Commit graph

10518 commits

Author SHA1 Message Date
Jouni Malinen
33dcced589 tests: SAE with number of different passwords
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-24 21:07:18 +03:00
Jouni Malinen
1965e1965e tests: SAE anti clogging protocol testing
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 23:10:36 +03:00
Jouni Malinen
8ec33326c1 SAE: Merge sae_derive_commit() error case return statements
These error cases have the exact same outcome, so a single return
statement can be used.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 23:10:36 +03:00
Jouni Malinen
d93abd4a97 SAE: Merge sae_get_rand() error case return statements
These error cases have the exact same outcome, so a single return
statement can be used.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 23:10:36 +03:00
Jouni Malinen
939527b59d tests: SAE reflection attack
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 23:10:36 +03:00
Jouni Malinen
6a58444d27 SAE: Verify that own/peer commit-scalar and COMMIT-ELEMENT are different
This check explicitly for reflection attack and stops authentication
immediately if that is detected instead of continuing to the following
4-way handshake that would fail due to the attacker not knowing the key
from the SAE exchange.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 23:10:36 +03:00
Jouni Malinen
4e7e68890a Add crypto_ec_point_cmp()
This is needed to allow SAE to check whether ECC elements are identical.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 22:29:23 +03:00
Jouni Malinen
8e2a3a4c21 dbus: Do not initialize variable twice
There is no point in initializing 'success' to FALSE when the actual
value is set just below this.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 21:16:35 +03:00
Jouni Malinen
c1a14ef42f Do not check unsigned size is less than zero
The variables here are unsigned and as such, cannot have a negative
value. Use == 0 instead of <= 0 to make this cleaner.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 21:16:15 +03:00
Jouni Malinen
fdc5608c12 OpenSSL: Remove SSL_CTX_{get,set}_app_data() compatibility wrapper
OpenSSL 0.9.8 (and newer) includes SSL_CTX_get_app_data() and
SSL_CTX_set_app_data(), so there is no need to maintain this old
OPENSSL_SUPPORTS_CTX_APP_DATA backwards compatibility design.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 21:05:02 +03:00
Jouni Malinen
ba54933f63 libtommath: Fix mp_init_multi() stdarg use on error path
Previously, it would have been possible for va_end(args) to be called
twice in case mp_init() fails. While that may not cause issues on number
of platforms, that is not how va_start()/va_end() are supposed to be
used. Fix this by returning from the function without using va_end()
twice on the same va_list args.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 20:39:08 +03:00
Jouni Malinen
f6332b0be1 wpa_gui: Initialize WpaGuiApp::w in the constructor
This gets rid of a static analyzer warning. The actual value for
WpaGuiApp::w will be set after the constructor has returned, so this
value was not really used uninitialized.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 20:23:31 +03:00
Jouni Malinen
f6df3f3a00 Use os_* wrapper more consistently
os_free() needs to be used when freeing memory that was allocated with
os_malloc()/os_zalloc()/os_calloc().

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 20:08:00 +03:00
Jouni Malinen
91b7a5e146 Use unsigned/signed printf format more consistently
These configuration parameters did not use matching printf format string
parameters (signed vs. unsigned). While these configuratin values are,
in practice, small unsigned integers, the implementation should use
matching types to write these.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 20:08:00 +03:00
Nishant Chaprana
59bae7463a HS 2.0R2: Fix memory leak on error path in hs20-osu-client
fqdn was not freed before return in case the server uses an unsupported
location for the PPS MO in the addMO command.

Signed-off-by: Nishant Chaprana <n.chaprana@samsung.com>
2015-06-23 18:51:41 +03:00
Jouni Malinen
c5ca73d1f7 P2P: Use offsetof() instead of local implementation
The construction used here to figure out the offset of variable length
IEs in Probe Request frames was a bit odd looking and resulted in a
warning from a static analyzer, so replace it with more standard use of
offsetof().

