: Jouni Malinen <j@w1.fi>
Print the SSID with printf escaping instead of wpa_hexdump_ascii()
format to clean up the debug log a bit. This was already done for number
of SSID debug prints.
Signed-off-by: Jouni Malinen <j@w1.fi>
p2p->find_start timer was updated on each p2p_find call irrespective of
p2p_find being successful/failed/rejected. For cases where p2p_find was
in progress/pending, another call to p2p_find would be rejected but
p2p->find_start timer would still be updated.
p2p->find_start is maintained in wpa_supplicant to reject the kernel
scan entries before the p2p->find_start time. In above scenario, some of
the scan entries could be discarded even if the Probe Respons frame(s)
were received during the last scan/p2p_find.
This commit changes this to update the p2p->find_start timer only when
call to p2p_find is successful, i.e., a new scan is actually started.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Do not use a separate enum for MBO WNM-Notification Request frame
subtype values since these share the same number space with the Hotspot
2.0 ones.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
The previously used value 2 was already assigned for another purpose
(MBO non-preferred channel report), so the newer T&C Acceptable
definition needs to be updated with a unique value.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This command has now been extended to include PMK for offload needs, so
the message buffer needs to be cleared explicitly after use to avoid
leaving such material in heap memory unnecessarily.
Fixes: 061a3d3d53 ("nl80211: Add support for FILS Cache Identifier in add/remove_pmkid()")
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This command can include keys (WEP or PSK for offload), so the message
buffer needs to be cleared explicitly after use to avoid leaving such
material in heap memory unnecessarily.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
We do not need such payload in the acknowledgment, so adding it uses
resources unnecessarily. Furthermore, the original request can include
key material (e.g., NL80211_ATTR_PMK). libnl does not explicitly clear
this received message buffer and it would be inconvenient for
wpa_supplicant/hostapd to try to clear it with the current libnl design
where a duplicated buffer is actually passed to the callback. This means
that keys might be left unnecessarily in heap memory. Avoid this by
requesting the kernel not to copy back the request payload.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
These buffers in TLS-based EAP methods might contain keys or password
(e.g., when using TTLS-PAP or PEAP-GTC), so clear them explicitly to
avoid leaving such material into heap memory unnecessarily.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
The case of PEAPv0 with crypto binding did not clear some of the
temporary keys from stack/heap when those keys were not needed anymore.
Clear those explicitly to avoid unnecessary caching of keying material.
Signed-off-by: Jouni Malinen <j@w1.fi>
Derive EMSK when using EAP-PEAP to enable ERP. In addition, change the
MSK derivation for EAP-PEAP to always derive 128 octets of key material
instead of the 64 octets to cover just the MSK. This is needed with the
PRF used in TLS 1.3 since the output length is mixed into the PRF
context.
Signed-off-by: Jouni Malinen <j@w1.fi>
The avoid channels are notified through
QCA_NL80211_VENDOR_SUBCMD_AVOID_FREQUENCY allow minimal traffic, so
enhance the P2P behavior accordingly by considering these avoid
frequencies for P2P discovery/negotiation as long as they are not in
disallowed frequencies list.
Additionally, do not return failure when none of social channels are
available as operation channel, rather, mark the op_channel/op_reg_class
to 0 as this would anyway get selected during the group formation in
p2p_prepare_channel.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
ieee802_11_rx_wnmsleep_req() might have been called for a short frame
that has no more payload after the Public Action field, i.e., with len
== 0. The bounds checking for the payload length was done only for the
information elements while the one octet Dialog Token field was read
unconditionally. In the original implementation, this could have
resulted in reading one octet beyond the end of the received frame data.
This case has not been reachable after the commit e0785ebbbd ("Use
more consistent Action frame RX handling in both AP mode paths"), but it
is better to address the specific issue in ieee802_11_rx_wnmsleep_req()
as well for additional protection against accidential removal of the
check and also to have something that can be merged into an older
version (pre-v2.7) if desired. The comments below apply for such older
versions where the case could have been reachable.
Depending on driver interface specific mechanism used for fetching the
frame, this could result in reading one octet beyond the end of a
stack/hash buffer or reading an uninitialized octet from within a
buffer. The actual value that was read as the Dialog Token field is not
used since the function returns immediately after having read this value
when there is no information elements following the field.
This issue was initially added in commit d32d94dbf4 ("WNM: Add
WNM-Sleep Mode implementation for AP") (with CONFIG_IEEE80211V=y build
option) and it remained in place during number of cleanup and fix
changes in this area and renaming of the build parameter to
CONFIG_WNM=y. The impacted function was not included in any default
build without one of the these optional build options being explicitly
enabled. CONFIG_WNM=y is still documented as "experimental and not
complete implementation" in hostapd/defconfig. In addition, commit
114f2830d2 ("WNM: Ignore WNM-Sleep Mode Request in wnm_sleep_mode=0
case") made this function exit before the impact read if WNM-Sleep Mode
support was not explicitly enabled in runtime configuration
(wnm_sleep_mode=1 in hostapd.conf). Commit e0785ebbbd ("Use more
consistent Action frame RX handling in both AP mode paths") made this
code unreachable in practice.
Add an explicit check that the frame has enough payload before reading
the Dialog Token field in ieee802_11_rx_wnmsleep_req().
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This patch adds a new flag 'crl_reload_interval' to reload CRL
periodically. This can be used to reload ca_cert file and the included
CRL information on every new TLS session if difference between the last
reload and the current time in seconds is greater than
crl_reload_interval.
This reloading is used for cases where check_crl is 1 or 2 and the CRL
is included in the ca_file.
Signed-off-by: Paresh Chaudhary <paresh.chaudhary@rockwellcollins.com>
Signed-off-by: Jared Bents <jared.bents@rockwellcollins.com>
It is now possible to optionally specify keyid for
each wpa_psk_file entry:
keyid=something 00:00:00:00:00:00 secretpassphrase
When station connects and the passphrase it used
has an associated keyid it will be appended to the
AP-STA-CONNECTED event string:
wlan0: AP-STA-CONNECTED 00:36:76:21:dc:7b keyid=something
It's also possible to retrieve it through the control interface:
$ hostapd_cli all_sta
Selected interface 'ap0'
00:36:76:21:dc:7b
...
keyid=something
New hostapd is able to read old wpa_psk_file. However, old hostapd will
not be able to read the new wpa_psk_file if it includes keyids.
