Sunil Dutt [Thu, 30 Oct 2014 15:21:52 +0000 (20:51 +0530)]
P2P: Do not consider the p2p_scan results after p2p_stop_find
If p2p_stop_find is issued after the p2p_scan request is triggered to
the host driver, the obtained scan results are also considered to update
the P2P peer entries. This is not always desired behavior, i.e., it can
be clearer if no P2P-DEVICE-FOUND events are generated based of that
final pending scan.
Sunil Dutt [Thu, 30 Oct 2014 12:07:46 +0000 (17:37 +0530)]
P2P: Delay scan operation only when P2P is not in search state
With the radio work interface in place, station interface SCAN command
was not scheduled (i.e., it got continously delayed with "Delay station
mode scan while P2P operation is in progress") when a p2p_find was
operational. Fix this be delaying station mode scan only when a P2P
operation is in progress, but not in search state.
Sunil Dutt [Thu, 30 Oct 2014 10:50:22 +0000 (16:20 +0530)]
P2P: Set p2p_scan_running based on driver scan request result
With the radio work interface, the actual request to start p2p_scan
operation is scheduled from a radio work and hence the initial return
value cannot provide the real result of the driver operation to trigger
a scan. Introduce a new notification API to indicate the scan trigger
status based on which the p2p_scan_running instance can be set using the
real return value from the driver operation.
Jouni Malinen [Thu, 30 Oct 2014 09:43:47 +0000 (11:43 +0200)]
MACsec: Fix policy configuration
macsec_validate variable was set incorrectly to FALSE(0) or TRUE(1)
instead of the enum validate_frames values (Disabled(0), Checked(1),
Strict(2). This ended up policy == SHOULD_SECURE to be mapped to
macsec_validate == Checked instead of Strict. This could have resulted
in unintended SecY forwarding of invalid packets rather than dropping
them.
Jouni Malinen [Wed, 29 Oct 2014 11:04:09 +0000 (13:04 +0200)]
P2P: Clean up p2p_go_neg_failed() calls
This function is always called with the peer argument equal to
p2p->go_neg_peer, so there is no need for that argument to be there. In
addition, p2p->go_neg_peer is not NULL in cases where there is an
ongoing GO Negotiation, so the function can be simplified to just check
once whether the peer pointer is set and if not, skip all processing.
Jouni Malinen [Tue, 28 Oct 2014 14:39:33 +0000 (16:39 +0200)]
tests: GO Negotiation and timeout while waiting for peer
grpform_cred_ready_timeout2 is similar to the grpform_cred_ready_timeout
test case with the difference being in initiating a P2P_FIND operation
during the wait.
Rashmi Ramanna [Tue, 28 Oct 2014 11:26:43 +0000 (16:56 +0530)]
P2P: Use timer to fail GO Negotation while waiting for peer
The timeout check while waiting for the peer to accept the GO
Negotiation depended on the WAIT_PEER_IDLE or WAIT_PEER_CONNECT states
being in use. Any P2P command to alter such states would have resulted
in the failure to time out GO Negotiation and thus ended up in not
indicating GO Negotiation failure or left the selected peer available
for new GO negotiation after the expected two minute timeout.
Fix this by using a separate timer to time out GO Negotiation
irrespective of the P2P state.
Jouni Malinen [Tue, 28 Oct 2014 12:01:16 +0000 (14:01 +0200)]
tests: Remove ap_ht40_csa and ap_ht40_csa3 workarounds
The issue in mac80211 was identified and proposed fix is now available,
so remove the workaround here and start reporting failures from
unexpected disconnection during CSA.
Jouni Malinen [Mon, 27 Oct 2014 23:47:17 +0000 (01:47 +0200)]
P2P: Do not call wpas_p2p_disconnect_safely() if no match is found
In theory, wpas_p2p_disconnect_safely() could have dereferenced the
wpa_s == NULL argument, but in practice, it won't due to the
calling_wpa_s == wpa_s check and wpas_p2p_disconnect() accepting NULL.