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 18:30:11 +03:00
Jouni Malinen
c3c5b5fe92 ERP server: Make erp_send_finish_reauth() easier for static analyzers
The flags argument is used to indicate a failure case (0x80) which
allows erp == NULL. This may be a bit too difficult combination for
static analyzers to understand, so add an explicit check for !erp as
another condition for returning from the function before the erp pointer
gets dereferenced without checking it.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 18:27:06 +03:00
Jouni Malinen
6ce1bea1ff bsd: Remove redundant NULL check in bsd_init()
drv cannot be NULL here (it is dereferenced even on the preceding line)
and anyway, os_free(NULL) is allowed, so remove the redundant check.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 18:20:24 +03:00
Jouni Malinen
c99df20192 Remove redundant NULL check in ieee802_1x_encapsulate_radius()
The eap argument to this function is never NULL and the earlier
ieee802_1x_learn_identity() call is dereferencing it anyway, so there is
no point in checking whether it is NULL later in the function.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 18:18:25 +03:00
Jouni Malinen
d5479878b8 tests: HT40 co-ex scan conflict (HT40-)
This verifies a case where the neighboring BSS is at the other end of
the band and has its PRI channel further away.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 11:25:22 +03:00
Jouni Malinen
2eb5967d99 AP: Add more 2.4 GHz channels for 20/40 MHz HT co-ex scan
This needs to find the PRI channel also in cases where the affected
channel is the SEC channel of a 40 MHz BSS, so need to include the
scanning coverage here to be 40 MHz from the center frequency. Without
this, it was possible to miss a neighboring 40 MHz BSS that was at the
other end of the 2.4 GHz band and had its PRI channel further away from
the local BSS.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 11:25:22 +03:00
Jouni Malinen
5ed6519625 hw_features: Merge similar return cases
There is no need to have separate return statements for these corner
cases that are unlikely to be hit in practice.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 11:24:52 +03:00
Jouni Malinen
24f4ad04f7 tests: Use different channel offset for the second 160 MHz VHT AP
This increases testing coverage a bit.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 02:42:17 +03:00
Jouni Malinen
5161ea8e30 tests: VHT 80 MHz with different HT40 channel options
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-23 02:30:24 +03:00
Jouni Malinen
4e37dd6c60 SAE: Simplify sae_prepare_commit() error path
There is no need to keep separate "return -1" statements for these error
cases.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-22 00:38:02 +03:00
Jouni Malinen
5527a3911a tests: SAE protocol tests
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-22 00:04:09 +03:00
Jouni Malinen
80dcede592 tests: HT40 co-ex scan and other BSS not affected
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-21 22:27:34 +03:00
Jouni Malinen
eb6d353270 tests: SAE and OOM in wpa_supplicant
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 23:31:12 +03:00
Jouni Malinen
04e6c4cc50 Fix SAE group selection in an error case
The sae_groups parameter is zero terminated array, not -1 terminated, so
must check the value against <= 0 to break out from the loop.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 23:29:57 +03:00
Jouni Malinen
3dce85ceb0 HS 2.0: Add WLAN RADIUS attributes in OSEN case
Previously, the common WLAN-* RADIUS attributes were added only when WPA
or WPA2 was used. These can be of use for OSEN as well, so include them
in that case, too.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 23:04:32 +03:00
Jouni Malinen
efd5d26d33 Remove unnecessary wpa_ie_len check from wpa_parse_wpa_ie_wpa()
There is no need to have a separate "fail silently" case for wpa_ie_len
== 0. That condition does not seem to be reachable and even if it were,
the following "ie len too short" case will result in the exact same
return value.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 23:04:31 +03:00
Jouni Malinen
6e3ee4c532 tests: Suite B with external RADIUS server
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 23:04:31 +03:00
Jouni Malinen
d9e2a057c8 tests: D-Bus AddNetwork and connection with WPA+WPA2-Enterprise AP
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 23:04:31 +03:00
Jouni Malinen
4b0e0c53fc tests: WPA IE parsing
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 23:04:31 +03:00
Jouni Malinen
1275208c82 tests: Update wpas_ctrl_network to match removal of WEP as group cipher
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 23:04:31 +03:00
Jouni Malinen
ce8963fc9f Remove WEP40/WEP104 cipher suite support for WPA/WPA2
As far as IEEE 802.11 standard is concerned, WEP is deprecated, but at
least in theory, allowed as a group cipher. This option is unlikely to
be deployed anywhere and to clean up the implementation, we might as
well remove all support for this combination.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 23:04:31 +03:00
Jouni Malinen
664093b55b tests: WPA2-PSK-FT AP and invalid response IEs
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 18:04:50 +03:00
Jouni Malinen
ee140ef98f FT: Stop association attempt if Auth response processing fails (SME)
Call the FT processing function directly instead of going through
wpa_supplicant_event() to process FT Authentication frame in SME case.
This allows parsing error to be used to trigger immediate failure for
the connection instead of trying to proceed to reassociation step that
cannot succeed.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 17:36:58 +03:00
Jouni Malinen
50a19b0107 tests: HT40 co-ex scan finding an AP advertising 40 MHz intolerant
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 16:16:11 +03:00
Jouni Malinen
1887be4fa7 Make check_20mhz_bss() static
This is not used outside this file anymore, so there is no need to
export the symbol either.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 16:05:54 +03:00
Jouni Malinen
f0b347b615 tests: HT40 co-ex scan conflict with HT 20 MHz AP
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 16:02:53 +03:00
Jouni Malinen
97e4688b2f tests: HT40 on 5 GHz with invalid channel pair
Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 15:51:41 +03:00
Jouni Malinen
5ce2a9ac50 tests: Make dbus_old_wps_pbc more robust
Flush the cfg80211 scan cache explicitly to avoid false failure reports
if a BSS entry from an earlier test case remain. Such a failure could be
hit, e.g., with the following test case sequence:
wpas_mesh_mode_scan p2p_channel_random_social dbus_old_wps_pbc

