Commit fb8984fd6f added a mechanism to
skip the Listen state when the peer is expected to be waiting for us to
initiate a new GO Negotiation. However, this flag was set when building
the GO Negotiation Response frame with status 1 regardless of whether we
managed to send that frame or peer receive it. This could result in GO
Negotiation failures in cases where the peer did not receive the
response and Listen channels of the devices were different. Fix this by
setting the flag only after TX status indicating success has been
received.
This fixes frequent failures shown for the test_grpform_pbc hwsim test
case.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
When Action frame TX is postponed until a pending p2p_scan completes,
there may be additional operations that need to be continued after the
postponed Action frame TX operation completes. Fix this by starting
pending operation (if any) from TX status event for after_scan_tx
frames.
This fixes common errors seen with the test_discovery hwsim test case.
Signed-hostap: Jouni Malinen <j@w1.fi>
If a peer replies to persistent group invitation with status code 8
(unknown group), remove the peer from the p2p_client_list if we are the
GO or remove the persistent group if we are the P2P client since it
looks like that the peer has dropped persistent group credentials and
the provisioning step needs to be executed again.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Commit 50285f5ca8 changed number of rules
in channel selection and among other things, it broke the design where
the currently used operating channel on a virtual interface that is
shared by the same radio is preferred to avoid costs related to
multi-channel concurrency. Fix this regression by making the P2P module
aware of the shared channel and using that preference as the highest
priority when re-selecting the channel during negotiation.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Peer device includes its list of allowed operating channels in the
Invitation Response frame. When we are becoming the GO, use that list
from the peer to filter out acceptable channels to avoid selecting a
channel that the peer is unable to use.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
If multi channel concurrency is supported, we have to populate the
p2p_channels with list of channels that we support. Use the same design
that was previously added for GO Negotiation.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Commit fb8984fd6f cleared wps_method to
WPS_NOT_READY in p2p_stop_find_for_freq() as an attempt to clear
authorization when a group formation is cancelled. However, this code
path is hit also in cases where the user did not actually cancel
anything (e.g., from p2p_process_go_neg_req()). As such, it is not fine
to clear wps_method here even if it could be proper for some cases. For
now, revert that part to avoid regressions and consider clearing
wps_method on cancel separately.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
In case we have replied to a peer's GO Negotiation Request frame with a
GO Negotiation Response frame using status code
info-currently-unavailable (1), the peer is likely going to wait for us
to initiate GO Negotiation on its Listen channel. We were previously
using alternativing send-GO-Neg-Req and Listen phase when providing that
response after the user had authorized the connection. However, the
Listen phase here is unnecessary in this case and will make the
connection take longer time to go through. Skip the Listen phase and
make the wait-for-GO-Neg-Resp timeout random between 100 and 200 ms to
avoid getting in sync with the peer. In practice, this will make us
retry GO Negotiation Request frames more frequently and remain on the
peer's Listen channel for most of the time when initiating GO
Negotiation after status=1 response.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
There may be environments in which large number of devices are operating
on the social channels. In such cases, it is possible for the Action
frame TX operation wait for quite long time before being able to get the
frame out. To avoid triggering GO Negotiation failures, increase the
timeouts for GO Neg Req (with TX ACK) and GO Neg Resp (with or without
TX ACK as long as status=0) to 500 ms.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
It is possible for the driver to indicate multiple Probe Request frames
that would be processed in a single loop. If those frames happen to be
from a peer which with we are trying to start GO Negotiation, multiple
timeouts to start GO Negotiation (p2p_go_neg_start) could end up being
scheduled. This would result in confusing burst of multiple GO
Negotiation Request frames being sent once the RX loop finally
concludes. Avoid this by scheduling only a single eloop timeout to
trigger GO Negotiation regardless of how many Probe Request frames from
the peer is received. In addition, make sure this timeout gets canceled
in p2p_deinit().
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
If we have already sent out GO Negotiation Response and are waiting for
the peer to reply with GO Negotiation Confirm, there is no point in
re-starting GO Negotiation based on Probe Request frame from the peer.