Anyway, it is cleaner to add an explicit check for this. (CID 74492)
Jouni Malinen [Mon, 27 Oct 2014 23:41:57 +0000 (01:41 +0200)]
nl80211: Fix a copy-paste error in key offload management
Commit b41f26845aaa7cf8aed6e4889e7041debc476ef9 ('Add support for
offloading key management operations to the driver') used incorrect
variable for determining the KCK length. This does not get triggered in
normal use cases since KCK and KEK are always included and of the same
length (in currently supported key management cases). Anyway, this needs
to be fixed to check the correct attribute. (CID 74491)
Kyeyoon Park [Fri, 26 Sep 2014 05:32:55 +0000 (22:32 -0700)]
AP: Add support for Proxy ARP, DHCP snooping mechanism
Proxy ARP allows the AP devices to keep track of the hardware address to
IP address mapping of the STA devices within the BSS. When a request for
such information is made (i.e., ARP request, Neighbor Solicitation), the
AP will respond on behalf of the STA device within the BSS. Such
requests could originate from a device within the BSS or also from the
bridge. In the process of the AP replying to the request (i.e., ARP
reply, Neighbor Advertisement), the AP will drop the original request
frame. The relevant STA will not even know that such information was
ever requested.
This feature is a requirement for Hotspot 2.0, and is defined in IEEE
Std 802.11-2012, 10.23.13. While the Proxy ARP support code mainly
resides in the kernel bridge code, in order to optimize the performance
and simplify kernel implementation, the DHCP snooping code was added to
the hostapd.
Signed-off-by: Kyeyoon Park <kyeyoonp@qca.qualcomm.com>
Kyeyoon Park [Sat, 27 Sep 2014 04:09:54 +0000 (21:09 -0700)]
AP: Add support for setting bridge network parameter
This allows setting a network parameter on the bridge that the BSS
belongs to.
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>
Kyeyoon Park [Fri, 26 Sep 2014 04:04:43 +0000 (21:04 -0700)]
AP: Add support for setting bridge port attributes
This allows setting a bridge port attribute. Specifically, the bridge
port in this context is the port 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>
Kyeyoon Park [Wed, 24 Sep 2014 06:02:04 +0000 (23:02 -0700)]
AP: Add support for IPv4 neighbor entry management to the BSS bridge
This allows adding/deleting an IPv4 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 [Mon, 27 Oct 2014 22:24:11 +0000 (00:24 +0200)]
tests: Scan result update changing last_scan_res and connect_work
This verifies that last_scan_res and connect work pointers to BSS
entries get updated when scan result update ends up removing or
reallocating a BSS entry.
Jouni Malinen [Mon, 27 Oct 2014 22:19:24 +0000 (00:19 +0200)]
Update pending connect radio work BSS pointer on scan update
It is possible for scan result processing or BSS entry removal to occur
while there is a pending connect or sme-connect radio work with a
previously selected BSS entry. The BSS pointer was previously verified
to be valid, i.e., still point to a BSS entry, at the time the actual
connection operation is started. However, that BSS entry could have
changed to point to another BSS if the old BSS entry was either removed
or reallocated and a new BSS entry was added at the same location in
memory. This could result in the connection attempt failing to configure
parameters properly due to different BSS information (e.g., different
BSSID).
Fix this by updated the pending connect radio work data on BSS entry
updates similarly to how the last_scan_res array was updated. If the
selected BSS entry is removed, this will still result in a failed
connection, but reallocated BSS entry is now followed properly and used
when the connection work starts.
Sunil Dutt [Mon, 20 Oct 2014 04:48:47 +0000 (10:18 +0530)]
Invoke connect work done for all the connection failure cases
This commit fixes couple of connection failure paths where
wpas_connect_work_done() was not called, thus enabling the radio work
interface to proceed with the other queued actions.
Jouni Malinen [Mon, 27 Oct 2014 18:13:39 +0000 (20:13 +0200)]
WPS: Fix ap_scan=2 special association case
Commit 22628eca3440976bf51846da0554099f7429b206 ('Support driver-based
BSS selection in ap_scan=1 mode') ended up disabling the special
ap_scan=2 WPS case where ap_scan=1 like scan followed by association is
used to find the WPS AP using wildcard SSID. Fix this by allowing
association request even with wpas_driver_bss_selection() when searching
for a WPS AP.
Tomasz Bursztyka [Thu, 23 Oct 2014 08:31:19 +0000 (11:31 +0300)]
wpa_supplicant: Do not register a P2P management interface on DBus
DBus client should always request the proper netdev interface. This will
be necessary to get a consistent behavior whatever driver is in use:
iwlwifi (which requires a P2P mgmt dev), ath9/10k (which does not),
etc...