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-20 15:35:52 +03:00
Ilan Peer
f43c1ae798 P2P: Handle P2P Device dedicated interface parent removal
In case of a network interface removal, check if the interface
was also the parent interface of the P2P Device dedicated interface.
If this is the case, then stop the P2P Device functionality, and
remove the P2P Device dedicated interface.

In case that the interface is added again and P2P Device
functionality can be enabled again, add a new P2P Device dedicated
interface and allow further P2P Device functionality.

In case that the P2P Device dedicated interface is re-created, the
original P2P Device configuration file is needed, so store it in
the global params (instead in the wpa_interface configuration).

Signed-off-by: Ilan Peer <ilan.peer@intel.com>
2015-06-19 16:22:26 +03:00
Ilan Peer
1ac977bdd5 nl8021: Allow sending wowlan configuration on any interface
Sending a wowlan configuration command can be done on any wireless
interface (not only netdev), as it is a device configuration and not
interface configuration specific. Fix the code to allow it to be
sent on any interface.

Signed-off-by: Ilan Peer <ilan.peer@intel.com>
2015-06-19 16:17:04 +03:00
Ilan Peer
489970270b nl80211: Remove android_genl_ctrl_resolve()
Android libnl_2 implementation added support for "nl80211" name in
commit 'libnl_2: Extend genl_ctrl_resolve() to support "nl80211" name'
in July 2012 which got included in Android 4.2. It is fine to drop this
old Android ICS workaround from wpa_supplicant now.

Signed-off-by: Ilan Peer <ilan.peer@intel.com>
2015-06-19 16:06:28 +03:00
Eliad Peller
38dcc86cb7 P2P: Consider ht/vht on P2P_GROUP_ADD command (with no params)
p2p_ctrl_group_add() takes care of various configuration options (such
as ht/vht) before calling wpas_p2p_group_add(), so use it (just like
when P2P_GROUP_ADD is called with additional params).

Signed-off-by: Eliad Peller <eliadx.peller@intel.com>
2015-06-19 15:57:16 +03:00
Eliad Peller
29292d53ee ctrl_iface: Make p2p_ctrl_group_add() more robust
Parse each parameter individually and combine all the function calls.
This will allow further patch to call it with no parameters (currently
this might result in failure).

Signed-off-by: Eliad Peller <eliadx.peller@intel.com>
2015-06-19 15:56:54 +03:00
Eliad Peller
e4a80d89bd P2P: Fix secondary channel selection for HT40
wpas_p2p_get_ht40_mode() used blacklist approach (bw != BW20) to find
the relevant op_class, but didn't take into account other non-BW40
cases, like BW80, that had been added to the bw enum after the initial
implementation. Fix this by looking for the specific BW40 bw cases.

Signed-off-by: Eliad Peller <eliadx.peller@intel.com>
2015-06-19 15:50:40 +03:00
Jouni Malinen
81f8e7e9fb tests: Increase HTTPConnection timeout in ap_wps_upnp_http_proto
It looks like the previous timeout of 0.1 seconds could be hit under
parallel VM load, so double this to 0.2 second to avoid hitting
unnecessary test failures.

Signed-off-by: Jouni Malinen <j@w1.fi>
2015-06-19 13:45:16 +03:00