owe_auth_req_process() can return NULL in error cases, but the caller
was not prepared for this. The p pointer cannot be overridden in such
cases since that would result in buffer length (p - buf) overflows. Fix
this by using a temporary variable to check the return value before
overriding p so that the hostapd_sta_assoc() ends up using correct
length for the IE buffer.
Fixes: 33c8bbd8ca ("OWE: Add AP mode handling of OWE with drivers that implement SME")
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
The BoringSSL version of crypto_ecdh_init() and dpp_gen_keypair() works
fine with OpenSSL as well, so use that same implementation for both to
avoid unnecessary maintanence of multiple versions.
Signed-off-by: Jouni Malinen <j@w1.fi>
There is no need to go through EC_GROUP_new_by_curve_name(),
EC_KEY_new(), and EC_KEY_set_group() when a single call to
EC_KEY_new_by_curve_name() takes care of all that.
Signed-off-by: Jouni Malinen <j@w1.fi>
This reverts commit 5548453a2d since
BoringSSL added ECDSA_SIG_set0() and ECDSA_SIG_get0() in commit
8dc226ca8f1ef60737e1c1bf8cfcabf51d4068c7 ('Add some missing OpenSSL
1.1.0 accessors.') and updated X509_ALGOR_get0() prototype to match
OpenSSL 1.1.0 changes in commit e3b2a5d30d309091cab3e6a19dee7323c40d968d
('Const-correct X509_ALGOR_get0.').
Signed-off-by: Jouni Malinen <j@w1.fi>
This allows wpa_supplicant configuration with phase1="tls_suiteb=1" to
use openssl_ciphers="ECDHE-RSA-AES256-GCM-SHA384" to further limit the
possible TLS cipher suites when using Suite B with RSA >3K keys. This
combination disables use of DHE and as such, mandates ECDHE to be used.
Signed-off-by: Jouni Malinen <j@w1.fi>
src/crypto/libcrypto.a and src/tls/libtls.a have circular references
and will need special handling with the linker at least for the time
being. This could be cleaned up eventually, but for now, provide a
mechanism to get the program linked.
This was already done in tests/Makefile, but tests/eapol-fuzzer/Makefile
needs the same.
Signed-off-by: Jouni Malinen <j@w1.fi>
These functions are a bit awkward to use for one-off file loads, as
suggested by the tls_clear_default_passwd_cb() logic. There was also
some historical mess with OpenSSL versions and either not having per-SSL
settings, having per-SSL settings but ignoring them, and requiring the
per-SSL settings.
Instead, loading the key with the lower-level functions seems a bit
tidier and also allows abstracting away trying both formats, one after
another.
Signed-off-by: David Benjamin <davidben@google.com>
There's no need to make an extra copy of private_key_passwd for
SSL_{CTX_,}set_default_passwd_cb().
Signed-off-by: David Benjamin <davidben@google.com>
Wpa_supplicant's random pool is not necessary on Android. Randomness
is already provided by the entropymixer service which ensures
sufficient entropy is maintained across reboots. Commit b410eb1913
'Initialize /dev/urandom earlier in boot' seeds /dev/urandom with
that entropy before either wpa_supplicant or hostapd are run.
Signed-off-by: Jeff Vander Stoep <jeffv@google.com>
Commit aab6612836 ('wlantest: Search
bss/sta entry more thoroughly for 4-address frames') allowed wlantest to
find a STA entry in this type of cases, but it was still possible for
that STA entry to be the one that has no derived PTK while the STA entry
for the other side of the link might have the derived PTK available.
Extend this BSS/STA selection mechanism to use sta->ptk_set to determine
which STA entry is more useful for decryption, i.e., select the one with
a known PTK.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
A new vendor attribute QCA_WLAN_VENDOR_ATTR_CONFIG_LATENCY_LEVEL is added
for vendor sub-command QCA_NL80211_VENDOR_SUBCMD_SET_WIFI_CONFIGURATION.
This attribute is for setting the level of WLM.