Signed-off-by: Michal Kazior <michal@plume.com>
This doesn't change any behavior on its own. It's going to be used to
expose per-station keyids and allow reloading passphrases in runtime.
Signed-off-by: Michal Kazior <michal@plume.com>
Support the new Extended Capabilities field bits 81 and 82 to indicate
whether SAe Password Identifiers are in use.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This new QCA vendor attribute indicates the EVM value in netlink.
Signed-off-by: stonez <stonez@codeaurora.org>
:100644 100644 ad5dac2... ede4fc8... M src/common/qca-vendor.h
When using LibreSSL build fails with:
../src/crypto/tls_openssl.o: in function `tls_connection_client_cert':
../src/crypto/tls_openssl.c:2817: undefined reference to `SSL_use_certificate_chain_file'
collect2: error: ld returned 1 exit status
make: *** [Makefile:1901: wpa_supplicant] Error 1
There is no such function in LibreSSL.
Signed-off-by: Stefan Strogin <stefan.strogin@gmail.com>
An optional parameter "he" is added to p2p_connect, p2p_group_add, and
p2p_invite to enable 11ax HE support. The new p2p_go_he=1 configuration
parameter can be used to request this to be enabled by default.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
The vendor command QCA_NL80211_VENDOR_SUBCMD_AVOID_FREQUENCY was defined
to carry the list of avoid frequencies that aim to avoid any
interference with other coexistencies. This recommendation was followed
strictly by trying to prevent WLAN traffic on the impacted channels.
This commit refines the expectation of the interface by defining this
avoid channel list to allow minimal traffic but not heavier one. For
example, P2P may still be able to use avoid list frequencies for P2P
discovery and GO negotiation if the actual group can be set up on a not
impact channel.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Add support for configuring parameters for the MU EDCA Parameter Set
element per IEEE P802.11ax/D3.0.
Signed-off-by: Siva Mullati <siva.mullati@intel.com>
If the libnl version is not specified explicitly with CONFIG_LIBNL*, try
to check for the most likely case today with pkg-config.
Signed-off-by: Jouni Malinen <j@w1.fi>
Fix compilation issue if we want to build wpa_supplicant without any
wireless connectivity but only with MACSec support via Linux kernel
driver.
Signed-off-by: Andrey Kartashev <a.s.kartashev@gmail.com>
Two recent changes to MKA create a situation where a new MI is generated
every time a SAK Use parameter set is decoded. The first change moved
invalid key detection from ieee802_1x_decode_basic_body() to
ieee802_1x_kay_decode_mpkdu():
commit db9ca18bbf ("mka: Do not ignore MKPDU parameter set decoding failures")
The second change forces the KaY to generate a new MI when an invalid
key is detected:
commit a8aeaf41df ("mka: Change MI if key invalid")
The fix is to move generation of a new MI from the old invalid key
detection location to the new location.
Fixes: a8aeaf41df ("mka: Change MI if key invalid")
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
Upon issuing a connect request we need to indicate that we want the
driver to offload the 802.1X 4-way handshake for us. Indicate it if
the driver capability supports the offload.
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Allow drivers to indicate support for offloading 4-way handshake for
either IEEE 802.1X (WPA2-Enterprise; EAP) and/or WPA/WPA2-PSK
(WPA2-Personal) by splitting the WPA_DRIVER_FLAGS_4WAY_HANDSHAKE flag
into two separate flags.
Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com>
When an interface is re-enabled after it was disabled during CAC, it
won't ever get active since hostapd is waiting for a CAC_FINISHED while
kernel side is waiting for a CMD_RADAR_DETECT to start a CAC.
This commit checks for a pending CAC when an interface is enabled and if
so restarts its DFS processing.
Signed-off-by: Zefir Kurtisi <zefir.kurtisi@neratec.com>
In the previous RADIUS client implementation, when there are multiple
RADIUS servers, we kept trying the next server when the current message
can not be acked. It leads to endless retry when all the RADIUS servers
are down.
Fix this by keeping a counter for the accumulated retransmit attempts
for the message, and guarantee that after all the servers failover
RADIUS_CLIENT_MAX_FAILOVER times the message will be dropped.
Another issue with the previous code was that the decision regarding
whether the server should fail over was made immediately after we send
out the message. This patch guarantees we consider whether a server
needs failover after pending ack times out.
Signed-off-by: Bo Chen<bochen@meraki.com>
Define QCA vendor command attributes to configure HE +HTC support and
HE operating mode control transmission. This is used to configure the
testbed device.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
These are not allowed in ISO C++ (and well, not really in ISO C either,
but that does not result in compiler warning without pedantic
compilation).
Since ieee802_11_common.h may end up getting pulled into C++ code for
some external interfaces, it is more convenient to keep it free of these
cases. Pull in ieee802_11_defs.h to get enum phy_type defined and move
enum chan_width to common/defs.h (which was already pulled in into
src/drivers/driver.h and src/common/ieee802_11_common.h).
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
While the AP is configured to enable both FT-PSK and WPA-PSK, an HP
printer request both AKMs (copied from AP?) in Association Request
frame, but don't add MDIE and don't use FT. This results in the
connection failing.
Next in logs we see:
RSN: Trying to use FT, but MDIE not included
IE - hexdump(len=26): 30 18 01 00 00 0f ac 04 01 00 00 0f ac 04
02 00 00 0f ac 02 00 0f ac 04 00 00
This is seen with some HP and Epson printers. Work around this by
stripping FT AKM(s) when MDE is not present and there is still a non-FT
AKM available.
Signed-off-by: Janusz Dziedzic <janusz@plumewifi.com>
The Android-specific chmod and chown operations on the client socket
(for communication with wpa_supplicant) did not protect against file
replacement between the bind() and chmod()/chown() calls. If the
directory in which the client socket is created (depends a bit on the
version and platform, but /data/misc/wifi/sockets is commonly used)
allows write access to processes that are different (less privileged)
compared to the process calling wpa_ctrl_open2(), it might be possible
to delete the socket file and replace it with something else (mainly, a
symlink) before the chmod/chown operations occur. This could have
resulted in the owner or permissions of the target of that symlink being
modified.
In general, it would be safest to use a directory which has more limited
write privileges (/data/misc/wifi/sockets normally has 'wifi' group
(AID_WIFI) with write access), but if that cannot be easily changed due
to other constraints, it is better to make wpa_ctrl_open2() less likely
to enable this type of race condition between the operations.