Doing that would just result in confusing GO Negotiation exchange with
multiple sessions running at the same time.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Commit 624b4d5a64 changed GO Negotiation
to use the same Dialog Token value for all retransmissions of the GO
Negotiation Request within the same session. However, it did leave the
tie breaker bit changing for each frame. While this should not have
caused issues for most cases, it looks like there are possible sequences
where the peer may end up replying to two GO Negotiation Request frames
with different tie breaker values. If in such a case the different GO
Negotiation Response frames are used at each device, GO role
determination may result in conflicting results when same GO intent is
used.
Fix this by assigning the tie breaker value at the same time with the
dialog token (i.e., when processing the p2p_connect command instead of
for each transmitted GO Negotiation Request frame) to avoid issues with
GO selection.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
The driver may have cached (e.g., in cfg80211 BSS table) the scan
results for relatively long time. To avoid reporting stale information,
update P2P peers only based on results that have based on frames
received after the last p2p_find operation was started.
This helps especially in detecting when a previously operating GO stops
the group since the BSS entry for that could live for 30 seconds in the
cfg80211 cache. Running p2p_flush followed by p2p_find will now allow
wpa_supplicant to not add a P2P peer entry for that GO if the group had
been terminated just before that p2p_flush command. Previously, that GO
could have been indicated as a newly found device for up to 30 seconds
after it had stopped the group.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
For various P2P use cases, it is useful to have more accurate timestamp
for the peer information update. This commit improves scan result
handling by using a single timestamp that is taken immediately after
fetching the results from the driver and then using that value to
calculate the time when the driver last updated the BSS entry. In
addition, more debug information is added for P2P peer updates to be
able to clearly see how old information is being used here.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
If we discover a P2P peer based on a Beacon frame from the GO role, we
do not get information about the supported configuration methods. This
can result in issues if the P2P managing entity above wpa_supplicant is
not prepared to handling config_methods=0x0. To avoid this, postpone
reporting of the P2P-DEVICE-FOUND event when this happens on one of the
social channels. It would be good to be able to this on all channels,
but that could result in issues of never indicating the event for a peer
that is operating a GO on a channel that requires passive scanning.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
It could be possible for the scan results to include two entries for a
peer, one from the Listen state and the second one from the GO role. The
latter could be based on a Beason frame. If that happens and the entry
from GO is processed last, the P2P peer config_methods value could
potentially get cleared since Beacon frames do not include this
information in either WPS or P2P element. Avoid this by allowing the
config_methods value for P2P peers to be updated only if the new value
is non-zero.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
When pending p2p_find fails we need to send p2p_stop_find event to
indicate the previous p2p_find command has been processed.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
The P2P-FIND-STOPPED event was sent only in the P2P_SEARCH state, but
this needs to be send also in the new continue-search-when-ready states
P2P_CONTINUE_SEARCH_WHEN_READY and P2P_SEARCH_WHEN_READY for consistent
behavior.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Each GO Negotiation Request is (re)tried with an unique dialog token and
a GO Negotiation Response / Confirmation from the peer with a mismatched
dialog token is ignored which could result in a failure in this group
formation attempt. Thus, the P2P device would continue retrying the GO
Negotiation Request frames till the GO Negotiation Response frame with a
matching dialog token is received. To avoid the failures due to the
dialog token mismatch in retry cases if the peer is too slow to reply
within the timeout, the same dialog token value is used for every retry
in the same group formation handshake.
It should be noted that this can result in different contents of the GO
Negotiation Request frame being sent with the same dialog token value
since the tie breaker bit in GO Intent is still toggled for each
attempt. The specification is not very clear on what would be the
correct behavior here. Tie breaker bit is not updated on
"retransmissions", but that is more likely referring to the layer 2
retransmission and not the retry at higher layer using a new MMPDU.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
This makes it easier to read the code for the two possible cases
(forced/preferred channel and automatic channel selection).
Signed-hostap: Jouni Malinen <j@w1.fi>
Both p2p_connect and p2p_authorize use the same functionality to select
the channel preferences for GO Negotiation. The part of setting this
device flag was copied to each function, but it can also be handled by
the shared function after some reordering of code.
