Assaf Krauss [Wed, 5 Nov 2014 08:42:51 +0000 (03:42 -0500)]
wpa_supplicant: Add NEIGHBOR_REP_REQUEST command to the control interface
Add NEIGHBOR_REP_REQUEST command to the wpa_supplicant ctrl_iface.
This command triggers the sending of a Neighbor Report Request to the
associated AP.
Assaf Krauss [Wed, 5 Nov 2014 08:42:48 +0000 (03:42 -0500)]
SME: Add RRM support to association request
In case the AP we are associating with advertises support for RRM,
advertise our own RRM support in the (Re)Association Request frame. This
is done by adding an RRM Capabilities IE. The underlying driver is
expected to further add a Power Capabilities IE to the request, and set
the Radio Measurement flag in the Capability Info field. At this point
the RRM Capabilities IE advertises no measurement support.
Assaf Krauss [Wed, 5 Nov 2014 08:42:45 +0000 (03:42 -0500)]
driver: Add RRM-related definitions to driver interface
Add definitions for RRM (Radio Resource Measurement) support:
1. Flags that specify the RRM capabilities of the underlying driver
2. Flag for RRM in Capability Info field in Management frames
3. Indication in association parameters regarding an RRM connection
Jouni Malinen [Sat, 22 Nov 2014 18:12:12 +0000 (20:12 +0200)]
WNM: Use recent scan results on BSS transition request
If the last scans are recent (for now, less than ten seconds old), use
them instead of triggering a new scan when a BSS Transition Management
Request frame is received. As a fallback, allow a new scan to be
triggered if no matches were found.
Jouni Malinen [Sat, 22 Nov 2014 17:50:16 +0000 (19:50 +0200)]
WNM: Optimize BSS transition management scans
When the list of preferred transition candidates is received, use the
identified channels to optimize the following scan so that no time is
wasted on other channels.
Jouni Malinen [Sat, 22 Nov 2014 17:48:08 +0000 (19:48 +0200)]
Add generic operating class and channel to frequency function
ieee80211_chan_to_freq() is a generic function that replaces and extends
the previous P2P-specific p2p_channel_to_freq(). The new function
supports both the global operating class table as well as the additional
US, EU, JP, and CN operating class tables.
Jouni Malinen [Sat, 22 Nov 2014 16:28:22 +0000 (18:28 +0200)]
tests: Fix wnm_bss_tm_req status code expectation
The implementation of WNM BSS transition management was extended to be
able to return a reject status code based on whether a matching entry is
found. The test case wnm_bss_tm_req was trying to enforce a different
status code to be used here based on old implementation.
Jouni Malinen [Sat, 22 Nov 2014 15:01:26 +0000 (17:01 +0200)]
WNM: Allow BSS transition request in same ESS even if RSSI is worse
This allows an AP to steer us to another BSS within the ESS even if that
results in reduced signal strength as long as the signal strength with
the target BSS is expected to provide some connectivity.
Jouni Malinen [Sat, 22 Nov 2014 13:37:27 +0000 (15:37 +0200)]
WNM: Simplify how candidate subelements are stored
There is no need to use a separately allocated data structures for this.
A bitfield indicating which information is present and variables within
struct neighbor_report are simpler to use and more efficient.
This implements minimal RSN 4-way handshake Supplicant in Python and
uses that to test hostapd Authenticator implementation in various
possible protocol sequencies.
Jouni Malinen [Fri, 21 Nov 2014 15:02:00 +0000 (17:02 +0200)]
AP: Extend EAPOL-Key msg 1/4 retry workaround for changing SNonce
If the 4-way handshake ends up having to retransmit the EAPOL-Key
message 1/4 due to a timeout on waiting for the response, it is possible
for the Supplicant to change SNonce between the first and second
EAPOL-Key message 2/4. This is not really desirable due to extra
complexities it causes on the Authenticator side, but some deployed
stations are doing this.