Replace chown() with lchown() (i.e., a version that does not dereference
symlinks) and chmod() with fchmod() on the socket before the bind() call
which is also not going to dereference a symlink (whereas chmod()
would). lchown() is a standard operation, but the fchmod() on the socket
is less so (unspecified behavior in some systems). However, it seems to
work on Linux and in particular, on Android, where this code is
executed.
Signed-off-by: Jouni Malinen <j@w1.fi>
While selecting a new channel as a reaction to radar event we need to
take into account supported bandwidth for each channel provided via
nl80211. Without this modification hostapd might select an unsupported
channel that would fail during AP startup.
Signed-off-by: Dmitry Lebed <dlebed@quantenna.com>
While doing automatic channel selection we need to take into account
supported bandwidth for each channel provided via nl80211. Without this
modification hostapd might select an unsupported channel which would
fail during AP startup.
Signed-off-by: Dmitry Lebed <dlebed@quantenna.com>
This adds checks to common code to verify supported bandwidth options
for each channel using nl80211-provided info. No support of additional
modes is added, just additional checks. Such checks are needed because
driver/hardware can declare more strict limitations than declared in the
IEEE 802.11 standard. Without this patch hostapd might select
unsupported channel and that will fail because Linux kernel does check
channel bandwidth limitations.
Signed-off-by: Dmitry Lebed <dlebed@quantenna.com>
Add NL80211_FREQUENCY_ATTR_NO_* channel attributes parsing. This is
needed for correct checking if channel is available in a particular
bandwidth.
Signed-off-by: Dmitry Lebed <dlebed@quantenna.com>
Add support for 160 MHz BW channels to automatic channel selection
algorithm. Only 36 and 100 channels are supported as 160 MHz channels.
Signed-off-by: Dmitry Lebed <lebed.dmitry@gmail.com>
Move to the version used in draft-ietf-emu-eap-tls13-03.txt, i.e.,
include the 0x0D prefix and use a different TLS-Exporter() label string.
Signed-off-by: Jouni Malinen <j@w1.fi>
Some distributions (e.g., Debian) have started introducting systemwide
OpenSSL policies to disable older protocol versions and ciphers
throughout all programs using OpenSSL. This can result in significant
number of interoperability issues with deployed EAP implementations.
Allow explicit wpa_supplicant (EAP peer) and hostapd (EAP server)
parameters to be used to request systemwide policies to be overridden if
older versions are needed to be able to interoperate with devices that
cannot be updated to support the newer protocol versions or keys. The
default behavior is not changed here, i.e., the systemwide policies will
be followed if no explicit override configuration is used. The overrides
should be used only if really needed since they can result in reduced
security.
In wpa_supplicant, tls_disable_tlsv1_?=0 value in the phase1 network
profile parameter can be used to explicitly enable TLS versions that are
disabled in the systemwide configuration. For example,
phase1="tls_disable_tlsv1_0=0 tls_disable_tlsv1_1=0" would request TLS
v1.0 and TLS v1.1 to be enabled even if the systemwide policy enforces
TLS v1.2 as the minimum version. Similarly, openssl_ciphers parameter
can be used to override systemwide policy, e.g., with
openssl_ciphers="DEFAULT@SECLEVEL=1" to drop from security level 2 to 1
in Debian to allow shorter keys to be used.
In hostapd, tls_flags parameter can be used to configure similar
options. E.g., tls_flags=[ENABLE-TLSv1.0][ENABLE-TLSv1.1]
Signed-off-by: Jouni Malinen <j@w1.fi>
TLS v1.3 was already disabled by default for EAP-FAST, EAP-TTLS,
EAP-PEAP, and EAP-TLS, but the unauthenticated client cases of EAP-TLS
-like functionality (e.g., the one used in OSEN) were missed. Address
those EAP types as well in the same way of disabling TLS v1.3 by default
for now to avoid functionality issues with TLS libraries that enable TLS
v1.3 by default.
Signed-off-by: Jouni Malinen <j@w1.fi>
SSL_use_certificate_chain_file() was added in OpenSSL 1.1.0, so need to
maintain the old version using SSL_use_certificate_file() for backwards
compatibility.
Fixes: 658c39809b ("OpenSSL: Load chain certificates from client_cert file")
Signed-off-by: Jouni Malinen <j@w1.fi>
FT-over-the-DS has a special case where the STA entry (and as such, the
TK) has not yet been configured to the driver depending on which driver
interface is used. For that case, allow add-STA operation to be used
(instead of set-STA). This is needed to allow mac80211-based drivers to
accept the STA parameter configuration. Since this is after a new
FT-over-DS exchange, a new TK has been derived after the last STA entry
was added to the driver, so key reinstallation is not a concern for this
case.
Fixes: 0e3bd7ac68 ("hostapd: Avoid key reinstallation in FT handshake")
Signed-off-by: Jouni Malinen <j@w1.fi>
The documentation in the hostapd.conf file says that the dynamic_vlan
variable is used to control whether VLAN assignments are accepted from a
RADIUS server. The implication seems to be that a static VLAN assignment
will come from the accept_mac_file if dynamic_vlan is set to 0, and a
dynamic assignment will come from the RADIUS server if dynamic_vlan is
set to 1. Instead, I'm seeing that the static settings from the
accept_mac_file are ignored if dynamic_vlan is set to 0, but used if
dynamic_vlan is set to 1. If dynamic_vlan is set to 1 and the RADIUS
server does not provide a VLAN, then the accept_mac_file assignment is
overridden and the STA is assigned to the default non-VLANed interface.
If my understanding of the expected behavior is correct, then I believe
the problem is in ap_sta_set_vlan(). That routine checks the
dynamic_vlan setting, but has no way of determining whether the incoming
vlan_desc is static (i.e., from accept_mac_file) or dynamic (i.e., from
a RADIUS server).
I've attached a patch that gets hostapd working as I believe it's meant
to, and updates the documentation to make the implicit behavior
explicit.
The functional changes are:
- hostapd_allowed_address() will always extract the vlan_id from the
accept_macs file. It will not update the vlan_id from the RADIUS cache
if dynamic_vlan is DISABLED.
- hostapd_acl_recv_radius() will not update the cached vlan_id if
dynamic_vlan is DISABLED.