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
Jouni Malinen [Sun, 26 Oct 2014 09:20:01 +0000 (11:20 +0200)]
nl80211: Store externally managed bridge name in driver status
This makes it easier to figure out from hostapd control interface
whether an interface had been added to a bridge externally at the time
hostapd interface was enabled or if the interface gets added during
hostapd operations.
Ilan Peer [Mon, 7 Jul 2014 11:21:01 +0000 (14:21 +0300)]
P2P: Save group common frequencies in invitation result
Save the group common frequencies when starting a GO due to
an invitation signaling requesting to re-invoke a persistent GO.
To do so, move the code that handles the translation of p2p_channels to
frequency list into a public function so it can be re-used both when GO
Negotiation is done and invitation signaling is done.
Ilan Peer [Mon, 7 Jul 2014 11:20:59 +0000 (14:20 +0300)]
P2P: Save group common frequencies
Once a P2P GO interface is configured, save the group common
frequencies, as this can be useful later for channel selection
considerations during channel switch, etc.
Ilan Peer [Mon, 7 Jul 2014 11:20:58 +0000 (14:20 +0300)]
nl80211: Clear beacon_set when deleting a beacon from deinit_ap()
When a beacon was deleted from the kernel in
wpa_driver_nl80211_deinit_ap(), bss->beacon_set was not cleared so
restarting the AP again was not possible.
Fix this by clearing the variable once the beacon was deleted.
Ilan Peer [Mon, 7 Jul 2014 11:20:55 +0000 (14:20 +0300)]
nl80211: Add indoor only and GO concurrent flags
Add the following channel flags:
* INDOOR_ONLY: The channel can be used if and only if there is a clear
assessment that the device is operating in an indoor environment, i.e.,
it is AC power.
* CONCURRENT_GO: The channel can be used for instantiating a GO if and
only if there is an additional station interface that is currently
connected to an AP on the same channel or on the same U-NII band
(assuming that the AP is an authorized master).
Jouni Malinen [Sat, 25 Oct 2014 19:27:24 +0000 (22:27 +0300)]
tests: Make ap_ht40_csa and ap_ht40_csa3 less likely to fail
It looks like cfg80211 can trigger disconnection even without disabling
HT, so the same issue may be hit with both of the channel switches in
these test cases.
Jouni Malinen [Sat, 25 Oct 2014 18:25:00 +0000 (21:25 +0300)]
Add test functionality to simulate driver increased ROC duration
The extra_roc_dur parameter can now be used in CONFIG_TESTING_OPTIONS=y
builds to simulate driver behavior where the ROC duration gets increased
without user space request.
Constantin Musca [Fri, 12 Sep 2014 07:22:19 +0000 (10:22 +0300)]
P2P: Stop driver listen in p2p_state_timeout()
When a P2P timeout occurs and p2p_state_timeout is executed, the
stop_listen function can be called besides setting in_listen to zero in
cases where the driver is still in ROC. That should not really happen in
normal cases, but it is possible for some drivers to extend the ROC
duration. If that happens, the next start_listen request may get
rejected with "P2P: Reject start_listen since p2p_listen_work already
exists".
Signed-off-by: Constantin Musca <constantin.musca@intel.com>
Mikael Kanstrup [Tue, 21 Oct 2014 10:08:30 +0000 (12:08 +0200)]
P2P: Use neg_freq when re-establish persistent group as GC
When GC receives invitation response and tries to re-establish
connection to a persistent group channels from passive list should
be allowed. A missing check for operation mode triggered reselection
of operating channel from active channels only to happen and thus fail
the connect attempt.
Add a check for operation mode and if GC instead use negotiated
frequency (i.e. GO operating channel from invitation response).
Signed-off-by: Mikael Kanstrup <mikael.kanstrup@sonymobile.com>
Thomas Pedersen [Mon, 1 Sep 2014 04:23:26 +0000 (00:23 -0400)]
mesh: Add user_mpm config option
Add user_mpm config parameter, when this is set to 1 (the default) the
peer link management is done on userspace, otherwise the peer management
will be done by the kernel.