Signed-hostap: Jouni Malinen <j@w1.fi>
cfg80211 caches the scan results according the channel number. Due to
the 15 sec aging this might cause the user mode to see more than one
scan result with the same BSSID, e.g. - one scan result for the
P2P Device and one for the P2P GO (once it's enabled).
Fix this by updating the device entry only if the new peer entry is
newer than the one previously stored.
Signed-off-by: Yoni Divinsky <yoni.divinsky@ti.com>
Signed-off-by: Victor Goldenshtein <victorg@ti.com>
Signed-off-by: Igal Chernobelsky <igalc@ti.com>
Signed-hostap: Arik Nemtsov <arik@wizery.com>
The GO may be in sleep when we send a PD Request frame to indicate that
we are about to join a running group. Previously, this frame was not
retried more than normal low level retries. This can result in the GO
not getting the frame especially in cases where concurrent multi-channel
operations or aggressive sleep schedule is used since most drivers do
not yet synchronize with the GO's NoA before association.
Increase the likelihood of the GO receiving the PD Request frame by
retransmitting it similarly to the PD-for-GO-Negotiation case. Start
the actual join operation only after these retries have failed to get
an acknowledgment from the GO to give the connection attempt a chance
to succeed if the driver implements better NoA synchronization for it.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
The new P2P_SET parameter disc_int can now be used to configure
discoverable interval for p2p_find operations. The format of the command
for setting the values is "P2P_SET disc_int <minDiscoverableInterval>
<maxDiscoverableInterval> <max TUs for discoverable interval>". The
first two parameters are given in units of 100 TUs (102.4 ms). The third
parameter can be used to further limit the interval into a specific TU
amount. If it is set to -1, no such additional limitation is enforced.
It should be noted that the P2P specification describes the random
Listen state interval to be in units of 100 TUs, so setting the max TU
value to anything else than -1 is not compliant with the specification
and should not be used in normal cases. The default parameters can be
set with "P2P_SET disc_int 1 3 -1".
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
If the driver indicates support for multi-channel concurrency, change
the p2p_connect behavior to not force the current operating channel, but
instead, just mark it as preferred for GO Negotiation. This change
applies only for the case when the freq parameter is not used with the
p2p_connect command.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
This commit adds control interface commands and internal storage of
Wi-Fi Display related configuration. In addition, WFD IE is now added
to various P2P frames, Probe Request/Response, and (Re)Association
Request/Response frames. WFD subelements from peers are stored in the
P2P peer table.
Following control interface commands are now available:
SET wifi_display <0/1>
GET wifi_display
WFD_SUBELEM_SET <subelem> [hexdump of length+body]
WFD_SUBELEM_GET <subelem>
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Previously, all station mode scan operations were either skipped or
delayed while any P2P operation was in progress. To make concurrent
operations easier to use, reduce this limitation by allowing a scan
operation to be completed in the middle of a p2p_find. In addition,
allow station mode association to be completed. When the station mode
operation is run to its completion (scan results not acted on,
connection to an AP completed, connection failed), resume the p2p_find
operation.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
A new optional delay=<search delay in milliseconds> parameter can now be
used with p2p_find command to request an extra delay between search
iterations. This can be used, e.g., to make p2p_find friendlier to
concurrent operations by avoiding it from taking 100% of the radio
resources.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Increase GO config timeout if HT40 is used since it takes some time
to scan channels for coex purposes before the BSS can be started.
Signed-hostap: Jouni Malinen <j@w1.fi>
p2p_set_timeout() calls in GO Neg Req/Resp TX callbacks used timeout of
100 ms which is the value given in the P2P specification for GO
Negotiation, but this was actually shorter than the
wait-for-offchannel-TX value (200 ms) used for the driver call. In
addition, it looks like some devices (e.g., Galaxy Nexus with JB image)
can take longer time to reply to GO Negotiation Response (somewhere
between 200 and 250 ms has been observed).
Increase the wait-for-GO-Neg-Resp timeout from 100 ms to 200 ms if GO
Negotiation Request frame was acknowledged (this matches with the
offchannel wait timeout that used previously). The no-ack case is left
at 100 ms since we use GO Negotiation Request frame also to discover
whether the peer is on its Listen channel.