- ieee802_1x_receive_auth() will not update the vlan_id if dynamic_vlan
is DISABLED.
More cosmetic:
Most of the delta is just moving code out of ieee802_1x_receive_auth()
into a new ieee802_1x_update_vlan() routine. While I initially did this
because the new DISABLED check introduced excessive indentation, it has
the added advantage of eliminating the vlan_description allocation and
os_memset() call for all DYNAMIC_VLAN_DISABLED configs.
I've done a couple rounds of review offline with Michael Braun (who has
done much of the work in this part of the code) and incorporated his
feedback.
If dynamic_vlan=0 (disabled), vlan assignments will be managed using the
local accept_mac_file ACL file, even if a RADIUS server is being used
for user authentication. This allows us to manage users and devices
independently.
Signed-off-by: Nils Nieuwejaar <nils.nieuwejaar@gmail.com>
The new ieee802_11_ext_capab() and wpa_bss_ext_capab() functions can be
used to check whether a specific extended capability bit is set instead
of having to implement bit parsing separately for each need.
Signed-off-by: Jouni Malinen <j@w1.fi>
The PADDING array used when adding padding bits in MD4 never change
so can be made const. Making it const puts the array in .rodata
section and can save a few bytes of RAM for systems running without
virtual memory.
Signed-off-by: Mikael Kanstrup <mikael.kanstrup@sony.com>
After performing a successful channel switch, the AP should update its
own neighbor report element, so do this from src/ap/drv_callbacks.c
after a successful switch.
Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
Move functions corresponding to neighbor report elements to
src/ap/neighbor_db.[c,h] in preparation to using them after channel
switch from src/ap/drv_callbacks.c.
Signed-off-by: Markus Theil <markus.theil@tu-ilmenau.de>
The function sha512_compress() has a local variable that consumes 640
bytes. This is very heavy for embedded devices that have limited stack
resources. Handle this by replacing the static allocation with a dynamic
one.
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
IEEE P802.11-REVmd/D2.0, 9.4.2.20.7 (Beacon request) and 9.4.2.21.7
(Beacon report) add the Last Beacon Report Indication subelement to
Beacon Request and Beacon Report elements.
Add the Last Beacon Report Indication subelement to all Beacon Report
elements if the Beacon Request indicated that this subelement is
requested.
Signed-off-by: Avraham Stern <avraham.stern@intel.com>
When the frame body subelement would cause the measurement report
element to exceed the maximum element size, the frame body subelement
used to be truncated. In addition, some elements were always truncated
in order to keep the reported frame body short (e.g. RSN IE).
Alternatively, IEEE P802.11-REVmd/D2.0, 9.4.2.21.7 extension to Beacon
reporting can be used: The frame body subelement is fragmented across
multiple beacon report elements, and the reported frame body fragment ID
subelement is added.
Use beacon report fragmentation instead of truncating the frame body
as this method gives the AP a more complete information about the
reported APs.
Signed-off-by: Avraham Stern <avraham.stern@intel.com>
EV_SET() for EV_ADD used a specific filter type, but that same filter
type was not provided to the matching EV_DELETE case. This resulted in
the kernel rejecting the deletion with "Invalid argument". Fix this by
setting the same filter type for both operations.
Fixes: f9982b3212 ("Implement kqueue(2) support via CONFIG_ELOOP_KQUEUE")
Signed-off-by: Jouni Malinen <j@w1.fi>
The previous implementation did not work if the first registered socket
had fd > 16 or if the fd was more than double the largest value used in
previous registrations. Those cases could result in too small a memory
allocation being used and writes/reads beyond the end of that buffer.
This fix is applicable to CONFIG_ELOOP_EPOLL=y and CONFIG_ELOOP_KQUEUE=y
builds.
Fixes: f0356ec85c ("eloop: Add epoll option for better performance")
Signed-off-by: Jouni Malinen <j@w1.fi>
According to random(4) manual, /dev/random is essentially deprecated on
Linux for quite some time:
"The /dev/random interface is considered a legacy interface, and
/dev/urandom is preferred and sufficient in all use cases, with the
exception of applications which require randomness during early boot
time; for these applications, getrandom(2) must be used instead, because
it will block until the entropy pool is initialized."
An attempt to use it would cause unnecessary blocking on machines
without a good hwrng even when it shouldn't be needed. Since Linux 3.17,
a getrandom(2) call is available that will block only until the
randomness pool has been seeded.
It is probably not a good default yet as it requires a fairly recent
kernel and glibc (3.17 and 2.25 respectively).
Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
In 2013 or so, IFNAME=foo was prepended to at least the Unix socket
communication from wpa_supplicant to wpa_cli. This broke the (fragile)
logic that made ping/pong work more often when wpa_supplicant is busy
sending logging info to wpa_cli.
Adding check for IFNAME=foo makes this work better.
Signed-off-by: Ben Greear <greearb@candelatech.com>
This helps the server to build the chain to trusted CA when PEM encoding
of client_cert is used with multiple listed certificates. This was
already done for the server certificate configuration, but the client
certificate was limited to using only the first certificate in the file.
Signed-off-by: Isaac Boukris <iboukris@gmail.com>
Legacy ioctl() through SIOCDEVPRIVATE are deprecated. Follow the
approach taken by bridge-utils and make use of new bridge ioctl's
whenever possible.
For example, using legacy ioctl() breaks dynamic VLAN mode on 32-bit
Linux systems running 64-bit kernels.
Signed-off-by: Sergey Matyukevich <sergey.matyukevich.os@quantenna.com>
If the internal EAP server is used instead of an external RADIUS server,
sm->identity does not get set. Use the identity from the internal EAP
server in such case to get the dot1xAuthSessionUserName value in STA MIB
information.
Signed-off-by: Jouni Malinen <j@w1.fi>
The SSL_METHOD patching hack to get proper OCSP validation for Hotspot
2.0 OSU needs cannot be used with OpenSSL 1.1.0 and newer since the
SSL_METHOD structure is not exposed anymore. Fall back to using the
incomplete CURLOPT_SSL_VERIFYSTATUS design to fix the build.
Signed-off-by: Ben Greear <greearb@candelatech.com>
SKM_sk_num() is not available anymore, so use DEFINE_STACK_OF() to get
the appropriate accessor functions.
Signed-off-by: Ben Greear <greearb@candelatech.com>
The direct ssl->ctx access are not allowed anymore in newer OpenSSL
versions, so use the SSL_get_SSL_CTX() helper for this.