Signed-off-by: Javier Lopez <jlopex@gmail.com> Signed-off-by: Jason Mobarak <x@jason.mobarak.name> Signed-off-by: Thomas Pedersen <thomas@noack.us>
Thomas Pedersen [Mon, 1 Sep 2014 04:23:21 +0000 (00:23 -0400)]
mesh: Add mesh mode routines
Add routines to (de)initialize mesh interface data structures and
join and leave mesh networks.
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>
Bob Copeland [Mon, 1 Sep 2014 04:23:23 +0000 (00:23 -0400)]
nl80211: Do not try to change supported rates
For mesh mode to work properly, set supported rates only once.
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: Bob Copeland <me@bobcopeland.com>
Bob Copeland [Mon, 1 Sep 2014 04:23:23 +0000 (00:23 -0400)]
nl80211: Add STA flag WPA_STA_AUTHENTICATED
This is needed for managing STA entries for mesh use cases.
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: Bob Copeland <me@bobcopeland.com>
Bob Copeland [Mon, 1 Sep 2014 04:23:23 +0000 (00:23 -0400)]
nl80211: Add plink_action_field to hostapd_sta_add_params
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: Bob Copeland <me@bobcopeland.com>
Bob Copeland [Mon, 1 Sep 2014 04:23:23 +0000 (00:23 -0400)]
nl80211: Reject deauth/disassoc for mesh interface
sta_info call these to send out the disassoc and deauth frame
which are both not relevent to mesh. So don't send them.
Signed-off-by: Chun-Yeow Yeoh <yeohchunyeow@gmail.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: Bob Copeland <me@bobcopeland.com>
Bob Copeland [Mon, 1 Sep 2014 04:23:23 +0000 (00:23 -0400)]
nl80211: Register to receive mesh frames
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: Bob Copeland <me@bobcopeland.com>
Jouni Malinen [Sat, 25 Oct 2014 09:09:02 +0000 (12:09 +0300)]
tests: Channel switch with HT40
ap_ht40_csa and ap_ht40_csa3 seem to show a cfg80211/mac80211 issue
where a CSA from HT40+ to HT40- channel results in HT getting disabled
due to cfg80211_chandef_compatible() check on c1->width == c2->width
claiming that the new channel is not compatible. For now, comment out
the FAIL case for these known issues to avoid constant failure for cases
that are not wpa_supplicant/hostapd issues. Once the kernel side design
has been reviewed, these error cases can be enabled here.
Jouni Malinen [Sat, 25 Oct 2014 08:27:48 +0000 (11:27 +0300)]
Fix channel switch to disable VHT with HT
If both HT and VHT was enabled on AP and channel switch event from the
driver indicated that HT was disabled, VHT was left enabled. This
resulted in the following channel configuration failing. Fix this by
disabling VHT if HT gets disabled.
Eliad Peller [Mon, 20 Oct 2014 03:21:39 +0000 (23:21 -0400)]
HT: Let the driver advertise its supported SMPS modes for AP mode
Add smps_modes field, and let the driver fill it with its supported SMPS
modes (static/dynamic). This will let us start an AP with specific SMPS
mode (e.g., dynamic) that will allow it to reduce its power usage.
Jouni Malinen [Thu, 23 Oct 2014 18:52:05 +0000 (21:52 +0300)]
P2P: Stop TX wait on SD query TX status failure
The previous TX operation could be on another channel if there are
multiple peers with pending SD queries. To avoid failing to send the
following query, stop the last one to allow any Listen channel to be
used for the following query during p2p_find iteration.
Jouni Malinen [Thu, 23 Oct 2014 18:49:42 +0000 (21:49 +0300)]
P2P: Limit number of SD retries during find
Commit 7139cf4a4f1fecfd03d0daff9bb33adb80cc3530 ('P2P: Decrement
sd_pending_bcast_queries when sd returns success') added support for
retrying P2P SD queries. However, it did this without limiting how many
retries are allowed. This can result in excessive number of retries if a
peer device does not show up on its Listen channel and there is a
pending SD query to it. Limit the maximum number of SD retries to 100
per p2p_find operation for each peer to avoid unlimited retries.