Increase the wait-for-GO-Neg-Conf timeout from 100 ms to 250 ms (and
increase the offchannel wait timeout to matching 250 ms) as a workaround
for devices that take over 200 ms to reply to GO Negotiation Response.
Signed-hostap: Jouni Malinen <j@w1.fi>
Commit 6b56cc2d97 added retries of
provision discovery request frames in IDLE state. However, it did not
make the p2p_find case behave consistently with the new limitied retry
behavior. This can result in way too many and frequent PD retries. Fix
this by extending the previous commit to address PD retries and maximum
retry limit consistently regardless of whether p2p_find is running.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
intended-for: hostap-1
There are separate states for these, so we can't really get into this
situation unless somebody tries to do multiple things at the same
time. p2p_connect stops find and CONNECT state is used to probe the peer
on its Listen channel with GO Negotiation Request frames. Similarly,
p2p_invite() stops find and INVITE state is used to probe the peer on
its Listen channel with Invitation Request frames. The older mechanism
of using Search state functionality to find the peer can be removed.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Currently, FullMAC Persistent GO can't use p2p_client_list because its
own hapd->p2p_group is NULL at ap_sta_set_authorized(). This patch
changes the processing to use sta->p2p_ie instead of
p2p_group_get_dev_addr() on FullMAC GO.
Signed-hostap: Masashi Honma <masashi.honma@gmail.com>
The P2P Client Discoverability bit is reserved in most frames and its
value in the local P2P peer table should only be updated based on P2P
Group Info attribute from a GO. Fix this by avoiding changes to this
dev_capab bit based on other P2P frames. It would be more correct to
track this separately for each group in which the peer is a member, but
since we do not do that for the other group specific information either,
this can do for now.
It should be noted that prior to commit
18485b5469 wpa_supplicant set this bit in
all P2P frames. However, that commit changed this to match the
specification, i.e., the bit is not set in frames which are received
from P2P Device role. As such, this fix is needed to be able to figure
out that a peer supports client discoverability capability after that
commit.
Signed-hostap: Jouni Malinen <j@w1.fi>
intended-for: hostap-1
In the P2P specification v1.1, the P2P Client Discoverability bit is
described in Table 12 "Device Capability Bitmap definition". The table
says "Within a P2P Group Info attribute and a (Re)association request
frame the P2P Client Discoverability field shall be set to 1 when the
P2P Device supports P2P Client Discoverability, and is set to 0
otherwise. This field shall be reserved and set to 0 in all other frames
or uses.". To match with this, filter out P2P Client Discoverability bit
from frames where its use is reserved.
Signed-hostap: Masashi Honma <masashi.honma@gmail.com>
If both peers initiate GO Negotiation at about the same time, it is
possible for the GO Negotiation Request frame from the peer to be
received between the local attempt to send the GO Negotiation Request
and TX status event for that. This could result in both devices sending
GO Negotiation Response frames even though one of them should have
skipped this based which device uses a higher MAC address.
Resolve this race by incrementing go_neg_req_sent when p2p_send_action()
returns success instead of doing this from the TX status callback. If
the frame is not acknowledged, go_neg_req_sent is cleared in TX status
handler.
Signed-off-by: Neeraj Garg <neerajkg@broadcom.com>
Stop the connection attempt if GO Negotiation Confirm is not received
within 100 ms of the GO Negotiation Response getting acknowledged.
Previously, we would have continued trying to connect to the peer even
in this case which could result in confusing second GO Negotiation
Request frame and unnecessarily long wait before indicating failure.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
This is a workaround for interoperability issues with some deployed P2P
implementations that require a Provision Discovery exchange to be used
before GO Negotiation. The new provdisc parameter for the p2p_connect
command can be used to request this behavior without having to run a
separate p2p_prov_disc command.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Some drivers may accept the remain-on-channel command, but instead of
indicating start event for remain-on-channel, just indicate that the
operation has been canceled immediately. This could result in continuous
loop of search/listen states with very limited time to do anything else
in wpa_supplicant if the scan command is also completed quickly (e.g.,
if the driver is unable to scan other channels than the current
operating channel).
As a workaround, do not start the next step (search) in P2P device
discovery if this type of rejection of listen operation is detected.