Signed-off-by: Ben Greear <greearb@candelatech.com>
Try to make RSSI-based rejection of associating stations a bit less
likely to trigger false rejections by considering RSSI from the last
received Authentication frame. Association is rejected only if both the
Authentication and (Re)Association Request frames are below the RSSI
threshold.
Signed-off-by: Jouni Malinen <j@w1.fi>
An AP might reject a STA association request due to low RSSI. In such
case, the AP informs the STA the desired RSSI improvement and a retry
timeout. The STA might retry to associate even if the RSSI hasn't
improved if the retry timeout expired.
Signed-off-by: Beni Lev <beni.lev@intel.com>
An AP might refuse to connect a STA if it has a low RSSI. In such case,
the AP informs the STA with the desired RSSI delta and a retry timeout.
Any subsequent association attempt with that AP (BSS) should be avoided,
unless the RSSI level improved by the desired delta or the timeout has
expired.
Defined in Wi-Fi Alliance Optimized Connectivity Experience technical
specification v1.0, section 3.14 (RSSI-based association rejection
information).
Signed-off-by: Beni Lev <beni.lev@intel.com>
Add the ability to ignore time-based CRL errors from OpenSSL by
specifying a new configuration parameter, check_crl_strict=0.
This causes the following:
- This setting does nothing when CRL checking is not enabled.
- When CRL is enabled, "strict mode" will cause CRL time errors to not
be ignored and will continue behaving as it currently does.
- When CRL is enabled, disabling strict mode will cause CRL time
errors to be ignored and will allow connections.
By default, check_crl_strict is set to 1, or strict mode, to keep
current functionality.
Signed-off-by: Sam Voss <sam.voss@rockwellcollins.com>
Build configurations with CONFIG_TLS=internal and NEED_SHA512 failed due
to missing sha512.c file. Add that file even though this is not really
used in the currently available configuration combinations since DPP and
OWE are the only users of it and the internal crypto implementation
supports neither.
Signed-off-by: Jouni Malinen <j@w1.fi>
One of the reset_participant_mi() callers did not log the error. Make
this more consistent with the other callers.
Signed-off-by: Jouni Malinen <j@w1.fi>
One of the linux_br_del_if() calls did not log nl80211-specific entry.
Make this more consistent with the other cases even though
linux_br_add_if() function itself is logging an error in the ioctl()
failure case (but not in the interface not found case).
Signed-off-by: Jouni Malinen <j@w1.fi>
This makes it possible to use ECDSA certificates with EAP-TLS/TTLS/etc.
It should be noted that when using Suite B, different mechanism is used
to specify the allowed ECDH curves and this new parameter must not be
used in such cases.
Signed-off-by: Hristo Venev <hristo@venev.name>
Some versions of OpenSSL need server support for ECDH to be explicitly
enabled, so provide a new parameter for doing so and all
SSL_{,CTX_}set_ecdh_auto() for versions that need it to enable automatic
selection.
Signed-off-by: Hristo Venev <hristo@venev.name>
handle_dhcp() was first trying to learn the IP address of an associated
STA before doing broadcast-to-unicast conversion. This could result in
not converting some DHCPACK messages since the address learning part
aborts processing by returning from the function in various cases.
Reorder these operations to allow broadcast-to-unicast conversion to
happen even if an associated STA entry is not updated based on a
DHCPACK.
Signed-off-by: Jouni Malinen <j@w1.fi>
IEEE Std 802.1X-2010, 11.11 describes that the ICV is separate from the
parameter sets before it. Due to its convenient layout the ICV Indicator
'body part' is used to encode the ICV as well.
IEEE Std 802.1X-2010, 11.11.3 describes the encoding of MKPDUs. In
bullet e) is desribed that the ICV Indicator itself is encoded when the
ICV is not 16 octets in length. IEEE Std 802.1Xbx-2014, Table 11-7 note
e) states that it will not be encoded unless the Algorithm Agility
parameter specifies the use of an ICV that is not 16 octets in length.
Therefore the length calculation for the ICV indicator body part must
take into account if the ICV Indicator is to be encoded or not. The
actual encoder of the ICV body already takes care of the rest.
In practice, this change will remove the ICV Indicator parameter set (4
octets before the ICV value itself) since the only defined algorithm
agility value uses an ICV of 16 octets. IEEE Std 802.1X-2010 MKPDU
validation and decoding rules in 11.11.2 and 11.11.4 require the
receipient to handle both cases of ICV Indicator being included or not.
Signed-off-by: Jaap Keuter <jaap.keuter@xs4all.nl>
It was possible for a participant to first be elected as a key server
and schedule a new SAK to be generated and distributed just to be
followed by another participant being elected as the key server. That
did not stop the participant that disabled key server functionality to
stop generating the new SAK and then trying to distribute it. That is
not correct behavior, so make these steps conditional on the participant
still being a key server when going through the timer.
Signed-off-by: Jouni Malinen <j@w1.fi>
This pointer needs to be cleared when the matching SAK is being removed
from the SAK list. The previous implementation was doing something
pretty strange in the loop by clearing the pointer for any non-matching
key that happened to be iterated through before finding the matching
key. This could probably result in incorrect behavior, but not clearing
the pointer for the matching key could do more harm by causing freed
memory to be referenced.
Signed-off-by: Jouni Malinen <j@w1.fi>
Instead of using a specifically set index value from table definition,
use the actual real index of the table entry. This removes need for
maintaining these index values separately. Furthermore, the
mka_alg_tbl[] index was already off-by-one (but not used anywhere).
Signed-off-by: Jouni Malinen <j@w1.fi>
When running wpa_supplicant (with logging for testing) the log output is
somewhat disorganized for KaY related items. E.g., items are not
aligned, inconsistent type handling, wrong wording, missing labels, etc.
This change tries to clean up the log output, so it is somewhat more
accessible.
Signed-off-by: Jaap Keuter <jaap.keuter@xs4all.nl>
Go through the SM_STEP_RUN() global transition to get into the INIT
state to follow the state machine design more closely.
Signed-off-by: Jouni Malinen <j@w1.fi>
While IEEE Std 802.1X-2010 talks about arbitrary authorization data that
could be passed to the CP from sources like RADIUS server, there is not
much point in trying to implement this as an arbitrary memory buffer in
wpa_supplicant. Should such data be supported in the future, it would
much more likely use more detailed data structures that encode the
received data in easier to use form.