Jouni Malinen [Thu, 23 Oct 2014 18:30:32 +0000 (21:30 +0300)]
P2P: Iterate through all peers in pending pre-find operation
Commit 7139cf4a4f1fecfd03d0daff9bb33adb80cc3530 ('P2P: Decrement
sd_pending_bcast_queries when sd returns success') changed P2P SD
behavior in a way that the P2P search loop ended up in continuing with
the first peer entry until it acknowledged receipt of a pending
broadcast SD request while the previous design went through all peers
once. While it is reasonable to retry SD, getting stuck with the first
peer is not really desirable. Change the p2p_continue_find() loop to
continue from the next peer in each iteration to allow progress through
all peers that have pending operations if any other peer is not
acknowledging frames (e.g., due to not being on Listen channel).
Chet Lanctot [Thu, 23 Oct 2014 15:21:49 +0000 (18:21 +0300)]
Add support for offloading key management operations to the driver
This commit introduces a QCA vendor command and event to provide an
option to use extended versions of the nl80211 connect/roam operations
in a way that allows drivers to offload key management operations to the
driver/firmware.
Jouni Malinen [Thu, 23 Oct 2014 13:33:25 +0000 (16:33 +0300)]
tests: Allow time for go_neg_pbc() GO Neg Resp to go out
It was possible for the not-pre-authorized GO Negotiation case to end up
starting new GO Negotiation before the GO Negotiation Response frame
with status=1 was transmitted. While this works for group formation, it
could reduce test coverage for the common case where that response is
received. At a small sleep to make it less likely for this unexpected
sequence to happen during testing.
Jouni Malinen [Wed, 22 Oct 2014 17:02:48 +0000 (20:02 +0300)]
P2P: Search all 2.4 GHz social channels regardless of disallow_freq
Commit 1595eb93aee0239de51f1090d4528e4b2cfe3add ('P2P: Add support for
60 GHz social channel') had an unintended change to how P2P search scans
2.4 GHz social channels. Use of p2p_supported_freq() to filter the list
of social channels ended up using the disallow_freq setting to remove
social channels from the P2P search scans. This is not desired since
peers need to be found on any of the social channels even if those
channels have been disabled from P2P operating channel use. Restore the
previous behavior by included all the 2.4 GHz social channels in P2P
search scans if the driver indicated support for the 2.4 GHz band.
Jouni Malinen [Tue, 21 Oct 2014 09:16:47 +0000 (12:16 +0300)]
GAS: Do not reply to P2P SD query from generic GAS/ANQP server
This avoids an issue where a wpa_supplicant build with CONFIG_P2P=y and
CONFIG_HS20=y ended up processing a P2P SD query twice when operating as
a GO and sending out two replies. Only the P2P SD implementation should
reply to P2P SD query in such a case.
The kernel now has full support for this (and it is turned off
by default for hwsim) so wpa_supplicant should really go back
to autodetecting this so clients don't have to figure out what
to do.
Also add a debug message stating that P2P_DEVICE support is used.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Johannes Berg [Mon, 20 Oct 2014 10:00:07 +0000 (12:00 +0200)]
tests: Disable support_p2p_device by default
When loading the hwsim module, disable support_p2p_device by default.
This will also become the default in the kernel, but until then it
makes sure it's not turned on by default.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Jouni Malinen [Wed, 22 Oct 2014 07:50:57 +0000 (10:50 +0300)]
tests: Run run-tests.py as root from run-all.sh
This will be needed to be able to control dynamic mac80211_hwsim
operations like adding and removing a phy. In the past, it has been
possible to start the main programs as root and then use non-root
account for run-tests.py. However, there is already a large number of
cases within the test scripts where sudo is needed. Moving that
requirement to execution of run-tests.py allows those to be simplified
as well.
Jouni Malinen [Tue, 21 Oct 2014 11:16:06 +0000 (14:16 +0300)]
browser-android: Use execv() directly instead of os_exec()
This allows the URL to be passed as a single argument to the program
instead of getting split into multiple by os_exec(). This makes the
operation more robust for cases where the URL could have been received
from an external source and could potentially add extra arguments to the
command line.
In addition, fix the /system/bin/input execution by using system() for
it instead of execv() through os_exec(). /system/bin/input is a script
that execv() won't be able to run. Since the full command line is
specified, system() can be used for this. The keycode is also changed
from 3 to KEYCODE_HOME to make this work with current Android version.