This gives some more time for wpa_supplicant to handle whatever else
may be needed at to be done at the same time and reduces the amount
of CPU used in a loop that does not really work correctly from the
view point of being discoverable.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
intended-for: hostap-1
The misplaced parenthesis caused Device ID matching to check only the
first octet of the P2P Device Address, i.e., we could have replied to
Probe Request frames that was searching for another device if any of
the last five octets of the address were different.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
intended-for: hostap-1
An existing persistent group information can now be used to force GO
Negotiation to use the previously used SSID/passphrase from a persistent
group if we become a GO. This can be used as an alternative to inviting
a new P2P peer to join the group (i.e., use GO Negotiation with GO
intent 15 instead of starting an autonomous GO and using invitation),
e.g., in case a GO Negotiation Request is received from a peer while we
are not running as a GO. The persistent group to use for parameters is
indicated with persistent=<network id> parameter to p2p_connect.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Commits 17bef1e97a and
ffe98dfb88 started using p2p_add_device()
with other frames than just Probe Response frames from scan results.
However, these changes did not take into account that the PD Request
and Invitation Request frames are normally received on the our own
Listen channel, not on the Listen channel of the peer. As such, these
frames must not be used to update Listen channel information for the
peer.
Fix this by letting p2p_add_device() know whether the results are from
scan results and if not, skip the peer table updates that are specific
to Probe Response frames.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
intended-for: hostap-1
Commit d9bdba9f86 was supposed to allow
p2p_listen to schedule new after scan operation only if a p2p_connect
operation was not pending. However, it used incorrect P2P_AFTER_SCAN_
value for this and did not actually prevent overriding of p2p_connect
and could have caused some p2p_listen operations to be skipped
unnecessarily.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
intended-for: hostap-1
Some applications require knowing about probe requests to identify
devices. This can be the case in AP mode to see the devices before they
connect, or even in P2P mode when operating as a P2P device to identify
non-P2P peers (P2P peers are identified via PeerFound signals).
As there are typically a lot of probe requests, require that an
interested application subscribes to this signal so the bus isn't always
flooded with these notifications. The notifications in DBus are then
unicast only to that application.
A small test script is also included.
Signed-hostap: Johannes Berg <johannes.berg@intel.com>
The "BSS p2p_dev_addr=address" command uses p2p_parse_dev_addr() to
figure out the P2P Device Address of the GO from scan results. This used
to work only if the P2P IE was received from Probe Response frames since
only those include the P2P Device Info attribute. Make this work with
Beacon frames, too, by using P2P Device ID attribute if the P2P Device
Info attribute is not present.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
p2p_pref_chan configuration parameter can now be used to set the
list of preferred channel for P2P GO Negotiation. This will be used
in the priority order if the peer does not support the channel we
are trying to use as the GO (configured operating channel or the
best 2.4 GHz/5 GHz channel) for the case where a forced channel is
not used.
p2p_pref_chan=<op class:channel>,...
For example:
p2p_pref_chan=81:1,81:2,81:3,81:4,81:5,81:6
This would configure 2.4 GHz channels 1-6 as the preferred ones with
channel 1 the most preferred option.
These configuration parameters can be set in wpa_supplicant.conf and
dynamically updated with "wpa_cli set <param> <value>".
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
Previously, this provisioning info was cleared using the P2P Interface
Address of the GO as the key. That did not always work in the case the
where we joined an already running group. This could result in the next
connection to that same GO skipping provision discovery. Fix this by
finding the peer entry based on its P2P Device Address instead of the
P2P Interface Address which may not always be set.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
intended-for: hostap-1
Use Device Password ID in WSC IE of Probe Request and Probe Response
frames to advertise immediate availability of WPS credentials per P2P
specification sections 3.1.2.1.1 (Listen State), 3.1.2.1.2 (Scan Phase),
and 3.1.2.1.3 (Find Phase).
For now, the Device Password ID is set only for the case where we are
active GO Negotiation with a specific peer. In practice, this means that
the Probe Response frames during pending GO Negotiation (whenever in
Listen state) indicate availability of the credential.
Signed-hostap: Jouni Malinen <jouni@qca.qualcomm.com>