Signed-off-by: Jouni Malinen <j@w1.fi>
This can be used to allow 256-bit key hierarchy to be derived from
EAP-based authentication. For now, the MSK length is hardcoded to 128
bits, so the previous behavior is maintained.
Signed-off-by: Jouni Malinen <j@w1.fi>
The CAK length is not hardcoded in the algorithm agility parameter, so
remove that from the table. Instead, allow both 16 (128-bit) and 32
(256-bit) CAK to be used so that the following key derivations use
appropriate key lengths based on the configured/derived CAK.
Signed-off-by: Jouni Malinen <j@w1.fi>
The ICK and KEK are derived from a CAK and the length of the CAK
determines the length of the KCK/ICK. Remove the separate ICK/KEK length
parameters from the algorithm agility table.
Signed-off-by: Jouni Malinen <j@w1.fi>
Extend the previously implemented KDF (IEEE Std 802.1X-2010, 6.2.1) to
support 256-bit input key and AES-CMAC-256. This does not change any
actual key derivation functionality yet, but is needed as a step towards
supporting 256-bit CAK.
Signed-off-by: Jouni Malinen <j@w1.fi>
It is possible to get a situation where a peer removes the Key Server
from its live peers list but the server still thinks that the peer is
alive (e.g., high packet loss in one direction). In such a case, the Key
Server will continue to advertise Last Key but this peer will not be
able to set up SA as it has already deleted its key.
Change the peer MI which will force the Key Server to distribute a new
SAK.
Signed-off-by: Andrey Kartashev <andrey.kartashev@afconsult.com>
Decrease timeout for a peer with duplicated SCI to speed up process in
case it is a valid peer after MI change.
Signed-off-by: Andrey Kartashev <andrey.kartashev@afconsult.com>
There is already partial support of GCM-AES-256. It is possible to
enable this mode by setting 'kay->macsec_csindex = 1;' in
ieee802_1x_kay_init() function, but the generated key contained only 128
bits of data while other 128 bits are in 0.
Enables KaY to generate full 256-bit SAK from the same 128-bit CAK. Note
that this does not support 256-bit CAK or AES-CMAC-256 -based KDF.
Signed-off-by: Andrey Kartashev <andrey.kartashev@afconsult.com>
According IEEE Std 802.1X-2010, 9.8 each participant shall record the
values of NextPN for last SAK accepted from each Key Server to use it in
case of a switch from one Key Server to another and back. Add LPN
recording and set saved value as the initial PN for the created channel.
Signed-off-by: Andrey Kartashev <andrey.kartashev@afconsult.com>
It is possible that the driver fails to create Secure Channel (due to
hardware limitations for example). Add checks of create_*_sc() result
codes and abort procedure in case of failure.
Signed-off-by: Andrey Kartashev <andrey.kartashev@afconsult.com>
Fix a minor memory leak in ieee802_1x_kay_create_mka() in
case of KEK/ICK derivation failure.
Signed-off-by: Andrey Kartashev <andrey.kartashev@afconsult.com>
Add new configuration parameters macsec_replay_protect and
macsec_replay_window to allow user to set up MACsec replay protection
feature. Note that according to IEEE Std 802.1X-2010 replay protection
and delay protection are different features: replay protection is
related only to SecY and does not appear on MKA level while delay
protection is something that KaY can use to manage SecY state.
Signed-off-by: Andrey Kartashev <andrey.kartashev@afconsult.com>
When syslog logging is used output from wpa_hexdump_ascii() was silently
discarded. This patch enables wpa_hexdump_ascii() to print data to
syslog but without ASCII decoding.
Signed-off-by: Andrey Kartashev <andrey.kartashev@afconsult.com>
To prevent a remote peer from getting stuck in a perpetual 'potential
peer' state, only update the peer liveness timer 'peer->expire' for live
peers and not for potential peers.
Per IEEE Std 802.1X-2010, 9.4.3 (Determining liveness), potential peers
need to show liveness by including our MI/MN in their transmitted MKPDU
(within potential or live parameter sets).
When a potential peer does include our MI/MN in an MKPDU, we respond by
moving the peer from 'potential_peers' to 'live_peers'.
If a potential peer does not include our MI/MN in an MKPDU within
MKPDU_LIFE_TIME, let the peer expire to facilitate getting back in sync
with the remote peer.
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
The previous commit introduced parameter set error checking. This commit
extends upon that by considering missing parameter sets a failure.
Two checks are added by this commit. First, verify that live peers start
encoding MKA_SAK_USE within a reasonable amount of time after going live
(10 MKPDUs). Second, verify that once a live peer starts encoding
MKA_SAK_USE it continues to do so indefinitely.
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
The status values returned by mka_param_body_handler.body_rx functions
are currently ignored by ieee802_1x_kay_decode_mkpdu(). If a failure is
detected the KaY should (a) stop processing the MKDPU and (b) do not
update the associated peer's liveliness.
IEEE Std 802.1X-2010, Table 11-7 (MKPDU parameter sets) and 11.11.3
(Encoding MKPDUs) dictate that MKA_SAK_USE (set type 3) will always be
encoded before MKA_DISTRIBUTED_SAK (set type 4) in MKPDUs. Due to
implementation of mka_param_body_handler, the code will always decode
MKA_SAK_USE before MKA_DISTRIBUTED_SAK. When MKA_DISTRUBUTED_SAK
contains a new SAK the code should decode MKA_DISTRUBUTED_SAK first so
that the latest SAK is in known before decoding MKA_SAK_USE.
The ideal solution would be to make two passes at MKDPU decoding: the
first pass decodes MKA_DISTRIBUTED_SAK, the second pass decodes all
other parameter sets.
A simpler and less risky solution is presented here: ignore MKA_SAK_USE
failures if MKA_DISTRIBUTED_SAK is also present. The new SAK will be
saved so that the next MKPDU's MKA_SAK_USE can be properly decoded. This
is basically what the code prior to this commit was doing (by ignoring
all errors).
Also, the only real recourse the KaY has when detecting any bad
parameter set is to ignore the MKPDU by not updating the corresponding
peer's liveliness timer, 'peer->expire'.
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
If a live peer ever changes its Member Identifier (MI), the KaY
correctly detects a "duplicated SCI" but then proceeds to delete the
peer without deleting the peer's resources (i.e., RxSC, RxSAs, TxSAs).