This message sequence looks like this:
AP->STA: EAPOL-Key 1/4 (replay counter 1, ANonce)
AP->STA: EAPOL-Key 1/4 (replay counter 2, ANonce)
STA->AP: EAPOL-Key 2/4 (replay counter 1, SNonce 1)
AP->STA: EAPOL-Key 3/4 (replay counter 3, ANonce)
STA->AP: EAPOL-Key 2/4 (replay counter 2, SNonce 2)
followed by either:
STA->AP: EAPOL-Key 4/4 (replay counter 3 using PTK from SNonce 1)
or:
AP->STA: EAPOL-Key 3/4 (replay counter 4, ANonce)
STA->AP: EAPOL-Key 4/4 (replay counter 4, using PTK from SNonce 2)
Previously, Authenticator implementation was able to handle the cases
where SNonce 1 and SNonce 2 were identifical (i.e., Supplicant did not
update SNonce which is the wpa_supplicant behavior) and where PTK
derived using SNonce 2 was used in EAPOL-Key 4/4. However, the case of
using PTK from SNonce 1 was rejected ("WPA: received EAPOL-Key 4/4
Pairwise with unexpected replay counter" since EAPOL-Key 3/4 TX and
following second EAPOL-Key 2/4 invalidated the Replay Counter that was
used previously with the first SNonce).
This commit extends the AP/Authenticator workaround to keep both SNonce
values in memory if two EAPOL-Key 2/4 messages are received with
different SNonce values. The following EAPOL-Key 4/4 message is then
accepted whether the MIC has been calculated with the latest SNonce (the
previously existing behavior) or with the earlier SNonce (the new
extension). This makes 4-way handshake more robust with stations that
update SNonce for each transmitted EAPOL-Key 2/4 message in cases where
EAPOL-Key message 1/4 needs to be retransmitted.
Javier Lopez [Mon, 1 Sep 2014 04:23:35 +0000 (00:23 -0400)]
tests: Add test_wpas_mesh test cases
This wpa_supplicant tests include basic tests for:
- Mesh scan
- Mesh group add/remove
- Mesh peer connected/disconnected
- Add/Set/Remove to test mesh mode network
- Open mesh connectivity test
- Secure mesh connectivity test
- no_auto_peer
Signed-off-by: Jason Mobarak <x@jason.mobarak.name>
[no_auto_peer test by: Javier Cardona <javier@cozybit.com> Signed-off-by: Javier Lopez <jlopex@gmail.com>
Jouni Malinen [Thu, 20 Nov 2014 14:25:00 +0000 (16:25 +0200)]
nl80211: Do not try to change STA capability on NL80211_CMD_SET_STATION
cfg80211 will reject any of these attributes regardless of whether the
values actually change or not when NL80211_CMD_SET_STATION is used for
something else than TDLS. This path is hit at least when setting up
mesh, so filter out the attributes from NL80211_CMD_SET_STATION to match
the rules in cfg80211.
Jouni Malinen [Wed, 19 Nov 2014 20:02:08 +0000 (22:02 +0200)]
tests: Fix regression in START line total count
The previous changes to enable stdin control broke the previous case of
showing the total number of test cases in the START lines. Fix that by
using a separate variable for the total number of test casess instead of
using length of the list of remaining test cases.
Rashmi Ramanna [Mon, 17 Nov 2014 10:37:24 +0000 (16:07 +0530)]
P2P: Do not change P2P state on GO Neg failure if it is P2P_SEARCH
Changing the P2P state to P2P_IDLE on GO Negotiation Failure would stop
the previously issued P2P_FIND operation without notifying the upper
layers. Leave the search operation running if in P2P_SEARCH state to
avoid unexpected behavior in case the upper layers issued a new P2P_FIND
while waiting for GO Negotiation to complete.
Kyeyoon Park [Sat, 1 Nov 2014 06:33:41 +0000 (23:33 -0700)]
AP: Add Neighbor Discovery snooping mechanism for Proxy ARP
This commit establishes the infrastructure, and handles the Neighbor
Solicitation and Neighbor Advertisement frames. This will be extended
in the future to handle other frames.
Signed-off-by: Kyeyoon Park <kyeyoonp@qca.qualcomm.com>
Kyeyoon Park [Thu, 6 Nov 2014 00:15:46 +0000 (16:15 -0800)]
AP: Extend the BSS bridge neighbor entry management to support IPv6
This allows adding/deleting an IPv6 neighbor entry to/from the bridge,
to which the BSS belongs. This commit adds the needed functionality in
driver_nl80211.c for the Linux bridge implementation. In theory, this
could be shared with multiple Linux driver interfaces, but for now, only
the main nl80211 interface is supported.
Signed-off-by: Kyeyoon Park <kyeyoonp@qca.qualcomm.com>
Jouni Malinen [Wed, 19 Nov 2014 00:03:39 +0000 (02:03 +0200)]
tests: Move parallel-vm.py test control to the host
This allows all VMs to be used at the end of a test sequence by
assigning test cases to VMs based on which VM is available for a new
test case rather than splitting the full task at the beginning and
potentially getting stuck with the last VM running long test cases for
significantly longer than another VM that gets shorter duration tests
assigned to it.