Signed-off-by: Paul Zhang <paulz@qti.qualcomm.com>
Do not include this argument in normal case, but add a test case to
cover the special extra check case.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
If a credential with encp type AES is received, add GCMP-256 and
CCMP-256 cipher options on station Enrollee based on local capabilities.
This is needed to allow connection with an AP using either of these
newer ciphers.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
This test case is not really realistic and the second connection attempt
would fail if additional AES-based ciphers get provisioned. Work this
around by dropping to CCMP only if other ciphers are present.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
This is needed to avoid false errors with GCMP-256 and CCMP-256 to be
added in the implementation.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
The credential MAC address is not necessarily that of the AP, i.e., it
is more likely to be that of the Enrollee. Check the scan results
against the current BSSID as well if match is not found otherwise when
going through the mixed mode workaround.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
This is needed to allow a credential to be built for GCMP-256/CCMP-256
networks that do not enable GCMP-128/CCMP-128.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Extend the check against WPA/TKIP only configuration by adding CCMP-256
and GCMP-256 to the list of allowed ciphers. This is needed to allow WPS
to be enabled in AP configurations where neither CCMP-128 nor GCMP-128
are enabled.
Signed-off-by: Shiva Sankar Gajula <sgajula@qti.qualcomm.com>
DPP Responder selects whether mutual authentication is used. This commit
adds information about that selection to upper layers (ctrl_iface event
DPP-AUTH-DIRECTION mutual=<0/1>) on the Initiator side.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
This extends dpp_test functionality to allow DPP exchanges to be stopped
when receiving a specified message.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
The previously used 10 second timer did not really make much sense since
the Initiator is not going to be waiting for the response that long.
Change this to 1 second based on the DPP tech spec change.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
The kernel started enforcing attribute lengths, and nl80211.py had been
doing it all wrong - the padding must be present, but not part of the
length.
Fix it to do it the right way.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This nl80211 attribute uses NLA_U8 policy in cfg80211 and wpa_supplicant
needs to use same size when writing the attribute.
This fixes mesh mode regression triggered by kernel commit "net:
netlink: Update attr validation to require exact length for some types"
in v4.15-rc1 that resulted in the following debug log entry when trying
to join a mesh:
nl80211: mesh join failed: ret=-22 (Invalid argument)
Fixes: 6c1664f605 ("nl80211: Add new commands to support mesh interfaces")
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
This nl80211 attribute uses NLA_U8 policy in cfg80211 and
hostapd/wpa_supplicant needs to use same size when writing the
attribute.
This fixes AP mode regression triggered by kernel commit "net: netlink:
Update attr validation to require exact length for some types" in
v4.15-rc1 that resulted in the following debug log entry when trying to
enable beaconing:
nl80211: Beacon set failed: -34 (Numerical result out of range)
Fixes: da1080d721 ("nl80211: Advertise and configure SMPS modes")
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Add QCA vendor command and attributes for vendor specific Representative
RF Operating Parameter (RROP) information. This information is intended
for optional use by external ACS. It provides guidance values for some
RF parameters that are used by the system during operation, so that
external ACS can utilize these to compare between channels, bands, etc.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
These test cases were failing when run immediately after
dpp_pkex_test_fail. It looks like timing of the TX status and the short
eloop wait were getting reordered in this cases. This ended up with some
of the DPP-TX-STATUS event messages missing. Instead of explicitly
checking for those message, simply count the number of DPP-TX messages
to verify that the correct number of retries are being sent.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
One of the event message for TX status was missing 'x' from the "0x"
prefix. Add that to make the used format consistent for all cookie debug
print cases.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
This makes the full DPP deinit operation more consistent with stopping
of a single operation. In practice, this adds the new GAS client
stopping functionality.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
sha256_vector() result was ignored apart from printing out the failure
in the debug log. This is not really a normal case and it is better to
reject the full operation rather than try to continue with an incorrect
public key hash value.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
If the second json_alloc_token() call failed to allocate memory,
json_parse() missed the first allocation on the error path. Assign the
root pointer earlier for that case to avoid the potential memory leak.
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