Note that a remote peer's MI will change if and when an
ieee8021XPaePortInitialize is executed on the remote port.
The solution here is to ignore all MKPDUs containing the new MI until
after the peer (that corresponds to the old MI) expires and cleans up
its resources. After the old peer is removed reception of the next MKPDU
containing the new MI will result in the creation of a new peer with the
new MI.
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
Per IEEE Std 802.1X-2010, Figure 12-2 (CP state machine), READY to
TRANSMIT transition includes !controlledPortEnabled condition.
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
Per IEEE Std 802.1X-2010, Figure 12-2 (CP state machine), READY should
move to ABANDON (not RECEIVE) when new_sak or changed_connect is true.
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
Per IEEE Std 802.1X-2010, Figure 12-2 (CP state machine), RECEIVING to
TRANSMIT transition includes !controlledPortEnabled condition.
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
Per IEEE Std 802.1X-2010, Figure 12-2 (CP state machine), deleteSAs(oki)
is used upon entering RETIRE. Do that in addition to freeing sm->oki.
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
The purpose of the Lowest Acceptable PN (lpn) parameters in the MACsec
SAK Use parameter set is to enforce delay protection. Per IEEE Std
802.1X-2010, Clause 9, "Each SecY uses MKA to communicate the lowest PN
used for transmission with the SAK within the last two seconds, allowing
receivers to bound transmission delays."
When encoding the SAK Use parameter set the KaY should set llpn and olpn
to the lowest PN transmitted by the latest SAK and oldest SAK (if
active) within the last two seconds. Because MKPDUs are transmitted
every 2 seconds (MKA_HELLO_TIME), the solution implemented here
calculates lpn based on the txsc->next_pn read during the previous MKPDU
transmit.
Upon receiving and decoding a SAK Use parameter set with delay
protection enabled, the KaY will update the SecY's lpn if the delay
protect lpn is greater than the SecY's current lpn (which is a product
of last PN received and replay protection and window size).
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
Delay Protect and Replay Protect are two separate and distinct features
of MKA. Per IEEE Std 802.1X-2010, 9.10.1 "Delay Protect, TRUE if LPNs
are being reported sufficiently frequently to allow the recipient to
provide data delay protection. If FALSE, the LPN can be reported as
zero", and per 9.10 "NOTE--Enforcement of bounded received delay
necessitates transmission of MKPDUs at frequent (0.5 s) intervals, to
meet a maximum data delay of 2 s while minimizing connectivity
interruption due to the possibility of lost or delayed MKPDUs."
This means struct ieee802_1x_mka_sak_use_body::delay_protect should only
be set TRUE when MKPDUs are being transmitted every 0.5 s (or faster).
By default the KaY sends MKPDUs every MKA_HELLO_TIME (2.0 s), so by
default delay_protect should be FALSE.
Add a new 'u32 mka_hello_time' parameter to struct ieee802_1x_kay. If
delay protection is desired, the KaY initialization code should set
kay->mka_hello_time to MKA_BOUNDED_HELLO_TIME (500 ms).
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
Commit 7b4d546e ("wpa_supplicant: Add macsec_integ_only setting for
MKA") introduced policy setting SHOULD_ENCRYPT (MACsec provides
integrity+confidentiality) in addition to SHOULD_SECURE (MACsec provides
integrity only). In both cases the KaY is populating the
"Confidentiality Offset" parameter within the "Distributed SAK parameter
set" with CONFIDENTIALITY_OFFSET_0=1. In the case of SHOULD_SECURE the
parameter should be populated with CONFIDENTIALITY_NONE=0.
IEEE Std 802.1X-2010, Table 11-6 and Figure 11-11 define how the two
Confidentiality Offset bits in the "Distributed SAK parameter set" must
be set: "0 if confidentiality not used" and "1 if confidentiality with
no offset". When policy is SHOULD_SECURE KaY should to send the former,
and when policy is SHOULD_ENCRYPT KaY should send the latter.
Fixes: 7b4d546e3d ("wpa_supplicant: Add macsec_integ_only setting for MKA")
Signed-off-by: Michael Siedzik <msiedzik@extremenetworks.com>
IEEE Std 802.1X-2010, Figure 11-7 explains that "Parameter set body
length" is exclusive of the suffix padding.
Fix variable length encoding and decoding when CKN length is not a
multiple of 4 bytes.
Signed-off-by: Michael Braun <michael-dev@fami-braun.de>
Mark the data structures used in construction/parsing frames packed to
prevent compiler from being able to pad them.
Signed-off-by: Jouni Malinen <j@w1.fi>
wpa_supplicant STATUS-DRIVER control interface command can now be used
to fetch the macsec_linux driver status information like parent
interface name.
Signed-off-by: Jouni Malinen <j@w1.fi>
These do not really get truncated in practice, but it looks like some
newer compilers warn about the prints, so silence those by checking the
result and do something a bit more useful if the output would actually
get truncated.
Signed-off-by: Jouni Malinen <j@w1.fi>
Try to fetch the list of supported AKM suite selectors from the driver
through the vendor interface
QCA_NL80211_VENDOR_SUBCMD_GET_SUPPORTED_AKMS. If that command is
available and succeeds, use the returned list to populate the
wpa_driver_capa key_mgmt information instead of assuming all
cfg80211-based drivers support all AKMs. If the driver does not support
this command, the previous behavior is maintained.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
This new QCA vendor command is used to query the supported AKM suite
selectors from the driver. There has been no such capability indication
from the driver and thus the current user space has to assume the driver
to support all the AKMs. This may be the case with some drivers (e.g.,
mac80211-based ones) but there are cfg80211-based drivers that implement
SME and have constraints on which AKMs can be supported (e.g., such
drivers may need an update to support SAE AKM using
NL80211_CMD_EXTERNAL_AUTH). Allow such drivers to specify the exact set
of supported AKMs so that user space tools can determine what network
profile options should be allowed to be configured. This command returns
the list of supported AKM suite selectors in the attribute
NL80211_ATTR_AKM_SUITES.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Looks like LibreSSL 2.8 pulled in the OpenSSL API change to mark the
first argument to X509_ALGOR_get0() const.