Jouni Malinen [Wed, 19 Nov 2014 00:01:27 +0000 (02:01 +0200)]
tests: Add run-tests.py -i to control execution from stdin
The new -i command line argument can be used to control test case
execution from stdin to run-tests.py and vm-run.sh. This can be used,
e.g., to run multiple repeated test sequences in a virtual machine
without havign to restart the VM between each iteration.
Jouni Malinen [Tue, 18 Nov 2014 22:41:45 +0000 (00:41 +0200)]
tests: Allow run-tests.py to be run from the vm directory
Some of the paths in run-tests.py are hardcoded and could not handle
tests/hwsim/vm as the working directory. Modify the design enough to
allow ../run-tests.py -L to be used from the vm directory.
Jouni Malinen [Tue, 18 Nov 2014 22:34:36 +0000 (00:34 +0200)]
tests: Make run-tests.py process test case selection arguments
This allows a list of matching test cases to be produced without having
to run the test cases. Previously, -L output included all defined test
cases regardless of what else was included on the command line.
Ilan Peer [Wed, 5 Nov 2014 08:50:34 +0000 (03:50 -0500)]
AP: Drop retransmitted auth/assoc/action frames
It is possible that a station device might miss an ACK for an
authentication, association, or action frame, and thus retransmit the
same frame although the frame is already being processed in the stack.
While the duplicated frame should really be dropped in the kernel or
firmware code where duplicate detection is implemented for data frames,
it is possible that pre-association cases are not fully addressed (which
is the case at least with mac80211 today) and the frame may be delivered
to upper layer stack.
In such a case, the local AP will process the retransmitted frame although
it has already handled the request, which might cause the station to get
confused and as a result disconnect from the AP, blacklist it, etc.
To avoid such a case, save the sequence control of the last processed
management frame and in case of retransmissions drop them.
This gets registered in tls_openssl.c from tls_init(), so there is no
need for EAP-pwd implementation to register explicitly. This avoids some
corner cases where OpenSSL resources do not get fully freed on exit.
Jouni Malinen [Sun, 16 Nov 2014 23:14:35 +0000 (01:14 +0200)]
DFS: Fix hostapd operation without hw_mode driver data
If DFS implementation was built in, some configurations with drivers
that do not provide mode information could end up dereferencing a NULL
pointer. Fix this by skipping DFS operations in such cases since not
having information about modes and channels means that hostapd could not
perform DFS anyway (i.e., either this is not a wireless driver or the
driver takes care of DFS internally).
Jouni Malinen [Sun, 16 Nov 2014 23:08:04 +0000 (01:08 +0200)]
Fix hostapd operation without hw_mode driver data
Commit 7f0303d5b0bb425f3e7318a7016b55ba9e67f9de ('hostapd: Verify VHT
160/80+80 MHz driver support') added couple of hapd->iface->current_mode
dereferences of which the one in hostapd_set_freq() can be hit with some
configuration files when using driver wrappers that do not have hw_mode
data, i.e., when current_mode is NULL. This could result in segmentation
fault when trying to use driver=wired. Fix this by checking that
current_mode is not NULL before dereferencing it to get vht_capab.
Johannes Berg [Wed, 22 Oct 2014 12:04:03 +0000 (08:04 -0400)]
WMM AC: Notify driver before sending DelTS
Switch the stop and sending DelTS to avoid sending data packets
for the session after the DelTS, which is otherwise possible.
This also helps the mac80211 implementation as it requires stopping the
traffic flow before sending the DelTS as it may modify the AC parameters
for the affected queue, and that may in turn affect management frames.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Moshe Benji [Wed, 22 Oct 2014 12:03:54 +0000 (08:03 -0400)]
WMM AC: Parse WMM IE on association
Initialize WMM AC data structures upon successful association
with an AP that publishes WMM support, and deinitialize the data
structure when the association is no longer valid.
Javier Lopez [Mon, 1 Sep 2014 04:23:33 +0000 (00:23 -0400)]
Add mesh_group_{add,remove} control interface commands
Parse MESH_GROUP_ADD/REMOVE commands on ctrl interface and call
wpa_supplicant routines. These commands are used to start or
join and leave a mesh network.
The mesh id is given in the configuration file, therefore there is
no need to scan before joining a mesh network. We reuse the
connect_without_scan construct used by P2P for that same purpose.
