Adding a new entry was returning FAIL on success due to incorrect
assumption of wpa_blacklist_add() returning 0 on success (it returns the
positive blacklist count).
Felix Fietkau [Sun, 6 Apr 2014 11:11:53 +0000 (13:11 +0200)]
nl80211: Handle multiple interface combinations for P2P
The first combination may allow single-channel concurrency for
P2P + managed, but there may be others that allow multi-channel
concurrency. Parse all of them to find the maximum number of channels.
Nirav Shah [Mon, 7 Apr 2014 22:12:19 +0000 (03:42 +0530)]
P2P: Add retry mechanism for GO Negotiation Confirmation
wpa_supplicant now retries for P2P_GO_NEG_CNF_MAX_RETRY_COUNT times if
it doesn't receive acknowledgement for GO Negotiation Confirmation
frame. Currently, P2P_GO_NEG_CNF_MAX_RETRY_COUNT is set to 1.
While this is not strictly speaking following the P2P specification,
this can improve robustness of GO Negotiation in environments with
interference and also with peer devices that do not behave properly
(e.g., by not remaining awake on the negotiation channel through the
full GO Negotiation).
P2P: Mark the scan in p2p_in_invitation as p2p_probe
Mark the scan performed by the P2P Client in search of the GO
during the persistant reinvocation as a p2p_probe to avoid
unnecessary use of 802.11b rates.
Some OpenSSL versions have vulnerability in TLS heartbeat request
processing. Check the processed message to determine if the attack has
been used and if so, do not send the response to the peer. This does not
prevent the buffer read overflow within OpenSSL, but this prevents the
attacker from receiving the information.
This change is an additional layer of protection if some yet to be
identified paths were to expose this OpenSSL vulnerability. However, the
way OpenSSL is used for EAP-TLS/TTLS/PEAP/FAST in hostapd/wpa_supplicant
was already rejecting the messages before the response goes out and as
such, this additional change is unlikely to be needed to avoid the
issue.
WNM: Fix neighbor report subelement parser to not leak memory
If a subelement is unexpectedly included multiple times, the parser must
not re-allocate memory for the entry without first freeing the old
allocation.
Only the Neighbor Report element should be included here, so verify that
the element id matches. In addition, verify that each subelement has
valid length before using the data.
Document and rename HT Capability/Operation fields
This makes the definitions match the terminology used in IEEE Std
802.11-2012 and makes it easier to understand how the HT Operation
element subfields are used.
This was used to fill in the "PSMP support" subfield that was defined
during P802.11n development. However, this subfield was marked reserved
in the published IEEE Std 802.11n-2009 and it is not supported by
current drivers that use hostapd for SME either. As such, there is not
much point in maintaining this field as ht_capab parameter within
hostapd either.
Commit 41d5ce9e0b7b37dd84fbf3c1aa5ed571c32321d4 added scan optimizations
for P2P invitation cases. However, it left a path where the invitation
state was not cleared and as such, introduced potential issues for
following scans that ended up getting incorrectly optimized for a single
channel regardless of parameters. Fix this by clearing p2p_in_invitation
more carefully, especially on p2p_cancel command.
p2p_cancel is now also run from the FLUSH command to avoid issues with
hwsim test cases (e.g., persistent_group_per_sta_psk followed by
ap_open_tdls resulted in an error due to p2p_in_invitation remaining set
after FLUSH).
WNM: Fix deinit path to clean neighbor report count
wnm_deallocate_memory() left wnm_num_neighbor_report set while freeing
the allocated buffer of neighbor reports. If this function was called
twice in a row without having went through new neighbor report parsing,
invalid pointers could have been freed resulted in segfault.
The newer NFC_REPORT_HANDOVER command is more general version that
handles the cases where NFC_RX_HANDOVER_SEL could potentially have been
used. In addition, with the new WPS NFC design, it is important to
receive the handover request message that was used in the handover in
addition to the select message, so the NFC_RX_HANDOVER_SEL command on
its own would not be sufficient for this.
tests: Add option for running test cases that take a long time
run-tests.py now takes an optional --long parameter that can be used to
enable running of test cases that take a long time (multiple minutes).
By default, such test cases are skipped to avoid making the normal test
run take excessive amounts of time.
As an initial long test case, verify WPS PBC walk time expiration (two
minutes).
Now that WPS 2.0 support is enabled unconditionally, WEP and Shared auth
type are not allowed. This made some of the older code unused and that
can now be removed to clean up the implementation. There is still one
place where WEP is allowed for testing purposes: wpa_supplicant as
Registrar trying to configure an AP to use WEP. That is now only allowed
in CONFIG_TESTING_OPTIONS=y builds, though.
EAP-pwd peer: Fix fragmentation of PWD-Confirm-Resp
This is somewhat of a corner case since there is no real point in using
so short a fragmentation threshold that it would result in this message
getting fragmented. Anyway, it is better be complete and support this
case as well.
WNM: Fix regression in Sleep Mode exit key data parsing
Commit dbfb8e82ff69e6c7969b7cd23e53fd39b3e896e7 changed the Action frame
RX payload pointer design to point to a different field. WNM Sleep Mode
Response handler updated one of the uses to accommodate this change, but
that commit missed another use for key data length. This resulted in GTK
and IGTK being ignored in many cases when waking up from WNM Sleep Mode
with PMF enabled.
Some of the allocated configuration parameter did not free the previous
value if a configuration file or ctrl_iface SET command updated a
previously set value.
Interworking: Add writing of forgotten cred parameters into config file
req_conn_capab and required_roaming_consortium parameters were not
written into the cred block when requesting wpa_supplicant to write the
configuration file.
"GET_CRED <id> <field>" can now be used to fetch credential parameters
over the control interface. This does not allow passwords etc. private
material to be retrieved ("*" is returned for those if the value is set
regardless of the value). FAIL is returned if the requested parameter
has not been set. For cred parameters that can have multiple values,
newline separated list of values is returned.
projects / thirdparty / hostap.git / log