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
When using LibreSSL instead of OpenSSL, linkage of hostapd executable
fails with the following error when using some LibreSSL versions
../src/crypto/tls_openssl.o: In function `tls_verify_cb':
tls_openssl.c:(.text+0x1273): undefined reference to `ASN1_STRING_get0_data'
../src/crypto/tls_openssl.o: In function `tls_connection_peer_serial_num':
tls_openssl.c:(.text+0x3023): undefined reference to `ASN1_STRING_get0_data'
collect2: error: ld returned 1 exit status
make: *** [Makefile:1278: hostapd] Error 1
ASN1_STRING_get0_data is present in recent OpenSSL, but absent in some
versions of LibreSSL (confirmed for version 2.6.5), so fallback needs to
be defined in this case, just like for old OpenSSL.
This patch was inspired by similar patches to other projects, such as
spice-gtk, pjsip.
Link: https://bugs.gentoo.org/672834
Signed-off-by: Andrey Utkin <andrey_utkin@gentoo.org>
This makes it easier to integrate dynamic VLANs in custom network
configurations. The bridge name is added after the interface name in the
vlan_file line, also separated by whitespace.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Add support for starting FTM responder when in AP mode. This just sends
the appropriate NEW/SET_BEACON command to the driver with the LCI/civic
location data.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Andrei Otcheretianski <andrei.otcheretianski@intel.com>
Enable FTM responder and configure LCI and civic if ftm_responder
configuration option is set. Since ftm_responder configuration existed
before and was used to set extended capability bits, don't fail AP setup
flow if ftm_responder is set, but the driver doesn't advertise FTM
responder support.
Signed-off-by: Andrei Otcheretianski <andrei.otcheretianski@intel.com>
Add configuration options to enable FTM responder and configure LCI and
civic parameters. In addition, introduce WPA_DRIVER_FLAGS_FTM_RESPONDER
flag, which can be used to indicate FTM responder support in AP mode.
Signed-off-by: Andrei Otcheretianski <andrei.otcheretianski@intel.com>
Advertise vendor specific Multi-AP IE in (Re)Association Request frames
and process Multi-AP IE from (Re)Association Response frames if the user
enables Multi-AP fuctionality. If the (Re)Association Response frame
does not contain the Multi-AP IE, disassociate.
This adds a new configuration parameter 'multi_ap_backhaul_sta' to
enable/disable Multi-AP functionality.
Enable 4-address mode after association (if the Association Response
frame contains the Multi-AP IE). Also enable the bridge in that case.
This is necessary because wpa_supplicant only enables the bridge in
wpa_drv_if_add(), which only gets called when an interface is added
through the control interface, not when it is configured from the
command line.
Signed-off-by: Venkateswara Naralasetty <vnaralas@codeaurora.org>
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
The purpose of Multi-AP specification is to enable inter-operability
across Wi-Fi access points (APs) from different vendors.
This patch introduces one new configuration parameter 'multi_ap' to
enable Multi-AP functionality and to configure the BSS as a backhaul
and/or fronthaul BSS.
Advertise vendor specific Multi-AP capabilities in (Re)Association
Response frame, if Multi-AP functionality is enabled through the
configuration parameter.
A backhaul AP must support receiving both 3addr and 4addr frames from a
backhaul STA, so create a VLAN for it just like is done for WDS, i.e.,
by calling hostapd_set_wds_sta(). Since Multi-AP requires WPA2 (never
WEP), we can safely call hostapd_set_wds_encryption() as well and we can
reuse the entire WDS condition.
To parse the Multi-AP Extension subelement, we use get_ie(): even though
that function is meant for parsing IEs, it works for subelements.
Signed-off-by: Venkateswara Naralasetty <vnaralas@codeaurora.org>
Signed-off-by: Jouni Malinen <jouni@codeaurora.org>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Include and verify the OCI element in WNM-Sleep Exit Request and
Response frames. In case verification fails, the frame is silently
ignored.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
After the network changed to a new channel, perform an SA Query with the
AP after a random delay if OCV was negotiated for the association. This
is used to confirm that we are still operating on the real operating
channel of the network. This commit is adding only the station side
functionality for this, i.e., the AP behavior is not changed to
disconnect stations with OCV that do not go through SA Query.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
Include an OCI element in SA Query Request and Response frames if OCV
has been negotiated.
On Linux, a kernel patch is needed to let clients correctly handle SA
Query Requests that contain an OCI element. Without this patch, the
kernel will reply to the SA Query Request itself, without verifying the
included OCI. Additionally, the SA Query Response sent by the kernel
will not include an OCI element. The correct operation of the AP does
not require a kernel patch.
Without the corresponding kernel patch, SA Query Requests sent by the
client are still valid, meaning they do include an OCI element.
Note that an AP does not require any kernel patches. In other words, SA
Query frames sent and received by the AP are properly handled, even
without a kernel patch.
As a result, the kernel patch is only required to make the client properly
process and respond to a SA Query Request from the AP. Without this
patch, the client will send a SA Query Response without an OCI element,
causing the AP to silently ignore the response and eventually disconnect
the client from the network if OCV has been negotiated to be used.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
Include and verify the the OCI element in (Re)Association Request and
Response frames of the FT handshake. In case verification fails, the
handshake message is silently ignored.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
Verify the received OCI element in the 4-way and group key handshakes.
If verification fails, the handshake message is silently dropped.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
Use the information elements that were present in the (Re)Association
Request frame to derive the maximum bandwidth the AP will use to
transmit frames to a specific STA. By using this approach, we don't need
to query the kernel for this information, and avoid having to add a
driver API for that.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
ocv_verify_tx_params() verifies that the receive OCI element includes
field values that are compatible with the local channel configuration.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
If Operating Channel Verification is negotiated, include the OCI KDE
element in EAPOL-Key msg 2/4 and 3/4 of the 4-way handshake and both
messages of the group key handshake.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
Set the OCV bit in RSN capabilities (RSNE) based on AP mode
configuration. Do the same for OSEN since it follows the RSNE field
definitions.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
APs and mesh peers use the VHT Operation element to advertise certain
channel properties (e.g., the bandwidth of the channel). Save this
information element so we can later access this information.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
This adds two utility functions to convert both operating classes and
and the chan_width enum to an integer representing the channel
bandwidth. This can then be used to compare bandwidth parameters in an
uniform manner.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
This function can be used to easily convert the parameters returned
by the channel_info driver API, into their corresponding operating
class and channel number.
Signed-off-by: Mathy Vanhoef <Mathy.Vanhoef@cs.kuleuven.be>