Signed-off-by: Javier Cardona <javier@cozybit.com> Signed-off-by: Javier Lopez <jlopex@gmail.com>
Jason Mobarak [Mon, 1 Sep 2014 04:23:36 +0000 (00:23 -0400)]
mesh: Enable mesh HT mode
Add a new option "mesh_ht_mode" that specifies the HT mode for the
mesh, with this option on, mesh beacons, actions frames, and probe
responses with include the appropriate HT information elements.
[original implementation by Chun-Yeow Yeoh <yeohchunyeow@gmail.com>]
[some fixes by Masashi Honma <masashi.honma@gmail.com>] Signed-off-by: Ashok Nagarajan <ashok.dragon@gmail.com> Signed-off-by: Javier Cardona <javier@cozybit.com> Signed-off-by: Jason Mobarak <x@jason.mobarak.name>
mesh: Add timer for SAE authentication in RSN mesh
Add timer to do SAE re-authentication with number of tries defined
by MESH_AUTH_RETRY and timeout defined by MESH_AUTH_TIMEOUT.
Ignoring the sending of reply message on "SAE confirm before commit"
to avoid "ping-pong" issues with other mesh nodes. This is obvious when
number of mesh nodes in MBSS reaching 6.
Signed-off-by: Chun-Yeow Yeoh <yeohchunyeow@gmail.com> Signed-off-by: Bob Copeland <me@bobcopeland.com>
Bob Copeland [Mon, 1 Sep 2014 04:23:30 +0000 (00:23 -0400)]
SAE: Enhance AP implementation to handle auth for mesh interfaces
Add state transition logic to the SAE frame handling in order to more
fully implement the state machine from the IEEE 802.11 standard. Special
cases are needed for infrastructure BSS case to avoid unexpected
Authentication frame sequence by postponing transmission of the second
Authentication frame untile the STA sends its Confirm.
[original patch by: Thomas Pedersen <thomas@noack.us>] Signed-off-by: Bob Copeland <me@bobcopeland.com>
Signed-off-by: Javier Lopez <jlopex@gmail.com> Signed-off-by: Javier Cardona <javier@cozybit.com> Signed-off-by: Jason Mobarak <x@jason.mobarak.name> Signed-off-by: Thomas Pedersen <thomas@noack.us>
New kernels in wiphy_suspend() will call cfg80211_leave_all() that will
eventually end up in cfg80211_stop_ap() unless wowlan_triggers were set.
For now, use the parameters from the station mode as-is. It may be
desirable to extend (or constraint) this in the future for specific AP
mode needs.
Jouni Malinen [Sun, 16 Nov 2014 13:40:02 +0000 (15:40 +0200)]
Suite B: Select EAPOL-Key integrity and key-wrap algorithms based on AKM
This adds support for AKM 00-0F-AC:11 to specify the integrity and
key-wrap algorithms for EAPOL-Key frames using the new design where
descriptor version is set to 0 and algorithms are determined based on
AKM.
Jouni Malinen [Sun, 16 Nov 2014 11:22:46 +0000 (13:22 +0200)]
Suite B: PMKID derivation for AKM 00-0F-AC:11
The new AKM uses a different mechanism of deriving the PMKID based on
KCK instead of PMK. hostapd was already doing this after the KCK had
been derived, but wpa_supplicant functionality needs to be moved from
processing of EAPOL-Key frame 1/4 to 3/4 to have the KCK available.
Jouni Malinen [Sun, 16 Nov 2014 11:20:51 +0000 (13:20 +0200)]
Suite B: Add AKM 00-0F-AC:11
This adds definitions for the 128-bit level Suite B AKM 00-0F-AC:11. The
functionality itself is not yet complete, i.e., this commit only
includes parts to negotiate the new AKM.
Jouni Malinen [Sat, 15 Nov 2014 10:56:02 +0000 (12:56 +0200)]
tests: Clear ignore_old_scan_res setting
This parameter was used in some of the NFC test cases to make scanning
more robust in case of changing AP configuration. However, the parameter
was not cleared anywhere, so it could have been left in use for other
test cases as well. To get more consistent behavior, clear the value
back to its default between test cases.
Jouni Malinen [Sat, 15 Nov 2014 10:35:10 +0000 (12:35 +0200)]
OpenSSL: Allow TLS v1.1 and v1.2 to be negotiated by default
Use SSLv23_method() to enable TLS version negotiation for any version
equal to or newer than 1.0. If the old behavior is needed as a
workaround for some broken authentication servers, it can be configured
with phase1="tls_disable_tlsv1_1=1 tls_disable_tlsv1_2=1".